Independent Economic Analysis Board:
Emery N. Castle, Chair
Joel R. Hamilton, Vice-chair
Kenneth C. Boire
Daniel D. Huppert
Lon L. Peters
Jack A. Richards
Anthony D. Scott
Paul C. Sorensen
University of Idaho Center for Business Development and Research
College of Business and Economics
Principal Authors: Steven Peterson and Lawrence H. Merk
Reviewed by Paul Sorensen, BST Associates
The Economic Impact of Steelhead Fishing and the Return of Salmon Fishing in Idaho
prepared by Don Reading, Consulting Economist
Ben Johnson Associates
Reviewed by Anthony Scott, University of British Columbia
1. General statement
2. Critique of the Tri-Port Study
3. Critique of the Economic Impact of Steelhead Fishing Study
1. General statement.
This report summarizes the reviews by members of the IEAB of the Tri-Port Economic Impact Study and the Economic Impact of Steelhead Fishing and the Return of Salmon Fishing in Idaho. These two reports provide glimpses at what is at stake in future decisions about the water management allocation of the Snake River. Our review starts with a comment on the place of such studies in the framework for the decisions that face the Council. Then we provide some information about the special assumptions and results of each of the studies, judged primarily from their local point of view.
1.2. Framework for Decisions.
Local economic impact studies, such as the Tri-Port Economic Impact Study and the Economic Impact of Steelhead Fishing and the Return of Salmon Fishing in Idaho, can provide useful comparisons of the relative importance of specific economic sectors within the local economy. However, while there is much useful information in each study, these types of studies do not provide decision makers with the information needed to evaluate system modifications for fish and wildlife mitigation or enhancement, or more specifically, dam removal.
Local economic studies can provide some measure of the "impact" on local business turnover and jobs of opportunities such as a port or a sports fishery. When everything else remains the same, the more port business, or the more anglers, the greater the measured impact. In this limited sense, calculation of the impact does provide a verifiable numerical estimate of the value of keeping the navigation and the angling opportunity. Furthermore, it is not unlikely that when the resource opportunity is changed by some act of river management, the local "impact" will change proportionately.
Calculating the impact can be useful to the locality. For example, if a community was trying to decide whether to spend tax money to improve the facilities for steelhead anglers or to improve a park for campers, it would be useful to know which kind of spending would have the greater impact on local taxpayers. Members of the IEAB, who have themselves participated in making such estimates, believe that these two reports are superior to many. In particular, the authors have, with great ingenuity, overcome some of the data deficiencies that can plague local impact studies. But data problems are not the main deficiency of local impact studies from the IEAB’s point of view.
The question is whether such a calculation has value for decision making when the economic benefits from alternative projects or policies are scattered across the whole Columbia/Snake River basin, or, indeed, the whole nation.
The impacts being calculated here are intended to measure the benefits from a particular set of Snake River levels and flows, and thus, basically, from a single selection among sets of proposed dam modifications or retirements. That is, the impacts were evaluated at the specific time that the surveys, on which they were based, were undertaken. Thus, to judge these calculations, we should consider whether the estimates can even be relevant to the selection procedures developed in the IEAB-reviewed Council staff report on Methods of Economic Analysis.
First. "With" and "without" project considerations
Economic selection must consider the benefits or costs in the locality "with" and "without" the "Natural River" or other proposals. The two studies attempt only the "with" calculation. They scarcely hint at how the various localities would make out "without." What would a later survey show, with different levels and flows? To what extent would local individuals and firms respond profitably? The implication of the two studies’ impact calculations is that there would be no response: the entire impact as calculated "with" the project would be lost. This means that even from a local point of view, the two calculations could over-estimate the predicted loss of impact in changing from with to without. Indeed, it is theoretically possible that the loss of "impact" from a change could be close to zero. A zero loss would result if all the economic actors in the locality made a smooth dynamic response and eventually undertook equally productive activities. For example, with a dam drawdown, Tri-port boat builders could well find an expanded market.
