City Limits - Ch. 3

This is an economic analysis of public policies adopted by city governments. Peterson first divides potential public policies governments can undertake into three types: developmental, redistributive and allocational. Here are briefly the characteristics of each.

Developmental - These policies enhance economic positions of a community in its competition with others. These policies enhance economic development of a community because they benefit those who have borne the cost directly. In more technical terms, these policies have benefit/tax ratio of more than 1, so that marginal benefits exceed the marginal cost to the average taxpayer.

Redistributive - In direct contrast, these policies hurt economic conditions of a community. These policies are characterized by the fact that recipients of the service are not the ones who have paid for it - with the result that those who are economically hurt by the policies must pay for these policies, usually through taxes.

Allocational - Neither developmental nor redistributive; these policies have neither much of a positive nor negative effect on the local economy. Police and fire personnel, as well as garbage and refuse collection, are examples of this type of policy. They do not quite pay for themselves, yet neither are they for the benefits of the needy who cannot pay for services they are using.
 

Now, with this categorization of types of policies kept in mind, the question for the author then is identifying variables that can account for governmental expenditure patterns for those three different types of policies. For this purpose, first, three factors - fiscal capacity, supply, and demand - are identified as crucially important economic variables affecting the supply of these three types of policies. For fiscal capacity, it is important to note that most of inter-city variation in this factor is cannot be attributed to difference in taxation rates, as drastically high taxation rates compared to competing regions drive away the residents. For the demand, it should be noted that Peterson is talking about demand strictly in the economics sense - that is, only those "demands" sought by ones who pay for them are considered as demands. Thus, redistributive policies do not answer demands, but rather what they are providing for will be regarded as "needs" by Peterson. After these three determining factors have been identified, Peterson proposes a hypothesized relationship between these determinants of expenditure and three types of policy. Briefly, it goes as follows. For the redistributive policy, since they do not respond to any "demands", demands are not the factor here. Instead, it is hypothesized that it is correlated highly with the fiscal capacity of the community - that is, these policies will be delivered only insofar as the city is rich enough to provide for them. For the developmental policies, since they respond most clearly to demands of those who pay for them and benefit from them, it is hypothesized that these policies are highly correlated with demand. On the other hand, since presumably these policies are funded by money coming from the potential beneficiaries, low correlation between fiscal capacity and these policies are proposed. For the allocational policy, moderate correlation for all three determinant factors of fiscal capacity, demand and supply are proposed, as this type of policy is the intermediate of the two. For the summary of this relationship, see pp. 48-9.
 

Now that the theoretical framework has been propounded, all that the task remains for Peterson is the empirical testing of this hypothesized relationship. And in general, the results of the regression analysis is very much consistent with what Peterson has proposed - albeit the operationalization of variables may seem somewhat crude. For the dependent variable side, redistributive policies are measured by old age assistance, aid to dependent children, unemployment benefits, welfare expenditures, and health and hospitals expenditure. Education is also categorized as redistributive, since in general it imposes so heavy a burden on those who are not necessarily the beneficiary. Police and Fire department expenditure are used as proxy for the allocational policies, while the highway construction is used to measure the developmental policies. On the independent variable side, income and property values are used as the index for fiscal capacity. For the demand - supply consideration, density, proportion of metropolitan population, proportion of non-agricultural employment, and average wage has been used as four indices for these factors. % poor and % black had been used to measure the "need", but not demand.

And the results on the whole seem to support Peterson's initial hypothesized relationship. Redistributive policies are most highly correlated with fiscal capacity, and ironically, they are negatively correlated with needs. This implies that governments of a community not burdened with the presence of so many of the needy and are rich enough in the first place are ones who can supply the greatest amount of policies aimed at the needy. Highway construction is most strongly correlated with the demand, on the other hand. Though, surprisingly to the author, it is also quite highly correlated with fiscal capacity. Presumably, even developmental policies need initial backing of tax-based governmental finance.