RETIREMENT RESEARCH August 2017, Number 17-15 ARE MANY RETIREES WITH DEMENTIA LACKING HELP? By Anek Belbase, Geoffrey T. Sanzenbacher, and Sara Ellen King* Introduction Many older individuals with cognitive impairment, The options considered include help from an infor- including the vast majority of people with demen- mal caregiver (e.g., a non-impaired spouse or child), tia, need help managing their finances. For retirees from the staff of a nursing home, or from a Power of receiving Social Security benefits, the Representative Attorney they may have assigned. Once these sources Payee Program can serve as one source of this help. of help are identified, the brief then focuses on the In the Representative Payee Program, a retiree’s types of people who seem to lack any observed form benefit is sent to another person (often a relative) who of aid in order to help policymakers and community- spends it on the retiree’s behalf and submits records based organizations better identify the most vulner- to Social Security documenting that the expendi- able individuals. tures were in the beneficiary’s best interest. But the The discussion proceeds as follows. The first sec- program seems to be seldom used by those with tion estimates the share of retirees with mild cogni- dementia: of those 65 and older, over 10 percent have tive impairment and dementia. The second section dementia, but just 1.5 percent have a payee. This lack documents what sources of help people have and of participation may not be a problem as long as the what share of retirees with impairment are lacking it. retiree has some other source of help. The third section identifies groups who are especially This brief – based on a recent paper – uses data likely to have no observed source of help available. from the Health and Retirement Study (HRS) linked The final section concludes that while few retirees to administrative Social Security records to first with dementia use the Representative Payee Program, document what share of retirees with mild cognitive the vast majority have some potential source of assis- impairment or dementia use the Representative Payee tance. Groups vulnerable to having no help available Program.1 Given that few use the program, the brief include those with less education, minorities, and then turns to the question of what they do instead. individuals living in densely populated areas. * All of the authors are with the Center for Retirement Research at Boston College. Anek Belbase is a research fellow. Geoffrey T. Sanzenbacher is a research economist. Sara Ellen King is a research associate. 2 Center for Retirement Research How Many Are Impaired? The onset of cognitive decline often leads to a decline in the ability to manage one’s finances. About To construct the sample of interest – retirees with 5 percent of those without any impairment need help cognitive impairment – the HRS is restricted to managing their finances, increasing to 18 percent for those ages 70 and over who receive a Social Security those with mild impairment and ultimately 80 per- retirement benefit.2 To identify those with cognitive cent for those with dementia.5 For those who need impairment, the sample is divided into three groups, help, the question is whether or not they have it. those with: 1) no observed cognitive impairment; 2) mild cognitive impairment; and 3) dementia. Because the HRS does not directly ask about Do the Impaired Have Help? dementia, it is necessary to rely on an imputation method developed in another study to create these Determining which retirees with cognitive impair- groups.3 That methodology starts with a subsample ment have access to financial assistance is based on of the HRS that did have a direct measure of demen- two measures of assistance. The first is access to a tia – a sample given detailed cognitive assessments representative payee. Individuals with a payee are during the Aging, Demographics, and Memory Study. identified using the restricted Respondent Cross-Year These sample members were either diagnosed as Benefits file, which records payee use and is linked to having no impairment, mild cognitive impairment, or the core HRS.6 Figure 2 shows that as cognitive im- dementia. The characteristics of these diagnosed sub- pairment worsens, the use of a Representative Payee sample members were then used to predict whether increases dramatically. Still, even for those with members of the full sample had these conditions.4 dementia, payee use is below 10 percent. Of course, this indirect method may fail to identify some people who have mild cognitive impairment but Figure 2. Percentage of Households Ages 70+ who do not have the characteristics identified that are Using a Payee, by Cognitive Impairment Status predictive of this condition. For this reason, the study also classifies any individuals as having mild cognitive 10% 9.1% impairment if they: 1) cannot remember the correct date; 2) failed in two attempts counting backwards 8% from 86; 3) failed in two attempts counting backwards from 20; or 4) remembered at most one word from 6% a list of ten words. As one would expect, the results show that mild cognitive impairment and dementia 4% become more common with age (see Figure 1). 2.0% 2% 0.5% Figure 1. Percentage of Households Ages 70+ 0% with Mild Cognitive Impairment or Dementia No cognitive Mild cognitive Dementia impairment impairment 50% 45.2% Sources: Authors’ calculations from 1992-2010 HRS; Social Mild cognitive impairment 42.2% Dementia Security Cross-Year Benefits Records; and Hurd et al. (2013). 