PB83-200634
Resource and Environmental
Impacts of Trends in U.S. Agriculture
Resources for the Future, Inc.
Washington, DC
Prepared for
Environmental Research Lab.
Athens, GA
U.S. DEPARTMENT OF COMMERCE
National Technical Information Service
NTS
-------
EPA-600/3-82-031
May 1983
PB83-200634
RESOURCE AND ENVIRONMENTAL IMPACTS
OF TRENDS IN U.S. AGRICULTURE
by
Pierre Crosson and Sterling Brubaker
Resources for the Future
Washington, D.C. 20036
Grant No. R8060236010
Project Officer
George W. Bailey
Environmental Research Laboratory
Athens, Georgia 30613
This study was partially funded by the
Rockefeller Foundation and Resources for the Future
ENVIRONMENTAL RESEARCH LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
ATHENS, GEORGIA 30613
-------
TECHNICAL REPORT DATA
(Please read hismictiuns on the reverse before completing)
1 R|pA-608/3-82-031 2" CRD Report
3. RECIPIENT'S ACCESSION NO.
an* x ? 0 0 6 3 k
<1. TITLE AND SUBTITLE .
Resource and Environmental Impacts of Trends
in U.S. Agriculture
5. REPO^T'fcf'Al'F
May 1983
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
Pierre Crosson and Sterling Brubaker
8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADORESS
Resources for the Future
Washington, D.C. 20036
10. PROGRAM-ELEMENT NO.
CARB1A
11. CONTRACT/GRANT NO.
R8060236010
12. SPONSORING AGENCY NAME AND ADDRESS
Environmental Research Laboratory—Athens, GA
Office of Research and Development
U.S. Environmental Protection Agency
Athens, Georgia 30613
13. TYPE OF REPORT AND PERIOD COVERED
Final, 10/78-12/80
14. SPONSORING AGENCY-CODE
EPA/600/01
15. SUPPLEMENTARY NOTES
,6' A8STF¥rends in demand for U.S. agricultural production and in agricultural technology
suggest increasing pressure on the nation's land and water resources over the next
several decades. The expected consequents would be rising economic costs of produc-
tion and damages to the environment. This study analyzes those trends, assesses their
economic and environmental impacts and discusses policies for dealing with their impact
The quantities of land, water and other resources farmers use to increase pro-
duction depend basically on the kinds of technologies they employ. Two categories of
technology are distinguished—land-using technologies and land-saving technologies.
Farmers' choices from the spectrum of technologies are conditioned by the prices and
productivities of the alternatives.
Analysis of trends indicates that an additional 60 to 70 million acres will be
brought into production and that erosion will emerge as the most serious environmental
problem of agriculture. A slower rise in inputs of fertilizer per acre is expected
and the total quantity o'f insecticide applied to crops should decline. Herbicide use
is expected to increase markedly. New land-saving technologies could reduce pressures
on the land.
17. KEV WORDS AND DOCUMENT ANALYSIS
a. DESCRIPTORS
b. 1DENTIFlERS/OPEN ENDED TERMS
c. COSATI Field/Group
18. DISTRIBUTION STATEMENT
RELEASE TO PUBLIC
19. SECURITY CLASS (This Report)
Unclassified
21. no. J?2^ages
20. SECURITY CLASS (This page)
Unclassified
22. PRICE
EPA Form 2220*1 (9-73).
-------
NOTICE
This document has been reviewed in accordance with
U.S. Environmental Protection Agency policy and
approved for publication. Mention of trade names
or commercial products does not constitute endorse-
ment or recommendation for use.
ii
-------
FOREWORD
Environmental protection efforts are increasingly directed towards
preventing adverse health and ecological effects associated with specific
compounds of natural or human origin. As part of this Laboratory's research
on the occurrence, movement, transformation, impact, and control of environ-
mental contaminants, management or engineering tools are developed for
addressing water quality problems with emphasis on assessing potential ex-
posures to chemical contaminants.
Crop production practices, already significant sources of water and air
pollution, represent areas of increasing environmental concern as agricultural
demands increase pressures on the nation's land and water resources. This
report identifies and assesses the relative importance of nation-wide trends
in agricultural production and technology that may have significant effects
on the environment over the next several decades. Study findings are intended
to aid environmental managers as they attempt to anticipate pollution problems
of the future.
David W. Duttweiler
Director
Environmental Research Laboratory
Athens, Georgia
iii
-------
PREFACE
This report is the culmination of several years of research undertaken
at Resources for the Future. The study originated out of concern that pros-
pective long-term trends in demand for U.S. agricultural production, in
prices of energy and other key inputs, and in agricultural technologies
would put increasing pressure on the nation's land and water resources and
on the environment. It appeared to us that these trends likely would shift
the agricultural situation of the United States from one of chronic surplus
to one of recurring if not chronic scarcity. Should such a shift occur the
nation likely will require major modification of the policies for resource
and environmental management which have evolved since the 1930s.
The study is in five parts: (1) projections to 1985, 1990, and 2010 of
production of wheat, feed grains (corn, grain sorghum, oats, and barley),
soybeans and cotton (section 1); (2) analysis of trends in input prices,
technology, and crop yields as a basis for projecting the demand for cropland
(sections 2-4); (3) analysis of trends* in fertilizer and, insect management
practices and in tillage technologies as a basis for projections of demand
for fertilizers and pesticides (section 5); (4) analysis of the environmental
impacts of the projected quantities of resources to reach judgments about the
future severity of the impacts (section 6); (5) assessment of the effective-
ness of current policies for dealing.with the prospective environmental
problems and discussion of alternative policies where these appear needed.
State planning done under Section 208 of the Federal Water Pollution Control
Act gets special attention in this part (section 7).
Given the objectives of the study, we aimed at comprehensiveness in two
senses: (1) we focused on the nation as a whole, although considerable
attention is given to regional and state issues; (2) we deal with all major
environmental impacts of crop production, meaning erosion, fertilizer, and
pesticide pollution, and soil and water salinity.associated with irrigation.
Being comprehensive and undertaken with a relatively modest commitment of
resources, the research necessarily is cast at a high level of generality.
The intent is to identify and assess the relative importance of trends in
agricultural production and technology that, from.a national perspective,
may have significant effects on the environment over the next several decades.
Given this perspective and time frame, the conclusions of the research neces-
sarily are subject to greater uncertainty than would be the case if the time
span were shorter and the focus more narrow.. The conclusions, therefore,
about trends in technology, resource use, environmental impacts, and policies
are not intended to apply to specific.local situations. The conclusions
should.be meaningful and useful, however, in providing insights as to emerg-
ing environmental problems of agriculture and the effectiveness of policy
and institutional modes of dealing with them.
iv
-------
With the exception of interviews conducted as part of our research on
tillage and pest management technologies, irrigation and 208 planning, we
have not developed new sources of data. Rather we have relied on systematic
surveys of data and other information already available elsewhere. Nor have
we developed formal models of the various relationships studied. We did,
however., make use of Iowa State University's model of U.S. agriculture to
project erosion and of a model developed at RFF for another purpose to trans-
form projections of erosion to projections.of sediment delivered. With these
exceptions the analysis is not cast in a formal modeling mode. This would
not have been appropriate, in our judgment, given the broad focus of the
research and its intent to identify and assess new patterns of resource use
and technology likely to emerge over several decades.
The research was partially funded by a grant from the Environmental
Protection Agency's Environmental Laboratory at Athens, Georgia as part of
the Laboratory's program of research on Environmental Implications of Trends
in Agriculture. Previously published research under this program are:
Development Planning and Research Associates, Inc., Environmental Implica-
tions of Trends in Agriculture and Silviculture, Vol. I: Trend Identifica-
tion and Evaluation by S. Unger, EPA-600/3-77-121 (1977); Vol. II:
Environmental Effects of Trends by S. Unger, EPA 600/3-78-102 (1978);
Vol. Ill: Regional Crop Production Trends by S. Unger, EPA 600/3-79-047
(1979). Also F. C. White et al., Environmental and Economic Impact of Agri-
cultural Land Use Conversion: an Evaluation Methodology, EPA 600/5-80-002
(1980); R. W. Skaggs, J. W. Gilliam, T. J. Sheets and J. S. Barnes, Effect
of Agricultural Land Development on Drainage Waters in the North Carolina
Tidewater Region, Water Resources Institute of North Carolina, WRR/Report
no. 154 and EPA 600/3-80-087.
Two other studies financed partially under the Athens Laboratory's
"Trends" program were undertaken at Resources for the Future simultaneously
with this one. One, with the general title Trends in U.S. Irrigation: Three
Regional Studies (Pierre Crosson, project leader, 1981) included Western
Irrigation: Its Past and Future .Growth, by Kenneth D. Frederick with James C.
Haiison; Growth and Prospects for Irrigation in the Eastern United States, by
James C. Hanson and James Pagano; The Economic Potential for the Expansion
of Irrigation in the Mississippi Delta Region, by R. N. Shulstad, R. May,
B. Herrington and J. Erstine. The study by Shulstad et al. was done at the
University of Arkansas under a grant from Resources for the Future.
The second companion RFF study was Conservation Tillage and Conventional
Tillage: a Comparative Assessment, by Pierre Crosson, published by the Soil
Conservation Society of America, Ankeny, Iowa (1981).
v
-------
ABSTRACT
Trends in demand for U.S. Agricultural production and in agricultural
technology suggest increasing pressure on the nation's land and water re-
sources over the next several decades. The consequence would be rising
economic costs of production and damages to the environment. The study
analyzes and projects those trends, assesses their economic and environmental
impacts and discusses policies for dealing with the impacts.
Rising demand for exports of grains and soybeans is a principal source
of increasing pressure on the resource base. Demand for corn to produce
ethanol will increase, but by a relatively small amount relative to export
demand. Other components of domestic demand will increase primarily in
response to population growth, expected to be less than 1 percent annually
from 1980 to 2010.
The quantities of land, water and other resources farmers use to in-
crease production depend basically on the kinds of technologies they employ.
Two categories of technology are distinguished, those combining relatively
large amounts of land with non-land inputs (land-using technologies) and
those for which the ratio of land to non-land inputs is relatively low (land-
saving technologies). Alternative technologies are thought of as lying along
a spectrum, with highly land-using technologies at one end and highly land-
saving technologies at the other.
Farmers' choices from the spectrum of technologies are conditioned by
the prices and productivities of the alternatives. From the end of World
War II to the early 1970s relatively low prices of energy and fertilizers
and the high productivity of these inputs favored choices- from the land-
saving end of the spectrum. Crop yields and total productivity rose at
unprecedented rates and the acreage of land in crops declined.
After the early 1970s real prices of energy and fertilizers rose and the
rate of increase of total productivity slowed. Farmers shifted toward the
land-using end of the spectrum, and crop yields increased much more slowly
than in the two decades previous to the early 1970s.
The consensus is that real energy prices will continue to rise, and
there is much evidence that fertilizer prices will also. These trends, com-
bined with increasing scarcity of water for irrigation in the West, imply
that farmers will continue to favor land-using technologies unless new land-
saving technologies are developed.
The anlysis of trends in technology indicates that an additional 60 to
70 million acres of land will have to be brought under crops to meet
vi
-------
projected crop demand in 2010. This increase, combined with rising demand
for agricultural land for urban and other non-agricultural uses, suggests
increasing real economic costs of agricultural land.
Inputs per acre of fertilizer are expected to rise, but at the relative-
ly slow pace experienced since the early 1970s rather than at the high rate
characteristic of earlier years. The total quantity of insecticides applied
to crops is expected to decline despite more acreage because of a sharp re-
duction in per acre applications of insecticides to cotton. Per acre appli-
cations to corn also are expected to decline, but not so rapidly as for
cotton. Herbicide usage is expected to increase markedly because of more
acres in crops and because of increased per acre applications reflecting the
spread of conservation tillage.
The projected quantities and pattern of resource use suggests that
erosion will emerge as the most serious environmental problem of agriculture.
Sheet and rill erosion from cropland is projected at 3.5 billion tons in
2010, compared with 1.9 billion tons in 1977. Per acre erosion would rise
from 4.7 tons to 7.4 tons, 2.4 tons more than the maximum set by the Soil
Conservation Service as consistent with maintaining long run productivity of
the land. Sediment delivered to the nation's rivers, lakes and reservoirs
would double. It is likely that erosion on the projected scale would be
viewed as posing a significant threat to national water quality as well as
to the productivity of the land.
The nation traditionally has relied on voluntary approaches to erosion
control, offering farmers financial inducements, e.g. cost sharing of erosion
control practices, to enlist their participation in control programs. This
approach works reasonably well when commodity markets are weak, but when they
are strong, as expected over the next several decades, then farmers have in-
centive for intensive use of their land. They are reluctant to adopt erosion
control measures if these require setting aside land which could be in pro-
duction.
Section 208 of the Clean Water Act Amendments of 1972 authorizes the
EPA, acting through the states, to regulate non-point pollution to achieve
improved water quality. Section 208 thus seems to convey authority to the
EPA to impose erosion controls where erosion threatens water quality. State
plans drawn up under Section 208, however, rely on traditional voluntary
approaches to erosion control, and the EPA has accepted this.
If the erosion problem emerges as projected, traditional voluntary'
approaches likely will be perceived as inadequate. New programs involving
penalties for farmers not in compliance with erosion standards may be needed;
and better targeting of programs on the worst offenders would have high pay-
off. Departures from the voluntary approach will be resisted, however, and
both their political and administrative costs likely will be high. This does
not mean that new departures should not be developed. It suggests, however,
that they be accompanied by a strategy of research to develop new technolo-
gies that serve both the farmers' economic interest and the social interest
in reducing environmental damages. New land-saving technologies that sub-
stitute other inputs for high priced fossil energy and fertilizer would
vii
-------
reduce pressure on the land. Enhanced photosynthetic efficiency in main
crops and expansion of their capacity to biologically fix nitrogen are exam-
ples. Development of a higher yielding variety of soybean is another.
Research to extend the economic limits of conservation tillage would make it
possible to use this technology on land where it now is not competitive,
thus significantly reducing the threat of erosion. However, the reliance of
conservation tillage on herbicides for weed control indicates that research
to extend the technology should include careful study of the long term en-
vironmental effects of high levels of herbicide use.
This report was submitted in fulfillment of Grant No. R8060236010 by
Resources for the Future under the sponsorship of the U.S. Environmental
Protection Agency. This report covers the period October 1978, to December
1980 , and work was completed as of June 1981.
viii
-------
CONTENTS
Preface . iv
Abstract ..... vi
Contents ix
Tables . ...... xiii
Acknowledgments . xvii
Conclusions xviii
Recommendations for Research xx
Section 1. PROJECTIONS OF PRODUCTION OF GRAINS, SOYBEANS AND
COTTON IN THE UNITED STATES 1
Introduction 1
Projections to 1985, 1990, and 2010 1
Sources of Error in Projections 4
Comparison with Others ........ 6
References 10
Section 2. FACTORS AFFECTING FARMERS' CHOICES AMONG TECHNOLOGIES:
PRICES OF INPUTS 11
Introduction 11
Types of Technology 11
Choices Among Technology: A Historical Perspective . 12
Prices of Inputs 14
Energy 14
Fertilizer 16
Water 17
Western irrigation . 18
Eastern Irrigation , 25
Conclusions on Prices of Inputs 27
References 28
Section 3. TRENDS IN PRODUCTIVITY AND CROP YIELDS IN THE
UNITED STATES 30
Introduction 30
Trends in Yields: All Crops 31
Yields of Grain and Soybeans 33
References 36
is
-------
Section 4. THE DEMAND FOR AND SUPPLY OF CROPLAND 37
Introduction 37
Regional Projections of Production 37
National and Regional Projections of Yields 38
Demand for and Supply of Cropland 43
Demand for Cropland 43
Supply of Cropland 45
Costs of Land Conversion 48
Conversion to Non-agricultural Uses 52
References 55
Section 5. PROJECTIONS OF FERTILIZERS AND PESTICIDES 56
Fertilizers 56
Corn 56
Wheat 62
Soybeans 62
Cotton 62
Other Uses of Fertilizers 63
Pesticides 63
Data Problems 63
Insecticides 65
Wheat 71
Soybeans 72
Corn 73
Cotton 75
Herbicides 80
Importance of Conservation Tillage 80
Economics of Conservation Tillage 82
Costs per acre 82
Yields 84
Summary on Economics 87
Future Trends in Herbicide Use 91
References 93
Section 6. ENVIRONMENTAL IMPACTS OF PROJECTED PRODUCTION
AND RESOURCE USE . . 97
Introduction 97
Fertilizer . . 97
Nature of the Environmental Impacts 97
Present Severity of the Problems . .' 98
Eutrophication 101
Future Severity of the Problems . 102
Movement of Nutrients 103
Effects of Increased Efficiency 103
Effects of Tillage Technologies 104
x
-------
Section 6 (Continued)
Pesticides • 107
Nature of the Environmental Impacts . ' 107
Present Severity of the Problems 108
Insecticides 109
Herbicides Ill
Future Severity of the Problems 113
Insecticides 113
Herbicides 115
Irrigation 116
Nature of the Environmental Impacts 116
Present Severity of the Problems 117
Future Severity of the Problems 119
Erosion 120
Nature of the Environmental Impacts 120
Present Severity of the Problems 122
Future Severity of the Problems . 126
Conclusion 130
References 131
Section 7. POLICY CONSIDERATIONS 136
Background 136
Current Environmental Approaches . 139
Soil Conservation Programs • • • 139
Water Pollution Control 141
Pesticide Controls 148
Alternative Policy Approaches . . . . 150
Demand and Soil Degradation . 150
Soil Loss and Social Responsibility . . . . 152
Soil Conservation Policies 154
Water Pollution Control Policies 157
Research as Policy . 161
A Final Word 162
References 164
Appendix A: DERIVATION OF PROJECTIONS OF PRODUCTION 166
Projections of World Trade .... 166
Wheat and Coarse Grains 166
Importing Developing Countries 166
Other Countries 166
Projections of Wheat Imports 171
Projections of Coarse Grain Imports 172
Oilmeal 175
U.S. Shares of World Trade in 2010 177
The Case for Constant Shares 177
The Case for Declining Shares 179
xi
-------
Appendix A (Continued)
U.S. Domestic Consumption . . . 180
Projections of Grains and Soybeans to 2010 180
Projections of Cotton 180
References ¦ . • 184
Appendix B: TRENDS IN TOTAL AGRICULTURAL PRODUCTIVITY 186
References 189
Appendix C: ANALYSIS OF TRENDS IN CROP YIELDS 190
Pure Yield and Shift Effects 190
Pure Yield Factors: Weather 191
Pure Yield Factors: Land Quality . . . . 201
Pure Yield Factors: Technology 205
Fertilizer . • 205
Irrigation 206
Summary 207
References 208
Appendix D: PROJECTIONS OF REGIONAL SHARES OF PRODUCTION AND
OF REGIONAL CROP YIELDS 209
Regional Shares 209
Yields 212
References . . 213
xii
-------
TABLES
Number Page
1-1 U.S. Production, Export, and Domestic Use of Wheat,
Feedgrains and cotton, 1978/80 and Projections to 1985,
1990 and 2010 2
1-2 Alternative Projections of U.S. Production of Wheat,
Feedgrains and Soybeans . 6
1-3 Crop Prices Received by Farmers, Dollars of 1979 9
2-1 Ratios of Inputs to Cropland Used for Crops in U.S.
Agriculture 12
2-2 Ratios of Input Prices to Prices of Land and Prices
of Fertilizer 13
2-3 Index of Real Prices Paid by Farmers-for Fuel and
Energy in the United States 15
2—4 Indexes of Real Prices of Fertilizers in the United States 18
2-5 1977 Irrigated and Dryland Agricultural Land Use by Farm
Production Region 19
2-6 Irrigated and Dryland Acreage of Corn, Sorghum, Wheat, and
Cotton for 1950 and 1977 . 20
2-7 Acres Irrigated with On-Farm Pumped Water in the United
States by Type of Energy, 1974 21
2-8 Energy Requirements for On-Farm Pumping of Irrigation
Water in United States by Type of Energy, 1974 . 22
2-9 Energy Costs to Pump 1 Acre-Foot of Water from Different
Depths with Alternative Fuels and Fuel Prices 24
2-10 Irrigated Land in the Eastern United States, 1967 and 1977 25
3-1 Yields for All Crops 32
4-1 Regional Shares of Crop Production 39
xiii
-------
4-2 Yields in the United States of Wheat, Feedgrains,
Soybeans and Cotton 41
4-3 Yields in the United States of Wheat, Feedgrains
and Soybeans 42
4-4 Projections of Demand for Cropland 44
4-5 Estimates of the Amount of Cropland in the United
States in 1977, by Major Use 46
4-6 Supply of and Demand for Cropland 47
4-7 Prices Received by Farmers for Grains and Soybeans 50
4-8 Amounts of Land Convertible to Crops in Iowa and the
Mississippi Delta 51
5-1' Applications of Fertilizer per Acre of Fertilized Land
in Corn, Wheat, Soybean and Cotton 57
5-2 Applications of Fertilizer per Acre of Harvested Land in
Corn, Wheat, Soybeans, and Cotton 58
5-3 Amounts of Fertilizer Applied to Land in Corn, Wheat,
Soybeans, Cotton and for All Other Uses 64
5-4 Contrasting Estimates of Amounts of Herbicides and
Insecticies Used in the United States 65
5-5 Pesticide Use on Crops in 1971 and 1976 . 66
5-6 Percentages of Acres on Which Pesticides were Used in
1976, Major Crops 66
5-7 Herbicides and Insecticides Applied to Major Crops in
1971 and 1976 . . 67
5-8 Types and Amounts of Insecticides Applied to Major Crops
in 1976 . 68
5-9 Herbicides and Insecticides Applied to Main Crops, by
Region in 1976 69
5-10 Types and Amounts of Herbicides Applied to Major Crops
in 1976 70
5-11 Land in Conservation Tillage in the United States 83
5-12 Conservation Tillage by Region 84
xiv
-------
5-13 Percentages of Land Apt for Conservation Tillage and in
Conservation Tillage, Ohio, Indiana, Illinois, and
Iowa 88
5-14 Percent Distribution of Cropland in Conservation
Tillage, Various States and Regions 1979 89
5-15 Herbicides Applied to Crops 92
6-1 Erosion from Cropland in the United States 124
6-2 Projections of Sheet and Rill Erosion from Cropland
in the United States in 2010 127
6-3 Sediment Delivered to Water Bodies in the United States . 127
7-1 Estimates of Gross Sheet and Rill Erosion and Sediment
Delivery from Cropland in the United States and
Selected States, 1977-2010 144
A-l Average Annual Growth Rates in Importing Developing
Countries, Wheat and Coarse Grains 167
A-2 Production, Consumption, and Imports of Wheat and Coarse
Grains in Importing Developing Countries 169
A-3 Wheat Imports 172
A-4 Population in Western Europe, Eastern Europe, USSR, and
Japan 174
A-5 Production, Consumption and Imports of Coarse Grains,
Selected Countries and Regions 175
A-6 World Oilmeal Trade 176
A-7 U.S. Production, Domestic Use, and Exports of Wheat,
Feedgrains and Soybeans 181
A-8 Production, Net Exports and Mill Consumption of Raw
Cotton in the United States 183
B-l Average Annual Rates of Growth of Total Agricultural
Productivity in the United States 187
B-2 Average Annual Rates of Increase in Production and
Selected Inputs in U.S. Agriculture 188
C-l Pure Yield, Pure Shift, and Interaction Effects on
Observed Yields of Corn, Wheat, Sorghum, and Soybeans . . . 192
xv
-------
C-2 Corn: Index of Effects of Weather on Yields, Actual
Yields, Weather-Adjusted Yields, and Trend Values of
Weather-Adjusted Yields . . 194
C-3 Soybeans: Index of Effects of Weather on Yields, Actual
Yields, Weather-Adjusted Yields, and Trend Values of
Weather-Adjusted Yields . ........ 196
C-4 Wheat: Index of Effects of Weather on Yields, Actual
Yields, Weather-Adjusted Yields and Trend Values of
Weather-Adjusted Yields . 198
C-5 Analysis of the Effects of Changing the Amount of
Land on Yields of Wheat and Corn 204
C-6 Fertilizer Use on Corn and Wheat in the United States . . . 206
D-l Distribution of Production of Selected Crops Among River .
Basins in the United States in 2010 211
xvi
-------
ACKNOWLEDGEMENTS
Many persons read and made valuable comments on all or parts of this
study. They are: George W. Bailey, Charles Benbrook, Norman A. Berg,
Robert F. Boxley, Daniel Bromley, Lee A. Christensen, Harold R. Cosper, Fred
T. Cooke, William H. Cross, Ray Frisbie, Roy M. Gray, Donald R. Griffith,
Richard R. Harwood, Earl Heady, Maureen K. Hinckle, Royce Hinton, Erik
Janson, J. M. Kennedy, Lawrence W. Libby, R. A. Olson, John H. Perkins,
David Pimentel, Fred Sanderson, Frank W. Schaller, Robert Shulstad,
Earl R. Swanson, C. Robert Taylor, Grant W. Thomas, John F. Timmons,
G. B. Triplett, Jr., David J. Walker, R. D. Wauchope, Allen F. Wiese,
M. Gordon Wolman, and two anonymous reviewers. Thanks go also to various
people interviewed in the course of doing the research. They are cited in
the text as appropriate.
In addition many others too numerous to list individually gave generous-
ly of their time in interviews.
We are grateful to all of these people for their help. All contributed
to such merit as the study may have. Of course none are responsible for
remaining defects.
Maybelle Frashure, ably assisted by John Mankin and Lorraine Van Dine,
was responsible for typing the manuscript. We appreciate their uncomplain-
ing (to us anyway) dedication to a sometimes tedious task.
Pierre Crosson
Sterling Brubaker
xvii
-------
CONCLUSIONS
Trends in demand for agricultural production and technology suggest that
pressure on the nation's land and water resources and on the environment will
increase over the next several decades. Erosion damages likely will be
viewed as_POSinfi a more serious threat to" the environmerr fhan fpri-i H zpt-s _
pesticides or salinity resulting troa irrigation. Except for a few "hot
spots" in California and Nebraska, concentrations of nitrate-N in groundwater
is not now a serious problem. Nitrates in rivers and streams generally are
less than the Public Health Service's standard of 10 ppm except occasionally
in some small mid-western streams.
Consumption of nitrogen fertilizer is expected to increase, but because
of its rising price farmers will have incentive to increase the efficiency
of use. Losses of nitrates to water bodies, therefore, should increase pro-
portionally much less than the increase in nitrogen applied. Nonetheless,
the projected increase in some regions, e.g. the Southeast, is so large that
nitrates entering water bodies may cause some problems.
Phosphorus is the main contributor to accelerated eutrophication.
Municipal and industrial sources contribute more phosphorus to surface waters
than agriculture. Steps already under way to reduce phosphorus discharges
by municipalities and industry are expected to ease the eutrophication
problem. Projections of increased use of phosphorus fertilizers suggest that
any additional phosphorus reaching surface waters from this source will be
small compared to reductons from municipal and industrial sources.
Cotton production is shifting from the Southeast and Mississippi Delta,
where per acre application of insecticides is high, to Texas, where it is
low. In addition, integrated pest.management (IPM), meaning primarily
increased use of consultants to advise on the timing and amount of insecti-
cide use, is spreading in the Delta and Southeast. IPM involves the marginal
substitution of knowledge of plant-insect interactions for rule-of-thumb
spraying, and aims at keeping insect damage below the economic threshold, not
at eliminating.damage completely. The result is more sparing use of insecti-
cides for a given level of control. IPM also has potential for dealing with'
corn insect pests.
Cotton and corn together accounted for 60 percent of all jnsprffp-iHps
applied to crops in 1976. Trends in the location of cotton production and
in insect management practices on cotton and corn thus suggest a significant
decline in total insecticide use, even though corn acreage (but not cotton)
is expected to increase.
xviii
-------
Since the late 1960s farmers have been substituting organophospherus and
carbamate insecticides for the organochlorines. This increases the danger
of acute toxic effects on humans and other non-target species but reduces the
more subtle and pervasive but less lethal effects of the organochlorines.
The outcome could be reduced environmental pressure because the effects of
the organophosphorus and carbamate compounds are readily detectable and
therefore easier to deal with.
Because of the projected decline in total insecticides applied and
change in the nature of the materials used the environmental threat of in-
secticides is expected to lessen.
Herbicide use will increase substantially because of increased acreage
and the spread of conservation tillage. On present evidence this increase
does not appear to pose a significantly greater environmental threat. How-
ever, some potential pathways by which herbicides may impact on the environ-
ment have not been investigated. Consequently the possibility of increased
environmental damages from herbicides cannot be dismissed.
The expansion of land in crops implies an increase in sheet and rill
erosion from cropland from 1.9 billion tons in 1977 to 3.5 billion tons in
2010. Per acre erosion would rise from 4.7 tons to 7.4 tons, jtnd_sediment
delivered to surface wat-firg wnnlH afrput double. Erosion on this scale
likely would be viewed as a major threat to water quality and perhaps also
to the long-term productivity of the land. The traditional voluntary
approaches to er^frfpn mntrr, 1 Hltp)y ftrn.l. he judged inadequate. Regulatory
programs involving penalties for farmers exceeding erosion standards may be
necessary. Such programs will incur high political and administrative costs.
Consequently a policy to invest in development of new technologies which
are both higher yielding and less environmental! damaging than current tech-
nologies could have high social pay-off.
xix
-------
RECOMMENDATIONS FOR RESEARCH
The study demonstrates that trends in demand for agricultural produc-
tion and in technology fundamentally determine the prospects for resource
use and environmental damage in American agriculture. Given the growth of
demand, the expected increase in pressure oh the resource base and on the
environment reflects the prospect that technological change will be rela-
tively land-using and total productivity growth relatively slow. This sug-
gests high social payoff to a research strategy to develop new technologies
that are higher yielding and have higher total productivity than those now
in prospect. Such technologies not only would restrain the rise in economic
costs of agricultural production which now seems in prospect, they also
would ease pressure on the environment by reducing the amount of land needed
to meet projected demand for production. The projected increase in cropland
implies a substantial increase in erosion, which this study concludes is the
major future threat to the environment from agriculture.
The social payoff to an appropriate research strategy would be even
higher if, in addition to high yields and high total productivity, the new
technologies also incorporate features specifically designed to reduce pres-
sure on the environment.
It is easier to make the general case for increased investment in re-
search to.develop new technologies than to specify the particular lines the
research should pursue. Four lines look promising:
1. Increasing yields and total productivity requires that the amount
of energy combined with the land must increase, or that the efficiency of
energy used should rise. The prospective increases in real prices of energy
and fertilizers indicate that simply increasing.the amounts of these re-
sources per acre of land is not the way to go. There is potential for
increasing the efficiency with which energy and fertilizer are used, but
this potential already is incorporated in the projections in this study.
Improvements in the photosynthetic capacity of main crops to cap-
ture and convert the sun's energy into valuable plant material would in-
crease yields without requiring proportional increases in inputs of pur-
chased energy and fertilizer. Improvements in photosynthesis also would
increase the attractiveness of enhancing the capacity of plants (and rela-
ted microorganisms) to biologically fix nitrogen. Biologically fixed nitro-
gen, of course, could substitute for increasingly expensive nitrogen ferti-
lizers. However, increasing the plant's capacity to biologically fix nitro-
gen would demand more plant energy. If photosynthetic efficiency were not
improved, meeting the nitrogen fixing demand for energy would reduce the
supply available for valuable plant growth. In this case, crop yields might
xx
-------
be less with enhanced biological nitrogen fixation than if purchased ferti-
lizers provided the necessary nitrogen.
Research on photosynthesis and biological nitrogen fixation was not
given high priority in the U.S. over the last several decades, perhaps be-
cause the main problem in American agriculture appeared to be too much pro-
ductivity rather than too little. A consequence is that the nation's ca-
pacity to markedly increase research in these two areas is limited by scar-
city of the necessary scientific talent. In addition, the modest amount of
research conducted so far means that there are large gaps in basic scien-
tific knowledge that must be filled before research can move to the stage
of developing new economically attractive technologies. Investment in re-
search in these two areas, therefore, does not promise quick returns.
2. Plant breeding to develop higher yielding varieties of main crops
has been a main source of higher yields and total productivity. There are
promising lines still to be pursued in this area of research. Soybeans are
a major, land using crop for which demand is rapidly increasing. In this
study land in soybeans is projected to exceed that in either corn or wheat
by 15 or 20 million acres by 2010. This occurs even though soybean yields
are projected to increase steadily at the rate established in the last
three decades. Development of a soybean variety with significantly higher
yield potential, while retaining other valuable characteristics, would make
a significant contribution to restraining the rising demand for agricul-
tural land and the attendent increases in economic and environmental costs.
3. On land with moderate to steep slopes conservation tillage greatly
reduces erosion compared to conventional tillage. The expansion of crop-
land projected in this study implies that much land with high erosion poten-
tial will be shifted out of pasture and forest and into crops. The use of
conservation tillage on this land would restrain the increase in erosion.
Conservation tillage presently is not competitive with conventional tillage
on poorly drained land, on land where weeds cannot be adequately controlled
with herbicides, and in areas where the growing season is short. Research
which would overcome these limitations to the spread of conservation tillage
likely would have high pay-off in reduced erosion damages.
A. The spread of conservation tillage implies increased use of herbi-
cides because the technology relies rflore on herbicides and less on culti-
vation than conventional tillage to control weeds. Present evidence does
not indicate that the herbicides in wide use present serious threats to the
environment. With a few exceptions they are not markedly toxic to animals
when applied in generally used amounts, nor do they persist in the environ-
ment for long periods. Studies of their effects on soil microorganisms
indicate that populations may be reduced immediately after application of
herbicides but they recover and show no permanent impairment of important
functions, e.g., fixing nitrogen.
Not all possible ways in which herbicides may impact on soil orga-
nisms have been adequately investigated, however, nor have their effects
on low trophic levels in aquatic environments. Moreover, the projected
increase in herbicide use is large enough to cause concern if there is a
-------
possibility of threshhold effects of herbicides on the environment. The
absence to date of evidence of strong adverse effects may be because these
threshholds have not been passed.
The advantages of conservation tillage in controlling erosion are
so strong that research to extend the technology should be pushed. To avoid
unpleasant surprises from spread of the technology, however, research on
the inadequately explored ways in which herbicides may adversely affect the
environment should be included as an integral part of the research on con-
servation tillage.
xxii
-------
SECTION 1
PROJECTIONS OF PRODUCTION OF GRAINS,
SOYBEANS AND COTTON IN THE UNITED STATES
INTRODUCTION
We are interested in these projections because the levels of production
of these crops will determine in a major way the quantities of land, water,
fertilizers, and pesticides that will be used in the United States. Alterna-
tive quantities of these resources, in turn, have varying economic and envi-
ronmental consequences, the analysis of which is the main purpose of this
study.
We deal with grains (wheat and feedgrains), soybeans and cotton for two
reasons: (1) these crops consistently account for 70-75 percent of the land
harvested in the United States and for high percentages of the fertilizers
and pesticides applied; (2) if the growth of agricultural production in the
United States over the next several decades puts substantial pressure on the
demand for resources and on the environment, it will be because of the growth
of production of these crops, particularly grains and soybeans. The growth
in demand for all other agricultural commodities will respond primarily to
growth in domestic population and income, and almost surely will be moderate.
Under some conditions, however, export demand for grains, soybeans and cotton
could increase enough to stimulate major increases in production of these
crops.
PROJECTIONS TO 1985, 1990 AND 2010
Table 1-1 shows the projections as well as actual figures for 1977-79.
The U.S. Department of Agriculture (USDA) projections were made by the Na-
tional Interregional Agricultural Projections (NIRAP) model, and reflect a
baseline situation. The principal assumptions underlying the baseline pro-
jections are that: (1) U.S. population will grow at the same rate from 1979
to 1985 as from 1970 to 1979 (.9 percent annually) and at a slightly slower
rate (.8 percent annually) from 1985 to 1990; (2) real GNP will grow 3.0
percent per year from 1979 to 1985 and from 1985 to 1990, slightly less than
in the 1970s (3.2 percent annually from 1969-71 to 1977-79); (3) export
growth will be "moderate," i.e., less than in the 1970s.
We accept the USDA projections as a reasonable set of outcomes for 1985
and 1990. We think the projections of exports of feedgrains and cotton may
be somewhat low and those for domestic use of feedgrains a bit high, but in
1
-------
TABLE 1-1. U.S. PRODUCTION, EXPORT, AND DOMESTIC USE OF WHEAT, FEEDGRAINS AND COTTON,
1978/80 AND PROJECTIONS TO 1985, 1990 AND 2010 (millions metric tons)
RFF: 2010
U.S. Share
Average
USDA
Constant
Reduced
1978
1979
1980
1978/80
1985
1990
(1)
(2)
(1)
(2)
Wheat
Prod.
48.3
58.1
64.5
57.0
67.6
77.1
98
100
84
85
Export
32.5
37.4
41.5
37.1
42.4
50.4
70
72
56
57
Dom. Use
22.8
21.3
22.9
22.3
.25.2
26.7
28
28
28
28
Feedgrains*
Prod.
222.1
238.8
198.7
219.9
253.7
282.0
354
428
316
372
Export
60.2
71.4
74.3
68.6
81.0
97.1
167
241
129
185
Dom. Use
157.2
161.9
155.8
158.3
172.7
184.9
187
187
187
187
Soybeans
Prod.
50.9
61.7
49.4
54.0
61.7
72.1
120
129
104
112
Exportt
27.7
32.9
29.5
30.0
27.8
33.8
76
85
60
68
Dom. Use
23.5
23.7
24.0
23.7
33.9
38.3
44
44
44
44
Cotton
(share
not
calculated)
Prod.
2.4
3.2
2.4
2.7
2.6
2.7
3.
,5-3.9
Export
1.4
2.0
1.2
1.5
1.1
1.2
Dom. Use
1.3
1.4
1.2
1.3
1.5
1.5
Sources: 1978-1980 from USDA, Jan. 28, 1981 for wheat and feedgrains; USDA, Mar. 1981 for soybeans
and cotton. 1985 and 1990, USDA projections provided by Leroy Quance, done in the summer of
1980. They represent a "baseline" situation, the assumptions of which are given in the
text. The projections are preliminary and not official.
(continued)
-------
TABLE 1-1. (Continued)
The RFF projections to 2010 are by Pierre Crosson. Constant U.S. shares
means that the U.S. maintains the same percentage of world trade in the
various commodities as in 1976/79; share reduced means a smaller percentage,
as described in the text. The columns (1) assume that the Common Agricultural
Policy of the European Community remains unchanged. The columns (2) assume
that the policy is changed to permit more imports, as described in the text.
For 1978-80 the difference between production and the sum of exports
and domestic use is the change in stocks. In the projections stock changes
are assumed to be zero.
* Corn and sorghum for grain, oats and barley.
t The USDA projections are beans only. The 1978-80 figures and RFF pro-
jections to 2010 are beans plus soybean meal and oil exports converted to the
bean equivalent.
3
-------
neither case are they implausible and the totals for 1985 agree closely with
previous projections by RFF."*"
There are no up-to-date USDA projections for 2010 so we made those shown
in Table 1 independently. The projections for grains and soybeans were made
in three steps: (1) project growth in world trade in each commodity; (2) pro-
ject the U.S. percentage of trade; (3) project domestic use. The projections
for cotton were derived from a USDA study (Collins e_t al. , 1979) described
below.
The detailed derivation of the projections is described in the appendix
to this section. Here we discuss some general considerations relevant to
evaluation of the projections.
Sources of Error in Projections
The projections game is one many can, and do, play, and there is no firm
basis for distinguishing the quality of the players. The range of equally
plausible outcomes is wide because the range of plausible assumptions deter-
mining outcomes is wide. Over periods of twenty or thirty years, small
differences in assumed rates of increase in population and per capita income,
in income elasticities of demand, or in rates of growth of food production
can yield large differences in projections of world trade, a key element in
our, and other, projections.
When projections depend upon extrapolation of trends—and all do to some
extent—problems arise if the trend is not clear. Which periods should be
selected to represent the trend if the data suggest the trend has shifted
over time? Ordinarily, the most recent period should be used, but one fre-
quently cannot be sure whether recent experience marks the beginning of a new
trend or only a momentary departure from established trend. This is an
especially difficult problem with trends in agricultural productions because
the data may reflect variations in the weather. For example, a combination
of unusually bad weather early in the period and unusually good weather later
in the period can give a quite marked, and misleading, upward tilt to the
trend in crop production.
Projections over periods as long as twenty or thirty years may also go
wide of the mark because of the emergence of wholly unforeseen factors.
Changes in dietary standards and habits could have a major impact on pro-
jections of food consumption and production. For example, our projections
of feedgrain consumption in high-income countries assume that per capita
Crosson and Brubaker, May, 1979. Our projections in the referenced
document were for 1985 and 2000 and were for grains and soybeans only. The
ratios of our production projections for 1985 to those of the USDA in Table 1
were as follows: wheat, .984; feedgrains, 1.016; soybeans, 1.036. There
was no trend in per capita domestic use of feedgrains between 1960 and 1979,
so we projected this component of feedgrains in accordance with population
growth. The USDA projections implicitly assume an increase in per capita
domestic use of feedgrains.
4
-------
consumption of meat in those countries will rise toward U.S. levels, thus in-
creasing per capita consumption of feedgrains and soybean meal. Trends in
meat, feedgrain and soybean meal consumption in those countries indicate this
is a plausible assumption, but concern about the health effects of high-meat
diets may dampen those trends, leading to slower than projected growth of .
demand for feedgrains and soybean meal.
The emergence of new uses for crops may have the opposite effect. The
prospect for long-term increase* in the real price of fossil fuels has stimu-
lated growing interest in alternative sources of energy. The use of crops to
produce ethanol for blending with gasoline to form gasohol is an example. At
present gasohol requires a subsidy to be competitive with gasoline, but con-
tinuing increases in the price of gasoline and progress in gasohol technology
could eliminate this need. This is discussed further below.
The development and spread of new, and unanticipated, production tech-
nologies also may cause outcomes to differ widely from projections.
Unexpected breakthroughs in research on photos.ynthetic efficiency or in
development of new seed varieties conceivably could lead to faster than pro-
jected growth in food production and slower than projected growth in import
demand in importing countries. Given the present state of knowledge of these
processes, such occurrences are not likely to affect five-to-ten-year pro-
jections, but for periods of twenty-to-thirty years they cannot be ruled out.
However, since there is no way of knowing when these developments may occur
or what their production, consumption, and trade consequences may be, there
is no point in trying to incorporate them in projections.
Policy changes also can cause projections to look foolish. The only
policy change incorporated in our projections assumes more liberal trade and
lower price policies in the European Community (EC) after the mid-1980s. We
did this because of evidence of increasing concern in the EC about the high
cost of present policies and the growing strength of consumer interests
seeking lower real prices of food. We, of course, do not project that EC
policies will change in the indicated direction. But, in our judgment, the
likelihood of such change is sufficiently high and its implications for trade,
particularly in feedgrains, sufficiently important that we ought to incor-
porate the change in our projections to 2010. Other policy changes, of
course, are conceivable, although none that we have thought about are suffi-
ciently likely, in our judgement, to warrant inclusion in the projections.
We may be wrong, however. If so, then our projections will prove to be in-
correct.
For all of these reasons we hold no brief that our projections are neces-
sarily more plausible than others that have been or might be done, and we are
confident that the projections will be wrong in ways we cannot now foresee.
We believe, however, that our projections are internally consistent and rea-
sonably in accord with present understanding about trends and policies
bearing on U.S. production of grains, soybeans and cotton. That is all we
claim for them.
5
-------
Comparison . with Others
Table 1-2 provides a basis for comparing our projections with those by
the USDA and by Martin Abel. The USDA projections are those for 1990 shown in
Table 1. Abel's are for 2005 and are in a paper prepared for a conference
held by RFF on the adequacy of agricultural land in the United States. For
comparison with our projections we extrapolated those of the USDA and Abel
assuming that the annual increments of production from 1977/79 to their tar-
get dates would continue to 2010. Thus the projections for 2010 labeled
USDA and Abel in Table 1-2 are not their work but extrapolations by us of
their projections.
TABLE 1-2. ALTERNATIVE PROJECTIONS OF U.S. PRODUCTION OF WHEAT,
FEEDGRAINS AND SOYBEANS (millions metric tons)
2010
USDA
Abel
Extrapolations
of
1977/79
1990
2005
USDA
¦ Abel
RFF
Wheat
54.2
77.1
83
115
88
98
Feedgrains
218.8
282.0
405-435
387
439-475
354
Soybeans
53.5
72.1
64-76
103
66-80
120
Sources: 1977/79 - USDA, January 23, 1981
USDA - from Table 1, this report
Abel - paper given at an RFF conference on the adequacy of
agricultural land (see Abel in references)
RFF - from Table 1, this report. See appendix to this
section for details on derivation.
Note: Abel projects oilseed production, not soybeans only. Typically,
soybeans account for 85-90 percent of oilseed production in the
United.States. We have assumed that soybeans are 85 percent of
Abel's projection for 2005.
The extrapolated USDA projections of feedgrains and soybeans are much
closer to ours than are Abel's. Abel assumes a major increase in production
of corn to produce ethanol for combination with gasoline to produce gasohol.
He also anticipates a significant increase in corn production for use as a
sweetener. Both uses of corn yield a high protein residue which competes
directly with soybeans as a feed supplement. Abel's assumptions about the
growth of demand for corn for production of ethanol and sweetener thus
accounts for much of the difference between his extrapolated projections and
ours for feedgrains and soybeans in 2010. The extrapolated USDA projections
for feedgrains and soybeans differ from ours in the same way as Abel's, but
the amount of the difference is much less. The USDA projections also assume
6
-------
increased production of corn for ethanol production, but evidently on a
smaller scale than Abel assumes.
Our projections make no special allowance for use of corn to produce
ethanol or sweeteners. In the 1980s ethanol production based on corn almost
surely will expand substantially. Whether this will constitute a permanently
higher demand for corn, however, is not clear. At a meeting of the Bio-
Energy World Congress and Exposition, held in Atlanta in April, 1980, most
speakers agreed that wood eventually will prove more economical than grain as
a source of biomass for energy production (Science, page 1018). Abel, in
discussion with one of the author's, agrees that there is great uncertainty
about the long-run course of ethanol production, and that by 2010 wood, or
more likely coal, may well replace corn as the main nontraditional feedstock
for production of liquid fuel. Abel believes, however, that the corn based
production capacity built in the 1980s will be maintained. We are not per-
suaded on this latter point. By the spring of 1981 much of the ardor for
gasohol evident in mid-1980 appeared to have cooled. Informal reports re-
ceived by the authors indicated that some of the plant expansion plans
announced in 1980 were being cut back or deferred. The Reagan administration
was taking a more cold-eyed look than its predecessor at the subsidies in the
gasohol program. All of this, combined with the prospect that by the 1990s
alternative feedstocks will prove more economical than grain, suggests to us
that by 2010 the demand for grain for ethanol will be insignificant. This is
the justification for our decision not to make special allowance for corn-
based production of ethanol in our projections of feedgrain production. It
follows that we would not make a corresponding adjustment in our projections
of soybean production.
With allowance for the difference with respect to treatment of the role
of corn in ethanol production we believe that Table 2 indicates that our pro-
jections are generally consistent with those of the USDA and Abel. The
differences in the total tonnage of grains and soybeans are small for thirty
year projections. This, of course, does not mean that we are right, but it
does suggest that our thinking about the long-run prospects for U.S. produc-
tion of grains and soybeans is not grossly different from that of others who
have labored in this field. In work of such great uncertainty only gross
differences should be taken as significant.^
2
Since this was written the National Agricultural Lands Study (NALS,
1981) published a set of USDA projections for 2000 which imply substantially
higher rates of growth in production and exports than any of those shown in
Table 1. Specifically, the NALS median projections (a range is given) are
2.7 percent annually for total production from 1980 to 2000 and 5.5 percent
annually for total exports. The RFF 2010 projections (Column 1, U.S. share
constant) in Table 1 imply the following annual growth rates from 1980:
Production
Exports
(percents)
Wheat
Feedgrains
Soybeans
Cotton
1.4
1.9
3.0
1.5
not projected
1.8
2.7
3.2
(continued)
7
-------
The RFF projections in Table 1-1 labeled "U.S. Share Constant" are based
on no explicit assumptions about crop prices except that their influence on
the growth of domestic and foreign demand will be small relative to the in-
fluence of world population and income growth. In fact, we think real prices
received by American farmers for grains, soybeans, and cotton are more likely
to rise over the next several decades than they are to remain at 1979 levels.
The reason is our expectation that real prices of key inputs—energy, ferti-
lizer, water and land—are likely to increase more than productivity. The
basis for these expectations is given in subsequent chapters.
If U.S. production costs rise more than our constant share projections
implicitly assume, U.S. shares of world trade in grains, soybeans, and cotton
could decline. To allow for this possibility we prepared projections of U.S.
production, exports, and domestic use assuming a diminished share of world
trade. These are labled "U.S. Share Reduced" in Table 1-1. Obviously we
think the likelihood of the United States losing ground in world trade is
sufficiently high to warrant presentation of the alternative projections.
However, in analyzing the demand by U.S. fanners for resources and the envir-
onmental impacts of production we focus only on the constant share projections
assuming no change in the EC's Common Agricultural Policy. This set of pro-
jections in our judgment is the most probable set. It would have been
interesting to explore the resource and environmental implications of the
alternative projections, but our resources did not permit this.
The USDA model which generated the projections for 1985 and 1990 shown
in Table 1-1 also generates prices for wheat, feedgrains and soybeans. . These
are shown in Table 1-3 along with 1979 prices. The model indicates a decline
in real prices of wheat and increases' for corn, sorghum, oats, barley, and
soybeans. Weighting the four feedgrains by their shares of total feedgrain
production of 1979 indicates an increase in average real prices of feed-
grains of about 22 percent from 1979 to 1985 and about 31 percent from 1979
to 1990. We have not explored in detail why the NIRAP model shows rising
real prices for feedgrains and soybeans, but the general reason is rising
real prices of key inputs not sufficiently offset by increases in produc-
tivity.
(Footnote 2 continued)
Demand is projected to grow more rapidly for grains and soybeans than for
other components of agricultural production. Consequently the differences
between the RFF projections of those commodities and the projections of them
embodied (but not shown separately) in the NALS projections are greater than
the above percentages indicate. The NALS projections assume constant rela-
tive prices and ours allow for some real increase in prices,- as noted below.
Nonetheless, it is likely that the constant price equivalents of the RFF pro-
jections would be less than the NALS projections. By that comparison,
therefore, the RFF projections probably are conservative.
8
-------
TABLE 1-3. CROP PRICES RECEIVED BY FARMERS, DOLLARS OF 1979 ($/bu.)
1979
1985
1990
Wheat
3.82
3.09
3.33
Corn
2.41
3.01
3.22
Grain sorghum
2.33
2.70
2.89
Oats
1.36
1.38
1.47
Barley
2.31
2.32
2.47
Soybeans
6.19
7.36
7.85
Sources: 1979, USDA, June 1980.
The data for 1985 and 1990, from print-out of projections of
the USDA s NIRAP model, provided by Leroy Quance (see note to
Table 1-1). The NIRAP model generates projections of nominal
crop prices and of the implicit price deflator for GNP. The
latter was used to convert nominal prices to "real" prices,
i.e. prices expressed in 1979 dollars. The NIRAP model does
not generate prices for cotton.
Martin Abel points out that wheat is not likely to sell for less
than its value as feed. In this case the projected prices of corn are
too low or those for wheat are too low.
Martin Abel's projections, shown in Table 1-2, explicitly assume no
change in real crop prices between 1980 and 2005. If real prices in fact
change;, as the NIRAP projections indicate and we think likely, Abel's pro-
jections would be modified accordingly. He asserts (Abel, p. 1) as follows:
Our projections for the U.S. and the world would have to
be modified if alternative real price assumptions were
used, and the modifications would depend upon whether the
price change was due to factors affecting either demand or
supply.
In summary, the projections we use to examine the resource and environ-
mental implications of increased agricultural production in the United States
do not exclude the possibility of changes in real crop prices. But they do
assume that the effect on demand of any such changes will be small relative
to the effects of world population and income growth, and that, in particular,
any such changes will not significantly alter the U.S. share of world trade
in wheat, feedgrains, and oilmeal.
9
-------
REFERENCES
Abel, Martin. 1980. Growth in Demand for U.S. Crop and Animal Production
by 2005. Paper given at a national conference on the. Adequacy of
Agricultural Land sponsored by Resources for the Future, June 19-20,
1980. Washington, D.C.
Collins, K. J., ;R. B. Evans and R. D. Barry. 1979. World Cotton Production
and Use: Projections for 1985 and 1990. USDA, Foreign Agricultural
Economic Report No. 154. Washington, D.C.
Crosson, Pierre and Sterling Brubaker. May 1979. Resource and Environmental
Issues of Agriculture in the United States: an Interim Report. Resources
for the Future, unpublished. Washington, D.C.
National Agricultural Lands Study. 1981. Final Report. A joing effort by
the U.S. Department of Agriculture, Council on Environmental Quality
and ten other federal government agencies. U.S. Government Printing
Office, Washington, D.C.
Science. May 30, 1980, Vol. 208.
USDA. January 15, 1980. Crop Production 1979 Annual Summary. CrPr 2-1(80).
Washington, D.C.
. June 1980. Agricultural Prices Annual Summary 1979. Pr 1-3(80).
Washington, D.C.
. January 28, 1981. Foreign Agricultural Circular Grains.
FG-4-81. Washington, D.C.
. March 1981. Agricultural Outlook. Washington, D.C.
10
-------
SECTION 2
FACTORS AFFECTING FARMERS' CHOICES AMONG
TECHNOLOGIES: PRICES OF INPUTS
INTRODUCTION
Meeting our projected levels of production of grains, soybeans and cot-
ton will increase the demand by American farmers for resources. The amount
of the increase for each type of resource depends upon the kinds of tech-
nologies farmers choose. We assume that in making these choices, farmers
seek to maximize their incomes, and that the choices are conditioned funda-
mentally by the prices and productivities of alternative technologies. In
this section we consider prices of certain key inputs. Productivity is con-
sidered in the next section. Farmers' choices also are affected by public
policies. In most cases these will be reflected in prices of technologies.
In some, however, it will not, as when the Environmental Protection Agency
(EPA) bans an insecticide. In these cases policy is a separate factor in-
fluencing farmers' choices, and we treat it as such where appropriate.
TYPES OF TECHNOLOGY
Technologies frequently are labeled according to the intensity of use
of some particular input relative to other inputs. For example, a technolo-
gy using much labor relative to capital is called labor-intensive, or labor-
using, to distinguish it from another•technology in which the ratio of labor
to capital is lower. Similarly, technological innovations may be described
as "saving" or "using" with respect to some particular input, depending upon
whether the proportion of that input to other inputs is less or more in the
new technology than in the one it replaces.
We find it useful to distinguish agricultural technologies as land-using
or land-saving, according to whether the ratio of land to other inputs is
high or low. The implication is that technologies should be thought of as
lying along a spectrum ranging from very land-using (high ratios of land to
other inputs) to very land-saving (low ratios of land to other inputs). A
given technology will be land-saving relative to some technologies and land-
using relative to others.
The distinction between the two kinds of technology lacks precision. It
is useful, however, for analysis of the land use and environmental implica-
tions of expanding agricultural production. To meet a given increase in
production, land-using technologies imply slower growth of yields, hence
11
-------
greater demand for land than land-saving technologies. The distinction is
useful also because the environmental implications of the two kinds of tech-
nology are different. With land-using technologies, problems of erosion
typically will be more important relative to problems of agricultural chemi-
cals than with land-saving technologies.
As with any analytical device, the distinction between land-using and
land-saving technologies will not catch all the nuances of reality. For
example, American farmers who expand crop production with land-using tech-
nologies typically will also apply fertilizers and pesticides to the land.
Consequently, the land-using mode of expansion may entail environmental costs
of fertilizer and pesticide pollution as well as costs of erosion. In the
same sense, farmers who adopt the land-saving mode may nonetheless cause
increased erosion by abandoning erosion control practices, such as strip-
cropping, in an effort to maximize production per unit of land. Since either
mode of expansion may entail environmental costs associated primarily with
the other, the two categories of technology should be regarded as polar
cases, the actual technologies employed being chosen from the spectrum in
between.
CHOICES AMONG TECHNOLOGY: A HISTORICAL PERSPECTIVE
Table 2-1 indicates that between 1951-1955 and 1972, the trend of Ameri-
can farmers' choices among technologies was weighted toward the land-saving
end of the spectrum. In this period the ratio of all purchased inputs (not
including labor) to cropland used for crops rose 61 percent. The increase
in the ratio of fertilizer to cropland was particularly high—266 percent.
TABLE 2-1. RATIOS OF INPUTS TO CROPLAND USED FOR CROPS
IN U.S. AGRICULTURE (1967=100)
All
Pur-
Fertilizer*
chased
Inputs
1951-55
35
67
1956-60
46
77
1961-65
70
92
1966-70
108
102
1971
123
105
1972
128
108
1973
123
105
1974
129
102
1975
115
99
1976
136
106
1977
141
109
1978
134
116
1979
140
119
Source: USDA, February 1981.
*Per acre of cropland harvested.
12
-------
This trend toward land-saving technologies from the beginning of the
1950s to 19.72 was consistent with the behavior of relative input prices in
this period. Table 2-2 supports this. The prices of all purchased inputs
and of fertilizers fell relative to the price of land, measuring the latter
by the per acre value of farmland and buildings (Land^ in Table 2-2).
The prices of purchased inputs and of fertilizer rose from 1951/55 to 1956/60
relative to the capitalized value of land rents (Landg) but fell thereafter
to the early 1970s. Relative to the Land^ price purchased input prices and
prices of fertilizer continued to fall after the early 1970s, but at a much
slower annual rate than in the two previous decades. By contrast, purchased
input prices and fertilizer prices rose relative to the capitalized value of
land rents after 1974, reversing the previous trend. The behavior of input
prices relative to land prices since the early 1970s thus is consistent with
the slower shift to land-saving technologies in that period, the price
effects being particularly marked when price is measured by the capitalized
value of land rents.
TABLE 2-2. RATIOS OF INPUT PRICES TO PRICES OF LAND
AND PRICES OF FERTILIZER
(1967 = 100)
Relative
to Land
Relative to
Land.
A
Landg
Fertilizer
All
All
All
purchased
purchased
purchased
inputs
Fertilizer
inputs
Fertilizer
inputs
1951-55
180
207
181
203
89
1956-60
147
167
294
322
91
1961-65
122
133 .
223
237
94
1966-70
95
87
159
145
110
1971
91
73
177
143
124
1972
89
69
190
148
128
1973
91
64
103
72
143
1974
85
85
57
57
100
1975
83
98
71
85
84
1976
76
73
94
90
104
1977
69
63
129
117
110
1978
68
56
163
136
120
1979
68
54
na
na
127
Sources: Land^, all purchased inputs and fertilizer from USDA, Agricultural
Statistics, 1972 and 1979. Landg from Castle and Hoch, 1980.
Land^ is an index of the value per acre of farmland and buildings,
Landg is the capitalized value of annual net rents earned by farm-
land, calculated by Castle and Hoch from USDA data.
13
-------
The implications for farmers' choices among technologies of the price
ratios in Table 2-2 are not as unambiguous as they may seem. The reason is
that land prices are themselves affected by farmers' choices\but prices of the
other inputs are not, at least not in the long run. Prices of fertilizer and
other farm inputs are determined primarily by the costs of producing these
inputs, or in any case by factors other than farmers' demand for them."*"
This is not true of land prices, however. The greater part of the demand
for farm land comes from farmers.^ There is interdependence, therefore, be-
tween the price of land and farmers' choices between land-using and land-
saving technologies. This is not true (or not true to nearly the same extent)
of nonland inputs, for which farmers are price takers.
The implication is that movements in the price of farm land reflect move-
ments in the demand and supply curves of land relative to one another. For
this reason our analysis of the role of land in farmers' future choices be-
tween land-saving and land-using technologies focuses,on factors affecting
the demand curve for land and the resulting supply response. Drawing on the
analysis in this and the following section of the future behavior of the
prices and productivity of non-land inputs we project the demand for land
needed to produce our projected amounts of crop output. In a subsequent
section we examine the supply response of cropland and make a judgment about
the effect on land price of the supply-demand interaction.
In this section we focus on the prices of fertilizer and energy, and on
the cost and availability of water for irrigation, because these are the
principal land-saving inputs.
