National Forest Timber Sale Losses in 1993 and 1994

by Randal O'Toole

Did the Forest Service make $412 million on timber sales in 1993, and $214 million in 1994, as it claims? Or did it lose $442 million in 1993 and $309 million in 1994, as cash-flow figures show? If it lost money, were the losses a subsidy to the timber industry--or to someone else?

These questions are raised by a Forest Service accounting system known as the Timber Sale Program Information Reporting System (TSPIRS). Each year, usually in March, the Forest Service publishes TSPIRS results for every national forest, accompanied by numerous press releases claiming that the agency made hundreds of millions of dollars in profits on timber sales.

TSPIRS is one program that hasn't been touched by reinvention. To understand this system, it is most important to understand what it is not:

TSPIRS does not compare returns to the Treasury with costs to the Treasury. Instead, it compares all timber revenues, only some of which go to the Treasury, with only some of the sale costs. Overestimated revenues and underestimated costs lead TSPIRS to greatly overestimate profitability.
TSPIRS does not apply to individual timber sales. Instead, it applies only to a national forest's entire sale program for one year.
TSPIRS is not used by Forest Service managers to judge the merits of proposed timber sales before selling them. Since it applies to sale programs, not individual sales, it couldn't be used this way. Even if it could, employees are not evaluated according to their forests' TSPIRS results, so they have no incentive to worry about them while planning sales.
If managers did use TSPIRS to judge sales, they would make the wrong decisions. TSPIRS does not count important costs, such as the costs of basic road construction other than culverts, bridges, and surfacing. Thus, a sale that required new road construction would be scored as favorably as one that merely resurfaced existing roads.

The Forest Service designed TSPIRS in response to Congressional worries about below-cost timber sales. (Typically, Congress gave the agency a supplemental appropriation to develop the accounting system.) When it was being designed, the people on the work team told timber industry and other agency employees that "the purpose of TSPIRS is to prove to the public that the benefits of timber sales exceed the costs."

The team's method was simple. It picked the forest that it thought had the worst cash flow of any in the country, which turned out to be the Black Hills Forest. It then designed a system that would make it appear that the Black Hills Forest makes money.

Not surprisingly, then, TSPIRS claims that the Black Hills Forest makes money. It also reports that other perennial money-losers, such as the Tongass Forest, make money. But if the team's mission was to prove that all forests made money, it failed. In every year since TSPIRS was first published in 1988, it concluded that a majority of the 122 forests lost money.

(For the purposes of counting, the Tongass is counted as three forests; for 1994 the count increases to 124 with the creation of the Columbia River Gorge as a separate counting unit and the division of the Wayne and Hoosier into two units.)

Nevertheless, Forest Service watchdogs such as Robert Wolf, the Wilderness Society, and CHEC (now the Thoreau Institute) took strong exception to TSPIRS's results. After reducing revenues to account for only those going to the Treasury and restoring the costs left out by TSPIRS, their results typically showed that 80 percent or more of all national forests lost money.

Different Drummer's predecessor, Forest Watch published TSPIRS results for every year from 1989 through 1992. We didn't publish results for 1993 for various reasons, so to be complete this issue includes results for both 1993 and 1994. The data in the tables on pages 32 through 35 come from three Forest Service computer files:

Basic TSPIRS data for every forest is contained in a file known as "TPIR01." This file shows how TSPIRS adjusts costs, such as road costs, but it also contains the raw accounting information on actual costs. I used these raw data.
Timber sale revenues go into several accounts, some of which are kept by the Forest Service in its Knutson-Vandenberg, salvage sale, and other funds, and some of which go to the Treasury. TSPIRS doesn't identify these accounts, so the Forest Service provided me with a special data file.
Some of the line items in the TPIR01 reports are paid by the Treasury, such as "appropriated reforestation, while others are paid out of timber sale receipts, such as "K-V reforestation." A few costs, such as "sale preparation," are paid by a mixture of Treasury dollars and salvage sale receipts. TSPIRS does not distinguish which is which, so the Forest Service prepared another data file showing salvage sale costs.

Table One
Forest Service Timber Sale Receipts and Costs
(millions of dollars except number of forests)

                          	 1989	 1990	 1991	 1992	 1993	 1994	Average
Numbers According to TSPIRS
TSPIRS Revenues           	1,518	1,376	1,157	1,077	1,017	  885	1,172
TSPIRS Costs              	  664	  746	  685	  601	  717	  672	  681
Net to Treasury           	  855	  630	  472	  475	  412	  214	  510
Number of Forests Losing  	   64	   65	   71	   68	   74	   68	   68
Losses of Forests Losing  	  -47	  -48	 -159	  -66	  -80	  -60	  -76
Numbers According to Cash Flow
Revenues to Treasury      	  872	  798	  641	  482	  443	  393	  605
Costs to Treasury         	  661	  327	  655	  630	  597	  421	  549
County Payments           	  339	  516	  301	  306	  288	  280	  338
Net to Treasury           	 -128	  -46	 -315	 -453	 -442	 -309	 -282
Number of Forests Losing  	  104	  101	  109	  109	  115	  104	  107
Losses of Forests Losing  	 -366	 -296	 -437	 -499	 -501	 -364	 -410

