This file was created by scanning the printed publication. Text errors identified by the software have been corrected;

however, some errors may remain.

AuthorsEDWARD C. WATERS is a private consultant and assistant professor (courtesy appointment),Department of Agricultural and Resource Economics, Oregon State University, Corvallis, OR97331-3601; DAVID W. HOLLAND is a professor, Department of Agricultural Economics,Washington State University, Pullman, WA 99164-6210; and RICHARD W. HAYNES is aresearch forester, U.S. Department of Agriculture, Forest Service, Pacific Northwest ResearchStation, Forestry Sciences Laboratory, P.O. Box 3890, Portland, OR 97208-3890.

The Economic Impact of Public Resource SupplyConstraints in Northeast Oregon

Edward C. Waters, David W. Holland, andRichard W. Haynes

Interior Columbia Basin Ecosystem ManagementProject: Scientific Assessment

Thomas M. Quigley, Editor

U.S. Department of AgricultureForest ServicePacific Northwest Research StationPortland, OregonGeneral Technical Report PNW-GTR-398April 1997

AbstractWaters, Edward C; Holland, David W.; Haynes, Richard W. 1997. The economic impact of

public resource supply constraints in northeast Oregon. Gen. Tech. Rep. PNW-GTR-398.Portland, OR: U.S. Department of Agriculture, Forest Service, Pacific Northwest ResearchStation. 23 p. (Quigley, Thomas M., ed.; Interior Columbia Basin Ecosystem ManagementProject: scientific assessment).

Traditional, fixed-price (input-output) economic models provide a useful framework for concep-tualizing links in a regional economy. Apparent shortcomings in these models, however, canseverely restrict our ability to deduce valid prescriptions for public policy and economic develop-ment. A more efficient approach using regional computable general equilibrium (CGE) models aswell as a brief survey of relevant literature is presented. Computable general equilibrium resultsunder several different resource policy scenarios are examined and contrasted with a fixed-priceanalysis. In the most severe CGE scenario, elimination of Federal range programs caused the lossof 1,371 jobs (2.3 percent of regional employment) and $29 million (1.6 percent) of householdincome; and an 80-percent reduction in Federal log supplies resulted in the loss of 3,329 jobs (5.5percent of regional employment), and $76 million (4.2 percent) of household income. Theseresults do not include positive economic impacts associated with improvement in salmon runs.Economic counter scenarios indicate that increases in tourism and high-technology manufactur-ing and growth in the population of retirees can largely offset total employment and incomelosses. Evidence suggests that a regional economic transformation, characterized by stable ordeclining employment in resource-based industries and growth in tourism and retirement-basedservices, is already underway.

Keywords: Regional economics, computable general equilibrium models, input-output models,economic base theory, public resource policy.

PrefaceThe Interior Columbia Basin Ecosystem Management Project was initiated by the Forest Serviceand the Bureau of Land Management to respond to several critical issues including, but notlimited to, forest and rangeland health, anadromous fish concerns, terrestrial species viabilityconcerns, and the recent decline in traditional commodity flows. The charter given to the projectwas to develop a scientifically sound, ecosystem-based strategy for managing the lands of theinterior Columbia River basin administered by the Forest Service and the Bureau of Land Man-agement. The Science Integration Team was organized to develop a framework for ecosystemmanagement, an assessment of the socioeconomic and biophysical systems in the basin, and anevaluation of alternative management strategies. This paper is one in a series of papers developedas background material for the framework, assessment, or evaluation of alternatives. It providesmore detail than was possible to disclose directly in the primary documents.

The Science Integration Team, although organized functionally, worked hard at integrating theapproaches, analyses, and conclusions. It is the collective effort of team members that providesdepth and understanding to the work of the project. The Science Integration Team leadershipincluded deputy team leaders Russel Graham and Sylvia Arbelbide; landscape ecology-WendelHann, Paul Hessburg, and Mark Jensen; aquatic—Jim Sedell, Kris Lee, Danny Lee, Jack Williams,Lynn Decker; economic—Richard Haynes, Amy Home, and Nick Reyna; social science-JimBurchfield, Steve McCool, and Jon Bumstead; terrestrial-Bruce Marcot, Kurt Nelson, JohnLehmkuhl, Richard Holthausen, and Randy Hickenbottom; spatial analysis-Becky Gravenmier,John Steffenson, and Andy Wilson.

Thomas M. QuigleyEditor

United StatesDepartment ofAgriculture

Forest Service

United StatesDepartment ofthe Interior

Bureau of LandManagement

Interior ColumbiaBasin EcosystemManagement Project

IntroductionDuring the 1980s, annual timber harvest frompublic lands in Oregon consistently approachedhistoric high levels. In 1990, timber harvests inwestern Oregon began a dramatic decline,largely reflecting increasing land use restrictionsbecause of the listing of the northern spottedowl (Strix occidentalis caurina) as an endangeredspecies. East of the Cascade Range, however,the decline has been less precipitous. Althoughtotal cut on west-side National Forests (USDAFS) decreased by 81 percent from 2,238 millionboard feet Scribner rule (MMBF) in 1988 to416 MMBF in 1993, over the same period,east-side Forest harvest declined by 45 percentfrom 1,250 MMBF to 685 MMBF (table 1).This trend also is evident on Bureau of LandManagement (BLM) lands, although on theeast side, BLM harvests are of relatively minorimportance (Oregon Department of Forestry1988-93).

Table 1 also shows an opposite, positive re-sponse for private timber. Private harvestincreased by 93 MMBF (3 percent) on the westside and by 257 MMBF (51 percent) on theeast side. As a result, total harvest from thesethree sources decreased by 44 percent on thewest side but by only 18 percent on the eastside, between 1988 and 1993.1

Historically, natural resource industries haveprovided a significant portion of Oregon'semployment and income. In 1992, livestockand timber-related industries together directlyprovided an estimated 105,000 jobs, represent-ing about 6.5 percent of total jobs and 7.7percent of employee compensation paid in theState (table 2). This total does not includeseveral thousand resource management jobs inFederal, State, and local governments.

In the five northeast Oregon counties focusedon for this study (that is, Baker, Grant,

1 Care should be taken when comparing historicalharvests. As cuts include more private and younger agedtrees, the ratio of total processed output to log input willchange, thereby creating a measurement problem. Thelog-scale measurement was established lor large trees andfor times when logs were less fully used. As a result,private timber harvest may contain 20 to 30 percentmore effective input than the log scale is measuring.

Umatilla, Union, and Wallowa Counties),employment in livestock and timber-relatedsectors is significantly more important. Ofabout 60,000 regional jobs in 1992, livestockdirectly provided 5,300 jobs. Logging, lumber,and wood products provided another 4,100jobs. Together, these industries contributedabout 1 5 percent of regional employment andemployee compensation (table 2).

Recent concern over the viability of severalnative anadromous fish runs has jeopardizedtraditional public land uses. Ecological disturb-ance caused by logging and riparian cattlegrazing has been shown to damage salmonspawning beds. In July 1994, the Ninth CircuitU.S. Court of Appeals issued an injunction thatwould effectively stop grazing activities (andeventually, perhaps, logging and road building)in the Umatilla and Wallowa-Whitman Na-tional Forests. To help maintain viable spawninghabitat for threatened or endangered salmon,considerable change in east-side public landmanagement policy may be necessary. Conse-quently, how much future access will be avail-able to much of the remaining east-side publicforest land for timber harvest and cattle grazingis uncertain.

Although it is unclear what changes in loggingand grazing policy will be introduced, therelative importance of livestock and timber-related industries in northeast Oregon suggeststhat any reduction in access to these naturalresource inputs could have significant impactson regional employment and income.

PurposeThe main purpose of this study was to estimatethe regional economic impact of alternativeForest Service policies regarding access to publicresources in northeast Oregon. As a prelude tothis analysis, we provide a historical overview ofthe economy of northeastern Oregon over thepast quarter of a century.

Estimates of total change in regional employ-ment and income are provided under differentassumptions about the nattire and size of thepolicy shift and length of time over which ad-justment to the new policy occurs. Analysis islimited to examining the impact of hypotheticalresource supply constraints resulting from

1

Table 1—Annual Oregon timber harvest volume3

Ownership

USDAFSBLMPrivate

Total

1988

West

2,2381,3992,754

6,391

East

1993

West

Million board feet

1,250 41640

505

1,795

3392,847

3,602 1

East

68623

762

,471

Change 1988 to 1993

West

MMBF

-1,822-1,060

+93

-2,789

Percent

-81-76+3

-44

East

MMBF

-564-17

+257

-324

Percent

-45-43+51

-18

' Volume is measured by using Scribner rule.

Source: Oregon Department of Forestry, 1988-93.

Table 2—The northeast Oregon regional economy, 1992a

Employment sectorsb

Livestock relatedOther agricultureFood processingLogging and wood processingOther agriculture and natural resourcesOther employment

Total

State

Numberof jobs

24,46142,45425,05381,07426,110

1,420,848

1,620,000

Percent

1.52.61.65.01.6

87.7

100.0

Numberof jobs

5,3533,7442,7924,1011,484

42,294

59,768

Region

Percent

8.96.34.76.82.5

70.8

100.0

aNortheast Oregon = Baker, Grant, Umatilla, Union, and Wallowa Counties.b Full- and part-time employment.

