Estimate of a Work and Save Plan in Georgia

Transcription

1 1 JUNE 6, 2017 Estimate of a Work and Save Plan in Georgia Wesley Jones Sally Wallace

2 2 Introduction AARP Georgia commissioned the Center for State and Local Finance at Georgia State University to estimate the impact of a Work and Save Plan in Georgia. We did this by estimating how a hypothetical increased savings rate for individuals retiring between 2018 and 2030 would affect state-level expenditures in Georgia. In 2013, 29 percent of Americans aged 55 and older had no retirement savings or defined benefit plan. 1 Additionally, 67.8 percent of people aged had insufficient retirement savings. 2 As a result, older adults are likely to be more dependent on state and federal aid during their retirement than during their working years. New estimates predict that with longer life expectancies nearly 46 percent of newly retired individuals will end up on long-term care paid for by Medicaid in their lifetime. 3 In the next two decades, spending for governmental aid programs such as Medicaid, Supplemental Nutrition Assistance Program (SNAP) and Supplemental Security Income (SSI) are expected to increase because of the growing number of people reaching retirement age and the low savings rate among retiring individuals. This report estimates how much the state would save if Georgians expecting to retire between 2018 and 2030 who currently have no savings for retirement were to save 3 percent of their income before retirement. If this savings rate were achieved, many individuals would no longer qualify for state aid due to asset threshold eligibility requirements. The asset threshold requirement for Medicaid and SSI is $2,000 in countable assets, which includes cash, annuities, stocks, bonds and other high liquidity assets. Thus, any individual with more than $2,000 in countable assets is automatically withdrawn from these programs. The SNAP asset requirement is $3,250 for households with an individual over the age of 60, but the program does not count most forms of retirement savings against that threshold. We assume that saving 3 percent of income will become a countable asset for each individual and that the individual will remain off Medicaid and SSI until their savings, or countable assets, drop below the $2,000 threshold. If the savings do not qualify as a retirement asset, their eligibility for SNAP would also be affected. Our estimates show that all individuals included in our estimate spend down their savings below the threshold between one and seven years after entering retirement. This report estimates the retirement savings of individuals between 2018 and State savings from the Work and Save Plan are predicted to increase every year over the duration of the study period. This occurs because with each additional year, retirees would have had more time to accumulate savings and therefore would spend more time off of state aid programs. Although this report only covers the 12-year period from 2018 to 2030, if the plan were enacted, we expect that state savings per year would continue to increase past U.S. Government Accountability Office, Retirement Security, National Institute of Retirement Security, Combination of two statistics: 1) Number of elderly who will live in a nursing home at some point in their life multiplied by the 2) Percentage of nursing home residents whose expenses are covered by Medicaid. 1) CDC.gov, Long-Term Care Providers and Services Users in the United States, 2016; 2) Center for Budget and Policy Priorities, 2015.

3 Note that savings could increase an individual s income above the income eligibility requirements by paying interest or dividends on investments. However, if the pre-retirement investment is larger than $2,000, the individual is already ineligible for state aid. Except for perhaps a few marginal cases, we believe that the vast majority of people who lose state aid due to increased savings will do so because they exceed the countable asset requirement. During the first few years of retirement, most low-income retirees will spend down savings and become dependent on state aid. Even this relatively short delay of retirees using state aid results in significant savings for Georgia. The remainder of this report outlines our estimation strategy for the number of retirees in Georgia, as well as the number of eligible retirees that will receive Medicaid, SNAP and SSI between 2018 and We then estimate the projected savings to Georgia and the federal government. 3 Georgia s Aging Demographics Georgia s population is rapidly aging. By 2030, it is projected to be the sixth oldest state by median age. Over the next 10 years, Georgia will have more people entering retirement than at any other time in the state s history. We began our analysis of the impact of a Work and Save Plan in Georgia by using estimates of the total number of people over the age of 65 in the state of Georgia through 2030 provided by the Kaiser Family Foundation. We took the total number of people over the age of 65 for each year between and then adjusted for death rates to yield the total number of people 66 and over for the following year. These figures were used to estimate the total number of people reaching retirement in each year between 2018 and Age 66 was used to estimate retirement in this study because it is the first year when full Social Security payments are available for the cohort born between 1945 and This cohort constitutes the majority of our study s focus; we note, though, that individuals born after 1960 will not receive full Social Security benefits until age 67. Individuals reaching the age of 66 between 2018 and 2030 represent those Georgians whose retirement would be affected by the policy change to savings. The number of people turning 66 each year was assumed to be equivalent to the number of people retiring in that same year. Using the projected number of people reaching the age of 66 between 2018 and 2030, we then created a base for the total number of people who will be retiring between 2018 and These individuals retirements would be affected by any changes to their savings rate. Individuals retiring in 2018 would only have one year to accumulate savings before retirement, while those retiring in 2030 would have 12 years to accumulate savings. We used the total number of people reaching retirement each year and the average death rates to calculate how many years each newly retired person will spend in retirement between 2018 and Georgia has positive elderly net migration, meaning more elderly people will be retiring in Georgia who will not be affected by a statewide savings plan than people moving out of the state at retirement. However, we believe that Georgia s net migration will only have marginal effects on the total estimate.

