expectancies gender. Six f the portion pa ignored. F for fem irticipating inc ve-year birth ales at or exam cohorts are sim retirem rease ple, an individ s to 48.3 percent (49. ent age, this can amount t ulated. ual ent The oldest grou ering the wor 4 percent for male o an appreciable decrease in retirement k force at age 21 and retiri p was born in t s and 47.2 percent for he period 1 ng at 65 93fe 6 to 1 ma coul lT-145 es). 940d Stochastic Retirement Expenditures Introduction 4 8 SSA states that the median earnings of full-time women workers in 2002 was $30,203 compared with $39,429 for When separat inclusive (currently ages 66 to 7 income. have left the work force for as ing these wage and salary 0), whil e the youn workers into many as nine fu gest grou the public and private sectors, the percentages ll p was born in t years without bei he period 1 ng penalized 961 towith respect to 1965 The second component of health expenditures is the result of simulated health events that would Although the Social Security retirement benefit program was put in place more than seven men; however, they also note that this disparity has closed substantially in the past four decades during which period (currently participating age this calculation. differ significantly. Slightly less than s 41 to 45), inclusive. In addition, the re 76 percent (79.9 percent for male lative income was reported by estimating lifetime s and 72.8 percent require long-term care in a nursing home or home-based setting for the elderly. Neither of these decades ago and the Employee Retirement Income Security Act (ERISA) was enacted more than three today's retirees would have generated their working histories necessary to calculate Social Security benefits. for fe income quartiles (fro ma • le s) The form of the public-sector worker m ula 2002 though retir used to convert the workers’ av s participat ement age) fo ed in an em r each of the co erage ind ployment-based r exed earnings to a m mbinations of birth cohort and etirement plan, com onthly retirement pared Written Statement simulated types of care would be reimbursed by Medicare because they would be for custodial (not decades ago, it appears there will still be substantial numbers of future retirees who will struggle with 5 Private pensions include survivor, disability, and retirement income pensions from corporate or union sponsors. In gender. with 43.0 percent (44.7 percent for benefit is heavily skewed toward lower-income wo males and 41.0 percent for fe rkers. For exam males) of the private-sector ple, if 2006 is the workers. year in 1 rehabilitative) care. Retirement Income Adequacy retirement security. One recent study by the Employee Benefit Research Institute (EBRI) found that addition, regular payments from individual retirement accounts (IRAs), Keoghs, and 401(k)-type accounts are It is im A mwhich the worker is first el ore restri portanctive definition of the work fo t to note that within each of igible for benef the groups m rce, which m its, his or her initial m odel ore closely ed there will undoubtedly be resem onthly b benefit at les the types of normal For determining whether an individual has these for the expenses, the following process is undertaken. Am erican reti While the previous m rees will have at least $45 billion less in aterial documented the com retirement income in 2030 t ponent parts of the accum hu an what they will need lation process, the included in private pensions. Public pensions include payments from survivor, disability, and retirement income Figure 1 workers who generally significant percentages in the zero categ retirement age would be equal to the sum must be covered by ory ERISA , as well as for a retire of: those at levels far higher than m ment plan offered by a private-sector ost individuals An individual reaching the Social Security normal retirement age has a probability of being in one of four to cover basic expenditures and any real question from a public policy perspective expense associ may be ated whether current and future with an episode of care in a nursing home or fro retirees will be able to m pensions from federal, U.S. military, and state or local sponsors. Other pension income includes survivor payments Importance of Social Security for Unmarried em could possibly save. These situations are accounted fo ployer or o union 90 percent of , is the work force of full-tim the first $656 of average indexed m e, full-y r in two wa ear wage and salary onthly earnings, pl ys: First, wor medi kans are report ers ages 21–64. us ed for each Figure 4 possible assumed “health” conditions, based on the estimates of the use of each type of care from the a hom afford an adequate standard of livi e health care provider. The aggregate defic U.S. Senate Special Committee on Aging ng in retirement. it in retiree inco Significant wo me during t rk has been d he decade ending 2030 will one by EBRI and others from U.S. railroad retirement, workers’ compensation, Black Lung, regular