United States
Environmental Protection
Agency
Office of
Drinking Water (WH-550)
Washington, DC 20460
EPA 570/9-84-004
June 1984
Wtttr
National Statistical
Assessment of Rural Water
Conditions
Technical Summary
-------
National Statistical Assessment
of Rural Water Conditions
Summary
-------
Summary
National Statistical Assessment
of Rural Water Conditions
Report prepared for
The Office of Drinking Water
US Environmental Protection Agency
by
DOE D. FRANCIS, Principal Investigator
BRUCE L. BROWER
WENDY F. GRAHAM
OSCAR W. LARSON III
JULIAN L. McCAULL
HELENE MORAN VIGORITA
Department of Rural Sociology
Cornell University
Illustrations and Graphics by:
Susan B. Lent
Word Processing by:
Betty L. Van Amburg
-------
Contents
Preface
Ackrr,wledgemefltS
I —
Summary of Chapter II I - 2
Summary of Chapter III 4
Demographic Profile of Rural US 5
Profile of Uses of Water 7
Water-Using Equipment and Wastewater Systems 8
Attitudes Toward Water Supplies 10
Summary of Chapter IV 13
Major Household Water Supplies 14
Drinking Water Supplies 15
Sources of Major Water Supplies 17
Summary of Chapter V 20
Quality 20
NSA Procedure 21
Bacteriological Findings 21
Physical and Chemical CharacteristicS 24
Inorganic Constituent Findings 25
Organic Constituent Findings 27
Radioactivity Findings 27
Proportions of Households Exceeding NSA
Reference Values 28
-------
(1 continued)
Perceived Water Quality I - 36
Odor 37
Taste 37
Cloudiness 38
Color 39
Sediment 40
Temperature 40
Water Quantity 41
Water Availability 43
Reliability 43
Accessibility 45
Effects of Quality, Quantity, and Availability 48
Cost 49
Affordability 51
Summary of Chapter VI 53
Summary of Chapter VII 57
Individual Systems 57
Intermediate Systems 64
Community Systems 70
Independent Systems 72
Consolidated Systems 77
Summary of Chapter VIII 81
Community Systems 82
Intermediate Systems 84
Summary of Chapter IX 87
Summary of Chapter X 90
Laboratory Indicators of Water Quality:
Regression Analyses 92
Total Coliform Organisms 92
Turbidity 93
Nitrate-N 94
Lead 94
Aesthetic/Economic Indicator 95
Health-Related Indicator 95
General Water Quality Indicator 96
Laboratory Indicators of Water Quality:
Conting ncy Tables 97
Tøtal Coliform Organisms 97
Perceived Water Quality Indicator 102
Summary of Chapter XI 103
Recorded Quantity 103
Perceived Quantity 105
Summary of Chapter XII 106
Accessibility 106
Reliability 108
Summary of Chapter XIII 109
-2-
-------
(I continued)
Monthly Cost I - 109
Unit Cost 110
-3-
-------
I Preface
This report presents the findings of a national research survey of rural
household water conditions. The purpose of the study, which was mandated by
Congress in the Safe Drinking Water Act of 1974, was to gather information about
the quantity, quality, availability, and cost of water supplies in rural America.
Under the direction of the US Environmental Protection Agency (EPA),
extensive developmental work was done in preparation for the survey, including
specialized studies on particular aspects of rural water supply. These efforts laid
the groundwork for what became the major thrust of the rural water survey: the
National Statistical Assessment of Rural Water Conditions (NSA). The NSA was
the first systematic national study of rural household water supplies. It was
designed as a comprehensive sampling of rural households so that it would
accurately reflect the full spectrum of rural domestic water conditions. Supplies
using any of three basic sources of water—groundwater, surface water, and
precipitation—were represented in the survey, as were all sizes of water supply
systems—individual supply systems which served a single household, intermediate
systems (defined in the NSA as any multiple-connection arrangement which served
from two through fourteen connections), and community systems of various sizes.
-------
Preface - 2
The NSA studied the physical and technical components of water supplies
and performed a laboratory water quality analysis on specimens of each household’s
major water supply. In addition, through interviews at households and at supply
systems, the NSA looked at the social framework that surrounds every rural
household’s water supply—the perceptions and attitudes of those who use the
water, the personal economics of securing a water supply, and the financing of
intermediate and community systems. The outcome of all this work is an
integrated picture of rural household water conditions in the continental United
States.
The NSA used an elaborate sampling scheme to ensure that its data would
be representative of the diverse situations that existed in various parts of the
country. It gathered data from every state except Alaska, Hawaii, and Wyoming.
Alaska and Hawaii were purposely excluded because their unique characteristics
would have interfered with generalizations to the rest of the United States. Since
the sample was drawn in proportion to rural population, the low population of
Wyoming precluded any of its counties from being selected for study. A total of
2,654 households were interviewed and studied; this sample was representative of
the roughly 22 million households in rural areas of the US in 1978. Data collection
took place from May 1978 through :January 1979.
This document and its attachments comprise the final report of the NSA
and the rural water survey. The report presents the results of the NSA and
incorporates findings from certain associated investigations that were done in
conjunction with the rural water survey. The authors have attempted to provide a
readable yet thorough and technically complete report of the massive survey
research effort.
The detailed findings constitute a sourcebook of information on many
aspects of rural households and rural water supply. Because of the size of the
report, however, descriptive and analytical findings are summarized in Chapter 1.
-------
Preface - 3
This technical summary may be useful to some readers as an overview of NSA
results.
The NSA used a multidisciplinary approach to describing and analyzing
rural household water conditions. It is hoped that this report will be a useful
reference document for many readers in the various disciplines, and that it will
serve as groundwork for further research and informed policy on rural water
supply.
-------
I Acknowledgments
Early planning of the NSA was done by a special research group in the
Department of Rural Sociology at Cornell University in consultation with EPA
staff and an advisory committee from the National Demonstration Water Project, a
Washington-based nonprofit organization concerned with the provision of adequate,
affordable water and wastewater disposal- services to the rural poor. As an
ancillary activity of the rural water survey, NDWP also commissioned several
studies on discrete topics dealing with rural water supply.
The Cornell research group, in concert with the Office of Drinking Water,
EPA, developed the conceptual framework of the NSA, designed the survey, and
constructed the interview schedule, which underwent extensive pretesting before
use. The Cornell group was aided in this effort by sanitary engineers, microbiol-
ogists, and water chemists from the EPA staff; consultants on water systems
technology from Engineering Enterprises, Inc., of Norman, Oklahoma; and several
social scientists from various universities. The Cornell researchers also received
valuable planning assistance from the Confer nce of State Sanitary Engineers, the
University Council on Water Resources, and the National Viater Well Association.
Cornell did not perform the data collection itself. Instead, it .ook an advisory role
-------
in both the data collection, which was done by TransCentury Corporation of
Washington, DC, and in the design of the sample, which was drawn by Survey
Design, Inc., a subcontractor of TransCentury. As part of its advisory capacity,
the Cornell research staff supervised and monitored the field work, coding, and
computerized data storage. The Cornell research group conducted all phases of the
data analysis and compiled the final report.
Several laboratory groups and organizations were involved in the analysis
of bacteriological, chemical, radiological, and physical properties of the drinking
water samples collected at each household. Under subcontract with TransCentury
to conduct the bacteriological assays were Energy Resources Company, a private
laboratory organization which was assigned the lead role and worked in cooperation
with the Madison County Environmental Control Department and the Colorado
State Health Department. Under a separate contractual arrangement with the
EPA Office of Drinking Water, analysis for the presence of herbicides and
pesticides was conducted by laboratories at the Medical University of South
Carolina and the Mississippi State Chemical Laboratory. The assays of chemical
and physical properties were conducted by the EPA-Municipal Environmental
Research Laboratory in Cincinnati and EPA-Environmental Monitoring and Support
Laboratory in Las Vegas. The latter organization also conducted all the radio-
logical assays. All results from these laboratories were transmitted to Trans-
Century which, in turn, transmitted them to Cornell.
In compiling the quality assurance report on the work at these laboratories,
several individuals from the Technical Support Group of the EPA Office of
Drinking Water were most helpful. Special th pks to Lowell VanDenBerg and Carl
Hirth from that group.
Throughout all phases of the NSA, Cornell and the EPA worked together
closely. To give further assurance that the NSA would adequately address the
-------
needs of the many interested parties, a review panel was brought together to assess
the data analysis efforts. Its recommendations on reporting the research results
were of critical importance in the final phase of the study and led to an improved,
more readable report.
The review team, selected by EPA, consisted of Martin Allen, Hugh
Holman, Arnold M. Kuzmack, Richard Rarnes, Philip Tate, and Wilbur Whitsell, all
of EPA; Joseph Delfino, University of Wisconsin - Madison; James Edzwald,
Clarkson College of Technology; John Herriandez, New Mexico State University at
Las Cruces;’ Tyler E. Gass, Bennett and Gass, Inc.; 2 Harvey Johnson, Mississippi
Institute for Small Towns; George Kiumb, Culligan International Co.; Robert L.
Lawrence, Director of Public Utilities, City of Alexandria, Louisiana; 3 William N.
Long, a private consultant on sanitary engineering from Boulder, Colorado; Robert
Murphy, National Center for Health Statistics; David L. Rogers, Colorado State
University; and Glenn Walden and Gary Morgan, Farmers Home Administration.
Special thanks to Phillip L. Tate of the Office of Drinking Water, EPA,
whose managerial skill and personal tact guided the study through its many
difficult phases.
The Cornell staff is indebted to several colleagues for their helpful advice
and consultation. David Allee, Norman Dondero, and Don Gates helped us through
many difficult analysis and reporting dilemmas. Dave Ingersoll and Ramish Vaidya
provided valuable help with the economics of water supply. Todd Quirilan was
indispensable to the u-aining and monitoring of interviewers, editors, and coders.
‘On March 19, 1981, Dr. Hernandez tendered his resignation from the review team,
as he became a Presidential designee for the position of Deputy Administrator of
EPA, to which he was confirmed May 6, 1981.
2 Formerly Director of Technical Services, National Water Well Association.
3 Formerly Director of Louisiana Rural Water Association.
-------
Thanks to Betty Van Amburg and 3anet Stauderman for valuable aid in
composing the report through its many drafts on the word processor and for
painstaking proofreading.
Though the Cornell staff is grateful to all of these individuals and
organizations which contributed to the enhancement of this important document,
ultimate responsibility for its accuracy and integrity resides with the Cornell
research group.
-------
I Executive Summary
Information about rural water conditions has long been fragmentary and
incomplete. The information that existed generally has been limited to the results
of unsystematic, localized surveys concentrated narrowly on a few prominent local
water quality problems. These surveys have not generally employed compatible
sample designs or methodologies. As a result, it has been difficult to formulate a
unified understanding of rural water conditions, and difficult to devise informed,
comprehensive policies and programs for rural water supplies.
To some extent, the absence of wide-ranging and detailed knowledge of
rural water conditions is related to the nature of rural water supplies. Rural water
systems are frequently individual and small multiple-connection systems, which fall
outside the regulatory and funding programs of the major “water agencies” of
government. With enactment of the Safe Drinking Water Act of 1974, Congress set
in motion two major efforts to develop systematic, current data on rural water
supplies across the nation. First, in response to growing concern with the quality
of drinking water and its effects on human health, the Safe Drinking Water Act
provided for a uniform, national set of water quality standards and extended the
monitoring and regulatory responsibility of the US government over smaller water
supplies. Second, the Act mandated a one-time national statistical assessment of
the current Status of rural domestic water characteristics.
The one-time study was to be an intensive examination of a wide range of
factors, but for only a sample of households and systems. The institutionalized
national monitoring program was to extensively cover a few factors on all
community systems. This document is concerned with the intensive, one-time
national statistical assessment of rural water conditions.
-------
Executive Summary - 2
TI-fE CALL FOR ASSESSMENT AND RESPONSE
The enabling legislation mandating the one-time assessment study was
contained in Section 3(a) of the Act. It reads:
Sec. 3(a). The Administrator of the Environmental Protection
Agency shall (after consultation with the Secretary of Agriculture
and the several States) enter into arrangements with public or
private entities as may be appropriate to conduct a survey of the
quantity, quality and availability of rural drinking water supplies.
Such survey shall include, but not be limited to, the consideration
of the number of residents in each rural area
(1) presently being inadequately served by a public or private
drinking water supply system; or by an individual home
drinking water supply system;
(2) presently having limited or otherwise inadequate access to
drinking water;
(3) who, due to the absence or inadequacy of a drinking water
supply system, are exposed to an increased health hazard; and
(4) who have experienced incidents of chronic or acute illness,
which may be attributed to the absence or inadequacy of a
drinking water supply system.
The study called for in this provision of the Act has been completed and
what follows is a summary of the five-volume, approximately 1,900 page report,
submitted to Congress in conformance with the stipulations of the Act. The study
and report, labelled the National Statistical Assessment of Rural Water Conditions
(NSA), focused primarily on the quality, quantity, availability, cost, and afford-
ability of domestic water in rural households throughout the continental United
States. Additionally, some corollary information was presented on water systems
serving these households.
RESPONSIBILITY FOR THE STUDY
Primary responsibility for oversight of the contents and conduct of the
study was allocated to the US Environmental Protection Agency. Within EPA, the
national assessment study was placed under the aegis of the Office of Drinking
Water. The planning of the content, scope, and execution of appropriate
procedures for conducting the study involved the Office of Drinking Water, various
private consulting firms, university researchers, and professional organizations.
Subsequently, proposals were solicited and a group of rural sociologists from the
Department of Rural Sociology at Cornell University was selected to further
conceptualize the study, oversee the sampling and data collection, analyze the
results, and write the report summarized herein.
In Section 3(a) of the Act, the Administrator of EPA was directed to
consult with the Secretary of Agriculture in the formulation and overview of the
study. During the formulation stage of study, two meetings were held with a
specially formed study group composed of representatives from Farmers’ Home
-------
Executive Summary - 3
Administration, Economic Research Service of the Economic Development
Division, and the Statistical Research Service of the Statistical Division of USDA.
The meetings were primarily devoted to discussions of (1) the general scope of the
study effort, (2) the types of data to be collected, (3) the manner in which this
study effort would complement and inform USDA programs and research, (4) a
critique of the conceptual foundations of the study, and (5) the ways in which the
data would be analyzed. In addition, FmHA representatives joined a review team
composed of experts representing a spectrum of professions, interest groups, and
government agencies. This review team met on three separate occasions to
critique the analysis and report writing in progress.
SIGNIFICANCE OF THE STUDY
The study summarized here is the first national study designed to investi-
gate the status of rural water quality at the point of use, in rural households. In
addition, it is the first systematic national assessment of quantity, availability,
cost, and affordability of rural domestic water supplies. It is the first national
assessment of the links between the status of domestic water in rural households
and characteristics of the supply which provides that water.
Prior to this study, the most complete information, at the national level,
about rural water was contained in the Community Water Supply Study: Analysis of
National Survey Findings , conducted by the Bureau of Water Hygiene, US Depart-
ment of Health, Education, and Welfare (1970), and Survey of Operating and
Financial Characteristics of Community Water Systems conducted by Temple,
Barker, and Sloane, Inc. for the US Environmental Protection Agency (1977).
Studies other than these were either local in scope or examined water pollution in
bodies of water located in rural areas, not the characteristics of the water at the
point of consumption.
This study complements and extends the information from the earlier
reports. But, its focus on the character of the water as it is actually consumed at
the household differs from earlier efforts. Additionally, it uniquely provides
complementary information about the water supply system. In short, it is the first
national portrayal of rural household water conditions and associated system
characteristics.
METHODOLOGY AND RESEARCH PROCESS
A review was undertaken of available information and procedures for
studying the status of rural domestic water conditions. This review pointed up
serious deficiencies in knowledge, conceptual maturity, and methodological tools.
A conceptual reformulation and development of data collection procedures was
undertaken by the Cornell research group. Members of this group developed the
interview schLdule in consultation with management and technical personnel at
EPA, and with various representatives from other governmental agencies, consult-
ing firms and research groups. The instrument was pretested and refined four
times, and the final version was submitted and approved by the Office of
Management and Budget.
-------
Executive Summary - 4
Data collection involved, at each selected household, personal interviews
with occupants, physical inspection of on-premises water supplies, drawing samples
of the major household water supply, various observations on sources of potential
contamination; and, where appropriate, separate interviews with managers and
operators of off-premises water supply systems which provided water to these rural
households.
TransCentury Corporation was selected by the EPA, after competitive bid,
to administer the data collection activities. Oversight and general coordination of
these activities was accomplished in coordination with the Cornell University
research group.
Interviewers underwent a two-week intensive training course on all aspects
of the data collection effort, including the drawing, packaging, labelling, and
transportation of specimens of “tap” water. Quality control of the field work
included constant and comprehensive monitoring. t ta coders were also inten-
sively trained and monitored.
Water samples collected at each household participating in the study were
shipped to central laboratories for analysis of as many as 40 separate biological,
physical, chemical, or radiological properties. Responsibility for these analyses
was shared by several organizations including the Energy Resources Corporation,
the Medical School of the University of South Carolina, the Mississippi State
Chemical Laboratory, the EPA-MERL Laboratory (Municipal Environmental
Research Laboratory) in Cincinnati, and the EPA-EMSL Laboratory (Environmental
Monitoring Systems Laboratory) in Las Vegas.
The sampling plan for the study defined the target population to conform
to the US Bureau of the Census’ definition of rural population and households:
civilian, noninstitutionalized persons located in unincorporated or incorporated
places of less than 2,500 population, or located in rural areas not designated as a
place. The sample was proportional to rural population (and households) in: (1) the
four broad census regions—Northeast, North Central, South, and West; (2) located
in areas designated as Standard Metropolitan Statistical Areas; (3) in rural
communities of 2,499 to 1,000 population, under 1,000 population, or open country.
The county was designated as the primary sampling unit. From the approximately
3,000 counties in the continental US, 400 were selected. Within those Counties, a
total of 2,654 households and their associated water supply systems were evalu-
ated. They represented an estimated 21,974,000 occupied rural households. Of
these, an estimated 8,765,000 households were served by individual, single connec-
tion systems; 2,228,000 households were served by 845,000 intermediate systems
(two -through fourteen connections); and 10,981,000 households were served by
34,000 Community systems (fifteen or more connections). Total counts of
households and systems were derived using standard statistical weighting pro-
cedures. Each unit’s weight, in broad terms, was the inverted sampling fraction
adjusted for nonresponse.
Collected data were compiled by the Cornell research group for statistical
analysis and report production. All aspects of the report underwent extensive
scrutiny by various review mechanisms.
-------
Executive Summary - 5
MAJOR FINDINGS
This study considered five dominant dimensions of the status of domestic
water: quality, quantity, availability, cost, and affordability. Rural residents were
asl
-------
Executive Summary - 6
Table I
Constituents Measured in NSA Survey
Measured in A Z!
Has Primary (P),
NSA Household
Secondary (S)
Samples or only
or
Category Constituent No (N) MCL
.
In Group U
Subsample
Microbial Total coUform P ALl
Fecal coliform N All
Fecal Streptococcus N All
Standard plate count N All
Fecal coliform/fecal -
streptococcus ratio N All
Physical Turbidity P All
arid Color S All
Chemical Temperature N AU
Specific conductance N All
Total dissolved solids
(as determined from
conductance) S All
Hardness
(as determined from
calcium and magnesium) N All
Inorganic Calcium N All
Magnesium N All
Nitrate-N P All
Sulfates S All
Iron S All
Manganese S All
Sodium N AU
Lead P All
Arsenic P Subsample
Selenium P Subsample
Fluoride P Subsample
Cadmium P Subsample
Mercury P Subsample
Chromium P Subsample
Barium P Subsample
Silver P Subsample
Organic Endrin P Subsample
Lindane P Subsample
Methoxychlor P Subsample
Toxaphene P Subsample
2,4-D P Subsample
2,4,5-TP P Subsample
Radioactive Gross alpha P Subsample
Gross beta P Subsample
Radtum 226 P
4 Radium 228 P
* Uranium P
Stontium-89 P
Strornium-90 P
*Ccsium -13 i
Tritium P
lodine-l3l P
Mezisured only if the laboratory analyst conidered gross alpha or gross beta
readings sufficient to warrant lurther Investigation.
-------
Executive Summary - 7
Among those contaminants which are covered by primary MCLs (because
they are potential health hazards), total coliform emerged as the most prevalent
problem, exceeding the reference value in 28.9 percent of all rural households.
Among constituents which tend to be aesthetically objectionable or impose
economic costs (e.g., degrading to plumbing, laundry, etc.), iron was the most
common.
The presence of total coliform organisms was the most common single
problem and received the greatest attention in the analysis. Coliform organisms
represent a problem in that they are used as indicators of the possible presence of
pathogenic bacteria. The rate of total coliform presence in excess of one coliform
per 100 milliliters of water (28.9 percent nationwide) was 15.5 percent among
households served by community systems, and over 40 percent among those served
by intermediate and individual systems (43.3 and 42.1 percent, respectively).
Households served by systems with fewer than fifteen connections that were not
wells (i.e., cisterns, springs, surface water, hauled, purchased bottled) had more
than one coliform per 100 milliliters of water in 77.7 percent of cases. In general,
households with low income (under $10,000) and low education (less than high
school) were more commonly found to have coliform problems than other house-
holds. Poorly accessible and privately owned supplies tended to have coliform
problems more often than other supplies. Households served by dug and bored
wells, wells in which the water leaves the casing above ground level, wells with
inadequate covers, inadequately maintained wells, and shallow wells all tended to
have high coliform levels more commonly than those served by wells without those
characteristics. Small systems, with few connections, repeatedly exceeded coil-
form levels more frequently than most large community systems.
Fecal coliforms were found among 12.2 percent of all rural households, but
among only 4.5 percent of households using community water systems. Fecal
-coliform counts were above 200 organisms per 100 milliliters of water-Ca suggested
upper limit in water used for swimming) for 1.6 percent of rural households.
Standard plate counts of more than 500 organisms per milliliter of water
were encountered at 19.3 percent of households. In this case, the percentage of
households with the problem, which used community water systems, was not
dramatically lower than for those using intermediate and individual systems.
Turbidity was measured above one NTU among 16.5 percent of rural
households. That rate varied from 23.8 percent in the North Central to 8.5 percent
in the West. Households using community systems were least likely to have more
than one NTU readings, 8.9 percent, compared to 24.0 percent for intermediate
systems, and 24.7 percent for individual systems.
Color was measured above fifteen standard color units among only 2.3
percent of all rural households.
Total dissolved solids, estimated from specific conductance readings, was
found above 500 milligrams per liter among 14.7 percent of all rural households.
Households in the North Central and West were over that level 23.9 and 22.2
percent, respectively. The Northeast, on the other hand, had only 5.0 percent of
households with levels above 500 milligrams per liter.
Magnesium was found above 125.0 milligrams per liter in only 0.1 percent
of rural households.
-------
Executive Summary - 8
Nitrate-N concentrations above 10.0 milligrams per liter occurred among
2.7 percent of households. Regionally, the proportion over that level was 0.3
percent in the Northeast, compared to 5.8 percent in the North Central.
Sulfates exceeded 250 milligrams per liter among only 4.0 percent of rural
households, but households in the West and the North Central were over at rates of
11.7 percent and 7.4 percent, respectively. The Northeast and South had household
supplies at that level in less than 1 percent of the cases.
Iron concentrations over 0.3 milligrams per liter were found in 18.7 percent
of all rural households. North Central households exceeded that level among 28.2
percent of households, compared to 7.0 percent in the West. Households served by
community water systems were over at the rate of 7.7 percent across the nation.
By contrast, the percent of households with iron concentrations over 0.3 milligrams
among intermediate and individual systems was 28.7 percent and 29.9 percent,
respectively.
The pattern of manganese occurrence was very similar to iron. Households
with more than 0.05 milligrams per liter occurred among 14.2 percent of the cases.
The rate in the North Central was highest (19.9 percent); and lowest (4.7 percent)
in the West. Households served by community systems were over that level at a
rate of 7.2 percent, compared to 23.3 percent and 20.7 percent, respectively,
among households served by intermediate and individual systems.
Sodium levels higher than 100 milligrams per liter appeared in 14.2 percent
of rural households nationwide. That rate was as low as 6.0 percent in the
Northeast, and as high as 19.2 percent in the North Central.
Lead was above 0.05 milligrams per liter at 16.6 percent of the households.
The percent of households exceeding that level was lowest in the Northeast (9.6
percent) and highest in the South (23.1 percent). It was discovered subsequent to
the data collection that a thin line of blue paint which marked the breakpoint on
the glass ampules of nitric acid preservative contained cadmium and lead. It is
likely that the average lead contamination due to the paint was 36 parts per billion
(0.036 milligrams per liter). Assuming relatively constant contamination from the
paint at that level, the estimated proportion of rural households over 0.05
milligrams dropped to 9.2 percent nationwide.
