Assessing Salinity-Control Programs on the Colorado River

Salinity in the Colorado River Basin adversely affects more than 12 million people and 1 million irrigated acres in the southwestern United States and a portion of Mexico. The river's salt content increases from about 50 milligrams per liter (mg per liter) at its headwaters to more than 800 mg per liter at Imperial Dam, Arizona, the last major diversion point before the river reaches Mexico. Although salt springs, water leaching through saline soil and rock formations, and other natural sources contribute to the salt load, salinity increases since 1900 mostly are of human origin. Society's increased consumption of relatively pure water causes salts to be concentrated in remaining stream flows, thus increasing salinity. Additional salts are added by irrigation return flows, which often percolate through saline rock and soil before rejoining the river.

In the 1960s, saline waters discharged from the Wellton-Mohawk project in Arizona and the filling of Lake Powell behind the Glen Canyon Dam sharply increased the salinity of water delivered to Mexico. Concurrently, the perception of salinity damages grew, and Upper and Lower Basin water interests united to form a strong coalition to push for a federally funded salinity-control program.

The result was the Salinity Control Act of 1974 authorizing construction of a major desalinization plant under Title I and four salinity-control units and the investigation of twelve others under Title II. Initially, $155 million was authorized for Title I and $125 million for Title II, but costs of Bureau of Reclamation and Soil Conservation Service salinity-control units under Title II ultimately may approach $500 million. Significantly, federal monies will pay entirely for the desalinization plant to meet commitments to Mexico and for 75 percent of the domestic salinity-control program. The remaining 25 percent will not be paid by project beneficiaries but by surplus power-generation revenues from the Colorado River Basin.

Despite this large financial commitment, considerable hydrologic uncertainty surrounds salinity control. At one time it was projected that by the year 2000 salinity levels at Imperial Dam would rise dramatically to 1,400 mg per liter from the 1972 level of 879 mg per liter. Subsequently, however, official projections have been reduced to a current expectation of 946 mg per liter by the turn of the century, without salinity control. Salinity at Imperial Dam has actually fallen in recent years to 816 mg per liter in 1981 and to 732 mg per liter in 1983. This, of course, leads one to speculate about the efficacy of salinity-control investments.

It also takes time for water and salts to pass from the Upper Basin through all the reservoirs to the Imperial Dam. The Bureau of Reclamation now assumes a hydraulic retention time of five to seven years, so that 90 percent of salinity-control benefits probably will be perceived at Imperial Dam in that time. This is important, because the economic measure of damages avoided by salinity control must be discounted to establish a net present value for comparison to program costs.

Finally, both the relative contributions to salinity by various sources and the amount of salt that must be removed from the river to achieve a given reduction remain uncertain. The accepted estimate of the sources of salinity are from a 1971 EPA study that attributes 47 percent of salt contributions to natural sources, 37 percent to irrigation, 12 percent to reservoir evaporation, 3 percent to water ports from the basin, and 1 percent to municipal and industrial use. The removal of 9,900 tons of salt from the river in the Upper Basin was believed to result in a 1-mg-per-liter decrease in salinity at Imperial Dam. (Since our research began, the Bureau of Reclamation has changed this assumption to 10,990 tons per milligram per liter, thereby further lowering the potential benefits of removing a given amount of salt.) However, substantial quantities of salt may precipitate naturally out of the water in Lake Powell, and other measures of the sources of salts may be problematical.

Research on salinity control

The changing view of the severity of the problem and the previously unconsidered lags between controlling salinity and the realization of benefits raise the question of whether federal investments in salinity-control programs are economically efficient. Furthermore, how might program costs be shared among participants? And does the method of assessing its costs affect a project's efficiency?

We first approached these questions empirically by estimating the agricultural benefits of salinity control through comparisons of net farm income at different salinity levels. Linear programs simulated farmers' efforts to maximize profits in the Imperial Valley of California with salinity levels of 800 mg per liter and 1,100 mg per liter in irrigation water. As salinity increases, it becomes necessary to undertake additional irrigations to leach salts or to switch to more salt-tolerant, but less-profitable, crops, even though these measures reduce farm net income. Benefits to municipal water users were adapted from previous federal studies, and the estimate of total benefits per ton of salt removed was compared to cost estimates from the Bureau of Reclamation and the Soil Conservation Service.

Another model simulated irrigated agriculture in a major area of salt loading in the Upper Basin—the Grand Valley of Colorado—where irrigation return flows pass through highly saline soils before reentering the river. We tested a variety of economic incentives for their ability to reduce salinity control-management techniques and irrigation technologies.

The salient points of the analysis are summarized below. The questions of efficiency and equity are addressed in turn, together with policy implications.

Benefits versus costs

Average annual damages from salinity to irrigated agriculture in the Lower Basin amount to $51,400 per milligram per liter in 1982 dollars. As noted, however, salinity reduction does not occur instantaneously, so the damage figure must be discounted by the hydraulic retention time of the river to obtain the present value of benefits of Upper Basin salinity control. We estimate benefits to agriculture from salinity control (using an 8 percent discount rate) at $39,100 per milligram per liter.

