Synopsis

Declines in National Forest timber harvests during the 1990s had a relatively modest impact on Washington's "timber dependent" communities. Any further reduction in federal timber harvests as a result of roadless area protection will have an even smaller impact on local communities, in part, because of the weak link between federal timber harvests and the local economy. The reasons, outlined in greater detail in this chapter, are listed below:

a. Most of these roadless areas were never part of the commercial timber base. Because very little harvest is planned from them, they provide a limited contribution to the total National Forest timber sale program.

b. National Forests are a minor source of Washington's total timber supply, about 4 percent statewide, 10 percent in the eastern Washington, and 3 percent in the southwest Washington. For some smaller areas the reliance on federal timber is much higher, but in general the National Forests are not the primary source of timber supply.

c. Other sources of timber supply can and do adjust to increases and decreases in federal timber supply, partially offsetting those federal changes.

d. Forest products employment is not rigidly tied to the level of timber harvest. When timber prices are high, additional labor can be employed to enhance wood fiber utilization. However, ongoing technological change tends to systematically reduce the employment associated with each mmbf (million board feet) harvested.

e. Fluctuation in the national demand for wood products is a primary determinant of the overall level of forest products production and employment. Timber supply is not the only or usual constraint.

f. Local harvests do not necessarily flow to local mills because logs are often hauled hundreds of miles to the mills that win the timber sale bids. As a result, any employment impact may well not be local.

g. Forest products workers do not necessarily live where forest products production facilities are located. Workers are mobile and take advantage of economic opportunities in the surrounding regional economy by commuting to work.

h. Washington's "timber dependent" communities have shown impressive economic vitality, continuously generating new sources of employment and income. This has allowed them to accommodate displaced forest products workers, diversify their economies, and productively adjust to a new economic reality.

The most common economic concern about setting the remaining National Forest roadless areas off limits to timber harvest is that this will reduce the amount of timber available for processing in local mills, triggering a reduction in lumber, plywood, particle board, and paper production and the employment associated with them. Limiting federal harvests will, it is argued, lead to lost jobs, reduced income, and a general deterioration in local economic vitality and well-being. In the previous chapter we saw that, in general, this did not happen in Washington during the 1990s despite a dramatic decline in federal timber harvests. This chapter will explore why catastrophic economic consequences did not follow from the loss of 70 to 90 percent of the National Forest harvests that had been available in the late 1980s. The study will follow the causal chain from reduced federal harvests through the economic impacts and adjustments to the final impact on the local communities. The study shows that there is not a firm connection between federal harvests and local economic vitality. This is important not only for understanding what the state experienced during the 1990s but for understanding what the impact of protecting Washington's remaining roadless areas is likely to be.

A. The Costs Associated with All Economic Choices

If restrictions on federal timber harvests were being proposed for reasons completely unassociated with local economic well-being and local economic vitality, we could analyze the employment impacts by simply focusing on the forest products industry since it would only be through that industry that employment would be affected. That, however, is not the case. The proposed restrictions on federal timber harvests are associated with efforts to protect other forest values. In that setting, the reduction in federal timber harvests is accompanied by enhanced production of other forest resources, goods, and services. There will be employment and income implications of that enhanced production of non-timber forest values. The net employment and income impacts, as well as the net impacts on local economic well-being can be calculated only by taking into account the economic implications of both types of changes.

The importance of looking at net impacts can be dramatized by considering a firm's decision to hire a worker or a consumer's decision to purchase a product. If one only focused on the costs side of the exchange, one could picture the firm as having suffered a loss by hiring the worker or the consumer as suffering a loss with each purchase. But both of those costs are incurred for a purpose: to obtain something that is expected to more than compensate for the cost incurred. The firm expects the value of the worker's output to more than cover the worker's wages and actually add to profits. Similarly the consumer expects the benefits associated with the thing being purchased to more than compensate for its cost. The costs are incurred in the pursuit of net benefits. One cannot evaluate economic decisions by only focusing on the costs.

