The retirement-consumption puzzle: a marital bargaining approach

https://doi.org/10.1016/S0047-2727(01)00169-4Get rights and content

Abstract

Evidence from several countries reveals a substantial drop in household consumption around the age of retirement that is difficult to explain with life-cycle models. Using food consumption data from more than 550 households from the Panel Study of Income Dynamics for the years 1979–1986 and 1989–1992, we find that married couple households decrease their expenditures on both food consumed at home and away from home by about 9% following the retirement of the male household head. No significant decrease in consumption is found for single households, either in a sample of males or a pooled sample of single males and females. These results are consistent with a model of marital bargaining in which wives prefer to save more than their husbands do to support an expected longer retirement period, and relative control over household decisions is affected by control over market income. The pattern of the consumption decline, which is increasing in the age gap between husband and wife, lends further support to this interpretation.

Introduction

Households appear to reduce their consumption expenditures substantially around the age of retirement. This pattern has been documented for the US by Hamermesh, 1984, Mariger, 1987, and Bernheim et al. (2001), for Canada by Robb and Burbridge (1989), and for the UK by Banks et al. (1998). The consumption decline appears to be fairly widespread across consumption categories, rather than concentrated on work-related expenses, and to take the form of a discrete drop at the year of retirement.

This behavior is puzzling, since life-cycle consumption models predict that households will want to smooth consumption (or rather, the marginal utility of consumption) when they experience a predictable drop in income, as at retirement. After examining alternative explanations that are consistent with forward-looking life-cycle behavior, other researchers have attributed this consumption drop to myopic behavior or to the systematic arrival of discouraging information at retirement. Understanding the cause of the consumption drop at retirement is important both to researchers who are trying to understand how individuals make complex decisions when the future is uncertain and to policy makers who are concerned about the adequacy of retirement savings by the baby boom generation.

In this paper, we explore an empirical hypothesis suggested by a non-unitary model of household behavior. If married couple households make decisions collectively (for example, by cooperative bargaining), and their ability to make binding agreements into the future is limited, then current consumption and savings decisions may be affected by each spouse’s current control over resources. Most wives expect to live several years longer than their husbands, and therefore prefer, in the absence of perfect altruism, to consume less as the couple ages than husbands do. If the husband’s bargaining power depends upon his current income or employment status, retirement from a career job will cause a relative deterioration in his influence on household decisions and a decline in the couple’s consumption spending.

This story generates a testable hypothesis: we should see a consumption drop at retirement for married couple households, but not for single households who, though they also experience a drop in income, can be expected to act in a way that is consistent with life-cycle utility maximization by the unitary consumer. We also expect the consumption drop to be more pronounced for couples with more divergent interests — i.e. for couples in which the husband is substantially older than his wife. We use food consumption data from the Panel Study of Income Dynamics for the years 1979–1986 and 1989–1992 to test these hypotheses and find that expenditures drop at retirement by 8 to 10% for married couples, but do not decrease significantly for single households. The magnitude of the consumption drop is increasing in the relative age of the husband. These results are robust with respect to alternative specifications of the consumption equation and definitions of retirement, and lend some support to a collective rather than unitary approach to the decisions of older couples.

Section snippets

Literature review

Household income falls substantially with retirement, and consumption expenditures fall as well. Yet standard economic models suggest that consumption should be smoothed over periods of predictably high and low income, and the permanent loss of income due to retirement is, for most, quite predictable as to both timing and magnitude. More formally, it is the marginal utility of consumption that should be held constant over the life-cycle, and changes coincident with retirement in family size,

Theory

In a collective model of household behavior, husbands and wives make joint decisions while attempting to maximize individual utility functions. Chiappori, 1988, Chiappori, 1992 analyzes a general framework in which household allocations are assumed to be efficient, and this implies the existence of a ‘sharing rule’ that divides total household resources among individual members. The sharing rule itself is not determined within the model, but it is described as a function of individual incomes.

Empirical model

Suppose now that couples differ only in their income. We can write the right hand side of Eq. (3) for the ith couple asαi=ln T1T1+T221θT3T21μ21θ−1T3T221θT31θYi.If we defineγ=ln T21μ21θ−1T3T221θT31θthen we have the following fixed effects equation for married couples:ln Citi+γ×AFTERitwhere t indexes time, ln Cit equals either (i) the log of total food expenditure; (ii) the log of expenditure on food at home; or (iii) the log of expenditure on food away from home, AFTERit=1 in all

Data

We examine the relationship between consumption behavior and retirement using food expenditure data reported by the representative SRC sample of the Panel Study of Income Dynamics (PSID). Our sub-samples include unmarried individuals who are between the age of 45 and 70 on July 1st, 1979, and matched husband/wife pairs in which at least one member is in this age range. The sample period for the analysis is 1979–1986 and 1989–1992, and is determined by the availability of consumption data for

Results

We first estimate Eq. (5) for the log of total food expenditures. Table 2 reports γ, the coefficient on AFTERit, for three alternative specifications. The control variables in all specifications include age dummy variables for the household head and household size dummy variables.15

Alternative specifications

If increases in leisure reduce the marginal utility of consumption, then retirement may cause a reduction in food consumption through a simple substitution mechanism. Other researchers have used non-retirement reductions in hours worked to investigate this mechanism and have argued that it cannot explain the entire consumption drop. In this paper, we are concerned with the differential responses of married couples and single households, and so must consider whether leisure–food substitution

Conclusion

In this paper, we re-examine the consumption decline at retirement using food expenditure data from the Panel Study of Income Dynamics. We find that the discrete drop in consumption at the retirement of the household head noted by other researchers is restricted to married couple households, and to the retirement of the husband. Single households exhibit either no significant change in their food expenditures at retirement, or a significant increase. These results are robust to alternative

Acknowledgements

We are indebted to Jonathan Skinner, Dan Hamermesh, Jan Ondrich, seminar participants at the University of British Columbia, Washington University, the University of Washington, and RAND, and two anonymous referees for helpful comments, but remain responsible for all errors and omissions. We gratefully acknowledge financial support from the National Institute on Aging (Lundberg and Startz) and the National Institute on Child Health and Human Development (Stillman).

References (22)

  • G.A. Akerlof

    Procrastination and obedience

    American Economic Review

    (1991)
  • O.P. Attanasio et al.

    Is consumption growth consistent with intertemporal optimization? Evidence from the Consumer Expenditure Survey

    Journal of Political Economy

    (1995)
  • S. Aura

    Does the balance of power within a family matter? The case of the retirement equity act

    (2000)
  • J. Banks et al.

    Is there a retirement savings puzzle?

    American Economic Review

    (1998)
  • B.D. Bernheim et al.

    What accounts for the variation in retirement wealth among households?

    American Economic Review

    (2001)
  • M. Browning

    The saving behavior of a two-person household

    Scandinavian Journal of Economics

    (2000)
  • M. Browning et al.

    The life cycle model of consumption and saving

    Journal of Economic Perspectives

    (2001)
  • P.-A. Chiappori

    Rational household labor supply

    Econometrica

    (1988)
  • P.-A. Chiappori

    Collective labor supply and welfare

    Journal of Political Economy

    (1992)
  • P. Diamond et al.

    Quasi-hyperbolic discounting and retirement

    (2000)
  • R. Euwals et al.

    The saving behaviour of two person households: evidence from Dutch panel data

    (2000)
  • Cited by (0)

    View full text