and participants at the NBER Summer Institute and RRC summer meetings provided many useful comments. NBER working papers are circulated for discussion and comment purposes. They have not been peerreviewed or been subject to the review by the NBER Board of Directors that accompanies official NBER publications.

This paper uses data from the Health and Retirement Study to explore the mechanism that underlies the robust relation found in the literature between cognitive ability, and in particular numeracy, and wealth, income constant. We have a number of findings. First, the more valuable the pension, the more knowledgeable are covered workers about their pensions. We suggest that causality is more likely to run from pension wealth to pension knowledge, rather than the other way around. Second, most measures of cognitive ability, including numeracy, are not significant determinants of pension and Social Security knowledge. Third, standardizing for incomes and other factors, a pension of higher value does not substitute for other forms of wealth. Rather, counting pensions in total wealth, those with more valuable pensions save more for retirement, other things the same. Fourth, there is no evidence that wealth held outside of pensions is influenced by knowledge of pensions.In sum, numeracy does not influence wealth in whole or in part by affecting financial knowledge of one's pension plan, where financial knowledge of the pension then influences other decisions about retirement saving.These findings raise questions about the mechanism that underlies the relation between cognition, especially numeracy, and wealth. From a policy perspective, they suggest that the numeracy-wealth relation should not be taken as evidence that increasing financial literacy will increase the wealth of households as they enter into retirement.

There is evidence of a relation between numeracy and wealth held outside of pensions and Social Security. With pensions and Social Security accounting for half of wealth at retirement, and evidence that those with pensions save more in other forms, one would expect to find knowledge of pensions and Social Security influencing retirement saving. Yet we find no evidence that knowledge of pensions and Social Security is related to nonpension, non-Social Security wealth, to numeracy, or that it plays an intermediate role in the numeracy-wealth relation. Our findings raise questions about policies that would enhance numeracy to increase retirement saving.

This paper investigates the effect of the current recession on the retirement age population. Data from the Health and Retirement Study suggest that those approaching retirement age (early boomers ages 53 to 58 in 2006) have only 15.2 percent of their wealth in stocks, held directly or in defined contribution plans or IRAs. Their vulnerability to a stock market decline is limited by the high value of their Social Security wealth, which represents over a quarter of the total household wealth of the early boomers. In addition, their defined contribution plans remain immature, so their defined benefit plans represent sixty five percent of their pension wealth. Simulations with a structural retirement model suggest the stock market decline will lead the early boomers to postpone their retirement by only 1.5 months on average. Health and Retirement Study data also show that those approaching retirement are not likely to be greatly or immediately affected by the decline in housing prices. We end with a discussion of important difficulties facing those who would use labor market policies to increase the employment of older workers.

NBER working papers are circulated for discussion and comment purposes. They have not been peerreviewed or been subject to the review by the NBER Board of Directors that accompanies official NBER publications.