Arsenic levels in rural household water were encountered above 0.05
milligrams per liter in only 0.8 percent of all rural households. The preponderance
of those households were found in small rural communities (population less than
1,000). Households in those communities were found with the high arsenic levels
among 6.6 percent of the cases.
Selenium, nationwide, was found above 0.01 milligrams per liter among
13.7 percent of households. Most of that occurred in the North Central and West.
Households in the Northeast and South had high rates in roughly 2 percent or fewer
of the cases. In the North Central, that rate was much greater at 25.7 percent, but
in the West a remarkable 41.3 percent of all rural households had more than 0.01
milligrams per liter of selenium.
Fluoride concentrations exceeded 1.4 milligrams per liter in only 2.5
percent of all rural households. The rate in the West, at 6.2 percent, was more
than double the rate found in any of the other regions.
-------
Executive Summary - 9
Cadmium was measured above 0.01 milligrams per liter among 16.8 percent
of all rural households. Only 1.6 percent of rural households in the Northeast were
over that level while 27.1 percent of those in the West were high. Households
served by both community and intermediate systems had the high cadmium rates in
more than 20 percent of the cases (21.2 percent and 26.9 percent, respectively)
while 7.9 percent of the households with individual systems registered high. Here
again, however, the paint on the acid preservative ampules apparently imparted
some cadmium background to the readings. The average contamination was
probably in the neighborhood of .92 parts per billion (0.00092 milligrams per liter).
With that assumption, the national rate of high cadmium values dropped to 15.9
percent.
Mercury was found among 24.1 percent of all rural households to be in
concentrations exceeding 0.002 milligrams per liter. That proportion climbed as
high as 31.8 percent in the North Central, and dropped as low as 10.4 percent in the
West.
Chromium was virtually nonexistent above concentrations of 0.05 milli-
grams per liter throughout the US.
Barium Concentrations over 1.0 milligram per liter were rarely found (0.3
percent of all rural households).
Silver exceeded 0.05 milligrams per liter among 4.7 percent of all rural
households. Among the types of systems serving .rural America, community
systems had the lowest rates (2.1 percent), and individual systems had the highest
rates (7.1 percent).
Endrin, lindane, methoxychior, toxaphene, 2,4-D, and 2,4,5-TP were vir-
tually never detected among rural household water supplies.
Gross alpha radiation in excess of acceptable exposure occurred among 0.5
percent of rural households—predominantly in the South. Unacceptable gross beta
radiation was never encountered in the survey.
While these findings are startling, they must be kept in perspective. The
large percentages of rural households with high contamination levels was a function
of the level chosen to represent “high”—in this case, the primary MCLs. Again,
these levels should not be interpreted as the proportion of households with
domestic water exceeding an MCL since no resampling for verification, as required
for MCLs, was performed. The levels established as MCLs generally incorporate
substantial safety margins. So, even though these levels were identified as “high,”
and they were higher than was generally expected, they should be evaluated
Concurrent with the fact that widespread water-related health problems were not
apparent throughout the rural US.*
*The Environmental Protection Agency has participated with otI er agencies on
another national research effort on drinking water quality. The national Health
and Nutrition Examination Survey 0-lANES I) has an augmentation survey of adults
aged 25 to 74 years in the United States in 1974-1975. The HANES I study differed
from the rural water survey in its purpose, in its statistical sampling plan, in the
-------
Executive Summary - 10
Of course, the long term effects of waterborne contamination at currently
encountered levels is not known. Even the relatively short-run impairment of
health from what it would be without the ingestion of these substances is unclear.
Rather than being clear indicators of possible health effects, these high percent-
ages of households with water quality problems, especially coliform bacteria,
reflected the marginality of rural water supplies against the backdrop of the
primary MCLs mandated for community water systems.
In terms of general water quality, considering all constituents, households
in the North Central had the poorest quality water followed by households in the
West. Households in the Northeast tended to have the best water quality.
Households served by community systems tended to have the best overall
water quality, followed closely by individual systems (usually wells), with those on
intermediate systems generally having the poorest quality. Households in SMSA
areas and those in large communities tended to have better quality water because
they had higher proportions of households on community water systems.
WATER QUANTITY
Relatively few rural households across the US reported a shortage of
domestic water. Perhaps because of the essential nature of water, most rural
households have made arrangements for adequate quantity, on a regular basis, to
satisfy most or all their needs. Moreover, the perception of occupants at most
rural households was that their water supplies were ample: about 0 percent
reported that the major household supply completely satisfied their water require-
ments, and another approximately 16 percent reported that it usually or almost
always provided sufficient water.
Though most households had adequate quantities of water, a significant
number did not. An estimated 700,000 households reported that their supply
usually or always provided an insufficient quantity. Most often the insufficiency
was attributable to deterioration or inadequate construction of the physical
facilities. However, at the extreme, roughly 370,000 rural households hauled water
from an off-premises supply, on a regular basis.
water collection, in the water preservation, and in the laboratory analytical
techniques. BLt some of the contaminants examined were the same. In general,
the HANES I Augmentation Study results for metals showed lower concentrations
than the rural water survey. Various inquiries into the differences have provided
no satisfactory resolution.
The Augmentation Study Survey of adults aged 25 to 74 years in the United
States of the national Health and Nutrition Examination Survey I (HANES I) is
currently in preparation and will be published jointly by the National Center for
Health Statistics and the National Heart, Lung, and Blood Institute.
-------
Executive Summary - 11
Among households not connected to community systems, problems of
insufficient quantity were more commonly experienced by those supplied by small
systems (two through fourteen connections) than by those with their own individual
system. Among households with their own supply, those with dug wells more
frequently experienced insufficient quantity.
Among rural households connected to a community water system, the
median daily household consumption was 664 liters (175 gallons). The median daily
per capita consumption was 227 liters (60 gallons).
Consumption of domestic water was not uniform across the US. Western
households had a consistently higher consumption pattern than rural households in
other regions. No table differences occurred between households located inside
and outside SMSAs, or among places of different size.
Among households connected to community systems, locational factors
were less important in accounting for variability of consumption than: (1) unit cost
(cost per 1,000 gallons); (2) number of people in the household; and (3) number of
water-using devices in the household. Although of less importance, the afford-
ability of water and type of ownership of the system were systematically related to
variation in consumption. In general, the higher the unit Cost, other things
constant, the lower the consumption. As would be expected, the greater the
number of occupants and the greater the number and usage of water-using devices
(dishwashers, washing machines, swamp coolers, as well as toilets and bathing
facilities), the greater the consumption. Likewise, the lower the ratio of cost of
water to the household’s total income, the larger the consumption. Somewhat
unexpectedly, households connected to privately owned systems tended to consume
approximately 50 percent more water than those connected to publicly owned
systems, even after controlling for a number of other, potentially confounding
factors.
AVAILABILITY
Most rural households had a readily available domestic water supply.
Quantity referred to the usual amount of water. Availability, on the other hand,
was defined, for purposes of this study, in terms of two components: (1) the supply’s
reliability, or uninterrupted service, and (2) the supply’s accessibility-—one which
provided water, under sufficient pressure, when needed and for which the source
was not at an inconvenient distance relative to the point-of-use by household
occupants.
A total of 5.6 million rural households (roughly 26 percent) experienced
water supply breakdowns during the year preceding the interview. Among those
households, the majority experienced only one or two interruptions of service.
About 15 percent of these households (3.2 million) indicated the breakdowns lasted
more than six hours.
In terms of regional differences, household supplies in the South seemed to
be somewhat less reliable than in the other regions. Although the reliability of
SMSA and nonSMSA supplies was similar, there were differences according to the
size of place classification. Generally, households in small rural communities
-------
Executive Summary - 12
reported supply breakdowns more frequently than households in large rural com-
munities or other rural areas. Similarly, water supply breakdowns which were
considered severe—lasting more than six hours—occurred more often among
households situated in small rural communities.
Compared with households using individual or community systems, a larger
proportion of households served by intermediate systems reported one or more
supply breakdowns. In addition, a greater proportion of households served by inter-
mediate systems reported breakdowns that were severe.
Regarding accessibility, sources were generally located at convenient
distances from the household. Among rural households connected to community
systems, Connections involved only piping from the edge of the property to the
house, generally a modest distance. Moreover, among households with their own,
on-premises supply, over half were located within 10 meters (33 feet) of the house.
However, comparatively greater distances were recorded at approximately 15
percent of the households—implying the need for more extensive piping or other
mode of conveyance. For the approximately 370,000 households that had to haul
water, accessibility imposed a far greater relative hardship.
Supplies were less accessible in the South and West than in the Northeast
and North Central. Though accessibility of SMSA and nonSMSA supplies was
similar, households located in open country were relatively less accessible than
those located in villages.
COST AND AFFORDABILITY
The cost of water was not estimated for rural households using their own
supply or for those connected to a system that didn’t explicitly charge for water on
a regular basis. Assessment of the cost of water was restricted to rural households
served by community systems which had a billing system. Perceptions of cost,
however, were assessed for all rural households.
Compared to many other consumable items in the US economy, water is
inexpensive. The median household monthly cost per thousand gallons was $1.35
nationally. The median total monthly bill for water was $7.00. A ratio of billed
cost to total household income (times 100), as a measure of affordability, indicated
that three-quarters of all rural households were paying less than 1 percent of their
income for water. Across the nation, water was found, at the extremes, to be as
inexpensive as $.08 and as expensive as $23.41 per thousand gallons. cupants in
the majority of rural households (79 percent) felt the water was reasonably priced
or inexpensive. Domestic water was perceived as expensive or very expensive at
about 14 percent of households.
Costs of domestic water were not uniform across the US. Regionally the
median cost per thousand gallons ranged from a high $2.00 in the West to $1.33 in
the South. Households within SMSAs had lower water costs than those located
outside SMSAs. The respective medians were $1.08 and $1.62. Median water costs
vai ied only slightly according to the size of community in which the household was
located.
-------
Executive Summary - 13
Despite these findings, a greater proportion of rural households in both the
South and West felt the water was expensive or very expensive. A greater
proportion of residents living within SMSAs reported their water costs were high
than was reported by those living outside SMSAs. Likewise, perceived costs tended
to rise the larger the size of place.
In addition to locational differences, several other factors were found to be
systematically related to household water costs. Regarding the amount billed
monthly, the single most important factor was, as expected, the amount of water
consumed. After adjusting for this factor, it was still discovered that total
monthly costs increased with the number of household occupants, and the education
level of the head of the house. Households connected to privately owned systems
paid more than those connected to publicly owned systems. On the other hand, the
longer the household had been connected to the system, the lower the total
monthly cost.
It was found that lower unit costs (price per thousand gallons) were
associated with (1) the size of the water system providing the water, (2) the
system’s source water, and (3) the amount of water consumed at the household.
Generally, the larger the system, the lower the unit cost. Systems using ground
water usually had lower unit costs, probably due to the lowered treatment and
storage costs. Some price break resulted for households with large monthly
consumption.
WATER SYSTEMS
In this study, water systems were examined in order to specify technical,
economic, and organizational features associated with delivering water to rural
households. Rural water systems were classified .according to three major-system-
size categories: individual systems (single connection), intermediate systems (two
through fourteen connections), and community systems (fifteen or more connec-
tions). The community systems were further classified as independent (with self
sufficient features) versus consolidated systems (with operating or organizational
features integrated with other systems or administrative bodies).
Overall, individual systems were by far the most prevalent, numbering
8,765,000. In contrast, there were 845,000 intermediate systems and only 34,000
community systems. Because of this multiple-connection feature, however,
intermediate systems and especially community systems were far more important
in rural water delivery than implied by their numbers. When taking into account
the number of rural households served by-each type of system, community systems
were most prominent (serving 10,981,000 households), followed by individual
systems (serving 8,765,000 households), and intermediate systems (serving
2,228,000 households). See Figure 1.
Community water systems were defined for purposes of the study as those
having fifteen or more connections. The standard EPA definition of public water
systems is somewhat different. It refers to fifteen or more connection systems
used by year-round residents or to systems which regularly provide service to 25 or
more year-round residents. The NSA study’s definition yielded a count of 34,000
community water systems. Had the study’s definition been altered to ten or more
-------
Executive Summary - 14
Figure!
Comparison of Number of Systems and
Numbers of Households Served in Rural America
SYSTEMS
(In millions)
SERVING
22
20
18
16
I
I
I
I
1
I
COMMUNITY-j ,‘
$0 ________
4
2
to
8
6
4
2
0
HOUSEHOLDS
(In millions)
-------
Executive Summary - 15
connections, the count of community water systems serving rural America would
have been about 47,200. The latter count is probably much closer to the standard
EPA definition.
Individual Systems
Of the 8.8 million individual systems in rural America, 90 percent relied on
wells. Other types of supply—surface water, springs, cisterns, hauled and
purchased supplies—comprised the balance of individual systems across the coun-
try. These supplies were more often associated with poor quality water.
Nearly six of every ten individual wells were installed by drilling, which is
one of the more highly recommended methods of installation because of the usually
good contamination protection it offers. Dug or bored wells, which were often
linked with bacterial contamination, represented nearly 18 percent of all individual
wells. Individual wells were generally located 50 feet or more from most potential
contamination sources, with the exception of sewage disposal systems. About one-
fifth of all wells were within 50 feet of sewage disposal. Field inspection classified
62 percent of rural wells as appearing to have weliheads which were adequately
sealed from solid or liquid pollutants.
Bacteriological tests and chemical (or physical) water tests by rural
residents were not common. Slightly more than one-third of all rural households
with individual systems had tested the water at least once, with bacteriological
tests being more frequent than chemical tests. Test results, when they could be
recalled, were generally classified broadly as “acceptable.” But, nearly 45 percent
of the bacteriological test results and nearly two-thirds of the chemical results
were not known. This lack of knowledge may be attributable to the respondent’s
lapse of memory, poor filing practices, the .length of time since the .test was
completed, or may have implied incomplete or uninterpretable reporting of test
results.
Treatment practices can in many situations alter the quality of the source
water provided to rural households. Among individual systems, the most common
treatment device was a water softener (used in 18 percent of systems). Other
water treating devices were rarely found.
Intermediate Systems
Intermediate systems were similar to individual systems. Over 90 percent
had only two or three connections and less than 2 percent were metered. About 88
percent relied on wells, the majority of which were installed by percussion or
rotary drilling (65 percent). Dug and bored wells comprised 17 percent of all
intermediate systems wells.
Intermediate system wells were generally not located close to potential
sources of contamination, except for roughly 16 percent which were within 50 feet
of a sewage disposal system. Field inspection classified 67 percent of intermediate
system wells as acceptably sealed against pollutants.
-------
Executive Summary - 16
Water quality tests were performed less frequently among intermediate
systems than individual systems. Bacteriological testing had been done at least
once during the lifetime of the system at 22 percent of systems. Testing for
chemical properties of the water had been done at 10 percent of systems. In
contrast to individual systems, among those intermediate systems which tested,
owners could usually provide information about the testing history as well as the
test results. Of the bacteriological tests performed, over 91 percent showed
acceptable counts of bacteriological content. As to treatment practices, among
inter mediate systems, softening—the most common treatment process—occurred
in only 6 percent of the systems.
Information on average daily use and maximum daily (design) capacity
collected at water systems provided some indication of the quantity of water
provided to rural households connected to those systems. About one-third of
intermediate systems had information on production and design capacity. Among
those with readings, the median for average daily use was 758 liters, or about 200
gallons. The median design capacity was 44,000 liters (11,600 gallons). The median
of the maximum daily design capacity was nearly 60 times the median “average
daily usage” level among intermediate systems serving rural households. This
difference suggests that some intermediate systems, similar to larger community
systems, overbuild to a certain degree, allowing for expansion, fire protection, or
increased consumption.
Availability indicators included the number and severity of system break-
downs, and the ease or difficulty in acquiring water from the source. About one-
quarter of intermediate systems reported breakdowns of any kind, and few reported
any difficulty obtaining water from ground or surface sources. The majority which
did report breakdowns had experienced only one or two during an entire year.
Despite these favorable findings, a very high proportion of households connected to
intermediate systems reported reliability problems in the sense that nearly all
breakdowns resulted in a loss of service to water users.
Most intermediate systems did not report revenues or any monetary
charging structure for the water service they provided. This fact suggested that
water was very cheap for intermediate system customers, but may have also
indicated that system owners bear the entire cost of installation, operation and
maintenance.
Independent Community Systems
Independent community systems were self-contained systems which
secured water from their own sources, treated the water, distributed it, and
managed their own organizational and financial affairs. They comprised about 88
percent of the 34,000 community systems in rural America.
Although independent community systems had more extensive extraction,
treatment, and conveyance arrangements than interme’liate systems, they still
were very small in comparison to the ‘arger, more complex systems such as those
based in large cities. The median numbe of connections was 59 and the median
length of distribution lines was 1.5 mile . About half of these systems metered
their connections. Similar to smaller ystems, the predominant source of water
-------
Executive Summary - 17
was ground water (90 percent). The labor-time demands involved in operation and
maintenance of the system were modest, with many requiring two or fewer man-
days of labor per month.
Water testing and treatment were much more frequent and extensive at
independent community systems than at systems with fewer than fifteen connec-
tions. Fully 91 percent had conducted bacteriological tests and about 54 percent
had done chemical testing. Treatment equipment was installed at nearly two-
thirds of all independent systems. Most independent systems relied on facilities
outside their own organizations for performing water quality tests. For the most
part, bacteriological tests conducted during the year prior to the study (1977)
reportedly met the levels prescribed by regulations affecting the systems. In
general, there were too few chemical and physical tests performed to suggest
meaningful patterns.
About 63 percent of independent community systems provided average
daily usage and design capacity figures. Among those which reported, the median
for average daily usage was 136,000 liters (36,000 gallons). The median design
capacity was 546,000 liters (144,000 gallons). Maximum daily production capacities
among independent systems were four times larger, on the average, than the
estimated average daily use. Though this ratio was less than for intermediate
systems, it was large enough to accommodate fire protection, increased usage, and
increasing the number of connections because of the larger volume being produced
daily.
Independent community systems were less reliable than intermediate
systems. A larger proportion of independent systems (55 percent, compared to 26
percent) had breakdowns, although fewer resulted in an interruption of service to
customers. As was true for intermediate systems, few had difficulties iri obtaining
water from any source.
Consolidated Community Systems
The consolidated community systems relied on interdependent links with
other organizations. One type of consolidated system had a number of separate,
individual facilities of which each served one community, but which were jointly
owned and administered by a central company. Another type consisted of a facility
which purchased water from another organization—a situation which was the
distinguishing feature at nearly 90 percent of the consolidated systems.
The consolidated systems tended to be the largest of the systems serving
rural America. The median number of connections for the 4,000 consolidated
systems was about 153—two and one-half times the median number for indepen-
dent community systems. Likewise, consolidated systems employed more system
operators and metered a larger proportion of all connections.
Testing for water quality was more extensive at consolidated community
systems than at independent community systems. Bacteriological testing was done
at almost all (98 percent) consolidated systems, and chemical testing was per-
formed at about 90 percent of systems. Results for bacteriological tests were not
uniformly interpretable. But, about 96 percent of the systems reporting tests had
acceptable results for either the previous twelve tests or for those tests conducted
-------
Executive Summary - 18
over the last year. The variety of Constituents tested, disparities in the age of the
tests, and the lack of information on analytic procedures combined to render the
chemical testing results uninterpretable.
Similar to independent community systems, consolidated community
systems could produce, on the average, over four times the quantity of water that
was consumed on a daily basis. The median for average daily use was 162,000 liters
(43,000 gallons) among the 88 percent of consolidated systems which had the
information. Design capacity was obtained for half of the consolidated systems.
The median design capacity was 719,000 liters (190,000 gallons).
About 50 percent of consolidated systems had breakdowns during the year
preceding the study. This proportion was slightly smaller than for independent
systems (55 percent), but still considerably greater than for intermediate systems
(26 percent). This finding could be interpreted to mean that problems of reliability
were more often associated with larger, more complex systems. However, it was
also the case that these larger systems more frequently had alternatives in the
event of some malfunctions (such as auxiliary pumps, storage capacity sufficient to
cover demand during a breakdown, and so forth).
-------
Executive Summary - 19
CONCLUSIONS
I. Most rural households had a domestic water supply which was acceptable
by most of the quality indicators used, was in sufficient quantity to meet
consumption demands, was readily available on a continuous basis, had a reasonable
cost, and was affordable. But, the exceptions were not rare and the difficulties
were not always minor.
2. The overwhelming majority of rural households had water judged to be of
acceptable quality for any particular characteristic studied. But, almost two-
thirds of all rural households had water judged unacceptable for at least one of the
constituents which have been given primary MCLs (excluding turbidity). Bacterial
contamination, in particular, was the predominant problem encountered.
3. The quality of the water available in most rural households was due to
numerous factors, but a few can be highlighted. Most rural households (on the
order of 90 percent) relied on ground water. High quality well technology and well
construction practices were generally evident throughout rural America. Their
employment apparently helped minimize alteration of the water from the quality it
had at the source until it was delivered to the tap.
4. Larger, more frequently monitored community water systems generally
delivered water of higher quality than smaller and less frequently monitored
systems.
5. By. comparison with community systems, individual and intermediate
systems employed less complex technologies and less frequently used water
treatment devices. Therefore the quality of the water delivered to the household
was mainly a reflection of the quality of the source water (which was ground water
for 90 percent of these systems). As long as the fundamental quality of the ground
water source remains intact, further improvements in well technology and con-
struction practices would beneficially affect the quality of delivered water by
ensuring the protection of the quality from source to tap. A larger benefit would
probably result from presently deficient wells being replaced or brought up to
current standards of good practice for well construction. Direct regulation of
small (usually individual) water systems would probably not dramatically alter
water quality because of the common lack of treatment devices and the improb-
ability of their installation due to the relatively large capital costs for the
individuals involved.
Therefore, perhaps the greatest protection for the continuing quality of
water from noncom munity water systems would be the preservation of ground
water quality, particularly the freedom from bacterial contamination. Acceptable
well construction practices for households using individual and intermediate wells
appeared to reduce bacterial contamination potential. Proper siting of wastewater
disposal, or wastewater removal by public sewer systems were also associated with
lower bacterial contamination for households served by individual systems. Addi-
tional improvement in water quality among water systems with fewer than fifteen
-------
Executive Summary - 20
Connections might result from wider availability of water testing services and
public education on their availability, their importance, and on possible response
alternatives to identified problems.
6. The more extreme the rural householder’s evaluation of the water supply,
whether good or bad, the better was the correspondence with the composite
indicators of laboratory-measured quality. Households which were very positive
about their supply usually rated well on the laboratory measures. Households in
which the water supply was given a poor rating also tended to have laboratory-
measured deficiencies.
Responses to questions regarding the relative cost of the supply and the
respondent’s willingness to pay more for an improved supply were both useful
indicators of the quality as measured by the laboratories. In general, households
reporting low relative costs or strong willingness to pay more for an improved
supply, tended to have poor water quality according to composites of laboratory
measures. Households which reported their supply was expensive or where the
respondents lacked a willingness to pay more for an improved supply tended to have
superior quality water according to the laboratory indicators.
7. It was noted that average daily consumption of water was greatest in the
West. This was so even though the unit cost of water was highest in the West.
Rural households of the West generally relied upon their major household supply—
usually a community water system—for virtually every water need around the
home, yard, and garden.
Over the last fifty years or so, there has been a federal subsidization of
Western water through low interest construction loans on large scale reclamation
projects. Water cost comparisons among large Northeastern cities and large
Western cities have frequently resulted in the observation that Western cities often
charged far less for domestic water than those in the Northeast. That effect was
not borne out by the NSA for systems serving rural Westerners. Most of the
community water systems in the West were relatively small and they were charging
on the average the highest rates among systems serving rural America. (Possibly,
there is an important economy of scale related to the acquisition and transport of
the water. Large systems are more likely to be primary customers, that is, to have
a direct arrangement in terms of payment and piping with large scale reclamation
projects. Smaller systems, on the other hand, may end up being secondary
customers which buy from large systems at some higher price. But even when a
small system obtains its water directly from a large scale reclamation project, or
from a smaller local project, or from a deep well, the relative cost per connection
would generally be higher than that experienced by large systems. Thus, even with
the federal subsidization of much of the developed water in the West, the average
unit cost for water among rural Western households was higher than for other rural
parts of the country.)
8. Intermediate systems were found to provide generally inferior service
compared to community and individual systems. It is suspected the reasons for this
lay with the nature of the design.
-------
Executive Summary - 2!
Apparently most of these small systems were not originally designed or
intended to be integrated, multiple-connection systems, but were originally
installed as individual systems to which additional households were later connected.
For instance, a relative, hired hand, or friend could build a home adjacent to a
house with an existing individual supply. Because of the cost of installing a new
system, occupants of the new household might ask, or be invited, to hook up to the
existing well rather than install a separate facility. It appeared that many small
intermediate systems seem to have started as individual systems and evo!ved to
become two and three connection systems without any accompanying redesign.
Many of these extended connections were probably homeowner installed.
