Traditional and Modern Pollution

Abstract

Pollution is a key concern in environmental economics, and economists have used their models to argue that this market failure can be overcome by harnessing market signals. The simple model of pollution control makes the argument clear, but in practice, developing countries use diverse instruments to control industrial pollution. In developing countries traditional pollution is a major cause of disease. We want to reduce pollution because we think it has harmful effects, but how do we establish our knowledge of its effects? Establishing the causality convincingly is difficult, but was done by John Snow in the middle of the nineteenth century in the case of cholera. The simple model of pollution control suggests that we consider the benefit of pollution control, but how do we establish them? For many, environmental economics is synonymous with monetary valuation. Monetary valuation is achieved by using a theoretical model along with an econometric one. In a developing country context, if we dig deeper, we may recognize the role of norms.

Appendix

In this appendix I compare a few features of the basic theoretical model used to value reductions in morbidity due to air pollution (Freeman 1993) with the theoretical model used by Das et al. (2009) (discussed in Extension 3).

In Freeman sickness S = S (C, a, b) where C is concentration, a is averting activity, and b is mitigating activities. On the other hand, in Das et al., sickness of an individual depends on total exposure to pollution of that individual (which can vary between persons in a household) E, consumption of cooked food (CF), doctor-visits (D), individual characteristics (Z_i), and household characteristics (Z_hh): S_i = S_i(E_i, CF_i, D_i; Z_i, Z_hh). Total exposure incorporates time spent in different micro-environments, but unlike in some western cities where people may avoid going outdoors on bad days, here the concentration indoors, particularly in the kitchen, may be high. Here the only mitigating activity is doctor visits.

A marked contrast between Freeman and Das et al. is the more detailed treatment of total exposure, E_i, by Das et al.:

E_i = t ⁱ_o C_o + t ⁱ_k C_k + t ⁱ_w C_w + t ⁱ_in C_in where d denotes time, i indexes individuals, o is for outdoor, k for the kitchen, w for work, in for indoor and C denotes concentration. So Co, i.e. outdoor concentration is not assumed to be the dose as in Freeman, but is only one component of total exposure.

Another difference in the two models is the utility functions: in Freeman, this is U(X, f, S) where X, f and s are consumption, leisure and sickness respectively. In Das et al., U = U(S^C, S^AM, S^AF, X^NF) where S is sickness and X denotes consumption. The key difference is that in Das et al. children (C), adult males (AM) and adult females (AF) are distinguished.

Finally, in Freeman, time spent working is a choice variable, but in Das et al. there is a choice between time spent in several activities and associated micro-environments, including, in the case of adult females, cooking and gathering biomass for fuel.