Source: IMPLAN (Alward and others 1989) and Bureau of Economic Analysis, Regional Economic Information System, U.S.Department of Commerce, 1969-92.

Table 3—Total and per capita personal income over time for

Category

Total personal income (thousand $)

Population (thousands)

Per capita personal income ($)

Total employment (jobs)

1969 1974 1979 1984

304,105.0 540,576.0 915,643.0 1,251,392.0

92.2 97.9 111.8 116.8

3,298.0 5.522 0 8,190 0 10,714.0

41,093 0 45,749 0 54,569.0 53,336.0

northeast Oregona

1989 1990 1991

1,542,637.0 1,632,475.0 1,717,510.0

112.0 1133 114.7

13,774.0 14,408.0 14,974.0

57,049.0 59,146.0 59,644.0

1992

1,818,414.0

116.1

15,662.0

59,768.0

aNortheast Oregon = Baker, Grant, Umatilla, Union, and Wallowa Counties.

Source: Bureau of Economic Analysis, Regional Economic Information System, U.S. Department of Commerce, 1969-92.

2

implementation of salmon habitat protectionand improvement measures on Federal lands.No attempt was made to estimate the effects ofother possible policy changes (for example, riverflow augmentation, commercial harvest restric-tion, and reduction in recreational access).

Besides a traditional analysis using a conven-tional, demand-driven fixed-price model,2 thisstudy also presents results obtained by usingflexible-price, computable general equilibrium(CGE) models of the regional economy. Thesemodels provide several advantages over moreconventional regional models. The structure ofa CGE model is more consistent with neoclassi-cal economic theory and flexible enough toincorporate factor and commodity substitutioninto the structure of production and demand.

A Profile of the NortheastOregon Regional Economy

The following discussion is based largely ondata obtained from the Regional EconomicInformation System (REIS) of the Bureau ofEconomic Analysis (BEA). The data describethe regional economy of Baker, Grant, Union,Umatilla, and Wallowa Counties in northeastOregon from 1969 to 1992. The region con-tains no metropolitan counties and is economi-cally related to larger urban centers in Lewiston,Idaho, and Walla Walla, Washington.

More than 116 thousand persons presentlyreside in the five-county, northeast Oregonregion. From 1969 to 1992, regional popula-tion grew from roughly 92,200 to 116,100persons (table 3). The number of full- andpart-time jobs increased from roughly 41,100in 1969 to 59,800 jobs in 1992.

According to our estimates for 1992, full- andpart-time employment in agriculture and ag-

2 For example, input-output (IO), economic base, andsocial accounting matrix (SAM) models. These types ofmodels produce constant, marginal multipliers (that is,invariant of the size of the economic shock) by assumingthat factor (labor and capital) supply constraints arenonbinding; and that factor demand ratios, commoditysupply proportions, and all prices are fixed. Multipliersderived from fixed-price models generally can be describedas providing an upper bound on the amount of economicimpact resulting from an exogenous demand shock.

ricultural processing was 13,373 jobs (table 2).The corresponding employment figure for thelogging and woods products industry was 4,101jobs. Agriculture and agricultural processingcontribute 22.4 percent of total jobs, whereaslumber and wood products constitute 6.9percent of total jobs. Combining the twoindustries, we find that 29.3 percent of totaljobs in the region are in traditional naturalresource-based industries. By any measure,resource-based industry is important to theregion. In the cases of cartle production andlogging and wood products industries, muchof the resource base is held as public lands. Be-hind these aggregate figures, however, stands adramatic, ongoing transformation of the re-gional economy.

We are especially interested in establishing thecurrent economic baseline for this economy andreviewing structural changes that have charac-terized this economy over the last two decades.As we move into the future, it will be importantto understand the past and the implications ofthat past for future patterns of economicdevelopment.

Sources of Personal Income

Personal income as defined by the BEA comesfrom five different sources: (1) employee com-pensation, (mainly wages and salaries); (2)interest, dividends, and rent (income derivedfrom property ownership); (3) transfers (mainlysocial security and medicare payments); (4) pro-prietor income (income to self-owned busi-nesses and farms); and finally (5) adjustmentsto account for people who receive income inthe area but do not live there (for example,commuters). The relative contribution of eachof these components to personal income issummarized in table 4.

The major components of earned income arcemployee compensation and proprietor income.In 1969, employee compensation made up56.47 percent of total personal income. By1992, the share had declined to 49.6 percent. In1969, proprietor income contributed 21.16percent of total personal income, 44 percentfrom farms and 56 percent from nonfarmsources. By 1992, the contribution of propri-etor income to personal income had declined to

3

Table 4—Components of per capita personal income over time for northeast Oregona

Personal income category

Employee compensation per capitaDividend interest and rent per capitaTransfers per capitaProprietors per capitaResidence adjustment per capita

Personal income per capita

Employee compensationDividends, interest and rentTransfersProprietorsResidence adjustment

Total percent

1969

1,922409365720-12

3,403

56.4712.0310.7121.16

-.37

100.00

1974

2,782688751

1,43945

5,705

48.7612.0713.1725.22

.78

100.00

1979

4,6361,1941,2231,150

260

8,462

54.7814.1114.4513.593.07

100.00aNortheast Oregon = Baker, Grant, Umatilla, Union, and Wallowa Counties.

1984 1989

Current dollars5,4782,0092,0081,374

243

11,112

6,9132,4352,7352,040

247

14,369

Percent

49.3018.0818.0712.372.18

100.00

48.1116.9519.0314.20

1.72

100.00

1990

7,4822,466

2,9851,823

264

15,021

49.8116.4219.8812.14

1.76

100.00

1991

7,7422,4073,2761,933

263

15,622

49.5615.4120.9712.38

1.69

100.00

1992

8,0922,3993,5482,005

286

16,329

49.5514.6921.7312.28

1.75

100.00

Source: Bureau of Economic Analysis, Regional Economic Information System, U.S. Department of Commerce, 1969-92.

Table 5—Transfer payments over time for northeast Oregona

Transfer payment category

Retirement and disability insurancebenefit

Medical paymentsIncome maintenance benefitsUnemployment insurance benefitVeterans benefitsFederal education and trainingassistance

Other payments to individual

Total government paymentsto individuals

Payments to nonprofit institutionsBusiness payments to individuals

Total transfers

Population (thousands)

Transfers per capita ($)

1969

20,902.03,628.01,628.01,621.03,387.0

72.065.0

31,379.0

1,484.0753.0

33,616.0

92.2

365.0

1974

45,439.07,797.05,319.05,070.05,899.0

291.0379.0

70,284.0

2,037.01,249.0

73,570.0

97.9

751.0

1979

84,076.019,178.09,961.06,553.07,539.0

1,167.0716.0

129,275.0

4,555.02,878.0

136,708.0

111.8

1,223.0

1984 1989 1990

Current thousand dollars

141,695.040,322.015,303.015,023.09,061.0

2,268.0359.0

224,075.0

5,647.04,816.0

234,538.0

116.8

2,008.0

172,305.071,190.022,071.013,687.09,558.0

3,702.0427.0

293,006.0

6,699.06,569.0

306,274.0

112.0

2,735.0

186,536.082,387.024,721.016,581.09,891.0

3,764.0553.0

324,519.0

6,973.06,761.0

338,253.0

113.3

2,985.0

1991

201,275.095,117.028,937.021,893.0

9,785.0

3,529.0507.0

361,095.0

8,015.06,601.0

375,711.0

114.7

3,276.0

1992

213,747.0106,395.032,766.028,100.010,519.0

3,897.0696.0

396,131.0

8,790.06,980.0

411,901.0

116.1

3,548.0

•' Northeast Oregon = Baker, Grant, Umatilla, Union, and Wallowa Counties.

Source: Bureau of Economic Analysis, Regional Economic Information System, U.S. Department of Commerce, 1969-92.

4

12.28 percent, 23 percent of which was fromfarm and 77 percent from nonfarm sources.

Two major trends emerge from examination ofregional earnings data: (1) growth in wage,salary, and proprietor income has not kept pacewith growth in transfer and property income;3

and (2) farm income has declined from roughlyone-half of proprietor income in 1969 toroughly one-fourth of proprietor income in 1992.

Probably the most striking feature of thestructural change in personal income in theregion is the tremendous growth in transferincome (table 4). Transfers grew from 10.71percent of personal income in 1969 to 21.73percent of personal income in 1992. Transferswere the fastest growing component of personalincome, increasing twelvefold over the period(in unadjusted terms). Per capita transfersincreased from $365 in 1969 to $3,548 in1992, a nearly tenfold increase. By comparison,total per capita personal income increased lessthan fivefold over the same period.

The components of transfer income are dis-played in table 5. In 1992, most of the transferswere retirement and disability insurance ben-efits and medical payments. These two compo-nents make up nearly 78 percent of totaltransfer income to the region. In 1992, incomemaintenance benefit payments (the welfare partof transfers) were $32.7 million, 7.9 percent oftotal transfers. The other major component oftransfer income is unemployment insurancebenefits that constituted 6.8 percent of the total.

Transfers have become the second most impor-tant source of income in the region (afteremployee compensation), and thus, should beviewed as a vital part of the new economic baseof the region. The main components of trans-fer income are those received by retired personswho settle in the region and bring with them asignificant amount of transfer and other prop-erty income.

