Which economic development policies work: determinants of state per capita income.

• Author: Trogen, Paul

Abstract

Economic development policies add to state economic efficiency and welfare if they compensate manufacturing firms for the positive externalities they produce. Incentives which try to alter business behavior, but do not produce positive externalities greater than their costs may, however, distort the market-place and result in reduced state economic efficiency and welfare. This article reports the results of a pooled cross sectional time series analysis that was conducted to estimate the influence of different types of economic development policies on one measure of overall welfare, change in state per capita income, for the years 1979 through 1995. Results suggest state development policies which offer tax breaks to all manufacturing firms, and programs which offer state loans and loan guarantees for all manufacturing firms, are positively related to growth in state per capita income. Programs which attempt to elicit specific firm behavior, such as incentives for new investment and incentives to create jobs, were negatively related to growth in per capita income. "Demand" side entrepreneurial state policies had no significant influence on per capita personal income.

Introduction

State policy makers invest in economic development policies as a means of reducing unemployment, attracting new capital investment, and building a larger tax base. But states have often invested blindly, not knowing which economic development policies actually achieve these goals. Proponents have made plausible cases for competing economic development policies, including: tax breaks for industry, tax breaks and/or subsidies for firms locating in the state, tax breaks and /or subsidies for existing plant expansion, and even social programs recast as investment in human capital. At the same time, critics have questioned the efficacy of economic development programs, suggesting states are competing against themselves in a zero sum game. The critics make an equally convincing case that converging state economic development programs cancel each other out, and only succeed in plundering state treasuries without any real benefit. Hampered by a lack of consensus on a theoretical model to explain economic development and the lack of a consistent body of evidence about which, if any, economic development policies have an impact on economic growth, states have been forced to rely on educated guesswork when adopting economic development policies.

Economic development policies can be theoretically justified on the grounds they improve economic efficiency and therefore the welfare of state citizens. Yet a lack of a clearly reasoned and empirically tested economic development strategy may cause states to squander scarce public resources on projects which cost more than the benefits those projects will deliver to its citizens. Inefficient economic development spending may reduce, rather than enhance, the efficiency of the state economy and the welfare of state residents. This study will first differentiate industrial development policies by what type of incentive they offer to encourage industry. Second, this study will test which strategies contribute to economic efficiency and welfare, and which do not.

If economic development policies can be justified on the grounds they improve economic efficiency, then a positive sum game is possible. State intervention may improve the efficiency of the market and total welfare because the market does not take externalities into account when setting prices and the quantity produced (Feiock, Dubnick and Mitchell, 1993: p. 61). The argument for state involvement in economic development is similar to that used to justify public aid to higher education. The state supports education because it generates positive externalities. Externalities occur whenever a private transaction creates either costs or benefits to a third party not involved in a transaction. The positive externalities which occur when a student receives higher education include: the student becomes a more informed citizen, a more productive employee, and is likely to contribute more in taxes during his/her lifetime. If the student had to pay the entire cost of education, he/she would not take into account the benefits to others in his/her calculation of whether to attend college. Without subsidies to higher education, some students would choose not to attend college, even though the benefits to them and society combined exceed the cost of education. Too little education would be produced because students and universities would not consider the benefits to third parties. Without state intervention, externalities cause markets to fail as a self regulating mechanism, and a sub-optimal amount of the good causing the positive externality is produced. Both market efficiency and total welfare can be improved by subsidizing those transactions which produce positive externalities, and taxing those which produce negative externalities (Weimer and Vining, 1992: pp. 152-162).

Positive externalities from the manufacture of exportable goods provide a theoretical justification for offering incentives to encourage the production of exportable goods. The manufacture of exportable goods creates positive externalities which extend far beyond the parties involved in the sale of the good. The export of goods bring new money into a state (Blair, 1995: pp. 123-130) and creates a multiplier effect which generates economic activity of up to 5 times the value of the original transaction (Peterson, 1981: p. 23). This economic activity creates jobs. Few state expenditures can contribute as much to the public welfare, yet alone at so little cost to the public. The state, therefore, can increase its economic efficiency and welfare by providing subsidies up to the amount of positive externalities generated by the manufacture of exportable goods. Current tax policies often create counter-productive disincentives for manufacturing exportable goods (Reich, 1983: pp. 3-9). The removal of these disincentives would also improve economic efficiency and state welfare. By adding to economic efficiency, economic development policies can create a positive sum game. Goods and jobs which otherwise would not exist would be created, not merely relocated.

The findings of this study suggest that incentives offered to all manufacturers, including tax breaks for manufacturing as a whole and financing aid for manufacturing, can increase economic efficiency and overall welfare, as measured by per capita personal income. Not every economic development policy, however, will make a positive contribution to economic efficiency and state welfare. Poorly conceived programs which offer subsidies in excess of the positive externalities of the goods manufactured may distort the market in the other direction, divert labor and capital from more productive alternative uses, may waste public funds which would have produced a better return if invested in other public programs or left in the hands of the taxpayers, and reduce economic efficiency and public welfare. Unfortunately, the economic development literature has done little to differentiate between types of economic programs and economic development strategies. Lacking a clear model, the efforts of practitioners and politicians have often been unfocused, causing them to "shoot at anything that flies and claim anything that falls" (Rubin, 1988: p. 236). The competition of states to attract large, highly visible plants (Grady, 1987: p. 87) has resulted in counter productive smokestack chasing and bidding wars, driving up the costs of incentives (Fulton, 1988: p. 39). In 1978, Pennsylvania gave Volkswagen a $71 million incentive package for a plant that was hoped to employ 20,000 workers, for a projected cost per $3,550, per job (Fulton, 1988; p. 32). The plant closed after 5 years and never employed more than 5,700 people, for a cost of $12,000 per job. In 1980, Tennessee won the competition for a Nissan plant with an incentive package that cost 33 million dollars (Wilson, 1989: p.8), or $11,000 per job (Fulton, 1988: p. 33). When Tennessee won a Saturn plant in 1985, the cost was 80 million dollars (Wilson, 1989: p. 8), or $26,000 per job (Fulton, 1988: p. 33). Shortly afterward, Kentucky won a new Toyota plant with an incentive package which monopolizes the states economic development budget with a price tag of between 125 and 150 million dollars, or a cost of about $50,000 per job (Fulton, 1988; p. 39). In 1993, Alabama won the bidding war for a Mercedes plan with an incentive package of 253 million dollars for a plant that will employ 1,500 workers (Zipser, 1995: p. 23), for a cost of nearly $165,000 per job (Kahan, 1996: p. 446). The costs of the incentives almost equal the cost of the 300 million dollar plant! As costs of these highly visible incentives packages surpass the value of the externalities they are supposed to correct, the act of granting these incentives will reduce economic efficiency and overall welfare. The potential for wasting scarce public dollars on counter productive economic development spending creates an urgent need to differentiate between policies which contribute to economic efficiency and public welfare, and those which detract from it.

The literature only imperfectly differentiates between types of programs. Much of the literature treats economic development policies as a monolithic block, without differentiating between different types of policies. Some authors use a variable "economic development policy" to identify the total number of economic development policies (Brace and Mucciaroni, 1990: p. 157; Brace, 1991: p. 300) or the percentage of listed policies offered (Brace, 1993: pp. 91-92), without differentiating by type of policy. Others use total economic development spending, without differentiating what incentives the funds are spend on. (Bingham and Bowen, 1994: pp. 501; Goss and Phillips, 1997...