cision to locate in Spring Hill, Tennessee. Spring Hill's central location and proximity to transportation and cost reducing interstate highways were the primary considerations. 55

Efficiency of tax incentives for enterprise zones

Even if tax incentives can significantly affect the location decisions of firms, it is unclear whether the induced investment in enterprise zones constitutes net new investment or whether it is merely investment shifted from another locale. If investment in enterprise zones replaces investment that would have taken place elsewhere (for instance, if investment moves away from established centers of economic activity and toward designated enterprise zones), the primary effect of the investment incentives would be redistributional. To the extent that investment in enterprise zones is investment which is redistributed from local labor markets with low unemployment to local labor markets with high unemployment, the enterprise zone programs may direct investment from expensive local labor markets to those with an excess of relatively less expensive, under-utilized labor. In this event, the enterprise zone programs may generate welfare gains for the economy as under-utilized resources are tapped. Efficiency gains also may result if reductions in unemployment lead to reductions in social ills such as crime, which some analysts view to be an externality associated with unemployment and blight.

Some attempts have been made to determine the extent to which investments in State enterprise zones represent relocations of existing businesses from outside the zone. A survey of businesses located in enterprise zones by the Department of Housing and Urban Development found that only 9.1 percent of the businesses had relocated from outside the zone. An additional 7.5 presented branches of an existing establishment located outside the zone, 26.4 represented new businesses, 2.2 percent represented businesses which had been kept from closing, and 54.8 percent of investments represented expansions of existing zone businesses. 56 The small percentage of business relocations understates the extent to which enterprise zone investments are potentially redistributional. The HUD urvey also revealed that such relocations of existing businesses averaged more new hires than either new firms or existing firm expansions. Moreover, the survey is not able to discern, except perhaps in the case of those businesses which would have otherwise shut their doors, whether these enterprise zone investments would have occurred outside the zone in the absence of the zone benefits. British enterprise zones have been found to experience a greater rate of relocations, 37 percent. 57

By design an enterprise zone will provide benefits to a business located within an enterprise zone while a similar business located just outside the enterprise zone boundary may not be eligible for such benefits. The potential for relocation, discussed above, could lead to areas located near enterprise zones losing businesses and economic vitality. For example, one study of British enterprise zones found that 86 percent of businesses which relocated into enterprise zones came from the same county as that which contained the enterprise zone. 58 A result could be that the social and economic problems which led to the establishment of an enterprise zone are shifted to neighboring areas. On the other hand, if the enterprise zone does experience economic growth, other taxpayers may find it advantageous to locate near the enterprise zone to either serve enterprise zone businesses or to live near their place of employment. If this occurs, some of the benefits of the enterprise zone spillover into neighboring areas that are not designated as a part of the enterprise zone.

In addition to providing incentives to locate existing businesses in particular geographical areas, the incentives could induce the creation of new businesses which would not otherwise have been initiated in any location. Such new businesses could produce taxable profits and incomes which might reduce the revenue cost of the incentives. On the other hand, the incentives could induce investments inside enterprise zones which would be uneconomic in the absence of the tax incentives. Such an outcome would reduce the efficiency of aggregate national investment.

Competition between communities for the location of business may reduce the efficiency of tax incentives and other inducements. Thus, even if enterprise zones provide sufficient incentives to affect the location decisions of firms, an additional question is whether these incentives are cost-effective. To be cost-effective, the tax subsidies should be the smallest subsidies needed to achieve the desired behavioral change. Moreover, the subsidies should be narrowly targeted so that the benefits go primarily to firms that change their economic behavior in the desired fashion. That is, a cost-effective tax incentive program would minimize the amount of subsidy going to investors who would have located in the enterprise zone even in the absence of the tax subsidy program. When communities compete with one another using financial incentives, the chosen community may spend, in tax and other benefits, more than is necessary to induce the business to locate in a particular location. 59

One measure of the efficiency of tax incentives for enterprise zones is the cost, in terms of direct spending and foregone revenue, per job created. Several studies have attempted to calculate the cost per job created by enterprise zone programs. Such calculations are inherently difficult. It is difficult to determine how many jobs in an enterprise zone represent net additions to employment. To the extent that there are net additions to employment within the enterprise zone, there may be other jobs created or lost outside the enterprise zone. Such employment changes are equally difficult to quantify. In addition, it may be important to distinguish all new jobs from those new jobs gained by residents of the enterprise zone.

A recent study has attempted to measure the cost to taxpayers of the new jobs created by the Indiana enterprise zone program. The study calculates the annual cost in 1988 per job created ranged across zones between $815 and $11,747 with a mean of $4,921. If, however, one looks at the cost per new job for an enterprise zone resident, the cost ranged between $1,722 and $173,539 with a mean of $33,543.60 A study of New Jersey's enterprise zone program put the costs per job at between $8,000 and $13,000, but if the program costs per year remain at the levels experienced in 1987 and 1988, then the costs per job would be $40,000 to $60,000 over a decade and continue to grow. 61 Analysis of British enterprise zones concluded that the cost of each additional job created in an enterprise zone was 8,500 pounds sterling (approximately $15,300 at current exchange rates) and the cost of each additional job in the wider area was between 23,000 and 30,000 pounds sterling ($41,000 to $54,000 at current exchange rates).62

Tax incentives and the type of business formation

The choice of tax incentives granted to enterprise zone businesses may influence the type of business that will take place in an enterprise zone. For example, tax incentives for investment may induce more capital intensive businesses to locate in enterprise zones. Alternatively, if only employment subsidies are offered, more labor intensive businesses may be expected to locate in enterprise zones. Size limitations may induce small rather than large businesses to locate in enterprise zones. When a number of tax incentives are offered, the relative value of the different tax preferences may influence the type of businesses which locate within an enterprise zone.

