v>EPA
United States
Environmental '
Agency
of Health and
Ecological Effects
Washington DC 20460
EPA-600/5-79-001c
February 1979
Rest'.'; merit
Methods Development
for Assessing Air
Pollution Control
Benefits
Volume III,
A Preliminary Assessment of
Air Pollution Damages for
Selected Crops Within
Southern California
-------
OTHER VOLUMES OF THIS STUDY
Volume I, Experiments in the Economics of Air Pollution Epidemiology,
EPA-600/5-79-001a.
This volume employs the analytical and empirical methods of economics to
develop hypotheses on disease etiologies and to value labor productivity and
consumer losses due to air pollution-induced mortality and morbidity.
Volume II, Experiments in Valuing Non-Market Goods; A Case Study of
Alternative Benefit Measures of Air Pollution Control in the South
Coast Air Basin of Southern California, EPA-600/5-79-001b.
This volume includes the empirical results obtained from two experiments
to measure the health and aesthetic benefits of air pollution control in the
South Coast Air Basin of Southern California.
Volume IV, Studies on Partial Equilibrium Approaches to Valuation of
Environmental Amenities, EPA-600/5-79-001d.
The research detailed in this volume explores various facets of the two
central project objectives that have not been given adequate attention in the
previous volumes.
Volume V, Executive Summary, EPA-600/5-79-001e.
This volume provides a 23 page summary of the findings of the first four
volumes of the study.
This document is available to the public through the National Technical
Information Service, Springfield, Virginia 22161.
-------
EPA-600/5-79-001c
February 1979
METHODS DEVELOPMENT FOR ASSESSING
AIR POLLUTION CONTROL BENEFITS
Volume III
A Preliminary Assessment of Air Pollution Damages for
Selected Crops Within Southern California
by
Richard M. Adams
University of Wyoming
Laramie, Wyoming 82071
Narongsakdi Thanavibulchai
University of Wyoming
Laramie, Wyoming 82071
Thomas D. Crocker
University of Wyoming
Laramie, Wyoming 82071
USEPA Grant No. R805059-01
Project Officer
Dr. Alan Carlin
Office of Health and Ecological Effects
Office of Research and Development
U.S. Environmental Protection Agency
Washington, D.C. 20460
OFFICE OF HEALTH AND ECOLOGICAL EFFECTS
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
WASHINGTON, D.C. 20460
EPA - RIP
-------
DISCLAIMER
This report has been reviewed by the Office of Research and Development,
U.S. Environmental Protection Agency, and approved for publication. Approval
does not signify that the contents necessarily reflect the views and policies
of the U.S. Environmental Prctection Agency, nor does mention of trade names
or commercial products constitute endorsement or recommendation for use.
-------
PREFACE
This project was initiated in the summer of 1977 when Dr. Alan Carlin
urged a research group from the Resource and Environmental Economics Labora-
tory at the University of Wyoming that was studying the economic benefits of
air pollution control in the South Coast Air Basin, to devote some attention
to the benefits that could accrue to the nonurban sector. Extensive use has
been made of library facilities at the Berkeley, Davis, and Riverside cam-
puses of the University of California. Drs. H. Johnson, O.C. Oshima, O.C.
Taylor, and C.R. Thompson of the University of California, Riverside, have
been extremely helpful in directing us to specific library materials. The
California Air Resources Board provided detailed air pollution data. Compu-
tational assistance has been provided by the University of Wyoming Computer
Center.
111
-------
ABSTRACT
This study investigates the economic benefits that would accrue from re-
ductions in oxidant/ozone air pollution-induced damages to 14 annual vegetable
and field crops in southern California. Southern California production of
many of these crops constitutes the bulk of national production.
Using the analytical perspective of economics, the study provides an up-
to-date review of the literature on the physical and economic damages to agri-
cultural crops from air pollution. In addition, methodologies are developed
permitting estimation of the impact of air pollution-induced price effects,
input and output substitution effects, and risk effects upon producer and con-
sumer losses. Estimates of the extent to which price effects contribute to
consumer losses are provided. These consumer losses are estimated to have
amounted to $14.8 million per year from 1972 to 1976. This loss is about
1.48% of the total value of production for the included crops in the area and
0.82% of the value of these crops produced in the State of California. Celery,
fresh tomatoes, and potatoes are the sources of most of these losses.
IV
-------
CONTENTS
Abstract iV
Tables vii
Chapter I Introduction 1
1.1 The Problem Setting 1
1.2 Scope of the Study Analysis 3
1.3 The Agricultural Sector: An Overview 4
1.4 Purpose and Objectives 9
1.5 Plan of Presentation 10
Chapter II Agricultural Crop Damages by Air Pollution -
A Review of Literature 12
2.1 Introduction 12
2.2 Physical Damages of Crops by Air Pollution 12
2.3 Economic Damages of Crc ps by Air Pollution 17
2.4 Measurement of Air Pollution Damages: Air
Pollution Response Functions 23
2.5 Air Pollution Response Functions and Crop
Loss Equations 24
2.6 Conclusion 27
Chapter III Some Methodological Considerations on the
Assessment of Air Pollution Damages: A
Proposed Mathematical Framework 29
3.1 Introduction 29
3.2 Methodological Framework 29
3.3 An Analytical Model for Measuring Impacts of
Air Pollution on Agricultural Crops 38
Chapter IV Air Pollution Yield Response Relationships 42
4.1 Introduction 42
4.2 A Hypothetical Relationship Between Air
Pollution and Yield 42
4.3 Methods of Estimating Effects of Air Pollution
Concentration on Yield 43
4.4 Estimated Results of Yield Reduction Due to
Air Pollution 50
Chapter V Price Forecasting Equation Estimation 59
5.1 Price Forecasting Equation Estimation Procedure .... 60
5.2 Price Forecasting Equations for Vegetable
and Field Crops 61
v
-------
Chapter VI An Economic Assessment of Crop Losses Due
to Air Pollution: The Consuming Sector 79
Chapter VII Implementation of the Complete Model;
An Assessment 89
7.1 Production Adjustments 89
7..2 Consumer Impacts 90
7.3 The Integrated Model • . . 91
7.4 Related Research Needs 91
7.5 Concluding Comment 92
References 93
VI
-------
TABLES
Number Page
1.1 United States and California Crop Production:
Specific Vegetable and Field Crops, 1976 5
1.2 Crop Acreage Harvested, by Region 1972-76
and 1976 6
1.3 Average Annual Crop Production and Market Shares,
by Region, 1972-76 ' 7
1.4 1976 Regional Crcp Production 8
4.1 Correlation Values between Level of Oxidant
and Yield 44
4.2a Regression Coefficients for Selected Crops,
Orange County, 1957-76 45
4.2b Regression Coefficients for Selected Crops,
Riverside County, 1957-76 46
4.2c Regression Coefficients for Selected Crops,
Kern County, 1957-76 47
4.3 A "Rule-of-Thumb" Relating Leaf Damage to
Yield Reduction 49
4.4 Levels of Oxidant/Ozone Concentration (pphm) 51
4.5 Percentage Yield Reduction for 12 Hour Exposure,
Using the Average Value of Oxidant/Ozone
Concentration from 1972-1976, "C" Level 52
4.6 Percentage Yield Reduction for 12 Hour Exposure,
Using the "C" Level of Oxidant/Ozone
Concentration for 1976 53
4.7 Percentage Yield Reduction Averaged Over County,
for each Region, by Time Period 54
VII
-------
Number Page
4.8 Actual Yield Per Acre (in the Presence of
Air Pollution) 55
4.9 Potential Yield Per Acre (without Air Pollution
Effects 56
5.1 Price Forecasting Equations for Lettuce and
Fresh Tomato, By Season 62
5.2 Estimated Price Flexibility for California
Processing Tomatoes, 1948-1971 64
5.3 Price Forecasting Equations for Potatoes,
By Season 66
5.4 Price Forecasting Equations for Celery, Cantaloupes,
and Broccoli, By Season 68
5.5 Price Forecasting Equations for Carrots, Cauliflower,
Onions and Beans, By Season 72
5.6 Summary of Price Forecasting Equations 76
Appendix
A Seasonal Patterns of Production for Selected
Vegetable Crcps in California 77
6.1 Production Without Air Pollution 80
6.2 Changes in Production Due to Air Pollution 82
6.3 Seasonal Vegetable Crop Production by Region
in California 83
6.4 Changes in Crop Price Due to Air Pollution,
1972-76 and 1976 84
6.5 Consumers' Surplus at Mean (1972-1976) Consumption,
Using the Mean Value (1972-1976) Level of Oxidant
Concentration 85
6.6 Consumers' Surplus at 1976 Consumption Levels,
Using the 1976 Level of Oxidant Concentration 86
vni
-------
CHAPTER I
INTRODUCTION
1.1 The Problem Setting
Agricultural production, even in the most advanced countries, is
heavily influenced by factors that are beyond the: producer's control.
Despite a tremendous increase in per unit agricultural yields during the
past three decades due, in part, to successful breeding of high yield and
disease resistant varieties of plants, favorable weather conditions,
substantial uses of fertilizer, insecticides, and modern farm machinery,
aggregate world food production has not kept pace with world population
growth. Further, within the more industrialized countries yield plateaus
appear to have been reached for specific crops. On a site specific basis,
such a leveling of yields may be partially attributed to man-induced
environmental factors, such as shifting production to soils of lower
inherent productivity and the general degradation of environmental quality,
including ambient air quality levels. The existence of such environmental
problems may not be critical in developing or non-industrialized countries
where agricultural production is still largely at a subsistence level.
However, within industrialized nations, the encroachment of urban and
industrial growth into regions of agricultural production bring attendant
problems for agriculture, including those associated with air pollution.
The problem of air quality and agricultural production is partially
pronounced on a regional basis.
Some agricultural crops, such as vegetables and fruits, tend to dis-
play highly concentrated geographical production patterns due to specific
climatological requirements. An example of such a region is the South
Coast Air Basin of California. Given the concentration of such production,
and the adverse effects of air pollution on vegetables and fruits (which
are highly perishable), one might expect price fluctuations for such
commodities in response to changes in air quality. Any depression of
yields due to the presence of air pollution may affect consumers and
producers of those commodities differentially, depending on the price
elasticity of demand (or the price flexibility coefficients, if emphasis
is-on direct price effects). That is, if the price elasticity of demand
for, say, celery is inelastic, consumers would suffer a net income loss,
while producers on the aggregate will benefit from the increase in price
of celery due to the reduction in celery supply.
The fact that air pollution poses problems in certain delineated
basins in California is well documented. Such air pollution problems
-------
appear most severe in the South Coastal Air Basin of the state. Injury
to vegetation from photochemical oxidants was first characterized in
in the Los Angeles area [Middleton, Kendrick and Schwalm, 1950], but was
soon recognized over a large part of Southern California as well as in the
San Francisco Bay area [Middleton, Barley and Brewer, 1958]. Moreover, the
high level of such potentially harmful photochemical oxidants and particu-
lates observed in the South Coast Air Basin are no longer confined to the
delineated area but rather extend east into the Mojave Desert and Imperial
Valley as well as northwest into the Ventura-Oxnard Plain. Areas of
previously low air pollution concentrations, such as the San Joaquin and
Central Coast valleys, are experiencing potentially damaging levels of
concentration.
The general effects of air pollution on vegetation are also well
documented._!/ While some effects, at the individual level, may be primar-
ily aesthetic, substantial economic costs to society in terms of deleter-
iou,s effects on production relationships are also incurred. These effects,
as applied to agricultural crops, may be pronounced in terms of depressed
yields and resultant increases in output prices.
Within agricultural crops, different species vary over a considerable
range in their susceptibility to injury by air pollution. These differ-
ences appear to be due primarily to differences in the absorption rate of
toxic substances by plant leaves. Succulent leaf plants (with the excep-
tion of corn) of high physiological activity are generally sensitive,
whereas those with fleshy leaves and needles are resistant. For these
reasons, it is necessary to find the appropriate air pollution response
function for each crop so that the level of yield reduction, if any, due
to different levels of air quality can be determined within the specified
area .
The physical effects of air pollutants on agricultural crops have
long been recognized [Brandt and Heck, 1968]. The adverse effects of air
pollution were recorded as early as 1874 [Cameron]. However, most research
in this area has concentrated on physical damages. There have been rela-
tively few research efforts directed at the economic impacts of air pol-
lution on agricultural crops. Perhaps one reason is that individuals who
traditionally carry out such studies are primarily biologists, biochemists,
plant pathologists, or other scientists more interested in physical rather
than economic or monetary losses to plants and agricultural crops due to
air pollution. Another reason is that it is more difficult to adequately
evaluate economic losses due to a wide range of stochastic factors, such as
possible input and output price fluctuation, for the commodities being
considered. To date, there does not appear to be a theoretically accept-
able means of measuring such economic losses. Of those studies directed at
economic losses, most employ the survey method and calculate the. damages
quantitatively by simply multiplying the estimated reduction of yield by a
fixed price [see Middleton and Paulus, 1973;2_l Lacasse, Weidensaul and
Carroll, 1969; Benedict, Miller and Smith, 1973; Thompson and Taylor, 1969;
Thompson, Kats and Hensel, 1971; Thompson, 1975].
-------
Given the importance of the South Coastal and contiguous regions in
the production of specific crops, increasing (or even constant) levels of
air pollution such as photochemical oxidants, may portend significant
changes in this regional agricultural production. Such agricultural
adjustments may adversely affect consumers, given the general range of
income elasticities and price flexibilities observed for many crops grown
in this area. The effects of air pollution on producers are uncertain, as
some compensating variation in the form of changes in output prices may
offset some production effects. Nevertheless, it is likely that resource
owners and input suppliers would experience lower rates of return.
As mentioned above, farm-level prices of some agricultural crops
fluctuate widely, due in part to changes in production levels. The prices
of some agricultural commodities may rise or drop more than 50% within a
certain time period [see Tomek and Robinson, 1972, p. 2], depending on the
magnitude of the price flexibility coefficient. Therefore, prices, under
such situations, cannot reasonably be taken as given. In addition, most
studies do not consider distributional effects due to air pollution, such
as welfare gains and losses across consumers and producers. Such effects
may be of more interest to policymakers than just the dollar value of
agricultural losses.
1.2 Scope of the Study Analysis
Vegetable production in the United States is dominated by California
in the aggregate and on a seasonal basis. Within certain regions of
California, air pollution in the form of oxidants has been a chronic
problem. This is particularly pronounced in parts of the South Coastal
region encompassing Los Angeles and surrounding areas. The South Coastal
region is also an important vegetable producing region on a seasonal basis.
In addition, levels of oxidants have been increasing in contiguous
production regions, such as the Imperial Valley, Southern San Joaquin
Valley and Central Coast (Salinas Valley). These regions, when combined
with the South Coast, constitute the principal fresh vegetable production
region in the U.S. These regions are included in this analysis in an
attempt to capture the comparative advantage across regions; i.e., in-
creasing levels of air .pollution in one region vis a vis contiguous regions
may result in structural changes in the agricultural sector as growers
attempt to ameliorate for the presence of air pollution. Such modifica-
tions in behavior may be in the form of changed cropping mixes, increased
costs or shifts in location of production. The net effect may be reduced
market shares for the affected region and altered producer revenues. Thus,
for the purpose of this study, the delineated study area contains four
production regions identified as the South Coast, Central Coast, Southern
San Joaquin and Southern Desert.^/ These regions appear to constitute an
appropriate area in which to analyze the interface between air pollution
and crop production.
At present, the economic analysis of crop damage is limited to 14
annual vegetable and field crops. Perennials, such as alfalfa, citrus and
3
-------
fruits, are excluded due to Che complex time horizons associated with such
crops. Also, from the standpoint of substitution possibilities (one aspect
of the analysis), annual crops offer a more diverse set of opportunity.
The annual crops selected for inclusion represent the major vegetable and
field crop commodities grown within the region. All had gross values in
excess of $8 million in 1976. The list of vegetable crops includes: beans
(lima), broccoli, cantaloupes, carrots, cauliflower, celery, lettuce (head),
onions (fresh and processed), potatoes and tomatoes (fresh and processed).
In addition to the 12 vegetable crops, two field crops are included: cotton
and sugarbeets. Acreage and production figures for the included crops, by
subregion and for the state, may be gleaned from Tables 1.1 through 1.4.
While a number of air pollutants are known to cause physical damage to
plants, the emphasis of this study is on one specific type of air pollutant
- oxidants/ozone. The selection of ozone concentration as the ambient air
quality parameter is based on the magnitude of ozone in terms of total air
pollutants. Within California, oxidants/ozone comprise approximately 50%
of total pollutants. Further, ozone appears to be the most significant
pollutant in terms of vegetation damage.
The procedures used within this analysis, while specific to the
included set of crops and type of pollutant, should be sufficiently general
to be applicable to a wide range of crops and pollutants. Further, in
terms of policy implications, results derived from the empirical analysis
concerning the included set of variables should fill the most pressing
informational needs of policymakers.
1.3 The Agricultural Sector: An Overview
The agricultural sector of California has experienced a significant
growth during the past few years. Gross on-farm revenues have increased
from $5.1 billion in 1972 to $9.1 billion in 1976 (U.S.D.A. Agricultural
Statistics). While due partly to higher prices for vegetables in the
period, there are several factors which continue to contribute to the over-
all growth of California agriculture. Among them are favorable environ-
mental and technological conditions. The temperate Mediterranean type
climate in California, a well-developed system for tapping the water
resource base, relatively productive soils in some areas, high application
of chemical fertilizers, pesticides and advanced mechanical aids enable
growers to harvest a diverse high yielding and high value crop mix [Adams].
As a result, 38 of California's 61 agricultural commodities rank number one
in the nation and only five of the 61 fail to rank nationally in the top
ten.47
Total economic values of California's principal vegetable crops_5/ for
1974, 1975 and 1976 are: $1.24 billion, $1.38 billion and $1.28 billion,
respectively. These values represent 44.23, 43.42 and 43.02% of total
national vegetable marketings. Total acreages for the same period are
808,470 acres (24.36% of the U.S.), 865,920 (25.46%) and 768,160 (24.19%).67
Value, acreage and percentages for specific crops are presented in Table 1.1.
-------
Table 1.1
United States and California Crop Production:
Specific Vegetable and Field Crops, 1976
Vegetable Crop
Beans, Green Lina
Broccoli
Cantalnpes
Car ro : s
Cauliflower
Celery
lettuce, Head
Onion, Fresh ,
Onion, Processing
Hot,', i. o us
Ton.ntoc'i;, Fresh
Tomatoes, 1'rocessing
Field Crop
Cotton
Sugarbeets
United Statesa
Acreage
(1000 acres)
48.0
53.8
73.2
75.5
33.8
33.3
222.5
31.7
n . a .
1,374.1
128.9
309.0
10,859.1
1,480.5
Production
(1000 cwt)
1
55.8
4,280.0
10,005.0
20,039.0
3,213.0
16,821.0
54,047.0
7,172.0
:l.a.
353,336.0
21 ,492,0
6, 4 71. o1
10,095.9,
29,427.0"
Value
(51000)
16,007
63,761
103,075
117,424
52,575
137,374
473,337
44,466
n. a .
1,182,816
425,837
375,407
3,267,560
582,655
Californicb
Acreage
(1000 acres)
15.7
50.4
39.0
33.0
26.5
J9.3
155.1
5.9
19.5
66.0
29.4
233.8
1,120.1
312.0
i
Production
(1000 cwt )
1
2.5.75
4,133.0
6,623.0
10,100.0
2,558.0
11,110.0
39,640.0
1,652.0
7,215.0
24,138.0
6,765.0
5,066.5
2.3S2.72
8.8921'3
California
Production
as 7. of U.S.
46.15
96.56
66.20
50.28
79.49
66.05
73,34
23.03
n . a .
6.85
31.48
78.29
23.60
30.22
Value
($1000)
8,317
63,123
70,442
58,291
40,400
78,922
327,685
7,814
27,524
110,161
137,904
284,734
835,192,
267,649
2,318,158
California
Value
as % of U.S.
51.96
99.00
65.13
49.64
76.84
57.45
69.16
17.57
n . a .
9.31
32.38
75.85
25.56
45.94
I
1000 tons
2
1000 bales of 500 Ibs eauh
J1975 figures since the 1976 figurca were not available ac the time of compiling the table.
'information on processing onions not readily available. Hovever, it is generally assumed that California produces the bulk of
U.S. processing onior. production.
Sources:
"L'.S.D.A. A.qrtci!! ; unij Si.it 1st ic:i and ""California Crop and Livestock Hupor ttnj;,.Service
-------
Table 1.2
Crop Acreage Harvested, by Region
1972-76 and 1976
Crop
Vegetable Crops
Beans, Green Liraa
Broccoli
Cantaicpes
CarrotK
Cauliflower
Celery
Lettuce, Head
Onion, Croon
Onion, Dohydracod
I'ol a toi/j
Tomatoes, Fresh
Tor.uCces, Processing
Field Crops
Cotton
Sug.i rbec cs
Southern Desert
1972-76
Average
_
-
9,330
5,102
-
-
44,380
1,678
2,007
-
1,765
1,110
54,400
62,600
1976
_
-
8,850
5,510
-
-
Southern Const
1972-76
Ave rage
10,778
2.91G
2,294
10,233
4,281
10,905
43,900 17,714
1,790
925
-
1,766
1,430
71,000
58,000
1 ,773
3,/94
8,339
8,S82
9,504
18,257
9,811
1976
6,911
3,497
3,067
11,302
5,419
11,852
18,939
952
4 , 000
9,433
9,924
8,776
23,562
9,015
Central Coast
1972-76
Average
2,847
18,712
-
4,803
8,676
7,273
69,206
1,279
1,682
4,803
3,895
10,094
_
18,258
1976
995
19,900
-
4,674
9,990
8,240
73,565
2,090
1,250
4,376
4,332
9,500
_
24,390
Southern San Joaquin
1972-76
Average
2,281
-
3,872
9,440
-
-
4,430
0
6,230
34,907
2,342
8,226
413,320
27,896
1976
3,000
-
2,600
- 10,000
-
-
5,100
0
6,500
36,023
2,023
7,950
447,000
29,891
Study Region
1972-76
Average
15,906
21,630
15,496
29,578
12,957
18,178
135,730
4,730
13,713
48,549
17,252
30,139
485,977
118,565
1976
10,906
23,397
14,517
31,486
15,409
20,092
141,504
4,832
12,675
49,837
18, 04 '3
27,6.'-
541,562
121,295
Sources: County Commissioner's Annual Reports
-------
Table 1.3
Average Annual Crop Production and Market Shares,
by Region, 1972-76
1/3/
Dolcad State.
Crop*
Vegetable Crops
Beans, Lima (tans)
Broccoli (Cvt.)
Cantaloupes (Cwt.)
Carrots (Cvt.)
Cauliflower (Cwt.)
Celery (Cut.)
Lettuce* (Cut.)
Oolonj, Fresh (Cwt.)
Onions. Processing
(Cut.)
Potatoes (Cvt.)
(Cut.)
(Tons)
Field Croos
Cotton (bales of
500 Ibs.)
Sugar beeta (tons)
(Total
Production)
86.010
3,959,800
10.759,600
20.648,200
3,055.600
16.J85.600
51,658,800
5.994.000
22.456,400
322. 129,000
20,406,400
7,140,680
10,869.817
26.832.600
SOUHCiS: 11 U.S.O.A. Agricultural
I/ County
3/ Csllfor
lettuce an
South Coast - Los An
Santa
Cerural Coast - Mont
SO'Uhvrn San loaqutn
California Southern &A*«re
Production
4:,930
3,597,800
7,155,600
10,321,200
2,383.800
10,579,400
37,079.800
1,788,200
7.562,200
22.720.000
6.882,000
5.514,440
2.024.640
7.842,000
Statistics.
* of I of I of
the U.S. Production the U.S. California
49.91 - -
90.86 - - -
66.50 1.199,600 11.15 16.76
49.99 1,703,«00 8.25 16.50
78.53 - - -
64.56 - - -
71.78 11,124,000 21.53 30.00
29.83 460,400 7.68 25.75
33.67 548,000 2.44 7.25
7.05 - -
33.72 384,400 1.88 5.59
77.23 24,060 .]> .44
18.59 124,568 1.14 6.15
29.23 1.598,400 5.96 20.38
So,,rh rn.Bt r*ntr*l Cj**ar Southern S»B Joaqul
Production
23,256
238,178
320,823
3.193,959
546,599
6.201,152
4.491.817
571,862
1,209,000
2.823.??4
4,174,195
235, 970
37.212
273,305
I Of I 0( 1 of t Of I «'
the U.S. California Production the U.S. Callfomli Production the U.S.
27.04 54.17 6,346 7.38 14.71 6.761 7.86
6.02- 6.62 1.012,180 25.56 28.11
2.98 4.48 - - - 728,400 6.77
15.47 30.95 1,402,620 6.79 13.59 J, 220. 000 15.59
18.01 22.93 861.370 28.38 36.13
37.84 58.61 4.086.580 24.94 33.63
8.69 12.11 18,349,364 35.52 49.49 1.151.600 2.23
9.54 31. )8 336.960 6.46 21.64
i.38 15,99 363.552 2.51 7.45 2,118,400 9.43
.88 12.1) 1.571.160 .49 6.91 9.611,452 2.98
20.45 60.46 1.19B.380 5.87 17.41 614.768 3.31
3,30 4.28 257,589 3.61 4.67 166.940 2.34
.34 1.84 - - - 648.268 7.79
1.02 3.49 600,102 J.2« 7.63 739.274 2.75
n Valley" Study Area
J of I of
California Production California
15.76 36,366 94.71
1,250,358 34.75
10.18 2,248,823 31.42
11.20 9.519,979 92.24
1,407,969 59.0*
10.287.7J2 97.24
3.11 35.116.781 94.71
1,419,222 79.37
28.01 4,438.952 58.70
42.30 14.006.386 61.64
9.80 6,431.74) 93.46
3.03 684.559 12.42
41.90 1.010,048 49.69
9.43 3,211,081 40.95
CovaUsloncr's Annual Crop Scport.
id 1.012.882 cu
igeles County.
Barbara County
Valle. - Kern
t. Ronfllne lett
Orange County,
, San Dltgo Cou
County. Tulare
uce.
nty. Ventura County
County
-------
Table 1.4
1976 Regional Crop Production
Region
Vegetable Crop
Beans, Green LiTna
Broccoli
Can [slopes
Ca rrots
Cauliflower
Co 1 o r y
Le 1 1 uce , Head
Union, I'rc.'ilt
OiUon, Dehydrated
Po ta tees
Tomatoes, Fresh
Toir.iroL'S , Processing
rtcld Crops
Cotton
Sugarbcets
Unit
(TON'S)
(COT)
(COT)
(CWT)
(CWT)
(CWT)
(r,'T)
(CWT)
(O.'T)
(CWT)
(C','T)
(TON'S)
(BALES)
(TO::S)
Reg lor.