PRICES OF INPUTS
Energy
American farmers use large amounts of energy to drive tractors and other
farm machinery, to pump irrigation water, to dry crops, and in the form of
nitrogen fertilizer and pesticides. With respect to farmers' choices among
land-using and land-saving technologies, the two most important uses of en-
ergy are for producing nitrogen fertilizer and for pumping irrigation water.
This probably is less true of fertilizer prices than of other' inputs.
The United States accounts for 20-25 percent of world consumption of nitrogen
and potash fertilizers, and for about 20 percent of consumption of phosphorus
fertilizers. These amounts suggest that short-term shifts in U.S. demand
likely would affect fertilizer prices.
7
Duncan (June 1977) indicates that in 1975, 63 percent of farm real
estate transfers in the United States were to farmers, about the same as in
the previous three decades.
^USDA (1976) shows that in 1975 fertilizer accounted for 30.8 percent of
energy used in farm operations in the United States and irrigation for 12.9
percent. Field operations took 25.9 percent, presumably for tractors and
other farm machinery.
14
-------
According to Dvoskin and Heady, about 30 cubic feet of natural gas are used
on average to produce one pound of nitrogen fertilizer. In the winter of
1979/80 farmers in the High Plains were paying about $2.00 per thousand cubic
feet (mcf) for natural gas.^ At that time, the nitrogen content of anhydrous
ammonia, the most widely used form of nitrogen fertilizer, was valued at
about $.13 per pound.At these prices, therefore, natural gas constituted
a little over 45 percent of the price of anhydrous ammonia. These numbers
are only illustrative, but they make the point that the cost of natural gas
has an important influence on the price of nitrogen fertilizer.
Table 2-3 shows an index of the real prices paid by farmers for fuel and
energy from 1965 to 1979. This price series dates only from 1965, but there
is little doubt that real energy prices paid by farmers declined more or less
steadily from the beginning of the 1950s to the early 1970s.® After 1973,
however, these prices rose sharply and in 1979 were 46 percent above 1973.
The decline in real energy prices from the 1950s to the early 1970s con-
tributed importantly to the decline in the real price of fertilizers in this
period (noted above) and to the spread of irrigation (detailed later in this
section). Should real energy prices continue to rise, the trends in prices
of nitrogen fertilizer and of irrigation are likely to be less favorable to
the adoption of land-saving technologies in the future than they were in the
two decades ending in the early 1970s.
TABLE 2-3 INDEX OF REAL PRICES PAID BY FARMERS FOR FUEL
AND ENERGY IN THE UNITED STATES (1967 = 100)
Year
Index
Year
Index
1965
104
1972
86
1966
101
1973
87
1967
100
1974
108
1968
97
1975
110
1969
93
1976
110
1970
89
1977
111
1971
88
1978
108
1979 127
Sources: USDA Agricultural Statistics, 1980. The "real" price is
the index of nominal prices divided by the Consumer Price
Index. The series for fuel and energy prices paid by farmers
begins in 1965.
^Information obtained from officials in Texas, Oklahoma, Kansas, and
Nebraska in January 1980.
"'The average price of anhydrous ammonia in the winter of 1979/80 was
about $.11 (USDA, May 30, 1980, p. 44). Nitrogen is 82 percent of anhydrous
ammonia so a pound of N in anhydrous ammonia cost $.13.
^In real terms the electricity, gas, fuel oil, and gasoline-components of
the Consumer Price Index (CPI) were all lower in the early 1970s than in
1950. Energy prices paid by farmers must have moved in a similar fashion.
15
-------
There is a consensus that the real price of energy will continue to rise.
This is evident from a major study done at Resources for the Future of energy
alternatives for the United States (Schurr, et al.). The study includes a
review of studies done by others of energy prospects for the United States to
the end of the century. All of the studies assume that real energy prices
will rise from 1975 to 2000. There are differences among the projections in
the amount of price increase, but all agree that real natural gas prices will
rise most and coal and electricity prices least. The average rate of in-
crease for all energy sources is roughly 2 percent per year from 1975 to 2000.
It should be noted that none of these projections of real energy prices
emerges directly from analyses of trends in world energy supply and demand.
Instead they are based on past trends and on informed judgments about the
effect on prices of trends in technology, of known reserves and new discover-
ies, OPEC policies, and so on. The authors of each of the studies obviously
believe that the price projections are consistent with projections of energy
demand and supply, but they are not derived from the analysis of demand and
supply. We make this point not to cast stones at the price projections—we
would do them in the same way—but to emphasize that the projections are
based on ad hoc judgments rather than on systematic analysis of underlying
relationships and tested hypotheses about the determinants of energy prices.
The large role of judgment relative to analysis means that faith (or lack of
it) weighs more heavily in acceptance or rejection of the projections than
it would if the roles of judgment and analysis were reversed. We have no
basis for questioning the judgments, but we recognize that we accordingly
accept the price projections largely on faith.
We assume, therefore, that real energy prices paid by American farmers
(directly or indirectly) will rise steadily to the end of the century and
beyond at a rate of some 2 percent annually, that prices of natural gas will
increase substantially faster than 2 percent (without specifying how much
faster), that electricity prices will rise by less than 2 percent annually,
and that prices of other energy sources will rise at rates somewhere between
those for electricity and natural gas.
Fertilizers^
While energy is an important component in the prices of fertilizers,
particularly nitrogen, other factors also are important. It now is generally
agreed that the run-up in fertilizer prices in 1974 and 1975 resulted more
from rising pressure of demand on supply than from the increase in energy
prices in those years. Similarly, the decline in fertilizer prices after
1975 occurred despite continued increases in energy prices. Thus, while our
This section is based primarily on USDA (Dec. 1978b) and on discussions
with people at USDA, the World Bank, and the International Fertilizer Devel-
opment Center at Muscle Shoals. The USDA does much work on projections of
quantities of fertilizers supplied and demanded, but little or nothing on
prices, at least for publication. Of the three institutions contacted, only
the World Bank currently projects world fertilizer prices, and these are
not published.
16
-------
projections of rising real prices of energy imply upward pressure on ferti-
lizer prices, particularly for nitrogen, other factors may offset this
pressure. Projections of fertilizer prices must take these other factors
into account as well as trends in energy prices.
Table 2-4 shows measures of the "real" price of fertilizers in the
United States. Prices are shown relative to the cost-of-living index and
relative to the price of corn. The relation to the price of corn probably
carries more weight in farmers' decisions about the use of fertilizer. The
significant point in the table is that since the mid-1970s the real price of
fertilizer, whether measured against the CPI or the price of corn, has been
sharply higher than in 1970. Projections in a study of demand and supply in
the world fertilizer industry indicate that the growth of world capacity to
produce nitrogen fertilizers would exceed the growth of consumption from 1978
to 1985, but that for phosphate and potash fertilizers, capacity would in-
crease less than consumption (Harris and Harre). The ratios of consumption
to capacity in 1978 and 1985 are as follows:
8
Ratios of World Fertilizer Consumption to Capacity
1978 1985
Nitrogen .97 .93
Phosphate .88 .94
Potash .92 1.04
These ratios suggest that, other things the same, nitrogen prices would
be about the same to slightly weaker in 1985 compared to 1978, but that
prices of phosphate and potash fertilizers would be higher. "Other things,"
however, are not likely to be the same. The anticipated increase in the real
price of natural gas would tend to increase the cost of production of nitro-
gen fertilizers, quite apart from trends in consumption and capacity in that
industry. Studies done at the World Bank conclude that rising costs of con-
struction of new plants also likely will put upward pressure on real prices
of nitrogen fertilizer, and also on prices of phosphates and potash. Accord-
ing to these studies, real.prices of the fertilizers in 1985 could be 15 to
20 percent above the levels of 1978. While projections beyond 1985 are not
available, we assume the increase in energy prices will put steady upward
pressure on fertilizer prices, particularly for nitrogen, and that these
prices will rise in real terms to the end of the century and beyond.
Water
Irrigation is a highly land-saving technology, substituting a controlled
water supply and other inputs for land. However, little of the water used
for irrigation in the United States is traded through markets. Most of it
either is pumped from aquifers or rivers by the farmers who use it or is pro-
vided by publicly funded surface water systems at highly subsidized prices.
Hence prices paid by farmers for irrigation water, where they exist at all,
bear little relationship to the true scarcity value of the resource. However,
where farmers are free to sell their rights to water to nonagricultural users
the water has an opportunity cost to the farmer, the amount depending upon
the strength of the nonagricultural demands. Hence, trends in those demands
^Harris and Harre, p. 23-27.
17
-------
TABLE 2-4. INDEXES OF REAL PRICES OF FERTILIZERS
IN THE UNITED STATES (1970 = 100)
Nitrogen* All Fertilizer
Relative
Relative
Relative
Relative
to
to price
to
to price
CPI
of corn
CPI
of corn
1970
100
100
100
100
1971
102
131
99
127
1972
100
94
99
91
1973
103
61
101
61
1974
193 ¦
108
149
84
1975
256
186
178
129
1976
174
158
i43
130
1977
155
160
132
135
1978
133
142
122
121
1979
134
148
119
122
Source: Nitrogen prices are from USDA (Dec. 1978b and June 1980);
1977, 1978 and 1979 are averages of prices of May 15 and
October 15 in each year; consumer price index, corn prices
and index of all fertilizer prices from USDA, Agricultural
Statistics 1980.
^Anhydrous ammonia, which accounts for almost 60 percent of nitrogen
fertilizer production in the United States.
are an indicator of trends in the "price" of water used for irrigation. The
other principal indicator is the cost of pumping groundwater. Accordingly,
in this discussion we focus on the increasing competititon for water for agri-
cultural uses and on the costs of pumping groundwater.
Table 2-5 gives information about irrigation in the United States in
1977, and Table 2-6 shows irrigated and rainfed land, by region, in corn,
soybeans, wheat, and cotton in 1950 and 1977. We deal first with western
irrigation and subsequently with irrigation in the East.
Western Irrigation—
Table 2-6 is of particular interest because of the substantial increase
it shows between 1950 and 1977 in irrigated acreage in the West of crops of
interest to this study.9 This increase was stimulated by easy access to low-
cost water. In the West water has commonly been treated as a free good to
9 !
The West means the seventeen westernmost states. This account of the
role of irrigation in the West draws heavily from Frederick. Irrigated
production of soybeans is trivial in the West.
18
-------
TABLE 2-5. 1977 IRRIGATED AND DRYLAND AGRICULTURAL LAND USE BY FARM PRODUCTION REGION
(millions of acres)
Farm
Production
Region
Irrigation
Dryland
Total
Cropland Pasture
Total
Cropland
Pasture
Total
Cropland
Pasture
Total
Northern Plains
10.6
0.1
10.7
83.9
9.4
93.3
94.6
9.5
104.1
Southern Plains
8.6
0.4
9.0
33.6
27.1
60.7
42.2
27.5
69.7
Mountain
15.2
2.0
17.2
27.1
5.5
32.6
42.2
7.4
49.6
Pacific
11.9
1.4
13.3
11.2
2.8
14.0
23.2
4.1
27.3
17 Western States
46.4
3.9
50.2
155.8
44.7
200.5
202.2
48.5
250.7
Northeast
0.4
0
0.4
16.5
5.8
22.3
16.9
5.8
22.7
Appalachian
0.4
0
0.4
20.4
18.5
38.9
20.8
18.5
39.3
Cornbelt
1.1
0
1.1
88.8
25.2
114.0
89.9
25.2
115.1
Lake States
1.0
0
1.0
43.2
6.9
50.1
44.1
6.9
51.0
Southeast
2.4
1.0
3.4
15.1
13.1
28.2
17.5
14.1
31.6
Delta States
4.0
0.1
4.1
17.2
12.6
29.8
21.2
12.7
33.9
31 Eastern States
9.3
1.1
10.4
201.2
82.1
283.3
.210.4
83.2
293.6
National Total
55.7
5.0
60.7
357.2
127.8
486.1
412.9
132.7
546.9
Source: Tables 3a and 4a, USDA, February 1980.
Note: Minor differences in totals are due to rounding.
-------
TABLE 2-6. IRRIGATED AND DRYLAND ACREAGE OF CORN, SORGHUM, WHEAT, AND
COTTON FOR 1950 AND 1977
(1000 harvested acres)
1950 1977
Corn:*
Irrigated West
46
8,838
Dryland West
16,933
4,572
East
55,486
56,596
U.S. Total
72,465
70,006
Sorghum:*
Irrigated West
894
1,847
Dryland West
9,435
10,774
East
13 ¦
1,444
U.S. Total
10,342
14,065
Wheat:
Irrigated West
911
3,899
Dryland West
48,555
49,387
East
12,141
12,930
U.S. Total
61,607
66,216
Cotton:
Irrigated West
1,563
3,868
Dryland West
6,949
5,212
East
9,331
4,199
U.S. Total
17,843 .
13,279
Sources: The total acreage estimates for the West and United States
in 1977 are from USDA, Agricultural Statistics 1978. The
1977 irrigated and dryland acreages are based on data from
the Agricultural Census of 1974, adjusted by information
from the National Resources Inventory (NRI). Details are
in Frederick, notes 4 and 5. The 1950 estimates also
are from Frederick.
*For grain only.
anyone able to transport it to the place of use. In addition, western irri-
gation benefitted from substantial subsidies from federally developed irri-
gation projects, which now water about one-fifth of the land irrigated in the
region. Cheap energy also spurred the growth of irrigation, especially from
groundwater.
The days of cheap water in the West appear to be over. Current demands
on western rivers commonly exceed available supplies, and groundwater tables
20
-------
are declining in some important areas. In addition, demands for nonagri-
cultural uses of water are rising rapidly. The value of water in these uses
typically is much higher than in agriculture, making it difficult for farmers
to compete on economic terms with nonfarm demanders of water. For example,
water costs of $100 per acre-foot are insignificant to the costs of mining
or processing western coal, but costs of even half this much are prohibitive
to most farmers. In areas of the West where nonagricultural demands for
water have risen sharply, some farmers already have found it profitable to
sell their water rights and many others are likely to do the same unless legal
constraints prohibit them. Because the amount of water used in western agri-
culture is so much greater than that in nonagricultural uses,, large percentage
increases in the latter could be met by relatively small diversions from
agriculture. Hence, no large general reduction in the amounts of water used
for irrigation in the West is expected. The shifts out of irrigation may be
significant in some local areas, however, and the general effect is likely to
be slower growth in irrigation than otherwise would occur.
Groundwater emerged as an important source of water for irrigation in
the West in the 1930s. In the 1970s the expansion of irrigation was based
largely on groundwater, and this now is the principal source of supply in
many areas of the West. Pumping commonly exceeds aquifer recharge, the
highest rates of such "mining" being in the High Plains of Texas, Oklahoma,
New Mexico, Colorado, and Kansas, and in Arizona. These areas account for
20 percent of the irrigated land in the West.
Mining does not threaten to exhaust groundwater supplies in the foresee-
able future, but it lowers water tables, thus increasing costs of pumping.
Pumping costs promise to rise also because of increasing energy costs.
Tables 2-7 and 2-8 show, respectively, the areas irrigated with pumped water
by type of energy in 1974, and energy requirements for on-farm pumping of
water. The tables refer to the whole United States, but most water pumped
for irrigation is used in the West. Whether in terms of area irrigated or
TABLE 2-7. ACRES IRRIGATED WITH ON-FARM PUMPED WATER
IN THE UNITED STATES BY TYPE OF ENERGY, 1974
(1,000 Acres)
Ground-
Surface
Percent of all
Energy
water
water
Total
energy types
Electricity
11,721
3,900
15,621
44.5
Diesel
2,693
1,242
3,935
11.2
Gasoline
1,025
523
1,548
4.4
Natural gas
9,687
948
10,635
30.4
LPG
2,585
753
3,338
9.5
Total
27,711
7,366
35,037
100.0
Source: Sloggett (1977), p. 7.
21
-------
TABLE 2-8. ENERGY REQUIREMENTS FOR ON-FARM PUMPING OF IRRIGATION
WATER IN UNITED STATES BY TYPE OF ENERGY, 1974
(Billion BTU)
Ground-
Surface
Percent of all
Energy
Water
Water
Total
energy types
Electricity
48,580
17,028
65,607
25.2
Diesel
22,342
2,471
24,813
9.5
Gasoline
7,379
1,451
8,830
3.4
Natural Gas
135,098
3,857
138,955
53.3
LPG
20,496
1,967
22,463
8.6
Total
233,895
26,773
260,668
100.0
Source:
Sloggett (1977), p. 8.
amounts of energy used, electricity and natural gas are by far the two most
important forms of energy used for irrigation.
Energy costs are an important component of costs of producing irrigated
grains. Corn production in the Southwest illustrates this. Most corn pro-
duced in that area is irrigated. In 1978 costs of fuel and lubrication were
20 percent of variable costs of production, and 12 percent of total nonland
costs (U.S. Senate). Among variable costs only fertilizer exceeded fuel and
lubrication, and that by a small margin. In Texas, which accounted for 80
percent of the corn produced in the Southwest in 1978, natural gas was the
energy source for 72 percent of all irrigated land (not just corn land).
Electricity provided 22 percent of the energy, with diesel fuel, gasoline,
and LPG making up the rest.10 Natural gas was relatively much more important
in irrigation in Texas than in the country as a whole (see Table 2-7).
Our discussion of trends in energy prices indicated a consensus that
real prices will increase by some 2 percent annually between 1975 and 2000,
with natural gas prices rising substantially more than this, electricity
prices somewhat less, and prices of other energy sources somewhere between
those of electricity and natural gas. In the studies referred to in our
discussion, the projections of increases in natural gas prices varied between
5.0 percent and 9.6 percent annually. Given the importance of natural gas
in on-farm requirements for pumping, these projections imply that the real
price of energy used for irrigation would rise more than 2 percent annually.
Table 2-9 gives estimates of the effects of higher energy prices and
declining water tables on costs of pumping one acre foot of water in the West.
"^Sloggett (1977). In the calculation of costs of corn production,
the Southwest consists of Texas and California.
22
-------
Thirty-three million acres of land in the West are irrigated by onfarm-pumped
water, and somewhat more than half of this is land in Texas, Oklahoma, Kansas,
and Nebraska irrigated with groundwater (Sloggett 1979). The feet of lift to
water that land are 100 in Nebraska, 180 in Kansas, and 200 in Texas and
Oklahoma. As noted above, aquifers in those states are being mined, so water
tables are declining over time.
Table 2-9 suggests that the combination of rising energy costs and de-
clining water tables would lead to sharply higher costs of pumping by 2000.
With no decline in the table pumping costs in 2000 would be double the 1980
level, and a 100-foot decline in the table, combined with rising energy
prices, would increase costs three to four times. These numbers probably
underestimate the increase in average pumping costs (given the assumed in-
crease in energy prices), because natural gas, the cheapest present source of
energy, is likely to decline as a percentage of total energy supply. A far-,
mer who found it necessary to shift from natural gas to electricity and whose
pumping depth increased from 100 to 200 feet would find his pumping costs
rising from $4.56 per acre-foot in 198.0 to $35.52 in 2000, an increase of
almost 700 percent (table 2-9).
Since energy costs are only a.part of total costs, the profitability
of pump irrigation is more sensitive to changes in crop prices than to
changes in energy costs. The USDA's NIRAP model projects a decline in real
prices of wheat between 1979 and 1990 (Table 1-3, previous section). Should
this occur an increase in pumping costs of the sort depicted in Table 2-9
would surely discourage, if not prohibit, the expansion of irrigated wheat
production based on groundwater. Indeed, the present 3.9 million acres of
western wheat irrigated by groundwater might well be forced out of pro-
duction.
The NIRAP model projects an increase of one-third in the real price of
corn from 1979 to 1990. If energy costs are 12 percent of total nonland
costs of irrigated corn production, as indicated above, then a one-third in-
crease in the price of corn would offset an almost fourfold increase in the
price of energy, assuming that all other costs remained the same. However,
if real energy costs double, as assumed here, other costs are not likely to
remain the same. We already have noted that rising real energy prices are
one of the principal reasons for projecting higher prices of nitrogen ferti-
lizer, also a major component in production costs of corn. In addition, as
noted above, the combination of falling water tables and a shift from lower-
cost natural gas to other higher-cost sources of energy for pumping could
increase average pumping costs sevenfold or even more.
We conclude that over the next several decades increasing competition
for water for nonagricultural uses and rising real energy prices will make
irrigation water in the West more expensive in real terms, without attempting
to project how much the increase may be.
23
-------
TABLE 2-9. ENERGY COSTS TO PUMP 1 ACRE-FOOT OF WATER FROM
DIFFERENT DEPTHS WITH ALTERNATIVE FUELS AND FUEL
PRICES* (1977 $)
Pump
Lift
(Ft. 0
Energy
1970
Costs Under Alternative
Fuel Prices
1980 2000
Natural gas
100
1.13
4.56
9.12
200
2.30
9.29
. 18.58
300
3.43
13.86
27.72
Electricity
100
7.52
8.88
17.76
200
15.03
17.76
35.52
300
22.55
26.64
53.28
LPG .
100
7.32
12.60
25.20
200
14.65
25.20
50.40
300
21.'98
37.80
75.59
Diesel
100
5.24
14.96
29.92
200
10.59
30.00
60.00
300
15.74
44.96
89.92
Source: Table 10 from Frederick.
*The technical assumptions such as the amount of fuel and the pressure
required to lift an acre-foot of water are based on information in Sloggett
(1979). Other assumptions include a 60 percent pumping efficiency and fuel
costs in 1977 constant dollars as follows: natural gas ($/mcf) .39 in 1970,
1.58 in 1980, and 3.15 in 2000; electricity ($kWh), 0.033 in 1980, 0.039 in
1980, and 0.078 in 2000; LPG ($/gal.) .25 in 1970, .43 in 1980, and .86 in
2000; diesel ($/gal.) .28 in 1970, .80 in 1980, and 1.60 in 2000. The 1970
prices for LPG and diesel are national averages obtained from U.S. Department
of Agriculture, June 1977. The 1970 price for electricity is from USDA,
October 1977. There was a wide range in the natural gas prices paid by High
Plains irrigators in 1970. The table assumes a price of 25c/mcf in 1970
dollars. The 1980 prices reflect average prices paid by farmers in Nebraska,
Kansas, Oklahoma, and Texas in January 1980. The prices were obtained
through phone conversations with several officials in those states. Based on
a conclusion of a recent major energy study (Landsberg et al., p. 71), it is
assumed that real fuel prices will double by the year 2000.
24
-------
Eastern Irrigation—
Table 2-5 indicates that in 1977 one-sixth of the nation's irrigated
land was in the thirty-one eastern states, 75 percent of that in the South-
east (Alabama, Florida, Georgia and South Carolina) and the Mississippi
Delta (Arkansas, Louisiana and Mississippi). The figures in Table 2-5 under-
state the amount of irrigated land in the Delta because most of the 776,000
acres of irrigated land in Missouri (a Cornbelt state) are in the southeast
corner of the state, a region counted as part of the Delta for soil classi-
fication purposes, but not for statistical purposes.
The percentage increase in eastern irrigation exceeded that in the West
from 1967 to 1977, with most of the increase in the Southeast, Delta (count-
ing Missouri as part.of the Delta) and Lake States (Michigan, Minnesota and
Wisconsin). This is evident in Table 2-10.
TABLE 2-10. IRRIGATED LAND IN THE EASTERN UNITED STATES,
1967 AND 1977 (1000 acres)
1967
1977
Northeast
440
371
Appalachia
—
414
Cornbelt
206
1,112
Missouri
118
776
Other Cornbelt
88
336
Lake States
129
966
Southeast
1,694
3,424
Delta
3,520
4,014
Total
5,989
10,301
Sources: Hanson and Pagano, September 1980. Basic data taken from
the USDA' Conservation Needs Inventory of 1967 and Na-
tional Resources Inventory of 1977
The expansion of eastern irrigation outside the Delta has been primarily
on droughty soils in the Lake States and Southeast. Because of low moisture
retention characteristics of these soils, plants come under moisture stress
very quickly, even though in an average year rainfall is ample to support
plant growth (Hanson and Pagano). The presence of these soils also accounts
for most of the expansion of irrigation in Indiana, the Cornbelt state,
This section is based on Hanson and Pagano (September 1980) on irri-
gation in the East outside the Mississippi Delta and on Shulstad et al.
(July 1980) on irrigation in the Delta.
25
-------
except for Missouri, with the greatest amount of irrigated land. Where these
soils prevail, supplemental irrigation is economical because it avoids yield
losses when the rains fail momentarily.
Hanson and Pagano conclude that most of the potential for future expan-
sion in eastern irrigation, apart from the Delta, is in the Southeast, par-
ticularly in Georgia and Florida. Aside from the droughty soils in those
states, their growing season is long enough to permit double-cropping, and
irrigation provides a timely water supply. Conservation tillage, of increas-
ing importance in the Southeast, also encourages double-cropping by saving
time between the harvest of the first crop and planting of the second.
Most of the irrigation in the Southeast is by onfarm pumping from ground-
water or small streams or ponds on the farmer's property. Both ground and
surface water supplies are ample relative to present and future competing
demands for them. Hence, farmers in the Southeast are not likely to feel the
competing pressures for water that their counterparts in the West must con-
tend with. While in Georgia the depth to groundwater is some 250 feet
(Sloggett, 1979), the aquifer is artesian, with enough pressure to make
pumping requirements relatively modest. Consequently, rising real costs of
energy will not inhibit the spread of irrigation by groundwater in the South-
east as they promise to do in the West.
Despite the prospect for increased irrigation in the Southeast, the im-
pact on the growth of production and yields of the crops of interest to this
study is likely to be small. Hanson and Pagano cite a 1977 federal govern-
ment inter-agency Task Force report on irrigation which estimates the maximum
irrigable area in the East, including the Delta, at 26 million acres. The
study by Shulstad et^ al. estimates that there are 17 million potentially
irrigable acres in the Delta. The implication is that there are some 9 mil-
lion potentially irrigable acres in the East, apart from the Delta. Table 2-10
indicates that in 1977 there already were some 6 million acres of eastern
irrigation outside the Delta. Even if all of the remaining irrigable but not
yet irrigated land were in the Southeast, and it is not, the potential for
expansion would appear to be only on the order of 3 million acres. Table 2-5
shows that in 1977 there were 31.6 million acres of cropland in the Southeast.
Hence an increase of 3 million irrigated acres would be only about 10 percent
of the cropland base of the region, not enough to have a major impact on the
growth of regional production.
Hanson and Pagano, drawing on the 1974 Census of Agriculture, show that
in that year only 5.5 percent of the irrigated land in the Southeast was in
crops of interest to this study. One-third of the land was in orchards,
25 percent was in pasture, 12 percent was in vegetables and the rest in
various minor crops. While corn and soybeans are expected to be much more
important in additional irrigation than at present, they are not likely to
account for all of the prospective growth.
In contrast, the study by Shulstad et al.'suggests that expansion of
irrigation in the Delta could have a significant impact on production and
yields in that region, particularly of soybeans. Shulstad et al. consider
numerous alternative combinations of crop prices, input costs, yields and
26
-------
crop rotations, and most indicate that it would be economical to irrigate all
of the some 17 million acres in the Delta judged by the authors to have the
physical characteristics consistent with irrigation. In the alternatives con-
sidered the amount of irrigated soybean land varies from 8 to 12 million
acres.
Table 2-5 indicates that there were 4.1 million acres of irrigated land
in the three Delta states in 1977. With most of Missouri's 776,000 acres
included in this, as they are in the study by Shulstad et al., presently irri-
gated land in the Delta is almost 5 million acres. If the estimate by Shul-
stad et al. of potential is correct, we can expect a substantial expansion of
irrigated land in the Delta, much of it in soybeans. The implications of
this for yield projections in the Delta are considered in Section 4.
CONCLUSIONS ON PRICES OF INPUTS
We conclude that real prices of energy will rise at an average annual
rate of about 2 percent from 1980 to the end of the century and beyond, with
natural gas prices increasing substantially more than this. This will put
upward pressure on prices of nitrogen fertilizer and on the costs of pumping
water for irrigation. Higher prices of natural gas combined with rising
costs of construction in the fertilizer industry likely will cause real prices
of fertilizer to rise by 1985, even though demand does not press hard on
capacity. We assume that real nitrogen prices will continue to rise to the
end of the century and beyond. Declining water tables in the High Plains
likely will reinforce the effect of higher energy prices on costs of pumping
in that region, and throughout the West growing nonagricultural demands for
water likely will increase its cost to agriculture.
The amount of irrigated land in the East, apart from the Mississippi
Delta, is expected to increase, particularly in the Southeast, but the impact
on trends in yields arid production of crops of interest to this study will
probably be small. The expansion of irrigated soybean production in the
Delta, however, could be significant.
In short, for most farmers real prices of energy, fertilizer and irriga-
tion water,.the key ingredients of land-saving technologies, are expected to
rise for the next several decades, in contrast to their behavior in the two
decades prior to the early 1970s. Other things the same, this would make
land-using technologies more attractive to farmers (except perhaps for those
in the Southeast and the Delta) than they were in those earlier decades.
However, farmers' choices will be influenced by the prospective productivity
of the alternative technologies and by the costs of increasing the supply of
cropland. Productivity is examined in the next section and the supply of
land in the one after that.
27
-------
REFERENCES
Castle, Emery and Irving Hoch. 1980. Explanation of Farm Real Estate Prices,
1920-1978. Resources for the Future (unpublished paper).
Duncan, Marvin. June 1977. "Farm Real Estate - Who Buys It and How."
Monthly Review, Federal Reserve Bank of Kansas City.
Dvoskin, D. and E. Heady. 1976. U.S. Agricultural Production Under Limited
Energy Supplies, High Energy Prices and Expanding Exports. Center for
Agricultural and Rural Development, Iowa State University. CARD Report
69. Ames.
Executive Office of the President. 1980. Economic Report of the President
1980. Washington, D.C.
Frederick, Kenneth D. Irrigation and the Adequacy of Agricultural Land.
Forthcoming in The Adequacy of Agricultural Land in the United States:
An Appraisal. Resources for the Future, Washington, D.C.
Hanson, James and James Pagano. September 1980. Growth and Prospects for
Irrigation in the Eastern United States. Resources for the Future.
Unpublished paper.
Harris, G. T. and E. A. Harre. 1979. World Fertilizer Situation and Outlook,
1978-1985. International Fertilizer Development Center and National
Fertilizer Development Center, Muscle Shoals, Alabama.
Landsberg, Hans et al. 1979. , Energy: The Next Twenty Years. A study
sponsored by the Ford Foundation and administered by Resources for the
Future. Ballinger, Cambridge, Mass. '
Schurr, S. H., J. Dannstadter, H. Perry, W. Ramsay, and M. Russell. 1979.
Energy in America's Future: The Choices Before Us. Published for
Resources for the Future by The Johns Hopkins University Press.
Baltimore.
Shulstad, R. N., R. D. May, B. E. Herrington, and J. M. Erstine. July 1980.
Expansion Potential for Irrigation Within the Mississippi Delta Region.
Department of Agricultural Economics and Rural Sociology, University of
Arkansas, Fayetteville.
Sloggett, Gordon. 1977. Energy in U.S. Agriculture: Irrigation Pumping.
USDA, Agricultural Economic Report 376, Washington, D.C.
28
-------
1978. Energy and U.S. Agriculture; Irrigation Pumping
1974-77. USDA, Agricultural Economics Report 436. Washington, D.C.
USDA. Various years. Agricultural Statistics. U.S. Government Printing
Office, Washington, D.C.
1976. Energy and U.S. Agriculture: 1974 Data Base. Washing-
ton, D.C.
. June 1977. Agricultural Prices. Washington, D.C.
. October 1977. Agricultural Prices. Washington, D.C.
. December 1978a. Agricultural Outlook AO-39. Washington, D.C.
. December 1978b. 1979 Fertilizer Situation. FS-9. Washing-
ton, D.C.
. February 1980. Basic Statistics: 1977 National Resources
Inventory, Revised. Soil Conservation Service, Washington, D.C.
. May 1980. Agricultural Outlook AO-54. Washington, D.C.
. May 30, 1980. Agricultural Prices. Prl (5-80). Washington,
D.C.
. June 1980. Agricultural Prices. Pr 1-3 (80). Washington,
D.C.
. February 1981. Economic Indicators of the Farm Sector:
Production and Efficiency Statistics, 1979. Economics and Statistics
Service, Stat. Bull. 657, Washington, D.C.
U.S. Senate. March 1978. Costs of Producing Selected Crops in the United
States—1976, 1977 and Projections for 1978. Prepared by the Economics,
Statistics and Cooperatives Service of the USDA for the Senate Committee
on Agriculture, Nutrition and Forestry. Government Printing Office,
Washington, D.C.
29
-------
SECTION 3
TRENDS IN PRODUCTIVITY
AND CROP YIELDS IN THE UNITED STATES
INTRODUCTION
We would like to be able to measure trends in the productivity of land-
using and land-saving technologies as a basis for projecting these trends
into the future. These projections, in combination with the projections of
input prices discussed in the previous chapter, would provide a basis for
judging how farmers in the future might choose between the two types of
technology.
We concluded in the last chapter that real prices of energy, fertili-
zer, and (for most farmers) irrigation water likely will rise over the next
several decades, in contrast to their movement in the two decades prior to
the early 1970s. This would make land-using technologies look more attrac-
tive in the future than in that earlier period, other things the same. How-
ever, if the productivity of land-saving technologies were to rise more than
in the earlier period, this would tend to offset the effect of higher input
prices on the relative attractiveness of land-using and land-saving tech-
nologies. Should the productivity of land-saving technologies rise less
than in the earlier period this, of course, would reinforce the effect of
input prices in making land-using technologies look more attractive.
We posit a situation in which the "typical" farmer, faced with rising
demand for crops, knowing present prices of inputs, and having expectations
about their future prices, asks himself the question, what mix of land and
other inputs should I choose in response to higher demand? We assume that
the desire to maximize profits will be the dominant criterion of choice.
Consequently, the farmer will tend to choose that combination of inputs
which, given input prices, yields the highest output per unit of total input,
i.e., that combination with the highest total productivity.
This suggests that examination of trends in total productivity would
provide insights about farmers' future choices among technologies. We have
not undertaken such an examination, however, because we have serious reser-
vations about the accuracy of the USDA's index of total productivity, espe-
cially in the period since 1972. The argument underlying these reservations
is given in appendix B. Suffice it to say here that the sharp decline in
the last several decades in the quantity and increase in price of farm labor
relative to the quantities and prices of other inputs, particularly ferti-
lizer and farm machinery, creates a fundamental problem in constructing an
30
-------
unambiguous index of total input, hence of movements in total productivity.
In addition, it appears to us that the way land is handled in the USDA's
index of total productivity significantly underestimates the role of land in
expanding production after 1972, thus overestimating the increase in total
productivity.^"
For these reasons we have focused our attention on crop yields instead
of on total productivity. This has several advantages. One is that our
interest in this study is in the resource and environmental impacts of crop
production, not total production. Another is that examination of trends in
yields provides a basis for projecting them, and therefore of converting our
projections of production into projections of land used for crops. A dis-
advantage of focusing on yields is that they provide only a partial measure
of productivity, but farmers' choices are conditioned by total productivity.
However, in attempting to explain the behavior of yields we necessarily deal
with trends in the productivity of fertilizers and other components of land-
saving technologies. This provides insights to movements in total produc-
tivity even though we do not measure it directly.
TRENDS IN YIELDS: ALL CROPS
Table 3-1 gives information about actual and trend yields of all crops
between 1960 and 1979. The trend values are based on actual yields in 1950-
1972. Column (3) of the table indicates that in each year after 1972 actual
yields were less than trend yields established in the period 1950-1972. Not
only did actual yields fall short of trend yields in every year 1973-1980
but the average annual deviation from trend in that period was 4.3 times as
large as in 1950-1972. In no period from 1950-1972 did actual yields fall
short of trend yields for as many as eight years in a row. (The maximum
was four years, 1954-1959.) The yield experience in 1973-1980, therefore,
suggests that the trend of yields in those years was lower than in 1950-1972.
Table 2-1 in the previous section indicates that the ratio of all pur-
chased inputs to cropland used for crops increased 60 percent from 1951 to
1972, then rose at a much slower rate from 1972 to 1979. The quantity of
fertilizer per acre of harvested cropland increased 7 percent annually from
1951-1955 to 1972, but by only 1.3 percent per year from 1972 to 1979. The
amount of harvested cropland declined from an average of 338 million acres
in 1951-1955 to 289 million acres in 1972, then rose to 331 million acres in
1978 and to 343 million acres in 1979 (USDA Feb. 1981).
Given these changes after 1972 in behavior of ratios of nonland inputs
to land and in the amount of harvested cropland, the slower growth of crop
yields after 1972 is not surprising. Indeed it is the expected result of
the reduced shift to land-saving technologies. With respect to post-1972
^"For a discussion of the index number problems, and others affecting
the USDA's measure of total productivity change, see USDA, Feb. 1980.
-------
TABLE 3-1. YIELDS FOR ALL CROPS (1967=100)
(1) (2) (3) (4) (5)
.Actual "Required"
Trend of Yields Yields
Actual
Actual
Trend
"Required"
Trend
Yields
Yields
Yields
Yields
Yields
1960
88
88
1.00
61
92
90
1.02
62
97
92
1.05
63
99
95
1.04
64
96
97
.99
1965
102
99
1.03
66
99
101
.98
67
100
103
.97
68
105
106
.99
69
110
108
1.02
1970
104
110
.95
71
112
112
1.00
72
118
114
1.04
73
113
117
.97
124
1.06
74
103
119
.87
115
.97
1975
110
121
.91
126
1.04
76
110
123
.89
126
1.02
77
115
126
.92
136
1.08
78
119
128
.93
137
1.07
79
126
130
.98
150
1.15
1980
114
132
.86
137
1.04
Average deviation from trend
1950-1972: 2.9
1973-1980: 12.5
Source: USDA, Feb. 1981 for 1950-1979. 1980: Harvested cropland from
Thomas Frey, USDA;
index
of crop production from USDA, Jan,-Feb., 1981.
Note: Column
(1):
Index of crop production divided by acres of crop'
land harvested, 1967=100.
Column
(2):
Arithmetic trend of actual yields 1950-1972, pro-
jected 1973-1980.
Column
(3):
Yields that would have been "required" to produce
each year's output 1973-1980 if the amount of crop-
land harvested had remained the same as in 1972.
32
-------
experience the fact that some of the land brought under crops after 1972 was
of lower natural fertility than land already in crops also contributed to
slower growth of yields. (More on this in Appendix C.)
For these reasons, the slower growth of crop yields after 1972 does not
necessarily indicate that the slower shift to land-saving technologies oc-
curred because the productivity of these technologies, as measured by yields,
had declined, or was rising less rapidly. At least some, and perhaps all of
the yield behavior was a result of the slower shift, not a cause of it.
To provide some insights into this cause-and-effeet relation between
yield behavior and the behavior of technology after 1972, we constructed an
index of "required" yields for 1973-1979. We asked the question, suppose
farmers had produced the actual amounts of crops in those years with the
same amount of cropland harvested as in 1972: what levels of yield would
have been required to do that? Those yields are in column (4) of table 3-1.
We then asked the question, given the yield experience of 1950-1972, could
farmers reasonably have expected to achieve the yields shown in column (4)?
We believe column (5) indicates that the farmers' response would have been
negative; that, with the exception of 1974 (when production was depressed by
exceptionally bad weather), the "required" yields would have appeared to be
out of reach, even if farmers had expected the pre-1972 trend in yields to
continue. To achieve the higher levels of production after 1972, therefore,
farmers would have concluded that more land would have to be brought into
production.
This argument indicates that the productivity of land-saving technolo-
gies, as measured by crop yields, was not rising fast enough before 1972 to
meet the demand for crop production in 1973-1979 without an increase in the
amount of land in production. The argument does not indicate, however, that
the rate of increase of productivity of these technologies was declining.
These conclusions are useful for our purposes, but they do not lead as far
as we would like because they deal with crop production as a whole. Our
interest is in specific crops, or groups of crops: wheat, feedgrains, soy-
beans, and cotton. Examination of the yield experience of these crops is
the next step.
2
YIELDS OF GRAIN AND SOYBEANS
The analysis is focused on the behavior of yields of principal grains
and soybeans in the United States from the late 1940s to the late 1970s,
seeking to measure the contribution of various factors to changes in yields
over this period. We identify two broad categories of factors. One we call
^«fe also are interested in production and yields of cotton. -However,
there has been no trend in cotton yields since the mid-1950s and we antici-
pate no dramatic change in this. Moveover, even if there were such a change
the amount of land in cotton would remain small relative to land in grains
and soybeans. For these reasons we do not consider trends in yields of cot-
ton.
33
-------
pure shift factors and the other pure yield factors. Yields for a given
crop differ among states. Should the relative distribution among states of
land in that crop change over time, national average yields will change even
if yields in each state remain the same. We say this change in national
average yields is attributable to pure shift factors. Changes over time
in yields within each state we say reflect pure yield factors.
We distinguish three pure yield factors: (1) weather, (2) land quality
and (3) technology, by which we mean everything affecting yields within
states except weather and land quality. We consider two components of tech-
nology: (a) fertilizers; and (b) irrigation.
We have made quantitative estimates of the contributions of pure shift
and pure yield factors to the behavior of corn, wheat, sorghum, and soybean
yields for selected periods from 1946-1950 to 1975-1976. Within the set of
pure yield factors we have made estimates of the effect of weather and of
land quality on yields of corn, wheat, and soybeans over these same periods.
We have not estimated the yield effects of fertilizer on these crops, and
our estimates of irrigation effects are incomplete. There is information
available, however, about both fertilizer and irrigation which permits some
judgments to be drawn about the effects of these inputs on trends in yields.
The analysis, which is quite detailed and technical, is in Appendix C.
Here we present only a summary of the principal conclusions.
Our analysis shows that growth of yields within states was far more
important than shifts among states in explaining the behavior of yields of
corn, sorghum, wheat and soybeans from 1946-1950 to 1971-1973. We have
examined corn and soybean yields in the Cornbelt and wheat yields in the
Plains states to isolate the effects on yields of (1) weather, (2) land
quality, and (3) technology. We particularly consider yield behavior
before and after 1972 because of the slower growth of yields after that
date. For soybeans we conclude that weather alone could account for post-
1972 yield behavior. Wheat yields, however, clearly cannot be attributed
to weather alone, and it is doubtful that corn yields can be.
We therefore examine the effect on yields of the expansion of land in
wheat and corn after 1972, on the assumption that the additional land is
inferior to land already in production. We conclude that the increment of
wheat land in the five Plains states probably would not explain much more
than 10 percent of the difference between actual wheat yields and the trend
value of weather-adjusted yields in 1977-1980, and that the weather could
not likely explain the remaining 80+ percent. The increment of corn land in
the Cornbelt explains roughly two-thirds of the yield shortfall, and the
remaining one-third may in fact be explained by weather.
For wheat at least it appears that technology contributed less to the
growth of yields after 1972 than before. This conclusion is supported by
the fact that per-acre applications of fertilizer to wheat land rose more
slowly after 1972 than before. The slower growth in fertilizer applications
per acre was even more marked for corn than for wheat. Although it appears
34
-------
4
that the weather and additional land may be sufficient to explain the short-
fall in corn yields in 1977-1980, the slower growth in per-acre fertilizer
use likely contributed something also.
The expansion of irrigated production of wheat and corn from 1950 to
1977 was substantial in percentage terms, and yields of irrigated land in
these crops rose also. However, irrigated production of these crops was
small in relation to dryland production, and dryland yields also rose sharp-
ly. Consequently, the contribution of irrigation to the growth of national
yields of corn and wheat was small. We cannot determine whether this con-
tribution declined after 1972 for lack of data, but qualitative information
from the Plains region suggests that it did not.
The analysis of yields of wheat, corn and soybeans leads to no definite
conclusion about the role of productivity in inducing the slower shift to
land-saving technologies after 1972. The slower growth of per-acre ferti-
lizer use on wheat and corn after that date does not necessarily reflect
diminishing marginal productivity of fertilizers. It could as well be ex-
plained by the increase in the real price of fertilizer. There is nothing
in our analysis, however, to suggest an upward shift in the trend of pro-
ductivity of land-saving technologies. The main question is whether the
pre-1972 trend has continued or whether it has shifted downward. . We con-
clude that continuation of the present trend of productivity of land-saving
technologies will not make these technologies look more attractive in the
future than in the past. Given our projected rise in real prices of energy,
fertilizer, and irrigation water, we conclude, therefore, that the slower
shift toward land-saving technologies that set in after 1972 will continue
over the indefinite future. In this event, our projections of increased pro-
duction likely mean that the demand for cropland will continue to rise, sug-
gesting two questions: how much might the demand for land rise, and what
will be the supply response? We examine these questions in the next section.
35
-------
REFERENCES
Feb. 1980. Measurement of U.S. Agricultural Productivity: A Review
of Current Statistics and Proposals for Change. Technical Bulletin No.
1614. Washington, D.C.
Jan.-Feb., 1981. Agricultural Outlook. Economics and Statistics
Service, Washington, D.C.
Feb. 1981. Economic Indicators of the Farm Sector: Production and
Efficiency Statistics, 1979. Economics and Statistics Service, Stat.
Bull. No. 657, Washington, D.C.
36
-------
SECTION 4
THE DEMAND FOR AND SUPPLY OF CROPLAND
INTRODUCTION
We have argued that prospects for the prices and productivity of land-
saving inputs suggest that farmers will continue to select technologies
from the land-using end of the spectrum. If this happens, how much land
would American farmers need to produce the levels of crop output projected
in Section 1? This is the first of two principal questions addressed in
this section. The second question is in two parts: (a) Would the present
supply of cropland be adequate to meet the projected demand? (b) If present
supply is not adequate, what would be the effect on the economic cost of
agricultural land of converting non-cropland to crops?
We proceed in three steps: (1) projection of regional production of
each crop; (2) projection of regional yields, and (3) analysis of the present
and potential supply of land, by region, in relation to projected demand. A
detailed account of the first two steps is in Appendix D. Here we give a
summary of the underlying arguments. The complete argument underlying the
third step is presented in this section, however.
REGIONAL PROJECTIONS OF PRODUCTION
These projections are derived from projections of regional shares of
national production. Table 4-1 shows actual shares for various years and
projected shares for 1985 and 2010. The table indicates that for the past
years shown the regional shares of wheat, feed grains, and soybeans were
quite stable in the sense that the absolute changes in percentages1 generally
were small. The maximum change was 7.5 percentage points in the share of the
Northern Plains in wheat production between 1970-72 and 1976-78. This de-
cline was almost exactly offset by an increase in the shares of the Southern
Plains and Lake States in wheat production.
A check of data for the 1950s, not shown here, indicates that regional
shares in those years were comparable to those shown in Table 4-1, with a
couple of exceptions. In the 1950s the Cornbelt had 14-15 percent of wheat
production, compared to the 10-12 percent shown in Table 4-1. Most of the
decline in the Cornbelt's share of wheat production since the 1950s was off-
set by increases in the shares of the Southern Plains and Pacific states.
37
-------
The other significant change since the 1950s was a decline in the Corn-
belt's share of soybean production from about 68 percent to the 56-60 percent
shown in Table 4-1. Most of this was offset by an increase in the Delta's
share of soybean production from about 9 percent in the 1950s to the 14-16
percent shown in the table and by an increase in Appalachia's share from
about 5 percent to 7-8 percent.
The distinctive feature about the regional distribution of cotton pro-
duction is its shift from the Southeast and Delta to the Southern Plains
(mostly Texas); Mountain states (mostly Arizona); and Pacific (California).
According to Collins, Evans, and Berry (p. 21) this shift reflects better
conditions for growing cotton in the West than in the Southeast and Delta.
In particular fewer pest problems and better water control through irrigation
have led to lower production costs in Texas, Arizona and California than in
the two more eastern regions.
The stability of regional shares in production of grains and soybeans
suggests that radical shifts in the regional location of production of these
crops are unlikely over the next several decades.-'- We do expect some changes,
however. These are: (1) an increase in the share of the Southeast in corn
and .soybean production and of the Mississippi Delta in soybean and wheat pro-
duction. In both cases the reasons for expecting increased shares are the
potential for supplemental irrigation, discussed in Section 2, and the avail-
ability of land now in pasture and forest with potential for conversion to
cropland, (2) An increase in the share of the Southern Plains in wheat pro-
duction is expected, continuing a trend evident for some years and consistent
with that region's relatively abundant supply of land with potential for
conversion to crops. (3) The decline in the share of the Mississippi Delta
in cotton production is expected to continue, with the Southern Plains
(mostly Texas) and the Pacific (California) gaining. As noted above, this
shift has been underway for some time and reflects lower costs of cotton
production in Texas and California. Texas is expected to gain also at the
expense of New Mexico and Arizona (the two Mountain states that grow cotton),
reflecting the increasing cost and scarcity of water for irrigation. Most
of the land in cotton in those states is irrigated compared with 40-45 per-
cent in Texas. Most cotton acreage in California (the Pacific state that
grows cotton) is irrigated, but California has other cost advantages suffi-
cient to maintain its share of production at a relatively high level.
NATIONAL AND REGIONAL PROJECTIONS OF YIELDS
If we are right in thinking that in the future farmers are likely to
continue to favor relatively land-using technologies, as they have since
1972, then the behavior of yields since 1972 should provide a guide for
projections of future yields. The trend of yields of the various crops
The effect of higher energy prices on transportation costs may cause
some shifts in the regional distribution by production not captured in our
projections. We were unable to explore this possibility in this study.
38
-------
Table 4-1. REGIONAL SHARKS OF CROP PRODUCTION (percent)*
North-
Lake
Corn-
Northern
Appa-
South-
Southern
Moun-
east
States
belt
Plains
lachia
east
Delta
Plains
tain
Pacific
Wheat
1967-69
1.9
4.3
12.1
40.5
1.9
.7
1.7
11.7
15.3
9.9
1970-72
1.3
4.3
10.2
45.2
2.1
.8
.9
8.7
15.3
11.2
1973-75
1.3
6.2
10.2
39.8
1.9
.6
.8
13.1
14.9
11.3
1976-78
1.1
7.5
11.1
37.7
1.6
.4
1.2
12.5
15.4
11.5
1985
1
7
10
36
1
b
1
15
16
13
2010
1
6
5
34
1
1
7
16
16
13
Feedgralns
1960-64
3
13
46
17
5
3
1
6
3
3
1965-69
3
13
48
17
4
3
1
6
3
2
1970-74
3
13
46
18
4
3
+
7
3
2
1975-78
3
13
47
18
4
3
t
6
3
2
1985
3
13
46
16
5
4
0
6
5
2
2010
3
15
46
15
5
4
0
6
4
2
Soybeans
1967-69
1.1
8.2
57.6
4.8
6.6
4.1
16.4
1.1
—
—
1970-72
1.0
7.9
59.7
3.9
6.8
4.5
15.5
.6
—
—
1973-75
1.2
8.5
56.8
4.7
7.8
5.9
14.2
.9
—
—
1976-78
1.3
8.4
56.2
4.2
7.8
5.7
14.9
1.4
—
—
1985
1
8
55
4
8
6
16
2
—
—
2010
1
8
48
5
8
11
16
4
—
—
Cotton
1967-69
2.0
—
4.6
9.4
28.1
34.7
8.1
13.1
1970-72
—¦
—
3.1
—
5.7
10.7
3i.9
30.9
6.2
11.6
1973-75
—
—
1.9
—
4
8.9
25.8
32
8.2
19.2
1976-78
—-
—
1.6
2.5
4.7
23.3
38
7.8
20.1
1985
—
—
1.6
—
1
3
19
40
io
25
2010
—
—
1
—
2
3
14
47
8
5
-------
Notes to Table 4—1.
Source: USDA, Agricultural Statistics, various years. Note that the
years shown for feedgrains are different from those for wheat, soybeans and
cotton. This reflects the way the data were compiled, and has no signifi-
cance for the projections.
May not add to 100 because of rounding.
"'"Less than .5 percent.
The states in each region are as follows: Northeast—all of New England
plus New York, New Jersey, Pennsylvania, Delaware, and Maryland; Lake states—
Michigan, Minnesota, and Wisconsin; Combelt—Ohio, Indiana, Illinois, Iowa,
Missouri; Northern Plains—North Dakota, South Dakota, Nebraska, Kansas;
Appalachia—Virginia, West Virginia, Kentucky, Tennessee, North Carolina;
Southeast—South Carolina, Georgia, Florida, Alabama; Delta—Arkansas,
Mississippi, Louisiana; Southern Plains—Oklahoma, Texas; Mountain—New
Mexico, Arizona, Nevada, Colorado, Wyoming, Montana, Idaho, Utah; Pacific—
California, Oregon, Washington.
since 1972 is obscured, however, by extreme variations in weather. The
weather effects are particularly troublesome because they severely depressed
yields early in the period (see the indexes of weather effects for' corn and
soybeans in Tables C-2 and C-3) and elevated them in 1973 and 1979. (The
effect of weather was strongly adverse in 1930, however.) This pattern of
weather effects gives a sharp upward tilt to the trend of yields of corn
and soybeans, and to a lesser extent of wheat, between 1973 and 1980.
Table 4-2 shows national yields of these crops, and of cotton, from 1971
to 1980. Table 4-3 shows average yields for 1975/78 and projections to 1985
and 2010 for the nation and for the ten USDA producing regions. The eight
years 1973 to 1980 obviously provide an uncertain base from which to project
national and regional yields. Consequently we saw little justification for
projecting trends derived from mathematical fits to the 1973-1980 data. In-
stead, for each crop we examined the behavior of national yields and yields
for the two or three major producing regions, and made judgmental projections
on that basis, assuming that the trend of national yields would be dominated
by trends in the major regions. Yields in other regions were then projected
proportionately to one another and so as to be consistent with the projec-
tions for the nation and the major regions.
These obviously are rough-and-ready procedures, and the resulting pro-
jections of yields are subject to an uncomfortably large margin of error.
The main issue with respect to yields, however, is whether their post-1972
trend behavior is more likely to presage the future than their pre-1972
trends. Believing as we do that the prices and productivities of yield-
increasing inputs will behave more as they did after 1972 than before that
40
-------
date, we have to conclude that the post-1972 trends in yields are more indi-
cative of the future than pre-1972 trends. If we are correct in this, our
projected yields, despite the large error to which they are subject, will be
closer to actual yields in 1985 and 2010 than if we were to assume that the
pre-1972 trends will reassert themselves.
The yield projections have enormous implications for projections of
land use and resulting economic and environmental pressures on the nation's
land and water resources. The effect of alternative yield projections is
particularly evident by 2010. Given our projections of crop production,
it is not too much to say that yield behavior will determine whether or not
the nation's agriculture will encounter serious economic and environmental
problems over the next several decades. We take the yield projections in
Table 4-3 as a point of departure in our subsequent analysis. We emphasize
here, however, as we did in the conclusions, that the results of that anal-
ysis, particularly as they bear on the severity of future environmental
pressures, depend crucially on our projections of yields.
Table 4-2. YIELDS IN THE UNITED STATES OF WHEAT, FEEDGRAINS, SOYBEANS
AND COTTON
Wheat
Feedgrains Soybeans .
Cotton
(metric tons/acre)
(lbs./acre)
1971
.92
1.77
. .75
438
1972
.89
1.93
.76
507
1973
.86
1.81
.75
520
1974
.74
1.49
.63
442
1975
.84
1.75
.78
453
1976
.82
1.82
.71
465
1977
.83
1.88
.83
520
1978
.86
2.08
.81
421
1979
.93
2.31
.87
551
1980
.89
1.95
.73
421
Sources: USDA, 1980 for 1971-1979, and May, 1981 for 1980.
41
-------
Table 4-3. YIELDS IN T11E UNITED STATES OF WHEAT, FEEDGRA1NS AND SOYBEANS (metric tons/acre)
1975/1978
1985
2010
Wheat Feedgrains Soybeans Wheat Feedgralns Soybeans Wheat Feedgralns Soybeans
Nation
.84
1.88
78
.91
2.28
.87
1.08
2.9A
1.10
Northeast
.94
1.96
.74
1.07
2.17
.80
1.14
2. 79
.98
Lake States
.95
1.9-5
.86
1.03
2.36
.93
1.15
3.04
1.14
Combe It
1.07
2.37
.90
1.16
2.87
.99
1.33
3.70
1.30
Northern Plains
.76
1.62
.74
.83
1.96
.80
.92
2.53
.98
Appalachia
.90
1.87
.67
.97
2.10
.72
1.09
2.72
.89
Southeast
—
1.35
.59
—
1.52
.70
A
1.93
.94
Delta
-95
—
.63
1.03
—
.74
*
—
1.00
Southern Plains
.66
1.62
.67
.71
1.80
.72
.81
2.30
.89
Mountain
.82
1.35
—
.89
1.52
—
.99
1.94
PaciEic
1.23
1.47
—
1.33
1.68
—
1.49
2.29
—
Sources; 1975/1978 from USDA. Afirlcultural Statistics, various years. 1985 and 2000 projected as
described in the text.
Assumes that all wheat In the Southeast and Delta is double-cropped with soybeans.
-------
DEMAND FOR AND SUPPLY OF CROPLAND
Demand for Cropland
The demand curve for cropland shifts to the right in response to rising
demand for crops and to farmers' choices of the most economical mix of land
and other inputs with which to meet crop demand.^ We take yields to reflect
the outcome of these choices. Taken together, therefore, projections of crop
production and yields define projected demands for cropland. These are shown
in Table 4~4. The table shows harvested cropland in main crops (feedgrains,
soybeans, wheat and cotton) and in other crops. It also shows other uses
of cropland. Since the late 1960s the amount of land in "other crops" has
been quite stable at around 90 to 95 million acres. About two-thirds of this
land consistently is in hay. Land in hay reached a peak of 77.6 million
acres in 1944, then declined steadily to 60.9 million acres in 1968. It has
varied narrowly around 60 million acres ever since. We have assumed that
land in hay will be 60 million acres in 1985, but that it will decline to 50
million acres in 2010 as hay finds increasing difficulty competing with other
higher-valued uses of the land. We assume that non-hay "other crops" will
occupy the same amount of land in 1985 and 2010 as in 1977. To do better
than this would require more analysis of future demand for these crops, and
of their yields, than we are in a position to make.
In each year from 1958 to~l977, crops failed on from 5 million to 10
million acres, the average being 7.4 million acres. In projecting 10 million
failed acres, the figure for 1977, we may be somewhat on the high side, but
any reasonable alternative would make no significant difference in our pro-
jections of total demand for cropland.
In the same twenty-year period just mentioned, land in fallow varied
between 30 million acres and 41 million acres, the average being 33.9 million
acres. Virtually all of the land in fallow is in the main wheat growing
areas. Fallowing restores soil moisture after harvest, and is essential to
economical production of wheat in the semi-arid West. The record suggests
that this imperative sets a minimum of about 30 million acres for land in
fallow. We have used that figure in both 1985 and 2010.
Data for idle cropland are available only for agricultural census years.
In the most recent four such years, the amounts of idle; cropland were as follows:
1954 33 million acres
1964 52 million acres
1969 51 million acres
1974 20 million acres
Source: USDA Agricultural Statistics 1973 and 1977. 1978 agricultural
census data for idle cropland were not available at this writing.
^we here leave aside other reasons for demanding cropland, e.g., as an
inflation hedge or tax haven.
43
-------
Table 4-4. PROJECTIONS OF DEMAND FOR CROPLAND (millions of acres)
1985 2010
Harvested Harvested
Main
Crops
Other
Crops
Failed
Fallow
Idle
Total
Main
Crops
Other
Crops
Failed
Fallow
Idle
Total
Northeast
5.1
8.0
.2
. —
1.0
14.3
5.9
7.1
.2
—
1.0
14.2
Lake States
23.9
13.0
.8
—
2.8
40.5
30.3
12.7
.8
—
2.8
46.6
Cornbelt
81.1
10.1
.7
—
2.9
94.8
92.3
5.1
.7
*
—
2.9
101.0
Northern Plains
53.1
20.4
3.3
15.9
3.0
95.7
63.8
20.6
3.3
15.9
3.0
106.6
Appalachia
13.8
6.4
.3
—
1.9
22.4
18.6
5.6
.3
—
1.9
26.4
Southeast
12.5
4.6
2.0
—
1.4
20.5
22.1
4.4
2.0
—
1.4
29.9
Delta
16.6
3.5
.3
—
1.4
21.8
20.6
3.2
.3
—
1.4
25.5
Southern Plains
32.3
6.2
1.2
.8
3.3
43.8
42.8
5.5
1.2
.8
3.3
53.6
Mountain
20.9
10.6
1.1
9.5
1.5
43.6
24.7
9.3
1.1
9.5
1.5
46.1
Paci fic
11.5
9.4
.2
3.6
.8
25.5
13.8
8.7
.2
3.6
.8
27.1
Nation
270.8
92.2
10.1
29.8
20.0
422.9
334.9
82.2
10.1
29.8
20.0
477.0
Notes: Main crops are wheat, feedgrains, soybeans, and cotton, projected as described in the text
and in Appendix D. Land in "other crops" is mostly hay (65 percent for the nation in 1985 and 60 percent
in 2010). The projection for non-hay "other crops" is the amount of land in these crops in 1977. The
projections for failed and fallow land also are the same as 19 77. Idle land is the same as 1974.
