Essays in households' earnings, portfolio choice and consumption

Macroeconomics relies on household data to calibrate and evaluate incomplete markets models. This dissertation examines a data set of Swedish households and individuals. The data set contains unique measures of skills and detailed information on asset holdings. In addition, it has a panel dimension. The dissertation exploits these features to improve our understanding of households earnings, portfolio choice and consumption expenditure.;Chapter 1 (with Erik Lindqvist) investigates the importance of cognitive skills and noncognitive skills for wages, earnings and unemployment. This study is the first to use a measure of noncognitive skill that is obtained from an interview conducted by a psychologist. We find that the effect of both kinds of skill varies substantially between workers with high and low productivity.;Chapter 2 studies the relationship between home ownership and stock market participation. In Sweden, as in the USA, households that own a home are about twice as likely to hold stocks or equity than households that rent a home. The goal of the chapter is to understand whether this gap in participation rates is a result of a positive effect of home ownership on stock market participation, or whether the gap is a result of other factors. The panel dimension of the data set is useful to separate the direct effect of home ownership on stock market participation from the effect of unobservable differences between renters and home owners. As an alternative way to isolate the effect of unobserved heterogeneity from the effect of home ownership, the data is compared to a portfolio choice model in which households are only ex-ante heterogeneous in terms of initial earnings and net worth. The main finding is that stock market participation is mainly determined by the net worth of the household whereas home ownership and unobserved heterogeneity matters less.;Finally, chapter 3 imputes consumption series of households between 2003 and 2007 from information on earnings, government transfers and taxes, asset holdings and asset returns. The final product is a data set with repeated household-level observations on earnings, asset holdings and consumption expenditure. This data set is potentially useful for many economic questions.