This is also true of public policy decisions. Any decision is likely to have costs associated with it. What is interesting is how those costs compare to the benefits that are achieved. Protecting roadless areas allow the attainment of certain forest values at the expense of others. The net economic impact can be determined only after considering both the costs and the benefits of the policy. In the current context, the interesting economic question is how the forest values protected by roadless status compare to the value of the economic opportunities foregone such as commercial timber harvest. To simplify the following analysis, we will set aside the forest values that roadless status seeks to protect and focus exclusively on the timber opportunity cost associated with protecting an area's roadless status. In the next chapter we will return to an analysis of the values that roadless area protection enhances. In this chapter, however, we will focus on only one side, the negative side, of roadless area protection.

B. The Complex and Weak Connection between National Forest Roadless Areas and Local Timber Jobs

Putting roadless areas off limits to logging will not necessarily lead to reductions in local forest products employment. Where management of these roadless areas for non-timber values does have a negative impact on local forest products employment, the impact is likely to be much smaller than the acreage of the protected roadless area might suggest.

The relationship between management decisions about our remaining roadless areas and local forest industry employment is actually quite complex. It is not simply a matter of each forested acre that is put off limits to logging causing some specific reduction in employment. Figure 3.1 provides a schematic outline of the complex connection between roadless area management and forest products employment. We will discuss each of these as they apply to the roadless areas of eastern and southwest Washington.

i. The Planned Commercial Timber Harvest in the Roadless Areas

Because of varying mixes of these economic, biological, and environmental values, the Forest Service never planned to harvest trees in many of these roadless areas. ⁽¹⁾ These areas have not been part of the National Forest commercial timber base for some time. Because of that, formally setting them off limits to logging now causes no reduction at all in the Forest Service's planned timber harvests. For other roadless areas the planned harvests have always been low and removing them now from the timber base causes only a small reduction in future timber harvests.
 

For example, the forest plans developed in the mid-1980s for the Colville and Umatilla National Forests indicated that approximately 53 percent of the Colville and 74 percent of the Umatilla inventoried roadless areas were not part of the commercial timber base. ⁽²⁾ That is, the majority of these inventoried roadless acres could be put off limits to timber harvest with a zero impact on federal timber harvests.

The Forest Service has estimated the historical harvests that have been coming from the inventoried roadless areas (IRAs) and the planned timber harvests associated with those roadless areas. In general, very little timber has been harvested from these roadless areas in the past and an even smaller harvest is planned over the next five years. The roadless areas in the past have been the source of zero to 2.5 million board feet of timber per year depending on the National Forest, most of it harvested as salvage sales. Planned annual harvests from these areas are in the same low range. If the actual National Forest timber harvest levels between 1994 and 1998 are used as the reference point, putting these roadless areas off limits to timber harvest would cause current planned harvest levels to decline between 1 and 15 percent over the next five years, depending on the forest. These potential declines in National Forest timber harvest can be contrasted with the declines that took place in the 1990s. The declines in timber harvest associated with losing access to the inventoried roadless areas for logging would only be a tiny fraction of the declines that the region has already had to digest: one-eighth to one-three hundredths of the past declines in National Forest timber harvest. See Table 3.1.
 

The longer term impact of eliminating timber harvests in the inventoried roadless areas on National Forest timber harvests in the more distant future could be somewhat higher than this five-year forecast, but still would be relatively small. The Forest Service bases its overall timber harvest objectives on the forest plans each National Forest develops. After determining the best use to which each parcel of land can be put, the timber productivity of the lands allocated to commercial timber management determines the total planned harvest. During its last planning cycle National Forests in Washington included some of the roadless areas in the commercial timber base. In that sense, the current "allowable" or "probable" sale quantity (ASQ or PSQ) was based on the assumed timber productivity of some inventoried roadless areas. If these areas are now set off limits to timber harvest, the National Forests will have to reduce their ASQs and PSQs appropriately. Since many of Washington's National Forests have not been harvesting timber at close to the ASQ (or, in some cases, even the PSQ) associated with their past forest plans, this may have very little practical effect on the ground. The ASQ or PSQ will simply be adjusted to reflect what is already the practice. Harvests themselves may not decline. We quantify this longer term perspective in Chapter 5.

Finally, in this discussion we have focused exclusively on the inventoried roadless areas in Washington because we only have detailed data on those areas. There may be an acreage almost as great in roadless areas of a thousand acres or more that were not part of the National Forests' past "inventories" of roadless acreage. At this point we do not know what harvests are planned for these areas or what role they play in setting the ASQs and PSQs.

