A premise of this volume is that the water economy of the western United
States is passing from the "expansionary" phase to the "mature" phase.
In the expansionary era, the incremental cost of water remained relatively constant over
time (in real terms), as water development project sites were available to meet growing
demands. The mature phase, brought on by change in the economy at large, is characterized
by rapidly rising incremental costs and greatly increased interdependencies among water
uses and users.[1]
In a maturing water economy, the high cost of new water brings about a
search for supplies from existing uses whose economic productivity is less than the cost
of acquiring new supplies. The largest use of water in the western U.S. is for crop
irrigation. Thus, competition for irrigation water is arising from industrial, energy, and
household water diversions, and from instream uses such as power generation, recreation,
fish and wildlife habitat, and navigation.
This chapter assesses the economic impacts of increased competition for
irrigation water on farmers and other direct water users, and on the local and regional
economies intertwined with the water using sectors. A second task is to consider the
policy changes needed to deal with anticipated impacts.
Economic Impacts:
Alternative Perspectives
Two competing hypotheses or viewpoints can be identified regarding the
economic impacts of reduced water supplies for irrigation.
The "Significant Impact" Hypothesis
One viewpoint, which appears to reflect the conventional wisdom in
political and media discussions of the "western water problem", contends that
irrigation has been a significant source of regional growth in the West. Supporters of
this perspective point to the sharply increased crop sales brought about by irrigation,
and claim substantial multiplier effects in jobs and spending in the related communities
and regional economies. From the "significant impact" hypothesis follows its
converse: removing water from the agricultural sector will have major, even intolerable,
negative effects on the regional economy.
The significant impact hypothesis draws its evidence from several
sources. One is our historic sense of the role of irrigation in developing the West.
Another is the obvious fact that irrigated lands yield enormously more product than
semidesert, or wheat-fallow rotations. Also supporting this view is the relatively high
income and employment associated with fresh vegetable and fruit crops in areas with milder
climates.
The "Limited Impact" Hypothesis
The alternative hypothesis conceptualizes the problem in what
economists would call "marginalist" terms. This viewpoint, invoking the law of
diminishing marginal returns, posits that the productivity of irrigation water ranges from
highly productive down to marginally productive uses. Similarly, the indirect regional
employment and income impacts range from significant in some sectors to minimal in others.
The marginalist position contends that water removed from irrigation will, in the long
run, be the least valuable portion. Even if the first-round impact is at the expense of
high-valued irrigation uses, subsequent economic adjustments will find the low
productivity uses giving way.
The limited impact hypothesis finds evidence for its viewpoint in the
large proportion of water diversions in the West which are devoted to irrigation (80-90
percent), and the high proportion of irrigation water use which is in relatively
low-valued uses
- 247 -
(forages, and food and feed grains). Proponents focus on the low
absolute amounts of indirect income and employment in the processing and input supply
sectors associated with most agricultural water use. This approach asks how much water
from agriculture will be needed to fuel expected nonagricultural growth and tries to
measure the incremental costs of such reallocations, compared with the costs of new water
supplies.[2]
The evidence, I believe, strongly supports the "limited
impact" hypothesis. This chapter presents the case for that viewpoint.
Procedures and Scenario for Impact Assessment
The general approach taken here is to compare the economic impacts of
potential reallocation of water from irrigation to other direct uses of water. Indirect
impacts of removal of irrigation water are similarly compared to indirect effects in
growing sectors which might acquire water. The net value foregone (or gained) is taken to
be the most suitable measure of direct economic impact. Payments to primary factors of
production (value added) and employment are the chosen measures of indirect impacts.
The assumptions of the analysis are as follows. The planning horizon is
taken to be the remainder of the century. The economic impacts presented below will be
expressed in 1982 price levels. As the economic background, there will be no general wars
or political upheavals which disrupt world production and trade in agricultural
commodities, and agricultural commodity prices will continue the trend which has been
observed through most of this century, with technology holding food prices down relative
to the cost of production inputs and to real consumer incomes. Hence, current price and
production relationships will be used in predicting future agricultural net income.
It seems unlikely that water diversion requirements in nonagricultural
sectors will grow so rapidly as to require any enormous reduction in water supplies to
agriculture. (For example, a 2.5 percent compounded rate of growth in nonagricultural
water demands in California would absorb only another 4 million acre-feet after twenty
years, or about 10 percent of current agricultural water use in that state.) The analysis
below posits at most a 20 percent reduction in irrigation water supply, which I regard as
an extreme outside limit of impact in the next two decades.
Direct Impact Measurement
Economic value has been a principal indicator of general value in our
society. In a market economy, economic values are measured by market prices. When the
market is working well, the price system measures the value that goods or services provide
to people and the value of resources used in production. Direct economic impacts of water
reallocation should be measured by the price of the commodity (or by a surrogate price
when markets are not operative), since the price measures the net value foregone or gained
in the respective use categories.
For various reasons, market prices are not generally available for
water. The physical barriers to markets for water stem from its flowing, mobile
characteristics, which make it difficult to establish and enforce the property rights
which are the essential foundation for any market system. Also, in certain uses, water is
a public rather than a private good, in that one individual's consumption does not
preclude use by others. Finally, water plays a special role in human society. Boulding[3] has pointed out that "the sacredness of water as a symbol of
ritual purity exempts it in some degree from the dirty rationality of the market."
