Title: Water Allocation And/Or Water Use Efficiency
Full Citation
Permanent Link: http://ufdc.ufl.edu/WL00003044/00001
 Material Information
Title: Water Allocation And/Or Water Use Efficiency
Physical Description: Book
Language: English
Publisher: American Society of Civil Engineers
Spatial Coverage: North America -- United States of America -- Florida
Abstract: Richard Hamann's Collection - Water Allocation And/Or Water Use Efficiency
General Note: Box 12, Folder 6 ( Legal, Institutional and Social Aspects of Irrigation and Drainage and Water Resources Planning and Management - 1979 ), Item 10
Funding: Digitized by the Legal Technology Institute in the Levin College of Law at the University of Florida.
 Record Information
Bibliographic ID: WL00003044
Volume ID: VID00001
Source Institution: Levin College of Law, University of Florida
Holding Location: Levin College of Law, University of Florida
Rights Management: All rights reserved by the source institution and holding location.

Full Text


The Value of Water for Irrigation in the Western United States

Unfortunately irrigation water is not often bought and sold in the
context of a freely operating market. Accordingly there are no readily
observable market prices to use as measures of value. What is needed,
then, is a proxy market price; i.e., an estimate of the value for water
that would be revealed if water were bought and sold through a freely
operating market.

Fortunately, in cases where a nonmarketed resource such as water
is used in the production of a good or service that is exchanged in the
marketplace (e.g., food and fiber), the development of a proxy or im-
puted market price for that resource is relatively straightforward.
One merely needs to determine the additional product attributable to an
additional unit(s) of the resource and multiply by the price of the
marketed product. In this manner, the value of (ability-to-pay for)
the resource in a particular use can be estimated.

During the last two years we have been involved in a study at
Texas A&M University concerned with the estimation of the economic
demand for (value of) irrigation water in the major irrigated subregions
of the western United States. Our principal data source was the 1969
Census of Agriculture with counties being the observational units. The
criteria used in delineating the study regions were (i) a significant
number of irrigated acres among total acres harvested, (ii) a relatively
homogenous type-of-farming area which in our case means mainly similar
crop-livestock mixes and (iii) similar climatic and topographic charac-
teristics. Agricultural production functions were estimated, and water
demand schedules derived, for a typical (composite) irrigated acre for
eleven different regions as shown in Figure 1.

For each of these regions a Cobb-Douglas type production function
of the following general form was estimated:

l1 22 9
Y = Bo XI X2 "'X

Y = Value of agricultural output--value of crops harvested and
livestock and livestock product sales (dollars/county)
X1 = Irrigation water applied (acre-feet/county)
X2 = Value of land and service buildings (dollars/county)
X = Hired labor expenditures (dollars/county)
X4 = Fuel and lubricant expenditures (dollars/county)
X5 = Fertilizer and lime expenditures (dollars/county)
X6 = Value of machinery inventory (dollars/county)
X7 = Value of livestock inventory (dollars/county)
X = Livestock feed expenditures (dollars/county)
X9 = Other operating expenses (dollars/county)


1 Snake-Columbia Basin
2 Central California
3 Desert Southwest
4 Upper Colorado Basin
5 Upper Rio Grande Basin
6 Lower Rio Grande Basin
7 Upper Missouri Basin
8 NW Ogallala
9 NE Ogallala
10 Central Ogallala
11 Southern Ogallata

FIGURE 1. Major Irrigated Regions of the 17 Western States.


A logrithmic transformation of this function was fitted using ordinary-
least-squares (OLS) regression and using ridge regression. As expected
parameter estimates under OLS were highly unstable due to high correla-
tions among the explanatory variable (multicollinearity). Several co-
efficients took on nonsensical negative signs and standard errors of
the estimated coefficients were generally high resulting in nonsignifi-
cance of many of the estimates. This posed a serious problem as the
quality of the estimation of the individual structural parameters in-
fluences greatly imputed values to irrigation water.

To mitigate this multicollinearity problem we employed ridge re-
gression (Hoerl & Kennard). While admittedly a biased estimation tech-
nique, ridge regression does permit substantial reduction in the stan-
dard error of estimation in the presence of multicollinearity, hopefully
resulting in lower mean-squared-error. Conditions for which ridge esti-
mates are likely of higher quality than OLS estimates have been noted by
Brown and the usefulness of ridge regression for agricultural production
function estimates has been demonstrated by Brown and Beattie.

A point estimate of the value of irrigation water for each region
is presented in Table 1. Before discussing these estimates the reader
is cautioned that they are point estimates representing the value of
irrigation water where all explanatory variables (input levels) are
fixed at their respective regional mean values. Lower water application
rates in these regions would result in higher imputed values to irriga-
tion water due to the downward sloping marginal value (benefit) function
for water. However, these estimates are probably quite reliable from a
longer-run (more than one production season) perspective because the
long-run input demand schedules tend to be quite flat for production
processes that exhibit near constant returns to scale.

Examination of the results in Table 1 indicates generally low mar-
ginal value per acre foot of irrigation water in the West. According
to our estimates the marginal value per acre foot of irrigation water
ranges from a low of $2.31 in Upper Rio Grande Basin to a high of
$24.35 in the Central Ogallala. One might reasonably pose the question:
Are irrigation water values really this low or is there some hidden con-
ceptual or estimation shortcoming in our analysis? We believe our esti-
mates are reliable. Not only are our estimates consistent with the find-
ings of other researchers (see Young and Gray), but an indirect check
suggests water values in agriculture were probably of this magnitude in
1969. In the Central California Region we have access to data on costs
of irrigation water. These data suggest average water prices in 1969 in
the order of $5 to $7 per acre foot (Shumway and Stults).1 If farmers
behave rationally then we would expect them to purchase and apply water
and/or bear pumping and delivery costs up to the point where the value
of water in production was greater than or near that cost. Indeed that
appears to be the case for the Central California Region.

Certainly agricultural irrigation water values are sensitive to
farm product prices and costs of procuring and applying irrigation
water. Other things equal, increases in irrigation costs will increase

These costs varied from a low of $0 to a high of $19.36 for
various subareas of the region.


Table 1. Value of Irrigation Water For The Major
Irrigated Regions of the Western United
States, 1969.*

Region Value ($/A.F.)

Snake-Columbia Basin 3.88

Central California 5.14

Desert Southwest 14.45

Upper Colorado Basin 6.53

Upper Rio Grande Basin 2.31

Lower Rio Grande Plain 14.44

Upper Missouri Basin 5.40

Northwestern Ogallala 15.24

Northeastern Ogallala 11.27

Central Ogallala 24.35

Southern Ogallala 5.98

product prices

*Based on regional mean values and
in 1969.


the marginal value of irrigation water associated with a reduction in
water applied. Similarly, increase in farm product prices will also
increase the value of irrigation water. For example, farm product
prices for crops grown on the Texas High Plains--principally, cotton,
grain sorghum and wheat--increased by 90 percent between 1969 and 1978.
Other things equal (e.g., no increase in pumping costs and other input
prices) this would increase the value of irrigation water by a like
amount. Obviously decreases in farm product prices have the opposite
effect of reducing agriculture's ability-to-pay for water.

Implications for Western Irrigated Agriculture

Given generally low values of water in irrigated agriculture and
the sensitivity of these values to unstable farm product prices and
rising irrigation costs, the prospects for a continued viable irrigated
agriculture in the West is of considerable concern to many. Fortunate-
ly, in the foreseeable future (30 to 40 years anyway) the major issue
confronting agriculture is not, in our view,'competition with other
uses. For surely in any economic sense, agriculture would be unable to
compete in terms of ability-to-pay with most other uses for water, if
it came to that.

In many irrigated agricultural areas, agriculture is far and away
the largest consumptive user of water. Thus, it seems unlikely that
the West will have any major or insurmountable problems in meeting its
demands for water for nonagricultural uses with the possible exception
of the needs for environmental and recreational uses. Agriculture's
real concern is whether it can afford the cost of using that water
which is in excess of that demanded by higher valued uses. Namely, can
agriculture bear the continuing rise in the cost of energy for pump
irrigation? Can agriculture bear the cost of the development of alter-
native or supplemental water supplies to replenish or replace declining

The picture provided by the estimates presented in Table 1 is
fairly grim. For example, it is apparent that current and anticipated
pumping costs have risen to the point where continued irrigation activi-
ty at present levels is vulnerable on the Texas High Plains and other
subregions of the Ogallala formation underlying much of the Great
Plains. Assuming 1971-74 average crop prices, Lacewell found that if
natural gas prices rise to about $2.25 per thousand cubic feet, cotton
will no longer be irrigated on the Texas High Plains. At a $3.00
natural gas price, grain sorghum shifts from irrigated to dryland pro-
duction and with a $4.67 gas price all production would revert to dry-
land (Lacewell, p. 61). It is clear that the situation is tight and
has the potential to get even tighter. As pumping depths and energy
prices continue to increase and/or if farm product prices tend to
stabilize at an unfavorable level, the decline of irrigation activity
in its present form seems inevitable. Further, the ability of agricul-
ture to pay for imported water to replace or substitute for groundwater
is constrained to the $5 to $24 per acre foot range. It seems unlikely
that there are any presently untapped sources of water that can be
delivered to agriculture at or near that price.


Is the Social Value of Irrigation Water Greater
Than Agriculture's Ability-To-Pay?

In view of agriculture's limited ability-to-pay for water, are
there considerations that would justify placing a higher value of the
use of water for irrigation than that implied by the estimates in Table
1? Or to put it another way, is the social value of irrigation water
now or in the foreseeable future greater than agriculture's ability-to-

A number of issues are often raised by the agricultural establish-
ment that would suggest the possibility that this might be the case.
Among others, claims for higher agricultural water values are often
argued based on (i) anticipated increases in farm product prices due to
increased domestic and export demand relative to supply capacity for
U.S. agriculture and (ii) indirect benefits or multiplier effects of
irrigated agriculture. We turn to a brief discussion of the validity
of these issues.

U.S. Domestic and Export Demand and Supply Situation

First, regarding the issue of the U.S. food and fiber situation
and prospects for export demand for the foreseeable future, the most
comprehensive work on this issue has been done by Earl Heady and his
associates at Iowa State University. Results of their rather massive
modeling efforts done under contract for the National Water Commision,
and significantly influencing the thinking of the Commission, suggest
that U.S. agriculture has the capacity to meet domestic and export
demands for the foreseeable future without significantly higher food
prices and with irrigation water supplies not being a constraint.

In discussing the implications of his modeling results, which
among the alternative scenarios included a higher than expected U.S.
population growth and reasonably ambitious expectations regarding export
demand, Heady suggests:

...the overall conclusion appears to be this. U.S. agricul-.
tural supply capacity will continue to be large in the future
in the absence of extremely binding environmental restraints.
...the United States could readily meet domestic demands in
year 2000 and have exports as large as 3,209 million bushels
of corn, 1,700 million bushels of soybeans, 1,500 million
bushels of wheat and some increases in other commodities.
...this nation does have great supply capacity and great
flexibility in how its grains are used and allocated among
uses and exports. Without increments in export demands
greater than those now in sight, my prediction is for supply
capacity which is periodically large relative to export demand;
then volatile price swings in some years when world crop short-
falls are large in comparison with our export surpluses. In
these cases, the problem is not so much a shortage of land
and water resources to produce in sufficient quantities. It
is more nearly one of lack of reserve stocks (Heady, pp. 27-28).


If one believes the Heady results, we do not find justification in
terms of supply shortfalls relative to expected domestic and export
demands for anticipating significantly higher than present real prices
for farm products and thus imputed value to irrigation water in the
case of irrigated agriculture.

Indirect Benefits of Irrigated Agriculture

Another issue that is often used to advocate the placing of a
higher social value on irrigation water than agriculture's ability-to-
pay involves the idea of indirect benefits spurred by increased econo-
mic activity attributable to irrigation. This idea is a particularly
illusive and dangerous one.

It is obvious that the indirect economic benefits resulting from
the regional multiplier or linkage effects of a thriving and prosperous
irrigation agriculture are high. Clearly in such cases indirect bene-
ficiaries (e.g., financial institutions, irrigation and farm equipment
suppliers, agricultural product processors, service sectors and other
businesses linked directly or indirectly to agriculture) have a consid-
erable stake and interest in the water resource base giving rise to a
substantial portion of their business activity. Accordingly, it is
understandable that these indirect beneficiaries are often among the
leaders in promoting water development in order to expand or maintain
that economic base.

However, it does not follow that the existence of regional multi-
plier effects justifies public investment in, or subsidization of,
water development or importation. In most cases cost spreading to
non-beneficiaries (public subsidization) amounts mostly to an inter-
regional transfer of income and not a net social economic gain. In
fact, if resources are reasonably fully employed, are mobile, and there
is an absence of excess capacity, then no real efficiency gains in the
form of net indirect benefits exist because gains in one region will
be nearly fully offset by indirect losses elsewhere. That is, there
will be indirect costs that must be counted on the cost side of the
equation as well--costs that will rival in magnitude the indirect
benefits. Accordingly, if indirect benefits are claimed in an economic
justification for public investment in water development, then only
those gains that exceed indirect losses elsewhere should be counted.
Unfortunately this excess is, as often as not, negligible or negative.
In most cases it seems unlikely that real net indirect benefits will
contribute significantly to a higher social value for irrigation water
than agriculture's ability-to-pay.


In summary, we have looked at the marginal social value of irriga-
tion water in the West,finding the value to be in the range of from $2
to $24 per acre foot. Since irrigated agriculture is a major consump-
tive user of water and the value in use is low relative to many other
uses, we considered two issues that are often suggested for claiming


a higher social value for irrigation water than agriculture's ability-
to-pay. Finding these issues either unlikely or invalid arguments
where does that leave us concerning the future of irrigated agriculture
in the West?

First, it is important to indicate that there are vast quantities
of groundwater in storage. This means that except for some local situa-
tions agriculture will not really be competing with other users.