The Amortization Debate

My analysis of a national forest's timber sale program is based on a strict cash flow in and out of the Treasury. The Forest Service argues that this is too simplistic because some costs, such as roads and reforestation, should be amortized over time. Amortization of long-term investments may seem appropriate, but even amortized costs have to eventually be covered by receipts. The Forest Service method makes no effort to make sure that this will ever happen; instead, it more or less wishes those costs away.

For example, between 50 and 60 percent of all road costs go into what the Forest Service calls the "prism," which includes all the work of excavating and grading the road. These costs are not counted in TSPIRS because, the Forest Service says, they "add to the capital value of the forest." But just because money is spent doesn't mean land is more valuable. If it did, a homeowner with a backyard swimming pool could increase the home's value by spending money to fill the pool with concrete.

The policy of neglecting prism costs meant that $54 million in appropriated funds were left out of the 1993 TSPIRS results and $24 million were left out in 1994. In addition, TSPIRS counts purchaser road credits among the receipts, but left out $43 million in road credit prism costs in 1993 and $63 million in 1994, greatly improving apparent profitability. The neglect of prism costs alone meant that TSPIRS overestimated profitability by $184 million for 1993 and 1994 together.

Reforestation and related costs are supposed to be amortized over a rotation, but the formula used miscalculates rotation ages. According to the formula, the average rotation age for acres reforested in 1994 is 137 years--far longer than realistic. In both 1993 and 1994, ten forests use rotation ages greater than 400 years and 35 use rotations more than 200 years. Actual rotations average under 100.

Who Is Being Subsidized?

Table one summarizes the results for the past several years, including the number of forests that lost money, how much the system lost overall, and how much the money-losing forests alone lost. The table shows both TSPIRS's results and the results of cash-flow comparisons of Treasury receipts and costs.

Environmentalists usually describe these results as direct subsidies to the industry. But these losses do not measure subsidies to the industry. Instead, they are mainly subsidies to the Forest Service bureaucracy. Agency managers have learned that a pork-hungry Congress will spend just about anything to convince managers to get out the cut. So the agency has run up most timber-related costs to levels that are far greater than any private landowner--or even any state agency--would spend.

The TSPIRS data show that a phenomenal share of the costs--more than $150 million per year, or nearly 40 percent of 1994's total timber-related costs--are overhead. Half of this is "general administration"--salaries of line officers and office support--at the forest level. The rest is money spent by the regional and Washington offices, including $25 to $35 million spent out of timber sale line items, $7 to $13 million spent out of road items, and $6 to $9 million from reforestation line items.

These overhead costs also appear to be rather inflexible. Although direct timber-related costs to the Treasury declined by 40 percent between 1993 and 1994, overhead costs declined by less than 3 percent. Thus, as the timber sale program shrinks, the Forest Service's bloated bureaucracy makes that program appear increasingly unprofitable.

In addition, some $280 million of payments to counties each year represent subsidies to a few counties. These payments are supposed to replace the property taxes that the counties would get if the lands were private. But many counties get less than property tax equivalency, while a few--mainly those in Pacific coast states--get far more.

The timber industry does receive some subsidies, but they are impossible to calculate with Forest Service data. The industry gets two different subsidies. First, due to limited competition, purchasers in some areas get to buy timber for well below the market value for similar timber in areas with competition. Forest Service timber pricing should, but does not, account for this. This subsidy is probably in the tens of millions of dollars per year.

The second subsidy is represented by the timber that the Forest Service wouldn't sell if it managed the national forests sensibly--which means for a profit and with due regard for the values of other resources. This subsidy is much more difficult to estimate but is probably larger than the first.

How do timber subsidies compare with subsidies to other resources? Table two compares subsidies estimated in two different ways. First is the subsidy based on costs to the Treasury minus receipts to the Treasury. The table's numbers are based on the Forest Service's 1994 budget, not TSPIRS, and do not include overhead or indirect expenses, so the numbers for timber differ from TSPIRS. The table shows that the subsidy to recreation is comparable to that for timber, while the wildlife subsidy is significantly greater than the grazing subsidy.

But subsidies based on costs are, as noted above, subsidies to the bureaucracy, not to the users. A good measure of user subisidies is the market value of the use minus the actual payments made by users. According to the Forest Service's 1990 RPA Program, the subsidies to recreation and wildlife when measured this way are overwhelmingly larger than the subsidies to timber and grazing.

