The Economics and Politics of Irrigation Projects on Indian Reservations, 1900-1940

Leonard A. Carlson, Emory University

In the West, water resources are critical for the development of agriculture and many Indian reservations are in areas with less than 20" of rainfall. This created a problem for the Bureau of Indian Affairs, since the task given it by Congress was to promote farming by Indians and many reservations were too dry for settled agriculture. In order to promote Indian farming, the BIA in the late nineteenth century began a series of irrigation projects on Indian reservations. One reason Congress emphasized programs to help Indians to become farmers was that agriculture was often the best available occupation on many reservations. Also, Congressmen and reformers in the 1880s and later thought that individual farming was the best way to assimilate Indians into white society. Despite the stated goal of promoting farming, however, Indian farming declined in the twentieth century. Irrigation projects also typically failed to yield as many benefits for Indians as expected. The failure of Indians irrigation projects is thus just one example of the broader failure of Indian policy to assist in the economic development of Indian reservations. In this paper, I model the behavior of the BIA to understand the BIA's irrigation program, and, more broadly, to understand the relationship of Indians with the federal government.

Congress first appropriated funds for an irrigation project on an Indian reservation in 1867 and made regular annual appropriations for irrigation projects on Indian reservations after 1893. Before 1902, no one had questioned the right of the BIA to the water needed to develop resources on Indian reservations. But Congress created a rival for the BIA when it passed the National Reclamation Act of 1902 (Newlands Act). The new agency created by the Act, the Bureau of Reclamation, potentially could appropriate water needed to develop Indian lands. This placed the BIA in the position of protecting Indians water rights. One strategy of the BIA was to construct irrigation projects on Indian reservations, which protected Indian rights under the doctrine of prior appropriation and beneficial use, discussed below. Over time, Congress changed the law so that Indian irrigation projects were administered under laws parallel to the National Reclamation Act. Under the National Reclamation Act land owners on a project were (at least in principle) expected to pay fees that would pay for the construction and maintenance costs of projects. This involved a substantial subsidy, since these costs did not include interest charges. In practice, farmers often avoided even paying all of these costs. (Mayhew and Delworth, 1994) In 1914, Congress amended the law governing Indian irrigation projects so that Indians were also liable for construction and maintenance costs. This change made Indians and Indian tribes liable for accrued charges from earlier projects. Reformers in the 1920s criticized this feature of the law, since many of these pre-1914 projects were built without tribal constant and with the understanding that costs would be paid by the federal government.

In the 1920's, projects on Indian reservations came under increasing criticism. The influential Meriam Report, inspired by the growing Indian reform movement in the 1920's, was highly critical of irrigation projects. However, the Meriam Report also stated that their staff lacked the technical expertise to fully investigate irrigation projects. In 1927, engineers Porter J. Preston and Charles Engle traveled to seventeen Indian irrigation projects in the west and their report is the first systematic investigation of Indian irrigation programs ever done.

What were the goals of the Bureau of Indian Affairs in developing the irrigation projects? The legal basis for the BIA's control over the lives and property of Indians was that the Office of Indian Affairs was acting as a guardian for the interest of Indians. One goal was to protect Indian water rights. As mentioned above, this led the BIA to lease Indian land to whites where Indians were not yet ready to use the water themselves. The BIA was establishing Indian water rights through the legal doctrine of prior appropriation and beneficial use. This legal principle holds that the first user has a right to water relative to later users, and it is the basis of water law in western states. After 1908, however, Indians had another legal claim to water. In the case of Winter v. United States the Supreme Court held in 1908 that in creating Indian reservations, Congress had guaranteed Indians a right to `sufficient' water, even where Indians were not currently using the water. In the first two decades after the decision, however, no one knew exactly how much protection this gave Indians, and the BIA continued to use the doctrine of prior appropriation. Thus, Indian water rights were unquestionably protected if the water was used, even if it was used by a white tenant. The second goal of the irrigation program was to make the projects self-funding. The third goal was to promote the independent agriculture among Indians, which was the general goal of all Indian agricultural policy in this period.