For the majority of the Washington inventoried roadless areas, restricting logging will have very small to zero impacts on Forest Service timber harvests. At the very least, in discussing the timber impacts of the Roadless Area Initiative, the roadless acres that have never been part of the Forest Service's commercial timber base should not be a subject of concern.

ii. How Important Is Federal Timber Supply to the Total Supply?

The National Forests were the source of between 5 and 7 percent of the total timber harvest for the major eastern Washington timber producing regions in 1998. In the southwest counties, the dependence on National Forests was 3 percent, in the northeast counties 7 percent, and in the central counties 5 percent. In the southeast, where the combined timber harvest of four counties was less than the timber harvest in 15 of the other individual counties listed in Table 2.1, federal harvests represented about a sixth of the total harvest. Two individual counties with substantial timber harvests showed a relatively high dependence on federal harvests: Chelan and Skamania. On the other hand, several counties with substantial timber harvests showed very low reliance on federal harvests: Lewis, Pacific, Cowlitz, and Yakima Counties, each of which harvested 200-400 million board feet, relied on federal harvests for zero to 2 percent of their harvest. See Table 3.2.
 

The National Forests of Washington are the source of only a very small part of the total timber harvested. The ICBEMP Supplemental Draft EIS estimated the direct employment associated with National Forest timber harvests in eastern Washington "non-owl" forests to be about 400. ⁽³⁾ Adding in the employment associated with the eastern Washington "owl" forest brings this to about 650. ⁽⁴⁾ Total employment in eastern Washington in 1997 was 692,000. ⁽⁵⁾ So the total direct employment associated with the entire federal harvest represented less than one-tenth of one percent of all jobs. Because of the relatively low employment associated with National Forest harvests, changes in that harvest can be expected to have only a very small impact on the local and state economies.

iii. How Do Federal Supplies Interact with Other Sources of Supply?

The reduction in federal timber harvest was partially offset by the expansion in timber and raw material available from other sources. This was not an accident or a coincidence or the result of conscious public policy. This is simply the way markets work. When supply from one source is reduced, prices rise, providing incentives for other sources of supply to expand and for a more efficient use of the available raw material.

Recognition of the interplay between various sources of wood fiber supply in the context of constantly changing national and international demand is the basis of any sophisticated projections of future forest product production and prices. Forest Service economic modeling for the 1993 national RPA timber assessment projected that during the 1990s National Forest timber harvests would decline almost 50 percent but that total timber harvests would expand slightly. ⁽⁶⁾ For eastern Washington and Oregon, a 200 million cubic feet reduction in National Forest Harvests (almost a 60 percent reduction) would lead to about a 100 million cubic feet reduction in total harvests as about half of the National Forest reduction was offset by expanded forest industry and other private harvests. ICBEMP projections of timber harvests in the BEA economic regions of eastern Washington showed the same type of partial offset of reductions in National Forest timber harvests. ⁽⁷⁾

This is not just wishful thinking. In many eastern Washington counties, as National Forest timber sales plummeted between the late 1980s and late 1990s, the decline in federal timber harvests was significantly offset by expansion in other harvests. For instance, in the northeast corner of the state, National Forest harvests declined from the peak years of 1987-1989 to the late 1990s by 90.9 mmbf. Timber harvests from other sources, however, expanded by 44.6 mmbf. If the decline in the federal harvests, in isolation from these other changes, had been used to estimate the job loss in the forest products industry, a total of over 1,300 jobs would have been projected. ⁽⁸⁾

Instead of this huge loss in forest products jobs, there was actually a gain in forest products jobs. This gain in forest products employment was partially due to the opening of the Ponderay Newsprint mill in Pend Oreille County (part of the pulp and paper industry total) and the expansion of wood products employment in Spokane County. If only wood products facilities are considered, there was a modest reduction in jobs, about 100. That, of course, is a tiny fraction of the 1,300 job loss that would be estimated if the National Forest harvest decline alone was focused on and a mechanical response coefficient applied to it.

This pattern of offsetting changes in other timber harvests can be found elsewhere in eastern Washington. See Table 3.3.

Statistical analysis that compares changes in National Forest harvests and total harvests confirms these average results. ⁽⁹⁾ For the northeast and southeast counties, a change in National Forest harvest is associated with a change only half as large in total harvest. For other regions there is often no statistically significant relationship between National Forest harvest and total harvests.

