Long-Term Care and Lifetime Earnings: Assessing the Potential to Pay

08/01/2008

U.S. Department of Health and Human Services

Long-Term Care and Lifetime Earnings: Assessing the Potential to Pay

Richard W. Johnson and Gordon B.T. Mermin

The Urban Institute

August 2008

PDF Version: http://aspe.hhs.gov/daltcp/reports/2008/ltearn.pdf (36 PDF pages)


This report was prepared under contract #HHSP23320095654WC between the U.S. Department of Health and Human Services (HHS), Office of Disability, Aging and Long-Term Care Policy (DALTCP) and the Urban Institute. For additional information about this subject, you can visit the DALTCP home page at http://aspe.hhs.gov/_/office_specific/daltcp.cfm or contact the office at HHS/ASPE/DALTCP, Room 424E, H.H. Humphrey Building, 200 Independence Avenue, S.W. , Washington , D.C. 20201 . The e-mail address is: webmaster.DALTCP@hhs.gov. The Project Officer was Hunter McKay.

The opinions and views expressed in this report are those of the authors. They do not necessarily reflect the views of the Department of Health and Human Services, the contractor or any other funding organization.


TABLE OF CONTENTS

INTRODUCTION
BACKGROUND
DATA AND METHODS
Measuring Disability
Measuring Long-Term Care Use
Measuring Baseline Income, Baseline Wealth, and Lifetime Earnings
Sample Characteristics
RESULTS
CONCLUSIONS
REFERENCES
TABLES
APPENDIX TABLES
NOTES
LIST OF TABLES
TABLE 1: Prevalence of Disability between 1993 and 2006, Adults Age 70-79 in 1993
TABLE 2: Incidence of Help with Disabilities between 1993 and 2006
TABLE 3: Demographic Characteristics by Disability Status and the Receipt of Medicaid-Financed Home Care between 1993 and 2006
TABLE 4: Economic Status by Disability Status and the Receipt of Medicaid-Financed Home Care between 1993 and 2006
TABLE 5: Incidence of Nursing Home Stays between 1993 and 2006
TABLE 6: Demographic Characteristics of Adults With and Without Medicaid-Financed Nursing Home Stays
TABLE 7: Economic Characteristics of Adults With and Without Medicaid-Financed Nursing Home Stays
TABLE A1: Coefficients from Ordinary Least Squares Regressions of Earnings Relative to the Social Security MTE Among Workers with Earnings Exceeding MTE in 1971, 1980, and 1990
TABLE A2: Coefficients from Ordinary Least Squares Regressions of Average Lifetime Annual Earnings Among Respondents with Social Security Earnings Records
TABLE A3: Coefficients from Ordinary Least Squares Regressions of Former Spouse's Average Lifetime Annual Earnings Among Widows with Deceased Spouse Social Security Earnings Records
TABLE A4: Coefficients from Ordinary Least Squares Regressions of Spouse's Average Lifetime Earnings Among Married Respondents
TABLE A5: Sample Characteristics by Household-Size-Adjusted Lifetime Earnings Quintile

INTRODUCTION

The aging of the population will sharply increase long-term care spending over the coming decades. The Centers for Medicare and Medicaid Services (CMS) project nursing home and home health costs will increase from $199 billion today to $337 billion by 2017 (CMS 2007b). Costs will probably grow even more rapidly in about 20 years when the Baby Boomers begin reaching their 80s and the demand for long-term care services will likely soar. According to Urban Institute projections the number of adults ages 65 and older with disabilities may double by 2040 (Johnson, Toohey, & Wiener 2007). Future increases in long-term care costs not only threaten individuals’ retirement security but also the financial stability of Medicaid, which now finances nearly half of all long-term care expenditures (Georgetown University Long-Term Care Financing Project 2007).

Enhanced private saving for long-term care needs might ease budgetary pressures by increasing the pool of funds that could finance future services and reducing reliance on public resources. Government-subsidized savings accounts might be an effective way to increase saving for long-term care needs if tax preferences and government matching contributions induce workers to set aside funds for the future. Experimental evidence suggests that matching contributions significantly increase savings in individual retirement accounts (IRAs) (Duflo et al. 2005), although those findings may not apply fully to savings accounts earmarked for long-term care, because they would restrict the future use of funds more than IRAs do.

An important question, however, is whether adults who receive publicly financed long-term care services at older ages earn enough over their lifetimes to accumulate much savings. If relatively few people with average or above-average lifetime earnings ever use Medicaid-financed long-term care services, encouraging private savings would have only limited potential for addressing the financial challenges facing the publicly funded long-term care system. However, if many people in the top three-fifths of the lifetime earnings distribution eventually receive Medicaid-paid services, enhancing savings incentives could make a significant difference, because many of these workers could likely afford to set money aside for the future.

This study examines the relationship between lifetime earnings and use of Medicaid-financed long-term care services in a nationally representative survey of older adults. We calculate the shares of older adults who are ever disabled, ever use any long-term care services, ever receive Medicaid-financed home care, and ever experience Medicaid-financed nursing home stays over a 13-year period. We show how these proportions vary by demographic characteristics and lifetime earnings, and examine the characteristics of adults who ever use Medicaid-financed home or institutional care. The study assesses how many Medicaid enrollees receiving long-term care earned enough when they were working to have potentially set aside sufficient savings to cover substantial portions of their future long-term care needs.

BACKGROUND

Many older adults eventually become frail and need help with everyday activities. About 10 million Americans ages 65 and older had long-term care needs in 2002, equal to about 29 percent of the older population (Johnson & Wiener 2006). Only about 1.4 million older adults (14 percent of the frail older population and 4 percent of the entire older population) resided in nursing homes (Spillman & Black 2005a). The chances that an older person ever becomes frail and enters a nursing home is much higher. According to one study, about 7 in 10 adults turning 65 will eventually develop long-term care needs, including about 8 in 10 older women (Kemper, Komisar, & Alecxih 2005/2006).

Most frail older people live at home and receive unpaid help from family and friends. Data from the National Long-Term Care Survey (NLTCS) show that only 25 percent of community-dwelling older adults with chronic disabilities received assistance from paid helpers in 1999, down from 33 percent in 1994 (Spillman & Black 2005b). Estimates from the Health and Retirement Study (HRS) are even lower, with only 14 percent of disabled older adults living at home receiving paid home care in 2002 (Johnson & Wiener 2006). By contrast, more than half of community-dwelling Americans ages 65 and older with disabilities receive unpaid care. Estimated shares range from 57 percent in the 2002 HRS to 66 percent in the 1999 NLTCS.1

Spouses and adult children (especially daughters) provide much of the unpaid care received by frail older adults. Spouses account for about two-thirds of the helpers assisting older married care recipients living at home (McGarry 1998). However, only about 4 in 10 at-home frail older adults are married and living with their spouses (Johnson & Wiener 2006). Children play the dominant caregiving role for older people who are not married. About three-fourths of unmarried older care recipients receive help from their children, with more than one-half receiving help from daughters and slightly more than one-quarter receiving help from sons (Johnson & Wiener 2006). However, children are more likely to provide help with errands and household chores than basic personal care (McGarry 1998). Although most care is provided by a solitary caregiver, many older people obtain assistance from a network of helpers. For example, nearly one-quarter of frail older care recipients with two or more children receive help from more than child (Checkovich & Stern 2002).

Caregiving responsibilities often involve long hours of arduous work that leaves caregivers feeling overwhelmed. On average, caregivers for frail adults spend 4.3 years providing care (National Alliance for Caregiving & American Association of Retired Persons (AARP) 2004). Those caring for older people average about 150 hours per month helping, and primary ADL caregivers average 260 hours per month (Johnson & Wiener 2006). Most caregivers are ill-prepared for their role and provide care with little or no support (Alzheimer’s Association & National Alliance for Caregiving 2004). Their care responsibilities often leave them feeling frustrated, angry, drained, guilty, or helpless (Center on an Aging Society 2005). These burdens take emotional, physical, and financial tolls on caregivers. They exhibit higher levels of depressive symptoms and mental health problems than their peers who do not provide care (Marks, Lambert, & Choi 2002; Pinquart & Sorensen 2003; Schulz, O’Brien, Bookwals, & Fleissner 1995). Caregiving responsibilities also appear to impair physical health. Caregivers are more likely to develop serious illness than noncaregivers (Shaw et al. 1997) and are less likely to engage in preventive health behaviors (Schulz et al. 1997). Stressed elderly spousal caregivers exhibit higher mortality rates than people of the same age who do not provide care (Schulz & Beach 1999). Caregiving responsibilities also impose financial costs. About 6 in 10 caregivers of disabled adults are employed (National Alliance for Caregiving & AARP 2004), and many middle-aged women participate less in the labor force when they spend time helping their frail parents (Johnson & Lo Sasso 2006).