40% 33.2% 34.3% 30% 26.5% Given that payee use is uncommon, a second more inclusive set of possible sources of financial 20% assistance is considered. The first source is a spouse 14.6% who has no sign of mild cognitive impairment or 10% 6.3% dementia. The second source is children who are 2.7% either living with their parent or are helping them 0% with either Activities of Daily Living or Instrumental 70-74 75-79 80-84 85+ Activities of Daily Living. The third source of help is Ages residence in nursing homes. Financial management Sources: Authors’ calculations from the Health and Retire- may be easier for those in a nursing home since, in ment Study (HRS), 1992-2010; and Hurd et al. (2013). Issue in Brief 3 some cases, their benefit may be applied directly to the stay. Fraud may also be less likely if individuals Table 1. Available Assistance, by Cognition Status are not exposed to the wider community. The final source considered is the designation of power of attor- Mild cognitive Percentage with: Dementia ney. With respect to power of attorney, our approach impairment is admittedly imperfect. The HRS does not ask its Representative payee 2.0 % 9.1 % full sample specifically about financial powers of at- Nursing home care 3.2 28.7 torney.7 Instead, it asks about whether the individual Non-impaired spouse 37.2 29.1 has designated a durable power of attorney in the context of health. The assumption is that a person Resident child 17.3 23.4 serving as a power of attorney for health is likely also Help from a child 12.2 36.6 available to provide other assistance. Although none Power of attorney 62.9 63.2 of these sources of help is sure to prevent financial fraud or abuse (for example, children themselves Any form of assistance 85.0 95.1 sometimes commit abuse),8 people without any of Number of observations 1,756 492 these forms of help are likely financially vulnerable. Fortunately, most individuals do have some help Sources: Authors’ calculation from 1992-2010 HRS; Social (see Table 1). Eighty-five percent of those with mild Security Cross-Year Benefits Records; and Hurd et al. (2013). cognitive impairment have at least one form of help, and 95 percent of those with dementia have help. As people transition from mild cognitive impairment to retiree’s children’s marital status is included since dementia, unimpaired spouses become less common single children may be more able to help. Other but children appear to step into the breach. variables include the retiree’s education, with separate Given that impaired individuals without assistance variables for status as a dropout or high school gradu- are vulnerable, the question is who they are. ate (with “at least some college” as the base case). Race and ethnicity is controlled for with an indicator for whether the person is non-white or Hispanic. Who Does Not Have Help? Finally, the analysis controls for the strength of an individual’s community with indicators for affiliation To determine the characteristics of those without any with a Catholic Church and residence in a county source of financial assistance, the analysis used the with fewer than 250,000 people.9 HRS to identify several likely correlates of lacking as- Figure 3 shows the effect of these variables on the sistance. The first variable reflects the extent to which probability of having at least one form of assistance the individual is isolated (not having a non-impaired (a negative effect indicates people with that charac- spouse and not having children within 10 miles). A teristic were less likely to have help).10 The results Figure 3. Estimated Effect of Select Variables on Probability of Having Some Form of Help for Those with Mild Cognitive Impairment or Dementia Isolated -9.7% Has unmarried children 2.9% High school dropout -6.7% High school degree -4.4% Non white -6.9% Catholic 3.4% County population under 250,000 2.7% -15.0% -10.0% -5.0% 0.0% 5.0% Note: All effects are statistically significant at least at the 10-percent level. Full results in Appendix. Sources: Authors’ calculations from 1992-2010 HRS; Social Security Cross-Year Benefits Records; and Hurd et al. (2013). 4 Center for Retirement Research show that people who are isolated from family, less Conclusion educated, and non-white are less likely to have help. In other words, those who are economically vulner- Retirees with cognitive impairment and especially able in the traditional sense are also more likely to dementia need help managing their finances. With- have little outside help and be vulnerable to financial out help, they are vulnerable to financial fraud, abuse, mismanagement. Having a strong community, as or mismanagement. So, at a first glance, it is con- indicated by involvement with a Catholic Church or cerning that only 9 percent of retirees with dementia residence in a small county, is associated with being have a Representative Payee. After all, this program more likely to have help. helps people manage what, for many, is their primary source of income: their Social Security benefit. Fortunately, this analysis shows that most people without a payee have some other source of assistance available. These arrangements may be preferable to having a payee because they allow individuals to maintain some autonomy for as long as possible. Still, social service organizations should recognize that people who are vulnerable in other ways – for example because they have little education or are isolated from family – may also lack help managing their money.   Issue in Brief 5 Endnotes References 1 Belbase and Sanzenbacher (2016). Ron Acierno, Melba A. Hernandez, Ananda B. Am- stadter, Heidi S. Resnick, Kenneth Steve, Wendy 2 Workers who have received non-retirement Social Muzzy, and Dean G. Kilpatrick. 