One of the reasons that forest products employment does not change as dramatically with National Forest and other timber harvests is that some elements of the industry are not particularly sensitive to the total level of harvests. This is especially true of value-added manufacturing that makes use of products from lumber mills; e.g., the milling of boards into molding and other products and, to a certain extent, particleboard and pulp and paper manufacturing. The latter two partially make use of the waste products of lumber mills for their raw material inputs. Paper also makes use of recycled wood fiber. Between the peak harvests in the late 1980s and 1998, total timber harvests in the state of Washington fell about 40 percent while employment in the paper industry fell by about 5 percent. ⁽¹⁰⁾ In eastern Washington pulp and paper industry employment actually increased by over 500 between 1988 and 1998 despite the decline in total timber harvests. Statistical analysis of changes in total harvest and employment in the paper industry shows almost no relationship between the two. ⁽¹¹⁾ Clearly this type of employment is not tied rigidly to either National Forest or other timber harvests.
 

For the state of Washington as a whole, the offsetting adjustments that ameliorated the impact of declining harvests was impressive. While total log utilization declined by about 37 percent between 1988 and 1996, log utilization in lumber mills fell only about 17 percent. This was made possible by a dramatic decline in log exports, over 50 percent, and a shift by pulp mills from the use of logs to the use of chips from lumber mills and recycled fiber. These shifts allowed the total production of lumber in Washington to remain relatively constant between 1988 and 1996 despite the decline in harvests. Plywood and veneer production was actually able to increase while pulp production was stable. ⁽¹²⁾

Given that National Forest timber harvests are linked by markets to other sources of wood fiber supply, it is not correct to assume that every change in federal harvest results in a similarly sized change in total harvest.

iv. Forest Products Jobs Depend on Forest Product Demand Not Just Timber Supply

The role of fluctuating demand for wood products can also be seen in the 1990s. As the national economy recovered from the recession and the lengthy expansion of the 1990s got underway, housing construction rebounded. National lumber prices shot upward between 1992 and 1994 as national demand for wood products boomed. Declining home construction and a close brush with a recession in 1995 led lumber prices to plunge. As the economy regained its strength lumber prices shot upward in 1996, only to plunge in 1997-98 again as the crisis in Asia led to a reduction in worldwide timber demand. That lingering "Asian flu" led lumber prices to fall for a record 14 straight weeks in 1997-1998. This weak demand led to reduced wood products production and employment. Low lumber prices discourage harvest and production; high prices encourage it. These familiar market cycles continue to play an important role in wood products employment.

Although it is true that reduction in timber harvests from National Forest lands has reduced employment in Washington's woods and lumber mills in the past, it is likely that most future reductions in employment will be tied to national market conditions and labor displacing technological change. The Washington State Office of the Forecast Council emphasized the importance of national and international market conditions in determining current and future employment in wood products in Washington recently:

…The U.S. recovery has not been reflected in Washington primarily as a result of timber supply constraints, which have reduced Washington's market share. In addition, exports of lumber and wood products have been particularly hard hit by the economic distress in Asia. Washington lumber and wood products employment is now 9,100 lower than the cyclical peak in the fourth quarter of 1988. The forecast assumes that the adjustment to reduced federal timber supplies is complete, but a cyclical downturn in housing and rising productivity will result in a loss of 1,200 jobs by the end of 2001. ⁽¹³⁾

Note the emphasis on weak Asian markets in addition to Washington supply constraints and the forecast that in the future it will be national market conditions and technological change that will be the source of significant additional layoffs, not timber supply limitations. This represents a return to the type of economic forces that have been responsible for most of the wood products layoffs in the past.

v. Local Harvest Does Not Necessarily Flow to Local Mills

vi. Technological Change, Rising Labor Productivity, and Jobs

The role of changes in production technology in reducing wood products employment were dramatized in Washington during the peak-to-peak years of 1979 to 1988, before the spotted owl decisions, when total timber harvests were about the same in both years, but wood products employment declined 21 percent and wood products real payrolls fell 34 percent. Clearly, powerful economic forces that had nothing to do with timber supply were at work in the wood products industry. Between the two peaks, wood products real payroll per million board feet harvested declined by a third and direct employment per million board feet harvested declined by a fifth. ⁽¹⁵⁾

