b'CHAPTER 3 | ECONOMIC PREPARATION FOR RETIREMENTFIGURE 3-2Fraction of total benefits re- One of the most important ways that house-distributed among deciles: 1992 and 2004holds can maximize their Social Security benefit Source: Gustman et al. (2013). is to delay claiming benefits until their full retirement age rather than filing for benefits early 12% at age 62. Even when income is the same between those who claim at 62 and those who claim at 65, 10%average household income for those who delay 8% claiming is higher at age 72 than for those who 6% claim early (Glickman and Hermes 2015).4% Gender Differences in2% Retirement Preparation0% Women face very different economic circumstanc-Individuals Households es in retirement than men. Single women andSingle women in general fare 1992 2004 African American women appear to be especiallymuch worse than their married vulnerable. Widowhood is one important risk but Others also consider the impact of changingdoes not account for all of the discrepancy. Othercounterparts in terms of work patterns of men and women on Socialfactors may be at play, like household bargainingretirement finances.Security replacement rates. Wu et al. (2013)power and gender differences in financial savvy calculate the replacement rate as the Socialand risk-taking. Single women in general fare much worse Security benefit amount divided by the AIME, soWomen live much longer than men onthan their married counterparts in terms of changes that result in higher earnings will tendaverage and tend to marry men who are olderretirement finances. Yet important differences to lower the replacement rate. The replacementthan them. As a result, most women live the lastexist even among single women. Lee and Rowley rate goes from 47% for the HRS cohort to 39%years of their lives as widows, many for as long as(2009) compare financial portfolios of widowed, for Early Baby Boomers. One-third of this decline15 to 20 years. Depending on how well the coupledivorced, and never married women between the is accounted for by the increase in female laborprepares, widowhood can put some women at riskages of 51 and 64 in 2004. Across all asset types, force participation and the resulting increaseof financial strain or even poverty. Sevak et al.widowed women hold higher asset values than in earnings between the cohorts. The decline is(2004) show that of women who were not living indivorced and never married women. Divorced larger for women than for men, and especially forpoverty when they were married in 1992, 10.8%women have especially low levels of stock women who are currently married, divorced orwere living in poverty by 1998. Women who wereholding. Total net worth is highest for widowed widowed, compared to women who never mar- poor before their husbands died face an evenwomen compared to divorced women and is ried. Increasingly, married women are earninggreater challenge because their husbands die atlowest for never married women. Other studies Social Security benefits through their own laboryounger ages than more affluent men. This leavesconfirm the risk associated with divorce for wom-force participation, which tends to reduce theirpoorer women widowed at younger ages and for aen. Ulker (2009) shows that, while divorce has eligibility for spousal benefits. longer time. Thus part of the association observednegative consequences for wealth accumulation between widowhood and poverty is explained byfor both men and women, declines for women are persistent poverty.more severe.55'