Industry Linkages and the Structure of the Local Economy
 By Daniel V. Rainey and Kevin T. McNamara
 
Introduction

State and local governments offer a variety of initiatives aimed at stimulating different segments of their economy (Smith et al.; Kriesel; Krmenec; Deaton and Gunter; and Herzog and Schlottmann). Of primary concern to local leaders is how to sustain or increase the local employment base. Even in localities where employment growth has been strong, strategies for maintaining a strong economic growth continue to be a concern (Bartik; Hasting and Brucker). Economic growth is the process of increasing the productive capacity of the economy.

The traditional economic growth or development strategy, industrial recruitment, first appeared in the United States in the 1930s (Smith and Fox). Recruitment remains the primary development strategy of most communities (Walker et al.; Steinnes; Smith et al.; Smith and Fox; Kriesel and McNamara; Kraybill). Industrial recruitment aims to attract new investment/firms or branch plants of existing firms to the recruiting region (McNamara).

The 1980s saw state and local government officials exploring development strategies to complement economic growth objectives. These alternative strategies focus on business retention and expansion (BR&E), entrepreneurship, technology transfer, and worker productivity (Ahlbrandt; Campbell and Allen; Clifton et al.; Feller; Morse; Shapira; and Kraybill). These programs provide assistance to existing and/or new firms to help them compete in regional, national, and global markets (Feller; Shapira).

Regional economic analysis has provided insights on the probability of communities attracting new establishments (Bartik; Schmenner et al.; Smith et al.; McNamara et al.; Herzog and Schlottmann) and on how a regionís economic structure influences its ability to grow (Plaut and Pluta; Carlino and Mills). Communities with better infrastructure, both physical and human (Bartik; Coughlin et al.; Woodward), market access (Friedman et al.; Bartik; Coughlin et al.), and amenities associated with a concentration of manufacturing and population (Woodward; Coughlin et al.) attract new investment and have higher economic growth rates as measured by the number of new manufacturing establishments and employment growth.

While new plant investment is a recognized source of manufacturing growth (Bartik, Kraybill), the share of employment growth associated with new firms versus existing firms is less known. This research analyzes the impacts of manufacturing employment growth from both new and existing establishments to assess if there are differences in the total employment growth impact associated with the employment growth from the two sources.

 

Conceptual Model

The economic base model postulates that the local economy can be divided into two sectors: basic and nonbasic (North). The basic sector consists of entities producing goods and/or services for sale outside of the local economy. The nonbasic sector consists of establishments and individuals producing goods and services primarily to be sold in the local economy. Revenue brought into the local economy by sales of basic sector firms stimulates expansion of the local economy, as nonbasis industry develops to support the needs of basic sector firms and workers (North, Shaffer).

The economic base model is a demand side model (Shaffer). Demand side models focus on issues that influence product demand. Demand side models assume availability of resources within the local economy to increase production capacity of basic sector firms. Expanding the demand of the basic sector=s products will return the economy toward a level of full employment.

The model is represented by the following equation:

DENB = a + bDEB + F

where DENB represents the change in nonbasic employment, DEB represents change in basic employment, a and b are parameters to be estimated, and F is a stochastic disturbance term.

The above equation implies the nonbasic sector employment change resulting from a change in basic employment is the same for all sectors. However, given the differences in production practices across industries the multipliers are likely to vary across industries. Thus, the basic sector is disaggregated to Standard Industrial Classification Divisions (Standard Industrial Classification Manual) to determine the influence of each of the industries on nonbasic employment (Rainey).

Research has shown that newly located manufacturing establishments, particularly foreign establishments, tend to purchase more of their inputs from suppliers outside the local economy (Barkley and McNamara). If new manufacturing firms purchase inputs from non local suppliers more than existing firms do, the total employment growth impact on the local economy associated with new firms will be lower than the total employment impacts associated with economic growth in existing manufacturing firms. To test whether there is a difference in the total employment growth associated with employment growth in new manufacturing establishments versus employment growth in existing manufacturing firms, the manufacturing sector is disaggregated into new and existing establishments.

Shahidsaless et al. concluded that economic base employment multipliers, the number that represents the total employment impact associated with an increase in basic sector employment, differ depending on the size of the market. Therefore, the economic base model for the Indiana economy is estimated for three county types: metro, adjacent, and nonadjacent. The county classifications are based on the Bureau of Censusí classification of Metropolitan Statistical Areas (MSAs)(U.S. Bureau of the Census). It is hypothesized that the larger markets in metro counties allow for more interactions of metro manufacturing establishments and more opportunities for households employed by these establishments to make purchases in the local economy than could occur in smaller markets. The potential for more interaction between establishments and households is hypothesized to increase the total employment multiplier for larger communities as compared to smaller communities because firms and individuals in larger economies have more opportunities to spend dollars in the local economy than firms and households in smaller communities.

In addition, the local resources available are likely to vary across counties within each county classification. To test the impact of the local economic structure (i.e. local resources) on industry linkages, the basic sector is interacted (entered in the estimated model as interaction terms)with attributes that describe the local economy. These attributes also are entered separately into the estimated model to determine how the structure of the local economy influences the total employment growth of the nonbasic sector (Rainey).