With costs of pumping approaching and exceeding value of water
used for irrigation, other adjustments can be anticipated. First,
farmers will tend to reduce the quantity of water applied per acre.
Although yields will probably decline somewhat, the production response
to the last unit of water will be larger, thus increasing the value of

Additionally, the opportunities afforded by new technology in both
rain-fed and irrigated agriculture must be considered. New innovative
crop production systems and irrigation distribution systems can be
expected which improve the efficiency of water used. For example, a
new cotton production system developed for South Texas which incorpo-
rated a new variety with changes in tillage practices, fertilizer
practices, irrigation applications and pest management strategies has
shown dramatic results (Sprott, et.al.'). This new cotton production
system has been widely adopted throughout the region. Per acre appli-
cation of irrigation water has been reduced 30 percent, overall reduc-
tion in energy use has been 33 percent while yield has increased over
30 percent.

The opportunity for new irrigation distribution systems is substan-
tial. Some economic impacts have been estimated based on a low pressure
mobile trickle system developed by Bill Lyle of the Texas Agricultural
Experiment Station. On 1.74 million sprinkler-irrigated acres on the
Texas High Plains, the increase in returns to water (present value of
the water supply) over a 20 year period would be $474 million for a
25 percent water efficiency improvement (Hardin, et.al.).

Clearly irrigated agriculture is in an economically precarious
position due to competition from other uses, declining aquifers and
increased pumping costs. Yet, considering the vast quantities of
groundwater still in reserve, opportunities to reduce irrigation appli-
cation rates and the evolution of new irrigation and rain-fed agricul-
tural technologies, there will be a continuing important irrigated
agriculture sector in the West.



Brown, William G., Effect of Omitting Relevant Variables Versus Use of
Ridge Regression in Economic Research, Oregon Agricultural Experi-
ment Station Special Report 394, October 1973.

Brown, William G. and B.R. Beattie, "Improving Estimates of Economics
Parameters by Use of Ridge Regression with Production Function
Applications," American Journal of Agricultural Economics 57
(1975): 21-32.

Hardin, Daniel C., Ronald D. Lacewell, and James A. Petty, "The Value
of Improved Irrigation Distribution Efficiency with a Declining
Groundwater Supply," Texas Agricultural Experiment Station
TA 14294 June 1978.

Heady, Earl O., "U.S. Supply Situation for Food and Fiber and the Role
of Irrigated Agriculture," In: The TAMU Centennial Year Water for
Texas Conference: Water for Food and Fiber Production, Texas
Water Resources Institute, Texas A&M University, College Station,

Hoerl, Arthur E. and R.W. Kennard, "Ridge Regression: Biased Estimation
for Nonorthogonal Problems," Technometrics 12 (1970a): 55-67.

Lacewell, Ronald D., "Impact of Energy Cost on Food and Fiber Produc-
tion," In: The TAMU Centennial Year Water for Texas Conference:
Water for Food and Fiber Production, Texas Water Resources
Institute, Texas A&M University, College Station, 1976.

National Water Commission, Water Policies for the Future: Final Report
to the President and to the Congress of the United States by the
National Water Commission, Water Information Center, Inc., Port
Washington, N.Y., 1973.

Shumway, C. Richard and H.M. Stults, Production Costs and Yields of
California Field Crops and Vegetables by Area, Average 1961-65
and Projected 1980, California Agricultural Experiment Station
Report No. 307, March 1970.

Sprott, J. Michael, Ronald D. Lacewell, G.A. Niles, J.K. Walker and
J.R. Gannaway, "Agronomic, Economics, Energy and Environmental
Implications of Short-Season Narrow-Row Cotton Production," Texas
Agricultural Experiment Station MP-1250C, February 1976.

U.S. Bureau of Census, U.S. Census of Agriculture, Vol. 1, State and
County Statistics, Washington, D.C., 1969.

Young, Robert A. and S.L. Gray, Economic Value of Water: Concepts and
Empirical Estimates, PB-210356, Dept. of Economics, Colorado State
University, March 1972.

Crop Response Information
for Water Institutions

Gary D. Lynne and Roy R. Carriker

It has been said that water institutions, on a world wide scale,"
.... almost universally fail to provide for a socially efficient water
use" [38]. This is a sweeping indictment in light of the ever increas-
ing scarcity of the water resource. One of the major reasons for such
inefficiency, if it exists, may be the lack of sufficient information
of satisfactory quality. As noted by Fox, the lack of appropriate
types of information may be one of the most difficult problems to
overcome in institutional arrangements [13, p. 753], and thus one of
the greatest contributors to "social inefficiency" in the performance
of a water institution.

This paper outlines a conceptual basis for integrating crop-water
response information into the decision making processes for water
allocation. It proceeds from a discussion of the economic role of
institutions for resource allocation and the nature of efficiency, with
particular reference to a system of administrative water law adopted by
the State of Florida. An appropriate treatment of crop-water response
information within this decision making framework is then examined, and
problems attending the incorporation of crop water response data in
water management decision making are noted.

Resource Scarcity and the Role of Institutions

Most natural resources, including water, are scarce at certain
times and certain places. Scarcity, in this context, occurs whenever
quantity demanded exceeds quantity supplied under existing terms of
trade. The condition of scarcity gives rise to the need for a decision
framework by which the scarce resource is allocated among competing
uses and users over time and space. This decision framework for
resource allocation is referred to as an institution. In customary
usage, the term institution refers either to an entity (an organization
or an individual), or a rule (a law, regulation, or establishment
custom) [13, p. 743].

Assistant Professors, Food and Resource Economics Department,
University of Florida, Gainesville, Florida.


Many goods, services, and resources are allocated by a process of
voluntary exchange through the market mechanism, which, in turn, is
based upon institutions of private property especially encouraged in
capitalistic economies and democratic societies. For example, coal,
lumber, feed grains, shoes, and haircuts are resources, goods, or
services allocated by voluntary exchange through market institutions.
Water (along with many other goods and services like police protection,
national defense, and lighthouses), is generally provided by a non-
market institution which is charged with resolving the supply demand

Water Management Institutions In Florida

The problem of water scarcity can be solved by some combination of
measures to augment supplies and/or curtail demand. Responsibility and
authority for deciding which combination of measures will be imple-
mented is a function of the institutional framework for water resource
allocation. Historically, for Florida (as with most of the Eastern
United States) the institutional framework for water management
(broadly speaking) consisted of common law water rights doctrines which
evolved over time in custom and case law. Under common law doctrines
the courts provided the definition and enforcement of an individual's
property right to water. When localized instances of scarcity produced
conflict between individuals over some aspect of water use, the case
was settled in court. As long as water supplies were relatively
abundant compared to demands, any given use was unlikely to interfere
with any other, and the water institution functioned fairly well.

With intensification of economic activities and corresponding
increases in water use, the potential for conflicts increased. The
doctrines resulting from case law and custom became cumbersome and were
not well suited to the task of reconciling water supplies with compet-
ing water demands in Florida. The Florida legislature, in light of this
problem, introduced a comprehensive administrative system of water
regulation. The Florida Water Resources Act of 1972 declares that
"... all the waters of the state are subject to regulation..." and
establishes an administrative structure to carry out the regulation
[6].1 The Department of Environmental Regulation provides overall
state coordination and five water management districts administer
regulations locally through appointed governing boards. This Act
assigned entitlement to the use of water to all the people of the
state, collectively, charging the districts to manage the waters of the
state in the public interest, authorizing them to require permits for
consumptive uses of water, for the drilling of wells, and for alter-
ations of drainage patterns.

The Model Water Code [28] was used as the basis for the Act.


For a consumptive use permit to be granted it must be established
that the proposed use is a "reasonable-beneficial" use, will not
interfere with any presently existing legal use (a permitted use), and
is consistent with the public interest. The Act defines reasonable-
beneficial use as "... the use of water in such quantities as is
necessary for economic and efficient utilization, for a purpose and in
a manner which is both reasonable and consistent with the public
interest ..." [6]. Other than these broad guidelines the Act provides
little guidance as to how private entitlements to use water and modify
drainage systems are to be established. It directs the districts to
develop "rules and regulations." The districts have, to this point in
time, been granted flexibility in establishing principles upon which
to base rules and regulations for the issuance of consumptive use
permits. As a result, they have certain information needs.

Information Needs of Water Districts

The type and extent of information sought by Water Management
Districts reflects the manner in which each district has defined its
role. None of the five districts in Florida publicly admits an
explicit economic role as an institution for resource allocation. In
areas where water is relatively scarce, much emphasis is placed on
conservation and avoidance of waste, yet the districts espouse
policies of permitting all legitimate uses of water. No explicit
attention is given the economic value of water in alternative uses,
and thus to social efficiency.

A series of seminars on the subject of information needs, with
staff personnel of four of the five districts in Florida, revealed that
information regarding physical water supply considerations received
highest priority.2 Interest was expressed in improved ways to measure
rainfall in particular locations over time, in better understanding the
storage and transmissivity of underground aquifers, and in computer
simulation of hydrologic systems.

Interest in water demand was usually expressed only indirectly as
a need to inventory irrigated and other land use types, to measure
evapotranspiration for particular uses over time, to identify "require-
ments" for fresh water releases to saline estuaries, and to improve
population projections. Staff at all four of the water management
districts expressed an interest in understanding factors affecting
economic demand and supply of water, but few resources in the various
water management districts are allocated to socio-economic measurement,
data collection, and analysis. It can be argued, however, that
economic information is important in the water allocation process.

This was also a concern of Milliman [31].


Economic Value as Useful Information

Resource scarcity by definition entails choice. Water management
districts with responsibility for granting consumptive use permits, and
for implementing emergency measures during periods of severe water
shortage, cannot escape the reality of the economic role in water
allocation. Arguments to the contrary stress opportunities for supply
augmentation, desalinization, interregional water pipelines, impound-
ments, etc. However, such measures deliver water, but only by incurring
costs. A decision to raise tax levels for supply augmentation could
result in sustainance of relatively trivial consumptive uses of water.
The tax money may have been better left in the other economic activity
from which it was drawn; i.e., there may be large opportunity costs of
supply augmentation.

Other arguments to the contrary stress opportunities for conserva-
tion, insisting that all foreseeable essential uses can be supplied if
only waste can be reduced. The efficacy of such arguments requires an
ability to distinguish "wasteful" uses from essential uses. Knowledge
of economic value is also important to this classification process.

The value of water to any user will be quite high for those small
amounts necessary forsurvival. However, there is a diminishing value
for each additional increment of water after "essential" needs are met.
For example, a certain minimum amount of water is essential to an
individual. Any additional water per day will have a lower value, and
probably a zero value as drinking water. In crop irrigation, that
quantity (per time period) which is necessary to keep perennial crops
alive during a drought is highly valued. Additional quantities which
improve yields may also be highly valued. However, there is likely a
point at which additional returns (yields) do not cover the additional
opportunity costs, or even the costs of applying the water.

The decline in value for each additional unit implies that the value
of most water is determined by its relative scarcity, not by its
essentiality. This has some inescapable consequences. When less water
can be made available than is demanded, given that the only prices for
water are nominal service charges, a decision to permit water for one
use means that it will not be available for another. If higher valued
uses are being denied in order that low valued units of water are
availabj. to another use, unnecessary hardship may be wrought to the
first user with little gain to the second, and a loss to the society at

The value of additional (marginal) units of water in alternative
uses is also helpful information when public investments in facilities
for supply augmentation are being considered. The opportunities to
compare the value in use of the additional water supplies with the cost
per unit of supplying the additional water enables public decision
makers to ascertain whether or not the investment of tax money yields a
net return to the community. If the intent of policy is to charge
users the cost of supplying additional units of water, the marginal
value of water in alternative uses serves as an indicator of which
uses and users will be willing and able to pay for the additional water


supply. The need to understand the economic value of water in crop
production, in turn, gives justification for water-crop response

The Relevance of Crop Response to Irrigation

Under an institutional framework characterized by administrative
water law (such as the Florida system), water management district
decisions concerning criteria for the granting of permits are the focus
for information on the value of water in alternative uses. There is
real competition for limited supplies. Special interest groups
typically participate at public hearings (and through private lobbying
efforts) to influence water district decisions concerning water
allocation, management, conservation, and development. Agricultural
firms seeking irrigation water must apply for consumptive use permits
along with municipal, commerical, and industrial users, and must
compete with other users. The concern of agricultural irrigators in
this decision process in Florida has been to insure that enough
irrigation water is permitted for them to profitably produce crops.
The reasons for their concern are straightforward and easily understood.
Distribution and "equity" are of primary concern in the current Florida
water allocation process.

As water supplies of good quality become more scarce, however,
agricultural irrigators (as well as other water users) may be asked to
demonstrate the value in use of additional allotments or to estimate
the loss in value of production from incremental curtailments of water.
Such estimates require two related sets of information. The soil-water-
plant relations and conditions affecting crop yield, must be quantified.
In particular, the increments to crop yield from increments in water
application must be known. These physical crop-water response rela-
tions must then be combined with several economic considerations to
arrive at an estimate of the value from use for crop production. This
information then becomes useful for regional water management decision

Recent Agronomic, Engineering and
Economic Research on Crop-Water Response

Issues of major journals for the period 1960-1977 were searched for
articles dealing with yield-water relations. The selection process was
limited to those studies which attempted to quantify the relationship
between crop yield and water applications. Over 40 such articles were

Much of the current research reflects a concern with determining
maximum yield and/or maximum yield per unit of water [1, 10, 11, 12, 15,
24, 30, 37, 40, 42, 43, 46]. Such results have limited relevance as
information for decision making to allocate scarce resources since they
focus on the relative productivity of inputs without accounting for the
equally important effects of relative input costs and output prices on



the firm's optimum solution.

Other studies were primarily concern with measuring functional
relationships between yield and various treatments of water, but gave
no explicit attention to identifying any sort of optimizing and/or
choice criterion [3, 9, 10, 17, 18, 20, 25, 28, 32, 34, 35, 39, 45, 49].
Other studies did integrate choice criterion and the physical soil-
water-plant relationships with varying degrees of success [2, 4, 5, 14,
16, 19, 22, 23, 26, 33, 35, 36, 47, 48, 51, 52]. These latter studies
may prove especially useful to future efforts at estimating the
marginal productivity and the marginal value of irrigation water. Some
studies recognized the role of economic considerations in crop-water
response models, but made no effort to incorporate them [9, 45, 46].3
A study by Mapp and Eidman [29] was of particular interest because it
integrated crop-water response estimates with economic information in
an analysis of possible alternative water allocation rules such as
those promulgated by regional water management authorities.