During the 1990s this decline in the forest industry jobs supported by each million board feet harvested reversed itself in Washington and employment per million board feet rose back to where it was at the beginning of the 1980s. ⁽¹⁶⁾ This does not represent a reversal of technological trends but the operation of a different set of forces. With much higher raw material and product prices, more fiber that would have been abandoned as waste was processed. The higher prices also allowed more labor intensive and environmentally benign harvest techniques to be used. In addition, high local prices reduced raw log exports and increased raw log imports. The high cost of raw material also increased the use of recycled wood fiber in the paper industry. Finally, there may have been an expansion of the value-added processing of wood products. The net result of these changes was to increase the employment associated with each million board feet of timber harvested in Washington during the 1990s. However, as the quote from the Washington Office of the Forecast Council above indicated, the expectation is that technology in the future will return to reducing overall employment in forest products.

When worker productivity in lumber mills is narrowly focused upon the long run, a national trend is clear: The labor content of wood products has continued to decline during the 1990s. Since 1958, the earliest year for which labor productivity figures are available, national wood products output has increased about 50 percent while wood products employment has decreased 50 percent. The net result is that output per employee has nearly tripled. Put the other way, the number of workers needed to produce a fixed level of output has been cut to a third of what it once was. See Figure 3.2.
 

The dramatic and divergent changes in the wood products jobs associated with Washington timber harvests in the 1980s and 1990s provide clear warning that there is no tight relationship between timber harvests and jobs. During the 1980s employment fell despite very high levels of harvest. During the 1990s employment was surprisingly stable despite dramatically declining harvests. During the 1980s as Washington moved from one peak harvest to another, each of almost identical size, employment declined significantly more than in did in the 1990s when harvests plunged 40%. See Figure 3.3.

This discussion of the relationship between timber harvests and jobs contains both good and bad news. The good news is that technological developments can lead to new products and additional processing of wood fiber that allows more jobs to be created without harvesting more timber. The use of lumber mill waste in both particle and fiberboard mills as well as in pulp and paper mills are good examples. The use of recycled paper and cardboard in pulp mills is another. Secondary manufacturing of specialized products from molding to trusses to prefabricated housing also adds to employment without requiring additional logging. Finally there are labor-intensive manufacturing processes, such as log homes.

The bad news is that technological change in the harvesting and processing of timber is relentless. If the manufacturing process does not diversify, a given level of harvest will support a steadily declining workforce. It would take exponentially increasing harvests and demand for wood products to hold off this loss of employment. Neither our forested mountains nor forest products markets would support such exponential increases in harvests.

vii. Where Do Wood Products Workers Live? Where Are Mills Located?

The forest products manufacturing facilities that provide most of Washington's forest products jobs are not, in general, located in isolated small towns. Instead they are located in the state's largest urbanized areas or within commuting distance of towns of 10,000 or more.

This is an important feature of forest products employment in Washington because the state's larger urban areas provide a broad range of employment opportunities and have been creating thousands of new jobs every year. These expanding employment opportunities allow workers who have lost jobs in forest products to find new employment relatively rapidly and with a relatively low loss of income. For instance, one of the primary industries into which wood products workers have shifted is construction. The ongoing economic vitality throughout Washington has increased the number of construction jobs by over 53,000 between 1988 and 1997. The average annual pay in wood products was $38,700 while the average pay in construction was $37,000 with general construction paying $38,700 and heavy construction paying $45,500. Such job shifts could be made with little or no loss in pay.

In 1998 about 60 percent of Washington's forest products employment was located in metropolitan areas, counties with at least 50,000 people in the core urban area and 100,000 in the surrounding settled area. In eastern Washington the percentage is a little lower; almost exactly half of the forest products employment is found in these large urban areas.

A worker does not have to be living in a metropolitan county to encounter a broad range of economic opportunities. A worker living within commuting distance of a metropolitan area can also take advantage of a more diversified economy. For instance, as discussed in Chapter 2 above, there is considerable commuting between Spokane and the non-metropolitan counties north of Spokane. The 1990 Census showed almost 2,200 workers commuting from Stevens County to Spokane to work. Over 500 workers in Pend Oreille commuted to Spokane. Meanwhile over 400 Spokane residents commuted to Stevens and 200 to Pend Oreille to work. Hundreds also commute between Ferry and Stevens Counties. In addition Pend Oreille shares hundreds of workers across the Idaho border with Boundary County.