4 Once the total number of people and the total number of years they will spend in retirement was calculated, we created a base for the total amount of government aid that would be spent on individuals in the study if the savings rate does not increase. The difference between the projected spending with the increased savings rate and the projected spending without the increased savings rate is the projected expenditure savings for the state of Georgia. We completed this analysis for Medicaid, SSI and SNAP. 4 Medicaid Medicaid is Georgia s largest safety net program for retiring low-income individuals. Based on current growth trends, total expenditures on Medicaid in Georgia are projected to be over $28.6 billion for individuals over the age of 65 between 2018 and Of that $28.6 billion, $9.31 billion is financed through the state budget (32.5 percent of the total Medicaid expenditure). Many of the elderly Georgians who enter the program are aged into Medicaid coverage. During retirement, individuals typically have less income. People ineligible for Medicaid prior to retirement may become eligible for the program after retiring. In Georgia, 13.1 percent of people 65 and over are enrolled in Medicaid. We break this percentage down into three main categories: low-income Medicaid, nursing home residents, and home and community care for the elderly. Older adults who qualify for low-income Medicaid have an income threshold of $235 per month for an individual. This means that any retiree must have had pre-retirement income below $4,500 per year in order for their Social Security payments to be below the $235 income threshold; therefore, older adults on low-income Medicaid would not be affected by changes in the savings rate because saving 3 percent of their income would not surpass the $2,000 threshold within the 12-year window of this analysis. We estimate the savings to the state from these Medicaid beneficiaries to be zero. Nursing home residents constitute a relatively small share of the total people on Medicaid, but they account for roughly 47 percent of the total Medicaid spending on the elderly. 5 Another roughly 35 percent of Medicaid spending for older adults is for home and community care for older adults. Both nursing and community care spending per capita quickly increases with age, and the majority of Medicaid spending is for individuals over the age of 85. However, by the time most people reach 85 years of age, they have exhausted their savings including the estimated additional savings from the proposed Work and Save initiative. Therefore, there is very little cost savings to Medicaid. NURSING HOME CARE An estimated 85 percent of nursing home residents are age 65 or over. In 2015, Medicaid was the primary payer for more than 60 percent of American nursing home residents and accounted for over 40 percent of all nursing home revenue. 6 In 2012, Medicaid spent more than $1.2 billion on Georgians in 5 From Kaiser Family Foundation, Medicaid Moving Forward. Calculated by the percentage of Medicaid spending on nursing homes multiplied by the percentage spent for the elderly, divided by the total percentage of Medicaid spending on the elderly. 6 Center for Budget and Policy Priorities, 2015.