Although not as dramatic as the trend in trans-fer income, income from dividends, interest,and rent also is significant. The portion ofpersonal income from these sources increasedfrom 12.03 percent in 1969 to 14.69 percent in

3 This is true for most rural Oregon counties.

1992 (table 4). On a per capita basis, theportion accounted for by dividends, interest,and rental income grew nearly sixfold from$409 in 1969 to $2,399 in 1992.

We are not able to distinguish property incomeoriginating from within the region from thatoriginating outside the region. We suspect,however, tnat the portion coining from outsidehas increased with the increasing attraction ofthe region as a vacation and retirement destina-tion. Thus, property-type income also must beviewed as a part of the new economic base. Liketransfer income, it will be an important deter-minant of the direction of economic change inthe future. In 1992, transfer income and property-type income combined to make up 36 percentof total personal income in northeast Oregon.

Sources of Employment

Several features of the change in the structure ofregional employment are worth reviewing. Em-ployment growth by all levels of governmentdid not keep pace with private sector employ-ment growth. In 1969, government employ-ment made up 24 percent of total nonfarmemployment. By 1992, the share declined to22 percent (table 6).

The major structural change in private sectoremployment was the growth in service-sectoremployment to the exclusion of nearly all othersectors. In 1969, service employment made up23 percent of private nonfarm employment inthe region. In 1992, service employment madeup 30 percent of private nonfarm employment.The number of service jobs more than doubledfrom 5,996 jobs in 1969 to 12,335 jobs in1992 (table 6).

In relative terms, job opportunities in manufac-turing increased the least of the various employ-ment sectors. Although manufacturing jobsincreased from 6,292 in 1969 to 8,506 in 1992,these jobs declined as a percentage of total jobsin the region. In 1969, manufacturing ac-counted for 24 percent of total private sectorjobs. In 1992, jobs in manufacturing declinedto 20 percent.

Our final observation regarding regional em-ployment trends concerns proprietors, (that is,self-owned businesses and farms), an economic

5

sectof in which another major change is evi-dent. From 1969 to 1992, the number of farmproprietors declined slightly from 4,469 to4,393 (table 7). In the same period, the numberof nonfarm proprietors grew from 5,937 to10,774. Farm proprietors have been relativelyconstant, whereas the number of nonfarmproprietors has nearly doubled. The result isthat proprietor employment and income is nowdominated by nonfarm business. This was notthe case two decades ago.

Summary of Regional Trends

Firms will remain under pressure to controllabor costs to remain competitive. This willinvolve recent labor market trends such ascapital-labor substitution, corporate down-sizing, and more competition for good jobs.Much of future job growth is likely to go toproprietors (self-owned business) rather thanwage labor jobs. For example, from 1989 to1992, the number of wage and salary jobs grew

Table 6—Sectoral composition of employment over time for northeast Oregona b

Employment category 1969

Private nonfarm employment:Agricultural services and

other ANRC

MiningConstructionManufacturingTransportation and public

utilitiesWholesale tradeRetail tradeFinance, insurance and

real estateServices+

Total private nonfarmemployment

Government employment:Federal, civilianMilitaryState and local

Total governmentemployment

Total nonfarm employmentFarm employment

Total employment

a Full- and part-time employment.

534109

1,3916,292

2,3761,2126,252

1,9385,996

26,112

1,970553

5,702

8,225

34,3376,756

41,093

1974

708142

1,6436,854

2,7011,4197,303

2,2086,963

29,951

1,534554

6,604

8,692

38,6437,106

45,749

1979

783180

2,2848,716

3,2982,0309,058

2,9638,820

38,140

2,027494

7,369

9,890

48,0306,539

54,569

1984 1989

Number of jobs

93568

1,7817,915

2,9821,8698,746

2,7849,601

36,904

2,005569

7,075

9,649

46,5536,783

53,336

1,205146

1,6008,191

3,1872,0879,471

2,53310,818

39,250

2,258561

8,025

10,844

50,0946,955

57,049

b Northeast Oregon = Baker, Grant, Umatilla, Union, and Wallowa Counties.c ANR = Agriculture and natural resources.

1990

1,231181

1,8388,681

3,2912,1659,446

2,49111,437

40,764

2,334554

8,325

11,213

51,9777,169

59,146

1991

1,336159

1,8788,642

3,2682,0689,528

2,48612,005

41,374

2,240552

8,480

11,272

52,6466,998

59,644

1992

1,308153

1,8108,506

3,2852,2129,637

2,49412,335

41,739

2,402542

8,592

11,536

53,2756,493

59,768

Source: Bureau of Economic Analysis, Regional Economic Information System, U.S. Department of Commerce, 1969-92.

by 1,530 jobs. Over the same period, thenumber of proprietors grew by 1,189, almostexclusively in nonfarm businesses (table 7). Italso seems likely that the growth of relativelyhigh-paying wage labor jobs will be limited,especially in sectors producing highly competi-tive tradable goods.

The reduction in communication cost broughtabout by computers and satellites favors rela-tively remote regions like northeast Oregon.Some businesses, professionals, and proprietorsthat were formerly unable to locate in theregion may now find it profitable to do so.Examples include firms providing producerservices, a component of the services sector intable 6. The service sector is usually thought ofas a nonbasic sector with sales mainly to thelocal population. Although this is true of manyservice businesses, it is also true that selectedservice firms do export most of their servicesoutside the region. Examples are tourist-ori-ented and producer service businesses. Thecommunication revolution has opened thelocation decision for export-oriented services.Many of these businesses are choosing to locatein high-amenity regions. It is likely that suchfirms will become an even more important partof the northeast Oregon economy in the future.

High-amenity value also is attractive to thoseseeking recreation and retirement destinations.The abundant natural amenities of the regionare becoming increasingly well known andappreciated. As more people take advantage ofthese attractions, we expect to continue to see

an increase in the proportion of transfer andproperty income in the region.4

The Issue of Regional Economic Base

Although we have previously referred to theterm "economic base," we will here broaden ourdiscussion of this key concept. Economic base(or export base) theory conceptually divides aregional economy into "basic" and "nonbasic"activities, postulating a causal relation betweenthe two. A basic activity is one that creates aninflow of money in return for goods and ser-vices sold to buyers outside the region (that is,"exported"). Regional economic activities thatare traditionally counted as basic include theextractive industries (agriculture, mining, for-estry, and fishing) and all primary and second-ary manufacturing.

Export dollars received by basic activities areused to purchase productive inputs of goodsand services, including labor. The degree towhich these goods and services are supplied byregional producers and households determinesthe amount of nonbasic activity in the region.It is assumed that nonbasic activity is induced,and therefore, could not exist without basicactivity. Traditional nonbasic sectors includemost services, retail trade, and state and localgovernment functions.

Multipliers are calculated as the ratio of totalactivity (that is, basic plus nonbasic) to basic

4 The future growth of transfer income also will dependon the outcome of the current political debate in Congress.

Table 7—Characteristics of regional employment over

Employment category

Total proprietors:FarmNonfarm

Total wage and salaryemployment

Total employment

1969

10,4064,4695,937

30,687

41,093

1974

11,1294,3046,825

34,620

45,749

1979

12,0233,8518,172

42,546

54,569

time for

1984

northeast Oregona b

1989

Number of jobs

13,5354,3539,182

39,801

53,336

13,9784,3779,601

43,071

57,049

1990

14,1864,2939,893

44,960

59,146

1991

14,8154,336

10,479

44,829

59,644

1992

15,1674,393

10,774

44,601

59,768

a Full- and part-time employment.b Northeast Oregon = Baker, Grant, Umatilla, Union, and Wallowa Counties.

Source: Bureau of Economic Analysis, Regional Economic Information System, U.S. Department of Commerce, 1969-92.

7

activity. In practice, estimates of industryemployment and payroll generally are used asindicators of regional economic activity. Thus,an economic base employment multiplier canbe interpreted as calculating the total numberof jobs generated by an increase in export salessufficient to require the addition of one job in aregional basic industry.

Although useful in understanding how regionaleconomics work, the concept of economic basehas been subject to some criticism and muchdiscussion over the years. Critics point out thatthe static nature of the hypothesized relationbetween basic and nonbasic sectors fails to ac-commodate structural changes that characterizea developing economy. Also, the linear relationbetween basic and total activity is assumedinvariant relative to the magnitude and direc-tion of the change. Thus an employmentmultiplier of 2.0 would be used to predict anincrease of 20 total jobs in response to theaddition of 10 basic jobs, and a loss of 2,000total jobs in response to the elimination of1,000 basic industry jobs.

A more fundamental criticism concerns thedefinition of basic activity. At the national level,balance of trade statistics are kept. These in-clude accounts of international flows of goods,services, and finances. There is no authoritativeaccounting of transboundary (that is, exportsand imports) flows, however, from subnationalregions. Consequently, it is necessary to esti-mate the contribution of the activities of variousindustries to the economic base of a region.Generally, this is done either by the simpleassignment method, in which all manufacturingand resource-based activity is defined as basic,or the location quotient method in which abasic industry provides a greater proportion orregional employment than is true of that indus-trv at the national level. Both of these methodscan create serious errors in the estimates of theexport base, especially if the structure of re-gional production and consumption is differentthan the national average. Also, neither methodaccounts for the growing prevalence of self-employed activity, the emergence of nonmanu-facturing exports (for example, producerservices), or the increasing importance oftransfer and property-type income.