Southern
Deserc
.
-
1,128,000
2,215,000
-
-
11,720,000
37', ,000
300,000
-
3S4 , 000
36,000
l/i 1,500
South Coast
14,087
292,770
461,332
2,908,021
617,877
6,473,100
4,950,130
277,328
1,.'. 00, 000
2,930,200
S.O.VO.-'.Ki
178,538
51,122
1,476,000 256,63'.
I
Central
Coast
2,505
1,207,400
-
1,416,800
975,850
4,529,800
20,535,170
596,600
393,260
1,428,600
872,000
188,980
_
867,020
Southern
San Joaquin
9,000
-
468,000
3,500,000
-
-
1,490,000
-
2,530,000
10,630,900
403,480
195,000
972,760
849,638
Total
25,592
1,500,170
- 2,057,322
10,039,821
1,593,727
11,007,900
38,695,300
1,247,928
4,673,260
15,039,700
6,679,896
598,518
1,165,382
3,449,292
7. of California
99.39
36.30
31.05
99.40
62.30
99.08
97.62
75.54
64.77
62.13
98.74
11.81
46.95
38.79
*I.eaf Lettuce: Southern Desert - 0
South Coast = 428,076 COT
Central Const = 526,200 COT
Southern San Joanuin - 0
*Ror.ane Lettuce:
Southern Desert • 0
South Coast = 233,320 CWT
Central Coast = 965,800 COT
Southern San Joaquir. = 0
-------
These aggregate characteristics of the California agricultural sector
tend to mask some rather sharp distinctions observed at the regional level.
Although the Central Valley and Central Coast (Salinas VaJley) are consi-
dered the most significant production regions in terms of value of pro-
duction, other regions such as the South Coast and Imperial Valley
(identified by the California Crop and Livestock Reporting Service as Crop
Reporting District No. 8) are nationally important in the production of
many specialty crops, on both a seasonal and annual basis. This is par-
ticularly pronounced in both winter and spring vegetables as well as hor-
ticulture crops such as cut flowers. Moreover, the South Coast and Imperial
Valley areas also produce significant quantities of avocadoes, strawberries
and sugarbeets. Table 1.2 presents a regional breakdown of crop acreages
for the period 1972-1976 and for 1976. Tables 1.3 and 1.4 provide regional
data on value of production and national market shares for the same periods.
This regional importance is primarily attributable to climatological
considerations concerning the product mix that growers may undertake in
these regions. For instance, crop production in some climatologically
distinct regions, while plagued by higher production costs, remains viable
due to higher output prices normally received for winter and spring season
production or for some specialty crops. However, in the presence of
environmental degradation which results in reduced production (yields) with-
in the region, one would expect the total output and cropping mix undertaken
by growers to be affected (if differential effects across crops are assumed)
through substitution effects (e.g., use of lower yielding but more resistant
crop varieties) or depressed per unit productivity (caused by diminished air
quality or sub-optimal changes in production location). The resultant
higher output prices and/or lower yield for certain seasonal production and
other specialty crops may then significantly affect consumers' and producers'
welfare.
1.4 Purpose and Objectives
The main purpose of this report is to convey the methodological and
empirical results realized to date for the agricultural phase of EPA
Benefits project. The intent of this project phase is to develop a tract-
ible methodology for the assessment of economic damages to agricultural
crops associated with air pollution (oxidants) and apply such a methodol-
ogy to an actual production region. The empirical basis of this study is
derived frcm the application of these methodological constructs to the
four delineated regions in the study area (South Coast, Desert, Central
Coast, Southern San Joaquin Valley).
Specific objectives of this report are to:
1. Present a current review of literature on physical and economic
damages as they pertain to the development of tractible research
approach;
2. Present an overview of the incorporated methodology;
-------
3. Estimate and discuss the results of air pollution yield response
functions and crop price-forecasting equations required for
damage estimation;
A. Present a measure of economic damages for consumers as measured
by the above yield and price parameters; and
5. Discuss areas in need of further research to fully capture the
effects of air pollution on crop production. These include
production substitution (both input and output effects) and risk
effects associated with crop production in areas of high levels
of oxidant.
1.5 Plan of Presentation
The report contains six major chapters, in addition to the intro-
duction. These include: Chapter II-review of literature; Chapter Ill-
methodological considerations; Chapter IV-yield response functions; Chapter
V-price forecasting equations; Chapter Vl-estimates of economic damages to
consumers; and Chapter Vll-areas in need of further research. Each chapter
is intended to be independent in content. Thus, readers may skip chapters,
depending upon area or extent of interest. Details concerning items with-
in the executive summary may be obtained from appropriate chapters.
10
-------
FOOTNOTES: CHAPTER I
~ For more details see Chapter II of this report.
?/
- Barrett and Waddell (1973).
3/
— The counties included in each region are as follows: South
Coast — San Diego, Orange, Los Angeles, Riverside, San Bernardino,
Santa Barbara, Ventura; Desert -- Imperial; Southern San Joaquin --
Tulare, Kern; Central Coast — Monterey, San Benito, San Luis Obispo,
Santa Cruz.
A/
- California County Fact Book 1976-1977, pp. 22-23. The ranking is
based on quantity produced. These five commodities are corn, for grain
(ranks 24th nationally), corn, sweet (llth), oats (16th), red clover seed
(17th) and wheat (13th)-
— For fresh market: artichokes, asparagus, snap beans, broccoli,
brussel sprouts, cabbage, cantalopes, carrots, cauliflower, celery, sweet
corn, cucumbers, eggplant, escarole, garlic, honeydew melons, lettuce,
onion, green peppers, spinach, tomatoes and watermelons. For processing:
lima beans, snap beans, beets, cabbage, sweet corn, cucumbers (pickles),
green peas, spinach, and tomatoes.
— All figures for 1976 are preliminary.
-------
CHAPTER II
AGRICULTURAL CROP DAMAGES BY AIR POLLUTION - A REVIEW OF LITERATURE
2.1 Introduction
The relationship between plant injury and levels of air pollutants
such as oxidants is a subject of significant research effort. The impor-
tance of the subject stems from the health and economic implications por-
tended by impacts of air pollution on plants. Also, the measurement of
such relationships is controversial because plant injury is due to a wide
range of factors. There does, however, appear to be general agreement that
plant injury is primarily dependent on the concentration of fumigant and
time of exposure, although environmental factors and meteorological condi-
tions also influence this relationship. Moreover, it has been discovered
that different varieties of each plant specie have different degrees of
susceptibility to air pollution concentration and thus display different
degrees of damages, both physically and economically.
The purpose of this section is to briefly review some recent studies
concerning both physical and economic damages of agricultural crops caused
by air pollution. Concentration will be on those studies dealing with such
air pollutants as photochemical oxidant and ozone, within the United States.
This literature review is thus not exhaustive. For a more detailed review,
the interested reader is urged to pursue the subject by going through the
bibliography cited in footnote 1. The review in this section will start
with those studies concerning physical damages and then proceed with a
review of literature dealing with economic damages.
2.2 Physical Damages of Crops by Air Pollution
Plant pathologists, biologists and other plant scientists have been
concerned with effects of air pollutants on vegetation for perhaps a cen-
tury or more but it was not until the early 1950's that extensive research
on the "physical" damages of air pollution on plants was carried out.
During the last 25 years the number of publications on the subject in
various professional journals has increased significantly.!./
Perhaps the first experimental evidence of effects of air pollutant on
vegetation was that done by Lea in 1864. In his experiment, Lea germinated
wheat seedlings on gauze under bell jars with and without ozone generators.
The seedlings without ozone developed normal roots but the roots subse-
quently became moldy. The seedlings in ozone, surprisingly, had very short
roots that grew upward and remained free of mold.
12
-------
Knight and Priestly in 1914 damaged seedlings with ozone during their
investigation on the effect of electrical discharges on respiration. Homan
in 1937 investigated the possibility that ionized air and ozone might be
capable of improving plant growth. In 1948, Schomer arid McColloch (1948)
attempted to use the antifungal properties of ozone to prolong the life of
apples in storage. From such an experiment, it was determined that one
could deter surface molds for seven months if the apples were kept in 3.25
ppm of ozone. Unfortunately, results obtained showed many of the ozone-
treated apples developed brown sunken areas around the lenticels [Rich,
1964, p. 154].
Middleton, e__t. al. (1950) were among the first to report that photo-
chemical air pollutants can damage field crops. Their initial concern was
with ozone damage, but later found the primary cause of damage to be PAN.
While ozone x^as not initially thought to be important as a crop damaging
pollutant, by 1957 Freebairn had established that crops could be adversely
affected by ozone injury. In 1958, grape stipple caused by ozone was
verified [Richards, et. al., 1958]. This type of injury had been a major
problem in California vineyards since 1954.
The nationwide distribution of ozone as a potential threat to agri-
culture became apparent in 1959, when ozone was reported to cause damage
to many crcps in New Jersey [Daines, et. al., I960]. Through fumigation
experiments, Ledbetter, et. _al_. (1959), and Hill, et. al. (1961) extended
the list of plants that can be injured by ozone. Thomas (1961) performed
a fairly comprehensive review of the available information on the effects
of photochemical oxidants on plants.
Middleton (1961) gave the first comprehensive coverage of the phyto-
toxic effects of photochemical oxidants. Rich (1964) presented an early
and detailed review of ozone effects on plants. The degree of injury to
susceptible plants is directly related to the concentration of ozone to
which plants are exposed and to the duration of the exposure [Rich].
Although symptoms of ozone injury may vary across species, there are several
symptoms that appear to be typical of the ozone syndrome. One of the first
symptoms of ozone injury is the appearance of "water-soaked" spots found on
tobacco leaves. If the damage is not severe, the injured cells may ulti-
mately recover. The following phase is usually bleaching. With more severe
injury, the chlorotic or discolored areas may become necrotic and then
collapse. Another symptom of ozone injury in plants is yellowing or pre-
mature senescence of older leaves, accompanied by abscission.
Once ozone gets inside the leaf, it attacks the palisade parenchyma
first. The symptoms of ozone injury to palisade cells vary. "In grapes,
the injured cells become darkly pigmented before they die" [Richards, et.
al., 1958, p. 257]. A similar type of pigmentation accompanied by thick-
ening of the cell walls is also found in ozone damaged palisade cells of
avocado and strawberry [Ledbetter, et. al., 1959]. In tobacco, sugarbeet,
and occasionally peanut and sweet potato, the ozone injured palisade cells
collapse and then become bleached. The surrounding tissue may be unaf-
fected if the ozone damage is not too severe. Otherwise, the adjoining
13
-------
mesophyll and upper epidermis may die [Povitailis, 1962]. In tomato and
potato there is complete collapse of the tissue within the lesion caused by
ozone.
A series of experiments about effects of air pollutants on citrus trees
carried out by Thompson and Taylor, Thompson and Taylor and Associates in
the 1960's in the Los Angeles Basin [Thompson and Ivie, 1965; Thompson and
Taylor, 1966; Thompson, et. al. 1967; Thompson and Taylor, 1969] show that
the photochemical smog complex present in that area reduced water use and
the apparent photosynthesis of citrus trees. Ambient levels of fluoride
had no significantly measurable effects.
"The smaller total leaf drop in trees which received filtered air
was compared to the unfiltered treatments is somewhat significant but
when measured for long periods tends to become equal in all trees be-
cause all leaves become senescent and fall eventually. The much more
revealing work was the study in which the separate lemon branches with
tagged, dated leaf flushed were counted periodically. These showed,
after 18 months that the trees receiving filtered air had lost 28% of
their leaves while the unfiltered treatments had lost 66%" [Thompson
and Taylor, 1969, p. 940].
Effects of ozone (comprising almost all the oxidants in the South
Coast Air Basin) on some crops such as corn, tomato, lettuce and cabbage in
the South Coast area have been studied and reported by Oshima (1973). In
that study, a short-term fumigation study was undertaken in order to deter-
mine oxidant effects on young seedlings. A long-term fumigation study was
then used to determine effects on crop quality and yield and to develop
criteria for field studies. Seedlings of the Golden Jubilee variety were
exposed to 0.24 ppm ozone concentrations for 1.5% of the growing period.
Fumigations were initiated upon emergence and discontinued after a 30-day
period. Results from the experiment indicated that ozone injury was
observed on the seedling corn leaves of the ozone treatment throughout the
fumigation. At harvest, the size and weight of the fumigated plants were
reduced when compared to controlled plants. In summary, Golden Jubilee
corn was seriously affected by ozone in the 0.20-0.35 ppm concentration
range under greenhouse conditions. The general effect of the ozone
exposures was a reduction in the size and weight of the corn plants. A
higher concentration of ozone, say, 0.35 ppm, reduced the dry weight of the
ears by 22.3% which is twice the 12.5% reduction found in the 0.20 ppm
treatment. However, ozone does not seem to influence the quality of field
grpx-m Golden Jubilee corn ears to any great extent. The only quality
criterion possibly associated x-/ith ambient oxidant dosages was the extent
of blemishes on harvested ears. This might be due to the fact that this
variety of corn is somewhat resista.nt to disease and air pollution injury.
The same procedure described above was used on tomato, lettuce and
cabbage; the results obtained are described below. Ozone exposures at a
moderate level (0..24 ppm) reduced the size and weight of H-ll variety
tomato seedlings. Reductions in height of plant, weight, and number of
leaves indicate that the fumigated seedlings were not as fully developed as
controlled plants. Higher levels of ozone concentrations (0.35 ppm)
14
-------
significantly affected fruit yield. Although ozone injury was observed on
field-grown tomato plants, no quality reductions attributed to ozone were
detectable on harvested fruit.
Prizehead lettuce was found to be resistant to ozone and other pol-
lutants at all stages of growth. Ozone fumigated seedlings were reduced in
percent solids from controlled plants, but only the high concentration
(0.35 ppm) level of ozone over a period of time produced detrimental effects
on the mature stages of growth. Dark green Boston lettuce was selected for
long-term fumigation studies as a comparison to Prizehead lettuce. This
variety proved to be far more susceptible to ozone than the Prizehead
variety. The percentage of leaves affected by oxidants would make these
plants unacceptable for marketing. Ozone also produced a reduction in the
overall size of plants in both fumigated treatments. It should be noted,
however, that lettuce is regarded as a cool weather crop and is thus gen-
erally grown in the spring or fall, a period when it would not be subjected
to the high exposures of ozone which affect summer grown crcps [Oshima,
1973, p. 80].
Long-term fumigations indicated that ozone does not affect the quality
of Copenhagen Market cabbage heads. Greenhouse grown Copenhagen Market
cabbage was found to be sensitive to ozone leaf injury at all stages of
growth. However, injury to wrapper leaves by ozone did not always reflect
reduced yields or quality. Plants exposed to a lower level of ozone (say,
0.20. ppm) displayed considerable leaf injury but no reduction in either the
size or the weight of harvested heads. Leaf injury was also observed in the
0.35 ppm level of ozone concentration but there were no significant yield
reductions. This variety apparently tolerates a degree of ozone leaf injury
without any significant effect on size or weight of the head. Jet Pack
cabbage, a commercial hybrid, was then introduced in the long-term fumi-
gation studies as a comparison with Copenhagen Market cabbage. Effects of
ozone injury were essentially the same as Copenhagen Market.
Brewer and Ferry (1974) carried out a study on effects of photo-
chemical air pollution (smog) on cotton in the San Joaquin Valley in 1972-
73. The experiment consisted of placing pairs of filtered and non-filtered
plastic covered gre.enhouse shelters over established plots of cotton in
some selected locations in the valley. All greenhouses were equipped with
electric motor driven blowers which changed the air in each house twice
every minute. One of each pair of blowers was equipped with activated
carbon filters which effectively removed oxidants, ozone and nitrcgen
dioxide. Plant height, squares, bloom and boll set were then recorded for
each plant at about two-week intervals. The experiment shows that one
obvious effect of the carbon-filtered air on cotton plant growth at all
locations was. the: retention of vigor and color during late summer and
early fall. Moreover, plants in the filtered air were green and continued
to bloom and mature bolls weeks after those in the outdoor plot and non-
filtered gre.enhouse had colored and become senescent.
Plant injury by air pollution not only depends on the level of concen-
tration of each pollutant and environmental factors but also depends on
differential variety of each crop. Many plant pathologists and vegetable
15
-------
crop specialists and other plant scientists have conducted studies in order
to test the degree of susceptibility of each variety of crop to air pollu-
tants at certain locations. Results from such experiments have then served
as suggestions to farmers as to which variety of crop should be used for
the next growing season. An experiment of this type was conducted on sweet
corn hybrids by Cameron, et. a_l_. (19.70) in Riverside and Los Angeles, Cali-
fornia. The study showed a marked differential in injury from air pollution
in different sweet corn hybrids; e.g., at Riverside, leaf damage by oxidants
ranged from nearly zero in 11 hybrids to slight to severe in 23 others.
This was also true in the Los Angeles area.
"Thus, it appears that among the cultivars there were great
differences in injury which cannot be attributed to cultural factors
such as fertilization or irrigation, or to high temperature alone.
Genetic resistance to air pollution damage is apparently present in
some cultivars, but not in others." [Cameron, et. al., 1970, p. 219]
Experiments by Thompson, et. al. (1976) on two varieties of sweet corn
in the Los Angeles Basin also showed different degrees of susceptibility to
ozone injury. Studies by Reinart, et. al. (1969), Clayberg (1971, 1972) and
Oshima, et. al. (1975) on different varieties of tomato found both resistant
and susceptible cultivars to ozone concentration. These varieties were then
ranked in order of degree of susceptibility. Finally, Davis and Kress (1974)
selected six varieties of bean from those recommended for commercial produc-
tion in Pennsylvania in their study concerning the relative susceptibility
of each variety to ozone. Plants were exposed to 0.25 ppm ozone for 4 hours
at a temperature of 21°C, 75% relative humidity, and a light intensity of
25,000 lux. In each variety, five plants were exposed from 8:00 am to
12:00 noon, and the remaining five from 1:00 pro to 5:00 pm on the same day.
Such exposures were conducted on three different days, each 30 days from
the respective planting date. Results showed that ozone symptoms differed
slightly across varieties, but were generally a dark stipple or a light tan
fleck on the upper surface of the leaf.
From the literature reviewed, one can conclude that air pollutants
such as oxidants or ozone cause damages to various plants and crops. The
degree of injury to susceptible plants depends directly on the concentra-
tion of ozone and the duration of the exposure. Minor injury may result
only in yellowing or premature senescence of older leaves and the injured
cells may ultimately recover. If, however, the damage is severe, the
chlorotic or discolored areas may become necrotic and collapse, followed
by leaf-drops, fruit-drops, reduction in growth and yield and may finally
result in the death of the plant.
Empirical studies indicate that various types of agriculturally
important vegetables such as beans, cabbage, corn, lettuce and tomatoes
and some field crops such as cotton are susceptible to ozone concentrations.
Selected exposures reduced the size and weight of fruit, the height of
plant and the number of leaves. Higher levels of ozone concentrations
significantly affected fruit yield. However, effects of ozone on the
quality of fruit is not well established. Finally, it is evident that
varieties of each crop respond differently in terms of degrees of
16
-------
susceptibility to a specific type of air pollution. Farmers, based mainly
en past experience, usually choose the variety that has the highest degree
of tolerance to air pollution in that area.
2.3 Economic Damages of Crops by Air Pollution
As reported in the preceding section, the effects of oxidant on
vegetation have been intensively studied over the past 25 years. Oxidant
or smog type symptoms were identified with the reaction product of ozone
and reactive hydrocarbons (automobile exhaustion). Generally speaking, the
adversary effects of air pollution on agricultural plants are reductions in
the quantity of output (yields) and/or degradation of the quality (nutri-
tional content) of the product. In terms of measurement of economic damages,
the scope and content of research efforts are somewhat more limited, parti-
cularly with respect to methodologies. Waddell (1974) identified some
genera] approaches for such measurement purposes. One approach is to
actually survey the damage loss on a statewide basis. This approach has
been used in studies by Middleton and Paulus (1956) , Weidensaul and Lacasse
(1970), Millecan (1971), Feliciano (1972), Naegele, e_t. al_. (1972), Pell
(1973), and Millecan (1976). Another approach is to construct predictive
models relating data on crop losses to crop values, pollution emission and
meteorological parameters. The most comprehensive attempts using such an
approach are studies done by Benedict and Associates (1970, 1971, 1973) at
the Stanford Research Institute (SRI). A third approach to assessing
economic damage of crops by air pollution is to estimate the "dose-response
function" and then relate it to the calculation of losses for each crop.
This approach has been attempted by O'Gara (1922), Guderian, Van Haut and
Stratmann (1960), Stratmann (1963), Zahn (1963), Larsen and Heck (1976),
Oshima (1975), Oshima, e_t. al, (1976, 1977), and Liu and Yu (1976). This
method will be described in the section on air pollution response function
estimation presented later.
Economic assessment of air pollution damages by investigators on a
site-specific basis was first done in a California survey conducted in 1949.
A somewhat similar survey in 19.55, reported by Middleton and Paulus (1956),
was designed to show the location of injury, the crops injured, and the
toxicant responsible, for the damage. Agricultural specialists throughout
the state were trained as crop survey reporters with the survey covering
four categories of crops: field, flower, fruit, and vegetable.
A program similar to that in California was established in Pennsylvania
in 1969 [Weidensaul and Lacasse, 1970]. The objectives of that survey were:
(1) to estimate the total cost of agricultural losses caused by air pollu-
tion in Pennsylvania; (2) to determine the relative importance of the
various pollutants in Pennsylvania; (3) to survey the extent of the air
pollution problem in Pennsylvania; (4) to provide a basis for estimating
the. nationwide impact of air pollution on vegetation; and (5) to provide a
basis for guiding research efforts.
The Pennsylvania study included both commerical and non-commercial
plants. Past air pollution episodes were investigated for purposes of
detecting possible trends. Estimates of losses obtained were based on
17
-------
crop value and production costs incurred by harvest time. Direct losses
to producers and growers included only production costs, whereas indirect
losses included profit losses, costs of reforestation, grower relocation
costs, and the cost of substituting lower value (highly resistant) crops
for higher value (but very sensitive) crops. Other costs such as those
associated with destruction of aesthetic values, erosion and resultant
stream silting, damage to watershed retention capacity, and farm aban-
donment were not considered.
Of the 92 field investigations made within the Pennsylvania study, 60
revealed damages that were attributable to air pollution. Damage resulting
from pollution was observed in 23 counties, primarily located in south-
eastern and western Pennsylvania. Direct losses estimated in the survey
exceeded $3.5 million. The air pollutants responsible for the damage, in
order of decreasing importance were oxidants, sulfur oxide, lead, hydrogen
chloride, particulates, herbicides, and ethylene. The vegetation most
affected (also in lawns, shrubs, woody ornamentals, timber, and commercial
flowers. Indirect losses were estimated at $8 million of which $7 million
reflects profit losses, $0.5 million reflects reforestation costs, and the
remainder reflects costs for grower relocation.
The approach used in the Pennsylvania study may be criticized on
several aspects. First, the method used in assessing losses is somewhat
questionable because grower profit losses are not included as direct costs
(since profit is normally the main objective of producers, such losses may
be direct). Second, methods of translating physical damage into economic
loss have not been standardized. Third, not much is known of the extent to
which home garden plantings and flowers are being affected by air pollution
and, if they are affected, then what value should be assigned to these
losses.
There are certain advantages, however, of this procedure, such as:
(1) existing manpower used in the initial survey can be used to achieve
continual coverage over an area; (2) local agents have rapport with growers
in that area, are familiar with crop peculiarities, and are probably
knowledgeable about local sources of pollution in the area; and (3) a field
coordinator supplies expertise to the reporting personnel and also provides
some degree of standardization in reporting losses.
A similar study was carried out for Pennsylvania in 1970 [Lacasse].
Using the same concepts of cost (direct and indirect) as in the previous
year's survey, Lacasse estimated direct losses to be $218,630 and indirect
losses of $4,000. The relatively low damage figure for that year was due
to:
"fewer inversions and to no unfavorable growing conditions when
air stagnation did occur." [Lacasse, 1971]
Similar surveys have also been carried out by Feliciano in New Jersey
and in the New England States in 19.71. Feliciano (1972) estimated that
agricultural losses due to air pollution in New Jersey were $1.19 million.
However, as in the Pennsylvania surveys profit losses were not included.
18
-------
In the course of the New Jersey surveys a total of 315 air pollution
incidences were investigated and documented. A "rule of thumb" evaluation
method developed by Millecan (1971) was used for estimating losses, i.e.,
if visual inspection of the overall leaf surface of the plants indicates
1-5% injury a 1% loss was applied for that crop. A leaf surface injury of
6-10% was assigned a 2% loss; 11-15% injury, a 4% loss; and 16-20% injury,
an 8% loss. Estimates of total losses were then based on the crop value of
the acreage affected.
Naegele, et. al. (1972) reported on a field survey of agricultural
losses in the New England region resulting from air pollution. The survey
contains 83 investigations in 40 counties covering the six New England
states. Direct economic losses for the 1971-72 season were estimated at
approximately $1.1 million. Economic loss estimates were based on grower
costs, crop value at the time of harvest and the possibility of crop re-
covery following the pollution incident. The direct losses in this study,
in contrast to the Pennsylvania and New Jersey cases, include grower profit
losses. Among the crops studied, fruit, vegetables, and agronomic crops
suffered the greatest losses, with over 90% of the damage being attributed
to oxidant air pollution.
An approach similar to that used in Pennsylvania, New Jersey and New
England was used by Millecan (1971) to survey and assess the damage of air
pollution to California vegetation in 1970. Prior knowledge about the dis-
tribution of air pollution problems placed concentration of the study in
the Los Angeles Basin, San Joaquin Valley, and the San Francisco Bay area.
Estimates of losses were confined to 15 of the 58 counties in the State,
even though plant injury from air pollution was observed in 22 counties.
Ventura County, on a county-wide basis, suffered the greatest economic crop
loss (approximately $11 million). Losses of citrus production in the Los
Angeles Air Basin accounted for over $19 million of a total monetary loss
of almost $26 million. Such a monetary loss estimate does not include
losses attributed to reduction in crop yield or growth (except for losses
of citrus and grapes) nor losses to native vegetation including forests,
nor to landscape (horticultural) plantings. Photochemical smog accounted
for most of the economic losses. Specifically, the percentages of plant
injury caused by each type of air pollutant are as follows: ozone, 50%;
PAN, 18%; fluorides, 15%; ethylene, 14%; sulfur dioxide, 2%; and particu-
lates, 1%.