-------
The high figures for 1964 and 1969 reflect the policy then in effect of
encouraging farmers to take land out of production; the low figure for 1974
reflects the abandonment of that policy and the influence of high-crop prices
in stimulating intensive use of the land. It is significant that despite
this stimulus there still were some 20 million acres of idle cropland in the
country in 1974. According to land use specialists at the USDA, some crop-
land inevitably is idle every year for reasons that vary widely from place
to place, and 20 million acres probably is close to the minimum amount of
such land. That is the basis for our projection of idle land.
Supply of Cropland
Table 4-5 shows two estimates of the amount of cropland in the United
States in 1977. We have no information that would account for the differ-
ences between the estimates, but take some comfort from the fact that the
difference between the totals is relatively small. We work with the USDA's
National Resources Inventory (Feb. 1980) because in addition to data on
the amount of present cropland it also includes estimates of the amount of
land potentially convertible to cropland, amounts of erosion from cropland,
and other useful information.
Table 4-6 shows data from the NRI for the present (1977) and potential
supply of cropland, and projections of the demand for cropland as percentages
of present and potential supply. The projections indicated that by 1985 the
nationwide demand for cropland would exceed the present supply by 2.5 per-
cent, or about 10 million acres. By region the situation would be quite
diverse, demand being weak relative to present supply in the Northeast and
Lake states and relatively strong in the Southeast. The greatest absolute
shortage of land would be about 5 million acres in the Cornbelt, indicating
that half of the projected nationwide deficit would be in that region.
By 2010 the nationwide demand for cropland would exceed the present
supply by 15.6 percent, or about 64 million acres. Only in the Northeast
would demand fall short of supply, and in the Southeast demand would exceed
present supply by over 70 percent. The largest absolute deficits would be
in the Northern Plains (12 million acres), the Southern Plains (11.4 million
acres) and the Cornbelt (11.1 million acres).
Given our projected levels of production and yields of main crops, there
would be two possible responses to the set of demand-supply relations indi-
cated in Table 4-6: (1) a reduction in the amount of land devoted to other
crops, or that is in fallow or that is idle; (2) conversion to crops of land
now in pasture, forest, or range.
These responses are not independent of one another, being linked by
the cost of each relative to the. other. We have more to say below about
the costs of land conversions, but we believe that conversions likely would
be much more important than shifts in present patterns of use of cropland.
It is not likely that idle land can be much reduced below the projected 20
million acres, for reasons given earlier. We indicated above that the main-
tenance of some 30 million acres in fallow seers an imperative of crop
45
-------
Table 4*5. ESTIMATES OF THE AMOUNT OF CROPLAND IN THE UNITED STATES IN 1977,
BY MAJOR USE*
(million acres)'
ESCSf
NRI*
Row crops and
close
grown crops
292.1§
308.1
All hay
60.7
63.4
Fallow
30.0
29.3
Tree and bush
crops
and vineyards
3.4
5.5
Idle
20.0^
—
Other
—
6.8
Total
406.2
413.2
'ft
Excludes cropland used only for pasture, which was 83.5 million acres
in 1974.
^The Economics, Statistics and Cooperatives Service of the USDA.
^The National Resources Inventory, conducted by the Soil Conservation
Service (USDA, Feb. 1980).
^Planted acres of all crops except hay, tree and bush crops and vine-
yards, from USDA, January 15, 1980. Includes land on which crops failed.
^From the 1974 Agricultural Census. The figure for 1977 is not known,
but for reasons given in the text it was unlikely to have been less than 20
million acres.
46
-------
Table in6. SUPPLY OF AND DEMAND FOR CROPLAND
Supply of Cropland
(million acres)
Present Plus
High and
High Medium
Present Potential Potential
Present
Demand for Cropland as Percent of Supply
1985
Present Plus
High and
High Medium
Potential Potential
Present
2010
Present Plus
High
Potential
lllgh and
Medium
Potential
Northeast
16.9
18.0
21.6
84.6
79.4
66.2
84.0
78.9
65.7
Lake States
44.1
4ft.4
52.3
91.8
87.3
77.4
105.7
100.4
89.1
Combe It
89.9
94.7
104.)
105.5
100.1
90.9
112.4
106.7
96.8
Northern Plains
94.6
99.7
112.5
101.2
96.0
85.1
112.7
106.9
94.8
Appalachla
20.8
25.5
35.1
107.7
87.8
63.8
126.9
103.5
75.2
Southeast
17.5
22.4
33.3
117.1
91.5
61.6
170.9
133.5
89.8
Delta
21.2
24.3
31.2
102.8
89.7
69.9
120.2
104.9
81.7
Southern Plains
42.2
47.4
62.2
103.8
92.4
70.4
127.0
113.1
86.2
Mountain
42.2
45.4
56.4
103.3
96.0
77.3
109.2
101.5
81.7
Pacific
23.2
24.8
28.4
109.9
102.8
89.8
116.8
109.3
95.4
Nation
412.6*
448.6
537.1
102.5
94. J
78.7
115.6
106.3
88.8
Sources: Present and potential cropland from USDA, Feb. 1980. Definitions of high and medium potential
land are given In the text. Demand for cropland from table 4.4.
Excludes 293,000 acre9 In Hawaii and 363,000 In the Caribbean.
-------
production, particularly wheat, in the semi-arid West. However, if pressure
on the land base in that region mounts, as Table 4-6 indicates, farmers might
find ways to reduce the amount of land in fallow without fatally compromising
the economics of wheat production. In this connection it is worth noting
that in a run of the Iowa State model in which the total amount of cropland
and its regional distribution corresponds closely to our projections, the
model puts 22.5 million acres iri fallow.
Conceivably, some of the expanded production of main crops could be on
land now in hay and other crops. In fact, we have projected a decline in
hayland from 60 million acres to 50 million acres in 2010. (The run of the
Iowa State model, mentioned above, put 55 million acres in hay in 2010.)
Apart from hay the most important land-using crops among "other crops" are
corn for silage (8.5 million acres in 1977-79), sunflower (2.2 to 5.4 mil-
lion acres), rice (2.2.to 3.0 million acres), peanuts (1.5 million acres) and
potatoes, sugar beets, flaxseed and edible beans (all 1 to 1.3 million acres).
No doubt, some of the land in these crops could be shifted to production of
main crops. How much flexibility there is in this connection would depend
upon cross-demand and -supply elasticities among main and minor crops,
matters which we cannot explore. Any significant shift of land out of minor
crops, however, surely would raise their prices, the sharper the rise the
more limited the shift.
The various substitution possibilities perhaps would be sufficient to
accommodate a good part of the 10 million acre deficit projected for 1985,
but they would not likely contribute much to cover the 64 million acre deficit
projected for 2010. Satisfying that excess demand surely would have to be
done primarily by converting to crops land now in pasture, forest, or range.
According to the NRI there were in 1977 36 million acres of land in
pasture, forest, and range with high potential for conversion to crops. An
additional 88.5 million acres of such land were judged to have medium poten-
tial for conversion. Table 4~6 indicates that the high potential land would
be sufficient to cover the excess demand for cropland in 1985, except perhaps
in the Pacific region, and that the combination of high and medium potential
land would be sufficient to cover the projected deficits in 2010 for every
region. Before concluding that the potential supply of cropland would, in
fact, be adequate, however, we must address two questions: (1) How much
would it cost to convert the potential land and would farmers be willing to
pay the cost? (2) How much of the present and potential supply of cropland
will likely be lost to non-crop uses over the next several decades?
Costs of Land Conversion—
In the NRI land with high potential for conversion to crops was defined
as that with "favorable physical characteristics" and of a sort which had
been converted in the vicinity in the three years preceding the survey (1974-
76). Commodity prices and development and production costs of 1976 were
assumed. Medium potential land also has favorable physical characteristics,
but conversion costs are estimated to be higher than for high potential land.
There is no indication in the NRI that the price-cost relationships of 19 76
48
-------
uniquely defined the economic conditions for conversion of land to crops,
i.e., the NRI does not state that had prices been slightly lower or costs
slightly higher conversions would not have occurred. The only statement is
that under price-cost conditions prevailing in 1976 land of the sort indi-
cated had been converted.
Table 4-7 shows prices received by farmers in 1976, 1979 and projections
to 1985 and 1990. The table suggests that crop prices in 1985 and 1990 would
favor conversion of land to corn and sorghum production but not to the other
crops. If prices of inputs and conversion costs rise faster than yields,
prospects for converting land to the other crops would appear to be even less
favorable, and the profitability of converting to corn and sorghum might also
become doubtful.
These statements must be taken with great caution, however. As noted
above, the NRI does not indicate that the 1976 cost-price relations uniquely
defined the terms of profitability of land conversion. Conceivably, about as
much land as was converted would have been even if the cost-price relations
had been less favorable than they were. In this connection it is noteworthy
that in 1976 yields of wheat, corn, sorghum and soybeans were all below our
projections for 1985.^ If yields in 1985 are as we have projected them,
this would offset some, if not all, of any increase in real prices of inputs
that might occur between 1976 and 1985.
Changes in the amount of land in the various crops and in crop prices
from 1976 and 1979 also suggest caution in interpreting the NRI's definition
of the economic conditions underlying the estimates of convertible land.
Table 4-7 indicates that changes in real crop prices from 1976 to 1979 would
have favored the expansion of land in wheat and cutbacks in land in all the
other crops. In fact, land planted to wheat in 1979 was 8.5 million acres
less than in 1976; land in corn in 1979 was about the same as in 1976; land
in sorghum, oats, and barley was down in aggregate about 7 million acres,
and soybean acreage was u£ 21.4 million acres, an increase of 42.6 percent
(USDA, 1979 and June 27, 1980). Total planted acreage in main crops rose
1.7 million acres, so the increase in land in soybeans could have been accom-
modated within the total amount of land in the main crops in 1976. Neverthe-
less, the sharp increase in the amount of land in soybeans despite the quite
3
Yields in 1976 and our projections for 1985 are as follows:
1976
1985
(metric tons/acre;
Wheat
Feedgrains
Soybeans
.82
.91
2.28
1.82
.71
.87
49
-------
Table 4-7. PRICES RECEIVED BY FARMERS FOR GRAINS AND SOYBEANS ($/bu.)
1976 in
NIRAP
(1979 $)
1976
1979 $
1979
1985
1990
Wheat
$ 2.73
$ 3.38
$ 3.82
$ 3.09
$ 3.33
Corn
2.15
2.66 .
2.41
3.01
3.22
Sorghum
2.03
2.51
2.33
2.70
2.89
Oats ,
1.56
1.93
1.36
1.38
1.47
Barley
2.25
2.78
2.31
2.32
2.47
Soybeans
6.81
8.42
6.19
7.36
7.85
Sources: 1976 and 1979 from USDA, 1979 and 1980, respectively. The
NIRAP projections are from Table 1-3 of this report. The GNP deflator was
used to convert 1976 prices to 1979 dollars.
unfavorable movement in real soybean prices indicates we should not read too
much into the price relations in Table 4-7 in judging whether farmers will
want to convert pasture, forest, and range land to crops on the scale indi-
cated in Table 4-6.
Studies done at Iowa State University and the University of Arkansas
provide insights to the supply of potential cropland in Iowa and the Missis-
sippi Delta (Shulstad, May, and Herrington for the Delta and Amos for Iowa).
Both studies are based on detailed analysis of land conversion costs, includ-
ing opportunity costs, and crop yields on various soil types in the Delta and
Iowa. Table 4-8 is based on these studies. The estimates of convertible
land in both regions assume that real prices of inputs will be one-third
higher in 1985 than in 1978 and that crop prices will be as shown. Yields
in Iowa are assumed to be 114 bushels per acre for corn and 40.4 bushels for
soybeans. Yields in the Delta are those achieved by the top 10 percent of
managers and are specified by soil type. Averages for each crop are not
given. The Delta study assumes a 10 percent discount rate. The two esti-
mates of convertible land in Iowa assume discount rates of 8 percent and 4
percent, respectively.
Table 4-8 suggests that NIRAP's projected prices for 1985 would not
favor conversion of land to corn in either the Delta or in Iowa or to wheat
in the Delta. The price comparisons tell only part of the story, however.
We have not projected corn yields in Iowa separately, but typically corn
yields in that state are 8-10 percent above feedgrain yields in the Cornbelt
as a whole. If we assume that in 1985 corn yields in Iowa will be 8 percent
above our projected yields for the Cornbelt, Iowa corn yields will be just
enough above those assumed in the Amos study to offset the difference between
the projected NIRAP price of corn and the price used by Amos.
50
-------
Our 1985 projections of soybean yields in the Cornbelt are virtually the
same as those used by Amos for Iowa. The NIRAP price for soybeans would,
therefore, imply expansion potential in Iowa greater than the Amos study
shows. Whether this would also be true in the Delta is not clear. The study
by Shulstad, May and Herrington indicated that in the scenario assuming out-
put prices one-third higher than in 1978 the amount of convertible land was
quite inelastic across the range of crop prices used. None of the prices for
soybeans was as high as the NIRAP projections, however.
We conclude that the NIRAP price projections for 1985 and our projec-
tions of corn and soybean yields in the Cornbelt and of soybean yields in
the Delta (we project no corn production in the Delta) are consistent with
the estimates of convertible land in the Iowa and Delta studies. The NRI
shows that there were 2 million acres of high and medium potential land in
Iowa in 19 77. The Amos study suggests that a quarter to half of this land
could be profitably converted to production of corn and soybeans. If these
projections hold for the Cornbelt as a whole, then 3.6 to 7.2 million acres
of the 14.4 million, acres of high-to-medium potential land in that region
could be profitably converted to crops under the cost-price conditions pro-
jected for 1985. Our projections of the demand for land (Table 4-6) indicate
that demand in the Cornbelt would exceed 1977 supply by 4.9 million acres in
1985. This would exceed the lower estimate of convertible land, but fall
Table 4-8. AMOUNTS OF LAND CONVERTIBLE TO CROPS IN IOWA AND THE MISSISSIPPI
DELTA
Iowa
Mississippi
Delta
Crop prices, 1979 $ ($/bu.)
Wheat
Corn
Soybeans
Convertible land (million acres)
NIRAP prices, 1985 in
1979 dollars ($/bu.)
Wheat
Corn
Soybeans
3.25
6.94
.5 to 1.0
3.09
3.01
7.36
3.79
3.25
6.94
2.6
Sources: Iowa from Amos; Delta from Shulstad, May, and Herrington,
NIRAP prices from Table 1-3. The estimates of convertible land assume real
input prices one-third above those of 1978. Other scenarios in the two
studies assumed no increase in real input prices.
*Wheat was not considered in Iowa.
51
-------
well within the upper bound. Table 4-6 indicates that by 2010 the demand
for cropland in the Cornbelt would exceed 1977 supply by 11 million acres,
well beyond the maximum amount indicated by the Amos study as convertible.
Our yield projections for feedgrains and soybeans in the Cornbelt show in-
creases of about 30 percent each from 1985 to 2010. This would tend to
increase the amount of convertible land beyond that estimated by Amos, but
the outcome would be affected also by the behavior of input and conversion
costs, and the effect of demand growth on crop prices. We expect prices of
inputs to rise more or less steadily from 1985 to 2010, but we have no basis
for projecting the amount of the increase. Nor are we able to estimate the
effects of crop demand growth on crop prices.
Table 4-6 indicates that the demand for cropland in the Delta would
exceed 1977 supply by .6 million acres in 1985, well under the 2.6 million
acres estimated as convertible by Shulstad et al. By 2010, however, demand
would exceed 1977 supply by 4.3 million acres.
In summary, it appears that if yields in the Cornbelt and Delta region
grow as we have projected and real prices behave as the NIRAP model indi-
cates, then our 1985 projections of the demand for cropland in those two
regions could be met. Whether this could be true in other regions is diffi-
cult to say, since we lack detailed studies of land conversion potential in
those regions.
Whether the demands for land projected for 2010 will be satisfied is
even less certain. However, if real prices of inputs rise faster than crop
yields, as we think likely, it is clear that real crop prices will have to
rise from 1985 to 2010 to make conversion of land on the required scale
profitable. Conceivably the increase in crop prices could erode the U.S.
position in world markets for grains, soybeans and cotton, thus causing the
demand for crop production to grow more slowly than we have projected. In
this case the demand for cropland also would grow more slowly, reducing
pressure to convert land now in pasture, forest, and range.
We assume that real crop prices will increase enough to justify con-
version of enough land to meet the demands for cropland projected for 2010,
but not so much as to reduce demand for crops significantly below the projec-
tions shown in Section 1.
Conversion to Non-agricultural Uses—
Table 4-6 indicates that by 2010 the demand for cropland would exceed
1977 supply by 64 million acres. The additional land would have to come
from land now in pasture, forest, and range. However, the amount of land
converted would exceed 64 million acres because some of the land now in
crops, pasture, forest, and range will be converted to urban and other uses
that will make it permanently unavailable as cropland. Between 1967 and
1975, about 2 million acres of land per year were so converted (Brewer and
Boxley). Less than 1 million of these acres, however, were cropland or
potential cropland.
52
-------
If this rate of conversion continues some 25-30 million acres of crop-
land and potential cropland would be lost to agriculture between 1977 and
2010. In fact, however, the rate of conversion is likely to be less than
that experienced in 1967-75. One of the principal factors underlying that
experience was urban population growth and its continued spread into for-
merly rural areas. While population will continue to grow and to demand
additional land in rural areas for residential purposes, the absolute rate
of growth is expected to decline, which should reduce pressure on the rural
land base relative to 1967-75.
Another important source of pressure on the rural land base in 1967-75
was construction of the interstate highway system. This system now is sub-
stantially completed.
Strip mining of coal likely will become more important over the next
several decades,in response to rising demands for energy and diminishing
supplies of petroleum and natural gas. However, the amount of land used
for this purpose is expected to be quite small, and not all of it will come
from land now constituting the present and potential supply of cropland.^
Thus the annual rate of conversion to urban and similar uses of present
and potential cropland is likely to be less in 1977-2010 than in 1967-75.
We have no reliable way of estimating how much less, but total conversion
over the thirty-three year period of 20-25 million acres is reasonable.^
If non-agricultural demand for present and potential cropland rises
20-25 million acres (implied by Table 4-6), then the 125 million acres of
land with potential for crop production would be reduced to 41 million acres
by 2010. We think it likely that conversion on this scale would increase
the real economic cost of agricultural land. There are two reasons: (1) over
half of the converted land would be that with medium potential, for which
conversion costs are higher than for high potential land (Brewer and Boxley),
(2) The conversion would reduce the supply of land in pasture, forest, and
range, likely increasing the price of these lands unless land-saving tech-
nologies for use of them are developed. Pasture land in particular could
be affected. The NRI shows that almost 40 percent of the land in pasture
in 1977 had potential for conversion to cropland. Shulstad et al. found
that in the Mississippi Delta pasture could be converted to crops more
cheaply than land in other uses, and this likely is generally true. Satis-
fying the projected demand for cropland, therefore, probably could put
Brewer and Boxley cite a study showing that by 2000 only about 1.8
million additional acres would be needed for strip-mining and siting of
coal-fired power plants and associated facilities.
^Robert Gray, Executive Director of the federal government's National
Agricultural Land Study, expects total conversion of 23-25 million acres
of cropland and potential cropland from 1980 to 2000, an annual rate of
1.15 to 1.25 million acres. (Journal of Soil and Water Conservation. 1981,
p. 63).
53
-------
greatest pressure on the supply of pasture. No doubt there is range and
forest land that could be converted to pasture, but this could occur only
at some cost.
We conclude that mounting pressure on the agricultural land base is
likely to increase the real economic costs of that land. We are not able
to say how much these costs might rise.
Any increase, however, would raise real costs of agricultural production
unless prices of non-land inputs or total productivity move to offset this.
We have argued, however, that real prices of non-land inputs are likely to
rise faster than productivity, in which case the effect would be to reinforce
rather than offset the upward pressure of land costs on production costs.
The clear implication of this set of conditions is rising real economic costs
of agricultural production.
54
-------
REFERENCES
Amos, 0. M. 1979. Supply of Potential Cropland in Iowa. Ph.D. thesis in
economics, Iowa States University, Ames. Partially supported by Re-
sources for the Future.
Brewer, M. and R. Boxley. "The Potential Supply of Cropland." In P. Crosson
(ed.), The Adequacy of Agricultural Land in the United States, forth-
coming from the Johns Hopkins University Press for Resources for the
Future.
Collins, K. J., R. B. Evans and R. D. Barry. 1979. World Cotton Production
and Use; Projections for 1985 and 1990. USDA, Foreign Agricultural
Economic Report No. 154. Washington, D.C.
Journal of Soil and Water Conservation." 1981. Vol. 36, No. 2 (March-April).
Shulstad, R. M:, R. D. May, and B. E. Herrington. 1979. Cropland Conversion
Study for the Mississippi Delta Region. A study done for Resources for
the Future. University of Arkansas, Fayetteville.
USDA. 1979. Agricultural Statistics. U.S. Government Printing Office,
Washington, D.C.
. .1980. Agricultural Statistics. U.S. Government Printing Office,
Washington, D.C.
. January 15, 1980. Crop Production 1979 Annual Summary. CrPr 2-1 (80).
Washington, D.C.
. February 1980. Basic Statistics: 1977 National Resources Inventory,
Revised. Soil Conservation Service, Washington, D.C.
. June 27, 1980. Acreage. Crop Reporting Board. CrPr 2-2 (6-80).
Washington, D.C.
. May 1981. Agricultural Outlook (AO-65). Economics and Statistics
Service, Washington, D.C.
55
-------
SECTION 5
PROJECTIONS OF FERTILIZERS AND PESTICIDES
FERTILIZERS
An implication of the shift by farmers toward the land-using end of the
spectrum of technologies is that fertilizer applications per acre will grow
more slowly in the future than in the past. Indeed, as Table 10 in Section
2 shows, per acre applications of fertilizer grew much more slowly after
197 2 than before, reflecting decreasing reliance on land-saving technologies
after that date. As noted in Section 3, the slower Increase in per acre
applications of fertilizer after 1972 was consistent with the post-1972
slowdown in yield growth. Correspondingly, our expectation that the trend
of yields to 2010 will more nearly resemble the post- than the pre-1972
experience is based in good part on our belief that per acre applications of
fertilizer will grow more slowly.
Tables 5-1 and 5-2 show, respectively, amounts of fertilizer applied
per fertilized acre and per harvested acre for corn, wheat, soybeans, and
cotton for selected years, with projections to 1985, 1990, and 2010. The
differences between the two tables, of course, reflect the percentages of
harvested acres which are fertilized.
The key elements in the projections are the percentages of land ferti-
lized and the amounts applied per fertilized acre. For corn and soybeans
in 1985 and 1990 we have used projections of these elements made by Doug-
las. 1 We modified Douglas' projections for wheat and cotton on the basis
of our judgment about recent trends. All of the projections for 2010 are
by us.
Corn
Since about 1970 nitrogen fertilizer has been applied to 94 to 97 per-
cent of corn acreage, phosphorous to 85 to 90 percent and potash to 80 to 85
percent. Our projections for 1985, 1990, and 2010 are 98 percent for nitro-
gen, 90 percent for phosphorous and 95 percent for potash.
Dr. John Douglas, Assistant to the Manager, Tennessee Valley Authori-
ty. Douglas' projections are in the paper listed in the reference's. In
conversation with one of the authors in the summer of'1980 Dr. Douglas indi-
cated he still was satisfied with his projections (they were made In 1978).
56
-------
TABLE 5-1. APPLICATIONS OF FERTILIZER PER ACRE OF FERTILIZED LAND IN
CORN, WHEAT, SOYBEAN AND COTTON (POUNDS)
Rate per receiving acra
Corn Wheat Soybeans Co_t ton
N P K N I. £ • H IE N P K
1965
73
47
42
31
29
13
10
28
34
77
52
34
1970
112
71
72
39
30
36
14
37
51
75
55
57
1972
115
66
69
46
37
38
14
42
51
75
55
61
1973
114
64
71
48
38
36
14
42
55
73
53
62
1974
103
62
73
46
38
37
15
41
55
78
53
55
1975
105
58
67
46
35
35
15
40
53
78
50
55
1976
127
67
78
51
37
37
14
42
60
81
52
56
1977
128
68
82
53
39
41
16
45
60
78
53
52
1978
126
68
80
52
35
34
17
45
62
76
54
54
1979
135
69
84
54
38
43
16
46
67
71
50
44
1980
130
66
86
58
39
40
17
46
70
72
46
46
1985
136
70
86
60
38
42
17
50
73
67
55
50
1990
140
70
87
65
38
43
18
53
80
65
55
50
2010
145
70
90
65
38
43
20
55
85
60
55
50
Sources: 1965-1979 from USDA 19.71, Jan. 1977 and December 1980.
Corn and Soybeans in 1985 and 1990 from Douglas. All
other projections as described in the text.
57
-------
5-
TABLE 5-2. APPLICATIONS OF FERTILIZER PER ACRE OF HARVESTED LAND IN
CORN, WHEAT, SOYBEANS, AND COTTON (POUNDS)
Rate per harvested acre
Com Wheat Soybeans Cotton
NPK NPK . N P K. N P K
1965
64
38
31
15
11
2
•k
3
3
60
30
14
1970
105
64
61
24
13
7
3
10
14
54
26
21
1972
110
59
59
29
16
6
3
12
16
58
30
25
1973
106
55
57
30
17
6
3
13
18
54
29
24
1974
97
54
61
30
17
7
3
11
15
62
31
25
1975
99
50
55
29
15
7
3
10
15
51
22
18
1976
123
60
66
36
19
8
3
12
14
61
28
21
1977
123
60
67
34
17
8
4
15
18
61
27
16
1978
120
59
65
32
13
5
4
16
20
52
24
17
1979
130
61
69
35
17
8
4
17
26
50
24
12
1980
125
57
70
39
17
7
4
16
25
51
22
14
1985
133
63
73
39
19
10
4
18
26
47
28
15
1990
138
63
74
46
20
10
5
20
31
42
25
14
2010
142
63
77
49
22
10
5
22
35
36
22
13
Less than one pound.
Sources: 1965-1979 from USDA 1971, January, 1977 and December, 1980.
Corn and Soybeans in 1985 and 1990 from Douglas. All other
projections as described in the text.
58
-------
Douglas projects very little increase in amounts of fertilizer applied
per acre of corn fertilized, arguing that much if not most land in corn
already is being fertilized at economically optimum rates. Heady expresses
a similair view. If real fertilizer prices, particularly for nitrogen, rise
as we expect, the economic optima will not likely change in the direction
of heavier per acre applications of fertilizer. On the contrary, price
movements likely will induce fanners to find ways to reduce fertilizer use
without commensurate sacrifice of yield. There evidently already are ways
of doing this which would be economically attractive with the higher ferti-
lizer prices now in prospect.^ Stanford (1978) believes that current yield
levels can be maintained with .an appreciable increase in efficiency" of
nitrogen fertilization. He identifies three areas for improvement: more
appropriate application rates, timing of application, and method of appli-
cation. Stanford argues that amounts of nitrogen applied often are higher
than necessary because farmers lack knowledge of the amount of soil organic
nitrogen that will be made available to the plant by mineralization. Stan-
ford is optimistic that current research will develop ways of predicting
mineralization rates, thus enabling the farmer to incorporate this source
of nitrogen into his plans and reduce the amount of purchased nitrogen
applied.
Stanford also sees potential for greater efficiency by splitting appli-
cations of nitrogen in two parts: one application at planting to act as a
"starter" to promote early plant growth, and a second "sidedressed" appli-
cation during the growing season to insure adequate nutrients to harvest.
The alternative is to apply all the nitrogen before or at the time of plant-
ing. The advantage of split applications is that it supplies nutrients at
times more in accord with the plant's need for them over the growing season,
thus reducing nutrient losses that otherwise result from soil processes
working on the nutrient material. However, sidedressing involves some risk.
Wet fields can delay the operation, possibly causing root damage from nitro-
gen deficiency, or allowing the crop to grow too high to drive the applica-
tion equipment through (Aldrich, 1980). Moreover, sidedressing is more
difficult and more time-consuming than pre-plant broadcasting. When anhy-
drous ammonia is knifed-in between the rows, special care must be taken not
to prune corn roots.
Split applications of nitrogen would have the highest pay-off where
leaching losses of nitrogen are greatest, such as in the sandy soils of the
Southeast. In the Midwest, where nitrogen fertilizer use is greatest, split
applications would be less attractive because soils in that area have high
clay content and leaching losses are relatively low. However, if real pri-
ces of nitrogen fertilizer rise as we expect, split applications should be-
come more attractive everywhere.
Stanford also sees some potential for more sparing use of nitrogen by
adoption of foliar application techniques. In field experiments with soy-
2
This account of alternatives permitting more sparing use of fertili-
zers is from Hemphill (1980). See also Aldrich, 1980, especially pages
263-270.
59
-------
beans in Iowa, foliar application outperformed conventional methods. How-
ever, with corn, which receives far more nitrogen than soybeans, problems
were encountered in supplying sufficient nitrogen by foliar application
without causing leaf burn.
Development of new fertilizer materials also offers a way of achieving
more efficient use of fertilizers. New forms of nitrogen hold the greatest
promise because an average crop now takes up only about 50 percent of nitro-
gen applied, less than for either P or K (Fertilizer Institute, pp. 36-39).
The relatively high losses of N result from nitrification, the conversion of
nitrogen fertilizer by soil bacteria to nitrate. Nitrate is soluble, hence
susceptible to movement beyond the reach of crop roots by water. Nitrifica-
tion results in losses of N in gaseous form also. Research is underway
which aims at developing nitrogen fertilizer materials which retard nitri-
fication, thus giving the crop more opportunity to use the nutrient before
it is washed out of reach or volatilized. There are three approaches to
slowing nitrification which presently seem to have most promise: (1) devel-
opment of materials of low solubility which release nitrogen into the soil
solution relatively slowly; (2) provide a temporary barrier between the
nitrogen fertilizer and the soil by coating the nitrogen with slowly decay-
ing materials such as plastic or sulfur; (3) apply nitrification inhibitors,
chemicals which kill enough of the soil bacteria responsible for nitrifica-
tion to slow the process. To date these three approaches have not proved
economical for general application. However, continued research, combined
with further increases in real prices of nitrogen fertilizers should make
them increasingly attractive, particularly in conditions where leaching and
volatilization losses of nitrogen are high.
Relatively greater use of so-called organic farming is a similar likely
response to rising prices of nitrogen fertilizer. Organic farmers substi-
tute animal manure and crop rotations which include a legume for inorganic
fertilizer to supply crop requirements for nitrogen. Since corn receives
far more nitrogen fertilizer than any other crop in the United States (about
AO percent of total applied N) the significance of organic farming with
respect to use of nitrogen fertilizer depends largely on its potential use
in corn production. Many farmers in the United States already practice
organic farming, and according to a USDA report (July 1980) the technology
has promise for more widespread use, particularly if real prices of nitro-
gen fertilizer rise. There is nothing in the report, however, to suggest
that organic farming is likely to substitute in a major way for current
fertilizer management practices. The USDA report examined the economics of
organic and conventional farming in the Midwest using crop budget data
developed by the USDA and Oklahoma State. University. Three organic farming
systems were examined: a four year alfalfa-corn-soybean-oa£s rotation; a
five year alfalfa-alfalfa-corn-soybean-oats rotation; and seven year alfal-
fa-alfalfa-corn-soybean-corn-soybean-alfalfa rotation. The conventional
system was a two year corn-soybean rotation. The conventional system pro-
duced a larger annual return above variable costs than any of the organic
systems. A main reason for this was that with the organic farming systems
only about 57 percent of total crop acreage on average was in corn and soy-
60
-------
S* 4
beans, while the conventional system had 100 percent of its acreage in these
relatively high valued crops.^
The USDA report concludes that the spread of organic farming is limited
by "...the lack of an adequate and economical supply of organic wastes and
residues and/or because soil nutrients and climatic conditions are not suit-
able for successful and profitable organic farming. Based on our observa-
tions, the greatest opportunity for organic farming will probably be on
small farms and on larger mixed crop/livestock farms with large numbers of
animal units" (USDA, July 1980, pp. 46-47).
A large scale shift to organic farming would have a major impact on
cropping patterns and land use. Meeting our crop production projections
would require substantially more cropland than we have projected because the
percentage of land in those crops would be substantially less in an average
year. Meeting our projections with organic farming systems thus would put
far more pressure on the nation's land base, implying higher land costs and
costs of crop production than are implicit in our projections. In fact,
however, our projections probably could not be met if there were a major
shift to organic farming systems. The USDA study examined the macro-economic
effects of such a shift in 1980, and found that corn and soybean prices
would have been significantly higher, while oats, hay and livestock prices
would have been lower. The higher corn and soybean prices would reduce the
volume of exports of these commodities, although the value of exports would
be higher because foreign demand for these crops is inelastic. Average
costs of farm production would have been higher, with consequent higher pri-
ces of food to consumers. Farm income would be higher, indicating that a
major consequence of a large scale shift to organic farming in the United
States would be a redistribution of income to American farmers from U.S. and
foreign consumers of food.
Lockeretz, Shearer and Kohl (1981) compared organic and conventional
farms in the Cornbelt over the 5 years 1974-1978. They explicitly did not
address the macro-economic implications of a large scale shift to organic
farming. Between the two groups of farms compared they found that conven-
tional farms generally had higher yields and gross revenues. However, vari-
able costs also were higher on the conventional farms so their advantage in
net revenues was small—only 2 percent over the 5 years as a whole. This is
a smaller difference than was found in the USDA study. A main reason, evi-
dently, is that many of the conventional farms studied by Lockeretz et. al.
were mixed crop-animal enterprises. These farms typically have some land in
relatively low value forage crops. As noted above, the USDA study assumed
a corn-soybean rotation.
We conclude that the economics of organic farming do not favor a large
scale shift to such systems, and that the macro-economic implications of
3
The study assumed that yields were 10 percent lower with the organic
systems. Even if no difference in yields were assumed, however, returns
above variable costs still would have been higher with the conventional
system.
61
-------
such a shift would inhibit development of political support for it. Accord-
ingly, we view the prospects for organic farming as sufficiently good to
restrain the increase in per acre application of nitrogen fertilizer, but
not so good as to cause a decline.^
Wheat
The percentage of wheat land fertilized is substantially less than that
for corn, and has shown no trend in the 1970s. Sixty to 65 percent of wheat
acreage receives nitrogen, 40 to 50 percent receives phosphorous and 15 to
20 percent receives potash. Douglas projects an increase in all these per-
centages, particularly for nitrogen. The rationale for this is not clear,
and we believe the outlook for fertilizer prices makes it doubtful. Accord-
ingly, we have projected a more modest increase in the percentages of wheat
land fertilized. .
Applications of nitrogen per acre of wheat land fertilized rose in the.
1970s, and we project a continuation, although at a slower rate than Douglas
Phosphorous applications, however, showed no trend and that for potash was
only slightly rising. Our projections of phosphorous and potash applica-
tions per acre of wheat land receiving these nutrients reflect these pat-
terns .
Soybeans
As a legume soybeans biologically fix most of the nitrogen they re-
quire. Consequently, the percentage of soybean acreage receiving nitrogen
fertilizer and the per acre amount applied to that which does are both small
Considerably greater quantities of P and K are applied to soybean land which
receives any of these materials, and application rates, particularly of K,
rose in the 1970s. Douglas projects a continuation of these increases.
However, the percentages of soybean land which receives P and K is small,
and Douglas projects little increase in it. Consequently, the projected
amounts of P and K per acre of land in soybeans are small and show little
increase. Douglas argues that this pattern will characterize fertilizer
use on soybeans absent a significant breakthrough in the agronomics of soy-
bean production.
Cotton
The percentage of cotton acres receiving N, P and K declined slightly
in the 1970s, and the amounts applied per receiving acre were stable to
4
According to the USDA report, organic farming may deplete soil stocks
of phosphorous and potash, indicating that after a few years of organic
farming, these nutrients would have to be applied on a regular basis.
Adoption of organic farming, therefore, particularly on the modest scale
we anticipate, would not restrain per acre applications of P and K.
62
-------
slightly declining. This behavior reflected the shift of cotton acreage to
the Southern Plains—Texas and Oklahoma—where percentages of acres receiv-
ing and amounts per receiving acre are low, from the Southeast and Missis-
sippi Delta, where percentages and amounts are high. Our projections of the
regional distribution of cotton production and acreage assume that the shift
to the Southern Plains will continue. Our projections of fertilizer use are
generally consistent with this shift.
Other Uses of Fertilizer
In addition to its use on corn, wheat, soybeans and cotton, fertilizer
also is applied to other field crops, hay and pasture, fruits and vegeta-
bles, forests, home gardens and lawns, golf courses, roadways and for other
miscellaneous uses (Douglas). In the second half of the 1970s these other
uses took about 43 percent of all the nitrogen applied, 39 percent of the
phosphorous and about 42 percent of the potash. Douglas' projections indi-
cate that the percentage of nitrogen taken by these other uses would rise
to 50 percent in 1985 and to 52 percent in 1990. For P and K Douglas pro-
jects only marginal increases in the percentages taken by these other uses.
We think it likely that the price elasticity of demand for these other
uses of fertilizer is lower than it is for corn, wheat, soybeans and cotton.
The average homeowner, for example, is likely to be less sensitive to the
price of fertilizer in deciding how much to apply to his garden or lawn than
the farmer in deciding how much to apply to his corn crop. Douglas' pro-
jections for the four main crops are consistent with this line of reasoning.
We project the amounts of N, P and K taken by other uses by assuming that
they will be in the same relation to our projected amounts of each used on
the main crops as in Douglas' projections. Table 5-3 shows the projections
for each crop and for all other uses.
PESTICIDES
Data Problems
Projections of pesticide use are on an uncertain footing because of
wide and unexplained discrepancies among the data showing present amounts
used and trends in amounts. Perhaps the most widely cited data are from
surveys of pesticide use conducted by the USDA in 1966, 1971 and 1976.
These data, however, are at variance with those collected by others. This
is indicated in Table 5-4. Von Rumker et. al. carefully examined the dif-
ferences between their estimates and those of the USDA. They concluded that
some of the differences could be accounted for by increased use from 1971
to 1972, e.g., methyl parathion and toxaphene on cotton and soybeans; also
alachlor. The greater part of the differences, however, could not be ex-
plained in this way. Von Rumker et. al. note that the USDA estimates are
from a survey using a "...mammoth form consisting of 56 pages and 487 ques-
tions..." (page 27) which required approximately 3 hours to complete. Twen-
ty-eight of the 487 master questions had to do with pesticides. According
to von Rumker et. al. even a person highly trained in the field of pesti-
63
-------
TABLE 5-3. AMOUNTS OF FERTILIZER APPLIED TO LAND IN CORN, WHEAT,
SOYBEANS, COTTON AND FOR ALL OTHER USES (MILLIONS METRIC
TONS)
1977/79
1985
2010
N
P
K
N
' P
K
N
P
K
Corn
3.97
1.93
2.15
4.83
2.29
2.65
5.96
2.64
3.24
Wheat
.87
.41
.18
1.32
.64
.34
2.01
.90
.41
Soybeans
' .10
.30
.61
.13
.58
.83
.25
1.09
1.73
Cotton
.31
. 16
.08
.30
.18
.10
.25
.15
.09
Total of 4
5.25
2.80
3.02
6.58
3.69
3.92
8.47
4.78
5.47
All Other
4.18
2.11
2.28
6.58
2.67
3.08
9.18
3.46
4.48
Total
9.43
4.91
5.30
13.16
6.36
7.00
17.65
8.24
9.95
Source: 1977/79 from USDA, December 1.980. Projections as described
in text.
cides, with knowledge of the many different kinds and formulations of ma-
terials used, would have difficulty completing the form accurately.
The estimates of von Rumker et. al. are based on comprehensive surveys
of agricultural extension specialists in the various states, directors of
EPA community pesticides studies projects in a number of states, and pesti-
cide manufacturers. For 25 most important pesticides which they studied
intensively von Rumker et. al. concluded that their estimates were accurate
within ± 10 percent for quantities over 10 million pounds and generally
within ± 1 million pounds for quantities less than 10 pounds.
Von Rumker et. al. conclude that their estimates are more accurate than
those of the USDA, and we accept this judgment. However, the USDA estimates
have the major advantage of covering several years (1964, 1966, 1971 and
1976), thus providing information about changes over time. This information
of course is particularly valuable in making judgments about trends in fu-
ture use of pesticides.
We resolved the dilemma posed by differences in the pesticide use data
by rejecting projections of specific quantities of various kinds of pesti-
cides. Instead we project directions of change, and marshal evidence for
judgments of whether the changes are likely to be much or little. For this
purpose we believe the USDA data are adequate and useful, especially since
they were collected on a consistent basis, even though the totals for each
year may be questionable. We focus only on insecticides and herbicides
64
-------
TABLE 5-4. CONTRASTING ESTIMATES OF AMOUNTS OF HERBICIDES AND INSECTI-
CIDES USED IN THE UNITED STATES (MILLIONS POUNDS ACTIVE
INGREDIENTS)
Insecticides
Carbaryl
Carbofuran
Disulfoton
Methyl parathion
Parathion
Toxaphene
Total
Herbicides
Alachlor
Atrazine
2,4-D
Trifluralin
Total
Insecticides on
Cotton
(1) (2) (3)
von Rumker
et al " USDA Col. (1) +
(1972) (1971) Col. (2)
19.0 11.2 1.70
5.0 2.8 1.79
4.9 2.8 1.75
39.7 27.1 1.46
10.0 7.0 1.43
57.0 31.9 1.79
135.6 82.8 1.64
21.0 14.0 1.50
72.0 53.9 1.34
36.0 30.5 1.18
16.8 10.3 1.63
145.8 108.7 1.34
105.0 (approx.) 73.4 1.43
Sources: von Rumker et al 1975; USDA from Andrilenas, 1974.
used in crop production.^
Insecticides
Tables 5-5 through 5-9 show data on insecticides used on crops. For
our purposes the most important information in the tables is the following:
Fungicides and other pesticides accounted for 14 percent of the quan-
tity of all pesticides used on crops in 1976, down from 19 percent in 1971.
The share of these materials among all pesticides used on crops of interest
to this study is much less than these percentages indicate. Moreover, these
materials in general pose fewer threats to the environment than insecticides
and herbicides.
65
-------
TABLE 5-5. PESTICIDE USE ON CROPS IN 1971 AND 1976
Amounts applied Acres treated
(millions lbs.
active ingredients) (millions)
1971
1976
1971
1976
Herbicides
224.0
394.3
157.8
196.6
Insecticides
153.8
162.1
56.7
74.9
Fungicides
39.6
43.2
8.5
10.5
Other
46.3
50.2
10.0
11.6
Total 463.7 649.8 n.a. n.a.
Source: Eichers, Andrilenas and Anderson, 1978, p. 6.
n.a. = not available
TABLE 5-6. PERCENTAGES OF ACRES ON WHICH PESTICIDES WERE USED IN
1976, MAJOR CROPS
Herbi-
cides
Insecti-
cides
Fungi-
cides
Other
pesticides
Any
pesticides
Corn
90
38
1
1
92
Cotton
84
60
9
34
95
Wheat
38
14
1
*
48
Sorghum
51
27
-
*
58
Other grainst
35
5
2
-
41
Soybeans
88
7
3
1
90
Source: Eichers, Andrilenas and
*Less than .5 percent .
Anderson,
1978, p. 7.
tOats, rye and barley
66
-------
TABLE 5-7. HERBICIDES AND INSECTICIDES APPLIED TO MAJOR CROPS IN 1971 AND 1976
(ACTIVE INGREDIENTS)
Herbicides Insecticides
1971
1976
1971
1976
Amount
Amount
Amount
Amount
Total
Per Acre
Total
Per Acre
Total
Per Acre
Total
Per Acre
(mill, lbs)
Treated
(mill, lbs)
Treated
(mill, lbs)
Treated
(mill, lbs)
Treated
(lbs)
(lbs)
(lbs)
(lbs)
Corn
101.1
1.7
207.1
2.7
25.5
1.2
32.0
1.0
Cotton
19.6
1.9
18.3
1.9
73.4
9.8
64.1
9.2
Wheat
11.6
.5
21.9
.7
1.7
.4
7.2
.6
Sorghum
11.5
1.2
15.7
1.7
5.7
.7
4.6
.9
Other grains
5.4
.5
5.5
.5
.8
.7
1.8
1.2
Soybeans
36.5
1.2
81.1
1.8
5.6
1.6
7.9
2.3
Total of these
crops 185.7
n.a.
349.6
n.a.
112.7
n.a.
117.6
n.a.
Total all crops
224.0
1.4
394.3
2.0
153.8
2.7
162.1
2.2
Source: Eichers, Andrilenas, and Anderson, 1978, pp. 9 and 15.
*Oats, rye, and barley.
-------
TABLE 5-8. TYPES AND AMOUNTS OF INSECTICIDES APPLIED TO MAJOR CROPS IN 1976
Millions Percent of
lbs total
Corn
Carbofuran 9.9 30.9
Phorate 5.8 18.1
Fonofos 5.0 15.6
Other 11.3 35.4
Total 32.0 100.0
Cotton
Toxaphene 26.3 41.0
Methyl parathion 20.0 31.2
Other 17.8 27.8
Total 64.1 100.0
Wheat
Parathion 3.1 43.1
Disulfoton 1.8 25.0
Methyl parathion 1.2 16.7
Other 1.1 15.2
Total 7.2 100.0
Soybeans
Carbaryl 3.7 46.8
Toxaphene 2.2 27.9
Other 2.0 25.3
Total 7.9 100.0
Sorghum
Parathion 1.2 26.1
Disulfoton 1.1 23.9
Toxaphene 1.0 21.7
Other 1.3 28.3
Total 4.6 100.0
Source: Eichers, Andrilenas and Anderson, 1978, p. 18.
68
-------
TABLE 5-9. HERBICIDES AND INSECTICIDES APPLIED TO MAIN CROPS, BY REGION IN 1976
(MILLION LBS.)
Corn Cotton Wheat Soybeans Sorghum
Herbi- Insecti- Herbi- Insecti- Herbi- Insecti- Herbi- Insecti- Herbi- Insect-
cides cides cides cides cides cides cides cities cides icides
Northeast
10.93
1
.02
. —
—
.01
.02
1.32
.35
.13
—
Appalachian
19.09
. 94
.75
4.09
.08
.17
8.21
.87
1
.50
—
Southeast
8.13
.97
1.04
20.58
—
—
6.37
6.18
.05
.10
Delta
.39
.02
11.56
32.65
.06
—
15.24
.17
; 42
.49
Lake
33.91
5
.00
—
—
2.41
.01
6.05
.02
.02
—
Corn Belt
108.04
14
.09
—
—
.06
.48
41.51
.12
1,
.30
.30
Northern Plains
22.81
8
.17
—
—
6.22
.40
2.35
*
7,
.94
2.20
Southern Plains
1.66
1
.25
2.76
2.46
.98
4
.49
.01
.15
4,
. 10
1.37
Mountain
1.19
.37
1.30
3.34
3.92
.41
—
—
.22
.06
Pacific
.91
.14
.90
1.02
8.15
1
.27
—
—
.05
.09
U.S.
207.06
31
.98
18.31
64.14
21.88
7
.24
81.06
7.87
15,
.72
4.60
Source: Eichers, Andrilenas and Anderson, 1978, pp. 13 and 19.
*Less than 5000 pounds.
-------
TABLE 5-10. TYPES AND AMOUNTS OF HERBICIDES- APPLIED TO MAJOR CROPS IN 1976
Millions Percent of
lbs ttotal
Corn
Atrazine 83.8 40.5
Alachlor 58.2 28.1
Butylute 24.3 11.7
Other 40.8 19.7
Total 207.1 100.0
Cotton
Trifluralin 7.0 38.3
Flurometuron 5.3 29.0
Other 6.0 32.7
Total 18.3 100.0
Wheat
2, 4-D 15.5 70.8
Other 6.4 29.2
Total 21.9 100.0
Soybeans
Alachlor 29.6 36.5
Trifluralin 21.1 26.0
Other 30.4 37.5
Total 81.1 100.0
Sorghum
Atrazine 6.5 41.4
Propazine 3.9 24.8
Propachlor 3.1 19.8
Other 2.2 14.0
Total 15.7 100.0
Other grain*
2, 4-D 3.8 69.1
Other 1.7 30.9
Total 5.5 100.0
*
Oats, rye and barley.
Source: Eichers, Andrilenas and Anderson, 1978, p. 12.
70
-------
i- /fe
1. The total amount of insecticides applied increased 5.4 percent from
1971 to 1976, much less than the percentage increase in quantity of herbi-
cides.
2. Insecticides applied to cotton accounted for 48 percent of total
crop use of insecticides in 1971 and for 40 percent in 1976.
3. Corn accounted for 17 percent of total crop use of insecticides in
1971 and for 20 percent in 1976.
4. In 1976 83 percent of the insecticides applied to cotton was used
in the Mississippi Delta (50.9 percent) and the Southeast (32.1 percent).
Together, these two regions accounted for one-third of all the insecticides
used on crops in the entire nation.
5. The Cornbelt, Northern Plains and Lake States accounted, respec-
tively, for 44.1 percent 25.5 percent and 15.6 percent of all insecticides
applied to corn in 1976.
6. Amounts of insecticides applied per acre of land in cotton and corn
declined from 1971 to 1976.
7. Organochlorine compounds (e.g. DDT, toxaphene) declined from 46
percent of all insecticides applied to crops in 1971 to 29 percent in 1976.
Organophosphorus compounds (e.g. methyl parathion) increased from 40 to 49
percent and carbamates (e.g. carbaryl, carbofuran) increased from 14 per-
cent to 19 percent.^
These statements suggest that unless there is reason to believe that
insecticides will be applied to sharply higher percentages of land in wheat
and soybeans, the direction of total insecticide use on crops in the future
will depend overwhelmingly on trends in use on cotton in the Southeast and
Delta and on corn in the three principal corn producing regions. Further,
if the substitution of organophosphorous for organochlorine compounds con-
tinues, problems resulting from persistence of insecticides in the environ-
ment will diminish in relative importance and those of acute toxic effects
on humans and animals will increase.
Wheat—
The percentage of wheat acres treated with insecticides and the amount
applied per treated acre are small (Tables 5-6 and 5-7) because wheat gener-
ally is not attacked by insects on a scale requiring a response. The prin-
cipal insect pests of wheat are the Hessian fly, greenbug, wheat stem saw-
fly, armyworms and cutworms (Office of Technology Assessment, OTA). Accord-
ing to the OTA, these pests present "serious to occasional" threats to
wheat. However, the damages typically are sufficiently small that it is
^These data are not in Tables 5-5 through 5-9. They are from Eichers,
Andrilenas and Anderson, 1978, page 16.
71
-------
5T-/7
not economical for farmers to invest heavily in insecticide treatments. By
1971 95 percent of the insecticides used on wheat were organophosphorous
compounds (OTA, p. 21). Use of cultivars with resistance to insects is a
major control practice in wheat production. According to the OTA (p. 27) it
is in fact the most effective means for controlling the Hessian fly and
wheat stem sawfly.
In judging the development of control strategies for wheat insects
into the 1990s the OTA concludes that the role of insecticides likely will
diminish. It is expected that more accurate determination of economic
thresholds will reduce unneeded applications and improvements in application
equipment will result in desired levels of control with smaller amounts of
insecticides.
The OTA report thus gives no reason to believe that the percentage of
wheat acres treated with insecticides or the amounts applied per treated
acre are likely to increase. On the contrary, the thrust of the argument
in the report is that insecticide use on wheat is likely to decline. In
this connection we think it significant that in a major study by the Na-
tional Academy of Sciences of pesticide practices in the United States
separate volumes were prepared dealing with cotton and with corn/soybeans,
but none with wheat. The EPA commissioned a study of alternatives for
reducing insecticides applied to cotton and corn (Pimentel et. al.) but
evidently has seen no need for a similar study for wheat.
We have projected harvested wheat land at 90 million acres in 2010,
compared with 70.8 million acres in 1976, the year to which the USDA's in-
secticide use data apply. If the percentage of wheat land receiving insec-
ticides and the amount per receiving acre in 2010 are the same as in 1976,
9.2 million pounds of these materials would be applied in 2010, up from 7.2
million pounds in 1976 (Table 5-7) . If the OTA is right in thinking that
insecticide per acre usage on wheat will decline, the increase from 1976,
of course, would be less.
Soybeans—
The low percentage of soybean land treated with insecticides (Table
5-6) masks a wide difference between the percentages for soybean land in the
Cornbelt and Delta on the one hand and in the Southeast on the other. Ac-
cording to the OTA report 1 to 10 percent of soybean land in the Cornbelt
is treated annually for insects. In Arkansas the figure in 1977 was 6 per-
cent and' in Louisiana and Mississippi 90 percent and 75 percent respective-
ly. However, the report notes that the figures for Louisiana and Missis-
sippi "were very high compared with normal usage." In the Southeast the
percentage of soybean land treated was 70-75 percent, which evidently is
typical for that region.
In 1977/79 the Southeast had 8.5 percent of the nation's land in soy-
beans. Our projections of production and land use imply that the South-
east's share of soybean land would rise to about 13 percent by 2010. The
Delta's share also would rise somewhat and that of the Cornbelt would
72
-------
S'i?
decline. These regional shifts in land in soybeans would imply a small in-
crease in the nationwide percentage of soybean land receiving insecticides,
assuming persistence of the regional differences in percentages noted above.
According to the OTA report there currently are only 2 major insect
pests of soybeans in the Cornbelt, the Mexican bean beetle and the green
cloverworm. Each now presents a "moderate" threat. By the 1990s the OTA
report judges that the Mexican bean beetle may constitute a "high" threat,
but that the threat from the green cloverworm would remain moderate. The
threats from all other insect pests of soybean pests in the Cornbelt are
judged now to be low to very low, and to remain that way into the 1990s.
The OTA report has little to say concerning future practices for control of
soybean insect pests in the Cornbelt. It notes, however, that insect-
tolerant varieties of soybeans have been identified and a strong effort is
underway to introduce this characteristic into commercial varieties of soy-
beans. According to the report, total resistance is not necessary for soy-
beans in the Cornbelt. Varieties with 50-75 percent tolerance to leaf-
feeding caterpillars and beetles would have little need for insecticides.
The high percentage of soybean land in the Southeast which receives
insecticides indicates that the insect problem in that region is more severe
than in the Cornbelt or Delta. The OTA report concludes that insecticides
will continue to play a role in control of insect pests of soybeans in the
Southeast and in the Delta. The report notes, however, that varieties with
resistance to soybean insects of those regions already have been developed
and asserts that a much greater commitment of resources to research along
this line would have high pay-off over the long term. The report also sees
promise in the present development of biological techniques for control of
soybean insects in the Southeast and Delta, and urges that further develop-
ment of these techniques should receive high priority.
In summary, although the small relative shift of soybean production to
the Southeast would tend to increase slightly the nationwide percentage of
soybean land receiving insecticides, the development of alternative control
practices in the Southeast, Delta, and Cornbelt looks promising enough to
offset this tendency. Unless there is an outbreak of some insect pest of
soybeans not presently identified as a serious threat, it appears unlikely
that the percentage of soybean acreage receiving insecticides or the amounts
applied per receiving acre will increase much, if at all. If the percentage
and amounts per acre remain the same, our projection of land in soybeans
would imply an increase in insecticides used on soybeans from 7.9 million
pounds in 1976 (Table 5-7) to 17.4 million pounds in 2010.
Corn—
The principal insect pests of corn are the northern and western root-
worm, European cornborers and the black cutworm (OTA 1980). Of these, the
corn rootworm is easily the most important, and most of the insecticides
applied to corn are for control of this insect. It is not too much to say
that the future trend in use of insecticides on corn will depend primarily
73
-------
S-fl
on development of practices which modify or substitute for insecticides
currently used to control the rootworm.
At present the principal alternative to insecticides for control of
the corn rootworm is rotation with another crop, primarily soybeans, since
the rootworm must have corn roots to survive. In deciding whether to adopt
this practice the farmer, of course, will consider and balance the economics
of the corn-soybean rotation against its advantages for control of the root-
worm. According to a study of Miranowski (1979) the corn-soybean rotation,
with no use of insecticides, was in fact the most economical means of con-
trolling the corn rootworm in the Cornbelt under cost-price relationships
prevailing in the late 1970s.
Part of the advantage of control by rotation is that it slows the
buildup of genetic resistance of the rootworm to insecticides, thus extend-
ing the useful economic life of currently used chemicals. This advantage
likely will increase in the future because of increasing costs of develop-
ing new insecticides (OTA, Part 3, p. 43).
So-called insect monitoring programs, or "scouting" are coming into
increasing use for control of insect pests of corn. Scouting is the key
component of so-called integrated pest management (IPM) in corn. The prin-
ciple of IPM as currently practiced is to apply insecticides only when,
where and in the amount needed to prevent insect pest damage from crossing
the economic threshold. The programs rely on "scouts" to provide informa-
tion about the number of insects in the field and about the threat they
present at various stages of their life cycle and that of the plant. Typi-
cally these programs result in more sparing use of insecticides for a given
level of insect control than practices-based on rules-of-thumb about when
and how much insecticide to apply. In effect, IPM substitutes knowledge at
the margin for chemicals to control insects.
Limited or experimental scouting programs have been mounted by the
Agricultural Extension Service in the Cornbelt and have been well received
by farmers. The spread of these programs could significantly reduce the
quantities of insecticides needed for effective control of the corn root-
worm (OTA, Part 3, pp. 37-38) .
The development of cultivars of corn with greater insect resistance
than those now used also has good potential as an insect control strategy.
The OTA expects that this strategy in fact will pay-off by the 1990s, re-
sulting in a reduction from 1976 levels in the amount of insecticides ap-
plied to corn (OTA, Part 3, p. 51).
We expect a considerable expansion of corn land in conservation till-
age, for reasons detailed below in the discussion of herbicides. Accord-
David Pimental (personal communication) is less sanguine than the OTA
about the potential for scouting to reduce insecticide use on corn. He
points out that only about one pound of active ingredient is used per acre,
indicating that the cost of treatment is low.
74
-------
S-2&
ing to the OTA report this could increase the threat to corn of insects now
only of minor importance. Conservation tillage leaves residue from the pre-
vious crop on the soil surface, creating a hospitable environment for in-
sects, particularly the black cutworm and armyworm, which could not flourish
in a clean tilled field. To the extent that this occurs, it would offset
the effects of increased corn-soybean rotation and scouting in reducing the
amount of insecticides applied to corn. It is worth noting in this connec-
tion, however, that the amount of land in conservation tillage doubled be-
tween 1972 and 1976 (No-Till Farmer, March 1979) but total insecticide use
increased less than 6 percent (Table 5-5). If conservation tillage required
significantly more insecticides than conventional tillage one would have
expected insecticide use to increase more. Conservation tillage definitely
requires more herbicides than conventional tillage, and harbicide use rose
sharply from 1971 to 1976 (Table 5-5).
Our projection of land in com indicates a 30 percent increase from
71.3 million acres in 1976 to 93 million acres in 2010. We expect a decline
in the amount of insecticides applied per receiving acre of corn, or in the
percentage of land receiving insecticides, or in some combination of the
two, to more than offset the increase in land in corn.® Accordingly, we
expect the total amount of insecticides applied to corn to decline over the
next several decades.
Cotton—
We expect a significant decline from 1976 to 2010 in the total amount
of insecticides applied to cotton. There are two reasons: (1) a marked
shift in the regional distribution of land in cotton from the Mississippi
Delta, where rates per receiving acre are high, to Texas where rates per
receiving acre are low; (2) a decline in present high rates per receiving
acre in the Delta arid in the Southeast.
Because the climate in the High Plains of Texas is hostile to cotton
insect pests, cotton farmers in that region can produce a crop in most years
without applying any insecticides (National Academy of Sciences 1975) .
Since 40-45 percent of the cotton acreage in.Texas is in the High Plains,
the low per acre amounts of insecticides applied in that area keep the aver-
age amount for the state as a whole"low.
