Macroeconomic Implications of Cross-Sectional Variation

National Bureau Of Economic Research Inc, Cambridge MA

Investigators

Abstract

This project's agenda is organized around two main ideas. First, that the risks associated with non-traded assets have an important impact on individual choices and, therefore on how individuals value the risks associated with traded assets. Second, that these non-traded risks have an important life-cycle component to them. In most of the investigators' work-to-date the 'non-traded asset' has been human capital: the value of all future labor income that an individual will receive. Much of the ongoing work involves expanding what we mean by non-traded assets to include housing and consumer durables. Human capital is non-tradable in the sense that it is difficult for individuals to borrow against future wage income and, just as importantly, to insure them against the uncertainty of what this income will be. Human capital has a fundamental life cycle component in that young people have more of it than old people and, therefore, face more of the risks associated with it. The investigators' work has shown that these factors can be important for many outstanding issues in economics. For instance, their work on asset pricing models shows that the life-cycle pattern in human-capital risk can inhibit the intergenerational sharing of aggregate risks and thereby helps to resolve the equity premium puzzle. Their work on consumption inequality exploits this life-cycle pattern to deduce strong restrictions on how the cross-sectional variance in income, consumption, and labor supply are related to one another. This proposal describes new work along similar lines, including the incorporation of cohort-specific risks and the intergenerational sharing of these risks, the associated role played by social security, and the impact of demographic shocks such as the Baby Boom. Like human capital, most non-human capital held by individuals is either non-tradable or is difficult to trade without incurring substantial costs. Housing, for instance, accounts for roughly one fourth of aggregate wealth and a much larger fraction of privately-held wealth. The costs of adjusting one's housing stock are large. Also like human capital, housing capital also has an important life-cycle component: young households hold much more of their wealth as real estate than older households. Unlike human capital the risks associated with housing are much more tangible and easily measured. The investigators develop a model, which incorporates housing into the existing framework, and they explore the implications. Where the previous work has emphasized idiosyncratic shocks, they model shocks to housing wealth as aggregate in nature, but non-uniformly distributed across the population. They argue that this, in conjunction with life-cycle issues, has important implications for asset pricing, portfolio choice, and various policy questions. Along similar lines, we discuss the incorporation of durable goods (more broadly defined than housing), and present some preliminary empirical evidence from the Consumer Expenditure Survey. An important part of this project has been, and will continue to be, the incorporation of large- scale panel datasets on labor market activity, consumption and wealth. The distinguishing feature of this work is restrictions -motivated by our life-cycle models - that are related to age. They show, for instance, that one can uncover much about how idiosyncratic and aggregate shocks interact by conditioning on the particular macroeconomic history experienced by each age-cohort in any panel dataset. They also show that age-dependent changes in the cross-sectional variance of consumption can reveal a great deal about the stochastic process that governs labor earnings.