In order to obtain a better understanding of the year-to-year variation
in plant losses caused by air pollution, Pell (1973) continued the research
initiated by Feliciano in 1971. The direct losses of agronomic crops and
ornamental plantings, estimated by Pell for the 1972-73 growing season were
approximately $130,0.00. As in the study by Feliciano, costs associated with
crop substitution and yield reductions were not considered. In decreasing
order of importance the damaging pollutants were: oxidants, 47% of crop
los.se.s; hydrogen fluoride, 18%; ethylene, 16%; sulfur dioxide, 4%; and
anhydrous ammonia, 1%. The damage reported in this sruvey, surprisingly,
was only 11% of that reported by Feliciano in the 1971-72 New Jersey survey.
Perha,ps one explanation is that the significant year-to-year variation
observed may be attributed to altered environmental conditions rather than
19
-------
to decreased air pollution concentrations. As an example, it is believed
that the unusual rainfall patterns in 1972 placed the plants under water
stress and thus protected them from air pollution injury.
A detailed survey and assessment of air pollution damages for Califor-
nia vegetation covering the period 1970-74 was again conducted by Millecan
(1976). The survey was done in 10 counties^/ and covered four types of
crops: fruit and nut, field crops, vegetable, and nursery and cut flowers.
Within the framework of this study, a method known as the crop-dose conver-
sion scale was developed to measure monetary losses to alfalfa. This method
is asserted to represent an improvement in determining monetary loss values
related to the effects of air pollution on agricultural crops. The conver-
sion scale method is viewed as providing accuracy since it utilizes actual
pollution doses within the growing areas in a county and does not have to
apply averaging techniques as are needed in the general survey method. In
addition, the conversion scale method is able to produce standardized annual
crop loss estimates, i.e., yearly estimates of crop losses taken from the
conversion scale would differ only from variations in ambient ozone dose and
would therefore provide a uniform basis of annual comparisons. In deriving
the loss figures three factors were considered: (1) the value of the crop,
taken from the respective County Agricultural Commissioner's annual crop
production reports or crop production reports of the California Department
of Food and Agriculture; (2) the pollution index, which represents a measure
of oxidant readings observed throughout the year, differences in air pollu-
tion levels among individual counties can then be compared by means of this
index; (3) the percentage of crop damage using the 1970 loss figures
[Millecan, 1971] as a reference point, as related to the increase or de-
crease in the air pollution index.
The overall monetary losses in the ten counties caused by air pollution
have increased from 1970 to 1974. Such losses are reported as about $16.1,
$19.1, $17.4, $35.2, and $55.1 million respectively. Such increases may be
due partly to the increased per unit value of agricultural crops in each
year, i.e., the physical damages to individual crops may not necessarily
have increased. The large increase in losses in 1973 and 1974 was attri-
buted to an increased level of air pollution, a larger crop and an increase
in crop value [Millecan, 1976, p. 7]. Almost half of the monetary loss in
1974 was in cotton in the San Joaquin Valley. In conclusion the author
noted that:
"Monetary loss from air pollution damage to agricultural crops
will generally increase yearly because of several factors such as: an
increase in knowledge of plant susceptibility, an increase in the
ability to assess more correctly the effects of air pollution, an in-
crease in population and possibly an increase in air pollution levels."
(P. 22)
Perhaps the most comprehensive research effort on economic damages was
performed by the Stanford Research Institute._3/ The objectives of the SRI
Nationwide Survey were to develop a model for estimating dollar losses to
vegetation resulting from the effects of pollutants, and to make such esti-
mates. The procedures and results of the study were as follows:
20
-------
1. Selection of those counties in the United States where major air
pollutants — oxidants (ozone, PAN, and oxides of nitrogen), sulfur dioxide,
and fluorides — were likely to reach plant-damaging concentrations. The
counties selected were those in the Standard Metropolitan Statistical Areas,
under the assumption that damaging concentrations of oxidants and sulfur
dioxide were more likely to occur in the most populous areas.
2. The potential relative severity of pollution in each county was
then estimated. The severity of oxidant pollution was then derived by
first estimating, from fuel consumption data, the emissions per square kil-
ometer per day of tons of hydrocarbons and oxides of nitrogen (the precur-
sors of oxidants). These emissions values were then multiplied by a con-
centration rate factor and a factor related to area of the county or SMSA.
The results obtained yielded a value indicative of the relative concentra-
tion of oxidant that might be reached in a single pollution episode. These
values were then multiplied by the number of days involved in pollution
episodes to obtain a value indicative of the overall plant-damaging poten-
tial for oxidant pollution in the various counties.
The same procedures were used for estimating the plant-damaging
potential for sulfur dioxide. In the case of fluorides, the relative
plant-damaging potential was based on the number, type, and size of large
single source emitters present.
The counties were then arranged and grouped into classes in order
of the severity of the plant-damaging pollution potential.
3. The dollar values of commercial crops, forests, and ornamental
plantings were then determined or calculated by the following procedures:
a. Commercial crop values for 1964 and 1969 were taken from data
in the Census of Agriculture and supplemented, for 1969, by yearly re-
ports of the states or individual counties involved.
b. Values of forests were calculated from Federal and State
records.
c. For ornamental plantings, maintenance and replacement costs
were the representative values. The dollar values for the states were
first determined and these values were then prorated to the polluted
counties based on their proportionate area, population, or combination
of area and population of the state.
4. To arrive at the loss to each plant that might occur in each class
of plant-damaging pollution potential, the following methods were used:
a. Each group of ornamentals were classified, based on litera-
ture reviews, as sensitive, intermediate or resistant to each pollutant.
They were also classified as to whether the part of the plant directly
affected by the pollutant (i.e., leaves, roots, fruit) had high,
medium, or no economic use.
21
-------
b. The next step was to obtain the percentage loss occurring to
the most sensitive plants in the high-use category in the most severely
polluted counties.
c. Using the above two types of information, tables were pre-
pared showing the percentage economic loss that would occur to plants
in each sensitivity use category in each pollution potential class for
each pollutant associated with those described in (2).
5. These factors were then applied to value of the crops, forests,
and ornamentals grown in the polluted counties, and recorded the dollar
loss value for each crop in each county. These values were added to arrive
at the state, regional, and national values.
6. In obtaining the 1969 estimates, 687 of the 3,078 counties in the
United States (excluding Alaska) were selected as having exposure to poten-
tially plant-damaging levels of oxidants, sulfur dioxide, and fluorides.
Of these counties, 493 would be exposed to oxidants, 410 to sulfur dioxide,
and 87 to fluorides (some counties would be exposed to damaging levels of
two or more pollutants). On the basis of area and population, about 14.6%
of the area and 68.9% of the population were likely to have plant-damaging
oxidant pollution. For sulfur dioxide, the respective values were 16.2%
and 53.0%, and 4.2% and 6.8% for fluorides. For the 1964 estimates, these
values were: 11% and 62% for oxidants, 13% and 54% for sulfur dioxide, and
4% and 9% for fluorides.
The analysis used in the 1969 estimates indicates that 40% of the gross
values of agricultural crops, 36% of the value of forests, and over 50% of
ornamental value lies within polluted areas of the United States. The study
also indicated that as much as 40% of the crops in a county could be lost
due to oxidants, 12% due to sulfur dioxide, and 12% due to fluorides.
When the loss factors for the various pollution intensities were applied
to the values of crops and ornamentals, the total annual dollar loss to
crops in the United States in 1969 was calculated to be about $87.5 million,
of which $77.3 million was due to oxidants, $4.97 million to sulfur dioxide,
and $5.25 million to fluorides. The value of loss to ornamentals was esti-
mated to be about $47.1 million, of which $42.8 million was attributable to
oxidants, $2.7 million to sulfur dioxide, and $1.7 million to fluorides.
These estimated values are not greatly different from those found for the
1964 estimates (total loss was $85.4 million, of which $78.0 million was due
to oxidants, $3.2 million to sulfur dioxide, and $4.2 million to fluorides).
•For 1971, it was estimated that the losses to vegetation for the United
States were $123.3 million due to oxidants and $8.2 million to sulfur dioxide.
No attempt was made to calculate losses due to fluorides in 1971.
In summary, the dollar loss as estimated for the 1969 and 1964 crop
values represented, respectively, 0.44 and 0.46% of the total crop value of
the United States in those years.4/
22
-------
"On a regional basis, the greatest percentage of crop losses
occurred in the heavily populated and industrialized areas of south-
western and middle Atlantic and midwestern states. The lowest percen-
tage loss occurred in the plains and mountain states." [Benedict,
Miller and Smith, 1975, p. 8]
2.4 Measurement of Air Pollution Damages: Air Pollution Response Functions
The approaches and estimates of air pollution crop damage outlined
above are representative of earlier research in this area. A more general
set of literature exists which deals with all types of air pollution damages.
This section briefly discusses the more important contributions in the area
and introduces explicitly the concept of an air pollution response function.
Such functions serve to quantify the relationship betxs'een a particular var-
iable and levels of air pollution. These relationships are extremely impor-
tant in the assessment of crop damages.
The literature on air pollution contains six general methods for esti-
mating damages from air pollution. These methods are: (1) technical coef-
ficients of production and consumption; (2) market studies; (3) opinion
surveys of air pollution sufferers; (4) litigation surveys; (5) political
expressions of social choice; and (6) the Delphi method. These methods have
been used with different degrees of success and are not necessarily mutually
exclusive [Waddell, 1974, p. 22]. Among these methods, the technical coef-
ficients of production and consumption and the Delphi methods have been used
substantially in agricultural studies in forecasting crop production levels
at different levels of air pollution. The market studies method is used
widely in determining the adverse effect of air pollution on human activity
and behavior such as the relationship between air quality and consumer
behavior or the consumption of recreation-related activities [Vars and
Sorenson, 1972]. Another type of market study is the use of the concept of
property values to estimate air pollution damages [Ridker and Henning, 1967;
Anderson and Crocker, 1970; Peckham, 1.970; Crocker, 1971; and Spore, 1972].
The method incorporating opinion surveys of air pollution sufferers is per-
haps closest to the classical economic approach in that it focuses on esti-
mating utility and demand functions for such individuals, but it also suffers
from at least two problems, i.e., the "free-rider" aspect and the possibility
that a respondent might not understand fully the consequences of air pollu-
tion on his health [Waddell, 1974, p. 30]. The litigation surveys and the
political expressions of social choice methods are rather subjective and
limited, since the information gathered represents opinions of special groups
of people such as lawyers, court clerks, state and local control officials,
politicians, and representatives. Their opinions might be quite different
from people who actually suffer from air pollution.
In general, the estimation of technical coefficients concerning pro-
duction and consumption is facilitated by: (1) the use of experimental data
on subjects under conditions simulating their natural environment; (2) esti-
mation of the physical or biological damage-function which relates damage to
different levels of air pollution; (3) translation of the physical damage
function into economic terms via "damage functions;" and (4) extrapolation
of the function to the population if an aggregate damage estimate is required.
23
-------
Because of a lack of adequate dose-response functions, a variation of the
basic method outlined above is typically followed. The researcher uses a
"damage factor approach" to estimate what proportion of a damage category
can be identified as being related to or caused by air pollution. Such a
proportionality factor will then be used to estimate the required air poll-
ution damage. However, one problem of this method is that, while the mag-
nitude of the physical and biological damages can be predicted with some
degree of accuracy, in many cases the attempts to translate these damages
into meaningful economic relationships are not very accurate. Perhaps one
reason is that controlled laboratory conditions are not usually represen-
tative of the real world. To solve such a problem, the normal practice is
to hold everything constant except one factor - a single pollutant or mix
of pollutants. Other problems are those of aggregation and substitution.
It is very unlikely that the aggregation process involves a straight arith-
metic summation over all individuals [Anderson and Crocker, 1971, p. 147].
Besides, the substitution of one factor of production by an individual will
not normally affect relative prices; but if the same substitution is carried
out'by all receptors, relative factors prices will often be changed.
The Delphi method is a method of combining the knowledge and abilities
of a diverse group of experts for the purpose of quantifying variables
which are either intangible or display a high level of uncertainty [Pill,
1971, p. 58]. Essentially, the method is a type of subjective decision-
making. It is an efficient way to arrive at "best judgments," where both
the knowledge and opinion of experts are extracted, i.e., those who are
considered experts in the relevant area are asked to give their best solu-
tion to any given problem. This method is one that has been used by the
U.S. Department of Agriculture in forecasting crop production levels [Wad-
dell, 1974, p. 34]. The Delphi method appears to be an approach that can
provide answers in a short period of time. However, due to the subjective
nature of this method, many of the air pollution damages created in this
manner have been questioned [Waddell, 1974, p. 35].
2.5 Air Pollution Response Functions and Crop Loss Equations
Several variants of air pollution response functions have been developed
for the purpose of measuring physical and economic damages of crops due to
air pollution. Perhaps the earliest one is that formulated by O'Gara (1922)
for alfalfa, taking the general form of:
(C - 0.33t) = 0.92 (2.1)
where C is the estimated concentration level and t is time in hours. The
constant 0.33 ppm is the concentration level (or the threshold level) that
a plant is presumably able to endure indefinitely.
In order to generalize O'Gara's equation, Thomas and Hill (1935) pro-
posed the following equation for measuring any degree of leaf destruction at
any degree of susceptibility:
t(c - a) = b (2.2)
where t = time in hours, c = pollution concentration level in ppm exceeding
a, a is the threshold concentration below which no injury occurs, and b is
the constant.
24
-------
The levels of leaf destruction are then given as follows:
if t(c - 0.24) = 0.94, only traces of leaf destruction are
observed
if t(c - 1.40) = 2.10, there is a 50% chance of leaf destruction
if t(c - 2.60) = 3.20, there is a 100% chance of leaf destruction.
Zahn (1963) modified the O'Gara equation and developed a new equation which
provides a better fit for a longer time period. This equation takes the
following form:
1 + Q.5C
t = ^TTpb (2-3)
where be is the dimensional resistance factor which includes effects of
environmental conditions.
An alternative experimental formula was proposed by Guderian, Van Haut
(1960) and Stratmann (1963). This formula provided a "best" fit to a set
of observations over both short or long periods of exposures. The proposed
formula is:
t = Ke (2.4)
where K = vegetation life time, in hours; t is time; and a, b, and C are the
same as in the Zahn equation. These parameters may vary with plant species,
environmental conditions, and degree of injury.
Benedict, et . al. (1973) derived crop loss estimates by the following
formulation :
Crop Loss = crop value x crop sensitivity to the pollutant
x regional pollution potential (2.5)
where the relative sensitivity of various plant species to the pollutant was
determined by using information provided in secondary sources. The regional
pollution potential is defined as a relative severity index of pollution
estimated for each county, arising from fuel consumption.
Larsen and Heck (1976) analyzed data on the foliar response of 14
plant species (two cultivars of corn) to ozone concentration. They used a
mathematical model with two characteristics: a constant percentage of leaf
surface injury caused by air pollution concentration level, that is, the
inverse proportion of exposure duration raised to an exponent and, for a
given length of exposure, the percentage leaf injury as a function of pol-
lution concentration level fit to a log-normal frequency distribution.
This relationship takes the following form:
c = m . sz tP (2.6)
& hr g
where c is pollutant concentration, in parts, per million, m is geometric
mean concentration for a one-hour exposure, s is the standard geometric
g
25
-------
deviation, t is time (hour), p is the slope of the line (logarithmic), and
z is the number of standard deviations that the percentage of leaf injury
is from the median.
In equation (2.6), m , s and p are known constants. They vary
g hr g
according to type of crop. Thus c is the function of two exogenous varia-
bles, z and t. By substituting different values of z and t into equation
(2.6), different values of c will then be obtained. Larsen and Heck (1976),
Table 2, p. 329, calculated injury threshold for exposure of 1, 3, and 8
hours of 14 plant species (two cultivars of corn).
Liu and Yu (1976) proposed a stepwise linear multivariate regression
model for determining the economic damage functions for selected crops and
plants as follows:
CROPL. = a + b(CROPV ) + c(TEMB) + d(TEMA) + e(SUN) + f(RHM)
i i
+ g(DTS) + h(SO ) + j(OXID) (2.7)
where CROPL denotes the economic loss (in $1000) of the ith type of crops
i
by a county; CROPV. is the crop value (in $1000) of the ith type of -crops;
TEMB and TEMA are, respectively, the number of days in a year with temper-
ature below 33°F and above 89°F; SUN denotes possible annual sunshine days;
RHM is the relative humidity; DTS represents the number of days with thun-
derstorm; SO is the level of sulfur dioxide concentration and OXID is the
relative severity index of oxidant.
Oshima (1975) and Oshima, et. al. (1976, 1977) calculated percentage
of yield reduction of alfalfa, tomatoes and cotton due to air pollution by
using the ozone dosage-crop loss conversion functions. These functions are
presented below.
Alfalfa
i. Yield function __
Percent reduction = 0 + (9.258 x 10 x dose) (2.8)
ii. Defoliation function __
Percent reduction = 0 + (3.030 x 10 x dose) (2.9)
Tomato
Percent reduction = 0 + (0.0232 x dose) (2.10)
Cotton
i. Uniformity Index _„
Percent reduction = 0 + (1.90 x 10 x dose) (2.11)
ii. Number of harvested bolls _
Percent reduction = 0 + (6.947 x 10~ x dose) (2.12)
26
-------
The ozone dose is derived from oxidant data measured by various types
of instrumentation. Hourly averages exceeding 10 pphm, the California
standard for oxidant air pollutants, were used in calculating the average
weekly dosage in pphm hours for any specified season. Since plants are
typically less sensitive to oxidants at night, only the hourly averages for
the daylight hours were used.
2.6 Conclusion
For policymakers, economic damage functions may be more relevant than
physical damage functions. An economic damage function, or a monetary
damage function, relates levels of pollution to the amount of compensation
which would be needed in order that society (i.e., consumers and producers)
not be worse off than before the deterioration of the air quality. The
economic damage function is useful to decisionmakers since the multiple
dimensions of the decision problem are reduced into one dimension only, i.e.,
money. It should be noted, however, that transformation of a physical
damage function into an economic damage function as has been tried by some
researchers, often involves value judgment on the. part of the policymaker
or researcher. A related question as to the degree of conformity of the
values of the policymaker with those of the consumer is largely unresolved
[Liu and Yu, 1976, p. 34] .
27
-------
FOOTNOTES: CHAPTER II
— For details see the bibliography at the end of Chapter II in
Committee on Medical and Biologic Effects of Environmental Pollutants,
Ozone and Other Photochemical Oxidants, Washington, D.C.: National
Academy of Sciences (1977).
2/
~ Alameda, Los Angeles, Marin, Orange, Riverside, San Bernardino,
San Joaquin Valley, San Mateo, Santa Clara, and Ventura.
— Prior to this study, two previous reports have appeared. The
first one [Benedict, 1970] was mainly devoted to description of the method
or model that was developed and the background information that led to
its development. The second report [Benedict, et. al_. , 1971] described
improvements in the model and gave vegetation loss estimates for 1964 crops
as related to 1963 emission data.
4/
~~ This loss is expressed as a percentage of the total crop value
in both polluted and unpolluted areas. The percentage of crop value lost
in the pollution threatened counties for the U.S. is 0.99 and 1.84% in 1969
and 1964 respectively.
28
-------
CHAPTER III
SOME METHODOLOGICAL CONSIDERATIONS ON THE ASSESSMENT
OF AIR POLLUTION DAMAGES: A PROPOSED
MATHEMATICAL FRAMEWORK
3.1 Introduction
One important aspect of economic analysis concerns the definition of
methods or procedures that may be used in addressing a problem or set of
problems at hand. When integrated with appropriate assumptions, methodol-
ogies constitute the conceptual framework within which to achieve possible
solution(s) or provide suggestions for solving such problem(s). The pro-
blem statement and justification for this study has been set forth in
Chapter I. Given these problem statements and objectives concerning the
relationship between air pollution and vegetation, the intent of the analy-
sis is to determine the consequences and the magnitude of such air pollution
damage. This quantitative assessment of air pollution damage occurs within
the methodological framework defined for this study. Thus, specification
of the appropriate technique is central to the success of the analysis.
A number of conceptual issues have been raised implicitly concerning
a methodology for estimating agricultural damages associated with air pol-
lution. The approach should have a general equilibrium flavor, in that
both producing and consuming sectors are assessed simultaneously. Further,
interregional competition and comparative advantage constructs are required,
given that all regions considered compete to some extent for shares of
national commodity markets. In addition, substitution effects on the pro-
duction side need to be considered. All of these relationships are depen-
dent to some degree on the physical environment surrounding crop production,
including ambient air quality. This section discusses these concepts or
components required for such an analysis. The concepts are then extended
into a tractible mathematical model.
3.2 Methodological Framework
The conceptual issues outlined above involve a wide range of economic
relationships suggested by theory. For the methodology to be tractable in
terms of empirical analysis., these relationships must be combined in a log-
ical sequence and given a quantitative interpretation. This section provides
a more detailed methodological framework with which the concepts discussed
earlier may be quantified.
29
-------
1. Production Section
a. Production Functions
Assume that a specified area is divided into r heterogeneous regions
where r = 1,2, . . .,R. Regions are differentiated by such factors as
climatological conditions, soil quality and levels of ambient air quality.
Climatological conditions and soil quality determine jointly or separately
type(s) of crop(s) suitable for each region, whereas ambient air quality is
assumed to have different effects (favorable or unfavorable) on crops. In
each region there are i(i =1,2, . . .,1) farmers (processes) producing
j (j = 1,2, . . . ,J) agricultural, crops. However, it is possible for a
region to produce one or more crops and many regions to produce the same
crop. Thus, two regions may be viewed as homogeneous; each has identical
cropping alternatives, i.e., the same set of crops. Perfect competition is
assumed to prevail in the sense that each producer and each consumer acting
alone cannot affect the market price of a commodity, regardless of the
amount each one supplies and demands; but aggregate supply put forth by all
farmers (processes) in the area, due to the nature of the commodity des-
cribed in the earlier section can affect the market price of that commodity.
Assume further that, in the short run, farmers use both fixed and variable
inputs. Fixed inputs are land (measured in acreage used in cultivation)
and irrigation water. The factor supply function for such inputs may be
assumed to be perfectly inelastic. Variable inputs include labor, seeds,
fertilizer, and insecticide. These inputs are used in different amounts
from one stage of production to another. The factor supply functions for
these inputs may be assumed to be perfectly elastic for some (e.g., seed).
Labor is a special case, since unskilled labor is assumed to be available
at any time and thus has a perfectly elastic supply curve, whereas skilled
labor required for some processes of production is relatively scarce. Con-
sequently, its factor supply curve is rather inelastic.
There is another type of input, ambient air quality, which enters into
the production function. It would appear reasonable to assume that if air
quality deteriorates, production (yield) may be reduced or the costs of
production increased. Some of the crops produced are assumed to be perish-
able and thus have to be sold within a certain period of time, limiting the
use of carryover or buffer stocks across seasons. Transportation cost is
excluded under the assumption that it is treated as a fixed cost of compar-
able magnitude for producers and regions within the analysis. Thus, its
exclusion from the model may not significantly alter the result of the
analysis.
L.et lower case letters denote individual units and capital letters
rm rm
aggregate units. Thus, q.. denotes total production of crop j at the end
of the current season by farmer i in region r using soil type m where m =
1,2, . . . ,M. Let 1, la, f, is, w, se, k and z be total land, labor,
fertilizer, insecticide, irrigation water, seed, capital, and environmental
quality, respectively, associated with the production of crop j. The pro-
duction function of crop j can then be expressed as:
30
-------
q... = q^ (l,la,f ,is,w,se,k,z) (3.1)
If one assumes chat the above production function is linear, one will,
by taking the first-order partial derivative of q with respect to each in-
put, obtain a constant marginal productivity of each input included in the
model. Such a result might be interpreted as the shadow price of each in-
put. This general class of production functions is said to be homogeneous
of degree one, i.e., a constant returns to scale production function in
which output will be increased by the same proportion as an increase in
inputs .
Let P. be the market price of commodity j. Assume one market price
across all regions. Q. = ZEEq.. is the aggregate production of commodity j.
•-1 irm -11
S , P , and 0 are stocks of commodity j, prices of all other commodities,
J 1 .1
and all other factors such as income associated with the price of commodity
j. We can express the price forecasting equation for commodity j as:
P. = P.(Q.,S.,P .0.), 1 = 1,2, . . . ,L. (3.2)
J J J J 1 J
For analytical purposes, assume that the effects of all variables ex-
cept Q. in the price forecasting equation can be summed together into the
intercept term (by using the mean value of each variable multiplied by its
corresponding estimated parameters) , yielding a new equation for the price
of commodity j :
P' = P' (Q ) = c + dp , d < 0 (3.3)
j J J j
i.e., it is strictly a function of quantity. Such an equation can then be
used to estimate changes in commodity price associated with changes in the
level of production.
Assuming that each farmer in the area has the same objective of maxi-
mizing total revenue (above variable costs) subject to certain constraints,
the analytical problem then becomes a quadratic objective function with
linear constraints as follows:
max TR = PQ = cQ. + dQT. (3.4)
subject to:
=aL + a N + a F +als + a W +aSe + a K
1 j 2 j 3 j 4 j 5 j 6 j 7 j
-f aDZ. (3.5)
o J
Y = EY = L 4- W. (3.6)
J Jg J J
o
31
(3.7)
Q. ^ 0 (3.8)
-------
Equation (3.5) represents the production function for commodity j.
All variables in that equation are the aggregate of those defined earlier.
The expectation is that all the estimated parameters will be positive ex-
cept those for Z. (environmental quality — ambient air quality). The sign
of the estimated coefficient of Z , as mentioned earlier, is uncertain.
j
Equation (3.6) states that the amount of fixed inputs available, Y , is
j
simply the summation of various fixed inputs (in this analysis only land
and irrigation water) used by producing commodity j. Equation (3.7) indi-
cates that environmental quality (as measured by the degree of concentration
of specific air quality parameters) is assumed to be given. Finally, Equa-
tion (3.8) states that the output of all commodities must be non-negative.
If each producer takes price as given, i.e., the economy is perfectly
competitive, the objective function must be modified by the use of the sea-
ler value of "1/2," i.e.,
max TR = cQ. + 1/2 dQ2 (3.9)
J J
which then yields the following first-order condition,
= MR = c + dQ. = P. (3.10)
dQ. J J
as required for perfect competition.
