The previous section made reference to the decision of governments to offer incentives. By contrast, this section refers to the decision made by firms to locate in particular areas, and why abatements might influence this decision. The traditional explanations are found in standard production and location theories.
10.4.1 Theoretical Arguments
According to production theory, a reduction in the price of capital (in this case, the PTA) will trigger two effects. The first is an increase in output, a parallel drop in price of the good produced by beneficiary firms, and an increase in demand for both capital and labor. The second effect is a substitution of capital (the factor made relatively cheaper) for labor. The output and substitution effects work in the same direction for capital.
Thus, demand for capital will invariably increase and put upward pressure
Promoting Economic Development with Tax Incentives g 271
on the price of capital. However, in the case of labor, the output and substitution effects work in opposite directions. If the substitution effect prevails, demand for, and the price of labor, could actually decrease.
The magnitude of the increase in capital investment and assessed value depends on the price elasticity of demand for the goods produced by beneficiary firms. The greater the price elasticity of demand, the greater the expected effect on capital investment and assessed value. According to Ihlanfeldt (1999), the price elasticity of demand is likely to be high for manufacturing industries that compete in the national or international market, because of opportunities for expansion, and low for locally mar- keted products. Along the same lines, James and Leslie Papke (1984) argue that investment tax incentives increase liquidity and influence the timing of capital acquisition, which encourages firms to retire and replace their plant and equipment more rapidly.
In addition to the standard supply-side, production theory rationale, a well-developed branch of the economic literature, known as location the- ory, has argued that profit-maximizing firms will choose the location that minimizes costs and thereby increases profits. PTAs and other incentives influence a firm’s cost function and its decision to locate within the juris- diction of the awarding government.
According to Nelson (1993), minimization of basic cost factors (trans- portation, access to markets, access to material inputs, and availability and cost of labor) was a good predictor of location in studies prior to 1960. In the period thereafter transportation costs decreased and innovations in technology added complexity to the economy. As a result, businesses gra- dually became ‘‘footloose.’’ James and Leslie Papke (1984) define footloose enterprises as those that are
not bound to particular locations because of resource and/or market availability. These are primarily manufacturers in pro- duction facilities for which they can obtain their raw materials and other supplies from a number of different locations and from which they can serve a broad market area. Moreover, they are primarily firms with multi-plant, multi-site operations whose products are sold in multi-state, national or international markets. (p. 65)
With these changes in the economy, other factors such as technical competence of the labor market, state and local taxes and expenditures, regional business climates, quality of life factors, inertia, agglomeration economies, coevolutionary development, serendipity and others became equally important (Nelson, 1993). All of these factors are considered to reduce firm costs in one way or another.
Businesses make location decisions in three stages (Schmenner, 1982).
In the first stage the decision to invest is made, and taxes are not considered to be an important factor. Next, the decision to locate in a region or state is taken (market stage), and in this stage taxes are expected to have a greater role, although inferior to variables that weigh more heavily in the firm’s cost function, such as labor cost differentials. Lastly, the decision to locate in a specific site is made (site stage) where it is considered that all other factors are relatively equal and hence tax incentives or tax differentials will have the greatest effect. In addition to taxes, government services enter into the equation, either because public spending substitutes for spending that private entrepreneurs would incur in the absence of public spending, because it enhances factor productivity, or because firms may not mind paying higher taxes as long as they receive valued services in exchange (Fox and Murray, 1990). Education is a good example.
Scholars debate whether taxes or tax incentives can play the role attributed by production and location theories. Some argue that taxes are a small fraction of business costs (2–3%) and should not be expected to have much impact (Lynch, 1996), especially if a jurisdiction is not competitive in uncontrollable factors such as transportation, labor and energy costs (Rubin and Zorn, 1985). However, others counter that, though small, they can amount to a significant part of total profits ( J. A. Papke and Papke, 1984). According to Oakland (1974), property tax rates often differ by as much as a factor of two and this implies cost differentials of approximately 10 percent of profits. Bartik (1991) is likewise not persuaded by the cost argument and affirms that, even if small, tax differentials may influence location decisions when all other factors are equal, i.e. in intraregional location decisions. Hall and Jorgenson (1991) equally stress going beyond assessing the effect of tax policy on costs to measuring empirically the effect on investment behavior. Finally, some argue that tax incentives may also contribute to economic development because of their positive impact on business climate ( J. A. Papke and Papke, 1984).
10.4.2 Empirical Literature
The empirical literature on location, taxation and economic development can be segmented according to different criteria. Several literature reviews divide studies according to the methodology employed, namely econo- metric, survey, case study, or hypothetical firm (P. S. Fisher and Peters, 1997;
J. A. Papke and Papke, 1984). Another common approach distinguishes between interstate, intermetropolitan, intrastate and intrametropolitan studies (Bartik, 1991; Wasylenko, 1997). A third way of segmenting is to differentiate between taxation and incentives, depending on whether the
Promoting Economic Development with Tax Incentives g 273
treatment or policy variable is an average measure of the level of taxation, such as the effective tax rate, or incentives such as PTAs and tax increment finance districts (P. S. Fisher and Peters, 1997; Wasylenko, 1997). Finally, further distinction draws the line between incentive packages, such as those that may be offered in an enterprise zone, and individual incentives such as tax increment financing or PTAs (P. S. Fisher and Peters, 1997). Most of the research has focused on the effects of taxation — as opposed to incentives
— on economic development, and on interstate and intermetropolitan loca- tion decisions. Intrastate, interlocal or intrametropolitan studies, as well as research on individual incentives or incentive packages, are in the minority.