The problem was that the first and second goals, utilizing the land on the irrigation projects to establish Indian water rights and generate revenue, was in conflict with the third goal, encouraging Indians to be farmers. The easiest way to be sure that someone farmed the irrigated land on a project was to lease it to white farmers, who had more capital, both physical and human, than did Indians. But, as I have argued extensively elsewhere, the leasing of land by Indians often caused a decline in farming by Indians. Since the BIA wanted to encourage Indians to farm their own land, leasing land to whites was a contradictory policy. The size of the allotments also proved to be a difficulty. For example, if an Indian family allotted 40 acres under the General Allotment Act could only use 10 acres, the remaining 30 acres would often be leased to non-Indians. This was wasteful, since some of the funds used to irrigate Indian lands might have been better used to aid Indian farmers (by offering loans or training), rather than irrigate lands Indians could not use. The leasing of allotments also encouraged a decline in Indian farming(Carlson, 1981).

My model of Indian policy starts with four major interest groups. The first is Congress-at-large. Most Congressmen had little direct interest in Indian Affairs, since the tax burden was small and few of their constituents had any direct interests in the outcome of Indian policy. Thus, indulging in a form of paternalism or altruism toward Indians was a relatively low cost activity for most members of Congress. But it also meant that they tended to establish vague goals that were hard to implement and monitor. The second interest group is white farmers who live near Indian reservations and the member of Congress who represented them. These farmers often wanted to acquire Indian land and receive subsidized water. Public choice theory shows how such small groups will organize to influence policy and capture rents. The third group is bureaucrats within the Bureau of Indian Affairs, who would seek to advance their career's subject to the imperfect monitoring provided by Congress at large by expanding their authorized expenditure. The last, and perhaps weakest group is Indian tribes. In many cases individual Indians could only sell or lease their land with the consent of the BIA and tribal lands were also subject to BIA supervision. This greatly limited the ability of many tribes to direct their own economic activities. Further, Indians typically did not have much influence with local legislators since their voting strength was diluted and, in some cases, Indians had restricted voting rights. In negotiating with the BIA, many tribes were represented by ad hoc councils formed by the agent. Nonetheless, sometimes the BIA needed consent from tribal representatives to undertake some actions and this gave Indians a voice in BIA decisions.

The explanations already discussed are powerful, but they miss part of the story. Simply put, the missing piece is why the administration of irrigation policy was so wasteful from any one's point of view. In a world of scarce resources, it always makes sense to economize. Thus, it is always better to have an efficient institution to distribute the benefits of federal programs to different groups, rather than to simply dissipate the resources.

In a recent paper, Paul David concludes that new institutions are often adaptations of previously existing forms to serve new purposes. The result is institutions that are `serviceable but inelegant.' The reason that existing institutions are adapted to new uses is that institutions are imbedded in a system with network externalities and learning effects. The metaphor that David uses is of the "panda's thumb," based on the metaphor for evolution used by biologist Stephen Gould.1 According to David, "...the `panda's thumb' metaphor offers institutional economics the paradigm of a serviceable, but inelegant, resultant of a path dependent process of evolutionary improvisation, a structure whose obvious limitations stem from its remote accidental origins."(David, 1992, p. 22)

Here I use the "Panda's thumb" metaphor in a somewhat different way. Congress passed laws for Indians on the premise that Indian tribes and Indian peoples need special institutions and laws to protect their interests. But, to do this, Congress almost inevitably adapted existing institutions to serve a new purpose, since Indians are a small minority interacting with the rest of society. Thus, laws governing Indians had to be consistent with laws governing whites, while at the same time trying to serve a special purpose. The result is that Indian policy is littered with institutions that are "contrivances." One example is the awkward and wasteful form of land tenure given to Indians by the General Allotment Act. Another is the special water rights of Indians recognized by the Winters Doctrine which Courts have struggled for years to reconcile non-Indian water rights.

Descriptively this model of interest groups jockeying for benefits plus a waste factor due to the contrived nature of Indian institutions fits Indian irrigation programs in the 1920s. The source of the data is the under-utilized survey of major Indian irrigation projects supervised by Preston and Engle in 1927.