payments from estates, trusts, annuities, Approxim of the groups; in other words, the num atel o y 57 32 percent of percent (55.4 the next $3 Individuals Age 62 and Older, 2003 percent for males and bers presente ,299 of average indexed m d in Figures 8 a 58.2 percent for fe nd 9 onthl provide y m earnings, ales) of these worke a num plus ber representing the rs surveys above and mortalit Distribution of W y: orkers, by Earnings or life insurance, and other survivor payments; disability payments from U.S. railroad retirement, accidental or be at least $400 billi to evaluate how much workers will need in or on. der to have the same after-tax an d after-savings amount for th th participated in a retirem estimate for the 75 o 15 percent of or 90 ent percentile when ranked by percentage plan. any average indexed monthly earnings in excess of $3,955. of compensation. Second, the reported 9 disability in • suran Not receiving either hom ce, Black Lung, workers’ co e health or nurs mpensation, istate ng hom teme p o car rary sick e. ness, and other disability payments; consumption in retirem While these num ent ber that the s are tr youbling in aggregat enjoyed prior to r ee, drilling down to s tirement. Althougee what the potential h this may be desirable from and Work Status, by Gender values are li • These gaps were significan mited to 25 percent of co In addition to retirement benefits tly larger in t mpensation unde based directly he late r the assu 1980s. For exam on one mption that few, 's own working history, an additional ple, in 1987, 40.7 percent of if any, family units Gender Distribution of Total Income Quartiles and retirement payments from U.S. railroad, regular payments from annuities or paid-up insurance policies, and • Home health care patient. consequences of toda the standpoint of financial y's programs planning m , it s ight be for certa ets a goal that in seg maym be u ents of the retiree population in nrealistically high for many segments of the future, Hearing on would be abl female Male wage benefit may e and salary to contribut workers ages 21-64 particip be available eq e in excess ofual to 50 percent of this peFemale rcentage on a conti ated in an em the spouse's benefit less the benefi pl nuous oy mebasis until retirem nt-base d retirement enplan t age. t the worker other retirement payments. These three sources of pension income are combined to determine the percentage of • Nursing home care patient. especially the population. Another st single females, shows that achieving retirem andard is used in this analysis to ent security assess the current state of the retire on a nearly universal basis will be a ment Overall Men Women st compared wit It is also i would have been entitled to based h 51.0 percent for m mportant to note athat these percen les. The gap decreased fro on his or her own tages mere m ly ju represent earnings. st over 10 per savings that need to be centage points to under 4 those 65 or older with pension income. 1 (lowest) Income Quartile 21.6% 78.4% • Death. particularly system. daunting challenge. Bridging the Gender Gap: Eliminating Retirement Income Disparity for Women Annual Earnings 6 generated percentage points. Further in addition to what retire more, while all fem ment incom ae le and/or wealth is sim wage and salary workers ages ulated by the 21–6 model. Therefore, if 4 were found to 2nd 22.2% 77.8% This analysis is based on the U.S. Census Bureau’s March 2005 Current Population Survey (CPS) and appeared in This testi Instead of attem mony begins by s pting to deu term mmari ine what zing the sources percentage of the population will of current retiree income, including Social be able to attain a Less than $5,000 12.1% 9.6% 14.9% participate in a retirement plan at a lower level than the family unit is already Social Security pla generating savings for retirement 10 ys a much larger role in males did, the percentage of full-tim total incom beyond what is inclu e for unmarried wo ded in define mee, full-y n over age 62 d benefit or ear Craig C3rd 24.8% opeland, "Employment-Based Retirement Plan P75.2% articipatio n: Geographic Differ ences and Tr ends, 2004," The individual is randomly assigned to each of these four categories with the likelihood of falling Security specified repl and em ace ploy ment ratio, ment-based retirement plans, and th this analysis attempts to model en reviewing the current literature wit what percentage of retirees will have h respect to $5,000–$9,999 8.2% 6.0% 10.5% th March 15, 2006 fem defined contribution plans, a than for their le workers who participated in a pla male counter IRAs, Social parts. Based n Security was higher on EBRI and/ estimates fro or net ho than for m usin ale m g equit s. In fact, acr the March 2005 Current Population y, that value needs to be oss all of the worker EBRI Issue Brief ,( October 2005). 