But application rates elsewhere in Texas also are low in comparison
with rates in the Southeast and Delta. The reasons for this are complex,
and what follows does not pretend to be a complete account. There appear
to be two key components to the explanation, however. One is that beginning
in the 1950s the key insect pests of cotton in Texas began to develop gene-
According to Table 5-7 a decline in the amount of insecticides per
receiving acre of corn already had occurred from 1971 to 1976.
75
-------
tic resistance to insecticidesFirst the boll weevil, then the cotton
fleahopper and thrips, became effectively immune to DDT and other organo-
chlorine compounds. Subsequently, the cotton bollworm and tobacco budworm
also developed resistance to these materials. Farmers switched to organo-
phosphorous compounds, at first with good success, but subsequently the
tobacco budworm developed resistance to these as well. The result was that
when farmers sprayed with organophosphates to control the boll weevil or
other primary insect pests, they also killed off predators of the budworm,
which then was able to, and often did, wreak considerable damage.
Researchers went to work on this problem and came up with insect manage-
ment practices which usually give adequate control with relatively little
reliance on insecticides. These practices are built arourid the fact that in
most instances the tobacco budworm is a secondary pest of cotton whose num-
bers can be kept below economically damaging amounts by reliance on insect
predators of the budworm. The key is not to destroy the predators. This
is avoided by spraying either very early in the cotton growing season with
one or two treatments to kill boll weevils surviving the winter or for flea-
hopper infestation, or very late in the season to get weevils preparing to
overwinter. These approaches avoid outbreaks of the budworm-bollworm com-
plex so well that multiapplications of insecticides no longer are needed or
considered good practice by most cotton producers in Texas.
This mode of attack has benefited also from the development and adoption
of "short-season" varieties of cotton in Texas. These varieties mature ear-
lier than the varieties they have replaced, getting them through the stage
of development when they are most vulnerable to insect attack before insect
populations swell to threatening numbers.
The other major component to the explanation of why per acre applica-
tions of insecticides to cotton are relatively low in Texas is that the yield
potential of cotton land in the state is less than in the Delta and South-
east (National Academy of Sciences 1975, p. 65). Consequently, the yield
increase that Texas cotton farmers would obtain if they put on as much
insecticides as farmers in the Delta and Southeast would not justify the
additional cost. Moreover, cotton in the Southeast and Delta encounters
stronger competition than in Texas from other high value crops, e.g., soy-
beans. To meet the competition farmers in those regions must get higher
cotton yields than in Texas, which implies higher per acre use of insecti-
cides.^
In 1976 .51 pounds of insecticides were applied per acre of cotton in
Texas and Oklahoma. The figures were 10.96 pounds in the Delta and 24.92
~
This account of developing insect resistance to insecticides is based
on OTA, Part 8, and on conversations with Ray Frisbie, Michael McWhorter,
Robert Metzer and Knox Walker, all of Texas A&M University.
^Cotton yields in Texas averaged 345 pounds per acre in 1977/79. In
the Delta they were 566 pounds and in the Southeast 482 pounds (USDA, Janu-
ary 15, 1980.
76
-------
pounds in the Southeast.^ The average application rate for the three re-
gions was 6.44 pounds per acre. Our projections of the regional distribu-
tion of land in cotton would reduce this average to 4.06 pounds in 1985 and
to 3.29 pounds in 2010, even if per acre applications in each region re-
mained at the 1976 level. In this case the amounts of insecticides used on
cotton in the three regions would decline from 55.7 million pounds in 1976
to 45.3 million pounds in 1985 and 38.8 million pounds in 2010.
Per acre applications of insecticides to cotton in Texas already are so
low that they probably will not go much lower. They could rise if the boll
weevil or other primary pests were to develop resistance to the organophos-
phorous compounds now used, requiring increased applications for a given
amount of control. This ultimately would be a self-defeating response, how-
ever. And the demonstrated capacity of the research establishment and cot-
ton producers in Texas to develop insect control strategies requiring only
sparing use of insecticides suggests they would not respond to increasing
boll weevil resistance simply by increasing per acre application rates. On
this ground we think it unlikely that rates will rise much if at all in
Texas. Even if they do increase somewhat, they are so low relative to those
in the Southeast and the Delta that the projected shift of cotton acreage
to Texas still would reduce the average application rate and total amount of
insecticides used on cotton in the three regions.
As noted above, the higher, yield potential of cotton land in the South-
east and Delta compared to Texas makes it economical for farmers in those
two regions to apply more insecticides per acre than Texas farmers do. This
suggests that per acre applications in the Southeast and Delta will continue
to be well above rates in Texas for the foreseeable future. Nonetheless,
there is reason to believe that they will decline from the 1976 rates. In-
deed, some observers believe they already have, at least in some places'.^
Where amounts applied per acre have declined the increased use of synthetic
pyrethroids is one of the reasons. This material, effective against the
budworm-bollworm complex where other insecticides are not, is applied in
much smaller amounts than the material it replaces. However, according to
^Total insecticides applied in each region are from Eichers, Andrilenas
and Anderson. Land in cotton is from USDA 1979,
12
At the beginning of this discussion of pesticides it was noted that
the USDA estimates of use differ significantly from others and are probably
too low. This would not likely affect the proportionate distribution of
use among regions, however. Consequently, while the above estimates of per
acre insecticide use on cotton in each region may be too low, estimates of
the effect of a shift to the Southern Plains on the average application rate
for the three regions should accurate enough.
13
In conversation with one of the authors in the winter of 1980, James
Hamer, Agricultural Extension specialist at Mississippi State University
involved in cotton pest management programs, stated that insecticide use on
cotton in Mississippi probably had declined from the 1976 level.
77
-------
Hamer (see footnote 13) the apparent decline after 1976 in pounds of insect-
icides applied to cotton in Mississippi was not just because of the increased
use of pyrethroids.
Cotton fanners in the Delta and Southeast have rapidly increased the
use of cotton pest management consultants in the last few years. According
to Hamer these consultants now provide pest management services to operators
of about 85 percent or 730,000 to 750,000 acres of cotton land in the Delta
region of Mississippi. About 75 percent of cotton acreage in Alabama is
said to be scouted for insects by extension trained people and about 18 per-
cent by professional consultants.^ Most farmers in South Carolina use
scouts, according to one source, and in Georgia about 85 percent of cotton
acreage is scouted, according to another.While data are not available
showing the effects of the spread of scouting on insecticide application
rates, the presumption is that rates required for a given amount of protec-
tion decline. As pointed out in the discussion of IPM in corn, scouting in-
volves substitution at the margin of knowledge for insecticides.
As noted above, one of the key components of cotton insect management
strategies in Texas is use of "short-season", or early maturing varieties,
the advantage of these varieties being that they pass through the stages
of greatest vulnerability to insect attack before insect populations reach
threatening size. It is natural to ask whether these varieties could not
also be used in the Delta and Southeast to reduce requirements for insecti-
cides.
Research is under way to develop short-season varieties adapted to
those regions. Some are now used in Mississippi but the acreage is very
small. There are some important obstacles to widespread adoption of these
varieties, however. The Texas varieties were developed for use in a dry
climate, a condition not met in the Delta and Southeast. Not only is the
climate wetter in those regions, but it also is less predictable. The tra-
ditional, long-season varieties used in the Delta and Southeast are more
resilient in recovering from unexpected weather .stress than the short-season
varieties of Texas. Where the short-season varieties have been tried in
experiment stations in the Southeast yields have been lower than with the
traditional varieties. This is not a serious disadvantage in Texas where
total costs per acre generally are lower than in the Southeast, but.in the
latter region high yields are necessary to offset higher costs. Finally, in
the Southeast and parts of the Delta the spring is colder than in south
Texas, which delays planting, thus negating some of the advantage of the
Communication with John French, pest management specialist with the
Cooperative Extension Service, Auburn University.
¦^Communication with Donald Johnson, Department of Entomology and Zool-
ogy, Clemson University, and with William Lambert, Coastal Plain Experiment
Station, Tifton, Georgia.
78
-------
short-season varieties.16
The USDA is sponsoring research on three approaches to management of
cotton insect pests: boll weevil eradication (BWE), optimum pest management
(OPM), and current insect control (CIC) practices. The objective is to de-
scribe relevant features of the alternative approaches and estimate their
economic and environmental impacts, both at the farm level and throughout
the cotton belt. The studies are expected to serve as a basis for choosing
among insect management strategies and for policies to promote the chosen
strategies, or mix of strategies. Since the results of the studies are not
available at this writing it is impossible to know what their implications
may be with respect to trends in quantities of insecticides applied to cot-
ton. At least one participant in the studies, however* asserts that the
main difference between the OPM strategy and the CIC strategy is that OPM
would increase the role of knowledge about plant-insect interactions and
reduce the role of insecticides, primarily because the rising costs of the
latter make it imperative to use them more sparingly.^
There is much other research underway, most of it pointed toward find-
ing alternatives to insecticides for insect control, alternatives which
would not likely eliminate the use of insecticides, but which would diminish
their role. One approach now receiving attention is development of insect
resistant varieties. A variety with some resistance to the boll weevil,
tobacco budworm, and fleahopper already has been developed in Texas.^ Some
use has been made of bacteria and viruses for control of the budworm complex
in the Delta, although not in' the Southeast, and research is continuing
along these lines.19
While some of the various alternative strategies now being explored
through research probably will not prove economically viable at the farm
level, others almost surely will. Indeed, as just noted, some already have.
We believe this outcome is sufficiently-likely to justify the conclusion
that per acre applications of insecticides to cotton in the Southeast and
Delta will decline from 1976 levels, probably substantially by 2010. Since
we project a reduction in total cotton acreage in those regions, we expect
the total insecticide load also to decline. When this is combined with the
projected shift of cotton acreage to Texas, the implication is a substantial
drop for the three regions combined both in per acre applications of insect-
icides to cotton and in the total quantity of insecticides used.
This account of obstacles to the adoption of short-season cotton vari-
eties in the Delta and Southeast is based on discussion with Donald John-
son, Clemson University.
¦^Donald Johnson, Clemson University, personal communication.
18
Raymond Frisbie, Texas A&M University, personal communication
19
Donald Johnson, Clemson University.
79
-------
On balance, we expect total insecticide use on grains, soybeans and
cotton to decline, primarily but not exclusively because of declining use on
cotton. A study by Headley (1978) supports this conclusion with respect to
grains and soybeans. (The study was not concerned with cotton.) As part
of the study Headley surveyed U.S. agricultural extension and research work-
ers familiar with pest management practices in grains and soybeans. The 39
responses indicated a consensus that chemical insecticides would continue
to play a major role -.into the 1990s, but that the trend of use would be
declining. Resistant varieties, already of major importance for control
of grain and soybean insects, would be on a rising trend. All other bio-
logical controls, e.g., deliberate use of parasites and predators, bacteria,
viruses and phermones would remain of minor importance.
Herbicides
Importance of Conservation Tillage—
Table 5-5 shows a substantial increase from 1971 to 1976 in the amounts
of herbicides applied to crops. Much of this increase must have resulted
from the spread of conservation tillage. Conservation tillage means any of
a variety of tillage practices which may differ in many details but which
are distinguished from conventional tillage by three common features: (1)
they rely on some instrument other than the moldboard plow to prepare the
land for planting; (2) they leave enough, residue from the previous crop on
the soil surface to significantly reduce erosion; (3) they rely more on
herbicides and less on mechanical cultivation to control weeds.
It is the third characteristic which is relevant for the present dis-
cussion.^® Conservation tillage relies heavily on herbicides to control
weeds. In no-till systems, a polar case of conservation tillage, the reli-
ance typically is completie, i.e. weed control relies exclusively on herbi-
cides. Other forms of conservation tillage may include some cultivation,
but all usually require more herbicides per acre for weed control than con-
ventional tillage. The amount more, however, is highly variable, judging
from the literature (e.g., University of Illinois, pp. 6-7; Taylor, Reneau,
and Trimble; Doster and Phillips; Walker; Amemiya; R.E. Phillips, et. al.) .
There are a number of reasons why any form of conservation tillage1may
require more herbicides than conventional tillage, and why no-till defi-
nitely does. One might be called the substitution effect. Under specific
conditions a given level of weed control may be achieved either by tillage
or by herbicides, or by some combination of them. If control by tillage is
reduced than a compensatory increase in the amount of herbicides applied is
necessary to maintain the same level of control.
A second reason why conservation tillage likely will require more
herbicides is the efficiency effect. Apart from the substitution effect,
20
The following discussion of conservation tillage is from Crosson,
1981.
80
-------
more herbicide must be applied to achieve a given level of weed control
because some of the herbicide gets tied up by the crop residue which is a
necessary concomitant of conservation tillage. Thus the efficiency of her-
bicide use is reduced (Siemens and Oschwald, citing Erbach and Lovely; Fens-
ter and McCalla; Griffith, Mannering and Moldenhauer; University of Illi-
nois; Robert Pope and Joseph Downs, University of Illinois, in discussion
with one of the authors).
The environmental effect is a third reason why conservation tillage may
require more herbicides than conventional tillage. The surface residue
which accompanies conservation tillage usually reduces evaporation of water,
so that typically soils are more moist with conservation tillage than with
conventional tillage. The increased moisture improves the conditions for
germination and growth of weeds (Griffith, Mannering and Moldenhauer; Pope
and Downs in conversation with one of the authors).
The substitution, efficiency, and environmental effects are independent
of one another, so their, contributions to increased herbicide requirements
with conservation tillage are additive. The literature does not indicate
that all three effects are present in all cases of conservation tillage,
and in fact there may be some circumstances, e.g., where weed problems are
not serious in any case, where none of the three effects is important. In
the usual case, however, it appears that one or more of the effects comes
into play to increase herbicide requirements for conservation tillage rela-
tive to conventional tillage.
A recurrent theme in the literature is that over time perennial weeds
become more important with conservation tillage. The reason is that herbi-
cides do not attack the root systems of these weeds as tillage does, thus
giving them a competitive advantage relative to annual weeds. It is not
clear that the relative shift from annual to perennial weeds requires hea-
vier applications of herbicides. It may mean instead that conservation
tillage becomes uneconomic relative to conventional tillage in places where
perennial weeds are a problem because some of these weeds cannot be ade-
quately controlled with any amount of herbicides. Johnsongrass and bermuda-
grass in particular are in this category (Unger, Wiese and Allen; McWhorter
and Jordan; Reicosky et. al.; Harrold, Triplett and Edwards; Pope and Downs
in conversation with one of the authors). In such instances the economic
disadvantage of conservation tillage results from unfavorable yields, not
from the need for heavier applications of herbicides.
Judgments about future trends in the amount of herbicides applied to
crops depend crucially upon prospects for the spread of conservation till-
age. These prospects will be shaped by the comparative economics of conser-
vation tillage and conventional tillage, and by public policies to deal with
their respective environmental consequences.
81
-------
Economics of Conservation Tillage—
Table 5-11 gives estimates of land in conservation tillage in 1965 and
subsequently, and Table 5-12 shows the regional distribution in 1978 and
1979. In the period covered by Table 5-11 there were no special public pro-
grams to promote adoption of conservation tillage. Its rapid spread since
the mid-1960s, therefore, suggests that conservation tillage has important
economic advantages over conventional tillage under a wide range of soil
and climate conditions.
Costs per acre—A. survey of the literature on tillage technologies
(Crosson, 1981) indicates that total costs per acre are roughly 5 to 10 per-
cent less for conservation tillage than for conventional tillage, the big-
gest differences being for sorghum and wheat and the smallest for corn and
cotton. The difference for soybeans is intermediate. These differences
reflect the net outcomes of lower costs for pre-harvest labor, for machin-
ery, for fuel, and higher costs for pesticides, mostly herbicides. The
lower labor and fuel costs result from the elimination of some (all in the
case of no-till) cultivation to control weeds. Machinery costs typically
are lower with conservation tillage because the practice involves less dis-
turbance of the soil, hence does not require as powerful, and expensive, a
tractor. We have not seen it stated in the literature but it seems likely
that tractor maintenance costs also would be lower since with conservation
tillage the farmer makes fewer passes over the field, which should reduce
tractor wear and tear. The literature is unclear as to whether conservation
tillage requires more or the same amount of fertilizer per acre to achieve
given yields. We assume the same amount is required.
There is a clear consensus in the literature that conservation tillage
requires "more" management than conventional tillage, but it is not so
clear what this means for the costs of production of the two systems. Whe--
ther "more" management increases costs depends on what it costs farmers to
acquire the additional management skills. These costs would be of three
sorts: (1) out-of-pocket expenses, e.g. for subscriptions to farm manage-
ment publications or tuition for a short course in conservation tillage
practices; (2) time spent in reading the publications, attending the course,
and talking with neighbors, extension agents, soil conservation technicians
and vendors of conservation tillage technology; (3) the difference between
the net income the farmer would earn if he stayed with conventional tillage
and what he would earn from conservation tillage, i.e. the costs of learn-
ing-by-doing. The first kind of cost surely would be trivial, and the
income foregone by engaging in the activities listed under (2) likely would
be small also. The psychic costs of these activities could be high, how-
ever, if the farmer places a high value on his leisure. The pecuniary costs
of learning-by-doing (the third kind of cost) clearly could be significant,
depending upon the specific circumstances the farmer faces and how quick he
is in acquiring the necessary knowledge through processes (1) and (2).
In estimating the future spread of conservation tillage increased
management costs obviously must be considered, but in our judgment they are
not likely to be of major importance. The history of American agriculture,
82
-------
TABLE 5-11. LAND IN CONSERVATION TILLAGE IN THE UNITED STATES
(MILLIONS ACRES)
% of No-Till Farmer % of
Harvested
No-Till+
Minimum
Harvested
USDA
Cropland
Till
Total
Cropland
1965
6.6
2.3
NA
NA
NA
NA
1973
29.5
9.3
4.9
39.1
44.0
13.9
1975
35.8
10.8
6.5
49.7
56.2
17.0
1976
39.2
11.8
7.5
52.1
59.6
18.0
1977
47.5
14.1
7.3
62.7
70.0
20.7
1978
51.7
15.6
7.1
67.7
74.8
22.6
1979*
55.0
16.1
7.6
71.6
79.2
23.2
Sources: USDA, from Gerald Darby, Conservation Agronomist with the
Soil Conservation Service. Based on reports from SCS county
field offices.
No-Till Farmer, March 1979, pp. 4-5. Estimates by state
agronomists of the Soil Conservation Service.
Harvested Cropland 1965-1979 from USDA, Feb. 1981.
*Preliminary
tDefined as "where only the intermediate seed zone is prepared. Up to
25% of surface area could be worked. Could be no-till, till-plant, chisel
plant rotary strip tillage, etc. Includes many forms of conservation til-
lage and mulch tillage."
NA = not available.
83
-------
TABLE 5-12. CONSERVATION TILLAGE BY REGION
I
1978 1979
Millions % of Cropland Millions % of Cropland
Acres harvested Acres harvested
Northeast
2.23
17.2
2.40
18.2
Lake states
6.78
18.2
6.93
17.9
Corn Belt
20.33
24.6
21.73
28.7
Northern Plains
23.65
32.2
25.01
33.3
Appalachia
5.42
29.8
5.83
30.7
Southeast
4.55
31.4
5.09
33.5
Delta
1.26
6.7
1.36
7.0
Southern Plains
2.24
7.5
1.98
6.1
Mountain
6.31
24.6
6.43
24.8
Pacific
1.90
10.9
2.35
13.5
Total
74.67
22.6
79.11
23.1
Sources: Land in conservation tillage from No-Till Farmer, March 1979.
Cropland harvested: USDA, Feb. 1981.
The No-Till Farmer estimates that in both 1978 and 1979 the
sum of land in no-tillage, minimum-tillage and conventional
tillage in the country as a whole was 300 million acres. The
USDA (Feb. 1981) shows 331 million acres of cropland harves-
ted in 1978 and 342 million acres in 1979. If some of the
"missing" cropland were in conservation tillage, the figures
in the table would be too low.
especially since the end of World War II, demonstrates that American farm-
ers can quickly learn to use new, complex technologies when it is in their
economic interest to do so. The extraordinary increases in this period in
the use of fertilizers, pesticides, larger and more complex machinery, new
seed varieties, new practices with respect to row spacing, and so on all
attest to this, as well as to the efficiency of the U.S. agricultural es-
tablishment in developing new technologies and spreading the knowledge of
how to use them. Indeed, the rapid spread of conservation tillage since
the mid-1960s is itself testimony to the managerial capacity of farmers
and the institutional structure which serves them. Accordingly, we do not
believe that the costs of acquiring the managerial skills required by con-
servation tillage will significantly impede its spread over the longest
term. Should conservation tillage not continue to expand, it likely will
be for reasons other than increased costs of management.
Yields—If there were no difference in yields between conservation
tillage and conventional tillage the cost advantage of conservation tillage
84
-------
5"- 3°
would indicate its continued rapid substitution for conventional tillage.
Comparative yields of the two systems thus must be considered.
It is important to distinguish between the short term (one or two years)
and long term (decades) yield effects of tillage technologies. Conservation
tillage reduces erosion by 50 to 90 percent compared to conventional tillage
(more on this in the next section), and over the long term this can give
conservation tillage a decisive yield advantage. Whether this occurs de-
pends upon (1) the differential advantage of conservation tillage in reduc-
ing erosion; (2) the amount of topsoil and the nature of the underlying
parent material; (3) the relation of changes in the amount of topsoil to
changes in yield over time.
Depending upon these three conditions, the effect of the erosion factor
in the choice of tillage technologies is either neutral or it favors conser-
vation tillage. It can never favor conventional tillage. Consequently,
where conservation tillage confers a short-term yield advantage farmers can
be expected to choose that technology, given its cost advantage. Even where
short-term yields of conservation tillage are less, however, farmers may yet
choose that technology over conventional tillage if the erosion advantage of
conservation tillage is sufficiently strong. The strength of the advantage
depends upon (1) the annual differences in yields between conservation and
conventional tillage as determined by the three factors listed above; (2)
the cost to the farmer of substituting fertilizers or other inputs for the
lost soil, thus offsetting the erosion induced decline in yields; (3) the
length of time over which the yield differences matter to the farmer; (4)
the farmer's expectations about future crop prices relative to current crop
prices; (5) the rate of discount the farmer applies to future earnings.
There is much speculation about these factors in the literature, but not
enough hard facts to warrant firm generalizations about effects on farmers'
choices between conservation tillage and conventional tillage.
The only confident statement we can make is one already made above:
since conservation tillage never produces more erosion than conventional
tillage and typically produces much less, the erosion effect on yields can
never be to the disadvantage of conservation tillage. At worst it will be
neutral, and it must often be positive, although without much additional
research we cannot identify the specific situations in which this would be
true or estimate the strength of the advantage.
The question of the long-term yield effects of erosion would be moot
if conservation tillage had a clear short-term yield advantage over con-
ventional tillage. The vast majority of the literature in fact deals with
short-term yield differences and shows no clear pattern of advantage for
either technology. Some studies show yields higher for conservation till-
age, others show them lower and still others show no significant differences.
The specific circumstances of soil, weather, and kind and quantity of weeds
are of crucial importance in determining relative yield response, and these
conditions differ widely, both spatially and even temporally in the same
location. Other factors less often mentioned, and apparently less impor-
tant are seed placement and soil compaction.
85
-------
The key factors appear to be soil moisture and temperature, length of
growing season and weeds. There is a strong consensus in the literature
that in the plant root zone soil moisture generally is higher with conserva-
tion tillage and that in the spring soil temperatures are lower. Both of
these conditions have a bearing on yields of conservation tillage relative
to conventional tillage. In areas where rainfall is the principal factor
limiting plant growth, as in part of the Northern Plains, the moisture con-
serving feature of conservation tillage is a distinct plus. Indeed, this
appears to be the principal reason for the relatively widespread use of
conservation tillage in that area (see Table 5-12) . The moisture retaining
characteristic of conservation tillage, however, is an advantage in any area
subject to periods of drought or where soils are droughty. The character-
istic is a distinct disadvantage, however, on poorly drained soils.
Cosper (1979) implicitly assuming that yield response is crucial in
determining the competitiveness of conservation tillage relative to conven-
tional tillage, finds the key to yield response in soil characteristics,
with moisture retention capacity most important. More specifically, he
asserts that "those physical features most significant in designating an
acceptable tillage system include inherent drainage and soil wetness levels,
structural stability, water percolation and surface soil texture. Among
these properties, soil wetness is perhaps the most restrictive in limiting
the range of usable tillage systems" (p. 17, emphasis added) . Cosper uses
these soil characteristics, especially wetness, to develop estimates of the
maximum amount of land that could be economically put in conservation till-
age in Ohio, Indiana, Illinois, and Iowa. His judgment that soil drainage .
is a key element in yield differences between conservation tillage and con-
ventional tillage is widely shared in the literature.
Apart from soil wetness soil temperature is important to seed germina-
tion and seedling emergence and affects nutrient uptake by plants. If soil
temperatures are too low, the availability and uptake of nutrients, parti-
cularly P and N, will be inhibited, slowing seed germination, seedling emer-
gence, and subsequent early growth (Willis and Amemiya). Soil temperature
generally is lower with conservation tillage than with conventional tillage,
especially in the spring, because the crop residue reflects some of the
sun's heat. On poorly drained soils temperatures may also be lower with
conservation tillage because of the greater moisture content of these soils
and the high heat capacity of water.
The delay in seed germination, emergence and early growth resulting
from low soil temperatures may cause the seedlings to be relatively unde-
veloped and vulnerable when populations of disease organisms and insects
begin to expand. The delay in early growth also is a disadvantage where
crops require a full season for best yield performance, as is true of corn
and soybeans in Illinois. Indeed, R.E. Phillips et. al. assert that the
delay in spring planting required by lower soil temperatures is a disadvant-
age to conservation tillage throughout the central and northern United
States. Amemiya (p. 33) also asserts that the low soil temperature asso-
ciated with conservation tillage is a "serious obstacle to adoption" of the
technology in the northern Cornbelt.
86
-------
The length of the growing season has a differential effect on yields
of conservation tillage and conventional tillage in two ways. One has just
been indicated. As the growing season shortens, the delay in planting im-
posed by the cooler soil temperatures of conservation tillage imposes an
increasing yield penalty on that practice. Hence the attractiveness of con-
servation tillage diminishes as one moves from south to north. However, as
one moves from north to south the growing season lengthens, and conserva-
tion tillage opens up opportunities for double-cropping not available with
conventional tillage. This is the other effect of the growing season on
yields. Because conservation tillage permits the planting of a crop direct-
ly in the stubble of a preceding crop immediately after harvest it saves
time. The time saved may be enough to take a second crop with conservation
tillage where this would be impossible with conventional tillage. Double-
cropping, of course, increases yields, thus reducing average costs of land
and other fixed inputs. Double-cropping with conservation tillage already
is practiced in the southern Cornbelt and Southeast, typically with wheat
followed by soybeans.
Inadequate weed control is a major factor adversely affecting yields
with conservation tillage. Weeds compete with crops for sunlight, moisture
and nutrients, and where they cannot be controlled yields will suffer. In
some areas and for some weeds, adequate control with herbicides is diffi-
cult or impossible. It was noted earlier that perennial weeds typically
become more important with uninterrupted use of conservation tillage (Rei-
chert; Davidson and Santelmann; Griffith, Mannering, and Moldenhauer; Ran-
dell and Swan; Amemiya; Wiese and Staniforth). A companion theme is that
there are no fully effective herbicides for controlling certain perennial
weeds and that conservation tillage will face severe difficulties or should
not be used at all in areas infested with these weeds. Johnsongrass, nut-
sedge, broomsedge, bermudagrass, bullnettle and bindweed are commonly men-
tioned.
Summary on Economics—
In production of corn, sorghum, wheat, soybeans, and cotton conserva-
tion tillage has a cost advantage of roughly 5 to 10 percent relative to
conventional tillage. It follows that in areas with well drained soils,
adequate control of weeds with herbicides, and potential for double-cropping
economics would clearly favor conservation tillage. Two questions arise:
(1) is the present geographical distribution of conservation tillage con-
sistent with this assertion? (2) Does the assertion provide any insights
to the role of economics in the future spread of conservation tillage?
There is little information with which to address either of these ques-
tions. With respect to the first, however, Table 5-13 is relevant. Column
(1) in the table is based on Cosper's analysis of soil types in the four
states, with special attention to soil drainage characteristics. The column
indicates that as one moves from the eastern to the western Cornbelt the
percentage of tillable land with drainage characteristics favorable to con-
servation tillage increases. Column (2) indicates a similar progression in
the percentage of land actually in conservation tillage. The two columns,
therefore, are consistent with the hypothesis that soil drainage is a key
87
-------
TABLE 5-13. PERCENTAGES OF LAND APT FOR CONSERVATION TILLAGE AND IN
CONSERVATION TILLAGE, OHIO, INDIANA, ILLINOIS, AND IOWA
(1)
Percent Apt
for conservation
Percent
in conservation
(2)
tillage*
tillage, 1979t
Ohio
Indiana
Illino is
Iowa
47.5
53.4
65.9
76.4
8.0
22.8
28.0
38.9
*From Harold Cosper (p. 26). Land is "tillable acres", and for each
state is almost exactly the same as the sum of cropland and pasture as
reported in the USDA's National Resources Inventory of 1977.
tFrom No-Till Farmer, March 1979.
factor in determining the yields and hence the economic feasibility of con-
servation tillage.
Cosper is extending his analysis of soil types to the Northern Plains,
but those estimates are not available as of this writing. Nor are there
similar estimates for any other region of the country. Consequently, we
are unable to judge whether the distribution of conservation tillage outside
the Cornbelt is consistent with the hypothesis that soil drainage is a key
factor.
There are other useful hypotheses, however, which can be examined.
Because of the semiarid climate in the Northern Plains (North Dakota, South
Dakota, Nebraska, and Kansas) one would expect the percentage of land in
conservation tillage in the region to be relatively high. However, because
conservation tillage lowers soil temperature in the spring, thus delaying
planting relative to conventional tillage, the percentage of land in con-
servation tillage would be expected to decline as one moves with the dimin-
ishing growing season from Kansas to North Dakota. For the same reason one
would expect the percentage of conservation-tilled land in the Lake States
(Michigan, Wisconsin, and Minnesota) to be less than in the Cornbelt. Be-
cause of the favorable prospects for double-cropping with conservation till-
age in the Southeast (Lewis) the percentage of conservation tilled land
ought to be relatively high in that region, while it ought to be relatively
low in the Mississippi Delta and the Southern Plains (Texas and Oklahoma)
because of the relatively large amounts of land in cotton in those regions
and the problems of controlling weeds, particularly perennial weeds, with
herbicides in cotton.
88
-------
5*-
TABLE 5-14. PERCENT DISTRIBUTION OF CROPLAND IN CONSERVATION
TILLAGE, VARIOUS STATES AND REGIONS, 1979
Percent
Kansas
50.1
Nebraska
47.1
South Dakota
34.3
North Dakota
13.2
Northern Plains
32.7
Corn Belt
27.7
Lake States
23.4
Southeast
50.7
Delta
7.8
Southern Plains
6.5
Source: No-Till Farmer. March 1979. The percentages are higher than
those in Table 5-12 because we used the No-Till Farmer esti-
mates of total cropland to calculate the percentages. These
estimates are lower than those of the USDA, used to calculate
the percentages in Table 5-12.
Table 5-14 shows that the regional distribution of conservation tillage
in 1979 was generally consistent with these expectations.
It appears that the available data support a positive answer to the
first question posed above: the present regional distribution of conser-
vation tillage is consistent with what the literature tells us about the
factors affecting yields of conservation tillage relative to yields of con-
ventional tillage. Given that, what can we say in answer to the second
questions—the role of economics in the future spread of conservation till-
age?
With respect to costs, the competitive position of conservation tillage
relative to conventional tillage probably will be strengthened somewhat if
energy prices rise over the long-term, as is now generally expected. How-
ever the overall impact probably will not be large. Differences in fuel
costs are a small part of the total cost differences between the.two tech-
nologies. Moreover, an increase in fuel costs is likely to trigger an in-
crease also in pesticide costs since most pesticides are petroleum based.
This would weaken the competitive position of conservation tillage. In the
discussion of costs we assumed no difference between the two technologies
with respect to fertilizer requirements, but if in fact conservation till-
age requires more fertilizer, particularly nitrogen, to achieve comparable
yields, as much of the literature suggests, rising energy prices would
weaken the competitive position of conservation tillage on this score also.
89
-------
On balance, we do not expect changes in resource requirements and
costs to significantly alter the present competitive position of conserva-
tion tillage relative to conventional tillage over the long-term. The small
present advantage of conservation tillage is expected to remain, neither
widening nor shrinking by much. So far as the future spread of the tech-
nology depends upon economics, therefore, factors affecting comparative
yields are likely to be most important.
If the demand for U.S. crop production and trends in crop technology
behave as we have projected, it is likely that the erosion advantage of
conservation tillage will favor the technology more strongly in the future
than it probably has so far. The reason is that much of the additional land
that would be needed for crop production iri 2010 would be more erosive than
land now in production.21 The negative effects of erosion on the yields of
such land likely would be more pronounced than on present cropland and
therefore of more immediate concern to the farmer. The use of conservation
tillage on these marginal lands should appear attractive.
Comparison of the amount of land now in conservation tillage in the
Cornbelt with the amount with potential for the technology, judging from
soil characteristics, indicates substantial room for expansion (see Table
5-13). How much of the potential might be realized is impossible to say,
but the table raises some questions in this regard. The small amount of
land in conservation tillage in Ohio (8.0 percent when the potential is
47.5 percent) suggests either that one or both of the numbers is wrong or
that something other than soil characteristics is limiting the spread of
conservation tillage in Ohio. Judging from entries in the literature as
much research on conservation tillage has been done in Ohio as in any other
state and more than in most.^ It is hard to believe that the failure of
conservation tillage to be more widely adopted in Ohio is because farmers
do not know enough about its advantages. A more plausible alternative ex-
planation (if we accept the numbers) is that for some reason weed control
with herbicides is especially difficult in Ohio. That problems of weed
control may be the principal obstacle to further spread of conservation
tillage, not only in Ohio but elsewhere in the Cornbelt, is a view held by
a number of knowledgeable observers.^
Much of any land converted to crops would come from class I-IV land
now in pasture. The USDA's Natural Resources Inventory of 1977 shows that
63 percent of present pasture land in classes I-IV has an erosion hazard.
The comparable figure for present cropland is 50 percent.
22
G.B. Triplett and his colleagues at the Ohio Agricultural Research
and Development Center at Wooster have been researching conservation till-
age for years and have published their results widely.
23
Griffith, Mannering and Moldenhauer. Also Robert Pope and Joseph
Downs, University of Illinois, personal communication.
90
-------
We already have suggested that the problem of controlling weeds, par-
ticularly perennials, with herbicides is the main reason for the small a-
mounts of land in conservation tillage in the Mississippi Delta and the
Southern Plains. If weed control is also the principal limiting factor in
the Cornbelt then it would be fair to say that for the country as a whole
the future expansion of conservation tillage depends more on improved tech-
niques for controlling weeds, particularly perennials, than on any other
single development. What is needed are new chemicals that would be effec-
tive against these weeds as well as new techniques of herbicide applications.
Much research and development now is underway to meet both of these needs.
For example, promising results have been obtained in Ohio with post-
emergence herbicides for controlling both perennial and annual grasses
soybeans. It is believed that these materials will also prove effective
in dealing with perennials in a corn-soybean rotation. Techniques for using
ropewick devices for post-emergence applications of contact herbicides also
are being developed.
We expect these research and development efforts to receive additional
impetus because of the value of conservation tillage in containing the ero- .
sion threat posed by the expansion of agricultural production onto marginal
lands. Consequently, we anticipate that the present limits imposed by weed
problems to the spread of conservation tillage will be pushed back. We have
no solid base for judging how far this may go, but we believe that economic
factors could induce farmers to adopt conservation tillage on 50 to 60 per-
cent of the nation's cropland by 2010. (In 1979 16 to 23 percent of crop-
land was in conservation tillage, depending upon the data sources—see Table
5-11.)
Future Trends in Herbicide Use—
Table 5-15 indicates how much herbicide would be used on main crops in
1985 and 2010, given our projections of crop acreage, if the percentages of
acres receiving herbicides and the amounts applied per acre are the same as
in 1976. If conservation tillage spreads as we expect and substitutes for
herbicides in conservation tillage systems are not.developed, per acre ap-
plications of herbicides for corn, soybeans and wheat will rise from 1976
levels. The continued spread of conservation tillage based on currently
available herbicides, therefore, implies larger applications of herbicides
in 1985 and 2010 than shown in Table 5-15. We have no basis for estimating
the additional amount, but it could be substantial, particularly if the per-
centage of wheat land receiving herbicides is increased. (Most corn, soy-
bean and cotton land already is treated with herbicides—see Table 5-6.)
The environmental implications of the trends in herbicide use, in use of
insecticides, and of our projections of crop production and land use are
discussed in the following section.
91
-------
V /
TABLE 5-15. HERBICIDES APPLIED TO CROPS
(MILLIONS LBS. ACTIVE INGREDIENTS)
1976
1985
Percent
increase
from 1976
2010
Percent
increase
from 1976
Corn
207.1
233.0
12.5
269.3
30.0
Cotton
18.3
23.8
30.1
25.5
39.3
Wheat
21.9
23.1
5.5
28.0
27.9
Soybeans
81.1
115.9
42.9
179.1
121.0
Total of above
328.4
395.8
20.5
501.9
52.8
Total of all crops
394.3
Sources: 1976 from Eichers, Andrilenas and Anderson. 1985 and 2010
assume that the percentages of each crop receiving herbicides and the
amounts applied per acre are the same as in 1976, and total acreage in
each crop is as we have projected. See text discussion for qualification
of these assumptions.
92
-------
REFERENCES
Aldrich, S. 1980. Nitrogen in Relation to Food, Environment and Energy.
Agricultural Experiment Station, University of Illinois, Urbana-
Chainpaign.
Amemiya, M. 1977. "Conservation Tillage in the Western Corn Belt," Conser-
vation Tillage: Problems and Potentials. Special publication no. 20,
Soil Conservation Society of America (Ankeny, Iowa).
Andrilenas, P. 1974. Farmers' Use of Pesticides in 1971 - Quantities.
Ag. Econ. Report no. 252, ERS-USDA (Washington, D.C.).
Cosper, Harold R. 1979. "Soil Taxonomy as a Guide to Economic Feasibility
of Soil Tillage Systems in Reducing Nonpoint Pollution." Natural
Resources Economic Division, ESCS, USDA (Washington, B.C.).
Crosson, Pierre. 198L Conservation Tillage and Conventional Tillage: A
Comparative Assessment. The Soil Conservation Society of America.
Ankeny, Iowa.
Davidson, J.M. and P.W. Santelmann. 1973. An Evaluation of Various Tillage
Systems for Wheat. Bulletin B-711, Agricultural Experiment Station
(Stillwater, Oklahoma, Oklahoma State University).
Doster, D.H. and J.A. Phillips. 1973. "Costs, Inputs and Returns: Humid
and Subhumid Regions," Conservation Tillage, the Proceedings of a
National Conference. The Soil Conservation Society of America (Ankeny,
Iowa).
Douglas, John. 1978. Remarks given at the Third World Fertilizer Confer-
ence (San Francisco).
Eichers, T.R. and P.A. Andrilenas and T.W. Anderson. 1978. Farmers' Use
of Pesticides in 1976. USDA-ESCS, Agricultural Economic Report no.
418 (Washington, D.C.).
Erbach, D.C. and W.G. Lovely. 1975. "Effect of Plant Residue on Herbicide
Performance in No-Tillage Corn," Weed Science vol. 23, no. 6, pp.
512-515.
Fenster, C.R. and T.M. McCalla. 1970. "Tillage Practices in Western
Nebraska with a Wheat-Fallow Rotation." College of Agriculture and
Home Economics, University of Nebraska, Agricultural Experiment
Station SB 507.
93
-------
Fertilizer Institute. 1976. The Fertilizer Handbook. Washington, D.C.
Griffith, D.R., J.V. Mannering and W.C. Moldenhauer. 1977. "Conservation
Tillage in the Eastern Cornbelt," Conservation Tillage: Problems and
Potential. Special publication no. 20, Soil Conservation Society of
America (Ankeny, Iowa).
Harris, B.L., E. Burnett and C.L. Williams. 1979. "Potentials for Conser-
vation Tillage Systems in Texas," in B.L. Harris and A.E. Colburn (eds.)
Conservation Tillage in Texas. Texas Agricultural Extension Service
(College Station, Texas A&M University System).
Harrold, L.L., G.B. Triplett, Jr. and W.M. Edwards. March, 1970. "No-
Tillage Corn: Characteristics of the System," Agricultural Engineering,
pp. 123-131.
Heady, Earl. "The Adequacy of Agricultural Land: A Demand-Supply Perspec-
tive." In P. Crosson (ed.), The Adequacy of Agricultural Land in the
United States, forthcoming from the Johns Hopkins University Press for
Resources for the Future.
Headley, J.C. 1978. "Pest Control as a Production Constraint for Grain,
Crops and Soybeans in the U.S. to 1990." Unpublished research sup-
ported by Resources for the Future and the Missouri Agricultural Ex-
periment Station. Headley is in the Department of Agricultural Eco-
nomics, University of Missouri, Columbia.
Hemphill, Gregory. 1980. "Fertilizer Use: Alternative Management Practices
and Nutrient Materials." Resources for the Future, unpublished memor-
andum .
Lewis, W.M. 1972. No-Tillage Production Systems for Double Cropping and
for Cotton and Other Crops (Raleigh, North Carolina.State University).
Lockeretz, W., G. Shearer and D.H. Kohl. 1981. "Organic Farming in the
Cornbelt", Science, Vol. 211 (Feb.), pp. 540-547.
McWhorter, C.G. and T.N. Jordan (no date). "Limited Tillage in Cotton
Production," forthcoming in Weed Control in Limited Tillage Systems,
the Weed Science Society of America.
Miranowski, J. 1979. "Integrated Pest Management in Corn Rootworm Control:
A Preliminary Technology Assessment." Paper given at the annual meet-
ing of the American Agricultural Economics Association, Pullman,
Washington, July 29-August 1, 1979.
National Academy of Sciences. 1975. Pest Control: An Assessment of Pre-
sent and Alternative Technologies. Vol. Ill', Cotton Pest Control
(Washington, D.C.).
No-Till Farmer. March, 1979, Renman Publications (Milwaukee, Wisconsin).
94
-------
Office of Technology Assessment. 1980. Pest Management Strategies, Vol. II
Working Papers, Government Printing Office (Washington, D.C.).
Phillips, R.E., R.L. Blevins, G.W. Thomas, W.W. Frye, and S.H. Phillips.
June 6, 1980. "No-Tillage Agriculture," Science, vol. 208.
Piraental, D., C.A. Shoemaker, E.L. LaDue, R.B. Rovinsky, and N.P. Russell.
1977. Alternatives for Reducing Insecticides on Cotton and Corn;
Economic and Environmental Impact. Report to the Environmental Pro-
tection Agency (Washington, D.C.).
Randall, G.W. and J.B. Swan. 1975. "Corn-Soybean Tillage," report avail-
able through the Agricultural Extension Service (St. Paul, University
of Minnesota).
Reicosky, D.C., D.K. Cassel, R.L. Blevin, W.R. Gill and G.C. Naderman.
1977. "Conservation Tillage in the Southeast," Conservation Tillage:
Problems and Potentials. Special Publication No. 20, Soil Conservation
Society of America (Ankeny, Iowa).
Reichert, Bonnie. 1979. No-Till Cultivation—Better Yields From Less Ex-
pense? Georgia Agricultural Experiment Station No. 20 (3) Winter.
Siemens, J.C. and W.R. Oschwald. 1978. "Corn-Soybean Tillage Systems:
Erosion Control, Effects on Crop Production, Costs," Transactions of
the American Society of Agricultural Engineers, vol. 21, no. 2.
Stanford, G. 1978. "Prospects for Using Nitrogen More Effectively in Crop
Production." Paper given at the 1978 meeting of the Northeastern
Branch, American Society of Agronomy, University of Connecticut, Storrs.
Taylor, C.F., D.R. Reneau and R. Trimble. 1979. "Economics of Conserva-
tion Tillage Systems," in B.L. Harris and A.E. Colburn (eds.), Conser-
vation Tillage in Texas. Texas Agricultural Extension Service (College
Station, Texas A&M University System).
Unger, P.W., Allen F. Wiese and Ronald R. Allen. 1977. "Conservation Till-
age in the Southern Plains," Conservation Tillage: Problems and Poten-
tial . Special Publication no. 20 of the Soil Conservation Society of
America (Ankeny, Iowa).
USDA. 1971. Cropping Practices: Corn, Cotton, Soybeans, Wheat 1964-70.
Statistical Reporting Service, SRS-17 (Washington, D.C.).
. Jan. 1977. 1977 Fertilizer Situation, Economic Research Service,
FS-5 (Washington, D.C.)
. 1979. Agricultural Statistics. Government Printing Office (Washing-
ton, D.C.)
. 1980. December 1980. .1981 Fertilizer Situation. Economics, Statis-.
tics and Cooperatives Service, FS-10 (Washington, D.C.)
95
-------
USDA. January 15, 1980. Crop Production 1979 Annual Summary. CrPr 2-1
(80). Crop Reporting Board (Washington, D.C.)
. July, 1980. Report and Recommendations on Organic Farming. Prepared
by a USDA study team on organic farming (Washington, D.C.)
. Feb., 1981. Economic Indicators of the Farm Sector: Production and
Efficiency Statistics, 1979. Economics and Statistics Service, Stat.
Bull. No. 657.
University of Illinois. 1979. Tillage Systems for Illinois. College of
Agriculture Cooperative Extension Service Circular 1172 (Urbana-Cham-
paign, University of Illinois) .
von Rumker, R., E.W. Lawless and A.F. Meines with K.A. Lawrence, G.L. Kelso
and F. Horay. 1975. Production, Distribution, Use and Environmental
Impact Potential of Selected Pesticides. For the Environmental Protec-
tion Agency Office of Pesticide Programs and the Council on Environ-
mental Quality (Washington, D.C.) .
Walker, D.J. 1977. "An Economic Analysis of Alternative Environmental and
Resource Policies for Controlling Soil Loss and Sedimentation from
Agriculture," Ph.D. Thesis (Iowa State University, Ames, Iowa).
Weaver, D.W. 1979. "Weed Control in Conservation Tillage Systems," in B.L.
Harris and A.E. Colburn (eds.), Conservation Tillage in Texas, Texas
Agricultural Extension Service (College Station, Texas A&M University
System).
Wiese, A.F. and D.W. Staniforth. 1973. "Weed Control in Conservation Till-
age," Conservation Tillage, Proceedings of a National Conference. Soil
Conservation Society of American (Ankeny, Iowa).
Willis, W.O. and M. Amemiya. 1973. "Tillage Management Principles: Soil
Temperature Effects," Conservation Tillage, Proceedings of a National
Conference. Soil Conservation Society of America (Ankeny, Iowa).
96
-------
SECTION 6
ENVIRONMENTAL IMPACTS OF PROJECTED PRODUCTION
AND RESOURCE USE
INTRODUCTION
The environmental impacts considered are those resulting from increased
use of fertilizers and pesticides, from the expansion of irrigation, and from
erosion. Comprehensive and generally acceptable quantitative estimates of
these impacts are not available, and perhaps not possible, even theoretically.
One reason is lack of basic knowledge, e.g., the contribution of phosphorus
fertilizer to eutrophication of water bodies compared to that of municipal
wastes; or the fate of pesticides after they are applied to farmers' fields.
A second major reason is the conceptual difficulty of valuing some of these
impacts, e.g., human illnesses or deaths from pesticides.
For these reasons our assessment of environmental impacts makes no
attempt at quantification in the sense of attempting to assign monetary
values to the impacts. Instead we seek to make judgments on two questions:
(1) are the impacts presently of such a nature as to require new policies or
modification of existing policies to deal adequately with them, (2) given our
projected increases in resource use, are the future impacts likely to be more
or less severe than at present? These judgments set the stage for the dis-
cussion in the following section of policy issues and options.
FERTILIZER
Nature of the Environmental Impacts
The principal concerns about fertilizers in the environment have to do
with effects of nitrogen in water on human and animal health and of nitrogen
and phosphorus in accelerating eutrophication of water bodies by stimulating
growth of aquatic plants.1
^"Denitrification of nitrate-N releases nitrogen oxides, one of which,
nitrous oxide (N2O), may attack the earth's ozone shield. The resulting in-
crease in solar radiation reaching the earth's surface would increase the
risk of skin cancer. This possibility received considerable popular atten-
tion in the mid-1970s, particularly after press reports of work done by
Michael McElroy at Harvard. According to Aldrich (1980, pp. 222-223) a sub-
97
-------
Nitrates in water may constitute a hazard to both humans and animals
that drink the water, not because the nitrates themselves are particularly
toxic but because of their potential for conversion to nitrites which are
toxic, or can be in sufficient concentration. Babies are more likely to
suffer from nitrite poisoning than adults. Deaths and illnesses among farm
animals also have been attributed to nitrite intoxication (Luhrs). Nitrogen
gas dissolved in water or nitrogen in ammonia may kill or injure fish
(Patrick).
Eutrophication is a process involving the nutrient enrichment of lakes
and reservoirs, the resultant growth of plant life, and the subsequent de-
cline in the water's dissolved oxygen supply because of the oxygen demand of
decaying plants. If severe, eutrophication makes the water incapable of
supporting fish and spoils its value for some kinds of recreation. Eutrophi-
cation is a natural process, but it can be accelerated by the addition of
nutrients supplied by man, including nitrogen and phosphorous carried by
run-off water and soil eroded from farmland.^ The amounts of these nutrients
may be very small yet have significant effects in stimulating growth of
aquatic plants. According to Holt, Johnson, and McDowell as little as 10
parts per billion (ppb) of phosphorus and 100 ppb of nitrogen are enough to
support growth of undesirable amounts of aquatic plants.
Present Severity of the Problems
Some years ago one of the authors assessed the severity of the nitrate
problem in the United States and agriculture's contribution to it in the
following words (Crosson and Frederick, pp. 192-195):
The evidence about the present seriousness of the environmental
hazards of fertilizers is not easy to interpret. Numerous studies
show the presence of nitrates in groundwater, for example, in parts
of Connecticut, Long Island, Illinois, the High Plains, Minnesota,
and California. (Martin, Fenster and Hanson; Miller, DeLuca, and
Tessier). In some of these places the nitrate-N level has exceeded
ten parts per million, the standard for drinking water set by the
Public Health Service. It is not clear, however, that fertilizer
used in agriculture is the principal source of tne nitrate in these
waters. Martin and coauthors cite a study by Smith of 6,000 rural
water supplies in Missouri which concluded that animal wastes and
(footnote continued from page 6-1)
sequent study by the National Academy of Sciences indicated a lag of about
100 years between the application of nitrogen fertilizer and effect upon the
ozone shield. Moreover, even a very large increase worldwide in nitrogen
fertilizer use would reduce the ozone shield only 1.5 to 3.5 percent by 2100.
Perhaps for these reasons the issue has faded from public discussion. We
do not consider it further in this report.
2
Municipal and industrial wastes also are major sources of nutrients
which contribute-importantly to eutrophication.
98
-------
septic tank drainage were the principal sources of nitrates.
Studies in Minnesota show that rural "wells were contaminated by
nitrate long before fertilizer use became important, and that
between 1899 and 1963 change in the quality of water in major
aquifers was minor. A study of an area near San Luis Obispo,
California, showed that nitrate in groundwater "...was substantial
and mostly associated with the native soil organic matter complex
supplemented with sewage waste from area homes and to a lesser
extent from lawn and farm fertilization." (Miller, DeLuca and
Tessier, P. 309) In their study of groundwater pollution in
the northeastern states Miller and coauthors assigned a "moderate"
priority to fertilizer for additional research and control measures
while giving high priority to domestic, municipal, and most other
sources of wastes.
The extent of fertilizer pollution of surface waters in the
United States also is unclear, although the problem seems to have
been less investigated. In a study of the southern High Plains
Goldberg concluded that "nitrogen fertilizer applied to the farm-
land adds little nitrate to the surface water." (Goldberg)
Aldrich says that the nitrate content of rivers in the central
Corn Belt for which long-term records are available shows an
upward trend. The nitrate content of the Mississippi River at the
point where it leaves the Corn Belt about doubled from the mid-1950s
to the mid-1970s but was still less than the maximum acceptable
amount set by the U.S. Public Health Service for drinking water.
The nitrate content of the Illinois River increased about one-
quarter in this period but still was only about one-half of the
Public Health Service's standard, while the nitrate in the
Missouri declined about 50 percent between 1950 and 1970. In
some small creeks and rivers in Illinois, however, the nitrate
content sometimes exceeds the standard. (Aldrich, 1976)
The National Academy of Sciences cites a report on fertilizer
pollution in Illinois by the Illinois Pollution Control Board (IPCB)
which concluded that "there is no factual basis for imposing re-
strictions on the use of fertilizer at this time." The conclusion
applied both to nitrogen and phosphate fertilizers....The EPA sub-
sequently issued a policy statement in general agreement with the
decision of the IPCB regarding fertilizer, but it also stressed
the need for additional information. (National Academy of Sciences,
1975b, pp. 92-93)
In our reading of the evidence on the environmental threats of
fertilizer, two important aspect stand out. One is that human
deaths or illnesses attributable to nitrite poisoning are extremely
rare. Animal poisonings are more common, but where caused by
drinking water the nitrate content of the water was seven to fifteen
times the standard set by the U.S. Public Health Service. (Aldrich,
1976) Enforcement of the standard should eliminate such instances.
Fish kills or injuries attributable to forms of nitrogen in surface
water also are of small overall importance.
99
-------
The second aspect is that much of the nitrogen and phosphorus
found in water bodies comes from municipal and industrial discharges
and from the natural leaching of these nutrients already in the soil.
Consideration of these two aspects of the evidence leads us
to conclude that with present levels of use fertilizers do not, pose
threats to the environment so severe that changes in present poli-
cies are required to deal with them.
We believe this assessment of the nitrate problem is still valid. After
a thorough review of the data on nitrates in surface waters, Aldrich (1980)
concludes:
The evidence is convincing that, based upon available data,
the nitrate issue in surface waters in the United States in rela-
tion to health is limited mainly to the midwest and to certain
rivers in intensively farmed regions of California, (p. 141)
Aldrich's review shows that even in the midwest nitrate concentrations in
rivers only occasionally exceed the 10 ppm nitrate-N concentration standard
set by the U.S. Public Health Service.
Nor have we found anything in the more recent literature to alter our
judgment about the contribution of nitrogen fertilizers to contamination of
groundwater. The Council on Environmental Quality (CEQ, 1979) has asserted
that surface and subsurface disposal of wastes are among the major sources
of groundwater contamination in the country. Disposal of industrial wastes
in this manner is mentioned in particular. In this context the CEQ (1979,
pp. 110-111) notes that infiltration of nitrates from nitrogen fertilizers
is another important source of groundwater contamination. However, earlier
in the same report (pp. 107-109) the CEQ presents a table summarizing ground-
water quality conditions in eighteen river basins making up the forty-eight
contiguous states which mentions fertilizer contamination in only two regions,
the Souris-Red-Rainey (northern Minnesota and North Dakota) and the Pacific
Northwest. Irrigation return flow is mentioned as a contaminant of ground-
water in the Great Basin, and this presumably includes nitrates from ferti-
lizer, although the CEQ report does not say so explicitly. The table conveys
the strong impression, however, that with the exception of local "hotspots"
in the regions mentioned, contamination of groundwater by nitrogen fertilizer
is not presently a major problem.
This interpretation is supported by a more recent publication by the CEQ
(Jan. 1981) dealing with contamination of groundwater. An EPA report (1977b)
is cited as identifying the disposal of industrial wastes at industrial im-
poundments and solid waste disposal sites as the most important source of
groundwater contamination. Septic tanks, municipal waste water, mining and
petroleum exploration and production are listed as of secondary importance.
Agriculture is not mentioned as of even secondary importance as a source of
groundwater pollution.
A study of the Santa Maria valley in California in 1976 showed that 39
percent of the nitrogen fertilizer applied to 57,000 acres of vegetable,
100
-------
field and fruit crops was leached below the root zone, but there is no indi-
cation that nitrate concentration in groundwater in the area exceeded 10 ppm
as a consequence (Lund et al.). A study of the Upper Santa Ana River basin
of California in the early 1970s revealed several "hotspots" where nitrate
concentrations were high (Ayers). In the dairy farm area of the Chino Basin,
a small part of the whole basin, nitrate-N concentrations at the top of the
water table underlying cropland sites averaged 45 ppm. Deep well water-in
the same vicinity averaged 6 ppm. Ayers infers from these numbers that the
dairy wastes had not yet had full impact on nitrate-N concentrations in the
groundwater. He evidently believes that animals rather than fertilizers are
the main contributors to the build-up of nitrate-N in the groundwater.
A problem area that has emerged since our previous review of the litera-
ture on fertilizer pollution (or which our review missed) is the Sandhills
region of Nebraska. In irrigated parts of that region nitrate-N concentra-
tions in groundwater have risen to 20 ppm. With good irrigation management
concentrations should not rise above 20 to 25 ppm, but with continued expan-
sion of irrigation in the area they are not likely to fall below those
levels.^
Eutrophication—
The Environmental Protection Agency conducted a nationwide eutrophica-
tion survey between 1972 and 1977 which found that two-thirds of the 800
lakes surveyed were eutrophic and another 4 percent were hypereutrophic.^ On
its face this would seem to suggest that eutrophication is a major national
problem. It is not so treated in the cited CEQ report, however, nor in recent
annual reports to Congress by the EPA on the condition of national water
quality. This is not to say that the CEQ and the EPA reports dismiss eutro-
phication as a national problem. They do not. But the reports convey no
sense that the problem is of "the magnitude suggested by the finding of the
EPA's eutrophication study that over 70 percent of the lakes surveyed were
eutrophic or hypereutrophic. The reason, perhaps, is that eutrophication is
a matter of degree, and a lake classed as eutrophic under the EPA definition
still can be fishable and swimmable and provide other recreational values.
Whatever the seriousness of the eutrophication problem, fertilizers
apparently contribute less to it in most instances than other sources of
nutrients. Phosphorous is usually the nutrient limiting plant growth in
lakes and reservoirs, and municipal, industrial, and other non-agricultura]
3
Darryl Watts, University of Nebraska, cited in Frederick.
^Council on Environmental Quality (1979, p. 92). According to the CEQ
a eutrophic lake usually has murky, greenish water and measurable amounts
of plant productivity, and a hypereutrophic lake has very murky water and
extremely high levels of plant biomass.
The EPA survey over-represented lakes receiving discharges from municipal
sewage treatment plants. Such lakes likely would be more eutrophic than
lakes not receiving such discharges (CEQ, 1979, p. 92, footnote 16).
101
-------
wastes typically contribute more phosphorus (and other nutrients) to these
water bodies than fertilizers applied in agriculture (CEQ, 1979, p. 95). In
discussing eutrophication in Lakes Erie and Ontario, the most seriously af-
fected of the five Great Lakes, the CEQ report puts high loads of nutrients
from municipal sewage ahead of those from agricultural and urban runoff as
the principal cause. According to the EPA (1977a, p. 9), control of phos-
phorous discharges from municipal waste treatment plants is expected to have
a major effect in reducing eutrophication in heavily populated regions.
Future Severity of the Problems
Our projections in the previous section show increases in total nitrogen
applied of 40 percent from 1977/79 to 1985 and of 87 percent from 1977/79 to
2010. For phosphorus, the increases are 30 percent and 68 percent respec-
tively (see Table 5-3)
The projected increases of fertilizers applied to the four main crops
(corn, wheat, soybeans, and cotton) from 1977/79 to 2010 are 60 percent for
nitrogen and 70 percent for phosphorus. Seventy to 75 percent of the in-
crease in nitrogen applied to these crops and 80 percent of the increase in
phosphorus is attributable to the increased acreage in corn, wheat, and
soybeans.6 The percentage increases for nitrogen are less for the four main
crops than for total applications because "all other" uses of nitrogen are
projected to increase faster than uses on the four main crops. "All other"
uses of fertilizer includes that put on crops other than the.main four, as
well as that used on golf courses, parks, home gardens, and lawns, and so on.
In Section 4 we assumed that land in crops other than main crops would
be the same in 1985 as in 1977, and that in 2010 land in these crops would be
16 million acres less than in 1977, reflecting a 10 million acre decline in '
land in hay and a 6 million acre decline in land in sorghum, oats, and bar-
le:y. We have not projected fertilizer applications per acre of land in
"othe.r crops," but we do not expect much increase, for the same reasons that
little increase is expected in per-acre applications to main crops. It
follows that most of the projected increase in "all other" uses of fertilizer
is for non-crop purposes. Since we are interested in this study in the en-
vironmental impacts of crop production, we give no further attention to "all
other" uses of fertilizer. We note, however, that should concern about
environmental impacts of fertilizers mount in the future the "all other"
uses likely will require special attention. .