As also discussed in Chapter 2, it is always possible that some areas are so isolated geographically that residents cannot practically commute to work in a larger urban area. In such isolated areas, the local economic opportunities might accurately represent all the opportunity available to residents. ⁽¹⁸⁾ However, the only isolated towns that were also significantly dependent upon forest products that were identified in Chapter 2 were found along the northeastern tier of counties: in Okanogan, Stevens, Ferry, and Pend Oreille Counties. Those counties contained 29 percent of all eastern Washington forest products employment in 1997, but all of the forest products employment in these counties is not found in the isolated northern valleys of these counties. Newport, Cusick, Usk, Chewelah, and Springdale, for instance, are within commuting distance of the Spokane-Coeur d'Alene economies. Thus, somewhere between a fifth and a quarter of the eastern Washington forest products jobs are located in small isolated towns where employment opportunities may be limited. The total population of the isolated "timber towns" in this northeastern tier of counties is about 19,000. The total population of eastern Washington's counties is about 1.3 million. Thus, about 98.5 percent of the eastern Washington population lives outside of these isolated "timber" towns and has access to a much more diversified set of economic opportunities. Although it is appropriate and important to provide isolated communities with the assistance they need to make the transition to a lower level of reliance on federal timber, it would not seem appropriate to change the management of federal forests in order to maintain or increase that dependence especially at the cost of damaging other forest values on which the entire population relies.

The conclusion we reach from this consideration of where forest products facilities are located is that for a large majority of forest products workers, economic opportunity is not dependent on local timber harvested from national forest lands. The facilities tend to be located in or near large urban areas with diversified economies. The forest products raw materials are hauled long distances to these facilities. Both of these facts limit workers' and communities' dependence on National Forest timber harvests.

C. Conclusion: Local Jobs and National Forest Timber Harvests

Declines in National Forest timber harvests during the 1990s had a relatively modest impact on Washington's "timber dependent" communities. Any reduction in federal timber harvests as a result of roadless area protection will have an even smaller impact on local communities. The reasons for this include:

a. Most of these roadless areas were never part of the commercial timber base. Very little harvest is planned from them.

b. National Forests are the source of only a small minority of total state timber harvests, about 4 percent statewide, 10 percent in eastern Washington, and 3 percent in the southwest. For some smaller areas the reliance on federal timber is much higher, but in general the National Forests are not the primary source of timber supply.

c. Other sources of supply adjust to increases and decreases in federal timber supply, partially offsetting those federal changes.

d. Forest products employment is not rigidly tied to the level of timber harvest. When timber prices are high, additional labor is employed to enhance wood fiber utilization. On the other hand, ongoing technological change tends to systematically reduce the employment associated with each mmbf harvested.

e. Fluctuation in the national demand for wood products is a primary determinant of the overall level of production and employment. Timber supply is not the only or usual constraint.

f. Local harvests do not flow to local mills because logs are often hauled hundreds of miles. As a result, any employment impact may well not be local.

g. Forest products workers do not necessarily live where forest products production facilities are located. Workers are mobile and take advantage of economic opportunities in the surrounding regional economy by commuting to work.

h. Washington's "timber dependent" communities have shown impressive economic vitality, continuously generating new sources of employment and income. This has allowed them to accommodate displaced forest products workers, diversify their economies, and productively adjust to a new economic reality.

2.  "The Timber Employment Impact of the Northern Rockies Ecosystem Protection Act in Washington and Oregon," Thomas M. Power, August 1992, Economics Department, University of Montana, p.9.

3.  Interior Columbia Basin Supplemental Draft Environmental Impact Statement, Volume 1, p. 159, Table 4-44, March 2000.

4.  See the discussion in Chapter 2, Section D.

5.  Employment covered by unemployment insurance in 1998 was 535,000. Total employment includes the self-employed, those working for very small businesses, and a few other types of jobs not covered by the unemployment insurance program. The advantage of covered employment is that it is based on actual reports by firms. Total employment has to be estimated based on survey data.

