Saving and Cohabitation: The Economic Consequences of Living with One's Parents in Italy and the Netherlands
Social Science Research Network
Introduction Much economic theory on saving takes the nuclear household as the benchmark for its analysis. The standard assumption is that children leave home as soon as they are of age, and that they become independent consumption units as soon as they do so. Yet, we have evidence of wide-spread cohabitation of at least two generations in some European countries, as well as in some Far Eastern countries. In these European countries the common pattern is not so much for the elderly to live with
... their children, rather for grown children to leave home well after they become of age. We define those households where grown children live with their parents as "composite households." We provide evidence from Italian survey data that composite households differ in their saving behavior: for all ages over 50, composite households' saving rates are higher than nuclear households' saving rates. One possible implication is that countries characterized by higher cohabitation have higher aggregate household saving. The economic issue we investigate is the link between cohabitation and saving. But this calls into question the reasons behind cohabitation, a topic that has been investigated in a number of papers, that point to imperfections in either the labor market or in the credit market (possibly in conjunction with the housing market). For instance, Fogli (2004) and Becker et al. (2004) stress the importance of lack of job security for the young in delaying the time of independence, while Guiso and Jappelli (2002) stress that rent controls and severe imperfections in the mortgage market make it hard for young Italians to move out of the parental home. However, recent work by Manacorda and Moretti (2006) 414 Alessie, Brugiavini, & Weber suggests that parental preference for living with their children may play a major role in explaining cohabitation, to the point that parents would choose to work harder to offer their grown children a higher standard of living if they remain in the parental home. In this paper we show how differences in saving rates found in micro data can shed light on the presence and nature of these imperfections or differences in preferences across generations. In particular, we want to assess the role played by transaction costs in the housing markets. If such costs are particularly high for both trading down (by the parents) and buying or renting (by the children), and capital markets are imperfect, cohabitation may be the optimal way for the young to accumulate liquid assets necessary for the down payment and in general for the purchase of their home. In the case where parents and grown children live together, household decisions are unlikely to be taken in a unitary way. Even if father and mother behave as one person, and they have only one child, key household decisions are likely to be the result of some form of bargaining between parents and child. For this reason, we extend Browning's (2000) "younger spouse" model to cover the case of two generations: in his model husband and wife have different survival probabilities, and therefore disagree on how much to save. In our model, the child may choose to leave home in the second period or to stay with her parents, and will base her saving decision on the preferred outcome. In either case, we can expect the child's income share to have an impact on the household saving rate. This paper is organized as follows. The second section presents some basic facts on cohabitation in Europe, and on its consequences on saving rates. The third section presents a simple theoretical model of how composite households jointly decide how much to save and whether to continue cohabitation. The fourth section describes the two data sets used in this paper: the Italian Survey on Household Income and Wealth (SHIW) and the Dutch Socio-Economic Panel (SEP). The fifth section presents estimates for both Italy and the Netherlands, while the sixth section concludes the paper. Cohabitation across European Countries Economists often assume that adult children live on their own. This probably reflects the most common living arrangement that prevails in Saving and Cohabitation 415 some Western countries, such as Germany, the UK, or the U.S., where children tend to leave the parental home soon after they become of age or at least complete their education. And yet, there is ample evidence that this is by no means the rule even within Europe. Important studies by demographers have pointed out that the age of leaving home varies dramatically across European countries (Kiernan 1986 and 1999; Fernandez Cordon 1997). In a recent and well-documented study, Billari et al. (2001) estimate that for the ten-year cohort born around 1960, for instance, median ages of leaving home were 22.5 for men and 20.5 for women in the Netherlands (very close to the UK or West Germany), as low as 20.1 for men and 19.8 for women in Lithuania, but much higher in Spain, Poland and particularly Italy (26.7 for men and 23.6 for women in Italy). This variability across countries is not a recent phenomenon, and suggests that institutional or cultural differences may play a lasting role in explaining international differences. An interesting summary on cohabitation in Europe as recently as 1998 is presented in Figure 1 , that shows the proportion of households in the European Community Household Panel headed by someone aged 50 or more with at least one child aged 25 or more in residence. This proportion is highest in Portugal, followed by Italy, Ireland, and Spain, and lowest in Denmark, followed by the Netherlands and the UK.1 However, even though in the Netherlands cohabitation of children aged 25+ with their parents is quite rare, the cohabitation of 18-25 is much more common. On the basis of this evidence, it makes sense to study the way cohabitation and saving decisions are taken across different countries but not necessarily to focus on a specific age group: we shall construct our empirical exercise in such a way that the leaving home decision could be taken mostly by children aged 25+ (like in Italy) or 18-25 (like in the Netherlands). Computing saving rates in household level data is difficult. Saving can be defined as income minus expenditure, or as the change in wealth. The two definitions are not even conceptually the same, because income does not normally include capital gains (see Brugiavini and Weber 2003 for a discussion of this and many other issues). But empirically, they are likely to differ for measurement problems too: the flow definition requires finding a survey that contains high quality data on both income and expenditure, the stock-based measure Alessie, R., A. Lusardi, and T.