آیا تعادل قدرت در خانواده مورد توجه است؟ در مورد قانون بازنشستگی حقوق صاحبان سهام

This paper studies within-family decision making regarding investment in income protection for surviving spouses using a simple and tractable Nash-bargaining model. A change in US pension law (the Retirement Equity Act of 1984) is used as an instrument to derive predictions from the bargaining model about the household demand for survivor annuities and life insurance and to contrast these with the predictions of the classical single-utility-function model of the household. In the empirical part of the paper, the predictions of the classical model are rejected in favor of the predictions of the Nash-bargaining model.

Most economic theory assumes that household behavior is determined by a rational agent maximizing a single household utility function. This means that the behavior of multiperson household can be described as decisions made by a (possibly benevolent) dictator within a household. For most purposes, this assumption has been proven to be powerful way of describing actual behavior, but in certain applications it is important to consider explicitly the multiperson nature of many households. This paper considers one such application: the analysis of a government policy intended to redistribute resources within a family. The specific issue analyzed in this paper is a married couple's choice of the amount of survivor protection to be provided to a surviving spouse after the death of her partner.1 The potential conflict of interest between spouses rises from the fact that providing protection to a surviving spouse is costly (e.g. life insurance is not free). This means that the more survivor protection is provided, the less resources the household has available in other states of the world. Thus there is the potential for conflicting interests between spouses. The application studied in this paper is the spousal signature requirement of the Retirement Equity Act (REA) of 1984. This requirement mandated that a married pension plan participant, when retiring, must choose his pension payment in a form of a joint-and-1/2 survivor annuity unless his spouse signs a notarized consent form waiving her right to this survivor protection.2 The mandate affected only pension plan participants who started receiving their pensions after January 1, 1985. In the theoretical part of the paper, a Nash-bargaining model of family decision making is used to analyze the specific effects of this law change for the selection of survivor annuities and life insurance holdings. The law change is interpreted as having changed spouses' relative outside options. The model predicts that the law change would increase the selection of the survivor annuities and increase life insurance holdings for most households. These predictions of the Nash-bargaining model are contrasted with the stark prediction of the classical model that the law would have had no effect since the household budget set is unchanged. Thus this exogenous law change provides a well-identified empirical strategy for testing the predictions of the bargaining model against the predictions of the classical model.3 In the empirical part of the paper, several cross-sectional datasets are used to study these predictions. The effect on the survivor annuity selection is studied using the Current Population Survey (CPS) December 1989 Pension Benefit Survey and a combination of the Health and Retirement Survey (HRS) and the Assets and Health Dynamics Among the Oldest Old (AHEAD).4 These results show that the law change increased the selection of survivor annuities by approximately 7 percentage points (a 10% increase). Results from HRS-AHEAD data indicate that the median increase in life insurance holdings for affected households was approximately US$5000. This corresponds to approximately 25% of median life insurance holdings of the affected group. These joint annuitization and life insurance findings support the Nash-bargaining theory over the classical single-utility maximization model.

The US spousal consent signature requirement studied here is not the only one of its kind that has been proposed and or implemented. Similar requirements have been implemented in Canada and have been proposed in the UK. In Canada, several provinces and the federal government (which oversees some private sector pension plans) passed a requirement very similar to REA to apply to private sector pension plans between 1987 and 1992 (Federal and Provincial Laws of Canada).24 In the UK, the direction of reform is more uncertain. Currently, occupational defined benefit pension plans provide survivor benefits as a condition of contracting-out of the State Earnings Related Pension Scheme. An independent government commissioned report called for elimination of this mandate (“the Pickering Report”; Pickering, 2002). On the other hand, the Equal Opportunities Commission has called for a mandate similar to REA to apply even to annuities purchased from other types of pension plans (Equal Opportunities Commission, 2004). Under current law, the primary annuitant can freely choose a single life annuity from these plans; only 19% married individuals choose survivor annuities (Association of British Insurers, 2002). In the US, extensions of REA to require spousal consent signature on all defined contribution pension plan payments have also been suggested (United States Senate, 2004), but no legislation has been passed. To summarize the paper, a tractable Nash-bargaining model for household decision making over survivor protection was presented. The model made two specific predictions regarding the effects of the Retirement Equity Act on the choices that households make: the selection of joint annuities would increase, and, for most households, life insurance holdings would increase. These predictions are in stark contrast with the predictions of the classical single-utility-function maximization model. Using several microdata sources, it is shown that the predictions of the Nash-bargaining model are confirmed. This constitutes a rejection of the single-utility maximization model of household behavior in this decision-making realm. These results imply that the change in the selection of survivor annuities were not the only effect of the Retirement Equity Act. The increase in life insurance holdings through the household bargaining mechanism, while increasing income security for widows, was neither foreseen nor intended by the legislation. In this, there is an important lesson for policy making that targets the resource allocation within a family. Because we do not yet fully understand the decision-making dynamics in the family, policies can have unanticipated effects due to the household decision-making process. The model and the empirical results presented here take the literature one step closer to understanding this process and its implications.