The environmental impacts of the projected increases in fertilizers
used on the four main crops depend on how much of the additional nutrients
move to ground and surface water and on existing nutrient concentrations in
the water. For the moment we focus on movement of the additional nutrients.
^There is little concern about the environmental impacts of potassium
fertilizers, 'and we do not consider them here.
*>We project a small decline in fertilizer applied to cotton by 2010, sc>
we do not consider use on cotton in this discussion (see Table 5-3 in Section 5).
102
-------
Movement of Nutrients—
For several reasons we expect the increase in movement of nutrients to
water bodies to be proportionally less than the increase in amounts of ferti-
lizers applied. As noted above, 70 to 80 percent of the increase in amounts
applied of both nitrogen and phosphorus from 1977/79 to 2010 is because of
the expansion of land in com, wheat, and soybeans. About 16 million acres
of the additional land is now in hay (10 million acres), sorghum, oats, and
barley (6 million acres). We have no completely reliable information on how
much fertilizer currently is applied to these 16 million acres, but several
sources suggest a conservative range of 50 to 75 pounds per acre of both ni-
trogen and phosphorus (International Minerals and Chemical Corporation). If
we assume 60 pounds, then the 16 million acres currently receive 960 million
pounds of nitrogen and phosphorus per year. Consequently the net increases
by 2010 in the amounts of these materials applied to corn, wheat, and soybeans
would be 960 million pounds less than the gross amounts we have projected.
In percentage terms, the net projected increase for nitrogen would be 53 per-
cent compared with a gross increase of 66 percent. For phosphorus the net
increase would be 51 percent compared with a gross increase of 75 percent.
The increase in movement of nutrients may not be proportional to the net
increases in amounts applied also because fertilizer technologies are likely
to become more efficient in response to higher prices, as discussed in the
previous section, and because of the expected increase in conservation till-
age relative to conventional tillage.
Effects of Increased Efficiency—The anticipated adoption of fertilizer
materials and practices which reduce nutrient losses would tend to hold the .
increase in movement of nutrients'below the increases in total amounts
applied. We have no solid basis, however, for estimating how much losses
might be reduced. Stanford cites a study in Nebraska indicating that losses
of sidedressed nitrogen were 30 percent less than losses of N applied in the
fall or spring. Stanford goes on to assert that an "appreciable" increase
in nitrogen efficiency can be achieved by more realistic recommendations of
application rates and proper timing of application, but he gives no numbers.
Nitrogen fertilizer losses vary widely from place to place, depending
upon soil moisture and temperature and other factors. According to Hinish,
losses can range from 20 to 60 percent. For illustrative purposes we assume
average nitrogen losses of 40 percent and that improvements in practices and
materials reduce losses by 20 percent between 1977/79 and 2010. With these
assumptions the projected 53 percent net increase in nitrogen applied to
corn, wheat, and soybeans would result in a little more than 20 percent in-
crease in nitrogen losses.^
^For example:
1977/79 2010
Amount of nitrogen applied 100 153
Percentage lost 40 32
Amount lost 40 49
The amount lost would increase a little more than 20 percent.
103
-------
Aldrich (1980, p. 119) expects use of nitrogen fertilizer in the Cornbelt
to increase. He argues, however, that higher prices, among other factors,
will induce farmers to use fertilizer more efficiently. Indeed, he asserts
that the increase in amount of nitrogen applied "...will closely follow the
rise in efficient utilization by crops."
Losses of phosphorous fertilizers range from 60 to 80 percent (Hinish).
Assuming an average of 70 percent and that losses fall by 20 percent by 2010,
the projected 51 percent net increase in phosphorus applied would result in
about a 20 percent increasfe in losses.
Effects of Tillage Technologies—Losses of fertilizers applied to land
in the three main crops in 2010 also will be influenced by the kind of til-
lage technologies used on this land. We concluded in the previous section
that the economics of conservation tillage would favor its spread to 50 to
60 percent of the nation's cropland by 2010 (increasing its share two to
three times). Most conservation tillage is on corn, soybean, and wheat land,
so we anticipate conservation tillage of land in these crops to increase
significantly by 2010.
The effects of the spread of conservation tillage on movement of
nutrients to water bodies is uncertain. It will be shown below that con-
servation tillage greatly reduces erosion compared to conventional tillage,
and a number of studies have demonstrated that most of the nitrogen and
phosphorus moved from farmers' fields to surface water bodies is carried
by eroded soil.® The nitrogen so moved is in the organic form and when
deposited with sediment in rivers, lakes, and reservoirs, is mineralized
to the nitrate form only very slowly. Consequently the advantage of con-
servation tillage in reducing erosion, and hence losses of organic nitrogen,
does not significantly reduce the' impact of nitrogen on water quality.
Wauchope, McDowell, and Hagen cite a number of studies indicating that,
unlike the nitrogen carried by eroded soil, 5 to AO percent of the phosphorus
so carried is available to support aquatic plant growth. However, the con-
centration of fertilizers in the top soil layer is higher with conservation
tillage than with conventional tillage, so nutrient concentrations in sedi-
ment from conservation tillage fields also are higher. Consequently, while
the reduction in erosion with conservation tillage tends to reduce losses of
available phosphorus, the higher concentration in eroded soil tends to
increase them. Studies by Barisas et al. and Johnson et al. showed, however,
that in the usual case the reduction in losses of available P because of
less erosion more than offset the increase in losses because of higher con-
centration in the eroded soil, so the net effect was smaller losses.
According to Wauchope, McDowell, and Hagen there is considerable evi-
dence that concentrations of nitrate nitrogen, ammonia, and available
phosphorus are higher in runoff water from conservation-tilled fields than
from conventionally-tilled fields. Barisas et al. found this also. Johnson
Barisas et al. and studies cited therein.
104
-------
et al. reported no significant differences among tillage practices with re-
spect to concentrations of nitrate nitrogen and ammonia in runoff water, but
concentrations of soluble P were higher with conservation tillage. Concen-
trations in runoff water are higher both because nutrient concentrations are
higher in the top soil layer and because of leaching of nutrients from the
crop residues found with conservation tillage.
The amount of runoff water typically is less with conservation tillage
because the surface residue and rougher soil texture characteristic of the
technology reduce the velocity of water flow, permitting more water to remain
on the field. Because the concentration of nutrients in runoff typically is
higher with conservation tillage, the reduction in runoff water does not
necessarily reduce the concentration of nutrients in receiving waters. Total
nutrients in runoff water conceivably could be greater with conservation
tillage, in which case the technology clearly increases the nutrient content
of the receiving waters compared to conventional tillage. Even if total
nutrient loss in runoff water is less with conservation tillage, the effect
on nutrient concentrations in receiving waters is uncertain. If, compared
to conventional tillage, conservation tillage reduces the amount of nutrient
delivered proportionally less than the amount of water delivered, conserva-
tion tillage will increase the nutrient content of the receiving waters. It
will decrease the nutrient content of the receiving waters if it reduces the
amount of nutrient delivered proportionally more than it reduces the amount
of water delivered. Both these conditions are consistent with higher con-
centrations of nutrients in runoff water with conservation tillage.
The reduction in runoff water with conservation tillage may result in
more infiltration of nitrate-N to graoundwater than with conventional tillage.
Unlike nitrate-N phosphorus is strongly adsorbed by soil and so does not
leach to groundwater in significant amounts.
The discussion suggests that conservation tillage may pose a greater
threat to groundwater quality than conventional tillage because of increased
leaching of nitrate-N, but that the comparative effects on delivery of ni-
trate-N to surface waters is too dependent on specific local conditions to
warrant a general conclusion.
The effect of the spread of conservation tillage on the threat of phos-
phorus to water quality also is not clear. The outcome depends upon the
reduction in available phosphorus carried by eroded soil relative to the
increase carried by runoff water. In general, the advantage of conservation
tillage with respect to the total amount of available phosphorus delivered
to water bodies will be greater (1) the greater the reduction in erosion:
(2) the smaller the difference between conservation and- conventional tillage
in concentration of total P in sediment; (3) the greater the ratio of avail-
able P to total P in sediment; (4) the higher the sediment delivery ratio;
(5) the smaller the difference between the two tillage systems in concentra-
tion of phosphorus in runoff water. While erosion typically will be sub-
stantially less with conservation tillage and the concentration of available
P in both sediment and runoff water higher, the rangeof difference in these
and the other variables is large. Consequently, no general conclusion about
105
-------
the effect of the spread of conservation tillage on the delivery of phosphorus
to water bodies seems warranted.
While we cannot say how the shift to conservation tillage would affect
delivery of phosphorus and nitrate-N to surface waters, it seems reasonably
well established that the shift would increase movement of nitrates to
groundwater. Thus the shift probably would offset to some extent the effect
of more efficient materials and practices in reducing nitrate losses. We
lack, data to precisely estimate the net outcome, but some plausible numbers
suggest that the increase in nitrate losses could be about 30 percent. This
would be the case if conservation tillage of land in corn, wheat and soybeans
increased from 30 percent in 1977/79 to 60 percent in 2010, if nitrate losses
were 50 percent from conservation-tilled land and 40 percent from convention-
ally-tilled land, and if losses with both kinds of tillage were 20 percent
less in 2010 than in 1977/79. ^
)
We conclude that from 1977/79 to 2010 losses of nitrogen fertilizer from
land in corn, wheat, and soybeans will increase significantly less than the
projected 66 percent gross increase in amounts applied. The increase in
phosphorus"losses also is likely to be significantly less than the gross
projected increase of 75 percent, although this outcome is clouded by un-
certainty about the effect of the shift to conservation tillage on phosphorus
losses.
Whether these increases in nutrient losses would cause water quality
problems depends upon nutrient concentrations in the receiving waters. Since
existing concentrations of nitrate-nitrogen do not appear to pose serious
problems except, perhaps, in a few local "hot spots" around the country,
some increase in concentrations evidently could be accommodated without
serious consequence. Whether an increase on the order of 30 percent would
be cause for concern we are unable to say.
Our earlier discussion indicated that municipal wastes are a major
source of phosphorus in water and that significant reductions in this source
are expected as a result of increased investment in municipal waste treat-
ment around the country. This suggests that the increased losses of phos-
phorus from land in corn, wheat, and soybeans would not cause major increased1,
damage to water quality. Accordingly, we do not further consider phosphorus
losses. - .
^For example:
1977/79 2010
"Conser-
Conven-
Com-
Conser-
Conven-
Com-
vation
tional
bined
vation
tional
bined
tillage
tillage
amount
tillage
tillage
amount
Amount of nitrogen
applied
30
70
100
91.8
61.2
153
Loss percentage
50
40
43
40
32
38.4
Amount lost
15
28
43
36.7
19.6
56.3
Total losses would increase about 30 percent.
106
-------
The estimates so far presented of fertilizer applications and losses are
for the nation as a whole. They mask widely varying regional increases in
application and losses reflecting primarily regional differences in projec-
tions of land in the three main crops. The smallest relative increases in
nitrogen applied—all about 40 percent—are in the Lake States, Cornbelt, and
Northern Plains. The largest nitrogen increases are in the Mountain States
(280 percent), Southeast (195 percent), Southern Plains (180 percent), Paci-
fic (150 percent), and Delta (100 percent).
We expect that in both the Cornbelt and Northern Plains about 2 million
acres of land now in sorghum, oats, and barley will be converted to corn,
wheat, and soybeans by 2010. Consequently, the net increases in nitrogen
applied in these regions would be somewhat less than the projected 40 percent
gross increase. Allowing for more efficient nitrogen materials and practices
and for the spread of conservation tillage, nitrate losses in these regions,
and in the Lake States, might increase 20 to 25 percent by 2010. Increases
of this magnitude perhaps would present few problems in the Lake States and
Cornbelt, where nitrate concentrations in water evidently are not now ser-
ious. The situation might be different, however, in the Northern Plains.
As noted earlier, nitrate concentrations in groundwater in some parts of
Nebraska now are about 20 ppm, twice the Public Health Service's standard
for safe water. In such areas, the prospect of increased nitrate losses of
20 to 25 percent indicates cause for concern.
Our projections indicate large relative increases in nitrogen applied
in the South, Southwest, Mountain Region and Pacific Coast. Even allowing
for the spread of more efficient nitrogen materials and practices, the pro-
jections imply substantial increases in nitrate losses in these regions.
Without information about present nitrate concentrations in regional water
bodies, there is no basis for judging whether the increased losses would
pose water quality problems. In our judgment, however, the increases are
sufficiently large to indicate that such problems may emerge by 2010. The
Southeast in particular would bear watching. Much of the expansion in crop
production and fertilizer use in that region will be with irrigation on the
sandy soils of Georgia and northwest Florida. The potential for increased
leaching of nitrate-N to the groundwaters of the region looks large.
PESTICIDES
Nature of the Environmental Impacts
Pesticides are chemicals designed to kill various species of plants and
animals considered to be pests by farmers and others. Their presence in the
environment arouses concern because often they act not only against the tar-
get pests but also against other organisms. They may pose threats to human
health and reproductive capacity; do damage to non-target species of plants,
insects, soil and water microorganisms, and wildlife; and cause the build-up
in pests, especially insects, of genetic resistance to pesticides.
107
-------
Present Severity of the Problems
Pimentel et al. (1980) have estimated the total environmental costs of
pesticide use in the United States at $839 million annually. This includes
seven categories of environmental cost ranging from reduced natural enemies
and increased genetic resistance to pesticices ($287 million) to fish and
wildlife losses ($11 million). Human pesticide poisonings, including those
that result in fatalities, are included ($184 million).
Pimentel et al. state that they were unable to estimate some important
classes of environmental costs of pesticides. Had these been included, total
costs would have been "several times" higher than those reported.
Total direct and environmental costs of pesticides are estimated by
Pimentel et al. to be about $3.35 billion and total benefits to be $10.9
billion. Benefits exceed total costs by a factor of 3. If environmental
costs are in fact three times as high as estimated by Pimentel et al., ben-
efits exceed costs by a factor of 2.
Pimentel et al. are aware of the tenuous underpinnings of these esti-
mates—indeed they emphasize them. We present the estimates here because
to our knowledge they are the only ones available, and they are made by
reputable scientists. Our presentation of the estimates does not imply
endorsement of them. In fact ws are skeptical that reliable estimates of
environmental costs of pesticides are possible at present.
There are two reasons for our skepticism. One is the conceptual diffi-
culty of valuing some of these damages, e.g., human deaths or genetic
abnormalities, even in cases where the damages clearly are attributable to
pesticides. Pimentel et al. assign a value of $1 million per life to the
fatalities from pesticide poisoning, but they acknowledge that there is no
really satisfactory way of valuing these losses.
The second reason for our skepticism is that little is known about the
paths by which pesticides move through the environment and their ultimate
fate, making it extraordinarily difficult to identify, let alone assign
values to the damages they do. In the literature on environmental impacts
of pesticides this second difficulty receives far more attention than the
conceptual problem. For example, von Rumker et al. (p. 96) cite two compre-
hensive reports on environmental impacts of pesticides (Mrak, 1979; and
Pimentel, 1971) as indicating
...that relatively little information is available on the
toxicity and hazards of pesticides and their residues to non-
target organisms under field conditions. Information is espe-
cially sparse concerning the effect, if any, of pesticide residues
on lower aquatic and terrestrial organisms. There is a copious
literature on the effects of individual pesticides on isolated
organisms or systems in the laboratory or greenhouse over short
periods of time. However, such studies are usually far removed
from field conditions, and their results do not answer the ques-
tion of their significance in regard to field conditions.
108
-------
...Pesticide monitoring studies have centered on chlorinated
hydrocarbon insecticides. The possibility that pesticides may affect
terrestrial or aquatic ecosystems in other ways than through bio-
accumulation and biomagnification has received less attention. By
virtue of their physical, chemical, and biological properties, funr
gicides and herbicides are more likely to affect the lower trophic
levels of food chains. It is not known whether or not currently
practiced monitoring and observation methods would detect such
effects prior to the occurrence of massive ecological damage.
.In a more recent review of the literature dealing with the effects of
pesticides on water quality, Wauchope (1978) concludes, that, at least in
principle, reasonable estimates can be made of edge-of-field losses of pesti-
cides to rivers and lakes; however, information is most needed about the fate
of pesticides after they leave the field. Wauchope argues (p. 471) that
...overall assessments of runoff impact must include judg-
ments on such factors as the time and distance of impact of a
given field runoff event and the ability of local ecosystems to
recover from temporary high concentrations of a pesticide. The
dynamics of dilution and sediment exchange, and uptake, transfer,
and metabolism by aquatic life of most of the pesticides presently
in use are not known. Without this knowledge, the impact of a
given pesticide input or the quality of water in a river or lake
cannot be predicted.
A convincing assessment of the current severity of the environmental
impacts of pesticides thus is not possible. Consideration of some of the
skimpy evidence about these impacts nonetheless is appropriate. We deal with
insecticides and herbicides. These accounted for 86 percent of the pesti-
cides applied to crops in 1976 (Eichers, Andrilenas, and Anderson). Fungi-
cides, the principal other class of pesticide, are nontoxic or only slightly
toxic to mammals. Consequently, in concentrating on insecticides and herbi-
cides we miss little of significance concerning environmental impacts.
Insecticides—
The initial concern with insecticides was primarily with the effects
of the organochlorine compounds, principally DDT and similar persistent
materials, on wildlife and humans. While generally not highly toxic to
vertebrates, the tendency of these materials to persist in the environment
and to increasingly concentrate in body tissue at higher levels in the food
chain made them suspect. When tests on laboratory animals indicated that,
in high enough dosages, some of the organochlorine compounds were carcino-
genic and/or teratogenic,10 concern mounted still higher. Eventually, be-
ginning with DDT in 1972, the EPA banned or tightly restricted the use of
the organochlorine compounds, and by 1976 the only one still in general use
was toxaphene. However, because of their persistence traces of these
materials still are lurking about in the environment. To the extent that
^Teratogenic materials cause fetal malformations.
109
-------
they pose a threat, the threat remains, although diminishing with time. If
the threat includes cancer in humans—this is suspected but not conclusively
demonstrated—the ultimate signs of it may not disappear for decades because
of the long latency period of the disease.
We noted in the previous section that while the use of organochlorine
insecticides has declined sharply since the early 1970s, use of organophos-
phorous and carbamate compounds increased enough to more than offset the
decline. By comparison with the organochlorines these compounds are not
persistent nor do they bioaccumulate. Unlike the organochlorines, however,
many of the organophosphorous and carbamate compounds are highly toxic to
humans and other nontarget organisms. Consequently the nature of the threat
of these compounds is quite different from that of the organochlorines.
Damages inflicted by the latter typically are subtle, diffused widely, both
geographically and among affected individuals, and long-term. Damages of
the organophosphorous and carbamate compounds typically are sharp, localized,
and short-term. They in fact have many of the characteristics of industrial
accidents rather than characteristics we typically associate with environ-
mental impacts.
According to Pimentel et .al. (1980) there were fifty-two accidental
deaths from pesticide poisonings in 1974, a significant decline over the
preceding two decades. The number of intentional deaths (suicides and homi-
cides) was about three times the number of accidental deaths. The total
number of human poisonings- from pesticides was estimated by Pimentel et al.
to be about 45,000 per year. These were attributable to all pesticides, not
just organophosphorous and carbamate insecticides.
One of the principal concerns with insecticides is the tendency for in-
sects to build genetic resistance to these materials. When this happens
farmers find themselves increasingly unable to control the resistant insect
and crop losses mount. Increasing amounts of insecticide are applied, costs
rise, damages to beneficial insects and other nontarget organisms increase,
and resistance in the target insect is strengthened even more. Ultimately
the insecticide may become completely useless and the farmer must fall back
on a substitute, if one is available. Perhaps the outstanding example of the
build-up of genetic resistance was the boll weevil's response to DDT. By the
1960s the insect had become so resistant that the use of DDT had begun to
decline sharply some years before the EPA's action to ban it. The tobacco
budworm and cotton boll worm also became resistant to DDT and subsequently
to some of the more important organophosphorous compounds. Indeed, the in-
creasing resistance of these insects gave major stimulus to development of
nonchemical means of control of cotton insect pests in Texas, described in
the previous section.
Since the mid-1970s use of synthetic pyrethroids to control cotton
insect pests, particularly the tobacco budworm, has spread widely. The
pyrethroids have low persistence—indeed their rapid rate of degradation
may weaken their effectiveness—and low toxicity to mammals. They may be
highly toxic to fish, but they are used in such small amounts and degrade
so rapidly that the probability of a significant threat to aquatic life is
small. The increasing adoption of synthetic pyrethroids by cotton farmers,
110
-------
therefore, has tended to ease the total impact of insecticides on the environ-
ment, especially to the.extent that the pyrethroids have substituted for the
highly toxic organophosphorous compounds.
11
Herbicides—
Most herbicides have low toxicity to people. Paraquat, widely used with
conservation tillage, is an exception. According to Giere, Johnson, and
Perkins a dose of 10 to 30 mg/kg is fatal to half the human population, and
continual contact in dilutions of 1 to 200 can cause skin Irritation and loss
or deformation of fingernails.
There is some evidence that a number of herbicides, including paraquat,
may be carcinogenic or mutagenic. Giere, Johnson, and Perkins cite two
studies, one of which raises a question about paraquat as a mutagen, and
another which failed to find mutagenic activity. Chemical and Engineering
News (April 11, 1980, p. 4) refers to studies in Europe indicating that
2,4,5-T and 2,4-D increase the risk o.f developing certain kinds of cancer.
2,4,5-T, the "agent orange" herbicide used in Vietnam for defoliation, is
thought by some to also produce birth defects and other reproductive abnor-
malities. And the press has carried accounts indicting 2,4-D as the probable
cause of miscarriages among women in Montana living in areas where 2,4-D was
regularly used. Chemical and Engineering News, however, reports a study by
the U.S. Government's National Toxicology Program of the effects of 2,4,5-T
and 2,4-D on male mice showing no significant effect on fertility, reproduc-
tion, germ cell toxicology or survival and development of the offspring of
the exposed animals.
Atrazine, which accounts for almost 25 percent of all herbicides applied
to crops in the United States, most of it on corn, has low toxicity to humans.
There is evidence, however, that atrazine may be transformed metabolically by
plants to form a substance which is mutagenic (Plewa and Gentile). Giere,
Johnson, and Perkins assert further that atrazine can be transformed in the
human stomach to a nitrogen derivative under strong suspicion as a carcinogen.
A study by M. J. S. Hsia of propanil, another low toxicity herbicide
used on rice, showed that in the soil propanil is metabolized by first fungi
and then microorganisms to a compound very similar to dioxin, the teratogen
found in Agent Orange (Science News).
Concern also has been expressed that extended application of herbicides
may damage soil microorganisms. There is disagreement in the literature on
this issue. Sommers and Biederbeck, after reviewing numerous laboratory and
field studies, assert that herbicides (and insecticides), when applied at
recommended field rates, generally have no lasting harmful effects on micro-
bial populations in the soil.
Greaves appears to be in at least partial agreement with Sommers and
Biederbeck. Summarizing a number of studies of the effects of herbicides
on soil microorganisms Greaves states that "While there is some evidence
•^This section is taken from Crosson (1981).
Ill
-------
that changes in the soil microflora do occur following long-term use of herb-
icides, the data suggest that these changes do not necessarily result in sig-
nificant changes in soil processes such as nitrogen transformation. Where
processes do apparently change, this is generally ascribable to decreases in
soil organic matter or to uptake of nutrients by surviving weeds" (p. 131).
Greaves does not take this conclusion as grounds for complacency, how-
ever. He points out that most of the work on the problem has ignored the
effects of herbicides on microorganisms living on roots. Herbicides affect
root morphology and may, therefore, damage these root dwelling microorga-
nisms. Greaves ends by noting that relatively little is known about the
long-term side effects of herbicides and calls for more research.
Eijsackers and van der Drift make a similar argument. They assert that
while field tests show that soil fauna recover quickly from the toxic effects
of herbicides, the processes underlying this are not well understood and "ex-
tensive research" is needed (p. 169). von Rumker et al. state that most
studies of the metabolism and degradation of herbicides have focused on
effects of herbicide residues in the soil on following.crops or other valu-
able vegetation. Consequently there is little information about other effects
of herbicide residues in the soil or in other elements of the environment.
There is enough uncertainty about the more subtle, long-term effects of
herbicides on soil and water dwelling organisms, and perhaps on humans, to
warn against complacency about present environmental impacts of these mater-
ials. Still, the available evidence does not support the argument that these
impacts impose major environmental damage. This judgment is supported by a
survey of thirty-six agricultural scientists from all around the country
(Barnes et ai.). ' The scientists were asked to describe cases of water pollu-
tion by pesticides which in their judgment constituted a problem. Only two
problems clearly involving farm-applied herbicides were listed. Both were
crop damage resulting from use of herbicide-contaminated water for irrigation.
Picloram, a particularly persistent and leachable herbicide, has been found
in the water of wells in two counties in Nebraska. The concentration was
only a few parts per billion, but enough to damage crops irrigated with the
well water. Similar damage was reported in cases where water from a pond
contaminated by herbicides in runoff was used for irrigation, and where herbi-
cide-contaminated runoff flows directly into fields.
The survey also brought mention of possible damage from atrazine and
alachlor residues in sediment in the Chesapeake Bay estuary and in the Bay
itself. A study suggested that these residues were killing bottom vegeta-
tion in the Bay. This is in dispute, however, because the scientific pro-
cedures used in the study are believed by some to be inadequate.
R. D. Wauchope, the person responsible for the survey of expert opinion
about present pesticide damages to water quality, summed up his findings as
follows:
It is unlikely that any significant observable, nonpoint
source water quality problem due to proper agricultural or sil-
vicultural use of pesticides would not be known to at least one
of the sources interviewed ...It is a reasonable conclusion,
112
-------
then, that with a few possible exceptions, currently-registered
agricultural and silvicultural pesticides are not observed to
be causing problems with respect to water quality. (Barnes et
al., p. 15. Emphasis in the original).
This conclusion must be taken with certain caveats. The experts sur-
veyed by Wauchope did not address the possibilities that some herbicides, as
noted above, may be carcinogens, mutagens or teratogens; and we already have
seen that the absence of evidence that herbicides now constitute a major
environmental threat may simply reflect the failure of researchers to look
in the right places, or the inability of present detection techniques to pick
up effects at very low trophic levels. Nevertheless, Wauchope's survey sup-
ports the judgment that, on the available evidence, present levels of use of
herbicides do not pose major threats to the environment.
Future Severity of the Problems
The future environmental impacts of pesticides will depend upon the
quantities of these materials used by farmers and upon their characteristics,
particularly their toxicity and longer-term effects on humans and other non-
target organisms. We discuss both quantities likely to be used and charac-
teristics .
Insecticides—
Judging from present quantities and nonpersistent characteristics of
insecticides used, the environmental impacts of these materials are concen-
trated primarily in the Southeast and Delta and secondarily in the Cornbelt.
This reflects the location of cotton production in the Southeast and Delta
and the difficulty of controlling cotton insect pests in those regions, and
the need to control corn insect pests, particularly the corn rootworm.
We concluded from our analysis in the last section of trends in tech-
nologies for controlling cotton and corn insect pests and of the continuing
shift of cotton production from the Delta to Texas that the quantity of
insecticides applied to cotton and corn would decline sharply from present
levels. These crops are so dominant in total insecticide use that their
decline would bring a significant reduction in the total quantity of insecti-
cides used unless there were an extraordinary increase in amounts used on
other crops. Our analysis of trends in insect management technologies for
wheat pointed to no such increase. On the contrary, these trends are toward
diminishing per acre applications of insecticides. While our projection of
land in wheat indicates an offsetting trend, the prospect, in our judgment,
is for little if any increase in total insecticide applications to wheat.
We projected a more than 100 percent increase in insecticide use on soy-
beans because of a relative shift of soybean acreage to the Southeast and
Mississippi Delta. This increase is small in absolute amount, however, com-
pared to the anticipated decline in insecticides used on cotton and corn.
113
-------
The implication is that so far as the environmental problems of insecti-
cides are a function of quantities used, the problems should diminish signi-
ficantly, both from the national perspective and from the perspective of the
three most affected regions.
We expect the characteristics of insecticides also to change in ways
tending to diminish environmental damage. The shift away from the organo-
chlorine compounds appears permanent, indicating that the particularly
troublesome problems associated with persistence, biomagnification, and bio-
accumulation will continue to decline.
The substitution of the organophosphorous and carbamate compounds for
the organochlorines substitutes problems of acute toxicity for those of per-
sistence, biomagnification, and bioaccumulation. To the extent that human
mortality and morbidity are measures of environmental impact, this substitu-
tion would appear to be for the worse. While this appears a plausible pre-
sumption, such data as are available—and they are quite incomplete—do not
indicate an increase in insecticide-related deaths associated with the shift
to organophosphates and carbamates (National Academy of Sciences, 1975,
pp. 87-89). Indeed, as noted earlier, the number of accidental deaths from
pesticide poisonings was less in 1974 than two decades previously, the period
when the shift from the organochlorines to organophosphorous and carbamate
compounds was occurring (Pimentel et al., 1980).
The shift away from the organochlorines may also have reduced water
pollution by insecticides. Pimentel et al. cite three studes of pesticide
concentrations in surface waters of the United States which show a steady
decline in concentrations from 1964 to 1978. Total pesticide use increased
sharply in this period. Pimentel et al. attribute the decline in concentra-
tions to the shift from the persistent organochlorines to the less-persistent
organophosphorous and carbamate compounds.
There is another, perhaps subtle, respect in which the shift to the
organophosphorous and carbamate compounds may be favorable to the environ-
ment. Injury to humans and other nontarget organisms by these materials
result from their acute toxicity, meaning that the injury is severe and
quickly evident. Consequently, both the victims and those responsible
usually can be readily identified. The damages caused by the organochlorine
compounds—by contrast, do not produce sudden, severe symptoms, but rather
show up as impaired reproductive processes, cancer, or other long-delayed
effects. In addition, because of the persistence, biomagnification and
bioaccumulation properties of the organochlorines, it may be difficult if
not impossible to determine the original source of the damaging materials
and those responsible for them. Because of these differences in the timing
and nature of the damages done, the organophosphorous and carbamate com-
pounds likely would be easier to manage than the organochlorines. The point
is not that the potential environmental damages of the organophosphorous and
carbamate compounds are less—they may in fact be greater—but that the
actual damages may be less because of greater ease in management of these
materials.
114
-------
The substitution of synthetic pyrethroids for some of the organophos-
phorous compounds in control of cotton insect pests seems likely to continue.
Since the pyrethroids have low toxicity to mammals and are not persistent,
this trend reduces pressure on the environment. Their apparent high toxicity
to fish is offset by the low probability that they will reach surface waters
in significant amounts. There are reports from the Southwest of emerging
resistance of the tobacco budworm to the pyrethroids, and if this becomes
general the usefulness of these materials will be greatly reduced. If this
happens, however, it would appear to be five to ten years ahead, time in
which to develop substitutes for the pyrethroids. While we do not now fore-
see what these substitutes may be, the whole thrust of research and develop-
ment in cotton pest management technologies is toward substitution of non-
chemical for chemical means of control. We consider it unlikely, therefore,
that whatever substitutes for pyrethroids are developed—if indeed they
become necessary—will pose a greater threat to the environment than the
pyrethroids.
We conclude that on balance changes in characteristics of insecticides
will reinforce the prospective decline in amounts applied to significantly
reduce the environmental damages of insecticides over the next several
decades.
Herbicides—
We concluded in the previous section that if per-acre amounts of herbi-
cides applied to land in corn, wheat, soybeans, and cotton remain the same as
in 1976, total herbicide applications to these crops (82 percent of all herbi-
cides applied to crops in 1976) would rise 21 percent from 1976 to 1985 and
53 percent to 2010 (see Table 5-15). We also noted that if conservation
tillage spreads to 50 to 60 percent of cropland by 2010, as seems likely,
the increase in total herbicide use would be significantly higher than these
numbers indicate.
While the shift to conservation tillage will increase the amounts of
herbicides applied more than otherwise would occur, the movement of herbi-
cides from farmers' fields probably will not rise in proportion to the
increase in amount of herbicides. The reason is that losses of herbicides
through erosion and runoff probably are less with conservation tillage than
with conventional tillage. This is the conclusion of a survey of the litera-
ture on tillage and losses of pesticides by Wauchope, McDowell and Hagen. It
was the key finding also of a study of small watersheds in Iowa by Baker and
Johnson.
Despite the apparent favorable effect of the shift to conservation till-
age in reducing herbicide losses, the projected increase in amounts applied
in 2010 is so large that a significant increase in the quantity of herbicides
in the environment is implied. To the extent that quantity is a measure of
environmental burden, that burden would increase.
The issue of the inherent threat of herbicides to the environment thus
is crucial in judging the severity of the future threat. We concluded from
our previous discussion of the issue that the evidence does not support the
115
-------
inference of a major inherent threat. We noted, however, that some aspects
of the herbicide-environment relationship which may contain a threat have not
been thoroughly investigated. The possibility of real or potential threats,
therefore, cannot be discounted. Moreover, we have no experience with the
use of herbicides on the scale projected for 2010. Conceivably there could
be thresholds of environmental damage which expansion on the projected scale
would pass.
The possibilities of presently undetected or potential future environ-
mental damages of herbicides are sufficiently likely, in our judgment, to
justify intensive, continuing investigation of herbicide-environment rela-
tionships, and a wary attitude toward the expanding use of herbicides.
Adopting such an attitude, we nevertheless conclude that present knowledge
does not suggest that the projected expansion of herbicide use will pose a
major threat to the environment.
IRRIGATION
Nature of the Environmental Impacts
Irrigation may, and in arid areas typically will, lead to an increase in
soil salinity. Where water is used several times for irrigation, as river
water usually is, the salt content of the water also rises. Too much salt
in the soil or in irrigation water inhibits plant growth, and in extreme
cases may render the land and water useless for agriculture.
All water used for irrigation carries salts. The increasing concentra-
tion of salt in the soil occurs because of evaporation of some of the water
applied to the land. Over time the amount of the salt will increase unless
it is periodically flushed out by rain or by water applied specifically for
that purpose. Flushing of the salt will protect the productivity of the land
so cleansed, but it increases the salt content of the water available to
farmers downstream. The salt content of the water also increases as it moves
downstream because irrigation return flows carry salt from the land to the
river.
If drainage is inadequate, repeated irrigation may also raise the
underground water table within reach of the plant root zone. Capillary
action then will carry water close to the soil surface where it evaporates,
leaving a salt residue. The accumulation of these residues eventually will
reduce the productivity of the land.
Stream flow is reduced by withdrawals of river water for irrigation.
The decline in stream flow increases pollutant concentrations and water
temperature. Both can have devastating effects on riverine flora and fauna.
Moreover, the flow of fresh water into estuaries and tidal marshes is dimin-
ished, reducing the productivity of these ecologically vital and often
commercially valuable resources.
116
-------
Irrigation may increase erosion relative to pre-irrigation treatment of
the land. It may also have the opposite effect, however, by increasing the
plant cover of the land, thus reducing its exposure to the erosive effects
of wind and rain.
Because irrigation increases the productivity of the land compared with
dryland fanning it justifies greater per-acre use of fertilizer and pesti-
cides, thus increasing the likelihood of environmental damage from these
materials.
Present Severity of the Problems
Salinity is the most pervasive environmental problem stemming from irri-
gation in the United States.H Except for the Columbia, all western river
basins are confronted with high and generally rising salt levels. In much
of the Lower Colorado, parts of the Rio Grande, and the western portion of
the San Joaquin river basins, the salt concentrations in either the water or
the soils are approaching levels that threaten the viability of traditional
forms of irrigated agriculture. Groundwater salinity is not now a widespread
problem in the West, However, the problem is growing and already is serious
in parts of California, New Mexico, Montana, and Texas (U.S. Department of
Interior, April 1975, pp. 116-118, and U.S. Water Resources Council, Vol. I,
Summary, p. 65). Although groundwater is less susceptible than surface waters
to pollution, the clean-up of groundwater is difficult.
von Schilfgaarde estimates roughly that 25 to 35 percent of the irri-
gated lands in the West have some type of salinity problem, and that the
problems are getting worse.^ Salinity already is a serious problem in two
large irrigated areas, the Lower Colorado River Basin and the west side of
the San Joaquin Valley in California. The underlying cause of the problem
in the two areas differs. In the Colorado the salt content of the river
increases progressively downstream due to the salt-concentrating effects of
irrigation and the addition of salts picked up as the water passes over
saline formations. About two-thirds of the salts delivered to the Colorado
are from natural sources. Elimination of these deliveries would require
expensive investments to divert the river and its tributaries around some
of the areas contributing the salts. Irrigation accounts for most of the
remaining salts. These could be curtailed greatly by reducing irrigation
return flows. Annual damages from salinity in the Colorado River were esti-
mated between $75 and $104 million in 1980. (U.S. Department of Interior
and U.S. Department of Agriculture, 1977).
In the San Joaquin Valley the principal salinity problem occurs because
poor drainage prevents salt-laden waters from being carried away from the
fields. High water tables already threaten the productivity of about 400,000
12This discussion of salinity in the West is taken from Frederick (1981),
Chapter 6, pp. 6-5/6-7.
13Jan van Schilfgaarde, Director, U.S. Salinity Laboratory, Riverside,
California, in an interview with Kenneth Frederick, February, 1980.
117
-------
acres. Ultimately more than 1 million acres in the Valley may be similarly
affected. A $1.26 billion drainage system has been proposed to carry irri-
gation runoff from the western side of the Valley to the Sacramento Delta.
Because farmers would have to -install their own underground drainage to get
the waters to the central drain, the total costs of an effective system would
be considerably higher than $1.26 billion (U.S. Department of Interior, Cali-
fornia Department of Water Resources and California State Water Resources
Control Board)
Salinity induced by irrigation is overwhelmingly a problem of the arid
and semiarid West. It is not a significant problem in the principal irri-
gated areas of the East: the Mississippi Delta region, the Great Lakes states
and the Southeast.
The contribution of irrigation to the nation's erosion problem is minor.
Although erosion and sedimentation are viewed as problems in many areas of
the West, it is not clear, with a few exceptions, that irrigation is an im-
portant contributor to these problems. A report by the U.S. Department of
Interior (April 1975, p. 127) states that in the eleven most western states
"most of the rapidly eroding range, grassland, and forest-covered soils
occur where natural geologic erosion is dominant." The human contribution
to erosion probably has been more important in the Plains States, but the
contribution there of irrigation is mixed. Center pivot irrigation is impor-
tant in that region, and the wheel tracks left with this system can lead to
gully erosion. However, erosion by wind is far more important than erosion
by water in the West, and irrigation, by increasing ground cover, helps to
contain wind erosion. The National Resources Inventory showed that soil loss
from wind erosion on cropland in the ten Great Plains states was 893 million
tons in 1977, or 5.3 tons per acre (USDA, Aug. 21, 1979 and Feb. 1980). Ero-
sion by sheet and rill erosion (that is, by water) in these states was 516.6
million tons, or 3.1 tons per acre. Only a small part of the sheet and rill
erosion was caused by irrigation, rain being the dominant cause. The Soil
Conservation Service sets 3 to 5 tons per acre per year as the maximum
amount of soil loss consistent with maintaining the productivity of the land
indefinitely. By this standard we can conclude that irrigation in the West
does not contribute importantly to the, region's erosion problem. Indeed, the
NRI data suggest that erosion of cropland by water, whatever the source, may
not be a problem in the West.
We point out below that sheet and rill erosion is a major problem in
parts of the East, including parts of the Mississippi Delta and the South-
east, the principal areas of irrigated farming in the East. However, in
those states where erosion is particularly serious—Mississippi, Alabama,
and Georgia—less than 10 percent of cropland is irrigated, and there is no
reason to believe that irrigated land contributes more erosion per acre than
nonirrigated land. Indeed, except when sprinkler systems are used, irriga-
tion typically requires that the land be leveled, suggesting that irrigation
would reduce erosion from such land.
The Second National Water Assessment (U.S. Water Resources Council)
identifies agricultural chemicals as a "significant" source of surface water
pollution in three areas in the West, but does not indicate whether irrigated
118
-------
or dryland farming is the principal culprit. In comparison with the rest of
the country this problem is not very extensive in the West. In our earlier
discussion of fertilizer pollution we noted certain "hot spots" in southern
California where nitrates in groundwater are perceived to be a problem, with
irrigation a probable contributor. We also noted nitrate levels of 20 to 25
ppm in groundwater in the Sandhills region of Nebraska, where irrigation is
extensively practiced. However, in other parts of the Plains states that
have been under irrigation for many years, e.g., the High Plains of Texas,
there still are no observable problems of groundwater pollution from irriga-
tion.
The Assessment indicates that agricultural chemicals present more water
quality problems in the East than in the West. However, since only a small
percentage of cropland in the East is irrigated, the contribution of irriga-
tion to these problems, such as they are, must be correspondingly small.
Future Severity of the Problems
We expect only limited further expansion of irrigation in the West, so
we do not expect the addition of irrigated land to significantly increase the
environmental damages of irrigation. However, the severity of the major
source of damage, salinity, is not just a function of the amount of irri-
gated land. Salinity tends to increase on a given amount of land with re-
peated irrigations. The salinity problem, therefore, is likely to increase
more than in proportion to the increase in irrigated land unless steps are
taken to deal with it. According to Frederick, improved farm management
techniques can go a long way toward reducing some salinity problems and ob-
viating the need for some of the costly structural solutions under considera-
tion.!^ Efficient water and agronomic management reduces evaporation losses,
permitting achievement of given yields with less water. The reduction in
water applied and evaporation slows the buildup of salt concentrations in
the soil. In addition, agricultural scientists have developed crops and
irrigation techniques which enable farmers to irrigate successfully with
surprisingly high salt levels. For example, van Schilfgaarde (1977) reports
research suggesting that salinity in the lower portion of the root zone can
be permitted to build up considerably more than suspected without decreasing
yields if good quality water is applied in the upper zone. Science has made
and is likely to continue to make breakthroughs such as the development of
crop varieties with higher tolerance to salt, which will expand the possi-
bilities for successful irrigation with saline water and soil. While such
results may postpone or mitigate the damages of high salt buildup, the long-
term implications of rising salt levels are uncertain. The soils can be
used in the short term as a depository for salts, but eventually additional
salts would have to be flushed to preserve productivity.
This discussion of techniques for managing salinity is based on
Frederick (1981), Chapter 6, pp. 6-16/6-17, as modified on the basis of
comments by Jan van Schilfgaarde. van Schilfgaarde is not responsible for
the analysis here presented.
119
-------
Improved basin-wide water management also has great potential for reduc-
ing salinity problems. Where poor drainage prevents salt-laden waters from
being carried away from the fields, artificial drainage is required to pre-
vent eventual loss of productivity. But the extent of the drainage needs can
be greatly reduced through good management, van Schilfgaarde suggests that
the San Joaquin Valley drainage problem could be reduced to about 5 percent
of current levels and a long-term equilibrium reached through an integrated
irrigation system whereby the best water is used first on salt-sensitive
crops with the increasingly salt-laden runoff applied to increasingly salt-
tolerant crops. The remaining highly saline waters would be reduced to
quantities that could be disposed of in evaporation ponds rather than requir-
ing a costly trans-basin drain.^
We concluded above that irrigation does not contribute importantly to
erosion at present, and we do not expect this to change over the next several
decades. Part of the considerable increase in fertilizer applications we
project for the Southeast, however, is associated with the prospective ex-
pansion of irrigation in that region (see Section 4). Environmental impacts
of much of the additional irrigation in the Southeast will be on sandy soils
in Georgia and Florida where percolation rates are high. The expansion of
irrigation in that region, therefore, likely will contribute to increasing
nitrate concentrations in groundwaters of the region. We are unable to say,
however, whether the concentrations are likely to reach troublesome levels.
EROSION
Nature of the Environmental Impacts
There are two sorts of impacts: on-farm and off-farm. The on-farm
impact is the socially undesirable loss of productivity resulting from ero-
sion of topsoil. The topsoil supports plant growth by providing storage
for nutrients, moisture and air, and a medium in which plant roots can take
hold. Erosion will eventually reduce the capacity of the topsoil to provide
¦^van Schilfgaarde in an interview with Kenneth Frederick, February 1980.
•^The socially undesirable productivity losses imposed by erosion per-
haps should not be called environmental costs as those costs are usually
defined. Environmental costs typically mean damages inflicted on society by
effluents of production and consumption activities. The social costs of
productivity losses are in higher, prices of food and fiber reflecting a di-
vergence between the social and private interest in maintaining the produc-
tivity of the land. However, dealing with the social costs of erosion-induced
productivity losses poses an issue of public policy in precisely the same
sense as the off-farm damages of erosion, which are "environmental" in the
usual sense of the word. Hence we treat the productivity losses as a kind
of environmental cost.
120
-------
these services to the plant unless new soil is formed at a rate sufficient
to offset erosion losses. The rate at which new soil is formed is highly
variable, depending upon the nature of the parent material underlying the
topsoil, climate, biological factors (plants, soil flora and fauna) and
topography (Cady, 1980). The U.S. Soil Conservation Service (SCS) has de-
fined the tolerable level of soil loss ("T" value) as the maximum amount of
loss per acre per year "that will permit a high level of crop productivity
to be sustained economically and indefinitely" (Wischmeier and Smith, p. 2).
Both physical and economic factors are considered in establishing "T" values,
although Wischmeier and Smith's discussion suggests that physical factors are
dominant. They assert, for example (p. 2), that "T" values in the United -
States range from 1 to. 5 tons per acre per year, depending upon soil proper-
ties, depth, prior erosion, and topography. Five tons of topsoil is about
one-thirtieth of an acre inch, suggesting that in the judgment of the SCS
that is the maximum amount of new soil that can be formed in a year in most
circumstances.
It is important to note that erosion-induced losses of productivity per
se do not impose on-farm environmental costs. Only those losses which im-
pinge on society's interest in the productivity of the land are counted.
These losses arise when there is a divergence between society's interest and
the farmer's interest in this respect. Society's interests include those of
future generations, but it is difficult for these interests to be registered
in ways that will make the farmer responsive to them. The farmer's concern
about erosion is strongly conditioned by his expectations about future crop
prices relative to current prices. Since the prices at which American far-
mers sell are highly unstable, the prudent farmer will apply a high rate of
discount to future prices. The effect is to shorten his planning horizon.
If current prices are high he will farm his land intensively, abandoning soil
conservation practices if necessary, because the future beyond the next year
or two is highly uncertain.
If commodity and land markets correctly measured the long-term social
interest in maintaining the productivity of the land, then the effect of
erosion on that interest would be reflected in land prices. Farmers who did
not hold erosion within socially acceptable limits would find the value of
their land reduced. They would have incentive, therefore, to take the mea-
sures necessary for erosion control. However, that markets correctly measure
the social interest in maintaining the productivity of the land is question-
able. Markets do not register prices expected to prevail generations into
the future. The reason is not simply that those who will people future
generations are not yet born or are too young to express their interests.
The fundamental difficulties are vast uncertainty about the substance of
the interest of future generations in conserving the land and about how best
to serve that interest, even if it were clearly identified. How big will
future generations be, and how much income will they have? How much of it
will they spend for food and fiber, and how can these demands be modified,
without damage to welfare, if they appear to exceed the capacity of the land
to meet them at reasonable cost? What are the possibilities for substituting
other inputs or techniques, e.g., fertilizer, increased photosynthetic effi-
ciency, hydroponic farming, for the land in meeting these demands?
121
-------
Fundamentally, it is the great uncertainty surrounding answers to these
questions which leads to a divergence between the farmer's interest and soci-
ety's interest in conserving the land, hence to socially undesirable amounts
of erosion. But if the uncertainty is so great, how can we be sure that
there is a conflict between the social interest and the fanner's interest
in the land? The fact is we cannot be sure. We can reasonably expect, how-
ever, that the combination of rising world population and income will sub-
stantially increase future demands for American produced food and fiber;
that continued erosion will diminish the capacity of the land to meet these
demands; that compensating technological advance is uncertain, and that if
it fails the resulting increase in costs of food and fiber may produce sig-
nificant social problems in the United States and in other countries where
U.S. interests are involved. We can be reasonably sure that the rise in
food and fiber prices would not be signaled to the farmer far enough in ad-
vance to induce him to take action to reduce erosion. It appears, therefore,
that for society, the prudent course is to assume a likely divergence between
the public interest and the farmer's interest in maintaining the productivity
of the land.
The off-farm costs of erosion result from the sedimentation of reser-
voirs and lakes, rivers, and harbors, and from the need to process the water
to make it drinkable or usable for industrial purposes. The suspended sedi-
ment reduces recreational values provided by water and may injure fish popu-
lations. When the sediment settles it shortens the useful life of reservoirs
for generation of power and in providing irrigation and flood control, and
of harbors in all the manifold services they provide. Removal of the sediment
from these water bodies becomes a continuing and costly burden.
In thinking about both the on-farm and off-farm costs of erosion it
should be remembered that some of the world's most productive land is in
flood plains built up by soil moved by water from upstream. We should be.
grateful for the existence of such land, but also recognize that most of it
was put in place at a time when the social cost of the process was low.
Under present conditions, building the productivity of the land by downstream
deposition of.eroded soil has got to be socially more costly than preserving
the productivity of the land threatened by erosion.
Present'Severity of the Problems
It is quite certain that the productivity of the nation's agricultural
land has been and is being reduced by erosion. There are no reliable, com-
prehensive estimates of past or current productivity losses, however. What-
ever past losses may have been, advances in technology more than compensated
for them. Production of crops per acre of cropland increased 110 percent
from 1910-14 to 1975-79, an annual rate of 1.14 percent. The rate of in-
crease was especially rapid after World War II: 1.83 percent.
This is not to say that the erosion-induced losses were, or are,
negligible from a social standpoint. Since by definition these losses do
not enter the farmer's private calculations,.they probably did not get the
weight society would have given them. Had this weight been assigned, the
122
-------
pattern of output growth likely would have veered more toward preserving
productivity by reducing erosion and less toward finding technological sub-
stitutes for the land.
All of this, however, is speculation. As noted earlier, the possibility
of divergence between the farmer's interest and society's interest in pre-
venting losses in productivity of the land is sufficiently great to justify
the presumption that the losses are excessive.
There are conflicting estimates of erosion from agricultural land,
implying conflict also in estimates of the off-farm costs of this erosion.
Pimentel et al. (1976) cite various sources indicating that costs of dredging
rivers and harbors, of reduced useful life of reservoirs, and of other sedi-
ment damages came to about $500 million annually in the 1960s. In 1980
prices these costs probably would not be less than $1 billion.
These costs reflect total erosion, not just that from agricultural land.
Pimentel et al. estimated total water-borne sediment at 4 billion tons per
year, three-quarters of it from agricultural land. Meister et al. use figures
which imply substantially smaller amounts of water-borne sediment from agri-
cultural land than estimated by Pimentel et al. In contrast, estimates from
the NRI (USDA, Feb. 1980) indicate that the Pimentel figures may be low.
Clearly there is great uncertainty about present on-farm and off-farm
environmental impacts of erosion. Nevertheless, consideration of current
rates of erosion provide some insights to the present severity of the prob-
lem. Since in this study we are interested in the environmental impacts of
crop production, we focus on erosion from cropland.
Table 6-1 shows national and regional estimates of erosion from cropland
in 1977. For the nation, total erosion was 2,799 million tons, about two-
thirds of it sheet and rill erosion and one-third erosion by wind. Sheet and
rill erosion accounted for a little more than two-thirds of the 6.8 tons per
acre of total erosion. Sheet and rill erosion also was important on range-
land, pasture, and grazed forestland (figures not shown in Table 6-1), but
per-acre loss rates were significantly lower in those cases. Moreover, the
heaviest losses occurred in less populated areas and on less valuable land,
making their economic costs and damages to the human environment clearly less
significant. Cropland erosion losses, by contrast, were concentrated in such
productive states as Iowa, Illinois and Missouri (parts of the Cornbelt),
Nebraska and Kansas (parts of the Northern Plains); Tennessee and Kentucky
(parts of Appalachia); and Mississippi and Alabama (parts of the Mississippi
Delta). Thus, the agricultural heartland of the country is most severely
affected.
In the Great Plains, cropland also is subject to significant wind ero-
sion. Texas, New Mexico, and Colorado are most vulnerable, with losses
amounting to 9 to 15 tons per acre in those states. Indeed, wind erosion
from cropland in Texas accounts for 16 percent of total cropland erosion in
the entire nation. In Plains states other than Texas, New Mexico and Colo-
rado, wind erosion rates fall below 5 tons per acre for all cropland.
123
-------
Table 6-1. EROSION FROM CROPLAND IN THE UNITED STATES
Wind
Sheet & Rill
Total
Percent
of Total
Region
Amount Tons per
(mil.tons) Acre
Amount Tons per
(mil.tons) Acre
Amount Tons per
(mil.tons) Acre
Erosion
Cropland
Nation
891
2.1
1,908
4.7
2, 799
6.8
100
100
Northeast
n.e.
82.9
5.0
82.9
5.0
3.0
4.0
Lake States
n.e.
117.5
2.7
117.5
2.7
4.2
10.7
Cornbelt
n.e.
688.3
7.7
688.3
7.7
24.6
21.8
Iowa
n.e.
261.3
9.9
261. 3
9.9
9.3
6.4
Northern Plains
212.3
2.2
322.4
3.4
534.7
5.6
19 .1
22.9
Nebraska
25.9
1.3
117.8
5.7
143.7
7.0
5.1
5.0
Appalachia
n. e.
186.3
9.0
186.3
9.0
6.7
5.0
Tennessee
n.e.
69.5
14.1
69.5
14.1
2.5
1.2
Southeast
n.e.
111.0
6.3
111.0
6.3
4.0
4.2
Georgia
n.e.
42.7
6.6
42.7
6.6
1.5
1.6
Delta
n.e.
154.9
7.3
154.9
7.3
5.5
5.1
Arkansas
n.e .
46.7
5.9
46.7
5.9
1.7
1.9
Southern Plains
488.8
11.6
141.4
3.4
630.2
15.0
22.5
10.2
Texas
453.5
14.9
99.5
3.3
553.0
18.2
19.6
7.4
Mountain
190.3
4.5
70.8
1.7
261.1
6.2
9.3
10.2
Pacific
n.e.
31.9
1.4
31.9
1.4
1.1
5.6
California
n.e.
8.6
.9
8.6
.9
.3
2.4
Note: n.e. means not estimated.
Source: USDA (Feb. 1980 and 1980). The data are for the 48 contiguous states. When Hawaii and the
Caribbean area are included, total sheet and rill erosion is 1,926 million tons.
Sheet and rill erosion are caused by water. Sheet erosion is the movement of continuous layers of
soil from the field. If the water moves fast enough it scours the soil and cuts small channels in the
soil surface. The soil moved in this fashion is rill erosion.
-------
The trend in erosion loss is difficult to assess. SCS surveys in 1958
and 1967 showed somewhat higher average losses from sheet and rill erosion
than the NRI and a greater percentage of cropland needing conservation treat-
ment. However, there are judgment factors involved in arriving at these
numbers, and it is difficult to know if they were treated in precisely the
same manner two decades apart. Even with this caveat in mind, it is useful
to note that reported erosion losses dropped about 1 ton per acre and the
amount of cropland needing conservation treatment fell by 3 percentage, points
between 196.7 and 1977—a period when the reserve of idle and often erosion-
prone acreage was brought back into cultivation and many decried the apparent
increase in erosion (USDA, 1974, USDA, February 1980). This reported im-
provement in the face of greater cropping pressure coincided with a signi-
ficant spread in the use of conservation tillage during the same period.
At the national level, sheet and rill erosion losses from cropland
approach the maximum of 5 tons per acre per year set by the SCS as consistent
with long-run maintenance of the productivity of the land, and erosion by
wind and water combined exceed the limit.
These national soil loss averages conceal more severe local and regional
problems. Many states, especially in the West and Lake states, have low
average losses from cropland, while in parts of the Cornbelt, Delta, South-
east, and Wouthern Plains, losses far exceed tolerable amounts. But erosion
is a still more localized problem. It may be severe in very small areas or
on individual fields or parts of fields, even in regions that generally are
not afflicted. Nationally, 10 percent of the cropland accounts for 90 per-
cent of all sheet and rill erosion in excess of 5 tons per acre (USDA, 1980,
Part II). Because the problem tends to be localized, it is amenable to
sharply focused programs of control, discussed in the next section.
The soil moved by wind and water erosion carries organic matter and
nutrients from the field. Erosion losses from the top layers of soil usual-
ly are the most damaging to soil quality because they are richest in the
nutrients and organic matter that are most accessible to plants. Damage on
the field can be measured in terms of the costs of replacing the nutrients
lost and the possible decrease in the productive potential of the land once
erosion exceeds "T" value. There are no confident measures of either effect.
The RCA study concluded that erosion at 1977 rates would reduce potential
corn and soybean yields in the Cornbelt by 15 to 30 percent over a fifty
year period (USDA, 1980, Part II). Other studies suggest that these figures
may be conservative. This does not necessarily mean that yields will de-
cline, but the land will become less responsive to other inputs and will
require more of them to sustain any given level of output.
The RCA estimates suggest that indefinite continuation of present rates
of erosion will in time impair the productivity of the land enough to in-
crease costs of agricultural production. Since trends in prices and produc-
tivities of nonland inputs also suggest rising costs, the additional impetus
to costs given by erosion is not to be taken lightly. In addition, as we
already have seen, estimates by Pimentel et al. (1976) indicate that off-
farm erosion damages to lakes, reservoirs, and harbors probably are not less
than $1 billion per year in 1980 dollars. There is basis for believing,
125
-------
therefore, that the on-farm and off-farm environmental impacts of present
rates of erosion are cause for concern. Given our projections of production,
land use and technology, what can we say about future impacts?
Future Severity of the Problems
¦ To obtain insights to future environmental impacts of erosion we engaged
the help of the Center for Agricultural and Rural Development at Iowa State
University. Specifically, we asked the Center to run our projections to 2010
of land in feedgrains, wheat, soybeans, and cotton through the ISU model of
the U.S. agricultural economy to obtain projections of erosion.^ ^e re-
sults are shown in Table 6-2. A description of how the results were obtained
will be useful for interpreting them. We provided ISU with an initial set of
projections to 2010 of harvested land in feedgrains, wheat, soybeans, cotton,
hay, and fallow, for the nation and the ten USDA producing regions. The
projections were fed into the model and the result was the set of projections
of erosion called run 1 in Table 6-2. (The model yields many other outputs,
but our interest was in erosion.) As indicated in Section 4, the model is
structured to give results by the 19 river basins shown in the table rather
than by the ten USDA producing regions. Moreover, the model had difficulty
in accommodating the RFF projections of harvested land precisely as we had
prepared them, so the run 1 results reflect a slightly different cropping
pattern than the one we had projected. There was no difference in the land
totals, however.
The ISU model allocates land and production by three kinds of tillage
technology and four kinds of conservation practices. The tillage technolo-
gies are (1) fall plowing with the moldboard plow, residue removed (here
called conventional tillage); (2) spring plowing, with or without the mold-
board, with residue left until spring (here called low conservation tillage);
(3) spring plowing with something other than the moldboard, with residue
left year around.(here called high conservation tillage). The four conser-
vation practices are straight row planting, contour farming, stripcropping
and terraces, each of which may be practiced with any of the tillage tech-
nologies.
Given projections of production, the model chooses, among tillage tech-
nologies and conservation practices so as to minimize production costs. In
run 1 the model allocated 17.2 percent of harvested land to conventional
tillage and 82.8 percent to the two forms of conservation tillage, 34.9
percent to the "low" form and 47.9 percent to the "high" form.
¦^As noted in footnote 3, Section 4, the ISU model is a cost minimizing
linear programming type. It permits analysis of a variety of resource and
environmental issues touching U.S. agriculture from both regional and na-
tional perspectives. The model was used extensively by the USDA in its RCA
work to explore the implications of various scenarios about future levels
of agricultural production and policies for controlling erosion.