6.  Haynes, R.W., Adams, D.M., Mills, J.R. 1995. The 1993 RPA Timber Assessment Update. Gen. Tech. Report RM-GTR-259. Fort Collins, CO: USDA Forest Service, Rocky Mountain Forest and Range Experiment Station.

7.  Table 6.32, An Assessment of Ecosystem Components in the Interior Columbia Basin, Volume IV, USDA Forest Service, Pacific Northwest Research Station, GTR PNW-GTR-405, Chapter 6, Richard W. Haynes and Amy L. Horne. June, 1997

8.  The Forest Service estimates that 14.96 jobs are associated with each million board feet harvested from National Forest land in the Inland Empire. See multiplier data from Dick Phillips, USFS Region 6 Regional Economist, February 15, 2000. This is the data used in TSPIR reports to Congress. The ICBEMP economic analysis appears to be using much lower employment multipliers in its estimates of the employment directly associated with federal timber harvests. See Interior Columbia Basin Supplemental Draft Environmental Impact Statement, Volume 1, p. 159, Table 4-44, March 2000. If the approach taken there is applied to the Inland Empire region, there would be about 9 jobs associated with each mmbf harvested and the job loss projection would be about 800 rather than 1,300. Job multipliers will be discussed further in Chapter 5.

9.  Some of the statistical results are highly dependent upon the time period used. Focusing on the 1990s usually leads to no statistically significant relationship. Carrying the analysis back to the late 1970s tends to show all harvests moving together.

10.  US BEA REIS data show a 1.4 percent or a 238 job decline in the paper industry (SIC 26) 1988-97. Estimated employment in 1998 was another 530 jobs lower for a total decline of 4.6 percent.

11.  A regression of paper industry employment on total timber harvest has an adjusted R-square that is less than zero. That is, changes in timber harvest can explain none of the variation in paper industry employment. Some pulp and paper mills have closed in Washington (e.g. Clallam and Grays Harbor Counties) during the 1990s. The closures were primarily tied to their operating costs relative to competing facilities.

12. Washington Mill Survey 1986-1996: Has the Sun Really Set on Washington State's Forest Products' Industry?, Dave Larsen and Phil Aust, Washington State Department of Natural Resources, paper presented at the 34^th Annual Pacific Northwest Regional Economic Conference, April 26, 2000.

13.  Quarterly Economic and Revenue Forecast, June, 1999, p. 10, www.wa.gov/ofc/pubs/jun99pub.pdf.

14.  Historical Statistics of the US, Bureau of the Census, Series L72-86; W.F. Freudenburg et al., "40 years of Spotted Owls? A Longitudinal Analysis of Logging-Industry Job Losses, University of Wisconsin, Madison, Figure 1.

15.  REIS SIC 24 real earnings and employment divided by total timber harvest. Two peak years are used for the comparison to avoid distortions tied to different points of wood products industry cycles. This is not the only way to measure changes in the employment supported by timber harvests. It was chosen because of the concern that reduced harvests will lead to mill closures and loss of jobs in the woods. Pulp and paper are not included because of the lack of connection between timber harvest and employment in that industry in Washington and because of the industry's reliance on recycled fiber and mill residue rather than primarily on round wood harvested for pulp.

16.  The Labor Market and Economic Analysis Branch of the Washington State Employment Security Department has measured worker productivity in lumber and wood products by real gross state product per worker. It confirms a steep rise in worker productivity during the 1980s followed by a decline of similar size during the 1990s. See "Worker Productivity Trends in Washington, 1977-1999," Gary Kamimura, October 1999, Figure 26, p. 14.

17.  The Washington Department of Natural Resources periodically studies the flow of logs between different regions in Washington. That data is published in the "Washington Mill Survey." Unfortunately the areas used in that log flow analysis are very large. For instance, one area is the "Inland Empire" which stretches from the Canadian to Oregon borders. "Central Washington" is even larger stretching from Canada to Oregon and just west of Spokane to Yakima. For instance, the log flow data for 1996 shows that 82 percent of the Colville National Forest logs stayed within the Inland Empire. That does not assure that the logs flow to the isolated mill towns in northern Ferry and Stevens Counties.

18. The issue of isolated timber towns was discussed in more detail in section D of Chapter 2.