Social and demographic changes will likely intensify caregiver burdens in coming years. Balancing employment and care responsibilities is becoming more challenging as more women join the labor force. Between 1980 and 2005, the labor force participation rate of married women ages 45-64 increased from 47 percent to 67 percent (U.S. Census Bureau 2007). Falling fertility rates will also limit the future availability of family caregivers, potentially increasing the burdens on those who provide care. Women born between 1956 and 1960 had only 1.9 children on average, compared with 3.2 children for women born between 1931 and 1935 (Redfoot & Pandya 2002), while the share of women ages 40-44 without any children almost doubled (to 19 percent) between 1980 and 1998 (Bachu & O’Connell 2001).

Families may respond to the rising burdens of informal care by turning to paid helpers. The use of paid home care increased during the 1980s and early 1990s, partly because of expansions in Medicare and Medicaid financing of home care services (Feder, Komisar, & Niefeld 2000; Liu, Manton, & Aragon 2000), but then declined somewhat (Spillman & Black 2005b). The share of older adults in nursing homes has steadily declined over the past two decades, falling from 4.5 percent in 1985 to 3.6 percent in 2004 (Bishop 1999; Lewin Group 2006). Nursing home rates fell most dramatically among adults ages 85 and older, declining from 21 percent to 14 percent between 1985 and 2004 (Lewin Group 2006). At the same time, the nursing home population has grown older and more frail. For example, 86 percent of residents needed help with three or more ADLs in 2004, up from 72 percent in 1987 (Lewin Group 2006; Rhodes & Krauss 1999). Some researchers expect future nursing home rates to increase in response to rising disability rates among younger adults, although others expect the downward trend to continue (Lakdawalla et al. 2003; Manton 2003). According to one estimate, the number of aged nursing home residents will more than double between 2000 and 2040 (Johnson, Toohey, & Wiener 2007).

Paid care, however, is expensive. The national average hourly rate for home health aides was $19 in 2007 (MetLife Mature Market Institute 2007a), which translates into nearly $14,000 per year for someone who receives 60 hours of paid care per month, the typical level of home care among recipients (Johnson & Wiener 2006). A year’s stay in a nursing home averaged $78,000 in 2007, and was much higher in certain parts of the country (MetLife Mature Market Institute 2007b).

Few families can afford to pay these costs out of pocket for long, and the lack of alternative financing options pushes many long-term care recipients--especially nursing home residents--onto Medicaid. Medicare covers nursing home costs only for stays that follow a hospitalization, and then only for a limited period. Private long-term care insurance is rare, covering only about 9 percent of adults ages 55 and older in 2002 (Johnson, Schaner, Toohey, & Uccello 2007). In 1999 Medicaid was the primary payer of nursing home costs for 57 percent of residents ages 65 and older (National Center for Health Statistics 2002). By comparison, 15 percent of elderly nursing home residents relied primarily on Medicare coverage, and 25 percent relied primarily on private funds (a small portion of which came from private insurers). Medicaid covers a smaller share of new nursing home residents, many of whom can cover the initial costs on their home or qualify for Medicare. At the time of admission, Medicaid-covered the majority of costs for 38 percent of elderly residents, compared with 33 percent by Medicare (National Center for Health Statistics 2002). Additionally, the share of nursing home residents covered by Medicaid exceeds the share of nursing home costs paid by Medicaid, because Medicaid-covered residents face lower rates than private-pay residents. In 2004, for example, Medicaid paid about 37 percent of nursing home costs for adults ages 65 and older, Medicare paid 16 percent, and private insurance paid 2 percent (Congressional Budget Office 2004). Out of pocket payments covered about 36 percent of total nursing home costs.

DATA AND METHODS

This study examines the relationship between lifetime earnings and use of Medicaid-financed long-term care services in a nationally representative survey of older adults. We first calculate the share of adults ages 70-79 in 1993 who are ever disabled over the next 13 years, ever receive any long-term care, ever receive Medicaid-financed home care, and ever experience a Medicaid-financed nursing home stay. The analysis compares the incidence of disability and Medicaid-financed long-term care by demographics, baseline income, baseline wealth, and lifetime earnings. We examine the characteristics of adults ever using Medicaid-financed home or nursing home care and assess how many of these adults had enough earnings when they were working to have potentially saved for future long-term care needs.

Our data come from the HRS, a longitudinal survey of older Americans conducted by the Survey Research Center at the University of Michigan with primary funding from the National Institute on Aging. The survey collects detailed information on employment, income, assets, and health status, including the presence of difficulties with particular ADLs and particular IADLs. It also collects information on long-term care use, including home care and nursing home care, and how respondents finance their paid care. All HRS respondents live in the community at baseline, but the survey follows respondents into nursing homes as necessary, and collects data from proxies for respondents unable to provide information themselves. The survey oversamples African Americans, Hispanics, and Florida residents but includes sample weights that we use to adjust the estimates so they represent the underlying national population.

We restrict our sample to adults ages 70-79 at the baseline 1993 interview.2 The HRS also collected information from community-dwelling adults age 80 and older in 1993, but we exclude these respondents because the likelihood of nursing home care increases rapidly after age 80. Including in the analysis a selected sample of very old adults who are not in nursing homes could bias our estimates of long-term care use. We use information from the baseline interview and the follow-ups in 1995, 1998, 2000, 2002, 2004, and 2006. We also incorporate information from exit interviews with next of kin for respondents who die by 2006. Surviving respondents who drop out of the survey before 2006 are eliminated from our sample. Our final sample consists of 4,471 adults.

Measuring Disability

We first compute the share of adults ages 70-79 in 1993 who are ever frail or severely disabled by 2006, when they range from ages 83-92.3 We classify respondents as frail if they report difficulty because of health problems lasting three or more months with one or more ADLs--bathing, dressing, eating, getting in or out of bed, walking across a room, or using the toilet--or IADLs--buying groceries, preparing hot meals, using the phone, taking medications, or managing money. We classify respondents as severely disabled if they report two or more ADL limitations, the threshold for triggering benefits under many private long-term care insurance policies.4

Measuring Long-Term Care Use

Next we calculate the incidence of long-term care use over the 13-year period for adults ages 70-79 at baseline. We classify adults as receiving at-home care if they live in the community and report receiving help with an ADL or IADL. We also identify adults who receive any care from paid caregivers and those who receive any help from caregivers paid by Medicaid. The analysis identifies respondents as Medicaid-financed home care recipients if they report that Medicaid or insurance paid their helpers and if they report elsewhere in the survey that they have Medicaid coverage. (The survey does not ask respondents whether Medicaid, not insurance, pays their caregivers.) We classify respondents as receiving nursing home care if they ever reside in a nursing home at the time of an interview or report a nursing home stay since the previous interview. Respondents are classified as receiving Medicaid-financed nursing home care if they ever report Medicaid coverage concurrent with a nursing home stay.

Rates of Medicaid coverage at older ages are similar in the HRS and other datasets. For example, our tabulations show that 7.2 percent of adults ages 65 and older receive Medicaid coverage in the 2006 HRS, compared with 6.8 percent in the 2006 National Health Interview Survey. Medicaid rates are somewhat higher in the Medical Panel Expenditure Survey, which shows about 10 percent of older adults with Medicaid coverage at the end of 2005, according to our calculations.