126
-------
Table 6-2. PROJECTIONS OF SHEET AND RILL EROSION FROM CROPLAND IN THE UNITED
STATES IN 2010
Run 1 Run 2
To tal
Per Acre
Total
Per Acre
(miLtons)
(tons)
(mil. tons)
(tons)
Nation
2,320
5.2*
3,537
8.3*
New England
2.6
3.3
1.5
2.2
Mid-Atlantic
123.9
10.6
121.5
10.7
South Atlantic-Gulf
465 .6
13.5
435.9
16.9
Great Lakes
50.5
2.0
83.2
3.4
Ohio
152.2
4.0
357.6
9.4
Tennessee
48.1
8.5
52.6
15.2
Upper Mississippi
284.7
4.1
617.8
9.0
Lower Mississippi
217.1
8.7
398.3
17.8
SourIs-Red-Rainy
52.1
2.7
48.0
2.6
Missouri
257.9
2.6
559.8
5.5
Arkansas-White-Red
287.9
5.5
408.8
7.8
Texas Gulf
262.8
8.6
336.7
11.8
Rio Grande
20.0
5.6
13.7
5.2
Upper Colorado
2.0
1.3
2.5
1.6
Lower Colorado
.8
.6
.3
.5
Great Basin
2.4
2.0
6.4
4.2
Columbia-North Pacific
85.2
4.9
88.2
5.2
California-South Pacific
4.4
.6
4.3
.6
Source: Runs of Iowa State University model of the U.S. agricultural
economy done for this study. The regions are the river basins draining the
48 contiguous states.
*The estimates of erosion per acre exclude 48 million acres of cropland
not treated in the ISU model. When this land is included, as it should be to
make these estimates comparable to those in the NRI, per acre erosion from
cropland is 4.7 tons in Run 1 and 7.4 tons in Run 2.
Table 6-3. SEDIMENT DELIVERED TO WATER BODIES IN THE UNITED STATES (mil./tons)
Region
1977
2010 (Run 2)
Northeast
26.6
39.2
Lake States
45.7
93.0
Cornbelt
250.9
319.9
Northern Plains
¦141.7
246.5
Appalachia
57.7
139.4
Southeast
41.9
101.2
Delta
73.8
175.1
Southern Plains
69.3
256.4
Mountain
29.5
48.3
Pacific
11.2
30.8
Nation
748.3
1,449.8
Source: Prepared for this study by Leonard Gianessi of Resources for the
Future, based on data and procedures developed in Gianessi, Peskin, and
Poles. Sediment delivered in each region Is estimated by multiplying gross
erosion in each region by estimates of sediment delivery ratios.
127
-------
Our analysis of tillage technologies led us to conclude that by 2010
some 50 to 60 percent of cropland would be in some form of conservation till-
age. Thus the model's projection of 83 percent appeared high to us. This
view was reinforced by the knowledge that the model takes account of the
long-run effects of erosion on crop yields, a characteristic favoring con-
servation tillage over conventional tillage in scenarios projected several
decades into the future. We think it probable that farmers give less weight
to the long-term effects of erosion than the model does. We were particularly
skeptical of the high percentages put in conservation tillage by run 1 of the
model in the Ohio basin (86 percent) and the Upper Mississippi basin (98 per-
cent) . Our analysis of tillage technologies suggested that in much of the
Ohio basin soils are too moist to support such a high percentage of land in
conservation tillage and much of the Upper Mississippi region (which includes
most of Minnesota and Wisconsin in addition to most of Iowa and Illinois) too
cool soil temperatures in the spring would be similarly limiting.
For these reasons we asked ISU to make a second run of the model, using
the same projections of production and land use as in run 1, but with two
changes in other respects: (1) let the model allocate production by region,
and (2) impose upper limits on the amount of land that could be in high con-
servation tillage of 60 percent in the Ohio basin and of 70 percent in the
Upper Mississippi. In all other regions the model was free to allocate land
among tillage technologies in accordance with its cost minimizing principle.
As Table 6-2 shows, the second run of the model projected substantially
more erosion, both in total and per acre, than the first run. This occurred
even though the total amount of land in crops was less than in run 1 by 17
million acres (4 percent) and the amount of land in some form of conservation
tillage was the same as in run 1, 83 percent. The difference was that the
amount of land in high conservation tillage fell from 48 percent in run 1 to
31 percent in run 2. In the Ohio basin land in high conservation tillage
fell from 86 percent to 47 percent, even though a maximum of 60 percent was
permitted. In the Upper Mississippi high conservation tillage fell from 98
percent to 62 percent, less than the permitted maximum of 70 percent.
We consider run 2 of the ISU model to be more consistent than run 1 with
our analysis of tillage technologies and our projections of total harvested
land and its regional distribution. Consequently, we consider the amounts
and regional distribution of erosion shown in run 2 of Table 6-2 to be more
As noted in the text, the model generated 17 million fewer acres of
harvested land in run 2 than in run 1. This was because in run 2 the model
was free to allocate production among regions while run 1 was constrained by
our original regional allocation. In the time between the two runs we re-
vised our initial projections of production and yields, resulting in a lower
projection of land in main crops, hay, and fallow. Our revised projection of
harvested land was almost the same as the corresponding projection in run 2:
417 million acres compared with 423 million acres. We indicated in Section 4
that we modified our initial allocations of regional production and land use
to bring them closer to the allocation in run 2. Consequently our final pro-
jections of both total.harvested land and its regional distribution are
closer to those of run 2 than of run 1.
128
-------
consistent with our projections of production and harvested land than the
run 1 estimates.
Both runs indicate more sheet and rill erosion in 2010 than in 1977
(compare Tables 6-1 and 6-2), both in total and per acre of cropland. Unless
wind erosion declines significantly, the projections imply an increase in
total erosion. We see no reason to expect a decline in wind erosion, given
our projected increases in harvested land in the Northern and Southern Plains
and the Mountain States, the regions most affected by wind erosion. Wind
erosion is particularly severe in the Southern Plains (see Table 6-1), and
our projections for that region show an increase in land in main crops of
15.5 million acres (57 percent) between 1977 and 2010.
These results strongly suggest that if production and yields behave as
we have projected them, there will be considerably more erosion from cropland
in the United States-in 2010 than in 1977, even with a significant expansion
of the amount of land in conservation tillage. We concluded earlier that
current and future costs of present amounts of erosion are cause for concern.
If erosion mounts as our projections indicate, the problem will become more
severe, especially if the situation which emerges is like that depicted in
run 2.
We have no basis for judging the precise effects of the projected levels
of erosion on productivity of the land, but they surely imply that produc-
tivity losses would be greater than those resulting from present erosion.
Nor are we able to estimate the off-farm damages of the projected .
amounts of erosion. We sdo have estimates, however, of the amounts of sedi-
ment delivered to water bodies around the country if erosion increases as
our projections indicate. These are shown in Table 6-3, with corresponding
estimates of sediment delivered in 1977. The estimates were made by multi-
1 Q
plying gross erosion in each year by estimates of sediment delivery ratios.
The estimates of sediment delivery ratios are very tenuous, so the estimates
in Table 6-3 of sediment delivered should be taken as no more than rough
approximations. Even as such, however, they leave little doubt that if ero-
sion increases by 2010 as we have projected, the sediment load in the
nation's rivers, streams, lakes, reservoirs, and harbors would increase
sharply. Indeed, Table 6-3 indicates it would almost double. The Cornbelt
would continue to contribute the greatest amount of sediment, followed by
the Southern Plains. The greatest relative increases, all in excess of 100
percent, would be in Appalachia, the Southeast, the Mississippi Delta, the
Southern Plains, and the Pacific Coast. (The absolute amount in the latter
region would be minor, however). This pattern of regional environmental
impacts is much the same as we found with respect to nitrogen fertilizers,
and for essentially the same reason. These are the regions with the largest
relative projected increases in crop production and harvested land.
19
As indicated in the note to Table 6-3, the source of the estimates
of sediment delivered and sediment delivery ratios is Gianessi, Peskin, and
Poles.
129
-------
Should amounts of sediment delivered increase as shown in Table 6-3, the
quality of the nation's surface waters surely would be worse than at present.
How serious the deterioration would be perceived to be we are unable to say,
but we suspect it would be considered of significant national concern, justi-
fying firmer measures to deal with erosion than any.previously adopted.
CONCLUSIONS
Our judgment is that the projections to 2010 of crop production and land
use do not imply an increasing problem of insecticide pollution. Indeed, it
appears that changes in the location of cotton production and innovation in
the management of pests of cotton and corn will significantly reduce the
quantities of insecticides used. Moreover, the shift toward less persistent
insecticides should reduce damages to water quality. Although these mater-
ials are more acutely toxic to humans than the persistent insecticides, man-
agement of their environmental impacts should be easier. We believe, there-
fore, that the insecticide problem will be of diminishing importance so far
as the environment is concerned.
Given present and foreseeable weed management practices, our projections
imply significant growth in the use of herbicides because of the projected
increases in both cropland and conservation tillage. Present evidence does
not demonstrate that the projected increase in herbicides applied will exact
serious environmental damage. There are gaps in the evidence, however,
relative to possible damages with present quantities of herbicides, and there
may be scale effects that would emerge with the sharply larger quantities
projected for the future. For these reasons there is no ground for compla-
cency about prospective environmental damages of herbicides.
Soil and water salinity likely will continue to be a problem in the
arid West, even though the amount of. irrigated land in the region is not
likely to grow much beyond present levels. However, presently known tech-
niques for dealing with salinity, e.g., improved water and agronomic manage-
ment, breeding more salt-tolerant varieties, appear sufficient to prevent
the problem from becoming significantly worse.
If planned measures to reduce municipal discharges of phosphorus are in
fact taken, then the projected increases in phosphorous fertilizers should
not make the eutrophication problem more serious than it is at present. The
projected increases in nitrogen in the Southeast, South, and Southern Plains,
however, are relatively large and may contribute to the build-up of water
quality problems in those regions.
The most serious threat to water quality by far, however, appears to be
from projected amounts of erosion and sediment. While we cannot attach a
value to the threat, a doubling of the amount of sediment delivered to the
nation's surface-waters surely would be cause for concern. When the effects
of erosion in reducing the productivity of the land are taken into account,
erosion, in our judgment, is easily the major threat to the nation's environ-
ment posed by the projected levels 6f crop production and land use.
130
-------
REFERENCES
Aldrich, .Samuel. 1976. "Perspectives on Nitrogen in Agriculture: Food
Production and Environmental Implications." Paper given at the annual
meeting of the American Association for the Advancement of Science,
Boston.
, 1980. Nitrogen in Relation to Food, Environment and Energy. Illinois
Agricultural Experiment Station, University of Illinois, Champaign-
Urbana.
Ayers, R. S. 1978. "A Case Study—Nitrates in the Upper Santa Ana River
Basin in Relation to Groundwater Pollution." In P. F. Pratt (ed.),
Management of Nitrogen in Irrigated Agriculture, Proceedings of a Na-
tional Conference sponsored by the National Science Foundation, U.S.
Environmental Protection Agency and the University of California.
Barisas, S. G., J. L. Baker, H. P. Johnson, and J. M. Laflen. 1978. "Effect
of Tillage Systems on Runoff Losses of Nutrients." Journal paper no.
J-8534 of the Iowa Agriculture and Home Economics Experiment Station,
Ames, Iowa, Project 1853.
Barnes, J. M., L. Heffner, R. F. Holt, R. G. Menzel and R. D. Wauchope. (no
date) "Influencing of Agricultural, Silvicultural, and Land Conversion
Activities to Help Meet State and National Water Quality Objectives."
SEA-USDA draft response to RCA. Unpublished paper. For information
contact R. D. Wauchope, USDA, Southern Weed Science Laboratory, Stone-
viile, Miss. 38776.
Cady, J. G. 1980. The Natural Background: Soil Formation and Erosion. Paper
given at a Workshop on Soil Transformation and Productivity sponsored by
the Commission on Natural REsources, National Research Council, Oct.
16-17, 1980.
Chemical and Engineering News. Aug. 11, 1980. "Agent Orange Health Issues
Raised Again."
Council on Environmental Quality. 1979. Environmental Quality. The Tenth
Annual Report of the CEQ, USGPO, Washington, D.C.
. Jan. 1981. Contamination of Ground Water by Toxic Organic Chemicals,
USGPO, Washington, D.C.
131
-------
Crosson, Pierre and Kenneth Frederick. 1977. The World Food Situation:
Resource and Environmental Issues for the Developing Countries and the
United States. Resources for the Future Research Paper R-6.
Crosson, Pierre. 1981. Conservation Tillage and Conventional Tillage: A
Comparative Assessment. The Soil Conservation Society of America,
Ankeny, Iowa.
Eichers, T. R., P. A. Andrilenas and T. W. Anderson. 1978- Farmers' Use of
Pesticides in 1976. USDA, Agricultural Economics Report No. 418.
Eijsackers, H. and J. van der Drift. 1976. "Effects on the Soil Fauna,"
Herbicides: Physiology, Biochemistry, Ecology. L. J. Andus (ed.) New
York, Academic Press.
Environmental Protection Agency. 1977a. National Water Quality Inventory.
1976 Report to Congress, Office of Water Planning and Standards, Wash-
ington, D.C.
. 1977b. The Report to Congress: Waste Disposal Practices and Their
Effects on Ground Water, Executive Summary. Washington, D.C.
Frederick, K. with J. Hanson. 1981. Western Irrigation: Its Importance to
the Growth and Environmental Impacts of U.S. Agriculture. A report to
the U.S. Environmental Protection Agency, Athens, Georgia.
Gianessi, L., H. M. Peskin and J. S. Poles. April 1980. Cropland Soil Ero-
sion and Sediment Discharge to Waterways in the United States. A report
done at RFF under contract with the U.S. Soil Conservation Service and
the U.S. Environmental Protection Agency.
Giere, J. P., K. M. Johnson and J. H. Perkins, 1980. "Is No-Till Farming the
Answer to Soil Erosion?" Environment, vol. 22, no. 6
Goldberg, M. 1970. "Sources of Nitrogen in Water Supplies," in Agricultural
Practices and Water Quality, Proceedings of a Conference Concerning the
Role of Agriculture in Clean Water, held at Iowa State University, Ames.
Greaves, M. P. 1979. "Long-term Effects of Herbicides on Soil Microorgan-
isms," Annals of Applied Biology, vol. 91, no. 1, pp. 129-131.
Hinish, W. W. Jan. 1980. "Soil Fertility," Crops and Soils Magazine, pub-
lished by the American Society of Agronomy.
Holt, R. F., H. P. Johnson and L. L. McDowell. 1973. "Surface Water Quality,"
in Conservation Tillage, Proceedings of a National Conference. Soil Con-
servation Society of America, Ankeny, Iowa.
132
-------
International Minerals and Chemical Corporation, (no date) "Fertilizer:
the Indispensable Profit Maker." A report by the Marketing Services
Department, Mundelein, Illinois.
Johnson, H. P., J. L. Baker, W. D. Shrader and J. M. Laflen. 1977. "Tillage
System Effects on Runoff Water Quality: Sediments and Nutrients." Paper
given at the meeting of the American Society of Agricultural Engineers,
Chicago, Dec. 13-16.
Luhrs, C. E. 1973. "Human Health," in Nitrogenous Compounds in the Environ-
ment. U.S. Environmental Protection Agency, Hazardous Materials Advisory
Committee, EPA-SAB-730001.
Lund, L. J., J. C. Ryden, R. J. Miller, A. E. Laag, and W. E. Bendixen. 1978.
"Nitrogen Balances for the Santa Maria Valley." In P. F. Pratt (ed.),
Management of Nitrogen in Irrigated Agriculture, Proceedings of a Na-
tional Conference sponsored by the National Science Foundation, U.S.
Environmental Protection Agency and the University of California.
Martin, W. P., W. E. Fenster and L. D. Hanson, 1970. "Fertilizer Management
for Pollution Control." In Agricultural Practices and Water Quality,
Proceedings of a Conference Concerning the Role of Agriculture in Clean
Water. Iowa State University, Ames.
Meister, A. D., E. 0. Heady, K. J. Nichol and R. W. Strohbehn. 1976. U.S.-
Agricultural Production in Relation to Alternative Water, Environmental
and Export Policies. Iowa State University Center for Agricultural and
Rural Development, Card Report 65, Ames.
Miller, D. W., F. A. DeLuca and T. L. Tessier. 1974. Groundwater Contamina-
tion in the Northeast States. Report prepared for the Office of Research
and Development, U.S. Environmental Protection Agency, Washington, D.C.
Mrak, J. M. 1969. Report of the Secretary's Commission on Pesticides and
Their Relationship to Environmental Health. U.S. Department of Health,
Education and WElfare, USGPO, Washington, D.C.
National Academy of Sciences. 1975a. Pest Control: An Assessment of Present
and Alternative Technologies, vol. I, Washington, D.C.
1975b. Agricultural Production Efficiency. Washington, D.C.
Patrick, Ruth. 1973. "Aquatic Systems," in Nitrogenous Compounds in the
Environment. U.S. Environmental Protection Agency, Hazardous Materials
Advisory Committee, EPA-SAB-730001.
Hmentel, D. 1971. Ecological Effects of Pesticides on Nontarget Species.
Executive Office of the President, Office of Science and Technology,
USGPO, Washington, D.C.
PLmentel, D., E. C. Terhune, R. Dyson-Hudson, S. Rochereau, R. Samis, E. A.
Smith, D. Denman, D. ReifSchneider and M. Shepard. 1976. "Land -¦
133
-------
Degradation: Effects on Food and Energy Resources," Science, vol. 194
(October 8), pp. 149-155.
Pimentel, D., D. Andow, R. Dyson-Hudson, D. Gallahan, S. Jacobson, M. Irish,
S. Kroop, A. Moss, IrSchreiner, M. Shepard, T. Thompson, B. Vinzart.
1980. "Environmental and Social Costs of Pesticides: A Preliminary
Assessment." OIKOS 34, no. 2, pp. 126-140.
Plewa, M. J. and J. M. Gentile. 1976. "Mutagenicity of Atrazine: A Maiz-
Microbe Bioassay," Mutation Research, vol. 38, pp. 287-292.
Science News. Sept. 1973. "The Worse for Biodegradation," vol. 114, no. 13.
Sommers, L. E. and V. 0. Biederbeck. 1973. "Tillage Management Principles:
Soil Microorganisms," Conservation Tillage, The Proceedings of a National
Conference, Soil Conservation Society of America, Ankeny, Iowa.
U.S. Department of Agriculture. 1974. Our Land and Water Resources.
. Aug. 21, 1979. "SCS Answers Critics of New Soil and Water Data."
Office of Governmental and Public Affairs, Issue Briefing Paper no. 20,
Washington, D.C.
. Feb. 1980. Basic Statistics, 1977 National Resources Inventory.
Soil Conservation Service, Washington, D.C.
. 1980. RCA Appraisal 1980 Review Draft, Parts I and II.
U.S. Department of Interior. April 1975. Westwide Study Report on Critical
Water Problems Facing the Eleven Western States.
U.S. Department of Interior and U.S. Department of AGriculture. 1977. Final
Environmental Statement: Colorado River Water Quality Improvement Pro-
gram, vol. I, Section 1. (May)
U.S. Department of Interior, California Department of Water Resources, and
California State Water Resources Control Board. June 1979. San Joaquin
Valley Interagency Drainage Program. Final Report.
U.S. Water Resources Council. Dec. 1978. The Nation's Water Resources,
1975-2000; Second National Water. Assessment.
van Schilfgaarde, Jan. 1977. "Minimizing Salt in Return Flow by Improving
Irrigation Efficiency," in J. Law and G. V. Skogerboe, Proceedings of
National Conference on Irrigation Return Flow Quality Management, spon-
sored by U.S. Environmental Protection Agency and Colorado State Univer-
sity, held at Fort Collins, Colorado, May 16-19, 1977.
von Rumker, Rosemarie, E. W. Lawless and A. F. Meiners, with K. A. Lawrence,
G. L. Kelso, and F. Horay. 19 75. Production, Distribution, Use, and
Environmental Impact Potential of Selected Pesticides. Environmental
134
-------
Protection Agnecy Office of Pesticide Programs and Council on Environ-
mental Quality, Washington, D.C.
Wischmeier, W. W. and D. D. Smith. 1965 (reprinted 1972). Predicting Rain-
fall-Erosion Losses from Cropland East of the Rocky Mountains. USDA-
Agricultural Research Service in Cooperation with Purdue Agricultural
Experiment Station. Agricultural Handbook no. 282.
Wauchope, R. D. 1978. "The Pesticide Content of Surface Water Draining
from Agricultural Fields - A Review," Journal of Environmental Quality,
vol. 7, no. 4, Oct-Dec.
Wauchope, R. D., L. L. McDowell and J. Hagen (no date) "Environmental
Effects of Limited Tillage," forthcoming in Weed Control in Limited
Tillage Systems, the Weed Science Society of America.
135
-------
SECTION 7
POLICY CONSIDERATIONS
BACKGROUND
The previous Section showed that expanding agricultural production
promises to generate significant new environmental stress, with implications
that extend beyond the farm. Policies to cope with that stress will be set
in the context of the long record of agricultural policy, but they also must
recognize changing circumstances and values outside of agriculture that
impinge upon it. A brief review of this changing context is needed.
Agricultural policy and agricultural land policy have been central to
the political and social history of our republic. The availability of free
or low-cost land was a primary impetus to early settlement and to the exten-
sion of the frontier. With land in ample supply, restrictions on its use
were not considered. Fee simple ownership and the virtually unrestricted
right to use owned property in any way desired were embedded early in the
nation's institutions and psyche. Thomas Jefferson's views on a democracy
of landowners gave intellectual sanction to this system. Early debates
over such matters as tariffs, transportation, banking and credit, policies
for the disposition of federal land, and even slavery, all reflected the
farmer's eagerness to acquire and work his own land and to sell his products
on a wider market. A largely agrarian society of free men in which wide-
spread access to land ownership played a central role was for decades our
major dynamic and was a beacon to the world.
The need to adjust to changing competitive, resource, and environmental
factors has been present from the beginning. Erosion and soil exhaustion
in some of the early settled areas of the South could not be countered by
techniques known at the time. Land was abandoned as operations, especially
plantations, moved further west. In Appalachia, similar problems often made
fanning unviable and resulted in populations stranded on unproductive land
where they remained for generations. New land to the west had better soils
and climate that also made it difficult for New England to compete, espe-
cially after improved transportation made the West more accessible. Thus,
regional shifts in agricultural production were an early consequence of both
economic and technical factors that were reinforced by policies affecting
land disposition and transportation.
By the end of World War I (WWI), the center of gravity of American
agriculture had moved definitively to the Midwest, and agricultural policy
issues had begun to assume their long-term form. Ultimately, two broad
136
-------
groups of policy issues predominated: 1) how could farmers be assisted to
hold onto their land and realize a decent income in the face of sharp price
fluctuations or chronic surpluses of production and the resulting depressed
prices? and 2) what assistance could government provide in helping farmers
maintain production on lands vulnerable to erosion and drought? Often the
two objectives merged, with fragile lands being withdrawn from crops under
government support programs to protect both the land and the market. Other
land was provided with subsidized water under conservation projects.
Prior to WWI, policy (except for land disposal policy and irrigation
projects) tended to operate at a general level through monetary, credit,
trade, and transportation measures. Thereafter a shift to the farm level
occurred, with policy now operating through government loans, support pay-
ments, farm conservation plans, and extension services. The once-independent
farmerhas become directly dependent on this policy net and very insistent
and influential in the halls of Congress on behalf of public support for the
farming sector.
Distress in the countryside, especially among grain farmers, was endem-
ic following WWI. The farming sector entered a period of stress in the
1920s and virtually collapsed in the following decade. What had been an
economic problem was now a major sociological issue as land and farmers were
compelled to leave farming while faced with poor opportunities elsewhere in
the economy. After WWII, the move away from the farm accelerated, propelled
now by a dramatic revolution in agricultural technology that increased
yields while greatly diminishing labor requirements. Unlike in pre-war
years, the swelling cities and growing service economy now could absorb the
people displaced, but land and farm commodities still remained in surplus.
During the decade of the 1970s the major human adjustment of the farm
economy was substantially completed. The number of low income farmers had
been reduced and the size of farm increased to the point where rural poverty
no longer was a dominant sociological problem. A series of events—weather,
a slower rate of yield increase, and strong foreign demand—evaporated
chronic surpluses, raised crop prices, and absorbed idled cropland back into
production. Familiarity with soil conservation techniques was now widespread
in the countryside, and conservation practices commonly were adopted when
profitable to the farmer or if sufficiently subsidized. Despite these
changes, the structure of farm programs retained much of the form acquired
during the time of surplus; those programs, especially the price support
structure, are jealously guarded by farm interests and have become reflected
in the value of farm enterprises. All are oriented toward assisting the
farmer. When they make him the instrument of conservation policy, they are
dependent on his voluntary cooperation. However, with demand for farm prod-
ucts strong again, the once most effective income support and conservation
practice—the return of cropland to cover—has lost its appeal.
Yet another important strand of agricultural policy has been the farm-,
er's continuing interest in the widest possible market for his products.
Cotton exports once sustained the South, and grain ggshipments built the mid-
continent rail system. Exports were quiescent during the protectionist
years of the Great Depression, and farm policy focused on domestic policies
137
-------
for averting surpous. With the return of foreign demand after WWII, driven
especially in recent years by prosperity in developed countries and by both
population and income growth elsewhere, farmers have regained' their keen
.interest in unrestricted exports. Government restrictions on the grain
trade are strongly resisted.
The surviving structure of agricultural policy does not seem one that
would be suggested de novo for the conditions of today. Farmers have now
become substantial businessmen attuned to international markets. They are
no more deserving of price supports than are other sectors of the economy
that must compete in world markets. Meanwhile, a new environmental con-
sciousness has been born that demands control of environmental insult. It
holds economic actors responsible for the off-site consequences of their
operations. Farm-related sediment, pesticide residues, and the like no
longer are viewed with indifference. At the same time, a new spirit of
conservation has arisen that shows concern about limits to the nation's
resources; it views preservation of the renewable productive potential of
the land as a matter of social interest that need not be left solely in
private hands, despite our long tradition of fee simple ownership. Thus,
new concerns have been introduced into the policy arena that are less farmer-
oriented. It is difficult to know which trends are ephemeral and which
will prove enduring. Environmental programs face a somewhat more skeptical
reception now than over the past decade. But concern for soil conservation
does not appear to abate very much.
Other societies sometimes have neglected soil conservation and environ-
mental matters to their own peril: failure to deal with these issues on a
timely basis has been a factor in the demise of some. However, in its long
denial of social responsibility for these matters that are of concern to all,
our own society has been somewhat of an aberration. In no small part this
goes back to our history of unfettered private ownership of land and to the
circumstances under which it arose whereby It was both an economically pro-
ductive and a politically liberating system. Carried to doctrinaire ex-
tremes, however, this viewpoint paralyzes the capacity of society to deal
with newly recognized and more acute problems of environmental control and
conservation.
There is a national interest in the maintenance of an efficient and
productive agriculture that permits moderate food prices and generates high
export earnings. Sometimes it is difficult to express that interest in ways
that are environmentally benign, for that may conflict with the private aims
of farmers. While regulation can restrict the way in which land is used by
forbidding offending practices, its positive thrust is weaker; it cannot
compel private owners to employ land in socially beneficial ways when it is
not profitable for them to do so. Therefore, a policy to control the en-
vironmental consequences of expanded agricultural production must acknowledge
both the legal standing and managerial role of farmers and must not threaten
the maintenance of a viable private agriculture. Realistically, policy also
must acknowledge the ingrained policy expectations of farmers, for these
expectations make abrupt or drastic change politically difficult.
138
-------
CURRENT ENVIRONMENTAL APPROACHES
Existing programs for soil conservation, water pollution control, and
pesticide regulation differ greatly in origin and form and are administered
by different agencies. The principal purpose of soil conservation programs
is to preserve and enhance the productive potential of the soil. At times,
such programs also have become a vehicle for farm income maintenance or for
commodity price stabilization, though that is not currently the case. The
programs, administered by U.S. Department of Agriculture, are overwhelmingly
voluntary. Water pollution control programs are administered by the Environ-
mental Protection Agency but, with respect to agriculturally-related nonpoint
source pollution, the agency delegates nearly all planning and control to
the states through its program under Section 208 of the Federal Water Pollu-
tion Control Act (Clean Water Act). In practice, 208 programs have tended
to become extensions of present soil conservation programs. Pesticides are
regulated in three ways: through the registration process administered by
EPA, through Food and Drug Administration, and through USDA rules on pesti-
cide application procedures. The other major concern of a resource and
environmental nature is the preservation of agricultural land. While USDA
and EPA have policy positions in favor of land preservation, most initiative
in this field lies with the states where it becomes entangled with other
nonagricultural land use considerations. Finally, various programs for
wetland preservation, upstream flood control, and fish and wildlife habitat
have implications for agricultural land use that may be restrictive in
character.
Are existing approaches well adapted to meet current needs and the
prospective needs suggested by our projections? In order to answer that
question each policy area must be examined in order, but it is apparent that
the disparate programs do not adequately consider their interrelationships.
For example, conservation tillage is a most promising means for reducing
erosion, but it implies greater use of pesticides and will be feasible only
if they remain available. Their use in turn may aggravate water pollution
problems and damage valued nontarget species.
Soil Conservation Programs
Three principal federal soil conservation programs operate at present.
The Soil Conservation Service (SCS) Conservation Operations Program provides
technical advice and conservation plans to farmers. The Agricultural Con-
servation Program is a means for channeling money to farmers to encourage
them to carry out conservation practices on their land. It is administered
by the Agricultural Stabilization and Conservation Service (ASCS) in cooper-
ation with local committees of farmers. The Great Plains Conservation
Program is a special effort by SCS to encourage farmers in that area to con-
tract for an agreed operational plan that will reduce wind and other erosion
from cropland and range.
The SCS Conservation Operations Program is strictly voluntary. Its
purpose is to provide technical assistance to farmers. The initiative lies
with the farmer, who must request assistance from SCS. The SCS representative
139
-------
prepares a conservation plan for the farm that incorporates a mix of farming
practices or the installation of more permanent features, such as terraces.
A review of this program by the Government Accounting Office (GAO) in
1977 criticized it as passive in not seeking out and concentrating on land
with the most severe erosion problems (GAO, 1977). Moreover, the GAO report
argued that many of the plans were too elaborate, were not used by farmers,
there was little follow-up by SCS, and that the soil loss results for those
cooperating with SCS were little different from those who did not. SCS ac-
knowledged the validity of these criticisms. But SCS plays a wider educa-
tional role in cooperating with Soil Conservation District members and in
other ways, so its impact probably is somewhat greater than the GAO report
admitted. The program has enlisted over one-half of all farm units as
cooperators. If conservation has not attainedits goals, that can hardly
be attributed to widespread unawareness of good conservation practice in
the countryside.
Soil Conservation Districts (SCDs) or related units play an important
role. These units were created under state legislation to promote conser-
vation. Their role is in all cases educational, but in some states they
have added powers of taxation and enforcement that permit them to do some
planning and to build control structures of wider benefit.
The Agricultural Conservation Program is a cost-sharing program to make
funds available to farmers for conservation practices. The program is under
the jurisdiction of ASCS and operates through state and local committees that
make the actual allocations of funds. SCS provides technical advice to the
committees and to farmers carrying out the practices. Again, the program is
entirely voluntary. The major objective of the program is to secure enduring
soil and water conservation practices that the farmer would not find profit-
able if financed from his own funds. The ASCS pleads that it has broader
responsibility for the protection of woodlands and wildlife as well as water
conservation and pollution control, so it should not be held narrowly to the
funding of erosion control measures.
In reviewing this program, the GAO found that it was not as strongly
focused on soil conservation as would be desirable; over one-half of the
available funds went for other measures such as tile drainage, irrigation,
land leveling, and liming that were primarily in the farmer's own interests.
Moreover, there appears to be a tendency on the part of local committees to
foster those conservation practices that are enduring—i.e., structural
works—over such nonstructural practices as conservation tillage or rotation
systems. The latter could be promoted by term contracts, but this device
has not been popular. As a consequence of programs designed by Congress,
the ASCS has a split personality. In one program it promotes conservation,
which might require the retirement of vulnerable acreage, while in another
it offers price supports based on historical acreage—a system that would
penalize premature land retirement.
A subsequent ASCS evaluation of the Agricultural Conservation Program
supports much of the GAO criticism of this program (USDA, 1980). Their
examination of recent experience on agricultural lands (not just cropland)
140
-------
found that the acreage on which uncontrolled erosion exceeds 5 tons per
acre per year received only 48 percent of all assistance in erosion control
practices. Conservation tillage, one of the most cost-effective practices
for erosion control, was the practice least often assisted (USDA-ASCS, 1980).
The Great Plains Conservation Program administered by SCS is an effort
to induce farmers to change cropping systems and land use to conserve soil
and water in that region. The chief practice to accomplish this is the pro-
vision of permanent cover on lands especially vulnerable to erosion. The
chosen policy instrument is long-term contracts in which the farmer agrees
to follow conservation practices in return for government payments. However,
when reviewed by the GAO, the program had been implemented on only one-fourth
of the intended acreage and had little prospect of reaching beyond half of
it by 1981, the terminal year of the program. Strong grain prices induced
farmers to return land to cultivation when contracts expired, or to decline
those offered. As with other programs, there was little evidence that
priority was being given to the most affected acres or to the most effective
practices.
Water Pollution Control
Agriculturally-related water pollution is subject to various provisions
of the Clean Water Act administered by EPA. Identifiable point sources,
such as large cattle feedlots, are controlled through a permit system that
specifies allowable operating practices. Most agricultural pollution does
not come from point sources, but rather in runoff from fields. Control of
nonpoint pollution is delegated to the states, subject to the approval of
state control plans by EPA. Under Section 208 of the Clean Water Act,
states are required to plan for area-wide waste-treatment management, and
the plan is to include "a process to (i) identify, if appropriate, agricul-
turally and silviculturally related nonpoint sources of pollution, including
return flows from irrigated agriculture, and their accumulative effects,
runoff from manure disposal areas, and from land used for livestock and crop
production, and (ii) set forth procedures and methods (including land use
requirements) to control to the extent feasible such sources." (The Clean
Water Act, Section 208 (b)(2) (F) . The state 208 planning effort thus set
in motion has been supported by federal grants administered by EPA. Com-
pletion of the plans has been laggard, but the first round is essentially
done and most plans have been approved by EPA.
While EPA provided financial assistance and administrative guidelines
for the preparation of 208 plans, they have allowed states great latitude
in their preparation. In particular, EPA has shunned the assertion of any
regulatory role for the agency, even though it has authority for it, and
it has encouraged states to favor voluntary approaches as well. The em-
phasis is on measures known as Best Management Practices (BMPs) that can
be adopted by individual farmers on their land to prevent pollution. In
most instances the BMPs prove to be familiar soil and water conservation
measures to control erosion and runoff, though there also is some emphasis
on those measures that inhibit sediment delivery to streams once erosion
has occurred. The plans, being relatively new and mostly voluntary, have
141
-------
not compiled any real record of implementation to date. However, most con-
template making use of existing administrative delivery systems for conser-
vation programs in the countryside; i.e., they will work thorugh the SCDs
and RCDs to contact farmers, and SCS personnel will provide technical help.
States may invest their own financial and administrative resources in these
programs as well, though few do so on any important scale.
In an effort to give impetus to agriculturally-related nonpoint source
control, the Rural Clean Water Program (RCWP) was established by amendment
to Section 208 to provide for a program of long-term (three to ten year)
cost-sharing contracts for the installation of BMPs to improve water quality.
The aim of this provision is to target more directly on water pollution con-
trol than is the case with other conservation programs. Responsibility for
the program is shared between EPA and the USDA, with the latter having
control over field implementation but sharing project approval and the ap-
proval of BMPs with EPA.
Funds are available to farmers who are operating under a water quality
plan approved by their SCD. Local committees determine priority among farm-
ers for funding. Areas and pollution sources being targeted are those with
the most significant effect on water quality, so the geographical scope of
this program is restricted to critical areas and sources. As of 1980,
thirteen projects were being funded across the country. For a project area
to qualify under the program farmer participation must be on the order of
75 percent (USDA-ASCS 1980 Handbook 1 RCWP).
The concept of this program is very promising. It is rather sharply
focused on areas with severe quality problems. The projects must be in con-
formance with 208 plans if they are to gain approval, and many states have
looked to RCWP as an important tool for eventual implementation of their
plans. Since RCWP provides for cultural as well as structural practices,
cost effective measures such as conservation tillage and the maintenance of
border strips can be encouraged. The program shuns measures intended to
enhance production in favor of those affecting water quality that farmers
would not undertake on their own because they are not profitable. The scale
of the program is very modest, with initial funding established at $50 mil-
lion per year. The program is referred to as experimental, and no operating
results were available at the time of writing. However, it is clear that
funding at this level will not make a major impact on farm-related water
pollution nationally.
EPA has provided 208 planning grants and guidelines, reviewed the plans
submitted by states, has made some effort to assess the character and sever-
ity of the problem nationally, and has approved lists of BMPs to be followed
in controlling water pollution. Except for this, most of the responsibility
and initiative has fallen to the states to control farm-related nonpoint
pollution. Their performance has been extremely variable but most often not
deserving of praise. Some states had to start from scratch to establish
planning staffs. In many cases they viewed the task as one imposed upon
them by federal government and it was not necessarily a response to a problem
that they felt deserved attention. Initially, there was concern that the
142
-------
federal government would use the law to gain sweeping control over land use
in the countryside. That fear has receded, but meanwhile, soil conserva-
tion—including its water quality aspects—has assumed a more prominent place
on the national agenda, and latent concern about a major federal role may yet
be justified.
The authors examined 208 planning in six representative states to learn
how it had been conducted:Arkansas, California, Georgia, Iowa, Nebraska,
and Texas. They were chosen as representative of major farming areas—not
because they are examples of either good or bad water quality or planning.
However, Iowa led the nation in sheet and rill erosion and sediment delivered
in 1977, and Texas led in both wind erosion and total erosion. Data showing
gross sheet and rill erosion and sediment delivered in each state in 1977 and
as projected in 2010 are shown in Table 7-1.
In conducting our survey of 208 planning in the six states we inter-
viewed responsible officials and knowledgeable academics in each case, and
collected documents wherever available. The survey was done in 1979-80.
Iowa is the preeminent Cornbelt state, with an intensive, productive,
and sophisticated agriculture. Corn and soybeans are erosive crops and that,
together with the rolling terrain in much of Iowa, and other factors, has
caused that state to lead all others in soil lost from cropland. Concern
about this problem brought state legislation in 1971 establishing soil con-
servancy districts and defining maximum allowable soil loss limits on agri-
cultural land comparable to SCS tolerance levels. The law has not been very
effective because the state cost sharing envisaged has been minimal. No
mandatory action can be taken against violators unless state cost-sharing
funds are available to help correct the problem. So far nearly all of the
small state appropriation has gone to voluntary participants (EPA, 1979) .
Moreover doubts have been raised about the wisdom of overly vigorous en-
forcement of the law because it would place Iowa at a competitive disadvan-
tage vis-a-vis other states (Nagadevara, Heady and Nicol).
The widespread concern about soil conservation has tended to overshadow
water quality measures and has presented some real dilemmas. As noted in
the last Section, conservation tillage reduces erosion but makes it hard to
incorporate some soluble chemicals in the soil, so they appear in runoff.
The widespread use of tile drains that carry much nitrogen leached through
the top layers of the soil further contaminates water in Iowa. Nonetheless,
most attention has gone to erosion and sediment reduction by means of the
usual conservation measures (Iowa State U. College of Agriculture, 1978).
While the Governor affirmed that Iowa had a nonpoint water quality prob-
lem, very little assessment of its severity was done. The approach was more
subjective. The aquatic-life potential of the water was taken as a standard
and then an attempt was made to identify the pollution sources that prevent
attainment of that standard or of the standard of body contact for recrea-
tional use.
With regard to 208 implementation, the plan is mostly voluntary in its
approach. The intention is to work with those who show a willingness to
143
-------
Table 7-1. ESTIMATES OF GROSS SHEET AND RILL EROSION AND SEDIMENT DELIVERY FROM CROPLAND IN THE
UNITED STATES AND SELECTED STATES, 19 77-2010
1977
2010
Run 1
Run 2
Gross
Erosion Sediment^
(000 tons)
Gross
Erosion
t
Sediment
(000 tons)
Gross
Erosion
t
Sediment
+
(000 tons)
UNITED STATES
1,925,849
765,113
2,320,210
1,001,110
3,537,031
1,486,418
ARKANSAS
46,711
22,421
36,572
17,555
68,686
32,969
CALIFORNIA
8,591
3,386
6,125
2,389
6,190
2,414
GEORGIA
42,662
14,777
83,445
29,205
92,729
32,455
IOWA
261,253
92,338
104,752
36,663
253,276
88,647
NEBRASKA
117,792
46,871
49,100
19,640
130,854
52,341
TEXAS
99,546
52,539
324,924
172,210
392,971
208,275
Sources: *USDA, February 1980, Table 16.
"f*
Provided by Leonard Gianessi, RFF. Runs 1 and 2 correspond to those
discussed in Section 6. See especially Table 6-3.
-------
participate rather than to apply pressure on the worst polluters. Any man-
datory action is to be left to the conservancy districts. Great hone was
placed on RCWP funds at the outset—a hope that now seems largely misplaced.
While Iowans seem acutely aware of water quality problems stemming from
agriculture and are disposed to act on them, elsewhere there often has been
much less attention to the problem. In neighboring Nebraska, where an ex-
panding irrigated agriculture faces rising energy costs, there was much less
attention to water quality and erosion. While a 208 plan was under prepara-
tion, the effort encountered great skepticism and frequent questions about
the need for it.
Perhaps because of that, unusual emphasis was placed on public parti-
cipation. This even extended to problem identification. Local advisory
committee members were invited to vote on which water quality problems they
considered deserving of highest priority. They rated soil erosion as the
top.priority problem, center pivot irrigation on marginal lands as second,
and leaching and runoff of fertilizer and chemicals as fourth and fifth.
This hardly impresses as a very scientific procedure, and the planning
agency acknowledged the need for further refinement of the list. Discus-
sions with academic experts supported the importance of erosion and leach-
ing, but there was agreement that sediment is not the problem in Nebrasks
that it is in Iowa. Many people, including some of those responsible for
208 planning, saw little need for planning, injart-Jjecausejnaj2l2<-smali_>i52jralo
sjreams have few beneficial uses in any case. However, contamination of
groundwaterwasacknowledge<^^o^K^J^robTeminjastate^that>i>relies<|ver£
heavily on underground suppliesT^^^
In view of widespread doubts about the need for 208 planning in
Nebraska, planners firmly intend to stress voluntary measures for implemen-
tation. Those which would be in the farmer's best interests will be pro-
moted. Thus, control of irrigation runoff where nutrients are applied with
the water can save the farmer money and should appeal to him. Contour farm-
ing and chisel plowing already are accepted in many parts of the state.
Nebraska is the site of one of a half-dozen USDA-EPA model implementa-
tion projects (MIPS) where a concentrated attack on rural water quality
problems is being attempted on a limited acreage. The problem confronted
in this instance is severe erosion from straight row corn cultivation in
hilly terrain. Expensive and heavily subsidized terraces are being built
on some farms to the neglect of more cost-effect contouring and conservation
tillage. Farmers in the area are not strongly motivated to arrest erosion
because they have deep soils. Even the water quality benefits seem uncer-
tain. for the small local streams have limited use at best.
what the projectwiii aemonstrate. wni-Le such concentrated effort may
succeed technicalT^""Tt™^^To^Tostly for widespread application.
Texas attempted a very decentralized approach to 208 planning, as
befits a state of its size and diversity. The state's Department of Water
Resources has primary responsibility for problem identification, which it
undertakes by river basin. Although streams in Texas generally carry a
heavy sediment load, the state has no standard for sediment and does not
145
-------
treat it as a pollutant. Sediment is attributed to natural conditions and
is not thought to be closely related to agriculture. In the High Plains area
there are no streams, and drainage is to evaporative playas. Sheet and rill
erosion is not severe on the rolling plains and is moderate on the black
plains; there seems little evidence that it affects productivity signifi-
cantly in those areas.
Erosion in Texas is mostly by wind, and the amount (14.9 tons per acre
of cropland per year- Table 6-1) is well in excess of "T" values set by the
SCS. Wind erosion evidently is not viewed as a serious threat to water
quality; in any event it does not fall within the purview of 208 planning
in Texas, or anywhere else. To the extent that wind erosion in Texas is'ad-
dressed it is primarily through the SCS's Great Plains Conservation Program.
The limitations of that program, as assessed by the GAO, were pointed out
above in the section on soil conservation programs.
Water pollution by agricultural chemicals is not regarded as a major
problem by 208 planners in Texas. Some leaching of chemicals to groundwater
is noted in the winter garden section of the state. In the coastal rice
growing areas, a heavy rain just after the application of pesticides may wash
enough into bays and estuaries to affect shellfish, though this point is
disputed.
The 208 screening process in Texas examined over twenty areas where
agriculture was suspected of contributing to failure to meet stream standards,
but only three or four involving arsenic, nitrate concentrations, and fecal
coliforms were finally attributed to agriculture. The State Soil and Water
Conservation Board, an overgroup of conservation districts, has responsibility
for developing plans to cope with the identified problem areas. They in turn
work with local conservation districts in that process. Land use surveys and
the designation of appropriate BMPs are the basis for plans in identified
problem areas, and the plans were seen as vehicles for seeking RCWP money for
Texas. The conclusion must be that most Texas agricultural land is not af-
fected by 208 planning and the state has minimal enthusiasm for the process.
Normal soil and water conservation programs continue, with stress on wind
erosion in the High Plains and on contour plowing and water retention in
other parts of the state.
Arkansas planners put their main emphasis on developing an information
system that should allow them to identify sources of nonpoint pollution.
They sought to develop detail on gross erosion and to apply specially tail-
ored sediment delivery ratios for each small watershed. Nutrient and pesti-
cide loads were derived from the recommended application rates for these
materials. Identified problem areas are to be more intensively studied.
Arkansas has much high quality water in the northwest part of the state,
where agricultural pressure on water quality is lower than in the more in-
tensively farmed southeastern delta area.
Problem identification appeared to exhaust the energy of the planners at
the time their state was examined. No statewide plan for implementation was
contemplated and they foresaw little other than an educational approach with
voluntary compliance as likely. It is conceded that streams in rice growing
146
-------
areas do not meet standards and are not likely to; little use is made of them
for fish or recreation, so there is a tendency to accommodate the convenience
of farmers.
Georgia's approach to 208 planning has been extremely cautious. The
responsible state planning agency notes that there has been little stream
monitoring in agricultural areas, so they cannot document the link between
water quality and agricultural activities. Instead they have identified
potential problems by assembling data on land use characteristics by county
and watershed. From this an erosion potential was computed and a sediment
delivery ratio applied so as to rank counties by sediment delivery potential.
Sales of pesticides by county were used as an indicator of that problem. The
inadequacies of this approach are obvious to planners, and they propose to
study in more detail over the next three years the links between potential
problems as identified by land use and actual problems seen in water quality.
They see sediment as their primary pollutant.
Implementation of the Georgia plan will be left to an overgroup of soil
conservation districts who propose to use education and demonstration projects
to encourage compliance. While the state has authority to regulate, it does
not feel that it has established a basis for doing so or that there would be
political support for it. Indeed, throughout the Southeast, while other
states have been more specific in identifying problems, nowhere is there
support for a regulatory approach. Farmers are not convinced that the prob-
lems are clearly enough identified and serious enough to warrant it, and
state officials appear to agree. Established state stream standards may be
met in Georgia, even though the larger goal of "fishable and swimmable" is
not.
California has a long history of water planning, including attention to
water quality. The state's agriculture and physical conditions are too
diverse to allow a single approach to quality management. Much agriculture
is conducted in metropolitan areas where its environmental effects are sub-
merged with those of other nonpoint sources. Intensive irrigated agriculture
predominates. While erosion is severe on some hillside orchards and grain
fields, much of the land in crops is flat, and erosion is not the major prob-
lem in California. Heavy use of agricultural chemicals and concentrated
livestock operations in some areas have threatened surface and groundwater
supplies, and the state's long history of irrigation has created salinity
problems, especially in some fertile areas of the San Joaquin Valley.
Planning under Section 208 in California has consisted of updating
already existing basin plans. This is done largely by the state's regional
water resources control boards. No single document summarizes their work.
The boards have responded to local concerns. Implementation again is left
to local conservation districts in the first instance, but the state seems
much more willing and able to intervene than in the other states examined.
Under the state's Porter Cologne Act they have considerable regulatory author-
ity, many counties have local statutes that can be invoked, and state funds
are available to finance water control programs (John Muir Institute, 1979) .
Moreover, agriculture in California is highly organized, and farm groups are
147
-------
willing to bring pressure on identified offenders in hopes of averting more
general regulations affecting everyone. As a result, there was a greater
sense of ability and willingness to cope with farm-related water pollution.
However, state officials and others attribute little of this to 208 planning;
rather it is considered to be the product of state efforts that antedate and
go beyond 208.
A review of 208 planning on the ground frequently disappoints the expec-
tations generated by federal law. Nonetheless, one should not discount what
has been accomplished. States have been forced to face farm generated water
problems explicitly and to build planning and administrative capacity to deal
with them. This can be an important step if the problems become more aggra-
vated, as our projections suggest they will. The states have had to establish
some kind of identification and priority system, however inadequate, and to
give evidence of legal authority to act where they deem it necessary. The
result has been educational for the states and has led a few to take state
initiatives to correct their problems without waiting for federal money. The
subtle hint of enforcement that underlies the law has made states, farm orga-
nizations, and farmers more self-conscious about the need to act on their own.
And the bait of RCWP funds has induced serious planning for some key problem
areas that may create its own momentum for action, even though federal money
is scarce.
Pesticide Controls
The use of pesticides in agriculture is regulated to prevent the con-
tamination of human foods, to safeguard the health of agricultural workers,
and to protect nontarget species (including man) from excessive damage. EPA
has primary responsibility for pesticide controls, but other agencies, es-
pecially the Food and Drug Administration and USDA, have assigned roles, and
states may also regulate to stricter standards if they choose.
Residues of pesticides in food may be directly toxic or they may be
suspected as carcinogenic, mutagenic, or teratogenic. Acutely toxic doses
would not be expected on food. But because some pesticides are persistent
in the environment and are concentrated by biological processes and may
accumulate in certain tissues and organs, the possibility of toxicity is
present at very low dose rates. Moreover, chronic low level exposure to
those chemicals that may be carcinogenic raises all of the murky issues
ahout thresholds and the validity of animal experiments or epidemiological
data that so bedevil all discussions of the etiology of cancer.
Responsibility for setting tolerance levels in food lies with EPA.
They establish guidelines for the amount of each pesticide that is allowed
in different foods. Where processing of raw products reduces the concen-
tration of pesticide, attention goes to the processed product. If there
is evidence that pesticides have not been used in accordance with the label,
action may be taken against treated food products even if no tolerance
levels have been exceeded. Actual enforcement of these regulations is left
to the FDA, which monitors food products about the country, and to the USDA,
which has responsibility for meat and poultry. Since agricultural commodities
148
-------
that do not meet standards are essentially unmarketable, farmers have a most
powerful incentive to avoid contamination.
The heart of federal regulation of pesticides lies in the registration
and classification process under the Federal Insecticide, Fungicide, and Ro-
denticide Act (FIFRA) as amended. Registration provides information about
the contents and effectiveness of the material and specifies the uses for
which it can be sold. In addition, EPA may classify pesticides according to
whether they are for general or restricted use. In the latter case, they can
be restricted to application by certified applicators only. In combination,
these provisions are the major protection to man and to other nontarget spe-
cies from pesticides in the environment. Human occupational exposure is
further protected by rules of the Occupational Safety and Health Administra-
tion (OSHA). States also play a role in regulating occupational exposure or,
in some circumstances, in registration.
The law has proved difficult to administer. The number of compounds on
the market is so large (about 35,000) that it has defied attempts at careful
evaluation of those already registered (NRC, 1980). As a consequence, EPA's
power to suspend or cancel existing registrations has become the focal point
of efforts to restrict use. Controversies over individual compounds and
classes of pesticide abound, and the scientific basis for decision often is
ambiguous. Nonetheless, Important agricultural pesticides, starting with
DDT, have been cancelled, suspended, or restricted in use.
Chemical companies complain that the cost of developing, testing, and
registering or defending a new compound frequently is prohibitive. Since
pests tend to develop resistance to insecticides, it is important that suc-
cessive generations of pesticides become available. Any system of pest
control, including IPM methods, will continue to make use of chemicals, even
if in smaller amounts. This problem presents a real dilemma for policy.
EPA has considerable latitude to interpret the law and increasingly they
have used risk/benefit analysis as a judgment tool (NRC, 1980) . It is recog-
nized that successful pest management cannot be conducted on a no-risk basis.
Progress has been made in the way in which chemicals are used, not only
because of the law but also because of greater environmental awareness.
While the standards for the certification of applicators have not been
stringent, a more responsible attitude toward pesticide application has
spread to farmers and other users. And ever-vigilant citizen environmental
groups have channels for complaint against careless users.
Modern agricultural technology has evolved into a dependence on pesti-
cides that is not easily controlled by regulation. Some otherwise benign
practices such as conservation tillage increase the need for pesticides.
Technologies that would reduce that need can be designed, but generally they
require a considerable lead time for development and their development may
not offer the same commercial possibilities that broad spectrum chemicals do.
Regulation to restrict present chemical use might stimulate alternatives if
they were profitable for private firms. Since most alternatives—resistant
varieties, attractants, pathogens, cultural practices, etc.—do not offer
149
-------
large-scale commercial opportunities, their development most often depends
on publicly supported research. The regulatory program of EPA is not well
coordinated with the provision of such alternatives and is weakened by their
absence. Clearly, there is need for a longer seeing government strategy in
this field.
We conclude that our projections of output and resource use threaten to
generate significantly greater environmental stress than current policies
have been asked to cope with. The added stress will occur in geographical
areas where current programs are weakest and in those areas will show up in
the form of erosion losses and attendant water pollution problems. However,
tension between the private objective of increased current income and the
social interest in maintaining the productive capacity of the soil poses the
most difficult problem for policy. It is important to consider alternatives
to those policies now in place.
ALTERNATIVE POLICY APPROACHES
The environmental consequences of expanding agricultural production
need not be met by environmental policies alone. A whole range of govern-
mental policies do affect or could affect the scale, location, and conduct
of agriculture, with implications for its environmental results. Our
analysis suggests that the preeminent environmental threat from agricultural
activities is not narrowly environmental at all—rather it is the long-run
threat to resource degradation from erosion and abuse of the soil. If agri-
culture cannot be conducted in a manner that holds this loss to acceptable
levels, then we are producing beyond sustainable capacity and may wish to
consider policies that restrict output. Restraints affecting demand would
be one course, although our projections do not indicate any near-term need
for this. There also appear to be ways to increase output while controlling
soil loss and abuse, but these may involve more intensive use of the best
lands and therefore imply other environmental pressures. Such measures
would likely change the character and location of production and would affect
both individual farmers and producing areas quite differently.
Very likely demand will be left to find its own level and attention
will go to better coordination of production, conservation, and environ-
mental policies to support one another. Water quality may be made a more
explicit concern of policy, and conservation and environmental policies
will be more vigorously pursued with programs focused more sharply on prob-
lem areas. Other agricultural programs may be asked to lend support to
these objectives. The way in which farming is conducted then may receive
more attention than its scale or location, but the result could still be
some reduction in output from what it otherwise would be.
Demand and Soil Degradation
In contemporary America agricultural output expands in response to
demand—to sales of farm commodities at attractive prices—and expanding
output increases environmental pressures of all kinds, but especially the
threat of erosion. This demand-driven output differs from that of decades
150
-------
past when individual farmers felt compelled by falling income to produce more
but, recognizing the futility of that, collectively were willing to restrict
output in return for income support. Conservation measures that idled land
or converted it to less intensive use in that context were consistent with
the policy of dampening output and removing resources from agriculture in
order to improve the economic position of those operators who remained.
Now the pressures are all in the other direction: farmers see their
welfare best served by access to growing world markets (where their price
must be competitive) and by measures that allow them to expand acreage or
intensify the use of presently cropped lands to serve those markets. Past
conservation policy was based on compensating farmers for production re-
straint, but it acted in a depressed market. In current markets, production
restraint is both unpopular and costly. Public compensation in return for
production curbs could be prohibitively expensive if it is to be effective.
If our soil resource is sufficiently threatened by the accommodation of
high demand—if it leads to production beyond long-term sustainable capacity—
then measures to restrict demand could be one response. Even though we do
not think that severe measures of this sort are justified at present, it is
useful to consider what might be done.
The most obvious candidate is exports. The United States exports over
one-third of its grain and aupplies about 60 percent of world trade in grains.
No one can accurately foresee the future, but if the United States maintains
its share of world markets, as our projections assume, then most of the large
domestic increase in production will go into exports. Restricting exports
would reduce output and greatly ease the pressure on U.S. agricultural land.
To do so would of course be very unpopular with domestic producers and
counter to the U.S. policy of promoting freer trade. It also would antago-.
nize buyers whose amity we cherish. In a world short of food, it might be
considered an inhumane policy. It also would be argued that we need the
exchange generated by farm exports to finance petroleum imports. In response,
it could be argued that the degradation of a renewable resource—land--in
order to satisfy a bloated and ephemeral demand for oil is a bad trade and
that energy conservation or domestic fuel alternatives should be sought in-
stead. Restrictions on exports could be managed so as to accommodate favored
friends and customers, though not without damage to the world's trading
system.. Humanitarian responsibilities are not unlimited, and, in a world of
national sovereignties, every nation must look after the health of its own
renewable resources, since no one else can. Yet restrictions on trade surely
do not deserve early attention as a conservation policy. Many other policies
to control erosion that are consistent with high output should be explored
first. If the problem becomes acute enough to warrant restrictions on trade, ,
they could be instituted quickly.
Nonetheless, it seems likely that American agricultural products are
sold too cheaply on world markets and, therefore, in greater quantity than
might otherwise be the case. Both the subsidies that abound throughout U.S.
agriculture and the environmental costs of production are not being reflected
in price; therefore, the United States subsidizes foreign buyers, including
151
-------
both developed countries and East Bloc adversaries. The increased domestic
environmental damage and soil degradation that result are real costs paid by
current and future generations of Americans. In principle, there can be no
objection to policies that correct for this underpricing or subsidy to con-
sumers of agricultural products, especially of those foreign buyers who bear
none of the domestic fiscal or environmental costs. The preferred solution
would be to root out the subsidies at the source or to reflect all external
costs in the price of agricultural commodities. However, that would be a
major undertaking—the present structure of agricultural policies is simply
too deeply embedded in our system. An export levy might be a more feasible
way of attacking the foreign component of the problem, but it carries the
disadvantage of penalizing those producers who may not use subsidies or cre-
ate environmental problems. Politically, and perhaps constitutionally as
well, it also would be difficult to accomplish.
Export demand also may be limited by fostering production abroad. To
the extent that demand is from developing countries and our concern is human-
itarian, there should be no objection to policies that assist them and spare
our own soil resources. This would involve transferring and developing tech-
nology suitable for their conditions and assisting in the development of new
land and water abroad. By creating the capacity for others to help themselves
we respond to the themes of self-reliance and independence now so strongly
asserted in the developing world.
Domestic demand is not a dynamic force for agricultural expansion, but
its composition could be altered so as to reduce pressure to produce by
shifting the diet away from grainfed meat animals. To do so need not result
in any loss of nutritional adequacy and might even produce health benefits.
There is growing evidence that meat eating on our present scale is not neces-
sary for a healthy diet and may even be harmful to it. The confirmation and
publicizing of this evidence could significantly alter our diet and demand
for grain. Recent price increases for feed grains, as they become reflected
in the price of meat, also serve to curb the domestic appetite for meat.
Finally, it should be possible ,to spare cropland by making better use of
forage resources and unconventional feeds. Improved management of public
rangelands is one obvious possibility. At the same time, much agricultural
and forest biomass not previously used by animals is proving adaptable as
feed.
Soil Loss and Social Responsibility
Most farmland is private property. So long as the farmer does not use
it in a manner that harms others, should he be free to use his property as he
chooses, even including the destruction of its productive potential? Note
that we do not ask that question about many kinds of property. A manufacturer
can allow his plant to go to ruin if it suits his operating strategy, and a
skilled professional trained at public expense may allow that investment to
go unused. On what grounds, then, can the public intervene when farmers find
it advantageous to treat their land assets as depletable?
152
-------
The narrowest argument is one of market failure—that there may be
differences between private and social time horizons and between private
and social discount rates. Studies have shown that farmers have little eco-
nomic incentive to control erosion and that they would need very long time
horizons and low discount rates for such control to be economically rational.