5 5 nursing homes, which equates to approximately 12 percent of all Medicaid spending in the state that year. 7 To estimate how much the state would save from a decrease in Medicaid nursing home care utilization, we calculated the number of people who are predicted to use Medicaid to pay for their nursing home in Georgia between 2018 and 2030 (Table 1). To do this, we first determined the total number of people using nursing homes in Georgia in 2016 (1). Using projections from the Kaiser Family Foundation, we estimated this number for each year from 2018 to We multiplied each year s total number of people in nursing homes by the likelihood that an individual residing in a nursing home would be enrolled in Medicaid (2). Then, we used age demographics for nursing home residents to find the total number of people in nursing homes paid for by Medicaid between the ages of 65 and 74 (15.5 percent) (3). We used ages because after age 74 all retirees will have exhausted their savings. We assumed a constant distribution for each discrete age between 65 and 74. This means we assumed that every age between 65 and 74 contained 1.55 percent of the total population in nursing homes. Table 1 shows a summary of these statistics. Table 1. Elderly Populations in Georgia Nursing Homes Paid by Medicaid (1) 65+ IN GEORGIA IN NURSING HOME (2) 65+ IN GEORGIA IN NURSING HOME PAID BY MEDICAID (3) IN GEORGIA IN NURSING HOME PAID BY MEDICAID AGE 66 IN GEORGIA IN NURSING HOME PAID BY MEDICAID 65+ IN YEAR GEORGIA ,490,045 34,271 22,276 3, ,559,339 35,865 23,312 3, ,628,633 37,459 24,348 3, ,697,928 39,052 25,384 3, ,767,222 40,646 26,420 4, ,836,516 42,240 27,456 4, ,910,428 43,940 28,561 4, ,984,341 45,640 29,666 4, ,064,312 47,479 30,861 4, ,147,505 49,393 32,105 4, ,234,052 51,383 33,399 5, ,324,086 53,454 34,745 5, ,336,267 53,734 34,927 5, Next, we created a distribution for the pre-retirement income for nursing home residents ages We did this by creating eight income categories, ranging from Georgians with no income to those within the 60th percentile of income. 8 This income range reflects those most likely to qualify for Medicaid when 7 Kaiser Family Foundation. 8 The income categories are based on U.S. income distributions.

6 entering a nursing home because higher income individuals are less likely to use Medicaid or be a nursing home resident. Individuals in each income group in our distribution have different likelihoods of enrolling in Medicaid and being a nursing home resident. To account for these differences, we multiplied the likelihood of a person being in any particular income group by the probability that a person in that group will have no retirement savings. The retiring poor have the highest probability of ending up in nursing homes paid for by Medicaid, and our distribution reflects the income characteristics of the retiring poor. We then normalized the probabilities and distributed the expected number of Medicaid nursing home residents across these groupings. We calculated this income distribution separately from our estimate of how many people would be utilizing Medicaid for nursing homes. The distribution reflects only how we calculated savings for those newly retired between 2018 and To calculate the savings, we took the median wage for each income category and assumed that each person in that group made the median wage. We assumed that each person saves 3 percent of his or her income and that all savings earn an average return of 3 percent per year. We used the median income to create a chart for total savings entering retirement for each income group for every year of the study ( ). We then calculated total savings entering retirement based on the income group and year of retirement for an individual. Table 2 shows total savings for each category. 6 Table 2. Savings Associated with a Work and Save Plan by Year of Retirement and Median Income Grouping ($) MEDIAN INCOME GROUPING* YEAR $2,943 $7,356 $10,299 $13,241 $15,478 $16,949 $29,296 $49, , ,784 3, ,005 1,100 2,716 4, ,163 1,360 1,489 3,677 6, ,148 1,476 1,726 1,890 4,666 7, ,399 1,799 2,102 2,302 5,685 9, ,184 1,657 2,131 2,491 2,727 6,734 11, ,374 1,923 2,473 2,890 3,165 7,815 13, ,569 2,197 2,825 3,302 3,616 8,928 15, ,771 2,479 3,188 3,726 4,080 10,075 17, ,978 2,770 3,561 4,163 4,559 11,256 19, ,192 3,069 3,946 4,613 5,051 12,473 21, ,413 3,378 4,343 5,076 5,559 13,726 23,412 *The dollar amount for median income grouping represents the median yearly salary for the eight different income categories. These categories range from 0 to the 60th percentile of U.S. household income. This range captures the vast majority of all people who will end up on Medicaid at some point in their retirement. Note: Savings entering retirement was calculated by having 3 percent of income saved and all savings receiving an annual 3 percent return on investment.

