In fact, the vote was so close--Wyden beat Smith by less than 1.5 percent--that almost anything can be and has been ascribed to the victory. Oregon is generally pro-choice, as Wyden clearly was while Smith was not. The candidates, especially Wyden, portrayed the election as a referendum on Clinton vs. Gingrich in the recent budget battle, so the points Clinton scored at the recent state-of-the-union address is thought to have given Wyden a few votes.
Some commentators say that voter reaction against negative advertising helped Wyden. Both candidates used negative ads, but three weeks before election day Wyden called off his negative ads and announced that he would only present positive ads from then on. Smith's ads continued to be negative. This may have tipped the balance in some voters minds, but no polls support this.
Oregon's vote-by-mail experiment probably played a much bigger role in helping Wyden. High voter participation usually benefits Democrats, and the vote by mail produced a 66 percent turnout--a record for a non-presidential election in Oregon, at least in recent decades.
Oregon's urban-rural split also influenced the election. Wyden, who represented Portland in Congress, was clearly the urban candidate, while Smith, who comes from eastern Oregon, represented ruralites. Wyden won 60 percent of the vote in the three counties in and around Portland, but just half the vote in the rest of the Willamette Valley--extending from liberal Eugene to conservative Salem. Smith took 61 percent of the rest of the state.
One more factor must be brought out that I haven't seen mentioned in other press reports: the presence of third parties on the ballot. Wyden and Smith actually shared the ballot with four "third party" candidates. Two of them, Socialist Party Vicki Valdez and Pacifica Party Lou Gold (a noted environmentalist), collected 15,000 votes. The other two, American Party (i.e., Perotistas) Karen Shilling and Libertarian Party Gene Nanni, collected 40,000 votes.
Final Count, Oregon U.S. Senate Election
Votes Percent Wyden 568,335 48.34 Smith 551,100 46.88 American 25,511 2.17 Libertarian 15,628 1.33 Socialist 7,826 0.67 Pacifica 7,187 0.61 Total 1,175,587 100.00
This means that a majority of voters selected parties or candidates who generally oppose big government and generally support trimming government spending. In other words, if the American and Libertarian party voters had gone for Smith, then Smith would have won the election even assuming that Socialist and Pacifica votes would have gone to Wyden.
Given all of these factors leaning toward Wyden, it is actually surprising that the election was so close. While Democrats may be heartened by this vote, the results should actually be chilling for environmentalists who support laws such as the Endangered Species Act. In the past, Oregon Republicans have usually won statewide office when they fielded moderate candidates such as Mark Hatfield or Bob Packwood. Conservative candidates such as Smith haven't come this close to winning in decades. It appears that even the Oregon electorate is getting more conservative.
Prognosis allows the user to enter variables to customize growth predictions for their forest. Using the wrong variables, however, can lead to wildly inaccurate predictions. This became a bone of contention during the Forest Service's planning process in the 1980s.
During my reviews of forest plans, I noticed serious discrepancies between growth predictions for the Idaho Panhandle and Clearwater national forests. These adjacent forests have similar plant communities with similar productive capacities. Inventory data indicated that similarities even extended to the average age and stocking levels for various parts of the forest.
Yet the growth predictions were wildly different. The Idaho Panhandle's plan was based on predictions that 200-year-old forests would double in volume in a few decades. One of the characteristics of older forests is that they do not grow as fast as they did when they were younger. But even if they did, a 200-year-old forest would require 200 years to double in volume. By comparison, the Clearwater's growth predictions were entirely reasonable.
Try as we might, we couldn't convince the Forest Service to change or even to publicly review the predictions. Even after Andy Stahl scrutinized the computer runs and found that some predicted that trees would grow more than 600 feet tall (the tallest trees in the world are only about 300 feet), the Forest Service defended the predictions. The final Idaho Panhandle plan was one of the few that significantly increased timber cutting.
Now, nearly a decade later, a new Forest Service report warns that Prognosis can lead to predictions "that are not biologically reasonable." The report's author, David Hamilton, Jr., says that he "examined its [the computer model's] sensitivity to variation in levels in the four primary component models." In other words, he run the program several times with slightly different user-defined variables in each run. With some variations, he found, he could cause the computer to predict that trees would grow 1,600 feet tall.