The economic structure of the local economy was represented by four measures: localization economies, urbanization economies, infrastructure, and labor quality. Localization economies refer to the cost savings that accrued to a firm from having a concentration of similar establishments in the local economy. It is hypothesized that a community with a concentration of similar economic activity is more likely have services and suppliers needed by establishments in that industry. Urbanization economies refer to the cost savings that accrue to an establishment from having access to a large and diverse local economic market. The diversity of products and services offered by larger markets provides more opportunities for purchasing goods in the local economy. Labor quality measures the productivity of the labor force. Communities with better labor quality typically are more productive and have greater growth. Transportation infrastructure allows for easier access to local and nonlocal markets. Greater access to markets allows establishments to lower their transaction costs.

The estimated model is:

DENB,i = a + Sb1jDEB,ij + SkSjmb2kjmEs,kDEB,ijm + Skb3kEs,k + Fi

where DENB,i is change in nonbasic employment for region i, DEB,ij is change in basic employment in region i and industry j, Es,kDEB,ijm is a matrix representing the interaction of the economic structure variables and the manufacturing sectors, and Es is the matrix measuring the structure of the local economy. The total employment multiplier for industry j will be equal to (1 + bj + SkSjmb2kjmEs,k)

 

Data

The employment data were obtained from the Indiana Department of Commerce (IDOC), the U.S. Bureau of the Census, the Economic Development Information Network at Indiana University, and the Indiana Department of Transportation. The data provided by IDOC were the ES202 data collected by the Indiana Department of Workforce Development. ES202 data from the second quarter of 1988 through the second quarter of 1994 were used in the analysis.

Basic and nonbasic employment was determined by using the location quotient (Mayer and Pleeter) and assumptions approach (Lewis) methods. All employment in manufacturing, agriculture, and mining was allocated to basic employment (Lewis; Shaffer). Employment in the other sectors was disaggregated between basic and nonbasic employment using the location quotient method with the U.S. economy being the benchmark economy (Mayer and Pleeter; Shaffer).

Employment in the manufacturing sector was disaggregated into new and existing employment based on the amount of time the establishment existed. If the establishment had been in the data set for 4 quarters or less, it was considered new. Establishments that had appeared in the data set for more than four quarters were considered existing.

Localization economies were measured by the percent of the local workforce in the manufacturing sector. The measure was calculated from the ES202 employment data by dividing total manufacturing employment by total employment for each quarter by county.

Population was used to represent urbanization economies. Annual U.S. Bureau of Census population estimates were obtained from the Economic Development Information Network developed by the Indiana University Business Research Center.

The percent of the population 25 years old or older with a high school diploma is used to represent labor quality. This data was obtained from the 1990 Census of Population.

The miles of interstate highway in the county are used to measure transportation infrastructure. The data on miles of interstate was obtained from the Indiana Department of Transportation and were compiled from 1992 data.

 

Variable Measurement

In estimating an economic base model, it is important to measure variables at the appropriate geographic or jurisdictional level. Since most local policy is confined to be within the county or a subset of the county it would appear the county is the appropriate unit of measure for the analysis. However, research suggests it may be more appropriate to measure market and labor characteristics that influence growth at the multi-county level (Henry and Drabenstott). The multi-county areas provide a better measure of the resources available to establishments within the area.

Economic structure variables, except the interstate variable, are measured at the labor market area (LMA) level. An LMA is a multicounty area designated by place of residence ó place of work commuting patterns (Tolbert and Killian). The LMA variables were calculated by taking a weighted average of county variables. The weights used were the size of workforce for LMA manufacturing and the population for LMA population and education. The employment and interstate variables are measured at the county level to account for concentration of employment in a county and to measure the countyís level of transportation infrastructure.

Summary statistics for each of the variables are presented in Table 2. The values in Table 2 represent county averages from 1988 to 1994 for each county classification. Metro counties have more employment in each of the basic sectors except agriculture and government, for which nonadjacent counties have the highest amount of basic employment. One would expect nonadjacent counties to have a larger agriculture sector, but the government sector being the largest in nonadjacent counties was not expected and is due to the location of a military base.

Metro counties have more than twice the amount of transportation infrastructure (20.74 miles) as adjacent counties (8.16 miles) and over four times the amount of nonadjacent counties (4.63 miles). The metro population (566,117) is forty-six percent larger than adjacent population (387,614) and twice the size of nonadjacent (280,618). Adjacent and nonadjacent counties have a slightly higher percentage of their workforce in manufacturing (29 percent) than metro (27 percent). There is only a minor difference in the percent of the adult population with a high school degree, with metro counties having 76.03, adjacent 74.63, and nonadjacent 74.08 percent.

 

Estimation Technique

A three equation model for Indiana county employment growth was estimated using the seemingly unrelated regressions (SUR) method in SAS (SAS Institute Inc.) with time series and cross-sectional data (Pindyck and Rubinfeld).

Equations were estimated for three county classifications: metro, adjacent, and nonadjacent, to test for differences across county types. National and international economic shocks likely affect all of the counties regardless of their classification which would lead to the three equations having correlated disturbances. To utilize the information captured in the correlation of the disturbances, the equations were estimated using the seemingly unrelated regression (SUR) method (Pindyck and Rubinfeld).