However, it is possible to estimate the net benefits conferred by water
even in the absence of markets. This process is sometimes termed "shadow
pricing." Shadow pricing can be understood as an attempt to establish an exchange
ratio in monetary terms which would be equivalent to that which would emerge from a
properly functioning market process. The basic concept is willingness to pay as the
indicator of economic value. Willingness to pay (WTP) reflects the amount which a
rational, fully informed consumer would be willing to forego rather than do without the
commodity in question. In accordance with the principles of diminishing marginal utility
or diminishing marginal productivity, willingness to pay falls as quantities increase.
Special Problems in Valuing the Water Resource
The hydrologic system must be considered in terms of its interactions
with climate, land, ecosystems, and the human social and economic systems. This intricacy
is further complicated by the highly variable nature of moisture supplies, the importance
of sequential uses as water flows from upper watersheds to its eventual destinations in
sea or sump, and the importance of transportation costs in establishing water value.
Concepts of the
- 249 -
economic value of water can be relevant only when explicit recognition
is given to quantity, location, quality, and time of supply. Put another way, the value of
water is highly site-specific, and varies directly with local conditions of supply and
demand for the resource.
Broadly speaking, there are three ways of determining net willingness to
pay for water.[4] One is to determine WTP by statistical analysis
of actual or hypothetical water use decisions by consumers. A second, called the
"change in net income" approach (applicable to agricultural or business users),
imputes the value of water as the increment to profits arising from an increment in water
supply. Finally, the "alternative cost" approach values water in terms of the
resource savings which would be achieved by water-intensive as opposed to alternative
production technologies.
Marginal versus Total Value
The correct concept is the incremental worth (value of the last unit)
rather than total value or revenue associated with all units of water and other resources.
The total value of product can be attributed to water only if all other factors of
production (i.e., labor, land, capital) have no known alternative beneficial use (an
extremely unlikely event).
Long-Run versus Short-Run Value
Short-run values apply to farmer decisions in the growing season (with
fixed resources not charged for), while long-run values reflect the net returns after all
resource input costs are deducted. Short-run values tend to be higher; as willingness to
pay rises, the fewer the fixed resources which need to be accounted for. Long-run values
are the most appropriate in the present context.
Comparability in Place, Form, and Time
As noted above, water is a bulky commodity, for which transportation
costs are often large relative to value at the place of use. Hence, value of water
declines rapidly with distances from site of use, and may even be negative at a potential
source. Quality (perhaps reflected in treatment costs) and timing are important specifics
as well.
Measuring Quantity:
Diversion versus Consumption
For off-stream uses, the usual choice of a measure of quantity is
between the amount diverted and the consumptive use (that
- 250 -
portion of diversion not returned to the stream or aquifer, and not
available for reuse). Although there are no firmly agreed upon conventions, most
economists prefer to deal with withdrawals. The discussion below follows that approach.
Annual Value versus Capitalized Value
Does the value estimate represent an annual rental equivalent (the
price of one acre-foot in a typical year), or the right to a certain flow each year into
the indefinite future? Clearly, these values are related, but not identical. The value of
the latter (property right) is conceptually equal to the discounted present value of the
stream of annual values. Hence, time horizon, interest rate, and expectations regarding
inflation, in addition to annual value, must be specified in order to reconcile the two.
The capitalized value will normally be in the range of twelve to twenty times as large as
the annual value.
In the subsequent discussion, "net benefits" will refer to the
willingness to pay for one acre-foot of raw water diverted in a particular year, unless
otherwise specified. Diversionary uses are treated first, followed by the nonwithdrawal or
instream categories.
Net Economic Benefits Foregone from Irrigation
The direct value of water from foregone irrigation is usually measured
in terms of the decrement of profit to the producer without irrigation (or with a
decrement of water supply) as compared to profits with irrigation. At the margin,
the value of a decrement is the same as that of an increment, so either formulation is
applicable.
One additional distinction is useful. The marginal net benefit measured
can be for the marginal crop or for the average net benefit for the marginal farm. I favor
the former measure as most generally applicable. But there may be some instances, where
entire farms (or even areas) are removed from irrigation, where the average benefit is
appropriate.
What are the net benefits of irrigation water? The previous discussion
implies that the value at the margin will reflect water scarcity and marginal cost of
supply. Local production conditions (i.e., rainfall, temperature, and growing season
length) and market situations will have an impact, so we would expect considerable
variation across the West.
A recent set of estimates has been developed by Beattie and Frank,[5] which involved statistical analysis of 1974 census data on
agricultural output, as influenced by resource inputs, including
- 251 -
land, labor, machinery, and chemicals, as well as irrigation water.
The results yielded values (converted to current 1982 dollars) of $10-$15 per acre-foot in
the intermountain valleys (Upper Colorado, Snake River Basin), $20-$25 in the desert
Southwest and central California, and $40-$45 per acre-foot in the Ogallala groundwater
region of the High Plains.
Howitt et al. recently reported rather similar results, using a much
different technique. Their interregional supply-demand model for California yielded shadow
prices at the margin of $23-$35 per acre-foot in the Central Valley and southern
California and $7 in the Imperial Valley.[6] Gollehon et al. show
shadow prices for irrigation water for eleven Rocky Mountain subregions. With a 20 percent
reduction in irrigation water supply, two regions were identified with water valued at the
margin in excess of $20 per acre-foot, while four were between $10 and $20 per acre-foot
and six were below $10.[7] Estimates based on conversions of water
rights sales to annual equivalent values in New Mexico[8] yield
similar results.