The result of these extensions and modifications to the original system is a
strain on the capacity of the system. These apparently stressed systems provided,
as a group, the poorest overall service of any system configuration studied. A
higher proportion of households connected to intermediate systems experienced
problems with water quality, they more frequently had insufficient quantity, and
they tended to report a greater number of breakdowns.
The trend in many parts of the country toward falling ground water levels
and increasing well construction costs may increase the likelihood of more rural
households entering into multiple connection arrangements on individual systems
not designed for that purpose.
* * *
This executive summary captures in only the broadest form the detail of
the full, 1,900 page, report. The five volume document describes attributes of
rural water users, households, supplies, and systems as they related to the quality,
quantity, availability, and cost of rural domestic water conditions. Each topic is
explored for the nation as well as for various subnational categorizations. The
study is a one-time effort which strives to be comprehensive, rather than
exhaustive on any particular topic. The report is the first systematic, nationally
representative examination of the broad issues related to rural domestic water.
-------
I Summary
As explained in the Preface, topics in this study range from domestic water
conditions in rural households to economic conditions at water systems serving the
households. This information is kept in perspective by reference to demographic
data regarding rural households. The voluminous details of the study are organized
in a systematic plan which is described in the Preface and in Chapter II. The
purpose of this chapter is to condense the findings in order to provide a concise
overview. The reader is cautioned, however, that the intent of the overview is to
provide orientation to a vast amount of data. Of necessity, the focus is only on the
salient aspects of the study. Thus, the reader should avoid making definitive
judgments about NSA findings on the basis of this summary alone. Conclusions
should be based on careful readings of the text, where the NSA findings are
amplified, qualified, and fully explained.
The major findings in each chapter in turn are presented in this summary.
The sequence in which the findings are given is not necessarily that used in the
respective chapters, however. The sequence ometime5 is altered in this chapter
to facilitate comparisons and to aid readability, but the reader generally is advised
if this is done.
-------
1-2
SUMMARY OF CHAPTER II
The formal introduction to the NSA is provided in Chapter II. The
introduction traces the development of the study: its scope, uniqueness among
other studies about domestic water conditions, specific terminology, objectives,
sampling and data collection methodology, and its analytical structure. Here, the
focus is limited to aspects of terminology, methodology, and analytic procedure
which are of importance in understanding this summary.
As to terminology, certain terms which have broad connotations in other
studies were defined more narrowly in the NSA. It was necessary to separate
source from water supply. “Source” was defined as any physical aggregation of
water prior to human manipulation. “Water supply” in the NSA was defined as the
set of facilities by which water was extracted from the source, treated, and
transported to the household—or as the water itself as it reached the household.
“System” referred to the social and organizational aspects of water supply, and was
represented by three specific types: individual systems, which- served a single
household; intermediate systems, which had from two through fourteen connec-
tions; and community systems, which had fifteen or more connections (roughly
corresponding to community systems affected by federal drinking water regula-
tions). The community systems were further classified as either “independent”
(generally self-sufficient) or “consolidated” (relying on certain interdependent links
with other organizations).
It also was necessary in the NSA to distinguish between a household’s
drinking water supply and its major water supply. The drinking water supply was
that one used exclusively for drinking, while the major supply was defined as that
one which satisfied most of the household needs (whether it was used for drinking
or not). In fact, it was discovered that in most rural households (about 97 percent)
-------
1-3
the major supply and the drinking water supply were the same, but the distinction
was important for analytic reasons described in Chapter II. It also was necessary in
the NSA to develop specialized connotations for six major aspects of water
conditions—quantity, quality, availability, potential health hazards, cost, and
affordability. Those connotations will be mentioned as needed in this summary, but
the full discussion of their meaning in the NSA is presented in Chapter II.
The results presented were for two types of entities: rural households and
supply systems. Some results pertained to rural households, the occupants, and the
status of the water they used. Other results pertained to the systems which
provide water to these households. For purposes of this study, rural households
were defined as those located in places less than 2,500 population or in open
country.
Approximately 2,600 rural households were selected by a carefully con-
trolled stratified, proportionate probability sampling design. These households and
the approximately 1,100 systems providing their water underwent intense observa-
tions and were the basis for all findings presented.
As to the NSA analytic procedure, findings were assessed first, for the
national as a whole. The findings then were assessed, as appropriate, according to
four subnational groupings. The first subnational grouping was by four regions as
determined by the US Bureau of the Census (Northeast, North Central, South,
West). The second grouping was by SMSA areas versus areas outside of SMSAs.
The third grouping was by size-of-place: large rural communities (places where the
population was 1,000 through 2,499); small rural Communities (places where the
population was 999 or less); or other rural areas. The fourth grouping was by size
of system (individual, intermediate, and community, as described above). The last
fot r chapters present systematic relationships between each of the major dimen-
sions of rural water supply considered in this study and selected attributes of the
household, its supply, and system.
-------
1-4
SUMMARY OF CHAPTER III
The rural household was the focal point for the collection of data in the
NSA and the terminal point for domestic water usage. Because of this, it was
important to develop a clear picture of the general socioeconomic conditions in
rural households, the nature of rural domestic water use, the types and number of
water-using appliances, and the residents’ general attitudes toward the water
supplies. This information, presented in Chapter III, provided the framework for
evaluation of the NSA results.
Another key aspect of the NSA report which is described in detail in
Chapter III, and analyzed systematically in Appendix B, is the use of projected data
for all rural households in the US. As is necessary in this type of national survey,
the projections in the NSA were made on the basis of investigations at a
scientifically selected sampling of rural households. Based on projections from the
NSA sample, it was estimated that at the survey’s midpoint (September 1978),
there were approximately 21,974,000 occupied housing units in rural America.
About 31 percent of Americans lived in rural households in 1978. Rural
households in the US were located predominately in the South and North Central:
42.3 percent of the households were in the South, 28.3 percent in the North
Central. Classified in a different way, 68.0 percent of rural households were
located outside of SMSAs, 32.0 percent within SMSAs. Consistent with the large
proportion of households outside of SMSAs, about 83 percent of households were
located in other rural areas, as opposed to much smaller proportions in large and
small rural communities.
-------
1-5
DEMOGRAPHIC PROFILE OF RURAL US
Summarized first are data about structural and occupancy characteristics
of rural dwelling units. Presented next is demographic information about the head
of household and about household members.
As to types of housing, the overwhelming majority of rural households
(about 81 percent) were in single-family, conventional homes. About 12 percent of
the households were in single-family mobile homes. The rest were in multi-
dwelling unit structures.
The highest proportion of single-family housing units (86.2 percent) was in
the North Central, the lowest (77.0 percent) in the West. In contrast, the South
had the largest proportion of mobile homes or trailers (16.3 percent), whereas the
Northeast had the smallest proportion (5.7 percent) of mobile dwellings.
As .to ownership of the households, about 79 percent were owner-occupied,
about 18 percent were renter-occupied, and the rest were occupied by people who
had other arrangements—e.g., provided by parents, occupied rent free. Ownership
arrangements involving farms associated with rural households also were studied in
the NSA. Accordingly, about 3.7 million (about 17 percent of rural households)
owned or leased farms, which were numerically most prevalent in the South and
North Central. However, of the 3.7 million households which owned or leased a
farm, 960,000 had no income from sale of farm products.
Regarding length of residence at the households, median length of resi-
dence for heads of rural households was 5.2 years nationally. Regionally, the
median was shortest in the West (3.8 years), reflecting the general pattern of
- recent migration to the Western states. In contrast, median length of residence
was greatest (6.3 years) iii the Northeast. In SMSA households, the median length
of residence was 4.1 years, ccrnpared to 5.9 years in noriSMSA households.
As to the number of prople in rural households, the NSA findings did not
support the traditional assumption that rural families are large. In fact, nearly
-------
1-6
half of all rural households had only one or two members; the median number of
households occupants was 2.65.
For the most part, men were reported to be the heads of rural households.
On the other hand, women were designated heads in 18.7 percent of rural
households, a 3.8 percent increase since the 1970 census.
The household heads were predominantly white—only 8.7 percent, or 1.9
million, were nonwhite. Their median age was 44 years, five years younger than
the median age at the time of the 1970 census. The household heads had completed
a median of twelve years of education, indicating a trend toward increased
education since 1970. Only 2.8 percent of household heads were unemployed (70.5
percent were working full or part time the week before the NSA interview, 21.1
percent were retired, and 5.5 percent were in some other employment category).
In 1978, the Northeast had the highest proportion of rural heads working full or
part time (78.9 percent) and the West had the lowest (66.9 percent). Rural
households in SMSAs had a higher proportion employed (78.8 percent) than
nonSMSAs (66.7 percent).
Of the approximately 15.8 million working household heads, about one-third
were employed in professional, technical, managerial, or administrative occupa-
tions (excluding farming). Another 20 percent were in crafts or related occupa-
tions. Only 7.3 percent described themselves as farmers or farm managers, and
only 1.3 percent said they were farm laborers or foremen.
The income available to rural households as a result of a nonfarm
occupation or farming may have had an influence on the chosen means for
providing basic needs like drinking water. Therefore, NSA analysts attempted to
determine total combined income of everyone in the household from nonfarm
sources. In addition, for those households that also owned or leased a farm, gross
farm income (sales) and net farm income also were tabulated. However, evaluation
of the household income__Particularly farm income—involved qualifications which
-------
1-7
are beyond the scope of this summary. Here, the focus is limited to a brief
summary of the findings for total household income and for household nonfarm
income. Thus, the median total household income for rural households during the
year preceding the NSA was $12,818—lower than the median for the US as a whole
in 1978 ($13,512). Median nonfarm income was $12,676. Regional, SMSA versus
nonSMSA, and size-of-place differences were similar for total household income
and rionfarm income, and the data for nonfarm income are cited here for
convenience. The subnational differences sometimes were prominent. For
example, median norifarm income was greatest in the Northeast and West (about
$15,000 in these regions), smallest in the South and North Central (about $12,000 in
these regions). Median nonf arm income was $18,000 in SMSA households, compared
to $12,000 in nonSMSA households. Median nonf arm income f or households in other
rural areas was $14,000, compared to $10,001 for households in large communities,
and to $10,995 for households in small communities.
PROFILE OF USES OF WATER
Shifting now from demographic data to patterns of household water use,
NSA data showed that almost all rural households (about 97 percent) used their
major water supply f or drinking. Five other indoor uses of the major water supply
(cooking, bathing, flushing toilets, washing clothes, and washing dishes) were
reported by at least 90 percent of all rural households (each household did not
necessarily report all uses), but other indoor uses (such as evaporative cooling)
were uncommon. Among outdoor uses of the major supply, the most significant
were periodic cleaning (of windows and cars, for example), reported by 80.6
percent of rural households; watering lawns and gardens (66.2 percent); water
livestock (12.0 percent); and recreational uses (7.6 perce. t).
In contrast to indoor major-supply use, outdoor uses showed substantial
regional variation, with households in the West reporting a variety of outdoor uses
-------
1-8
much more frequently than households in other regions. In the SMSA/nonSMSA
comparison, SMSA households, more frequently than nonSMSA households, reported
two uses—watering lawns and gardens and recreational use. In the size-of-place
comparison, only one outdoor use showed appreciable variation—watering live-
stock, which was reported almost exclusively by households in other rural areas.
Two uses—watering lawns and gardens, and periodic outdoor cleaning—were
reported most often by households served by Community systems and least often by
households served by intermediate systems. Having no outdoor uses for the major
water supply also was most Common among households served by intermediate
systems.
WATER-USING EQUIPMENT AND WASTEWATER SYSTEMS
Closely related to household uses of water are the use of certain water-
using fixtures and of wastewater systems. About 95 percent of all rural households
in the US had complete plumbing facilities as defined by the US Bureau of Census
(piped hot and cold water, flush toilet, and bathtub or shower). However, the
regional proportions of complete facilities varied from 98.8 percent in the West to
91.5 percent in the South.
In a more specific tabulation of water-using fixtures and devices, it was
found that at least 95 percent of all rural households had all of the five most
important fixtures: kitchen sink, bathing facility, flush toilet, bathroom sink, water
heater, or shared access to such facilities. As with the data on dompleteness of
plumbing facilities, these data showed substantial regional variation. In general,
the West’s rural households were the most completely equipped, and the South’s
rural households least completely equipped.
With regard to wastewater systems, almost two-thirds of rural households
used either a septic tank, cesspool, or similar arrangement for disposal of
-------
1-9
wastewater, while about 28 percent were connected to public sewer systems.
Direct discharges of wastewater were made at about 6 percent of rural households.
The West had the greatest proportion of households using public sewer
systems (42.7 percent), and the smallest proportions of septic tank and similar
arrangements (55.2 percent) and direct discharges (1.5 percent). The Northeast had
the greatest proportion of rural households using septic tanks and the like (74.7
percent), and the smallest proportion using public sewer systems (22.6 percent).
Direct discharges were most common in the South, occurring at 9.7 percent of
rural households—more than double the proportion in any other region.
Only 23.1 percent of nonSMSA households used public sewer systems,
compared to 39.3 percent of SMSA households. Conversely, 67.4 percent of
nonSMSA households used septic tanks, cesspools, and the like, compared to 57.6
percert within SMSAs. Direct discharges were far more common among rionSMSA
households (8.2 percent, compared to only 1.6 percent of SMSA households).
The size-of-place comparison showed quite different patterns of waste-
water disposal. Public sewer systems were used by 71.9 percent of households in
large communities, 50.0 percent in small communities, but by only 20.8 percent in
other rural areas. Septic tank and similar arrangements revealed almost the
opposite pattern being used in 25.8 percent of households in large communities,
44.0 percent of households in small Communities, and 70.9 percent of households in
other rural areas. Direct discharges, very infrequent in large rural communities,
occurred at 4.8 percent of households in small rural communities and 6.1 percent of
households in other rural areas.
About 87 percent of households served by individual or intermediate water
supply systems used septic tanks or cesspool arrangements. More than half of the
households served by community systems (54.3 percent) used public sewer systems,
while 41.5 percent used septic tank arrangements. Direct discharges occurred at
9.0 percent of households served by individual water supply systems, compared to
-------
I - 10
6.4 percent of those using intermediate systems and 2.5 percent using community
water supply systems.
Reliability of the wastewater systems was assessed in the NSA according
to household experience with breakdowns. Breakdowns in wastewater systems had
occurred in the year prior to the NSA at 6.1 percent of all rural households in the
US (1.3 million). For the most part, the breakdowns were limited to only one or
two occurrences. At 4.3 percent of all rural households, the breakdown was severe
in that it lasted for six hours or longer. At nearly two-thirds of the households
with breakdowns, the incidents involved septic tanks or similar arrangements;
breakdowns at about one-quarter of households involved toilets, pipes, or related
equipment; breakdowns at 7.4 percent of the rural households were associated with
public sewer facilities.
ATTITUDES TOWARD WATER SUPPLIES
The final section of Chapter III describes household attitudes toward water
supplies. Specifically, the NSA asked a series of questions about liked and disliked
supply characteristics, respondents’ general satisfaction with their water supplies,
and possible preferences for changing to a different supply.
About two-thirds of respondents identified a specific characteristic of the
water supply which they liked—most often, taste. About 44 percent identified a
specific characteristic which they disliked—again, taste was mentioned frequently,
along with hardness. Regional variation was not striking, though temperature was
mentioned as a liked characteristic most often in the West (14.8 percent) and least
often in the South (5.6 percent). Hardness was disliked at about 18 percent of rural
households in both the North Central and West, compared to 5.0 percent of
households in the South.
In the size-of-place comparison, the water supply’s taste was liked far
more often at households in small rural communities and other rural areas than at
-------
I — 11
households in large rural communities. The temperature was liked far more often
in other rural areas than in large or small communities.
The size-of-system comparison showed greater variation than occurred in
the other subnational categories. Specifically—in comparison to households served
by individual or intermediate systems—a much smaller proportion of households
served by community water systems liked the water’s taste (21.5 percent) and
temperature (4.4 percent), though a much larger proportion liked the supply’s
reliability (11.0 percent). Conversely, taste was mentioned as a disagreeable
characteristic more than twice as frequently among households served by commun-
ity systems (13.8 percent). Hardness was mentioned as a disagreeable characteris-
tic most often among individual-system households (15.3 percent).
When respondents were asked to rate their Current water supply against an
ideal supply, 72 percent reported their present supply to be essentially equal to the
idealized supply. However, in over five million households (23.9 perCent), the ideal
supply was rated superior to the existing one, while the remaining 4.1 percent
(900,000 households) rated their present supply superior to the idealized one. The
proportion of households rating their present supply equal to the ideal one varied
substantially from region to region, with the largest proportion so rating their
system in the North Central (76.2 percent), and the lowest proportion in the West
(68.9 percent). Correspondingly, there was regional variation in the proportion of
rural households rating the idealized supply superior, which ranged from a low of 20
percent in the North Central to a high of 27 percent in the West.
Over four-fifths of the rural households preferred to continue receiving
water from their present system. However, almost four million rural households
(17.8 percent) stated a preference for changing to a different water system, most
often because of the existing supply’s taste. About 46 percent of those desiring a
change stated a preference for a community system, and 40.0 percent stated a
preference for an individual system. The form of ownership that was preferred was
-------
1- 12
consistent with size preferences. Over 90 percent of the households which
preferred changing to an individual water system also preferred having an in-house
ownership of the system. At households indicating a preference for a community
system, preferences were for public ownership (72.6 percent), cooperative owner-
ship (13.6 percent), and ownership by a private firm (9.2 percent).
Preferences for staying with their present water system were most domi-
nant in the North Central (86.3 percent) and least dominant in the West (77.4
percent). Conversely, preferences for changing to a different water system were
reported most commonly in the West (22.6 percent) and least commonly in the
North Central (13.6 percent). Though taste was everywhere the most prominent
factor in the desire for a change, it was cited far more frequently in the West (36.6
percent). Other leading factors in the West were the softness and healthfulness of
the water supply. In the Northeast, reliability and healthfulness were the most
frequently mentioned reasons; and in the South, they were convenience, healthful-
ness, and quantity.
A community system was the most frequently preferred size of alternative
system in all regions (preferred by approximately 47 percent) except the Northeast
with about 43 percent. There, individual systems were preferred most frequently
(by about half). Ownership preferences varied substantially from region to region,
with cooperative ownership being mentioned far more frequently in the West than
in any other region.
Among nonSMSA households, the most important factors for those house-
holds desiring a change were convenience, reliability, and the quantity afforded by
the supply, while taste and softness were mentioned more often by SMSA
households than by nonSMSA households. As to ownership preferences, cooperative
ownership and ownership by a private firm were preferred much more often among
SMSA households than among nonSMSA households.
-------
I - 13
SUMMARY OF CHAPTER IV
Although a central objective of the NSA was to investigate various aspects
of drinking water supplies, it became apparent that a study of drinking water
supplies in isolation from supplies that served other domestic needs would be
inadequate. For example, health and economic issues were involved in the use of
water for cooking, bathing, and cleaning as well as in the use of water for drinking.
Furthermore, an exclusive research focus on drinking water supplies would not
reveal why major water supplies—those used for most domestic needs—were not
necessarily always used for drinking. In view of these considerations, the NSA
research was focused on aspects of the major water supply, although data also were
collected on the types of drinking water supplies in rural households.
Chapter IV is organized so that information about drinking water sypplies is
presented first, information - about major household supplies second. Separate
consideration then is given to households which do not use the major supply for
drinking, and to households which have supplementary supplies. Finally, the types
of water sources are enumerated. In this summary, however, the situation for
major supplies is addressed first since that information is the basis for analysis in
subsequent chapters. The situation for drinking water supplies is taken up second,
and that is followed by a summary of types of water sources.
As to the typology used in the NSA, characterizing supplies in terms of
their physical arrangements resulted in seven supply types: (1) multiple-connection
system which served fifteen or more households or more than 25 people (synony-
mous with “community supplies” or “community systems” in the NSA report), (2)
wells, (3) springs, (4) surface water (pond, lake, river, or stream), (5) cisterns, (6)
hauled water, and (7) purchased bottled water.
-------
I - 14
As to classification of sources, the NSA centered on conditions at the rural
household, and there was no provision for geological assessment of the source of
household water supplies. However, on the basis of questions asked of household
respondents and of representatives of multiple-connection systems, supplemented
by examination of household premises by the interviewers, NSA investigators were
able to classify most sources accurately. In addition to the three conventional
classifications—_groundwater, surface water, or precipitation—three others were
included in the NSA for purposes of analysis: “purchased,” which designated water
which a multiple-connection system had obtained from another supplier (which
included purchased bottled water); “mixed,” which designated water from a
combination of sources (such as groundwater and surface water); and “residual,”
which designated all water supplies for which the source could not be determined.
MAJOR HOUSEHOLD WATER SUPPLIES
Although the NSA distinguished between the major water supply and the
drinking water supply (see Chapter II), the two actually were the same in about 97
percent of rural households. In other words, drinking water was obtained from a
supply which was not the major household supply in only about 3 percent (676,000)
of all rural households. In all of the other households (21,298,000), drinking water
was obtained from the same supply (the major household supply) which served other
domestic needs.
As to the type of supply, the major water supply most often was a
community supply, particularly in the West, in SMSAS, and in large and small rural
communities. Whereas the major supplies were wells in 44.8 percent of systems
nationally, wells were the dominant major supply for 60.7 percent of households in
the North Central—compared to only 29.1 percent of households in the West.
Wells were also the most common major supply for households located outside of
SMSAS (about 48 percent) and in other rural areas (approximately 52 percent).
-------
I - 15
Together, community supplies and wells account for approximately 95 percent
(20,831,000 households) of the water supplies of rural households nationally, and
subnational variations from this condition ranged from a low of 92.6 to a high of
99.0 percent. Clearly, community supplies and wells are the overwhelmingly
dominant types of major household water supplies in rural America.
Regarding the remaining types of major household supplies, springs pro-.
vided the major supplies for 2.2 percent of households nationally, and surface water
provided the supplies for only 0.6 percent of households. Cisterns provided the
supplies for 0.6 percent of households, hauled water for 1.2 percent of households,
and purchased bottled water for 0.5 percent of households. Although hauled water
was the major suppl’y for only about 269,000 rural households, and purchased
bottled water (which was sold by a commercial distributor either in small
containers or in bulk) was the major supply at only 102,000 rural households, these
situations represented potential hardship since most of the households apparently
did not have access to a satisfactory alternative.
DRINKING WATER SUPPLIES
Although it was anticipated that many rural households might have more
than one drinking water supply, 97.8 percent of the households had only one supply.
Moreover, though the NSA distinguished between the household’s major and
drinking water supply, they were the same for 97 percent of the households.
Because of this, the data for drinking water supplies were similar to those for
major supplies. Thus, the most common type of drinking water supply (in nearly
half of rural households) was a community system. The next most common type (in
43.6 percent of rural households) was a well. Far less common supply types were
purchased bottled water, springs, surface water, cisterns, and hauled water—
which, taken collecti ely, supplied drinking water to less than 10 percent of rural
households. One prominent aspect of these findings was the prevalence of
-------
I - 16
community supplies in comparison to wells, a reversal of the pattern observed prior
to 1975. Possible reasons for this trend are presented in Chapter IV. Parallel to
the pattern for major household supplies, wells and community supplies together
account for 93.0 percent of the drinking water supplies.
As to findings in subnational classifications, community supplies for
drinking were most common in Western households (61.6 percent), least common in
North Central households (36.8 percent). Conversely, well supplies were most
common in North Central households (59.0 percent), least common in Western
households (27.8 percent). Within SMSAs, community supplies were used at about
58 percent of rural households, compared to about 45 percent of households outside
of SMSAS. (Outside SMSAs, 46.7 percent used wells.) In large rural communities,
91.8 percent of households used a community supply for drinking water, as did 80.4
percent of households in small rural communities. In other rural areas, however,
41.0 percent of households used community supplies, and 50.8 percent used wells.
As to the scope of potential problems at households with hauled or
purchased bottled drinking water supplies, the most common reason for having
either type of supply was that other types of supplies at the households did not
provide water of sufficient quality to drink. Among the 733,000 households with at
least one hauled drinking water supply the median quantity of water hauled each
month was 228 liters (60 gallons). Among the 545,000 households with at least one
purchased bottled water supply for drinking, the median number of times that
water was purchased was about 3.3 times each month. Purchased bottled water
was bought from a grocery store for the majority of households, and 20 liters (five
gallons) or less at a time was purchased for about 54 percent of the households.
The median quantity of water purchased each month was about 64 liters (seventeen
gallons).
Although most rural households had only one drinking water supply, 2.2
percent (about 494,000) had a second supply. Overall, although no rural household
-------
I — 17
had more than two drinking water supplies, 0.6 percent of all rural households had
three or more supplies for domestic purposes, including drinking. Thus, approx-
imately 1.1 million rural households—5.0 percent—had multiple water supplies (for
all domestic purposes), and the remaining 95.0 percent had only one supply for all
domestic purposes. In addition, as already explained, of the 5.0 percent of
households which had multiple supplies, only about 3 percent (676,000) households
took drinking water from a supply which was not the major supply source. Thus, as
already explained, at 97 percent of all rural households, the major household supply
was the same as the drinking water supply, and the information in subsequent
chapters of this report which refers to major household supplies therefore
generally—but not always—includes data relevant to drinking water as well as to
water used for other domestic purposes.