 
Results

The model results (Table 3) indicate the expansion of the basic sector will lead to employment growth in the rest of the local economy. There is little change in which basic sectors are significant across county types. Thus, all communities experiencing growth in basic sector employment will experience additional employment growth from the nonbasic sector in the local economy. The results are consistent with export base theory and results from previous studies (Shahidsaless et al.; Kraybill and Dorfman; and Lesage and Reed).

The influence basic sector employment growth has on nonbasic sector employment growth depends on which basic sector increases employment. Multipliers computed from the SUR results (Rainey) estimate nonbasic employment growth associated with employment growth in each basis sector (Table 4). The magnitude of the multiplier is related to basic sector firmsí marginal propensity to consume local products as well as householdsí consumption patterns. Larger multipliers imply that firms in a basic sector have established more linkages with the local economy than firms in basic sectors with smaller multipliers.

The results of the study are conflicting concerning the influence of employment growth in new and existing manufacturing establishments on nonbasic employment growth. Each new manufacturing job in a new establishment in metro county is associated with 1.58 new jobs in the nonbasic sector. Each new manufacturing job in an existing metro manufacturing establishment stimulates 1.26 jobs in the nonbasic sector. Employment growth in existing manufacturing establishments is associated with more nonbasic employment growth than new jobs in new manufacturing establishments.

In adjacent and nonadjacent counties a new job in an existing manufacturing establishment is associated with .54 and .37 nonbasic jobs, respectively. A new job in a new manufacturing establishment is associated with .52 and .25 nonbasic jobs for adjacent and nonadjacent counties, respectively. This suggests that employment growth by new and existing firms has similar impacts on total employment growth in adjacent and nonadjacent counties.

The differences in nonbasic employment impacts associated with new and existing manufacturing firm employment growth suggest that manufacturing employment growth has a larger employment growth impact the larger the community where the growth occurs. Manufacturing location and growth literature (Herzog and Schlottmann; Carlino and Mills) indicates that firms tend to locate and grow in communities that have infrastructure, population, and business and consumer services. The results of this research with regard to manufacturing employment growth suggest that firms and their employees use local business and consumer suppliers if they are available.

The difference of the size of the multipliers associated with new and existing manufacturing employment growth in metro counties suggests that new firms and their employees purchase a larger share of their service and product needs locally than existing firms. This is different from research on rural manufacturing that concluded existing manufacturing firms purchased a larger share of inputs in the local market than new manufacturing firms (Barkley and McNamara). The results of this study suggest no differences in the input purchasing behavior of existing and new firms in adjacent and nonadjacent counties.

Of the other basic sectors, the transportation and utilities sector is associated with the largest nonbasic employment multiplier and the FIRE sector the smallest. This implies the transportation and utilities sector has the highest marginal propensity to consume locally. The high local consumption of the transportation sector could be due to the amount of service industries, fueling, maintenance, insurance, and weighting used by this sector. The FIRE sector has the lowest propensity to consume local products, suggesting that finance, insurance and real estate firms and their employees purchase fewer business and consumer goods and services in the local economy than the firms and their employees in other basic sectors.

Employment growth in the agriculture and mining sector is not associated with nonbasic employment growth in metro or nonadjacent counties. The expansion of the wholesale sector caused nonbasic employment to contract in all counties. This is not consistent with our hypothesis or export base theory. It could be that use of the location quotient method does not correctly allocate basic wholesale employment activity to the basic sector.

In general, nonbasic employment multipliers are larger for metro counties than for adjacent counties, and employment multipliers in adjacent counties tend to be larger than multipliers in nonadjacent counties. This supports the hypothesis that communities with more urbanization economies will receive a larger stimulus from expansion of the basic sectors. The larger size and diversity of markets in metro areas allows for more interactions between industries and households.

The trend for multipliers to be larger in metro counties suggests that metro counties may benefit from basic sector growth in adjacent and nonadjacent counties. To the extent the basic sector firms and their employees from adjacent and nonadjacent communities purchase business and consumer goods and services in metro counties, the metro economy benefits. These spillovers stimulate employment growth in nonbasic employment in nearby metro counties.

 

Impact of Economic Structure

We hypothesized the structure of the local economy would influence the growth of nonbasic employment and impact the way in which the basic sectors interact with the nonbasic sector. The results support both of these hypotheses. In metro and nonadjacent counties, the economic structure influences the relationship between basic and nonbasic sectors. In adjacent and nonadjacent counties, the economic structure influences nonbasic employment growth.

 

Metro Counties

In metro counties, the structure of the economy influences how the manufacturing sectors interact with the nonbasic sector but does not have a significant relationship with nonbasic employment growth. The structure of the economy decreases the impact of employment growth in new establishments on nonbasic employment, while increasing the impact of employment growth in existing establishments.

The marginal impact of urbanization economies positively impact both new and existing establishments, implying there is more interaction between both types of establishments in larger communities. Urbanization economies refer to the cost savings associated with the concentration of economic activity in a particular location. This supports the hypothesis that larger communities provide more opportunity for establishments to interact and create linkages.