Lately, the question of what activities constituteOregon's economic base has received muchattention. Until recently, the answer seemedobvious: lumber and wood products industriesclaimed a major share of employment, income,and export sales of most counties. Links withforestry, logging, transportation, and secondarywood products industries, as well as substantialindustry payrolls, ensured that growth in woodproducts had significant multiplier impacts onregional economies. The almost pervasivenature of employment in wood products is alsokey. Mills have traditionally located near majortimber sources, often providing the mainemployment in small towns and rural areas.

In a recent paper, Beuter (1995) reported thatin 1990, wood products industries providedmore than one-third of Oregon's economicbase, with forestry, fisheries, agriculture, andfood processing contributing another one-third.According to Beuter, in eastern Oregon thecontribution of wood products was less, aboutone-quarter, with forestry, agriculture, and foodprocessing contributing more than 50 percentof the economic base of the region. High-technology manufacturing, much touted as thefoundation of Oregon's economic future,comprised about 11 percent of the economicbase of the Portland metropolitan region butless than 1 percent of the economic base state-wide. Beuter's analysis stresses the importanceto the State economy of maintaining a viablewood products industry. He also doubtswhether high-technology manufacturing willever become significant outside the Interstate 5corridor because of its locational requirements,which differ from those of the timber industry.

Beuter used 1990 location quotients to estimateexport base indices and defined basic activitiesin a limited way, which tends to exaggerate theimportance of resource-based industries. Hisanalysis thus overestimates the current impor-tance of wood products relative to other com-ponents of the regional economy. As evidence,consider that since 1990, Oregon wood prod-ucts employment lost more than 10,000 jobs,whereas total employment in the State increasedby more than 50,000 jobs. Also between 1990and 1994, Oregon's exports to foreign countriesincreased 40 percent. This was largely due todramatic increases in high technology and agri-

8

culture sales, whereas foreign exports of lumberand wood products remained about constant.

Although these aggregate numbers ignore theimportant geographical aspects of the employ-ment shift, they do indicate a trend that is notpredicted by a simple economic-base approach.A major problem with standard economic baseanalysis is that it excludes "nontraditional"components of regional economic base. Forexample, in 1992, combined transfer andproperty-type income contributed 36 percentof total personal income in northeast Oregon.Most of this is derived from sources outside theregion. Federal Government spending is an-other important source of outside income oftenoverlooked by standard economic-base analysis.Dollars from these outside sources directlyfinance a large portion of industry purchases,household consumption, and local governmentactivity, and thus, should properly be includedin the regional economic base.

Economic-base models provide a useful frame-work for conceptualizing economic links in aregional economy. Apparent shortcomings,however, can severely restrict our ability to for-mulate relevant prescriptions for public policyand economic development. One alternative isto use IO or SAM-based models. These fixed-price models provide internally consistentrepresentations of regional economic structurebut under restrictive assumptions (for example,fixed-proportion production functions, uncon-strained factor and commodity supplies, andfixed or price-inelastic demand for goods andservices). Ideally suited to estimating the impactof changes in final demand, fixed-price modelsare severely limited in their applicability tosupply-side issues. A more efficient approach,using regional CGE models is described below.

Natural Resource Constraintsand Regional Economic Analysis

For examples of innovative uses of fixed-pricemodels to analyze the impacts of natural re-source constraints, the reader is referred toPetkovich and Ching (1978) (mining), andWaters and others (1994) (timber). Examples ofCGE applications to resource policy issues canbe found in Despotakis and Fisher (1988)(petroleum) and Berck and others (1991) (water).

In the Pacific Northwest, several recent studieshave used fixed-price (constant multiplier)methods to estimate impacts resulting fromimplementation of hypothetical salmon recov-ery plans. To our knowledge, there has been nocomparable flexible-price analysis of this issue.

Aillery and others (1994) examine direct andtotal impacts resulting from flow modificationson the Snake River to facilitate downstreammigration of salmon. They estimate that re-duced irrigation capacity could reduce agricul-tural employment in the Upper Snake Riverbasin by 2,500 jobs and total employment by4,100 jobs.

Haynes and others (1992) compare the costsassociated with implementing salmon recoverymeasures on National Forest land in the Colum-bia and Snake River basins with impacts ofeliminating timber, range, and recreationprograms in sensitive anadramous habitat areas.The authors conclude that even under theirworst case scenario, any resulting job losseswould seem minor spread over tne entire region,but would be significant in certain localities. Jobloss resulting from reduced timber availability inthe Umatilla and Wallowa-Whitman NationalForests is estimated at between 94 and 925 jobs,depending on availability of private log supplies.

In a later study, by using similar methods, Bolonand others (1994) compare incremental costs ofadopting the proposed interim PACFISPPstrategy on affected anadramous fish habitatwith estimates of foregone sales if timber, range,and recreation programs on Western publiclands were eliminated. Their baseline scenarioincludes any de facto restrictions on logging,grazing, and recreation activities resulting fromrecent legal and administrative actions. Conse-quently, the additional costs of implementingPACFISH seem minor compared with the valueof timber and recreation programs on Federallands. It seems, however, that the costs incurredin implementing the guidelines on public rangemay exceed the value of the Federal grazingprogram, particularly on BLM land. No esti-

5 Comprehensive strategy for improved Pacific salmonand steclhead habitat management developed by theUSDA Forest Service in 1992-93. PACFISH specifiesriparian management objectives to assure maintenanceor restoration offish habitat.

9

mate of regional employment or income im-pacts was provided.

A detailed study by Robison and Freitag (1994)presents estimates of job and earnings impactsresulting from timber supply reductions in thethree Oregon counties associated with theWallowa- Whitman National Forest. The au-thors used nine community-level, economic-base models to derive employment and earningsresponse assuming a 56-percent (-167 MMBF)reduction in available log supplies and a par-ticular geographical distribution of mill cut-backs and closures. In aggregate, they estimatethat the three-county region would lose 1,943jobs accounting for $56 million in earnings.The authors estimate it would take a fourteen-fold increase in recreation activity to replacethese lost jobs, or a twenty-eightfold increase toreplace lost earnings,6 Finally, Robison andFreitag emphasize the uneven distribution oflocal economic impacts. They estimate thatUnion County would lose 15 percent of earn-ings and 13 percent of jobs, with the commu-nity of North Powder losing 70 percent of itsemployment. Wallowa County would lose 11percent of earnings and 13 percent of jobs, withthe communities of Wallowa and Joseph eachlosing more than 40 percent ot their job base.

Obermiller examined impacts of the rangerestrictions by using an updated, survey-based,IO-type model of the Wallowa Countyeconomy (Obermiller and West 1983). Heassumes that the direct effect of closing theWallowa-Whitman National Forest to livestockgrazing would reduce the cattle herd by 38percent. Total impacts resulting from theclosure include a $2.5 million (42 percent of1981 baseline) reduction in cattle sales, $6.3million (2.8 percent of 1981 baseline) reductionin total regional sales, and $1.25 million (2

4 T h e bas is f o r this c a l c u l a t i o n , h o w e v e r , s e e m s c u r i o u s

given their description ot" the regional export base inwhich tourism was estimated to generate 4 percent ofe a r n i n g s i n t h e t h r e e - c o u n r y r e g i o n .

O b r m i l l e r , E 1994. The local monetary costs ofclosure of the Wallowa-Whitman National Forest inWallowa County, Oregon, to domestic livestock grazing;report to the Wallowa Coun ty Cour t . On file with: FredObermiller, Oregon State University, Depar tment ofAgricultural and Resource Economics, Corvallis, OR9 7 3 3 1 - 3 6 0 1 .

percent of 1981 baseline) reduction in localhousehold income.

Niemi and others8 estimated economic impactsof legal restrictions on public grazing. Theyjuxtapose the relatively small localized impactsof the grazing restrictions against the backdropof a diverse, 13-county economy spread acrossparrs of three states (Oregon, Washington, andIdaho). The authors offer severe criticism ofObermiller's results but do not provide quanti-tative counter estimates. They predict that "forthe region as a whole, the positive impacts ofthe injunction should outweigh the negativeones, especially in the long run." Niemi andothers (1994) emphasize the inherent dyna-mism of this regional economy, stating thateconomic growth is being driven by populationgrowth, which is strongly linked to quality oflife and natural environmental amenities.Traditional resource-based industries, whichtend to consume or alter these natural ameni-ties, are shrinking and have become relativelyunimportant as a source of regional growth.These themes are supported by both Power9

(1995) and Rasker (1995).

Also, the cost estimates associated with salmonrecovery efforts reported above should be placedin proper perspective against the potentialgains. Radtke and Davis10 (1995) estimated theannual state-level economic contribution of a fullyrestored salmon run (10-16 million fish) at $136million to $273 million in exvessel sales, providing13,000 to 25,000 total jobs and $254 millionto $507 million personal income state-wide.

8 Niemi, E.; MacMullan, E.; Whitelaw, E. 1994.Economic consequences of an injunction to protectsalmon habitat on the Wallowa-Whitman and UmatillaNational Foresrs; report prepared for Pacific RiversCouncil. On file with: ECO Northwest, 99 W. TenthSt., Suite 400, Eugene, OR 97401. (Note: The PacificRivers Council is the lead plaintiff in the lawsuit thatbrought the Ninth Circuit US Court injunctionrestricting grazing in the Umatilla and Wallowa-Whitman National Forests).9 Power, T. 1995. Economic well-being and environmen-tal protection in the Pacific Northwest. Unpublishedreport. On file with: University of Montana, EconomicsDepartment, Missoula, MT 59812.10 Radtke, H.; Davis, S. 1995. An estimate of the assetvalue for historic Columbia River salmon runs; reportprepared for the Institute for Fisheries Resources. On filewith: Hans Radtke, P.O. Box 244, Yachats, OR 97498.