To shift back to the NSA findings regarding the 2.2 percent of rural
households which had a second drinking water supply, it was most likely that one of
the supplies was a community supply, but that the second supply was preferred for
drinking. Furthermore, even though one of the two drinking water supplies was less
likely to be a well than a community system, those two-system households with
wells were even more likely to satisfy most drinking water needs from the second
supply. In either situation, either purchased bottled water or hauled water were
most likely to be the second supply—the one used most often for drinking.
- _ SOURCES OF MA3OR WATER SUPPLIES
As was true for most of the data gathered in the NSA, data on water
sources referred to sources of the major household supplies. Since there was only
one major supply at each household, there was a direct correspondence between the
number of major supplies—21,974,000---and tl.e number of rural households.
As to the NSA findings, the water in 72.3 percent of all major household
supplies (15.9 million) was extracted from a groundwater source, while only 11.4
-------
I - 18
percent of supplies (2.5 million) came from surface water sources. About 8 percent
of all major supplies (1.8 million) came from mixed sources (a combination of
sources), and about 6 percent (1.3 million) came from purchased water (water
which a multiple-connection system obtained from another supplier). Relatively
few cisterns were used by rural households, and as a result precipitation was the
source for only 0.6 percent of all major supplies. The residual category accounted
for 1.4 percent of supply sources.
As to subnational classification findings concerning major supply sources,
groundwater use still predominated, but there was considerable variation among
the various classifications. For example, groundwater was the source for the major
supply in 87.5 percent of North Central households, but in only 56.7 percent of
Western households. In contrast, surface water was the source for the major supply
in 17.2 percent of Southern households, 11.0 percent of Western households, 10.4
percent of Northeast households, but only 3.5 percent of North Central households.
Purchased water was the source at 14.4 percent of Western households, but at less
than 6 percent of households in each of the other regions. Similarly, mixed sources
provided the major supplies at 17.0 percent of Western households, and around 13
percent of Northeast households, but only at less than 8 percent of households in
the other two regions.
About 60 percent of SMSA major household supplies were dependent on
groundwater, compared to about 78 percent of nonSMSA major household supplies.
Pronounced differences also occurred among the size-of-place findings. For
example, although groundwater was still the most prevalent source for major
supplies in large rural communities (in 58.9 percent), surface water was the source
for 21.4 percent and mixed for 10.2 percent of the major supplies in those
communities. In contrast, groundwater was the source for 81.4 percent of major
supplies in small rural communities, whereas surface water was the source at only
11.0 percent of major supplies in those communities. Groundwater was the source
-------
I - 19
for 73.3 percent of major supplies in other rural areas, and surface water was the
source for only 10.1 percent of supplies in those communities.
-------
- 20
SUMMARY OF CHAPTER V
The status of rural water supplies was described in the NSA in terms of
five primary factors: quality, quantity, availability, cost, and affordability. In
addition, the survey questioned rural water users about the effects of water
quality, quantity, and availability on their households. Each of these broad subjects
is explored in detail in Chapter V.
The status of rural water supplies was determined in the rural household .
This emphasis was in keeping with the Congressional directive to obtain informa-
tion on the number of rural residents who had inadequate service, limited access to
supply, or exposure to waterborne health risks. New federal drinking water
regulations reiterated this concern by requiring that most quality standards be met
in the consumer’s household, at his tap, rather than just at the supply facility or at
the source. In the NSA, variables describing status included both laboratory-
measured values (for water quality, for example) and perceived values (such as the
user’s evaluation of the water’s taste and appearance).
QUALITY
In terms of laboratory measurement of water quality, major emphasis in
the NSA was given to bacteriological, physical, and chemical water constituents
which traditionally have characterized water quality. A total of 43 constituents
were studied in the NSA (see Table I-I). The major focus was on measurements of
health-related Constituents of water, particularly in the subsample of 10 percent of
the NSA water specimens, but some determinations were related more to aesthetic
and economic considerations.
-------
I - 21
NSA PROCEDURE
The NSA results for laboratory—determined water quality were grouped in
four broad categories—(l) bacterial content, (2) physical or chemical properties
(turbidity, color, temperature, specific conductance, or hardness), (3) inorganic and
organic constituents, and (4) radioactivity. Summaries of the findings in each of
these categories are given below. Then a separate section is presented which
summarizes the percentages of rural households which exceeded the NSA reference
values established for 28 of the 43 constituents. As to the reference values
themselves (listed in Table I-I), it was desirable to develop them in order to
quantify households water quality. The federal primary and secondary drinking
water regulations composed one set of standards which provided appropriate bases
f or the reference values, but other standards and criteria also were consulted in
developing the values. The overall considerations behind selection of the values
are too complex to summarize: the reader will find full details in Chapter V.
Bacteriological findings
Among bacteriological tests performed in the NSA, the total coliform
count is regarded by many professionals as the best available general indicator of
bacteriological water quality—despite technical shortcomings discussed in Chapter
V. Accordingly, 28.9 percent of all rural households at the time of the NSA survey
had supplies with total coliform levels which exceeded the NSA reference value
(one bacterium per 100 milliliters of water) and which therefore probably required
further assessment and possible remedial action. Rural supplies were above the
reference value more often in the South and West than in other regions, more often
outside of SMSAs than inside SMSAs, and more often in other rural areas than in
large or small rural communities. High values also were much more common
among supplies served by individual or intermediate systems than among supplies
provided by community systems.
-------
I - 22
Table 1-1
Constituents Measured in NSA Survey
NSA Reference Value
Measured in All
NSA Household
(milligrams per
Samples or only
liter of water, unless
in Group H
Category Constituent otherwise noted)
Subsample
Microbial Total coliform Not more than one All
bacterium per 100
milliliters of water
Fecal coliform Complete absence of All
bacteria in a
1 O0-milliliter sample
Fecal streptococcus None All
Standard plate count 500 colony-forming units All
per one milliliter of water
Fecal coliform/fecal
streptococcus ratio None All
Physical Turbidity None All
and Color 15 color units All
Chemical Temperature None All
Specific conductance None All
Total dissolved solids 500
(as determined from
conductance) All
Hard ness None
(as determined from
calcium and magnesium) All
Inorganic Calcium None All
Magnesium 125 All
Nitrate-N 10 All
Sulfates 250 All
Iron 0.3 All
Manganese 0.05 All
Sodium More stringent: 20 All
Less stringent: 100
Lead 0.05 All
Arsenic 0.05 Subsample
Selenium 0.01 Subsample
Fluoride 1.4 Subsample
Cadmium 0.01 Subsample
-------
I - 23
Table I-i continued
NSA Reference Value
Measured in All
NSA Household
(milligrams per
Samples or only
liter of water, unless
in Group II
Category Constituent otherwise noted)
Subsample
Inorganic (continued)
Mercury 0.002 Subsample
Chromium 0.05 Subsample
Barium 1 Subsample
Silver 0.05 Subsample
Organic Endrin 0.0002 Subsample
Lindane 0.004 Subsample
Methoxychlor 0.1 Subsample
Toxaphene 0.005 Subsample
2,4 -D 0.1 Subsample
2,4,5-TP 0.01 Subsample
Radioactive
Gross alpha See Figure V-28 Subsample
Gross beta 50 pCi Subsample
* Radium 226 These constituents
* Radium 228 were not measured
* Uranium frequently enough
* Stontjum_89 to provide
*Strontium_9 0 independent
* Cesium- 134 national estimates.
*Trjtium (See text for details
*Iodine_13 1 about NSA reference
values.)
*Measured only if the laboratory analyst considered gross alpha or gross beta
readings sufficient to warrant further investigation.
-------
I - 24
Furthermore, on the basis of NSA findings, 12.2 percent of all rural
households had supplies which may have required attention because of the presence
of at least one fecal coliform bacterium. In 1.6 percent of rural households
(350,000), the fecal coliform count was higher than 200 colonies per 100 milliliters
of water, a level which has been associated with increased occurrence of at least
one organism with disease potential, indicating the possibility of heightened threat
of disease. High values were proportionately most common outside of SMSAS, in
other rural areas, and among households served by individual or intermediate
systems.
Fecal streptococci were found in 1.8 million rural water supplies which also
had fecal coliform bacteria. The ratio of these two organisms in the supplies
suggested that contamination from animal wastes alone (a potential in 4.9 percent
of all rural households) outweighed contamination from human wastes alone (a
potential in 1.5 percent of all rural supplies). This trend was apparent in all regions
of the US and also in households grouped by size of place and size of system.
In regard to general levels of bacteria as determined by the Standard Plate
Count, values exceeding 500 per milliliter of water were found most often in the
West and South. The high values were more frequent among nonSMSA households,
among households located in other rural areas, and among households served by
individual systems.
Physical and chemical characteristics
Of the physical and chemical characteristics studied in the NSA, turbidity
probably was the most comprehensive, but least specific, indicator of water
quality. Generally, domestic water with low turbidity is easier to disinfect, offers
less opportunity for proliferation of bacteria, and may be less suscejtible to taste
and odor problems. Nearly 84 percent of rural household supplies had readings
which were quite low (less than one nephelometric turbidity unit). Higher values
-------
I - 25
were particularly prominent in the North Central and South, and in households
served by individual or intermediate systems as opposed to community systems.
Households which had more than 500 milligrams per liter of total dissolved
solids (as derived from specific conductance) also were most prominent in the
North Central, where about one of every four households had concentrations of this
magnitude. The proportion above this reference value was nearly as large in the
West, but much smaller in the South and Northeast.
As to water hardness, the North Central once again had more than its share
of supplies with very hard water. The median in the North Central (255.6 calcium
carbonate equivalent units) was more than twice the median value for the nation,
and considerably larger than the median for any other region.
Color was not a problem in rural supplies. Water temperature, on the other
hand, was a potential aesthetic problem. Measured temperature was above 200 C
at 43.8 percent of all rural households. The proportion of households with warm
water was most prominent in the South, where about seven of every ten households
had water temperatures over 200 C. The proportion over 200 C in the West was
less than that in the South, but large in comparison to the Northeast and North
Central. An important consideration, however, is that the 200 C mark refers to
conventional abstract measures of desirable water temperature. Two additional
considerations should be kept in mind. First, the temperature readings reflect not
a yearly average but more likely the season in which they were taken—late
summer and early fall. Second, household residents may consider water that is
warmer than 200 C acceptable.
Inorganic constituent findings
iorganic substances studied in NSA specimens ranged from those largely
with aesthc tic effects, such as manganese, to those predominately with health
effects, such as lead. Certain of the constituents have received particular
-------
I - 26
attention from investigators in the past, and it was anticipated that they might be
associated with significant findings in the NSA. Among those constituents were
calcium and magnesium, key elements in water hardness; nitrates, substances
which pose a special health risk for infants; sulfates, substances which may make
water distasteful and even cause diarrhea; and iron and manganese, elements which
cause aesthetic and economic problems.
Calcium and magnesium were present in sufficient amounts to produce
moderately hard water in rural US supplies, but the substances themselves were not
implicated in any direct health effects by the NSA findings. Nitrates were not
discovered in large concentrations in rural supplies—fully 97.3 percent of rural
households were below the NSA reference value. Sulfates were potential problems
in 4.0 percent of rural households, but more often for aesthetic rather than for
serious health reasons. Iron and manganese concentrations posed aesthetic
problems in a number of rural households, but about eight out of ten rural
households met the NSA reference value for each metal.
Although households which exceeded the NSA reference value for any of
these common constituents faced potential problems which required assessment
and attention, the difficulties were not beyond the scope anticipated.
There were potentially important problems posed by several other inor-
ganic substances, however. Those substances, all heavy metals, were lead,
cadmium, and mercury. Lead was studied in all NSA sample households, and on the
basis of the results, 16.6 percent of all rural American households (numbering 3.6
million) were above the NSA reference value for lead. Cadmium and mercury were
studied in NSA subsample households, and on the basis of the results, 16.8 percent
and 24.1 percent of US rural households, respectively, exceeded the NSA reference
value.
The possible health implications of these findings are important. The three
metals have different physiological effects, but the effects are known to be
-------
I - 27
potentially serious, as is detailed in the separate reports on each of the heavy
metals in Chapter V.
Study of a number of other inorganic substances in NSA subsample
households showed other persistent, but less consequential, problems. Supplies at
13.7 percent of rural households surpassed the NSA reference value for selenium,
although the concentrations were not large enough to pose health threats. About 4
percent of rural households had values beyond the reference value for silver, a
situation which posed no immediate health threat but which involved a substantial
number of rural households.
As to some of the other NSA constituents, amounts of arsenic, chromium,
and barium in rural households generally were very small and posed few health
concerns. Levels of fluoride also generally were well within the NSA reference
value.
Organic constituent findings
Organic constituents studied in the NSA were limited to four chlorinated
- hydrocarbon insecticides and two chlorophenoxy herbicides. Of these six sub-
stances, only two were detected at all in NSA households. Those two substances
were lindane—found only in small proportions of households in the West and
South—and methoxychlor—found only in a small proportion of households in the
West. None of the values for the substances exceeded or even closely approached
the respective NSA reference values.
Radioactivity findings
Rural household supplies showed low levels of both gross alpha and gross
beta radiation. The presence of bAckground gross alpha radiation, in particular,
was not surprising, since it is produ:ed by natural sources commonly found in
groundwater. Despite the prevalence of the radiation, levels of radioactivity were
-------
I - 28
so low in the NSA that radiation in drinking water was not shown to be a national
problem.
PROPORTIONS OF HOUSEHOLDS EXCEEDING NSA REFERENCE VALUES
Although 43 constituents were studied in the NSA, only 28 were assigned
NSA reference values (see Table [ -I). (The complex reasons for not assigning
reference values for certain constituents are given in Chapter V.) The focus here
is exclusively on the proportions of households exceeding the NSA reference values
(Table 1-2). As detailed in Chapter V, the reference values provide important
benchmarks regarding health, aesthetic, and economic implications associated with
the relevant concentrations of the Constituents.
In examining Table 1-2, it should be noted that for many of the 28
constituents, results were available for only a 10 percent subsample of the NSA
sample. Specifically, the following constituents were analyzed for the subsample
only: arsenic, selenium, fluoride, cadr iium, mercury, chromium, barium, silver,
endrin, lindane, methoxychlor, toxaphene, 2,4-D, 2,4,5-TP, gross alpha radiation,
and gross beta radiation. All other constituent results in Table 1-2 are based on the
full NSA sample.
For the nation as a whole, concentrations sometimes exceeded the ref er-
ence values for twenty out of 28 constituents. The constituents which were never
found to exceed the respective reference values at any household were chromium,
all six organic pesticides and gross beta radioactivity. Table 1-2 indicates that the
most frequent problems were caused by bacterial contamination, iron, lead,
cadmium, and mercury. The NSA reference values were exceeded for each of
these constituents in 15.0 percent or more of all rural households.
It is apparent from Table 1-2 that there were distinct regional differences
in the results. For example, households in the Northeast exceeded reference values
for the smallest number of substances—fourteen. In contrast, households in each
-------
I - 29
Table 1-2
Percentage of Rural US Households Exceeding NSA Reference Values
Established for 28 of 43 Constituents Studied in the NSA
(Reference Values in Parentheses)
Total
Fecal
Standard
Color
Colif arm
Coliform
Plate Count
(15
standard
(1/100 ml)
(0/100 ml)
(500/mi)
color units)
2.3
REGION
Northeast
North Central
South
West
28.3
24.4
31.7
30.6
14.0
8.0
13.9
13.6
10.2
17.1
22.8
24.8
0.5
3.4
2.6
1.6
SMSA/NonSMSA
SMSA
NonSMSA
18.3
33.9
6.8
14.7
13.8
22.0
1.4
2.8
SIZE OF PLACE
Large rural
communities
17.7
4.9
15.0
1.5
Small rural
commun Ities
19.6
4.0
11.7
5.4
Other rural
areas
31.2
13.8
20.5
2.2
SIZE OF SYSTEM
Community
Intermediate
Individual
15.5
43.3
42.1
4.5
20.2
19.8
13.9
17.8
26.6
1.9
1.9
3.0
-------
I - 30
Table 1-2 continued
Estimated
Total
Dissolved
Solids
Magnesium
Nitrate-N
Sulfates
(500 mg/I)
(125 mg/I)
(10 mg/I)
(250 mg/I)
NATION 14.7 0.1 2.7 4.0
REGION
Northeast 5.0 0.0 0.3 0.5
North Central 23.9 0.1 5.8 7.4
South 10.2 0.0 1.3 0.7
West 22.2 0.5 4.0 11.7
SMSA/N0nSM5A
SMSA 15.1 0.1 1.7 2.2
NonSMSA 14.5 0.1 3.2 4.8
SIZE OF PLACE
Large rural
communities 15.8 0.0 4.2 2.6
Small rural
communIties 17.7 0.0 4.7 7.5
Other rural
areas 14.3 0.1 2.4 3.8
SEZE OF SYSTEM
Community 15.0 0.0 1.6 4.2
Intermediate 13.4 0.4 3.0 1.7
Individual 14.7 0.1 4.1 4.2
-------
I - 31
Table 1-2 continued
Iron Manganese
(0.3 mg/I) (0.05 mg/I)
Sodium
(100 mg/I)
Lead Arsenic*
(0.05 mg/I) (0.05 mg/I)
NATION
REGION
18.7 14.2
14.3
16.6 0.8
Northeast
NorthCentral
South
West
16.0
28.2
17.0
7.0
16.9
19.9
12.3
4.7
6.0
19.2
14.1
15.0
9.6
10.8
23.1
16.9
0.0
1.8
0.0
2.1
SMSA/NonSMSA
SMSA
NonSMSA
13.8
21.0
9.9
16.3
14.9
13.9
12.9
18.3
0.0
1.2
SIZE OF PLACE
Large rural
communities
9.4
11.4
15.7
15.1
0.0
Small rural
communities
23.3
21.7
17.0
18.1
6.6
Other rural
areas
19.5
14.0
13.8
16.6
0.4
SIZE OF SYSTEM
Community
Intermediate
Individual
7.7
28.7
29.9
7.2
23.3
20.7
15.8
10.3
13.3
17.7
20.5
14.1
0.9
0.0
0.8
-------
I - 32
Table 1-2 continued
Selenium* Fluoride* Cadmium* Mercury* Chromium*
( 0.01 mg/I) (1.4 mg/I) (0.01 mg/I) (0.002 mg/I) (0.05 mg/I )
NATION 13.7 2.5 16.8 24.1 0.0
REGION
Northeast 0.0 0.0 1.6 22.0 0.0
North Central 25.7 1.8 20.7 31.8 0.0
South 2.1 2.7 17.3 25.0 0.0
West 41.3 6.2 27.1 10.4 0.0
SMSA/N 0n SM SA
SMSA 14.4 1.6 21.4 21.5 0.0
NonSMSA 13.3 2.9 14.3 25.5 0.0
SIZE OF PLACE
Large rural
communities 16.5 2.9 19.8 16.2 0.0
Small rural
communities 6.6 6.6 7.3 27.6 0.0
Other rural
areas 14.0 2.1 17.3 24.6 0.0
SIZE OF SYSTEM
Community 12.5 2.9 21.2 23.3 0.0
Intermediate 21.7 6.7 26.9 36.0 0.0
Individual 13.6 0.8 7.9 22.3 0.0
-------
I - 33
Table 1-2 continued
Barium*
Silver*
Endrin*
Liridane*
Methoxychlor*
(1.0
(0.05
(0.0002
(0.004
(0.1
mg/I)
mg/I)
mg/I)
mg/I)
mg/I)
NATION 0.3 4.1 0.0 0.0 0.0
REGION
Northeast 0.0 4.8 0.0 0.0 0.0
North Central 0.0 3.7 0.0 0.0 0.0
South 0.7 4.8 0.0 0.0 0.0
West 0.0 2.1 0.0 0.0 0.0
SMSA/NonSMSA
SMSA 0.0 5.1 0.0 0.0 0.0
Nor iSMSA 0.4 3.6 0.0 0.0 0.0
SIZE OF PLACE
Large rural
communities 0.0 3.3 0.0 0.0 0.0
Small rural
communities 0.0 3.6 0.0 0.0 0.0
Other rural
areas 0.3 4.2 0.0 0.0 0.0
SIZE OF SYSTEM
Community 0.0 2.1 0.0 0.0 0.0
intermediate 0.0 3.4 0.0 0.0 0.0
Individual 0.8 7.1 0.0 0.0 0.0
-------
I - 34
Table 1-2 continued
Toxaphene* 2,4_D* 2,4,5_TP* Gross Alpha*Gross Beta*
( 0.005 mg/I) (0.1 mg/I) (0.01 mg/I) (see Fig.V-28) (50 pCi/I )
NATION 0.0 0.0 0.0 0.5 0.0
REGION
Northeast 0.0 0.0 0.0 0.0 0.0
North Central 0.0 0.0 0.0 0.0 0.0
South 0.0 0.0 0.0 1.3 0.0
West 0.0 0.0 0.0 0.0 0.0
SMSA/N0nSMSA
SMSA 0.0 0.0 0.0 0.0 0.0
NonSMSA 0.0 0.0 0.0 0.8 0.0
SIZE OF PLACE
Large rural
communities 0.0 0.0 0.0 0.0 0.0
Small rural
communities 0.0 0.0 0.0 0.0 0.0
Other rural
areas 0.0 0.0 0.0 0.6 0.0
SIZE OF SYSTEM
Community 0.0 0.0 0.0 1.0 0.0
Intermediate C.0 0.0 0.0 0.0 0.0
Individual 0.0 0.0 0.0 0.0 0.0
*Constjtuent analyzed for only the 10 percent NSA subsample.
-------
I - 35
of the other three regions surpassed the reference values for eighteen of the
twenty Constituents.
In terms of general regional water quality implications, households in the
Northeast clearly were least likely to have problems. Households in the North
Central generally were most likely to have potential water quality problems—
whether the problems were related to health, aesthetic, or economic effects.
However, potential bacteriological problems, as indicated by the findings for total
coliform and fecal coliform, were most likely to occur in the South and the West.
Households in the South were most likely to exceed the reference value for lead;
households in the North Central were most likely to be above the reference value
for mercury; households in the West were most likely to surpass the reference
value for cadmium and selenium.
NonSMSA households had a higher potential for water quality problems
than did SMSA households. In all, nonSMSA households had higher percentages
above the reference values for fourteen of the twenty constituents. In Contrast,
SMSA households had higher percentages above the reference values for only four
of the twenty constituents. The situation among nonSMSA households was
particularly serious because of the numbers of households involved: about two-
thirds of the total US rural population lived in nonSMSA areas.
Values for total coliform bacteria, lead, and mercury exceeded the respec-
tive reference values in over 15.0 percent of the households in every size-of-place
category. Iron problems tended to be more prominent in small communities and in
other rural areas. Total dissolved solids and sodium problems were slightly higher
in large and small rural communities than in other rural areas. Manganese
problems seemed to be concentrated in small rural communities, while selenium
and cadmium problems occurred most frequently in other rural areas .nd large
rural communities.
-------
I - 36
Among all categories of system size, values for mercury and totai coliform
bacteria exceeded the reference values in over 15.0 percent of the households.
Sodium appeared over the reference value most often among households served by
community systems. Selenium was prominent among households served by inter-
mediate systems. Five constituents were found in concentrations exceeding the
reference values in more than 15.0 percent of rural households served by
community water systems or by individual systems. This situation was the same
for seven constituents among households served by intermediate systems. More
than 15.0 percent of rural households exceeded the , eference value for total
coliform, regardless of the size of supply system serving the household.
Following the completion of the water collection and laboratory assess-
ment, it was discovered that the ampules containing the acid preservatives could
induce an artificial lead and cadmium background in the water specimens by an
average of 36 parts per billion lead and 0.92 parts per billion cadmium. If the NSA
specimens were thus contaminated, and it is likely they were, the estimate of
households above the lead reference value would be reduced from 16.6 percent to
9.2 percent. The reduction in the estimated proportion of households above the
cadmium reference value would be reduced from 16.8 percent to 15.9 percent.
PERCEIVED WATER QUALITY
In addition to laboratory analysis, people’s perceptions provided supplemen-
tary information concerning the quality of water supplies in rural America.
Subjective judgments revealed the prevailing conditions of the household water
supply. Specifically, NSA respondents were asked about the odor, taste, clarity,
color, sediment content, and temperature of the major household water supply. In
those instances where problems were identified, further questioning revealed the
severity, frequency, and possible reasons for the problem. The summary which
follows focuses on selected highlights of the NSA findings, since, as was true for
-------
I - 37
the laboratory findings, many of the details were too complex to be summarized
briefly.