4 (highest) Income Quartile 40.2% 59.8% into one of the four categories based upon the estimated probabilities of each event. If the individual does sufficient retirem gender disparities in the percentage of w ent wealth to pay for a basket of orkers with an em non-luxur ploy y goo med nst in retirem -based retire enm t for the remainder of ent plan, the $10,000–$14,999 8.8% 7.2% 10.7% 7 status categor deducted from Survey (CPS), Social Security represents ies, fe the esti males mated percentage were more likely s. to partic an averag ipate in a retire e of 65.5 percent of total income for unmar ment plan than males (Figure 2). ried 11 Craig Copeland, "Retirement Plan Participation and Retirees’ Pe rception of Their Stan dard of Livi ng, EBRI Issue not need long-term care, no stochastic expenses are incurred. Each year, the individual will again face their si participation mulated life-paths. rates of those eligible to par The expenditticipate in ures used in t a plan, and the c he model for ontribut the elderly ion le covels for those who have nsist of two $15,000–$19,999 8.5% 7.5% 9.6% Furtherm fema After the retir lore, es over age 62 but only 56.1 percent for unm when exam ement inco ining the partici me and wealth was si pation by earnings level, t m arr uie lad males ove ted for each family he proport r age 62. As seen in Figure 1, this ion unit, of females p 1,000 observations articipating Brief, January Percentag 2006. e of Total Income Social Security Represents, by Total Income Quartiles these probabilities (the probabilities of being in the different conditions will change as the individual co a plan allowing a contribution. The im mponents—deterministic (unchanging) and stoc pact of the cha onti stic (variable) nuing evoluti expense assu on from defined benefit (pension) to mptions. The $20,000–$29,999 th 17.0% th 15.6% 18.5% th th 8 in a plan was also higher than it was for male were si is largely mulated (from a function of income— retirement age until death of th when results ar s (Figur e reported by e e3). Conseque individual for incom ntly single males , it appears that fe e quartile, these differences and single females or males’ lower Mean 5 25 Median 75 95 Sheila Campbell and Alicia H. Munnell, “Sex and 401(K) Plans,” Just The Facts On Retirement Issues (No. 4, becomes older after reaching age 75 then again at age 85). This continues until death or the need for deter defined contri ministic expenses incl bution (401(k)-type) plans ude those expenses that the among private plan sponsors on future r elderly incur from a basic e need or want of daily tirees’ sources of $30,000–$39,999 14.0% 14.1% 13.8% probabilit the second person to die f decrease significantly y of participation o . r fam For in the aggregate was a resu indivi ilies), an duals i d the present va n the first (l lt of female workers’ overall lower earnings and/or owest) inco lue of the aggregated deficits at retirement age me quartile (total income of $9,199 or May 2002), Center for Retirement Research at Boston College. All long-term care. life, while the stochastic ex income is demonstrated. This evolution transfer penses in this model ar s longevit e ex clusively y risk health-event related—e.g., an adm to future retirees— and may have a ission $40,000–$49,999 9.6% 10.7% 8.4% 9 lower rates of full-tim were co less), Social mputed. At that point, the observations were Security e work in co represents an ave mparison with m rage of 79 percent of total income for women versus 75 a rank-ordered les (Figure 4). in terms of the present value of the Total 63.2% 0.0% 32.7% 72.6% 100.0% 100.0% See Jack L. VanDerhei, "Measuring Retirement Income Adequacy, Part One: Traditional Replacement Ratios and For those who have a resulting status of home health care or nursing hom e care, the duration of to a nursing home or the commen disproportionate impact on women. cement of an episode of home health care—that occur only for a th th $50,000 or more 21.9% 29.3% 13.6% deficits, and the 75 percent for men. This num Figure 5 shows the averag and 90 percentiles of the distribution were dete ber decreases significantl e employee contribution ra y for both genders in the four te to salary rmined. Next, the fut reduction plans for th (highest) quartile ure simulated Results fo Male r Workers at Large Companies, 56.1% " EBRI Notes 0.0% no. 9, Sept 22.5% ember 20055.9% 4. 