10

The Northeast Oregon CGEModelA regional CGE model consists of a system ofequations representing the equilibrium behavioror factor and commodity markets and otherrelevant economic institutions. The system cansimulate economic response to changes invarious policy, management, and behavioralvariables. A kev feature is the inclusion ofrelative prices that reflect the economic scarcityof all commodities and productive factors in themodel. Endogenous prices adjust until factorand commodity market equilibrium conditionsare satisfied. Compared with fixed-price mod-els, the CGE method is more consistent withmodern economic theory, allows greater flex-ibility in the specification of economic behav-ioral relations, and generally produces moremoderate estimates of economic impact.

Implementing a regional CGE model does notrequire a priori designation of an economicbase. Basic activity is not necessarily limited toa few traditional manufacturing sectors. In aCGE model, economic change is governed bysupply-side (for example, available quantities ofproductive goods and services) and trade-related constraints, rather than by backwardlinks transmitted via changes in final demand.Also, because the opportunity to substituteamong resource inputs is subject to diminishingreturns in a CGE model, measures such asemployment multipliers are variable dependingon the level of resource inputs.

Figure 1 traces the basic steps of a regionalCGE modeling effort. First, data are collected,organized, and reconciled into a benchmarkequilibrium data sci (the social accountingmatrix).11 Next, behavioral and accountingrelations are specified, and the model param-eters are calibrated given the benchmark data.Finally, the policy-change scenarios are intro-duced, and counterfactual equilibria represent-

11 The data set for the northeast Oregon CGE wasassembled mainly from IMPLAN-generated regionalproduct accounts, state of Oregon tax and expendituredata, and RE1S county income and employmentestimates. Timber harvest estimates (Oregon Departmentof Forestry) and livestock grazing statistics (Bedell 1984,Bedell and Stringham 1984, Hewlett and others 1987)also were used to estimate benchmark resource flow levels.

ing the situation under the new policy regimes iscalculated. Impacts are estimated by comparingthe counterfactual equilibria against the bench-mark scenario.

Structure of the Northeast OregonCGE ModelAllocation of all resources and commodities mthe northeast Oregon CGE model is a functionof economic scarcity as reflected by the relativeprices of all goods, services, and productivefactors. Thus, price variables assume a preemi-nent role in the model. Key determinants ofrelative prices include (1) factor supply andproduction constraints; (2) ability of regionalconsumers to substitute among alternativesources of commodity supply (that is, regionalindustrial supply, regional nonindustrial supply,and imported supply); (3) ability of regionalproducers to supply alternative markets (that is,regional versus outside the region, or "export");and (4) demand conditions affecting regionaland export markets.

Figure 2 traces the links among components ofthe regional CGE model.12 First, value is addedto inputs of labor, services of proprietors, andcapital via linearly homogeneous Cobb-Douglasproduction functions, and combined with inter-mediate inputs to produce output for each sector(X). Behavioral assumptions ensure that produc-ers maximize economic returns by equating themarginal factor cost with the value of thecontribution of each factor to marginal product.

Each unit of X is either sold to local buyers(XXD) or exported outside the region (E). Aconstant elasticity transformation function(CET) governs the ease with which regionalproducers can switch between regional andexport market destinations. Revenue maximiz-ing behavior by producers determines theproportion of output supplied to satisfy regionaldemand versus export markets. Export demandis assumed to be perfectly elastic (that is, worldcommodity prices are fixed), whereas regionaldemand is influenced by endogenous price andincome effects.

12 A list of variables, parameters and equations, anddetailed descriptions of the CGE and fixed-price modelsare available on request.

11

Figure 1—Flowchart of CGE modeling procedures.

Commodities produced for regional use (XXD)are combined with available nonindustrialsupplies (GS, "government sales") to form totalaggregate supply from regional sources (XD).This aggregation occurs via a constant elasticitysubstitution (CES) function. The treatment ofgovernment supplies as imperfect substitutes forprivate supplies in generating total regionalcommodity supply illustrates the flexibility ofthe regional CGE framework. In constructingthe analysis for this paper, we imposed quantity

restrictions on government supplies of logs andgrazing.13 Regional supply (XD) is, in turn,combined with competitive imports (M) via aCES aggregation to form a composite absorp-tion good (or service) for each class of commod-ity (Q). The role of the nested CES functions isto allow partial substitution of public for private

13 By using this framework, we also could have examinedthe potential impact of public grazing or logging fee(price) changes as an alternative policy.

12

Figure 2—Schematic of the northeast Oregon CGE model.

ment sales are assumed to be offset by transfers00from other sources (that is, State and FederalGovernment).15 Consumption by each of threehousehold income classes is driven by changesin endogenous factor incomes and relativecommodity prices. Finally, business investmentspending can be treated as either exogenous orendogenous, depending on the model specifica-tion and implied length of run.

Resource Shock ScenariosThe baseline for this study is the aggregateestimated economic structure of the five-county, northeast Oregon region in 1992. Alldirect-impact scenarios are measured as reduc-

15 Although this assumption may tend to underestimatethe extent of regional economic adjustment, it is notunreasonable given the currently changing relationbetween state and local Fiscal responsibilities.

13

(and imported) sources of logs and forage.14

Expenditure minimization at both stages ofaggregation determines substitution betweenXXD and GS, and between XD and M, respec-tively. The use of CET and CES functions inthe model accommodates the observed phe-nomenon of "crosshauling" in which simulta-neous imports and exports appear in highlyaggregated commodity classifications.

Total supply of Q supports intermediate de-mand (ND) and final demand for consumergoods (C), investment needs (IT), and govern-ment purchases (G). In this model, all spendingby local agencies of Federal, State, and localgovernment is fixed at baseline levels. Anyreduction in revenues from taxation or govern-

14 This treatment contrasts with a fixed-price analyticalapproach in which the relative proportions of inputs frompublic, private, and imported sources is assumed to be fixed.

tions in resource supply relative to the 1992baseline. It can be argued that in anticipation ofde facto or perceived changes in resourcemanagement policy, considerable adjustmentalready has occurred. Thus, using 1992 as thebase year will tend to overestimate the actualimpact of a future policy shift. It is also useful,however, to view the effects of a major shift inpublic policy and attitude against the backdropof a real level of resource use in the recent past.Whether this level of resource use was sustain-able in the longer term or would have beenreduced for other reasons (for example, foresthealth, fire) is an important question but onethat is not directly addressed here. Positiveimpacts associated with the expected improve-ment in salmon spawning and migration success(for example, commercial and recreationalfishing) are not directly considered here.16

The main scenarios examined are as follows:

1. Moderate resource shocks:A. A 25-percent reduction in Federal

grazing animal unit months (AUMs)B. A 50-percent reduction in logging on

Federal lands

2. Severe resource shocks:A. A 99-percent reduction in Federal

grazing AUMsB. An 80-percent reduction in logging on

Federal lands

Each scenario is examined by using fixed-priceIO and flexible-price CGE methods. The CGEmodel results also are examined under differentassumptions regarding the length of time overwhich economic adjustment occurs. This isdone by assuming differing mobility of primaryfactors (labor, proprietors, and capital) asfollows:

• Short run (1-2 years):• Only labor is able to adjust. Workers who

have been laid off remain in the region andreceive an average unemployment benefit.

• Intermediate run (2-5 years):• Unemployed labor leaves the region. Propri-

etors can move into other businesses but donot leave the region.

• Long run (5-10 years):• Unemployed labor and unprofitable propri-

etors are able to leave the region.

In all three CGE formulations, sectoral (that is,"corporate") capital is assumed to remain fixedat baseline levels. In the CGE models, the re-source shocks are implemented as proportionalreductions in the baseline availability of GScommodities. Thus, the direct impacts ofresource supply shocks are treated as supply-side phenomena. This treatment has logicalappeal and also is consistent with neoclassicaleconomic theory.

A key feature of the CGE models is the abilityto substititute among public, private, andimported sources of logs and livestock foragevia CES aggregation functions as described inthe previous section. The ease of substitution isdetermined by exogenously specified elasticitiesand endogenous variation in the relative cost ofinputs obtainable from the three alternativesources of supply. An implication of the con-stant elasticity specification used here is that,because of the law of diminishing returns,relatively small shifts in the availability ofalternative supplies are much more easilyaccommodated than are major shifts. In prac-tice, we have assumed that the substitution ofprivate or imported logs and forage to replacereduced public supplies is difficult because ofhigh transportation costs, seasonal grazingconstraints, and the length of time necessary toalter timber growth and harvest rotations.17

In contrast to the CGE models, fixed-pricemodels regard all factor supplies as interregion-ally and intersectorally elastic by holding allcommodity and factor prices fixed. Constantmultipliers result from the assumption of fixedproportion, column-normalized expenditurecoefficients.18 The magnitude of these multipli-ers reflects the strength of backward links in theregional economy.

16 Enhancement of environmental amenities understricter land use policies, however, may be responsiblefor some of the positive impacts illustrated in the nextsection (economic counter scenarios).

