The Effectiveness and Costs of Hatcheries

The basic management flaw was (and is) in measuring hatchery success in “pounds of product” where product was defined as smolts released alive into the river. But the mission of hatcheries is to produce harvestable adults. Success in that endeavor depends not only upon the pounds of smolts, but also their quality, and how, when and where they are released.


Only rarely did fishery managers attempt to figure out whether hatcheries were succeeding in their basic mission. A series of cost-benefit studies conducted on fall chinook releases from 1961 to 1964 showed cost-benefit ratios ranging from 2:1 to 7.2:1.34 Another evaluation based on releases from 1978 to 1982 showed a cost-benefit ratio of 5.7:1.35 Average survival (or percentage of returning adults) had fallen from 0.7 percent to 0.33 percent, but no one paid much attention to that problem.


By the 1980s, assessing the effectiveness of hatcheries was all but taboo among fishery managers. When Reagan’s Office of Management and Budget began an effort to remove Mitchell Act hatchery funding through appropriations and substitute Bonneville Power Administration funding through surcharges on electricity (an effort started by the Nixon OMB in 1973), someone mailed our clients an anonymous assessment of the Mitchell Act hatcheries.


It reported that by the late 1980s funding for the Mitchell Act hatcheries was running in the $7-8 million range. The hatcheries were producing over 100 million smolts annually. From 1973 to 1987, the cost per pound of smolts produced ranged from $2.03 to $3.04. The report did not provide the cost per harvestable adult, but if 1% of the smolts had survived to harvest, the cost would have been $7-8 per fish—still a bargain by today’s standards.


There has been little oversight of Mitchell Act and most other hatchery spending. In 1996, Bonneville engaged a consultant to audit hatchery operations, producing comprehensive reports on many of the hatcheries in the Columbia River Basin. But the reports are marred by the lack of the most basic data needed to assess effectiveness; what data are available suggests that the cost per fish has risen substantially. By my calculation, we spent $51 per adult for Bonneville Hatchery fall chinook and $273 per fish for Rapid River Hatchery spring chinook for a few years of releases in the late 1980s; no data is available since then.36


Some of these hatcheries have had years in which none of their fish were recovered in fisheries at all. The worst upriver spring chinook hatcheries appear to be spending over a thousand dollars per harvested adult salmon. There is no government report, anywhere, which presents the cost hatcheries per returning adult salmon for all hatcheries in the Columbia River Basin.


Professor Ray Hilborn of the University of Washington recently analyzed the effectiveness, in terms of returning adults, of hatcheries in the Columbia River Basin. Since hatcheries often produce more than one salmon stock, he divided his analysis between fall chinook production and spring chinook production.


Professor Ray Hilborn's data, presented in Chapter 4, show the huge variation in the effectiveness of hatcheries. Some hatcheries are consistently more effective. For example, somebody should take a closer look at the Shuswap Hatchery on the Fraser River to see what its managers are doing right. Unfortunately, fishery managers have made no effort to distinguish the hatcheries that work from hatcheries that don’t, and the most inefficient Snake River hatcheries remain in operation.


Many Northwest salmon decisionmakers will admit, in private, that running failing hatcheries no longer makes economic sense. But they do nothing in public, because of the political imperative to produce salmon. The fishery managers have no incentive to make these hatcheries work. As hatchery opponent Bruce Brown pointed out in 1982, “[f]or government agencies with resource responsibility, allowing the degradation of wild salmon has been a self-fulfilling prophesy of greater hatchery production, expanding budgetary allocations and increased agency power”.37


By 1996, hatcheries have come under so much criticism that but for their pork barrel attributes, they probably would have failed to secure funding in Congress. The Clinton/Gore Administration has continued to try on several occasions to eliminate Mitchell Act funding, in favor of sticking the Bonneville Power Administration with the bill.


The biggest push in hatchery management right now is to move more production upriver. Driven by tribal demands and increased pressure from upriver interests on the Northwest Power Planning Council, BPA has been spending more and more money on hatcheries:



Figure 17: BPA Hatchery and Hatchery-Related Spending38
The fish produced at these hatcheries will be some of the most expensive animals ever raised in captivity. With this much money, we could start captive breeding programs for dozens of truly endangered species.


One of the reasons spending is rising so fast is that politicians are pursuing inherently conflicting objectives. In directing expansion of upriver hatcheries, Congress “emphasize[d] that only projects which protect, maintain or enhance biological diversity of existing wild salmon stocks should be pursued”.39 As the U.S. Army Corps of Engineers noted, Congress’ direction had an inherent “conflict”, in that to “initiate hatchery construction projects” would not “increase the biodiversity of the listed [endangered Snake River salmon] stocks”.40


The concept of cost benefit analysis has begun to lose all meaning, because no one can measure the benefits of making fish more “wild”. Instead, we assume that this is of incalculable value, and spend immense sums of money building hatcheries that are supposed to make "wild" fish. When the U.S. Army Corps of Engineers first proposed the Lower Snake River Compensation Plan, they reported that they expected to pay $46,000,000, with operation and maintenance costs of $3,000,000, for a benefit to cost ratio of two-to-one.41 Twenty-one years later, they reported having spent $215,696,000, with an estimate of $268,288,000 to complete the project, since the Corps has only completed five of nine hatcheries.42 Contrary to the Corps’ own regulations, its 1996 progress report does not even mention benefits—other than to report on the poundage of smolts produced at the hatcheries. There is no cost-benefit analysis whatsoever.


34 Id. at 385-86; see also J. Richards, “An Economic Evaluation of Columbia River Anadromous Fish Programs”, Ph.D. Thesis, Department of Agricultural Economics, Oregon State University, 1968, at 212, cited in C. Smith, Salmon Fishers of the Columbia 79 (showing cost-benefit ratio of only 1.6 to 1).

35 ISG, Return to the River 386.

36 Montgomery Watson, "Hatchery Evaluation Report: Bonneville Hatchery—Tule Fall Chinook ", USDOE Contract No. 95AC49468, May 1996 (using minimum program cost from  6 divided by total average returns from  5); Montgomery Watson, "Hatchery Evaluation Report: Rapid River Hatchery—Spring Chinook", USDOE Contract No. 95AC49468, May 1996 (same).

37 B. Brown, Mountain in the Clouds 231-32.

38 This data comes from BPA’s contract management computer database and reflects contract “obligations”; actual spending figures may vary slightly.

39 Conference Report for H.R. 4506, P.L. 103-316 (Aug. 26, 1994), quoted in “Interim Report: Lower Snake River Fish and Wildlife Compensation Plan, Lower Snake River, Washington and Idaho”, at 2 (USACE Walla Walla Dist. April 1996).

40 Interim Report: Lower Snake River Fish and Wildlife Compensation Plan, at 14.

41 “Special Report, Lower Snake River Fish and Wildlife Compensation Plan, Lower Snake River, Washington and Idaho”, Syllabus (USACE Walla Walla Dist. June 1975).

42 “Interim Report, Supplement to Special Report, Lower Snake River Compensation Plan, Lower Snake River, Washington and Idaho, June 1975”, at 26 (USACE Walla Walla Dist. April 1996).

Previous PageTable Of ContentsNext Page

This Web page was created using a Trial Version of HTML Transit 3.0.