Odor
About one-quarter of all rural households (approximately 5.6 million)
reported odor in the major household water supplies. Although strong odors usually
were only detected occasionally in these households, slight odors were present most
or all of the time. The reported odors were most often associated with deliberate
or planned activities and they most often were described as smelling like chlorine.
Little variation was detected among households in the four regions of the
United States, or between SMSA and nonSMSA households, in either the intensity or
the duration of water supply odors. However, in terms of the size-of-place and
size-of-system comparisons, odor conditions were somewhat different. Specif i-
cally, more households in large rural communities reported odors in their water
supplies (35.6 percent, compared to 33.2 percent for small rural communities and
23.5 percent for other rural areas). Not only were water supply odors more
prominent in large rural communities, but odors were also more intense and of
longer duration. Similarly, odors were more prominent in household water supplies
served by community systems (32.3 percent, compared to 19.8 percent for
intermediate systems and 18.3 percent for individual systems).
Taste
Although odor and taste generally would be expected to occur togeth . r,
water supply tastes were reported more often than were odors. In all, 5.6 million
households reported water supply odors; 7.8 million households reported tastes.
These tastes were most ‘f ten described as slight, but persistent throughout the
year, or occasionally str ng. As with odors, water supply tastes were most often
-------
I - 38
attributed to planned activities. While water supplies sometimes tasted like
chlorine, they most often tasted like mineral water.
The subnational variations in water supply tastes as described by rural
residents followed somewhat the same pattern as seen for water supply odors.
While little variation was detected among households in the four US regions and
between SMSA and nonSMSA households, taste conditions varied f or the size-of-
place and size-of-system comparisons. In particular, more households in large rural
communities reported tastes in their water supplies (45.4 percent, compared to
38.8 percent for small rural communities and 33.9 percent for other rural areas).
Although water supply tastes occurred more frequently among households located
in large rural communities, intense tastes of longer duration were no more
prominent here than in small communities or other rural areas. Among households
served by supply systems of different sizes, 42.4 percent of those served by
community systems reported water supply tastes, compared to 32.1 percent of
households served by intermediate systems and 27.8 percent of households served
by individual systems.
Cloudiness
Components of water as perceived by sight—cloudiness, color, and sedi-
ment content—have potential implications for water quality. Furthermore, these
characteristics may affect people’s attitudes towards their water supplies.
Across the nation, 5.9 million households noticed a cloudy condition in their
water supplies. The cloudiness was most often characterized as slight and only
occasionally present. Water supply cloudiness was most often attributed to
breakdowns in supply technology or to planned activities.
Differences in cloudy conditions were observed for all of the subnational
groupings. Comparing across regions, problems with cloudiness in the water supply
were most often reported among households in the South and West (29.9 percent
-------
I - 39
and 27.4 percent, respectively); cloudy conditions were reported least often among
households rn the North Central and Northeast (24.7 percent and 22.6 percent,
respectively). These conditions were slightly more intense and of longer duration
in the North Central and West.
The SMSA/nonSMSA comparison showed that proportionately more non-
SMSA households noticed cloudiness in their water supplies (28.9 percent) as
compared to SMSA households (22.7 percent). These differences reflected the
higher proportion of nonSMSA households which reported infrequent slight cloudi-
ness, as opposed to more persistent or intense cloudy conditions.
Households situated in large rural communities reported cloudy conditions
more often than those in small rural communities or other rural areas. Also,
cloudy conditions were slightly more persistent and intense in large rural communi-
ties. Consistent with the variations in the size-of-place comparison, community
systems were least often free of any cloudiness.
Color
Color in rural water supplies was noticed less often than cloudiness. In
total, 4.5 million households across the nation reported that the water supply had
color. While the colored condition was most often characterized as slight and only
occasionally noticeable, 1.4 million households reported a prevalent color in the
water supply. The presence of color was most often attributed to breakdowns in
supply technology although mineral content of the water and planned activities
were frequently cited.
Water supply color varied somewhat according to region, size of place, and
size of system. While equal proportions of households in each of the four regions
reported that color was present in the water supply, proportionately more
households in the West characterized the colored C( ndition as severe. Compared to
-------
1 - 40
households in small rural communities arid other rural areas, a larger proportion of
households in large rural communities reported color in the household water supply.
Sediment
Approximately one-third of all rural households (7.3 million) experienced
some degree of sedimentation. For most households reporting sediment, the
condition was an occasional problem, and did not involve heavy sediment. Rural
residents often thought that the sedimentation resulted from either a breakdown in
the physical facilities of the water system or the inadequacy of the physical
components of the system.
Except for regional differences, subnational comparisons showed no appre-
ciable variation in the frequency and intensity of sediment conditions. The chief
regional difference was the much higher proportion of households in the West that
repo ted sediment in the water supply. In the West, 43.5 percent of households
reported sediment in the water, compared to 34.2 percent in the North Central,
33.1 percent in the Northeast, and 29.5 percent in the South.
Temperature
As expected, substantial regional differences were observed in the per-
ceived temperature of rural water supplies. Water supplies in the Northeast and
North Central were thought to be predominately cold (63.3 percent and 53.8
percent, respectively) while supplies in the South and West were most often
characterized as cool (74.1 percent and 54.8 percent, respectively). Overall, warm
water supplies were infrequent, reported by less than 10 percent of the households
in each of the regions.
A comparison of SMSA and nonSMSA households showed little variation
irom national estimates. With respect to the size-of-place comparison, however,
slight differences were observed in water supply temperatures. Generally, with
-------
I - 41
decreasing place size, the proportion of cold supplies increased and the proportion
of warm supplies decreased. More specifically, cold supplies ranged from 30.5
percent in large communities through 39.2 percent in other rural areas; warm
supplies ranged from 11.4 percent through 5.2 percent.
Consistent with the pattern seen in the size-of-place comparison, cold
water temperatures were reported by 53.8 percent of households served by
individual systems, 44.2 percent served by intermediate systems, and 24.1 percent
served by community systems. About 2.1 percent of households served by
individual systems reported warm water temperatures, as did 4.9 percent of those
served by intermediate systems, and 9.2 percent served by community systems.
WATER QUANTITY
Although water quality is a paramount concern, water quantity also is an
urgent consideration. However, water consumption is difficult to measure unless
the household supply is metered. In the NSA, meter reading was done primarily at
households served by community systems (those with fifteen or more connections).
Water usage at households served by individual systems was more difficult to
ascertain since devices for metering were not employed—although persons who
used hauled or purchased bottled water could usually provide fairly accurate
information. NSA investigators thus assessed potential water use from individual
supplies by measuring pump capacity, the effective volume of pressure tanks
associated with the pump, or the capacity of storage tanks. In addition to the
above assessments, a sense of water usage was obtained by asking about the
respondent’s perception of the amount of water available to the household.
The major findings about recorded and perceived water use were:
• For the 4.5 million rural households which could provide bills ‘,ith
sufficient infurmation, the median daily household Consumption was 664 liters (175
gallons). Median daily per capita consumption was 227 liters (60 gallons).
-------
I - 42
• There were roughly 9.8 million rural households which had on-premise
pumps. The median pump capacity was 22 liters (6 gallons) per minute. Most of
these households (9.5 million) also had at least one pressure tank with a median
capacity of 114 liters (30 gallons).
• Among the approximately nine million rural households where the effective
volume of the pressure tank was measured, the median effective volume was 13
liters (3 gallons). About 4 percent of these tanks, however, were completely
waterlogged (the cushion of compressed air in the tank was entirely dissipated).
• Auxiliary storage tanks were relatively rare: they were reported at only
about 945,000 households (or 4.3 percent of all rural households). Although rare,
these tanks usually were large; the median size was about 760 liters (200 gallons).
• While few households relied upon hauled water (269,000) and purchased
bottled water (1o2;000) as the major supply, those which did used smaller quantities
of water than households served by community systems. The median quantity
hauled each month was 609 liters (161 gallons); the median quantity purchased each
month was 5,900 liters (1,557 gallons).
• The perception of most rural households was that water supplies were
ample: fully 17.7 million households reported that the major household supply
completely satisfied water requirements, and another 3.6 million households
reported that it usually or almost always provided sufficient water. Only about
700,000 households reported that the supply usually or always provided an un-
acceptable quantity of water; insufficient quantity was most often attributed to an
inadequacy of the physical facilities of the supply.
• As to regional variations among households metered and billed for water,
the West and South had the highest median per capita daily consumption (234 liters,
or 62 gallons, each) whie the North Central had the lowest (1g8 liters, or 50
gallons).
-------
I -
• Consistent with the pattern seen in the regional comparison of metered
quantity, the potential for greater per capita consumption was evident in the West
based on pump capacity, effective volume, and tank size. The median pump
capacity in the West was at least twice the capacity of pumps in other regions (48
liters, or 13 gallons per minute); pressure tanks had nearly double the effective
volume (23 liters, or 6 gallons); and storage tanks were at least seven times as
large, with a median size of 3,777 liters (997 gallons). The only exception was in
the South, where median storage tank size approximated that in the West.
• As to regional differences in household perception of water quantity,
Western households reported more frequent problems with quantity than households
in the other three regions. For example, 7.2 percent of households in the West
reported either that the water supply usually did not provide enough water or that
it never did, compared to 2.4 percent in the Northeast, 1.9 percent in the North
Central, and 2.8 percent in the South.
WAtER AVAILABILITY
In the NSA, availability was defined in terms of the supply’s reliability, or
its uninterrupted service, and the supply’s accessibility, or the difficulty of
obtaining water from the supply.
RELIABILITY
On the whole, supplies which served rural households were fairly reliable.
Of the estimat d 22 million rural households across the nation, only 5.6 million
reported water supply breakdowns. Among the households reporting breakdowns,
the majority had only one or two during the year preceding the NSA, and only a
small proportion (less than 2 percent) reported as many s six or more.
Slightly more than 3.2 million iural households (about 1.5 percent of all
rural households) reported breakdowns which lasted six or more hours (considered
-------
I - 44
severe by NSA standards). Regardless of the severity of the breakdowns, the vast
majority of the 5.6 million households that reported breakdowns (4.6 million)
attributed them to problems outside of the household. When a breakdown occurred,
most households (4.5 million) waited until the supply was repaired or borrowed
water from relative, friends, or neighbors.
Iii terms of regional differences, household supplies in the South seemed to
be somewhat less reliable than supplies in the other regions (29.2 percent reported
breakdowns, compared to 25.3 percent in the North Central, 21.8 percent in the
Northeast, and 20.5 percent in the West). Similarly, severe water supply
breakdowns tended to be more prevalent in the South, where they were reported by
17.1 percent of the households. While the majority of households in each region
indicated that supply breakdowns originated outside of the structure, a much larger
proportion of households in the West (about 98 percent) attributed breakdowns to
that source.
Although the reliability of SMSA and nonSMSA supplies was similar, there
were differences according to the size-of-place classification. Generally, house-
holds in small rural communities reported supply breakdowns more often than did
households in large rural communities or other rural areas (29.7 percent, compared
to 22.2 percent in large rural communities and 26.0 percent in other rural areas).
Similarly, water supply breakdowns which were considered severe by NSA standards
occurred more often among households situated in small rural communities
(reported at 16.6 percent).
Compared with households using individual or community systems, a larger
proportion of households served by intermediate systems reported one or more
supply breakdowns (33.1 percent, compared to 26.4 percent served by community
systems and 22.8 percent served by individual systeris). Also, a greater proportion
of households served by intermediate systems reprted breakdowns that were
severe. In response to supply breakdowns, households served by community systems
-------
- 45
generally waited until service was restored, while those with intermediate systems
borrowed water or waited, and those with individual systems borrowed water or
took action to repair the supply.
ACCESSIBILITY
An accessible water supply was defined as one which provided water when
needed and was not in an inconvenient location. Furthermore, the supply had to
provide water at an appropriate pressure—as measured by a pressure gauge which
the NSA interviewer applied to the tap.
The convenience of the supply’s location was determined for nearly half of
rural households (48.2 percent) by distance measurements between the household
and the point of withdrawal from the source. These measurements were made at
households served by wells, springs, surface water, or cisterns—but not at
households served by hauled supplies, purchased bottled water, or community water
systems (the last of which was assumed to provide maximum convenience). The
NSA results indicated that supply sources generally were at a convenient distance
from the 10.6 million households at which measurements were made. However,
comparatively great distances were recorded at a certain proportion of households.
This generally meant the need for more extensive piping or other transmission
systems to transport water from the source to the households. Specifically, about
52 percent of the supplies for which a distance measurement was recorded
withdrew water from a point that was within ten meters (33 feet) of the household.
For another 31.9 percent, the point of wit’idrawal was eleven to 50 meters away
(36 to 164 feet). In contrast, less than 1 percent of rural households reported
distances which exceeded 1,000 meters (more than half a mile).
Water pressure at rural households also was generally sufficient. The
median measurement was 39 pounds per square inch, although 1.4 million house-
holds had supplies with a pressure of 20 pounds per square inch or less. This
-------
I - 46
corresponded closely to the number of rural households that perceived the usual
water supply pressure to be too low (1.6 million households). Another 2.0 million
rural households with pressurized supplies reported either that the pressure was too
high or that it fluctuated. Pressure difficulties were most often attributed to some
unspecified inadequacy of the water supply’s physical facilities.
As to regional differences, there was a great deal of variation in the
distance between the dwelling unit and the point where the water was withdrawn
from the source. Distance measurements were greater in the South and West than
in the Northeast or North Central.
Water supply pressure also varied from region to region. The lowest
pressure readings (20 psi or less) were most common in the North Central, and the
highest readings (61 through 100 psi) were most common in the West. In all
regions, the great bulk of supplies were perceived to have satisfactory water
pressure.
Supplies serving rural America tended to be less accessible for nonSMSA
households than for SMSA households. Among nonSMSA households, 49.2 percent of
supplies had the point of withdrawal within ten meters of the dwelling unit,
compared to 58.4 percent of supplies among SMSA households. Distances which
exceeded 100 meters were recorded f or about 7 percent of nonSMSA supplies,
compared to roughly 3 percent of SMSA supplies. Also, a similar pattern was seen
in terms of the water pressure readings. That is, water pressure tended to be lower
among nonSMSA households and therefore the supply was considered less acces-
sible.
The size-of-place comparison showed a great deal of variation in each of
the indicators of accesiibility. NSA results showed that water supplies in other
rural areas were least accessible in terms of distance to source, while t ose
situated in small rural communities were most accessible. Roughly three-fourth5
of all supply sources in small rural communities were located within ten meters of
-------
I - 47
the dwelling unit, compared to about half of the supplies in other rural areas. At
the other extreme, the only water supplies that withdrew water from a source over
1,000 meters away from the household were located in other rural areas.
Higher pressure readings were more common in large rural Communities
than in small rural communities or other rural areas. In large communities, only
3.1 percent of supplies had pressure readings between one and twenty pounds per
square inch, compared to 5.7 percent of supplies in small rural communities and 8.2
percent of supplies in other rural areas. Readings of from 61 through 100 pounds
per square inch were found in 22.4 percent of supplies in large communities,
compared to 14.9 percent of supplies in small rural communities and 12.0 percent
of supplies in other rural areas.
Finally, water pressure as determined by household representatives was
thought to be “about right” less often among households in large communities.
Although the differences were only slight, 79.3 percent of households in large
communities reported that the pressure was about right, compared to 82.7 percent
in small communities and 83.8 percent in other rural areas.
Size-of-system differences were also observed when comparing distances
to the ource. (Since the information gathered did not apply to community
systems, the comparison was restricted to individual and intermediate systems.)
Supplies for households served by intermediate systems had greater distances
between the point of withdrawal from the source and the dwelling unit. Among
households served by individual systems, 57.8 percent of supplies had the point of
withdrawal either on the premises or within ten meters of the structure, while this
was true for only 27.0 percent of supplies at intermediate-system households. With
respect to greater distances, 8.1 percent of households served by intermediate
systems were s wated 101 meters or more from the source, while only 3.0 percent
of those serve I by individual systems were that distance.
-------
I - 48
Pressure readings also varied according to the size of the water system.
Among households served by individual and intermediate systems, pressure was
virtually never more than 60 pounds per square inch, and most often between 21
and 40 pounds per square inch. In contrast, almost one-quarter of community-
system households had pressure readings between 61 and 100 psi, and close to 50
percent had readings between 41 and 60 psi. Nevertheless, more households served
by community systems expressed dissatisfaction with the supply’s water pressure,
even though on the average their pressure readings were higher. To summarize,
80.8 percent of households served by community systems reported that the usual
pressure was about right, compared to 85.2 percent of households served by
individual systems and 87.9 percent served by intermediate systems.
EFFECTS OF QUALITY, QUANTITY, AND AVAILABILITY
In the NSA interviews, residents were questioned about problems or
inconveniences which they associated with inadequacies of various conditions of
their water supplies. Additionally, residents were specifically asked whether
anyone living in the household or any visitors to the household had become ill from
drinking the household water.
On the basis of inquiries at the household, illnesses among rural residents,
visitors, or both, during the year prior to the NSA study were thought to be
associated with the water supply in only 2.3 percent (or 505,000 households) of all
rural households. Diarrhea was the most common illness although a variety of
maladies were reported. As to problems not related to health, two million
households reported inconveniences resulting from water supply conditions ranging
from discolored laundry to deposits on household fixtures.
-------
I - 49
COST
The cost of water to households is determined by a number of factors.
Though the NSA focused on the household, there was no reliable way to collect
operating and maintenance Cost data for individual supply systems. Also, although
some households served by intermediate systems were billed on a regular basis, rio
cost information was available. Therefore the cost analysis was restricted to
community-system households which could provide the relevant billing data. (The
operating and maintenance Cost of the community systems supplying the water was
a separate consideration, as explained in Chapter VIII.) Although Cost could be
quantified only f or community-system households, perceived cost was assessed for
all rural households.
Another difficulty in assessing household water cost was that water bills
sometimes were unavailable at households served by community systems. There-
fore, it was possible to calculate costs for only 4.6 million households, essentially
one-fifth of the nation’s rural households. The NSA findings showed that the
median household water cost for every thousand gallons consumed was $1.35
nationally, while the median monthly household cost was $7.00.
When NSA respondents were queried about water costs, the majority felt
the cost was reasonable, or that the water was inexpensive (17.3 million,
combined). On the other hand, the water was considered expensive or very
expensive at about three million households. Although a significant number of
households reported that costs had increased during the year prior to the survey
(7.9 million), the majority (12.2 million) judged costs to have remained the same.
In slightly less than one million rural households, it was reported that no cost was
associated with the water.
As to costs in the different reg ons, median costs per thousand gallons
ranged from $1.33 in the South to $2.00 in the West. In the Northeast and North
Central, the medians were $1.34 and $1.50, respectively.
-------
I - 50
Although the cost of water was thought to be reasonable or inexpensive by
residents at the great majority of rural households in all four regions, water was
reportedly expensive or very expensive for a greater proportion of households in
the South and West. In a related finding, the cost of water was reported to have
increased in a large proportion of households in the West (42.6 percent, Compared
to about 35 percent in the other three regions). However, water costs appeared to
have remained the same over the past year for the majority of households in each
of the four regions.
In general, households within SMSAs had lower water Costs per thousand
gallons than did those outside SMSAs. Half of the households within SMSAs were
charged $1.08 or less for every thousand gallons, whereas half of nonSMSA
households were charged $1.62 or less. Despite these findings, residents living
within SMSAs believed that their water costs were higher than residents living in
the more rural areas. (However, it should be recalled that billing costs could be
calculated f or only 4.6 million households, whereas all households responded to
perceived cost.) The most noticeable difference in the SMSA/nonSMSA comparison
was that 19.1 percent of SMSA households reported that the water supply was
expensive or very expensive, compared to 12.3 percent of nonSMSA households.
Median water costs varied only slightly according to the size of community
in which the household was located. In large rural communities, the median cost
per thousand gallons was $1.45; in small communities, it was $1.60; and in other
rural areas, it was $1.33. Likewise, perceived costs tended to rise with population
size. For example, about 14 percent of households in small rural communities and
other rural areas reported that the water supply was expensive or very expensive,
compared to 21.6 percent of households in large rural communities. In addition,
proportionately more households in large rural communities reported cost increases
during the precedLng year (44.4 percent, compared to 38.7 percent in other rural
areas and 31.8 percent in small rural communities).
-------
I - 51
The size-of-system comparison yielded differences which were not un-
expected. For example, households served by community water systems predomin-
ately paid f or their water by a regular billing (83.0 percent). The most common
way of paying for water among individual- and intermediate-system households was
by direct household operation and maintenance of the supply (about 95 percent and
65 percent, respectively).
Cost comparisons were restricted to perceived costs since only community
water systems in the NSA billed for their services. 3ust as perceived costs were
highest at households in large rural communities, they were also highest at
households served by community water systems. It was possible, however, that
rural residents were not aware of the costs involved in the operation and
maintenance of individual supplies, whereas a regular billing was a reminder of the
direct cost associated with water consumption.
AFFORDABILITY
In the NSA, affordability was measured in two ways. First, a ratio of billed
cost to total household income was multiplied by 100 to calculate the percent of
household income paid for water. Second, NSA respondents were questioned about
whether—if offered a supply that was good to drink and provided as much water as
needed—they would be willing to pay more, or the same amount they were
currently paying for their existing supply. The assumption was that respondents
who would be willing to pay either more or the same felt that their present water
supply was affordable.
For the 4.6 million households which could provide billing information, the
proportion of household income paid for water ranged from 0.04 percent through
15.60 percent. One-quarter of these households paid 0.30 perce. t or less of
house ’ old income for water. Half paid 0.60 percent or less. Thre ‘-quarters paid
0.99 percent or less, and one-quarter paid 1 percent or more.
-------
I - 52
In terms of the second measure of affordability, a willingness to pay either
the same or more for the ideal supply was reported in 95.8 percent of all rural
households (21 million). A willingness to pay less was expressed at 3.6 percent of
rural households (0.8 million). This response may have signalled a supply which was
not affordable in light of other unidentified household expenditures.
With respect to regional differences, households in the Northeast paid a
smaller proportion of income for water than those in the South, West, or North
Central. Sepcifically, the median cost-to-income ratio in the Northeast was 0.38,
compared to about 0.60 percent in the South and North Central, and 0.73 in the
West. The SMSA/rtonSMSA comparison showed a more striking difference in the
cost of water in relation to household income, the ratio being much higher for
households located outside SMSAs. In fact, the median ratio for nonSMSA
households was twice as high as it was for households located within SMSAs (0.72
versus 0.36). Finally, the median cost-to-income ratio was much higher for large
and small rural communities (0.82 and 0.76, respectively), than for other rural
areas (0.52).
The regional comparison of perceived affordability showed that households
in the Northeast more often indicated a willingness to pay more for an ideal water
supply than households in the other three regions. Perhaps this was an indication
that they could afford to pay for a supply which was superior to the one which they
currently received. Likewise, a larger proportion of SMSA households were willing
to pay more for water.
-------
- 53
SUMMARY OF CHAPTER VI
Chapter VI considers selected aspects of the material presented in Chapter
V and incorporates this information into a number of composites or summary
measures. These composites or indices are designed to produce a set of values or
scores that can be used to distinguish rural households with respect to (1) the
quality of the water provided by their major supply and (2) the supply’s availability.
Indices for quantity and cost were not constructed because the data for these
status factors were not extensive enough to summarize collectively. Besides
disclosing patterns of variation among rural households, the indices developed in
Chapter VI also function as analytical devices in Chapters IX and XI. In short, the
indices discussed in Chapter VI were developed to reduce the information on
quality and availability to a level that was manageable within the NSA’S descriptive
and analytical framework.
The formation of each composite involved combining variables according to
a particular set of principles. These principles suggested how the variables were to
be aggregated, given the substantive purpose of the index. The purpose of each
index was determined by the information that was being summarized and the
complexity of the factor being measured. Referents for water quality included
concentrations of physical substances that were assayed by laboratories as well as
selected perceptions of rural people. Availability, on the other hand, encompassed
material about recorded water supply pressure, the distance between the household
and the point at which water was withdrawn from the source, and supply
breakdowns. To emphasize the distinctive aspects of these data, four indices were
developed for water quality, two composites for availability.
Three of the four qualitj indices were based on NSA laboratory findings.
The first of the three indices was keyed to the potential health signilicance of
-------
I - 54
relevant findings. The second was geared to aesthetic and economic implications
of the findings. The third was keyed to the general expression of water quality,
whether related to health, aesthetic, or economic considerations. In contrast, the
fourth water quality index was keyed to perceptual data obtained in the NSA.
Except f or the general index, which comprised the complete set of indicator
constituents chosen for the indexing effort, the composites were independent or
mutually exclusive. Therefore, each index was unique and satisfied a need that
could not be fulfilled by the other three. Each index was further distinguished by
the method used to combine the constituents and the interpretation of the index
scores, although all four quality indices permitted the same relative comparisons.