Infrastructure negatively impacts the interaction of employment in new establishments and nonbasic employment, but positively impacts the interaction of employment in existing establishments and nonbasic employment. High levels of infrastructure allow existing firms easier access to other establishments within the metropolitan area, providing better opportunities for increased interaction. However, infrastructure also provides better access to markets outside of the metro area. Thus, new firms can locate in a metro area and still have easy access to other markets.

Localization economies and labor quality decrease the relationship between manufacturing and nonbasic employment. These findings are not consistent with the hypotheses that higher levels of the structural measures would improve the interaction between basic and nonbasic employment. Localization economies refer to the cost savings associated with the concentration of manufacturing activity in a particular location. The negative relationship of localization economies may indicate that labor markets are tight in areas that already have a high concentration of manufacturing activity. Thus, when employment in the manufacturing sector expands it must pull employment out of the nonbasic sector.

 

Adjacent Counties

The structural variables do not affect the way in which manufacturing and nonbasic employment interacts in adjacent counties. This is not consistent with our hypothesis or previous findings. It implies that the linkages between industries will be the same regardless of the size of local markets. It could be that the industry located in adjacent communities are located in the community that can best provide it with the needed resources so that the linkages are the same across the communities.

The structure of the economy does impact the growth potential of nonbasic establishments, which supports our hypothesis. Urbanization economies have a positive impact on nonbasic employment changes. The size of the markets in larger communities provides more customers for nonbasic establishments to establish linkages. This is consistent with the hypothesis that as the size of the economy grows there is a need for more nonbasic products and services.

The level of infrastructure in the community has a positive relationship with the nonbasic sector, which supports our hypothesis. The presence of more transportation infrastructure provides better access to local markets to people outside the local economy. The infrastructure provides the nonbasic sector more opportunity for sales.

 

Nonadjacent Counties

The structural variables in nonadjacent counties influence the size of the nonbasic sector and the relationship between the manufacturing and nonbasic sectors. Urbanization economies in nonadjacent counties increase the relationship of nonbasic employment with both new and existing basic employment. This supports the hypothesis that larger communities provide more opportunity for industries to interact. Increased levels of infrastructure allow for easier transactions between industries. Infrastructure decreased the relationship between new and existing basic and nonbasic employment, implying better infrastructure allows for easier access to outside markets as well.

Localization economies increase the impact between new and existing basic sector employment and nonbasic employment, while labor quality decreases the impact. The localization economies finding supports the hypothesis that communities that have an existing presence in manufacturing activity can better serve the manufacturing sector. The negative relationship between labor quality and nonbasic employment may indicate that these industries do not require highly skilled labor.

 

Implications

The model results show that expansion of the basic sector does lead to greater employment in the nonbasic sector. Policies that lead to employment growth in the basic sector will lead to growth in the rest of the economy.

The impact from the expansion of basic employment on nonbasic employment varies by industry and by location. Traditionally, economic development policy focuses on the recruitment of manufacturing establishments. This strategy appears to have been successful in Indiana, particularly for metro counties. However, the results of this study indicate several of the other industrial sectors are equal to or better than the manufacturing sector at stimulating nonbasic employment. Community leaders might expand their development efforts to support industries in addition to manufacturing.

The manufacturing sector accounted for more than 60 percent of basic employment in each county classification. Relative to the other basic sectors, the manufacturing sector obtains more outside resources. Because the manufacturing sector can bring in larger amounts of outside resources, it is more capable of providing a large stimulus to nonbasic employment.

The impact from the expansion of basic sector employment on nonbasic employment growth tended to be highest in metro counties and lowest in nonadjacent counties. Policies aimed at stimulating the basic sector will affect rural economic growth differently than similar policies to encourage basic sector growth in metro area. Policy makers must keep this in mind when comparing the effectiveness of programs between metro and adjacent communities.

The larger leakages in adjacent and nonadjacent counties imply potential benefits for metro counties that provide goods and services to firms and households from adjacent and nonadjacent counties. Metro counties may experience employment growth if their leadership cooperates with leads in adjacent and nonadjacent counties to encourage economic growth. Wherever the growth occurs within the region, the metro county should experience employment growth.

The influence of infrastructure means that communities with higher infrastructure levels will receive more of an impact from the expansion of the basic sector. The results also indicate that increasing the level of infrastructure will lead to greater overall economic growth in the nonbasic sector. Therefore, policies that increase the level of infrastructure will lead to greater employment growth.

 

Summary

Basic sector employment leads to expansion of the nonbasic sector. The magnitude of the stimulus to nonbasic employment depends on which basic sector is expanding and how well the sector is linked with the local economy. The transportation and utilities sector has the largest multiplier in metro and nonadjacent communities, and the FIRE sector has the largest multiplier in adjacent counties. However, most communities are likely to obtain more employment in the manufacturing sector.

The results did not indicate difference between the nonbasic employment growth associated with new and existing manufacturing establishments in adjacent and nonadjacent counties. This implies new and existing manufacturing establishments are equally as good in stimulating nonbasic employment in these counties. In metro counties, however, employment growth in new manufacturing establishments is associated with higher growth than that associated with employment growth in existing manufacturing. Successful manufacturing recruitment adds more jobs than expansion of existing manufacturing firms in metro areas.

