October 16, 2008, Number 8-15 THE IMPACT OF INFLATION ON SOCIAL SECURITY BENEFITS By Alicia H. Munnell and Dan Muldoon* Introduction Today, the Social Security Administration announced for joint returns) above which taxes are levied are not that benefits payable in December 2008 would be adjusted for wage growth or even for inflation, rising increased 5.8 percent beginning January 1, 2009. benefit levels mean that taxation reaches further and This cost-of-living-adjustment (COLA) – the largest further down the income distribution. in 26 years – is an important reminder that keeping This brief explores the interaction of inflation and pace with inflation is one of the attributes that makes Social Security benefits. The first section describes Social Security benefits such a unique source of the nature of the cost-of-living adjustment. The sec- income. (The other is that the payments continue for ond section looks at the interaction of Medicare pre- life.) Higher inflation raises two other issues, how- miums and the cost-of-living adjustment. The third ever, that diminish the impact of the COLA. The first section explores how inflation affects the taxation of issue pertains to Medicare Part B premiums, which benefits. The final section concludes. are deducted automatically from Social Security The overall finding is that, while the inflation benefits. To the extent that premium costs rise faster adjustment in Social Security is extremely valuable, than the COLA, the net benefit will not keep pace with the rise in Medicare premiums and the extension of inflation. Historically, premiums have gone up much taxation under the personal income tax mitigate the faster than the COLA, although this year is an excep- ability of beneficiaries to maintain their purchasing tion as premiums for 2009 will be unchanged from power. This erosion of retiree purchasing power is their current level.1 The second issue pertains to serious given that virtually all other sources of retire- taxation under the personal income tax. Because the ment income have no inflation protection at all. thresholds ($25,000 for single taxpayers and $32,000 * Alicia H. Munnell is the Director of the Center for Retirement Research at Boston College (CRR) and the Peter F. Drucker Professor of Management Sciences at Boston College’s Carroll School of Management. Dan Muldoon is a research associate at the CRR. The authors wish to thank Stephen Goss for helpful comments. 2 Center for Retirement Research Social Security’s Cost-of- Figure 1. Social Security Cost-of-Living Adjustment, 1980-2008 Living Adjustment 16% Social Security benefits are subject each year to a cost-of-living adjustment (COLA).2 The adjustment, which is based on the change in the Consumer Price 12% Index for Urban Wage Earners and Clerical Work- ers (CPI-W) over the last year, protects beneficiaries 8% against the effects of inflation. Since the COLA first affects benefits paid after January 1, Social Security needs to have figures avail- 4% able before the end of the year. As a result, the adjust- ment for January 1, 2009 is based on the increase in 0% the CPI for the third quarter of 2008 over the third 1980 1985 1990 1995 2000 2005 quarter of 2007. This calculation produces a COLA of 5.8 percent, the largest since 1982 (see Figure 1). Automatic indexing is generally viewed as a posi- Sources: U.S. Social Security Administration (2008a); and tive feature of social security systems, both in the U.S. Bureau of Labor Statistics (2008). United States and abroad. Without such automatic adjustments, the government would have to make fre- quent changes to benefits to prevent retirees’ standard of living from eroding as they age.3 Box 1. What’s the Right percent versus 30.2 percent), and the cost of these items rose faster than prices in general.5 Price Index? The results from the CPI-E need to be interpreted with caution, because it is not constructed from The only controversy that sometimes arises with scratch but rather is derived from the CPI-U. As a respect to the adjustment is the nature of the index result, it suffers from several flaws. First, a relatively used. Social Security bases its adjustment on the small number of households is used to determine Consumer Price Index for Urban Wage Earners and the expenditure patterns, so the weights are subject Clerical Workers (CPI-W), which is a subset of the to much greater sampling error than those in the population covered by the Consumer Price Index for broad index. Second, prices