99.2% 100.0% care is simulated. After the duration of care for a nursing hom e stay or episode of home health care, the portion , if ev The last secti er, during reti on of the testim rement, not o ony n ties all of this an annual basis. together by showing the results of a simulation retirement income accu nonagricultur 10 (total incom al wage and salary e of $61,398 or m mulated to retirement age was workers age 16 and ore): Social Security ov determ represents an av er. On average, ined, and the infor erage of 26.8 males co m ntributed ap ation used to determ percent of total proximately ine Female 65.5% 0.0% 37.1% 79.0% 100.0% 100.0% indivi Replacem dual will have a probabilit ent ratios typically attem y of pt to provide an indica being discharged to tion of t one of the he perce other three conditions. The ntage of income earned just prior t stochastic o model that projects whether individuals will have sufficient money in retirement to pay for basic expenses Work Status the percentage of com 0.3 percentag income for e points of co males and 27.6 pensation that wo mpensation m percent for females. uld need to ore than females in 199 be saved to have sufficient additional inc 3. This difference decreased to ome to offset retiremFirst ent thQuartile at will be replaced in retiremen t. This typica lly involves a numerator that comb ines annuity p ayments expenses incurred are then determined by the length of the stay/number of days of care times the per diem plus the potential costs of long-term health care costs not typically covered by health insurance. The Deterministic Retirement Expense Assumptions th th Full-time, full-year 66.6% 73.6% 58.8% the present value of accu approximately 0.1 percentage points of mulated deficits for the 75 compensation and in 1998 90 percentiles of the distribution. before increasing to 0.4 percentage points from Social Security and defined benefit plans with an annuitized amount from defined contribution and IRAs. The Total 78.3% 0.0% 79.1% 100.0% 100.0% 100.0% charge estimated for the nursing home care and home health care, respectively, in each region. additional amounts needed to be saved to provide a specified level of retirement security are generated to The deterministic expenses are broken down into seven categories—food, apparel and services Full-time, part-year 13.1% 13.2% 12.9% in 2003. It appears much of this differential is due to the propensity of those with larger incomes to denominator will be based on an average of final earnings just prior to retirement age. Private Pensions Male 75.4% 0.0% 57.4% 100.0% 100.0% 100.0% For any person without the need for long-term care, this process repeats annually. The process show the values needed by birth cohorts, income quartiles, and gender. (dry cleaning, haircuts), transportation, entertainment, reading and education, housing, and basic health Part-time, full-year 10.2% 6.2% 14.7% 11 contribute a larger percentage of compensation. Results Unlike m Based on EBRI esti any other models, tm he m ates f oderl used om the Mar in this an ch 2005 CPS, alysis does NOT m 34.6 percent of those age 65 or older had erely assume that a retiree will survive to Female 79.0% 0.0% 83.8% 99.9% 100.0% 100.0% repeats for individuals receiving home health care or nursing home care at the end of their duration of 7 expenditures. Each of these expenses is 5 estimated for the elderly (age 65 or older) by family size (single Part-time, part-year 10.1% 7.0% 13.6% In a similar fashion, Copeland shows the participation rate for salary reduction plans among pension inco his or her average life e me in 2004. xpectancy. The mean amount was $13 Unfortunately, results gen ,951 an eratedd the under th median was $9,60 ese assumptions wou 0. Am ld prov ong m ide t ah les e stay/care and subsequently Second Figure 8 shows the Quartile if not receiving the specia median percentage of com lized care pensation that m again at their next birthda ust be saved each y year until . Those who or couple) and family income (less than $15,000, $15,000 to $29,999, and $30,000, or more in 2002 Source: EBRI estimates from the March 2005 Current Population Survey. those eligible is higher for amount necessary to pay for retirem males (84.3 percent) than ent expenditures only fe approxim males (79.1 percent). Again, these difference ately 50 percent of the time. Instead, this msodel can 44.7 percent had pension income, as opposed to 27.0 percent for females. The mean amount for males retirement for a 75 percent Total chance that there will be 86.8% 23.5% adequate retirement inco 83.9% 99.7% me 100.0 when co% mbined with 100.0% are simulated to die, of course, are not further simulated. dollars) of the family/individual. Overview of Gender Disparities in Current Retiree Income Sources be explained considers the en in large part b tire distributioy n of the gender i possible fu ntcome disparities. ure lifetimes on a gender-specific basis and allows the concept of “Gender Disparities in Retirement Security” was $17,175 while females averaged only $10,035. There was an even larger disparity in the median simulated reti Male rement wealth, assum84.9% ing current Social S 0.0% 83.9% ecurity benefits and that 99.9% housing equity is never 100.0% 100.0% As with the basic health care expenses, the qualification of Medicaid by income and asset levels The estimates are derived from the 2000 Consumer Expenditure Survey (CES) conducted by the longevity risk to be explicitly modeled. Social Security amounts: $12,012 for males and $6,600 for females. liquidated. F is considered to see how much of the stochastic expe Female or example, both gend87.3% ers in the first tw 34.5% 83.8% 99.6% o incom nses mue st be c quartovered by iles for