In spite of their dissimilarities, the three water quality indices constructed
from the laboratory data revealed generally similar trends within the subnational
groupings (the findings for the nation were difficult to compare since the purpose
of each index was different). Scores for all three indices tended to be higher for
households in the North Central, lower for households in the Northeast. These
results suggested that household water supplies in the North Central were more
likely to have health risk, adverse aesthetic and economic effects, and lower
general quality than household supplies in the South, West, or Northeast. Another
implication of these findings was that households in the North Central and West
tended to have water quality which was lower than the nationai average.
Two of the three composites also showed generally similar trends in the
size-of-system grouping. Households using intermediate systems had higher values
in the health risk and general water quality indices, while households using
individual systems or community systems had lower scores. This indicated that
households using intermediate systems received water of lower quality than
households served by individual or commLruty systems—’although there were no
appreciable differences on the aesthetic/econumic index according to the size-of-
system grouping. Despite these trends, scores for the three indices generally
-------
I - 55
varied little in the SMSA/nonSMSA and size-of-place groupings. The one exception
was that scores on the general water quality index indicated that households in
small rural Communities received lower quality water than did households in large
rural communities or other rural areas.
The preceding patterns of variation were different from those associated
with the perceived quality index. Regional distributions suggested that, on the
average, households in the South tended to have somewhat lower scores than
househQlds in the Northeast and North Central, and that households in the West
tended to have higher scores. This indicated that households in the South
apparently perceived the quality of their water supplies to be slightly better than
did households in other regions; households in the West perceived the quality of
their water supplies to be lower than did households in other regions. Additionally,
while the perceived quality composite revealed no appreciable differences among
SMSA versus nonSMSA households, the composite indicated that households in other
rural areas perceived the quality of their water supplies to be better than did
households in small and large rural communities. Finally, the direction and
magnitude of differences on the index showed a pronounced tendency for the
quality of water supplies to be perceived as better in rural households with
individual systems than in households using intermediate or community systems.
The two composites of availability were developed to measure the reliabil-
ity and accessibility of rural water supplies. Consistent with the approach in
Chapter V, the reliability index drew on information about minor and severe supply
breakdowns, while the accessibility index used data on water pressure and the
distance between the household and the point of withdrawal. Since each index
measured a unique aspect of availability, they were independent.
According to the reliability index, rural water su?plies tended to be very
reliable when considered nationally. Since the scores on he accessibility index
were generally low, rural water supplies were also accessible. Subnationaily,
-------
I - 56
however, the indices generally disclosed different patterns of variation. The
supplies of households in the South were less reliable than the supplies of
households in the other regions, but supplies in the North Central were most
inaccessible relative to the supplies of households in the Northeast, South, and
West. Similarly, the supplies of households in large rural communities were more
accessible, while the supplies of households in small rural communities were
somewhat less reliable than those in large rural communities or other rural areas.
The only parallels between the two indices were in the size-of-system comparison,
where the supplies of households on intermediate systems were least reliable and
accessible, and for the SMSA/noriSMSA contrast, where only trivial differences
were detected.
-------
I - 57
SUMMARY OF CHAPTER VII
In Chapter VII, the scope of the NSA report is expanded to include various
physical and technical aspects of water supplies and their organizational features.
This examination is done in the context of water systems, since the systems have a
strong influence on the size and technical features of the supplies. Chapter VIII,
dealing with the economics of water supply, complements this chapter and the two
chapters represent the extent of the NSA investigation of supply systems.
The systems are typified according to three major system-size categories
—individual, intermediate, and community systems. As noted in Chapter II, the
individual systems served single households, the intermediate systems had two
through fourteen connections, and the community systems had fifteen or more
connections. i s to numbers of the respective systems at the time of the NSA
survey, there were 8,765,000 individual systems, 845,000 intermediate systems, and
34,000 community systems providing water to rural US. Although individual
systems were by far the most numerous, intermediate and community systems
served proportionately more rural households, since each of those systems had
multiple connections, whereas each individual system had only one connection.
Specifically, community systems served 10,981,000 households, compared to
8,765,000 households served by individual systems and 2,228,000 households served
by intermediate systems.
INDIVIDUAL SYSTEMS
The profile f individual systems was developed so that certain overall
feati’ es, such as source and ownership, were presented first. Then engineering
fea ures which generally were external to the nousehold and which were related to
selected aspects of water supply (principally to water quality) were described.
-------
I - 58
Next, the system was traced from the external engineering features to findings
regarding water transmission to considerations within the household structure, such
as types of water treatement devices, which seemed relevant to the status of rural
water supplies. Finally, the household experience with laboratory assessment of
water quality was investigated.
Major findings of these investigations are presented in this summary. The
diverse nature of the material in Chapter V II requres considerable selectivity in
the summary here. Thus, only the major findings are presented emphasizing the
nation as a whole, although some references are made to regional, SMSA versus
nonSMSA, and size-of-place findings. Furthermore, findings which require detailed
qualification and explanation generally are omitted here to avoid the possibility of
oversimplification —a problem which otherwise would be particularly acute in
relation to the subnational findings. A further qualification is that, as was true in
Chapters V and VI, the water supplies described in Chapter VII are the major
household supplies.
As to the findings related to overall features of individual systems:
• Groundwater (from wells or springs) was the source for fully 93 percent of
the 8.8 million individual systems in rural America. Groundwater use ‘was
particularly prevalent in the Northeast and North Central.
• Although groundwater usually was associated with wells, 275,000 individual
systems had springs. Most (about 93 percent) of the springs were located either in
the Northeast or the South, and all of the springs were in other rural areas.
• About 2 percent of all individual systems—those with cisterns—had
precipitation as a source.
• Only about 1 percent of all individual systems had a surface water source.
• About 4 percent of individual systems used purchased bottled or hauled
water I or which the sources could not be determined.
-------
I - 59
• Nearly all individual systems—97 percent—were owned by individuals. A
few systems—about 1 percent—were owned by private firms. For the remaining 2
percent of systems, there was a variety of arrangements, ranging from no
ownership to ownership by a cooperative.
• The median number of years during which households were served by
individual systems (for those systems for which information could be obtained) was
7.1 years. The majority of systems had served households for ten years or less.
As to specific external engineering features, those associated with wells
were reported here first, those associated with springs next, those relevant to
cisterns after that, and those relevant to surface water supplies last. Features of
individual systems using purchased bottled or hauled water were not considered
since the sources could not be determined. Major findings about features
associated with wells were:
• About 90 percent of the 8.8 million individual systems in rural America
relied on wells.
• Individual weliheads most often terminated above ground—frequently
under protective covering. However, although above-ground-termination was used
for about 60 percent of wells across the nation, this arrangement was found at only
about 47 percent of wells in the Northeast and 51 percent of wells in the North
Central. In the South, on the other hand, 77.0 percent of wells terminated above
ground.
• Less than one-fifth of all weliheads were in well pits, an arrangement
which was particularly rare in the South, where only 10.9 percent of wells had well
pits.
• Less than one-fifth of wellheads in the nation were entirely buried,
although the proportion of bun ‘d wells rose to 31.0 percent ti -i the Northea ,t.
-------
I - 60
• Nearly six of every ten individual wells were drilled. However, the
proportion of drilled wells was considerably smaller in large rural Communities
(about 38 percent of wells there) than in small communities or other rural areas
(both of which had about the national average).
• Only 12.7 percent of wells were driven, and a similar proportion were dug.
The proportion of driven wells, however, was much greater in large rural
communities (about 42 percent of wells there) than in small communities or in
other rural places (about 10 and 12 percent of wells in those places, respectively).
As to dug wells, the proportion was larger for nonSMSA systems than for SMSA
systems—14.3 percent versus 8 percent.
• About 6 percent of wells were either bored or jetted; the type of
construction could not be determined for nearly 11 percent of individual we1ls
• The median depth of individual wells f or which information was available
(about 88 percent of all wells) was 26.8 meters (87.9 feet) from ground level to
bottom of shaft.
• About 39 percent of the 6.8 million drilled, driven, bored, or jetted wells in
rural America were not grouted; about 35 percent were grouted; grouting status for
the other 26 percent of wells could not be determined. (Grouting is a pollution
countermeasure employing cement to fill part of the space between the casing and
the surrounding earthen bore hole. However, groutirig may not be necessary for
driven or jetted wells.)
• About 62 percent of all individual wells (whether drilled, driven, dug,
bored, or jetted) had adequate arrangements for sealing the wellhead from solid or
liquid pollutants. Overall, sealing was adequate in a somewhat larger proportion of
wellheads in the North Central than in the other regions.
• It appeared that potential pollution sources generally presented little
threat to individual rural well systems. About 90 percent of individual wells were
-------
I - 61
located more than sixteen meters (about 50 feet) from most potential pollution
sources.
As to individual springs (for which engineering information was much more
limited than for wells):
• The majority of the 275,000 individual rural springs used for major
domestic water supplies were inadequately protected from Contamination: reser-
voirs often had inadequate covers, the source tended to be inadequately shielded
from pollution, and the spring pool itself tended to be subject to contamination
which could be passed on to the domestic supply.
As to engineering features associated with individual cistern systems:
• Many of the 133,000 individual cistern systems used for major domestic
water supplies in rural America appeared to be pollution prone. For example,
nearly six of ten systems had no first-water diverter (to protect the supply from
debris in incoming water). Then, cleaning drains were not installed in 76 percent of
the systems. Finally, covers were unacceptable for excluding contamination in
30.4 percent of the systems.
As to engineering features associated with individual surface water
systems:
• Of the 93,000 individual systems in rural America supplied by lakes,
streams, ponds, and the like, about 41 percent had adequate source protection, and
the balance had inadequate protection.
As to general source-to-household considerations or all .8 million mdi-
‘idual systems (regardless of source or type of supply), ma or NSA findings were:
-------
I - 62
• In 91.0 percent of all individual systems, water was piped to the households
by electric pump. Only 0.8 percent of supplies had hand pumps. Natural pressure
(gravity flow or artesian pressure) provided the motive power in about 2 percent of
systems. In 5.5 percent of all individual supplies, water was either transported in
bulk to the household or else carried there by hand. Other methods, or unknown
methods, were utilized in less than 1 percent of all systems. Whereas nearly all
SMSA systems (about 98 percent) used pumps for bansport to the household, other
methods were employed for about 10 percent of nonSMSA systems. Propulsion by
electric pump was the only method for bringing water into the house in large-rural-
community systems.
• Among the various possible devices used to propel, pressurize, store, or
treat water, the two devices found most frequently in individual systems were the
water pump (in 92.5 percent of all systems) and the pressure tank (in 89.2 percent
of all systems). As opposed to the pressure tanks, which enhanced the efficiency of
power pumps, storage tanks were rare (in only 7.8 percent of systems). Among the
various devices used to treat water, only the water softener (in 17.6 percent of
systems) was used in more than a small percentage of systems. About 75 percent
of all softeners in individual systems were in the North Central—a finding which
was consistent with the discovery that the median water hardness value in the
major supplies of the North Central was larger than for- any other region (see
Chapter V). Also, about 28 percent of SMSA individual systems had one water
softener, but only about half that proportion of nonSMSA systems had softeners.
• Some type of free assistance or advice related to water systems was
obtained for 22 percent of all individual systems. The most frequent type of
assistance was free chemical or biological testing of the supply. Among the
subnatio al findings, the most striking discovery was that nearly two-thirds of all
large-rural community systems had received some type of aid: this proportion was
-------
1 - 63
almost three times that for systems in other rural areas and sixteen times that for
systems in small rural communities.
• Assistance or advice related to water systems were obtained most often
from local or state agencies or health departments, from individuals, or from water
softening companies. Assistance was rarely obtained from federal sources, stores,
contractors, and local plumbing companies.
As to tests done professionally for water quality in individual systems:
• Approximately 37 percent of all rural households with individual systems
had had the water tested at least once. In the South, however, water tests had
been conducted in only 27.6 percent of the households—compared to at least 40
percent in each of the other three region . Water quality testing also occurred
somewhat more often in SMSA (43 percent) than in nonSMSA households (35
percent). In addition, testing was done much more often in large communities (44.4
percent) and in other rural areas (37.5 percent) than in small Communities (20.2
percent).
• The frequency of testing varied considerably: in about 0.6 percent of all
households in which testing had been done, the water had been tested 70 times or
more, while in 83.5 percent, the water had been tested only once or twice.
• Nearly 70 percent of the most recent tests (that is, the last test performed
in each household) were done during 1970-79.
• Two prominent reasons for testing were (1) to determine if a system
alteration was needed (in 49.5 percent of all households testing) and (2) to check a
newly installed system (in 26.2 percent of households testing).
• Tests of water quality most often were requested by members of the
household (in about S percent of all households testing), and this was the
prominent source of reruests for testing regardless of the households’ regional,
SMSA versus nonSMSA. and size-of-place status.
-------
I - 64
• As to the specific type of the most recent test, bacteriological tests,
either singly or in combination with chemical tests, were done in 53.0 percent of
all households testing; chemical tests, either singly or in combination with
bacteriological tests, were done in 38.8 percent of all households testing. The most
prominent regional differences were in the Northeast, where the proportion of
households with single or combined bacteriological tests was 68.6 percent, and in
the South, where the proportion of single or combined chemical tests was only 23.3
percent.
• According to the verbal reports from household members, the levels of
bacteria were within applicable standards at 52.3 percent of households at which
the relevant tests were done: the levels exceeded the standards at 3.1 percent of
the households; the results were not known at 44.5 percent of the households.
Similarly, results of the chemical tests were unknown at about 64 percent of
households at which the tests were done. At the households where the test results
were known, 2 percent or less reported unacceptable levels of nitrate, sulfur,
calcium, sodium, and arsenic, while 5.5 percent of these households reported
unacceptable iron levels.
INTERMEDIATE SYSTEMS
The intermediate systems (those with two through fourteen connections)
usually were centered at some rural household rather than at a separate facility.
Intermediate systems thus were similar to individual systems rather than to
community systems. On the other hand, intermediate systems potentially had some
of the characteristics associated with community systems (those with more than
fourteen connections). The NSA profile of intermediate systems thus was
developed from data obtained at bot i the household and the system.
The order of presentation in this summary combines features of the
individual-system summary, just completed, and features of the community-system
-------
I - 65
summary, which follows. As is true in the summaries for individual and for
community systems, only selected findings are given here. In particular, external
engineering findings are omitted for intermediate systems which used springs or
surface water since the source of that information must be qualified in some
detail—a task beyond the scope of this summary. Furthermore, as was true for
individual systems, external engineering features of intermediate systems using
purchased water were not considered since the source could not be determined. As
a result of these considerations, external engineering features are summarized here
only for intermediate systems with wells.
As to findings related to overall features of all intermediate systems:
• Of the 845,000 intermediate systems serving rural America, the largest
proportion (46.3 percent) was in the South (the regional distribution generally
paralleled the regional distribution of rural households).
• About 91 percent of intermediate systems had only two or three connec-
tions. Larger intermediate systems (with four through fourteen Connections) were
uncommon, and were found more often in the Northeast (12.1 percent of inter-
mediate systems there) and West (13.8 percent) than in the North Central (6.4
percent) and South (7.4 percent).
• About 80 percent of all intermediate systems had only one kilometer or
less of distribution pipelines, and about 1 percent of systems metered service
connections.
• Groundwater, primarily from wells rather than from springs, was used by
92.1 percent of intermediate systems. About 8 percent of intermediate systems
used surface water, and about 1 percent used either precipitation or purchased
water. Only about 1 percent of all intermediat ’ systems had a combination of
sources. The most striking subnational-data finding was that all systems serving
households in either large or small rural communities used only groundwater.
-------
- 66
• About 90 percent of intermediate systems were owned by individuals (one
of many characteristics shared with individual systems). Despite the predominant
individual ownership, there was some regional variation: cooperative ownership was
reported by 12.9 percent of systems in the Northeast, compared to 2.5 percent of
systems in the South. Ownership by private firm was reported by 8.1 percent of
systems in the West, compared to no more than 1 percent of systems in each of the
other regions.
• As to age (determined either by the number of years since original
construction or by the number of years since subsequent major alteration), the
median for intermediate systems was seven years. Systems in the South tended to
be newer than systems in other regions, and the median age of systems serving
large-community households was only 1.8 years, compared to 7.3 years for systems
serving other-rural-area households.
• For the 20 percent of intermediate systems which could provide the
necessary information, the median average daily use (production) of water was 758
liters, or about 200 gallons, and the median maximum daily design capacity was
44,000 liters (11,609 gallons), nearly 60 times the median daily average production.
As to specific external engineering features of intermediate-system wells
(examined at the household in the NSA, as for individual systems): -
• About 88 percent of the intermediate systems relied on wells. An
estimated 68 percent of the wellheads terminated above ground, generally under
the protection of a structure such as a pumphouse. Only 39.0 percent of Northeast
wells terminated above ground, however, compared to 78.5 percent of Southern
wells.
• About 12 percent of intermediate wells were in pits only about half of
which had watertight covers. The well pit arrangement was mo t common in the
Northeast, least common in the South.
-------
I - 67
• About 1 2 percent of intermediate wells were buried, and about 3 percent
were located inside the dwelling unit.
• About 65 percent of all intermediate wells were drilled, and much smaller
proportions were bored (8.9 percent) or dug (8.3 percent). Proportionately few
wells were driven (3.0 percent) or jetted (2.1 percent), and the type of construction
could not be determined for about 13 percent. Bored wells were most common in
the South (in 16.8 percent of wells there), and dug wells were most common in the
Northeast (in 17.2 percent of wells there).
• Median depth of intermediate wells for which information was available
(about 85 percent of all wells) was about 28 meters (92 feet)— .-similar to the
median depth for individual wells. The wells tended to be deepest in the West (a
median depth of 38.5 meters, or about 126 feet), and shal.lowest in the North
Central (a median depth of 23.9 meters, or about 78 feet). On the average, SMSA
wells were deeper than nonSMSA wells: the median for SMSA wells was 38.2 meters
(about 125 feet), compared to 29.8 meters (about 98 feet) for nonSMSA wells.
• Of the 683,000 drilled, driven, bored, or jetted intermediate wells, almost
40 percent were not grouted, 35 percent were grouted, and the grouting status of
25 percent could not be determined. These percentages are essentially the same as
found for individual systems. Again, grouting may not be necessary for driven or
jetted wells.
• According to three separate considerations explained in Chapter VII,
arrangements for sealing wells against external contamination were adequate in
two-thirds of all intermediate wells. The North Central had the highest proportion
of wells judged acceptable for all three considerations (83.6 percent), the South had
the lowest proportion (57.6 percent).
• R-ughly 85 percent of all intermediate wells were more than ixteen
meters (about 50 feet) from potential pollution sources.
-------
I - 68
As to general source-to-household considerations for all 845,000 inter-
mediate systems (regardless of source), major NSA findings were:
• Most intermediate systems (about 91 percent) distributed water by electric
pump. A relatively small number of systems (about 50,000) relied on gravity flow,
and a few systems (about 3,000) used artesian pressure.
• As to treatment of the water, the supply in intermediate systems generally
was untreated. Softening—the most common treatment process—occurred in only
6.0 percent of systems. Activated carbon or anthracite filters were installed and
functioning in about 1 percent of intermediate systems. Only 2.8 percent of
intermediate systems chlorinated the water, and the chlorination equipment was
not operable at all of the systems which did chlorinate. Some type of treatment
was practiced most often in the Northeast—at 15.4 percent of systems in that
region. Comparable proportions were 10.0 percent in the South, 8.2 percent in the
West, and 7.6 percent in the North Central.
As to manpower used to operate intermediate systems and system perfor-
mance:
• About one-third of the systems had no equipment operators, and nearly half
had only one operator. More than hall of the intermediate systems in the North
Central had no operators, compared to about one-quarter of systems in the South
and West. Whereas more than half of systems serving large-community and small-
community households had no operators, only about 34 percent of systems serving
other-rural-area households were in this status.
• Of the 537,000 intermediate systems with operators, only 7,000 had
operators who were paid, full-time employees. Most systems (84.6 percent) had at
least one volunteer operator, who received no payment. Where the work was not
done by volunteers, it most often was undertaken by paid, part-time operators.
-------
I - 69
• Since most systems with operators relied on volunteers, special training in
the operation of system equipment was rare. Special training was reported for the
operators at only about 3 percent of systems with one or more operators.
• Intermediate systems required little operation time: about 83 percent of all
intermediate systems needed less than one man-day per month (less than eight
hours).
• As to system performance, about three of every four intermediate systems
reported no breakdowns during the year before the NSA. The breakdowns which did
happen most often were attributed to pump malfunctions or to ruptured lines.
Those systems reporting breakdowns usually had no more than two, but nearly all of
the breakdowns were severe: that is, they resulted in loss of service to water users
for more than six hours. The most prominent subnational finding was that systems
serving households in other rural areas were more reliable than those serving
households in communities. Specifically, nearly 76 percent of the former system
had been free of breakdowns, compared to about 57 percent of systems serving
small-community households and about 34 percent of systems serving large-
community households.
• Only about 7 percent of intermediate systems in the nation had difficulty
obtaining water from the source. As to further detail, fully 94.0 percent of
groundwater systems had no difficulty, compared to 79.7 percent of surface-water
systems—indicating greater reliability for groundwater.
As to water quality tests done at intermediate systems:
• Bacteriological testing was infrequent at intermediate systems; the testing
was done at only 22.4 percent of all eystems, and although the percentage rose to
more than twice thet in Western systems, testing was never done at systems
serving large-comm”nity households. Of those systems that test, testing was done
no more than once a year at about 61 percent of systems testing; 29 percent tested
-------
I - 70
more frequently; and the frequency of testing could not be determined for an
additional 10 percent.
• Most bacteriological tests for intermediate systems were done by state or
local health departments, by state environmental departments, or by private
laboratories.
• At the 131,000 intermediate systems which had performed bacteriological
tests within a year of the NSA and which could provide test results, 91.1 percent
had results which met prescribed levels for all tests (a maximum of twelve, as
explained in Chapter VII), and 2.8 percent had results which met prescribed levels
for more than 90 percent of the tests.
• Only 9.7 percent of intermediate systems conducted chemical and physical
tests. The Northeast had the greatest proportion of intermediate systems
performing chemical tests (23.9 percent), the North Central had the smallest
proportion (4.6 percent).
• At nearly 60 percent of intermediate systems at which chemical or
physical tests had been done, the tests had been performed during the preceding
three years. The earliest year for testing was 1955, which may have been the year
of original construction. Results for the tests were too few to assess in the NSA.
COMMUNITY SYSTEMS
In the NSA, community systems initially were defined as those which had
fifteen or more service connections or which served more than 25 people. In
Chapter VII, however, and in the NSA analysis generally, the definition was limited
to systems which had fifteen or more service connections, since it was considered
difficult for system representatives to report accurately on the number of people
served by the systems.
The physical and economic configurations of the community systems were
too complex to b conveyed by the one dimension of size, however. For this and
-------
I - 71
other reasons, community systems were classified as either “independent” or
“consolidated.” The independent systems, which comprised nearly 90 percent of all
Commurutv Systems, were those self-contained systems which performed all
functions associated with the provision of water. The independent system thus
secured water from its own sources, treated the water, distributed it, and managed
its own organizational and financial affairs.
The consolidated system, on the other hand, relied on interdependent links
with other organizations. In one type of consolidated system, for example, the sets
of facilities would be located far enough apart from each other to require local
water supply arrangements, but the facilities would be considered components of a
consolidated system if their management were accomplished within the overall
structure of a water district or similar administrative body. In another type of
consolidated system, one facility would purchase water from another organization
—a situation which existed at nearly 90 percent of the approximately 4,000
consolidated systems.
In this summary, as in Chapter VII, the independent and consolidated
systems are considered separately. The general order of presentation for each type
of system is similar to that used for individual and intermediate systems. Details
about supply-related engineering features are not addressed here or in the text,
however. Such an analysis for community systems would have been beyond the
scope of the NSA. Thus, overall organizational and physical features are described
first. Then the discussion shifts immediately to overall technical, administrative,
and water quality considerations.
As was true for the preceding summaries of findings for both individual and
intermediate systems, the main focus here is on the national findings. Some
references are made to regional, SMSA versus norSMSA, and s lzL-of -place findings,
but considerable selectivity is necessary because of the diverse, qualified nature of
many findings. Furthermore, no references are made to subnational findings
-------
I - 72
regarding consolidated systems. Those findings are omitted from Chapter VII for
reasons discussed in detail in the text.
INDEPENDENT SYSTEMS
As to findings related to overall features of independent systems:
• Of the 34,000 community systems serving rural households, nearly nine of
ten were independent. The distribution of these systems generally paralleled the
distribution of rural households; 80.5 percent of the systems served households
outside of SMSAs; 74.5 percent of the systems served households in other rural
areas; likewise, about 42 percent of the systems were in the South, 28 percent in
the North Central, and so forth.
• The median number of independent-system connections was about 59
nationally. Regionally, however, medians ranged from about 30 for systems serving
Western households through about 79 for systems serving North Central households.