The main point here is that a single snap-shot of one situation tells the Council next to nothing about what would be lost in changing to another management strategy. What is needed is a measure of the loss or change.
Second. Impact change versus income change.
Even from a local point of view the best measure of the benefit from a change in a set of river levels and flows is the change in local income. The economic impact of a river change includes spending and re-spending on inputs and can rise or fall whether or not local incomes rise. For example, when the price of fuel rises, truckers and guides must charge more. Their higher charges show up as an increase in the "impact "of port activity and the "impact" of steelhead fishing. But the actual incomes in these activities will remain about the same. The Council will find approximations to calculations of income changes stemming from changes in the management of major parts of the Columbia basin (i.e., studies surveyed in the IEAB’s Task 2 report). The IEAB’s report suggests the need for more such studies, each linked to particular proposed changes in projects.
Third. The Council’s point of view versus the local point of view.
If the local studies reviewed had calculated changes in income stemming from a new river regime, they would have real value for local decision-making. They would also be valuable from the Councils’ river-oriented point of view, especially if such calculations were available for most localities. For the whole basin, the net costs of a change in river management are the total of the income decreases in the places that are hurt (i.e., where economic activity declines), minus the increases in places where there is an improvement (i.e., where there is an increase in economic activity).
However, most of the Council’s changes in river management must be judged by their cost effectiveness or perhaps net benefit as seen from a national economic point of view. This point of view requires that significant income and welfare changes for individuals, businesses and communities outside the basin also be taken into account. These could be very numerous. In the case of river changes in the Tri-port region and the steelhead fishery, they may include cost or income changes for users of electricity, users of transportation services, tourists, sports fishermen, plus changes for outside suppliers of inputs and equipment for these activities. It is worth noting that where markets for these things are not perfect, for some river projects the gains or losses outside the river region could actually offset those experienced locally.
2. Critique of the Tri-Port Study
With this overview in mind, we would first like to briefly comment on the chief strengths and weaknesses of the Tri-Port Economic Impact Study, which was sponsored by the Ports of Clarkston, Lewiston, Whitman County and the NPPC.
2.1. Strengths and weaknesses of the Tri-Port Study
The authors provide a useful assessment of the existing trends that affect the tri-ports’ area economy and the impact from major industries. The following observations are intended to pinpoint the report’s strengths and weaknesses.
2.1.1. Survey Methodology and Process
Conducting a mailed survey with follow-up telephone calls is a legitimate process to obtain the variables required to undertake an economic impact report. However, the results of the individual surveys are considered "confidential", which makes it difficult to verify direct employment calculations.
In addition, there is little explanation of how survey results were categorized and classified, which raises a concern about double counting. For example, many jobs are included in business services, boat repair, and construction, among other sub-sectors. These jobs may be indirectly related to manufacturing, water transportation or other sectors, in which case they should not be included in the direct employment calculations. Rather, they should be included in the indirect employment calculations. The discussion of methodology does not investigate categorization/classification issues in sufficient detail.
The report presented some very useful insights into the definition of sub-regions of the economy. The linkage between expenditures in the economy is directly related to the location of major retail/service centers. The authors present a good description of attraction centers and how this influenced creation of the community-based input-output models.
The multipliers seem to be reasonable (+/- 2 for total impacts) but, there are different community-based models used for each port region. Because the community-based regions are defined differently (some are larger regions), a job in Whitman County may create more total impact than a job in Lewiston simply because it is being judged in a larger region.
2.1.2. Water Dependency Considerations
An attempt to determine the dependency of firms to water transportation is considered in the report. However, the linkage to dependence on water transportation is not explicitly stated. Rather, it is implied in an indirect manner. The central questions that should be asked (and answered) are:
- How much of the local economy is dependent on water transportation at the present time and after implementation of the Natural River alternative?
- What is the net impact on the community of the services provided by the dams?
This report was not intended to address the net loss (or gain) of jobs with and without the dams. As a result, it does not address some of the central issues that the NPPC and IEAB is concerned with.