the oldest birth cohort are at the individual to deter 100.0% 100.0% mine 12 2 Bureau of Labor Statistics of the U.S. Department of Labor. The survey targets the total The Nort Based on num heast region incl bers published by udes the states of Maine, New the Social Security Hampshire, Adm Ve inistration, rmont, Massach at the end of 200 usetts, Rhode 3, wo Islandmen , 's By the 25 percent of compensation threshold. For those in the highest income quartile for this birth cohort, the individual’Third Quartile s final expenditures for the care. Only those expenditures attributable to the noninstitutionalized population (urban and rural) of the United States and is the basic source of data for Connecticut, New York, New Jersey, and Pennsylvania. The Midwest region includes the states of Ohio, Indiana, average monthly Social Security retirement benefit was $798, compared with $1,039 for their male Jack VanDerhei the percentages of additional annual com Total 62.9% pensation need 0.0% ed to be saved are 23.8 percent for singe fem 47.3% 65.8% 86.5% 100.0% ales individual ?not the Medicaid program ?are considered as expenses to the individual and as a result are 3 revising the items and weights in the market basket of consumer purchases to be priced for the Consumer Illinois, Michigan, Wisconsin, Minnesota, Iowa, Missouri, North Dakota, South Dakota, Nebraska, and Kansas. The counterparts. Given the gender-neutral calculation of retirement benefits under Social Security, this Temple University and 13. Increasing Importance of Indi 9 percent for single Male males. 64.0% vidual Account Plans for Future 0.0% 49.2% 66.0% 91.2% 100.0% included any of the “deficit” calculations. 4 Price Index. South region in CES data provide de cludes the states of Delaware, Maryland tailed data on expen , District o ditures and in f Columco bia, Virg me of consumers, as inia, West Virginwel ia, No l as the rth Gender Dispariti disparity is largely due to te he differences s in Retirem in average earnings prior ent Program Parti to retirem cent ipation Among and the number of years Figure 9 shows the additional savings required to provide retirement adequacy in 9 out of 10 Female 62.6% 0.0% 46.7% and 65.7% 85.3% 100.0% Carolina, South Carolina, Georgia, Florida, Kentucky, Tennessee, Alabama, Mississippi, Arkansas, Louisiana, de Reti mographic characte rement Income Secu ristics of those consu rity mers. The survey does not provide state estimates, but it does the individual worked prior to receiving benefits. However, the current calculation method used for most simulated life paths. In this case, all the medians for both genders in the first three income quartiles are at Fourth Quartile Current Workers Total Retirement Expenses Research Directo r of the Employee Benefit Research Institute Fellows Program Oklahoma, and Texas; while the West region includes the states of Montana, Idaho, Wyoming, Colorado, New provide regional estimates. Thus, the estimates are broken down into four regions—Northeast, Midwest, The increasing importance of this shift from defined benefit to defined contribution retirement workers provides that the full disparity between earnings and number of years worked prior to retirement the threshold. Total Those in the highest income quartile 27.3% 0.0% for this birt4.7% 26.8% h cohort all have requirem 42.8% ents that would 64.2% 12 Mexico, Arizona, Utah, Nevada, Washington, Oregon, California, Alaska, and Hawaii. Although there is a substantial amount of gender disparity am ong those already age 65 with The elderly individual or families’ expenses are then the sum of their assumed deterministic South, and West—to account for the differences in the cost of living across various parts of the country. plans can be seen in Figures 6 and 7. Figure 6 provides the composition of estimated retirement wealth is mitigated to a certain extent: prove difficult if not im 13 possible to implement: Single females are estimated to now need to save more Male 26.8% 0.0% 6.1% 25.8% 43.2% 62.0% respect to receiving pension incom For more detail see Jack L. VanDerhei e, it appears this and Craig Copedisparity land, “Ca will n Amdecrease sharp erica Afford Tomo ly and for some worker rrow's Retirees: Results expenses based upon their demographic characteristics plus any simulated stochastic expenses that they Consequently, an expense value is calculated using actual experience of the elderly for each region, for males at Social Security normal retirement age, by birth cohort. Similar figures for females are • The value of the worker's career earnings used to calculate the monthly benefits for most 6 than 25 percent of compensation and single males 22.1 percent of compensation. Given that most Female 27.6% 0.0% 3.1% 27.1% 42.6% 65.2% types will act From the EBRI-ERF Retiremen ually reverse. As a case in point, C t Security Projection M opeland odel,” EBR analy I Iss zes 2004 em ue Brief no.263 ploy , N m ov ee nt-based participation mber 2003. may have incurred. In