The Evolution of Irrigation Institutions in California: The Rise of the Irrigation District, 1910-1930

Edward McDevitt
California State University-Northridge

Abstract. This paper addresses the phenomenon of the sudden decline of private irrigation enterprises and the corresponding rise of public irrigation enterprises in early twentieth-century California. The existing literature contains explanations of this phenomenon that views public irrigation enterprises either as solutions to market failure or as non-efficiency enhancing, redistribution mechanisms. An alternative explanation is proposed. This explanation does not rely on the usual market failure arguments nor on rent-seeking theories, but rather it explains the rise of the irrigation district as a result of a complex combination of agricultural, regulatory, and legal changes during the early years of this century. Using newly collected primary data from irrigation districts and contemporary newspapers, as well as data from state and federal sources, these alternative hypotheses will be tested.

The first significant movement towards public ownership of water in California occurred with the passage of the Wright Act in 1887. This act created the irrigation district, a local governmental unit with the authority to issue bonds, levy taxes, and condemn property. The first wave of district formation(1887-1893) resulted in the creation of 49 irrigation districts. Faced with strong opposition from large landowners, the district movement quickly collapsed. In contrast, the second wave of district formation(1909-1927) was to meet with long-term success. During this second wave, 112 irrigation districts were formed, several of which remain active up to the present. (Table 1.) This period witnessed the decline of the private irrigation company as the irrigation works of many of the largest and most important private companies were acquired by irrigation districts.

Several questions present themselves: What factors explain the sudden rise of the irrigation district and the corresponding decline of the private irrigation company? Why was the first wave of district formation a failure, while the second wave of formation resulted in success? How do we explain the timing of the second wave of formation? The answers to these questions should provide some insight on the broader issue of institutional change.

The origin and development of the irrigation district has traditionally been explained in terms of market failure: private irrigation companies were not capable of raising large amounts of capital to finance large-scale irrigation projects, there were problems of monopoly, or of post- contractual opportunistic behavior. It will be one of the contentions of this paper that these traditional arguments fail to provide convincing explanations of the rise of the irrigation district.

Recently, this traditional view has also been challenged by Rodney Smith(1983). Smith contends that the irrigation district was a mechanism for facilitating intra-district income transfers. Smith develops a median-voter model to demonstrate how a majority of farmers in an irrigation district can manipulate the water toll-tax structure to grant themselves a subsidy at the expense of a minority of farmers. Any district costs not covered by water toll revenue must be covered by property tax revenue. Therefore, any given subsidy (the amount by which the water price is below the marginal cost of the water supply) implies a unique property tax rate. If the water-use intensity (water used per acre) of the "median voter" exceeds the mean water-use intensity for the district, then the median voter will support a pricing policy which sets price below marginal cost. The lower price will benefit the median voter in proportion to his water-use intensity, yet the property tax increase will be diffused over all district landowners.

This hypothesis, while useful in explaining much about the first wave of district formation, does not succeed in explaining the second wave of formation. The Smith argument implies that we should at least expect to observe some patterns of opposition during this second period similar to the patterns of opposition observed in the earlier period of district formation. However, there is little in district histories that would indicate any substantial litigation between landowners within districts, nor do we find the other forms of opposition so common in the first wave.

If this rent-seeking hypothesis were correct, we should expect to observe district rules which would have made it relatively costly for district landowners to exclude their land from the district. Yet the evidence points in the opposite direction. The Bridgeford Act of 1897 made it easier to have one's land excluded from the district.

The wide voting margins in favor of district organization also cast doubt on the Smith hypothesis. In the 46 districts for which voting data was available, the average percentage "yes" vote was 92.2 percent. (Tables 2a and 2b.) Although not conclusive, this finding does not indicate any strong potential for intra-district income distribution; rather, it suggests a nearly unanimous expectation of some net gain to those within the district. A more rigorous test would compare the actual vote with the vote that would be predicted by the Smith model. Using assessment data gathered from the Consolidated Irrigation District, such a test is performed in the paper. Again, the results do not support Smith's redistribution theory.

A further implication of this redistribution argument is that districts will set water prices below marginal cost. Using cost data from 1929, an estimate of the district cost function has been made to test this prediction. The evidence indicates that price was not significantly different from marginal cost.