The size_of-place range was even greater—the medians ranged from about 37 for
systems serving other-rural-area households through 544 for systems serving large-
community households.
• Most independent systems had very limited distribution networks (which, in
the NSA, included main pipelines but excluded service lines): 31.3 percent of all
independent systems had only one kilometer (0.6 miles) or less of pipeline, and the
national median length was 2.4 kilometers (1.5 miles). Again, the most notable
discovery among the subnational findings was that systems serving large-commun-
ity households had a median distribution line length of 15.3 kilometers (9.5
miles)—still limited in comparison to large urban systems, but relatively large in
comparison to systems serving small rural communities or other rural areas.
• Groundwater was a source for about 90 percent of systems, surface water
for about 1 2 percent (some systems used both sources, so the percentage is more
than 100). The use of surface water wa . most prominent in systems serving
-------
I - 73
households in the West (in 21.3 percent of those systems), systems serving
households outside of SMSAs (in 12.7 percent of those systems), and systems
serving large-community households (in 20.2 percent of those systems).
• Whereas individuals owned most of the individual and intermediate sys-
tems, public bodies owned about 37 percent of independent systems, and individuals
owned about 31 percent. Independent systems serving Western households,
however, were more often owned by individuals (42.9 percent of systems) than by
public bodies (20.9 percent). Independent systems serving other-rural-area house-
holds also were more often owned by individuals (41.9 percent of systems) than by
public bodies (18.7 percent of systems).
• Median age of independent community systems—as determined either by
the number of years since original construction or by the number of years since
subsequent major alteration of the systems—was only three years, less than half
the median age of intermediate systems.
• The median average daily use (production) of water was about 136,000
liters (35,928 gallons) f or those 19,000 independent systems for which information
was available. The median maximum daily production capacity of independent
systems was four times as large as the median average daily use, about 546,000
liters (144,000 gallons).
• The NSA results revealed a striking dichotomy in metering arrangements:
in nearly half (48 percent) of all independent community systems, none of the
connections were metered, while in a proportion nearly equivalent to the other half
(46.7 percent), 95 percent or more of the connections were metered. (Lesser
proportions were metered at about 5 percent of systems.)
• As to treatment processes, it was discovered that some type of teatmerit
was employed ‘t 60.6 percent of independent systems. Regionally, some treatment
was aone in near’y 71 percent of systems serving Southern households, but in only
about 41 percent of systems serving Western households. The most decisive
-------
I - 74
difference in the subnational findings was that treatment was used at 81.6 percent
of systems serving small-cqmmunity households, but only at 54.0 percent of
systems serving other-rural-area households.
• Although chlorination was the most frequently used of various specific
treatment processes, even chlorination was employed at only 57.5 percent of
independent systems. Fluoridation, aeration, coagulation, and flocculation were
used infrequently. Treatment was most prominent at independent systems which
served large-community and small-community households. For example, chlorine
was being added to the water at about 75 percent of those systems, but only at
about 51 percent of systems serving other-rural-area households.
• As to manpower used to operate equipment at the systems, most indepen-
dent systems (about 93 percent) had at least one operator, and nearly half had more
than one operator. All of the systems serving households in the Northeast had
operators, whereas some of the systems serving households in each of the other
regions did not.
• Most independent systems with operators (85.7 percent) had at least one
paid part-time or paid full-time operator, and only 25 percent had volunteer
operators. Jr -i contrast, most intermediate systems (84.6 percent) had at least one
volunteer.
• About 45 percent of all 30,000 independent community systems had at least
one operator who had received special training or schooling in the operation of
equipment at the system. Regionally, at least one operator had received special
training at about 60 percent of all independent systems serving Northeast house-
holds, but only at about 27 percent of all independent systems serving Western
households. At least one operator had received the training at 70.1 percent of all
systems serving large-comm’ nity households, compared to 42.8 percent of systems
serving other-rural-area househ3lds.
-------
I - 75
• The monthly operation of about 69 percent of all independent community
systems required at least one man-day of labor (equivalent to the work of one
person during eight hours). The median for independent systems (excluding those
for which information was not available) was 1.6 man-days per month. The most
prominent subnational finding was that, in keeping with their larger operations,
independent systems serving large-community households required the most labor
time. Specifically, the median number of man-days per month was 18.3 at systems
serving large-community households, 7.2 at systems serving small-community
households, and 1.4 at systems serving other-rural-area households.
• As to system performance, about 55 percent of independent systems
reported one or more breakdowns during the preceding year. The breakdowns were
caused most often by pump malfunctions or ruptured lines. Although a few systems
had numerous breakdowns, most had five or fewer. The incidence of breakdowns
was particularly prominent among systems serving Southern households: roughly 72
percent of those systems reported one or more breakdowns during the preceding
year.
• Of the approximately 16,000 independent systems with one or more
breakdowns during the preceding year, about 11,000 (approximately 70 percent)
reported that at least one of the breakdowns caused interruption of service to
customers. Systems serving Northeast households were more likely than systems
serving households elsewhere to have had service interruptions.
• The NSA results showed that 3,700 independent community systems (about
9 percent of all independent systems) sometimes had difficulties which interfered
with acquisition of water from the source. A total of 3,000 of the 27,000
independent systems (11.1 percent) which used groundwater had such difficulties
compared to 700 of the 3,000 systems (23.3 percent) which used surface water. In
other words, by this measure, groundwater was about twice as reliable a source as
was surface water.
-------
I - 76
As to water quality tests done at independent systems:
• Bacteriological testing was done at about 91 percent of independent
community systems (27,400 systems) compared to only 22 percent of intermediate
systems (189,000 Systems). Bacteriological testing was done at a much smaller
proportion of independent systems serving households in the West (74 percent) than
elsewhere. Tests were performed at more than 93 percent of systems serving
households in each of the other regions.
• Most bacteriological testing was done at least once every month. In
contrast, most testing at intermediate systems was done irregularly or no more
than once a year.
• In regard tc organizations conducting the bacteriological tests, most
independent systems relied on facilities outside of their own organizaitons, as was
true for intermediate systems. Most tests for both independent and intermediate
systerns were done ‘by state or local public health departments, by state environ-
mental departments, or by private laboratories.
• At 23,000 (92.4 percent) of the 25,000 independent systems for which test
results were available, all bacteriological tests (a maximum of twelve, as explained
in Chapter VII) showed results which met the levels prescribed by regulations
affecting the systems. At nearly all the rest of the systems (6.4 percent), more
than 90 percent of the tests showed results which met those levels.
• About 54 percent of the systems (16,000) reported that chemical tests
(including physical tests) were performed, but about 46 percent (14,000) reported
either that chemical tests were not performed, or that it was not known whether
tests were done. Chemical tests were performed more often in systems serving
households in the Northeast than in other regions, and in systems serving
households in large communities than households in small communities or other
rural areas.
-------
I - 77
• At nearly 70 percent of the systems at which chemical tests had been done,
tne tests had been performed during the preceding two years. The indication of
increased testing since 1975 was consistent with promulgation of the EPA’s
National Interim Primary Drinking Water Regulations on December 24, 1975.
Although results of the most recent tests were obtained in the NSA, those results
were not analyzed for this report f or reasons outlined in Chapter VII.
CONSOLIDATED SYSTEMS
As to findings related to overall features of consolidated systems (national
findings only):
• Although only about one of ten community systems was consolidated, the
consolidated systems tended to be considerably larger than independent systems in
terms of ni mber of connections. The median number of connections for the 4,000
consolidated systems was about 153—two and one-hall times the median number
for independent systems.
• Excluding the 450 consolidated systems (11.3 percent) for which the
information could not be obtained, the median length of consolidated-system
distribution lines was about 4.3 kilometers (2.8 miles).
• By definition, as explained in Chapter VII, purchased water was by far the
most common source f or consolidated community systems (in about 88 percent of
systems). About 11 percent of consolidated systems used more than one source of
water.
• Consolidated systems most often were owned by a public body. In fact,
proportionately more consolidated than independent systems (48.7 percent, corn-
r)ared to 36.8 percent) were owned by the public.
• The median age of consolidated systems—as determined c.ither by the
number of years since c.riginal Construction or by the number of y ars since
-------
I - 78
subsequent alteration of the system—was 5.5 years, compared to seven years f or
intermediate systems and three years for independent systems.
• In terms of water usage, the consolidated systems were only slightly larger
than the independent systems. Median average daily use of water was about
162,000 liters (42,797 gallons) for consolidated systems, compared to 136,000 liters
(35,928 gallons) at independent systems.
• About 74 percent of all consolidated systems, compared to only about 52
percent of independent systems, metered some connections. As was true for
independent systems, most consolidated systems which did meter connections did
so for 95 percent or more of the connections.
• Concerning treatment processes, only about 5 percent of consolidated
systems (about 200) reported using untreated water. In contrast, nearly 40 percent
of all independent community systems did not treat their water at all. The
difference may in large part have been determined by treatment practices at the
facilities which sold water within consolidated systems.
• Chlorination, the most frequently used treatment process, was utilized by
about 92 percent of consolidated systems—compared to about 58 percent of
independent systems.
• As to manpower requirements, about 96 percent of all consolidated systems
had at least one equipment operator. Furthermore, about 79 percent of consol-
idated systems had two or more operators, compared to about 49 percent of
independent systems and 16 percent of intermediate systems.
• The work Status of operators in consolidated systems also indicated a more
formal organizational arrangement than in independent or in intermediate systems.
Only 2.4 percent of consolidated systems had volunteer operators, compared to
about 25 percent of independent systems and 86 percent of intermediate systems.
The balance of the consolidated systems had operators who were paid either part
time or full time.
-------
I - 79
• Although a greater proportion of consolidated than independent systems
had part-time or full-time operators, only about 41 percent of consolidated
systems—compared to 45 percent of independent systems—had operators with
special training or schooling in the operation of equipment.
• Consistent with the higher operational staffing level in consolidated versus
independent systems, the consolidated systems required more labor time for their
operation. Excluding the systems for which information was not available, the
median man-days per month required to operate consolidated systems was 6.5—or
four times the median for independent community systems.
• As to system performance, about 50 percent of consolidated community
systems had breakdowns during the preceding year. This proportion was slightly
smaller than for independent systems (55 percent), but still considerably greater
than for intermediate systems (26 percent). The majority of consolidated systems
which did have breakdowns had five or fewer. Ruptured lines and pump mall unc-
tions were among the leading causes at consolidated systems, as they were at
intermediate and independent systems. About 1,800 of the 2,000 consolidated
systems which had breakdowns (about 88 percent) reported that at least one of the
breakdowns resulted in loss of service to customers.
As to water quality tests done at independent systems:
• Bacteriological testing was done at essentially all consolidated systems.
Also, testing was done more frequently at consolidated community systems than at
other multiple-connection systems serving rural America. T sting at monthly
intervals —or more frequently—was done at almost all consolidated systems which
tested, at about 85 percent of independent systems which tested, and at about 15
percent of ntermediate systems which tested.
• Ir ferms of organizations conducting the t sts, the majority of systems
relied on state and local health departments and state environmental departments,
-------
I - 80
as did both intermediate and independent systems. On the other hand, the
proportion of consolidated systems using their own laboratories exclusively (11 .4
percent) was larger than the proportion of independent systems (1.9 percent) or of
intermediate systems (0.0 percent).
• At about 3,800 consolidated systems (96.1 percent of those performing
tests), the results of all of the bacteriological tests met the prescribed levels. At
nearly all the rest of the systems (3.4 percent), more than 90 percent of the tests
had results which met the prescribed levels.
• Although essentially all consolidated systems tested for bacteriological
water quality, about 15 percent did not test for chemical (or physical) water
quality. In addition, it was not known whether chemical testing was done at 5.5
percent of systems. Most consolidated systems which tested for chemical water
quality performed the tests within a year of the NSA survey. Results of the
chemical tests were not analyzed in the NSA for reasons outlined in Chapter VII.
-------
I - 81
SUMMARY OF CHAPTER VIII
Potentially, there are a large number of attributes that may be used to
characterize the economic conditions of water systems. The NSA compiled data on
those aspects judged to be most important to the financial viability of the systems,
and the results are presented in Chapter VIII. The subjects addressed include
capital expenditures either for initial construction or for the most recent major
alteration, the general and particular arrangements for the financing of these
activities, revenues from water production, and operating and debt retirement
expenses. Although the information about revenue and expenses is limited to dollar
amounts, more detail is provided for financing arrangements. This detail Consists
of the source of the capital, the amount of funds acquired, the interest rate for the
funds, and the length of repayment periods. As to the sources, information is
provided about internal mechanisms, such as reserve funds or accumulations of
surplus, and external modes of support, such as loans and grants from public
entities, municipal bonds, and commercial loans. An effort also is made to
establish the relative prominence of each source and the extent to which each
subsidizes the operations of water systems serving rural America.
For a number of reasons, but most importantly the limited amount of
information available to describe intermediate water systems, the economics of
community systems are presented first, followed by similar data on intermediate
systems. Because of the relatively small size of the NSA case base, no distinction
is made between independent and consolidated commuruty systems, a discrimina-
tion made in Chapter VII. For the same reason, the information is limited to the
nation as a whole: subriational groupin;s are not considered.
-------
I - 82
COMMUNITY SYSTEMS
Construction costs, which varied substantially among community systems,
ranged from a low of $90 to a high of $35 million when expressed in 1978 constant
collars. (Construction costs were made either for original Construction or for
subsequent major alterations, as already noted.) Systems most often financed
these activities with capital secured from internal sources (47.0 percent of
systems), less often from external sources, (39.2 percent of systems), and least
often from a combination of internal and external sources (9.7 percent of systems).
Although community systems financed capital projects with internally generated
funds, the amounts invested tended to be smaller, as was indicated by the median
amount of $6,000, compared to much larger medians for capital from external
sources, as noted below. When the systems had a larger requirement for capital,
they were inclined to seek some external source. Of these external sources, loans
from government agencies were most prevalent, followed by commercial loans,
government agency grants, and municipal bonds.
Community systems that supported capital expenditures with loans and
grants from public entities tended to receive them from some organization
affiliated with the federal government. State agencies and other public bodies,
such as special districts or water authorities, also were involved in providing loans
to community water systems. Grants were made available by all these sources as
well as by regional agencies. Of these federal organizations, the Farmers Home
Administration (FmHA) was the most prominent source of loans arid grants,
followed by the Economic Development Administration (EDA). Some systems also
secured loans or grants from the Department of Housing and Urban Development
(HUD), and a few others received grants from the Department of Defense (DOD).
On the average, loans from HUD (mean of $609,000, median of $280,000)
were the largest, followed by those from FmHA (mean of $288,000, median of
$175,000), and EDA (mean of $228 000, median of $237,000). Generally, state
-------
I - 83
agencies lent smaller amounts of funds in comparison to the other organizations
from which community systems borrowed capital. Data on standardized rates
indicated that systems with loans from FmHA paid a slightly higher mean interest
rate than systems with loans from HUD or EDA, although a lower median interest
rate. Another aspect of the situation was that a larger proportion of systems with
FmHA loans were being subsidized (paying less annually than the value of the loan).
At the other extreme, however, a higher percentage of other systems that had
loans from FmHA appeared to be paying very high capital costs ($50 to $100 per
every $100 of loan value).
Regarding grants, although they did not provide support to as many
community systems, grants awarded by regional agencies tended to be for larger
amounts than those received from the three federal agencies (FmHA, EDA, and
HUD). Considering only these three organizations, grants from HUD (median of
$164,000) and EDA (median of $159,000) were slightly larger than those obtained
from FmHA (median of $105,000). Grants from regional agencies (median of
$172,000) were larger overall than those acquired from the federal organizations,
while grants from state agencies (median of $61,000) were the smallest.
Many community systems supported their capital projects by issuing bonds
or securing loans from commercial establishments. The bonds tended to be issued
for relatively large amounts, with some exceeding $10.0 million. Standardized
interest rates were comparable to those for federal loans, but a smaller proportion
of systems received the equivalent of a subsidy from the bond—holders (determined
by an estimated discount rate which, in simplest terms, compensated for the fact
that a dollar currently in hand had a greater present worth than one promised after
a time lapse—such as after a bond matured). Commercial loans were smaller ‘n
the average, and the systems that obtained capital from this scurce also paid the
highest interest rates.
-------
I - 84
One objective of the NSA was to determine the amount of water-related
revenue obtained by systems. The revenues (which unintentionally may have
included other income, such as from taxes) differed substantially among community
systems. Systems which charged for water, and for which information was
available, reported revenues as low as $400 a year and as high as $12.0 million (the
median was $6,000 yearly). About 18 percent of the systems did not charge for
water, and therefore they had no water-related revenues. Annual revenues could
not be determined for a substantial proportion of the systems (about 15 percent).
While they varied appreciably, revenues were occasionally not sufficient to
support all expenditures. Fully 10 percent of community systems received less in
revenue than was allocated to the combination of operating and maintenance Costs
and debt retirement expenses. Operating and maintenance costs tended to be
substantial, although 3 percent of the systems reported that no funds were
expended to support these activities. Debt retirement expenses tended to be small
since about 33 percent of the systems did not allocate any money for this cost.
Consistent with the considerable size differences among these systems, some had
annual operating and maintenance costs of $7.3 million, while others had debt
obligations of $ .0 million. Most significantly, information on operating and
maintenance costs was not available for about 37 percent of the community
systems. Similarly, data on debt costs were not available for about 40 percent of
the systems.
INTERMEDIATE SYSTEMS
Construction costs, which varied substantially across intermediate systems,
ranged from a low of $20 to a high of $1.5 million when expressed in 1978 constant
dollars. Average costs were about $5,100, while the median of the distribution
centered on $1,400. Co. struction activity among intermediate systems was mainly
financed through reserve funds (about 66 percent), although assistance was
-------
I - 85
sometimes secured from an external source (6 percent), or a combination of
internal and external sources (less than 1 percent). The source of financing could
not be determined for the remaining 28 percent.
The amounts of capital invested from internal sources were not as wide-
ranging when compared to construction costs without regard to the specific source
of financing. These internal amounts, which are also expressed in 1978 constant
dollars, ranged from a minimum of $20 through a maximum of $63,000. The
median amount invested from internal sources was $1,200, while the average
amount invested was about $2,100.
Among the 30,000 intermediate systems that received external monies for
water system development, funds were most often secured from commercial
establishments (about 60 percent), but were sometimes obtained from federal
agencies in the form of loans (about 37 percent) or grants (about 3 percent).
Commercial financing was modest in scale, ranging from $600 through about
$15,000. The mean and median amounts borrowed from banks ($3,300 and $2,300,
respectively) were somewhat larger than the comparable amounts invested from
internal sources (see above). The amounts of funds secured from federal agencies
were not detailed here or elsewhere, since only three intermediate systems
appearing in the NSA sample were involved.
As to water-related revenues, the most significant finding was that fully
93.5 percent of all intermediate water systems did not charge for the provision of
water. Those that did, reported revenues between $1 and $1,500, although only a
few systems claimed revenues in excess of $100. Likewise, operation and
maintenance costs rarely exceeded $100. However, a substantial proportion of
systems (40 percent) which reported revenues indicated that, at least during the
year preceding the NSA survey, no funds were expended to support operation and
maintenance activities. Finally, few intermediate systems were retiring any debt.
Among those that reported annual revenues, about 88 percent had no debt costs; 9
-------
- 86
percent could not provide any information on this Cost Component; and for the few
systems that allocated a proportion of annual revenues to debt retirement, small
amounts were expended.
-------
I - 87
SUMMARY OF CHAPTER IX
The purpose of Chapter IX was to introduce the analytical segment of the
NSA report. Each of the dimensions of status of rural drinking water was
introduced along with the specific measures used in the analysis of each of these
dimensions. A rationale was given to support the use of regression analysis for
identifying independent variables which had an association with each of the
dimensions of rural water supply. Finally, the structure of analysis was detailed.
Several constraints and desired objectives for the analysis dictated the
selection of multiple regression as the common analytical framework. Some of
these constraints were: (1) wide range of content of the dependent variables, (2)
complex phenomena encapsulated in the dependent variables, (3) irregularity of the
distributions of some of the dependent variables, (4) use of several independent
variables in modeling, (5) use of both categorical and Continuous independent
variables, (6) unequal number of cases in various categories of the independent
variables, arid (7) need for a standard, common analytical framework to ease
communication of results.
As to the structure of analysis, independent variables were selected based
on their expected relationships with the dependent variable under investigation.
Bivariate relationships between the candidate independent variables and dependent
variable were examined and interpretations made. Only those variables which met
or exceeded the ref eren:e value used to determine analytical significance
(R 2 . 0.03) were retained for intermediate model development. Intermediate
models were constructed for the dependent variable with each of the three sets of
independent variables—household characteristics, supply attributes, and system
properties. Th last stage of the analysis was to develop a final overall regression
model which incorporated all the variables contained in the intermediate models
-------
• 88
and the variables used in the design of the NSA sampling frame (region, SMSA/non-
SMSA, size of place). The stages of analysis are illustrated in Figure I-I.
In the case of quality, extended analyses were done on households served by
wells and contingency tables were used to detail the conditions associated with
totai coliform contamination.
-------
- 89
Figure 1-1
Strategy for Analysis of Each Dependent Variable
STEP I:
STEP 2:
STEP 3:
STEP 4;
Select
for Each
Dependent Variables
Dimension
Assess Bivariate Relationships
Between
the Dependent Variable and Independent
Variables in Each of Three Subsets
(House-
hold, Supply, System)
*
Develop intermediate Model
Using All Retained Household
independent Variables
Develop
intermediate
Model
Using All
Independent
Retained
Variab..es
Supply
Select independent Variables
nudged to be Meaningfully Re-
lated to Dependent Variables
Group Independent Variables
According to Whether They
Mainly Reflect an Attribute of
the Household, Supply, or
System
Retain for Intermediate Model
Development
If
Exclude Independent Variable
from Further Model Develop-
ment
Develop Intermediate Model
Using All Retained Supply
System independent Variables
Specification of Final Model for Dependent
Variable Using All Retained independent
Variables from Each Intermediate Mode!.
Final Model Also Includes Sampling Design
Variables (Pegion, SMSA/nonSMSA, Place
Size)
-------
- 90
SUMMARY OF CHAPTER X
In Chapter X, the dependent variables were eight indicators of water
quality (see Table 1-3). Seven of the dependent variables were derived from NSA
laboratory findings and one was derived from the evaluation of household residents.
Each of these variables was combined with selected independent variables in three
major groupings. The first grouping encompassed household sociodemographic
characteristics such as income and education. The second grouping took in supply
characteristics such as type of supply and reliability of supply. The third grouping
included system characteristics such as number of service connections and water
test history.
The household, supply, and system characteristics (the independent vari-
ables) were selected on the basis of those which would potentially identify national
patterns of widescale association between one or more of the eight water quality
indicators (the dependent variables). In the summary which follows, the NSA
findings are reviewed selectively. The statistical measures which describe the
relationships among the variables are omitted in this summary, since there is not
space here to provide the qualifications which would be necessary for interpreta-
tion of the indicated relationships. Furthermore, the relationships are summarized
without systematic reference to the point at which they were developed in the
three-step analytic procedure. Thus, this summary is limited to generalized
statements relevant to potential national patterns regarding water quality. The
indicators of water quality (the eight dependent variables) are addressed in the
order listed in Table 1-3.
-------
I - 91
Table 1-3
Indicators of Household Water Quality
Household Water
Quality Indicators Description
Total coliform Count of total coliform organisms per 100
milliliters in household water specimens
Nitrate-N Measured level of nitrogen in the household
water
Lead Measured level of lead in the household water
Turbidity Measured level of turbidity in household water
Aesthetic/economic Standardized combination of sulfates, iron,
indicator manganese, color, and total dissolved solids
(Index 2 in Chapter VI)
Health-related indicator Combination of the primary MCL-regulated
(Index 1 in Chapter VI) contaminants: arsenic, barium, cadmium,
chromium, mercury, selenium, silver, fluoride,
gross alpha radiation, lindane, and
methoxychlor, available for the 10 percent
subsample of households
General water quality A gross measure of general water quality with
indicator standard scored contributions from total
(Index 3 in Chapter VI) coliform, nitrate-N, sulfates, hardness, iron,
manganese, sodium, lead, turbidity, color, total
dissolved solids, arsenic, barium, cadmium,
- chromium, mercury, selenium, silver, fluoride,
gross alpha radiation, gross beta radiation,
lindane, and methoxychior
Perceived water quality A composite of household resident evaluations
indicator of the taste, odor, color, sediment, and clarity
(Perceived water quality of the water supply
index in Chapter vi)
-------
I - 92
LABORATORY INDICATORS OF WATER QUALITY: REGRESSION ANALYSES
TOTAL COLIFORM ORGANISMS
The concentration of coliform organisms in rural supplies did not have a
strong linear relationship with any particular household, supply, or system charac-
teristic studied in the NSA analysis. The most useful independent variable
appeared to be the type of household supply, which had a tendency to be associated
with variation in coliform concentration. For purposes of the analysis in Chapters
X through XIII, the household supply type was classified as community (served by a
system with fifteen or more connections), well, or other (served by surface water,
springs, cisterns, hauled, or purchased bottled water). The NSA findings showed
that lower coliform concentrations generally were associated with community
supplies rather than with wells, and lower with wells than with other types of
supplies. In other words, the association indicated a greater potential for
contamination in wells— and particularly in other supplies—than in Community
supplies. Similarly, the correlation between the number of connections (a system
characteristic) and coliform concentrations suggested that the larger the system,
the lower the concentration.