From an interest group perspective, the easiest sorts of projects to under stand are those which benefit white farmers and ranchers. It is easy to understand why federal funds would be spent on projects beneficial to white land users. White farmers and ranchers had already organized to capture rents from Bureau of Reclamation projects. It is easy to imagine how BIA Officials would come to administer irrigation Indian projects so as to serve the interests of white farmers, thereby gaining the support of local members of Congress. Leasing the lands to whites also meant that more project lands were used and not idle, which helped placate members of Congress concerned with waste.

The largest irrigation project was on the Yakima Reservation in Washington. The Yakima project fits the example of an irrigation project that primarily served the interests of white farmers who moved on to the reservation to take advantage of subsidized water. By 1927, the BIA had spent $3,698,994 on irrigation programs at Yakima. In that year, Indians farmed 6,250 acres out of 89,000 irrigated acres on the reservation, with the rest farmed by white lessors or owners. Who benefited from this project? White farmers benefited in using subsidized irrigated lands. Irrigation project officials benefited from the collection of fees from water users and in being able to demonstrate to Congress that project lands were used productively. Indians also benefited from the voluntary lease or sale of their lands to whites.

But this is not necessarily what Indians might have wanted nor what a guardian interested in the interests of Indian should have done. Jim Fitch concluded in his study of the Yakima reservation that the Yakima were far better equipped to be ranchers or dry farmers than to farm irrigated lands. Thus, while the Yakima benefited from the project, the BIA might have better served the interests of the Yakima by letting the land be used for grazing and dry farming; or, at least, building smaller irrigation projects. Indeed, according to Preston and Engle, lease payments made to the Yakima were sufficiently large that the Yakima could engage in casual day labor and refrain from agriculture altogether. (Preston and Engle, pp. 2655-6) If the goal of federal policy-makers was to encourage the Yakima to farm their own land, the policy of irrigating the Yakima was at best a limited success.

The leading historian of Indian policy, Francis Prucha, reached a similar conclusion about irrigation projects on Indian reservations. According to Prucha, "It became increasingly clear throughout the 1920's that irrigation projects were authorized for Indian reservations not with the primary intent of aiding the Indian to advance toward self-sufficiency, but to develop the arid west for the benefit of white interests. And, in fact, the whites were the ones who most profited from irrigation." (Prucha, Volume II, p. 894)

It is easy to see why Prucha reached such a pessimistic conclusion. Out of a total of 362,818 acres of irrigated land on Indian reservations, only 131,427 acres were farmed by Indians. A total of 113,402 acres was farmed by whites leasing Indian lands and 131,427 acres used by white owners. If the goal, was to promote Indian farming, other programs might have served Indians better.

A related problem was that often the BIA did little to train Indians to farm irrigated lands. Preston and Engle complained that "With a peculiar idea of economy, we expend some millions of dollars in building an irrigation plant to enable Indians to support themselves by farming, and employ a man called a `farmer,' but who, in reality, is a chore boy or field clerk at a salary of $1500 a year, to instruct the Indians on how to use this intricate and complicated plant." (Preston and Engle, p. 221) Teaching Indians to farm was a long term goal proposition that yielded few immediate benefits that the BIA could show to Congress. In addition, such programs were not in the interest of white farmers. Why bother?

There is more to the story, however, than a cozy relationship between the BIA and white landowners at the expense of taxpayers and Indians. Some projects provided very little benefits for whites and primarily served Indians living on reservations so far from markets that no white farmers or ranchers wanted to buy or lease Indian land. It appears likely that the BIA built these projects knowing that most of the benefits from the project would go to the Indians living there and, of course, those whites who built and administered the projects. Many of these were smaller projects targeted to the needs to Indians. Indeed, Indians farmed 46,199 acres out of 55,375 acres on the projects deemed too small for Preston and Engle to visit. Other examples of projects that largely benefited Indian land users include the Walker River project and the Carson Sinks allotments in Nevada.