The reasons why firms value water-dependence from both the transportation and aesthetic points of view are briefly considered. However, it is a much more difficult issue to evaluate how much dependency exists and how firms would actually respond to the lack of water transportation, recreation, aesthetics or other amenities provided by the dams.
In the case of Potlatch, the authors state that they would not cease business due to the loss of barging because the company would divert the shipments to truck and rail. No detailed analysis is presented to illustrate how grain shippers would respond. This discussion and the consequences of the shift (i.e., factors such as higher rail rates, potential car shortages and other impacts) are alluded to but not sufficiently addressed or quantified.
In addition, some "water-dependent" firms may have an improvement in sales without the dams (e.g., some manufacturers of water craft may sell more boats with white water conditions).
There is an implication that there will be a net loss of jobs (i.e., especially water dependent jobs and potentially other jobs) as a result of the loss of the dams. However, the report does not really evaluate the possible economic scenarios if the dams are removed. As a result, discussion of net impacts after dam removal are not properly defined or verified.
2.1.3. Consideration of Dam Removal Impacts
The weakest portion of the report was the consideration of impacts from potential drawdowns in the final section of the report. Considerations appear at the end of this section which have different levels of justification, including:
- One useful observation is the differentiation between the average wage and multipliers in manufacturing versus tourism sectors. Manufacturing has a much larger average wage and higher multipliers than tourism-related jobs. This difference is well documented but must be tempered by the relative increase in retail/service sector jobs compared with those in the agriculture/manufacturing sectors.
- The authors present an estimate of the impacts of the "Natural River Option" but these are not properly documented:
- loss of $35.6 million in income from water transportation (it is unclear whether this is a net loss or whether there will be a diversion to other forms of transportation which would mitigate or outweigh the loss),
- loss of $81.3 million to port industrial parks and economic development (it is unclear how some firms that are non-water dependent would be impacted by dam removal)
- loss of 1,580 jobs in water transportation and 3,249 jobs in industrial parks and economic development (again, it is unclear how this impact is calculated and there is no justification)
- several legitimate considerations are raised but not discussed in sufficient detail, such as: potentially increased rail rates if capacity is unavailable, additional capital costs to ship grain efficiently from river elevators by rail, additional road maintenance costs for grain to elevator locations, aesthetic concerns about the appearance of Lewiston without the Lower Granite Dam, loss of economic competitiveness without water transportation options for both grain growers and Potlatch Corporation, concern that the improvements would take substantial sums of dollars that may not be available, potential for environmental damage from taking out the dams.
As stated above, these impacts are discussed but not fully developed. Hence, it is difficult to relate the economic impact estimates to the potential impacts of the Natural River option.
2.2. Overall Review of the Tri-Port Economic Impact Study
This report was intended to provide an economic impact snapshot of the tri-port region. It was not intended to address the economic impact of the region under existing conditions versus those that might occur with the Natural River option. It is useful to provide an insight into the importance of the ports to the local region. However, it does not address the more difficult questions about how the community will change in the future as a result of potential water budget allocations.
Unfortunately, the report has been used in a manner that is unwarranted. To some degree, this mis-representation was engendered by the report’s authors discussing potential impacts without fully developing (and quantifying) the changes in the communities. In any event, the report does not provide decision-makers with the level of information that is required to address this difficult issue.
3. Critique of the Economic Impact of Steelhead Fishing Study
3.1. Purpose, Surveys and Data.
The study was commissioned by the Idaho Fish & Wildlife Association. The Association is chiefly interested in showing that the fishery is economically worth maintaining at its present level. To this end Don Reading has estimated the "economic value" or impact of steelhead by a calculation of the payments and jobs stemming from the spending of steelhead anglers.