Generally, studies were found to be lacking in their suitability
for integration into institutional decision making. Consideration must
be given both physical and economic considerations if such integration
is to be successful, and the approach of Mapp and Eidman is encouraging
in this regard. The following conceptual model should guide the crop-
water estimation process if such integration is to be attained.

A Conceptual Framework for Crop
Response Models and Agricultural
Water Demand Estimation

The following formulation recognizes that the agricultural
irrigator not only has an economic demand for water, but that he must
also generally operate his own water supply service with its concomi-
tant investment and operation costs. The formulation describes,
conceptually, the demand and supply relationships for crop irrigation
water, and broadly outlines the decision calculus of the firm manager
who wishes to maximize his net returns to crop production.

Long run water demand by an irrigation firm can be presented
generally as follows:

Wi = g(rwid, rl, ..., rn' p, Wp T) (1)


Wi = irrigation water available at the field

A noted Natural Resource Economist has chastised social scientists
for not giving due regard to fundamental concepts of the biological and
physical sciences in social science [7, p. 3]. We see a double edged
sword as appropriate to this issue.


r .d marginal value per unit pumped (at field)

p = price of output q

r, ..., rn = marginal cost of other inputs X used to produce output q,
where j = 1, ..., n

W = water from other sources reaching the field, including
P precipitation4

T other environmental conditions, such as temperature.

In short-run periods, the demand function becomes

Wi = g(rid Wp, T/X1, ..., X p) (2)

where the variables are defined as before, except that

X, ... Xn = the actual level of other factors (irrigation
system type, fertilizer levels, labor inputs,
etc.) assumed fixed (invariant) during the time
period to which the analysis applies.

The water supply side of the firm's operation is described as follows:

Wi = h(rwis, Wa I, rl, ... rk) (3)

W = water supplies to ET

rwis = marginal water cost

W = a water availability factor, recognizing such natural
conditions as aquifer recharge rate, surface stream
flow, etc.

I = limitations, conditions or other regulations imposed by
the institutional framework within which individuals make
private water use decisions

r ..., rk= prices of other inputs used to "produce" Wi (irrigation
system costs per unit of water delivered, etc.).

4It is assumed in this formulation that yield (q) responds to total
water (Wt) available at the field level where Wt = Wi + Wp. All
functions and costs are expressed in a manner consistent with this
assumption. The term Wt is generally referred to as evapotranspiration
(ET) in the agronomic, crop science, soils, and engineering literature.


In the short run, the supply equation is

W = h(rwis /Wa I, i...vk) (4)

The variables are defined as in (3) and

vi, ..., vk = those features and input levels of the water supply
portion of the firm that are not varied easily in day to
day operations.

The actual quantity of water demanded in the long run by agricul-
tural irrigators, seeking to maximize their net returns to crop
production, will be that quantity which satisfies simultaneously
equations (1) and (3). In the short run, effective demand will be
given by simultaneous solution of Equations (2) and (4). In both
cases, the demand on the water resources of the area becomes a function
of the internal (to the firm) demand/supply reconciliation process.
The constraints imposed by the institutional framework within which
irrigators make their own water use decisions are felt through their
effects on the water supply costs of the irrigator the farm firm.
The ultimate demand for water by the firm becomes apparent, not simply
as a result of biological plant-water relationships, but through the
process by which farm managers combine their knowledge of plant water
relationships, relative input costs, and product price to maximize net
returns to crop production. Water managers and decision makers in the
water institution must-consider all of these elements in the decision

Other Problems in Integrating Crop Water
Response Information

While appropriate physical-economic models are a necessary condi-
tion for total integration into institutional decision making there
are other considerations equally as important. At least three other
reasons why successful integration may be difficult exist:

1. the decision body and/or the staff of the water institution may
have goals other than those related to economic (and social)
efficiency. This may also be the case for the special interest
groups served by the institution.

2. the very structure of the institution may preclude the inte-
gration of crop-water response information.

3. the individuals in the water institution may lack the training
and understanding of the concepts, even though they may be
receptive to the idea of using crop response models.

A discussion of each of these follows:

Goal Functions. The existence of multiple objective planning functions


relative to water resources management is certainly a reality, and
need not be discussed thoroughly here. It must be remembered, however,
that lack of concern (lack of goal) for economic efficiency may make
integration of crop-water models nearly impossible and, perhaps,
irrelevant. If the opening paragraph of this paper is indeed correct,
this factor may be pervasive throughout the water institutions of the
world, and should not be discounted. Thus, estimates of the crop-water
relation, while an interesting research effort, may never be used in
the real world of water allocation.

A related possibility is that those currently in the power
structure and having control over the water management process simply
wish to keep things as they currently are, thus feeling no need for new
types of information. Wolman [50] notes this may occur if the new
alternative might reduce the likelihood of outcomes desired by the
ruling parties. It is probable that some water institutions fit this
description relative to the integration of crop-water response infor-

The goals and objectives of the various special interest or
referent groups must also be considered. The preference functions of
the members in the referent groups determine which alternatives are
relevant and what information is necessary [13, p. 747]. If
agricultural interests (and others) are primarily concerned with the
distribution of the resource on some criterion of "equity" (or fair-
ness"), value in use information becomes less important.

As Wolman has stated, regarding the illumination of alternatives in
water planning, analysts may have to become inured "... to the shock of
having their inquiries ignored" [50, p. 789]. The researcher in crop-
water response studies may benefit from the same inurement process.

Institutional Structure. Water institutions that evolved under common
law doctrines, dependent heavily on custom and case law, may have no
place and use for crop-water response functions. This may be a severe
roadblock to such integration in those prior appropriation and
riparian states which use little administrative law. However, the
administrative system of water law (such as in Florida) is potentially
amenable to such integration. The Florida water districts have yet to
establish rigid water permit allocation rules and regulations, for
example. Again, however, inurementt of the analyst" may be appropriate
in many areas of the United States.

Training and Understanding. Many of the Florida water management
personnel were found to be interested in the concepts of economic
demand and economic supply, but also expressed difficulty in some cases
to effectively communicate this interest. This is not an indictment of
current personnel in the Florida water institutions (primarily
engineers, hydrologists, and planners), but rather a real challenge for

See for example, Steiner [44], for a discussion of the weighting
process in the multiple objective function.


the natural resource and/or production economist. The only long run
solution to this problem is appropriate training and educational
sessions, plus a greater role of individuals trained in the social and
economic sciences in the water institutions of the United States.

Summary and Conclusions

A function of the institutional framework within which individuals
make water-use decisions is to facilitate the allocation of water among
users and uses. When water is a truly scarce resource, the process and
the criteria by which limited supplies are apportioned among competing
uses are of critical importance.

In Florida, regional water management districts are charged with
responsibility for managing all of Florida's water in the public
interest, and are authorized to require permits for consumptive uses of
water, among other things. In the event that chronic water shortages
occur in some parts of Florida in the future, the value of additional
(marginal) increments of water in alternative uses may prove to be use-
ful information, both to assure equitable and efficient allocation of
limited water supplies, and for evaluation of public investment in
measures to augment supplies. The value of a marginal unit of water
for crop irrigation is a function of

1. soil-water-plant relationships underlying crop yield response
to irrigation, and

2. the relative prices of water and other inputs in the crop
production function, and the price per unit of harvested crop.

The price per unit of irrigation water is usually the marginal cost
of delivering water to the crop. Typically, irrigators must, in
effect, operate their own water supply service. Agriculturalists
seeking to maximize returns to crop production must make investment and
production decisions concerning their irrigation systems in light of
and in addition to, investment and production decisions on all other
aspects of their irrigated agricultural enterprise.

By influencing the cost of supplying irrigation water, or by
constraining rates of water use, water management districts affect the
irrigating firm through the firm's irrigation water supply function.
Should the water management district install well-head flow meters on
all irrigation pumps and impose a price on irrigation water, profit-
maximizing irrigators would adjust their rates of water use in a
manner which equates the firm's water supply function with its own
water demand function. On the other hand, other things being equal,
irrigators attempting to maximize net returns to crop production would
reduce their water application rates so that the return to the last
gallon of water applied would just equal the marginal cost of applying
it including the district's charge for water.

By rationing water for irrigation at a level lower than irrigators
currently find optimum, the district would cause a reduction in net


returns to irrigated crop production.6 Other things being equal, the
decline in net returns (after marginal water supply costs have been
subtracted) associated with each incremental reduction in water allot-
ment is a measure of the marginal value of water for irrigation a
figure appropriately comparable to the marginal value of water in other

The significance of information on the demand for crop irrigation
water and the ability to integrate such information into institutional
decision making depends on several considerations.

1. The ability to derive reliable estimates of the marginal value
of water for crop irrigation depends upon successful modeling
of the physical and biological relationships underlying crop
yield response to irrigation, and the economic relationships
characterizing the irrigator's economic supply and demand for
irrigation water. Such efforts require high levels of techni-
cal competence on the part of the investigators, and substan-
tial costs for data acquisition and analysis.

2. The marginal value of water in alternative uses is a
potentially important body of information, but will not be used
unless water management staff and decision makers understand the
concept and consciously adopt social efficiency of water use as
one criterion against which to evaluate applications for
consumptive use permits and for allocation during water short

3. Agriculturalists may fear that better information concerning
the marginal value of water for crop irrigation may hurt their
bid for consumptive use permits, if their use of water turns
out to be of lower value than some competing use. This fear is
probably unfounded in much of Florida. The goals and objec-
tives of the referent groups must be considered, however.

4. The institutional structure, often based in custom and case
law, may not be capable of using information concerning the
marginal value of water in alternative uses. Florida's system
of administrative water law is potentially capable of effec-
tively using such information in the context of both planning
and permitting.

The integration of crop-water response information into institu-
tional decision making appears desirable on social efficiency grounds.
It also is possible if the problems and misunderstandings are made
explicit in the integration process.

Assuming no irrigation water is being "wasted."



[1] Ahmed, J., C.H.M. van Bavel and E.A. Hiler. "Optimization of
Crop Irrigation Strategy Under a Stochastic Weather Regime:
A Simulation Study," Water Resources Research, Volume 12, No.
6, December 1976, pp. 1241-1247.

[2] Anderson, Raymond L., Dan Yaron and Robert Young. "Models
Designed to Efficiently Allocate Irrigation Water Use Based
on Crop Response to Soil Moisture Stress," Technical Report
No. 8, Environmental Resources Center, Colorado State
University, Ft. Collins, May 1977, 83 pp.

[3] Bauder, J.W., R.J. Hanks and D.W. James. "Crop Production Func-
tion Determinations as Influenced by Irrigation and Nitrogen
Fertilization Using a Continuous Variable Design," Soil
Science Society of America, Proceedings, Volume 39, 1975,
pp. 1187-1192.

[4] Beringer, Christoph. "An Economic Model for Determining the
Production Function for Water in Agriculture," Giannini
Foundation Research Report No. 240, University of California,
Berkeley, 1961, 20 pp.

[5] Bresler, E. "Soil Water Regime in Economic Evaluation of
Salinity in Irrigation," Water Resources Research, Volume 8,
No. 4, August 1972, pp. 791-800.

[6] Carriker, Roy R., and Gary D. Lynne. "The Florida Water
Resources Act of 1972: A Synopsis." IFAS Water Resources
Council Fact Sheet (in press), University of Florida,
Gainesville, Florida, 1978.

[7] Castle, Emery N. "Property Rights and the Political Economy of
Resource Scarcity," American Journal of Agricultural
Economics, Volume 60, February 1978, pp. 1-9.

[8] Central and Southern Florida Flood Control District. Annual
Report for 1974-1975 and 1975-1976. 32 pp.

[9] hangn, Jen-hu, R.B. Campbell, and F.E. Robinson. "On the
Relationship Between Water and Sugar Cane Yield in Hawaii,"
Agronomy Journal, Volume 55, 1963, pp. 450-452.

[10] Downey, Lloyd A. "Water Yield Relations for Non Forage Crops,"
Journal of the Irrigation and Drainage Division, Proceedings
of the American Society of Civil Engineers, March 1972, pp.

[I1] Ehlig, C.F. and R.D. LeMert. "Water Use and Productivity of
Wheat under Five Irrigation Treatments." Soil Science
Society of America Journal, Volume 40, 1976, pp. 750-755.


[12] Fischer, Ralph A. and Robert M. Hagan. "Plant Water Relations,
Irrigation Management and Crop Yield." Expl. Agriculture,
1965, pp. 161-177.

[13] Fox, Irving K. "Institutions for Water Management in a Changing
World." Natural Resources Journal. Volume 16, November
1976, pp. 743-754.

[14] Gisser, Micha and Abraham Mercado. "Integration of the Agricul-
tural Demand Function for Water and the Hydrologic Model of
the Pecos Basin." Water Resources Research, Volume 8, No. 6,
December 1972, pp. 1373-1384.

[15] Grimes, Donald, H. Yamada, and W.L. Dickens. "Functions for
Cotton (Gossypium hirsutum L.) Production from Irrigation and
Nitrogen Fertilization Variables: I. Yield and Evapotrans-
piration." Agronomy Journal, Volume 61, 1969, pp. 769-776.

[16] Hall, Warren A., M. ASCE, and William S. Butcher. "Optimal
Timing of Irrigation," Journal of the Irrigation and Drainage
Division, Proceedings of the ASCE, Volume 94, June 1968, pp.

[17] Hanks, R.J. "Model for Predicting Plant Yield as Influenced by
Water Use." Agronomy Journal, Volume 66, 1977, pp. 660-665.

[18] Hanks, R.J. H.R. Gardner, and R.L. Florian. "Plant Growth-
Evapotranspiration Relations for Several Crops in the
Central Great Plains." Agronomy Journal, Volume 61, 1969,
pp. 30-34.

[19] Hexem, Roger W., V.A. Sposito, and Earl O. Heady. "Application
of a Two-Variable Mitscherlich Function in the Analysis of
Yield-Water-Fertilizer Relationships for Corn." Water
Resources Research, Volume 12, No. 1, February 1976, pp. 6-

[203 Hiler, E.A., and R.N. Clark. "Stress Day Index to Characterize
Effects of Water Stress on Crop Yields." Am. Soc. Ag. Eng.,
Volume 14, 1971, pp. 751-761.