are based on the same All Urban Consumers (CPI-U).4 Some contend that geographic areas and retail outlets used by younger since the CPI-W does not represent the price changes people and may not be representative of the location faced by retirees, another index should be used. In and types of stores frequented by the older popula- 1987, Congress directed the Bureau of Labor Statis- tion. Third, the items sampled may not be the same tics to calculate a separate price index for the elderly as those bought by the elderly. Finally, the prices (persons 62 and older). This index, called the CPI-E, used are the same as those reported for younger has been constructed back to 1982. Over the period people and do not reflect any type of senior discounts. 1982-2007, the average annual increase for the CPI-E Thus, if the decision were made to employ an index was 3.3 percent, compared to 3.0 percent for the CPI- for the elderly, a new index would be needed that W. The index rose faster for older Americans than used a larger sample of older households and gath- for their younger counterparts because older people ered prices for the products that older people buy at devote a larger share of their budgets to medical care the places they shop. (10.8 percent versus 5.3 percent) and shelter (36.9 Issue in Brief 3 How Medicare Premiums COLA is 5.8 percent while the Medicare Part B pre- mium remains unchanged). As a result, the benefi- Affect the Outcome ciary would receive a net benefit of $917.20 ($1027- $109.80), or 1.9 percent more than the original $900. Medicare premiums for Part B (physician and out- Thus, the increase in the Medicare premium offsets patient services) and Part D (prescription drugs) are some of the cost-of-living adjustment on the net deducted from Social Security benefits before they are benefit. sent to the recipient.6 Typically, the Medicare Part B Figure 3 shows what happens if this process were premium is increased each year in line with Part B to continue for 30 years. During the first ten years, per capita expenditures.7 the $900 net benefit grows at an average annual rate Although the 2009 Part B premium will remain of 1.6 percent. During the second ten years, the net unchanged from its 2008 level, over the past three benefit would continue to grow until year 16 and then decades the average annual adjustment for the Part would start to decline, and the decline would continue B premium has been 9.0 percent compared to an throughout the third ten-year period.8 Legislation, average annual Social Security COLA of 3.8 percent; however, prevents nominal benefits from declining. since 2000, the comparable numbers have been 9.8 Specifically, section 1839(f) of the Act ensures that percent and 2.7 percent (see Figure 2). the dollar value increases in the premium for current beneficiaries do not exceed the dollar value of the an- Figure 2. Average Social Security Cost-of- Living nual Social Security cost-of-living adjustment. Adjustment and Annual Increase in Medicare Part B Premium, 1980-2007 and 2000-2007 Figure 3. Hypothetical Growth of OASI Benefit 10% 9.8% of $1000 and Part B Premium of $100 and Path 9.0% of Net Benefit, with and without Protective 8% Legislation $2,500 $2,255 6% 3.8% $2,000 4% $1,670 2.7% $1,500 2% $1,000 $1,094 $1,000 0% $900 1980-2007 2000-2007 $500 $584 Medicare Part B Premium $100 SSA COLA $0 0 5 10 15 20 25 30 Sources: U.S. Social Security Administration (2008a); and Centers for Medicare and Medicaid Services (2008b). OASI benefit: 2.7% Effective benefit Part B premium: 9.8% Without protective legislation Sources: Authors’ calculations based on U.S. Social Security To see the impact of the rapidly rising Medicare Administration (2008a); and Centers for Medicare and premium, assume that the average benefit is about Medicaid Services (2008b). $1,000 per month and the Medicare Part B premium is $100, so the beneficiary receives a net benefit of $900 to spend on non-health items, such as food, Figure 4 on the next page shows what the change shelter, and clothing. Assuming the average annual in the net benefit would be with and without the increases in the two amounts over the period 2000- protective provision. Although the provision prevents 2007, in 2008 the benefit would increase to $1027 dollar declines in the net benefits, beneficiaries still and the Medicare premium to $109.80. (As stated experience erosion in the real purchasing power of before, 2008 is unusual in that the Social Security their net benefit. 