Consumption and labour supply : relaxing the intertemporal and intratemporal separability assumptions
In the first chapter of the thesis, the consumption insurance hypothesis is tested using the Euler equation framework and applying it on a utility function that is nonseparable in consumption and leisure. The aim is to investigate whether the restrictive way in which leisure enters the equation for the growth of consumption affects the acceptance or rejection of the theory. In general, research on risk sharing in consumption investigates a number of explicit and implicit insurance mechanisms under which the consumption growth rate between households would differ because of changes in the earnings capacity and income, whether anticipated or not. In the case of complete markets and identical tastes and isoelastic preferences, it is implied that there is no idiosyncratic variation in the growth of individual consumption. However, the results are not altogether conclusive. In the first chapter, a test of the consumption insurance hypothesis is conducted, controlling for nonseparability between male leisure and household consumption in the utility function. Because the proper modelling of labour supply is beyond the scope of this study, leisure is treated as a conditional variable in the equation for consumption. An Euler equation approach is adopted where the results from the Euler equation of consumption are used to construct the growth of the log of marginal utility to be finally be used to test the consumption insurance hypothesis. The results indicate towards a rejection of the consumption insurance hypothesis for the quarterly series. Intratemporal nonseparability between consumption and leisure is one issue, but not the only one. Other important issues that should be taken into account are the intratemporal nonseparability between durable and nondurable consumption and the intertemporal nonseparability implied by durability and habits. The second chapter investigates the effect of durability and habit formation on consumption in a dynamic almost ideal demand system, and shows that time effects are still significant when household heterogeneity is taken into account. The third chapter models commodity demand and leisure jointly, taking an unconditional approach, separating male and female leisure in a dynamic context. The results indicate that nonseparability between consumption and leisure cannot be rejected, and that durability and habit effects are once again significant not only for commodity demands but also for male leisure.