The Lagrangian equation is:
L = cQ. + 1/2 dQ2 + A[Q. - (•)] + ufY. - EY. ] + <5lZ. - Z.] (3.11)
J J J JJ8 J J
o
Revenue maximization requires that the following first-order partial
derivatives
dQ + =o (3.12)
Y = Q. - (0 = Q (3.13)
d A J
— = Y - £Y =0 (3.14)
3U j jg
O
|j = Z. - Z. = 0 (3.15)
35 J J
be fulfilled and the Bordered Hessian be negative definite or negative semi-
definite.
Using the above procedure, variations in the level of Z should trans-
late into different levels of Q and thus changes in TR, as a result of
changes in prices due to such changes in Q. Moreover, it might also be
possible to calculate changes in Q resulting from tradeoffs between or among
inputs and Z, e.g., mitigative effects of fertilizer. Thus, air pollution
32
-------
damages, as measured by changes in output and price, may be assessed by
crop and region.
Another means of measuring damages of air pollution to an agricultural
crop is the use of a cost concept. In the presence of air pollution, farmers
may increase some other variable inputs such as fertilizer or labor to com-
pensate for the yield depressing effect of some pollutants. Under such a
situation, marginal cost and thus total cost will increase, while total
yield might decrease, remain constant, or increase, depending upon whether
such input adjustments are less than, equal to, or more than offsetting in
terms of the impact of air pollution. Assume that the objective function of
producers is to maximize total profit, which is defined as the difference
between total revenue and total cost where total revenue remains as defined
earlier, but total cost is a function of total production and different
levels of ambient air pollution concentration in the specified area. In
other words, the higher such concentrations, the greater the additional
costs producers must bear, in addition to the "normal" cost of production.
Mathematically, this again may be expressed as a non-linear objective
function with linear constraints, i.e..
subject to:
max ir = TR - TC = cO. + 1/2 dQ2 - C(Q ,Z.) (3.16)
'3 3 3 3
= a L + a N + a F + a Is + a W + a Se + a K
1 j 2 j 3 j 4 j 5 j 6 j 7 j
Y = ):Y (3.18)
j g jg
C = b A +bB +bZ (3.19)
j 1 J 2 j 3 j
Q. > 0 (3.20)
where C = total cost of producing commodity j
j
A = total fixed cost of producing commodity j
j
B. = total variable cost of producing commodity j
Z = levels of air pollution concentration (per unit of measurement)
j
b ,b ,b are estimated parameters where b ,b > 0, b > 0.
All other variables and parameters are as defined earlier.
The Lagrangian equation is:
L = cQ + 1/2 dQ2 - C(Q.,Z.) + X[Q. - (a L . + a N + a F.
J J 33 J 1 .] 2 j 3 j
+ a Is + a W + a Se + a K + a Z )] + p[Y. - ZY. ]
A j 5 j 6 j 7 j 8 j 3 18
o
+ Y[C. - (b^. + b2B. + b3Z.)] (3.21)
33
-------
The first-order conditions for profit maximization require that the fol-
lowing partial derivatives:
fi=c + dQ. -f^A=0 (3.22)
i. = -t-Aas~YV° <3-23)
.1 J
and the constant maximization equations stated above be fulfilled. The
second-order conditions for profit maximization require that the Bordered
Hessian be either negative definite or negative semi-definite.
From the first-order partial derivatives obtained above, one will then
be able to solve for Q. by varying the value of Z.. After solving for the
value of Q., the values of cost, revenue, and profit can then be obtained.
Such results will provide a measure of the damages of air pollution to
commodity j.
Risk and Uncertainty
Quadratic risk programming is usually regarded as a theoretically
appealing technique for analyzing impacts, of risk aversion on farm planning.
Let M be the gross income associated with agricultural crop j. Then
M = P q where P is the market price of commodity j and it is assumed to be
j j j 2 2
distributed normally with mean y and variance a , i.e., P Z N(u,cr ).
Let the utility-of-income function be exponential in the form:
U(M) = « - Bexp(-AM) where =, 3, A > 0 (3.25)
<*, 6 are estimated parameters and A is an arbitrarily assigned degree of
risk aversion of the decisionmaker(s) toward commodity j. However, it is
possible to directly estimate the value of A. Wiens (1976) has suggested
the following procedure to estimate A:
Define a quadratic programming problem of maximizing:
W = y'X - (A/2) X'EX = E(R) - (A/2)o(R)2 (3.25)
subject to:
AX <_ C* (3.26)
where X is a vector of activities; R is net income; A is the technology
matrix relating units of inputs to one unit of output (activity); C* is the
level of resource use; A is degree of risk aversion; and y is the mean of
income.
The Kuhn-Tucker conditions for an optimal solution to the above quad-
ratic programming problem require that:
-------
p. - XIX - A'. 0 = 0 (3.27)
i
for all non-zero activities i in the solution, where E, V and A are, re-
i i
i
spectively, the ith row of £, the ith element of u, and the ith column of A.
Substituting X", the actual activity level, for X and r, the actual market
prices, for 0, A can be estimated as:
X* = E(u - A'r)/IX* (3.28)
i i
i
While this should hold for each production activity if all assumptions are
exactly fulfilled, empirically an average overall production activity will
suffice.
Following the method suggested by Wiens (1976), the expected value of
the utility of income function is:
E(U(M)] = « - Bexp[-Xuq. + (A"/2)o~q7] (3.29)
To maximize equation (3.29) with respect to q. is equivalent to maximizing:
W = uq. - (A/2) var (q.) = E(M) - (A/2) var (M) (3.30)
where P can be interpreted as the shadow price of q.. This is a conventional
E,V objective function. Applying Weins" method described above, the values
of A can then be estimated. Hazell (1971) points out the following advan-
tages of using the EV criterion for farm management research:
"(a) The criterion is consistent with probability statements with re-
spect to the likelihood of occurrence of different levels of in-
come for any given farm plan. If total gross margins can be ex-
pected to be approximate].)' normally distributed, and if the var-
iance-covariance coefficients used can be regarded as non-stoch-
astic or subjective parameters, then such probability statements
are easily derived by using tables for the normal deviate statis-
tic . . .
(b) The variance V is totally specified by the variance-covariance
coefficients; and when subjective values of these parameters are
available or can be found, the variance is no longer estimated
from the sample of observed gross margin outcomes . . .
(c) The criterion is consistent with the Separation Theorem (see
Johnson, 1967, pp. 614-620) and allows more general solution to
the farm diversification problem given a riskless option (for the
decisionmakers-)." [pp. 55-56, expression in the parentheses is
added]
Due to the fact that use of EV method requires a special computer
algorithm, Thomson and Hazell (1972) suggest that it be replaced by the mean
absolute income deviation (MAD) and used to obtain a solution through
35
-------
standard linear programming codes with the parametric option. Hov/ever, when
the sample mean absolute deviation is used rather than the sample variance,
the reliability of the estimated efficient, EV, farm plans is necessarily
weakened [Thomson and Hazell, 1972, p. 503]. Nevertheless, MAD is still a
best substitute when access to such a special computer algorithm is not
possible and provided that certain adjustments are also used in order to re-
duce error due to the use of sample MAD rather than the sample variance.
Such an adjustment has been suggested by Thomson and Hazell (1972) by using:
. - P0.)U -
li 2i
Z(l - Pn .)(! - V.)
li i
i
x 100 percent (3.31)
where P and P are the estimated probabilities of the correct ranking of
li 2i ,'
the ith farm plan for the sample MAD -'and variance respectively and V is the
/ i
ithe variance ratio as a weight in the MAD model.
Consumer Sector
In aggregate models of the consuming sector, it is convenient to assume
that there are n individuals x^ith similar taste and preferences. Each
individual has a utility function which is concave and is the function of
various goods and services consumed, i.e.,
u = u (q q . . .,q ), n = 1,2, . . .,N (3.32)
n n nl n2 nJ
where u is the utility of individual n and q (j - 1,2, . . . ,J) is the
n nj
j th commodity or service consumed by individual n. q is, of course, a
nj
function of the price of commodity j, prices of all other commodities or
services, and income associated with individual n, i.e.,
qnj = qnj(Pj'?o'V °= X»2> ' ' "° (3'33)
Total demand for commodity j is given by:
N
Q. = £ q .(P.,P ,M ) (3.34)
J n=1 V) J o n
Individual n then maximizes his utility subject to his budget constraint,
i.e. ,
max u = (q , . . . ,q .) (3.35)
n nl nj
subject to:
P.q + P<1 + • • • P.q . = M (3.36)
1 nl 2 n2 j nj n
The Lagrangian equation is:
L = Un(qnl' ' • "qnj + u[Mn ~ (Piq i + P9q ? + • • • P-q -)] (3'37)
n nj- nj n I nl 2 nl j nj
36
-------
The necessary conditions for utility maximization require that:
3L/3q = 9u /3q - uP = 0
nj n nj j
J
9L/3u = E P.q . = M
j=1 J nj n
(3.38)
(3.39)
and the associated sufficient conditions will be fulfilled if the Bordered
Hessian is either negative definite or negative semi-definite.
With changes in price due to effects of air pollution on crop produc-
tion, individual n's demand function will change. If there is no change in
his income level (i.e., uncompensated in the case of price increase or taxed
in the case of price decrease), his level of well-being will be altered.
This alteration of consumer welfare may then be approximated by changes in
"consumer's surplus" which will be introduced in the next section.
Consumer's and Producer's Surpluses
Of the three concepts of economic rent the Marshallian concept of
consumer's and producer's surpluses is most applicable if one assumes that
all other prices and incomes of all individuals concerned are constant.
Let:
P. = F(.q.,e.)
J J J
(.3.40)
be the demand function of commodity j denoted by d. in the following dia-
gram where P and q are the price and quantity respectively. 8 is the
j J j
shift parameter denoting changes in price of commodity j due to, say, changes
in total supply arising from air pollution. It is assumed that the demand
curve, d , is downward sloping.
j
37
-------
Let the supply function of commodity j be represented by:
P = G(q ,8.) (3.41)
3 ' J J
The supply curve, s., is assumed to be upward sloping. It starts from the
origin under the assumption that farmers will not supply any of commodity j
if its price is zero. For simplicity the subscript j will be dropped.
Market equilibrium for commodity j, as shown in the above diagram, will be
obtained when the quantity supplied and demanded is q and the market price
is P . Consumer's surplus is then defined as the area under the demand
o
curve d , and above the equilibrium price, P , or the triangle P ab. Pro-
j oo
ducer's surplus or net return to factor owners is the area under the equil-
ibrium price but above the supply curve or the triangle OP b. The sum of
consumer's and producer's surpluses is given by:
^q
R(Q) = j ° !F(q,6) - G(q,6)]dq (3.42)
o'
= Oabq - Obq = Oab (3.43)
o o
From the above equation one can compare the value of R associated with
different values of 6, e.g., if we let 9 = 6(0) be the initial situation
o
when there is no air pollution and 8 be the situation with some level of
air pollution then the difference between R(0 ) - R(o) will measure changes
in consumer's and producer's surpluses due to increase in level of air pol-
lution.
This method when applied via the mathematical concepts developed earlier
in this report is analogous to Samuelson's (1952) "net social payoff" theory
in which he relates Enke's formulation [Enke, 1951] to a standard problem in
linear programming, the so-called Koopmans-Hitchcock (1941, 1949) minimum
transport cost problem. Basically, the Samuelson's net social payoff is
defined as. the sum of the algebraic area under the excess demand curves of
n individuals minus the total transport costs of all shipments [Takayama and
Judge, 1964a, 1964b]. The objective is to artificially convert the descrip-
tive price behavior into a maximization problem which can then be solved by
using trial and error or a systematic procedure of varying shipments in the
direction of increasing social payoff [Takayama and Judge, 1964b, p. 510].
However, in the formulation outlined in this section, quadratic programming
can be used to approximate (see the subsequent section on analytical model)
such.a payoff.
3.3 An Analytical Model for Measuring Impacts of Air Pollution on Agricul-
tural Crops
The conceptual model and mathematical concepts developed earlier in
these sections can be used to construct a mathematical programming model
capable of achieving some of the goals set forth in this study. This model
can be explanded further by incorporating into it some additional concepts
38
-------
such as technical substitution possibilities, endogenous air pollution re-
sponse functions and risk. The derived model will, it is hoped, present a
realistic example of the agricultural sector within the constraints imposed
by data limitations. Data requirements are extensive for such programming
techniques and some of them, such as quadratic programming, require special
computer algorithms. Nevertheless, the incorporated model should be analy-
tically feasible and mathematically tractable. The degree of sophistication
will be dependent on the availability of the required data and computer
software.
In order to simplify some of the notations given in the earlier sec-
tions, matrix notation will be used in the models proposed below. However,
all notations will remain as described earlier. It is assumed that air
pollution in the specified area adversely affects crop production, and,
consequently, may affect producers and consumers. Mathematically, the ob-
jective of the model is to maximize a "quasi-net social payoff" which is
defined as the summation of consumers' and producers' surpluses and subject
to certain constraints, i.e.,
T T T
Max QNSP = C Q + 1/2 Q DQ - E Q (3.44)
subject to:
AQ = b (3.45)
Q _> Q (3.46)
where
QNSP = quasi-net social payoff (a sealer),
Q = jxl column vector of agricultural crop production, where pro-
duction equals yield per acre times acreage planted.
C = jxl column vector of constants (intercepts in a linear demand
structure).
D = jxj negative diagonal matrix (negative definite of coefficients
representing slope values with the linear demand structure).
E = jxl column vector of unit cost of production.
A = gxj technology matrix relating units of inputs to one unit of
output (g constraints and j variables).
b = gxl column vector of fixed inputs (land, water).
and T denotes matrix transportation.
The. summation of the first two terms on the right-hand side of equation
(3.44) is the total revenue for commodities Q. When integrated it represents
the area under the demand curve but above the horizontal axis from the ori-
gin to the equilibrium amount demanded. The last term in the right-hand
side of equation (3.44) is the total cost of production whose first-order
derivative is the marginal cost. The rising portion of the marginal cost
curve can then be treated as the short-run supply curve. Therefore, total
cost of production can be considered as- the returns to factor owners. The
difference between the sum of the first two terms and the third term is the
sum of producers' and consumers' surpluses over all commodities. It is
equivalent to the quasi-net social payoff. Maximizing the objective func-
tion is analogous to maximizing a quadratic "quasi-net social payoff" sub-
ject to linear constraints.
39
-------
The above price endogenous model can be expressed as a quadratic pro-
gramming economic model. Such a model formulation will result in solution
values for price and quantity of each commodity which maximize the value of
QNSP.
In order to assess the impacts of air pollution upon agricultural crops,
one may either introduce a variable, Z*(0 < Z* < 1), defined as an index of
j - j ~
crop yield reduction (% of yield reduction divided by 100) associated with
crop j, into the production function (yield) or the cost function. If Z*
j
enters directly into the production function, it may affect crop yield but
not necessarily total cost. Alternatively, Z* may be treated through the
cost function as an investment in ameliorating air pollution effects on
agricultural crops by means of either increasing use of other variable in-
puts or relocating the site of production. The former involves the problem
of technical substitution possibilities. For example, can fertilizer appli-
,cation rates be increased to partially offset the negative impact of air
pollution? The latter can be achieved by comparing two neighboring areas,
one with and the other without air pollution, using the same technique of
measuring crop yield for same type(s) of crop(s). If total yield in the
area with air pollution is lower than yields in adjacent areas, (keeping all
other factors constant) one might suspect that such a reduction in yield is
caused by air pollution. Thus, it might be possible to compare cost of re-
location vs. investment in air pollution abatement.
Consider the case when Z* enters directly into the production function.
If one lets Z. represent air pollution concentration, in parts per hundred
million, associated with agricultural crop j, Z. can then be calculated,
by using the formulas given by Larsen and Heck (1976), Oshima (1975), Oshima,
et. al. (1976, 1977) or others. Applying this method to all other crops
under study will provide a vector of Z* which is a jxl column vector for j
crops. Then calculate total production of each crop by the following
formula:
Q* = (I - Z*)LTY (3.47)
where
Q* = jxl column vector of total production of j crops with air pollution.
Z* = jxl column vector of yield reduction index.
I = jxl vector of unity.
L = jxl column vector of acre of land used for cultivating j crops.
Y - jxl column vector of yield of j crops.
The model given under equations (3.44) through (3.46) will then be modified
to be:
Max QNSP = CTQ* +1/2 Q*TDQ* - ETQ* (3.48)
subject to:
40
-------
AQ = b (3.53)
Q 1 0 (3.54)
As mentioned -earlier, agricultural production involves various degrees
of uncertainty. If farmers are assumed risk averse then their production
decisions will reflect the uncertainty of climatological conditions for the
coming season, and hence the total quantity supplied, and thus prices in the
next season. Thus, production risks are multiplicative in nature, meaning
that it is the slope of the supply curve which contains the important stoch-
astics [Hazell and Scandizzo, 1975, p. 641]. Therefore, the traditional
method of treating the risky component of supply as a constant added to the
intercept term may not be appropriate. Hazell and Scandizzo (1974) use
linear programming models with multiplicative supply function. Other methods
frequently used in the empirical supply analysis are the econometric estima-
tion of constant elasticity of substitution and Cobb-Douglas functions.
Another method that is widely used in farm management is the EV criter-
ion described earlier. The analytical model may be further modified to be
the following form:
Max QNSP = C Q* + 1/2 Q*TDQ* - ETQ* - A(Q*TftQ*) (3.55)
AQ* = b (3.56)
subject to:
/
Q* _> 0 (3.57)
where
A = a constant value of risk aversion coefficient.
Q = jxj matrix of variance-covariance of income associated with each
type of crop.
if Z* enters only into the yield function.
If Z" enters directly into the cost function, then the formulation be-
comes:
Max QNSP = CTQ + 1/2 QTDQ - E*TQ - MQTfiQ) (3.58)
subject to:
AQ = b (3.59)
Q _> 0 (3.60)
Thus, the quasi-net social payoff will be lower for higher values of
the risk aversion coefficient given no change in income and vice versa. In
the above formulations all but A can be observed. However, the values of A
can be assigned arbitrarily or can also be estimated by using the method
suggested earlier.
41
-------
CHAPTER IV
AIR POLLUTION YIELD RESPONSE RELATIONSHIPS
4.1 Introduction
Effects of air pollution on agricultural crops, such as vegetables and
field crops, have been well documented, as discussed in Chapter II. Although
results obtained by various researchers are mixed, depending in part upon
different methodologies and varieties of crops chosen for each study, the
'effects on some vegetable crops appear to be particularly acute. Controlled
experiments performed in laboratories and greenhouse tests tend to indicate
consistent adversary effects of air pollution on crop yield. However, simi-
lar results may not always be obtained in actual field tests due to the fact
that various factors such as climatological conditions are either difficult
or impossible to control and are capable of moderating impacts of air pol-
lution on yield. Given the importance of selected vegetable and field crops
to the agricultural sector of the study region and the significant share of
the national market held by the region, the relationship between ozone and
vegetable yields is a critical component of this analysis.
This chapter discusses the development of a set of yield-ozone rela-
tionships for the study area in general and the four production regions in
particular. These relationships are derived from research discussed in
Chapter II and some specific concepts presented in this chapter. The fol-
lowing subsection presents a hypothetical relationship between air pollution
and yield. A quantitative relationship is obtained by using methods which
will be more fully described in another subsection. The last subsection in
this chapter provides estimated yield effects by crop and production region.
4.2 A Hypothetical Relationship Between Air Pollution and Yield
Most studies concerning the effects of air pollution on agricultural
crops concentrate largely on physical damages such as leaf-drop and growth
retardation of plants. Analyses of the specific relationship between air
pollution concentration levels and yield reduction have been limited. Ob-
viously, such a relationship is important in economic analysis of the impact
of air pollution, given the need to directly estimate a market value or loss
associated with air pollution.
Based on research discussed above, one may hypothesize that a negative
relationship exists between ozone concentration and crop yields. A simple
method for testing such a relationship is to examine the correlation coef-
ficient between the level of air pollution and yield over a certain period
of time for each crop. Again, one would hypothesize that cet. par, an
42
-------
increase in the level of air pollution concentration will lower yield. In
other words the correlation coefficient between air pollution and yield
should be negative. Further, the higher the coefficient the greater is the
degree of relationship between air pollution and the crop, assuming the re-
lationship is statistically significant.
To obtain correlation coefficients for the entire set of crops and re-
gions, data on yield per acre for the 12 vegetable and 2 field crops were
collected, covering the period from 1957 to 1976 for each county in the 4
major vegetable growing regions.!/ The yield per acre was then correlated
against the maximum level of oxidant/ozone concentration taken from the
publication "California Air Quality Data" for each county for the same
period. However, due to a lack of complete data on ozone concentration and
crop yields, only three counties (Orange, Riverside and Kern) and some crops
were included in the correlation analysis. Orange and Riverside counties
represent an area of more severe air pollution, whereas Kern County, a major
agricultural county, was selected to represent an area of relatively low
ozone concentration. The correlation coefficients between air pollution
concentration (annual maximum level of oxidant/ozone concentration in parts
per hundred million) and yield for these three counties are given in Table
4.1.
The correlation coefficients presented in Table 4.1 tend to conform to
a^ priori expectations; that is, most of the coefficients have the expected
sign although not all are statistically significant. This tends to suggest
that air pollution in these areas has had some adverse effects on yield.
However, correlation analysis does not imply causality, thus these results
only lend support to the earlier supposition concerning yield and air pol-
lution.
In order to further test the effects of air pollution on yield a simple
production function relationship between yield per acre, the hourly maximum
of oxidant/ozone concentrations and the crop acreage harvested for the three
counties from 1957 to 1976 was estimated for some crops. The relationship
was estimated via ordinary least squares, assuming a linear functional rela-
tionship. The production relationship was first estimated as strictly a
function of ozone concentration, then acreage only and finally as a function
of both variables.
Results obtained, as shown in Tables 4.2a, 4.2b, and 4.2c were gener-
ally not statistically significant, although the estimated coefficients for
ozone had the expected negative sign in most equations. The coefficients of
determination (R2) are very low with insignificant F-statistics. The Durbin-
Watson in some equations are inconclusive. This means that variations in
crop yield per acre can only be slightly explained by changes in the levels
of oxidant/ozone concentration. As expected, the multiple regression had
slightly higher levels of significance than the simple regressions.
4.3 Methods of Estimating Effects of Air Pollution Concentration on Yield
Earlier analysis suggests that air pollution (ozone) does indeed have
a negative effect on yield. In order to estimate more precisely the effect
of air pollution on yield, the Larsen and Heck and the Oshima equations as
43
-------
Table 4.1
Correlation Values between Level of Oxidant and Yield
Crops
Oxidant, average of hourly maximum (pphm)
Orange County
Riverside County
Kern County
Beans, Lima
Cantalopes
Carrots
Cauliflower
Celery
Lettuce
Onion, Green
Potatoes
Tomatoes, Fresh
Tomatoes, Process
Cotton
Sugarbeets
-0.13626
0.62585*
-0.29654*
-0.19098*
-0.13500
-0.30188-
0.05014
-0.23151*
0.42106*
0.10914
0.01628
0.01481
-0.28838*
-0.23089*
0.19790*
-0.16470*
-0.33044*
-0.39337*
-0.05636
-0.30327*
*Denotes those coefficients significant at the 20% level.
44
-------
Table 4.2a
Regression Coefficients for Selected Crops, Orange County, 1957-76
Dependent Variables
Beans, Lima
Yield (Tons/acre)
Cauliflower
Yield (CWT/acre)
Celery
Yield (CWT/acre)
Lettuce, Head
Yield (GUI'/acre)
Tomato, Fresh
Yield (CWT/acre)
Constant
Independent Variables
Oxidant (pphm)
Est. Coef. T-Value
4.614 -0.0187 -0.58
5.018
5.159 -0.0052 -0.15
144.119
235.003
155.092
589.803
499.655
557.367
295.236
231.671
283.851
524.710
740.489
2.1054 3.40
-
2.0625 3.24
-1.6592 -1.32
_
-1.7191 -1.35
-1.2048 -0.83
-
-1.5784 -1.09
-1.5756 -0.58
-
744.570 1 -0.1349 -0.07
Acreage
Est. Coef. T-Value
-0.0012 -1.38
-0.0011 -1.21
-0.0150 -0.77
-0.0087 -0.55
0.0189 0.03
0.0208 0.84
0.0325 1.18
0.0383 1.37
-0.4807 -4.78
-0.4794 -4.57
Summary Statistics
R2
0.0186
0.0954
0.0966
0.3917
0.0321
0.4023
0.0879
0.0299
0.1240
0.0365
0.0713
0.1317
0.0182 |
F
0.34
1.90
0.91
11.59
0.60
5.72
1.74
0.55
1.20
0.68
1.38
1.29
0.33
0.5592 i 22.84
0.5594 ! 10.79
i
DW
! 0.59
; 0.72
0.71
! 1.51
1.37
1.53
1.00
1.05
1.14
0.88
0-74
0.99
1.25
1.15
1.15
Tomatoe,
Yield
Processing
(Tons/acre)
24.838
19.928
24.085
-0.1205
-0.1215
-1.34 :
! 0.0002
-1.32 ! 0.0003
0.22
0.28
0.0911
0.0028
0.0953
I
1.80
0.05
0.90
1.26
0.89
1.25
-------
Table 4.2b
Regression Coefficients for Selected Crops, Riverside County, 1957-76
Dependent Variables
Beans, Green Lima
Yield (Tons/acre)
Cantalope
Yield (CWT/acre)
Carrots
Yield (CWT/acre)
Lettuce, Head
Yield (CWT/acre)
Onion, Green
Yield (CUT/acre)
Potatoes
Yield (CVJT/acre)
Tomato, Fresh
Yield (CWT7acre)
Cotton
Yield (Ibs/acre)
Sugarbeets
Yield (Tons/acre)
Constant
2.4161
5.0358
5.6386
164.457
163.975
213.048
122.741
212.491
45.293
185.394
211.929
189.710
251.105
119.844
110.930
270.514
332.465
366.201
340.654
162.343
L95.421
1316.92
478.02
392.62
16. 1] 26
22.3874
16.1528
Independent Variables
Oxidant (pphm)
Est. Coef. T-Value
0.0159 0.21
-0.0126 -0.1?
