On the measurement of risk aversion from experimental data

Attitudes towards risk are measured for households in Northern Zambia using an experimental gambling approach with real payoffs that at maximum were equal to 30% of average total annual income per capita. The results of the experiment show decreasing absolute risk aversion and increasing partial risk aversion. Determinants of risk aversion are investigated using random effects interval regression model exploiting the panel data structure of the repeated experiments. Wealth indicator variables are found to be significant, and partial relative risk aversion decreases as wealth increases. Females are found to be more risk averse than males.     

History-dependent Quantity Regulation

We consider an industry that becomes quantity regulated if production is too large. The threshold level of production at which regulation is implemented is unknown to the firms in the industry, and future rights to produce in the regulated situation are made dependent on current production. Hence, a firm produces in order to earn profit in the current period, and to secure property rights in the future; this latter motive is a form of rent-seeking behavior, and will tend to raise current production even though the aim of regulation may be to reduce it. The beneficial effects of an increase in current production must be weighed against the raised probability that regulation will be implemented in order to determine equilibrium production. We characterize this equilibrium, and look at the comparative static effects of changing various policy variables. We consider how a regulator can achieve a current production target and at the same time pursue his own private goals. For example, a regulator who cares about his image or probability of re-election may not wish to pursue a strict regulation policy for the industry in question. We show that a production target can be achieved whilst following a seemingly relaxed regulation policy due to the history dependence of the allocation rule. This provides a new justification for the use of history-dependent allocations in quantity regulation; these are widespread in practice, even though they may lead to various undesirable phenomena.revised version received October 17, 2003     

A contingent grouping approach for stated preferences

This paper introduces the theoretical and statistical foundations of contingent grouping, a variant of attribute-based stated preference methods, and checks for its implementation difficulty. Respondents are asked to simply group a number of alternatives as better or worse than a baseline or the status quo situation. A Monte Carlo approach shows that it performs statistically better than contingent choice according to the MSE criterion, and worse than contingent ranking, while a case study suggests that contingent grouping is less difficult to answer than contingent ranking. Furthermore, contingent choice shows a greater consistency in preferences with contingent grouping than with contingent ranking.