Our HRS estimates of the incidence of Medicaid-financed home care and Medicaid-financed nursing home stays over the 13-year period are roughly consistent with administrative data. Our measure indicates that 400,000 adults ages 65 and older received Medicaid home care in 2006, whereas estimates based on CMS’ Medicaid Statistical Information System indicate that about 650,000 older adults received Medicaid home care in 2002 (Sommers, Cohen & O'Malley 2006). Our estimates may be lower than the administrative totals because our sample is restricted to older adults who were not in nursing homes at study baseline, so we are excluding long-term nursing home residents, who are most likely to receive Medicaid financing. Additionally, we estimate that about 800,000 adults ages 65 and older received Medicaid nursing home care in 2006, whereas CMS (2007a) counts about 1,160,000 aged Medicaid nursing home residents in 2004. Any undercount in Medicaid-financed long-term care use is unlikely to affect our principal findings much, because we focus on the characteristics of the recipients of Medicaid-financed care, and there is no evidence that our sample is biased.

Measuring Baseline Income, Baseline Wealth, and Lifetime Earnings

Baseline income, baseline wealth, and lifetime earnings are key measures in our study. They are all computed at the household level and expressed in constant 2006 dollars, adjusted by the change in the consumer price index. When grouping respondents into income, wealth, and earnings quintiles we divide these financial amounts by 1.62 for married and partnered respondents to account for couples’ high consumption needs relative to unmarried adults.5

Household income reported in 1993 includes 1992 Social Security retirement benefits, employer-sponsored pension benefits, disability and Supplemental Security Income benefits, asset income, and earnings.6 Household wealth includes the value of bank accounts, stocks, bonds, IRAs, housing, other real estate, vehicles, and businesses, net of mortgage and other debt.

Our measure of lifetime earnings is based on Social Security administrative earnings records that can be linked to many, but not all, respondents in our sample.7 To compute average lifetime annual earnings we calculate mean earnings from 1951 (when our sample members were ages 28-37) to the year they attain age 64.8 The estimated average includes years of zero earnings. We adjust earnings each year by the growth in economy-wide average wages from that year to 1993, to make earnings comparable over time. This approach is similar to the one used by the Social Security Administration when computing average indexed monthly earnings to determine benefits. To assess better the total earnings pool from which respondents could have saved during their working years, we calculate household average lifetime annual earnings, including earnings of respondents, current spouses, and former spouses. Our lifetime earnings measures require numerous imputations to account for the truncation of observed earnings, employment outside the Social Security system, missing own earnings records, and missing spouse earnings, as described below.

Workers and their employers pay Social Security taxes only on earnings up to a certain amount--known as maximum taxable earnings (MTE)--and Social Security earnings records report only those earnings that do not exceed this amount in any given year. The MTE was $102,000 in 2008 and $57,600 in 1993, but it was much lower in earlier years. In 1961, for example, it was only $4,800, just 9 percent higher than average Social Security-covered earnings. As a result, most men in our sample with administrative earnings records (84 percent) have some truncated earnings years. Because women born between 1914 and 1923 generally earned much less than men, only 19 percent of women with administrative earnings records have at least one year of truncated earnings. Among respondents with truncated earnings, the average number of truncated earnings years is 18 for men and 7 for women. We impute total annual earnings in truncated years based on ordinary least square regression models of earnings estimated on a sample of workers with earnings exceeding MTE in the 1971, 1980, and 1990 March Current Population Surveys (CPS). (See Appendix Table A1.)9

We impute average lifetime annual earnings (as opposed to annual earnings in select years) for respondents without links to administrative earnings records and for those who appear to have worked for much of their careers in fields that were not always covered by Social Security. Administrative earnings records are available for only 62 percent of the respondents in our sample. Additionally, the earnings records do not record earnings from jobs that are not covered by Social Security, so ignoring uncovered work could lead us to underestimate lifetime earnings. Although nearly all workers are now covered by Social Security, the system excluded many workers in the past. In particular, most government employees were not covered by Social Security before 1983 (and some state and local government workers remain outside the system today). We identify adults who spent much of their careers working outside of Social Security as those whose self-reported estimate of lifetime employment years in the HRS survey exceeds the number of positive-earning years in their linked administrative records by 50 percent or more. This approach implies that 8 percent of our sample worked outside of Social Security. Another 38 percent of respondents lack any administrative earnings records, and 28 percent are married to spouses who lack administrative records or worked outside of Social Security. We impute average lifetime annual earnings for these respondents and spouses based on an ordinary least squares regression of average lifetime annual earnings estimated on a sample of HRS respondents with administrative earnings records who worked in covered employment for their entire careers. (See Appendix Table A2.)

We also impute average lifetime annual earnings for most former spouses. Former spouses’ earnings are important for the analysis because the capacity to save for future long-term care needs depends on the lifetime earnings of the entire household, including current and former spouses. Social Security earnings records are available for about 200 deceased spouses of widowed female HRS respondents in 1993. We must impute the former spouse’s lifetime earnings for the remaining 1,300 widows, widowers, and divorced respondents. Imputations for widowed women are based on an ordinary least squares regression model of the former spouse’s average lifetime annual earnings estimated on a sample of widows linked to their deceased spouse’s earnings records. (See Appendix Table A3.) The model is estimated as functions of the widow’s characteristics. To impute lifetime earnings for former spouses of widowers and divorced respondents, we use regression models of spouse’s average lifetime annual earnings estimated on a sample of married respondents linked to their spouse’s administrative earnings records. We estimate separate models for men and women. (See Appendix Table A4).

Sample Characteristics

Appendix Table A5 reports the characteristics of our sample and show how they vary by household-size-adjusted lifetime earnings quintile. Women make up about 57 percent of the sample, and 60 percent of the sample is married in 1993. More than one-third (about 36 percent) did not complete high school, and more than one-half (53 percent) are ever widowed between 1993 and 2006. About 46 percent of observations rely on imputed own lifetime earnings, and 63 percent rely on imputed lifetime earnings for current and former spouses. Respondents in the lower lifetime earnings quintiles are more likely to be female, unmarried, African American, and Hispanic than those in the upper quintiles. They also have less education and less baseline income and wealth.

RESULTS

Disability is quite common at older ages. More than three-quarters of adults (about 77 percent) living at home at ages 70-79 are frail at some point by ages 83-92, and more than half (54 percent) experience severe disability at some point during the 13-year period (Table 1). Although disability is common among all groups, prevalence rates vary substantially by demographic characteristics. Women, widows, adults with limited education, and African Americans are more likely to experience disability than other groups. Differences by educational attainment are particularly striking, with about three-fifths of adults who lack high school diplomas experiencing severe disability, compared with only about two-fifths of college graduates. (About 38 percent of community-dwelling adults ages 70-79 in 1993 did not complete high school, and only about 11 percent completed college.) Disability rates vary considerably by economic group as well. Adults with baseline incomes below the federal poverty line (FPL) are about 24 percentage points more likely to be severely disabled at some point than adults with incomes above 400 percent of FPL. Similarly, the prevalence of severe disability is about 21 percentage points higher for adults in the bottom quintile of lifetime earnings than for those in the top quintile.

About half of adults ages 70-79 and living at home in 1993 ever receive at-home assistance with ADLs or IADLs between 1993 and 2006 (Table 2). Not surprisingly, adults with ADL or IADL limitations are more likely to receive help than those who do not need assistance. About 65 percent of those who are ever frail and 70 percent of those who are severely disabled ever receive ADL or IADL help. Women are more likely to receive help than men and African Americans and Hispanics are more likely to receive help than Whites, largely because women and racial minority groups are more likely to be frail. The likelihood of receiving help also declines as educational attainment, income, and lifetime earnings increase.

Most of the at-home care received by older adults comes from unpaid family members and friends. Only about 18 percent of adults in our sample ever receive paid home care during the 13-year observation period, and only about 4 percent ever receive Medicaid-financed home care. Among adults ever severely disabled, about 30 percent receive paid help and about 6 percent receive Medicaid-financed home care. Women are nearly twice as likely as men to receive paid help and more than three times as likely to receive home care paid by Medicare. The likelihood of receiving Medicaid-financed home care is much lower among college graduates than those who did not complete high school, and falls rapidly as income and lifetime earnings increase. For example, about 11 percent of adults in the bottom fifth of the household-size-adjusted lifetime earnings distribution receive home care paid by Medicaid at some point over a 13-year period, compared with only about 1 percent of those in the top three-fifths of the distribution. Similar patterns exist for the receipt of any type of paid help, regardless of funding source, but they are much less pronounced.