In general, nonbasic employment multipliers decline from metro to nonadjacent communities. The larger size and diversity of markets in metro areas allow for more interactions between metro industries. Metro areas capture a larger share of business and consumer purchases associated with the expansion of basic sector employment. This suggests policies that stimulate basic employment will have a greater impact in metro areas than in adjacent and nonadjacent areas.

Within each class of counties, those with smaller markets tend to receive less of an impact from the expansion of the basic sector. Larger metro and adjacent communities provide more opportunities for business and household linkages than smaller adjacent and nonadjacent communities. Communities with more infrastructure tend to receive more of a stimulus from existing manufacturing establishments, but less from new establishments. Infrastructure allows for better market access within the local economy and better access to outside markets. Communities with smaller markets and less infrastructure tend to have slower nonbasic employment growth associated with basic sector employment growth.

 

 
List of References
Ahlbrandt, R.S. Jr. "Adjusting to Changes in Traditional Markets: The Problems of Small Manufacturers in Older Industrial Regions." Economic Development Quarterly, August 1988, pp. 252-264.

Barkley, D.L. and K.T. McNamara. "Local Input Linkages: A Comparison of Foreign-owned and Domestic Manufacturers in Georgia and South Carolina." Regional Studies, 1994, pp. 725-737.

Bartik, T. "Small Business Start-Ups in the United States: Estimates of the Effects of Characteristics of States." Southern Economic Journal, Vol. 55, 1989, pp. 1004-1018.

Bartik, T. Who benefits from State and Local Economic Development Policies. Kalamazoo, MI. W.E. Upjohn Institute for Employment Research. 1991.

Bernat, G.A. Jr. "The Rural Economy in the 1990ís: The Role of Manufacturing." Paper presented at the Agricultural Outlook Conference USDA, Washington, D.C., December 3, 1992.

Campbell, C. and D.N. Allen. "The Small Business Incubator: Micro-Level Economic Development." Economic Development Quarterly. 1(2) 1987, pp. 178-191.

Carlino, G. and E. Mills. "The Determinants of County Growth." The Journal of Regional Science, 1987, 27(1), pp. 39-54.

Clifton, D.S. Jr., L.R. Edens, H.T. Johnson, and R.W. Springfield. "Elements of an Effective Technology Assistance Policy to Stimulate Economic Development." Economic Development Quarterly, 1989, pp. 52-57.

Coughlin, C.C., J.V. Terza, and V. Arromdee. "State Characteristics and the Location of FDI within the United States." Review of Economics and Statistics, 73(4), 1991, pp. 675-683.

Deaton, B. and D. Gunter. "The Influence of Community Characteristics on Industrial Plant Location and Expansion: A Preliminary View." Tennessee Farm and Home Science, 1974, pp.34-36.

Feller, I. "Evaluating State Advanced Technology Programs," Evaluation Review, June 1988, pp. 232-252.

Friedman, J., D. Gerlowski, and J. Silberman. "What Attracts Foreign Multinational Corporations? Evidence from Branch Plant Locations in the United States." Journal of Regional Science, 32(4), November 1992, pp. 403-418.

Hastings, S.E. and S.M. Bruckner. "An Introduction to Regional Input-Output Analysis." In Microcomputer-Based Input-Output Modeling. Ed by Daniel Otto and Thomas G. Johnson, Westview Press, 1993.

Henry, M. and M. Drabenstott. "A New Micro View of the U.S. Rural Economy." Economic Review. Federal Reserve Bank of Kansas City, 1996, pp. 53-70.

Herzog, H. and A. Schlottmann. Industry Location and Public Policy. Knoxville, TN. The University of Tennessee Press. 1991.

Kraybill, D.S. "Existing Business Create Most New Jobs in Ohio." Ohio's Challenge. Forthcoming.

Kraybill, D.S. and J.H. Dorfman. "A Dynamic Intersectoral Model of Regional Economic Growth." Journal of Regional Science. 1992, pp. 1-17.

Kriesel, W. Manufacturing Location within Economic Development. Masters Thesis, Virginia Tech University, 1983.

Kriesel, W. and K. McNamara. "A County-Level Model of Manufacturing Plant Recruitment with Improved Industrial Site Quality Measurement." Southern Journal of Agricultural Economics, 1991 pp. 121-127.

Krmenec, J. "The Employment Impacts of an Investment Incentive: Differential Efficiency of the Industrial Revenue Bond." Regional Studies, April 1990, 24(2) pp. 95-107.

Lesage, J.P. and J.D. Reed. "The Dynamic Relationship Between Export, Local, and Total Area Employment." Regional Science and Urban Economics, 1989, pp. 615-636.

Lewis, W.C. "Export Base Theory and Multiplier Estimation: A Critique." Annals of Regional Science, 1976, pp. 58-70.

Mayer, W. and S. Pleeter. "A Theoretical Justification for the use of Location Quotients." Regional Science and Urban Economics, 1975, pp. 343-355.

McNamara, K. "Recruiting Manufacturing Firms as a Community Development Strategy." West Lafayette, IN, Purdue University, Cooperative Extension Service, EC 659, 1991.

