EXAMINING THE ASSUMPTIONS UNDERLYING
THE NATIONAL FOREST PLANS

DOES MORE LOGGING EQUAL MORE JOBS?

ABOUT THIS REPORT

This report was authored by Michael T. Garrity, a fifth generation Montanan from Helena. This report was developed as part of a graduate internship project for the Alliance for the Wild Rockies through the University of Montana in Missoula.

Critical peer review was provided by Dr. Thomas Power, Chairman of the University of Montana Economics Department; and Dr. Douglas Dalenberg, professor of economics, University of Montana. Funding was provided by the Alliance for the Wild Rockies.

The Alliance for the Wild Rockies is an international alliance of hundreds of member organizations and businesses, and thousands of individuals. Based in Missoula, Montana, AWR is organized as a non-profit corporation to promote and protect the remaining wildlands and native species in the five-state, two-province Northern Rockies Bioregion through research, education, and citizen involvement.

EXAMINING THE ASSUMPTIONS
UNDERLYING THE NATIONAL FOREST PLANS

The Forest Plans that govern the management of our National Forests are very controversial. Environmentalists believe we are destroying our forests by cutting too much. The timber industry believes that we are destroying our wood products jobs by cutting too little.
The National Forest Management Act requires the Forest Service to prepare management plans with alternative management scenarios. The Forest Service is required to choose the alternative which maximizes the public net benefits amortized over the next 150 years. The preferred alternative will guide the Forest Service in management of the forest until the plan is revised. They are required to revise their plans at least once every 15 years. Timber constitutes the largest proportion of the resource base in the 10 National Forests in Montana. Therefore timber management is the most important contributing factor in the present net value of the resources on the forest.

The wood products industry constitutes the third largest industry in Montana. The decisions implemented in the forest plans can have a major impact on the economy in Montana. The Forest Service recognizes this and therefore makes the goal of the plans to stabilize timber dependent communities and the aggregate Montana economy. The wood products industry is expected to become more dependent upon National Forest timber as the harvest from private lands declines. Harvest is expected to decline on private lands due to overharvesting and the decision to withdraw some land from the timber base and convert it to other uses such as cattle grazing.

The Forest Service sees timber related jobs as the driving force for community stabilization. As a result the number of jobs that result from each alternative is forecast at the end of the first decade in the life of the fifty year plan and at the end of the 5th decade. The more jobs each option creates the more weight assigned to it. The Forest Service assumes increases in timber sales will result in more jobs and therefore help stabilize timber dependent communities.

The Forest Service takes two big steps in this assumption. The first is that they assume all timber offered for sale is sold and harvested. For the period 1969 through 1989 this has not been the case. Even today in a time that the Montana wood products industry and the general public believe that the timber industry is being threatened because of a decrease in the supply of timber, the federal government has offered timber sales which it has determined to be profitable and has received no bids for them.

The Bureau of Land Management offered two million board feet of timber in the Upper Blackfoot salvage sale in the fall of 1990 and did not receive any bids for it. While two million board feet does not seem like a lot of timber when compared to the 1.279 billion board feet harvested in the entire state of Montana last year, it is a lot when compared to the .3 million board feet of suitable timber that is estimated to be in all of the roadless lands along the Rocky Mountain Front on the Lewis and Clark National Forest. It is assumed that all the timber industry's economic woes in Montana will be cured if our roadless lands are unlocked and opened up for logging.

Charles Keegan of the University of Montana Bureau of Business and Economic Research believes that slumping public timber sales may drastically shrink Montana's timber industry by 30 percent within the next five to seven years. If this were true we should expect a loss of over 2700 jobs in the wood products industry. This view of a restricted timber supply costing jobs is also held by people on the other end of the spectrum. Congressman Pat Williams (D-MT), generally considered a strong supporter of the environment, recently asserted in a speech to the Montana Mining Association that Montana's timber industry has been crippled in the last decade by administrative appeals of timber sales, which have limited the supply of timber. This assumption has led to a sense of urgency in resolving the status of Montana's remaining 6 million acres of unprotected roadless land.