Table 3 compares the demographic characteristics of adults ages 70-79 in 1993 who are never-disabled by 2006, who are disabled at some point but never receive Medicaid-financed home care, and who receive Medicaid-financed home care at some point during the 13-year period. Adults who ever receive Medicaid-financed home care are predominantly female and widowed. For example, women make up about 8 in 10 Medicaid home care recipients and widows and widowers make up about 7 in 10. By contrast, only about half of the adults in our sample who are never-disabled are female, and less than half are ever widowed. Medicaid-financed home care recipients also tend to have less education than those who do not receive these services, and they are more likely to be African American or Hispanic. About 2 percent of adults who ever receive Medicaid home care have college degrees, compared with 18 percent of never-disabled adults. African Americans and Hispanics together account for nearly 4 in 10 Medicaid home care recipients but less than 1 in 10 never-disabled adults.

Medicaid home care recipients have much lower baseline incomes and wealth and lifetime earnings than other older adults (Table 4). For example, unmarried older adults who ever receive Medicaid home care between 1993 and 2006 receive less than half as much income in 1992 as never-disabled adults, and only about 55 percent as much income as those who are disabled at some point but never receive home care paid by Medicaid. Household annual earnings averaged over a lifetime is only about two-thirds as high for unmarried users of Medicaid-financed home care as for older unmarried adults with disabilities who do not use Medicaid home care. Median wealth for single adults who ever receive Medicaid home care is only about $450 in 1993, compared with about $84,000 for those with disabilities who do not use Medicaid home care and about $136,000 for never-disabled adults.

Relatively few recipients of Medicaid-financed home care earned enough at younger ages to have saved much for future long-term care needs. Only 15 percent of older Medicaid home care recipients are in the top three-fifths of the lifetime household earnings distribution, a plausible criterion for being able to afford to put aside substantial funds. About 63 percent of Medicaid home care recipients are in the bottom fifth of the distribution. By comparison, nearly 7 in 10 older adults who are never-disabled are in the top three-fifths of the lifetime earnings distribution.

Table 5 shows the incidence of nursing home care between 1993 and 2006. Overall about 3 in 10 adults ages 70-79 and living at home in 1993 enter a nursing home by 2006, including nearly half of adults experiencing a spell of severe disability. Women and widows are more likely than men and other marital status groups to receive nursing home care over the 13-year period. Hispanics and adults with baseline income in excess of twice the poverty line are less likely to enter nursing homes. The likelihood of entering a nursing home does not vary much by education or lifetime earnings.

About 9 percent of adults experience Medicaid-financed nursing home stays during the sample period, slightly more than one-quarter of those who receive any nursing home care. About one-third of severely disabled adults who enter nursing homes (and 15 percent of severely disabled adults overall) eventually go on to Medicaid during the 13-year period. Women, widowed adults, and African Americans are about twice as likely to receive Medicaid-financed nursing home care as men, those not widowed, and nonHispanic Whites, and adults without high school diplomas are more than five times as likely to receive these services as college graduates. About one in five adults with baseline income below the poverty line eventually receive Medicaid-financed nursing home care during the sample period. Although the incidence of Medicaid nursing home stays is relatively low for adults in the top three-fifths of the household lifetime earnings distribution, they are much more likely to receive Medicaid financing for nursing home stays than home care.

Table 6 compares the demographic characteristics for older adults who ever have a Medicaid-financed nursing home stays, those who never enter a nursing, and those who enter a nursing home but never receive Medicaid financing. Recipients of Medicaid-financed nursing home care are predominantly female, widows or widowers, and adults who did not complete high school. For example, women account for three-quarters of older adults with Medicaid-financed nursing home stays, and adults widowed at some point over the 13-year observation period account for nearly the same share. By contrast, only about half of older adults who never enter nursing homes are female or widowed. Fewer than 4 percent of adults with Medicaid-financed nursing home stays are college graduates, compared with 13 percent of adults who never receive nursing home care. Additionally, 19 percent of adults with Medicaid-financed nursing home stays are African American or Hispanic, compared with about 12 percent of adults without nursing home stays. Adults with nursing home stays financed by sources other than Medicaid do not differ much from adults who never enter nursing homes. In fact, adults with nonMedicaid-financed nursing home stays are somewhat more likely to have high school diplomas or college degrees than adults without nursing home stays.

Adults with Medicaid-financed nursing home stays have less baseline income and wealth than adults with no nursing home stays (Table 7). For example, median 1993 household wealth for single adults with Medicaid nursing home stays is about $22,000 (measured in 2006 dollars), compared with about $95,000 for single adults with no nursing home stays. Median 1992 income is less than two-thirds as high for single adults who enter nursing homes and eventually go on to Medicaid than for single adults who never enter nursing homes. (The relative difference is smaller for married adults, but still substantial.)

However, many older adults with Medicaid-financed nursing home stays had significant earnings when they were younger. About two-fifths of these adults are among the top 60 percent of household lifetime earners, suggesting they could have set money aside when they were younger to finance their nursing home stays. (Married adults in our sample needed to average more than $35,000 per year in household earnings--measured in today’s dollars--between ages 20 and 64 to be in the top 60 percent of lifetime earnings.) Nearly one-fifth of these adults are among the 40 percent of household lifetime earners. Of course, many older adults with Medicaid-financed nursing home stays have very low lifetime earnings; nearly two-fifths are in the bottom of the household-size-adjusted lifetime earnings distribution.

CONCLUSIONS

Although many older nursing home residents on Medicaid have limited income and little wealth, a substantial share received substantial earnings when they were younger. About 45 percent of older adults with Medicaid-financed nursing home stays between 1993 and 2006 fell into the bottom fifth of the 1992 household income distribution for their age group, and about 46 percent fell into the bottom fifth of the 1993 household wealth distribution, when both distributions are adjusted for household size. Only about 27 percent were in the top three-fifths of the 1992 income distribution, and only 29 percent were in the top three-fifths of the 1993 wealth distribution. However, nearly 40 percent of older adults with Medicaid-financed nursing home stays between 1993 and 2006 fell into the top three-fifths of the lifetimeearnings distribution, adjusted for household size, and nearly 20 percent fell into the top 40 percent of the distribution.

Our results suggest that many older adults who end up in nursing homes and receive Medicaid benefits could have put aside funds to cover part of their future long-term care costs when they were younger. Policymakers could reduce Medicaid costs, about one-third of which now cover long-term care (Kasper, Lyons, & O’Malley 2007), by convincing families to earmark some of their earnings when they are relatively young for future long-term care costs. The challenge, of course, is to create the necessary incentives to boost private savings. Efforts to promote private long-term care insurance coverage, mostly through tax incentives, have not been particularly successful, perhaps because of the existence of the Medicaid safety net (Brown & Finkelstein 2008). An alternative approach would be to create tax incentives for people to invest in special accounts that they could later use to purchase private long-term care insurance (Mermin, Johnson, & Lewis 2008). Although these tax breaks would accrue disproportionately to high-income families, those are the families that could most plausibly be kept off Medicaid when long-term care needs arise, and thus offer the best prospects for Medicaid savings.

REFERENCES

Alzheimer's Association & National Alliance for Caregiving.2004. Families Care: Alzheimer's Caregiving in the United States. Chicago, IL: Alzheimer's Association.

Bachu, Amara, & Martin O'Connell. 2001. "Fertility of American Women: June 2000." Current Population Reports.Washington, DC: U.S. Census Bureau.

Bishop, Christine E. 1999. "Where Are the Missing Elders? The Decline in Nursing Home Use, 1985 and 1996. Health Affairs 18(4):146-155.

Brown, Jeffrey R., & Amy Finkelstein. 2008. "The Interaction of Public and Private Insurance: Medicaid and the Long-Term Care Insurance Market." American Economic Review 98(3):1083-02.

Center on an Aging Society. 2005. "How Do Family Caregivers Fare: A Closer Look at their Experiences." Washington, DC: Georgetown University. http://ihcrp.georgetown.edu/agingsociety/pdfs/CAREGIVERS3.pdf.

CMS. 2007a. Health Care Financing Review 2007 Statistical Supplement.Baltimore, MD: CMS.