This paper introduces a new hypothesis: a combination of agricultural, regulatory, and legal changes converged to increase the institutional advantages of the irrigation district relative to private irrigation companies. As large farm holdings were increasingly subdivided and settled during the late 19th and early 20th centuries, and with the imposition of water rate regulation in 1913, it became increasingly difficult for private irrigation companies (which were commonly joint land-water development enterprises that relied on land sales to turn a profit) to capture a sufficient amount of the benefits of new irrigation projects to make them profitable. (Table 3 records the drop in average farm size over this period.) With much of the land now settled, most of the benefits of new irrigation projects would have accrued to the new landowners. Hold-out problems would likely have prevented irrigation companies from buying out these landowners, completing the project, and then selling the land at higher values. An irrigation district, on the other hand, would have been able to cover its costs by taxing the beneficiaries.

It might be argued that private irrigation companies would have been able to capture the gains of new irrigation works by engaging is some form of price discrimination. However, even ignoring the problem of the high costs of determining the proper set of prices, price discrimination was not a possibility since it was disallowed by the State Railroad Commission (the predecessor of the Public Utilities Commission). The Railroad Commission was established in 1911 and two years later it was granted the authority to set water rates for commercial irrigation companies. Irrigation districts were not subject to this regulation.

The combination of these factors–the intensification of agriculture and water rate regulation–imply that private irrigation companies would find it increasingly difficult to capture the benefits of new irrigation projects. With increasing water scarcity over this period (from about 1900 to the mid 1920s), we would expect to observe projects with positive net benefits which private irrigation companies were unwilling to undertake. To test this claim, a comparison of the costs and benefits of specific irrigation projects are examined and presented in this paper. The preliminary evidence indicates the existence of several socially-beneficial irrigation projects which were developed only after the creation of an irrigation district. In general, the evidence suggests that irrigation projects were becoming increasingly profitable from about 1900 to the mid 1920s.

The irrigation district movement can be seen as an attempt to capture these untapped social gains. However, irrigation districts had their own difficulties: the massive failure of districts in the 1890s raised considerable uncertainty about the soundness of district bonds. In 1911 and 1913, however, important legislation was passed which was to greatly enhance the standing of district bonds. In particular, the creation of the Bond Certification Commission(BCC) in 1911 was to have important consequences for the ability of districts to sell bonds.

The BCC was composed of the state engineer, the state attorney general, and the state superintendent of banks. Before the directors of the irrigation district could call a bond election, they had to submit to the commission the district plans, along with cost estimates of any construction work of purchases and the amount of bonds to be issued. If the commission did not approve the bond issue, this would not legally prevent a district from calling a bond election or selling bonds, although an adverse report would effectively prevent the sale of bonds.

The BCC was required to carry out a detailed investigation into the legal, financial, and engineering aspects of the project to be funded by the bond issue. Authorized bonds could not exceed 60% of the aggregate market value of lands in the district. Once the bond issue was approved by the commission, the district was not allowed to spend the proceeds on plans other than the plans originally submitted to the commission (unless commission approval was received).

How are we to interpret the role of the BCC? The well-known principal-agent problem associated with debt finance gives us a clue about a possible interpretation. Highly-leveraged firms have incentives to engage in a greater-than-optimal level of risky activities. Anticipating this, bondholders have incentives to impose restrictions on the use of the borrowed funds ("bond covenants") and to monitor the behavior of the firm. With high monitoring costs, bondholders would require a large discount to induce them to purchase the bond. In the case of the irrigation district, it is likely that the BCC could monitor and control district activities at a lower cost than the bondholders themselves (if we consider monitoring activity a public good, then bondholders had incentives to free ride on monitoring activities of other bondholders). If it is, in fact, true that the BCC had lower monitoring and enforcement costs, then a district could lower its capital costs by seeking certification from the BCC.

Abundant evidence is presented in the paper which demonstrates the significant impact that bond certification had on the marketability of district bonds. A comparison between district bond rates and several "benchmark" rates indicates a substantial drop in district rates as a result of certification by the BCC. There was a substantial increase in district bond sales in the years following the establishment of the BCC. Finally, evidence from Moody's bond ratings provides yet another indication of the improved marketability of district bonds as a result of certification.