[21] Hogg, Howard C., Jack R. Davidson, M. ASCE, and Jen-hu Chang..
"Economics of a Water-Yield Function for Sugar Cane." Journal
of the Irrigation and Drainage Division, Proceedings of the
ASCE, Volume 95, March 1969, pp. 127-138.

[22] Hogg, Howard C., Lloyd B. Rankine, Jack R. Davidson. "Estimating
the Productivity of Irrigation Water." Agricultural
Economics Research, Volume 22, No. 1, January 1972, pp. 12-

[23] Hogg, Howard C. and Gary R. Vieth. "Method for Evaluating
Irrigation Projections." Journal of the Irrigation and
Drainage Division, Proceedings of the ASCE, Volume 103,


March 1977, pp. 43-52.

[24] Howell, T.A. and E.A. Hiler. "Optimization of Water Use
Efficiency Under High Frequency Irrigation I. Evapotrans-
piration and Yield Relationship." Transactions of the ASAE,
1975, pp. 873-878.

[25] Kissel, D.E., J.T. Ritchie and C.W. Richardson. "A Stress Day
Concept to Improve Nitrogen Fertilizer Utilization: Dryland
Grain Sorghum in the Texas Blackland Prairie." MP-1201,
The Texas Agricultural Experiment Station, College Station
Texas, June 1975, 15 pp.

[26] Kloster, Larry D. and Norman K. Whittlesey. "Production Function
Analysis of Irrigation Water and Nitrogen Fertilizer in Wheat
Production." Washington Ag. Exp. Station Bulletin 746, 1971,
9 pp.

[27] Lomas, J. and Y. Shashoua. "The Effect of Rainfall on Wheat
Yields in an Arid Region." In Slatyer, R.O.(ed.) Plant
Response to Climatic Factors. Proceedings Uppsala Synposium,
UNESCO, Paris, 1973, pp. 531-537.

[28] Maloney, F.E., R.C. Ausness, and J.S. Morris. "A Model Water
Code." University of Florida Press, Gainesville, 1972.

[29] Mapp, Harry P. and Vernon R. Eidman. "An Economic Analysis of
Regulating Water Use in the Central Ogllala Formation."
Agriculture Experiment Station Technical Bulletin T-141.
Oklahoma State University, Stillwater, January 1976, 99 pp.

[30] Miller, D.E. and J.S. Aarstad. "Yields and Sugar Content of
Sugarbeets as Affected by Deficit High-Frequency Irrigation."
Agronomy Journal, Volume 68, March-April 1976, pp. 231-234.

[31] Milliman, J.E. "Water Supply Planning and the Long-Term Outlook
for Southeast Florida." The Florida Outlook, Volume 1, No.
2, University of Florida, Gainesville, 1977.

[32] Minhas, B.S., K.S. Pakikh, and T.N. Srinivasan. "Toward the
Structure of a Production Function for Wheat Yields with
Dated Inputs of Irrigation Water." Water Resources Research,
Volume 10, No. 1, June 1974, pp. 383-393.

[33] Moore, Charles V. "A General Analytical Framework for Estimating
the Production Function for Crops Using Irrigation Water."
J. Farm Economics, Volume 61, 1961, pp. 876-891.

[34] Nairizi, S. and J.R. Rydzewski. "Effects of Dated Soil Moisture
Stress on Crop Yields." Exp. Agric., 1977, pp. 51-59.

[35] Neghassi, Habte Mariam, Dale F, Heermann, and Darrly E. Smika.
"Wheat Yield Models with Limited Soil Water." Transactions
of the ASAE, Volume 18, 1975, pp. 549-557.


[36] Parvin, D.W., Jr. "Estimation of Irrigation Response from Time-
Series Data on Nonirrigated Crops." AJAE, Volume 55, 1973,
pp. 73-76.

[37] Quisenberry, J.E. and Bruce Roork. "Influence of Indeterminate
Growth Habit on Yield and Irrigation Water-Use Efficiency in
Upland Cotton." Crop Science, Volume 16, pp. 762-765.

[38] Resources for the Future. Resources (newsletter). No. 55,
April-June 1977, p. 1.

[39] Ritchie, J.T. "Atmospheric and Soil Water Influences on the
Plant-Water Balance." Agric. Meteorol. Volume 14, pp. 183-198.

[40] Schneider, A.D., J.T. Musick and D.A. Dusek. "Efficient Wheat
Irrigation with Limited Water." Transactions of the ASAE,
Volume 14, 1969, pp. 23-26.

[41] Southwest Florida Water Management District. "Hydroscope."
Volume 6, No. 5, Brooksville, Florida, May 1975.

[42] Splinter, W.E. "Modeling of Plant Growth for Yield Prediction."
Agric. Meteorol., Volume 14, 1974, pp. 243-253.

[43] Stegman, E.C., L.H. Schiele, A. Bauer. "Plant Water Stress
Criteria for Irrigation Scheduling." Transactions of the Am.
Soc. Ag. Eng. Volume 19, 1976, pp. 850-855.

[44] Steiner, Peter O. "The Public Sector and the Public Interest" in
The Analysis and Evaluation of Public Expenditures: The PPB
System. Washington: U.S. Government Printing Office, 1969,
pp. 13-46.

[45] Stewart, J. Ian and Robert M. Hagan. "Functions to Predict of
Crop Water Deficits." Journal of the Irrigation and Drainage
Division of ASCE, Volume 99, December 1973, pp. 421-439.

[46] Stewart, J. Ian, Robert M. Hagan and William 0. Pruitt. "Func-
tions to Predict Optimal Irrigation Programs." Journal of
the Irrigation and Drainage Division of the ASCE, Volume 100,
June 1974, pp. 179-199.

[47] Stewart, J.I., R.D. Misra, W.O. Pruitt, R.M. Hagan. "Irrigation
Corn and Grain Sorghum with a Deficient Water Supply."
Transactions of the ASAE, Volume 18, 1975, pp. 270-280.

[48] VanDeman, John M., Robert Sowell, and Ronald E. Sneed.
"Optimilization of Water Use for Irrigation." Winter meeting
of American Society of Agricultural Engineers, 1976, pp. 1-20.

[49] Williams, D.W. and W.J. Chancellor. "Irrigated Agricultural
Production Response to Constraints in Energy-Related Inputs."
Transactions of the ASAE, Volume 18, 1975.


[50] Wolman, M. Gordon. "SelecTing Alternatives in Water Resources
Planning and the Politics of Agendas." Natural Resources
Journal, Volume 16, October 1976, pp. 773-789.

[51] Yaron, Dan. "Estimation and Use of Water Production Functions in
Crops." Journal of the Irrigation and Drainage Division of
the ASCE, Volume 97, June 1971, pp. 291-303.

[52] Yaron, D., H. gielorai, J. Shalhevet, and Y. Gavish. "Estimation
Procedures for Response Functions of Crops to Soil Water
Content and Salinity." Water Resources Research, Volume 8,
No. 2, April 1972, pp. 291-300.


James F. Dunne and David J. Allee*

Public cost sharing for water resources projects is an institu-
tional device necessitated by various types of market failure, the
most important of which involve public goods, common property resources
and externalities. Public goods are said to exist when, once pro-
vided, the cost incurred per additional user is small or non-existent,
or when it is impractical to exclude users who don't pay a fee. An ex-
ample might be flood protection. A resource is said to have common
property characteristics when, if competitively extracted, over-use or
depletion occurs because each competitor has an incentive to extract as
much as possible. Extraction of groundwater for irrigation provides a
good example. Externalities occur when production of one good has ei-
ther beneficial or adverse effects on production of others, but the ef-
fects do not pass through the market and therefore are not considered
explicitly in production decisions. An example is the factory which
dumps sewage into a river, adversely affecting fishing and other rec-

In these special cases competitive markets fail to achieve a so-
cially optimum result; goods are not produced in optimum quantities and
resources are not optimally allocated to different uses. As a result,
some modification of the market mechanism is necessary for efficiency.
While this point is usually stated with the public-private dichotomy in
mind -- it applies almost as well between levels of government. Like-
wise, cost sharing is sometimes thought of as restricted to a partial
payment of out-of-pocket costs, but the principles involved apply to
any change in incentive brought about by a relevant unit of government.

The problems associated with common property resources and extern-
alities could be solved without public action, at least in theory, if
a detailed assignment of property rights, including effective markets
for those rights, could be brought about. For resources such as water,
however, this is a formidable task; for others, such as air, it is
surely impossible. In 1960 Coase pointed out the natural tendency of
market forces to resolve the problem t ough negotiation and suitable
payments between the involved parties.- Transaction costs, including
imperfect information, however, limit the scope for negotiation when
large numbers of users are involved. Public action, in such an in-
stance, must be instituted to bring about socially optimal results.

* Graduate Research Assistant and Professor, Department of Agricul-
tural Economics, Cornell University. Many persons made helpful com-
ments on an earlier draft of this paper; we wish to especially
thank R. N. Boisvert, D. U. Fisher, and R. A. Milligan.


The appropriate public action from the economist's point of view
is the imposition of a tax on the resource user or on the source of the
externality which amounts to the difference between social and private
marginal cost. Such a tax leaves the choice to the decision maker who
is likely to take the most efficient action. Because it is often diff-
icult or impossible in practice to specify the optimal level of such a
tax, or to devise necessary institutional arrangements for implementing
it, second-best solutions in the form of standards or incentives have
been used to achieve similar if less perfect results. Air pollution
standards have been get which force industries and private motorists to
bear the direct costs of abatement. Limits in the form of net mesh
sizes and special seasons have been imposed in connection with fishery
resources. For problems of soil erosion, cost-sharing funds are pro-
vided to encourage individual landowners to undertake soil conservation
improvements which also benefit society as a whole.

For soil conservation practices in New York State and elsewhere,
varying levels of cost sharing are provided to landowners through the
Agricultural Stabilization and Conservation Service (ASCS) of the
United States Department of Agriculture (USDA). This agency operates
according to conservation priorities broadly established at the na-
tional level and supplemented at the state and county levels through
committee action. Until 1978, ASCS cost-sharing policy has been ori-
ented toward "soil conservation practices", and state and county com-
mittees have selected practices and cost-sharing levels to reflect lo-
cal conservation needs which were defined to include such goals as in-
creasing agricultural output, prevention of soil loss, and regional

This setting of priorities at the local level, coupled with the
focus on "practices", has partially diverted the program away from
problems of soil loss and pollution. In many areas, such as northern
New York, drainage which has a negligible soil conservation component
is undertaken with public cost-sharing funds. Since local committees
influence the setting of priorities for funds, the outcome reflects lo-
cal concern with improving output and incomes for the agricultural sec-
tor as a means of increasing the prosperity of the local region. As a
result, environmental protection and regional agricultural development
have, in many cases, become competing uses for scarce cost-sharing

Recognizing these problems with the Agricultural Conservation Pro-
gram's (ACP) operation, the USDA in late 1977 issued notices to state
and county committees informing them of a "new ACP development proced-
ure which is needed to redirect ACP from q practice oriented program
to one of conservation-problem solving"X27 This action followed con-
tinued criticism of ACP in recent years, including Presidential concern
over the cost-effectiveness of the program in solving "continuing con-
servation problems".

For certain areas of the country, this shift in priority will have
significant regional development impacts. In Clinton County, New York,
as in many parts of the nation, tile drainage of agricultural land has
in the past been supported by ACP cost-sharing funds. Since this prac-
tice does not rank high on the list of permanent solutions to soil loss


problems and given a continuation of apparent trends in policy emphasis,
less financial support may be expected in the future.

In 1977, for instance, Clinton County allocated 51 percent of its
ACP cost-sharing funds to tile and open ditch drainage, but this figure
was reduced to 40 percent for the 1978 program year to reflect national
priorities. If farmers depend on cost-sharing funds to drain flat-
lying land and increase output and income, as they do in northern New
York, the new policy emphasis will affect them adversely.

Looking back on several years of high agricultural output, one
might conclude that the shift in national priorities for ACP cost-
sharing funds is a reflection of the successful solution of the "pro-
duction problem" in agriculture. Since surpluses are now more problem-
atic than shortages, the wisdom of encouraging greater output might be
questioned. However, regional impacts cannot be ignored, as many areas
where incomes are lowest are heavily dependent on agriculture and fre-
quently have much land with potential for improvement. Such areas
would bear a dispropor donate share of the burden of non-availability
of cost-sharing funds.-'

To demonstrate the regional importance of agricultural cost-
sharing funds, the study on which this paper is based involved a com-
prehensive analysis of the economic benefits which accrue to the re-
gion. Identification of more than the direct costs and benefits is
needed because, through its multiplier effects, the dollar invested can
become more than a dollar of income through successive responding
rounds. Once the magnitude of such benefits is known, investment in
agriculture can be compared with similar growth stimulating investments
in other sectors for its relative income-producing capability. Further-
more, those sectors which are indirect beneficiaries of the investment
may be identified. The distribution of cost shares may then be reviewed
in light of the distribution of benefits.

In the remainder of this paper we focus on the estimation of in-
come benefits from a farm drainage project and the implications of
these benefits for cost-sharing arrangements. The next section briefly
discusses some of the problems in measuring benefits and presents a
methodology which is capable of estimating both direct and indirect
benefits. This technique is then applied to a current drainage project
in northern New York. Finally, the relevance of the results for public
cost-sharing policy is discussed.

Evaluation of Benefits

The evaluation of benefits of water resources projects has been
the subject of much discussion. The existence of intangible or non-
monetary benefits has placed obstacles in the path of traditional cost-
benefit analyses and has caused such analyses to be subject to much
criticism. Even monetary benefits, which are more easily measurable
(and it is this type with which we concern ourselves here), have not
failed to provoke considerable controversy.