Anyone who is monitoring the Forest Service's new round of forest planning, particularly if the forests they are interested in are using Prognosis, should get this paper. Titled Uses and Abuses of Multipliers in the Stand Prognosis Model, the report is numbered INT-GTR-310 and is available from the Intermountain Forest and Range Research Station at 324 25th Street, Ogden, Utah 84401.
Q. What is RPA?
"RPA" is short for Resources Planning Act, which is short for the Forest and Rangelands Renewable Resources Planning Act of 1974. This soviet-style law required the Forest Service to present five-year plans to Congress which would become a part of the budgetary process. As amended by the National Forest Management Act of 1976, these plans come complete with targets for individual forests that are based on physical outputs, not economic efficiency.
Q. What is the RPA Assessment?
RPA requires the Forest Service prepare a detailed "assessment" of U.S. forest resources--on both public and private lands--in 1975, 1980, and every ten years thereafter. This assessment is supposed to consider the condition of the lands, supplies of resources, and the demand for those resources.
The most detailed assessment to date, done in 1989, produced huge volumes for timber, wildlife, recreation, and other resources. The timber assessment that year was the first that the Forest Service had ever done that assumed that rising prices would stem the volumes of timber people cut from the forests. All past assessments (including those done long before RPA) assumed constant prices and found that demand for timber quickly outstripped supply.
Q. What is the RPA Program?
RPA also directed the Forest Service to write a five-year "program" for the national forests and its other divisions (research, state & private forestry, and most recent international forestry) that responds to the needs identified in the assessment. In other words, the program identifies the volume of timber and other resources will the Forest Service produce to meet the demands estimated in the assessments.
Q. What is the difference between draft and final RPA Programs?
Like forest plans and other environmental impact statements, RPA programs come out in two editions: a draft, which is available for 90 days of public comment; and a final. The 1995 draft was published in October, and the deadline for public comment was January 16. A final will probably come out late in 1996. (The Forest Service is always late.)
Only a few score people typically comment on the programs; the Thoreau Institute prepared detailed comments on the 1980, 1985, and 1990 programs, which had no observable effects on the finals. I didn't comment this year.
Congress has had more influence on the programs in some years. At least two different programs were essentially scrapped by Congress, which wrote new targets for the Forest Service. These were duly ignored by on-the-ground managers.
Q. So what good is RPA?
The program is probably a waste of time. But the assessments contain huge piles of data, and some of it is fairly reliable. Some of the most interesting pertains to recreation values.
Q. What does the RPA assessment say about recreation values?
As with timber, the 1989 RPA assessment attempted to be more economically sophisticated. Forest Service economists combed the literature for estimates of the values of different kinds of recreation in different parts of the country. In the 1990 Program, these values were then multiplied by the actual and projected numbers of recreationists using the national forests to calculate the total annual value of national forest recreation.
Q. What kind of values were estimated by the assessment?
The 1989 assessment estimated three different values for recreation as well as other resources:
Typically, RPA counts all receipts, even though most of these receipts do not end up in the U.S. Treasury. Purchaser road credits, which are a form of in-kind payments, are counted, as are Knutson-Vandenberg and other funds retained by the Forest Service. Receipts actually retained by the Treasury are not counted in RPA.
Q. What is "willingness to pay"?
Willingness to pay is an estimate of how much people are willing to pay for recreation experiences and other goods and services.
Q. How do economists calculate willingness to pay for resources such as recreation that have no established markets?
Economists use two methods to estimate willingness to pay for recreation resources. The first is called "contingent value." Basically, this method comes down to surveys asking recreationists how much they are willing to pay. This question is supposed to be asked in a way that insures that recreationists do not overestimate the values. The second is called the "travel cost method," which is based on how much people actually do spend in traveling to recreation sites.
Q. What is "market clearing price"?
The market clearing price is the price that the Forest Service estimates it could actually collect from recreationists and other forest users.
Q. Why is the market clearing price different from willingness to pay?
Some people might be willing to pay $1 for a particular experience such as a hike. Others might be willing to pay $3, and others $10. The calculation of willingness to pay adds all of these together. But if the Forest Service were to actually charge, it could not collect all of these.