We adapted a previous study to estimate discounted municipal salinity-control benefits of $218,700 per milligram per liter. We did not include environmental, health, and aesthetic benefits in this analysis, but most observers do not see them as significant in the current salinity range. We also omitted secondary, or indirect, economic benefits as they normally are not included in project analysis from a national accounting perspective. Under normal economic conditions, these regional impacts are offset by losses or gains elsewhere in the economy.

We thus estimate the total benefits of salinity control to be $257,800 per milligram per liter of salinity at Imperial Dam, or $26 per ton of salt removed from the upper stretch of the Colorado River. The Bureau of Reclamation's total estimate of $513,300 is almost double because it includes secondary benefits and fails to allow for the river's time lag.

This benefit estimate can be compared with projected salinity-control costs of existing and future projects. Cost estimates were available for nine Soil Conservation Service (SCS) on-farm projects that decrease salt loading mostly by reducing deep percolation of irrigation water on Upper Basin fields, although they also include some lining of irrigation canals. Benefits exceed costs for only three of the nine SCS projects—the Virgin Valley, Price—San Rafael, and Upper Virgin Valley units.

Similarly, of ten Bureau of Reclamation projects, only the Paradox Valley and Las Vegas Wash units are economically feasible. Benefit-cost ratios for the other eight projects generally are lower than the SCS on-farm projects. This is because the bureau's projects emphasize capital-intensive solutions, while substantial salt-load reductions can be obtained through relatively inexpensive changes in irrigation-management techniques.

Cost-sharing mechanisms

The Soil Conservation Service subsidizes the capital cost of improvements to the irrigation system to achieve salt reductions. How farmers are required to pay their share of costs affects the cost per ton of salt removed through on-farm programs. We analyzed the effect of a wide variety of cost-sharing mechanisms in the Grand Valley of Colorado, with the net social cost per ton of salt removed as a common measuring stick. These social costs are net of the value of irrigation labor saved and do not include the cost of administering the program.

We found the economic efficiency of cost-sharing mechanisms to vary with the degree to which an economic incentive is directed toward salt discharges. As economists would predict, an effluent tax on the salt discharged by each farm achieves salinity control at the lowest social cost. Unfortunately, it is not practical to measure the amount of salt or any other non-point-source pollution coming from every irrigated field. Average discharges could be estimated given soil type, crops, and irrigation techniques, but even this approach is an administrative nightmare.

A more feasible alternative is to tax production inputs whose use correlates directly with the level of pollution. For salinity, the influent tax should be placed on irrigation water. This cost-sharing mechanism is not as efficient as an effluent tax—the relationship between reduced water use and lower salt loads is not exact—but a tax on water would be much easier to administer because charges for water already are assessed. In addition, subsidies for conserving water will have essentially the same effect on salinity control as taxes on water consumption.

Subsidizing irrigation system improvements moves the cost-sharing mechanism one step farther from the goal of reduced salt loads. Capital-intensive investments in water-conserving technologies are encouraged, but this type of subsidy ignores irrigation management options. This results in an increased social cost per ton of salt removed. Property taxes are the least efficient cost-sharing mechanism of all. Raising property taxes creates no incentive to change salt loads, but it could raise revenues for the local cost share.

Who pays?

In assessing the worth of a project or program, economists tend to focus almost exclusively on the investment's efficiency, but the equity question—how costs and benefits are distributed—is equally important. In this case, the costs of salinity control are measured by the net cost to society per ton of salt removed; this may be defined as the farmers' plus the government's cost. It includes the cost of capital improvements to irrigation systems, the value of labor saved by these investments, and the net farm income forgone by changing cropping patterns or retiring cropland altogether. However, the amount of taxes or subsidies is not counted, because they are transfer payments that balance out across the economy. Thus, two cost-sharing methods that have the same net social costs may distribute them very differently between parties.

For instance, subsidies to improve irrigation systems or payments for water conservation would cost the government far more than the net social cost. Farm income is increased by these policies. In contrast, the charges on salt or for water required to induce lower salt discharges reduce farm income by more than the cost to society. In an extreme case, farms might be forced out of business.

This unintended hardship on Upper Basin farmers could be reduced by the use of increasing block rates, much as lifeline electricity rates allow the poor a minimum amount of low-cost electricity. Some basic amount of water—say, 3 acre-feet per acre—could be sold at its distribution cost with a surcharge placed on consumption in excess of this amount. This significantly reduces the cost to farmers while retaining the economic incentive to lower salt loads.

The equity or distributional problem highlights the value of flexible cost-sharing mechanisms, which can be adjusted to apportion costs in a variety of ways. The "carrot-and-stick" approach is a good way to ensure flexibility in setting explicit cost shares. Tax and subsidy combinations may have higher administrative costs, but they may provide a fairer solution.

Who should pay?