17 This assumption also would result in larger estimatesof economic impact in response to a reduction in resourcesupplies than would a less restrictive assumption.18 In the fixed-price model used for this study, factorincomes and household expenditures also are assumedendogenous.

14

The fixed-price specification embodies tradi-tional behavioral assumptions and proceduresregarding regional economic systems. Conse-quently, the resource supply shocks are modeledas "equivalent" reductions in exogenous de-mand for the output of the directly impactedindustrial sectors. Although this approach islogically less appealing and less consistent withmodern notions of microeconomic behavior,arguably reasonable results can be generatedgiven careful construction of the direct-shockscenario. There are also several supply-sidefixed-priced modeling techniques described inthe literature that allow a more consistenttreatment of the direct-shock scenario butusually at the cost of adopting even more

questionable assumptions regarding the behav-ior of regional economic systems.

ResultsResults obtained under each scenario by usingCGE and IO methods are presented below. Foreach shock scenario, CGE model results underthe three length-of-run assumptions (short run,intermediate run, and long run) are presentedfirst, followed by results obtained by using theIO model. Aggregate employment results alsoare displayed in table 8. Income and revenueresults are summarized in table 9.19

19 More detailed sectoral results for other variables in themodel are available on request.

15

Table 8—Results: comparative aggregate employment impacts

Shocks

Range shocks:Moderate (-25 percent)—

Range and livestockOther

TotalEmployment multipilier

Severe (-100%)—Range and livestockOther

Total employment changeEmployment multiplier

Logging shocks:Moderate (-50 percent)—

Logging and wood productsOther

TotalEmployment multiplier

Severe (-80 percent)—Logging and wood productsOther

Total employment changeEmployment multplier

Short run

-9-3

-121.33

-516-78

-5941.15

-928-300

-1,2281.32

-1,786-580

-2,3661.32

Model configuration

Medium run Long run

Change in number of jobs

-3237

5-.16

-1,3131,366

53-.04

-1,141-359

-1,5001.31

-2,118-681

-2,7991.32

-29-12

-411.41

-1,036-335

-1,3711.32

-1,138-628

-1,7661.55

-2,115-1,214

-3,3291.57

Fixed price

-236-72

-3081.31

-947-290

-1,2371.31

-1,246-914

-2,1601.73

-1,993-1,460

-3,4531.73

16

Table 9—Results: comparative aggregate income and revenue impacts

Shocks

Range shocks:Moderate (-25 percent)—

Household incomeLow income householdsMedium income househouldsHigh income households

Total change in household income

State local government revenues:Range feesOther revenues

Total change in government revenue

Severe (-100 percent)—Household incomeLow income householdsMedium income householdsHigh income households

Total change in household income

State and local government revenues:Range feesOther revenues

Total change in government revenue

Logging shocks:Moderate (-50 percent)—

Household incomeLow income householdsMedium income householdsHigh income households

Total change in household income

State and local government revenues:Timber feesOther revenues

Total change in government revenue

Severe (-80 percent)—Household incomeLow income householdsMedium income householdsHigh income households

Total change in household income

State and local government revenues:Timber feesOther revenues

Total change in government revenue

Short run

-0.006-.130-.143

-.279

-.093-.035

-.128

-.217-4.891-5.450

-10.558

-.371-1.100

-1.471

-.932-8.025-5.032

-13.989

-9.105-2.051

-11.156

-1.837-15.556

-9.796

-27.189

-14.568-4.059

-18.627

Model configuration

Medium run Long run

+0.019-.052-.186

-.220

-.093-.009

-.102

+.688-1.083-5.897

-6.292

-.371-.478

-.849

-2.618-21.300-11.084

-35.002

-9.105-3.080

-12.185

-4.901-39.614-20.663

-65.178

-14.568-6.839

-21.407

Million dollars

-0.030-.489-.470

-.989

-.093-.094

-.187

-.914-14.541-13.993

-29.448

-.371-3.23

-3.601

-2.814-24.060-13.579

-40.452

-9.105-4.222

-13.327

-5.289-45.080-25.603

-75.971

-14.568-8.031

-22.599

Fixed price

-0.240-3.390-3.220

-6.850

-.960-13.610-12.940

-27.510

-3.920-30.390-17.920

-52.230

-6.260-48.590-28.650

-83.500

Moderate Resource Shocks

Shortrun range shock—In the short run, withonly labor adjusting, the 25-percent reductionin public range supply has only a slight impacton the regional economy. Regional privateoutput of hay and pasture decreases by 1.74percent, but imports rise by 13.9 percent.Consequently, total supplies of range, hay, andpasture remain about constant. Total employ-ment decreases by only 12 jobs, including a lossof 9 livestock-related jobs. Changes in house-hold incomes are slight with low, medium, andhigh income households experiencing declinesof less than one-tenth of 1 percent (0.1 per-cent). Total state and local government revenuesdecline by $128,000.

Intermediate-run range shock-—In the inter-mediate run, proprietors are free to pursueother business opportunities in the region. The25-percent reduction in public range induces areallocation of economic activity. Imports ofhay and pasture slow somewhat compared withthe shortrun scenario to a rate 8 percent abovebaseline levels. Overall, there is a 0.82-percentdecrease in total supply of range, hay, and pas-ture. A loss of 32 livestock-related jobs is offsetby small gains in some other sectors, and sototal employment remains virtually unchanged.Total household income declines by $220,000.

Longrun range shock—In the long run, unem-ployed labor and unprofitable proprietors leavethe region. Production of private hay and pas-ture recovers and increases 5.7 percent abovebaseline, adding 17 new jobs. Hay importssettle at a rate less than 1 percent above baselinelevels. Total availability or range, hay, and pas-ture declines by only 0.87 percent. Employ-ment declines for almost all other sectors for anet reduction of 41 jobs. Total householdincome falls by just less than $1 million (0.05percent). Total state and local governmentrevenues are down $187,000.

Fixed-price range shock—Under fixed-priceanalysis, output and employment changes aredirectly proportional. Hence, no sector exhibitsemployment increases. Livestock-related indus-tries lose 236 jobs, and another 72 jobs are lostin the rest of the economy. Total householdincome falls by $6.85 million, with low, me-dium, and high income households each losing

less than 1 percent of baseline income. Esti-mated fixed-price employment impacts areroughly seven to eight times greater thanestimates using the regional CGE model.

Shortrun log shock—The shortrun impact of a50-percent reduction in logging of publictimber is more significant. Total employmentdecreases by 1,228 jobs. Output of private logsremains about constant but is allocated awayfrom export markets to supply regional mills.Log imports increase 34.7 percent, but totalsupply of logs in the region falls by 13.9 per-cent. Logging and wood-processing shedsalmost one-fourth of baseline employment (928jobs) largely because of the increased cost ofpurchasing higher priced logs. Incomes of low,medium, and high income households declineby 0.2 percent, 0.91 percent, and 1.02 percent,respectively. Revenues of local governmentsdecline by at least $11 million, mostly becauseof reduced Federal timber sales.20

Intermediate-run log shock—The reallocationof activity induced by the 50-percent reductionin public logging fails to insulate the economyfrom negative impacts. Compared with theshortrun scenario, log imports slow consider-ably. In the intermediate run, logs are importedat a rate of 17.7 percent above the baseline level,but total log supply falls 17.9 percent belowbaseline levels. Despite an increase in livestock,agriculture, and food-processing employment of218 jobs, large decreases in logging and woodprocessing (-1,141 jobs) induce a net loss of1,500 total jobs. Household income declines by$35 million, and incomes for low, medium, andhigh income households decrease by 0.58percent, 2.42 percent, and 2.25 percent, respec-tively. State and local government revenues aredown more than $12 million.

Longrun log shock—Regional log suppliesremain about the same as in the intermediaterun (that is, 17.98 percent below baseline).Employment shows the same trend as in theintermediate run, although more negative (lesspositive). Slight increases in livestock, agricul-ture, and food-processing employment are

20 Note that reductions in local government revenuesresulting from reduced Federal timber sales may actuallybe mitigated through revenue guarantees under TheNorthwest Forest Plan.

17

canceled by large reductions in logging andwood products (-1,138 jobs), services, andtrade. Total employment decreases by 1,766jobs. Total household income declines by$40.45 million, with reductions of 0.62 per-cent, 2.73 percent, and 2.76 percent in incomesof low, medium, and high income households,respectively.

Fixed-price log shock—Logging and woodproducts shed 1,246 of the total 2,160 jobs lostin the economy. A total of $52.23 millionhousehold income is lost. Low, medium, andhigh income households lose 0.86 percent, 3.45percent, and 3.64 percent of baseline income,respectively. Here the fixed-price employmentimpacts range from 75 percent greater in theshort run to 22 percent greater in the long runthan estimates from the regional CGE model.

Severe Resource Shocks

Shortrun range shock—Even in the short run,the impact of the virtual elimination of publicrange in the region is unambiguously negative.Private hay and pasture decreases by 18.3percent, but total supply of range, hay, andpasture in the region decreases by only 2.65percent because of a sevenfold increase in hayimports. The livestock sector decreases outputby 2.36 percent, and sheds 516 jobs. Alto-gether, 594 regional jobs are lost, along with$10,558 million in household income. Thistranslates into a reduction in the incomes oflow, medium, and high income households of0.05 percent, 0.56 percent, and 1.11 percent,respectively. Local government revenues fromsales of Federal grazing permits decline by$371,000, and total state and local governmentrevenues are down $1,471 million.