CMS. 2007b. "National Health Expenditure Projections 2007-2017." Baltimore, MD: CMS. http://www.cms.hhs.gov/NationalHealthExpendData/Downloads/proj2007.pdf.

Checkovich, Tennille J., & Steven Stern. 2002. "Shared Caregiving Responsibilities of Adult Siblings with Elderly Parents." Journal of Human Resources37(3):441-78.

Citro, Constance F., & Robert T. Michael. 1995. Measuring Poverty: A New Approach. Washington, DC: National Academy Press.

Congressional Budget Office. 2004. "Financing Long-Term Care for the Elderly." Washington, DC: Congressional Budget Office. http://www.cbo.gov/ftpdocs/54xx/doc5400/04-26-LongTermCare.pdf.

Duflo, Esther, William Gale, Jeffrey Liebman, Peter Orszag, & Emmanuel Saez. 2005. "Saving Incentives for Low- and Middle-Income Families: Evidence from a Field Experiment with H&R Block." NBER Working Paper 11680. Cambridge, MA: National Bureau of Economic Research.

Feder, Judith, Harriet L. Komisar, & Marlene Niefeld.2000. "Long-Term Care in the United State: An Overview." Health Affairs 19(3):40-56.

Georgetown University Long-Term Care Financing Project.2007. "National Spending for Long-Term Care." http://ltc.georgetown.edu/pdfs/natspendfeb07.pdf.

Johnson, Richard W., & Anthony T. Lo Sasso. 2006. "The Impact of Elder Care on Women's Labor Supply at Midlife." Inquiry 43(3):195-210.

Johnson, Richard W., & Joshua M. Wiener. 2006. "A Profile of Frail Older Americans and Their Caregivers." Retirement Project Occasional Paper No. 8. Washington, DC: Urban Institute. http://www.urban.org/url.cfm?ID=311284.

Johnson, Richard W., Desmond Toohey, & Joshua M. Wiener. 2007. "Meeting the Long-Term Care Needs of the Baby Boomers: How Changing Families Will Affect Paid Helpers and Institutions." Retirement Project Discussion Paper 07-04. Washington, DC: Urban Institute. http://www.urban.org/retirement_policy/url.cfm?ID=311451.

Johnson, Richard W., Simone G. Schaner, Desmond Toohey, & Cori E. Uccello. 2007. Modeling the Decision to Purchase Private Long-Term Care Insurance. Final Report to the Office of the Assistant Secretary for Planning and Evaluation, U.S. Department of Health and Human Services. Washington, DC: Urban Institute. http://aspe.hhs.gov/daltcp/reports/2007/LTCImod.htm.

Kasper, Judith, Barbara Lyons, & Molly O'Malley. 2007. "Long-Term Care Services and Supports: The Future Role and Challenges for Medicaid." Washington, DC: Kaiser Commission on Medicaid and the Uninsured. http://www.kff.org/medicaid/upload/7671.pdf.

Kemper, Peter, Harriet L. Komisar, & Lisa Alecxih. 2005/2006. "Long-Term Care Over an Uncertain Future: What Can Current Retirees Expect?" Inquiry42(4):335-350.

Lakdawalla, Darius, Dana P. Goldman, Jay Bhattacharya, Michael D. Hurd, Geoffrey F. Joyce, & Constantijn W.A. Panis. 2003. "Forecasting the Nursing Home Population." Medical Care41(a):21-24.

Lewin Group. 2006. "Nursing Home Use by 'Oldest Old' Sharply Declines." http://www.lewin.com/NR/rdonlyres/9A0A92A2-4D76-4397-A0A2-04EB20700795/0....

Liu, Korbin, Kenneth G. Manton, & Cynthia Aragon. 2000. "Changes in Home Care Use by Disabled Elderly Persons: 1982-1994." Journal of Gerontology: Social Sciences 55b(4):S245-53.

Manton, Kenneth G. 2003. "Forecasting the Nursing Home Population."Medical Care 41(1):8-20.

Marks, Nadine, James D. Lambert, & Heejeong Choi. 2002. "Transitions to Caregiving, Gender, and Psychological Well-Being: A Prospective U.S. National Study." Journal of Marriage and Family 64:657-67.

McGarry, Kathleen. 1998. "Caring for the Elderly: The role of Adult Children."In Frontiers in the Economics of Aging, edited by David A. Wise (133-63). Chicago, IL: University of Chicago Press.

Mermin, Gordon B.T., Richard W. Johnson, & Eric Lewis. 2007. Assessing the Potential of Subsidized Retirement and Health Savings Accounts. Final Report to the Office of the Assistant Secretary for Planning and Evaluation, U.S. Department of Health and Human Services. Washington, DC: Urban Institute. http://aspe.hhs.gov/daltcp/reports/2008/apHRSA.htm.

MetLife Mature Market Institute. 2007a. "The MetLife Market Survey of Adult Day Services and Home Care Costs." New York, NY: Metropolitan Life Insurance Company.

MetLife Mature Market Institute. 2007b. "The MetLife Market Survey of Nursing Home and Assisted Living Costs." New York, NY: Metropolitan Life Insurance Company.

National Alliance for Caregiving & AARP. 2004. "Caregiving in the U.S." http://www.caregiving.org/data/04finalreport.pdf.

National Center for Health Statistics. 2002. "The National Nursing Home Survey: 1999 Summary." Vital and Health Statistics 13(152):1-69.

Pinquart, M., & S. Sorensen. 2003. "Differences Between Caregivers and Noncaregivers in Psychological Health and Physical Health: A Meta-Analysis." Psychology and Aging 18(2):250-267.

Redfoot, Donald L., & Sheel M. Pandya. 2002. "Before the Boom: Trends in Long-term Supportive Services for Older Americans with Disabilities." AARP Public Policy Institute Report No. 2002-15. Washington, DC: AARP.

Rhoades, Jeffrey A., & Nancy A. Krauss. 1999. Nursing Home Trends, 1987 and 1996. AHCPR Pub. No. 99-00032. Rockville, MD: Agency for Health Care Policy and Research.

Schulz, R., & S.R. Beach. 1999. "Caregiving as a Risk Factor for Mortality: The Caregiver Health Effects Study." Journal of the American Medical Association282:2215-2219.

Schulz, R., A.T. O'Brien, J. Bookwals, & K. Fleissner.1995. "Psychiatric and Physical Morbidity Effects of Dementia Caregiving: Prevalence, Correlates, and Causes." Gerontologist35:771-791.

Schultz, R., J. Newson, M. Mittelmark, L. Burton, C. Hirsch, & S. Jackson. 1997. "Health Effects of Caregiving: The Caregiver Health Effects Study." Annals of Behavioral Medicine 19:110-116.

Shaw, W.S., T.L. Patterson, S.J. Semple, S. Ho, M.R. Irwin, R.L. Hauger, & I. Grant. 1997. "Longitudinal Analysis of Multiple Indicators of Health Decline Among Spousal Caregivers." Annals of Behavioral Medicine 19:101-109.

Sommers, Anna, Mindy Cohen, & Molly O'Malley. 2006. "Medicaid's Long-Term Care Beneficiaries: An Analysis of Spending Patterns." Washington, DC: Kaiser Commission on Medicaid and the Uninsured.

Spillman, Brenda C., & Kristen J. Black. 2005a. The Size of the Long-Term Care Population in Residential Care: A Review of Estimates and Methodology. Washington, DC: Office of the Assistant Secretary for Planning and Evaluation, U.S. Department of Health and Human Services. http://aspe.hhs.gov/daltcp/reports/2005/ltcpopsz.htm.

Spillman, Brenda C., & Kristen J. Black. 2005b. "Staying the Course: Trends in Family Caregiving." Washington, DC: AARP. http://assets.aarp.org/rgcenter/il/2005_17_caregiving.pdf.

U.S. Census Bureau. 2007. Statistical Abstract of the United States (127th Edition). Washington, DC: U.S. Department of Commerce.