McNamara, K., W. Kriesel, and D. Rainey. "Manufacturing Recruitment as A Rural Development Strategy." Chapter 7, pp. 119-136, in David W. Sears and J. Norman Reid, editors, Rural Development Policy, Greenwood Press, 1996.

Morse, G.W. The Retention and Expansion of Existing Businesses: Theory and Practice in Business Visitation Programs. Iowa State University Press, Ames, Iowa, 1990.

North, D.C. "Location Theory and Regional Economic Growth." Journal of Political Economy, 1955, pp. 243-258.

Pindyck, R.S. and D.L. Rubinfeld. Econometric Models and Economic Forecasts. New York, McGraw-Hill, Inc., 1991.

Plaut, T. and J. Pluta. "Business Climate, Taxes and Expenditures and State Industrial Growth in the United States." Southern Economic Journal, 1983, pp. 99-119.

Rainey, D.V. "Growth Impacts of New Firm Investment Versus Existing Firmsí Expansion." Unpublished Ph.D. dissertation, Department of Agricultural Economics, Purdue University, West Lafayette, IN, May 1997.

SAS Institute Inc. SAS/ETS Userís Guide, Version 6, SAS Institute Inc., Cary, N.C., 1990.

Schmenner, R., J. Huber and R. Cook. "Geographic Differences and the Location of New Manufacturing Facilities." Journal of Urban Economics, 1987, pp. 83-104.

Shaffer, R. Community Economics: Economic Structure and Change in Smaller Communities. Ames, IA, Iowa State University Press, 1989.

Shahidsaless, S.W. Gillis, and R. Shaffer. "Community Characteristics and Employment Multipliers in Nonmetropolitan Counties, 1950-1970." Land Economics, 1983, pp. 84-93.

Shapira, P. "Modern Times: Learning from State Initiatives in Industrial Extension and Technology Transfer," Economic Development Quarterly, August 1990a, pp. 186-202.

Smith, E.B. Deaton and D. Kelch. "Location Determinants of Manufacturing Industry in Rural Areas." Southern Journal of Agricultural Economics, July 1978, pp. 23-32.

Smith, T. and W. Fox. "Economic Development Programs for States in the 1990's." Economic Review, Federal Reserve Bank of Kansas City, July/August 1990, pp. 25-35.

Standard Industrial Classification Manual. Prentice Hall Information Service, 1988.

Steinnes, D. "Business Climate, Tax Incentives, and Regional Economic Development." Growth and Change, April 1984, pp.38-47.

Tolbert, C. and M. Killian. Labor Market Areas for the United States. United States Department of Agriculture, Economic Research Service. AGES870771, 1987

U.S. Bureau of the Census. Statistical Abstracts of the United States: 1996 (116th edition), Washington, D.C., 1996.

Walker, R., D. Greenstreet and F. Calzonetti. An Analysis of New Manufacturing Plants in Appalachia: Why Do They Locate Here and What Makes Them Successful? Final Report to the Appalachian Regional Commission, July 1989.

Woodward, D. "Location Determinants of Japanese Manufacturing Start-ups in the United States." Southern Economic Journal, 58(3), 1992, pp. 690-708.

TABLE 1. Variable Definitions and Sources

 
 
Name Definition Source
     
NB Nonbasic employment Computed from ES202 data files
AG Agriculture basic employment; includes agriculture, forestry, fisheries, and mining Computed from ES202 data files
MANE Existing manufacturing basic employment Computed from ES202 data files
MANN New manufacturing basic employment Computed from ES202 data files
RETAIL Retail basic employment Computed from ES202 data files
SERV Service sector basic employment Computed from ES202 data files
GOV Government sector basic employment Computed from ES202 data files
CON Construction sector basic employment Computed from ES202 data files
TRAN Transportation and Public Utilities sector basic employment Computed from ES202 data files
WHOLE Wholesale trade sector basic employment Computed from ES202 data files
F.I.R.E. F.I.R.E. sector basic employment Computed from ES202 data files
HWY Miles of interstate highway in the county Compiled by Dadang Mohamad
LMAHS Percent of population 25 years or older that graduated high school in the LMA Computed from the 1990 Census of Population
LMAPOP Total population of the LMA Computed from the Economic Development Information Network population estimates
LMAPCMAN Percent of LMA workforce employed in manufacturing Computed from ES202 data files
Q1-3 Seasonal dummy variables for first three quarters of year Computed by Dan Rainey
EXTPOP Interaction term of MANE and LMAPOP Computed by Dan Rainey
EXTHWY Interaction term of MANE and HWY Computed by Dan Rainey
EXTMAN Interaction term of MANE and LMAPCMAN Computed by Dan Rainey
EXTHS Interaction term of MANE and LMAHS Computed by Dan Rainey
NEWPOP Interaction term of MANN and LMAPOP Computed by Dan Rainey
NEWHWY Interaction term of MANN and HWY Computed by Dan Rainey
NEWMAN Interaction term of MANN and LMAPCMAN Computed by Dan Rainey
NEWHS Interaction term of MANN and LMAHS Computed by Dan Rainey
COLMAPOP Interaction term of LMAPOP and POP Computed by Dan Rainey
POP County population Economic Development Information Network (EDIN)
 
Note: Employment data covers the second quarter of 1989 to the second quarter of 1994; Population values are annual estimates; Education measures are for 1990; and Interstate highway miles are for 1992.