A distinction is made in the literature between "real" benefits,
or those which represent a net increase in the available amount of some
good, from a societal point of view, and "pecuniary" benefits, or those
which represent merely a change in relative prices or a locational
shift in economic activity. When a project is viewed from a national
standpoint, the inappropriateness of including pecuniary benefits is
clear, for local increases or decreases serve only to cancel opposite
effects elsewhere. However, if the perspective is local or regional,
they may be of major importance, as they might similarly be if the na-
tional outlook included development priority for the region in ques-

A further distinction is made between "direct" or "primary" bene-
fits and "indirect" or "secondary" benefits. Direct benefits are those
net increases in the available quantity of a certain good which the
project was specifically intended to produce, whereas indirect benefits
are complementary net increases in the availability of other goods,
which arise as a result of the project. Other types of indirect bene-
fits which are sometimes pointed out are increases in income which
arise in conjunction with the direct benefits of the project or simply
changes in income distribution if this is thought desirable.

The major criticism of the inclusion of indirect benefits in pro-
ject evaluation has been that this process can involve double counting;
in the case of a drainage project for instance, where the value of in-
creased crop yield is regarded as the direct benefit, the income in-
creases in other sectors which result from higher output made necessary
by production increases in the farming sector may not be added to this
direct benefit. To do so would be similar to counting both the value
of the additional crop produced and the value of the animal to which
it was fed.

Several writers, however, while criticizing the inclusion of indi-
rect benefits in project evaluation, point out the relevance of these 4
benefits when assessing the regional development impacts of a project.-
Ciriacy-Wantrup distinguishes between the separate processes of select-
ing thq best project and repayment by the beneficiaries to the govern-
ment.5/ For the sake of efficient resource allocation, he indicates,
benefits which are essentially transfers between regions, individuals,
or groups should not be included, even though such transfers may be a
desirable policy objective. However, when seeking an appropriate form-
ula for repayment purposes, such transfers are of primary importance,
because allocational efficiency of resources is no longer the relevant
criterion. Instead, the criterion is now clearly distributional and
political in nature; the distribution of benefits should be relevant to
the distribution of repayment costs.

Once the relevance of indirect benefits for regional development
and formulation of cost-sharing policies is recognized, they may be es-
timated through use of techniques which have been refined over the past
twenty years. The most important and useful of these techniques is
interindustry analysis, often referred to as input-output analysis, de-
veloped from Leontief's pioneering work on trade problems in the
1930' s._


The heart of the interindustry model is the "transactions" matrix
which displays the sales and purchases which take place between sectors
in a given year. Different economic sectors may be relegated to the
endogenous part of the model -- their output levels are determined by
levels of production and demand elsewhere in the model -- or they can
be relegated to the exogenously determined final demand categories
which may include such sectoral classifications as government, house-
holds, and exports. From the transactions matrix, a "technical coeffi-
cients" matrix is constructed which gives the intersectoral flows per
unit of gross output of the purchasing sector, and is interpreted as
showing the direct needs of a given purchasing sector from all other
sectors in producing a dollar's worth of its output. In order to get
the total (direct and indirect) effect, the technical coefficient ma-
trix may be subtracted from an identity matrix of the same dimensions
and the result inverted to yield an "interdependence matrix" whose ele-
ments are now interpreted as being the direct and indirect requirements
of the purchasing sector to produce a dollar's worth of output for
final demand sectors.

A useful variation of the interindustry model, generally known as
the closed system, classifies the household sector endogenously. By
use of this technique, changes in sectoral output made necessary by
changes in consumer demand, known as "induced" effects, may be identi-
fied and added to those changes brought about by production interdepen-
dence. In this case, the interdependence matrix gives the direct, in-
direct, and induced changes in sectoral output made necessary by deliv-
ery of one dollar of the purchasing sector's output to final demand.

Two types of income multipliers may be computed from the interde-
pendence tables. They differ only insofar as the household sector is
endogenous or exogenous to the model. They are interpreted as the to-
tal increase in income per dollar of direct income increase in the sec-
tor for which the change in shipments to final demand takes place. The
general concept is similar to that of the Keynesian national income
multiplier; an invested dollar moves through successive responding
rounds of continually decreasing magnitude.

Even though several limiting assumptions such as fixed coefficient-
constant returns to scale production functions are inherent in the
interindustry model, its usefulness has long been recognized. Although
data requirements are extensive, a number of models have been construc-
ted for smaller areas suc as counties directly from primary data ob-
tained through sampling.- For purposes of this study, a twenty-five
sector model for Clinton County, New York, developed by survey methods,
is used to evaluate the local impact of agricultural drainage in one
of the county's watersheds../

Benefits From Farm Drainage -- Corbeau Creek Watershed

Clinton County is the northern-most county in New York State and
is situated on the shore of Lake Champlain. Corbeau Creek Watershed
is in the eastern part of the county, near the lake shore. Incomes in
the watershed are low; of all families living in its four townships,
6.6 percent were in poverty status in 1970, and this figure was as high


as 15 percent in one township. Median family income in the same year
was $7,978, and the county unemployment rate varied between extremes
of 10.3 percent and 12.1 percent during the March '73 January '77 pe-
riod. Approximately four percent of all employed persons in the water-
shed are employed in agriculture.

Part of this watershed -- "Corbeau Flts" -- is at an elevation of
220 feet m.s.l. and demonstrates little or no gradient. Agricultural
yields in the flats are low due to poor drainage conditions; corn sil-
age yields on its 17 active dairy farms average 10.5 tons/acre and hay
yields average 1.5 tons/acre. These figures are well below county ave-
rages of 15 tons/acre and 2.2 tons/acre for corn silage and hay, re-

The differences in yields are in part due to a lack of suitable
outlets which has traditionally hampered drainage efforts, and farmers
have adapted to physical conditions by practicing extensive agriculture
and following restrictive crop rotations. Technical solutions to the
drainage problem are clearly available, but since all farms do not
border the creek, outlets must cross property lines. The success or
failure of a drainage effort, therefore, depends on cooperation of all
landowners, most particularly those whose farms are nearer to the creek.

Proposed drainage in Corbeau Flats involves some 2,292 acres of
cropland. Soil Conservation Service (SCS) and Cornell farm management
experts' projections indicate tt cropping patterns and yields will
change as outlined in Table 1.-n'

Table 1. Changes in Yield and Cropping Patterns

Before Drainage After Drainage

Acres Yield/Acre Acres Yield/Acre

Hay (timothy) 843 1.5 tons 567 2.5 tons
Hay (alfalfa) -- 566 3.0 tons
Corn silage 1,064 10.5 tons 1,159 16 tons
Pasture 1,019 1.0 tons 634 1.0 tons

Using Cornell Dairy Farm Enterprise Budgets, production costs other 1/
than labor were estimated for the pre- and post-drainage situations.-
Along with other increases in costs for higher yields already assumed
in the budgets, an increase in payments to land, property taxes, depre-
ciation, and maintenance of $35 per acre per year is included to cover
drainage costs.-2 Improved drainage results in an estimated increase
in forage crop output of $222,976 annually (valued at market prices)
after all cost increases including yearly per acre cost of on-farm
drainage, were deducted.-- As a result of this increase in crop out-
put, direct labor and management income to farmers in the Corbeau Flats
area would rise by an estimated $79,554 annually.

Applying the income multiplier (see Table 2) for the dairy farming
sector to this increase in direct income payments, a direct plus indi-
rect plus induced income increase of $153,937 per year within the


Table 2. Income Multipliers (Dollars)

Type II
Sector Income Multiplier*

Dairy farms 1.9349
Other farms 1.6676
Farm supply 1.5424
Dairy processing 8.6043
Paper manufacturing 1.3945
Construction 1.5617
Chemical, metal & plastic mfg. 1.3476
Other manufacturing 1.6118
Petroleum sales 1.6691
Financial 1.3808
Automotive 1.6154
Food wholesale & retail 2.4528
Recreation 1.6024
Trade 1.6269
Professional services 1.3196
Other services 1.4635
Mobile home 1.5990
Transportation & utilities 1.4491
Education 1.3093
Local government 1.4564

Source: "An Interindustry Analysis of Clinton County, New
York", Michel Lee Hiser, unpublished Ph.D. thesis, Cornell
University, 1977.

Households endogenous.

county is seen to result from the drainage improvement. As evident
from the size of the multiplier (1.94), the total indirect and induced
income effects are nearly equal to the direct effects -- a result which
is due to the wide usage of non-labor inputs in dairy farming and ex-
tensive sectoral ties with other sectors, particularly those which are
labor-intensive. Those sectors where non-direct income benefits exceed
$5,000 are given in Table 3.

Table 3. Sectors in Which Indirect Plus Induced
Income Benefits Exceeded $5,000

Sector Income Benefits

Farm supply $14,316
Professional services 11,494
Transport & utilities 7,750
Finance 7,679
Education 7,550
Local government 6,453


In order to view this figure ($153,937) as the total effect on
county income of the investment in drainage, it must be assumed that
total increased output is exported or otherwise purchased by final de-
mand sectors. As indicated in the Clinton County interindustry model,
81 percent of dairy farm output is already being exported from the
county for processing elsewhere. Thus, this assumption is not too un-
realistic, but the increase should be viewed clearly as an upper bound
which would obtain in the event of no idle processing capacity in the
county. In the long run, this figure is probably somewhat high, as
capacity adjustments may indeed take place, and more dairy sector out-
put may be processed locally.

While we do not measure them here, in many instances it may be
useful to determine the level of benefits occurring during the con-
struction phase of the project. These benefits would be especially
high if construction services and materials were largely supplied by
the local economy rather than imported from outside the county. Past
evidence shows that some of the more vocal proponents of projects have
been those who reap benefits primarily during the construction phase,
and while such benefits would certainly not be as large for a farm
drainage project as for dan construction, they will be significant to
some sectors of the local economy.V/

Benefits and Cost Sharing

What are the implications of these figures for cost sharing?
First, the existence of very significant local benefits which do not
simply accrue to the direct beneficiaries has been established. A
strategy for extracting a contribution toward the cost of the drainage
improvement in proportion to the income benefits received would there-
fore seem to be in order. Assessment for such contributions is no
simple task, for it depends on such factors as sector definition and
aggregation. Because firms produce and sell a variety of products,
there is no single agreed-upon method of sector delineation. Further-
more, resistance may be encountered in the political sphere, as indi-
rect benefits are not as easily observed as direct benefits, and the
levying of special taxes at the state or local level may not be prac-

Cost-sharing grants equal to all or part of the increase in state
and local tax revenues due to the drainage improvement may provide one
solution to this dilemma. With an average state tax rate of about six
percent on income, $9,236 per year would accrue to the state treasury
as a result of the drainage improvement, if it is assumed that all ex-
tra output goes to final demand.15/ If the state were to make avail-
able to farmers, in the early stages of the drainage project, a sum
equal to the present value of the increased income taxes due to the
project over its useful life, this would represent a significant con-
tribution when compared with the federal share. For a useful life of
20 years and a five percent discount rate, for instance, this would
amount to $115,101.

Other tax revenue also increases following the drainage improve-
ment. Approximately 1.8 percent of personal income is currently spent


on state sales tax in New York.1/ Following the income increase
brought about by drainage in Corbeau Creek Watershed, the state share
of sales tax collections should increase by about $2,770 per year.
Similarly, if approximately $2 per drained acre in increased local
property taxes is paid by farmers, $4,584 of additional revenue per
year will accrue to local government as well as over $1,000 in sales
tax revenue returned by the state. As in the case of income taxes,
the state or local government would be "no worse off" if it offered
the present value of increased tax revenue over the project life as a
non-federal cost share. If a state and county were both to offer cost
shares on the same project, they might find it useful to isolate in-
come benefits which accrued within the state but outside the county.
This could be accomplished through use of a state interindustry model
in conjunction with the county model to capture interregional transfer

A state or local cost share in proportion to indirect benefits be-
comes even more important when seen from the perspective of uncertainty
of continued federal support at previous cost-sharing levels which
changing federal conservation priorities make necessary. Even at pres-
ent levels, the federal share is a major limitation for projects of
this type which involve drainage of whole farms or major portions of
them; because of the $10,000 limitation per farm, it is more suited to
small scale projects where 50 percent of the cost (in the case of
Clinton County) is provided if the limit is.not exceeded.

How does estimation of total benefits and investigation of cost-
sharing alternatives relate to the problem of drainage outlet provision?
Although SCS technicians have not yet determined outlet requirements,
it is not expected that outlet costs will be large when compared with
on-farm drainage costs. It is clear that where outlets are necessary,
cooperative arrangements between farmers must evolve because outlets
are essentially joint ventures with public good characteristics. Farm-
ers have an economic incentive not to share in the cost of an outlet,
but to use it after it has been installed through the efforts of others.
Also, the cost to participating farmers rises as the number of farmer
"hold-outs" increases. If total income benefits are high enough, it
may be economical to pay the cost share for outlet construction, of a
farmer who is not interested in drainage, by use of public funds. In
cases where such a farmer changed his mind at a later date and decided
to install on-farm drainage, he could at that time be assessed for out-
let costs. In this way, drainage projects with a small number of hold-
outs would not be held up permanently. Action of this type is facili-
tated in New York by the existence of legal provisions for the granting
of easements so that necessary outlets might be constructed through
land owned by unwilling proprietors. However, institutional arrange-
ments to finance voluntary hold-outs have not been developed and there
is reluctance to use involuntary means to take easements, much less to
force payment. Any justification for the use of local general revenue
sources is therefore of interest.


Summary and Conclusions

Shifting federal priorities make the future availability of cost-
sharing funds for farm drainage uncertain. Since drainage has public
good qualities, a less than optimal amount will be carried out in the
absence of public incentives. The regional development impact of farm
drainage on local economies can be effectively measured through inter-
industry techniques. Once this impact is known, it can provide a basis
for discussion of appropriate cost-sharing levels, the effects of
lesser federal funding, and the likely results of increased state and
local participation.


1/ R. H. Coase, "The Problem of Social Cost". Journal of Law and
Economics, Vol. 3 (October 1960).

2/ U. S. Department of Agriculture, Agricultural Stabilization and
Conservation Service, Wire Notice RE-183 to State and County Of-
fices (December 1, 1977).

2/ It is beyond the scope of this paper to attempt a judgment of whe-
ther incremental social returns would be greater in expending
Clinton County's share of the funds on environmental enhancement
practices instead of those which emphasize development. The tech-
niques discussed provide a means of estimating part of the oppor-
tunity foregone in choosing one over the other.