Who should pay for controlling salinity in the Colorado? The three likely candidates are upstream farmers who add salt to the river, downstream water users who are affected by water quality, and the government. Maximum economic efficiency calls for sharing costs in proportion to the benefits received; this is called the benefit principle. Defining the beneficiaries of salinity control in turn depends on who holds the property right for water quality. If it is held by the public, then polluters would be liable for salinity-control costs. From the EPA estimates cited earlier, all Upper Basin irrigators who contribute to salt load would be liable for 37 percent of total salinity-control costs. Most reservoir evaporation occurs for the benefit of the Lower Basin, suggesting that downstream irrigators and urban water users should be liable for 12 percent of costs. A tax on water exported from the basin would cover the costs of controlling the 3 percent of salinity it causes, and municipal and industrial water users would pay 1 percent. The federal government would pay the remaining 47 percent of costs, since most natural salt loading comes from publicly owned lands.

On the other hand, if landowners have the right to pollute, then salinity control costs should be shared among those who suffer damages. Our research indicates that downstream irrigators would be willing to pay up to $3.95 per ton of salt removed, while municipal interests could collectively pay as much as $22.05 per ton. Upper Basin farmers would pay only for labor savings and production increases resulting from the program. Under this property rights scenario, the federal government would pay for benefits accruing to the nation as a whole. This might include recreation, fish and wildlife benefits, and the value of meeting our salinity agreements with Mexico.

We believe that some explicit share of the costs of salinity control should be paid by both upstream and downstream water users. Upstream irrigators who developed their lands long ago had no inkling that they someday would be liable for then contributions to the salinity of the river. Indeed, the federal government not only encouraged but also often financed further development of irrigated acreage. The recognition that some water-quality degradation is a natural consequence of irrigation should set this parameter as part of a defined water right. Thus, the liability of upstream irrigators for salt loading would be limited below the 37 percent figure. However, economic incentives should be used to collect their cost share, while encouraging reductions in salt discharges to the "natural" minimum.

Downstream water users should bear responsibility for salinity resulting from reservoir evaporation, but what is more important is that they are in a position to reap large and tangible benefits from salinity control. Beneficiaries sharing the cost would further encourage some rational limit to the quest for salinity-control projects. Cost shares of 15 to 25 percent for both upstream irrigators and downstream beneficiaries seem reasonable. A tax on water exports to cover that 3 percent of salt loading also is desirable. The federal government should bear the remainder of salinity control costs on the grounds that publicly owned lands contribute as much as one-half of the total salt load, some salinity control benefits may be considered public goods, and the government is partially responsible for irrigation development in the Upper Basin.

Prognosis

Cost sharing may make good economic and common sense, but actual implementation of such a plan is highly unlikely. Considerable resistance would be mounted by all the beneficiaries of salinity control now being subsidized. As Ralph Johnson says, "It is clear that economic efficiency has not been the dominant, or even a highly important, goal in the design of the Colorado Basin salinity management program. This goal has been subordinated to the twin goals of political equity and income distribution."

Water users can point to a long tradition of federal financing for water projects. Precedents for federal funding of salinity control can be found in the low effective cost shares established for agricultural water supply projects (19 percent), nonpoint-source water-quality management (3 percent), and runoff-control projects (34 percent). Various interests have come to rely on federal solutions to water problems, solutions that usually increase the supply of water as well. In turn, the abundant water supply has contributed to the overwatering of Upper Basin crops, one of the causes of salinity.

Yet with the alternative being an end to all federal largesse, the budget deficit creates pressure for cost sharing that is more in line with the benefit principle. Evidence of this is seen in the action by Congress last year that raised the regional share (paid from power revenues) of salinity control from 25 to 30 percent of costs. In addition, at least 30 percent of the SCS on-farm salinity-control improvements will be recovered from participating farmers.

Should economic considerations ever become a priority, several concrete measures can be employed to get more bang from the federal buck invested in salinity control. First, only those projects for which the present value of benefits exceeds costs should be undertaken. Hydrologic research and modeling efforts repeatedly have reduced the dimensions of the salinity problem and the potential benefits of control. Constructing the five efficient projects cited above would provide some reduction in salinity while offering an opportunity to advance on the learning curve of project design, but postponing the remaining projects would avoid irreversible commitments of funds to nonproductive social investments.

Each element of USDA-SCS on-farm programs should be scrutinized. Less productive measures, such as mainline canal lining, often are added to gain farmer approval. This local support may be gained at less cost by the innovative use of economic incentives. Similarly, cost-sharing and irrigation-management programs should be extended to residential areas, which may be among the worst culprits for deep percolation and salt discharge per acre. An education program for residential water users about evapotranspiration needs and irrigation concepts should be a very cost-effective salinity control measure.

Finally, there is the opportunity for increased efficiency and fairness through the choice of cost-sharing mechanisms and economic incentives. The greatest gain probably would come from establishing a well-defined water market. Market prices eventually would reflect both the value of water saved and salt removed, creating a strong incentive to conserve irrigation water. This is tantamount to the conservation subsidies discussed earlier. Of course, legal and institutional obstacles to water markets first must be surmounted.

The means of cost sharing should be more thoroughly considered in all areas of public investment. The returns to investments in salinity control have been affected substantially by haphazard and implicit decisions made in the political process. If today's policymakers would explicitly consider equity impacts and eliminate unintended subsidies, they could increase both the efficiency of public projects and fairness among those most affected by salinity in the Colorado River.