TABLES

TABLE 1. Prevalence of Disability between 1993 and 2006, Adults Ages 70-79 in 1993
  N   Ever Frail     Ever Severely  Disabled
All   4,471   76.6 54.0
Gender
   Male 1,848 74.4 50.6
   Female 2,623 78.4 56.5
Marital Status in 1993
   Married 2,585 74.7 50.2
   Divorced 261 78.2 57.0
   Widowed 1,495 80.3 60.8
   Never married 130 73.5 52.2
Ever Widowed
   Yes 2,454 78.5 55.7
   No 2,017 74.5 52.1
Education
   Not high school graduate   1,718 81.5 60.4
   High school graduate 2,249 76.1 52.6
   College graduate 504 64.7 40.8
Race
   White, other 3,619 75.6 52.5
   African American 598 87.0 66.9
   Hispanic 254 78.9 61.2
1992 Household Income as Percent of Poverty Line
   Less than 100 1,102 83.8 66.2
   100-124 452 80.9 57.8
   125-199 1,167 79.9 55.9
   200-400 1,202 71.9 47.0
   More than 400 548 65.2 42.4
Household-Size-Adjusted Lifetime Earnings Quintile
   Bottom 1,068 84.2 65.8
   2nd 897 80.0 57.3
   3rd 848 77.5 51.3
   4th 847 74.1 50.8
   Top 811 67.5 45.0
SOURCE: Authors' estimates from the HRS.NOTE: The sample is restricted to adults ages 70-79 and living at home in 1993. Respondents are classified as frail if they ever report any difficulty with an ADL or IADL, and are classified as severely disabled if they ever report two or more ADL limitations. Lifetime earnings include the earnings of spouses and former spouses, and are divided by 1.62 for respondents who have ever been married.
TABLE 2. Incidents of Help with Disabilities between 1993 and 2006 (%)
  N   Ever Receiving  Help withADL/IADL   Ever Receiving  Paid Help   Ever Receiving  Medicaid-FinancedHome Care
All   4,471   49.8 17.9 3.6
Disability History
   Ever frail 3,455 64.9 23.4 4.7
   Ever severely disabled 2,468 70.0 30.1 6.2
Gender
   Male 1,848 46.7 12.9 1.6
   Female 2,623 52.1 21.7 5.1
Marital Status in 1993
   Married 2,585 48.0 15.3 2.0
   Divorced 261 49.1 22.2 7.6
   Widowed 1,495 53.5 21.7 5.7
   Never married 130 47.0 22.5 5.7
Ever Widowed by 2006
   Yes 2,454 51.8 21.2 4.7
   No 2,017 47.5 14.2 2.3
Education
   Not high school graduate   1,718 56.6 19.7 7.0
   High school graduate 2,249 47.7 17.3 1.9
   College graduate 504 38.6 15.4 0.6
Race
   White, other 3,619 48.4 17.1 2.5
   African American 598 60.9 23.1 9.1
   Hispanic 254 59.3 26.8 17.5
1992 Household Income as Percent of Poverty Line
   Less than 100 1,102 58.6 22.6 11.0
   100-124 452 54.9 21.8 5.7
   125-199 1,167 52.1 17.3 1.6
   200-400 1,202 43.3 14.5 0.8
   More than 400 548 40.7 15.8 0.0
Household-Size-Adjusted Lifetime Earnings Quintile
   Bottom 1,068 57.9 23.0 11.4
   2nd 897 52.3 19.8 3.9
   3rd 848 53.0 16.1 1.1
   4th 847 47.5 15.6 1.1
   Top 811 38.3 15.1 0.5
SOURCE: Authors' estimates form the HRS.NOTE: The sample is restricted to adults ages 70-79 and living at home in 1993. Respondents are classified as frail if they ever report any difficulty with an ADL or IADL, and are classified as severely disabled if they ever report two or more ADL limitations. Lifetime earnings include the earnings of spouses and former spouses, and are divided by 1.62 for respondents who have ever been married.
TABLE 3. Demographic Characteristics by Disability Status and the Receipt of Medicaid-Financed Home Care between 1993 and 2006 (%)
    Never Disabled     Ever Disabled,  No MedicaidHome Care   Ever Received  Medicaid-FinancedHome Care
N 1,062 3,221 188
Gender
   Male 47.6 42.9 19.6
   Female 52.4 57.1 80.4
Marital Status in 1993
   Married 64.9 59.7 33.3
   Divorced 4.7 4.8 10.7
   Widowed 27.0 32.7 51.2
   Never married 3.4 2.7 4.8
Ever Widowed
   Yes 48.6 53.7 69.3
   No 51.5 46.3 30.7
Education
   Not high school graduate   29.0 36.3 69.9
   High school graduate 53.0 52.9 28.2
   College graduate 18.0 10.8 1.9
Race
   White, other 92.3 88.5 61.8
   African American 4.4 8.4 20.0
   Hispanic 3.4 3.1 18.2
SOURCE: Authors' estimates from the HRS.NOTE: The sample is restricted to adults ages 70-79 and living at home in 1993. Respondents are classified as disabled if they ever report any difficulty with an ADL or IADL.
TABLE 4. Economic Status by Disability Status and the Receipt of Medicaid-Financed Home Care between 1993 and 2006
    Never Disabled     Ever Disabled,  No MedicaidHome Care   Ever Received  Medicaid-FinancedHome Care
N 1,062 3,221 188
Median (2006 Dollars)
All
   1992 household income $33,517 $26,820 $10,899
   1993 household wealth $213,459 $137,423 $3,376
   Household lifetime earnings   $58,881 $53,823 $32,357
Married
   1992 household income $42,266 $34,321 $19,393
   1993 household wealth $267,870 $182,401 $28,084
   Household lifetime earnings $62,791 $57,401 $31,108
Single
   1992 household income $20,090 $16,742 $9,208
   1993 household wealth $135,888 $83,709 $446
   Household lifetime earnings $52,716 $48,718 $33,359
Share in Quintile (%)
1992 Household-Size-Adjusted Income Quintile
   Bottom 14.2 19.6 62.3
   2nd 16.2 21.1 24.4
   3rd 17.8 21.5 8.0
   4th 23.0 19.8 4.9
   Top 28.8 18.1 0.4
1993 Household-Size-Adjusted Income Quintile
   Bottom 13.2 19.7 68.5
   2nd 15.0 21.6 17.0
   3rd 19.2 20.9 11.6
   4th 24.9 19.3 3.0
   Top 27.6 18.5  
Household-Size-Adjusted Lifetime Earnings Quintile
   Bottom 13.8 19.6 63.3
   2nd 16.9 20.8 21.7
   3rd 19.0 21.2 6.2
   4th 22.6 20.1 6.1
   Top 27.8 18.4 2.7
SOURCE: Authors' estimates from the HRS.NOTE: The sample is restricted to adults ages 70-79 and living at home in 1993. Respondents are classified as disabled if they ever report any difficulty with an ADL or IADL. Lifetime earnings include the earnings of spouses and former spouses. Income, wealth, and lifetime earnings quintiles are adjusted for household size, by dividing by 1.62 for couples.
TABLE 5. Incidence of Nursing Home Stays between 1993 and 2006 (%)
  N   Ever Had Nursing  Home Stay Ever HadMedicaid-  Financed Nursing  Home Stay
All   4,471   30.8 8.6
Disability History
   Never disabled 1,016 8.4 0.2
   Ever frail 3,455 37.6 11.0
   Ever severely disabled 2,468 46.2 14.8
Gender
   Male 1,848 26.0 5.1
   Female 2,623 34.4 11.3
Marital Status in 1993
   Married 2,585 27.4 5.0
   Divorced 261 33.6 16.1
   Widowed 1,495 36.5 14.2
   Never married 130 32.4 10.5
Ever Widowed
   Yes 2,454 35.7 12.0
   No 2,017 25.2 4.8
Education
   Not high school graduate   1,718 29.5 12.6
   High school graduate 2,249 32.5 7.3
   College graduate 504 27.2 2.5
Race
   White, other 3,619 31.3 7.9
   African American 598 29.2 15.7
   Hispanic 254 21.3 11.1
1992 Household Income as Percent of Poverty Line
   Less than 100 1,102 34.3 20.5
   100-124 452 33.0 13.0
   125-199 1,167 32.1 6.5
   200-400 1,202 27.7 3.4
   More than 400 548 27.3 1.8
Household-Size-Adjusted Lifetime Earnings Quintile
   Bottom 1,068 30.6 16.7
   2nd 897 30.8 9.7
   3rd 848 31.1 8.6
   4th 847 30.9 4.8
   Top 811 30.4 3.5
SOURCE: Authors' estimates from the HRS.NOTE: The sample is restricted to adults ages 70-79 and living at home in 1993. Respondents are classified as frail if they ever report any difficulty with an ADL or IADL, and are classified as severely disabled if they ever report two or more ADL limitations. Lifetime earnings include the earnings of spouses and former spouses, and are divided by 1.62 for respondents who have ever been married.
TABLE 6. Demographic Characteristics of Adults With and Without Medicaid-Financed Nursing Home Stays (%)
  Never in  Nursing Home     Ever Had Nursing  Home Stay, NotFinanced byMedicaid Ever HadMedicaid-  Financed Nursing  Home Stay
N 3,108 955 408
Gender
   Male 46.3 40.7 25.4
   Female 53.7 59.3 74.6
Marital Status in 1993
   Married 63.0 60.6 34.7
   Divorced 4.8 4.1 9.2
   Widowed 29.3 32.4 52.5
   Never married 2.9 3.0 3.5
Ever Widowed
   Yes 49.3 56.7 74.0
   No 50.7 43.3 26.0
Education
   Not high school graduate   36.3 27.5 52.0
   High school graduate 50.8 58.9 44.4
   College graduate 12.9 13.6 3.6
Race
   White, other 87.8 93.5 81.0
   African American 8.0 4.8 14.2
   Hispanic 4.2 1.7 4.8
SOURCE: Authors' estimates from the HRS.NOTE: The sample is restricted to adults ages 70-79 and living at home in 1993. Respondents are classified as frail if they ever report any difficulty with an ADL or IADL.
TABLE 7. Economic Characteristics of Adults With and Without Medicaid-Financed Nursing Home Stay between 1993 and 2006
  Never in  Nursing Home     Ever Had Nursing  Home Stay, NotFinanced byMedicaid Ever HadMedicaid-  Financed Nursing  Home Stay
N 3,108 955 408
Median (2006 Dollars)
All
   1992 household income $28,461 $30,693 $14,192
   1993 household wealth $152,072 $171,604 $36,274
   Household lifetime earnings   $54,112 $57,758 $44,186
   Married
   1992 household income $35,574 $38,077 $24,661
   1993 household wealth $201,600 $228,806 $71,327
   Household lifetime earnings $58,487 $61,160 $46,265
Single
   1992 household income $17,043 $19,301 $10,765
   1993 household wealth $95,150 $115,798 $21,904
   Household lifetime earnings $48,092 $53,800 4$3,935
Share in Quintile (%)
1992 Household-Size-Adjusted Income Quintile
   Bottom 19.1 12.2 45.4
   2nd 19.4 19.0 27.6
   3rd 19.8 23.6 13.7
   4th 21.1 21.3 7.9
   Top 20.7 23.9 5.4
1993 Household-Size-Adjusted Income Quintile
   Bottom 19.4 11.7 45.8
   2nd 19.3 19.7 24.8
   3rd 19.6 23.7 15.2
   4th 20.0 24.2 10.1
   Top 21.8 20.8 4.1
Household-Size-Adjusted Lifetime Earnings Quintile
   Bottom 19.8 12.6 38.2
   2nd 19.9 19.0 22.4
   3rd 20.0 20.5 19.9
   4th 20.1 23.6 11.4
   Top 20.2 24.2 8.2
SOURCE: Authors' estimates from the HRS.NOTE: The sample is restricted to adults ages 70-79 and living at home in 1993. Respondents are classified as disabled if they ever report any difficulty with an ADL or IADL. Lifetime earnings include the earnings of spouses and former spouses. Income, wealth, and lifetime earnings quintiles are adjusted for household size, by dividing by 1.62 for couples.