4/ See criticism of practices of the Bureau of Reclamation in: R.
McKean, Efficiency in Government Through Systems Analysis (1958),
Chapter 9, and Otto Eckstein, Water Resource Development (1961),
Chapter 7.

5/ S. V. Ciriacy-Wantrup, "Benefit-Cost Analysis and Public Resource
Development". Journal of Farm Economics, Vol. XXXVII, No. 4 (Nov-
ember 1955).

6/ For a good introduction to interindustry analysis, see: Chiou-
Shuang Yan, Introduction to Input-Output Economics (Holt, Rinehart
and Winston, 1969).

7/ More recently, techniques have been developed at Cornell and else-
where for the construction of models for smaller areas from second-
ary data. See, for instance, R. N. Boisvert and N. L. Bills, A
Non-Survey Technique for Regional 1-0 Models. Agricultural Econ-
omics Research 76-19, Cornell University, Ithaca, New York (August

8/ M. L. Hiser, "An Interindustry Analysis of Clinton County, New
York". Unpublished Ph.D. thesis, Cornell University, 1977.

9/ Corbeau Creek RC&D Measure, Environmental Assessment of Resources
and Related Problems, USDA/SCS, Cornell University, Miner Institute
(October 1977).


10/ Professors Robert A. Milligan and Fred Swader of Cornell University
projected post-drainage yields and provided other valuable assist-

11/ Wayne A. Knoblauch, Robert A. Milligan and Merri L. Woodell, An
Economic Analysis of New York Dairy Farm Enterprises. Agricultural
Economics Research 78-1, Cornell University, Ithaca, N. Y. (Jan.

12/ For a per acre on-farm drainage cost of $300, yearly depreciation
and interest were projected at 10 percent or $30, maintenance at
one percent or $3, and increased property taxes at .66 percent or
$2. Total non-labor costs were subtracted from total crop output
valued at market prices in the pre- and post-drainage situations.
The results were compared to arrive at a net change in labor and
management income attributable to drainage. As is usually the case
some discrepancy exists between this direct income figure and the
one which could be calculated from the interindustry tables. The
budget-based income change is used here, however, because it is
thought to be more accurate.

13/ Cornell farm management experts argue that output and income would
probably increase even more than this. The dairy cow enterprise
provides further value added to the crops, and land expansion or
switching to higher return crops such as corn grain would be feas-

14/ David J. Allee, "Government Facilitation of Irrigation in New York".
Unpublished Ph.D. thesis, Cornell University, 1961.

15/ Estimated with data from: New York Department of Taxation and Fin-
ance, Annual Report 1976-77, p. 7.

16/ Estimated with data from: New York Department of Taxation and Fin-
ance, Annual Report 1976-77, p. 7 and Office of the Governor, State
of New York Executive Budget, Albany, N. Y. (Jan. 1976), p. a5.


By Richard Greenhalgh and Fred Stewart*

Economists, engineers, and other social scientists continue the
search for more refined methods to make resource allocation decisions.
The thrust has been toward a comprehensive objective function which
quantifies an optimum allocation of resources for maximizing social
welfare. A completely satisfactory criterion has not been found, and
the economic profession is in general agreement that one probably does
not exist. Substantial strides have been made, however, toward devel-
oping and improving the resource allocation tools we use.

Economists feel that when market prices reflect the value of goods
and services they provide a basic data set for making decisions in our
decentralized society. Economists' infatuation with market prices stems
from the.fact that under assumptions of a purely competitive market,
such prices used as value signals lead to maximum aggregate social wel-
fare given the existing distribution of wealth. But some goods and
resources are not traded in the market place, while others are traded in
imperfect markets resulting in prices that are deemed inappropriate for
measuring social costs and benefits. Some economists argue for adjusting
prices from imperfect markets and imputing prices when none exist.
These new prices are referred to as "shadow prices". Economic theorists
are quick to point out problems with adjusted or imputed prices and
caution that they represent mere approximations of value per unit.

As professional economists, or engineers, who provide planning
assistance, we need to proceed in the best manner available. Sometimes
this is not completely consistent with our theoretical training, and it
is appropriate that we review some of the problems with imputed, ad-
justed, or even observed prices. The first section of the paper is
devoted to the development of shadow prices; the second considers prob-
lems associated with using shadow and market prices for decisions con-
cerning resource development.

*Richard Greenhalgh and Fred Stewart are agricultural economists
with the Natural Resource Economics Division, Economics, Statistics, and
Cooperatives Service, U.S. Department of Agriculture, stationed at the
University of Florida, Gainesville, Florida. This paper represents the
authors' personal views and are not necessarily those of ESCS.


Developing Shadow Prices

Broadly speaking a shadow price is "the price economists attribute
to a good or factor on the argument that it is more appropriate for the
purpose of economic calculation than its existing price" if one exists

Mishan discusses in detail the use of shadow prices for valuing
imports in countries where domestic prices do not reflect their relative
scarcity (7). Shadow prices and accounting prices are equated and a
wide variety of logical price adjustments of a general nature are dis-
cussed. Although the application does not relate directly to natural
resource development, the basic analogies can be made to natural resource
development issues. Governmental influence on natural resource prices,
agricultural product price supports, and low opportunity costs for labor
are examples of corollary issues where the application of shadow prices
would be informative.

Current Normalized Prices

In the natural resources field, current normalized agricultural
prices are provided by the Water Resource Council for use in all economic
evaluations of agricultural development (1). These prices are updated
annually to correct for short term fluctuations. They reflect a weighted
average of seasonal prices observed in the preceding five years. They
are shadow prices because they reflect a modification of daily prices
and are an improved basis for evaluating resource development. They are
not necessarily the best shadow prices to use in reflecting relative
values which lead to optimum social welfare. In fact, a strong case can
be made that current normalized prices are inflated, at least in some
years, through government price support programs. The difference
between actual price and what market price would have been without the
support price, in years it is effective, represents an income subsidy
to producers. To the extent that this exists, use of current normalized
agricultural prices builds in a liability for future income maintenance
in recommended resource development proposals. The social implications
may not be significant, but a procedure could be devised to estimate
market-clearing prices without the price support program. Normalized
prices based upon such an adjustment would reflect a different set of
shadow prices.

Target Prices

Target prices are becoming an important tool in farm policy. Costs
of production are a very important information base for guiding policy
makers in developing target prices. In the process of developing target
prices for products, it is necessary to develop shadow prices for some
unpriced inputs, such as operator labor and management services.
Further, priced inputs are not the same for all producers, and some
representative price has to be selected.

A very interesting issue exists concerning the appropriate price to
use for land in this type of budget analysis (5). The use of market
price results in a cyclical upward bias for future product prices and
some alternate, social optimal shadow price may be desirable.


Programming Shadow Prices

A more specific interpretation is held in mathematical programming
literature where shadow prices refer to values imputed to resources when
optimizing techniques are used (4). An objective function is maximized
or minimized subject to certain quantities of inputs or restricted
factor combinations. Specific prices are imputed at the solution level
of resource use which reflect the marginal value of inputs and output.

Individual farm operations can be optimized with programming models
and the resulting shadow price for limiting resources indicates what
operators can pay for additional units of resource over some finite
range. Similar analyses have been done on a national and regional scale
to evaluate present and future production potential. Shadow prices from
these models provide information useful in examining resource develop-
ment needs.

Shadow prices for agricultural products and land resources devel-
oped using linear programming in a regional study of Northwest Florida
by Greenhalgh and Stewart is an example (3). The land resource is
defined in the study as a package of several joint characteristics,
reflecting both quantitative and qualitative features. Conventional
interpretation of the linear programming output would reflect the mar-
ginal value of each individual characteristic. Joint characteristics
can not be provided individually, however, and a summation of these
values is not appropriate. A weighted sum of the shadow prices is
presented as the appropriate marginal value for the package of charac-
teristics (8).

The interpretation of programming shadow prices for land in the
analysis parallels that of the individual farm model. It is the annual
payment which could be paid to add an additional unit of land of spe-
cific quality. In resource development, new lands are generally not
added to the land base as individual farm operators might do, but
instead land resources are upgraded from a quality standpoint. To
determine the value of resource development, the shadow price for the
developed land must be decreased by the opportunity cost of comparable
undeveloped land in its present use. For example, well drained land
may have a shadow price of $85.00 whereas the same soil, poorly drained,
and in a less intensive use may have a shadow price of only $25.00.
The annual payment to compare with the cost for resource development
would be $60.00 rather than the full $85.00.

This type of analysis is not without inherent problems. Regional
and national models treat the entire area as a single large farm and
aggregation bias is possible (2). Further, programming shadow prices
reflect an undervaluation compared to actual resource and product prices.
The shadow prices for land do not include any nonagricultural potential
and fail to incorporate any expected appreciation perspective land
owners may hold. The product prices are undervalued at least partially
due to the undervaluation of the land input. Trends in shadow prices,
however, can be very helpful as relative values with directional impli-
cations regarding future prices.


Future Programming Shadow Prices

Agricultural resource developments are investments in capital stock
that are useful for many years, and future price ratios over the life of
the project are needed to analyze economic feasibility. Current normal-
ized prices are presently used as the best indicator available for
future prices and are considered sufficient estimates of future prices
in many cases. In most situations, however, expected future price
trends are desirable. This is especially true during years of rapid
change such as the period we have just encountered. The energy situa-
tion and society's concern for the environment is projected to have a
substantial impact on prices related to agricultural development such
as irrigation and drainage. The problem is that specifying future price
expectations is more risky during such periods of uncertainty. It is
during these periods, however, that the information is most valuable.

Future product and land resource prices can be estimated by simu-
lating various regional production levels and holding future technologi-
cal considerations at some assumed level. Given a specific future
technological scenario, expanded production levels would be expected
to have a positive correlation with shadow prices for products and land
resources. The theoretical rationale is that if demand expands in the
future resulting in production expanding onto poorer soils, marginal
unit cost will increase causing product prices and land values to
increase under normal market conditions. Alternately, increases in pro-
duct prices will bid land prices up and draw more resources, including
land, into production.

The Northwest Florida study included an analysis for future time
periods with two different assumed levels of future total production and
two different rates of yield increases. The shadow prices from each
future scenario were compared to suggest the influence of different
production and yield assumptions on the value of resource development
and agricultural product prices. The analysis suggested very modest
increases in the value of development and product prices with the lower
production and yield scenario. With the high production, low yield,
scenario, these values increased substantially, but dropped off again
if the high level of yields was also assumed.

The value of resource developments such as irrigation and drainage
are very sensitive to product prices used in the evaluation. We pre-
sently assume that current prices (i.e. current normalized agricultural
prices) will prevail throughout the project life. The appropriateness
of such an assumption depends upon future price trends and can be veri-
fied only in retrospect. It is interesting to note, however, that the
Water Resource Council has recommended switching from the low to a high
agricultural production scenario used in the analysis. The assumption
that present prices will prevail through the life of a project needs to
be carefully evaluated. Price prediction capabilities need to be
refined and the impact on resource development with expected price
trends considered.



Applications of Shadow and Market Prices

It is necessary to know the value society places on resources in
order to allocate the resources in a manner that will maximize society's
welfare. Under some conditions, market prices do reflect the value
society places on resources. At other times, a shadow price may allow
a superior allocation. Inherent strengths and weaknesses of alternative
prices must be kept in mind when performing analysis of natural resource
development. McKean pessimistically points out some important theoreti-
cal considerations that should be carefully reviewed before an evalua-
tion of resource development is undertaken.

Choice of a Decision Criterion

One problem is selection of a criterion for making social choices.
McKean argues that "there is no preference function that is inherently
correct in decisions affecting several persons. One person can have a
utility function that is unambiguously correct for him...but there is no
right choice or ultimate correct group preference function to be maxi-
mized" (6).

One solution to group decisions is to maximize one individual's
utility with some constraint on disutility suffered by other group
members as characterized by a dictatorship. Another solution, at the
other end of the spectrum, is to allow majority rule and voluntary
exchange. The problem is that logic does not compel preference for any
one particular model. There is no fundamentally correct criterion for
choosing between alternate courses of group behavior, and this is the
same as saying there is no unique set of shadow prices appropriate for
making ultimate social choices.

Our democratic system of government and market economy leads us to
prefer prices reflective of voluntary exchange over prices set by one
or a few individuals regardless of his benevolent tendencies. But, as
practitioners, having selected this criterion, theoretical problems
still exist. Equity-efficiency considerations cannot be adequately
analyzed jointly. The conventional approach is to address issues on
efficiency grounds through the pareto optimality criterion. Pareto-
optimality exists when each individual has become as well off as possi-
ble in his own view without making anyone else worse off based on their
view of their own well-being.

This type of solution is reflected in a perfectly competitive indus-
try where prices tend to equilibrate at the marginal cost of production.
Each individual maximizes his well-being to the extent possible through
market interaction. The pure competition model, however, takes as given
the existing distribution of wealth. There exists a whole family of
pareto-optimal points associated with various wealth distributions but
these can not be addressed without a judgement concerning some desired
wealth distribution.

Market Prices

Even if voluntary exchange and market prices in general are assumed
appropriate as the measure of the relative values of goods and services


to society, many existing conditions contribute to inaccurate price
ratios. These are discussed by McKean under the following topics:

1. imperfect markets
2. resource use constraints
3. price-support programs
4. anticipated changes in supply and demand
5. unemployed resources
6. external effects,

Prices in markets not characterized by pure competition deviate
from the marginal cost of production. Such conditions generally result
in underproduction of goods with product prices exceeding their marginal
cost. Resource use constraints, such as labor unions, barriers to entry,
and import restrictions, which restrict input use below a competitive
equilibrium have the same general impact on product prices.

Subsidies for the production of some products have essentially the
opposite effect. Resources are used beyond the competitive equilibrium
level reflecting inappropriate prices for movements toward pareto-
optimality. The example of agricultural price supports is a case in
point. Resource use constraints and subsidies may reflect a political
preference toward a specific equity redistribution--domestic vs. foreign
populations or farm vs. nonfarm.