TABLE 2. Summary Statistics for Model Variables
Metro Counties (37 counties)
Standard
Variable N Mean Deviation Minimum Maximum

NB 740 32,881 63,225.34 258.00 389,215

AG 740 149 169.17 0.00 1,035

MANE 740 11,419 16,248.34 7.00 84,073

BN 740 250 703.71 0.00 7,536

RETAIL 740 1,481 1,751.82 0.00 9,469

SERV 740 2,126 3,858.88 0.00 24,027

CON 740 510 866.22 0.00 5,169

TRAN 740 917 2,063.34 0.00 12,154

WHOLE 740 441 1,423.33 0.00 9,328

FIRE 740 463 1,753.17 0.00 11,106

GOV 740 244 1,251.20 0.00 8,031

HWY 740 20.74 20.82 0.00 112.18

LMAHS 740 76.03 3.06 70.22 81.56

LMAPOP 740 566,117 407,243 148,500 1.3E+6

LMAPCMAN 740 0.27 0.08 0.15 0.48

Q1 740 0.25 0.43 0.00 1.00

Q2 740 0.25 0.43 0.00 1.00

Q3 740 0.25 0.43 0.00 1.00

EXTPOP 740 7.4E+9 1.7E+10 1.1E+6 1.1E+11

EXTHWY 740 504,580 1,524,708 0.00 9,431,309

EXTMAN 740 3,225.51 4,829.73 2.04 28,188.84

EXTHS 740 869,443 1,247,331 491 6,585,887

NEWPOP 740 201,766,368 871,861,360 0.00 9.8E+9

NEWHWY 740 14,511.40 75,155.40 0.00 845,388.48

NEWMAN 740 63.93 145.33 0.00 1,218.27

NEWHS 740 19,214.17 54,895.14 0.00 590,335.17

COLMAPOP 740 7.8E+10 1.7E+11 7.9E8 1.1E+12

POP 740 108,534.02 148,205.84 5,300.00 817,207.00

 

 

TABLE 2. Continued

Adjacent Counties (42 counties)
Standard
Variable N Mean Deviation Minimum Maximum

NB 840 5,587 5,510 548 29,486

AG 840 143 153 0 846

BE 840 3,501 3,730 46 15,990

BN 840 72 146 0 1,203

RETAIL 840 259 222 0 1,199

SERV 840 310 372 0 1,762

CON 840 43 79 0 526

TRAN 840 146 226 0 949

WHOLE 840 24 51 0 241

FIRE 840 21 24 0 141

GOV 840 57 152 0 859

HWY 840 8.16 11.67 0 41.19

LMAHS 840 74.63 3.20 69.40 81.56

LMAPOP 840 387,614 264,852 104,942 1,341,712

LMAPCMAN 840 0.29 0.08 0.15 0.48

Q1 840 0.25 0.43 0.00 1.00

Q2 840 0.25 0.43 0.00 1.00

Q3 840 0.25 0.43 0.00 1.00

EXTPOP 840 1.4E+9 1.9E+9 4,8E+6 1.0E+10

EXTHWY 840 46,206.22 99,683.91 0.00 496,069.33

EXTMAN 840 1,150.92 1,433.45 10.01 6,719.32

EXTHS 840 259,915 276,251 3,192 1,195,801

NEWPOP 840 27,341,393 58,446,823 0.00 6.1E8

NEWHWY 840 816.07 2,700.79 0.00 29,891.89

NEWMAN 840 24.88 53.63 0.00 372.40

NEWHS 840 5,401.16 10,772.18 0.00 91,401.47

COLMAPOP 840 1.2E+10 1.4E+10 7.3E+8 8.7E+10

POP 840 29,346.29 19,758.35 6,976.00 110,008.00

 

TABLE 2. Continued

 

Nonadjacent Counties (13 counties)

 

Standard

Variable N Mean Deviation Minimum Maximum

 

NB 260 6,668 5,036 1,885.00 22,227.00

AG 260 198 171 1.00 735.00

BE 260 3,761 2,970 512.00 11,946.00

BN 260 104 246 0.00 1,995.00

RETAIL 260 329 426 0.00 1,563.00

SERV 260 463 667 0.00 2,748.00

CON 260 85 103 0.00 414.00

TRAN 260 86 115 0.00 545.00

WHOLE 260 51 74 0.00 266.00

FIRE 260 23 33 0.00 165.00

GOV 260 380 1,254 0.00 4,905.00

HWY 260 4.63 8.22 0.00 23.03

LMAHS 260 74.08 2.19 69.40 76.45

LMAPOP 260 280,618 90,890 104,942 388,023

LMAPCMAN 260 0.29 0.06 0.20 0.38

Q1 260 0.25 0.43 0.00 1.00

Q2 260 0.25 0.43 0.00 1.00

Q3 260 0.25 0.43 0.00 1.00

EXTPOP 260 1.0E+9 1.0E+9 1.6E+8 4.6E+9

EXTHWY 260 27,456.23 55,845.82 0.00 209,573.00

EXTMAN 260 1,124.59 883.48 108.08 3,025.58

EXTHS 260 276,278.82 216,000.98 37,940.98 889,113.37

NEWPOP 260 2.8E+7 7.9E+7 0.00 7.5E+8

NEWHWY 260 630.53 1,777.71 0.00 10,593.80

NEWMAN 260 30.71 72.11 0.00 541.30

NEWHS 260 7,686.03 18,356.16 0.00 148,483.27

COLMAPOP 260 7.7E+9 3.9E+9 2.7E+9 1.5E+10

POP 260 29,308.35 15,514.60 10,361.00 72,557.00

 

 

Source: Calculated from ES202 data.