The timber industry would like all of the 6 million acres of roadless lands in Montana opened up for timber sales. They have been campaigning with the theme that if these lands were harvested, more jobs would result. The Forest Service uses this same assumption in their forest plans. Using an input-output model the Forest Service has come up with an estimate of the number of jobs which result when they harvest an additional unit of timber. The input-output model's underlying principle is that a unit of timber, one million board feet, is followed through the production process until it reaches its final form as processed lumber. The additional jobs that result from the additional unit of timber harvested is the figure they assign for every additional unit harvested over the next 150 years. The Forest Service estimates that 5 jobs will be created for every one million board feet of timber harvested.

Fred Stewart, the economist for the Lolo National Forest, informed me that the data used in the most recent input-output model is from 1972. The problem with this is that Montana's timber industry has changed drastically in the last 20 years. In the 1970's the labor intensive plywood and fiberboard industry expanded resulting in record employment. Employment in the wood products industry peaked in 1979 at 11,606 workers. Beginning in late 1979 the forest products industry experienced its worst slump since the Great Depression. This slump started with a general decline in housing construction and was extended in the mid 1980's by the very low lumber prices that were caused in large part by the very high U.S. dollar. Lumber imports from Canada greatly increased. In the late 1980's a lowered U.S. dollar and higher consumption resulted in the timber industry rebounding. In 1987 a record amount, 1.376 billion board feet of timber was harvested in Montana but only 9,093 people were employed. The lower employment was due primarily to technological changes instigated by the very competitive markets of the early 1980's and smaller diameter logs harvested. The smaller logs are due to the fact that most of the old growth forest in low elevation lands has already been cut. The smaller diameter logs have forced the timber industry to shift from one made up primarily of sawmills in the early 1970's to one made up primarily of stud, plywood and particle board mills today. Charles Keegan estimates that a million board feet of timber for a stud mill results in only 2.68 jobs today. There is no reason to believe that the timber industry will not change again in the next 20 years let alone the next 50 years, but the Forest Service uses this constant figure of 5 jobs per million board feet to state exactly how many additional jobs will result a half a century from now.

To try and check the Forest Service estimate of 5 jobs per million board feet of timber harvested I ran a regression of the number of employees in the wood products industry as a function of the amount of timber harvested. To account for technological change I added a time variable. The Forest Service Region One office in Missoula provided me with data on the total amount harvested and the total employment in the wood products industry for the years 1969 through 1989.

TABLE 1

From 1909 to 1968 the Forest Service recorded the amount harvested from National Forest lands in Montana but this did not include the harvest from any other lands. Even this data is not estimated to be very accurate. Because of the unreliability of these numbers and the unstable percent the Forest Service harvest is of the total harvest there was no reliable way to estimate the total timber harvest in Montana for that period. The figure for total harvest from 1969 through 1989 includes the harvest from state lands, private lands, Native American Reservation lands, and Bureau of Land Management lands.

I assumed a regression would show a strong relationship between the amount harvested and the number of jobs in the wood products industry. By the very nature of the business common sense tells me to agree with the Forest Service. The more timber that is harvested the more workers should be needed to process it. I only questioned the magnitude of the figure they arrive at and the fact that it is a constant.

I did not include any figures for timber harvested in Montana but processed out of state or for timber processed in Montana but harvested out of state. The Bureau of Business and Economic Research found that the timber we export without processing was offset during this period by an equal amount of timber imported into Montana to be processed.

An OLS regression of Employment on Year and Harvest resulted in an R squared of .01 with none of the variables being significant. The Durbin Watson D statistic was .99 which was significant at 97 percent level of confidence. To correct for this autocorrelation I ran the canned Auto package on Shazam. The canned Auto package gave me a Rho of .73 which means that there is positive auto correlation. A disturbance in the industry effects more than one year. A shut down of a mill does not just have an effect on the year the plant shut down. Just as when a mill is retooled it does not just mill timber more efficiently the year the investment in it was made.

Year was not significant because changes in technology have not been constant. When I took Year out of the model my R squared was .428 and my adjusted R squared was .398. Major investments in capital have instead come in spurts. Smaller diameter trees require less labor intensive processing. I assume trees have gradually become thinner but the industry has processed the larger diameter trees until they ran out. When they did, they retooled the mills to handle smaller diameter logs. Stud mills are less labor intensive because a laser takes a picture of the log and automatically adjusts the saws to make the most efficient cut.