Irrigation water is seen to be most valuable when it helps to confer a
special comparative advantage on crop production in a particular locale. This advantage is
most likely to occur in the desert valleys of the Southwest which are adapted to
perishable fruit and vegetable production. Forage, feed grain, and food grains are
storable and readily produced with natural moisture elsewhere in the nation, so that net
value productivity of water use in these categories is relatively limited. Net benefit
estimates obtained for certain specialty crops may therefore be somewhat higher than the
figures cited above. However, such uses will probably account for no more than 10 to 20
percent of total irrigation water use in the West in the period under discussion, and are
not relevant for the present analysis. Eighty percent of irrigation demand probably lies
below $40 per acre-foot, and the last 10-20 percent subject to transfer would be valued
below that level.
Net Economic Benefits in Nonagricultural Uses
The Value of Water in Industry
While water is used throughout the industrial sector, the major
consumer, particularly in the arid West, is in the energy sector, particularly for cooling
steam-electric power plants.
- 252 -
Several processes can be used for cooling, depending on water scarcity
and price. Young and Gray[9] show with an alternative cost approach
that it is economical to convert from a passthrough system to evaporative cooling towers
when water costs rise above about $5 per acre-foot (1982 price levels). Methods which
conserve more water are much more expensive. Gold et al.,[10] in a
study for EPA, report that breakeven points for a combination wet-dry system run around
$600 per acre-foot, while the shift to a completely dry cooling system would be economical
only if water was priced above about $1400 per acre-foot. Abbey's analysis of the
water/energy problems in the Colorado River Basin provides similar estimates.[11] Leigh has estimated the net benefits to water for coal slurry
pipelines, using the cost saving from the alternative of rail transportation as the
measure.[12] The value of water in a Colorado to Texas system is
estimated to exceed $1600 per acre-foot. Hence, the large-scale energy projects proposed
for several areas of the West could, if necessary, be willing to pay an amount many times
the price of water in neighboring agricultural uses. Other manufacturing processes, i.e.,
electronics, would likely be willing to pay even higher amounts for the relatively small
amounts required.
Value of Water in Households
While willingness to pay for water delivered to households is readily
observed and much studied, deriving a marginal value of water to households which is
comparable and commensurate with estimates of raw water values in streams is relatively
difficult. Household water, which is treated (filtered, chlorinated), stored, and
delivered to the user on demand, is a much different economic commodity than the raw river
water used in irrigation or in many industries. Hence, a deduction for treatment, storage,
and delivery costs must be made to achieve comparability. A method suggested by Young and
Gray using data developed by Howe and Lineaweaver[13] estimates
that lawn sprinkling is valued at about $150 per acre-foot and in-house uses as $250 per
acre-foot (in 1982 dollars). A weighted average would be about $220 per acre-foot. Howitt
et al.[14] do not distinguish between industrial and household
demand. Their municipal and household sector estimates for 1980 (in 1982 prices) are about
$160-$200 per acre-foot.
Gardner and Miller report the price of water rights in the Colorado-Big
Thompson project (in northeastern Colorado) transferable to urban uses to have averaged
$2450 per acre-foot
- 253 -
in 1981.[15] Converting this figure to an
annual acre-foot value requires assumptions regarding the appropriate capitalization rate
and expectations about future inflation. However, at an interest rate of 8 to 9 percent
(which seems plausible), and a long planning horizon, the figure is practically equivalent
to the figures given above.
Hydroelectric Power Generation
Evaluation of hydroelectric projects has usually proceeded on the
assumption that water is a free good, so that recorded efforts to value water in this use
are rare. The procedure which has been developed is to value electricity in terms of cost
of production by the alternative process of a steam-powered plant (alternative cost
method). The value of water is then derived by deducting capital and operating costs of
the generation and transmission system. The residual, if any, is attributed to the water
resource (change in net income method). Specific value estimates vary according to the
differences in head (the distance the water falls before turning turbines) but also with
distance to load centers, energy costs of the stream alternative, and the cost of dam and
storage reservoir construction relative to power output. Values also may be expressed for
one site only or for several sites on a given river reach. Young and Gray report single
site values ranging from $3.30 to $10 per acre-foot in 1982 prices in the western states,
the higher values associated with sites with relatively large head on the Colorado River.[16] Whittlesey and Gibbs report values in the Columbia Basin of over
$30 per acre-foot (1982 prices) for water going through all dams below Franklin Reservoir,
including Grand Coulee.[17]
Valuing Water in Waste Load Dilution
Most analysts have adopted the concept that the value of a unit of
dilution water is equivalent to the cost of treating effluent to achieve an improvement in
water quality equivalent to the specified quantity and quality of dilution water. The
results of these studies generally imply that dilution benefits, for the most part, are
not large. Merritt and Mar[18] showed dilution water in the
Willamette Basin (Oregon) to be about $1.30 per acre-foot (1982 price levels). Gray and
Young[19] applied this technique for several regions, deriving
estimates for diluting urban effluents ranging from $0.08 per acre-foot (Colorado Basin)
to $3.25 in the lower Missouri. However, they derived a value of high quality water in the
Colorado Basin for dilution of salinity at about $15 per acre-foot.
The Value of Water in Water-Based Recreation
Water-based recreational services, by tradition and policy, are not
often priced by market processes. Hence, the normal problems of valuing water are
compounded, since in this case, the value of water for recreation must be derived from a
prior synthetic imputation of the value of the recreational services themselves. Moreover,
recreational uses of water are often complementary to other uses, rather than competing
with them. Thus, water stored for irrigation or flood control can often by enjoyed without
diminishing its usefulness in alternative uses. However, growing recreation demand is
creating situations in which these uses are competitive with other classes of instream and
offstream use, but economic analysts have only recently begun work on measuring values
which are suitable for comparing allocations among alternative uses.