Prices respond to changing supply and demand conditions, and eco-
nomic analysis of the feasibility of a resource development of a rela-
tively permanent nature requires consideration of future price ratios
as well as present prices. It is possible that the act of development
itself may have an impact on prices. McKean notes these problems but
suggests that shadow prices from "programming and econometric models do
not so far have a good record when used to make predictions" (6).

Resources are not always as mobile as we assume in theory. Unem-
ployed or underemployed labor resources are a special case where the
opportunity cost of use in the immediate area may be zero especially in
the short to intermediate run. The existence of external effects from
resource use for which no compensation is made also distorts market
prices. Such prices do not reflect the desired prices for making pareto-
optimal choices.

Further Considerations

Even when reflective of pure competition and void of any other
inaccuracies market prices fail to consider future generational prefer-
ences. Society's tastes and preferences based upon complete knowledge
would probably not change significantly over any planning horizon we
Sould consider. But in cases of incomplete or inaccurate information
this may not be true. Unfortunately, we are faced with many uncertain-
ties; our knowledge base is incomplete at best and probably has sections
of inaccuracy imbedded in it.

In addition to these problems with shadow prices and their applica-
tion, the cost of developing such values can be very costly. But as



practitioners in the realm of the real world we still have a task to do
in the best way we know how. In many cases the best information we can
provide to decision makers may be shadow prices with all their inaccu-
racies1. We feel that marginal shadow prices are a very important tool
in our role as analyst and advisor to decision makers. When and where
these values are appropriate is judgemental.

Concluding Remarks

Prices are a fact of life easily understood in our market oriented
society and provide a good numeraire for communication with policy
makers as well as an effective calculus for economic decision making.
Some economic valuation is either explicitly defined or implicitly
inferred in most natural resource decisions. Shadow prices are a basic
tool useful in providing information to decision makers when market
prices are inappropriate or do not exist. Development of such prices
for goods and services not sold in a market place, and adjustment of
existing market prices which diverge from the purely competitive norm
are a means of improving the natural resource decision making process.

Unfortunately, the framework for using prices is not pure--deter-
mination of optimum social welfare is not exact. The rationale for its
use has to be based upon the belief that it is one of the best tools
we have. In such an environment the decision to use shadow prices is
judgemental and the extent to which shadow prices should be used can
not be unequivocally stated or demonstrated.

This position is supported by the reviewers by McKean's article.


1. Agricultural Price Standards, U.S. Water Resources Council, October

2. Egbert, Albin C. and Hyung M. Kim, "Analysis of Aggregation Errors
in Linear Programming Planning Models", American Journal of
Agricultural Economics, 57(1975).

3. Greenhalgh, Richard and Fred Stewart, The Use of Shadow Prices in
Determining Potential for Natural Resource Development Programs,
Food and Resource Economics Department, Econ. Report #90,
January 1978.

4. Heady, Earl 0. and Wilford Candler, Linear Programming Methods,
Iowa State Press,' 1969.

5. Krenz, Ronald D., "Current Efforts at Estimation of Costs of Pro-
duction in ERS", American Journal of Agricultural Economics,


6. McKean, Roland N., "The Use of Shadow Prices", Problems in Public
Expenditure Analysis, The Brookings Institution, Washington,
D.C. 1966, 33-65.
7. Mishan, E. J., Cost Benefit Analysis, New and Expanded edition,
Praeger Publishers, New York, 1976.
8. Stewart, Fred and Richard Greenhalgh, "The Use of Shadow Prices in
Determining Marginal Values", Paper presented at the annual
meeting of the American Agricultural Economics Association,




by Richard A. Schoneya/

Considerable recent interest has been shown in irrigation in the
humid Midwest. Many soils are well suited to irrigation because there
is an abundance of water relatively close to the surface and the soils
are amenable to an irrigated culture-high water infiltration rates
and low moisture holding capacity. In Wisconsin, the estimated
irrigation potential is 1.5 million acres. About 260,000 acres are
irrigated currently, about 17 percent of the total. Since 1949, the
growth in irrigated acres has averaged about 12.8 percent per year.
At that growth rate, it would take about 9 years to reach 50 percent
of the potential and 15 years to reach full potential. The number of
acres irrigated and growth rates of irrigated acres for states
surrounding Wisconsin are even higher. Because irrigation is a
comparatively recent phenomenon and irrigation in Wisconsin is some-
what unique, the economic impact of irrigation upon land values and
ownership is not well established.

Further assessment of the economic impact is complicated by a
wide diversity of factors affecting economic costs and returns.
These factors include soil type, field shape, ownership patterns,
topographical features, crops irrigated, management, water source,
and pumping depth.

The valuation and appraisal of soil irrigability is an important
problem. A variety of people, often with conflicting goals, have an
interest in it. They include private individuals such as buyers,
sellers and taxpayers; credit agencies and those agencies providing
transfer services; and government institutions which are involved
in the regulation, use and taxation of farm land and water rights.
In addition, because land and water valuation and their corresponding
rents are intertwined, annual rents are likely to be affected by
their valuation. Hence, tenants and landlords are also likely to be
interested in the economic valuation of irrigability.

The Assessment of the Impact of Irrigability on Land Values

In a region where irrigation is well established and irrigated
areas are relatively homogeneous, the value of irrigability can be
determined by comparing the values of land with water rights against

/Assistant Professor, Department of Agricultural Economics,
University of Wisconsin-Madison.


land with no water rights. Where water rights are legally negotiable,
the value of those rights can be observed directly. In Wisconsin,
these rights are not often sold or transferred. Assessment of
irrigability based on land sales is confronted with two difficulties.
First, there may be few parcels sold which are "comparable" to the
land being evaluated. Second, and more significantly, because of
the newness of irrigation, land market values may not have yet
correctly evaluated the value of irrigale soils; the markets may be
erratic and confound irrigability with other land characteristics.

An alternative is to impute an economic value for irrigability
based on its discounted increases in annual net returns due to
increased land productivity. Real estate appraisers have referred
traditionally to this method of appraisal as the "income approach",
"capitalized value approach" or the "earning method". Appraisers
have long recognized that the values derived by the "income approach"
represent a theoretical maximum and may differ considerably from
those values generated by the open market.

With no capital constraints, continuous compounding, and an
infinite planning horizon, equation (1.0) is the appropriate estimate.

(1.0) V = where: V = Imputed economic value of irrigability
r R = Annual net cash returns to irrigability
r = Discount rate

The present economic value of irrigability is a function of the
annual net cash returns and the discount rate. The increased net
cash returns are those due to irrigability after payment of all cash
costs including cash operating costs, taxes and finance charges
associated with equipment and can be considered as "economic rent".

The choice of the appropriate discount is always troublesome,
particularly in times of inflation. Table 1 displays recent trends
in Consumer and Industrial Commodities, Interest Rates and Wisconsin
Land Prices. Inflationary forces have induced sizeable appreciation
in the nominal prices of commodities, land values, and interest
rates. Somewhat surprisingly, recent agricultural land values in
Wisconsin (and in the North Central Region) have been appreciating
at an average annual rate considerably above the consumer and
industrial commodity trends. The rise in agricultural land values
reflects 1) its value as an inflationary hedge, 2) its associated
tax advantages and 3) the competition for land among farmers.

The impact of inflation on R, net cash returns, can be considered
by adjusting the denominator to (r i) where i is the rate of
inflation in R. However, a number of factors are still not
considered. Equation (1.0) estimates the value of land based on its
continual use over infinity. However, most land is held by private
individuals whose planning horizons are relatively short. Moreover,
they may be investing in land because of the capital gains provisions,
or as a hedge against inflation, driving up the value of land above
its earning value based on R. Because of the finite planning
horizons, income tax provisions, inflation and timing of payments, a
second model was formulated and computerized.


Table 1. Trends in Consumer and Industrial Commodities, Interest
Rates and Wisconsin Land Prices

Short Term
Change in Value From Prior Year Interest
Consumer Wisconsin Rates to
Price Industrial Agricultural Small
Year Index Commodities Land Prices Businessmen

1978 18

1977 8.0 9.9 19

1976 5.8 6.4 13 7.5

1975 9.1 11.5 12 8.7

1974 11.0 22.2 19 11.3

1973 6.2 6.8 18 8.3

1972 3.3 3.4 7 5.8


The Economic Model

An economic, computer model was developed to impute the value
of irrigability of soils in corn production based on increases in
returns for a on-going Wisconsin farm. The economic model is a
traditional discounted net present value of the net income stream
over n years (1.2).

n _t
NPV = E [Rt(1 + ) Pt(1 + r)-] E + NTV(1 + r)
t = 1 + r

where: Rt = Annual Net Cash Returns
i = Annual Increase in Rt
r = Discount rate on equity capital
t = Time period
n = Ending period
NTV = Net terminal value
E = Owner's equity

Pt = Annual property taxes

Equation (1.0) discounts annual net cash returns (Rt) and
includes property taxes at 2 percent of the current market price,
equity (E) and net terminal value.

Annual Net Cash Returns

The annual net cash returns (R) are those which can be attributed
to the adoption of irrigation. The cash returns (G) are adjusted
for cash costs of operation (C), financial costs associated with the
irrigation equipment and development (F), income taxes (T), a manage-
ment charge and a machine replacement charge (S).

(1.3) Rt =Gt -Ct Tt Mt St

where: Rt = Annual net cash returns

Gt = Increased annual cash revenues
Ct = Increased annual cash operating
Tt = Income taxes associated with the
irrigation investment

Ft = Finance costs associated with
the irrigation investment in
equipment and development
Mt = Management charge
St = Replacement cost of components


Annual gross returns are particularly difficult to estimate
because shifts in cropping patterns are likely to occur over time.
Most newly developed irrigated soils are devoted initially to corn
production and later shifted to include specialty crops in a
rotation. Corn is normally continued in the rotation, however,
because of risk, crop diseases and pests.

The cash costs of operation include those costs associated with
operating the equipment. Other cash cost components include
financial costs and taxes. Financial costs include all cash costs
associated with securing and repaying investment and operating
capital. Repayment costs include both principal and interest

Taxes include income tax credits, and State, Federal and Self
Employment Income taxes. Property taxes are not included in Rt,
because they are a function of NPV. It is particularly important
that taxes be included because of their relative magnitude. In
Wisconsin, the combined Federal, State and Self Employment taxes can
very easily exceed 40 percent at the margin. In times of inflation,
the impact of income taxes is particularly important.

An important part of income tax management is the use of
investment credit and the judicious use of fast depreciation methods.
Investment credit is a particularly powerful incentive to invest
because 1) it effectively decreases the acquisition cost without
affecting the associated depreciation as a normal tax deduction, and
2) it is "after tax" income. Fast depreciation and additional
first-year depreciation methods allow rapid write-off in the early
portion of the matching life cycle, when the time value of money is
high. Repayment of excess depreciation, if any, occurs when the
machine is sold, when the time value of money is relatively low.
Finally, the interest components of an amortized loan are also heavily
weighted to the early part of the life cycle.

One other tax-related factor becomes important, particularly in
evaluating the net terminal value (NTV) of land. Currently IRS tax
procedures distinguish between ordinary income and capital gains
income. Investments which can generate capital gains income become
relatively more profitable because only half of the gain is taxable.
Thus land becomes extremely attractive to investors because of its
relatively steady appreciation in value.

Trend Variables and Inflation

Incorporated in the model are trend variables: 1) a factor, i,
increasing annual net returns (Rt), 2) a factor, a, reflecting
the time value of money imputed to equity capital. In this model
the discount rate is calculated on an after-tax basis. The discount
rate reflects the reservation price on equity capital and incorporates
three components: the "real" rate of interest or opportunity cost,
capital risk allowance, and inflation rates Once the inflation rate
is incorporated in r, its presence in R and land values also must be
acknowledged. Thus trends in R and land values are also included.


Property Taxes

In equation (1.2) two terms are a function of NPV: property
taxes and the net terminal value of the investment. Property taxes
are assumed to be 2% of the NPV and increasing at an annual rate, a,
the annual appreciation in land values. Equation (1.4) expresses
net property taxes, adjusted to an after-income tax basis, as a
function of NPV. Appreciation in land values, a, acts as an inflator.
Equation (1.5) converts an annuity to a present value.

(1.4) Pt = .02(1 + a) NPV (1 B)
n A
(1.5) E Pt (l + r) =.02 rn NPV (1 B)
t = 1 a,n

where: a = Annual appreciation in land
B = Marginal tax bracket
An = Amortized annual payment at
r interest and n years

Terminal Value of the Investment: Future Value of Irrigability

After the end of n years, it is assumed that the investment is
sold and that the value of irrigability has a residual value which
may become incorporated with the land. The NTV (net terminal value)
of irrigability is assumed to be a function of NPV and income taxes.
Since the value of irrigability should appreciate in value in a
manner similar to land, the terminal value of irrigability is assumed
to be equation (1.6).

(1.6) TV = NPV (1 + a)n where: TV = Terminal value of
Appreciation in irrigability is treated as capital gains,
therefore only one-half of the amount is taxed (1.7).

(1.7) NTV = TV (.5 (TV NPV) B) where: NTV = Net terminal

Final Equation

The final equation can be expressed as equation (1.8), where
the NPV is the sum of discounted net returns, corrected for property
taxes and capital gains which are based on NPV.


n a
(1.8) NPV = ( Rt E) / [1 + .02 *- (1 B) {(1 .5B) *
t = 1 Ai,n

1 + r,-n
(-.- + .5B(1 + r)-n}]

The Computer Model

In addition to the basic discounting equation (1.7) the computer
model incorporates the following features and subroutines: 1) a
learning curve, 2) tax subroutines, 3) various types of loans,
4) replacement of system components and 5) economic and financial
breakeven points for prices and yields for various equity positions
and an estimated minimum repayment period for the machinery loan.

