|
Article Excerpt
With the revolution in social welfare policy at the federal level, responsibilities for more services have been moved back to state and local governments. This growing service burden and the continued decline in intergovernmental revenue have placed increased pressure on cities to protect and expand their tax bases. Previous research on local development policy focused primarily on the negative effect of high tax rates on business growth and the need for tax relief. Only recently has attention shifted to include the opposite problem--that unnecessary concessions give away too much of the tax base, providing corporate welfare at the expense of other taxpayers. As financial pressures on local governments build, so will pressure to use location incentives such as tax abatements with care.
Unfortunately, existing work on development policy does not provide much guidance to city officials in deciding the appropriate level of location incentives to offer firms in particular cases. Most research has tried to explain the macro-level city decision to adopt general economic development programs--either traditional `Type I' tax and financial incentives or `Type II' programs which require concessions from developers, such as hiring low-income city residents (Bowman, 1988; Clarke and Gaile, 1992; Dreier and Ehrlich, 1991; Elkins, 1995; Goetz, 1990; Green and Fleischmann, 1989, 1991; Reese, 1998; Rubin and Rubin, 1987). These studies do not address the micro-level decision on the value of incentives to offer in particular deals. Additionally, most of the significant explanatory factors have been city demographic conditions, such as the poverty rate or population growth, which municipal officials cannot control. These characteristics seem to act as constraints on the range of development policies a city will use, but there is still a great deal of variation within the constraints that the models cannot explain.
A few authors have incorporated cost-benefit analysis in their analysis of city economic development policies (Blair and Kumar, 1997). This approach does provide a decision criteria for cities to use for individual incentive packages, but it is a very conservative one. The analysis only determines the maximum incentive package that the city should offer--one equal to the total economic benefits generated by the development. It does not help the city determine the minimum concession package that a firm would accept or when to offer something between these two extremes.
A potentially more useful framework for city officials might be provided by a simple bargaining model. This approach provides a cohesive framework for considering the factors that influence the negotiations, incorporating those that earlier studies have found to be significant. It also demonstrates how differences in firm characteristics influence the final negotiated outcome and, therefore, how they should influence the city's choice of negotiation strategy. Cities do not consistently need to agree to everything that a firm requests or even to the maximum value derived from a cost/benefit analysis of the deal. Previous work has rarely addressed the firm's role in these negotiations. Primarily authors have argued that every firm holds an information advantage over every city--they know the level of incentives necessary for them to choose a particular location while city officials can never be certain (Bachelor and Jones, 1984). Since all firms hold this powerful advantage, any differences among them on other characteristics have been considered relatively unimportant and not included in studies of incentive policies.
However, firms do vary on several characteristics that can influence the negotiated outcome despite their information advantage. Most importantly they value specific locations differently, based on such factors as the quality of the workforce available or access to their markets. A firm with few substitutes for a particular location will be willing to pay a higher price for that site. These firms are in a weaker bargaining position than a business with many acceptable locations. The value that the firm attaches to the specific locale determines the minimum concession package that city must offer to beat competing locations and win the development, a relationship the bargaining model makes clear. City incentive offers can be made dependent on the characteristics of the firm in each deal, and this greater variation in bargaining strategy can result in the city retaining more of the benefits from its growing tax base.
BARGAINING MODELS
In this section I describe the simplest bargaining game, the Nash model. Because this is a fairly basic model, it does not capture every feature of the city-firm negotiation process, but it does provide a perspective for evaluating a city's strategy in specific incentive deals. If this framework appears useful in understanding the situation, then more complex bargaining games that capture additional features of the real world could be applied.