Intermediate-run range shock—In the inter-mediate run, hay imports remain high, aboutsix times baseline levels. Large cuts in livestock-related employment (-1,313 jobs) are counteredby an increase of 1,366 jobs in other sectors(mainly trade, services, and other agriculture) sothat net regional employment increases by 53jobs. Total household income declines by $6.29million, decreasing by 1.12 percent for mediumincome households and 1.2 percent for highincome households, but increasing by 0.15percent for the low income group. Total state

and local government revenues decline by$849,000.

Longrun range shock—In the longer term,private production of hay and pasture returnsto about baseline levels. Hay imports settle at alevel 21 percent above baseline. Total supply ofrange, pasture, and hay is 19.8 percent belowthe original level. A large cut in livestock-related employment (-1,036 jobs) is accompa-nied by decreases in most other sectors, so thattotal regional employment decreases by 1,371jobs. Total household income declines by$29.45 million, falling by 0.2 percent for lowincome households, 1.65 percent for mediumincome households, and 2.84 percent for thehigh income group. Total state and local gov-ernment revenues are down by $3.6 million.

Fixed-price range shock—Livestock-relatedindustries lose 947 jobs, and another 290 jobsare lost in the rest of the economy. In house-hold income, $27.51 million is lost, with low,medium, and high income households losing0.21 percent, 1.55 percent, and 2.63 percent ofbaseline household income, respectively. Underthe severe range shock, the fixed-price modelnow underestimates economic impact relativeto the regional CGE model.

Shortrun log shock—The response to an 80-percent reduction in baseline Federal logsupplies is decidedly negative. A 68.5-percentincrease in log imports partially counters the80-percent decline in public log supply. But an8.85-percent reduction in private log outputcombines to reduce total regional log supply by28.36 percent from baseline levels. This con-tributes to a 48.38-percent reduction in woodproducts activity, and the loss of 1,786 loggingand wood-processing jobs. Total employmentfalls by 2,366 jobs, with the loss of $27.19million in household income. Incomes declineby 0.4 percent, 1.77 percent, and 1.99 percentfor low, medium, and high income households,respectively. State and local government rev-enues decline by $18,627 million, $14,568million of which is due to reduced Federaltimber sales.

Intermediate-run log shock—Despite a 30.3-percent increase in log imports, total logsupplies fall by 35.1 percent, including a 10-percent reduction in private logs. Logging and

18

wood products employment declines by one-half (-2,118 jobs). Total regional employmentdeclines by 2,799 jobs, despite an increase of411 jobs in livestock, other agriculture, andfood processing. Household income declines by$65-18 million, therby resulting in decreases forlow, medium, and high income households of1.08 percent, 4.5 percent, and 4.2 percent,respectively. State and local government rev-enues are down more than $21 million.

Longrun log shock—Total log supplies remainabout 35 percent below baseline levels as in theintermediate scenario. Logging and woodproducts employment remains weak (-2,115jobs). Employment in services, trade, and mostother sectors also weakens, so that total regionalemployment is 3,329 jobs below the baselinelevel. Household income declines by almost$76 million, thereby resulting in decreases forlow, medium, and high income households of1.17 percent, 5.12 percent, and 5.2 percent,respectively. There is also a loss of $22.6 millionin state and local government revenues.

Fixed-price log shock—Logging and woodproducts shed 1,993 of the total3,453 jobs lostin the economy. In household income, $83.5million is lost, with low, medium, and highincome households losing 1.38 percent, 5.52percent, and 5.82 percent of baseline householdincome, respectively.

Combined intermediate-run range and logshocks—It is interesting to note that if bothrange and log shocks occur simultaneously, theintermediate-term impact seems significantlyworse than the sum of the two individualshocks. Under the combined shock scenario,2,847 jobs are lost compared with 2,746 jobs ifthe separate intermediate-run impacts of the logand range scenarios are summed. For reference,the fixed-price model indicates a loss of 4,690jobs. There is also an additional $2.52 millionloss in household income under the combinedscenario. The difference can be attributed toincreased pressure on relative prices (particularlythe rate or return to proprietors), which exacer-bates output and employment impacts underthe combined shock scenario.

Economic Counter Scenarios

In addition to the range and logging supplyshocks, several expansionary "economic counterscenarios" also were examined. Results of thesescenarios are presented below and in table 10.21

The basic counter scenarios examined includethe following:

"High-technology" assumes the location orexpansion of a regional facility for manufactur-ing high-technology products, and the resultinghiring of 330 new workers at the facility.

"Inmigration" assumes 1,000 households settlein the region, bringing with them an average of$91,000 in transfer and property income (total$91 million). The inmigrants also build newhouses at the rate of 200 per year, spending$100,000 per home in construction costs (total$20 million).

"Tourism" assumes there is roughly a doublingin nonresident tourist and recreation activity inthe region (that is, enough to generate 2,200service-sector jobs, or about 4 percent of the1992 nonfarm employment base).

"Combined" refers to simultaneous impact ofall three counter scenarios (that is, high-technol-ogy, inmigration, and tourism).

All counter scenarios include the "new baseline,"which assumes the simultaneous, longrunimpact of the severe public logging (-80 per-cent) and severe public range-reduction (-100-percent) scenarios. Thus, relative to the original1992 situation, there are now 4,679 fewer jobsavailable in the new baseline, including 3,199jobs lost from resource-based industries and1,480 jobs lost from other sectors.22 Totalhousehold income is $105 million less than itwas in 1992.

The counter scenarios should be regarded asillustrative only, and not viewed as prescriptions

21 In cable 10, counter scenario impacts are expressed aschanges relative to the original, 1992 baseline levels.22 This is slightly less severe than the straight sum ofindividual impacts under the severe logging and severerange scenarios (-4,700). The difference (21 jobs) is dueto the need to slightly modify the standard CGE modelto simultaneously accommodate the combined resourceshocks and economic counter scenarios.

19

or predictions of future circumstances. Al-though the scenarios were selected assuming,arguably, reasonable growth prospects for thenortheast Oregon regional economy over thenext 5 to 10 years, they are not based on anyknowledge of pending industrial locations orextrapolated from any current demographic oreconomic trends.

Expansion of high-technology manufacturing inthe region adds 332 manufacturing jobs (236plus 96 jobs previously lost in the baselinescenario) and 109 jobs in services industries.Altogether there are 433 more jobs than in thenew baseline, and an additional $9 million inhousehold income.

Inmigration results in the addition of 243construction jobs and 860 service jobs. Totalemployment increases by 1,082 jobs but is still3,597 jobs lower than in 1992. Largely becauseof the influx of exogenous, unearned transferand property income, household income is$109 million greater than in the new baseline,and $4 million higher than it was in 1992.

Tourism expansion adds 2,179 service-sectorjobs (eating, drinking, and lodging; trade and

other services). Total employment increases by2,229 jobs compared with the baseline in table10. Household income increases by $26 millionbut remains $79 million lower than in 1992.

If all three expansion scenarios occur simulta-neously, 3,707 total jobs are added, including250 construction jobs, 330 manufacturing jobs,and 3,117 service-sector jobs. Total employ-ment, however, is still 972 jobs lower than itwas before the impact of the severe resourceshocks. Household income is $38 milliongreater than it was in 1992, with medium andnigh income households adding $30 millionand $11 million, respectively; low incomehouseholds are still about $3 million worse offthan they were in 1992.

ConclusionsThe results presented here emphasize thecomplexity of the answer to the question Howdependent is Northeast Oregon on resource-based industries? Data show that in 1992, 29percent of regional employment was naturalresource based. Consequently, by most meas-ures, a sustained, severe reduction in the avail-

20

Table 10—Results: employment and income impacts under possible scenarios, change relativeto original 1992 economy

Employment

Resource-based:b

ConstructionOther manufacturingc

Servicesd

Total jobs

Household income:Low income householdsMedium income householdsHigh income households

Total household income

Newbaseline3

-3,199-50-96

-1,334

-4,679

-6-59-40

-105

High-technology Inmigration Tourism

Change in number of jobs-3,208

-49236

-1,225

-4,246

-5-54-37

-96

-3,218193-98

-474

-3,597

Million dollars-510-1

4

-3,154-45-96

845

-2,450

-5-44-30

-79

Combined

-3,189200234

1,783

-972

-33011

38

a New baseline—includes the simultaneous longrun impact of the severe logging (-80 percent) and severe range reduction (-100percent) scenarios.b Resource based—includes livestock-related, logging and wood products, other agriculture, other natural resources, andagriculture processing.c Other manufacturing—includes all manufacturing industry other than agricultural processing and wood productsd Services—includes transportation, communications, and utilities; wholesale and retail trade; eating, drinking, and lodging;finance, insurance, and real estate; other services; and government industry and government enterprise.

ability of once abundant public range andtimber resources will have dramatic impacts onthe existing regional economic structure. In themost severe cases examined (1) elimination ofFederal range programs caused the loss of 1,371jobs (2.3 percent of regional employment) and$29 million (1.6 percent) of household income;and (2) an 80-percent reduction in Federal logsupplies resulted in the loss of 3,329 jobs (5.5percent of regional employment) and $76million (4.2 percent) of household income.