The study uses a 1992-3 Idaho Fish and Game survey of anglers’ (a) steelhead catches and (b) spending on food, transport, guides etc. It also uses two sets of input-output models of the economic impacts of spending for river communities (a) in Northcentral Idaho and (b) upper Salmon region. These models, developed by Robison and associates, were also used in the Tri-Port study. Ingeniously combined by Reading, they report direct and indirect impacts almost mile by mile, town by town, along the Snake River and in the rest of Idaho.
3.2. Methods and results.
Reading first assigned the anglers’ spending to the nearest community/city. Then, using non-Idaho data, he made two adjustments. First, he added to the expenditures reported above a guess about the anglers’ "hard’ or equipment expenditures. Second, he assigned a percentage share of each expenditure to the larger cities in the state as a whole rather than to the community nearest the reach where the angler had fished/camped.
Next he used the input-output models to estimate the further spending of the original direct expenditures. For example his data showed that anglers had spent $ 4.5 million in the thirteen steelhead regional cities. Using an I-O model, he estimated that the spending and re-spending of such outlays had resulted in a further $8.8 million impact in those same cities.
For the state as a whole, $35 million in anglers direct spending had multiplied to a further $100 m impact on the state economy. Reading also estimated the total of direct spending and further impact as creating almost 2,700 jobs. About ten percent of the jobs were located in the angling region; and the rest elsewhere in the state.
The author also uses the steelhead results as the basis for a rough estimate of the community impact of a possible revived recreational salmon fishery in the Snake region. State-wide, the salmon fishermen might create about 1800 more jobs, with, again, ten percent in the riverside communities..
3.3. Criticisms of the steelhead study.
It can be argued that such small-community estimates as in this study are better than whole-state estimates, because state lines cut across natural market areas, and across river-basins. To meet this small-community need, Reading has been resourceful. His adaptation of the state-of-the-art Robison input-output model means that from many points of view it would be difficult to produce better estimates.
So much local and community detail means that assumptions and short cuts are unavoidable. IEAB members feel that if some of the flaws criticized were dealt with, this study could be of even greater local interest, and also moderately helpful to the Council. It must be noted that the report’s brevity prevented IEAB members from understanding all the assumptions and from confidently tracking the calculations. So several of the following points might turn out to be over-severe.
3.3.1. Idaho community estimates.
Unfortunately Reading’s Idaho geographical estimates could be substantially in error. For example, the study predicts that anglers fishing steelhead nearby must have spent about $100,000 in Grangeville. For this single allocation, a wide range of error would not be surprising. From this sum, the study goes on to predict how and where the supposed Grangeville recipients will spend on "inputs." As the underlying linear models have never been verified, these further geographical predictions could be wildly wrong. In short, Reading’s local estimates are much shakier than is his estimate of the impact of steelhead angler’s spending on the state or region as a whole.
3.3.2. In-state versus out-of-state impacts.
The study’s methods seem not to handle correctly the anglers’ impact on spending that crosses state borders. Long tables are presented, but do not clarify these three questions.
a. Original spending. In the Reading study, the anglers’ original expenditures are attributed either to on-river communities or to the rest of Idaho. It is not explained why some of this spending, for car travel for example, was not made in other states. The answer may be that in the original Idaho questionnaire, anglers were told to mention only their in-Idaho spending. If so, the reliability of their responses needs to be touched on.
b. Indirect job impacts. The study’s input-output modeling led to an estimate of secondary or indirect impacts throughout Idaho. Again, it is not made clear why the impacts shown ( sales of inputs to the original suppliers to the anglers) were only within the state. One would imagine that Idaho eating-places and lodging places must be buying some current supplies from sources out of state. The answer is probably that the underlying input-output models did measure this "leakage" of local spending to other states and regions. But the subject is not dealt with here.
c. Out-of-state visitors. The data used by the Reading study apparently do not disclose whether the anglers were from Idaho or out of state. To the extent they were from Idaho, the study’s estimate of the total dollar or job impact could be a serious exaggeration. To see why, imagine that the steelhead fishery was closed. Then standard economic and business reasoning asks how the Idaho fishermen will "adjust" to this lost recreational opportunity. The likely answer distinguishes between Idaho and other anglers.