A plot of the data shows that harvest was high in the beginning of the period and again at the end. But employment shows just the opposite: it was low when harvest was high and it peaked in the late 1970's when the harvest was low. This can be explained by the increased use of residual wood in the late 1970's when improvements were made at the linear board mill in Frenchtown, Louisiana Pacific built their fiberboard mill in Missoula and Champion International constructed plywood mills in Libby and Bonner. This is reflected in the fact that the 95 percent of the wood fiber residue from Montana's mills was utilized in 1981 compared to 63 percent in 1969. Data for 1979, the year employment peaked, was not available.

The 1980's contradicted this trend for an increase in employment. Forty-six sawmills closed between 1981 and 1988. The industry now consists of eighty-seven sawmills, four plywood plants, a particleboard plant, a medium density fiberboard plant, a pulp and paper mill, two planing mills, thirty-seven post and pole plants, thirty-five house log plants, four wood pellet plants, three cedar product plants, two utility pole plants, one facility generating electricity, and a roundwood dipping plant.

Lumber production has become concentrated in larger mills. The average output per sawmill rose from 7.5 million board feet in 1981 to 17.9 million board feet in 1986. These new, more capital intensive mills require less labor than the smaller mills they replaced. The data from 1987 provides a good example of this. Though we had a record harvest, employment was down by more than 2400 people from its peak in 1979. After checking for heteroscedasdicity and finding none, the model indicates 3.3 jobs are created for every additional million board feet of timber harvested.

If I leave out the most influential outlier, 1982, when only 7900 workers harvested and processed 828 million board feet of timber, the coefficient drops to 1.67 workers for every additional million board feet of timber harvested. The R squared for this new model increases to .447 and the adjusted R squared rises to .416. Both models over-estimate the number of workers in 1989 by about 500.

The tremendous investment that we have seen in capital in Montana's wood products industry allows the mills to fluctuate the amount of timber they harvest without fluctuating the amount of labor. The 1970's showed a more labor intensive trend but the 1980's contradicted this with extraordinary investment in capital intensive stud mills. Where will the 1990's take us? My coefficients do not point toward using 5 workers for every additional million board feet cut. But my low R squares state that my coefficient is not reliable. Therefore it is ludicrous to try and use it to predicate the number of jobs at the end of 50 years as the Forest Service does. As absurd as it sounds to state that we can stop all harvesting with a result of no loss of timber industry jobs, it is equally absurd to state that if we harvest more timber it will result in 5 more jobs per million board feet now or any more jobs 10 or 50 years from now. The wood products industry constantly changes. There is no reason to believe that the industry will not become more capital intensive over the next decade. The components of the industry seem to have more effect on the number of jobs than the supply of timber. But I cannot predicate that another industry might skyrocket such as the log home industry. The Bureau of Business and Economic Research has found that the log home industry has a labor ratio of 36 workers per million board feet of timber it processes. Perhaps the Forest Service should base their estimate of the number of wood products jobs on the make up of the industry and not on the size of the timber harvest.

The regression shows that increased timber harvesting is not a guaranteed solution to Montana's economic problems or a guarantee that new timber jobs will result or even that existing ones can be maintained. This implies that the National Forest Plans should not be written as if such gurantees exist. The use of our remaining 6 million acres of unprotected roadless lands should be based on other assumptions rather than a very questionable one that more jobs will result if we cut all our remaining pristine forests down.

BIBLIOGRAPHY

Keegan, Charles E. et al. Montana's Forest Products Industry: A Descriptive Analysis 1969-1988. Missoula, Montana. Bureau of Business and Economic Research, School of Business Administration, University of Montana, 1990.

Power, Thomas Michael. "To Be or Not To Be? The Economics of Development Along the Rocky Mountain Front." Western Wildlands, Missoula, Montana. University of Montana. Fall 1987.

U.S. Department of Agriculture. Flathead National Forest Plan. Kalispell, Montana 1985.

U.S. Department of Agriculture. Lewis and Clark National Forest Plan. Great Falls, Montana. 1986.

U.S. Department of Agriculture. Lolo National Forest Plan. Missoula, Montana. 1986.

White, Kenneth J. et al. Shazam, Econometrics Computer Program. New York, McGraw-Hill, 1990.