7 All enrollees above the $2,000 threshold are expected to remain off Medicaid until their savings fall below this threshold. We estimate the spending of savings to be the difference between retirement consumption and Social Security payments, which are typically the only sources of income for poor, retired individuals. Retirement consumption is based on pre-retirement consumption levels. We assume that the retired maintain some level of their standard of living; however, consumption is still less than it would be had the individual not retired. 9 We, therefore, use a replacement rate of 70 percent, assuming that the average retiree will consume 70 percent of their pre-retirement consumption. In all cases, retirement consumption is greater than Social Security payments. The individual is expected to use savings to make up the difference between retirement consumption and retirement income. This difference represents how quickly individuals will spend down their retirement savings. Once retirement savings is close to the $2,000 threshold, we allow for the intentional spending down of savings to qualify for Medicaid. Individuals with six months of savings above the threshold will intentionally spend down savings to qualify for Medicaid and lower their monthly expenses. Table 3 summarizes the difference between retirement consumption and Social Security payments for the eight income groups. 7 Table 3. Savings Loss per Year by Median Income Grouping ($ per year) MEDIAN INCOME GROUPING $2,943 $7,356 $10,299 $13,241 $15,478 $16,949 $29,296 $49,968 Consumption at Retirement* 2,060 5,149 7,209 9,268 10,834 11,864 20,506 34,977 Yearly Social Security Income 1,896 4,716 6,648 8,520 9,876 10,212 17,796 13,044 Loss in Savings Each Year ,652 2,710 21,933 *Consumption at retirement is assumed to be 70 percent of pre-retirement consumption. Note: For individuals with small retirement savings, Social Security and savings are assumed to be the only sources of income. Social Security less retirement consumption represents how much savings will be spent down each year. By calculating how quickly retirees will spend down their savings, we estimated how long each group will remain above the $2,000 threshold and, therefore, off Medicaid due to the Work and Save Plan. Groups with relatively low income and groups with fewer years before retirement are less likely to spend time off Medicaid. A significant portion of the retiring poor will spend between one and three years off Medicaid, with a majority of those delaying state aid for only a year. For example, those retiring in 2030 who make between 125 and 150 percent of the federal poverty level are expected to save $4,343 before retirement. Their retirement income minus their retirement consumption nets to -$748 per year, meaning individuals in that category are expected to delay their use of Medicaid benefits by 3.1 years. We calculated the total savings to the state by multiplying the total amount of time each older adult delays Medicaid coverage by their expected expenditures on nursing home care over that time. Table 4 represents the government savings for each non-zero year between 2018 and The table also displays the total government 9 For example, we assume retirees will have lower transportation costs due to not having to get to work.

8 8 savings (federal and state) and total savings to the state of Georgia alone, which accounts for 32.5 percent of total Medicaid spending. An important step when examining nursing home consumption is how individuals will intentionally shield, gift or spend down assets in order to subsidize their nursing home care with Medicaid. In general, people will intentionally spend down savings to receive state aid earlier, which we account for throughout our analysis. However, we believe this effect may be greatly increased in regard to nursing home and home care services due to the high out-of-pocket costs to individuals who pay for their own care. Also, the combination of few private insurers willing to cover nursing homes, and the increased rates for older adults who are covered under private insurance makes it unlikely that a low-income retiree s nursing home will be covered by a private insurer. In 2015, the median out-of-pocket nursing home expense for Georgians was $71,135 per year. 10 Considering that the highest aggregate pre-retirement savings among all groups was $23,412, we believe that the spend-down effect will have significant implications for state savings. It is still possible that people will delay nursing home residency because of savings or the time it takes to spend down savings. We estimate that between 40 and 80 percent of people will intentionally shield, gift or spend down saving to have Medicaid pay for nursing home care. Table 4 reflects the high and low bounds for both total government and state-only savings for each non-zero year of Medicaid nursing home expenses between 2018 and Table 4. Medicaid Savings for Nursing Home Care from a Work and Save Plan ($ millions) YEAR LOWER BOUND OF SHIELDING SAVINGS TOTAL GOVERNMENT UPPER BOUND OF SHIELDING SAVINGS LOWER BOUND OF SHIELDING SAVINGS STATE OF GEORGIA UPPER BOUND OF SHIELDING SAVINGS Total Note: This chart displays non-zero years. During , the expected government savings are zero. A Work and Save Plan that increases retirement savings by 3 percent of income would create between $71.7 and $143.4 million in government Medicaid savings from nursing home expenses between 2018 and Over this period, Georgia would save between $23.3 and $46.6 million on nursing homes. Note that each year the government savings will increase from the previous year because people have longer to accumulate savings before retirement, resulting in more people being above the $2,000 asset 10 Genworth, Cost of Care Survey, Georgia, 2015.