Instead, it might charge, say, a flat $3 fee for this hike. Those willing to pay only $1 wouldn't go. So the amount they are willing to pay isn't included in the market clearing price. Those willing to pay $3 would go as would those willing to pay $10. But they would only have to pay $3, so the extra $7 wouldn't be included in the market clearing price. (Economists call this extra $7 "consumer surplus" because this part of the economic value is retained by the consumers.)
Q. How did the Forest Service calculate market clearing price?
This calculation was partly based on the surveys estimating willingness to pay, some of which may have also estimated the market rates. The Forest Service tempered these estimates with a survey of actual prices charged by private landowners who offer various forms of recreation on their land.
Q. What recreation values were calculated by the 1990 Program?
The 1990 Program estimated values for 1989 (based on actual recreation use), 1995, 2000, 2005, and 2040 (all based on projected recreation use). Table one shows these values.
Table One 1989 1995 2000 2005 2040 Recreation values in millions of dollars Returns to Government 38 47 51 56 76 Market Clearing Price 2,943 4,002 4,632 5,195 7,416 Willingness to Pay 4,749 6,505 7,589 8,544 12,194
Source: Page 6-53 of The Forest Service Program for Forest and Rangeland Resources: A long-Term Strategic plan (the long title for the 1990 RPA Program).
The 1990 Program listed separate values for fish & wildlife recreation (table two). Unfortunately, these were grouped with other minor returns and labeled "Other." But charts on page 6-54 clearly showed that wildlife and fish made up nearly all of "Other." In addition to hunting and fishing, wildlife/fish includes "nonconsumptive wildlife uses," but this is a small component of the total.
Table Two 1989 1995 2000 2005 2040 Wildlife/fish and other values in millions of dollars Returns to Government 9 15 16 18 28 Market Clearing Price 2,281 2,713 3,116 3,514 7,525 Willingness to Pay 4,046 4,820 5,539 6,251 13,457
We can estimate the share of table two values attributable to wildlife and fish. Wildlife/fish returns to the government are 0, so the returns shown in table two are attributable to other resources. A graph on page 6-54 shows that, in 1995, wildlife/fish accounted for 98.2 percent of the market clearing price and 98.1 percent willingness to pay. Table three applies these multipliers to all years.
Table Three 1989 1995 2000 2005 2040 Wildlife/fish values in millions of dollars Returns to Government 9 15 16 18 28 Market Clearing Price 2,240 2,664 3,060 3,451 7,389 Willingness to Pay 3,968 4,727 5,432 6,130 13,197
Q. How do recreation values compare with commodity values?
Actual receipts from commodities--timber, range, and minerals--greatly exceed recreation and wildlife/fish receipts (table four). But recreation and wildlife/fish market values are three times greater than commodity market values.
Table Four 1989 1995 2000 2005 2040 Timber, range, and mineral values in millions of dollars Returns to Government 1,187 1,738 1,876 2,272 2,458 Market Clearing Price 1,550 2,386 2,595 3,860 4,141 Willingness to Pay 7,072 9,153 9,778 11,029 16,523
The willingness to pay for commodities is supposedly closer to (but still less than) that for recreation. This, however, did not seem reasonable and the Thoreau Institute's comments on the 1990 RPA argued that the commodity willingness-to-pay numbers were overestimated.
Q. Which of the three values should be used in economic analysis?
None of them. They are all wrong. The returns to the government are incomparable because of legal restrictions on the Forest Service's ability to charge fair market value for any resource but timber and some minerals. The market clearing prices and willingness to pay figures are the products of various economists' imaginations.
In theory, willingness-to-pay values should be used in economic analyses because these are the best indicators of total social good. But there are two good reasons why this is pointless.
First, if an economic analysis were done with these numbers, the results would be wrong because the numbers are almost certainly wrong. Second, even if the numbers were right, the results of the analysis would be ignored unless the results matched the incentives facing the person making decisions.
Instead of an economic analysis, the Forest Service should be allowed to charge fair market value. This would do two things. First, we would find out for sure what people are willing to pay (or at least the market clearing price) for various resources. Second, and much more important, the resulting income could provide sound incentives for federal land managers, giving them rewards for emphasizing recreation, scenery, wildlife, and other amenities.
