wordpress-themes.org wordpress themes wordpress themes

INTERTEMPORAL CHOICE AND THE CROSS-SECTIONAL VARIANCE OF MARGINAL UTILITY: Regression results 3

As we have repeatedly stressed, our test does not assume certainty equivalence. Rather, we are able to confront with the data a flexible representation of preferences, which allows for departures from certainty equivalence, dependence on family size, labor supply and other demographic variables. This increased generality, however, has a price: we loose the ability to relate the evolution of the cross-sectional variance of consumption to the evolution of the cross-sectional variance of earnings. This relation has proved to be particularly useful in welfare comparisons among households.

Blundell and Preston (1997) study under which conditions consumption is a better indicator of welfare inequality than income. They note that in the life-cycle model, cross-sectional comparisons of consumption inequality are informative about welfare inequality only on restrictive assumptions concerning the form of the utility function. They also show that under certainty equivalence the evolution of the variance of consumption should reflect permanent income innovations; whereas the variance of earnings should reflect permanent as well as transitory changes in uncertainty. Our study is complementary to theirs in that it shows that one of the most general implications of individuals’ intertemporal choices for the evolution of the cross sectional variance is borne out by the data, particularly in the case of the UK.

Dropping certainty equivalence, only simulation analysis can give an idea of the structural relation between the cross sectional variances of earnings and consumption (examples of such simulation models are in Hubbard, Skinner and Zeldes, 1995). In principle, these models could simulate the consumption behavior of a generation and compute the evolution of the cross sectional variance of marginal utility. By changing the stochastic properties of the income generating process it would then be possible to study the relation between income and consumption inequality. We regard this as an interesting topic for future research.

This post was written by , posted on June 17, 2014 Tuesday at 2:57 pm