9 threshold and remaining above the threshold for longer during retirement. The state savings per year is expected to plateau only when those entering the workforce in 2018 (after the policy change) begin to retire. Home care and hospice care was estimated using the same methodology as nursing home care. Elderly people represent a smaller proportion of total home and community care users (56.6 percent). We also suspect that the same increased spend down effect will occur for home and hospice care, as Georgia s median home health aide services cost $41,184 in We constructed the same bounds as in the nursing home estimate to account for this effect. Table 5 shows an overview of the savings from home care and hospice care. 9 Table 5. Medicaid Savings for Home and Hospice Care from a Work and Save Plan ($ millions) YEAR TOTAL GOVERNMENT SAVINGS STATE OF GEORGIA SAVINGS LOWER BOUND UPPER BOUND LOWER BOUND UPPER BOUND Total Note: This chart displays non-zero years. During , the expected government savings are zero. Total government savings from the policy change is estimated to be between $38.5 and $77.0 million for the years , with Georgia s savings expected to be between $12.5 and $25.0 million over that period. Supplemental Security Income SSI is an aid program for disabled and low-income elderly individuals. The federal government is expected to spend $480 million on SSI for Georgians over age 65 from 2018 to SSI has the same $2,000 countable assets limit as Medicaid. The estimate for SSI was conducted similarly to Medicaid: 1. We first estimated the average cost and total number of people expected to receive payments from SSI from 2018 to We then projected the number of people who would retire and be affected by the policy change. 3. We estimated income, savings and retirement consumption. 11 Ibid.

10 4. We used the difference between retirement income and retirement consumption to spend down savings. 5. The amount of time each qualified individual was not eligible for SSI because they exceeded the $2,000 asset threshold was estimated for each year and each income category. 6. To determine state savings, we calculated the total amount of time spent unenrolled from SSI multiplied by the average yearly state expense for each SSI enrollee. The analyses of Medicaid and SSI have two notable differences. First, the income distribution used for retirees is different. For the Medicaid estimate, an income distribution between zero and the 60th percentile was used. SSI has a significantly lower unearned income threshold than Medicaid at $8,820 per year. Therefore, many people who are eligible for Medicaid would not be eligible for SSI. To account for the smaller distribution, we use four income groupings between zero and 125 percent of the federal poverty level. Social Security payments to this group of retirees would still allow them to be eligible for SSI. The second significant difference is how we controlled for the spending down of savings. Benefits from SSI are relatively small compared to some benefits from Medicaid (i.e., nursing and home care). Therefore, we estimated that people would spend down savings that were close to the $2,000 threshold (the exact amount is dependent on the income group); however, people are unlikely to immediately spend down savings in order to qualify for SSI. Table 6 shows the total amount of governmental savings on SSI for every non-zero year between 2018 and Table 6. SSI Savings from a Work and Save Plan ($) YEAR FEDERAL GOVERNMENT SAVINGS , ,492, ,293, ,813, ,729, ,395, ,583, ,249,061 Total 35,247,437 Note: This chart displays non-zero years. During , the expected government savings are zero. The total amount of savings for SSI between 2018 and 2030 is $35.2 million. This number represents savings for the federal government, not for the state of Georgia, as SSI is 100 percent federally funded. Supplemental Nutrition Assistance Program The Supplemental Nutrition Assistance Program (SNAP) boosts the food budget for low-income individuals. SNAP s countable asset threshold is $3,250 for households with an individual over the age of