Q. So what good are the RPA numbers?
The RPA numbers may be wrong, but at least the provide an indication of the relative values of recreation and other resources. Even if RPA overestimates recreation values by 200 or 300 percent, those values still outweigh commodity values. In fact, since most of the commodity values are confined to a few national forests, recreation values will exceed commodity values on most forests even if the RPA numbers are five or six times too high.
Q. How did the 1995 RPA modify the 1990 numbers?
Although RPA does not require an assessment for 1995, the Forest Service did a "mini-assessment" of various resources. Most significantly, the timber data was completely redone by Pacific Northwest Experiment Station economist Richard Haynes. An assessment was also done for recreation, but its results were not used in the 1995 RPA Program.
The new timber numbers in the 1995 Program partly reflect changes in projected cutting levels. But, in addition, the Forest Service has revised the willingness-to-pay timber numbers so that they are almost identical to, rather than four times, the market clearing prices. Table five compares 1990 and 1995 timber projections for the year 2000, which is the earliest year both projections have in common.
Table Five 90 RPA 95 RPA Projected National Forest Timber Values in the Year 2000 Returns to Government 1,536 1,094 Market Clearing Price 1,537 1,094 Willingness to Pay 6,440 1,166
The recreation values are complicated by the fact that, in the 1995 Program, recreation values are lumped together with the wildlife/fish values. This is particularly confusing because the wildlife/fish values are listed separately and there is not indication in the tables that recreation includes wildlife/fish.
Furthermore, the wildlife/fish portion of the recreation values was calculated by the recreation staff, while the wildlife/fish values were calculated by the wildlife staff. So they probably differed and we can't simply subtract the wildlife/fish values from the recreation total to get non-wildlife/fish recreation values.
The 1995 recreation values also differed from the 1990 values in that the 1995 values are based on national averages of all recreation while 1990's are based on regional breakdowns of nine different categories of recreation. So the 1995 values are cruder.
Although it may be inaccurate to do so, table six shows the non-wildlife/fish recreation values, calculated by subtracting wildlife/fish values from recreation values. These are compared with the 1990 recreation values. The table suggests that the 1995 update projects slightly lower values than the 1990 Program.
Table Six 90 RPA 95 RPA Projected National Forest Recreation Values in the Year 2000 Returns to Government 51 77 Market Clearing Price 4,632 4,448 Willingness to Pay 7,589 7,168
Wildlife/fish values were slightly reduced to account for more recent Fish & Wildlife Service estimates of hunting and fishing. Table seven compares the 1995 wildlife/fish values with the 1990 "other" values which I adjusted to include only wildlife and fish values.
Table Seven 90 RPA 95 RPA Projected National Forest Wildlife/fish Values in 2000 Returns to Government 0 0 Market Clearing Price 3,060 3,039 Willingness to Pay 5,432 5,167
The minerals and range values from the 1990 and 1995 RPA Programs are shown in tables eight and nine.
Table Eight 90 RPA 95 RPA Projected National Forest Mineral Values in 2000 Returns to Government 330 255 Market Clearing Price 1,007 786 Willingness to Pay 3,252 3,301Table Nine 90 RPA 95 RPA Projected National Forest Range Values in 2000 Returns to Government 10 18 Market Clearing Price 51 55 Willingness to Pay 86 67
In short, the 1995 RPA significantly reduced timber and mineral values, and slightly reduced most other values. Overall, this has boosted recreation and wildlife/fish values as a proportion of total national forest values (table ten), particularly in the willingness-to-pay category.
Table Ten 90 RPA 95 RPA Recreation & wildlife/fish values as percent of total in 2000 Returns to Government 3 5 Market Clearing Price 73 79 Willingness to Pay 55 73
Q. If it could charge recreation fees, how much does the Forest Service expect to collect from recreationists?
The Forest Service estimates that the actual numbers of recreation visits in 1993 were 730 million of which 110 million were wildlife/fish. By the year 2000, these are projected to increase to 767 million recreation visits, including 119 million wildlife/fish visits.
Table Eleven Market Clearing Price Willingness to Pay
1993 2000 1993 2000
Estimated market price per visit
Recreation 6.74 6.87 11.37 11.06
Wildlife/fish 24.11 25.47 41.16 43.31
Combined 15.87 16.08 16.24 16.39In other words, the Forest Service thinks it can collect about $7 per visit for non-wildlife recreation and about $25 per visit for wildlife recreation.