-0.5960 -1.01
-1.0226 -2.25
3.6983 1.97
3.7706 2.20
0.4628 0.47
0.4541 0.44
0.1616 0.07
0.2024 0.16
0.0575 0.06
-0.6614 -0.87
-2.6267 -1.28
-0.7081 -0.46
-5.9587 -1.01
1.3758 0.24
0.1604 0.86
0.1579 0.75
Acreage
Est. Coef. T-Value
-0.0078 -1.65
-0.0080 -1.60
-0.0069 -3.13
-0.0080 -3.91
0.0137 1.92
0.0140 2.16
-0.0009 -0.13
-0.0006 -0.08
0.1657 7.03
0.1657 6.84
-0.0061 -3.27
-0.0066 -3.36
0.1470 4.62
0.1424 4.19
0.0310 3.02
0.0323 2.71
0.0003 0.36
0.00003 0.03
R2
0.0025
0.1320
0.1335
0.0536
0.3529
0.5016
0.1773
0.1706
0.35C8
0.0119
0.0009
0.0123
0.0003
0.7330
0.7334
0.0002
0.3723
0.3995
0.0832
0.5430
0.54U5
0.0533
0.3362
0.3384
0.0392
0.0072
0.0392
F
0.0454
2.74
1.31
1.02
9.82
8.55
3.88
3.70
4.67
0.22
0.02
0.11
0.01
49.41
23.38
0.01
10.70
5.66
1.63 '
21.38
10.32
1.01
9.12
4.35 '
0. 73 :
0.13 '
0. 35 !
DW
2.30
2.42
2.49
1.56
2.03
2.53
1.62
1.07
1.95
1.18
1.16
1.16
0.56
2.00
2.02
1.06
1.74
1.91
1.10
2.18
2.12
1.65
1.58
1.56
1.63
1.58
1.64
46
-------
Table 4.2c
Regression Coefficients for Selected Crops, Kern County, 1957-76
Dependent Variable
Cantalope
Yield (CWT/acre)
Carrots
Yield (CWT/acre)
Potato
Yield (CWT/acre)
Cotton
Yield (Ibs/acre)
Sugarbeets
Yield (Tons/acre)
Constant
189.936
225.148
247.578
415.033
303-150
370.558
396.505
515.002
518.229
1122.86
1146.15
1189.35
25.827
19.102
19.823
Independent Variables
,
Oxidant (pphm) Acreage
Est. Coef. T-Value Est. Coef. T-Value
i
-1.2713 -0.71
-
-1.2876 -0.75
-4.9290 -1.49
-3.3722 -0.88
-3.9781 -1.82
„
-0.6357 -0.33
-1.5612 -0.24
_
-2.0374 -0.30
-0.1965 -1.35
_
-0.0314 -0.19
i
'
R2 F
DW
0.0271 0.50 1.91
-0.0149 -1.63 0.1290 2.67 1.97
-0.0149 -1.62 , 0.1568; 1.58 . 2.14
0.1092; 2.21 1.43
0.0067 1-45 : 0.1050
0.0044 0.83 j 0.1439
0.1547
-0.0042 -4.44 0.5225
-0.0040 -3.65
-0.0002 -0.35
-0.0003 -0.39
0.0002 2.40
0.0002 1.85
0.5255
0.0032
0.0069
0.0121
0.0920
0.2428
0.2445
2.11 1.36
1.43 1.44
3.30
19.70
9.42
0.06
0.13
0.10
1.82
5.77
2.75
0.73
0.94
0.94
1.02
1.08
1.03
0.82
1.03
1.03
-------
described in Section 4.2 are used. The Larsen and Heck relationships measure
percentage leaf damage associated with different levels of air pollution
concentration (ozone) and hours of exposure. They thus take into account
the intuitively obvious fact that leaf damages may be more serious if a
given plant is exposed to either higher levels of air pollution or a constant
level for longer duration. There is one difficulty attendant to the use of
the Larsen and Heck relationship, that being leaf damage may not correspond
to yield reduction, especially for fruit or root crops. Thus, certain ad-
justments must then be made to translate leaf damage to yield reduction.
Based on empirical results reported by Millecan, a general "rule-of-thumb"
can be used to relate leaf damage to percentage of yield reduction. These
translations are presented in Table 4.3.
An additional problem concerning the use of the Larsen and Heck method-
ology is that only a limited set of equations has been estimated. Of these,
very few correspond to the set of crops included in the study. To circum-
vent this problem certain equations have been selected from the Larsen and
Heck set to serve as "proxies" for general classes of crops. This assign-
ment of equations to represent groups of included crops is based on a review
of secondary information concerning degree of susceptibility of each plant
or plant group to air pollution to establish some consistency of response.
The representative crop equations used in this study are presented below.
Larsen and Heck equation Study Crops
1. Pinto Beans approximates Green Lima Beans
Celery (times 0.8)
2. Radish " Onion, Fresh (times 1.2)
Onion, Processing (times 1.2)
Sugarbeets
3. Spinach " Head Lettuce (times 0.6)
4. Summer Squash " Broccoli
Cantalopes
Carrots
Cauliflower
5. Tomato " Tomato, Fresh
Tomato, Processing
Potato
After the selection of a specific equation to serve as a proxy for a
particular study crop, a table of leaf damage (percent) associated with
actual levels of air pollution concentration (ozone) as measured at repre-
sentative air monitoring stations for each county and hours of exposure (8,
10, 12 hours) are calculated. For the purposes of this study, the level of
air pollution concentration is classified into three categories: (1) Air
pollution concentration level A represents the annual hourly maximum re-
corded at the county monitoring station. It is thus the highest level of
oxidant/ozone concentration in each year; (2) Level B is the annual average
48
-------
Table 4.3
A "Rule-of-Thumb"* Relating Leaf
Damage to Yield Reduction
% of Leaf Damage
% of Yield Reduction
1
6
11
16
2.1
26
31
36
41
46
51
56
2
7
12
17
22
27
32
37
42
47
52
57
3
o
O
13
18
23
28
33
38
43
48
53
58
4
9
14
19
24
29
34
39
44
49
54
59
5
10
15
20
25
30
35
40
45
50
55
60
corresponds to
1.
2 .
4.
9.
15.
19.
23.
27.
30.
33.
36.
0
1
3
5
0
0
9
8
6
6
6
6
0.
1.
2.
5.
10.
16.
20.
24.
28.
31.
34.
37.
1
9
7
1
1
0
9
7
?
2
2
2
0.
1.
3.
6.
11.
17.
21.
25.
28.
31.
34.
37.
4
5
0
0
2
0
7
5
8
7
8
8
0.
1.
3.
7.
12.
18.
22.
26.
29.
32.
35.
38.
8
7
5
0
5
0
5
3
4
3
4
4
1.
2.
4.
8.
14.
19.
23.
27.
30.
33.
36.
39.
0
0
0
0
0
0
0
0
0
0
0
0
*It should be noted that Millecan's "rule-of-thumb," as cited, applies
only to 20% leaf damage. For damage in excess of 20%, yield reduction was
derived from secondary sources concerning genera] crop sensitivity as well
as information relating to yield reduction at high levels of physical damage.
49
-------
of the hourly maximum;^/ (3) Level C is the annual average of the average
hourly maximum._3/ Table 4.4 contains the levels of oxidant/ozone concen-
tration by station and region for the period 1972 to 1976 and the average
of that period classified according to the three levels mentioned above.
The second type of equation used in the yield response analysis is
that developed by Oshima, et. al_. This equational structure is used to
measure yield reductions in cotton, California's major field crop. The
Oshima, et. al. equations, unlike those of Larsen and Heck, relate ozone
doses directly to percentage of yield reduction. To date, this type of
equation has been estimated for only three crops; alfalfa, cotton and toma-
to. In order to obtain an estimated percentage yield reduction, a cumula-
tive ozone dose greater than 10 parts per hundred million (the required
California standard) over the growing season in each year for each county
is needed. Such data for 1976 were not available at the time of this study.
However, the cumulative dose can be calculated for alternative levels (e.g.,
8 and 20 parts per hundred million). The 8 pphm level was selected for use,
with levels measured at air monitoring stations in or close, to the growing
regions for cotton. The stations include Indio-Oasis for Riverside and San
Bernardino Counties; half the level of ozone doses observed in Indio-Oasis
for Imperial County;4/ and Delano for Kern and Tulare Counties. The cumu-
lative ozone dose is obtained by adding up total doses exceeding 8 pphm
from March to September 1976 (representing the growing season) for each
station, as reported in Table 1 in "California Air Quality Data, Summary of
1976 Air Quality Data Gaseous Pollutants." The average value of yield re-
duction across county is then used for each region producing cotton.
4.4 Estimated Results of Yield Reduction Due to Air Pollution
From the three levels of concentration in Table 4.4, concentration
level C was selected for use in estimating the degree of yield reduction to
be used in the study. Such a level is the most conservative level of the
three, thus perhaps representing a lower bound on yield damage. In the
South Coast region, the Pasadena, Anaheim, Indio-Oasis, San Bernardino,
Santa Maria, San Diego, and Ventura air monitoring stations are used to cal-
culate air pollution concentration for their respective counties. The Mon-
terey station in the Central Coas.t was eliminated, as was the Bakerfield
station in the Southern San Joaquin Valley on the assumption that levels at
these stations are not representative of the levels in the actual growing
areas.
In calculating the effect of air pollution on yield, two values of air
ppllution concentration (both representing level "C") are used. One is the
average of 1972-1976 and the other is the 1976 level (for level C). The
estimated yield reduction for a 12 hour exposure for the study crops is
given in Tables 4.5 and 4.6. Table 4.7 is the average percentage of yield
reduction across- county in each region attributed to the presence of air
pollution. Table 4.8 gives the actual yield per acre for the average of
1972-1976 and the 1976 crop year derived from Tables 1.2, 1.3 and 1.4 of
Chapter I. These yield figures thus represent actual yields, i.e., yields
in the presence of air pollution. Finally, Table 4.7 is used to estimate
Table 4.9, the production or yield per acre in the absence of air pollution
50
-------
Table 4.4
Levels of Oxidant/Ozone Concentration (pphra)
Area/Station
Imperial Valley
El Centre
South Coast
Down to '.on
Pasodona
Anaheim
Rivcrside-Robidoux
Indio-Oasis
San Bernardino
Santa B.irbara
Santa M.iria
San Dirgo
Ventura-Telegraph Rd.
Central Coast
MoiitL-rey
Salinas
HolUsion
San Luis Obispo
Southi;r:i S.^n Joaouin
Uakorf ield
Delano
Visalia-Old Jail
County
Imperial
Los Angeles
Los Angeles
Orange
Riverside
Riverside
San Bernardino
Sunta Barbara
Santa Barbara
S.in Ulvgo
Ventura
Hinueroy
Monterey
S.ir. Benito
San Luis Obispo
Kern
Kurn
Tulare
72 73
-
25 52
38 45
35 32
50 39
25 22
42 42
13 24
15 13
17 24
-
11 14
9 15
- 13
12 11
18 17
-
20 19
Level
74 75
17 13
25 25
34 32
25 17
39 35
22 20
33 38
21 25
15 6
hS 15
20 16
14 8
12 8
14 13
15 U
17 12
- 12
20 13
A
76 Av.
8 13
34 32
34 36
30 28
36 40
16 21
30 37
17 20
12 12
16 18
19 18
7 11
11 11
15 14
11 12
12 15
11 11
13 17
72 73
-
18 20
23 23
18 19
27 24
16 18
18 22
9 13
8 7
12 11
-
7 9
7 9
- 10
3 8
10 11
-
13 11
Level
74 75
11 9
17 18
24 22
16 13
26 24
14 12
22 22
12 12
7 5
12 J.1
14 11
7 6
8 5
10 9
9 7
12 8
- 10
12 10
B
76 Av.
6 7
22 19
24 23
16 16
23 25
11 14
16 20
10 11
9 7
13 12
13 13
5 7
6 7
9 10
8 8
8 10
10 10
10 11
72
-
7
10
5
12
8
8
4
4
4
-
3
4
-
4
6
-
7
73
-
7_
10'
6
11
7
11
6
4
5
-
4
4
5
4
6
-
7
Level
74 75
6 6
8 8
U 10
6 5
13 10
7 7
9 10
6 6
4 3
6 6
7 5
4 3
4 3
5 5
5 4
7 5
4
8 5
C
76 Av.
3 5
8 8
11 10
6 6
10 11
6 7
7 9
5 4
5 4
6 5
5 6
3 3
4 4
5 5
4 4
5 6
6 5
5 6
A « M.ixlaium value for the year. It is the rcaxirum of t_he month c.r.d also the maximum of each day. Ic is obtained by first obtaining
the l.uurly n.ixlr.un for oncii day (24 values) then pick the maxicun to represent the maximum for each day and pick the ciaxinum to
rL'proaoiit each month (12 values). Then pick the maximum to represent each year (5 values) and the average over a 5-year period.
B « Average; of the B.-ixlctu-n of hourly mix. It is obtained by the same procedure as in A but the final value for each year is obtained
by averaging the r.o:i:lily n;ixir.i:rc and then the average o/cr a 5-yo.ir puriod.
C • Av.;r.iRo of the .iVL-r.ifx- of hourly m-mlnum. Il is vhroliK-J by avtr;i(;IiitJ the uvor.it-i: of tho huurly-mximuro. Then nverage over a
-
-------
Table A.5
Percentage Yield Reduction for 12 Hour Exposure,
Using the Average Value of Oxiciant/Ozor.c
Concentration from 1972-1976, "C" Level
R^s Ion/Sent Ion County
n C^-ruru Imperial
<* j cTi/r1" Los An g e 1 C 8
Ir.J i-3-O.ts ! tf Riverside
far i-rrurjlno San Bernardino
Sj-j.-t . .-rl.i _jrtr_ ^.»rbJra
£'»£".:-<»>>• pa. ";;:S:r
i:j 1 1 is : vr S.in iii 1. 1 1 o
San LuU OOU;>-> S-m !."'* obispo
S5.:'h;rn Sj-\ Jjdruln
Oxidont/Ozonc
Concentration
5
10
6
7
9
5
6
5
6
Green
Lisa Ef.lr.3
3.1
49.9
15.8
28.2
45.0
0.8
3.1
15.8
3.1
0.8
15.8
Erocc
0
0
0
0
0
0
0
C
0
Q
0
ill Csn.tnlopos
Q
0
C-
0
0
0
0
0
0
0
0
Carrot <
0
0
0
0
Q
C
I;
0
0
0
0
•.U"f
-------
Table 4.6
Percentage Yield Reduction, for 12 Hour Exposure,
Using the "C" Level of Oxiciant/Ozone Concentration for 1976
•
ti'*on
I-.r«r!al
£1 Ccntro
SPU'> f-visE
Pj j-Jvr.3
A:IJ' -.-^3
In-',', o -Oasis
Sj". 5cr:: Jr*l Ino
Sji-.;j X-irI.i
53:1 :'ior'^
Co- .:r •'. r.-.-st
S.i : lr..ii
!iol '. i >', .T
S j:; Luis. OJl spo
CJl *.'...
County
Imperial
Los Anpeles
CrJing*
Kivcrsldc
S^r. SLT r.nrdlno
Ssnu £jrbaro
San D!o>;o
M^ni ercy
S.in Uon 1 co
SR:I Luis Oblri^o
>;cr:i
Tul.irc
Cor.centrtTtlon
3
7
6
6
7
5
6
4
5
4
6
5
•
Creer.
0
28.2
15.8
15.3
2S.2
3.1
15.8
0.8
3.1
0.8
15.8
3.1
0
0
C
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
C
0
0
C rroc
C
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
C
0
0
0
22.6
12.6
12.6
22.6
2. 5
J2.6
0.6
2.5
0.6
12.6
2.5
Crop
Hend
y Le.tuce On. on
0 0
0.1 3.4
0 1.7
0 1.7
0.1 3.4
o :.o
0 1.7
0 0.1
0 1.0
0 0. 1
0 1.7
0 1.0
0
9.4
2.S
2.3
9.4
1.1
2.8
o.:
1 . 1
0. 1
2.8
1.1
0 9.4 0
9.4 - 2.8
: . 8 - j . 4
2.8 18.7 1.4
S.4 18.7 2.8
1.1 - O.S
2.8 - 1.4
C.I - 0.1
1.1 - O.S
0.1 - 0.1
2.8 6.9 1.4
1.1 6.9 0.8
So',*: 1h\* "3j\('.\M Station It alau used fpr SjntJ Cry;
-------
Table 4.7
Percentage Yield Reduction Averaged Over County,
for each Region, by Time Period
Crop
Vegetable Crop
Beans
Broccoli
Cantalopes
Carrots
Cauliflower
Celery
\Lettuce
Onion, Fresh
Onion, Process
Potato
Tomato, Fresh
Tomato, Process
Field Crop
Cotton
Sugarbeets
Southern Desert
1972-76
Average
-
0
0
-
0
1.00
1.00
-
1.10
1.10
9.40
0.80
1976
-
0
0
-
0
0
0
-
0
0
9.40
0
South Coast
1972-76
Averspe
22.66
0
0
0
0
18.11
0.27
6.80
6.80
11.24
11.24
11.24
18.70
5.66
1976
15.71
0
0
0
0
12.57
0.03
1.99
1.99
4.20
4.20
4.20
18.70
1.63
Central Coast
1972-76 j
Average ! 1976
1.57
0
-
0
0
1.23
0
0.40
0.40
0.43
0.43
0.43
0.33
1.57
0
-
0
0
1.23
0
0.40
0.40
0.43
0.43
0.43
0.33
Southern
San Joaquin
1972-76
Average
9.45
-
0
0
-
0
-
1.35
1.95
1.95
1.95
6.90
1.10
1976
9.45
-
0
0
_
-
0
-
1.35
1.95
1.95
1.95
6.90
1.10
Total
1972-76
Average
11.23
0
0
0
0
9.67
0.068
3.60
2.387
k.Stt
3.68
3.68
11.67
1.97
1976
8.91
0
0
0
0
6.90
0.01
0.60
0.94
2.19
1.65
1.65
11.67
0.77
-------
Table 4.8
Actual Yield Per Acre (in the Presence o£ Air Pollution)
Crop
Vegetable Crops
Beans, Creer. Lima
Broccoli
Cnntalopes
Carrots
C.iul 1 f 1 oi;er
CeK-ry
Lotcuci.1, Head
0:iior:, Crocn
Onion, Dehydrated
Point ooy
To;n;:LO, Fresh
Tonui to, j'roccss
Flo Id Crops
Cut ron
Sui^jrbctt s
Unit
Tons
CVT
CWT
CV.T
CUT
CWT
CKT
CUT
CV.T
CUT
CUT
Tons
I.bs
Tons
Southern Desert
1972-76
Avo rage
-
-
123.57
333.87
-
-
250.65
27/,.37
273.0-'.
-
217.72
21.66
1,144.93
25.53
1976
-
_
127.46
402.00
_
-
266.97
208.94
324.32
-
217.44
25.17
996.48
25.45
Southern Coast
1972-76
Average
2.16
31.62
139. S5
212.12
127. GS
568.65
253.57
322.54
3! 8. 66
3 1 0 . A 7
469.96
24.83
1,019.12
27.86
19.76
2.04
S3. 72
150.42
257.30
114.02
546.58
261. 37
291. 3]
350.00
3! 5. 77
505.89
20.34
1,084.84
28.47
Central Coast
1972-76
Average .
2.23
54.09
-
292.03
99.28
56 J .88
265. 14
302.55
335.05
327. 12
307.67
25. 52
-
32.87
1976
2.52
60.67
-
303.12
97.68
549.73
279.14
285.45
314.61
326.46
201.29
19.89
-
35.55
Southern San Joaquin
1972-76
Average
2.96
-
.183.12
341.10
-
-
259.95
-
340.03
275.34
288.12
20.29
1,026.16
26.50
1
1976
3.00
-
180.00
350.00
-
-
292.16
-
396.92
295.11
199.45
24.53
1,033. 10
28.42
Study Region
1972-76
Average
2.29
57. SI
145.12
321.86
108.66
565.94
258.73
300.05
323.70
2£8. 50
372.81
22.71
1,039.19
27.08
1976
2.35
64.12
141.72
318.87
103.43
547.87
273.46
25S.26
368.70
301 . 78
370.18
21.64
1,075.95
28.44
-------
Table 4.9
Potential Yield Per Acre (without Air Pollution Effects)
Crop
Vegetable Crops
Beans, Green Limn
Broccol 1
Can :a lopes
Carrots
Caul iflower
Ceiury
Lettuce, Head
Onion, Green
Onion, Dehydrated
!'o L j t o c s
Tu:;:a i o , Fresh
Tor.;: to, 1'rocess
Field C'. ops
Cot con
Suf1,.irbcers
Unit
Tons
CUT
CUT
CV.T
CUT
CUT
CUT
CWT
CUT
CUT
CUT
Tons
Lbs
Tons
Southern Desert
1972-76
Average
-
-
128.57
333.87
-
-
230.65
277.11
273.77
-
220.11
21.90
1 ,252.55
25. 73
1976
-
-
127.46
402.00
-
-
266.97
203.94
32'.. 32
-
217.44
25.17
1,090.15
25.45
Southern Coast
1972-76
Average
2.65
S] .62
139.85
312.12
127.68
G71.63
254.25
344.47
340. :n
355. 3S
522. 78
27.62
1,209.70
29.44
1976
2.36
83.72
150.42
257.30
114.02
635.28
261.45
297. !1
356. 96
320.03
527.14
21.19
1,287.70
28. 93
Central Coast
1972-76
Average
2.26
54.09
-
292.03
99.28
568.79
265. 14
303. 76
330.3'.)
328.53
308. 'J9
25.63
1976
2.56
60.67
-
303.12
97.68
556.49
279.14
286.59
315.07
327.86
202. 16
19.98
-
32.98
35.67
;
Southern San Joaquln
1972-76
Average
3.24
188.12
341.10
-
-
259.95
-
344.62
280.71
293.74
20.69
1,096.97
26.79
1976
3.28
-
180.00
350.00
-
-
292.16
-
402.23
300.86
203.34
25.01
1,163.18
28.73
Study Region
1972-76
Average
2.55
57.81
145. 12
321.86
10S.66
620.67
258.91
310.85
331.43
301.60
386.53
23.55.:
1,160.46
27.61
1976
2.56
64.12
141.72
318.87
103.43
585.67
273.49
259.81
370.91
308.39
376.29
;, . 22.00
1,201.51
28.66
Southern Dosert includes Imperial County
Southern Coast includes Los Angeles, Orange, JUverslde, San Bernardino, Santa Barbara, San Diego, and Ventura Counties
Centra! Coast Includes Monterey, Sap. Eenlto, San Luis Obl.spo, and S.irUa Cruz Counties
Southorn San Joaqutr. Includes Kern and Tulare Counties
Sources: County Agricultural CormLisioner Annual Crop Reports
-------
effects in the study area. Table 4.9 thus represents the potential or hy-
pothetical yield that could be realized if the negative effects of air pol-
lution were removed from the crop environment.
57
-------
FOOTNOTES: CHAPTER IV
— The counties in each region as well as included crops in this
study are discussed in Chapter I.
21
~ See the explanation on these levels at the bottom of Table 4.4.
4/
~ El Centre, the monitoring station for Imperial County, typically
has approximately one-half the ozone level observed at Indio-Oasis. Hence,
the one-half value for cumulative doses.
58
-------
CHAPTER V
PRICE-FORECASTING EQUATION ESTIMATION^
As discussed in Chapter I, price fluctuations observed in most agri-
cultural crops depend on a wide range of economic and physical factors,
such as climatological conditions, which may affect per-unit and aggregate
yield of a crop in a particular year and thus translate into subsequent
changes in crop price. For most agricultural crops supply is inelastic in
the short run. In other words, changes in crop price cannot affect the
quantity supplied in the short run, since supply cannot respond immediately
to such changes in price. Furthermore, some agricultural crops (e.g.,
vegetables) are highly perishable; thus, the quantity produced must be sold
immediately after harvest. These characteristics of agricultural production
imply that quantity produced, perhaps more than other factors, determines
the overall level and variation in prices._2/
Due to these characteristics, price cannot reasonably be assumed to be
predetermined for many crops; consequently, a price endogenous structure of
demand is needed to correctly capture the structure of the market. There
are, however, some exceptions; e.g., prices of some vegetable or field crops
are predetermined, as in the case of "contractual" crops or crops affected
by institutional arrangements- such as payments, subsidies and quotas on pro-
duction. Processing tomatoes and market (dehydrated) onions are examples in
the first case (contracts) and sugarbeets the second case (government support
and quota program).
The mathematical model developed in Chapter III features linear demand
functions incorporated into a quadratic objective function, with the intent
of determining prices endogenously. The objective of such a model is to
capture the price effect of air pollution. The purpose of this chapter is
to discuss the estimation procedure and present the statistical results
associated with the derivation of price forecasting equations for the 12
vegetable and 2 field crops included in the study, on a seasonal basis. As
pointed out by Adams (1975) and as discussed earlier, seasonality of produc-
tion for vegetables is particularly important. Given the regional production
patterns observed in California, correct analysis of comparative advantage,
on a regional basis, requires a suitable set of seasonal demand function
estimates.
The following subsection describes the procedure for estimating general
price forecasting equations. The actual results of price forecasting equa-
tions for the 12 vegetable and 2 field crops are then presented. A brief
summary of the overall estimation will then be given in the concluding sub-
section.
59
-------
5.1 Price Forecasting Equation Estimation Procedure
The concept of a price forecasting equation was discussed in Chapter
III with respect to a general formulation. In this section, the actual
procedure used in estimating such equations will be described briefly. The
linear demand functions included in the model have the following form:
p = c + dQ (5.1)
where p is a vector (j x 1) of commodity prices, c is a vector (j x 1) of
constants, d is a negative diagonal matrix (j x j) of price-quantity slope
coefficients and Q is a vector (j x 1) of agricultural crop production.
The above equational form assumes that price of a particular crop is affected
only by its quantity supplied, i.e., a diagonal "d" matrix implies zero
cross-effects for competing commodities.
Consider the following functional specification of a price endogenous
demand equation:
PC = f(Q^, Qr, S., Y) (5.2)
J J J J
Q
where P. = annual seasonal average price received by farmers in California
for commodity j.
c
Q. = seasonal production of commodity j in California.
;£•
Q. = seasonal production of commodity j, the rest of the United States.
S = existing stocks for commodity j for the United States.
j
Y = U.S. personal aggregate disposable income.
A priori one would expect that quantity produced and existing stocks
would have a negative sign whe.reas income would be positively correlated
with changes in price. That is, an increase in quantity produced of crop j
has a negative effect on its own price regardless of where it is sold, as-
suming the crop is homogeneous. An increase in stock tends to indicate a
reduction in price since stocks tend to be positively correlated with pro-
duction and producers tend to increase the level of stocks (for sale in a
later period when price is higher) during periods of lower price. An in-
crease in income enables one to consume more (if a good is assumed to be
normal) and thus affects the; price. To keep the assessment problem tract-
able, it is assumed that the price of commodity j is not affected by price
or quantity of other commodities, i.e., cross-price effects are zero.