Daubert, Young, and Gray[20] formulated a direct
interview procedure which elicits bids from recreationists on the value of water in
flowing streams. Applied to a sample of visitors to the Poudre Canyon in northeastern
Colorado, this approach yielded estimates of economic value related to flow in fishing,
whitewater kayaking, and noncontact streamside recreation (i.e., picnicking). The
resulting marginal values were (at a typical summer flow rate of 200 cfs) converted to
dollars per acre-foot: $9 per acre-foot for fishing, $5 for whitewater sports, and $7 for
the noncontact group. Walsh et al. performed similar analyses on western Colorado streams,
reporting $13 per acre-foot for fishing, $4 for kayaking, and $2 for rafting, when flows
were maintained at 35 percent of maximum.[21]
These findings lend support to the notion that nonconsumptive uses, even
though they are nonmarketed, have economic value to users. While many are skeptical of the
validity of benefit estimates based on responses to questions regarding hypothetical
consumption situations, the values cited seem plausible, and a preferable alternative
technique to generate quantitative estimates of instream flow values has not been
developed.
Summary of Direct Impact Analysis
Up to this point, it has been shown that:
1) The direct net economic value foregone from partially reduced
irrigation water supplies will mostly fall in the range of $5-$30 per acre-foot, depending
on location and type of use.
- 255 -
2) The gain in net value of product or willingness to pay in
industries and households absorbing water previously in agricultural use is five to ten
times or more as high as the losses in the agricultural sector. Important direct economic
values also are found in instream uses, particularly in power generation and recreation.
(These latter uses are less often in direct conflict with irrigation, since they are
largely nonconsumptive.)
These findings are summarized graphically in Figure 10.1. The horizontal
axis represents the fixed supply of developed water in a representative river basin.
Agricultural water values are shown in the left vertical scale, and urban-industrial
values on the right. The step-function Da represents the demand
(value-quantity) relationship for agriculture. Maximum WTP would be in the $50-$100/AF
range, but most of the demand is below that range. The step-function Du , drawn
in reverse from the right-hand axis, represents demand in nonagricultural uses. Maximum
WTP is much higher than in agriculture. Du intercepts Da at a point
such that 10 percent (more or less) of the water is most profitably used in
nonagricultural pursuits, while the balance is profitably employed in irrigation. The
function D'u represents a hypothetical future nonagricultural demand,
reflecting growth in those sectors. The gross gains to the regional economy from the shift
from Du to D'u is shown by the area between the curves, MNOP, while
the losses foregone in agriculture were MNQR. The net gain to the economy from the shift
is then RQOP, and is likely to be quite large.
Indirect Impacts from Reduced Irrigation
Indirect income effects, often called secondary impacts, are the
impacts on related economic sectors which are associated with changes in the level of
irrigation. They are conventionally divided into forward-linked activities ("stemming
from" effects), those which involve processing, marketing, and transportation of the
farm products, and backward-linked activities ("induced" effects) which include
supply of inputs (seed, fertilizer, machinery, etc.) to the farm sector.
- 256 -
[Full Size]
Figure 10.1
Marginal Willingness to Pay for Water as Related
to Percent of Average Annual Supply: Agriculture
and Municipal and Industrial Combined
(for a Representative Western River Basin)
Indirect impact measures must not be confused with direct impact
measures. Indirect income measures usually refer to either gross revenue charges or
payments to all primary resources, rather than the net revenue shifts measured in direct
impact analysis above. Therefore, direct economic impact measures and indirect economic
impact measures, even though both are expressed in dollars, are not strictly commensurate.
The Limited Indirect Impact Hypothesis
For purposes of this chapter, the main question regarding indirect
impacts is the relative magnitudes in losing sectors versus gaining sectors. Concern over
the magnitude of potential indirect effects in irrigation-based subregions is the basis
for public action to avoid loss of irrigation in areas where supply depletion is imminent.
My conclusion regarding the importance of irrigation to regional
economies has, in some circles, proven to be
- 257 -
controversial. Stated simply, my belief is that irrigation
developments have had a relatively minor impact on regional economies in the
post-industrial era. The converse proposition is that loss of 10 to 20 percent of the
irrigation supply in the West would not have an appreciable effect on regional income or
employment.
Evidence for this proposition can be put forward in three classes:
casual observation, statistical analysis of growth impacts on water development, and
detailed studies of the structure of regional economies.
Casual Observation
Consider any of a number of irrigated areas with no other industry or
governmental installation to bolster the economy, beyond the suppliers and processors
linked to agriculture. Many tens of thousands of acres of irrigation, particularly when
the products are forages, grains, or cotton, are required to support even a small town.
The numerous small communities dotting the Ogallala-High Plains region provide one
example; Yuma and Pinal Counties in Arizona, another.
Econometric Growth Analyses
More systematic statistical analysis of growth impacts also have not
been able to identify significant regional growth impacts from irrigation. Only a few
detailed ex post analyses of regional growth impacts associated with irrigation
projects have been published. Cicchetti et al., under contract to the Bureau of
Reclamation, employed regression analysis to study the effect on various indices of
regional economic growth of a number of variables representing Bureau of Reclamation
investments.[22] Census data were obtained for numerous economic
subregions in five arid western states, for 1950, 1960, and 1970. Variables representing
USBR investments in irrigation facilities were not found to have any significant impact on
subregion income, and only a small and not convincingly significant impact (t-value =
1.62) on the value of farm output.
In a similar study, Fullerton et al. used econometric techniques to
estimate the quantitative impacts of federal water resource development on economic growth
in 246 counties in 7 western states.[23] The authors summed up (p.