Some argue that the need for enterprise zones grows out of the persistence of areas of pervasive unemployment and poverty, and therefore it is more appropriate to induce labor intensive businesses to locate in enterprise zones. It is argued that the resulting demand for labor simultaneously will attack both unemployment and poverty. Critics of this view note that there are no guarantees that the jobs created will be filled by residents of the enterprise zone. When jobs are filled by individuals from outside the enterprise zone, the objectives of reducing poverty and unemployment within the enterprise zone are not accomplished. Critics also observe that many labor intensive businesses are low wage employers which, by the nature of their business, offer little training to enhance the skills of employees. As a result, while employment might increase, poverty may only be somewhat mitigated and individuals' further economic advancement may still be limited by a lack of marketable skills. Proponents counter that society gains by reducing government welfare payments and individuals gain by establishing positive employment histories.

Others argue that it may be more appropriate to induce capital intensive businesses to locate within enterprise zones. Capital intensive businesses often require skilled workers and pay higher wages. Proponents also argue that businesses which make large investments are less likely to move once available subsidies expire or when another community offers financial inducements. As a consequence, such businesses may provide a more stable economic base to area development. Critics of this view observe that the residents of many areas which might qualify as enterprise zones may not have the skills necessary to gain employment in many capital intensive businesses, and little employment gain among residents may result. They also note that often individuals do not choose to reside in close proximity to many capital intensive businesses (for example, a steel mill) and large capital intensive businesses may, by locating within an enterprise zone, displace current residents. Such displacement, if it occurs, may only disperse the problems of economic development which were manifest in the area's designation as an enterprise zone.

Some assert that only small businesses should be permitted to take advantage of the economic inducements offered within enterprise zones. They note that small businesses are responsible for many of the jobs created within the United States and that small businesses are often innovators. Critics of this view observe that small businesses frequently fail. Consequently, small business may provide an unstable employment base for an enterprise zone. They contend that large employers are more stable employers. They further note that many small businesses need large businesses to purchase their products and often find it most economical to locate near major customers. Proponents counter that fostering many small businesses rather than one or two large businesses creates a broad economic base which is not subject to the business fluctuations of a single industry.

Others would like to encourage venture capital investments and the location of high technology businesses within enterprise zones. They contend that such businesses are the source of future economic growth and it is appropriate to direct some of this growth to economically under-developed areas. As discussed regarding capital intensive businesses above, high technology businesses may require skills not possessed by residents of an enterprise zone. Such businesses may find non-tax factors such as the proximity of scientists at a research university more important to their location decisions. Venture capital investments generally are high risk investments. Because a high risk implies a higher probability of failure, such investments may not provide stable employment opportunities within the enterprise zones.

Neutrality

Tax incentives for employment and investment in enterprise zones are intended to affect employment and investment decisions. However, these incentives can create an inefficient allocation of resources because the preferences can make it more profitable, on an after-tax basis, to locate property at site A rather than site B, even though site B would produce greater pre-tax profits. On the other hand, the incentives may be necessary to promote the social goal of more economic growth and opportunity in distressed areas. If the enterprise zone does help reduce crime and other social ills in an area, then an efficiency gain can result despite the non-neutral treatment of employment and investment decisions.

Targeting tax incentives within an enterprise zone can also reduce economic welfare below that which might be attainable under proposals with broad based incentives. For example, an employment tax credit may skew the allocation of resources within an enterprise zone to labor-intensive industries. Similarly, tax incentives for capital may skew the allocation of resources within an enterprise zone to capital-intensive industries. A proposal which provides incentives of similar magnitudes for all types of capital and all types of labor is likely to result in larger economic benefit per dollar of revenue cost than more narrowly targeted incentives.

Incidence of enterprise zone benefits

The tax benefits associated with enterprise zones are aimed at creating investment, employment, and business activity within the enterprise zones. However, as with any tax or subsidy, the ultimate division of these tax reductions among various classes of potential beneficiaries depends on demand and supply conditions in the affected markets and the particular characteristics of the proposals. In general, the incidence of a tax (or of a tax subsidy) falls most heavily on the factor of production that is least mobile, that is, the factor that is least able to escape the burden of the tax by changing behavior. Because enterprise zones distribute tax benefits according to geographic location, factors which are relatively immobile across geographic locations are more likely to receive the benefit of proposed tax incentives than are factors which are geographically mobile.

Among the groups that may benefit from the establishment of enterprise zones are those owning land in the zone, those who may gain employment in the zone, those who invest in the zone, and the entrepreneurs who organize businesses within the zone. 63 Land is an immobile factor. It may be expected that tax benefits granted for economic activity undertaken in enterprise zones will tend to result in higher prices for land in the enterprise zone. If other factors of production are to some extent immobile, some of the value of enterprise zone benefits may accrue to these factors. For example, if residents of other areas are unable to commute easily to jobs in enterprise zones, then residents of enterprise zones may receive some of the benefits granted for employment or investment in enterprise zones. 64 Likewise, if entrepreneurs possess specific knowl