The above formulation was. used for all seasonal demand relationships
for all crops included in the study, except processing tomatoes, cotton and
sugarbeets which cannot be estimated by the same procedure due to suspected
simultaneity. As a result, a single equation estimation would not be valid;
thus, some secondary estimates were used* Ordinary least squares was used
in estimating the coefficients for all the variables in the above equation,
for all the s.tudy crops on a seasonal basis. Once coefficients, are obtained
for the variables, in the price equation, coefficients of all independent
variables (except quantity produced in California) are then used to calculate
60
-------
an "adjusted intercept." This, then, results in a price forecasting equa-
tion featuring an adjusted intercept and the slope coefficient with respect
to California quantity. Results of the estimations, including price-flexi-
bility coefficients, are given in the next section.
3 /
5.2 Price Forecasting Equations for Vegetable and Field Crops-
Vegetables
The seasonal patterns and magnitudes of production for the 12 vegetable
crops included in this study are described in Adams (1975) and King, et. al.
(1978). The period covered in estimating the price forecasting equations
for the 12 vegetable crops in this study is from 1955 to 1976, using data
from Adams (1975) for the period 1955 to 1972. There is a problem attendant
to quantifying seasonal production for these 12 crops in California after
1972 due to changes in seasonal patterns as reported by the U.S. Department
of Agriculture, i.e., the twelve reporting seasons used in the earlier per-
iod were collapsed into four. As a result, this required disaggregating
some seasonal estimates for the period 1973 to 1976 into the more numerous
seasonal classification employed in the earlier time period. Such adjust-
ments were made for the period 1973 to 1976 to ensure consistency with data
from 1955 to 1972. The adjustments, by season, are given in Appendix Table
A. The net result is the estimation of 28 equations for the 12 vegetable
crops. These estimated equations will be presented below in order of impor-
tance, as measured by gross income received in 1976.
1. Lettuce. Lettuce contributes the second highest gross income to
California growers (behind tomatoes—fresh and processing), with a total
gross value of $327.7 million in 1976. This value is almost 70% of the total
revenue from U.S. lettuce production. The leading counties are Monterey,
San Benito, San Luis Obispo and Santa Cruz in the Central Coast, and Santa
Barbara in the South Coast for summer lettuce, spring and fall lettuce.
Winter lettuce is produced mostly in Imperial and Riverside counties.
Imperial County also dominates production of fall lettuce. The nature and
marketing patterns of this and other crops are more completely described in
Adams (1975).
Following Adams (1975), the. four seasonal price forecasting equations
for lettuce xoere estimated and presented in Table 5.1. Results of the esti-
mation were not totally satisfactory, even though the signs of all variables
except that of "other production" in the winter lettuce were as expected.
The estimated coefficients of all variables in the winter lettuce are sta-
tistically insignificant (5%) and test of autocorrelation among error terms
is. inconclusive at 5% levels of significance in all but one equation. Com-
paring the: results obtained with those in Adams (1975) shows that the coef-
ficients of determination (R2) and the price flexibility coefficient with
respect to California production are higher in all equations of the same
seasons. However, as is true in Adams (1975), the estimated California pro-
duction slope coefficient in this study is higher than that associated with
"other production" in the same season except for fall lettuce. This result
tends to suggest that lettuce sold in California vis-a-vis "other" U.S. pro-
duction is not homogeneous. Evidence from other researchers (Johnston and
61
-------
Table 5.1
Prite-Forecasting Equations for Lettuce and Fresh Tomato, By Seasoiic
Crop/
Season
Lettuce
Winter
Early Spring
Summer
Fall
Tomato, Fresh
Early Spring
1
b
Constant
2.12
5.67
6.60
2.71
0.30
Estimated Coefficient with Respect To:
t
California
Production
(1000 cwt.)
-0.59E-3
(0.48)
-1.27E-3
(-2.27)
-0.84E-3
(-2.59)
-0.50E-3
(-1.54)
-5.49E-3
; (-0.82)
Early Summer -3.29 -1.07E-3
(-1.01)
i
"Other"
Production j Stock
(1000 cwt.)
0.20E-3
(0.63)
-0.47E-3
(-1.19)
-0.31E-3
(-0.95)
-0.82E-3
(-2.99)
0.47E-3
(0.30)
2.34E-3
_
_
_
-
-
-
Personal
Aggregated
Disposable
Income
Average
California
_ _ . . Production
Summary Statistics ln-,-, -,e
„ ' i 1? 1 i / 0
R : D.W. (Actual)
:
($ billion) ! (1000 cwt.)
2.78E-3 i
(0.67)
10.00E-3
0.54
0.52
(3.22)
10.11E-3 j 0.75
(5.24) j
11.90E-3 i 0.79
(4.7D ;
19.89E-3 ,' 0.70
(4.83) :
18.76E-3 0.93
(2.95) j 1 (6.44) i
Early Fall 7.10 -1.27E-3 - \ -
i
\
2.86e 11903
2.55s 6953
2.02d 10580
1.506 7617
2.45e 378
1.89d 3887
Price
Flexibility
With
Respect to
California
Production
1972-76
c
-1.50
-1.30
-0.55
c
-0.19
1
14.09E-3 0.93 ; 2.46e 2529 j -0.18
(-1.23) i ' (7.65)
\ \
! i i __
3 Data cover period for 1955 to 1976 crop year with quantity produced expressed in units of 1000 hundredweight (cwt.)
and price on actual dollars per cwt. Personal aggregate disposable income (in billion dollars) is for the fiscal
year. Numbers in parentheses are estimated t-stacistics.
Dollar per cwt.
Not available due to statistically insignificant and/or wrong expected sign for the estimated coefficient.
No autocorrelation among error terms at 5% levels of significance.
Test of autocorrelation among error terms is inconclusive at 5% levels of significance.
-------
Dean, 1969; Zusman, 1962) indicates that fresh vegetables produced in Cali-
fornia have somewhat higher quality compared to that produced elsewhere;
hence, it may not be unreasonable to expect a divergence across such coef-
ficients .
2. Processing tomatoes. Processing tomatoes in California have a
gross value of $284.7 million in 1976. This value is about 75% of the na-
tional total. The processing tomatoes industry is one of the most rapidly
growing subsectors in California agriculture over the last two decades.
Several factors such as a favorable climate, advances in production techno-
logy, harvesting systems and a progressive canning industry attribute to
such growth. Major production areas are Solano, Sutter and Yolo counties
in the Sacramento Valley; and Fresno and San Joaquin counties in the San
Joaquin Valley. Total state production in 1976 exceeded 230,000 acres, down
from almost 300,000 acres in 1975. This reduction in production is partially
attributable to drought conditions in 1976.
It is more difficult to estimate a reasonable price forecasting equation
for processing tomatoes, given that processing tomatoes are generally grown
under contract between growers and processors. Prices are usually deter-
mined prior to planting based on several factors, most important being the
carryover of tomato products and the existing market situation, characteris-
tics which suggest simultaneity. Moreover, the estimation of a price fore-
casting equation for such a crop is further complicated by the fact that
processing tomatoes are marketed in various forms such as catsup, juice,
canned whole tomatoes, paste and puree, and other concentrated products.
Each form does not have the same price flexibility coefficient, as is evi-
dence from the secondary information presented in Table 5.2.
Given these problems, it was decided that the values given in Adams
(1975), derived via a weighting procedure of flexibilities presented in
Table 5.2, will be used for the price-forecasting equation for processing
tomatoes in this study.
3. Fresh market tomatoes-. Gross income for California fresh tomatoes
in 1976 exceeded $137 million, 32.4% of the national total. Early spring
fresh tomatoes are produced mostly in Imperial and San Diego counties. Early
summer tomatoes come almost exclusively from the Central Coast (Monterey
County), San Joaquin Valley (Fresno, Merced, San Joaquin, Stanislaus and
Tulare Counties) and the South Coast (San Diego County). San Diego and Ven-
tura Counties in the South Coast are the main suppliers of early fall fresh
tomatoes in California. California fresh tomato production has to compete
with other major production regions such as Florida, Texas, New York, Mich-
igan and Mexico.
The. estimated price forecasting equations for fresh tomatoes are given
in Table 5.1. The sign attached to the coefficient on early spring Califor-
nia production was not consistent with expectations, i.e., it had a positive
sign. In such, a case, the coefficient had to be reestimated by using a
weighting procedure, utilizing the price flexibilities for other seasons
weighted by th.e volume of production from 1972 to 1976. The estimated coef-
ficients of "other production" have positive signs, perhaps due to the con-
founding effects- of California production. From the table, it is evident
63
-------
Table 5.2
Estimated Price Flexibility for California
Processing Tomatoes, 1948-1971a
Product
Canned whole
Juice
Catsup and Chile
Puree
Paste and other
Total
Weighted average
Price
Flexibility
Coefficient
-0.33
-0.23
-0.33
-0.10
-0.28
-
-0.277
California total Shipments
of the Processing Tomatoes,
1975b, (Thousand Tons)
566
290
369
333
1,979
3,537
Total shipments = beginning stocks plus pack minus ending stocks.
Source: King, Jesse and French (1973), and Adams (1975).
Brandt, French and Jesse (1978).
64
-------
that income was the most significant explanatory variable. The price flex-
ibility with respect to California production obtained in this study is of
the same magnitude of that obtained by Shuffett (1954) and Adams (1975).
4. Potatoes. Although California's current potato production is only
about 9% of the national total, it contributed more than $110 million to to-
toal state gross income in 1976. Kern County supplies most of the California
winter and spring potatoes, whereas Riverside County is the major producer
of summer potatoes. Fall potatoes are produced mostly in the Central Coast
and Siskiyou and Modoc Counties in extreme northern California.
Potatoes are marketed in either fresh and/or processed forms; thus, in
estimating the price forecasting equations stock is also included as an ex-
planatory variable. Results obtained are presented in Table 5.3.
From Table 5.3 it is evident that most of the estimated equations are
somewhat disappointing with respect to statistical robustness although the
estimated coefficients attached to the California production have the ex-
pected signs. A divergence of sign is noticed on the disposable income var-
iable for winter and early summer potatoes. One would expect that an in-
crease in personal income will tend to reduce potato consumption and thus
depress price since potatoes are usually assumed to be an inferior good.
The estimated price flexibility coefficients are somewhat lower than
those estimated by Adams (1975). However, the coefficients of determination
in all equations are higher than those of Adams'.
5. Celery. California celery production in 1976 constituted about 66%
of the total U.S. production. The gross income in that year is $78.9 million
which is about 60% of the U.S. value of celery production. Of the four mar-
keting periods, Ventura County supplies most of the winter and spring celery.
Monterey County, on the other hand., produces most of the early summer and
late fall celery. Nationally, California celery faces some competition from
other states such as Florida (for winter celery) and Michigan and New York
(for early summer celery).
Celery is highly perishable and is marketed only in its fresh form.
Thus, in estimating the price forecasting equation only three explanatory
variables were used. These variables are California production, "other
production," and personal aggregate disposable income. The estimated re-
sults are presented in Table 5.4.
As is evident from the table, all the estimated coefficients have the
right expected signs and most are statistically significant. Income is the
most important variable in explaining the variation of price. Only one
equation has an inconclusive test of autocorrelation whereas the rest indi-
cate no autocorrelation among error terras. In terms of competition from
other states, the magnitude of the estimated coefficient of production from
other- areas is higher than that of California for spring celery and vice
versa for winter celery. This tends to suggest that cet. par, production
outside California has an influence on the price of celery sold in California
in the spring season but not in the winter market. The magnitude of the
65
-------
Table 5.3
casL'inj?. Equations f°r PoL'.iroc.1;, By Sc-ason1"
L
Crop/Season Constant0
Esti
California
Production
1(1000 cwt.)
Potatoes |
Winter -0.49
-0.85E-3
(-i.'J'J)
Late Spring 2.79 -0.30E-3
(-1.89)
Early Suir.mer 8.56 . -1.29E-3
(-1.63)
Late Summer 7.27 -0.15E-3
(-0.35)
Fall 4.14 i -0.04E-3
(-0.33)
nated Coefficient
"Other"
Production
(1000 cwt.)
0.31K-3
(0.82)
0.26E-3
(1.60)
-0.34E-3
(-2.65)
-0.15H-3
(-2.26)
-0.03E-3
(-1.90)
with Resoect
Stock
As at
Dec. 1
(1000 cwt.)
0.06E-3
(3.03)
0.02Z-3
(0.71)
0.02E-3
(1-01)
„
_
to:
Personal
Aggregated
Disposable
Income
($ billion)
-4.62K-3
(-1.61)
0.22E-3
(0.07)
Statistics
^
R D.W.
Ac cual
Average
1 California
Production
1972-76
i (1000 cwt.)
I
0.71 1.49
0.62 1.723
-4.38E-3 i 0.65 2.4Sd
(-1.09)
0.06E-3
(0.03)
7.38E-3
(4.09)
0.66 1.69d
10S2
12066
894
1761
|
0.77 1.30e 6574
i
i
Price
Flexibility
vith
KGSpGCL CO
California
Production
for
1972-76
-0.18
-0.69
-0.23
-0.05
-0.05
(continued)
-------
Table 5.3
(continued)
Data cover period from 1955 to 1976 crop year with quantity produced expressed in units of 1000 hundred-
weight (cwt.) and price in actual dollars per cwt. Stock is in units of 1000 pounds. Personal aggre-
gate disposable income (in billion dollars) is for the fiscal year. Numbers in parentheses are estimated
t-statistics.
b
Dollars per cwt.
Q
No autocorrelation among error terms at 5% levels of significance.
Test of autocorrelation among error terms is inconclusive at 5% levels of significance.
-------
Table 5.4
Price-Forecasting Equations for Celery, Cantalopes and Broccoli, By Season'
Estimated Coefficient with Resoect to:
b
Crop/Season Constant
Celery:
Winter 6.19
Spring 10.70
00 Early Summer 3.29
Lace Fall 6.35
Cantaloupes:
Spring 6.58
Summer 6.53
Broccoli:
Early Spring 5.32
Fall 4.68
California
Production
(1000 cwt.)
-1.35E-3
(-2.22)
-1.76E-3
(-2.49)
-0.62E-3
(-0.71)
-1.62E-3
(-1.88)
-1.63E-3
(-2.49)
-0.54E-3
(-2.69)
-0.72E-3
(-0.76)
-2.97E-3
(-1.73)
"Other"
Production
(1000 cwt.)
-0.35E-3
(-0.57)
-2.89E-3
(-3.41)
-
_
-0.77E-3
(-1.61)
-0.52E-3
(-1.27)
-
1.76E-3
(0.60)
Price
Flexibility
With
Average 5
Frozen Personal California (
Stock Aggregated Production I
A- ir ni - nn « -i K 1 r- StfltlStlCS T r, -i-, -if
t\:^ oU L'.ibpUociL'JLt; 7,
Dec. 31 Income R
(1000 Ibs.) ($ billion)
4.53E-3 0.68
(5.24)
4.18E-3 0.67
(5-35)
4.05E-3 0.65
(3.81)
6.42E-3 0.69
(6.15)
7.83E-3 0.89
(7.82)
5.73E-3 0.90
(5.78)
-0.02E-3 12.28E-3 0.93
(-1.92) (.6.80)
-0.02E-3 17.03E-3 0.96
(-1.65) (9.13)
J-3 I £- 1 \J
D.W. (Actual)
(1000 cwt.)
2.618 2459
1.83d 2421
2.11d 1961
1.96d 3667
2.20d 1197
2,56e 5870
l,20e 2000
d
2.14 1615
lespect to
California
'reduction
for
1972-76
-0.48
-0.69
-0.20
-0.88
-0.18
-0.40
-0.11
-0.34
(continued)
-------
Table 5.4
(continued)
Data cover period from 1955 to 1976 crop year with quantity produced expressed in units of
1000 hundredweight (cwt) and price in actual dollars per cwt. Stock is in units of 1000 Ibs.
Personal aggregate disposable income (in billion dollars) is for the fiscal year. Numbers in
parentheses are estimated t-statistics.
Dollars per cwt.
No autocorrelation among error terms at 5% levels of significance.
d
Test of autocorrelation among error terms in inconclusive at 5% levels of significance.
-------
price flexibility coefficients obtained in this study are similar to those
obtained by Adams (1975).
6. Cantaloupes. California produces about two-thirds of the total
cantaloupes produced in the United States. In 1976, gross income from can-
taloupes in California amounted to about $70.4 million (65.2% of the U.S.).
Prior to 1972, cantaloupes were marketed in two seasons: spring and summer.
After 1972, three seasons were recognized with the third season being fall.
Imperial County is the leading production area for spring and fall canta-
loupes, whereas Fresno and Kern Counties supply most of the California sum-
mer cantaloupes. Of the three seasons in the present system, summer season
accounts for more than 75% of annual production. California cantaloupes
face strong competition from other areas such as Texas and Mexico, especially
for the summer market. Disease and labor problems and a decline in the
price of cantaloupes relative to other less labor-intensive commodities
caused a sharp reduction in the spring crop over the past decade [Adams,
1975, p.
Since cantaloupes are highly perishable and are marketed only in fresh
form, the formulated price forecasting equations for this crop consist only
of three explanatory variables. The estimated results are presented in
Table 5.4.
The estimated coefficients for the explanatory variables in all equa-
tions have the right expected signs and are statistically significant at not
less than 10% levels of significance (except the coefficient for "other pro-
duction" in summer cantaloupes). Income is significant and the coefficients
of determination are quite high. The price flexibility coefficient is con-
sistent with that obtained by Adams (1975).
7. Broccoli. California produces about 97% of total U.S. broccoli
production. Gross income from broccoli production in 1976 was $65.6 million
(99% of the U.S.). Broccoli is marketed in two forms: fresh and frozen.
Fresh market broccoli was previously reported for two market seasons, early
spring and fall. After 1972, however, the market had been broadened to four
seasons: winter, spring, summer, and fall. Monterey and Santa Barbara
Counties are the main production areas for broccoli in California.
The estimated price forecasting equations for broccoli are given in
Table 5.4. All but one variable had the expected signs, the exception being
the estimated coefficient for "other production," which is also statistically
insignificant. Once again, income is the most important explanatory variable
in explaining the variations in price of broccoli. The price flexibility
coefficients obtained in this study again are similar to those obtained by
Adams (1975).
8. Carrots. The average production of carrots in California over the
last 5 years represents about 50% of the national total. In 1976, California's
market share of carrots was 50.3% with a gross income of $58.3 million (49.6%
of the U.S.). Winter carrots are produced mostly in Riverside and Kern
Counties, whereas Monterey, Kern and Imperial Counties supply most of the
early summer carrots. Monterey, Kern and Riverside Counties are also impor-
tant producers of late fall carrots.
70
-------
Since carrots are marketed in both fresh and frozen forms, the frozen
pack is included in the price forecasting equation estimations. The esti-
mated results are presented in Table 5.5.
Of the three estimated equations, winter carrots have the wrong ex-
pected sign on the stock variable. The magnitude of the price flexibility
coefficient obtained in this study displays a wider range of values than
those obtained by Adams (1975).
9. Cauliflower. California is a major producer of cauliflower, sup-
plying about 80% of the national total in 1976. The gross income from cau-
liflower production in that year exceeded $50 million (76.8% of the U.S.).
Cauliflower is marketed in fresh and frozen forms. Frozen pack accounts
for about 36.5% of the total production and 19% of the gross income from
California cauliflower production in 1976. Early spring cauliflower is pro-
duced mostly in Alameda and Monterey Counties. Kern, Monterey and Santa
Barbara Counties are main producers of late fall cauliflower.
The fact that California cauliflower production faces little signifi-
cant competition in any season frcm other sources resulted in only three
variables being included in the equation; California production, frozen pack
and aggregate income. The estimated equations are given in Table 5.5.
The estimated equations obtained do not have the expected signs for all
variables. Most significantly, the estimated coefficient attached to the
California production of late fall cauliflower has the wrong expected sign.
The slope, coefficient for this variable was reestimated by using the price
flexibility coefficient for early spring production, adjusted to fall quan-
tities and prices.
1Q. Processing onions. California produces the bulk of the supply of
processing (dehydrated) onions in the U.S., due to the state's long growing
season. Processing onions in California are marketed in summer (late).
Total production in 1976 was 7.2 million hundredweight, with a gross income
cf $27.5 million. Kern, Fresno, Riverside and Monterey Counties are the
main producers of processing onions.
Processing onions are grown mostly under contract to specific proces-
sors. These institutional arrangements influence the fluctuations in price
and thus the causality of price-quantity relationship; hence, a single equa-
tion estimation may not be appropriate. In estimating the price forecasting
equation for processing onions, four explanatory variables are included in
the model. Results obtained, shown in Table 5.5, are not entirely satis-
factory, given that the. estimated coefficients are either statistically in-
significant (10%) or do not have the right expected signs. This tends to
confirm the hypothesis stated above. Lack of alternative estimates from
more detailed econometric analysts mandated the use of this equation, as
estimated.
11. Fresh market onions. California fresh onion production contri-
buted only about 23.0% in volume and 17.6% in value to the national totals
in 1976. The other states that produce late spring (or spring) onions are
71
-------
Table 5.5
Price-Forecasting Equations for Carrots, Cauliflower, Onions and Beans, by Season
Crop/Season Constant
Carrots:
Winter 7.71
Early Summer 3.10
Late Fall 2.63
Cauliflower:
Early Spring 5.64
Late Fall 3.38
Onions:
Late Spring 3.84
Late Summer -1.04
Processing
Green Lima
Beans 69.61
Estimated Coefficient with Respect to:
California
Production
•1000 cwt.)
-1.48E-3
(-2.13)
-0.15E-3
(-0.21)
-0.18E-3
(-0.39)
-6.40E-3
(-2.43)
-2.40E-3
(1.69)
-0.60E-3
C-0.29)
-0.01E-3
(-0.03)
-0.15E-3
(-0.08)
"Other"
Production
(1000 cwt.)
-0.54E-3
(-1.91)
_
_
_
_
-0.14E-3
C-0.21)
0.13E-3
(1.40)
-1.40E-3
(-1.20)
"Frozen
Stock
As at
Dec. 1
(1000 Ibs)
0.01E-3
(0.77)
-0.01E-3
(-1-39)
-0.02E-3
(-2.42)
-0.03E-3
(-1.19)
-0.07E-3
(-4.28)
-0.33E-3
(-0.29)
0.12E-3
(0.49)
13.61E-3
(0.79)
Personal
Aggregated
Disposable
Income
($ billion)
2.02E-3
(1. 12)
5.54E-3
(2.27)
7.85E-3
(5.00)
18.47E-3
(9.75)
10.91E-3
(9.31)
6.23E-3
(1.46)
1.77E-3
(1.21)
218.35E-3
Summary
Statistics
R2 D.W.
0.56 2.01d
0.47 2.28e
0.68 1.59e
0.93 1.22e
0.96 1.21e
0.36 2.63e
0.71 1.446
0.91 1.52e
(10.42)
1
Average
California
Production
1972-76
(Actual)
(1000 cwt.)
3438
4072
3501
792
1594
1788
7555
42930
Price
Flexibility
with
Respect to
California
Production
for
1972-76
-0.83
-0.10
-0.10
-0.30
c
-0.14
-0.01
-0.02
(continued)
-------
Table 5.5
(continued)
Data cover period from 1955 to 1976 crop year with quantity produced expressed in units of
1000 hundredweight (cwt.) except for processing green lima beans which is in tons. Prices are in
actual dollars per cwt. except for processing green lima beans which are in dollars per ton.
Frozen stock is in 1000 Ibs. except processing green lima beans which is in tons. Stock for
onion is expressed as stock in storage, January 1, in 1000 cwt. Personal aggregate disposable
income (in billion dollars) is for the fiscal year. Numbers in parentheses are estimated
t-statistics.
Dollars per cwt. except for processing green lima beans which is in dollars per ton.
Not applicable due to either insignificant and/or wrong expected sign of the estimated
coefficient.
d
No autocorrelation among error terms at 5% levels of significance.
£>
Test of autocorrelation during error terms is inconclusive at 5% levels of significance.
-------
Texas (66.8%) and Arizona (1Q.2%). Gross income from California fresh
onion production in 1976 amounted to $7.8 million. San Joaquin and Imperial
Counties are the leading counties for spring onion production, with Kern and
Fresno Counties supplying the remainder of the production.
The variables estimated in the price forecasting equation for late
spring onions, shown in Table 5.5, are not statistically significant at the
10% level of significance except for personal aggregate disposable income,
although the estimated coefficients of all variables have the right expected
signs. The test of autocorrelation among error terms is inconclusive at the
5% level of significance.
12. Processing Green Lima Beans. Processing green lima bean produc-
tion in California currently is about 45% of the national total. In 1976,
California produced 25,750 tons at a gross income, of $8.3 million (52% of
the U.S. value). Processing green lima beans in California includes two
varieties, Fordhooks and baby limas. Leading producing counties for pro-
cessing green lima beans are Ventura and Stanislaus.
In estimating the price forecasting equation for processing green lima
beans, four explanatory variables, were used. They were production in Cali-
fornia, production elsewhere, frozen pack and personal aggregate disposable
income. Results of the estimation are given in Table 5.5.
It is somewhat surprising that although. California's share of proces-
sing green lima beans represents about 45% of the national total, the esti-
mated coefficient for California production is significantly smaller than
that of "other production." This might be due to the fact that about 50% of
annual production of processing green lima beans in California are used as
dry edible beans, implying a somewhat different demand structure. Only the
estimated coefficient for personal aggregate disposable income is statisti-
cally significant at the 10% level. The test of autocorrelation among error
term ia inconclusive at the 5% level of significance.
Field Crops
As mentioned in the introductory subsection of this chapter, the mar-
ket structure of some agricultural crops may not be adequately represented
by a single equation model due to institutional arrangements and other fac-
tors. Thus, the estimation of price forecasting equations- for these crops
is more unwieldly than vegetable crops, requiring a multiple equation econ-
ometric model. The two field crops included in this study are examples of
these types of crops. Cotton prices were usually muted by government inter-
vention, whereas sugarbeet prices were affected by a combination of proces-
sor capacity scheduling, government quotas, payments and subsidies [Adams,
1975]. Therefore, the specified price forecasting equation estimation for
vegetables discussed above was deemed inappropriate for these two crops.
Consequently, estimates obtained from more detailed econometric sources will
be used in this study.