The Nash model is similar to a game of "splitting the pie". There is a valued item (the pie) that will be divided between two players. Each side wants more of the item and the more that any one player gets the less that the other will receive. Not all of the item needs to be given to the players--some could be left on the table--but that outcome is undesirable from the perspective of both participants. Either player could choose to walk away from the negotiations at some point because he could get a better deal with someone else. In the case of location incentives, the pie is equal to the economic value of the new development--including both the public sector benefits of increased taxes and the private sector benefits of increased jobs and payroll income. The two players are the city and the firm trying to reach an agreement over what percentage of this value the city will get to keep. The Nash model of the negotiation leads to the following solution:
Player One: [x.sub.1] = [b.sub.1] + [log [d.sub.2]]/[log [d.sub.1] + log [d.sub.2]] * (1 - [b.sub.1] - [b.sub.2])
Player Two: [x.sub.2] = [b.sub.2] + [log [d.sub.1]]/[log [d.sub.1] + log [d.sub.2]] * (1 - [b.sub.1] - [b.sub.2])
where [x.sub.i] = the proportion of the `pie' kept by player i
[b.sub.i] = the breakdown point or outside option of player i
[d.sub.i] = [1 - [q.sub.i]]/[1 + [r.sub.i]] = the discount factor of player i
[q.sub.i] = risk aversion of player i
[r.sub.i] = time preference of player i
The breakdown points ([b.sub.i]) represent the benefits that each player would get if this specific negotiation collapses and each party then pursues his next best option. These are the minimum values that each player will demand in the current negotiation, because each could do at least that well in a different deal. These breakdown points are measured as proportions of the development benefits (the pie) in the current deal. If the sum of these breakdown points is greater than one, each player is demanding more than the other can give, and no deal is possible. Negotiations break down and both players pursue their outside options in other deals. If the sum is less than one, there is a split acceptable to both parties. Each player gets the value of his outside option, and any amount remaining above the total of these values (1 - [b.sub.1] - [b.sub.2]) is divided based on the two players' levels of time preference and risk aversion (Binmore and Dasgupta, 1987; Binmore, Rubinstein, and Wolinsky, 1986; Rubinstein, 1982; Sutton, 1986).
DIRECT EFFECTS OF THE PARAMETERS
From the model it is fairly easy to see the direct impact that the three parameters will have on the negotiated outcome. The higher the value of a player's outside option ([b.sub.i]), the more he can command in any given deal. He must get at least the value of his breakdown point or he would prefer a deal with another partner where he is sure he can get that amount. A higher outside option does decrease the number of partners with whom a mutually acceptable arrangement can be made, but having fewer potential partners does not hurt this player. The only deals he loses are those less valuable than the certain benefits of his option value.
The impacts of time preference and risk aversion are slightly more difficult to see since they are combined in the discount factors ([d.sub.i]). The ratio of the log values of these discount factors determines the allocation of the remainder after each player has satisfied his outside option demands. The ratios in the two equations must sum to one, representing complete allocation of the remaining portion. Each player's payoff depends on the level of the discount factor for the other player. The higher the discount factor, the more impatient the player is, and the more willing he is to accept less in order to ratify the deal immediately. Breaking apart the allocation fraction (log [d.sub.i]/(log [d.sub.1] + log [d.sub.2]) illustrates the separate effects of time preference and risk aversion.
A player's time preference reflects how important it is to him that a deal be completed today rather than in the future. This is simply a present value discount rate, where higher rates mean the value of the item (in this case the development deal) declines quickly. The impact of time preference on the negotiated outcome can be seen most clearly if the value of the other factor in the allocation fraction (risk aversion) is set to a constant value. Using zero for that constant simplifies the algebra by letting the factor drop out of the formula, but comparable results would be obtained with any constant.
Setting risk aversion ([g.sub.i]) equal to zero, [d.sub.i] = 1/(1 + [r.sub.i]). As [r.sub.i] (the discount rate) increases, [d.sub.i] decreases. Because [d.sub.i] is always less than one, its logged value is always negative, and as [d.sub.i] decreases the log of [d.sub.i] takes on a larger negative value. Negative values in both the numerator and denominator of the allocation fraction cancel, so as log [d.sub.i] becomes a larger negative number the allocation fraction as a whole increases in value. The higher allocation fraction means the other player receives a larger percentage of the remaining good. The player with the higher discount rate does worse in the negotiations. Intuitively, whichever player sees the value of the deal dropping faster if it is delayed would be more eager to get a deal now, even if that means taking a smaller portion of the value.
Risk aversion measures the player's assessment that if the deal is not concluded immediately the negotiations might end for some reason, usually that the other player changes his mind about the investment, and the opportunity to make any deal disappears. Very risk averse players place a high probability on not seeing another round of negotiations. At the extreme, they...
|
