Publications by Karen E. Smith for Retirement Policy

More about Karen E. Smith's areas of expertise can be found on this Urban Institute expert's page.

We use the Health and Retirement Study to investigate household assets at older ages. We find a notable increase in the net worth of older households between 1998 and 2006, with most of the growth due to housing. The age pattern of asset accumulation and decumulation varies considerably by income. High-income seniors increase assets at older ages. Middle-income seniors reduce assets in retirement, but at a rate that will not deplete assets within their expected life. Low-income seniors spend their financial assets at a rate that will mostly deplete them, leaving only Social Security and defined-benefit pensions at older ages.

The paper provides an independent evaluation of the SIPP Synthetic Beta File. This file, created by the Bureau of the Census, is intended to provide a public use database with similar statistical properties as the confidential Social Security Administration's earnings and benefit data linked to the SIPP. There is much to praise in the Census work. Many univariate distributions were "spot on." Unweighted regression analyses had some problems and results for them were mixed. In policy simulation modeling there were many instances of differences between the Synthetic and actual data that would have led researchers to wrong conclusions.

The sharp decline in the stock market in 2008 placed the retirement security of many Americans at risk. Although the market has rebounded this year after bottoming out in March 2009, as of mid-October 2009 the S&P 500 Index remained 30 percent below its peak level two years earlier. This brief simulates the impact of the 2008 stock market crash on future retirement savings under alternative scenarios. The results show that by delaying retirement one additional year, many mid- and late-career workers could increase their income at age 67 enough to offset some or all of their stock market losses.

This paper simulates the impact of the 2008 stock market crash on future retirement savings under alternative scenarios. If stocks remain depressed as after the 1974 crash, 20 percent of preboomers born 1941-45 and 22 percent of late boomers born 1961-65 would see their retirement incomes drop 10 percent or more. Working another year would reduce the share of these big losers to 14 percent for late boomers. Because most pre-boomers were already retired, their share of big losers would decline slightly to 19 percent. Delaying retirement would disproportionately benefit low-income people because their additional earnings exceed their stock market losses.

Urban Institute projections suggest the stock market collapse will have small effects on most Americans' retirement incomes. It's estimated that 37 percent of Americans born between 1941 and 1965 owned no stocks when the market crashed in 2008 and that income from assets will account for a small share of retirement income, even for those with stocks. For most retirees, Social Security provides the majority of income. Had Social Security been invested in private accounts with equities, the impact of the crash would have been much larger—positive or negative, depending on one's birth cohort and on future market performance.

The one-third drop in the S&P 500 index between year-end 2007 and 2008 raises concerns about retirement security since Americans now hold more equities through their retirement plans. Those near retirement will fare the worst because they have no time to recoup their losses. Midcareer workers will fare better because they have more time to rebuild their wealth. They may even gain income if they buy stocks at low prices and get above-average rates of return. High-income groups will be the most affected because they are most likely to have financial assets and to be invested in the stock market.

This report compares projections of income and assets from the Model of Income in the Near Term (MINT) with data from the Survey of Income and Program Participation (SIPP), the Health and Retirement Study (HRS), the Survey of Consumer Finances (SCF), and the Current Population Survey (CPS). The comparison reveals a fair amount of variability in population characteristics and reported income and assets among these data files. There is no "right" answer, but rather a range of possible answers. For most statistics we compare, MINT's projected values fall between the highest and lowest values among the survey data.

Over the last three decades there has been a steady shift from DB to DC pensions. The Pension Protection Act of 2006 may accelerate this trend. This paper examines the impact of an accelerated freeze on the retirement income of boomers. Simulations suggest that such a scenario would produce more losers than winners and reduce average retirement incomes. Income changes will be substantial among high-income workers, who have the highest DB coverage and pension incomes. Late boomers will experience the largest impacts, as they lose their high DB accrual years and have inadequate time to accumulate DC wealth before retirement.

Income tax provisions affect the buildup of retirement assets during workers' careers and after-tax income following retirement. This paper uses the Urban Institute's DYNASIM model to simulate how potential changes in the tax treatment of retirement saving, Social Security benefits, and income from assets outside retirement accounts may affect boomers' retirement incomes. Changes in the income thresholds for taxing Social Security benefits have the largest impact on middle-income boomers, while changes in contribution limits for retirement saving plans and tax rates on capital gains and dividends have the largest impact on the highest-income boomers.

This report describes the work the Urban Institute performed to generate the Model of Income in the Near Term, Version 5 (MINT5). MINT is a tool developed for The Division of Policy Evaluation (DPE) of the Social Security Administration (SSA) to analyze the distributional consequences of Social Security reform proposals. MINT is a micro-level data file of individuals born between 1926 and 2018. It starts with a rich set of income and demographic characteristics from the 1990 to 1996 Survey of Income and Program Particpation (SIPP) data linked to SSA data on earnings and benefits. MINT then projects these characteristics until death or the year 2099.

Return to list of authors.