2010. “Prevalence Security income, like disability or SSI, are excluded. and Correlates of Emotional, Physical, Sexual, In the analysis, one observation per person is used, and Financial Abuse and Potential Neglect in the either their last year in the sample or their observa- United States: The National Elder Mistreatment tion from Wave 10 of the HRS. For more detail on Study.” American Journal of Public Health 100(2): the sample, see the full paper. 292-297. 3 Hurd et al. (2013). In the most recent wave, the Belbase, Anek and Geoffrey T. Sanzenbacher. 2016. HRS did add a question on Alzheimer’s. “Cognitive Impairment and Social Security’s Representative Payee Program.” Working Paper 4 In practice, this prediction was accomplished in 2016-12. Chestnut Hill, MA: Center for Retire- two steps. First, ordered probit models were run ment Research at Boston College. using the subsample. In this regression, the inde- pendent variables were available in the full sample, Michael D. Hurd, Paco Martorell, Adeline Delavande, and the dependent variable was no impairment, mild Kathleen J. Mullen, and Kenneth M. Langa. 2013. cognitive impairment, or dementia. The second step “Monetary Costs of Dementia in the United used the coefficients from the regression to form States.” The New England Journal of Medicine predictions for the full sample. 368(14): 1326-1334. 5 Marson et al. (2009). Marson, Daniel C., Roy C. Martin, Virginia Wad- ley, H. Randall Griffith, Scott Snyder, Patricia S. 6 Retirees are said to have a payee if they have a non- Goode, F. Cleveland Kinney, Anthony P. Nicholas, missing value for the “Custody Code” of their payee Terri Steele, and Britt Anderson. 2009. “Clinical (e.g., spouse, natural or adopted child or stepchild, or Interview Assessment of Financial Capacity in non-profit non-mental institution, etc.). Older Adults with Mild Cognitive Impairment and Alzheimer’s Disease.” Journal of the American Geri- 7 A new experimental module does ask about fi- atrics Society 57(5): 806-814. nancial power of attorney, but it was asked to only a subset of the population and the number of individu- als in our sample who were asked the question is very small. 8 Acierno et al. (2010). 9 Affiliation with a Catholic church was included as a proxy for having access to social services. Other religions were tested, but in general the sample sizes were too small to yield a significant result. 10 Other controls were included in the regression but are not shown in Figure 3. These include the individ- ual’s history of being divorced or widowed, purchase of long-term care insurance, Social Security Primary Insurance Amount, gender, and region of residence. For full results, see the Appendix. APPENDIX Issue in Brief 7 Appendix Table 1. Marginal Effects for Some Assistance with Financial Management Variable Coefficient Isolated -0.097 *** (0.025) Has unmarried children 0.029 * (0.016) High school dropout -0.067 *** (0.026) High school degree -0.044 ** (0.021) Non white -0.069 *** (0.020) Catholic 0.034 ** (0.016) County population under 250,000 0.027 * (0.015) Age 0.005 *** (0.001) History of divorce/being widowed -0.027 (0.027) Has children 0.002 (0.033) Long-term care insurance -0.028 (0.024) Social Security PIA -0.000 (0.000) Male -0.002 (0.016) Midwest -0.022 (0.025) South -0.010 (0.023) West 0.023 (0.024) Number of observations 2,615 R-squared 0.038 Notes: Statistically significant at 10-percent (*), 5-percent (**), or 1-percent level (***). Significance is calculated using robust standard errors, which are reported in parentheses. Source: Authors’ calculations from 1992-2010 HRS. RETIREMENT RESEARCH About the Center Affiliated Institutions The mission of the Center for Retirement Research The Brookings Institution at Boston College is to produce first-class research Syracuse University and educational tools and forge a strong link between Urban Institute the academic community and decision-makers in the public and private sectors around an issue of criti- cal importance to the nation’s future. To achieve Contact Information Center for Retirement Research this mission, the Center sponsors a wide variety of Boston College research projects, transmits new findings to a broad Hovey House audience, trains new scholars, and broadens access to 140 Commonwealth Avenue valuable data sources. Since its inception in 1998, the Chestnut Hill, MA 02467-3808 Center has established a reputation as an authorita- Phone: (617) 552-1762 tive source of information on all major aspects of the Fax: (617) 552-0191 retirement income debate. E-mail: crr@bc.edu Website: http://crr.bc.edu © 2017, by Trustees of Boston College, Center for Retirement Research. All rights reserved. Short sections of text, not to exceed two paragraphs, may be quoted without explicit permission provided that the authors are identified and full credit, including copyright notice, is given to Trustees of Boston College, Center for Retirement Research. The research reported herein was performed pursuant to a grant from the U.S. Social Security Administration (SSA) funded as part of the Retirement Research Consortium. The opinions and conclusions expressed are solely those of the authors and do not represent the opinions or policy of SSA, any agency of the federal government, Boston College, or the Center for Retirement Research. Neither the United States Government nor any agency thereof, nor any of their employees, makes any warranty, express or implied, or assumes any legal liability or responsibility for the accuracy, completeness, or usefulness of the contents of this report. Reference herein to any specific commercial product, process or service by trade name, trade- mark, manufacturer, or otherwise does not necessarily constitute or imply endorsement, recommendation or favoring by the United States Government or any agency thereof.