Following a period in which empirical research suggested that taxes or incentives were relatively unimportant location factors, recent reviews presented by Greenhut (1956), Bartik (1991) and Wasylenko (1997) have provided encouraging summaries of the effects of taxation on economic development in both interregional and intraregional studies. However, voices continue to expressed caution (Lynch, 1996) and concern over the time dependency and irreplicability of results (McGuire, 1992), and recent evidence has suggested that states may actually be converging and making their tax systems similar (Annala, 2003). Fisher and Peters (1997) have presented a summary of the literature on incentive packages and enterprise zones, and likewise presented a mixed picture.
In the specific area of PTAs, the earliest attempt at verifying whether abatements induce investment was done by Ross (1953) who applied the survey approach to firms receiving abatements in Louisiana and concluded that abatements were ineffective, given that only 7% of investments would not have been made without the abatements. Similarly, Morse and Farmer (1986) used a survey approach on firms receiving abatements in Ohio and found that the percentage of investment influenced was 25%. Royse (1994) also used a survey approach and found that actual jobs and investment created generally exceeded those promised by companies. The problems with the survey approach (biased answers, etc.) are well known and will not be elaborated upon here.
Other approaches have been used. Morgan and Hackbart (1974) concluded that if abatements account for 5 to 10 percent of the increase in property value, a positive net present value can be achieved. Coffin (1982) found evidence that abatements slow the exodus or relocation of firms from the inner city. Wolkoff (1985) estimated that full abatement could decrease the price of capital by 4%, assuming a tax rate of $80 per
$1,000 and an assessment ratio of 0.5. However, this percentage fell considerably short of the 75% reduction required to increase the probability of investment from 0.23 to 0.39. Severn (1992) demonstrated that abate- ments approximate permanent reductions of tax rates on buildings as the abatement period increases.
Leslie Papke (1991; L. E. Papke, 1993) conducted an evaluation of the Indiana Enterprise Zone Program, which exempts inventories from property taxation for a period of 10 years, using three different specifications to control for selection bias. Findings included an 8% increase in the value of inventories in the zones relative to what it would have been without the program, and a decline in unemployment claims by about 19%.
Chang (2001), using similar methodology on the Indiana Economic Revitalization Area program, determined that the effect of PTAs on job creation varies by sector. Abatements given to the service sector are more effective at creating jobs than those given to the industrial sector due to the higher capital/labor ratios of the latter. Finally, the most recent peer reviewed study indicates that industrial abatements are only effective at increasing the tax base in the first years of the program; industrial abatements given at a later stage and commercial abatements are found to decrease the tax base of local jurisdictions (Wassmer and Anderson, 2001).
Although not in the specific area of PTAs, a recent study using a novel research design found encouraging results concerning the effects of suc- cessfully bidding for plants. Greenstone and Moretti (2003a) use articles from the journal Site Selection reporting on the winning and runner-up counties from a competition over some large plant. These authors used the runner-ups as the revealed (by profit maximizing firms) counterfactual for what would have happened in the winner counties in the absence of the plant opening. They found that a plant opening is associated with a 1.5% trend break in labor earnings in the new plant’s industry in winning counties (relative to losing ones). They also found a relative trend break of 1.1% in property values and considered this as evidence that the net effect on welfare was positive.
10.4.3 Methodological Lessons from Previous Empirical Research
The literature review provides a mixed picture and begs the question of why some studies produce significant results while others do not? One answer to this question is time, i.e., recent studies are more successful at finding significant effects. This has been attributed generally to increased methodological sophistication. Scholars have derived lessons from previous research in the area of taxation and economic development that take the form of design recommendations for future research. One of these recommendations is the need to perform natural experiments (Bartik, 1997).
A second is the necessity to control for unobserved variables (Bartik, 1991;
R. C. Fisher, 1997; Phillips and Goss, 1995). A third recommendation is the inclusion of public service controls, in addition to tax variables
Promoting Economic Development with Tax Incentives g 275
(Bartik, 1991; R. C. Fisher, 1997; Phillips and Goss, 1995). This third recommendation has evolved over time. Earlier studies that relied on public expenditure data as proxies for public services recommended that the fullest range of expenditures by function be utilized, including public welfare expenditures that may not directly benefit enterprises (Helms, 1985; Mofidi and Stone, 1990). More recently it has been recognized that expenditure data are poor measures of public service quality, and that both revenue and expenditure data are directly and immediately endogenous relative to local economic development (Bartik, 1997). Thus the recommendation subsists but in the sense of including quality measures of public services that matter to business location decisions. A fourth suggestion is to account for the possible endogeneity of independent variables (Bartik, 1991) and jointly model the behavior of firms and host communities (Oakland, 1974).