Analysis of economic counterscenarios suggeststhat plausible, simultaneous increases in tour-ism, high-technology manufacturing, and thepopulation of independently employed orretired persons can partially offset employmentand income losses resulting from resourcepolicy changes. The industry mix of the newjobs and the distribution of household incomewill both be different than they were originally.

Evidence suggests that the regional economyhas already begun a transformation, character-ized by stable or declining employment inresource-based industries and employmentgrowth in tourism and retirement-based services.These trends are likely to continue with orwithout major changes in natural resource policy.

Proprietors currently provide the major input inthe regional livestock industry. It was notedunder "Sources of Employment" chat nonfarmproprietors have become a fast-growing compo-nent of the northeast Oregon regional economy.If access to public range becomes restricted, wewould expect to see an acceleration of this trendas small- and medium-sized livestock operationstransfer soine or all of their resources into otherkinds of businesses.

In this analysis, commodity prices outside theregion for logs, lumber, hay, and livestock wereassumed to remain fixed. This assumptionseems reasonable given that changes in theregional output of these commodities wouldnot significantly affect world supplies. Worldcommodity prices, however, have fluctuatedduring the period under consideration. If theeffect of recently higher lumber prices had beenincluded, the response of regional and importedlog supplies would have been greater, therebyresulting in a smaller estimated decline inregional employment and income. On the

other hand, incorporating the effect of fallingprices, such as was recently observed for beefproducts, would have resulted in larger esti-mated reductions in the output and employ-ment of livestock and related industries.

Our treatment of all government activity asfixed allows comparison of preshock andpostshock scenarios in "revenue-neutral" terms(that is, assuming no change in governmentexpenditures in the region). Although thisassumption may tend to underestimate theamount of regional impact resulting from aneconomic shock, it is not unreasonable giventhe currently unstable relation between stateand local fiscal responsibilities in Oregon. Ifreductions in government revenues had beentranslated directly into reduced expenditurelevels, then the impact of the resource shockswould have appeared more severe.

Moderate shock results under the fixed-pricemodel are more pessimistic than under CGEassumptions. In the case of the severe range orlog shock, however, certain impacts appear lesssevere under the fixed-price treatment than inthe corresponding longrun CGE model sce-nario.23 This is because adjustment to severerestrictions on Federal land use is considerablymore difficult when modeled as a CGE supply-side constraint than when modeled as a fixed-price reduction in demand. In the flexible-pricemodel, increase in the regional cost of feedmakes the livestock sector less profitable. Forthe severe log shock, increase in regional logprices hampers adjustment in the wood-process-ing sector. Clearly the ehminarion of publicresource supply has differed, and severe conse-quences occur that are difficult to capture byusing conventional, fixed-price models. It isinteresting, however, that both the fixed-priceand flexible-price models tell much the samestory regarding total employment impactsunder severe resource shocks in the long run.The difference between models is in estimatedimpact on the directly affected sectors (that is,range and livestock, or logging and woodproducts).

23 For example, under the severe range shock scenario,the longrun CGE model estimates loss of an additional89 livestock-related jobs and 45 other jobs comparedwith the fixed-price model.

21

Literature Cited

Aillery, M.; Betels, P.; Cooper, J. [and others]. 1994. Salmon recovery in the Pacific Northwest: asummary of agricultural and other effects. AIB-699. Washington, DC: U.S. Department ofAgriculture, Economic Research Service. 10 p.

Alward, G. [and others]. 1989. Micro IMPLAN software manual. Fort Collins, CO: Colorado

State University. [Irregular pagination].

Bedell, T. 1984. Dependency on Federal grazing in eastern Oregon. Rangelands. 6(4): 152-155.

Bedell, X; Sringham, T. 1984. Forage sources for eastern Oregon cattle ranches with Federalgrazing permits. In: Research in rangeland management. Spec. Rep. 715. Corvallis, OR: OregonState University, Agricultural Experiment Station: 7-14. In cooperation with: U.S. Departmentof Agriculture, Agricultural Research Service.

Berck, P.; Robinson, S.; Goldman, G. 1991. The use of computable general equilibrium models toassess water policies. In: Dinar, A.; Zilberman, D., eds. The economics and management of waterand drainage in agriculture. Amsterdam, NY: Kluwer Publishing Co: 489-509.

Beuter, J. 1995. Legacy and promise: Oregon's forests and wood products industry report for theOregon Business Council and the Oregon Forest Resources Institute. Portland, OR: OregonBusiness Council. 56 p.

Bolon, N.; Hansen-Murray, C; Haynes, R. 1994. Estimated economic impacts on the timber,range, and recreation programs on NFS and BLM public lands from adopting the proposedinterim PACFISH strategy. Gen. Tech. Rep. PNW-GTR-344. Portland, OR: U.S. Departmentof Agriculture, Forest Service, Pacific Northwest Research Station. 80 p.

Despotakis, K.; Fisher, A. 1988. Energy in a regional economy: a computable general equilibriummodel for California. Journal of Environmental Economics and Management. 15: 313-330.

Haynes, R.; Bolon, N.; Hormaechea, D. 1992. The economic impact on the forest sector of criticalhabitat delineation for salmon in the Columbia and Snake River basins. Gen. Tech. Rep. PNW-GTR-307. Portland, OR: U.S. Department of Agriculture, Forest Service, Pacific NorthwestResearch Station. 33 p.

Hewlett, J.; Cross, T.; Hart, R. 1987. Enterprise budget: cow-calf, La Grande area. EM 8337.Corvallis, OR: Oregon State University Extension Service. 4 p.

Niemi, E.; MacMullan, E.; Whitelaw, E. 1994. Economic consequences of an injunction to protectsalmon habitat on the Wallowa-Whitman and Umatilla National Forests. Eugene, OR: ECONorthwest; preliminary report. 20 p.

Obermiller, E; West, P. 1983. The structure of a resource dependent economy: Wallowa County,Oregon. Corvallis, OR: Department of Agricultural and Resource Economics, Oregon StateUniversity. 105 p.

Petkovich, M.; Ching, C. 1978. Modifying a one region Leontief input-output model to showsector capacity constraints. Western Journal of Agricultural Economics. 3: 173-179.

Rasker, R. 1995. A new home on the range: economic realities in the Columbia River basin. Wash-ington, DC: The Wilderness Society. 59 p.

Robison, M.; Freitag, J. 1994. The economic impact to local communities of eliminating theWallowa-Whitman National Forest timber program. Moscow, ID: Robison and Associates. 57 p.

State of Oregon, Department of Forestry. 1993. Oregon timber harvest reports for 1988-1993.Bourhill, R., comp. Salem, OR: Oregon Department of Forestry.

22

U.S. Department of Commerce, Bureau of Economic Analysis. 1992. Employment and incomedata tor Oregon counties: 1969-1992. Washington, DC: U.S. Government Printing Office.Regional Economic Information System: on CD-ROM.

U.S. 9th Circuit Court. 1994. Pacific Rivers Council, Plaintiff, vs. Jack Ward Thomas, Chief,United States Forest Service, Defendants. Order dated July 7, 1994. No. 30F-3d 1050: 7425-7442.

Waters, E.; Holland, D.; Weber, B. 1994. Interregional effects of reduced timber harvests: theimpact of the northern spotted owl listing in rural and urban Oregon. Journal of Agriculturaland Resource Economics. 19(1): 141-160.

23

Waters, Edward C; Holland, David W.; Haynes, Richard W. 1997. The eco-nomic impact of public resource supply constraints in northeast Oregon. Gen.Tech. Rep. PNW-GTR-398. Portland, OR: U.S. Department of Agriculture,Forest Service, Pacific Northwest Research Station. 23 p. (Quigley, Thomas M..ed.; Inferior Columbia Basin Ecosystem Management: scientific assessment).

Traditional, fixed-price (input-output) economic models provide a useful frame-work for conceptualizing links in a regional economy. Apparent shortcomings inthese models, however, can severely restrict our ability to deduce valid prescrip-tions for public policy and economic development. A more efficient approachusing regional computable general equilibrium (CGE) models as well as a briefsurvey of relevant literature is presented. Evidence suggests that a regionaleconomic transformation, characterized by stable or declining employment inresource-based industries and growth in tourism and retirement-based services, isalready underway.

Keywords: Regional economics, computable general equilibrium models, input-output models, economic base theory, public resource policy.

The Forest Service of the U.S. Department ofAgriculture is dedicated to the principle of multipleuse management of the Nation's forest resourcesfor sustained yields of wood, water, forage, wildlife,and recreation. Through forestry research,cooperation with the States and private forestowners, and management of the National Forestsand National Grasslands, it strives-as directed byCongress—to provide increasingly greater serviceto a growing Nation.

The United States Department of Agriculture (USDA)prohibits discrimination in its programs on the basisof race, color, national origin, sex, religion, age, disability,political beliefs, and marital or familial status. (Not allprohibited bases apply to all programs.) Persons withdisabilities who require alternative means of communica-tion of program information (Braille, large print, audiotape,etc.) should contact the USDA Office of Communicationsat (202) 720-2791 (voice), or (800) 855-1234 (TDD).

To file a complaint, write the Secretary of Agriculture,U.S. Department of Agriculture, Washington, DC 20250,or call (800) 245-6340 (voice), or (800) 855-1234 (TDD).USDA is an equal employment opportunity employer.

Pacific Northwest Research Station333 S.W. First AvenueP.O. Box 3890Portland, Oregon 97208-3890