Finally, there are a third class of project that are notable because they were simply so wasteful that no one benefited very much. In the extreme, Preston and Engle recommended the closure of three of the 17 projects they visited. In their opinion, there was simply too little water to justify further expenditure to develop those three projects. This is not to say that no one benefited. Administrators, builders or some users must have benefited, or BIA would never have built them. One of the three projects, Tongue River, had already been shut down by mutual consent by the time of Preston and Engle's visit in 1927. However, the other two projects that they recommended for closure, Blackfeet and Fort Peck, continued to operate. Indeed, the fact that the Preston and Engle Report was buried in the Congressional Record and never explicitly acted upon suggests that some in Congress did not care for the advisors recommendations.

Even where there was enough water, agents were willing to justify extending irrigation projects beyond their useful radius. According to the Preston-Engle report,

Particularly, there is a tendency on the part of some of the irrigation employees to extend `irrigation works' over areas of land where there is no need for them, and apparently without regard to whether or not there is a water supply for such land. . . These so-called irrigation works sometimes consist of the mere skeleton - a furrow or a ditch - but they justify a request for increased appropriations `for extending the irrigation system to the lands of Indians who are anxious to farm,' and after extensions are made, they form a justification for increased appropriations for operation and maintenance. (Preston and Engle, p. 2316)

What is the unifying theme? In designing a program to build irrigation projects on Indian reservations, Congress adapted laws designed for other purposes. The resultant bureaucratic system was not really efficient for the task at hand. Given such a "contrivance," I conclude that : 1) bureaucratic entrepreneurs within the BIA appear to have tried to build as many projects as possible in order to justify bigger budgets and promotions. Indeed, most large reservations in the regions with less than 20" of rainfall were the site of major projects. 2) It was best to have allies and to have the land used. At those locations where white farmers and ranchers were in a position to enter the reservation, white farmers owned or leased most irrigated land. This may have served the interests of Indians in protecting their water rights, and giving them higher lease incomes, but this was a by product. 3) But in areas where there was little interest by whites, Indian farmers did use most of the project lands. 4) Few of the projects appear to have been justified by any sort of cost benefit criteria and like other reclamation projects of the period, there were transfers from the general tax base to western water users. But someone must have benefited from even the poorest designed projects.

The contrived nature of laws shaping Indian policy may also help explain other puzzles in Indian policy. One example discussed in the larger paper is the form of tribal governments established by the Indian reorganization Act of 1934.


Carlson, Leonard A., (1981) "Land Allotment and the Decline of American Indian Farming," Explorations in Economic History, vol. 18. no 1 (April 1981), pp. 128-154.

David, Paul A.,(1992) "Why Are Institutions the "Carriers of History," working paper presented to Stanford Institute for Theoretical Economics, July 1992, revised version, October 1992.

Fitch, James B., (1974), "Economic Development in a Minority Enclave: The Case of the Yakima Indian Nation, Washington," Ph.D. dissertation, Stanford University, May, 1974.

Gould, Stephen Jay (1980), The Panda's Thumb: More Reflections on Natural History, New York, W.W. Norton, 1980.

Mayhew, Stewart, and Gardner, B. Delworth, (1994) "The Political Economy of Early Federal Reclamation in the West," in Terry L. Anderson and Peter J. Hill (eds), The Political Economy of the American West, Lanham, Md, Rowan and Littlefield Publishers.

Preston, Porter J. and Engle, Charles A., "Report of the Advisors on Irrigation on Indian Reservations," Junes 8, 1928, printed in Survey of Conditions of the Indians in the United States," Hearings before a Subcommittee of the Committee of Indians Affairs, United States Senate , 2d Session (1930), part 6, pp. 2210-2661.

Prucha, Francis Paul, (1984, vol. 2) The Great Father: The United States Government and the American Indians, Vol. 2, Lincoln, The University of Nebraska Press, 1984.


1The panda has evolved a crude sixth "finger" out of its wrist bone to serve as a thumb to allow it to eat bamboo, since the usual thumb has evolved in bears for running and climbing and is no longer suited to the task of eating bamboo.