1. Cotton. Total acreage harvested of cotton in California in 19.76
exceeded 1.1 million acres, yielding about 2.3 million 50.0-lh. bales. Gross
74
-------
income for that year exceeded $835 million, which is about 25.6% of the
total U.S. value. San Joaquin (Fresno, Kern, King and Tulare Counties) and
Imperial Valley are two major cotton producing areas in California. The
average yield per acre for California cotton production currently is about
1,000 pounds of cotton lint. This yield is higher than the U.S. average
(almost twice the U.S. average in 1976). Over the period 1972-1976, Cali-
fornia cotton production averaged about 18.6% of U.S. total production.
California's share in 1976 increased to 23%, due primarily to the higher
yields obtainable under irrigation, the high quality of cotton planted, and
the adaptability of mechanical harvesting systems [Adams, 1975, p. 101].
The price forecasting equation chosen for this study is taken from
Adams (1975) and is given in Table 5.6.
2. Sugarbeets. The production and marketing mechanism for sugar in
the U.S. are discussed in Adams (1975) and elsewhere. Sugarbeet production
in California has increased each year since 1967 with the exception of 1973
and 1974. Total production in 1975 was 8.9 million tons. Gross income
received (including government payments and subsidies) in 1975 exceeded $267
million which is about 46% of the U.S. value (1976 figures were not availa-
ble at the time of this study). Annual yield per acre of Sugarbeets in Cal-
ifornia is higher than the U.S. average (about 40% higher, 1972-1976).
Sugarbeets are grown in 31 counties in California. The leading producing
counties are Imperial, Fresno, Kern and San Joaquin, and Monterey.
The estimated slope coefficient for Sugarbeets used in this study is
also taken from Adams (1975) and is given in Table 5.6.
Summary of Price Forecasting Equations
The estimated price forecasting equations for the 12 vegetable and 2
field crops discussed above are needed to obtain the linear price structure
discussed earlier (see equation 5.1). The slope coefficient for California
production was obtained directly from the equations, except where the signs
were deemed inappropriate. Two procedures for the calculation of the inter-
cept term were employed. The first, identified as "calculated" intercept in
Table 5.6, was derived by adding a value to the constant term which would
ensure that the "actual" price for 1976 would be predicted when 1976 quanti-
ties were used in the price forecasting equation. The second procedure re-
sulted in the obtaining of an "adjusted" intercept. The "adjusted" inter-
cept term reported in Table 5.6 is derived by adding to the estimated con-
stant term all explanatory variables (at mean and 1976 levels) except Cali-
fornia production. Additionally, price flexibility coefficients vere esti-
mated with respect to California production as a means of establishing gen-
eral credibility of the slope coefficients and as a point of comparison with
other studies. A summary of the various intercept calculations and the
price flexibility coefficients for each crcp and season are presented in
Table 5.6. For the purposes of calculating "price effects" of air pollution,
those equations employing the "adjusted" intercept were used.
75
-------
Table 5.6
Summary of Price Forecasting Equations
Intercept Tens
Calculated
Crop/Season (1976)
Vegetable Crops
Processing Green
Litza Beans
Broccoli:
Early Spring
Foil
Cantaloupes:
Spring
SuBiaer
Carrots:
Winter
Eorly Suraer
Late Fall
Cauliflower:
Early Spring
Late Fall
Celery:
Winter
Spring
Early Sunnier
Late Tall
Lettuce:
Winter
Early Spring
Sunnner
Fall
Onions:
Late Spring
Late Summer
Potatoes:
Winter
Late Spring
r^iriw Rti-mer
into Surr»cr
Fall
Tomato, Fresh:
Early Spring
Early Stairjner
Early Fall
Tomato, Processing
Field Crops:
Cotton
Su£ar beets
and be.lns, vhich are
Units in the slope of
vhtch ar« in Billion
326.97
16.57
22.64
14.40
12.62
9.05
6.25
9.48
25.91
12.04
10.53
10.88
7.56
14.00
5.98
16.55
19.68
14.01
5.71
4.00
6.86
8.64
5.23
4.13
4.79
20.29
29.60
26.34
: 68.00
70.17
32.46
dollars per
coeffirtcni
tons, beans
Slope
CocI f Icient
u ( * ;.
AJItisted F
(1955-76) C
(1976) Hear, £
333.29 215.20
15.65
20.85
14.62
12.40
9.22
7.94
6.32
25.51
11.57
10.83
11.43
8.09
13.97
6.36
16.72
17.75
12.57
8.97
4.27
6.50
9.95
5.12
5.27
4.00
26.04
29.41
23.81
-
-
-
dollars per hu
ton. The unit
ts
9.30
11.39
9.16
8.46
7.20
5.11
4.80
14.56
5.72
7.86
8.59
5.61
10.04
4.57
9.75
11.60
8.00
5.61
2.55
5.04
7.69
5.11
J.49
2.07
13.21
14.72
15.18
-
-
-
tespccC to
:aliforui.-i
'reduction
-0.1543
-0.7247
-2.9696
-1.6286
-0.5355
-1.4781
-0.1467
-0.1808
-6.3986
-2.4036C
-1.3500
-1.7608
-0.6228
-1.6232
-O.S3570
-1.2690
-0.8376
-0.5047
-0.5951
-0.0053
-0,8493
-0.2997
-l.Wi
-0.1512
-0.0377
-5.4866C
-1.0693
-1.2692
-2.4800
-0.0296
-0.2655
indreJu-cftfhts for all
for cotton is cents
Hc.-vn Value for Prli-o Flexibility
Quantity Mvldcd by Coefficient!
He.in Value ft.r Price Uilh K"!^" [°
r«lif. Production
Q/P
1955-76
207.32
138.56
100.51
160.88
1048.61
418.40
563.38
596.55
69.59
124.81
476.57
400.40
319.02
703.35
1877.87
1003.26
1846.05
1137.20
308.02
1958.13
691.72
4712. IS
700.33
870.37
i!73.06
33.62
218.76
293.04
-
.
-
vegetables
per pound.
in thousand tons and cotton In oLIHon SOO-lo.
Mean Value
1972-76
139.66
151.51
115.69
110.22
708.03
561.76
684.37
534.50
47.17
134.46
355.86
389.85
322.53
544.07
1845.43
1184.50
1555.88
1031.96
239.04
2098.61
210.51
2315.93
173. SO
352.20
13S6.92
15.87
181.30
142.88
_
-
1955-76
-0.03
-0,10
-0.30
-0.26
-0.56
-0,62
-O.OS
-0.11
-0.45
d
-0.64
-0.71
-0.20
-1.15
d
-1.27
-1.55
-0.57
-0.18
-0.01
-0.59
-1.41
-0.50
-0.13
-0.03
d
-0.23
-0.37
„
-
1972-76
-0.02
-0.11
-0.34
-0.18
-0.40
-0.83
-0.10
-0.10
-0.30
d
-0.48
-0.69
-0.20
-0.88
d
-1.50
-1.30
-0.55
-0.14
-0.01
-0.1S
-0.69
-0.23
-0.05
-0.05
d
-0.19
-0.18
~
-
except processing tomatoes and
. bales.
processing t
ii
0.91
0.93
0.96
O.S9
0.90
0.56
0,47
0.6.8
0.93
0.96
0.68
0.6S
0.65
0.69
0.53
0.52
0.75
0.79
0.36
0.71
0.71
0.62
0.65
0.66
0.77
0.70
0.93
0.93
-
sugur beets
sugar be-tfts
en*ff fclonr IN Xnr4u»j t f ,i " ""' " "''•"" "' ""• *--'••"""•• ^ **upr coefficient, the incorporated slope
e-oenicunt is derived from other season prlce-flexihlllt 1... for the »a». crop, .c relevant pric. .nd quanciv level..
Hot applicable due to reasons given under Footnote c.
76
-------
Appendix Table A
Seasonal Patterns of Production for Selected Vegetable Crops in California
Period To 1972
Crop Season
Period
Actual
Season
After 1972
Adjustments
Broccoli: Early Spring
Fall
Cantaloupes: Spring
Summer
Carrots: Winter
Early Summer
Late Fall
Caulif lower : Early Spring
Late Fall
Winter
Spring
Summer
Fall
Spring
Summer
Fall
Winter
Spring
Summer
Fall
Winter
Spring
Summer
Winter + Spring
Summer + Fall
Spring
Summer + Fall
Winter (Desert) +
Winter (Other)
Spring + 1/2 (Summer) i
1/2 (Summer) + Fall
Winter + Spring
Summer + Fall
Caulif lower: Early Spring
Late Fall
]
4
'; Celery: Winter
ii
1 Spring
Early Summer
Late Fall
|
Lettuce: Winter
Early Spring
Summer
Fall
Onions: Late Spring
Late Summer
Potatoes: Winter
Late Spring
Early Summer
Late Summer
Fall
Tomatoes, Fresh: Early Spring
Early Summer
Early Fall
Winter
Spring
Summer
Fall
Winter
Spring
Summer
Fall
Winter
Spring
Summer
Fall
Spring
Summer
Winter
Spring
Summer
Fall
Spring
Summer
Fall
Winter + Spring
Summer + Fall
Winter (South Coast) +
Spring (Central Coast)
Spring (South Coast)
Summer (Central Coast)
Fall (South Coast) +
Fall (Central Coast)
Winter + 1/3 (Spring)
2/3 (Spring)
Summer j
Fall
Spring
Summer
Winter
Spring
0.3 (Summer)
0.7 (Summer)
Fall
Spring (Desert)
Spring (Others) +
Summer (Others)
Fall
77
-------
FOOTNOTES: CHAPTER V
~ The material presented in this chapter, including the estimation
procedure, is borrowed from Adams (1975) and King, et. al. (1978). The
interested reader is referred to these references for a more complete
discussion.
2/
— As an example, consider the events of spring lettuce of 1978. During
that period, the retail price of head lettuce throughout the country increased
sharply over prices in the preceding period. This sharp increase was
attributed to the reduction of supply caused by heavy rains in the-Central
Coast region of California, the major source of lettuce supply during spring.
However, within a few months, supply conditions improved, reflected in a
gradual drop in the price of lettuce.
~ It should be emphasized that these estimated equations are for
California, but the regions included in this study only encompass a part
of California. Nevertheless, the included regions together constitute a
major share of production of the study crops in the state.
78
-------
CHAPTER VI
AN ECONOMIC ASSESSMENT OF CROP LOSSES
DUE TO AIR POLLUTION: THE
CON SWUNG SECTOR
As mentioned earlier, past economic assessments of crop losses due to
air pollution were obtained simply by multiplying the estimated reduction
in yield by the respective prices associated with each crop. Such an ap-
proach is not appropriate for most vegetable and specialty crops where
prices may be affected by the reduction in supplies, whether due to air
pollution or other factors. Thus, variations in quantity produced due to
the presence of air pollution may subsequently alter the existing price of
that crop.
This chapter describes a simple procedure used in arriving at an eco-
nomic assessment of crop losses due to air pollution in the study area for
some selected vegetable and field crops. The procedure takes into account
variations in prices due to yield depression and thus the effect on consu-
mers' well-being. Several steps were involved in the procedure yielding
the estimated results presented at the end of this chapter. It should be
emphasized that this procedure is only a "first-step" approach; a more
elegant and detailed analysis of both the consuming and producing sectors
is planned for "Phase 2" as discussed in Chapter VII.
Two levels of production, the annual average from 1972 to 1976 and
that for 1976, were determined by region for each of the included annual
vegetable and field crops. These are presented in Table 1.2 and 1.3 of
Chapter I. These levels of production should reflect the effects of air
pollution (oxidant/ozone concentrations) in those regions observed during
the production periods, given that the values represent actual production.
In the absence of such air pollution, one might expect to observe higher
production, yields, at least for the. more sensitive crops. This "potential"
level of production can be calculated after determining the percentage of
yield reduction due to air pollution for each crop in each region. Such a
degree of yield reduction has been calculated and discussed in Chapter IV
and is presented in Table 4.7 of that chapter. The "potential" levels of
production in the absence of air pollution were then calculated as shown in
Table 6.1 of this chapter.
The next step involved is to calculate the changes in production due
to air pollution. Such changes, by region and by crop, are derived by
79
-------
Table 6.1
Production without Air Pollution
Crop
Vesetablo
Pro. Lina Beans
Broccoli
Unit '
Southern Desert
1972-76
Tons |
Cut.
Cantaloupes Cwt.
Carrots
Cut.
1,199,600
1,703,400
Cauliflower Cvt.
Celery Cvt.
Let tucc
Onion, Fresh
Onion, Process
-
1
Cut.
Cut.
11,124,800
464,990
Cvt. 553,470
Potato : Cvt. :
i
Tomato, Fresh : Cut. : 388,494
Tomato, Process j Tons ' 24,309
Field Crop ;
Cotton ' Bales 136,277
Suj-arhcct Tons 1,610,698
1976
1,128,000
2,215,000
So-th
1972-76
(Average)
28,562
238,178
320,823
3,193,959
! 546,599
-
7,324,125
11,720,000
374,000
,300,000
_
4,503,785
610,745
1,291,212
3,141.204
384,000 4,643,332
36,000 262,500
154,801 44,171
1,476,000 288,835
Coast
1976
16,310
292,770
461,332
2,908,021
617,877
7,292,298
4,951,602
282, S49
1,427,840
Central Coast
1972-76
(Average)
6,434
1,012,180
-
1,402,620
861,370
4,136,810
18,349,364
388,509
565.806
1976
Southern San Joaguin
1972-76, : 1976
(Average)
1 :
2,547
1,207,400
-
1,416,800
975,850
4,585,473
20,535,170
598,973
394,833
i i
3,105,385
1.577.930
5,231,337 I 1,203,516
185,963 j 258,709
60,682 :
1,434,715
875,757
189,810
-
7,390 ; 9,840
i
728,400 i 468,000
3,220,000 i 3,500,000
!
I
1,151,600 ! 1,490,000
i
2,146,983 ! 2,614,820
I
9,795,744 • 10,837,879
687,939 | 411,357
170,196 . 198',830
906,799 1,039,883
260,804 602,149 i 869,991 : 747,334 858,768
00
o
NOTE: Dash indicates no production of that crop in that region.
-------
taking the differences between production with and without air pollution
and are given in Table 6.2. These changes in production were then used to
calculate changes in price. Such changes in price were obtained by using
the price forecasting equations discussed and presented in Chapter V. Sea-
sonal as well as annual price forecasting equations were required, due to
the fact that each region, because of distinct climatic conditions, produces
vegetable crops for different market periods. Appropriate seasonal price
forecasting equations were assigned to each region based on actual marketing
patterns!./ and are given in Table 6.3.
Table 6.4 contains changes in prices due to air pollution by crop and
by region for two periods of time — the average for the period 1972 to
1976 and the 1976 periods. These are the increases in price per unit due
to the reduction of production caused by the adversary effect of air pollu-
tion in that area. Table 6.4 is thus a measure of the overall price effect
due to air pollution. Such price effects were then used to calculate a
measure of consumers' surplus (or compensating variation).^/ Due to the
absence of regional consumption data on the study crops, it is assumed that
production in each year is totally consumed. Such an assumption does not
appear to be unrealistic for most vegetable crops which are highly perish-
able and thus have to be consumed in a relatively short period of time.
However, some vegetable crops are consumed in processing forms and thus
have some carryover stock. Nevertheless, total consumption and total pro-
duction for those crcps in each year should be somewhat consistent. Total
production for each crop in each region was then used to calculate the
compensating variations as given in Table 6.5 (for the mean of 1972-1976)
and Table 6.6 (for 1976).
Results obtained in Table 6.5 show that the most severe economic dam-
age is associated with celery (65.6% of the total crop loss), fresh toma-
toes (16.9%) and potatoes (11.4%). On a regional basis, as expected, the
South Coast region suffers the heaviest crop loss among the study regions,
almost 90%, of total crop loss. Most of the damage in the Southern San
Joaquin Valley is on cotton and potatoes, whereas celery contributes almost
all crop losses in the Central Coast region. Th«; Southern Desert (includes
only Imperial County in this study) shows very minimal crop loss. The to-
tal crop loss per year during 1972 to 1976 is $14.8 million. This loss is
about 1.48% of the total value of production for the included crops in the
four regions and 0.82% of the value of these crops produced in the entire
state.
Table 6.6 shows the total crop loss due to air pollution by crop and
county in 1976. As is true in the case of Table 6.5, celery, fresh toma-
toes and potatoes contribute most of the losses and are followed by cotton
lint. Th«: South Coast and the Southern San Joaquin Valley suffer the most
severe crop losses. Total crop loss in 1976 is $11.1 million (0.9% of the
value of production in the study regions and 0.48% on the state basis).
Note that this total crop loss for 1976 is lower than the crop loss ob-
served for the average of the past five years. This might be due partly
to improvement in the air quality in the study regions, especially in the
Southern Desert region.
81
-------
Table 6.2
h;!nj>,i's in Production Dae to Air Pollution
| Sou them
Crop Unit !1972-76
Vej?.etable i j
Pro. Gr. Lima Beans 1000 Tons ! -
Broccoli ! 1000 Cwt. i
; i
Cantaloupes : 1000 Cwt. j 0
': |
Carrots : 1000 Cwt. ; 0
Cauliflower ' 1000 Cwt. j -
Celery , 1000 Cwt. j -
Lettuce ! 1000 Cwt. ! 0
Onion, Fresh ' 1000 CwL. i 4.590
Onion, Process 1000 Cwt. I 5.470
! 1
Potato 1000 Cwt. 1 -
Tor.ato, Fresh 1000 Cwt. ! 4.094
Toxato, Process 1000 Tons ' .249
;
Field Crop
Cotton j 1000 Bales 111.709
Sugarbeet | 1000 Tons :12.298
;
Desert i South Coast
1976 1972-76 | 1976
. !
|
; 5.306 1 2.223
1 0 j 0
i |
0 '-. 0 i 0
1
0 . 0 | 0
i
1122.973 j 814.198
0 ; 11.968 | 1.472
0 ; 38.883 j 5.521
0 > 82.212 i 27.840
; i
: 317.430 I 125.185
0 ; 469.137 I 210.921
0 ! 26.530 i 7.425
i
: i
i
13.301 . 6.959 j 9.560
0 . 15.531 'j 4.170
Central Coast
1972-76
.088
0
-
1976
.042
0
-
00 0
0
50.230
0
1.549
2.256
0
55.678
0
2.373
1.578
6.770 j 6.115
5.136 3.757
1.120 .830
!
i
Southern San Joaauin
1972-76
.626
_
0
0
-
-
0
-
28.583
187.292
13.171
3.256
1976
.840
_
0
0
-
-
0
-
34.820
206.979
7.877
3.830
!
1
58.531 3 67.123
2.047 ! 2.971 ! 8.060 j 9.130
j :
CO
K3
NOTE: Dash indicates no production of that crop in that region.
Zero indicates no change in production (due to insignificant
effect of air pollution on that crop).
-------
Table 6.3
Seasonal Vegetable Crop Production by Region in California
Crop
Broccoli
Cantaloupes
Carrots
Cauliflower
Celery
Lettuce, head
Onion, fresh
Onion, process.
Potatoes
Tomatoes, fresh
Southern
Desert
—
Spring
Winter
—
Winter
Late Spring
Late Summer
'
Early Spring
Region
i
South Coast
Early Spring
Spring
Late Fall :
Late Fall
Winter
Early Spring
Late Spring
Late Summer
Early Summer
Early Fall
1
I
Central j
Coast |
>
Fall ;
:
Early Summer
Early Spring ]
Late Fall ;
Summer :
Late Spring i
Late Summer
Late Summer •'
Early Summer ;
Southern
San Joaquin
—
Summer
Early Summer
—
Early Spring
—
Late Summer
Late Spring
Early Summer
NOTE: Dash indicates no production in that region.
83
-------
Table 6.4
in Crop Price ))uc Co Air Pollution, 1972-76 and 1976
Crop
Vegetable
Pr. Cr. Lima
Beans
Broccoli
Cantaloupes
Carrots
Cauliflower
Celery
Lettuce
Onion, Fresh
Onion, Proc.
Potato
To~ato, Fresh
Tomato, Proc.
rialci Crop
Cotton
Sugarbeet
Unit
Southern Desert j South Coast
1972-76 1976
• j
! I
I
I
S/Ton i -
$/Cwt.
$/Cwt.
$/Cwt.
$/Cwt.
$/Cwt.
$/Cvt.
$/Cvt.
$/Cvt.
$/Cvt.
$/CwC.
S/Ton
-
0 0
t
0 0
-
1
!
0 0
.002754 i 0
.0000547 ] 0
_ i _
.022476 | 0
.0006175 0
i
$/Lb. i .0003512 ; .000399
1972-76 j 1976
1
Central Coast Southern San Joacuin
1972-76 ; 1976 19
i i
i 1
i
.3187158
0
i i
.3430089 .0135784 .0064806 .0
0
I
0 0
0
0
1.5160135
.0151993
.0233298
.OOOS221
.4094847
.5958039
.0657944
0
0
00-
- - 0
000
.0 o i -
j
1.0991673 .0813726 .0901983
.0018694
0 00
( i
.0033126 .0009294 j .0014238
.00027S4 .0000225 I .0000157 .0
i
i
.1614386
.2678696
.018414
.0010155 . .0009172 .0
.0054955 .0040199 i .0
72-76 1976
1
96591S .129612
-
0
1
0
t ~~
i
0
; -
302S5S ,' .0003482
561876 i .0620937
L40929 .OOS42S3
i
.0027776 j .0020534 .0080743 .0094984
|
I i
i
.0002087 i .0002868 j - - ; .0017559 ! .0020136
S/Ton ' .0033204 0 i .0041933 i .0011259
.0005526 j .0008021 ! .0021762 .0024651
NOTE: Dash indicates no production of that crop in that region.
Zero indicates no changes in price due to no effect from air pollution on that crop in that region.
-------
Table 6.5
Consumers' Surplus at Mean (1972-1976) Consumption
Using the Mean Value (1972-1976) Level of Oxidant Concentration
Crop
Vegetable Crops
Beans, Pro. Gr.
Lima
Broccoli
Cantaloupes
Carrots
Cauliflower
Celery
Lettuce, Head
Onion, Fresh
Onion, Processing
Potato
Tomato, Fresh
Tomato, Processing ;
Field Crops
Cotton, Lint |
Sugarbeets :
Total :
Percent of Total
Southern South
Desert Coast
19,040
! o
0 . 0
0 0
', o
9,401,030
0 68,272
1,268 13,341
30 994
1 1,156,292
8,640 2,487,002
15 15,526
22,000 4,000
5,307 1,146
37,260 13;166,643
0.25 88.70
; Southern Percent
San of Total
Central ! Joaquin : : Consumer
Coast ! Valley Total ! Surplus :
! '\'>
<
V
'.
1
86 ' 653 19,779 0.13 j
0 ; - 00;
0 00,
0 0 00.
0 - CO..
332,536 - 9,733,566 65.57 .
0 0 68,272 .46 1
360 - 14,969 0.10 ;
13 605 1,642 0.01 :
1,596 540,044 1,697,932 : 11.44 :
6,586 9,509 2,511,737 ; 16.92 i
715 1,348 17,604 , 0.12
,
744,500 770,500 5.19
332 1,609 8,394 1 0.06 !
342,224 1,298,268 U, 844, 395
2.30 8.75 100.00
85
-------
Table 6.6
Consumers' Surplus at 1976 Consumption' Levels,
Using the 1976 Level of Oxidanc Concentration
Crop
Vegetable Crops
Beans, Pro. Gr.
Lima
Broccoli
Cantaloupes
Carrots
Cauliflower
Celery
Lettuce, Head
Onion, Fresh
Onion, Process
Potato
Tomato, Fresh
Tomato, Process
Field Crops
Cotton, Lint
Sugarbeets
Total
Percent of
Total
Southern
Desert
—
-
0
0
-
-
0
0
0
-
0
0
I
1
28,000
1
0
28 ,000
i 0.25
i
t
Region
Southern
San
South Central Joaquin
Coast Coast Valley Total
1
i C t
? . !
i
4,832 i 16 1,167 6,015
0 | 0 - 0
0-0 0
0 0 0 I 0
00- 0
7,120,516 408,580 - 7,529,096
9,254 0 0 ! 9,254
919 849 - 1,768
1 ;
390 6 898 1,294
481,268 1,310 | 660,112 ; 1,142,690
1,344,817 3,505 3,401 ' 1,351,723
! 3,288 389 ! 1,852 5,529
I |
1 <
i
i
! l
i 1
: 7,500 ; - ! 979,500 1,015,000
: 289 i 695 2,094 • 3,078
8,973,073 ! 415,350 1,649,024 . 11,065,447
i •
81,09 '• 3;76 14.90 !
; i
! i
Percent
of
Total
0.05
0
0
0
.
0
68.04
0.08
0.02
0.01
10.33
12.22
; 0.05
i
!
i
9.17
0.03
100.00
86
-------
As a benchmark on the magnitude of these results, the results obtained
can be compared with those obtained by Millecan (1976)3/ although the meth-
odologies used are quite different. In the Millecan study, the total crop
loss (obtained by multiplying the reduction in yield with prices (for vege-
tables^/ due to air pollution in the South Coast region (includes Los Ange-
les, Orange, Riverside, San Bernardino and Ventura Counties) has an average
value of $1,400,308 per annum from 1970 to 1974. Total loss for field
cropsV in that region for the same period is $964,047 per year. For Los
Angeles and Orange Counties, the Millecan study did not specify the types of
vegetable and field crops included, thus it is not possible to compare re-
sults on an individual crop basis. Nevertheless, one common finding is that
celery suffers the heaviest loss among included vegetable crops in Ventura
County. It should be noted that the Millecan study did not include some
counties selected for this study, e.g., Kern, Tulare, Imperial and the Cen-
tral Coast. The magnitude of the difference in total damages realized under
the two approaches suggests that damages (in terms of "costs" to consumers)
may be underestimated in earlier research.
It should also be noted that the results of this study, as presented in
this section, do not include effects of air pollution on producers (growers).
Such effects may be reflected in higher cost of production and/or lower
revenue, depending upon the price elasticity for each crop. These effects
will be addressed in the second phase of the analysis via the mathematical
model presented earlier. In addition, this study includes only selected
types of vegetable and field crops; thus, the value of crop losses derived
above represents only a portion of total crop losses in these regions. One
would expect to have a much higher value of crop losses if other types of
agricultural crops, such as citrus and horticultural crops, were also in-
cluded in the analysis.
87
-------
FOOTNOTES: CHAPTER VI
— For details see Johnston and Dean (1969).