For example, a study by Seitz, et al. shows that conservation practices in a
Cornbelt study area would not be rational, even at zero discount rates, un-
less the farmer's time horizon was forty years or longer. Even a modest 5
percent discount rate extends this period to sixty years (Seitz, et al.,
February 1979). Given realistic private discount rates, the farmer is not
likely to act, and over a 100-year period the above authors found that over
half of the area studied would lose all topsoil. Thus, purely private moti-
vations may not surface to protect society's interest in soil conservation.
Society may choose to operate with an infinite time horizon (we do hope
to endure as a society) and, lacking visible substitutes for soil, may deter-
mine to manage the land as a perpetual rather than a depletable resource..
The logic for doing so is that (unlike with depletable energy or mineral
resources) the option exists and that future generations should be given
access to this resource equal to our own. Note that this becomes an ethical
rather than an economic choice. Thus, no discounting is appropriate to such
a decision.
Others might reject a noneconomic rationale but still concede that the
enormous uncertainties concerning future demand for land and the irreversible
character of soil loss (within an economic time horizon) argue for a very
low social discount rate that would justify conservation measures. Of course
uncertainty applies to all aspects of the problem, and it is conceivable that
future technical progress may diminish the importance of the soil resource.
Sufficient confidence that "something may turn up" would be a basis for
treating soil like any other input to be used according to current market
dictates. Nonetheless, there is a broad public commitment to the idea of
soil conservation, and few would be comfortable with a pure market solution
that treats soil as a minable resource (Harris, 1980).
A society with an infinite time horizon has an interest in restraining
the loss of the soil's productive capacity, for that loss imposes a cost on
future generations. But is there any basis for making the cost of preven-
tion a responsibility of current private holders when they impose no current
off-site costs on others? The legal foundation for restricting the use of
property is the need to protect public health and safety, but it stretches
the police power severely if it is asked to regulate matters that are so
remote in time and that affect no identifiable current victims. To the
extent that such problems are dealt with at all, it is through an implicit
intergenerational bargain involving the whole society rather than individual
actors. For example, the current consumption of exhaustible resources that
may deprive a future generation of their use is not something for which
each present individual is held directly accountable. Rather we view
broadly borne restraint and social investment in resource-replacing tech-
nology as the way to meet our obligation to the future. Likewise, if we
wish to protect the rights of a future generation in a productive soil, the
153
-------
equitable way of meeting the current costs of doing so would be to make them
broadly borne.
The farmer has both operational control and constitutionally protected
property rights on his land. So long as damage is confined to his own acres,
it is not clear that he can legally be compelled to correct it. Further, it
might appear inequitable for society to require that its interest in preserv-
ing the soil be accomplished at his particular expense rather than being
broadly borne. From an equity standpoint, it appears that the farmer should
be compensated for his expenses in income forgone in return for achieving
some socially desired level of soil conservation that lies beyond his own
best interest. This principle is even more important because of the farmer's
operational control over the land and the fact that he must be the instrument
for effecting whatever measures are taken.
Soil Conservation Policies
The accomplishments of the present system should not be denigrated.
Farmers are well informed on what can be done to limit soil loss. They
appear to have a decent appreciation of what conservation steps are in their
own best interests and have gone far toward applying them. Allegiance to a
conservation ethic is widespread; under its influence many farmers undoubt-
edly undertake conservation measures not justified by private economic cal-
culations. In effect, they pay a self-assessed tax toward a public goal.
Other farmers have accepted contracts and installed conservation practices
on a cost-shared basis, thereby helping to restrain soil loss. In this case
they may also have enhanced the value of their own property at public ex-
pense—an outcome that can result from any public subsidy to private action.
Under a low or moderate demand scenario, current policies might suffice
to achieve a socially acceptable level of soil loss, especially if they were
targeted better on key areas. However, the demand scenario we have projected
carries the risk of soil losses far higher than at present. The public is
not likely to accept the 84 percent increase in soil lost from cropland or
average per acre losses of over 7 tons per year from the expanded area in
crops that appears in our run 2 (see Table 6-2) .
Whether or not the pressure becomes this acute, existing programs can
be better focused. The GAO study comments are especially applicable here
(GAO, 1977). Detailed farm conservation plans are not needed in many cir-
cumstances, yet SCS spends an inordinate amount of time on them to the
neglect of more critical problems. The passive approach of SCS means that
scarce manpower resources go to those requesting help rather than to those
who need it. Meanwhile, because ASCS allotment programs are based on his-
torical acreage, they encourage farmers to plant crops on land that should
be in other use. The GAO report recommended that SCS systematically seek
out farmers with the most critical erosion problems, tailor its conservation
plans to their major problems without being overly elaborate in the plans
prepared, and coordinate with other agencies to relieve program conflicts.
Likewise, the GAO criticized the lack of priorities in assigning ACP money
by area and form.
154
-------
Not all of the GAO criticism seems well directed, however. While the
government should not be paying for measures that the farmer would find
profitable to increase his productive capacity, from a social standpoint,
why does increasing capacity differ from preserving capacity for which we
so readily accept a government role? Perhaps the criterion for government
support in either case ought to be whether it is profitable to the farmer.
If not, then we should aim at supporting those lands and measures that offer
the highest favorable social benefit/cost ratio for public investments at
whatever discount rate, if any, is selected. In particular, short-term
measures that are highly cost effective in controlling erosion should not
be dismissed in favor of longer-term structural measures that are less so.
The present system has had only moderate success in dealing with exist-
ing pressures on the land. It contains a social element in appealing for the
practice of good husbandry, but mostly it has relied on the farmer's acting
in his own interest and has provided him with information and technical sup-
port to do so. In addition, it has included subsidy elements in the form of
cost sharing for installing practices that are not justified by the farmer's
private interest. Fine tuning of this system would be a help, but when com-
modity prices make it attractive to abandon conservation measures, it is
difficult to resist the pressure. The experience with the Great Plains pro-
gram cited earlier is the best evidence of this, and elsewhere the widespread
return to crops of acreage once idled by conservation programs is additional
evidence. If a high export scenario prevails, more compelling inducements
to. maintain conservation practices will be required.
The draft RCA study establishes targets (essentially the achievement and
maintenance of "T" values) and lists alternative strategies for achieving
them (USDA, RCA II, 1980). Since the export projections used in that study
are lower than ours, their policy recommendations are directed at a less
acute problem. Thus, while their alternatives of organizational reform and
reassigning responsibilities to make current programs more effective could
help, the crunch comes in the choice of devices to induce compliance.
Inducements may be either positive or negative—ri.e., desired behavior
may be rewarded or undesired behavior penalized. One proposal that has re-
ceived much.attention is the idea of cross compliance (Benbrook, Journal of
Soil and Water Conservation. 1979). In negative form this would penalize
farmers who fail to employ good conservation practice by withdrawing from
them the benefits of other federal programs, such as price supports. In
positive form it would offer favored treatment in benefits from these pro-
grams to farmers using good practices. Farmers prefer the second approach
and have been highly critical of the penalty version. While that preference
is understandable, it is hard to see why fanners should be granted a vested
right to public benefits when they ignore public policy.
Yet the idea of cross compliance has several other drawbacks. Not all
farmers use the benefit programs, so the reach of this device is incomplete.
It is also unselective and would not focus on those lands most in need of
attention. Moreover, in a high output scenario, federal price support pro-
grams would become of less interest to farmers at the very time when erosion
control becomes a more acute problem. Finally, from the standpoint of social
155
-------
strategy, it may be unwise to tie a long-term concern with soil erosion to
an array of agricultural policies that may not deserve continued support on
their own merits. To do so would tend to embed those policies further into
the system at the very time when they otherwise might be modified.
Taxation can be used either as a reward or penalty. Property taxes
could be lowered for those following good practices, for example. This would
require state/federal cooperative arrangements to encourage states to parti-
cipate. But manipulation of general purpose taxes, for special purposes,
while widespread, should not be encouraged. A punitive tax on socially
undesired behavior could be quite effective in theory and would have certain
efficiency benefits. Thus, a soil loss tax would provide powerful incentives
for the farmer to employ conservation measures while allowing him full lati-
tude to do so in the next most efficient way (Seitz et al., 1979). It is
objected that such a tax would be difficult to administer, though it is hard
to see why it should be more so than a cross compliance scheme. The tax
could be made operative only when losses exceed the "T" level. Note that
a soil loss tax, if unrelated to sediment delivery beyond the farm, places
the burden on the farmer to achieve the social objectives of preserving his
land's productive capacity. Therefore, while the tax may be employed as an
incentive, it could logically be combined with subsidy programs that relieve
the burden on those willing to cooperate.
Present subsidies for conservation practices are in the form of cost
sharing programs. Being voluntary, they do not attract all farmers and
perhaps not those with the worst problems. Since the farmer still must pay
some part of the cost, he has little incentive to act unless the measure
increases his capacity in proportion to his own investment. Further, he is
tempted to install those practices for which money is available, even though
these may not be the most efficient ones in his case. As a consequence, the
public gets poor value for its money.
A system of contracts between farmers and government to reduce soil loss
has much appeal. It would allow the program to target on those areas and on
the specific acres that most deserve attention. The contract could be writ-
ten to specify performance, allowing the operator to arrive at the most
efficient means of achieving it. Good cultivation would not be at a disad-
vantage in this sytem; thus the public not only would get what it pays for
but would not pay more than necessary. As the RCA study points out, the
contracts could be flexible, with provisions for the owner to "buy out" if
market conditions change, and he could trade off other conservation measures
for those he wishes to abandon if the government agrees (USDA, RCA II, 1980).
A possible weakness is that the fanner might be able to exact payments on
land that he would not crop for his own reasons, but that defect has been
present in past conservation programs. Administrative costs are expected
to be high.
The Great Plains Conservation Program operates on a contract basis but,
as was noted by the GAO, it has not enlisted the desired participation rates
or been able to restrain the reversion of grassland to crops at the end of
contract periods. Evidently contract prices must be realistic in meeting
the cost to the farmer of doing society's work for it, yet they should not
156
-------
exceed the value of the social benefit derived. If the farmer is fully com-
pensated for all costs, including opportunity costs, there is no reason for
him to resist the contract, and the same applies to renewal if the contract
reflects the changed conditions prevailing at that time.
Nonetheless, the program need not be entirely voluntary. Where reason-
able contracts are offered, the operator's defense for declining to take
conservation measures vanishes. If he still resists, he could fairly be
coerced, either by tax or by regulatory measures. A soil loss tax paid by
those who decline to contract would encourage participation. It seems pref-
erable to regulation, for it gives the operator yet one more chance to do the
job in his own way while still protecting the public interest.
The threat of regulation to encourage farmers to contract is consistent
with protection of the public interest. Major reliance on regulation alone
is apt to disappoint, however, for it is very difficult to compel a farmer to
undertake specific measures on his own land. Moreover, if the purpose is
primarily conservation beyond that which is profitable to the farmer, then
the equity of imposing the costs on him is questionable when damage is con-
fined to his own land. Thus, regulation appears better suited as a backstop
to a system of contracts, but it would appear inferior in that role to a soil
loss tax.
Any subsidy scheme to induce good soil conservation is premised on ac-
knowledgment of the sanctity of an owner's property rights. In effect, it
concedes his right to use his property in a socially undesirable way unless
he is offered compensation to refrain. This is a familiar but not unchal-
lenged position. Where property is not involved, we do not allow uncon-
strained pursuit of private purposes without regard to social welfare. It
is troublesome to contemplate a society where antisocial behavior is re-
strained by a complex system of compensating payments in perpetuity. But a
purely voluntary ethic does not work or works very unevenly. The legal right
of an individual to dissipate his capital (soil) in favor of current returns
may come into basic conflict with society's preference (or obligation?) to
maintain the resource base intact. This entire dilemma could be avoided by
acquisition of certain property rights, but if we want to "take" some aspect
of property rights, then it should be done cleanly, openly, and once for all.
A contract system does not by itself ensure that an optimum amount of
erosion control is purchased by public funds. In some cases the cost of soil
conservation may be so high that even from a social accounting standpoint it
is rational to mine the soil rather than preserve it. Before conceding that
to the operator, he should be made to face the off-site consequences of such
action.
Water Pollution Control Policies
The preceding section has dealt with policy to control soil loss so as
to maintain productive capacity. Such policy confronts a landowner who can
argue that his property is his to manage as he sees fit so long as he does
not currently harm others; a legitimate public interest here must accommodate
157
-------
equally legitimate property rights. But if the owner's actions do harm
others, then it is usually accepted that he is accountable and may be re-
strained by appropriate measures. Farm-related water pollution fits the
latter case. Since much farm-related water pollution is entrained with
sediment, the two problems of erosion and water quality often are treated
as one, though the distinction can be important in designing program emphasis
or in assigning cost incidence.
Throughout the private sector, measures, for environmental amelioration
have been made the responsibility of the offending party. Auto manufacturers
must design cars to meet standards at their own expense, utilities pay to
control emissions at standard levels, and, best available control technology
is required of manufacturers discharging to streams. The final incidence of
these costs depends on market structure, but the costs do generally become
reflected in prices paid for the products and, therefore, affect production
and consumption decisions.
Nonpoint sources other than agriculture are subject to a variety of con-
trols, but the agricultural sector generally has been accorded more lenient
treatment. It is hard to understand on what principle this should be so.
Sometimes it is argued that the costs of such measures are difficult to pass
on to consumers because markets are competitive and therefore costs come to
rest on producers. What kind of an argument is this? It only serves to per-
petuate the problem. If costs were promptly reflected in decisions on how
much, what, and how to produce and ultimately in commodity prices, this
would help to reduce farm-related pollution.
Administratively and politically, however, it is very difficult to im-
pose both the cost and the responsibility for control on the farming sector.
The number of units to be controlled is large, data on effluents from a given
farm are lacking, and the consequences of their discharge hard to establish.
Moreover, EPA has no bureaucracy in the countryside and would not be welcomed
there, and USDA, which does have what they call "delivery systems" is loath
to be perceived as regulator rather than servant of their clientele. Exist-
ing programs for farm-related water pollution control are mostly based on
education and voluntary measures, with the latter sometimes induced by cost
sharing. Indeed, the cooperative and voluntary tradition employed by insti-
tutions in touch with farmers is so much a part of farmers' expectations
that abrupt change seems quite infeasible. Selective regulation can be
admitted—in some states where problems are acute local authorities can act
through regulation—but the enforcement of generally applicable rules is
much harder to envisage.
The RCWP projects now getting under way represent a concentrated attack
on some of the more acutely affected areas, albeit with the usual congress-
ional sensitivity to regional distribution. Since one can have reservations
about the equity and the economic consequences of public subsidy to the
abatement of privately generated pollution, it might be argued that this
program should be kept small. In any case, a limited budget should help to
focus the program on the most troubled areas where the returns from public
investment should be high. Given the rather poor documentation of farm water
158
-------
problems and the damages they cause, this seems to be a politically realistic
approach.
Even within this basic framework, state and local authorities, and Soil
Conservation Districts could be more aggressive in disciplining blatant of-
fenders. Direct counselling and local enforcement seem to have had some
effect in California, and there is no need to await ponderous federal legal
processes before attempting these simpler approaches.
Another possibility within the existing framework is to make farm con-
servation plans more explicitly water quality plans as well. The plans
already are criticized as overly elaborate, but most of them were prepared
with an eye to soil conservation rather than water quality, and they may
have neglected simple and effective measures conducive to quality. As new
plans are prepared or old ones modified, this defect should be repaired.
Experience with RCWP and with some experimental model implementation projects
(MIP) should give SCS an opportunity to sharpen its approach to water quality
control.
If our projected scenario prevails, however, then water quality prob-
lems attributable to agriculture likely will be much more severe than at
present. They will be mostly sediment-related and will increase in all
regions, but particularly in the nation's agricultural heartland—the Corn-
belt, Southern and Northern Plains, Mississippi Delta, Appalachia and the
Southeast (see Table 6.3). There will be wide variations among regions,
however. Measured by the increase in amount of sediment delivered, the
Southern Plains will be by far the most seriously affected. This is a re-
gion now little concerned about or prepared to cope with increased pressure.
Among the major producing regions the Cornbelt will be least affected. The
Northeast, Mountain, and Pacific regions will all experience substantial
percentage increases in sediment delivered, but the absolute amounts will
remain relatively small. These disparate trends, while stressing the need
for programs that are responsive to local conditions, also suggest that it
may be necessary to stiffen local resolve in some cases.
Whenever federal programs differentially affect some areas there is a
cry of unfairness and a call for equalizing payments. If there is unfair-
ness, it is in the needless application of uniform national point source
regulations that cause differential effects. However, water quality is a
local responsibility. Except for the broad goal of fishable and swimmable
waters, nonpoint pollution control will be pursued in relation to state
established stream standards. There are no effluent standards for nonpoint
sources. Differential regional impacts could occur through simple nonen-
forcement or through competitive degradation of streams standards in order
to retain local production. If stream standards are set in some objective
way and are enforced, then any depressing effect that water quality control
may have on local farming activities is a reflection of the comparative
advantage of local land resources and need not be equalized by national
payments. Some areas may be disadvantaged, and local farmers may lose in-
come and capital values, just as any segment of society may do by change,
but they are not unlawfully or unreasonably deprived according to any
standard of equity.
159
-------
It is clear that under our projected scenario EPA may need to take a
sterner view of 208 plans and, especially, of their implementation in some
districts. States could be encouraged to identify quality problems more pre-
cisely and to coordinate with federal programs so that water quality measures
could more effectively go piggy-back on conservation measures where conserva-
tion receives high priority, or so that quality can receive direct attention
if conservation does not.
The affected states will have the responsibility to implement stronger
measures for farm water quality control. They cannot expect equalizing
federal grants to protect their agriculture from adjustment to the need to
protect quality. However, federal money is likely to be available for mea-
sures primarily aimed at conservation, and fairly inexpensive adaptations
to such measures could have important quality implications. Thus, states
that are threatened by severe quality problems would be advised to strongly
support conservation programs for which federal money is available and to
backstop such programs with provisions for penalizing those farmers who fail
to cooperate. If the contract/soil loss tax combination is available, it
should provide a state with such an opportunity. However, the water quality
dimensions of conservation plans must be developed in cooperation with the
federal technical services.
It is worrisome to note the asymmetrical character of conservation pro-
grams, which we have argued should be federally supported to the extent that
they protect a social rather than a private interest, and water quality pro-
grams, that, being an external cost, should be the farmer's responsibility
and become reflected in price and production decisions. The implications of
these two positions for the location and profitability of production are
quite different. Were there no vested rights to consider, conservation could
be required up to a socially established level just like quality control,
with the operator absorbing any loss of income and capital that might imply,
but in our view equity argues against that. Our position on responsibility
for environmental protection is consistent with practice elsewhere in the
society, however. If we choose inconsistency with other environmental rules
and are willing to ignore the budgetary consequences, we can subsidize
quality control just as we do conservation.
In practice the two principles can support one another. A soil conser-
vation program makes sense from a social standpoint on much private land. It
may require the retirement of some land from crops and the less profitable
use of other land. The farmer must be given inducement to do that which is
socially rational. A realistic subsidy is at the center of this. But his
sense of responsibility can be accentuated by making him responsible for the
off-site consequences of soil loss. The conservation objective pursued by
performance contract or like device can be backstopped by a soil loss or a
sediment tax grounded in quality control for those unwilling to accept the
conservation objective freely. Soil miners at least would be deterred to the
extent of paying for off-site damages. And in all cases the incentive to use
the most efficient methods for obtaining target objectives would be preserved.
Moreover, both subsidy and tax could be, and should be, varied by area to
reflect the real social value of the objective sought.
160
-------
Research as Policy
Increased environmental stress from agriculture originates both in the
expanded scale of production and in the technologies employed. New technol-
ogies often are seen as posing added threats, but if they are pursued with
the need in mind to accommodate production while diminishing environmental
Impacts, then the outcome can be quite different. Moreover, if we are to
avoid both surprises and needless fears, there is a need for better informa-
tion on the consequences of changes in technology and scale of production.
Thus, research on which to base both control actions and environmentally
benign new technologies can be an important policy tool. The principal pol-
icy issues are those of the scale and the focus of the research.
Concern about research has been implciit in much of the previous dis-
cussion in this volume. We noted that knowledge of where to focus programs
for environmental amelioration is skimpy. State 208 plans, intended above
all to identify nonpoint sources of water pollution, have not increased the
precision of this information as much as was hoped. In general, excessive
reliance is placed on estimates derived from application of the Universal
Soil Loss Equation and on large models with obscure and coarse assumptions
rather than on selective direct measurement. EPA could help by identifying
and publicizing the more successful problem identification approaches found
in state plans.
There is always a temptation for a single mission regulator to attack
those problems that he perceives directly without full attention to the
possibly adverse displacements that this may cause elsewhere for other ob-
jectives or, indeed, even for the one intended. Although EPA has the prin-
cipal environmental regulatory authority, they do not have the detailed
knowledge of agriculture that would permit them to assess these tradeoffs
and they have little incentive to research them thoroughly or to seek tech-
nologies that may restructure them. One course might be for EPA to give
firmer long-range guidance on ambient needs and then allow the states and/or
USDA to structure controls and related research programs to meet those needs.
However it is triggered, detailed research is needed on the most effec-
tive ways of meeting ambient goals. Several possible directions for research
come to mind. Alternative pest management strategy is an obvious candidate.
The need is both to develop new chemicals and to perfect management tech-
niques that rely less on chemicals. Conservation tillage implies greater
use of herbicides and perhaps insecticides. The chemicals may move and
degrade differently under sustained use in this system—a matter that should
be better understood. The use of fertilizer also needs study. While the
leaching of nitrogen presently is a spotty problem, our projections suggest
it may become more severe and more general.
Environmental concerns need to be a more active component of agricul-
tural research aiming at yield increasing or cost reducing technologies,
for if the latter are not environmentally tolerable, they will fail. In-
creasing yields, especially on the best and least vulnerable lands, can
relieve pressure on the environment. The rapid strides in the biological
sciences of recent years so far have not been translated into new varieties
161
-------
on any great scale; there is an uncertain but possibly large potential to do
so. Targets might include plants that are more efficient synthesizers, more
pest or drought resistant varieties, and varieties with greater protein con-
tent. Conventional plant breeding and cultivation research might give atten-
tion to soybeans—an important crop whose yield is not impressive. Advances
in these areas should permit higher yields without greater environmental
hazard. A related possibility is the development of more unconventional
feeds for animals using resources presently unutilized. Even the development
of new markets can be an important innovation if they allow more intensive
(viz. corn/soybean combinations) use of good lands or the substitution of
higher yielding crops (e.g., sunflowers for wheat).
It would be helpful for agencies to be more aware of the productivity
and research implications of regulatory actions. Regulation that restricts
the use of current technology (e.g., pesticide use) may divert scarce re-
search resources into compensating for the lost techniques. And regulation
may discourage private spending on research (e.g., in pesticides or genetic
engineering) if it is not sensitively designed.
Agricultural research is especially susceptible to policy because so
much of it is the province of government. Agricultural commodities are un-
differentiated and are sold by small firms that are unable to undertake
costly scientific research of the kind needed. Research by private vendors
of supplies and services can be an important contribution in some areas, but
the government has an uncommon opportunity in this field to set the priori-
ties and the scale of the program in response to an integrated appraisal of
social needs.
A Final Word
American agriculture has always been one of our nation's greatest
strengths. From the first Thanksgiving to the factory farm, it has responded
to the needs placed upon it by our society. The challenge posed in this
study is whether agriculture can meet the high level export demands that it
may face in the future without irreparable damage to the productive base or
unacceptable cost to environmental quality. On the whole, we conclude that
it probably can, but that social controls are likely to intrude more into
the countryside than in the past if major objectives are to be met. If we
wish to avoid land use controls, then pressure on the land can only be re-
lieved by reducing demand or finding land substitutes. If we want to pursue
all objectives simultaneously, then conservation and rural water quality
programs must be operated more efficiently and in a mutually supportive man-
ner .
Any set of distant projections, such as those attempted here, is subject
to wide error. If the demand projections are too exuberant, much of the po-
tential problem vanishes. Others have been more conservative than we in
projecting demand, but we think that our scenario has sufficient plausibility
to deserve attention. The implications of such a level of production for the
use of land and other resources is also subject to interpretation. Per acre
yields must rise if demand is to be satisfied within the land resources that
162
-------
we are likely to be able to devote to crops. We think that condition can be
met, but we do not expect the past yield trend to be matched. Fertilizer use
will grow, but if better use is made of that applied, it will not outpace
production growth. Future pesticide use will depend on the location of pro-
duction and on the techniques of pest management and cultivation employed,
but declining use of insecticides on cotton and moderation of use on corn
suggest that no explosion will occur in this area.
The primary consequence of expanded production that has held our atten-
tion is damage to the soil resource from erosion and the consequent damage
to water quality as this is discharged as sediment to streams. Preserving
the productive capacity of the land seems to deserve the highest priority.
But damage to the soil is a function not only of the scale of production but
also of the cultivation practices and the location of production as well.
The spread of conservation practices, especially conservation tillage, gives
promise that any given output can be produced with less damage to the soil.
However, not all soils or crops are suited to this practice, and we have
been more conservative in projecting its adoption than some other projections
have been. As a consequence of this (together with our demand projections),
we foresee potentially grave problems from agriculture. They will not be
moderated much by geographical production shifts, for any such moves will be
toward vulnerable areas of the country. Explicit and vigorous soil conser-
vation tillage implies increased pesticide use, especially herbicides, with
consequences that cannot be entirely foreseen; therefore other p.est control
techniques must continuously be explored.
Our cautious optimism is meant to reflect confidence in the physical
capacity to meet demands and our belief that the environmental consequences
are on a scale that is manageable in principle. However, the latter will
not occur automatically. There must be recognition that a more heavily
stressed agricultural system may demand closer control if it is to continue
to expand without damage to productive capacity and environmental quality.
Much better monitoring and data on farm-related pollution should be sought
so that adverse trends can be spotted early and policy designed to meet
them. With these significant caveats, we believe that American agriculture
will continue to meet its domestic and international challenges.
163
-------
REFERENCES
Benbrook, Charles. Juiy-August 1979. "Integrating Soil Conservation and
Commodity Programs: A Policy Proposal," Journal of Soil and Water
Conservation.
The Clean Water Act. (Public Law 92-500).
EPA. November 1979. Institutional Bases for Control of Nonpoint Source
Pollution. Beatrice H. Holmes.
GAO (Comptroller General of the United States). February 1977. "To Protect
Tomorrow's Food Supply, Soil Conservation Needs Priority Attention."
Haith, Douglas A. and Raymond C. Loehr (eds.). 1978. Effectiveness of Soil
and Water Conservation Practices for Pollution Control. Draft manu-
script (Ithaca, New York).
Harris, Louis and Associates, Inc. January 17, 1980. "Survey of the Pub-
lic's Attitudes Towards Soil, Water, and Renewable Resources Conserva-
tion Policy."
Iowa State University College of Agriculture. March 1978. A Technical
Assessment of Nonpoint Pollution in Iowa (Harmon & Duncan).
John Muir Institute. June 1979. Erosion and Sediment in California Water-
sheds: A Study of Ins-titutional Controls.
Loehr, Raymond C., Douglas A. Haith, Michael F. Walter, Colleen S. Martin
(eds.). 1979. Best Management Practices for Agriculture and Silvicul-
ture. Proceedings of the 1978 Cornell Agricultural Waste Management
Conference (Ann Arbor, Ann Arbor Science Publishers Inc.).
Nagadevara, Vishnuprasad, S.S., Earl 0. Heady, Kenneth J. Nicol. June 1975.
Implications of Application of Soil Conservancy and Environmental Regu-
lations in Iowa Within a National Framework. CARD Report #57, Iowa
State University.
National Resources Council, Committee on Prototype Explicit Analyses for
Pesticides. 1980. Regulating Pesticides. (Washington, D.C., National
Academy of Sciences).
USDA. February 1980. Basic Statistics 1977, National Resources Inventory
(NRI). Mimeo revised.
164
-------
USDA. May 1974. Our Land and Water Resource.
USDA-ASCS. 1980. National Summary Evaluation of the Agricultural Conser-
vation Program - Phase I.
USDA. March 1980. 1980 Rural Clean Water Program (RCWP).
USDA. 1980. Soil and Water Resources Conservation Act, (RCA) Summary,
Parts I, II and Environmental Impact Statement. Review draft.
Seitz, W. D., D. K. Gardner, S. K. Gove, K. L. Gunterman, J. P. Karr,
R. C. F. Spitze, E. R. Swanson, C. R. Taylor, D. L. Uchtman, J. C.
vanEss. 1979. Alternative Policies for Controlling Nonpoint Agri-
cultural Sources of Water Pollution. EAP-60015-78-005. April.
Seitz, Wesley D., C. Robert Taylor, Robert C. F. Spitze, Craig Osteen, Mack
C. Nelson. February 1979. "Economic Impacts of Soil Erosion," Land
Economics, 55,1.
Wischmeier, Walter H. and Dwight D. Smith. May 1965. "Predicting Rainfall-
Erosion Losses From Cropland East of the Rocky Mountains" USDA-ARS
Agricultural Handbook No. 282.
Young, Keith K. 1978. "The Impact of Erosion on the Productivity of Soils
in the United States," paper presented at workshop in Ghent, Belgium.
165
-------
APPENDIX A: DERIVATION OF PROJECTIONS OF PRODUCTION
PROJECTIONS OF WORLD TRADE
WHEAT AND COARSE GRAINS1
The projections of world trade in these commodities are based on analy-
sis of trends in production and consumption in importing countries and groups
of countries, namely the importing developing countries, Western Europe,
Eastern Europe, the Soviet Union, the People's Republic of China (PRC) and
Japan.
2
Importing Developing Countries
Much of the relative increase in d-emand for food in the less developed
countries (LDCs) has come from relatively few among them with higher than
average income and income elasticities of demand. This is an argument for
considering these countries apart from the other LDCs. We decided that for
our purposes the improvement in our projections from doing this would not
justify the extra effort required.
Projections of food consumption typically are based on projections of
population, per capita income, and income elasticities of demand. The World
Bank (1980, pages 142-143) projects population growth in the importing devel-
oping countries as defined here at 2.2 percent from 1980 to 2000. Although
the Bank report does not indicate this, rates of population growth in these
countries are declining, and one recent projection is that by 1995-2000 ^
growth in the developing countries will be 1.62 to 2.00 percent annually.
Accordingly, we assume that in the countries of interest to us population
growth between 2000 and 2010 will average 1.7 percent per year. In this case
average annual growth over the full period 1980 to 2010 would be 2.0 percent.
The World Bank (1980, page 99) shows that in the 1970s real per capita
GNP in the developing countries increased 2.8 percent annually, down from 3.1
percent in the 1960s. The Bank expects per capita GNP in these countries to
grow only 2.6 percent from 1980 to 1985 beca!use of the steep increase in oil
prices in 1979. Between 1985 and 1990, however, the Bank expects GNP per
World trade data are for coarse grains rather than feedgrains as de-
fined in Table 1-1. Coarse grains as defined by the USDA are feedgrains plus
millet and "mixed grains."
2
For wheat these are all countries of Asia, Africa and Latin America
except the PRC, Japan, South Africa and Argentina. For feedgrains these coun-
tries as well as Brazil and Thailand are excluded.
3
Projections by the Community and Family Study Center, World Bank, United
Nations, and U.S. Bureau of the Census, reported in Tsui and Bogue.
166
-------
capita in these countries to grow 3.3 percent.annually, reflecting a success-
ful transition to a world of permanently higher, and continually increasing,
prices of oil. We assume that growth at 3.3 percent annually will continue
to 2010. In this case real per capita GNP over the entire period 1980-2010
will average 3.2 percent per year.4
The income elasticity of demand for grains in the developing countries
is between .3 and .5, being higher for wheat and lower for coarse grains for
direct consumption (USDA, 1974, page 77). However, in countries with incomes
sufficient to support rising demand for meat, the income elasticity of demand
for coarse grains for livestock is relatively high. If real per capita in-
come in the developing countries increases by 3.2 percent annually, the
demand for animal products is likely also to grow steadily, maintaining
steady growth in demand for coarse grains.
In summary, if population in the developing countries grows 2.0 percent
annually from the late 1970s to 2010 and per capita income grows 3.2 percent,
then demand for grains in those countries could reasonably be expected to
grow 3.0-3.5 percent per year over the period as a whole.
Table A-l shows production and consumption of wheat and coarse grains in
the importing developing countries in two periods since the mid-1960s. The
periods were selected to reflect the impact of the Green Revolution in these
countries on production and of the increase in oil prices on both consumption
TABLE A-l AVERAGE ANNUAL GROWTH RATES IN IMPORTING DEVELOPING COUNTRIES,
WHEAT AND COARSE GRAINS*
1966-1979 1972-1979
(percentages)
Wheat
Production 4.9 3.4
Consumption 4.8 3.9
Coarse Grains
Production 1.0 3.4
Consumption 3.1^ 4.3
^Calculated from the logarithmic time trends. Sources are various issues
of USDA, Foreign Agriculture Circular Grains. Data are world totals of pro-
duction and consumption less Western Europe, Eastern Europe, USSR, PRC, Japan,
U.S., Canada, Australia, South Africa, and Argentina (for wheat); the above
countries plus Brazil and Thailand (for coarse grains). Coarse grains are
corn and sorghum for grain, barley, oats, rye, millet and mixed grains.
+1968-1979
Schnittker Associates and the Food and Agriculture Organization project
per capita income growth in the developing countries at 3.2 percent annually
from the mid-1970s to 2000.
167
-------
arid production. With respect to wheat the most notable feature of Table A-l
is the slowdown after 1972 in growth of both production and consumption. The
slower growth of production likely reflects the relatively slower growth of
the Green Revolution after the initial burst in the mid-1960s. The heavy de-
pendence of the Green Revolution technology on irrigation and the difficulty
in rapidly expanding the irrigated area in these countries probably is a major
reason for the slower spread of the Green Revolution since the early 1970s.
Higher prices, and occasional disruptions in supply, of nitrogen fertilizer
in this period may also have contributed.
The slower growth in wheat consumption after 1972 likely reflects the
slowdown in growth of production. While there is no necessary connection
between the growth of consumption and production of wheat in these countries,
balance of payments constraints imposed by sharply higher oil prices after
1973 likely would have prevented wheat consumption from growing at the pre-
1972 rate, given the slowdown in production growth. In addition, real wheat
prices may have been higher after 1972, although we have not pursued this
possibility because of the complexity of internal pricing policies in these
countries.
We believe that the tendency toward slower growth of wheat production
and consumption will continue from 1979 to 1985 and from 1985 to 2010. Con-
sumption likely will grow more slowly because of lower population and per
capita income growth. Slower production growth seems likely, at least for
another decade, because of increasing real prices of energy and the expense
of extending irrigation. We think it not unlikely that by the 1990s high
yielding varieties of wheat will be developed which are less dependent on
irrigation, thus easing an important constraint on expansion of wheat pro-
duction. This is not predictable, however, and we have not tried to take it
into account in making our projections of wheat production in the developing
countries.
There is, in fact, no firm basis for projecting specific rates of growth
in wheat production and consumption in these countries. We assume that pro-
duction will grow 3.3 percent annually from 1979/80 to 2010. We assume that
consumption will grow yearly by 3.2 percent to 1985 and by 3.0 percent to
2010. The resulting projections of wheat production, consumption and imports
are shown in Table A-2.
Table A-l shows that, unlike wheat, production, and consumption of
coarse grains grew more rapidly after 1972 than in the preceding five or six
years-. The behavior of coarse grain production reflects unusually low pro-
duction in 1972 and 1973 because of bad weather. Production growth since
1972, therefore, cannot be used as a guide for projecting growth after 1979.
The fact is that production of coarse grains in the importing developing
countries has increased very sluggishly since at least the early 1960s. For
the period 1961-1979 production grew 1.3 percent annually. From 1974 to 1979
the annual rate was 1.4 percent.-'
These and subsequent figures on production and consumption of coarse
grains are from various issues of USDA, Foreign Agricultural Circular
Grains.
168
-------
TABLE A-2. PRODUCTION, CONSUMPTION, AND IMPORTS OF WHEAT AND COARSE GRAINS
IN IMPORTING DEVELOPING COUNTRIES
(millions metric tons)
1979/80 2010
Wheat
Production 80.8 217
Consumption 133.1 332
Imports 52.3 115
Coarse grains
Production 117.0 273
Consumption 139.9 399
Imports 22.9 126
Source: 1979/80 from USDA, August 13, 1980. Projections for 2010 de-
scribed in the text. Changes in stocks assumed to be zero. For wheat
includes all countries of Asia, Africa and Latin America except the PRC,
Japan, South Africa and Argentina. For coarse grains these countries plus
Brazil and Thailand are excluded.
The acceleration in the growth of coarse grain consumption after 1972
reflects a rapid increase in consumption by the OPEC countries. These coun-
tries consumed about 11.0 million metric tons of coarse grains in 1968 and
about 11.5 million metric tons in 1972. Thereafter, their consumption grew
sharply, reaching 18.9 million metric tons in 1978.
While the growth of coarse grain consumption in the OPEC countries was
especially dramatic, consumption in developing countries not members of OPEC
increased 2.7 percent annually between 1968 and 1978, with even faster
growth—4.1 percent—after 1972.
The growth of coarse grain consumption in the developing countries is
consistent with their rising income and a consequent shift in diet toward
more animal protein. If per capita income in these countries grows as we
have projected it, then the proportion of animal protein in their diets
should continue to increase, with a consequent steady expansion in con-
sumption of coarse grains. Consumption is not likely to continue to grow
at the pace set since 1972, however, if population and per capita income
growth in these countries slow down in accordance with our projections. We
assume that from 1979/80 to 2010 coarse grain consumption will grow 3.5
percent annually. At this rate of growth consumption would be 399 million
tons in 2010.
We think it likely that this rate of growth in coarse grain consumption
in the importing developing countries will stimulate coarse grain production
to grow faster than in the past. Should production continue to grow at the
sluggish pace experienced in 1961-1979 (1.3 percent annually), coarse grain
169
-------
imports in 2010 would have to be 220 million tons to satisfy the projected
levels of consumption. Valued at 1979 FOB Gulf Ports prices of corn and
grain sorghum, this volume of coarse grain imports would cost the importing
developing countries approximately $26 billion. The comparable value of
1979/80 imports was about $2.7 billion. We argue below that these countries
will be able to finance a substantial increase in their grain imports over
the next several decades, and perhaps they would be able to manage an import
bill for coarse grains of $27 billion. Long before the bill reached this
magnitude, however, it is likely that these countries would seek ways to
stimulate their own production of coarse grains ,so as to keep the bill to
more manageable size. We think their search would be rewarded. Until the
last few years, most of the effort of international and national food re-
search institutions was devoted to development of higher yielding varieties
of rice and wheat. Research on coarse grains was relatively neglected.
Recently the balance has begun to shift toward coarse grains, with the Re-
search Center for Improvement of Maize and Wheat (CIMMYT) in Mexico taking the
lead. We expect that an increasing proportion of national and international
research resources will be devoted to developing new technologies for coarse-
grain production. While we cannot now point to any major breakthroughs, we
believe the effort will soon begin to pay off in higher rates of coarse-grain
production in the importing developing countries. Accordingly, we assume
that production will grow 2 percent annually from 1979/80 to 1985, and that
the yearly rate will average 3 percent from 1985 to 2010.
Table A-2 shows the projections of production, consumption and imports of
wheat and coarse grains in the importing developing countries in 2010. Could
these countries afford the projected levels of imports? We think they could.
Valued at FOB Gulf Port prices of 1979, average wheat and coarse grain im-
ports of these countries in 1979/80 cost about $11.3 billion. Valued at the
same prices, our projections imply that the import bill for wheat and coarse
grains would rise to $34 billion in 2010. In 1978, merchandise exports of
the importing developing countries were $253 billion (World Bank 1980,
pages 124-125). Their imports of wheat and feedgrains in 1979 prices ($11.3
billion) were 4.5 percent of 1978 exports. Exclusive of the OPEC countries
grain imports of importing developing countries in 1979/80 were about 7 per-
cent of their merchandise exports in 1978.
Our projections of growth in population and per capita income in the
importing developing countries imply that their real GNP will increase 5.0-
5.5 percent annually. Export growth likely will keep pace, or approximately
so. Our projections of grain imports of these countries, valued in 1979
prices, indicate annual growth of 3.7 percent. Thus the value of these im-
ports would decline relative to the value of exports. Since these countries
were able to finance the grain imports of 1979/80 (food aid being relatively
minor), it is reasonable to assume that they should be able to finance them
also in 2010 when the imports likely will be smaller relative to exports than
in 1979/80. Of course, if real grain prices are higher in 2010 than in 1979
the financing problem would be more difficult, and debt service may lay a
relatively higher claim on export earnings than in 1979. Moreover, the prob-
lem of financing imports will not be equally distributed, those countries with
low export-grain import ratios having a more difficult time than countries
with high ratios (e.g., OPEC). Nonetheless, we believe the importing
170
-------
developing countries as a group will be able and willing to manage the fi-
nancial burden implied by our projections of grains.
Other Countries^
Projections of Wheat Imports—
For Western Europe, Eastern Europe and Japan our projections of wheat
imports are in two parts: to 1985 and then to 2010. Per capita consumption
of wheat in these countries was constant in the 1970s, and there was a small
declining trend in imports, reflecting expansion of wheat production rela-
tive to consumption in the European.Community (EC). We assume that these
tendencies continue to 1985, i.e., that wheat consumption grows only with
population and that imports continue to decline at the same rate as in the
1970s. In this case, wheat imports in these countries would be 16 million
metric tons in 1985, compared with 17 million in 1979/80 (USDA, August 13,
1980).
For 2010 we make two projections of wheat imports by these countries.
One assumes that trends in production, consumption and imports established
in the 1970s continue beyond 1985 to 2010. In that case these countries
would import 13 million metric tons of wheat in 2010. The alternative pro-
jection assumes that the high cost of the EC's Common Agricultural Policy
will increase pressure to liberalize the policy, and that after the mid-
1980s barriers to imports of wheat and feedgrains will gradually be lowered.
We assume this will restrain the growth of wheat production in the EC after
1985 enough to end the slow decline in imports for the group of countries
as a whole. Wheat imports by these countries in 2010 thus would be 16
million metric tons, the same as in 1985.
In the 1970s the Soviet Union shifted from being a net exporter of wheat
to being a net importer and over the decade imports fluctuated widely, from
.5 million metric tons in 1970 to 14.9 million in 1972. The average for
1970-1978 was 5.8 million tons. In the year which ended in June 1980 the
USSR imported 11.9 million metric tons of wheat, and the USDA expected the
figure for 1980/81 to be 13.0 million tons (USDA, August 13, 1980).
The behavior of wheat imports by the USSR reflected both year-to-year
fluctuations in the weather and a political decision by the Soviet authori-
ties to improve the diet of the Russian people. This decision is not
likely to be reversed. It suggests that the USSR will continue to be a
net importer of wheat.
While the shift of the Soviet Union from being a net exporter of wheat
to being a net importer is clear, the annual fluctuations in imports have
been so marked that we have no clear basis for extrapolating the trend.
Accordingly, we arbitarily assume that in 2010 the USSR will import 12
million metric tons of wheat.
Western Europe, Eastern Europe, USSR, Japan and the PRC.
171
-------
From the early 1960s through the mid-1970s the PRC regularly imported 2
to 6 million metric tons of wheat per year (USDA, May 1976). Since the mid-
1970s, imports have risen, reaching 8.8 million tons in the year ending
mid-1980, with 11.5 million tons expected by the USDA in 1980/81 (USDA,
August 13, 1980). This increase in wheat imports by the PRC evidently re-
flected a generally more open trade policy adopted by the country in the late
1970s. Given the formidable problems faced by the PRC government in feeding
its own people adequately, we believe the tendency toward rising wheat imports
will continue. We have no sound basis for projecting these imports to 2010,
however, so we arbitrarily assume they will come to 20 million metric tons
that year.
Projected imports of wheat by Western Europe, Eastern Europe, Japan, the
USSR and the PRC in 2010 are shown in Table A-3. Corresponding imports for
1979/80 also are shown for comparison.
TABLE A-3. WHEAT IMPORTS
(Millions Metric Tons)
2010
1979/80 (1) (2)*
Western Europe, etc.
¦17.2
13
16
USSR
. 12.5
12
12
PRC
10.2
20
20
Total
39.9
45
48
"Reflects more liberal agricultural trade policies by the EC.
1979/80: USDA, August 13, 1980.
Projections of Coarse Grain Imports—
For Western Europe, Eastern Europe, the USSR and Japan these projections
are differences between projections of production and consumption. We make
two projections for Western Europe, one assuming a continuation of the EC's
Common Agricultural Policy (CAP), and the other assuming that after 1985 the
CAP is liberalized, permitting lower grain prices within the community and
lower barriers to grain imports. On the assumption that the CAP remains
intact, coarse-grain production is projected to increase in accord with the
trend established in 1970-1980.^ In this case production in 2010 would be
139 million metric tons. (Production in 1978-1980 averaged 92.6 million MT).
^The trend equation for this period is Y = 74.2 + 1.62T (r = .73) where
T^ = 1970. Y is millions of metric tons of coarse-grain production. The
source is USDA, August 13, 1980.
172
-------
On the assumption that the CAP is liberalized and internal EC grain
prices fall after 1985 we project a gradual slowing of growth in coarse-grain
production, with a maximum of 110 million MT reached in the 1990s.
Coarse grain production in the USSR is projected to increase in accord
with the trend established in 1960-1980,® in which case it would be 179
million MT in 2010. (Production in 1978-1980 averaged 93.8 million MT.)
For Eastern Europe we assume that production of coarse grains reaches
65 million metric tons in 1985 and then shows no further increase. Produc-
tion in this region increased about 1 million metric tons per year on average
between 1971 and 1980 (USDA, August 13, 1980). But from 1975 to 1979 the
annual increase slowed to about .5 million MT. We assume the slower growth
reflects underlying economic and physical constraints affecting feedgrain
production in Eastern Europe and that these constraints will persist, allow-
ing only limited additional amounts of production.
The projections of consumption of feedgrains in Western Europe, Eastern
Europe, the USSR, and Japan^ are based on projections of feedgrain consump-
tion per capita. Per capita consumption in all of these areas now is less
than in the United States, the greatest difference being between Japan and
the United States (.15 metric tons per capita per year in Japan and .67 metric
tons in the United States) and the smallest difference being between the
United States and Eastern Europe (.53 metric tons per capita per year).in
the United States between 1960 and 1978 per capita consumption of feedgrains
fluctuated between .57 metric tons (in 1974) and .76 metric tons (in 1972),
but there was no trend. The average for the period was .67 metric tons.
In Western Europe, Eastern Europe, USSR, and Japan, however, per capita con-
sumption was rising, primarily in response to rising demand for grain-fed
animals for meat.
We assume that this trend will continue to 2010, with each region
approaching the current U.S. level of per capita consumption of feedgrains.
Accordingly, per capita consumption is set at .30 metric tons in Japan,
.65 metric tons in the USSR, and .60 metric tons in Eastern Europe. For
Western Europe we assume the figure will be .45 metric tons if the CAP con-
tinues unchanged. If it is liberalized, however, we assume that lower
coarse grain prices would stimulate the demand for grain-fed meat animals,
and that per capita coarse-grain consumption would then rise to the current
(and projected) U.S. level of .67 MT.
®The trend equation is Y = 44.6 + 2.63T (r = .83) where T^ « 1960 and
Y ¦ millions of metric tons of coarse-grain production. Sources are USDA,
May, 1976 and August 13, 1980.
9
Japanese production of coarse grains is insignificant. We treat it as
zero. Consequently for Japan projected consumption equals imports.
^All estimates of per capita feedgrains consumption are for 1978.
173
-------
The projections of per capita consumption of coarse grains were multi-
plied by projections of population in 2010 to derive the projections of total
consumption in each region or country. Table A-4 shows the population data.
TABLE A-4. POPULATION IN WESTERN EUROPE, EASTERN EUROPE, USSR, AND JAPAN
(millions)
1976
2010
Average Annual
1970-1976
Growth ($
1976-2010
Western Europe
343.9
381
.55
.3
Eastern Europe
129.2
148
.68
.4
USSR
256.7
304
.90
.5
Japan
112.8
138
1.30
.6
Source: Population in 1976 and growth rates 1970-1976 are from the 1978
World Bank Atlas. The projected growth rates are ours. Another source pro-
jects growth rates for the developed countries generally at .45 to .60 in
1995-2000 (Tsui and Bogue).
From the first half of the 1960s to the second half of the 1970s consump-
tion of coarse grains in the PRC increased a little more than 3 percent
annually, and until 1978 imports were negligible. In 1978, 1979 and 1980,
however, coarse grain imports averaged 2.5 million MT per year (USDA, May 1976
and August 13, 1980).
As noted in the discussion of wheat imports by the PRC, the country
evidently adopted a more open trade policy in the late 1970s, and the in-
crease in coarse grain imports after 1977 may be a reflection of this.
Whether this portends further growth in imports of coarse grains by the PRC
is unknown since Chinese policy in this regard is unpredictable. Given the
size of the country, however, the potential for increased imports of coarse
grains clearly is substantial if the authorities pursue a policy of upgrading
the quality of the Chinese diet. In this connection a member of a team of
USDA and farm industry officials who visited China in the spring of 1979 said
that the Chinese plan a major effort to increase animal production, and that
this "...will dramatically increase the PRC's demand for feedgrairis."
Although we do not necessarily subscribe to this view we nonetheless believe
some increase in PRC imports of coarse grains is likely. Without pretending
to a solid base for doing so, we project these imports at 15 million MT in
2010.
Our projections of coarse grains for Western Europe, Eastern Europe,.
Japan, the USSR and the PRC are shown in Table A-5. Data for 1979/80 are
shown for comparison.
¦^Darwin S*tolte, President of the U.S. Feed Grains Council, quoted in
Ag World, p. 20
174
-------
TABLE A-5. PRODUCTION, CONSUMPTION AND IMPORTS OF COARSE GRAINS, SELECTED
COUNTRIES AND REGIONS (millions MT)
1978/79
2010
Prod.
Cons.
Imports
Prod.
Cons.
Imports
(1) (2)
(1) (2)
(1) (2)
Western Eur.
92.
110.7
23.0
139 110*
171 255*
32 145*
Eastern Eur.
62.0
71.4
10.7
65
89
24
Japan
.5
19.0
18.4
—
41
41
USSR
92.7
106.9
14.3
179
198
19
PRC
77.0
79.6
2.6
np
np
15
Total
324.5
387.6
59.8
131 244
np: not projected
*Assumes that the EC's Common Agricultural Policy is liberalized
1979/80 from USDA, August 13, 1980. Changes in stocks not shown.
OILMEAL
We are interested in U.S. production and exports of soybeans. However,
in world markets soybeans are competitive with other sources of oilmeal, such
as oilpalm. In recognition of this, USDA data on world production and trade
of oilmeal products is given in metric tons of soybean meal equivalent.^
The ratio of oilmeal weight to weight of soybeans in the United States regu-
larly falls between .78 and .80. We have converted the USDA's data on world
trade in oilmeal products to a soybean basis by dividing them by .79. Our
procedure, then, for projecting U.S. production of soybeans is in three steps:
(1) project world trade in oilmeal products converted to soybean equivalents;
(2) project the U.S. share of this trade; (3) project U.S. domestic use of
soybeans.
World trade in oilmeal increased rapidly in the 1970s. In terms of soy-
bean meal the annual rate of increase was 7.8 percent between 1972 and 1980
(USDA, November 1977 and December 1979).^ Since 1973, that is, in the
period since the run-up in energy prices, world trade increased at an average
annual rate of 9.0 percent. It is unlikely that the oilmeal trade could
continue to expand very much longer at such fast rates, and we assume that
growth will slow considerably between 1979/80 and 1985, namely to 5.5 percent
per year. In this case the world oilmeal trade in 1985, expressed in soybean
12
These data appear monthly in Agricultural Outlook.
13
The trade figure for 1980 is a projection by the USDA based on 1979
production and analysis of previous relationships between production in one
year and trade in the following year. In recent years these projections have
understated the actual increase in trade.
175
-------
weight, will be 72 million metric tons. We expect continued growth in trade
from 1985 to 2010, but have no good.basis for projecting a specific amount.
The demand for ollmeals, particularly those with high protein content such
as soybean meal, is stimulated by high and rising income because those are
the conditions which encourage high-protein diets. Hence the major con-
sumers of high-protein oilmeals at the present time are the high-income
countries. We expect continued expansion of oilmeal consumption in those
countries as the demand for meat continues to rise. If the developing
countries achieve per capita income growth rates on the order of 3.0-3.5
percent annually, as we have projected, those countries also likely will be-
come important consumers of oilmeal by the end of the century.
At present, none of the high-income countries except the United States
are significant producers of soybeans or other sources of oilmeal. We see
no reason to expect this to change over the balance of the century. The
French have experimented with soybean production, but the results do not
suggest that France will become a significant producer of soybeans. Nor does
this appear likely for any of the other countries of Europe. Discussions
with persons in Europe and the United States knowledgeable about European
agriculture indicate a consensus on this issue.^
The prospect of rising demand for oilmeal in both developed and develop-
ing countries with production confined to a relatively small number of coun-
tires implies increasing trade in oilmeal between 1985 and 2010. We assume
that the trend toward slower growth in trade which we projected for 1979/80
to 1985 will continue from 1985 to 2010. We make two projections of annual
growth for this period: 3.0 percent and 3.5 percent. The higher figure
reflects the assumption of liberalization of EC price and trade policies
after 1985, leading to faster growth in demand for meat. This would not only
stimulate the demand for feedgrains, as noted earlier, but also the demand
for high-protein oilmeal as a feed supplement.
Table A-6 shows the projections of world oilmeal trade.
TABLE A-6. WORLD OILMEAL TRADE
(Millions metric tons, soybean equivalent)
1979/80* 2000 —
(1) (21
53.9 151 170
"USDA (April 1980, p. 40). Figure is average trade for the two years
divided by .79 to convert from soybean meal equivalent to soybean equivalent.
Column (1) in 2010 assumes continuation of the EC's Common Agricultural
Policy. Column (2) assumes the policy is liberalized.
14 "
In a listing of soybean producers in 1976-1978, Romania and Yugoslavia
are the only European countries to appear. Romania was the more important
with average annual production of about 200,000 tons. The USSR produced an
average of 600,000 tons in the three years (USDA, 1978, p. 133).
176
-------
U.S. SHARES OF WORLD TRADE IN 2010
In the four years 1976-79 the United States accounted for 44 percent of
world exports of wheat and for 65 percent of world coarse grain exports. Be-
tween 1972 and 1980 the U.S. share of world exports of oilmeal varied from
46.0 percent in 1975 to 59.5 percent in 1974. In 1979 it was 53.4 percent
and the USDA expected it to be 52.7 percent in 1980. For 1972-80 as a whole
the U.S. share averaged 51.5.
We have made two projections of U.S. shares in world exports of these
commodities in 2010. One assumes that the U.S. shares remain constant at
present levels. The other assumes that the U.S. competitive position in
world markets for these commodities weakens so that the U.S. share of each
declines.
THE CASE FOR CONSTANT SHARES
The case for believing the United States can maintain present shares is
based primarily on recent performance and lack of evidence for a major change
in the competitive position of other exporters relative to the United States.
Even good production performance in the developing importing countries and
the PRC seems likely at most to hold the growth of their imports of grains
within manageable limits. We already have indicated our belief that Eastern
Europe has little potential for additional production of grains and that the
Soviet Union's shift in the 1970s from being a net exporter of wheat to
being a net importer will continue. Continuation of the trend in coarse
grain production in the USSR, according to our projections, would enable that
country to hold imports to modest growth, but would not generate an export-
able surplus.
Since the mid-1970s the EC has been a net exporter of wheat, and while
the Community has been and continues to be a net importer of coarse grains,
its import deficit in these commodities declined from an average of 13.6
million MT in 1970/71 to 9.9 million MT in 1978/79. This performance could
be used to argue that the EC could emerge as an important challenger of the
United States in world grain markets. We find the argument unconvincing,
however. For both wheat and coarse grains the performance of the EC was
stimulated powerfully by the Community's Common Agricultural Policy (CAP).
The high internal prices adopted under the CAP both restricted the growth of
internal demand for grains and gave impulse to production. Continuation of
the CAP likely would increase the EC's exports of wheat, and we allowed for
that implicitly in our projections of wheat imports of Western Europe and
other regions shown in Table A-4. In our projections of coarse grains, we
dealt with Western Europe as a whole rather than the EC. Our projections
showed that Western Europe would continue to be a net importer of coarse
grains even if the CAP stays intact and coarse grain production continues to
grow at the rate set in the 1970s.
The expansion of wheat exports by the EC does not necessarily mean that
the U.S. share of the world wheat trade would be diminished. The emergence
177
-------
of the EC as a net exporter of wheat in the 1970s was at the expense of
Canada, Australia and the Soviet Union, not the United States. We have not
undertaken a detailed analysis of wheat production and costs in these coun-
tries so we are unable to explain why they lost ground in the world wheat
trade relative to the United States and the EC. Wheat yields in Canada and
Australia are well below those in the United States, although yield is a
quite imperfect measure of production costs. However, there is nothing in
the trend of wheat yields in Canada or Australia to suggest that those coun-
tries may improve their positions relative to the United States. As noted in
our discussion of wheat projections, the shift of the Soviet Union from wheat
exporter to wheat importer apparently reflects a policy decision to upgrade
the diet of the Soviet people, a decision unlikely to be reversed.
With respect to coarse grains, the declines in the EC's net imports in
the 1970s meant that other countries' exports, probably including those of
the United States, increased less than they would have otherwise. However,
as in the case of wheat, U.S. exports of coarse grains were less affected
by developments in the EC than exports of other countries. Indeed, the
U.S. share of world trade in coarse grains increased from 40-45 percent early
in the 1970s to 65 percent at the end of the decade.
Thus the reduction in the EC's net imports of coarse grains already is
reflected in the U.S. share of world trade in these crops. Continuation of
the trend toward smaller net imports in the EC, therefore, in itself would
not threaten maintenance of the U.S. share. Acceleration of the trend would
be a threat, but we see no reason to expect this. All of the increase in EC
production of coarse grains in the 1970s was due to rising yields. The
amount of land in these crops was constant, and knowledgeable people in
Europe believe there is little potential for bringing in additional land.
Yields of coarse grains in the Community were almost 15 percent less than in
the United States in the 1970s and rose at a somewhat slower rate.
In short, there is nothing apparent in the performance of EC production
and trade in coarse grains in the 1970s to suggest that the Community might
threaten the present U.S. share of world trade in these crops. A detailed
study might reveal undercurrents that our admittedly superficial analysis
has missed, but in the absence of any such evidence we conclude that the
United States could reasonably be expected to retain its 65 percent of
world trade in coarse grains.
The ability of the United States in the 1970s to maintain its position
in the world oilmeal trade is impressive since in this period Brazil, and
more recently Argentina, emerged as major exporters of soybeans and soybean
products. At the beginning of the 1970s Brazil and Argentina's share of
the trade in soybeans and products was negligible, but by 1977 their exports
of these commodities were almost one-third of world trade while the share of
the United States had declined to two-thirds (USDA, January 1979, p. 10).
The fact that the United States maintained its position in the oilmead trade
generally even while losing ground to Brazil and Argentina in soybeans indi-
cates that trade in soybeans and products was expanding faster than trade
in other oilseed products.
178
-------
Since the mid-1970s the relative positions of the United States and
Brazil-Argentina in the world soybean trade appear to have stabilized.
Should U.S. exports maintain the current relation to Brazilian and Argentine
exports, the U.S. share of the world trade in oilmeal would begin to rise if
trade in soybeans and products continues to increase more rapidly than trade
in other oilseed products. Should the U.S. share of trade in soybeans and
products remain constant, its share of total oilmeal trade also would remain
constant even if trade in soybeans and products increases no more rapidly
than trade in other oilmeal products. Some combination of a slower rate of
increase in soybean trade relative to trade in other oilmeal products and
slower growth in Brazilian-Argentine exports of soybeans and products rela-
tive to those of the United States also would vield stability in the U.S.
share of total oilmeal.trade.
On the basis of these considerations, we believe that U.S. exports of
soybeans and products could reasonably be expected to continue to take half
of total world trade in oilmeal.
