ABSTRACT This article summarizes the range of joint-life products that are currently available to married couples, and it explores the potential utility gain that such couples receive from access to actuarially fair markets. It is more difficult to estimate this utility gain for couples than for individuals because a couple's value of annuitization will depend in part on the survivor benefits that are available after one spouse has died but while the other is still alive. Valuing joint and survivor annuities also requires recognition of the potentially important interactions between the members of a married couple, such as joint consumption, interdependent utilities, and correlated mortality rates. This article considers each of these issues. It develops an valuation model for married couples and it estimates their annuity equivalent wealth, the amount of wealth that couples would need in the absence of actuarially fair markets in order to achieve the same utility level that they rece ive when such markets are available. The utility gain from annuitization is smaller for couples than for single individuals. Since most potential buyers are married, this finding may help to explain the limited size of the market for single premium annuities in the United States. INTRODUCTION Annuities play an important role in the standard theory of consumer choice when life length is uncertain. Yaari (1965) showed that an individual with a fixed stock of resources and an uncertain lifetime should purchase an contract to insure against the risk of outliving his resources. More recent work, such as Mitchell, Poterba, Warshawsky, and Brown (1999) (hereafter MPWB), finds substantial gains to annuitization for individual life-cycle consumers. Typical results suggest that a 65-year-old man who does not have access to an actuarially fair market would be willing to forgo roughly one-third of his wealth if by doing so he could purchase an actuarially fair with his remaining wealth. In spite of Yaari's (1965) insight, a number of studies have observed that the market for individual contracts in the United States is very small. Standard explanations for the limited flow of new purchases are adverse selection in the market, individual bequest motives, and the presence of other annuitized resources. Several empirical studies, including Friedman and Warshawsky (1988, 1990) and MPWB (1999), have explored the extent of adverse selection. Numerous other studies surveyed in Laitner (1997) and (more briefly) in Altonji, Hayashi, and Kotlikoff (1997) and Laitner and Juster (1996) have focused on intergenerational altruism as a potential explanation for limited demand. Simple altruistic models do not appear to provide a satisfactory explanation for observed patterns of intergenerational transfers. Auerbach, Kotlikoff, and Well (1992) show that when Social Security benefits, private defined benefit pension plan payments, and Medicare benefits are added together, more than half of the resources of the current elderly in the United States take the form of life-contingent payouts. Thus many households are already annuitized. This may also explain the limited demand for additional private annuities. The functioning of markets has recently attracted attention as part of the global policy debate on individual accounts Social Security systems. A central design issue in such systems concerns the way a retiree would spread accumulated resources over his or her remaining lifetime, or his or her lifetime and that of his or her spouse. Private annuities offer one way of spreading such resources; mandatory government-provided annuities are another. Although the treatment of married couples is an important issue in retirement income security, virtually all of the previous research on annuities has focused on individuals rather than couples as decision-making units. …