2/
— The concept of compensating variation (or price compensating)
popularized by R. Hicks, is the amount of money the consumer of a commodity
would have to gain (lose) in order to offset the loss (gain) in utility
due to the rise (fall) in price of that commodity (caused by, say, reduction
in quantity supplied due to yield depression in the presence of air
pollution) in order to be as well off as before. It differs from
"equivalent variation" (or price equivalent) in that the level of utility,
after being compensated, in the case of compensating variation is unchanged
whereas in the case of equivalent variation, it is the amount of money
paid to (or received from) the consumer in order to make him as well off
as before after the changes in utility level caused by the rise (or fall)
in price of that commodity.
3/
— Details of that study had already been discussed in Chapter II
of this analysis.
4/ 57
— '~~ The mix of vegetable and field crops included in the Millecan
study do not coincide with those in this study. Also, Millecan includes
more crops in the analysis.
88
-------
CHAPTER VII
IMPLEMENTATION OF THE COMPLETE MODEL: AN ASSESSMENT
The preceding six chapters have dealt with numerous conceptual and em-
pirical issues relevant to the assessment of air pollution damages to crops.
As is evident, the analysis to date has not integrated and empiricized the
complete set of components. Specifically, the economic costs at the pro-
ducer's level have not been measured. Included under this general area of
producer's impacts are such issues as changes in cropping mix and location
in response to air pollution, substitution effects on the input side and
other mitagative strategies. Also, impacts of air pollution on non-inclu-
ded crops (e.g., perennials and horticultural crops) are not addressed.
This concluding chapter will deal with these areas, with an emphasis on de-
tailing the approaches to be used in their assessment in the second phase
of the agricultural impact study.
7.1 Production Adjustments
Agricultural producers are capable of modifying their production de-
cisions and/or plans in the face of change. California agriculture has
demonstrated a high degree of resilency in dealing with such adjustments as
energy shortages, labor disruptions or natural phenomena such as drought.
Typical response patterns have been reflected in adjustments in cropping
patterns and input use to minimize the effects of the "shock" to the agri-
cultural system. Similar mitagative procedures would be expected in the
presence of air quality degradation. While increasing levels of oxidants
may not be viewed as a "shock," the response pattern should be similar, if
somewhat more gradual. As an indication of such adjustments, it appears
that producers of vegetable crops are planting crop varieties with greater
resistance to certain air pollutants.
The range of mitagative procedures open to producers within southern
California includes the following set of responses:
1. in situ adjustments in cropping mix, substituting more resistant
crops into current cropping systems;
2. in situ increase in input use rates to offset adverse effects of
air pollution (reflected in an increase in firm's cost structure);
and
89
-------
3. locational adjustments in production whereby production is shifted
from areas of high oxidant levels to areas of relatively low
levels (timing of such adjustments will obviously be determined
by land market considerations).
In addition to such mitagative procedures, which entail either in-
creased costs or reduced returns for total produce sold, producers also
face the possibility of revenue losses due to quality degradation, even in
the absence of yield reductions. As a result of quality degradation, prices
received for selected commodities may be discounted. A further decision-
affecting phenomenon associated with air pollution is the effect on produ-
cer risk-bearing. If ambient air quality experiences a continuous or
abrupt degradation over time, crop yield variation (a major source of farm
risk) may be increased. Thus, the inherent riskiness of crop production
decisions may be exacerbated.
'It should be noted that the potential exists for net increases in the
revenue of producers in the face of yield reductions, given the price elas-
ticity of demand for some agricultural crops. Such an outcome would be de-
pendent upon the price elasticity of each crop in the crop mix and the mag-
nitude of changes in the firm's cost structure due to mitagation.. Given
the price endogenous nature of the proposed mathematical model, this poten-
tial outcome would be tested directly within the analysis.
The mathematical model formulated in Chapter III of this report is
intended to deal with the production decision variables outlined above.
The data for such an analysis has been obtained and risk measures have been
calculated. The overall integration effort will be discussed below.
7.2 Consumer Impacts
Chapter VI of this report presented a somewhat simplistic assessment
of consumer effects of air pollution. The economic cost of air pollution
(compensating variation) was captured via the use of price forecasting
equations for each included crop. However, given that production adjust-
ments in the form of cropping mix changes' or relocation will also affect
quantities supplied, an integration of producer and consumer sectors is
desired and needed to capture future economic effects of air pollution.
This can be accomplished through the price endogenous model outlined in
Chapter III.
Indirect impacts on a third group, input suppliers, could also be sub-
stantial, if the derived demand for inputs were altered as a result of such
mitagative procedures as changes in cropping mix or input use. Major crop
adjustments could also portend significant disruptions to agricultural land
markets as well as the demand for irrigation water, given a differential in
production coefficients across crops. While input suppliers are not inclu-
ded within the scope of this analysis, the resource usage and shadow price
values generated by the model should suggest potential input supply disrup-
tions.
90
-------
7.3 The Integrated Model
As discussed in Chapter III, the complete model will assess a wide
range of possible outcomes associated with actual and projected levels of
air pollution, with emphasis on approximating current damages (under actual
air quality parameters) as well as potential damages under a range of pos-
sible air quality changes.
The model output will feature the surplus maximizing (producer's and
consumer's) levels of commodity production (for the included crops) in the
face of alternative levels of oxidant concentration. The programming algo-
rithm employed will optimize, based on the relationship between commodity
prices, yield sensitivity and resource availabilities. Additional output
from the model should be regional production, equilibrium prices, resource
usage and resource shadow prices as well as the relevant surpluses.
While most data necessary for the construction of the model has been
collected, additional programming assistance is needed to develop sub-
routines for existing software. This programming is needed to:
1. allow for multiple regions in the analysis (test of locational
adjustments in production between the South Coast and the three
contiguous regions);
2. introduce risk directly into the objective function; and
3. include cost vectors directly in the objective function.
While current economic damages can be approximated in the absence of the
programming effort, the full general equilibrium flavor of the analysis will
be lacking without such an effort.
7.4 Related Research Needs
The yield-oxidant relationships used in this analysis have been out-
lined in Chapter IV. The correlation analysis and production function esti-
mation serve to establish a possible negative relationship between oxidants
and selected crops, over the last 20 years. The significance and signs
attached to oxidants suggest a range of sensitivities across crops. How-
ever, to further test the relationship and to establish consistency with re-
sults obtained under controlled conditions, a more complete production
function is required. A more complete specification of the production
function would serve to further define the nature and magnitude of the oxi-
dant-yield interface under actual production conditions.
The included crops in this study have been limited to annual vegetables
and field crops. Some measure of damages experienced by perennials such as
fruits and nuts, as well as horticultural crops, is needed to complete the
analysis. While their complex time horizons make assessment more difficult
(in a dynamic sense), damages can be approximated via more pedestrian
91
-------
approaches such as survey techniques. These results would be needed for a
complete agricultural assessment.
7.5 Concluding Comment
The primary purpose of the agricultural assessment component of the EPA
Benefits project is to address some conceptual and empirical limitations of
earlier studies concerning agricultural damages. The first specific objec-
tive of the agricultural study is to define a methodology capable of dealing
with some of the weaknesses inherent in previous research. Thus, this study
should not be viewed as a definitive empirical assessment of agricultural
damages within southern California, but rather an initial inquiry into crop
damage assessment methodologies.
The analytical framework, conceptual issues and preliminary results re-
ported in this report offer support to the use of more complete models in the
measurement of air pollution damages/benefits. While this report and results
obtained in the next phase of the project will not resolve all relevant issues
in assessment methodologies, it is hoped that the study output will be sug-
gestive of more fertile areas for investigation.
92
-------
REFERENCES
Adams, R.M., A Quadratic Programming Approach to the Production of California
Field and Vegetable Crops Emphasizing Land, Water and Energy Use, Ph.D.
Dissertation, University of California, Davis, 1975 (mimeo).
Anderson, R.J., Jr. and T.D. Crocker, Air Pollution and Housing: Some Findings,
Herman C. Krannert Graduate School of Industrial Administration, Purdue
University, Lafayette, Indiana, Paper No. 264 (January 1970).
Barrett, L.B. and I.E. Waddell, Cost of Air Pollution Damages: A Status
Report, EPA Research Triangle Park, N.C., February .1973, Chapter 4,
27-31.
Benedict, H.M., Economic Impact of Air Pollutants on Plants in the United
States, Vols. 1 & 2, Stanford Research Institute, Menlo Park, California,
1970.
Benedict, H.M., C.J. Miller, and R.E. Olson, Economic Impact of Air Pollutants
on Plants in the United States, Final Report, Contract CRC-APRAC, CAPA-2-68
(1-70) Stanford Research Institute, Menlo Park, California, November 1971.
Benedict, H.M., C.J. Miller and J.S. Smith, Assessment of Economic Impact
of Air Pollutants on Vegetation in the United States—1969 and 1971,
Stanford Research Institute, Menlo Park, California, July 1973.
Brandt, C.S. and W.W. Heck, "Effects of Air Pollutants on Vegetation,"
Air Pollution Vol. 1: Air Pollution and Its Effects, A.C. Stern (ed.),
Academic Press: New York, 1968, 401-443.
Brandt, J.A., B.C. French and E.V. Jesse, Economic Performance of the Processing
Tomato Indusj:jr/, Giannini Foundation of Agricultural Economics Information
Series No. 78-1, University of California, April 1978.
Brewer, R.F. and G. Ferry, "Effects of Air Pollution on Cotton in the San
Joaquin Valley," California Agriculture, June 1974.
Cameron, C.A., Garden Chronicle 1, 274 (1874).
Cameron, J.W., H. Johnson, Jr., O.C. Taylor and H.W. Otto, "Differential
Susceptibility of Sweet Corn Hybrids to Field Injury by Air Pollution,"
HortScience: 5(4), 217-21.9 (1970).
93
-------
Clayberg, C.D., "Screening Tomatoes for Ozone Resistance," HortScience 6(4),
396-397 (1971).
Clayberg, C.D., "Evaluation of Tomato Varieties for Resistance to Ozone,"
Connecticut Agricultural Experimentation Circulation #246, 3-11 (1972).
County Supervisors Association of California, California County Fact Book:
1976-1977.
Crocker, T.D., Urban Air Pollution Damage Functions: Theory and Measurement,
Environmental Protection Agency, Washington, D.C., Final report, Contract
No. CPA 22-69-52 (January 1971).
Daines, R.H., I.A. Leone, and E. Brennan, "Air Pollution as it Affects
Agriculture in New Jersey," New Jersey Agricultural Experiment Station
Bulletin #794, 1960.
Davis,-D.D. and L. Kress, "The Relative Susceptibility of Ten Bean Varieties
to Ozone," Plant Dis.ease Reporter 58(1), 14-16 (1974).
Enke, S., "Equilibrium Among Spatially Separated Markets: Solution by Electric
Analogue," Econometrica, Vol. 19 (1951), 40-47.
Feliciano, A., 1971 Survey and Assessment of Air Pollution Damage to Vegetation
in New Jersey, EPA-R5-72-010, Cooperative Extension Service, CAES, Rutgers
State University, New Brunswick, N.J. (1972).
Freebairn, H.T., "Reversal of Inhibitory Effects of Ozone on Oxygen Uptake of
Mitochondria," Science 126, 303-304 (1957).
Gudarian, R., H. van Haut and H. Stratmann, "Probleme der Erfassung und
Beurteilung von Wirkungen gasfoermiger Luftverunreinigungen auf die
Vegetation," Z. Pflanzenkrankh Pflanzenschutz 67: 257 (1960).
Hazell, P.B.R., "A Linear Alternative to Quadratic and Semivariance Programming
for Farm Planning Under Uncertainty," American Journal of Agricultural
Economics 53(1), February 1971, 53-62.
Hazell, P.B.R. and P.L. Scadizzo, "Competitive Demand Structures under Risk
in Agricultural Linear Programming Models," American Journal of Agricultural
Economics 56(2), May 1974, 235-244.
Hazell, P.B.R. and P.L. Scadizzo, "Market Intervention Policies When Production
is Risky," American Journal of Agricultural Economics 57(4), November 1975,
641-649.
Henderson, J.M. and R.E. Quandt, Microeccnomic Theory, A Mathematical Approach,
New York: McGraw-Hill (1971).
Hill, A.C., M.R. Pack, M. Treshow, R.J. Downs, and L.G. Transtrurn, "Plant
Injury Induced by Ozone," Phytopathology 51, 356-363 (1961).
94
-------
Hitchcock, F.L,, "Distribution of a Product from Several Sources to Numerous
Localities," Journal of Mathematics and Physics, Vol. 21 (1941), 224-230.
Roman, C., "Effects of Ionized Air and Ozone in Plants," Plant Physiology 12,
957-958 (1937). "
Johnson, S.R., "A Reexamination of the Farm Diversification Problem,"
Journal of Farm Economics 49(3), August 1967, 610-621.
Johnston, W.E. and G.W. Dean, California Crop Trends: Yields, Acreages,
and Production Areas, California Agricultural Experiment Station,
Extension Service Circular 551, Berkeley: University of California,
November 1969.
King, G.A., E.V. Jesse, and B.C. French, Economic Trends in the Processing
Tomato Industry;, California Agricultural Experiment Station, Extension
Service Information Series in Agricultural Economics No. 73-4, University
of California, 1973.
King, G.A., R.M. Adams, and W.E. Johnston, "Selected California Vegetable
and Field Ci'op Price-Forecasting Equations," supplement to Some Effects
of Alternative Energy Policies on California Annual Crop Production,
Giannini Foundation Research Report No. 326, July 1978.
Knight, R.C. and J.E. Priestly, "The Respiration of Plants under Various
Electrical Conditions," Annals of Botany (London) 28, 135-161 (1914).
Koopmans, T.C., "Optimum Utilization of the Transportation System," Econometrica
Vol. 71, Supplement (1949), 136-146.
Lacasse, N.L., T.C. Weidensaul and J.W. Carroll, Statewide Survey of Air
Pollution Damage to Vegetation, 1.969, Center for Air Environment Studies
(CAES), State College, Penn. CAES Publication No. 148-70, Jan. 1970.
Lacasse, N.L., Assessment of Air Pollution Damage to Vegetation in Pennsylvania,
Center for Air Environment Studies (AAES), Publication #209-71, University
Park, Pennsylvania State University (1971).
Larsen, R.I. and W.W. Heck, "An Air Quality Data Analysis System for
Interrelating Effects, Standards, and Needed Source Reductions: Part 3
Vegetation Injury," Journal of the Air Pollution Control Association 26(4)
April 1976, 325-333.
Lea, M.C., "On the: Influence of Ozone and Other Chemical Agents on Germination
and Vegetation," American Journal o£ Science and Arts, 37, 373-376 (1864).
Ledbetter, M.C., P.W. Zimmerman and A.E. Hitchcock, "The Histopathological
Effects of Ozone on Plant Foliage," Contribution Boyce Thompson Institute
20, 275-282 (1959).
Liu, Ben-Chieh and Eden Siu-Hung Yu, Physical and Economic Damage Functions
for Air Pollutants by Receptors, Midwest Research Institute, Kansas City,
Missouri, USEPA, September 1976.
95
-------
Middleton, J.T., "Photochemical Air Pollution Damage to Plants," Annual Review
of Plant Physiology 12: 431-448 (1961).
Middleton, J.T., J.B. Kendrick, Jr., and H.W. Schwalm, "Injury to Herbaceous
Plants hy Smog or Air Pollution," Plant Disease Reporter 34, 245-252,
(1950).
Middleton, J.T. and A.O. Paulus, "The Identification and Distribution of Air
Pollution through Plant Responses," AMA Archive of Industrial Health 14
December 1956, 526-532.
Millecan, A.A., A Survey and Assessment of Air Pollution Damage to California
Vegetation, California Department of Agriculture, June 1971.
Millecan, A.A., A Survey and Assessment of Air Pollution Damage to California
Vegetation, Department of Food and Agriculture, Sacramento, California
(1976).
Middleton, J.T., E.F. Darley and R.F. Brewer, "Damage to Vegetation from
Polluted Atmospheres," Journal of the Air Pollution Control Association,
8, 9-15, 1958.
Naegele, J.A., W.A. Feder and C.J. Brandt, Assessment of Air Pollution Damage:
to Vegetation in New England, July 1971-July 1972, Final report, EPA-R5-
72-009, Surburban Experiment Station, University of Massachusetts,
Waltham, Mass., 1972.
National Academy of Science, Committee on Medical and Biologic Effects of
Enviornmental Pollutants, Ozone and Other Photochemical Oxidants,
Washington, D.C. (1977).
O'Gara, P.J., "Sulphur Dioxide and Fume Problems and their Solutions,"
Journal of Industrial and Engineering Chemistry 14, 744 (1922).
Oshima, R.J., Development of a System for Evaluating and Reporting Economic
Crop Losses caused by Air Pollution in California: I Quality Study,
Final report to the California Air Resources Board under the Agreement
ARB-287 (1973).
Oshima, R.J., Final report to the California Air Resources Board under the
Agreement ARB-3-690 Development of a System for Evaluating and Reporting
Economic Crop Losses in California III: Ozone Dosage-Crop Loss
Conversion Function—Alfalfa, Sweet Corn (1975).
Oshima, R.J., O.C. Taylor, P.K. Braegelmann and D.W. Baldwin, "Effect of
Ozone on the Yield and Plant Biomass of a Commercial Variety of Tomato,"
Journal of Environmental Quality 4(4), 463-464 (1975).
Oshima, R.J., M.P. Pie, P.K. Braegelmann, D.W. Baldwin and V.V. Way, "Ozone
Dosage-Crop Loss Function for Alfalfa: A Standardized Method for
Assessing Crop Losses for Air Pollutants," Journal of the Air Pollution
Control Association 26(9), September 1976, 861-865.
9.6
-------
Oshima, R.J., P.K. Braegelraann, D.W. Baldwin, V.V. Way and O.C. Taylor,
"Reduction of Tomato Fruit Size and Yield by Ozone," Journal of the
American Society for Horticultural Science 102(3), May 1977, 289-293.
Peckham, B.W., Air Pollution and Residential Property Values in Philadelphia,
HEW, PHS, NAPCA, Division of Economic Effects Research, Raleigh, N.C.,
unpublished report (September 1970).
Pell, E.J., 1972 Survey and Assessment of Air Pollution Damage to Vegetation
in New Jersey, EPA-R5-73-022, Cooperative Extension Service, CAES,
Rutgers State University, New Brunswick, N.J. (1973).
Pill, J., "The Delphi Method: Substance, Context, A Critique and an
Annotated Bibliography," Socio-Econ. Plan. Sci. 5(1), February 1971.
Povitaitis, B., "A Historical Study of the Effects of Weather Fleck on Leaf
Tissues of Flue-curred Tobacco," Canadian Journal of Botany 40, 327-330
(1962).
Reinert, R.A., D.T. Tingey and H.C. Carter, "Varietal Sensitivity of Tomato
and Radish to Ozone," HortScience 4, 189 (1969).
Rich, S., "Ozone Damage to Plants," Annual Review of Phytopathology 2,
253-266 (1964).
Richards, B.L., J.T. Middleton and W.B. Hewitt, "Air Pollution with Relation
to Agronomic Crops: V, Oxidant Stipple of Grape," Agronomic Journal
50, 539-561 (1958).
Ridker, R.B. and J. Henning, "The Determinants of Residential Property Values
with Special Reference to Air Pollution," Review of Economics and
Statistics 49, May 1967, 246-257.
Samuelson, P.A., "Spatial Price Equilibrium and Linear Programming,"
American Economic Review 42(3), June 1952, 283-303.
Schomer, H.A. and L.P. McColloch, "Ozone in Relation to Storage of Apples,"
U.S. Department of Agriculture Circulation #768 (1948).
Shuffett, D.M., The Demand and Price Structure for Selected Vegetables,
U.S. Department of Agriculture Technical Bulletin No. 1105,
Washington, D.C. (1954).
Spore, R.L., Property Value Differential as a Measure of the Economic Costs
of Air Pollution, Ph.D. Dissertation, Center for Air Environment Studies,
Pennsylvania State University, University Park, Pennsylvania, CAES
Publication No. 254-72 (June 1972).
Stratmann, H., "Field Experiments to Determine the Effects of SO on Vegetation,"
Forschungsberichte des Landes Nordrhein-Westfallen, Essen, Rest Germany,
No. 1984 (19.63).
97
-------
Takayama, T. and G.G. Judge, "Spatial Equilibrium and Quadratic Programming,"
Journal of Farm Economics 46(1), February 1964a, 67-93.
Takayama, T. and G.G. Judge, "Equilibrium Among Spatially Separated Markets:
A Reformulation," Econometrica 32, October 1964b, 510-524.
Thomas, M.D. and G.R. Hill, Jr., "Absorption of Sulfur Dioxide by Alfalfa
and Its Relation to Leaf Injury," Plant Physiology 10 (1935), 291.
Thomas, M.D., "Effects of Air Pollution on Plants," in Air Pollution World
Health Organization Monograph Series No. 46, New York: Columbia
University Press, 1961, 233-278.
Thompson, C.R., "Economic Effects of Smog Injury to Agricultural Crops,"
Final report to Cal ARE Contract 2-650, June 30, 1975 (mimeo).
Thompson, C.R. and J.O. Ivie, International Journal Air Water Pollution 9(3)
799-805 (1965).
Thompson, C.R. and O.C. Taylor, "Plastic-Covered Greenhouse Supply Controlled
Atmospheres to Citrus Trees," Transactions of the American Society of
Agricultural Engineers 9(3), 338-339, 342 (1966).
Thompson, C.R., O.C. Taylor, M.D. Thomas, and J.O. Ivie, "Effects of Air
Pollutants on Apparent Photosynthesis and Water Use by Citrus Trees,"
Environmental Science & Technology 1(8), 644-650 (1967).
Thompson, C.R. and O.C. Taylor, "Effects of Air Pollutants on Growth, Leaf
Drop, Fruit Drops, and Yield of Citrus Trees," Environmental Science
& Technology, Vol. 3, No. 10, October 1969, 934-940.
Thompson, C.R., G. Kats and E.O. Hensel, "Effects of Ambient Levels of NO
on Navel Oranges," Environmental Science & Technology, Vol. 5, No. 10,
October 1971, 1017-1019.
Thompson, C.R., G. Kats and J.W. Cameron, "Effects of Ambient Photochemical
Air Pollutants on Growth, Yield, and Ear Characters of Two Sweet Corn
Hybrids," Journal of Environmental Quality 5(4), 410-412 (1976).
Thompson, K.J. and P.B..R. Hazell, "Reliability of Using the Mean Absolute
Deviation to Derive Efficient E,V, Farm Plans," American Journal of
Agricultural Economics 54(3), August 1972, 5Q3-5Q6.
Tomek, W.G. and K.L. Robinson, Agricultural Product Prices, Cornell University
Press, Ithaca, New York (1972).
U.S. Department of Agriculture, Agricultural Statistics, various issues.
Vars, C.R., Jr. and G.W. Sorrenson, Study of the Economic Effects of Changes
in Air Quality, EPA, Research Triangle Park., N.C., final report, Contract
Number CPA 70-117, June 1972.
98
-------
Waddell, I.E., The Economic Damages of Air Pollution, Washington, D.C.,
Environmental Research Center, Office of Research and Development,
USEPA, May 1974.
Weidensaul, T.C. and N.L. Lacasse, Statewide Survey o£ Air Pollution Damage
to Vegetation—1969, Center for Air Environment Studies, Pennsylvania
State University, Publication No. 148-70 (1970).
Wiens, T.B., "Peasant Risk Aversion and Allocation Behavior: A Quadratic
Programming Experiment," American Journal of Agricultural Economics
58(4), November 1976, Part I, 629-635.
Zahn, R., "Untersuchungen uber die bedeutung Kontinuierlicher and intermittierender
Schwefeldioxideinwirkung fuer die Pflanzenreaktion," Staub 23, 343 (1963).
Zusman, P., "Econometric Analysis of the Market for California Early Potatoes,
Hilgardia, Vol. 33, No. 11 (1962).
!l
99
-------
TECHNICAL REPORT DATA
(I'lcasc read Instructions on the reverse before coniplctiii);)
1. REPORT NO.
4. TITLE AND SUBTITLE Methods Development for Assessing Air
Pollution Control Benefits: Volume III, A Preliminary
Assessment of Air Pollution Damages for Selected Crops
Within Southern California
3. RECIPIENT'S ACCESSION NO.
5. REPORT DATE
February 1979
6. PERFORMING ORGANIZATION CODE
7.AUTHORS Richard M. Adams
Narongsakdi Thanavibulchai
Thomas D. Crocker
8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAMB AND ADDRESS
University of Wyoming
Laramie, Wyoming 82071
10. PROGRAM ELEMENT NO.
1HA616 and 630
11. CONTRACT/GRANT NO.
R805059-01
12. SPONSORING AGENCY NAME AND ADDRESS
Office of Health and Ecological Effects
Office of Research and Development
U.S. Environmental Protection Agency
Washington, DC 20460
13. TYPE OF REPORT AND PERIOD COVERED
Interim Final, Oct 76 - Oct 78
14. SPONSORING AGENCY CODE
EPA-600/18
15. SUPPLEMENTARY NOTES
16. ABSTRACT
This volume of a five volume study of the economic benefits of air pollution
control investigates the economic benefits that would accrue from reductions in
oxidant/ozone air pollution-induced damages to 14 annual vegetable and field
crops in southern California. Southern California production of many of these
crops constitutes the bulk of national production.
Using the analytical perspective of economics, the study provides an up-
to-date review of the literature on the physical and economic damages to
agricultural crops from air pollution. In addition, methodologies are developed
permitting estimation of the impact of air pollution-induced price effects
input and output substitution effects, and risk effects upon producer and
consumer losses. Estimates of the extent to which price effects contribute to
consumer losses are provided. These consumer losses are estimated to have
amounted to $14.8 million per year from 1972 to 1976. This loss is about
1.48% of the total value of production for the included crops in the area and
0.82/0 of the value of these crops produced in the State of California Celerv
fresh tomatoes, and potatoes are the sources of most of these losses
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.EDENTIFIERS/OPEN ENDEDTERMS
COSATI Field/Group
Economic analysis
Air pollution
Agricultural economics
Economic benefits of
pollution control
Los Angeles
13B
18. DISTRIBUTION STATEMENT
Release unlimited
19. SECURITY CLASS (This Report/
Unclassified
21. NO. OF PAGES
108
20. SECURITY CLASS (This page}
Unclassified
22. PRICE
EPA Form 2220-1 (9-73)
-------
United States
Environmental Protection
Agency
Rfl-fifl.T
Official Business
Penalty for Private Use
$300
Special Fourth-Class Rate
Book
Postage and Fees Paid
EPA
Permit No. G-35
Washington DC 20460
£•
2'
.a
<
Q_
LU
!?
; in
: o
i CM
------- |