Figure 3.--Alternative harvest levels for the Umatilla National Forest.
Harvests can be converted to board feet, local scale, by using the
board-foot/cubic-foot ratio 6.4.

Figure 3 also shows that the potential yield was reduced from a base of 31 million cubic feet per year to 24 million, or 22 percent when 100 percent of the roadless area was withdrawn. Notice that potential yield when 100 percent of the roadless area was withdrawn is lower than the base programed harvest. With 50 percent of the roadless area withdrawn, potential yield is reduced 13 percent but still remains at a higher level than the base programed harvest.

GENERALIZING STUDY RESULTS TO OTHER NATIONAL FORESTS

The study Forests are not a scientific "sample." Therefore, no firm, quantitative conclusion can be drawn about other Pacific Northwest National Forests. This study, however, does suggest that there may be some ability to substitute management intensification for roadless area volume on other National Forests, especially if withdrawals include up to half rather than all the roadless area.

Identifying National Forests that are likely candidates for this substitution is made difficult by the nature of the constraints that hold down the harvest as the land base changes. For example, when all roadless areas are included in the Willamette's harvest calculation, the budget for management intensification is the primary constraint holding down harvest; however, when all roadless areas are excluded from the Willamette's harvest calculation, the number of acres allowed for regeneration harvest becomes the primary constraint holding down the harvest. This situation makes it extremely difficult or perhaps impossible to use the harvest results from the study Forests to pre

dict the result of withdrawals of roadless areas on other National Forests. To predict the result of withdrawing roadless areas and intensifying timber management on the remaining land, we must be able to determine what the principal constraint on harvest would be with the reduced land base. In many cases that information can only be obtained by a careful consideration of some harvest calculations. For most Forests, calculations of that sort do not now exist. If one could determine that funds for intensive management would be the principal constraints with the reduced land base, it would still be necessary to have some harvest calculations to indicate the amount that the harvest could be increased through reallocation. The earned harvest from management activities on the full land base will not provide a reliable estimate because opportunities for applying management activities are normally reduced when roadless areas are withdrawn. Therefore, aside from the general conclusion that the opportunities to offset harvest reductions through investments in more intensive timber management appear to be limited, it does not appear possible to draw sound conclusions about other Forests without making harvest calculations for the alternatives as we have done.

Financial and Employment Consequences of
Roadless Withdrawals

ASSUMPTIONS USED IN FINANCIAL ANALYSIS

Purpose of the Analysis

The purpose of our analysis is to explore the financial implications of some broad alternatives for allocation of roadless areas on the Siskiyou, Umatilla, and Willamette National Forests. One factor in particular, the 25 percent of gross revenues that National Forests are required to pay to counties, is a valuable financial indicator of the impact of roadless area allocations on local economies. Present net worth calculations are essential components of a national efficiency analysis. Finally, revenue and cost consequences are useful indicators of the effect of the alternatives on Forest budgets and Forest Service receipts to the United States Treasury.

We have no financial results for individual roadless areas on the Forests. We know, however, that within the broad averages used for each Forest individual roadless areas vary tremendously in timber inventories, amount of commercial forest land, productivity, and financial value.

Projecting Future Trends

Perhaps the most perplexing problem in quantifying the financial consequences of the harvest alternatives is how to account for the considerable uncertainty which exists regarding prices and costs in the future. To investigate the sensitivity of the financial results to alternative views of the future, we calculated the results with two interest rates, three assumptions about the course of stumpage prices, and two assumptions about future management costs. Real (deflated) prices, costs, and interest rates were used in the financial analysis. In terms of projecting future economic trends, "real" means that we make no attempt at projecting inflationary trends.

Our procedure for projecting future trends involves determining a current value for the item we are projecting and then estimating its future real increase, if any. The initial stumpage prices for each National Forest come from the trend in high bid prices from recent sales on the Forests. The changes in real prices come from an early version of the Resources Planning Act Timber Assessment Softwood Market Model.9

The average annual compound growth rate in real prices for the period 1978-2030 on Region 6 Forests is: West-side (Douglas-fir) region, 1.9 percent; east-side (ponderosa pine) region, 1.5 percent. The growth rate for prices is higher than average for the west side and east side until the year 2000, after which it slows considerably. After the year 2030, real stumpage prices are assumed to be constant.

Price Assumptions

Table 2 shows the values for initial stumpage prices and road items for each study Forest. Our assessment of stumpage prices starts with the high bid price for the accessible area. High bid is the value of stumpage as if the roads were in place. It is appropriate to use the high bid since we are accounting for road costs separately. The prices received for stumpage will be different in the roadless area from those in the accessible area because of differences in species mix, timber quality, and logging and hauling costs. factors were accounted for when we developed separate prices on each Forest for each half of the roadless area and for the accessible area. The high bid prices for each area (table 2) were provided by each Forest and represent 1977 stumpage prices which were trended to average out recent fluctuations in stumpage markets.

In the financial analysis, we use three assumptions about the future course of stumpage prices.