The RPA Program probably overestimates the market value of recreation and underestimates the timber revenues lost due to lack of competition. But if the timber subsidy were 100 times greater and the recreation-wildlife subsidy were ten times smaller, the latter would still be six times greater than the former.

Table Two
Resource Subsidies Measured Two Ways
(millions of dollars per year)

                 	By Cost	By Lost Revenue
Timber           	$222	   $1
Grazing          	  13	   40
Recreation       	 220	3,955
Fish & Wildlife  	  84	2,664

Subsidies by cost measure the 1994 returns to the Treasury minus the 1994 costs to the Treasury, not counting overhead. Subsidies by lost revenue measure total market value minus actual payments by resource users and are based on the Forest Service's 1990 RPA Program, p. 6-53, using projections for 1995. When measured by cost, amenity subsidies are comparable to commodity subsidies. When measured by lost revenue, amenity subsidies are overwhelmingly larger.

Some Curious Results

Without counting any costs at all, Treasury receipts were negative on 9 forests in 1994 and 13 in 1993. (This is indicated in the tables when county payments plus "kept by Treasury" is less than zero.) This can happen when the Forest Service decides to keep more of a previous year's receipts that it had originally kept.
When deducting only 25 percent county payments and no other costs, 57 forests lost money in 1994 and 60 in 1993.
When deducting only county payments and direct sale costs--no roads, reforestation, or overhead--90 forests lost money in 1994 and 101 in 1993.
The Tongass Forest lost money before counting costs because of the peculiar way it handles road credits under the Ketchikan Pulp Company long-term timber sale contract. This contract began in the 1950s but did not incorporate purchaser road credits until 1975. Since then, the purchaser has been allowed to credit its road costs against payments it made for the timber before 1975. As a result, the Forest Service paid the purchaser millions of dollars for cutting national forest timber in 1992 and 1993.
Since 1989, average timber values have increased steadily from $127 per thousand board feet to $185 (see figure). But the share retained by the Forest Service (including purchaser credits) has nearly doubled, from $54 to $103 per thousand. Since 25 percent county payments have also increased, this means that the share retained by the Treasury has fallen from $45 to $23 per thousand.

Explanation of Tables

The downloadable data tables (25K, tab delimited) present timber sale results by national forest for 1993 and 1994. Each year has sixteen columns of data, as follows:

Total Revenue is one number that the Forest Service and the Thoreau Institute agree upon. It includes purchaser road credits and payments going into the K-V and other Forest Service funds, with the Treasury getting the leftovers. TSPIRS ignores these distinctions.
TSPIRS Cost is the Forest Service's sum total of costs. This includes sale, road, reforestation, general administration, and NEPA/appeals/litigation costs. It also includes costs paid out of the Treasury as well as costs paid out of timber receipts such as the K-V fund. This sounds comprehensive, but many of the costs are "amortized" or, in some cases, simply left out. For example, all subsurface road costs--the most expensive part of road construction--are left out.
TSPIRS Net is total revenues minus TSPIRS costs.
Kept by FS shows how much of the revenue was retained by the Forest Service in its K-V, salvage sale, brush disposal, or other funds.
Road Credits shows how much of the revenue was actually paid in kind in the form of road construction.
County Payments shows how much counties get at 25 percent of the gross (including K-V, salvage sale funds, and purchaser credits but not brush disposal or cooperative road maintenance). Actually, some counties get more than 25 percent, but that is another story.
Kept by Treasury shows how much is left over after deducting road credits, county payments, and the Forest Service's share from total revenues. At this point it turns out that 60 forests lost money in 1993 and 57 in 1994 without even counting any costs.
Sale Costs includes inventory, planning, sale preparation, harvest administration, and resource support paid out of tax dollars, but not any costs paid out of timber receipts. Costs paid out of the salvage sale fund have been carefully deducted.
Road Costs includes road engineering and construction costs paid out of tax dollars but not purchaser credits.
Ref & TSI includes reforestation and timber stand improvement costs paid out of tax dollars but not out of the K-V fund.
Forest Overhead includes general administration related to timber paid at the forest level.
RO-WO Overhead includes timber costs paid out of tax dollars at the regional, Washington, and other non-forest offices and disaggregated to forests based on volume of timber cut.
NEPA Appeals includes the costs of NEPA planning and appeals and litigation.
Volume Cut shows the volume of timber cut on each forest in millions of board feet.
Total Costs shows the costs to the Treasury, including all the costs above from sale costs to NEPA appeal costs.
Net to Treasury equals the dollars kept by the Treasury minus the costs to the Treasury. Turns out that 115 forests lost money in 1993 and 99 lost in 1994. Total losses in 1993 were $442 million, but the losses of just the forests losing money were $501 million. In 1994 the losses were $309 million and $364 million respectively.

Return to listing of Reinventing the Forest Service excerpts.