Many from Idaho will find something else to do close to home, and will spend money on it --- perhaps even more than they had spent on steelhead fishing. The final job and spending impact of closing the fishery to them may be small or negative. The rest of the anglers from Idaho will now go fishing and spending in other states. The impact of closing the fishery to them will be roughly as suggested in the Reading study. Similarly, anglers from other states will probably now not visit Idaho at all. The impact of closing the fishery to them will also be roughly as in the Reading study. To summarize: if most steelhead anglers are from out of state, the Reading estimates may be relevant. But if most are from Idaho, the economic impact on the whole state of opening or closing the steelhead fishery could be small, even trivial. Only the river communities might suffer.
d. Summary. If much of the anglers’ direct and indirect spending is for goods and services procured outside Idaho, Reading may have over-stated the impact of steelhead fishing. One hopes the study took care of these problems in a professional fashion. If many of the anglers are from Idaho, and would stay there whatever the fishing situation, Reading may again have over-stated the long-run impact of steelhead fishing. One suspects that the study did not deal with this source of error.
3.3.3. Expenditures on boats and major equipment.
The study apparently assumes that the guessed-at purchases of boats and major equipment were spending caused by the anglers’ fishing activities. Since the number used is only a rounded add-on to other spending, it is hard to criticize. But it should be noted that people in Idaho make such expenditures in order to go camping, hunting, and fishing in other places. In other words, even if Reading’s guessed value of steelhead anglers’ boat spending is right, that value may over-state the amount that should be credited to the impact of steelhead fishing.
3.3.4. The regional economy.
The Reading study, having defined the angler-derived jobs as benefits, goes on to sketch how the Snake region’s other industries are declining. The implication is that steelhead dollars and jobs are especially valuable in that they can replace farming and logging in the local economy. This seems reasonable, but depends on two implicit assumptions.
(a) The angler spending and its re-spending goes to former farmers and loggers, not to newcomers who are attracted by the recreation industry. Nothing is said on this matter.
(b) The anglers are from out-of-state. The old community industries were exporters and pulled in income from outside. If the new anglers are local people, their spending does little to offset a fall in exports. Their income is just transferred to other Idahoans. This question, partly technical, is passed over. Of course, even if all the steelhead anglers are from Idaho, the river communities will welcome these transfers.
3.3.5. River-management choices.
Unlike the references to dam drawdowns in the Tri-port study, the steelhead study says nothing to relate the size of the steelhead catch to dam retention or retirement, or to other changes in levels and flows. Nor does it address the question of increasing or decreasing hatchery investment.
The same is true of the study’s calculations for a restored salmon fishery (discussed below). Nothing is said about the changes in the downstream fishery, hatcheries, dams, habitat or flows that would give the greatest restored-salmon fishery economic impact, or even whether these would be compatible with a continued steelhead fishery.
3.3.6. A restored salmon fishery.
An attractive goal of the Reading study is its attempt to estimate the jobs and dollars that would follow the restoration of a harvestable salmon run. Chinook is taken as an example.
22.214.171.124. Data and Method.
Reading assumes that the 1960s fishery, on which there is some data, is resumed. A catch of 23,000 chinook annually from the river banks involved 150,000 fishing days. He reasons that the direct and indirect economic impacts of a chinook angler’s day should resemble those of a 1990 steelhead angler.
Reading rightly refrains from in-depth explorations of this unresearched salmon fishery. Where would it be? Could it employ the same people and capacity as the steelhead (hotels, guiding, boats ?) Assuming that the two fisheries would be independent, the salmon jobs created would turn out to be about two-thirds of the steelhead jobs mentioned above (2,700) leading to 4,400 fishing jobs altogether in the state. Of these, about 10 percent would be in the regional river-side communities. It is fair to say that this calculation of the impact of a restored salmon fishery is, so far, just an easy, though suggestive, by-product of the steelhead study.