22):
The null hypothesis that regional economic growth is caused by
investment in water resources of various types is given virtually no support from these
empirical results.
Studies of Regional Economic Structure
Other detailed regional studies, such as Kelso et al.,[24]
yield similar inferences. The recent investigation of the Colorado Ogallala-High Plains
region found that the 600,000 acres irrigated in the area directly employed about 1200
workers (one man-year = 2000 hours), while withdrawing 1.1 million acre-feet of water
annually. Indirect employment in the region associated with irrigation from the Colorado
Ogallala accounted for another 1800 workers.[25] Our conclusion was
that the 40 percent reduction in irrigated crop production, employment, and income
anticipated over the next four decades would have an imperceptible effect on the state
economy, since the impact would amount to less than one-tenth of one percent of the
state's work force. Gollehon et al. studied the effects of reduced irrigation due to
energy development on regional employment and income in eleven Rocky Mountain region
subareas, in Montana, Wyoming, Colorado, and New Mexico.[26] The
area studied encompassed nearly one million irrigated acres, producing mainly forages, and
is supplied by 3.1 million acre-feet of water. A 20 percent reduction in water supply to
this group of subregions would cost the area about 450 jobs directly in farming and about
900 jobs in the region as a whole.
Indirect effects are often measured by reference to
"multipliers" derived from a regional input-output (interindustry) model which
indicates the monetary value of income generated elsewhere in the economy in relation to a
dollar's worth of increased income in the sector of interest (i.e., irrigated crop
production). Applying the multiplier to estimates of increased (or reduced) crop sales
yields an estimate of increased (reduced) economic activity in the region represented by
the model.
The income multipliers for the irrigated agriculture sector are among
the highest of all economic sectors, since each added dollar's worth of crop output
generates economic activity in processing sectors, such as feedlots, dairies, and packing
plants. To project what would be the net regional effect of reallocation of the water
resource, however, the analysis must be carried further. I have computed the predicted
income effects in California of an additional unit of water in several selected irrigated
agriculture sectors and in four of the rapidly growing sectors in the state's economy
(Table 10.1). California was selected because of the size and importance of its irrigated
sector, and because the changes being examined are likely to be registered early in that
state. The income projections would probably not differ much in other western states.
- 259 -
| |
Table 10.1
Direct, Indirect, and Induced Income and Employment Impacts from Water Reallocation,
Selected Agricultural and Industrial Sectors, California, 1977 |
| Part A: Income Impacts |
|
|
|
|
|
| Sector Name (Number) |
Water Use
(1,000 AF) |
Total Incomea
($ million) |
Total Income
per 1,000 AF
($ million
per 1,000 AF) |
Direct Plus
Indirect Income
per 1,000 AF
($ million
per 1,000 AF) |
Direct, Indirect,
Plus Induced Income
per 1,000 AF |
| Selected Agricultural Sectors |
|
|
|
|
|
| Hay and pasture (14) |
11,440 |
319 |
.03 |
.06 |
.09 |
| Cotton (9) |
4,714 |
253 |
.05 |
.13 |
.21 |
| Noncitrus fruits (22) |
3,478 |
710 |
.20 |
.33 |
.52 |
| Vegetables (25) |
1,988 |
767 |
.39 |
.63 |
.93 |
| All agricultural sectors |
34,460 |
5,000 |
|
|
|
Selected Industrial Sectors |
|
|
|
|
|
| Printing, publishing (70) |
4 |
1,404 |
351 |
611 |
1,130 |
| Aircraft (114) |
15 |
3,247 |
216 |
492 |
957 |
| Communication equip. (110) |
5 |
2,578 |
516 |
836 |
1,579 |
| Computers, office equip. (103) |
6 |
1,029 |
172 |
396 |
734 |
|
- 260 -
| |
| Part B: Employment Impacts |
|
|
|
|
|
| Sector Name (Number) |
Water Use
(1,000 AF) |
Labor Use
(person
years) |
Direct Workers
per 1,000 AF |
Direct Plus
Indirect
Workers
per 1,000 AF |
Direct, Indirect, Plus
Induced Workers
per 1,000 AF |
| Selected Agricultural Sectors |
|
|
|
|
|
| Hay and pasture (14) |
11,440 |
7,891 |
0.7 |
1.9 |
3.0 |
| Cotton (9) |
4,714 |
10,187 |
2.2 |
7.9 |
11.7 |
| Noncitrus fruits (22) |
3,478 |
114,408 |
32.9 |
43.8 |
52.6 |
| Vegetables (25) |
1,988 |
59,183 |
29.8 |
48.0 |
62.0 |
| All agricultural sectors |
34,460 |
451,000 |
|
|
|
Selected Industrial Sectors |
|
|
|
|
|
| Printing, publishing (70) |
4 |
68,540 |
18,037 |
29,941 |
56,095 |
| Aircraft (114) |
15 |
119,930 |
8,156 |
20,064 |
42,091 |
| Communication equip. (110) |
5 |
101,492 |
20,298 |
33,635 |
69,420 |
| Computers, office equip. (103) |
6 |
63,224 |
11,290 |
22,128 |
39,515 |
| All nonagricultural sectors |
5,604 |
8,831,000 |
|
|
|
| Source: Computed from California Department of
Water Resources, Measuring Economic Impacts: The Application of Input-Output Analysis
to California Water Resource Problems, Bulletin 210, Sacramento (reprinted 1981),
(Tables 3, 15, 16, 17). |
a Income = payments to primary factors of production ("value
added"). |
|
- 261 -
The income measured here refers to payments to primary resources. An
acre-foot of water used in hay and pasture production yielded in 1977 about $30 of direct
income, $60 of direct-plus-indirect income, and $90 of direct, indirect, and induced
incomes. Other agricultural sectors are considerably larger. Comparing the industrial
sectors, it is seen that total annual income per acre-foot is several hundred to several
thousand times as large as in irrigation.
