First Symposium on Policy Research

1.10

Indentifying a desirable air quality for Atlanta in 2050

Ed Weber, Georgia Institute of Technology, Atlanta, GA; and D. Noonan

This paper presents preliminary results from a study to identify an ideal air quality for the Atlanta metropolitan area in the year 2050. Identifying an ideal spatial distribution of ambient concentrations of pollutants requires establishing the criteria for optimal air quality and applying that to the Atlanta context. Even a simple goal of maximizing net benefits encounters challenges in empirically estimating benefits and costs of air quality changes robust to changes in behavior and adaptation of urban structure possible in the long term. We propose alternative empirical methods to account for these possible changes and estimate ideal air quality measures for Atlanta under various assumptions. We report results that can give guidance to policymakers and urban planners in moving toward such a desirable goal.

Importantly, we seek to ascertain an ideal air quality, rather than the expected air quality. We establish "ideal" air quality as what level of air quality is optimal or most-preferred given our constraints. A variety of methods to determine future air quality desires within the metropolitan Atlanta region are outlined. This optimal level of air quality might be identified, on a basic level, as the intersection between the overall supply of the good (air quality) and its demand. We can define the demand for air quality as the amount of resources that the populace would be willing to trade for differing levels of air quality. A necessary condition for an economically efficient equilibrium is where air quality is supplied at a marginal cost to society that intersects with demand (either at the individual or community level). The intersection point between supply and demand establishes the optimal quantity of the good.

In this simple model the efficient equilibrium possesses several important characteristics. For our purposes, the most important one is that it maximizes net benefits. When marginal costs (MC) equal marginal benefits (MB) and the supply and demand curves intersect, net social benefits are maximized. Obviously, any deviation from this equilibrium must leave someone worse off. This theoretical characterization of an ideal or efficient air quality is assessed in light of the Pareto and the Hicks-Kaldor criteria. Yet devising an institution or policy to achieve the optimum is beyond the scope of this paper. We just seek to identify the optimal air quality levels.

To empirically apply this theoretical model of optimal air quality requires estimates of values, such as social benefits and costs. Partial equilibrium (PE) models are often used to identify the value of environmental quality. PE models analyze changes in prices associated with environmental amenities. For most private goods like TVs or apples, the market price represents both the marginal benefit and the marginal cost of that good. Its marginal value is readily apparent to the researcher. Yet environmental amenities like air quality generally have no price in a market. Our ability to go down to the store and buy clean air for our home or neighborhood is limited. PE models have been developed to help assess the value of these nonmarket goods. One common method is to examine the values placed on complementary goods that allow access to enjoying an environmental amenity. For example, we can compare the price of housing across different areas with differing levels of air quality.

Initially, we report the results of a technique known as the "benefit transfer method," which exploits this variation in valuation estimates in the literature (on air quality benefits) to estimate the values for our particular study site. In this approach, we make use of the relevant prior research to produce air quality valuation estimates, which we will then apply to our context. The variation or inconsistency in the air quality valuation literature is not seen as a weakness or limitation in this approach – instead it is a crucial strength. By understanding how benefit estimates vary by context, we are able to predict benefits for our particular context.

A more comprehensive approach to valuation of community amenities is the use of a general equilibrium (GE) model. This differs significantly in complexity from a partial equilibrium (PE) model. PE models can account for the costs and benefits directly arising from air quality changes. GE models capture the propagation of improving environmental amenities by allowing relocation and other responses to air quality changes, making it a particularly useful method for this analysis. For the short term, PE models may serve to capture with a reasonable degree of accuracy the effects of regulation. However, over the long term like 50 years, the possibility of adjustments affecting the overall equilibrium makes GE models much more robust. For air quality changes, where the main source of our information about the value of air quality is derived from households paying more to live in areas with cleaner air, a GE model allows us to estimate benefits that are robust to changes in the housing market.

This paper implements a computable GE model known as a locational equilibrium model to estimate the value of air quality to individuals within the Atlanta Metropolitan area over a long time horizon. Based on work by Smith et al. (2004) and Sieg et al. (2004), we develop a locational equilibrium model that allows for the simulated "sorting" of individuals within differing jurisdiction in the Atlanta area. The consumer choice identified by that simulated sorting allows for a valuation of air quality among the populace. Combined with general estimates of future cost of pollution abatement, these revealed preferences allow for an estimation of ideal air quality. The results presented will offer a spatial distribution of ambient concentrations consistent with an optimal air quality mix.

Session 1, Policy Research in the Earth System Sciences
Wednesday, 1 February 2006, 8:30 AM-5:30 PM, A307

Previous paper  Next paper

Browse or search entire meeting

AMS Home Page