The social choice of alternative corrective devices

The social preference of corrective taxation to corrective subsidy or to a policy of no action at all is examined within the context of a simple general general equilibrium framework in which an external production diseconomy exists. Equilibrium positions are obtained for a number of different parameter sets and resource endowments. For each equilibrium position, different mechanisms are introduced as procedures for distributing revenues accruing to a central authority which administers the corrective tax, or for collecting revenues to finance a subsidy programme. The only income redistribution which occurs in this economy is through these distribution mechanisms. Any change in policy regarding the correction of allocation inefficiencies resulting from the existence of externalities in the economy will be reflected in changes in prices (of both final goods and resources) and the resulting income redistribution necessitated by the need for revenues by the central authority or the need to reduce tax collections from individuals (if a corrective tax yields revenues from an industry) to maintain a balanced budget constraint. In an economy in which no government goods are produced, a tax reduction implies transfers to individuals. Within this framework, the following results emerge. 1. If the distribution mechanism is regressive, proportional or progressive, the corrective tax will always be preferred to the corrective subsidy as a method for internalizing external costs. Only under extreme progressivity does this support become unanimous. In all other situations only the agents controlling the one particular resource which identifies the firm will oppose the choice of the tax. 2. The status quo equilibrium will also be preferred to the equilibrium resulting from the use of a corrective subsidy. Again this result occurs consistently, and with the same coalition of agents supporting the status quo as supported the corrective tax. 3. Only under the proportional mechanism is the corrective tax preferred consistently to a ‘do nothing’ policy. If the distribution mechanism is progressive or regressive, there exist situations in which the status quo equilibrium state will be preferred by a majority of individuals to the state resulting after corrective action is taken, and resources are efficiently allocated. As progression increases, the likelihood of the corrective tax finding support from a majority of the agents in this economy tends to fall. 4. While majorities may exist, they appear to be tenuous, at best, in many situations. The difference between utility levels is at times very small, and with the introduction of transaction costs associated with casting a vote, some individuals may find it less costly to abstain from voting. In cases such as these, a majority of those voting may oppose an alternative which would have been supported had transaction costs not been introduced. There is also the problem of whether or not individuals actually perceive small differences in utility from state to state. However, since the model has demand functions based upon individuals who maximize a particular utility function, it would be inconsistent to assume that as consumers they are aware of their utility when choosing between different commodity bundles, but when comparing alternate states, they suddenly have a perception problem. The assumption of threshold perceptions for utility differences would require the development of an entirely different demand function. This extension does not appear to be as profitable as the introduction of transaction costs. 5. The distribution mechanism and the progressivity of the mechanism are exogenously determined. The convention that voters choose from among competing corrective devices but do not choose the degree of progression or the specific form of the mechanism could be relaxed. The easiest extension of this work can include both the choice of the progressivity of the distribution mechanism and the choice of the corrective device. 6. A further extension of this work could include a role for the central authority within the equilibrium framework as a user of resources. Not only may voting carry a cost, but the implementation and monitoring of corrective policies require resources. This results in incentives in conflicting directions. The costs involved in administering a programme reduce the potential gains to consumers, while the expansion of an industry (the central authority) provides benefits to others (bureaucrats). To the extent that the resources used in the production of corrective policies gain, particular groups may find income increases which offset price increases and make a costly corrective scheme more desirable than no action at all. 7. Another valuable extension of this work would be to generalize the production and utility functions to examine the implications of the model with variable elasticities of substitution between the inputs used in the two industries. A further change within this framework would be to introduce an externality which results in increasing marginal damages, rather than the decreasing marginal damages exhibited in this model. The earlier conjecture in Mestelman (1982b), suggesting that corrective subsidies may be preferred to corrective taxes under some distribution mechanisms, has not been demonstrated with this model of an economy. However, the work of Buchanan and Tullock (1975) which shows that efficiency need not imply social preference, and that the effect of any policy on the agents who make up the society must be identified if the failure of positive prescriptions is to be understood, is confirmed by this work, within a general equilibrium framework. The results of the experiments reported here suggest that for policy changes which require that resources be either collected from consumers or returned to consumers in a fashion that does not distort relative prices (in this model the leisure-labour decision is not considered), a distribution mechanism which is progressive may impede the adoption of policies directed at correcting allocation inefficiencies attributed to externalities. The model from which these results follow is certainly not an accurate representation of any actual economy, yet the results do point to the relevance of the financing of government regulatory policies. The distribution of income, and the effect on this distribution by government policies not necessarily intended to redistribute income, may significantly effect the success of those policies. Copyright Martinus Nijhoff Publishers 1984