by Shannon Sietz and Arthur Lewbel

This application proposes a new methodology to study how expenditures in elderly households translate into standards of living for elderly individuals and how standards of living change with retirement, illness, and widowhood. Our proposed research addresses two related questions: 1. How do retirement-related changes in time allocations and expenditures affect individual consumption and wellbeing? Declines in consumption expenditures observed at retirement may be offset by increases in leisure or a shift from market produced goods to home produced goods. For example, retirees may cook more and eat in restaurants less frequently. Clearly, such a change will result in a reduction in expenditures. But, does it also imply a fall in welfare? We will study expenditure and time use data from a supplement to the Health and Retirement Survey (HRS), the Consumption and Activities Mail Survey (CAMS), to estimate how expenditures and time translate into the final goods consumed by households.

2. How can household level data give us information about individual living standards? After analyzing the way final goods are produced within the household, we propose new methods for inferring individual living standards, which are unobserved, from observed household expenditure and time use data. Our methods are able to separate out three factors that determine welfare: (a) the preferences of each household member, (b) the extent to which final goods, including leisure activities, are shared by husbands and wives, and (c) how much consumption and leisure is allocated to the wife versus the husband. The proposed methodology allows us to determine how each component changes with retirement, health status and widowhood. Broader impact. Our proposed research can provide quantitative answers to questions such as: How costly is it for widows to maintain the standard of living they had when their spouses were alive? How much do retired couples save by producing goods at home? Many policy and welfare calculations depend on answering these questions correctly. Our proposed research can ultimately be used in constructing poverty lines, for determining appropriate levels of social and private insurance, for indexing welfare, pension, or social security payments, and for calculating appropriate measures of welfare inequality in the population.