APPENDIX TABLES

TABLE A1. Coefficients from Ordinary Least Squares Regressions of Earnings Relative to the Social Security MTE Among Workers with Earnings Exceeding MTE in 1971, 1980, and 1990
  1971 1980 1990
Men Women Men Women Men Women
Age
   Age spline 14-55 0.016***(0.001)   0.005***  (0.001) 0.007***(0.000) 0.000(0.002) 0.004***(0.001) 0.003*(0.002)
   Age spline 56+   -0.013***  (0.002) 0.004(0.004) -0.002(0.002) 0.005(0.006) -0.0002(0.0021) -0.001(0.006)
Education
   Not high school graduate   -0.138***(0.015) -0.005(0.042)   -0.045***  (0.015) 0.201**(0.087) -0.033(0.040) 0.002(0.117)
   Some college 0.190***(0.016) 0.067**(0.030) 0.075***(0.011) 0.076*(0.043) 0.028(0.019) -0.009(0.050)
   College graduate 0.640***(0.015) 0.111***(0.025) 0.250***(0.010) 0.042(0.036) 0.151***(0.016) -0.014(0.042)
Race and Ethnicity
   African American -0.212***(0.027) -0.023(0.040) -0.104***(0.023) -0.101**(0.050)   -0.108***  (0.036) -0.040(0.052)
   Hispanic -0.078(0.036) 0.066(0.084) -0.006(0.026) -0.071(0.081) 0.003(0.035) -0.008(0.072)
   Other race (nonWhite) -0.136**(0.061) -0.093(0.084) 0.011(0.032) 0.150**(0.075) -0.025(0.031) 0.013(0.058)
   Constant 0.894***(0.025) 1.079***(0.050) 0.995***(0.021)   1.275***  (0.072) 1.139***(0.035)   1.210***  (0.084)
   R2 0.15 0.02 0.11 0.03 0.04 0.01
   N 19,045 2,618 9,416 618 5,016 688
SOURCE: Authors' estimates from the 1971, 1980, and 1990 March CPS.NOTE: Standard errors are in parentheses. We use these coefficients to impute annual earnings from workers in our HRS sample with administrative earnings records truncated at MTE. We use the 1971 coefficients to impute earnings in years before 1975, the 1980 coefficients for 1976-1984, and the 1990 coefficients for 1985-1993. * p < 0.10; ** p < 0.05; *** p < 0.01.
TABLE A2. Coefficients from Ordinary Least Squares Regressions of Average Lifetime Annual Earnings Among Respondents with Social Security Earnings Records
  Men Women
Married 2,776***(826) -253(358)
Age 156**(77) -25(41)
Health Status
   Excellent 880(780) -35(414)
   Poor -2,751(807) 334(420)
Education
   Not high school graduate -3,705***(719) -64(385)
   Some college 7,080***(2,923) -31(1,110)
   College graduate   12,693***  (1,017)   4,905***  (654)
Race and Ethnicity
   African American -8,827***(1,351) -1,119*(672)
   Hispanic -11,382***(1,963) 8(944)
Annual Household Income 0.09***(0.02) 0.09***(0.02)
Household Wealth 0.0005(0.002) -0.0011(0.001)
Total Years Worked 177***(23) 295***(12)
Receiving Defined Benefit Pension Benefits   5,463***(658) 4,335***(390)
Constant 9,451(5,895) 2,100(3,116)
R2 0.42 0.47
N 1,254 1,478
SOURCE: Authors' estimates from the HRS.NOTE: Standard errors are in parentheses. Time-varying explanatory variables refer to 1993. We use these coefficients to impute average lifetime earnings for workers who lack Social Security earnings records or appear to have worked most of their careers outside of Social Security-covered employment.* p < 0.10; ** p < 0.05; *** p < 0.01.
TABLE A3. Coefficients from Ordinary Least Squares Regressions of Former Spouse's Average Lifetime Annual Earnings Among Widows with Deceased Spouse Social Security Earnings Records
Explanatory Variables Coefficients  (Standard Error)  
Age 596*(315)
Health Status
   Excellent 1,897(1,943)
   Poor -3,381(2,139)
Education
   Not high school graduate -4,673***(1,885)
   Some college 382(5,059)
   College graduate 8,662***(3,468)
Race and Ethnicity
   African American -9,551***(3,890)
   Hispanic -10,816*(5,637)
Annual Household Income 0.05(0.09)
Household Wealth 0.01(0.01)
Total Years Worked -1.3(56.1)
Defined Benefit Pension Coverage   2,700(1,711)
Constant -15,928(23,781)
R2 0.278
N 205
SOURCE: Authors' estimates from the HRS.NOTE: Standard errors are in parentheses. The dependent variable is average lifetime earnings for deceased spouses. The explanatory variables refer to widows' characteristics in 1993. We use these coefficients to impute the former spouse's average lifetime earnings for widows without former spouse Social Security earnings records.* p < 0.10; ** p < 0.05; *** p < 0.01.
TABLE A4. Coefficients from Ordinary Least Squares Regressions of Spouse's Average Lifetime Earnings Among Married Respondents
  Men Women
Age 147(105) -249(192)
Health Status
   Excellent 870(694) 99(1,235)
   Poor 299(732) -1,830(1,322)
Education
   Not high school graduate     -1,843***  (648)   -5,517***  (1,169)
   Some college 7,693***(2,985) 2,064(3,286)
   College graduate 1,012(952) 5,135***(1,835)
Race and Ethnicity
   African American 359(1,307) -13,254***(2,426)
   Hispanic -2,334(1,695.1) -13,560***(2,966.4)
Own Lifetime Earnings -0.02(0.03) -0.17*(0.10)
Annual Household Income 0.002(0.02) 0.167***(0.06)
Household Wealth -0.003(0.002) 0.014***(0.003)
Total Years Worked 1.0(22.0) -30.7(42.9)
Defined Benefit Pension Coverage   142(609) -417(1,363)
Constant -2,527(7,799) 53,898***(14,164)
R2 0.03 0.23
N 814 705
SOURCE: Authors' estimates from the HRS.NOTE: Standard errors are in parentheses. The dependent variable is average lifetime earnings for spouses. The explanatory variables refer to the respondents' characteristics in 1993. We use these coefficients to impute the former spouse's average lifetime earnings for widowers and divorced respondents. We adjust imputed lifetime earnings downwards for former spouses when marriages ended before respondents attained age 64.* p < 0.10; ** p < 0.05; *** p < 0.01.
TABLE A5. Sample Characteristics by Household-Size-Adjusted Lifetime Earnings Quintile
  All Quintile
Bottom 2nd 3rd 4th Top
