Measuring the equity burden in public service provision: the case of New Jersey toll roads

Economic Papers (Economic Society of Australia), Dec, 2008 by Jonathan Peters, Cameron Gordon

1 Introduction

The basic research questions addressed in this paper are: (1) How can equity in public services be measured? and (2) How equitable is the provision of particular public services by particular agencies?

We begin to answer these questions by focusing on a case study: an analysis of existing tolls in the US State of New Jersey based upon electronic toll collection data. Through this case study some of the data and methodological issues surrounding the measurement of equity in public services, as well as some policy issues that arise, are illustrated.

Economic theory recognises two major dimensions of social welfare: efficiency and equity. Equity broadly refers to the distribution of resources across different groups and (more subjectively) whether that distribution matches some socially preferred ideal. Technical efficiency implies being on the boundary of a production possibilities frontier and delimits a technically feasible set of outcomes that can be expanded through exchange; it is a necessary but not sufficient condition for a Pareto optimum. If exchange occurs from a technically efficient point and through which a Pareto optimum is reached then allocative efficiency has been achieved. If both allocative efficiency and equity are achieved then a complete social optimum has been reached.

Both efficiency and equity are important when it comes to the provision of public services, such as police protection, garbage collection, schools and, the area examined in this article, transportation. There has been a great deal of study of the technical efficiency of various public services (Fried et al., 1993), though less so for allocative efficiency, which is harder to measure (Brueckner, 1982). There is relatively little analysis of whether the provision of existing services is equitable, or even of 'who gets what' (Litman, 1996). This statement does not apply, of course, to taxation, where there is a large body of work on 'burden' and 'impact' (Sorensen, 2004).

Technical and allocative efficiency is a prerequisite to achieving better equity. The more that there is to distribute generally, the more each individual can, potentially, receive. To measure technical efficiency requires input and output data only. But allocative efficiency requires additional data about prices and market conditions, which may not be readily available for public services. To measure equity requires knowledge of both sorts of efficiency and also information about which parties are providing the inputs and which are getting the outputs. This difficulty in measuring equity may explain why there has been relatively little actual measurement of it.

However, equity is often the critical element of many debates about how to allocate public resources. Moreover, while efficiency should clearly be sought before equity in a first-best world, in a second- or even third-best world, there may be cases where efficient sub-optima will in fact be socially inferior to less efficient but more equitable alternatives. Thus equity in public services is, in general, an area of analysis that deserves much more attention than it currently gets.

Roads are an interesting case in this regard. A road can be 'consumed' by many people at the same time and has a degree of 'publicness' to it. After a certain point, however, congestion sets in and motorists cannot travel as fast or as reliably as they did when there was little or no congestion. Before that point roads are similar to pure private good with complete rivalness and excludability in consumption. Once congestion occurs rivalness suffers as one person's consumption of the road degrades another's. Hence roads are a 'club good' (Buchanan, 1965).

Because of this, roads will generally have some sort of public policy dimension, even if, as is often the case, the road itself is privately owned and/or operated. Moreover, in cases where roads connect key points and there are few or no feasible alternative routes, there is a situation of natural monopoly. This potential for market failure is also a reason why equity can be a concern in the provision of roads, especially toll-roads.

There are two standard dimensions of equity: horizontal and vertical. Vertical equity is defined as fairness in the treatment of people in differing social position (usually different levels of income or consumption). Horizontal equity is defined as fairness in the treatment of people at the same levels.

A standard measure of equity (or more properly speaking relative equality) is the Lorenz Curve in which the proportion of population is matched against the proportion of a metric of some benefit or cost. A perfectly equal distribution of income across a population would produce a linear Lorenz curve in which the proportion of the population exactly matched the proportion of available income.

With transport facilities and services a metric of such equality would be access or mobility. In this case there might be a Lorenz Curve mapping total travel-time to and from work, or, for more detailed network analyses a network matrix showing relative differences in travel time between specific origins and destinations (Levinson, 2002). Other metrics, such as who bears external environmental costs are also possible, the latter being specially important in the environmental justice literature (Bullard and Johnson, 1997).