Learning Curve. A learning curve was included which adjusted
yield increases according to the year of operation. The learning
curve reflected 1) gains in irrigator experience and productivity and
2) the adjustment of a developed soil to an irrigated culture.
Irrigation skills were assumed to be "learned" at the rate of 60, 85,
90 and 100 percent in the first, second, third and fourth years

Tax Subroutines. Self employment, State and Federal income
taxes are estimated each year based on 1) other taxable income,
2) number of exemption and 3) irrigation taxable income. Because of
the magnitude of the irrigation investment, the interest payments on
borrowed capital, which are tax deductible expenses, can generate
sizeable changes in taxable income. Since both Wisconsin and U.S.
income taxes are very progressive, changes in taxable income can
generate correspondingly greater than proportionate changes in taxes.
Thus, the timing of interest payments and changing marginal tax
brackets have a profound impact on annual after-tax cash flows.
Accordingly, tax subroutines were incorporated in the model because
of the potential problems and corresponding bias in selecting a
single-valued marginal tax rate.

The appropriate amount of investment credit was determined
according to eligibility and anticipated component lifetime. The
model applies investment credits to Federal income taxes as it
becomes available.

Loan Type. Three types of loans are generated according to user
selection: operating, machinery and real estate loans. The model
determines the appropriate amortized payment for the latter two and
maintains interest and principal payments for tax calculations.
Operating loans are made and repaid according to the net cash
position. Short run cash surpluses are given a credit.

Machine Replacements. Various components of the irrigation
system are replaced at prespecified intervals. Their investment
requirements are met either by cash surpluses or additional debt.


Breakeven Points. Additional information is generated which aids
the decisionmaker in surveying and assessing both economic and
financial feasibility. Breakeven yields (product price fixed) and
prices (yield increase fixed) are generated to reveal economic
feasibility (all inputs priced at their opportunity cost) and
financial feasibility under various equity positions (meeting all
cash demands in a timely manner). In addition, because of differing
practices in establishing machinery loan repayment periods, the
minimum machinery loan repayment period was estimated based on the
annual equivalent cash flows.

Representative Systems and Data

In Wisconsin as in many other areas, the imputed value of
irrigation is heavily influenced by the following variables: 1) field
size and shape, 2) anticipated yield increases, 3) markets available,
and 4) depth to water. Economies of size (decreasing average costs)
exist up to a certain point in irrigation. Economies of size are
generated by the basic nature of the center pivot and by the s
specialized nature of many machines associated with the specialty
crops. The cost of ownership (fixed costs) and operation (variable
costs) are least with the standard size center pivot (CP) on a square,
160 acre field. However, only 132 acres are covered effectively
because the corners are left unirrigated.

Smaller or irregular-shaped fields are caused by topographical
features, roads or ownership patterns. They generate higher invest-
ments and operating costs per irrigated acre. Two systems commonly
employed to irrigate subsized or irregular shaped fields are the
towable center pivot and the traveling gun. These systems trade
higher labor requirements for greater flexibility and mobility. Each
of the two operate in a semi-automatic mode, operating automatically
on location but requiring towing and manual set-up at the next

The towable center pivot system (TCP) is a subsized center pivot
equipped with a towing package. The TCP is towed from location
to location by a farm tractor and is assumed able to cover 111 acres.
Once in place, the TCP operates like a conventional center pivot.
The traveling gun system (TG) operates in an entirely different
manner, by pulling itself across the field via winch and cable. The
TG is assumed able to adequately cover 80 acres.

The investments in deep well irrigation are very substantial.
The center pivot, towable center pivot and traveling gun require
total investments of $51627, $40836 and $38531 which generate invest-
ments of $391, $368 and $482 per irrigated acre, respectively
(Table 2). The investment requirements generate annual fixed costs
of $63, $58 and $85 per acre for the CP, TCP and TG systems.


Table 2. Investment Requirements and Associated Cost Characteristics
Of Three Typical Irrigation Systems, 1978.

Center Center Traveling
Item Pivot Pivot Gun

1. Acres Irrigated (A) 132 111 80
2. Total Investment ($) 51,627 40,836 38,531
3. Investment Per Acre ($/A) 391 368 482
4. Amount Borrowed 38,720 30,627 28,898
5. Economic Breakeven Yield
Increase (Bu/A) 53 59 79
6. Fixed Costs Per Acre (S/A)!/ 63 58 85
7. Annual Variable Costs
Per Acre ($/A)1 / 20.40 31.44 36.00

a Includes a planning horizon of 12 years and an interest charge of
9.5% on equipment; a 25 year planning horizon and 9.0% interest
on well and land development; a 5 year 9.5% planning horizon on
hoses and repair charges of 2.5% on center pivots and 3.0% on
traveling guns.
/ Assumes 12 inches of water applied per acre.


Owner Equity and Terms of Financing

It was assumed that 75 percent pf the investment was debt
financed and the remaining 25 percent was financed by the equity
capital, and that 100 percent of the operating capital was furnished
by the investor. Financing was based on a real estate loan (25 years
and 9.0% interest) and a machinery loan (7 years and 9.5% interest).
Additional emergency operating loans were available if needed at
12 percent.

Discount Rates and Trend Variables

The selection of parameter values for the variables which
generate future trends is particularly problematic. However, the
relationship between the trend variables associated with R, TV and
the discounted rate, r, is more stable than the absolute values. In
the following analysis it was assumed that R increased at an annual
rate of 4 percent and that TV increased at an annual rate of 6 per-
cent. By ranging the discount rate from 7 to 11 percent, the
relative relationship between the variables can be examined. The
appropriate discount rate will vary considerably from individual to
individual according to investment opportunities and marginal tax

Annual variable costs include charges for labor, tractor usage
and power. It was assumed that 12 inches of water were applied per
acre. The TCP and TG incur greater energy costs, reflecting longer
piping distances and potentially higher pressures, and greater labor
charges associated with moving and set-up. Thus the annual variable
costs are higher for the TCP ($31.44) and TG ($36.00) systems than
the CP ($20.44).

Land Productivity

Annual gross returns were estimated for two corn yield increase
levels, 70 and 85 bushels per acre, based on $2.40 per bushel of
corn. With exception of the TG system, these yield increases exceed
the economic breakeven yields of 53, 58 and 79 for the CP, TCP and TG
systems, respectively. However, the economic breakeven yields are
based on a slightly different set of assumptions, mainly due to the
lack of consideration in the appreciation of the value of irriga-
bility. Hence a different charge to investment is used.

It was assumed that corn was already being produced. Thus most
of the costs are fixed, such as operation and ownership of tillage
machines and most agricultural chemicals. The costs associated with
increased yields due to irrigation are those generated by higher
plant and fertility levels, harvesting, handling, storage and
marketing. The costs associated with increased production are
estimated to be $.70 per bushel.


The Results

The imputed values or NPV of future earnings, based on a 25 year
planning horizon, are presented in Table 3. The imputed values ranged
from -$57 per irrigated acre (11 percent discount rate, traveling gun
and 70 bushel yield increase) to $572 per irrigated acre (7 percent,
full-size center pivot and an 85 bushel yield increase). At a 9
percent rate and a 70 bushel yield increase, the full-size center
pivot generates a NPV of $280 per irrigated acre. The NPV values are
increased by 1) decreasing the discount rate, 2) increasing the
productivity or 3) moving to optimally shaped fields. Further
discussion of the results will be divided into three sections: 1) the
impact of field shape and size, 2) the impact of added productivity
and 3) the impact of varying discount rates.

The Impact of Field Shape and Size

Three systems were identified according to their ability to
accommodate various sizes and shapes of fields. In order of efficiency
of use, the most efficient systems are the following: 1) full size
center pivot (132 acres irrigated per system), 2) towable center
pivot (111 acres irrigated) and 3) the traveling gun (80 acres
irrigated). As can be readily seen field shape and size has a
tremendous impact on NPV per irrigated acre--if the shift is to the
traveling gun. At a 9 percent discount rate and a 70 bushel yield
increase the shift to a towable center pivot reduces the NPV per acre
by $293 per acre over full size pivot operation. Reducing the
discount rate increases the difference while increasing the yield
production decreases the differences.

The Impact of Added Productivity

As would be expected, differences in added productivity can
generate sizeable differences in the imputed value of irrigability,
the greatest difference being associated with the lowest discount
rates. At an annual discount rate of 7 percent, the differences
ranged from $192 per acre (CP) to $228 (TG). If the annual discount
rate is raised to 11 percent, the differences due to productivity
ranged from $148 (CP) to $168 (TG). Increased productivities added
relatively more to the TG system because of its precarious nature:
it is not nearly as profitable.as the other two systems. As new
technologies emerge and increase profitability, the value of
irrigability will increase. If production shifts to more profitable
crops such as potatoes, the value of irrigability will also increase.

The Impact of Varying Discount Rates

The appropriate discount rate would vary from person to person
depending upon their investment alternatives, tax brackets, time
preference for money and-risk bearing ability. In addition it would


Table 3. Imputed Value to Irrigability and Water, NPV Model

Full-Sizeb/ Towable / Traveling
Center Pivot- Center Pivot-' Gun!/

Dis- 70 Bu. 85 Bu. 70 Bu. 85 Bu. 70 Bu. 85 Bu.
count In- a In- a In- a In-a/ In- In-
Rate crease- crease- crease- crease- crease- crease-

(Per- (Dollars Per Acre) (Dollars Per Acre) (Dollars Per Acre)

7 380 572 303 508 52 280

8 325 506 258 449 16 224

9 280 449 220 398 -13 179

10 242 400 188 353 -37 142

11 209 357 161 315 -57 111

/Yield increase is in bushels per

irrigated acre over the present

SValues based on 132 acres irrigated.
cValues based on 111 acres irrigated.

- Values based on 80 acres irrigated.


Table 4. Reduction in Imputed Value of Suboptimal

Field Size and Shape~a/, by Production Increase

Towable Traveling
Center Pivot Gun

Discount 70 Bushel 85 Bushel 70 Bushel 85 Bushel
Rate Increase/ Increase Increase Increase

(%) ($/Irrigated Acre)
7 77 64 328 292
8 67 57 309 282

9 60 51 293 270

10 54 47 279 258

11 48 42 266 246

As compared

Increase in

to the full-size center pivot.

corn yield over the present yield.


Table 5. Increase in the Imputed Value of
Irrigability with a 15 Bushel Increase in Yieldsa/

Discount Towable Traveling
Rate Center Pivot Center Pivot Gun

(%) ($/Acre)
7 192 205 228
8 181 191 208

9 169 178 192
10 158 165 179

11 148 154 168

/Increase in added yields from 70 to 85 bushels per acre.


vary from crop to crop depending on the comparative riskiness of the
crop. Again as would be anticipated, varying discount rates have a
significant impact on imputed values. An investor with limited long
run investment opportunities might find a 7 to 8 percent discount
rate appropriate. Another investor with greater investment oppor-
tunities or who perceives irrigation as being risky might choose an
11 percent discount rate. Investment in a center pivot and an 85
bushel yield increase with different discount rates would lead one
investor to place a maximum imputed value of $572 (7%) while another
investor would impute a value of $357 (11%).

Differences in discount rates might also lead one investor to
invest in a traveling gun irrigation system (7-8%) and others to
reject it (9-11%) if the yield increase in production is 70 bushels
per acre.


Although their impact on the imputed values of irrigability
differed, for the range of parameters evaluated, the value of
irrigability ranged from -$57 to $572 per irrigated acre. Variables
examined were productivity increases, field shape and size, and
discount rate.

Field shape and size are the most important variables examined in
determining values because they determine the appropriate irrigation
system and thus its cost structure. Differences as much as $292 per
irrigated acre were generated.

Increases in productivity are also important in determining
annual net returns. Although seemingly not great, an increase from
70 to 85 additional bushels can generate an additional $228 of
imputed value per irrigated acre.

Finally, although the imputed values were sensitive to the
discount rate, over the range covered, its impact is similar to the
others. Its impact on the center pivot was between $357 per acre
(11%) to $572 per acre (7%).


The imputed value to irrigability was estimated based on the
discounted net cash flows over a 25 year planning horizon with corn
production. The imputed values were estimated for three variables:
field shape and size, added productivity and discount rates. Invest-
ment requirements were based on Wisconsin conditions and estimated
for three systems: a full-size center pivot, a towable center pivot
and a traveling gun. Investment requirements ranged from $391 per
acre to $482. Based on a 9 percent discount rate, a full-size center
pivot and an 85 bushel yield increase, the present.value of
irrigability was $449 per acre. Field shape and size, productivity
increases and disco-nt rates, in that order, were very important in
determining the value of irrigability.


Bruce R. Beattie, Michael D. Frank and Ronald D. Lacewell

Water resources for irrigated agriculture in the western United
States have always been limited and in some cases their development
costly. However in recent years problems associated with drought,
declining aquifers, rising energy prices and competition with other
uses have focused more than the usual attention on the future viability
of western irrigated agriculture.

The purpose of this paper is to address several issues pertinent
to the economics of western irrigated agriculture. We begin by consid-
ering the value of water for the major irrigated regions of the 17
contiguous western states. The sensitivity of these values to changes
in farm product prices and pumping costs is discussed. Several myths
for assigning agricultural water a social value in excess of farmer's
ability-to-pay are presented and dispelled. The paper concludes with
some thoughts on the future viability of irrigated agriculture in the

The Social Value of Water

As a general rule, economists tend to be outspoken and fairly
stubborn concerning the role of the marketplace in assessing the social
value of resources in alternative uses and in allocating resources
among those uses. Not surprisingly we find ourselves in agreement with
those who hold (i) that it is the market value of the products produced
by irrigated agriculture that best serves as an indicator of the bene-
fit (social value) of water used in agriculture and (ii) that this value
must be measured against its acquisition cost including its value in
foregone uses. The enormous capacity of the market place to synthesize
the preferences of consumers and producers and to reflect these prefer-
ences in terms of prices (value) of goods and resources is indeed im-
pressive. Thus, it seems appropriate that we begin today by taking a
look at the market value of water in agriculture.

Texas Agricultural Experiment Station Technical Article No. 14514.
This research was funded in part by the office of Water Research and
Technology and the Texas Water Resources Institute.

Bruce R. Beattie is an, associate professor of agricultural eco-
nomics; Michael D. Frank is a research associate in agricultural eco-
nomics; and Ronald D. Lacewell is an associate professor of agricultural
economics at Texas A&M University.

University of Florida Home Page
© 2004 - 2010 University of Florida George A. Smathers Libraries.
All rights reserved.

Acceptable Use, Copyright, and Disclaimer Statement
Last updated October 10, 2010 - - mvs