by R. Alessi, A. Brugiavini and G. Weber in NBER International Seminar on Macroeconomics, 2004.
The Alessi, Brugiavini and Weber (ABW) paper provides an illuminating introduction to the consequences of income shares for savings decisions within composite households. The increasing prevalence of such households and the importance of savings decisions to broader macroeconomic questions justify a much larger literature on the topic than is currently available. The ABW paper begins the process of understanding the myriad issues that these trends raise. As it happens, the ABW paper also succeeds on a different, more personal level. Anyone with significant experience as a child or a parent will find several opportunities to reflect on their own experiences with the ties that bind.
The relevance of such composite households extends beyond the stereotypical mammoni (“mama’s boys”) of Italy. Young women living at home in Japan have been christened parasaitu shinguru (“parasite singles”) and “boomerang” children are increasingly common in the U.K.. These populations have grown sufficiently such that marketers target this new customer segment, given their high levels of disposable income, and sociologists have begun to debate the consequences of the blurred distinction between adolescence and adulthood. Of course, such composite households are just one example of the rich variety of households (joint households, single parent families, childless families) that are increasingly relevant and that the profession often abstracts from. ABW begin the process of trying to understand how the dynamics of composite households can influence savings decisions. While ABW raise several interesting questions (Do composite households save more and, if so, why? How do housing market characteristics give rise to cohabitation decisions?), the question that their paper addresses best is somewhat more narrow – should income shares matter for savings decisions of composite households and, if so, how?
The model has a fairly straightforward intuition with a few critical ingredients. Children face a transaction cost upon departure from the family, children and parents have conflicting preferences over their departure with parents favoring cohabitation, some fraction of consumption is public within the composite household and there is neither uncertainty nor multiple periods. The results depend on whether the children choose to stay or leave and the solution method. For the case of children who choose to stay, a Nash solution provides, unsurprisingly, a special case of the predictions of the younger spouse model presented in Browning (1995) – the distinctive horizons of parents (older husbands) and children (younger wives) create distinct preferences over savings and overall saving decisions depend, unlike the typical unitary family model, on income shares of children as a result. Specifically, higher income shares for children who stay leads to reduced savings. The Nash solution for children who choose to leave provides for a non-Pareto outcome so ABW emphasize the cooperative solution. This cooperative solution reflects the nature of the pareto weights assumed in the model. The parent’s pareto weight is modeled as a decreasing function of the child’s income share so higher income shares for leavers are associated with higher savings.
The predictions emerging from the model – that higher income shares for stayers lead to reduced savings and higher income shares for leavers lead to higher savings – are then tested with data from Italy and the Netherlands. The Italian data indicates that higher child income shares are associated with higher savings. This result is difficult to reconcile with the unitary family model but, while intriguing, cannot be used to affirm the ABW model given the inability to identify stayers and leavers in the Italian data. Said another way, coefficients of either sign could be regarded as bolstering the ABW model in this data. The Dutch data is more helpful in this regard as the panel-nature of the data allows identification of the cohabitation decisions of children. Here, ABW find not only that income shares matter for household savings decisions but that they matter in the way their model predicts – higher income shares are associated with lower (higher) savings for stayers (leavers) – though it would be useful to know if the coefficients are different in a statistically significant or if a pooled setting would provide statistically significant results. Nonetheless, the paper successfully marries together an intuitive model and interesting data with fruitful results.
Several aspects of the model and empirical results deserve comment. First, bribes by parents to children often take the form of private consumption (eg. an automobile) and this would clearly have distinctive implications for savings decisions. While there is little doubt that public consumption is important in this setting, bribes of private consumption could reverse the implications of the model. While the model incorporates a variety of interesting features (eg. differing preferences for cohabitation, transaction costs upon exit), it is difficult to view the empirical tests as confirming the model without explicit consideration of alternative hypotheses. For example, could differential riskiness of labor income across generations (as in Becker et al. (2002)) explain these empirical patterns rather than differing preferences over cohabitation? Indeed, are simply distinctive housing transaction costs for the generations (as in Guiso and Jappelli (2002)) sufficient to generate these results? In other words, it would be useful to consider these readily available alternative in order to bolster ABW’s interpretation of the results as either evidence of exit costs or conflicting preferences over cohabitation. Finally, this setting seems particularly ripe for considering mainstays of the savings literature – the effects of pension wealth and inter vivo transfers – given the magnitude of such wealth and the intergenerational dynamics of these decisions.
The crux of the matter in the ABW paper, of course, is the conflicting preferences of children and parents. ABW rely heavily, though not exclusively, on the intuition that children seek independence and parents seek dependence. It seems equally, if not more, plausible that parents are seeking independence and children are enjoying dependence. Survey evidence is hardly the last word on this given the ambivalence prompted in parents faced with the departure of their children. Even in the Italian setting it is not clear who is enjoying cohabitation more. Consider the case of Giuseppe Andreoloni – a noted Neapolitan doctor and legislator. Subsequent to his divorce, he was compelled to pay monthly support payments (€800) to his thirty year old son (already living off a large trust fund) as the court found that "You cannot blame a young person, particularly from a well-off family, who refuses a job that does not fit his aspirations. The parents have to pay for their upkeep." Such decisions suggest that composite households might not reflect the romantic notion of a daring child and doting parents but rather a dilatory child and dutybound parents. Further research might usefully devise tests of which mapping of preferences is borne out by the data rather than assuming the source of this conflict.
Identification of these effects, ultimately, will hinge on exogenous sources of variation in key attributes of the model. While beyond the scope of the ABW paper, such shocks might include labor responses to changes in retirement ages, as in Manacorda and Moretti (2003), regional variation in financial development (and presumably varying transaction costs to exit) as in Guiso, Sapienza and Zingales (2004), or changed (if they ever come) educational fee structures. Hopefully, future generations within this stream of scholarship will thank ABW for the foundation they have laid but also begin to stray by undertaking such econometric investigations.
Becker, Sascha O., Samuel Bentolila, Ana Fernandes and Andrea Ichino (2002). "Job insecurity and children’s emancipation: The Italian puzzle," CESIfo Working Paper.
Browning, Martin (1995). "The saving behavior of a two-person household,"Scandinavian Journal of Economics, 102:235-251.
Guiso, Luigi and Tullio Jappelli (2002). "Private transfers, borrowing constraints and the timing of homeownership," Journal of Money, Credit and Banking, 34:315-339
Guiso, Luigi, Paola Sapienza, and Luigi Zingales (2004). “The Role of Social Capital in Financial Development,” American Economic Review, forthcoming.
Manacorda, Marco and Enrico Moretti (2003). "Why do most Italian youths live with their parents? Intergenerational transfers and household structure", UCLA Working Paper.