4 Center for Retirement Research Figure 4. Hypothetical Annual Growth Rates of Net OASI Benefits over a 30-Year Horizon The Impact of Taxes on Social Security Benefits 5% With legislation The other way that inflation affects Social Security 3% Without legislation benefits is through the federal personal income tax. 1.6% 1.6% 1% 0.4% 0.1% Under current law, individuals with less than $25,000 0.0% and married couples filing jointly with less than -1% $32,000 of “combined income” do not have to pay taxes on their Social Security benefits (see Table 2 on -3% the next page).9 Above those thresholds, recipients -5% must pay taxes on up to 85 percent of their benefits. -5.8% Today, about 30 percent of people who receive So- -7% cial Security have to pay taxes on their benefits, so the 1-10 11-20 21-30 Years beneficiary at 65 with a history of medium earnings – and thus about $15,700 of Social Security benefits Note: Figure assumes an OASI benefit of $1000 in year 0 – probably does not pay any taxes on benefits. But that grows annually at 2.7 percent and a Medicare Part B the thresholds are not indexed for growth in average premium of $100 in year 0 that grows annually at 9.8 per- wages or even for inflation, so in the future a signifi- cent. The net benefit for each year is the difference between cantly higher percentage of recipients will be subject the two. Sources: Authors’ calculations based on U.S. Social Security Administration (2008a); and Centers for Medicare and Medicaid Services (2008b). BOX 2. Income-Related Premiums for Part B increase by only $46 for an effective adjustment of Between 2007 and 2009, the Medicare program is 2.3 percent as their Medicare Part B premium went transitioning to income-related Part B premiums (see up almost $70 ($308.30-$238.40). After 2009, the Table 1). Thus while most Part B enrollees will see income-related Part B premiums will increase at the no increase in their premium in 2009, high-income same rate as the base premium and high-income peo- people will see a sizeable one. For example, persons ple will be protected from declines in their nominal with income in excess of $213,000 who had the So- benefit as a result of increases in Medicare Part B pre- cial Security maximum benefit of about $2,000 per miums. This protection will be particularly valuable month at age 65 in 2008 and received the COLA of given the large base to which the annual increases in 5.8 percent, or $116, would see their monthly benefit the Part B premium will apply. Table 1. Income-Related Premiums for Part B Income thresholds (for 2009) Year Less than $85,000 $85,000-$107,000 $107,000-$160,000 $160,000-$213,000 Over $213,000 2008 $96.40 $122.20 $160.90 $199.70 $238.40 2009 96.40 134.90 192.70 250.50 308.30 2010 and later Standard 1.4 x standard 2.0 x standard 2.6 x standard 3.2 x standard premium premium premium premium premium Note: This provision was effective in 2007. The amount of the Part B premium above the standard premium will be phased in at 33, 67, and 100 percent for 2007 to 2009 and later. The income thresholds are indexed to the CPI. The thresholds shown are for beneficiaries filing tax returns as individuals; the thresholds for couples filing jointly are twice these amounts. Sources: Centers for Medicare and Medicaid Services (2008a and 2008c). Issue in Brief 5 Table 2. Percent of Social Security Benefits Conclusion Subject to Personal Income Taxation Social Security is an extremely valuable source of “Combined” Percent Family type retirement income. It is payable for life and benefits income limits are adjusted to keep pace with inflation. The 5.8 Individual Less than 25,000 0% percent COLA announced for 2008 highlights the $25,000-$34,000 50 importance of the automatic indexing provision. Above $34,000 85 However, two factors undermine much of the in- flation protection offered by Social Security. First, in Couple Less than 32,000 0 most years, increasing Medicare premiums mean that $32,000-$44,000 50 a larger and larger chunk of the benefit goes to health Above $44,000 85 insurance, so the net benefit available for non-health expenditures does not keep pace with inflation. In Source: U.S. Social Security Administration (2008b). the future, without the protective legislation in place, benefits would actually decrease for many after retire- to tax (see Figure 5). By 2030, the Social Security ment. With the protective legislation, the nominal benefit for the worker with a history of medium earn- benefits remain at least constant, but their purchasing ings will more than double to about $34,100. Assum- power will be eroded by inflation. Second, a personal ing other income increases similarly, many medium income tax with unindexed thresholds for benefit earners will pay tax on half of their benefits. taxation means that wage growth and inflation will subject an increasing portion of the income distri- Figure 5. Percentage of Social Security bution to taxation. Taxation further reduces the net Recipients Paying Income Tax on Their Benefits, benefit that people will receive. 