Those familiar with Forest Service recreation data are most used to "recreation visitor day" than to "visits." One visitor day is twelve hours of recreation use. For RPA, wildlife staff assumed that one visit lasted one day (twelve hours), so visits and visitor days are identical.
Other recreationists, however, were assumed to stay an average of only 3.6 hours per visit. This may seem too short, but relates to the way the way the Forest Service accounts for recreation. If someone goes camping, driving, and wilderness hiking in one trip, that may be counted as four visits. The drive is one visit that may last three hours. The camping is a visit that may last 10 hours. The wilderness hike is a visit that may last 4 hours.
In any case, at 3.6 hours per visit, the average market price per recreation visitor day in 1993 is $22. The willingness to pay is $38 per visitor day. These figures seem high. Some clarification is provided by the 1990 RPA.
The 1990 RPA published estimated values for each of nine kinds of recreation for each of the nine Forest Service regions. Market prices ranged from $1.13 per visit or $3.41 per visitor day for "hiking, horseback riding, and water travel" in Region 8 (the Southeast) to $28.50 per visit or $51.83 per visitor day for hunting in Region 2 (the central Rockies).
According to the 1990 numbers, the market prices per visitor day are roughly (averaged across the regions):
Even so, I routinely discount the RPA numbers by 67 to 75 percent. Even at a fourth to a third of the claims--which works out to $5 to $7 per visitor day--recreation values equal or outweigh all commodity values on the national forests combined. On the vast majority of forests, recreation values would swamp all other values.
Q. If the Forest Service were to charge fees, how would dispersed recreation fee collection be enforced?
State fish and wildlife agencies have nearly a century of experience in collecting and enforcing user fees for fishing and hunting, which are both highly dispersed forms of recreation. Although some poaching takes place, the agencies have been highly successful in collecting fees and in funding themselves--including habitat improvement as well as enforcement--out of the fees they collect.
In fact, as the Thoreau Institute's 1995 survey of fish and wildlife agencies found, nearly all state agencies fund their game programs entirely out of user fees plus federal grants from the Pittman-Robertson and similar funds. These federal funds all come from taxes hunters and anglers pay on guns, ammunition, and fishing tackle. State agencies that collect general funds or other tax dollars generally use those funds exclusively on non-game wildlife and other non-game activities.
State game fee collection is enforced in two main ways. First, of course, are the "game wardens" who check hunters and anglers to see that they have licenses. Most hunters I've talked to say that they have been checked by wardens a few times during their careers, but most anglers have rarely been checked.
The second enforcement method is peer pressure. Hunters and anglers know that part of the fees they pay go to help the wildlife and fish populations. So they frown on anyone who is not doing their share.
The Forest Service could enhance the methods used by state agencies with the use of a visible permit. Such a permit could be a bumper sticker, such as are found in many reserved parking lots. Or it could be a ski-lift-like tag worn by the user. Either way, anyone could quickly observe who had paid their permits. Peer pressure would thus play an even more important role in national forest fee collection than in hunting and fishing.
Q. Wouldn't recreation fees encourage the "Disneyfication" of national forests?
I hope so. But to understand why, we need to understand the history of Disneyland and DisneyWorld. Disneyland was built on about a hundred acres of orange groves. Disney couldn't afford to buy any more land, so as Disneyland became more popular, nearby landowners surrounded it with a sea of motels, fast food stores, and other developments. It is this sprawl, as much as anything, that disturbs visitors.
Disney vowed not to make the same mistake again. So when he planned DisneyWorld, he purchased nearly 30,000 acres in Florida. Only about 5 percent of these are developed; the rest are left in their natural state, partly as a preserve, partly as a buffer.
This proportion--5 percent development to 95 percent natural--is probably higher than would be found in the national forests. Yellowstone Park is less than 2 percent developed, and that 2 percent receives 5 million visitor days of recreation use--twice that of a typical national forest.
Q. But wouldn't recreation fees tend to bias managers in favor of developed recreation over dispersed recreation?
No, not if managers were funded out of their own net receipts rather than out of tax dollars. Developed recreation is distinct from dispersed recreation in two important ways.