THE CASE FOR DECLINING SHARES
The argument that the United States could lose ground in world markets
for grains and soybeans is based on .an analysis of long-run costs of produc-
tion in U.S. agriculture (Crosson). The analysis shows that from the end of
World War II until the early 1970s real prices (i.e., nominal prices deflated
by the cost-of-living index) of U.S. agricultural commodities were declining,
reflecting rapid gains in productivity, relatively low prices for fertilizer,
pesticides, and fuel, and a declining share of returns to land, capital, and
management in total farm income. Since the early 1970s the rate of increase
of productivity has slowed, and there is some evidence that this reflects
approaching exhaustion of the productivity potential of the technologies on
which U.S. farmers have relied since the end of World War II (National Acad-
emy of Sciences). Real energy and fertilizer prices in 1978 were 20-to-30
percent higher than in the early 1970s, and there is a consensus that energy
prices, and probably fertilizer prices, will be higher in 2010 in real terms
than they were in 1978. Finally, the share of returns to land, capital, and
management in total farm income now is only some 5 or 6 percent, down from
around 25 percent in the late 1940s. The room for further decline, without
restricting production, is more limited now than it was then.
The implication of these underlying movements is that the trend toward
lower real prices of U.S. agricultural production may slow, or give way to a
rising trend. Should this occur, and we believe it likely, then the compe-
titive position of U.S. agriculture in the world markets for grains and
oilmeal may be weakened by 2010. We do not argue that this necessarily will
occur. New technologies may come into use which would stimulate more rapid
productivity growth, and farmers abroad will be as exposed to higher energy
prices as American farmers. Still, the likelihood of some deterioration in
the U.S. competitive position seems sufficiently high that we ought to take
it into account in our projections.
179
-------
We have no basis for judging how much the U.S. share of world trade in
grains and oilmeal may decline by 2010. We arbitrarily set these shares at
35 percent for wheat, 50 percent for feedgrains, and 40 percent for oilmeal.
U.S. DOMESTIC CONSUMPTION
Our projections for 2010 are found by multiplying projections of per
capita consumption by projected population of 279.6 million.15 Annual per
capita consumption of wheat in the United States fluctuated between 85 and
113 kilos between 1963 and 1978. The average was 100 kilos and there was no
trend. We project 100 kilos for 2010.
As indicated earlier, per capita consumption of feedgrains in the United
States was more variable in this period, fluctuating around an average of
.67 metric tons, but there was no trend. We project .67 tons for 2010.
Per capita consumption of soybeans increased from 64.4 kilos in 1963/65
to 104.4 kilos in 1977/79, an average annual rate of increase of 2.9 kilos.
However, there was a tendency for the rate of increase to slow in the 1970s,
the annual growth being 2.1 kilos from 1970/72 to 1976/78. We assume that
the tendency for the rate of increase to slow will continue, and that on
average per capita consumption from 1977/79 to 2010 will rise 1.7 kilos
annually. Soybean consumption per capita in 2010 then would be 159 kilos.
PROJECTIONS OF GRAINS AND SOYBEANS TO 2010
The projections of U.S. production, exports and domestic use of wheat,
feedgrains and soybeans are shown in table A-7.
PROJECTIONS OF COTTON
These projections are based on a USDA study by Collins, Evans, and
Barry (C-E-B), with certain adjustments by us. The C-E-B projections are
for U.S. production, export, and domestic use of cotton in 1985 and 1990.
The basic assumptions underlying the projections are as follows (C-E-B),
p. 4):
(1) There will be no major wars in the world between now and 1990.
(2) Cotton is viewed as a homogenous commodity; no distinction
is made among various staples and grades.
"^Found by interpolating between the Department of Commerce's Series E
estimates for 2000 and 2025.
180
-------
TABLE A-7. U.S. PRODUCTION, DOMESTIC USE, AND EXPORTS OF WHEAT, FEEDGRAINS
AND SOYBEANS
(million metric tons)
2010
1979/80*
share
U.S.
constant
U,
share
,S.
reduced
(1)
(2)
(1)
(2)
Wheat
Production
60.8
98
100
84
85
Domestic use
22.1
28
28
28
28
Exports
38.4
70
72
56
57
Feedgrains
Production
207.7
354
428
316
372
Domestic Use
153.0
187
187
187
137
Exports
72.6
167
241
129
185
Soybeans
Production
56.3
120
129
104
112
Domestic use
24.3
44
44
44
44
Exports
29.3
76
85
60
68
*Averages for the two years, except for soybeans, the years are 1978/79.
The differences between production and the sum of utilization and exports is
the change in stocks. Exports of soybean meal and oil are expressed in
soybean equivalents.
Sources: USDA, August 13, 1980 for wheat and feedgrains. USDA April
1980 for soybeans. In the projections, changes in stocks are assumed to be
zero. Projection (1) for 2010 assumes continuation of EC trade and price
policies. Projection (2) assumes more liberal trade and lower price policies.
Export projections derived by multiplying projections of world trade by U.S.
shares. For wheat and feedgrains the projections of world trade are the sums
of imports from tables A-2 and A-3 (wheat) and A-2 and A-5 (coarse grains).
World trade in soybeans is from Table A-6. Projections of domestic use are
described in the previous section.
(3) A synthetic noncellulosic fiber combining the best properties
of both cotton and current state of the art synthetics will
not become widely available.
(4) A low-cost process that permits cotton to assume all the easy-
care and abrasion-resistant properties of synthetics will
not become widely available.
181
-------
(5) There will be no technological advances in cotton production
that permit dramatic yield increases worldwide (such as a multi-
adversity seed variety).
(6) Any sustained worldwide food production shortages that might
occur would not result in a major shift of land from cotton
production to food production.
(7) The price of manmade fiber will increase relative to the price
of an equivalent amount of cotton, compared with price averages
over 1974-76.
(8) Government policies on fiber consumption, production and trade
will remain unchanged from current policies unless specified
otherwise in certain markets.
(9) Cotton and manmade fiber will compete in a mature market be-
tween now and 1990, especially in developed regions. This
means the relative price of cotton and manmade fiber will
become more important in determining the market share of each
fiber.
Within this framework of assumption two sets of projections to 1990 are
developed, one assuming slightly slower growth of world income than in the
1960s and 1970s and the second assuming that income grows 25 percent less
than in the 1960s and 1970s. The growth of world demand and supply of fiber
is projected in accord with these projections of income growth. Cotton's
share of the world fiber market is assumed to continue to decline, as it has
for several decades, but at a slower rate than in the 1960s and early 1970s.
The main reason for cotton's declining share, according to C-E-B, has been
consumer preference for easy-care and abrasion-resistant fabrics competitive
with cotton in price.
Projections of world trade in raw cotton are based on analysis of
demand-supply balances among developed, developing and centrally planned
economies. Under the higher income growth alternative world trade in raw
cotton is 20.9 million bales in 1985 and 23.3 million bales in 1990. With
the lower income growth alternative trade is 18.8 million bales in both 1985
and 1990. In 1974-76 world trade averaged 18 million bales.
The U.S. share of world trade with the higher income alternative is
27 percent. With the lower income alternative it is 19 percent. In 1974-76
the U.S. share averaged 22 percent. The decline in the U.S. share with
slower growth is attributed by C-E-B to the position of the United States as
one of the most price-elastic members among major cotton exporting countries
for both supply and demand.
Domestic demand for cotton in the United States is projected in two
steps: (1) a projection of demand for fiber, based primarily on projections
of income and the income elasticity of demand for fiber; (2) a projection of
cotton's share of total demand for fiber. The share is expected to decline
gradually from 27 percent in 1977 to 22 percent in 1990.
182
-------
Within the two income growth scenarios alternative projections of U.S.
production, exports and domestic production are made in accordance with
different assumptions about ratios of cotton prices to prices of polyester
fibers. These price assumptions have little effect on the projections, how-
ever (C-E-B, p. 23).
The various assumptions made by C-E-B result in a variety of projections
of U.S. production, exports and domestic consumption of raw cotton. The
production projections for 1990 range from 10.0 million bales to 13.4 million
bales. We have elected to base our projection on a figure of 13.1 million
for 1990 plus an increment of 2.5 million bales. The 13.1 million bale
figure is the 1990 projection of C-E-B corresponding to the higher income
growth scenario and a cotton-to-polyester price ratio of 1.15:1. We selected
this scenario because it is in line with the World Bank's projections of
world income growth (World Bank, 1980). We added the increment of 2.5 mil-
lion bales because in discussion with one of the authors in the summer of
1980, Collins (of C-E-B) said he then believed that the 13.1 million bale
projection was low by about that amount. The reason was that the 13.1 mil-
lion bale figure did not reflect changed expectations about the growth of
demand in the PRC and Western Europe for U.S. cotton exports. Accordingly,
we project U.S. cotton production in 1990 at 15.6 million bales, and obtain
our projection for 2010 by extrapolation. These and related figures are
shown in Table A-8. The projections to 2010 assume that the maximum growth
of cotton production after 1990 will be half the annual rate projected by
C-E-B for 1976/79 to 1990. At a minimum cotton production would grow only
.4 million bales from 1990 to 2010.
TABLE A-8. PRODUCTION, NET EXPORTS AND MILL CONSUMPTION OF RAW COTTON IN
THE UNITED STATES (million 480 lb. bales)
1976/79
1990
2010
Millions
Millions
Millions
480 lb.
Bales
Metric
Tons
480 lb.
Bales
Metric
Tons
480 lb. Metric
Bales Tons
Production
12.6
2.7
15.6
3.4
16-18 3.5-3.9
U.S. Mill Con-
sumption
6.4
1.4
6.8
1.5
np np
Exports
5.7
1.2
8.8
1.9
np np
Sources: 1976/79 from USDA, November 1979.
1990 from C-E-B» p. 23. The projections of production and
net exports were increased by 2.3 million bales. See text
for explanation.
2010. See text for explanation.
183
-------
REFERENCES
Ag World. May/June 1979. Vol. 5, No. A.
Collins, K. J., R. B. Evans and R. D. Barry. 1979. World Cotton Production
and Use: Projections for 1985 and 1990. USDA, Foreign Agricultural
Economic Report No. 154. Washington,. D.C.
Crosson, Pierre. February 1979. Long-run Costs of Production in U.S. Agri-
culture. Resources for the Future. Unpublished paper.
Food and Agriculture Organization. 1979. Agriculture: Toward 2000. C 79/24.
Rome
National Academy of Sciences. 1975. Agricultural Production Efficiency.
Washington, D.C.
Schnitker Associates. March 30, 1979. Trade Issues Relating to World Hunger.
A report prepared for the Presidential Commission on World Hunger.
Washington, D.C.
Tsui, Amy Ong and D. J. Bogue. Oct. 1978. "Declining World Fertility:
Trends, Causes, Implications," Population Bulletin, Vol. 3, No. 4.
Population Reference Bureau, Washington, D.C.
USDA. Dec. 1974. The World Food Situation and Prospects to 1985. Foreign
Agricultural Economics Report No..98. Washington, D.C.
. May 1976. Foreign Agriculture Circular Grains. FG 9-76.
Washington, D.C.
. August 3, 1976. Foreign Agricultural Circular Grains. FG-17-76.
Washington, D.C.
. November 1977. Agricultural Outlook. Washington, D.C.
. 1978. Agricultural Statistics. Washington, D.C.
. January 1979. Foreign Agriculture Circular FOP 2079.
Washington, D.C.
. November 1979. Foreign Agriculture Circular Cotton. FC 17-79.
Washington, D.C.
184
-------
. . December 1979. Agricultural Outlook. Washington, D.C.
. April 1980. Agricultural Outlook. Washington, D.C.
. June 1980. Agricultural Prices Annual Summary 1979.
Pr 1-3 (80). Washington, D.C.
. August 13, 1980. Foreign Agricultural Circular Grains.
FG 23-80. Washington, D.C.
. January 28, 1981. Foreign Agricultural Circular Grains.
FG-4-81. Washington, D.C.
World Bank. 1978. 1978 World Bank Atlas. Washington, D.C.
. 1980. World Development Report. Washington, D.C.
185
-------
APPENDIX B
TRENDS IN TOTAL AGRICULTURAL PRODUCTIVITY
Table B-l shows rates of growth in total agricultural productivity over
various periods between 1950 and 1972 and between 1972 and 1980. As the
years advanced from 1950 to 1960, the annual rate of increase in productivity
declined. From 1960 to 1972 it was stable to slightly increasing and from
1972 to 1980 it again declined. It can be argued that between 1950 and 1972
American farmers were engaged in exploiting the high-productivity potential
of a new technology consisting basically of fertilizers, high-yielding seed
varieties, mechanical power substituting for labor and animals, and to a les-
ser extent, irrigation. If this statement captures the essence of what was
happening, then the decline and subsequent leveling off from 1960 to 1972 in
the rate of increase of productivity is plausible. The argument runs as fol-
lows. When the new technology began to spread after World War II there was
a large difference in productivity between it and the technologies then in
use. Consequently the rapid spread of the new technology produced a fast
increase also in productivity. By about 1960, however, the substitution of
the new for the old technology was substantially completed. Subsequent in-
creases in productivity, therefore, reflected incremental improvements in
the new technology rather than wholesale substitution of the new for the old.
In this circumstance a slower rate of increase in productivity after about
1960 would be expected.
Whether this account explains the apparent decline in productivity
growth after 1972 is not clear, and the issue cannot be pursued here. In-
deed, the discussion of trends in agricultural productivity is vexed by
serious problems in the measurement of total productivity. There is strong
evidence that failure to include quality improvements in inputs resulted in
significant underestimates of increases in the quantities of inputs, thus
overestimating the increase in productivity.^ Perhaps more important, the
By 1960 high-yielding varieties of corn and sorghum had almost 100 per-
cent replaced lower yielding varieties previously In use, and a major part
of the substitution of mechanical power for animal power had been completed
by the mid-1950s. (Inputs of mechanical power and machinery increased at an
annual rate of 6 percent from 1946 to 1955, but by only .24 percent annually
from 1955 to 1972. Fertilizer consumption, however, grew even faster after
1960 than it did before that date.)
^National Academy of Sciences, 1975, pp. 28-29. Also USDA, Feb. 1980,
especially pp. 28-32.
186
-------
TABLE B-l. AVERAGE ANNUAL RATES OF GROWTH OF TOTAL AGRI-
CULTURAL PRODUCTIVITY IN THE UNITED STATES
Percent Index Points
1950-1972: 2.05% 1.80
1955-1972: 1.91 1.77
1960-1972: 1.62 1.59
1965-1972: 1.64 1.69
1972-1980: .97 1.08
Source: USDA, Feb. 1981 for 1950-1979. 1980 from Economic Report of
the President, 1981.
Note: The growth rates are derived from least squares analysis of an-
nual index numbers of total productivity. Logarithms of these
numbers were used to calculate percentage rates of change.
vast technological changes in American agriculture since the end of World
War II make it difficult to construct an index which unambiguously measures
the growth of inputs; consequently measurement of productivity change also is
ambiguous. The input index currently used by the USDA weights inputs by
their average prices in 1971-1973. Wages were much higher in those years
relative to prices of other inputs than at the end of World War II, but the
amount of labor was sharply reduced relative to those inputs, particularly
fertilizer and other agricultural chemicals. Use of 1971-1973 weights gives
relatively great importance to labor and hence dampens the rise of the total
input index. If price weights of earlier years were used—say 1946-1949—
labor would weigh less and other Inputs more, and the input index would rise
much more steeply than the one used by the USDA. The index of total produc-
tivity, of course, would also rise less.^
There is no generally satisfactory solution to this index number prob-
lem, and the USDA is not to be faulted for not having solved it. The point
is that the ambiguity in the productivity index requires that movements in
it be interpreted cautiously, and that conclusions about those movements be
checked for consistency with other indicators of agricultural performance.
A question arises also about the behavior of the index specifically in
the period after 1972. Two features of Table B-2 prompt the question: (1)
the fact that most of the growth in total output after 1972 was attributable
to crops, and (2) the contrasting behavior in this period'of farm real estate
and harvested cropland. Farm real estate includes all land in farms, service
3
A study by Barton and Durost, cited in National Academy of Sciences,
1975, p. 31, indicated that for the period 1940-1942 to 1956-1958, the index
of total inputs increased 23 percent when prices of 1935-1939 were used as
weights. When 1957 prices were used the index declined 1 percent.
187
-------
TABLE B-2. AVERAGE ANNUAL RATES OF INCREASE IN PRODUCTION AND
SELECTED INPUTS IN U.S. AGRICULTURE (PERCENT)
1960-1972 1972-1980
Production
Crops 1.6 2.7
Animal Products 1.7 .7
Total 1.6 2.1
Inputs*
Fertilizer 7.3 3.7
Labor -4.5 -3.3
Machinery .3 ,3.5
Farm real estate .2 .0
Harvested cropland -.2 1.8
Source: USDA, Feb. 1981 for 1960-1979; 1980 from USDA, Jan.-Feb. 1981.
Note: Rates are derived from least squares analysis of logarithms of
indexes of inputs and production.
^Harvested cropland in 1972-1980. All other inputs are 1972-1979.
buildings, grazing fees, and repairs on service buildings. It thus includes
harvested cropland. The amount of harvested cropland increased after 1972
even though farm real estate was unchanged because some farmland previously
held idle or in other uses was shifted to crops. It is farm real estate
which is included in the index of inputs, but it was the increase in harves-
ted cropland which accounted for most of the increase in crop production,
and hence total production, after 1972. It appears, therefore, that because
of the way land is handled in the index of inputs, the rise in the index
after 1972 was underestimated, with consequent overestimation of the rise
in productivity.
188
-------
REFERENCES
Economic Report of the President. 1981. Government Printing Office.
Washington, D.C.
National Academy of Sciences. 1975. Agricultural Production Efficiency.
Washington, D.C.
USDA. Feb. 1980. Measurement of U.S. Agricultural Productivity: a Review
of Current Statistics and Proposals for Change. ESCS Tech. Bull.
No. 1614. Washington, D.C.
. Jan-Feb. 1981. Economic Outlook.
. Feb. 1981. Economic Indicators of the Farm Sector: Production
and Efficiency Statistics. 1979. Economics and Statistics Service.
Stat. Bull. No. 657.
189
-------
APPENDIX C
ANALYSIS OF TRENDS IN CROP YIELDS
PURE YIELD AND SHIFT EFFECTS
If ^ is the weighted average of yields for a given crop in period
L1 L
1 relative to period 0 across a number of states, then
1
0
T
1
0
i.
0
ft)
la
+ % *
lb + ...
+ a
"n
in
i
~
-------
We have calculated these measures with yields in period 0 equal to 100.
Consequently, the yield effect minus 100 plus the shift effect minus 100 plus
an interaction effect (found as a residual) minus 100 equals the observed
percent change in yields.
We calculated these effects for corn, wheat, sorghum, and soybeans in
those states which accounted for 80 to 90 percent of total production in the
period since 1946-1950. Table C-l gives the results. The table shows that
except for soybeans pure yield effects accounted for 90 percent or more of
observed changes in yields. The shift effect for yields of soybeans was
negative because of a marked relative shift of land in that crop to Arkansas,
Tennessee, Mississippi, and Louisiana, where yields are less than in the
Cornbelt, the other principal producing region. Even for soybeans, however,
the pure yield effect was by far the most important factor in observed yield
changes over most of the period considered.
Perhaps the most significant aspect of table C-l is the decline in the
rate of increase of yields of corn, sorghum and soybeans, and the decline in
wheat yields, from 1971-1973 to 1977-1979 compared with the earlier periods
shown. This apparent break in trend is generally consistent with the beha-
vior of the index of all crop yields (see table 3-1). We noted that the
behavior of that index was generally consistent with the slower shift to
land-saving technologies which occurred after 1972. This must also partially
explain the behavior of yields of grains and soybeans. We know, however,
that weather has a powerful effect on yields in any given year, and that it
may also affect the trend of yields. Before we can make statements about
the effect of land quality and technology on yields, therefore, we must take
out the effects of weather.
PURE YIELD FACTORS: WEATHER
We have estimates of the effects of weather on yields of corn and soy-
beans in the Corn Belt and of wheat in the Great Plains. The estimates are
based on the work of Louis Thompson of Iowa State University (National Oce-
anic and Atmospheric Administration, 1973 and Thompson, 1977) . For corn
and soybeans the estimates refer to Ohio, Indiana, Illinois, Iowa, and Mis-
souri. For wheat the estimates are for North and South Dakota, Nebraska,
Kansas and Oklahoma.
For corn and soybeans Thompson's analysis covers the years 1891-1976.
For wheat the period is 1893-1973. For corn the share of the five states
in total land in corn nationwide was about 40 percent in the late forties
and fifties, rising to a little over 50 percent in the early 1970s. The
share of these states in total soybean land was about 75 percent in the
late forties, falling to about 50 percent in the early 1970s. The share of
the five wheat states in total wheat land was about 50 percent in the for-
ties and fifties, rising to about 65 percent in the early 1970s. Thompson
presents graphs showing that yields of each crop in each set of states were
more variable than national average yields but that the trends of yields
in the states were very similar to national trends.
191
-------
TABLE C-l. PURE YIELD, PURE SHIFT, AND INTERACTION EFFECTS ON
OBSERVED YIELDS OF CORN, WHEAT, SORGHUM, AND SOYBEANS
1946/50
1966/70
¦ 1971/73
1971/73
to
to
to
to
1966/70
1971/73
1975/76
1977/79
Corn
% change in
yields
100.7
21.6
-8.2
8.9
Pure yield
91.0
21.5
-9.1
n.a.
Pure shift
10.5
2.1
.8
n.a.
Interaction
-.8
-2.0
.1
n.a.
Wheat
% change in
yields
66.5
15.0
-5.9
h-*
00
Pure yield
65.3
15.7
-6.3
n. a.
Pure shift
1.0
-. 6
.5
n.a.
Interaction
.2
-.1
-.1
n. a.
Sorghum
•
% change in
yields
180.0
15.0
-15.9
.9
Pure yield
172.2
14.5
-16.3
n.a.
Pure shift
-1.3
-.1
.1
n.a.
Interaction
9.1
.6
.3
n.a.
Soybeans
% change in
yields
31.0
6.0
-1.4
8.1
Pure yield
39.2
5.7
-.6
n .a.
Pure shift
-7.8
.6
-.6
n.a.
Interact ion
-.4
-.3
. -.2
n.a.
Sources: Data on yields are from U.S. Department of Agriculture, Agri-
cultural Statistics, various years. Procedures for calculating pure yield,
pure shift, and interaction effects are described in the text.
n.a. means not available because time did not permit calculation of
these numbers. However, visual inspection of the data indicates that the
pure yield effect accounted for most of the change in yields for each crop
between 1971/73 and 1977/79.
192
-------
Thompson relies on a regression of yields against time and weather.
Time is measured in years and weather in deviations of monthly precipitation
and temperature values in particular crop season months from long-term aver-
age values of precipitation and temperature for those months. Time is as-
sumed to measure technology. The analysis gives annual estimates of yields
for each state, reflecting technology and weather. The estimates for each
state are weighted by the harvested area of the state in a given year to get
estimates for groups of states. The weighting procedure eliminates the ef-
fects on yields of shifts in the share of each state in total harvested
area. The effects of shifts within states to land of differing quality
would show up, however, as an effect of technology.
To get at the effects of weather over time, Thompson holds technology
constant by selecting a given year, 1973, and then substituting weather var-
iables in his equations for each year in his time series. The result is an
estimate for each year of what yields would have been, given 1973 technology
and the weather for that year.
We calculated an index of these technology-adjusted yields, setting
yields with "normal" (i.e., long-term average) weather equal to 100. Annual
differences in this yield index reflect, in principle, the effects of year-
to-year differences in the weather. By dividing this yield index for each
year into actual yields, we get an estimate of yields for each year which
is independent of the effect of weather. Movements in the yield series de-
rived in this way thus reflect only the effect of technology.
Tables C-2, C-3 and C-4 show for corn, soybeans and wheat, the index
of weather effects on yields, actual yields, actual yields adjusted for wea-
ther, and two sets of weather-adjusted trend values for yields. Examination
of weather-adjusted yields in the tables indicates clearly that those series
include influences other than technology. With respect to corn, for example,
we know that in 1970 yields were greatly reduced by a kind of blight, an
occurrence independent of weather, or in any case, not picked up by Thomp-
son's method of measuring weather effects. Nevertheless, for all three
crops, weather-adjusted yields are less variable from year to year than
actual yields, the expected result. In any case, we have nothing better.
Accordingly, we assume that the weather-adjusted series roughly approximate
the effects of technology on the trend of yields of corn, wheat, and soy-
beans .
We ask the following question about the data in the tables: is the
behavior of actual yields of corn, soybeans and wheat from 1972 to 1980
consistent with the hypothesis that, after adjustment for weather, the trend
of yields in that period was the same as the pre-1972 trend? If the answer
is yes we would conclude that there was no change in the effect of land
quality or technology on the growth of yields after 1972. If the answer is
no we would conclude that yield growth had been affected by changes in land
quality, or in technology, or both.
We are handicapped by the lack of data which would permit us to adjust
yields for weather for most years after 1972, as the tables indicate. For
wheat we can make the adjustment only for 1973 and for corn and soybeans
193
-------
TABLE C-2. CORN: INDEX OF EFFECTS OF WEATHER ON YIELDS, ACTUAL YIELDS,
WEATHER-ADJUSTED YIELDS, AND TREND VALUES OF WEATHER-ADJUSTED
YIELDS*
Y ield s
(bu.- per
acre)
Trend of
Index of
Weather-Ad 1usted
Weather
Weather-
Arith-
Loga-
Year
Effects^
Actual
Adjusted
metic
rithmic
1950
96.8
49.0
50.6
47.1
50.0
51
95.2
48.1
50.5
49.4
51.1
52
92.8
55.7
60.0
51.7
53.3
53
95.6
51.0
53.4
.54.0
55.0
54
92.5
51.1
55.3
56.3
56.7
1955
90.7
51.8
57.1
58.7
58.6
56
97.9
58.4
59.7
61.0
60.4
57
97.4
58.7
60.3
63.3
62.4
58
99.5
64.7
65.0
65.6
64.4
59
99.4
64.0
64.4
67.9
66.4
1960
97.1
64.2
66.1
70.3
68.6
61
102.9
74.2
72.1
72.6
70.8
62
103.6
77.0
74.4
74.9
73.0
63
102.2
80.8
79.0
77.2
75.4
64
98.1
72.6
74.0
79.5
77.8
1965
101.5
85.9
81.1
81.9
80.3 .
66
96.6
81.3
84.2
84.2
82.8
67
97.8
87.8
89.8
86.5
85.5
68
99.1
88.6
89.4
88.8
88.2
69
100.9
94.0
92.1
91.1
91.0
1970
98.1
77.5
79.0
93.5
94.0
71
97.1
100.6
103.6
95.8
97.0
72
101.4
107.7
106.2
98.1
100.1
73
99.3
101.1
101.8
100.4
103.3
74
78.8
76.7
, 97.3
102.7
106.6
1975
91.5
97.3
106.3
105.1
110.0
76
93.0
98.0
105.4
107.4
113.5
77
95.6
109.7
117.1
78
109.9
112.0
120.9
79
122.0
114.3
124.7
1980
103.6
116.7
. 128.7
194
-------
C~ -
TABLE C-2 (continued)
Sources: The index of weather effects 1950-1973 is derived from Na-
tional Oceanic and Atmospheric Administration 1973, p. 26. The indexes for
1974, 1975 and 1976 are from Louis Thompson (1977). Actual Yields: 1950-
1978 from USDA, Agricultural Statistics, various years, 1979-1980 from USDA
September 11, 1980. Weather-adjusted yields are actual yields divided by
the index.
*For Ohio, Indiana, Illinois, Iowa, Missouri.
t"Normal" yields = 100, 1973 "technology." Values above 100 indicate
better than "normal" weather.
195
-------
TABLE C-3. SOYBEANS: INDEX OF EFFECTS OF WEATHER ON YIELDS, ACTUAL YIELDS,
WEATHER-ADJUSTED YIELDS, AND TREND VALUES OF WEATHER-ADJUSTED
YIELDS*
Yields (bu.
per acre)
Trend of
Index of
Weather-Adjusted
Weather .
Weather-
Arith-
Loga-
Year
Ef fects+
Actual
Adjusted
metic
rithmic
1950
99.8
23.1
23.1
21.8
22.1
51
96.6
23.1
23.9
22.2
22.4
52
97.7
23.1
23.6
22.7
22.8
53
91.9
19.6
21.3
23.1
23.2
54
97.1
22.1
22.7
23.5
23.5
1955
93.6
21.1
22.6
23.9
23.9
56
96.8
23.1
23.8
24.3
24.3
57
98.2
24.7
25.1
24.8
24.7
58
105.7
26.8
25.3
25.2
25.1
59
97.8
25.8
26.3
25.6
25.5
1960
97.5
25.2
25.9
26.0
25.9
61
101.6
27.6
27.2
26.4
26.3
62
102.6
26.8
26.1
26.9
26.7
63
99.8
27.8
27.9
27.3
27.1
64
96.8
24.7
25.5
27.7
27.5
1965
101.0
27.3
27.0
28.1
28.0
66
98.0
27.3
27.9
.28.5
28.4
67
97.0
26.5
27.3
29.0
28.9
68
99.3
31.0
31.2
29.4
29.3
69
102.7
31.5
30.7
29.8
29.8
1970
98.4
30.2
30.7
30.2
30.2
71
97.2
31.6
32.5
30.6
30.7
72
100.6
32.3
32.1
.31.1
31.2
73
102.6
30.7
29.9
31.5
31.7
74
84.0
24.9
29.6
31.9
32.2
1975
97.6
33.1
33.9
32.3
32.7
76
93.7
29.9
31.9
32.7
33.2
77
35.9
33.2
33.8
78
34.0
33.6
34.3
79
36.4
34.0
34.8
1980
32.7
34.4
35.4
196
-------
TABLE C-3 (continued)
Sources: The index of weather effects 1950-1973 is derived from Na-
tional Oceanic and Atmospheric Administration 1973, page 28. The indexes
for 1974, 1975 and 1976 are from Louis Thompson.(1977). Actual Yields:
1950-78 from USDA Agricultural Statistics, various years; 1979-80 from USDA,
Sept. 11, 1980. Weather-adjusted yields are actual yields divided by the
index.
*For Ohio, Indiana, Illinois, Iowa, Missouri.
fNormal" yields = 100, 1973 "technology." Values above 100 indicate
better than "normal" weather.
197
-------
Year
1950
51
52
53
54
1955
56
.57
58
59
1960
61
62
63
64
1965
66
67
68
69
1970
71
72
73
74
1975
76
77
c-
WHEAT: INDEX OF EFFECTS OF WEATHER ON YIELDS, ACTUAL YIELDS,
WEATHER-ADJUSTED YIELDS AND TREND VALUES OF WEATHER-ADJUSTED
YIELDS*
Index of
Weather
Effectst
.Yields- (bu.' per. acre)
Weather-
Actual Adjusted
Trend of
Weather-Adjusted
Arith- Loga-
metic rithmic
96.3
14.0 .
14.5
14.1
14.7
93.7
13.4
14.3
14.8
15.2
98.6
16.7
17.0
15.5
15.7
92.4
12.6
13.6
16.3
16.3
97.9
14.3
14.6
17.0
16.8
95.7
15.3
15.9
17.7
17.4
87.5
16.1
18.4
18.4
18.0
89.9
19.1
21.2
19.1
18.6
110.1
26.8
24.3
19.8
19.3
91.9
18.1
19.7
20.6
19.9
104.8
24.9
23.7
21.3
20.6
94.7
21.6
22.8
22.0
21.3
98.9
23.1
23.3
22.7
22.1
93.2
21.0
22.5
23.4
22.8
95.3
22.9
24.0
24.1
23.6
96.8
24.1
24.9
24.9
24.4
94.1
22.3
23.7
25.6
25.2
90.5
21.5
23.7
26.3
26.1
100.7
26.5
26.3
27.0
27.0
108.7
29.7
27.3
27.7
27.9
95.7
28.9
30.2
28.4
28.9
104.1
31.9
30.7
29.2
29.9
101. 6
30.5
30.0
29.9
30.9
108.5
32.1
29.6
30.6
32.0
23.9 „
31.3
33.1
26.7
32.0
34.2
25.9
32.7
35.4
27.4
33.4
36.6
28.7
34.2
37.9
32.6
34.9
39.2
27.6
35.6
40.5
198
-------
TABLE C-4 (continued)
Sources: The index of weather effects is derived from National Oceanic
and Atmospheric Administration, 1973, page 27. Actual yields: 1950-1978
from USDA Agricultural Statistics, various years; 1979-1980 from USDA,
Sept. 11, 1980. Weather-adjusted yields are actual yields divided by the
index.
*For North Dakota, South Dakota, Nebraska, Kansas, Oklahoma.
+"Normal" yields = 100, 1973 "technology." Values above 100 indicate
better than "normal" weather.
199
-------
c-^i
only for 1973, 1974, 1975 and 1976. For these years the differences between
weather-adjusted yields and trend values of weather-adjusted yields do not
suggest a break in trend after 1972. For each crop, differences as' large as
those observed in 1973-1976 also occurred occasionally in the years 1950-
1972.
Because we cannot adjust for weather after 1976 (1973 in the case of
wheat) we adopt a different approach in assessing yield behavior in the
more recent years. We ask the "following question: given the historical
effects of weather on yields, what is the probability that weather could
have caused the observed differences between actual yields and the trend
values of weather-adjusted yields? In asking this question we assume that
weather is the main cause of annual differences between actual yields and
the trend values of weather-adjusted yields in 1950-1972, and that other
causes are randomly distributed over the period. This, of course, is at
best an approximation to the truth. We believe it is a sufficiently good
approximation, however, to provide insights into the behavior of actual
yields in recent years. If, for example, in every recent year actual yields
differed from trend values of weather-adjusted yields by more than the
amount observed in any year in the historical record, we would suspect
strongly that something had happened to the trend of weather-adjusted yields
in recent years.
For those years for which we are unable to adjust for weather we take
the difference between actual yields and the trend value of weather-adjusted
yields and apply a chi-square test to judge whether the observed differences
could be due to weather, given the effects of weather on yields in 1950-
1972. For wheat the years considered are 1974-1980. For com and soybeans
they are 1977-1980. For wheat the chi-square test is unequivocal that the
observed yield differences could not be due to weather alone.^ Not only
were the differences greater than would be expected on the basis of weather
experienced from 1950-1972, but observed yields were less than the trend
value of weather-adjusted yields in each year 1974-1980. It is highly prob-
able that there was a tilt downward in the effect of land quality or tech-
nology on wheat yields after 1973. We return to this below.
The chi-square test also indicates that the differences in 1977-1980
between actual yields of corn and the trend values of weather-adjusted
yields could not be attributable to weather alone.^ For corn, however, the
test result is not as clearcut as that for wheat because in 1979 actual corn
In considering the question we use differences between actual yields
and the arithmetic trend values of weather-adjusted yields. For all three
crops the arithmetic trend gave as good a fit to the weather- adjusted yield
data as the logarithmic trend.
2
The test indicated that the probablility of differences as large as
those observed if weather were responsible was much less than 1 in 100.
3
The probability that weather alone was responsible is less than 1 in
100.
200
-------
yields exceeded trend yields by a considerable margin. It seems unlikely
that this would have happened if there had been a definite downward tilt in
the effect of land quality or technology on corn yields in the Cornbelt.
We conclude that there may have been such a tilt, but we shall look for
additional evidence of it.
The chi-square test indicates that.the differences between actual soy-
bean yields and trend values of weather-adjusted yields in 1977-1980 could
reasonably have been caused by the weather.^ There is no convincing evi-
dence, therefore, of a downward tilt in the effect of either land or tech-
nology on soybean yields. Indeed if there was a change in quality of soy-
bean land or technology it more likely gave an upward tilt, since actual
yields exceeded trend yields in three of the four years 1977-1980. However,
we assume no change in land quality or technology effects on the trend of
soybean yields in the Cornbelt. Instead we concentrate on corn and wheat,
attempting to isolate the effects of land quality and technology on yields
of these crops after 1972.
PURE YIELD FACTORS: LAND QUALITY
It is reasonable to assume that for any given level of production far-
mers will use that land which gives the highest net return per acre. This
is not necessarily the land with the highest physical yield per acre, but
we expect there is a high correlation between net return and physical yield.
Without putting too fine a point on it, we assume accordingly that if land
is taken out of production, average yields will rise. If more land is
brought into production, average yields will fall.
Measurement of the effect on yields of changes in the amount of land
is hampered—in fact made possible—by the absence of data showing yield
differences between land in production at any given time and increments or
decrements of land between that time and any other. In the absence of such
data we have taken two approaches to try to establish limits to the effects
of increments or decrements of land on observed yields. In one approach we
ask, what would production have been in period t+1 on land in production in
t if the trend of yields on that land between t-1 and t had continued. The
hypothetical production on this land in t+1 is subtracted from actual pro-
duction in t+1 and the difference is divided by the increment (or decrement)
of land between t and t+1. This gives an estimate of yields on the incre-
ment (or decrement) assuming that yields on land in production in t in-
creased between t and t+1 at the same rate as between t-1 and t. One can
then ask whether the yield estimate on the increment (or decrement) seems
reasonable, for example by comparison with estimated yields on land in pro-
duction in t. If it seems too low, the implication is that yields on land
in production in t grew more slowly between t and t+1 than between t-1 and
t.
4
The probability that weather alone was responsible is about 15 in 100,
201
-------
The second approach begins with an assumption about yields on the incre-
ment (or decrement) of land in relation to yields on land in production in t.
If the amount of land increased between t and t+1 then we assume that yields
on the increment were less than on land in production in t. We then multi-
ply the assumed yield by the amount of the increment and subtract the result
from total production in t+1. The result is an estimate of production in
t+1 on land in production in t. Dividing that estimate by the amount of
land in production in t gives an estimate of yields on that land in t+1.
Comparisons of that yield estimate with yields in t are then free of the in-
fluence of the increment of land on observed yields, assuming, of course,
that the procedure correctly measures that influence.
These procedures, of course, are very rough. Aside from the arbitrary
nature of the assumptions about yields on the increment (or decrement) of
land, it also is assumed that the gross and net changes in land between t
and t+1 are the same. For example, if land in corn is 50 million acres in
t and 60 million acres in t+1, the procedure assumes that the same 50 million
acres is in production in both t and t+1 with an addition of 10 million acres
between the two periods. This may not be the case. It is possible, for
example, that 5 million of the 50, acres is shifted to some other use while
15 million acres formerly in some other use are shifted to corn. We really
should measure the effect on yields of this 15 million acres, but the pro-
cedures we have used cannot do it. The problem probably is not too serious,
however, at least for the period between 1972 and 1977-1980, the one of
greatest interest to us. Harvested land in wheat and corn in the states we
are analyzing increased by a net of about 13 million acres in this period.
The amount of land previously in these crops which was shifted to other uses
in this period probably was small compared to the increment of land. Hence
the difference'between the net and the gross increment likely was small.
Changes in the amount of land in wheat and corn in the states of inter-
est here may have significantly affected the pre- and post-1972 trends of
yields of those crops. Wheat in particular may have been affected. Between
1946-1950 and 1972 the amount of land in wheat in the five wheat states—
North Dakota, South Dakota, Nebraska, Kansas and Oklahome—declined from
36.8 million acres to 25.2 million acres. If in 1946-1950 yields on the
11.6 million acres subsequently taken out of production were 90 percent of
yields on the land remaining in production, then removal of the lower yield-
ing land would have contributed a little over 3 percent to the growth of
weather-adjusted wheat yields from 1946-1950 to 1972. Between 1972 and
1977-1980 an average of 7.9 million additional acres of land in these states
were brought into wheat production, an increase of almost one-third. This
reversal in the declining trend of land in wheat may have significantly re-
duced wheat yields in 1977-1980 below what they otherwise would have been.
There was no significant change between 1946-1950 and 1972 in the
amount of land in corn in the five corn states—Ohio, Indiana, Illinois,
"*P. Weisgerber (1969) estimated that yields on wheat land held in re-
serve in the late 1960s were about 90 percent of land in wheat production.
202
-------
Iowa and Missouri. Consequently, changes in the amount of land could not
have affected the trend of corn yields in that period. Between 1972 and
1977-1980, however, the average amount of land in com in these states in-
creased by 5.4 million acres, or about 18 percent. This may have depressed
somewhat the growth of corn yields between 1972 and 1977-1980.
We have used the procedures described above to estimate the effect on
wheat and corn yields in 1977-1980 of bringing more land under these crops
after 1972. The results are in table C-5. The purpose of the exercise is
to estimate the contribution of additional land to the differences between
actual yields and weather-adjusted trend yields in 1977-1980 (i.e., the dif-
ferences between lines 1 and 2 in table C-5). If all of the difference were
attributable to the additional land, then yields on that land would have
been as shown in line 7, i.e. yields on the increment of wheat land would
have averaged 11.9 bushels per acre in 1977-1980, and yields in the incre-
ment of corn land would have averaged 71.3 bushels. These figures are so
low relative to 1972 actual yields (107.7 bushels for corn and 30.5 bushels
for wheat, tables C-2 and C-4) that we can conclude quite confidently that
while yields on the increments of land might well have been lower than
yields on 1972 land, they could not have been so much lower as to explain
all the difference between actual yields in 1977/80 and the trend values of
weather-adjusted yields.
Lines 8, 9, and 10 in table C-5 present an alternative approach to
estimating the effect of the increments of land in wheat and corn on yields
in 1977-1980. For wheat this approach indicates that yields on 1972 land
would have been 29.6 bushels in 1977-1980 (line 9), 4.9 bushels less than
the trend value of weather-adjusted yields (line 10), and .5 bushels more
than actual yields on all wheat land (line 9 minus line 2). This suggests
that the increment of wheat land was responsible for only about 10 percent
of the difference between actual wheat yields and the trend value of wea-
ther-adjusted yields in 1977-1980 (.5 divided by 5.4, the total yield short-
fall) . This understates the effect of the increment of wheat land on yields
because the pre-1972 trend of yields was raised slightly by the reduction in
land in wheat in that period. That effect was small, however, as noted
above. So the switch from decreasing the amount of wheat land to increas-
ing it probably was not responsible for much more than 10 percent of the
shortfall between actual yields and the trend value of weather-adjusted
yields. Could the other 80+ percent be attributable to weather alone or did
technology also play a role? We return to this question below.
A comparable analysis for corn indicates that the increment of corn
land was responsible for about two-thirds of the shortfall in corn yields,
given the various underlying assumptions. In the analysis of weather ef-
fects on corn yields in 1977-1980 we concluded that weather could not likely
explain all the difference between actual yields and trend values of wea-
ther-adjusted years. If the increment of land in corn from 1972 to 1977—
1980 explains as much as two-thirds of the difference, however, we think it
likely that weather could explain the other one-third. It seems unneces-
sary, therefore, to search further for an explanation of the yield short-
fall in the five Cornbelt states in 1977-1980.
203
-------
TABLE C-5. ANALYSIS OF THE EFFECTS OF CHANGING THE AMOUNT OF
LAND ON YIELDS OF WHEAT AND CORN
Wheat
1977/80
Corn
1977/80
1) Yields if pre-1972 trend in weather-
adjusted yields had continued 34.5
2) Actual yields (bu.) 29.1
3) Production in 1977/80 on 1972 land if
yields had grown at pre-1972 rate (mill, bu.) 869.4
4) Actual production in 1977/80 (mill, bu.) 963.2
5) Line 4 minus line 3 (mill, bu.) 93.8
6) Increment of land 1972 to 1977/80 (mill, bu.) 7.9
7) Yield in 1977/80 on increment of land
(line 5 divided by line 6 -bu.) 11.9
8) Production on increment of land if yield
on the wheat increment were 90% of 1972
yield and on the corn increment were 82%
(mill, bu.)* 216.9
9) Yield in 19 77-80 on 1972 land if yield
on the increments of land were 90% of
yield on 1972 land for wheat and 82%
for com (bu.) 29.6
10) The yield shortfall on 1972 land in
1977/80 if yield on increments of land \
were as stated in line 9 (bu.) 4.9
113.2
107.7
3452.6
3866.4
413.8
5.8
71.3
512.2
111.4
1.8
*The percentages are from Weisgerber (1969). Wheat yields are
averages for Oklahoma, Kansas, Nebraska, South Dakota and
North Dakota. For corn the states are Ohio, Indiana, Illinois,
Iowa and Missouri.
204
-------
The analysis of corn yields has focused on the Cornbelt because in that
region we were able to adjust yield data for effects of weather. Corn
yields in the rest of the nation, however, moved much as they did in the
Cornbelt after 1972, that is, the rate of increase slowed considerably. The
amount of harvested land in corn outside the Cornbelt increased 8.3 million
acres, or 30 percent, from 1972 to 1977-1980, substantially more than in the
Cornbelt both in absolute amount and in percentage. Since our analysis of
Cornbelt yields suggests that increased acreage was responsible for a major
part of the slowdown in yield growth after 1972, it seems likely that the
slower growth of yields outside the Cornbelt also may be attributable large-
ly to the additional acreage.
PURE YIELD FACTORS: TECHNOLOGY
Fertilizer
Table C-6 shows fertilizer use data for corn and wheat. The data are
consistent with substantially slower growth in yields of both crops after
1970-1972. From 1964-1966 to 1970-1972 wheat yields increased 1.05 bushels
per acre per year. From 1970-1972 to 1977-1979 wheat yields declined
slightly. Corn yields grew 3.2 bushels per acre per year from 1964-1966 to
1971-1973 and by 1.4 bushels annually from 1971-1973 to 1977-1979.
In the discussion of the effects of weather and additional land on
corn yields in the Cornbelt we concluded that those two factors together
could have been responsible for the slower growth of corn yields after 1972.
The data in table C-6 suggest, however, that the slower growth in amount of
fertilizer applied per acre in corn must also have contributed.
We concluded that additional land did not explain much more than 10
percent of the shortfall in wheat yields in 1977-1980 and asked the ques-
tion: could weather explain the remaining 80+ percent or did technology
also play a role? Table C-6 suggests that technology did play a role but
the suggestion is not as strong as in the case of corn. The percentage of
land receiving any fertilizer is much less for wheat than for corn, and so
is the amount applied per harvested acre. Consequently the contribution of
fertilizer to the growth of wheat yields must have been much less decisive
than for corn. It follows that the slower growth of fertilizer applied to
wheat land after 1972 would have had less impact in slowing the growth of
wheat yields after that date.
£
We used 1971-1973 for corn instead of 1970-1972 because 1970 yields
were severely depressed by an attack of leaf blight.
205
-------
TABLE C-6. FERTILIZER USE ON CORN AND WHEAT IN THE UNITED STATES
Amount per re- Amount per
Acres
receiving
ceiving acre
Harvested
acre
(percent)
(lbs.)
(lbs,
.)
N
P
K
N
P
K
N
P
K
Tota]
Corn
1964-1966
88
82
76
73
22
39
64
18
30
112
1970-1972
95
89
84
111
66
68
105
59
57
221
1977-1979
96
88
82
130
68
82
125
60
67
252
Wheat
1964-1966
48
37
15
30
13
25
14
5
4
23
1970-1972
60
43
16
42
34
37
25
15
6
46
1977-1979
63
42
18
53
37
39
33
16
7
56
Sources:
USDA, Jan.
1972,
Dec.
, 1974
and Dec.
1979.
The data begin
with 1964
N = nitrogen
P = phosphorus
K = potassium
Irrigation^
Irrigated production of corn and wheat grew rapidly from 1950 to 1977—
4748 percent for corn and 470 percent for wheat—reflecting both an increase
in irrigated acreage in these crops and rising yields. However, total
irrigated production of corn and wheat was small relative to dryland pro-
duction. Thus irrigation accounted for only 28 percent of the national
increase in corn production and for 12 percent of the increase in wheat
production. Yields of irrigated corn and wheat exceed dryland yields, and
irrigated yields increased from 1950 to 1977. Consequently, irrigation con-
tributed to rising yields of corn and wheat both because of the expansion
of irrigated acreage relative to dryland acreage in these crops and because
of rising yields on the irrigated land. Nonetheless, because irrigated
acreage in and production of corn and wheat were small relative to dryland
acreage and production, the contribution of irrigation to the increase in
national yields between 1950 and 1977 was small—13 percent for corn and 7
percent for wheat.
This discussion is based on Frederick. We deal only with western
irrigation, since to date the contribution of eastern irrigation to the
growth of yields of crops of interest here has been negligible.
206
-------
Reliable data on changes in the amount of irrigated land in corn and
wheat since 1972 are not available. However, most of the expansion of irri-
gated land in these crops for at least ten or fifteen years before 1972 was
in the Plains region, particularly Nebraska and Kansas. Discussion with
people in the region who follow irrigation suggests little slowdown in the
rate of expansion since the early 1970s. If this is true, irrigation would
not likely have contributed to slower growth of national corn and wheat
yields after 1972. The effect on yields of even relatively large changes
in irrigated land would be difficult to detect in any case because of the
relatively small contribution of irrigation to the change in national yields.
SUMMARY
There is no evidence of a change in the trend of yields of soybeans in
the 5 Cornbelt states after 1972. The observed differences between actual
yields and trend values of weather-adjusted yields would easily have re-
sulted from variations in the weather.
Actual corn yields fell short of the trend values of weather-adjusted
yields in three of the four years 1977-1980. This cannot be attributed
solely to the weather. The expansion of corn acreage after 1972 probably
accounts for a substantial proportion—perhaps as much as two-thirds—of the
yield short fall. Weather might explain much of the remainder, but slower
growth in fertilizer applied per acre after 1972 likely contributed also.
Actual wheat yields in the Great Plains were less than the trend values
of weather-adjusted yields in every year from 1974 to 1980. Weather could
not have caused this and the expansion of land in wheat probably contributed
only about 10 percent to the shortfall. Par acre applications of fertilizer
to wheat land increased more slowly after 1972 than before, but the per-
centage-of wheat land receiving fertilizer is too small for this factor to
have been very important. There is no evidence to suggest that slower
growth in irrigated wheat acreage contributed to the yield shortfall. We
are left without an adequate explanation of the failure of wheat yields to
grow after 1973 in accordance with the pre-1973 trend of weather-adjusted
yields.
207
-------
REFERENCES
Frederick, Kenneth. "Irrigation and the Adequacy of Agricultural Land,"
forthcoming in The Adequacy of Agricultural Land in the United States,
Resources for the Future, Washington, D.C.
National Oceanic and Atmospheric Administration. 1973. The Influence of
Weather and Climate on United States Grain Yields: Bumper Crops or
Droughts. A Report to the Administrator of NOAA, December 14, 1973.
Washington, D.C.
Thompson, Louis M. 1977. "Climatic Variability and World Grain Production,"
paper given at a meeting of the American Seed Trade-, Kansas City, Mo.,
November 8, 1977.
USDA, Agricultural Statistics, various years.- U.S. Government Printing
Office, Washington, D.C.
, Jan. 1972. Fertilizer Situation. FS-2. Washington, D.C.
, Dec. 1974. Fertilizer, Situation., .FS-5. Washington, D.C.
, Dec. 1979. Fertilizer Situation. FS-10. Washington, D.C.
, Sept. 11, 1980. Crop Production. Crop,; Reporting Board. CrPr 2-2
(9-80). Washington, D.C.
Weisgerber, P. 1969. Productivity of Converted Cropland. USDA, Economic
Research Service, ERS-398. Washington, D.C.
208
-------
APPENDIX D
PROJECTIONS OF REGIONAL SHARES OF PRODUCTION
AND OF REGIONAL CROP YIELDS
REGIONAL SHARES
We noted in Section 2 that there is considerable potential for in-
creased production of irrigated soybeans in the Mississippi Delta, and some
potential for expansion of irrigated corn and soybean production in the
Southeast. Accordingly, we have projected an increase in the share of the
Delta's production of soybeans and in the Southeast's share of both soy-
beans and corn. Most of the increase would occur after 1985. The South-
east is allocated the greatest percentage increase in production of soy-
beans because that region has more land in pasture and forest with potential
fnr conversion to cropland than the Delta, both in absolute amount and in
relation to its present cropland base.-*- The increased shares of the Delta
and the Southeast in soybean production would come at the expense of the
Cornbelt, although that region would continue to be by far the nation's
leading producer of soybeans. However, as will be pointed out below, our
projections of crop production and yields in the Cornbelt imply heavy pres-
sure on the region's supply of cropland by 2010, including land now in
forest, pasture, and range with potential for conversion to cropland. In-
deed, our projections of crop production and yields imply that the amount
of land in crops in the Cornbelt would exceed the present and potential sup-
ply by 2010 unless the region's shares of production of some crops decline.
We project a small decline in the share of feedgrain production, but we
expect a much larger decline in the ;share of soybeans because of the demon-
strated potential for expanded soybean production in both the Delta and the
Southeast.
The study by Shulstad et. al. (1980) of irrigation potential in the
Mississippi Delta indicates that double-cropping of wheat and soybeans would
be one of the more profitable options. If soybean production in the Delta
expands as we expect, it is likely that wheat production would increase
also. This is reflected in Table 4-1 (Section 4) in the sharp relative
increase in wheat production projected for the Delta.
We expect continued growth in the share of the Southern Plains in wheat
production, reflecting a process evident for some years, and consistent with
the region's relative abundance of land In range and pasture with potential
^Data on potential cropland are from USDA, Feb. 1980. These data are
presented in Table 4-6 in Section 4.
209
-------
(TN —
for conversion to cropland. The principal losers in shares of wheat produc-
tion would be the Cornbelt and Northern Plains, although the later region
would remain by far the most important wheat producer in the nation.
We expect the Southern Plains to increase its share of cotton produc-
tion also, continuing the trend evident over the last couple of decades.
However, future conditions are likely to favor the Southern Plains (mainly
Texas) relative to Arizona and California as well as relative to the South-
east and the Mississippi Delta. Most cotton production in the two western
states is irrigated and will be exposed, therefore, to the increasing con-
straints on western irrigation noted in Section 2. By contrast only 40-45
percent of 4:he cotton land in Texas is irrigated. Moreover, research in
Texas has developed cotton varieties especially well adapted to growing con-
ditions in that state, and which reduce the pest management problem relative
to its dimensions in the Southeast and Delta.^
Apart from factors already discussed, our projections of regional
shares of production also were influenced by comparison of two runs of Iowa
State University's model of the agricultural economy of the United States.
The runs were done especially for this study.^ In one of the runs we sup-
plied our projections of crop production and a preliminary allocation of
production among the ten USDA producing regions." In the second run we used
the same nationwide projections of production, but asked" the model to allo-
cate the production among regions. The-results o,f the two runs are shown
in Table D-l. If we accept the model's allocation as that which would occur
as farmers seek to minimize costs of production and transport among markets,
then we would conclude that the regional distribution of production result-
ing from run 2 is more likely than that'resultirig from our initial distri-
bution imposed upon the model, represented by run 1. . We do not accept the
model's distribution as the ultimate truth, but we do believe it should be
given some weight. Accordingly, where, in our judgment, our initial distri-
bution differed significantly from'the model's distribution we adjusted the
initial distribution to bring it closer to the model's. These, adjustments
were for feedgrains among the Great Lakes,' Ohio,"Upper Mississippi, and
Missouri regions;-for wheat among the Upper and Lower Mississippi'and the
Arkansas-White-Red: and for oilmeal (soybeans) among the.Lower Mississippi,
the Missouri", and the Arkansas^hi'tiS-Red"We judged-the-distrltoufrten-or
cotton in the two runs to be similar enough to need no adjustment.
The various adjustments are incorporated in the regional distribution
shown in Table 4-1 (Section 4).
2
There is more on this in the discussion in Section 5 of trends in
insecticide use oh: cot-toil.
3
The model was'developed by- E&rl Heady and'associates;. .It^is ...a. linear
programming type which allocates production of major crops among regions in
such fashion as to minimize production and transportation costs. The mo-
del's structure does not permit allocation of production among USDA produc-
tion regions. Instead the regions are those shown in Table D-l.
210
-------
0-3
TABLE D-l. DISTRIBUTION OF PRODUCTION OF SELECTED CROPS AMONG RIVER BASINS
IN- THE UNITED STATES IN 2010 (PERCENT)
Region or Basia
Run
1
Run
2
Feed-
grains
Wheat
0il-t
meal
Cotton
Feed—
grains
Wheat
0il- +
'meal
Cotton
New England
*
0
0
0
0
0
0
0
Middle Atlantic
2.6
.7
3.1
0
1.4
2.2
3.2
0
South. Arlaneic-Gulf
6.7
.5
15.8
8.0
4.1
1.9
13.0
9.2 "
Great Lakes
3.8
5.2
5.7
0
16.1
2.7
1.8
0
Ohio.
14.1
16.7
0
10.7
4.3
20.0
0
Tennessee
.4
*
3.7
0
.3
0
1.5
0
Upper Mississippi
27.1
9 .'4
24.7
0
37.1
2.4
22.0
0
Lower Mississippi
.1
1.3
14.4
1.2
1.5
7.7
a.6
1.4
Souris-Red-Raiaey
.1
8.6
2-9
0
0
8.5
2.0
0
Missouri
25.6
31.3
i.9
0i
16.3
34.5
14.2
0
Arkansas-White-Red
7.3
20.4
3.2
9.6
5.5
14.4
9.2
11.1
Texas-Gulf
3.4
1.0
3.9
48.7
4.0
1.7
3.6
44.6
Rio Grande
.5
.2
'.2
¦ 6:5
.!8
.1
5.3
Upper Colorado
.1
1.2
Q ,
9
.1
;L,5,
.0
0
Lower Colorado
.1
.6
.1
'5/3
*
.0"
.1
4.5
Great Basin
.L
0;:
:Q,
. 0
0
O
0
Columbia-North. Pacific
.6
13.2
0
o-
.3
14.4
0
0
Calif ornia-Soutlu -Pacific-; 2.1
.6
. 20.7
'2*2;
3.2
• 6
23.9
Source: Iowa State University model of the U.S. agricultural economy.
*Less than 1 percent
tVirtually all o£ this is soybean.
Note: In run l the distribution among the USDA producing regions was
imposed by RFF. The model then allocated among the regions in the tablfe.
In run 2 the nibdel allocated national production-among .the regions.
211
-------
\>Lj
YIELDS
We expect prices and productivities of inputs to favor more land-using
technologies, as they have since the early 1970s. In general, therefore, we
take yield behavior in the 1970s as a guide in projecting future yields.
For feedgrains, however, we assumed that yields would not grow as rapidly
as experience in the 1970s might suggest. The main reason is that we expect
the price of fertilizer in the future to behave differently relative to the
price of corn then it did in 1973/1979. From 1973 to 1975 the price of
nitrogen rose sharply relative to the price of corn and the amount of.nitro-
gen applied per acre of corn declined.^ From 1975-to 1979 the price of
nitrogen declined relative to the price of corn and per acre applications
rose. This pattern must have tended to depress corn yields in 1973/75 and
then raise them to 1979. As indicated in Section.2, we expect the price .of
nitrogen to rise relative to the price of corn over the next several decades,
in contrast to its behavior in 1975/1979, and that,the amount of nitrogen
applied per acre of corn will rise less rapidly than from 1975 to 1979.-*
Accordingly, we expect corn yields to rise more.; slowly from 1979; to 2010
than from 1973 to 1979. Since corn currently accounts for about-70 percent
of the land in feedgrains, and for an even larger percentage in our projec-
tions, the slower growth of corn yields implies slower growth of feedgrain
yields.
We expect soybean yields in the Cornbelt, Delta and Southeast to grow
more rapidly than national yields. Our analysis in Section 3 of soybean
yields in the Cornbelt indicated that the growth of yields in that region
did not slow after 1972, as they did nationwide, and we see no reason to
expect different performance in the future. Consequently we projected soy-
bean yields in the Cornbelt to grow to 1985 and 2010 in accord with the
arithmetic trend of weather-adjusted yields as shown in Table C-3. ,
Our discussion of eastern irrigation indicated considerable expansion
potential for irrigated soybeans in the Mississippi Delta and, to a lesser
extent, in the Southeast. To reflect this we projected a significant in-
crease in soybean yields in those regions, using yield estimates in Shul-
stad et. al. as a guide. Their study also indicates greatly increased po-
tential for double-cropping wheat with soybeans in the Delta, and this looks
promising also in the Southeast. We increased national wheat yields over
what the trend since 1973 would suggest to reflect this potential.
4
See Table 2-4 for the ratio of nitrogen prices to corn prices. Data
on nitrogen applied to corn are from USDA December 1974 and December 1979.
^Projections of fertilizer are developed in Section 5.
212
-------
REFERENCES
Shulstad, R. M., R.; D. May, arid B. E. Herrington. 1979. Cropland Gonver-r
sion Study for the Mississippi-Delta Region. A study- done;' for ¦
Resources for the Future. University of Arkansas, Fayetteville-.
USDA. Dec. 1974. Fertilizer Situation. FS-5. Washington, D.C.
. Dec. 1979. Fertilizer-Situation. FS-10. Washington, D:Co
. Feb. 1980. Basic Statistics-: 1977 National Resources. Inventory,
Revised. Soil CoriSerVcUiion Service, Washington, D.C.
213
------- |