Job creation is another aspect of regional growth policy. The water
requirements per worker for the same sectors are shown in Part B of the table. Less than
one worker per year was directly employed in association with 1,000 acre-feet of water in
hay and pasture production in California. This compares with 8,000 to 20,000 workers per
1,000 acre-feet in the selected industrial sectors. Considering indirect and induced as
well as direct employment shows similar relationships.
Indirect Impacts:
Summing Up
The analysis of indirect regional economic impacts yields similar
inferences to those reached concerning on-farm impacts.
1) The indirect losses to a region giving up irrigation water, while not
insignificant in terms of either monetary flows or employment, will be dwarfed by the
gains in nonagricultural sectors.
2) As in direct impact analysis, there are stair-steps of impacts, when
analyzed on the basis of returns per acre-foot. These steps parallel the steps in the
direct analysis, in that forages and food and feed grains, which account for over half of
water use in western states, yield relatively small indirect employment and income
effects, while the emerging manufacturing and service sectors yield relatively large
increases per unit water of employed.
Caveats
Space and time limitations preclude discussion of several important
aspects of the subject. Specific localities are likely to feel large proportional impacts
of increased urban demands for water. The use of state data may mask the seriousness of
these effects of water reallocation. Where the water is transferred to distant uses,
rather than in the locale of agricultural application,
- 262 -
the community left behind may suffer significant proportional losses
of income and employment; the Owens Valley is an example. Also, the question of indirect
impacts on public sector activities and investments (schools, roads, health care, public
safety) is not considered here. Finally, little information is available on indirect
impacts from instream uses, and that topic has not been touched on in this paper.
Policy Implications
The preceding analysis has shown that as the western states have been
transformed from an agriculturally-based economy toward more manufacturing and eventually
to a primarily service-based economy, the proportion of the irrigation-agriculture sector
to total income and employment has declined. In particular, the proportion of direct and
indirect employment and income generated by the last 10-20 percent of water in irrigation
represents an imperceptible portion of the economy of any of the western states.
We should recognize that changes between sectors are the natural
consequence of an evolving economy.
Policy Implications Regarding Farmers Facing Losses of
Agricultural Water Supplies
In the case of farmers who have a renewable source of supply (usually
surface water or aquifers interrelated with streams), the existence of a problem turns on
the degree to which property rights in water and land are protected by state and federal
law, and hence, whether or not due compensation will be received by the farmers losing the
water.
I perceive a relatively limited threat in this instance. Most farmers
who have sold water rights (either directly or with associated lands) have not only been
amply repaid for foregone productivity of their water, but have shared liberally in the
benefits of alternative uses. Land and water values have been greatly bid up in the face
of anticipated urban, industrial, and energy demands. The fact is that large acreages with
associated water rights in regions of urban growth are held speculatively (by farmers and
others) in anticipation of further asset appreciation. Those who are forced out of farming
are "crying all the way to the bank," and to a subsequent reentry to farming
where land and water is cheaper, or, if desired, to a comfortable retirement
- 263 -
in Sun City, Honolulu, or Acapulco. Chapter 18 in this volume reports
on trends in formalizing property rights in water throughout the West. This trend to
firmer property rights in agricultural water should be encouraged, both to aid in
reallocation to higher valued uses and to assure adequate recompense to resource owners.
Policy Implications Regarding Losses to Indirect
Beneficiaries of Irrigation
The protections afforded by property rights to primary users of water
against the loss of assets is not available to the indirect beneficiaries, who are linked
to irrigated agriculture as input suppliers or processors of products. Even so, at the
risk of appearing insensitive, I can see only a limited basis for concern, and not much
need for formal public action in response.
Most individual transfers of irrigation water supply are neither large
nor unexpected, enabling those indirectly impacted to adapt to new conditions. As seen
above, a small amount of water from agriculture can fuel a large change in a region's
population and industrial base. Even in rapidly growing metropolitan areas, such as near
Denver, Phoenix, or Los Angeles, irrigation continues, and the associated indirect
economic activity and employment decline only slowly. In the face of slowly declining
demand, workers generally have time to plan for career change, and business and public
sectors have time to depreciate their investments without suffering severe economic
losses.
Finally, it might be observed that relatively few instances outside of
irrigated agriculture can be identified where secondary impactees are the subject of
formal public policy concern. Risks are inherent in a changing market economy, as
testified by the changes affecting millions of workers in the industrial Midwest. We need
to think carefully about the justification for public intervention in this case, unless it
is a part of a more general response to the structural changes throughout the economy.
Conclusions
The evidence regarding the role of irrigation in regional economies in
the semiarid West suggests that under modern conditions of production, irrigation accounts
for a relatively minor portion of employment and income. This is particularly true for the
half or more of the irrigation water diversions used for forage, and food and feed grain
production. Second, significant
- 264 -
growth in the nonagricultural sector can be accommodated with
relatively minor shifts from irrigation. Thus, we can expect only a relatively small
impact on local economies by the anticipated limited reduction in irrigation water use.