The lack of strong association among the variables was conditioned by the
fact that so many household supplies were free of coliform organisms, or else had
only one organism. By focusing on those households which had two or more
coliiorm organisms per 100 milliliters of water (the NSA water quality reference
value, as explained in Chapter V), it was possible to develop a tentative profile of
households which tended to be most susceptible to potential contamination.
Specifically, the profile suggested increased potential for contamination in house-
holds with: (1) heads having no more than a high school education, (2) total
household income of less than $15,000 yearly, (3) supplies served by systems with
fewer than 500 connectIons, (4) more than one treatment device, (5) less accessible
supplies, and (6) long-term service from the supply (more than eight years). Across
-------
- 93
the nation, the NSA coliform reference value was exceeded at 28.9 percent of all
rural households, but at 47.9 percent of households which had the six characteris-
tics of the profile.
In considering the profile, it is essential to keep in mind that the features
of the profile only help identify households in which bacteriologicai contamination
might be more likely than in other households. The profile features do not,
however, imply causality. For example, although the number of treatment devices
is one profile feature, there is no indication in the NSA data that the devices
themselves are associated with increased pollution potential—although such might
be the case if, for example, one of the devices were a filter contaminated with
bacteria. On the other hand, it may well be that households with pollution-prone
source water are more likely to have several treatment devices, and that the water
would be of even worse quality tithe devices were not in use.
Among households served by wells which had fewer than fifteen connec-
tions, the only associations which proved useful for modeling were the well
attributes. They included well type, grouting, splashproof cover, and well exit.
The resulting final model for households served by wells related to total coliform
contamination with an adjusted R 2 of only 0.17.
TURBIDITY
Levels of turbidity were not strongly associated with household, supply, or
system variables. However, mean turbidity levels suggested that well supplies
tended to be more susceptible to turbidity than did community or other supplies.
At the same time, the standard deviation for turbidity levels in wells reiterated the
finding that well turbidity levels varied greatly—from low to very ‘ugh—across
rural America.
The impression that community supplies tenred to have the lowest turbid-
iL was consistent with certain findings regarding the system variables.
-------
I - 94
Specifically, household supplies served by systems with fewer Connections (such as
single-connection individual systems) had a slight tendency to have greater
turbidity. In addition, supplies served by privately owned systems (which often
were individual wells) had the largest mean turbidity levels—a level which was
nearly four times the mean level for supplies served by systems owned by the
public or by private firms.
Among households served by wells, none of the considered variables
demonstrated a systematic association with turbidity.
NITRATE-N
Concentrations of nitrate-N in rural household supplies showed very little
association with household, supply, or system variables. This finding indicated that
none of the specific variables—either alone or in combination—could be relied on
as screening devices to identify households with nitrate-contaminated water
supplies. The final linear regression analysis showed that regional variation in
nitrate levels, reported in Chapter V, remained the only rough guide in the NSA to
locating rural supplies prone to nitrate contamination.
For households using wells, the weli type showed a weak association the
eta 2 equaling 0.03. No other considered variable showed a useful linear association
with nitrate-N.
LEAD
As with Concentrations of nitrate-N, concentrations of lead were not
meaningfully associated with household, supply, or system variables. Thus, since
lead contamination was fairly pervasive in rural America (see Chapter V), it was
possible for the metal to affect domestic supplies regardless of type of household,
supply, or system as classified in the NSA.
-------
- 95
Among households using wells, three attributes of wells did show some
association with lead levels. But, in a final model they combined for an R 2 of only
0.07.
AESTHETIC/ECONOMIC INDICATOR
The aesthetic/economic indicator (which is based on certain laboratory
findings, and which is described in detail in Chapter VI as Index 2) showed no strong
linear relationship with household, supply, or system variables. An examination of
the mean values, however, did suggest that households which were willing to pay
more for a better supply tended to have poorer aesthetic/economic indicator
scores. This implied that household members sometimes perceived aesthetic or
economic problems with the supply (in agreement with laboratory findings) and
were willing to pay more for an improved supply.
A study of the mean values also indicated that the aesthetic/economic
condition of weU supplies was poorer than that of community or other supplies at
rural households. Also, there was a slight tendency toward a less favorable
condition in household supplies served by privately owned systems.
There was no indication of any systematic national association between the
aesthetic/economic indicator and the considered variables among households served
by wells.
I-IEALTH-RELATED INDICATOR
The health-related indicator (which is based on certain laboratory findings,
and which is described in detail in Chapter VI as Index 1) showed no strong
association with household, supply, or system variables. In fact, the associations
were so weak that all of the independent variables were screened out in the fir .t
step of the analytic procedure, and linear association thus was not examined.
-------
- 96
Among households served by wells, there were associations with the health-
related composite. Well type, the number of connections, willingness to pay, total
treatment, contamination distance, education, and well age were all variables
sufficiently related to the composite to be included in the final regression model.
In concert with the sample design variables, they registered an R 2 of 0.39 with the
health-related composite.
GENERAL WATER QUALITY INDICATOR
Of the eight NSA water quality indicators, the most useful proved to be the
general water quality indicator (which is based on laboratory findings related to a
combination of aesthetic, economic, and health considerations, and which is
described in detail in Chapter VI as Index 3). The final linear regression ñiodel
showed that 27 percent of the variability of the indicator could be accounted for by
certain household characteristics, one supply characteristic, and the sample design
factors (the regional, SMSA/nonSMSA, and size-of-place groupings). The overall,
three-step analysis showed that the most important associations occurred between
the indicator and: (1) household willingness to pay more for a better supply, (2) the
perceived relative cost of the water in the household, (3) household income, (4) the
number of water treatment devices in the household, (5) supply accessibility, and
(6) the regional location of the household.
Specifically, households which were willing to pay more for an improved
supply tended to have the poorest quality water (on the basis of the indicator
scores). As a corollary, a desire to pay less f or the water tended to be expressed in
households which had the best quality water. At the same time, the more costly
the supply was perceived to be (on the basis of whether the supply cost was
reported to be expensive, all right, or nonexistent), the better the quality tended to
be. Household income appeared to have some association with variation in the
indicator scores, but this independent variable was not as important as the other
-------
- 97
household variables. Finally, households with a larger number of treatment devices
tended to have poorer quality water. (As noted before in the summary of findings
for total coliform organisms, such a finding does not imply that the devices
themselves in any way caused the poorer quality.)
Among the supply characteristics, accessibility, as measured by the access-
ibility index described in Chapter VI, showed some association with water quality.
That is, the less accessible water supplies tended to provide poorer water quality.
However, accessibility was too weakly associated to provide a reliable clue to
anticipated water quality. Among the sample design factors, regional location was
the most useful: water quality tended to be superior in Northeast households (as
also reported in Chapter V).
A similar level of association was encountered for households using wells
(R 2 of 0.25), but the contributing variables were substantially different. Most
important was the type of well, followed by the reliability and the well exit
variables.
The most useful variable among all the indicators of water quality for
households served by wells was the well construction.
LABORATORY INDICATORS OF WATER QUALITY: CONTINGENCY TABLES
TOTAL COLIFORM ORGANISMS
Households served by community water supplies were above the NSA total
coliform reference value at a rate of 15.5 percent, compared to those served by
wells at 38.2 percent. Other supply types (hauled, springs, etc.) served households
which were over the reference value in 77.7 percent of the cases.
Publicly owned systems served households which were found to be above
the NSA toni coliform reference value in 15.2 percent of the cases, compared to
19.3 percent of those served by systems owned by private firms and 40.6 percent
for those owned by a private individual. Households served by privately owned
-------
I - 98
systems were over six times more likely to be at or above 100 coliforms per 100
milliliters of water than those served by publicly owned systems.
Households which had higher incomes were less likely to have a total
coliform problem (more than one coliform per 100 milliliters of water) than
households with low income. Households which earned less than $10,000 per year
were above the NSA total coliform reference value in 16.3 percent of the cases, as
compared to 4.5 percent of households earning $30,000 or more per year. Under
$10,000 income households served by wells were over the total coliform reference
value at a rate of 44.3 percent, compared to 25.7 percent among the $30,000-plus
income group. High income households had 7.5 percent, compared to 22.2 percent
for low income households at the level of 100 or more coliforrns per 100 milliliters
of water.
Households where the head had less than a high school education were
above the NSA total coliform reference value in about 15 percent more cases than
among households where the head had more than a high school education. The low
education households were at or above the count of 100 coliforms per 100
milliliters of water more than twice as often as the high education group. The
relative pattern among education groups for households on wells was about the
same as for the nation as a whole.
The more expensive rural residents perceived their water supply, the less
likely they were to have a coliform problem. Over half (57.8 percent) of
households where it was reported the water cost nothing were above the NSA total
coiLform reference value. Only 7.0 percent of househoids where the supply was
reported to be very expensive were over the reference value. Among well users,
47.4 percent of those reporting no cost were over the reference value, as compared
to 4.1 percent of those households which reported water costs as very expensive.
Households which w’ re willing to pay more for an improved supply were
more likely to be above the total coliform reference value than those wishing to
-------
- 99
pay less (32.2 percent, compared to 21 .6 percent). Well users who were willing to
pay more were over the reference value at a rate of 41.7 percent, compared to
30.5 percent among households willing to pay less.
Households served by supplies judged as poorly accessible were above the
total coliform reference value in 59.9 percent of the cases. Those judged to have
good accessibility were over the reference value in 23.2 percent of the cases. In
terms of having counts of more than 100 coltforms per 100 milliliters, poorly
accessible supplies were over four times more likely to be at that level than
households served by supplies with good accessibility. Almost the same rate of
over reference value households was found among well users as among all
households with poorly accessible supplies (58.8 percent). A higher rate of wells
judged to have good accessibility (37.5 percent) were over the reference value than
all rural households judged to have supplies with good accessibility.
Renters were over the total coliform reference value among 47.3 percent
of rural households on wells, while people who owned (or were buying) their home
were over in 37.0 percent of the cases.
Households served by wells judged to be of poor or medium reliability were
above the reference value in almost 49 percent of the cases, compared to 35.9
percent for those in the good reliability category.
Mobile home dwellers served by wells were over the reference value at a
rate of 46.0 percent, compared to 37.1 percent for other well-served households.
Well-supplied households which thd not treat the water were above the
total coliform reference value at a rate of 52.5 percent. That problem rate
declined to 32.6 percent for households using wells with three treatment devices.
The rate climbed to 41.6 percent for those with four or more devices. The
difference in source quality among categories was unknown.
-------
I - 100
While 40.2 percent of well-supplied households reported having tested the
water at least once, their rate of exceeding the reference value, 36.0 percent, was
close to the rate among households which had never tested, 39.7 percent.
Households using dug arid bored wells exceeded the reference value most
frequently among categories of well construction techniques (76.4 and 50.3
percent, respectively). Indeed, dug wells were over the high 100 coliform count
level in almost hall (48.5 percent) of the cases. Households on drilled and driven
wells had the best records, but even they had a rate of exceeding the reference
value on the order of 30 percent.
K flowing whether or not a well was grouted, in the absence of other
information about its construction and condition, was not useful for discriminating
rates of total coliform contamination among ho4seholds on wells. Similarly, the
age of the well did not account for much difference in rates of coliform problems.
Households on wells which either had a buried seal or a pitless adapter
were over the total coliform reference value at about a 30 percent rate, compared
to 45.0 percent for those on wells which had the water exit the well above ground
1evel This latter group was over twice as likely to have high (100-plus) total
coliform Counts.
The rates of total cohform problems were importantly different among
well-using households depending upon the condition of the well cover. Cases where
the cover was judged splashproof and otherwise adequate were over the reference
value at 33.6 and 33.7 percent, respectively. On the other hand, cases where the
cover was not splashproof, and inadequate in other ways, were over at rates of 57.1
and 54.4 percent, respectively. Similarly, households using adequately maintained
wells were over the reference value at the rate of 35.3 percent, as opposed to 49.6
percent among those inadequately maintained.
-------
I - 101
Households on very shallow wells (3 meters or less) were over the reference
value in 61 .0 percent of the cases. Well depths from 30 to 60 meters showed a rate
of 27.4 percent. Curiously, deeper wells had somewhat higher rates.
Households on wells which used public sewers were over the total coliform
reference value at the rate of 23.9 percent. When wastewater was unconfined,
that rate jumped to 53.2 percent.
Households served by supplies of all types which had all of three selected
system characteristics (less than fifteen connections, privately owned, poorly
accessible) were over the NSA total coliform reference value in 63.5 percent of the
cases. Households served by systems with none of those attributes were over in
1 5.4 percent of the cases.
Total coliform problems among well-using households varied dramatically,
based upon six selected well characteristics: the well was dug or bored, the water
exited the casing above ground, the cover was not splashproof, the cover was
generally inadequate, the maintenance was inadequate, the well was ten meters or
less in depth. Households using wells with none of the selected attributes were
over the total coliform reference value in 28.1 percent of the cases. The more of
the attributes present, the greater the likelihood of a problem until the rate
reached 77.4 percent when all six were present. When the list of attributes was
enlarged to include four socioeconomic characteristics (less than $10,000 income,
less than a high school education, mobile home, renting) and four supply and
wastewater items (poor accessibility, medium or poor reliability, no treatment
devices, uncontrolled wastewater), the results ranged from 21.4 per ent among
households with none of those characteristics to 94.3 precent when ten or more
were present.
-------
I - 102
PERCEIVED WATER QUALITY INDICATOR
Before associations were tested between the perceived water quality
indicator (described in Chapter VI as the perceived water quality index) and the
independent variables, the indicator was checked for consistency with other
household responses. The indicator was designed to reflect household perceptions
about the combined intensity and duration of undesirable properties of odor, taste,
lack of clarity, color, and sediment. Various correlations and comparisons
indicated that respondents were consistent in the in.fo mation which they provided
to the NSA. In other words, the household respondents were able to focus their
attention on the condition of the household supply and to give consistent responses
about the nature of that supply. Furthermore, as another indication of the
usefulness of the perceived water quality indicator, scores on that index tended to
be consistent with scores on the aesthetic/economic index, which was based on
certain relevant laboratory findings, as described previously.
As to relationships between the perceived water quality indicator and
independent variables, only three variables among the household characteristics,
and none among the supply or system characteristics, proved useful. Specifically,
the indicator showed some association with willingness to pay more for the supply,
age of the head of household, and length of residence. That is, there was a slight
tendency for younger heads and for heads who had been in residence for shorter
periods of time to have a poorer opinion of the quality of the water supply. The
strongest association, however, was with the willingness to pay more for the
supply. Residents who reported poorer quality water also tended to be more likely
to be willing to pay more money for a better supply. This finding was consistent
with the finding (reported above) that households which were willing to pay more
for a better supply tended to have the poorest quality water as determined by the
general water quality indicator based on a number of laboratory assessments.
-------
- 103
SUMMARY OF CHAPTER XI
Two measures of quantity are assessed in Chapter XI. The first measure is
the monthly quantity of water used in rural households, as calculated from
household water bills. The second measure is of the NSA respondents’ perception
of the quantity of water usually provided by the household supply. The develop-
ment of the two measures—called recorded quantity and perceived quantity,
respectively—is described in Chapter V. As in Chapter X, a three-step procedure
is used to identify potentially useful associations between the two dependent
measures and selected independent variables. The independent variables are
organized into three groups according to whether they are relevant to household,
supply, or system characteristics.
RECORDED QUANTITY
Special conditions pertained to the development of the dependent variable
representing recorded quantity. In brief (see Chapter V), both quantity of water
consumed and the amount charged for water were available for only 567 of the
2,65L households surveyed in the NSA. Those 567 households were served by
community systems, and they had sufficient billing information on which to base
estimates about water consumption and cost. An additional consideration was that
the quantity information itself had certain limitations. For example, data were
recorded from only one water bill per household, so possible seasonal variation was
not taken into account.
Overall, then, results of the NSA analysis were relevant only to a subset of
sampled households, a subset which represe ited 5.4 million households across rural
America (nearly 60 percent of all house! olds which received water bills). As to
the analytic findings relevant t those 5.4 million households, a total of five
-------
I - 104
independent variables—four in the household_characteristics group and one in the
system-characteristics group—had the minimum association with recorded quantity
necessary for further analysis. As the analysis was completed, the most useful
variables proved to be the four related to household characteristics—the unit cost
of water, the number of household residents, the ratio of water charges to total
household income (affordability), and the number of water-using fixtures and
devices in the household. The one variable in the system-characteristics group—
system ownership—was not as useful, and the emphasis thus was placed on the four
household characteristics.
The strongest association was between the unit cost and the quantity
consumed, with unit cost accounting for fully 22 percent of the screening-model
variation in consumption. The unit cost was the cost in dollars per 1,000 gallons of
water, but to ascertain the nature of its relationship to consumption, it was
necessary to transform dollar amounts into the natural log of dollars. This
transformation made it possible to perform the necessary linear regression
analysis, but it meant that interpretation had to be stated in corresponding terms.
Specifically, a unit increase in price (natural log of dollars) resulted in a decrease
in demand of 13,911 liters (3,672 gallons) per month.
Although the effect of unit cost was pronounced, the exact influence of
cost could not be ascertained because the effect of various pricing strategies
(ranging from decreasing block rates to uniform rates) could not be incorporated
into the variable representing the unit cost of water. Instead, it was necessary to
assume a uniform set of charges by which water was priced at a constant rate for
every Thousand gallons consumed.
After unit cost, the number of household residents showed the strongest
relationship to recorded quantity. As would be expected, household consumption
increased a certain amount with each additional occupant. Consumption also was
affected, to a lesser extent, by the ratio of water charges to total household
-------
I - 105
income. Since this ratio was a measure of affordability, as explained in Chapter V,
one might anticipate that as the ratio increased, consumption would decline.
However, the opposite was true. That is, consumption actually increased as water
charges took up a greater proportion of household income. This positive relation-
ship suggested that families which allocated larger proportions of their budgets to
water use were families which used more water. One possible explanation for this
situation was that the cost-to-income ratio actually represented the consumer’s
willingness to pay for water, rather than a constraint imposed by a fixed budget.
As to the fourth independent variable (the number of water-using fixtures and
devices), water use increased moderately with each additional device, although
with twelve or more devices, water use declined.
PERCEIVED QUANTITY
The dependent variable representing perceived quantity was developed
from information about peoples’ beliefs about the volume of water readily provided
by the household supply. In Chapter V, this information was coordinated in a five-
category scale. However, as reported there, about eight of every ten rural
households reported having ample supplies, and only two of every ten reported
some deficiency in the supply. Thus, in Chapter XI, because there was so little
variability in the distribution of perceived quantity values, the scale scores were
collapsed into two groups: in one group were all households where the supply
quantity was reported to be completely satisfactory; in the other group were all
households in which the supply quantity was in some measure deficient.
As to the analytic results, only one independent variable had an association
with perceived quantity which was above the scr.eening level. The variable, which
was in the group of supply characteristics, represented the type of well supply.
Specifically, dug wells more often were associated with insufficient quantity than
were drilled, driven, jetted, or bored wells.
-------
I - 106
SUMMARY OF CHAPTER XII
Two indices which quantify the concept of availability are assessed in
Chapter XII. The indices—representing accessibility and reliability—are described
in detail in Chapter VI. In Chapter XII, however, the accessibility index is altered
slightly, and certain modifications are made to counteract the effect of the
unusual distributional properties of the reliability index.
The analytic procedure is the same as that used in Chapters X and XI.
That is, relationships between the two indices and selected independent variables
are assessed in a three-step process which is designed to identify factors which
could account for variation in the indices. The independent variables are organized
into three groups according to whether they are relevant to household, supply, or
system characteristics.
ACCESSIBILITY
In the first step of the accessibility analysis, a total of ten independent
variables—five in the household-characteristics group, four in the supply-charac-
teristics group, and one in the system-characteristics group—had the minimum
association with accessibility required for inclusion in the second step of the
analysis. In the second step, the list of independent variables was reduced to nine,
since one did not contribute appreciably to the explanation of variability in the
accessibility index.
Although the nine independent variables, plus design-effect variables (for
classification by region, SMSA/nonSMSA, and size of place), were included in the
final linear regression model (step three of the analysis), the model did not prove to
be more effective than the intermediate model f or association of supply character-
istics. That is, the interrnediaw supply model had about the same aegree of
-------
- 107
explanatory ability as the final model, an it had the advantage of focusing on
three key independent variables which were associated with accessibility. Those
variables were: the type of conveyance by which water was transported from the
source to the household (pressurized public supply, electric pump, or other method
such as gravity flow or hand-carrying), the total number of treatment devices in
the supply and system which served it, and the type of supply (community, well, or
other). These three variables, combined, explained 52 percent of the variability in
the accessibility index—compared to 54 percent of the variability which was
explained by combining all of the variables in the final model.
The association between each of the three supply characteristics and
accessibility was consistent with the nature of the variables involved. That is,
accessibility in the NSA was determined by the distance over which water was
transported and by the supply pressure. Thus, it was not surprising to discover that
the most useful explanatory variable was conveyance—with water most accessible
at households served by pressurized public supplies, less accessible at households
served by individual or intermediate systems with electric pumps, and least
accessible at households served by other types of supply. The next most useful
explanatory variable was the type of supply—with water most accessible at
households served by community supplies, less accessible at households served by
wells, and least accessible at households served by other types of supply. Finally,
the variable with the weakest association was the total number of treatment
devices in the supply and system serving the supply—with water more accessible
where there were more devices.
-------
I - 108
RELIABILITY
In contrast to the accessibility analysis, the reliability analysis uncovered
no characteristics which had sufficient association to meet the screening level
criterion. Thus, analysis of the correlates of reliability (based on the frequency of
breakdowns and supply interruptions) did not proceed past the initial procedural
step.
-------
I - 109
SUMMARY OF CHAPTER XIII
Two measures of cost are assessed in Chapter XIII. The first measure is
the monthly Cost of water to rural households across the nation. The second
measure is the cost of every thousand gallons of water consumed at rural
households. The detailed discussion of the two measures of cost appears in
Chapter V. As in previous chapters, a variety of relationships are explored
between the two dependent measures and independent variables organized accord-
ing to whether they represent household, supply, or system characteristics.
Similar to the restrictions imposed on the analysis of the quantity of water
consumed at rural households (see Chapter XI), the cost analysis was also restricted
to households that were served by community systems and that received a bill for
services. Even when a water bill was available, the necessary information was not
always included on the btll. Hence, not all households served by community
systems were represented in the analysis presented herein. The subset data
actually represented 5.4 million rural households—about 60 percent of all house-
holds which reported receiving a regular water billing.
MONTHLY COST
The analysis pertinent to the 5.4 million rural households revealed the
importance of the household set of attributes in estimating monthly water costs.
More specifically, although the amount of water consumed during a month was the
single most important factor in predicting monthly water expenditures, other
household attributes of importance included the number of household occupants,
the number of years served by the water system, the education of the household
head, and the respondenVs perceptions concerning the cost of the water supply.
With the exception of the length of time served by the water system, all of the
-------
1-110
important household attributes showed a positive association with monthly water
costs. That is, higher monthly water costs were associated with more of the
attribute. A different pattern was evidenced between water Costs and the number
of years served by the present water supply system: monthly expenditures were
lower in households served for longer periods of time. Perhaps this relationship
was driven by the age of rural dwellers, since there is some evidence to support
diminished water consumption among the elderly.
The analysis of monthly water costs identified one other variable which had
an association with the dependent measure. The variable, which was in the group
water system attributes, reflected the type of system ownership. Specifically,
private firms were more often associated with higher monthly water costs, while
public bodies were more often associated with lower costs.
UNIT COST
A similar set of factors emerged from the analysis of unit water costs.
The volume of water consumed, as well as the respondent’s perceptions about the
cost of the supply, were identified as important correlates of unit water costs.
While higher perceived costs were linked with higher unit costs, water consumption
and the unit cost of water were inversely related, indicating that higher unit costs
resulted in reduced consumption.
In addition to the household set of attributes, two variables in the water
system group were important in determining per thousand gallon costs. The first
corroborated the “economies of scale” theory which states that the cost of water is
lower among households supplied by water systems larger in size. Here, the size of
the system was evaluated in terms of the number of service connections. Secondly,
the source from which the water was withdrawn had a bearing on the unit cost of
water. Households which purchased water from systems using a groundwater
-------
I—ill
source paid considerably less per unit of water consumed than those purchasing
from systems which withdrew from a surface source.
------- |