wordpress-themes.org wordpress themes wordpress themes

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

We consider two possible methods of constructing the cross-sectional variances. The first is to estimate the “calendar year” consumption figure for each household on the basis of the months available, with seasonal adjustment. By this method, most households would provide figures to be used in the computation of the variance in two different years, and this should be taken into account in the choice of instruments. The second procedure involves considering only one monthly observation per household; after seasonal adjustment one can compute the cross sectional variance for all the households in a given quarter or year.
We use the second procedure. Its drawback is that by limiting itself to one observation per household, it wastes useful information; on the other hand, it is very simple to apply and yields a sample whose structure is very similar to that of the FES. To minimize the effect of non-random attrition, the monthly observation we select is the first available. The procedure for seasonal adjustment is the same one used for the FES (detailed in the appendix).

The SHIW

The primary purpose of the Bank of Italy Survey of Household Income and Wealth (SHIW) is to collect detailed data on demographics and households’ income and wealth. It uses a representative sample of the Italian resident population; probability selection is enforced at every stage of sampling. Like the FES and the CEX, it samples the household, defined as all persons residing in the same dwelling and related by blood, marriage or adoption. The SHIW consumption data are collected by retrospective questions on broad categories of durable and non-durable consumption during the previous year. This implies that seasonal adjustment is not necessary. Given the different purpose of the survey, the detail and quality of the consumption data is not as good as in the FES or the CEX. Nonetheless, the SHIW consumption data match the trends in the national accounts data reasonably well. The sample size is slightly larger than that of the other two surveys (about 8,000 households).

Until 1984 the age of the household head is available only in wide bands precluding the construction of year-of-birth cohorts. In 1986 the data for expenditures on durable goods is not distinguishable from non-durable consumption. Since 1987 the survey was conducted every other year. For our purpose, therefore, the only usable surveys are the five surveys taken from 1987 to 1995.