Share with Characteristic (%)
Gender
   Male 43.3 33.1 38.0 45.0 48.7 51.3
   Female 56.7 66.9 62.0 55.0 51.3 48.7
Marital Status in 1993
   Married 60.0 48.8 49.0 61.2 64.6 76.2
   Divorced 5.0 8.1 6.4 6.4 2.9 1.4
   Widowed 32.0 37.1 41.7 30.9 30.7 19.7
   Never married 3.0 5.9 2.9 1.6 1.8 2.7
Ever Widowed
   Yes 53.0 59.4 61.7 51.6 50.5 42.2
   No 47.0 40.6 38.3 48.4 49.6 57.8
Education
   Not high school graduate 35.7 64.6 49.4 33.2 24.1 7.8
   High school graduate 52.0 33.1 47.4 63.1 66.6 49.6
   College graduate 12.3 2.3 3.2 3.7 9.3 42.6
Race
   White, other 88.5 62.2 88.4 94.9 97.5 99.0
   African American 7.8 23.9 9.1 4.0 1.7 0.6
   Hispanic 3.7 13.9 2.5 1.1 0.8 0.4
Respondent Imputation Type
   Not imputed 54.1 58.8 48.8 49.0 57.0 56.9
   Imputed (no Social Security earnings record) 38.2 36.7 44.6 45.2 33.3 31.1
   Imputed (worked outside of Social Security-covered emp.)   7.7 4.5 6.7 5.8 9.6 12.0
Spouse Imputation Type
   Never married in 1993 3.0 5.9 2.9 1.6 1.8 2.7
   Not imputed 34.2 32.1 24.1 32.8 39.6 42.4
   Imputed (no Social Security earnings record) 23.0 15.6 25.5 26.4 21.2 26.1
   Imputed (worked outside of Social Security-covered emp.) 5.4 1.9 3.3 3.9 6.8 11.1
   Imputed (former spouse) 34.4 44.6 44.2 35.4 30.7 17.7
Median (2006 Dollars)
1992 Household-size-adjusted income 19,997 11,083 15,929 18,364 23,503 35,569
1993 Household-size-adjusted wealth   108,082     34,018     85,346     102,742     140,911     216,163  
Household-size-adjusted average lifetime annual earnings 33,766 17,075 27,699 33,632 40,448 51,523
N 4,471 1,068 897 848 847 811
SOURCE: Authors' estimates from the HRS.NOTE: Lifetime earnings include the earnings of spouses and former spouses. Income, wealth, and lifetime earnings quintiles are adjusted for household size by dividing by 1.62 for couples.

NOTES

  1. Paid and unpaid help for older disabled adults may be more common in the NLTCS than the HRS because the disability criteria is stricter in the NLTCS. Spillman & Black, who classify older adults in the NLTCS as disabled if they receive help with an activity of daily living (ADL), use an assistive device to perform in ADL, or are unable to perform instrumental activities of daily living (IADLs), find that 16 percent of community-dwelling adults ages 65 and older in the NLTCS are disabled. Johnson & Wiener estimate that about 27 percent of older at-home adults are disabled,defined as those who report any difficulty with an ADL or IADL.

  2. The HRS began interviewing a younger cohort in 1992, when they were ages 51-61. We do not include these respondents in our sample because they are too young to experience high rates of disability.

  3. The share of adults with disabilities includes those who are disabled in 1993.

  4. Under the 1996 Health Insurance Portability and Accountability Act, long-term care insurance policies are tax qualified if they restrict benefit payouts to policyholders with at least two ADL limitations or with moderate or severe cognitive impairment.

  5. This figure implies that living expenses for two adults living separately are 62 percent more than those for two adults living together. It is the midpoint of the range of household equivalence scales recommended by a recent National Academy of Science expert panel (Citro & Michael 1995).

  6. The income data collected in the HRS refer to payments received in the previous calendar year.

  7. Social Security earnings records are not available in public-use HRS data. Only researchers who have been granted special permission by HRS can gain access to these restricted data.

  8. Social Security earnings records are not available before 1951.

  9. We estimate separate models for men and women and impute earnings based on the coefficients from the CPS models and respondent characteristics in our HRS sample. We use the 1971 coefficients to impute earnings in years before 1975, the 1980 coefficients for earnings from 1976 to 1984, and the 1990 coefficients for earnings from 1985 to 1993.


SYMPOSIUM ON HEALTH AND RETIREMENT SAVINGS ACCOUNTS REPORTS AVAILABLE

Assessing the Potential of Subsidized Health and Retirement Savings Accounts
Full HTML Version   http://aspe.hhs.gov/daltcp/reports/2008/apHRSA.htm
Full PDF Version   http://aspe.hhs.gov/daltcp/reports/2008/apHRSA.pdf
Long-Term Care and Lifetime Earnings: Assessing the Potential to Pay
Full HTML Version   http://aspe.hhs.gov/daltcp/reports/2008/ltearn.htm
Full PDF Version   http://aspe.hhs.gov/daltcp/reports/2008/ltearn.pdf
Symposium on Health and Retirement Savings Accounts: Summary and Recommendations
Full HTML Version   http://aspe.hhs.gov/daltcp/reports/2010/HRSAsym.htm
Full PDF Version   http://aspe.hhs.gov/daltcp/reports/2010/HRSAsym.pdf

To obtain a printed copy of this report, send the full report title and your mailing information to:

U.S. Department of Health and Human ServicesOffice of Disability, Aging and Long-Term Care PolicyRoom 424E, H.H. Humphrey Building200 Independence Avenue, S.W.Washington, D.C. 20201FAX:  202-401-7733Email:  webmaster.DALTCP@hhs.gov


RETURN TO:

Office of Disability, Aging and Long-Term Care Policy (DALTCP) Home [http://aspe.hhs.gov/_/office_specific/daltcp.cfm]Assistant Secretary for Planning and Evaluation (ASPE) Home [http://aspe.hhs.gov]U.S. Department of Health and Human Services (HHS) Home [http://www.hhs.gov]

Preview
Download

"ltearn.pdf" (pdf, 915.57Kb)

Note: Documents in PDF format require the Adobe Acrobat Reader®. If you experience problems with PDF documents, please download the latest version of the Reader®