First, developed recreation is capital intensive, for both the users and the providers. Since the users have paid so much for their equipment, they have less left over to pay for recreation fees. The Forest Service estimates that dispersed recreation market values are several times as large as developed values. Since the providers have to pay so much to service developed recreation users, they will be cautious to invest only where it pays off.
Second, disperse recreation is land intensive, while developed is not. Dispersed recreationists are willing to pay more for solitude; developed recreationists aren't, or at least not as much.
This means that it will pay national forest managers to leave the vast majority of their forests for dispersed recreation. Federal lands are also probably more suitable for dispersed recreation than most private lands, while many private lands are just as suitable for developed recreation as federal. So federal land managers will have an extra incentive to emphasize dispersed since they will have less competition from the private sector.
Of course, the answer would be different if forest managers received part of their funding from tax dollars or even were funded out of their gross receipts. If they received tax dollars, they would tend to use those tax dollars for expensive developments because such developments are politically more valuable as pork.
If managers were funded out of gross receipts and they made large profits from dispersed recreation, they would have to divert some of those profits to below-cost developed recreation or risk losing them to the Treasury. We see the same thing happening with Knutson-Vandenberg funds, which forest managers keep out of timber receipts: to maximize these funds, they cross-subsidize worthless timber with valuable timber.
Recreation fees will produce a truly balanced situation only if managers are funded out of the net income they earn from those and other user fees.
For the 1996 appropriations bill, Senator Mark Hatfield has proposed a provision that would require each agency to set up at least ten, and up to fifty, fee collection projects. To give the agencies an incentive as well as a mandate, Hatfield would allow them to keep any fees they collect over-and-above 104 percent of the fees collected in 1995. Of the fees they keep, 80 percent would be retained by the unit collecting the fees while the other 20 percent would be used by the agency as a whole.
In addition, the Forest Service, Park Service, and BLM would be allowed to keep up to 15 percent of their 1995 collections, but not more than their actual fee collection costs. To make things even more complicated, the 104 percent of 1995 collections is to be increased in future years by 4 percent per year.
This proposal is a step in the right direction. But the approach it takes is complicated and unwieldy. In addition, many parts of it create bad incentives for managers and others.
First, managers will get to keep a percentage of the gross receipts, giving them an incentive to maximize gross, not net. This will encourage cross-subsidization as noted above. Second, since the "up to 15 percent" share cannot exceed actual fee costs, some managers may have an incentive to bloat their fee collection costs in order to keep the whole 15 percent.
Managers will also continue to receive funds out of tax dollars. This will add to the incentive to emphasize developed recreation.
Finally, Hatfield's proposal would exempt the fees collected from the county 25 percent fund. This would lead county officials to resent any recreation fees because they would get no share of those fees.
Andy Stahl, director of the Association of Forest Service Employees for Environmental Ethics, has rewritten Hatfield's proposal to simplify and correct some of these problems. Under Stahl's proposal, on-the-ground managers would get to keep all the fees they collect. In exchange, sites collecting fees would no longer get any tax dollars for recreation.
Unlike Hatfield's proposal, none of the fees would be dedicated to higher levels of the bureaucracy. Local managers would be allowed to use some of the fees they collect to purchase goods and services from higher levels or from outside sources--promoting the "internal markets" advocated by Gifford Pinchot, III, in the "Reinventing the Forest Service" issue of Different Drummer.
Stahl considered, but rejected, the idea of letting counties keep 25 percent of recreation fees, as they do most other fees. "Why burden Forest Service managers with an additional subsidy when they are initially afraid of not covering their costs?" asks Stahl. One answer is to gain the political support of counties, which may or may not be needed.
The biggest problem remains that funding managers out of their gross, rather than net, income will encourage cross-subsidization. But this can be fixed only as a part of a much larger reform, one that changes the funding for all other forest resources as well.
An appropriations bill may be the wrong place to reform recreation fees in the first place. Such bills encourage a piecemeal approach that presumes that minor changes are sufficient to fix federal land management. Only the authorizing committees will be able to make a more serious reform. Public land supporters should work closely with those committees to endorse recreation fees and funding of land management out of net income.