The general perception that irrigation has been an engine of economic growth, and
conversely that loss of irrigation would have major economic consequences, is not
supported by a close examination of the structure of the economy. This misperception
probably arises from what Boulding terms the mythical role of water in human society,
abetted by the public relations activities of the "iron triangle" of
bureaucracy, construction firms, and legislators who have a stake in "business as
usual" rather than adjusting to the imperatives of a maturing water economy. I see a
limited need for special public policy to solve problems which are similar to, but less
severe than, those in other sectors of the changing economy.
Discussion:
Charles V. Moore
Professor Young has presented a succinct overview of the impacts of
shifting water supplies between agriculture and other uses. I find very little with which
to disagree, although a couple of important points were made that may have been missed by
the noneconomist in the rush of jargon that we economists tend to use in communicating
with each other. I would like to amplify and expand on three points made or alluded to in
the chapter.
First is the possible regional benefits from a transfer of water between
agriculture and municipal and industrial users. Young clearly shows that water transferred
from agriculture to the printing and publishing, aircraft, communications, or computer and
office equipment manufacturing industries will generate increase in regional income and
employment.
Significant benefits can also accrue to the region if water is
transferred within agriculture. The institutional arrangements under which water was
originally allocated in the western states (appropriative water rights, riparian rights,
or long-term contract) did not take into account the productivity of this water as a
criterion.
"First in time" only meant that lands closest to a water
source received the largest and most reliable water supplies. Riparian lands may in fact
contain some of the poorest soils in a river basin. Service areas for governmental water
projects are more closely related to the political power of the local elected
representative than to the productivity of the soils to which the water is to be applied.
Thus the probability that lands with the highest productivity also have the largest and
most secure water supply is very small indeed. The probability that existing institutions
have allocated water in the exact same manner as a free market, where all potential water
users have an equal opportunity to bid for the supply of water, is almost nil. When one
irrigation district has rights to seven acre-feet per acre and another district not too
many miles away has rights to only one acre-foot, this seems prima facie evidence
to prove my point.
If a generously endowed district is applying the last acre-foot to a
field of grain sorghum with a return of $5 per acre-foot and next door a
"Johnny-come-lately" district with very junior water rights has highly
productive cotton or perennial crop land left idle with a potential return of $40 to $50
per acre-foot, both the region and the nation would benefit from a transfer.
- 267 -
Howitt, Mann, and Vaux developed an interregional programming model of
California to analyze the potential for water transfers between subregions of the state
and between agricultural and urban users.[1] Assuming that water
laws were allowed to evolve so that a competitive market for water rights was established,
these authors found that, compared to maintaining existing water laws and allowing new
water supplies to be developed only when users are willing to pay their full costs, the
quasi water market saved up to 2 million acre-feet of water annually, with net benefits to
buyers and sellers of over $70 million. Annual benefits would increase with time and
population growth to $83 million by the year 2020.
I would like to comment on another point alluded to by both Young and
Whittlesey: how price will be determined in water sales. Both provide estimates of the
incremental value in use for irrigation and industrial water. The range of these values is
quite wide, varying from zero to over $40 per acre-foot at the margin for agriculture, and
up to $1,600 per acre-foot for municipal and industrial users.
To make their models workable, economists make some assumptions with
respect to supply and demand functions. For one thing, we assume that demand and supply
functions are continuous, and that both buyers and sellers know and understand these
relations (our assumption of perfect knowledge). However, in the big, cruel world things
don't always work that way. The incremental values in agriculture provided by Young and
Whittlesey become the lower bounds or reservation prices those growers would be willing to
accept. The incremental values for municipal and industrial water then become the upper
limits to price offers by municipalities. The final market clearing price will be
somewhere in between, after allowances are made for transportation costs.
Given that in most states there are many landowners and only a
relatively few large metropolitan areas interested in purchasing water rights, the water
market would be one characterized by an economist as ologopolistic. In other words, the
bargaining power or market power will be in the hands of the urban areas. A market
structure with most of the bargaining power on the side of the buyer will tend to reduce
prices paid to sellers, and shift many of the benefits to urban areas and away from
irrigated agriculture.
One suggestion for equalizing the bargaining power in the marketplace
and thus making the market more competitive
- 268 -
would be for landowners to create a water broker or bargaining agent
to represent their interests. This broker or agent could also serve an important function
by consolidating small lots, or to coin a word, "dribbles," into units large
enough to be attractive to municipal buyers. This function would be especially important
in California, where the legislature has recently passed a bill allowing landowners to
sell water saved through water conservation efforts. Adjustment costs to limited regional
water supplies will be minimized through water transfers away from marginal soils or
marginal farms, rather than wholesale shifts of entire districts out of agriculture.
Every time the subject of water transfers out of agriculture comes up in
California, the question of the Owens Valley is raised. The Owens Valley is located on the
east side of the southern Sierras; in the 1920s, the City of Los Angeles purchased land
and water rights in the valley and subsequently exported water, with farming in the valley
reverting to dryland agriculture. Although books and movies have described the "rape
of Owens Valley," it is my opinion that the expressed anger stems from the feeling
that valley landowners sold out too cheap, rather than that water rights were sold per se.
The crux of the matter is that the sale price was based on the agricultural value of the
water, not on its value to the buyer. If Owens water rights had been sold at something in
the neighborhood of the 1980 equivalent of $150 per acre-foot per year, a much smaller
fuss might have ensued.
One final point is related to the secondary impacts of water transfers.
Young is correct in saying that, from the national point of view, secondary impacts
"wash out". I would suspect that in a state as large as California this relation
would also hold. In states with a smaller economic base than California, however, I would
expect to find measurable impacts.