The first assumption is that the stumpage price on each study Forest remains constant over time at the recent high bid levels reported by the Forests. The second and third assumptions are that real stumpage prices will rise until 2030; the difference between the second and third assumptions is in their treatment of Region 6 Forest Service harvest levels. With the second assumption no roadless areas would be withdrawn on other Region 6 National Forests and Forest Service harvest levels would follow trends assumed in the Resources Planning Act Timber Assessment Softwood Market Model (see footnote 9). The third price assumption for the Willamette and Siskiyou National Forests is the expected trend in prices if changes in harvest levels on all west-side Region 6 National Forests occur simultaneously and are proportional to the change on the Forest being analyzed. The third price assumption for the Umatilla is the expected trend in prices if there are simultaneous changes in harvest levels on all east-side Region 6 National Forests that are proportional to the changes on the Umatilla.

9

Adams, Darius M., and Richard W. Haynes. 1978. A preliminary description of the 1980 Timber Assessment Softwood Market Model. Report on file at the Pacific Northwest Forest and Range Experiment Station, Portland, Oregon.

For the second and third assumptions, the stumpage price for each study Forest depends on three factors: (1) regional stumpage price, (2) the quantity of stumpage harvested on the Forest, and (3) the land base (roadless areas included or not) associated with harvest. Regional price effects are determined from demand relationships which take into consideration harvest changes on private lands.

In the absence of data on harvest changes on other Forests, we consider the second assumption most realistic, and results using that assumption are highlighted in the financial results section. Results for the third price assumption are included in appendix D. These results must be viewed as a rough approximation of the financial effects on each study Forest of withdrawing roadless areas simultaneously from the other National Forests because the actual harvest changes that would take place on other Forests would not likely be proportional to the change on the Forest being analyzed. Since we cannot estimate the Region 6 change in harvest from our limited number of study Forests, we use the assumption of proportional harvest changes to illustrate the effect of this alternative.

The actual prices used in present net worth calculations and other financial results are shown for each price assumption and for each study alternative in appendix A.

Interest Rates and Discounting

Gregersen (1975) points out the importance of not confusing market rates of interest (which incorporate expected inflation) with real rates of interest. Referring to investments in private forestry, Klemperer (1976) concludes that when inflationary effects are removed from interest rates, an after-tax rate of 5 to 6 percent is competitive. Two interest rates, 5 and 10 percent, were used in present net worth (PNW) calculations; 5 to 10 percent represent a range in interest rates which is sufficiently wide to reveal the sensitivity of the financial results to the cost of capital. The 5- to 10-percent range also avoids the difficulties of attempting to identify a single "correct" interest rate for public investment evaluation.

Present net worths shown in the financial results section and in appendixes D and E are calculated for 10 decades by the following relationship:

average annual net revenue received in the n-th decade,

Cost Assumptions

The success of dealing with uncertainty through sensitivity analysis depends on thoughtful selection of alternative views of the future. Two assumptions about the future course of real costs were used in the financial analysis. The first assumption is that all costs will remain constant at their present levels. The second assumption is that the cost per acre for labor intensive practices will increase at the same rate as real per capita income in Oregon and Washington. Specifically, costs for regeneration, precommercial thinning, and timber sale preparation are assumed to increase at the same rate as the U.S. Water Resources Council (1974) projections of real per capita income in Oregon and Washington to the year 2020. The increase in real costs of these items is approximately 2.7 percent per year over the period 1980-2020. Costs are assumed constant after that.

All other cost items including road construction, reconstruction, and maintenance are assumed to remain constant in real terms; i.e., they will increase at the same rate as the general price level. The management and road costs for each Forest are found in appendix C. Road costs are also summarized in table 2.

The compound annual growth rate of real per capita income in Oregon and Washington for selected years between 1970 and 2020 is shown in appendix B. These growth rates are applied directly to the costs provided by each study Forest to obtain the future cost of the labor intensive practices.

ASSUMPTIONS AND PROCEDURES USED IN
EMPLOYMENT ANALYSIS

We quantified the employment consequences of the harvest alternatives with the input-output (1-0) models developed as a part of the RARE II analysis. The RARE II I-0 model for each Forest is based on employment data from a multicounty area encompassing the Forest. These employment data are used to scale the national I-0 model to reflect the characteristics of the local economy.

Changes in harvest levels are directly and fully translated into changes in sales to final demand from the local wood products processing and timber supply sectors. No compensating adjustments in harvest flows from other local ownerships or nonlocal sources are recognized.

The consequences of the harvest alternatives on employment apply only to the local economies--economies for which the study Forests are an important source of forest-related goods and services. The reported impacts are not the only consequences of the harvest alternatives on employment, and another choice for the area of employment impact would lead to a different set of results. It is at the local level, however, that the effects of harvest changes and land allocation decisions take place and will be felt most heavily and the concern for impacts on employment is likely to be intense.

The employment results represent initial effects only. The difficulty of accurately assessing the future course of labor productivity and structural change within the local economy precludes a projection of the consequences for employment over several decades.