1980-2030 In short, even Social Security does not fully insulate older households from the erosive impact of 60% inflation and this is a serious concern given that other Historical Estimated 51.5% sources of retirement income offer virtually no infla- 50% 41.5% tion protection. 40% 31.5% 30% 23.4% 20% 12.8% 10% 0.0% 0% 1980 1990 2000 2010 2020 2030 Note: Legislation introduced in 1983 created a provision for the taxation of Social Security benefits. Estimates for 2010-2030 were linearly extrapolated based on data from 1990-2006. Sources: Authors’ calculations based on Parisi and Strudler (2007); Internal Revenue Service (2007, 2008); and U.S. Social Security Administration (2008c). 6 Center for Retirement Research Endnotes 1 This situation does not imply that total medical 8 However, this exercise is not likely to materialize costs will not increase in 2009. Under current leg- since the relationship between the Social Security islation, the Part B premium is typically set to cover COLA and increases in Part B premiums is unlikely 25 percent of the estimated average Part B per capita to remain constant over time. Rather, increases in expenditure. This year, however, a reduction in the Part B premiums and other medical costs should lead premium “margin” needed to sustain adequate levels to an increase in the COLA. in Part B’s contingency reserve has offset some of the In a perfectly indexed world, medical care’s increasing costs of Medicare, which mainly accounts relative importance in the index would rebalance for the lack of increase in the premium (Centers for each year based on how many dollars were spent on Medicare and Medicaid Services, 2008a). medical care. If medical costs continued to grow at a much faster pace than prices of other goods, medical 2 In calculating initial benefits, past earnings are care would account for a larger fraction of all goods indexed not to inflation but to past earnings in the purchased. This, in turn, would cause growth in economy so that Social Security benefits keep pace medical costs to have a larger impact on the growth with wage growth over time and the replacement rate of the index, or the price of all goods purchased. (benefits as a percent of pre-retirement earnings) Suppose we want to create an index from a situation remains stable. where a person spends a total of Yt dollars on n goods in each period t. Let Xi,t be the amount of dollars 3 Indeed, this was the case with the U.S. Social spent on good i in period t. Then good Xi,t’s relative Security program from its origin until 1975 when importance is: automatic indexing was adopted. Xi,t Xi,t n = , and growth from Yt to Yt+1 equals: 4 The CPI-U reflects the spending habits of about 87 percent of the population and the CPI-W 32 percent. ΣX i=1 i,t Yt ( Interestingly, despite its use for Social Security index- ing, the CPI-W explicitly excludes older Americans. (Stewart 2008). Yt+1 = n Xi,t+1 Σ Xi,t * Y , ( Yt i=1 Xi,t t 5 Stewart (2008). which is the sum across all goods of each good’s growth rate times its relative importance. So as the 6 Part D enrollees may elect to waive this deduction price of any good, Xi, goes up faster than other goods, and pay their premiums via other mechanisms. its relative importance increases, and it has a larger effect on the index as a whole. 7 At the inception of Medicare in 1966, the Part B premium was set to cover 50 percent of the per capita 9 Combined income is adjusted gross income as re- costs of the program. Legislation in 1972 linked ported on tax forms plus nontaxable interest income increases in the Part B premium to Social Security’s plus one half of Social Security benefits. annual cost-of-living adjustment. In several years during the 1980s, Congress overruled this legislation and voted to make the Part B premium 25 percent of the per capita costs of the program. In the early 1990s, the Omnibus Budget Reconciliation Acts of 1990 and 1993 set the premium at 25 percent of the program’s costs through 1998. Finally, the Balanced Budget Act of 1997 permanently set the Part B pre- mium at 25 percent of the program’s per capita costs. See O’Sullivan (2004) for a more detailed history of the Part B premium. Issue in Brief 7 References Centers for Medicare and Medicaid Services. 