The question I would like to raise also, without appearing crass, is
"Why all the fuss?" What is the difference between the water transfer case and
the state building a freeway which bypasses a small town and leaves its commercial section
to wither on the vine, or the federal government opening or closing a military facility?
Did anyone in the southwestern states offer compensation to the thousands of sharecroppers
growing cotton in the southeastern U.S., when subsidized irrigation water favored the
shift of the location of cotton production westward?
Traditionally in this country, people injured by these types of
structural changes (what economists call "pecuniary
- 269 -
externalities" or damages) have not been compensated, nor has
government felt it necessary to intervene to prevent adjustments from occurring. When a
movie theater shuts down due to lack of business, government is not expected to move in to
prevent the closing just because the person operating the popcorn stand next door will be
faced with a significantly reduced income.
Discussion:
Norman K. Whittlesey
I agree with Young's findings that reallocation of water from
agriculture to competing uses is unlikely to cause serious problems for the agricultural
community or for other regional and national economies. Both the farmer and the nation's
economy will profit from reallocations of water to higher valued uses. The total amount of
water exchanged in any given area will usually be relatively small, providing ample room
for adjustment by all affected parties.
Though water is a "mobile resource" leading to problems of
capture for private property management, it is difficult and expensive to move far from
present locations. Hence, the values given to water are likely to be site-specific and
highly influenced by other resources, competing water uses, the assumptions of the
analyst, plus many other factors.
Young acknowledges that different assumptions can lead to different
estimates of agricultural water value. However, he implies that the differences between
agricultural and nonagricultural water values are so great as to render the correct value
unimportant. This may not always be true.
I have developed an example in Table 10.2 to show how assumptions can
affect the estimated value of irrigation water. These values are based on consumptive use,
so that an appropriate adjustment would be required to obtain values for total water
diversions.
The first value is a very short run measure called "returns above
variable costs." It could be the rental price for water in an
- 270 -
| |
Table 10.2
Estimation of Alternative Agricultural Water Values
for an Average Crop Rotation |
| Consumptive Water Requirement (AF/A) |
2.50 |
| Gross Revenue ($/A) |
397 |
| Costs ($/A) |
|
| Variable production (VC) |
137 |
| Fixed excluding land and water (FC) |
64 |
| Irrigated land (IL) |
115 |
| Nonirrigated land (DL) |
20 |
| Water delivery and application (WC) |
80 |
| Net Returns to Water Above ($/AF): |
|
| VC |
104 |
| VC + FC + DL |
70 |
| VC+FC +DL +W |
38 |
| VC + FL +IL +W |
1 |
|
emergency drought situation, where it could go even higher for
the rescue of perennial crops like orchards or vineyards. This value of water is sometimes
called "value added" to indicate the economic impact of agriculture in a
regional or state economy as estimated by an input-output model. In this context, it is a
measure of payments for fixed resources used in the production process. This value would
not likely be used to determine a market price for permanent exchange of water rights, nor
would it provide any indication of the profitability of agriculture.
For water exchanges that leave land and undepreciated irrigation
facilities idle, we move close to the average value of water to determine its market price
and measure the social impact of exchange. This value is probably best reflected by the
returns above costs including dryland rent ($70/acre-foot). But if the irrigation
facilities are allowed to depreciate normally (or salvaged) before the water is exchanged,
we should probably move to the next water value ($38) which deducts the current payment
for the irrigation facility.
If the land and irrigation facility usefulness are not reduced, the
value of both resources must be deducted from the value of water sold. At this point the
marginal value of water will be
- 271 -
near zero, as reflected by the last line of the table. No rational
farmer is going to sell water at this price, even though it probably reflects the social
impact of small quantities of water diversions to alternative uses. This is Young's
conclusion.
This exercise illustrates that alternative agricultural water values can
be used to compare with values in competitive uses. The result is site-specific,
crop-specific, and operator-specific and, to a large extent, depends upon which side hires
the "best" economists and lawyers in the bargaining process.
Even after deciding which is the proper agricultural value of water to
capitalize, we are left with questions about discount rates and planning horizons in
establishing market exchange values. I believe that it would be possible to argue for
lower discount rates and longer planning periods in agriculture than for industry. Such
assumptions can move the values of agricultural and competing uses closer together than is
apparent when only comparing annual rental values. In any case, the derived demand for
water in most industrial uses is very inelastic and can generally be satisfied with little
impact on agriculture.
Generally, I would agree that small incremental adjustments in water use
are unlikely to cause any serious secondary economic impacts. We should always expect
measures of impact on the national economy to be positive for water reallocations.
However, we must recognize that rather large economic communities throughout the West are
based solely upon irrigation. As the quantity of water and distance to the new use are
increased, the negative local impacts will also increase, regardless of how well the
farmer is compensated. There are examples of whole communities being significantly reduced
by the thirst of Los Angeles and the California water system-not to deny that the state
and society have been made better off after the reallocation.
A recent study in Washington State provides some estimates of secondary
impacts of irrigation.[1] Though subject to all of the vagueness of
input-output analysis, the results do show greater secondary impacts than the studies
quoted by Young. Increased value added at the secondary level was approximately $540 per
acre or about $140 per acre-foot of water diversion. Additional employment created at the
secondary level equaled one job per 26 acres or 100 acre-feet of water diverted.
To conclude, we should not be unduly concerned about reallocation of
water from agriculture to competing uses. In fact, we should aid that transition whenever
possible through better economic studies and improvements in the legal and institutional
- 272 -
system for water exchange. The net societal impact will be positive.
We must, however, be prepared to deal with problems of primary value and secondary impact
when they do arise.