 

TABLE 3. SUR Regression Results

Dependent Variable: Change in Nonbasic Employment

 

Basic Sector Metro Adjacent Nonadjacent

AG -0.54 0.35 *** -0.01

MANE 0.77 1.41 ** 2.62 ***

MANN 17.56 *** 1.64 4.62 ***

RETAIL 1.40 *** 0.75 *** -0.80 ***

SERV 0.77 *** 0.43 *** 0.34 ***

CON 1.60 *** 0.90 *** 0.14

TRAN 1.78 *** 0.61 *** 1.42 ***

WHOLE -2.27 *** -0.54 * -0.70 ***

F.I.R.E. 0.71 *** 1.05 *** 0.25

GOV 0.89 -0.21 0.08

 

Economic Structure

HWY 2.22 0.87 ** 1.12 *

LMAHS -0.28 -0.09 -0.08

LMAPOP 5.0E-4 1.2E-3 *** 5.5E-4

LMAPCMAN -119.64 -20.74 10.47

 

Interaction Terms

EXTPOP 2.0E-6 *** 1.3E-6 *** 2.4E-6 ***

EXTHWY 0.01 *** 3.4E-3 -9.0E-3 ***

EXTMAN -2.47 *** 0.32 1.32 **

EXTHS -2.6E-3 -0.02 *** -0.04 ***

NEWPOP 4.4E-6 *** 1.1E-6 3.8E-6 ***

NEWHWY -0.02 *** 0.01 *** -0.04 ***

NEWMAN -7.23 *** -0.56 *** 1.41 **

NEWHS -0.21 *** -0.02 -0.08 ***

LMACOPOP 1.8E-9 *** 1.7E-10 2.8E-9 ***

 

Seasonal Dummies and intercept

Q1 -183.56 *** -28.38 *** -11.04 ***

Q2 62.77 * 47.52 *** 8.50 ***

Q3 -27.79 4.55 10.10 ***

 

Intercept 36.80 -6.09 -1.91

 

Inverse Covariance Matrix

 

Metro Adjacent Nonadjacent

 

Metro 3.55E-6 5.81E-7 6.59E-7 Adjacent 5.78E-7 7.1E-5 3.66E-6

Nonadjacent 6.59E-7 3.66E-6 2.78E-4

 

System Weighted MSE: 0.99999 with 5439 degrees of freedom.

System Weighted R-Square: 0.6070

 

Durbin-Watson 2.65 2.70 2.30

Estimated Rho -0.34 -0.38 -0.16

Note: *** is significant at the .01 level.

** is significant at the .05 level.

* is significant at the .10 level.

 

TABLE 4. Multipliers from SUR Regression Results

 

Dependent Variable: Change in Nonbasic Employment

 

Basic Sector Metro Adjacent Nonadjacent

AG -0.54 0.35 *** -1.91

MANE 1.26 * 0.54 *** 0.37 ***

MANN 1.58 *** 0.52 0.25 ***

RETAIL 1.40 *** 0.75 *** -0.80 ***

SERV 0.77 *** 0.43 *** 0.35 ***

CON 1.60 *** 0.90 *** 0.14

TRAN 1.78 *** 0.61 *** 1.42 ***

WHOLE -2.27 *** -0.54 * -0.70 ***

F.I.R.E. 0.71 *** 1.05 *** 0.25

GOV 0.89 -0.21 0.08

 

Economic Structure

HWY 126.69 12.66 *** -37.52 *

LMAHS -83.99 -71.96 -173.16

LMAPOP 0.02 0.01 *** 0.01

LMAPCMAN -30166.40 1052.01 5140.53

 

Interaction Terms

EXTPOP 2.0E-6 *** 1.3E-6 *** 2.4E-6 ***

EXTHWY 0.01 *** 3.4E-3 -9.0E-3 ***

EXTMAN -2.47 *** 0.32 1.32 **

EXTHS -2.6E-3 -0.02 *** -0.04 ***

NEWPOP 4.4E-6 *** 1.1E-6 3.8E-6 ***

NEWHWY -0.02 *** 0.01 *** -0.04 ***

NEWMAN -7.23 *** -0.56 *** 1.41 **

NEWHS -0.21 *** -0.02 -0.08 ***

LMACOPOP 1.8E-9 *** 1.7E-10 2.8E-9 ***

 

Seasonal Dummies and intercept

Q1 -183.56 *** -28.38 *** -11.04 ***

Q2 62.77 * 47.52 *** 8.50 ***

Q3 -27.79 4.55 10.10 ***

Intercept 36.80 -6.09 -1.91

Note:

*** is significant at the .01 level.
  ** is significant at the .05 level.
    * is significant at the .10 level.