11 60. SNAP does not count most forms of retirement savings against the countable asset threshold. Assuming that the Work and Save Plan is structured to invest assets in an exempted retirement savings plan, state savings from SNAP would be zero. This is because the savings accrued from the Work and Save Plan would not affect the eligibility of any SNAP participants. However, if Work and Save Plan savings were invested in a non-exempt form, savings will count against the SNAP asset threshold. To estimate state savings in this scenario we used similar methodology as our SSI estimate with the exception of having a higher asset threshold and a wider income distribution to estimate savings. We expect all people who qualify for SNAP to have pre-retirement income between zero and 200 percent of the federal poverty line (as opposed to zero and 125 percent for SSI). This wider distribution reflects wider income eligibility requirements between SNAP and SSI. Table 7 summarizes the findings for every non-zero year. Note that any savings greater than zero in this scenario is contingent on the Work and Save Plan participants investing in a non-exempt countable asset. If Work and Save Plan participants invest in an exempt countable asset, state savings would be zero, thus our lower bound estimate for savings from SNAP is zero. 11 Table 7. SNAP Savings from a Work and Save Plan ($) YEAR FEDERAL GOVERNMENT SAVINGS ,490, ,056, ,827, ,329, ,181, ,943, ,839,413 Total 140,666,949 Notes: The lower bound for total state savings from SNAP is zero. This chart displays non-zero years. During , the expected government savings are zero. Conclusion If a Work and Save Plan in Georgia were successful and the retiring poor saved 3 percent of their income before retirement, many older individuals would delay their dependence on state support, resulting in significant savings for the state. Total government savings from Medicaid, SSI and SNAP for the years are estimated to be over $145 million. Savings for the state of Georgia are estimated to be between $35.8 and $71.6 million. The total savings per year is expected to continue to increase beyond 2030 as people can accumulate savings over a longer duration of their careers. A summary of total government and state of Georgia savings for non-zero years is found in Table 8.

12 12 Table 8. Summary of Government Savings from a Work and Save Plan ($ thousands) YEAR SSI SNAP MEDICAID TOTAL GOVERNMENT SAVINGS LOWER UPPER LOWER UPPER BOUND BOUND BOUND BOUND STATE OF GEORGIA SAVINGS LOWER UPPER BOUND BOUND ,493 7,490 7,061 14,122 8,554 23,105 2,295 4, ,294 10,056 12,861 25,722 15,155 38,072 4,180 8, ,813 12,828 13,324 26,648 17,137 43,289 4,330 8, ,730 18,329 15,542 31,084 20,272 54,143 5,051 10, ,396 19,181 18,309 36,617 24,704 62,194 5,950 11, ,583 34,943 19,505 39,009 26,088 80,536 6,339 12, ,249 37,839 23,558 47,117 32,807 94,205 7,656 15,313 Total 35, , , , , ,234 35,802 71,604 Notes: The lower bound for SNAP and SSI is zero, reflected in the total government savings. This chart displays non-zero years. During , the expected government savings are zero. About the Center for State and Local Finance The Center for State and Local Finance s (CSLF) mission is to develop the people and ideas for next generation public finance by bringing together the Andrew Young School s nationally-ranked faculty and the broader public finance community. CSLF conducts innovative, nonpartisan research on tax policy and reform, budget and financial management, education finance, and economic development and urban policy. Additionally, it provides premier executive education in public finance for state and local finance officials and works with local and state partners on technical assistance projects on fiscal and economic policy. CSLF maintains a position of neutrality on public policy issues. However, in order to protect their academic freedom, authors may express a wide range of viewpoints in CSLF s publications. The research, interpretations or conclusions in CSLF publications should be understood to be solely those of the author(s). For more information on the Center for State and Local Finance, visit our website at:.