2008a. U.S. Bureau of Labor Statistics. 2008. Consumer Price “CMS Announces Medicare Premiums, Deduct- Index: Wage and Clerical Workers. Washington, DC. ibles for 2009.” Fact Sheet from Centers for Medicare and Medicaid Services Media Release U.S. Social Security Administration. 2008a. Annual Database. Washington, DC: Department of Health Report of the Board of Trustees of the Federal Old- and Human Services. Age and Survivors Insurance and Federal Disability Insurance Trust Funds. Washington, DC: U.S. Centers for Medicare and Medicaid Services. 2008b. Government Printing Office. Annual Report of the Boards of Trustees of the Federal Hospital Insurance and Federal Supplementary Medi- U.S. Social Security Administration. 2008b. Social cal Insurance Trust Funds. Washington, DC: U.S. Security Retirement Benefits. SSA Publication No. Government Printing Office. 05-10035. Washington, DC: U.S. Government Printing Office. Centers for Medicare and Medicaid Services. 2008c. “Additional Information Regarding Comparisons U.S. Social Security Administration. 2008c. “Number of Beneficiary Income and Out-of-Pocket Costs of Beneficiaries Receiving Benefits on December for Medicare Supplementary Medical Insurance.” 31, 1970-2007.” Social Security Beneficiary Statis- Memorandum from Richard S. Foster and M. tics. Washington, DC. Kent Clemens. Washington, DC: Department of Health and Human Services. Internal Revenue Service. 2007. Statistics of Income – 2005 Individual Income Tax Returns. Washington, DC: U.S. Government Printing Office. Internal Revenue Service. 2008. Statistics of Income – 2006 Individual Income Tax Returns. Washington, DC: U.S. Government Printing Office. O’Sullivan, Jennifer. 2004. “Medicare: Part B Premi- ums.” Congressional Research Service Report for Congress No. RL32582. Washington, DC: Library of Congress. Parisi, Michael and Michael Strudler. 2007. “Selected Income and Tax Items from Inflation-Indexed Individual Tax Returns, 1990-2004, Tables 1a-1o.” Statistics on Income Bulletin. Washington, DC: Internal Revenue Service. Stewart, Kenneth J. 2008. “The Experimental Con- sumer Price Index for Elderly Americans (CPI-E): 1982-2007.” Monthly Labor Review 131(4): 19-24. About the Center Affiliated Institutions The Center for Retirement Research at Boston Col- American Enterprise Institute lege was established in 1998 through a grant from the The Brookings Institution Social Security Administration. The Center’s mission Massachusetts Institute of Technology is to produce first-class research and forge a strong Syracuse University link between the academic community and decision Urban Institute makers in the public and private sectors around an issue of critical importance to the nation’s future. To achieve this mission, the Center sponsors a wide Contact Information Center for Retirement Research variety of research projects, transmits new findings to Boston College a broad audience, trains new scholars, and broadens Hovey House access to valuable data sources. Since its inception, 140 Commonwealth Avenue the Center has established a reputation as an authori- Chestnut Hill, MA 02467-3808 tative source of information on all major aspects of Phone: (617) 552-1762 the retirement income debate. Fax: (617) 552-0191 E-mail: crr@bc.edu Website: http://www.bc.edu/crr The Center for Retirement Research thanks AARP, AIM Investments, Bank of America, CitiStreet, Deloitte Consulting LLP, ING, John Hancock, MetLife, Nationwide Mutual Insurance Company, Prudential Financial, State Street, TIAA-CREF Institute, and T. Rowe Price for support of this project. © 2008, by Trustees of Boston College, Center for Retire- The research reported herein was pursuant to a grant from ment Research. All rights reserved. Short sections of text, the Center’s Partnership Program. The findings and not to exceed two paragraphs, may be quoted without ex- conclusions expressed are solely those of the authors and plicit permission provided that the authors are identified and do not represent the views of the partners or the Center for full credit, including copyright notice, is given to Trustees of Retirement Research at Boston College. Boston College, Center for Retirement Research.