each subsequent year of life, the total expenditures are again calculated in this family size, and income level by averaging the observed expenses for the elderly within each category provided in Figure 7. It is readily apparent from these graphs that both genders have an appreciable drop workers ignores any earnings greater than the maximum taxable wage base in that year. For individuals would be unlikely to choose a situation that would provide them with adequate retirement 14 Source: EBRI estimates from the March 2005 Current Population Survey. levels from the March 2005 CPS and finds that among the 152.7 million Americans who worked in 2004, manner. The base y The 2003 OASDI Tru ear’ stees rep s expenditure ort subsequ value esti ently redu mced ates th excluding the health care e e assumed general inflation rate to xpenses are a 3.0 percen djusted t. The meeting the above criteria. The housing expenses are further broken down by whether the elderly own or in the percentage of private retirement income that is attributable to defined benefit plans (other than cash example, in 2006 any earnings greater than $94,200 would be not be subject to Social income only 50 percent of the time, this analysis focuses only on the 75 percent and 90 percent 13 81.2 m actuaries at t illion he worked for an em Center for Medicare ploy & Medi er or union that s caid Services ponsored a pe developed a per nsion or retir sonal health ca em re c ent plan, and 63.9 hain-type index that is a annually using the assumed general inflation rate of 3.3 percent from the 2001 OASDI Trustees Report, rent their home. The basic health expenditure category has additional data needs beyond the CES. balance). Females start with a slightly higher defined benefit concentration than men (49.7 percent vs. 16 Security payroll tax and thus would not be used in calculating the Social Security benefits for confidence levels. composite index of health care prices in the overall health care economy, which they predict will rise at a 3.5 percent million participated in the plan. This translates into a sponsorship rate (the percentage of workers while the health care expenses are adjusted annually using the 4.0 percent medical consumer price index The total deterministic expenses for elderly individual or family are then the sum of the value in 39.0 percent for the 1936 cohort), and the difference remains fairly constant over time (37.2 percent vs. the individual. 14, 15 level annually from 2004 -2008 and 3.9 percent annually from 2009 -2012. working for an employer or union that sponsored a plan) of 53.2 percent (52.3 percent for males and 54.2 all the expense categories f that corresponds to the June o 2002 r fam - ily June 2003 size, fam level. ily income level, and region of the individual or family. 26.4 percent for the 1964 cohort). • The value of the worker's career earnings used to calculate the monthly benefits for most 15 percent for females) and a participation level (fraction of the workforce who participates in a plan While the medical consumer price index only accounts for the increases in prices of the health care services, it These expenses make up the basic annual (recurring) expenses for the individual or family. However, if These results show a clear increase in the incom e retirees will receive that will have to be workers only uses the 35 highest indexed values. This may be useful to mitigating the gender Endnotes regardless does not accof oueligibility nt for the c) h of 41.9 percent anges in the numb (42.5 percen er and/or inten t for males and sity of services ob 41.2 tain percent for fem ed. Thus, with in ales). However, creased the individual or family meet the income and asset tests for Medicaid, Medicaid is assumed to cover the Comparison of Retirement Income and Retirement Expenses managed by the retiree. This makes the risk of longevity more central to retirees’ expenditure decisions. gap that otherwise would exist in two ways. First, the amount of earnings (below the longevity, the rate of health care expenditure growth will be significantly higher than the 4.0 percent medical this measure of the work force contains the unincorporated self-employed and those typically with a basic health c 1 are expenses (both parts), not those of the individual or family. Furthermore, Part B Therefore, they will have to understand that life expectancies are merely averages, and that wide variation The prim maxim ary um objective of t taxable wage ba his analy se) is indexed for sis is to combthe increase in average ine the simulated retirement national wages between income and wealth Jack L. VanDerhei and Craig Copeland, , “Can America Afford Tomorrow's Retirees: Results From the EBRI-ERF inflation rate, as has been the case in recent years. looser connection to the work force—individuals under age 21 and older than age 64. Therefore, a premium relief for the low-income elderly (not qualifying for Medicaid) is also incorporated. beyond the average is possible. Moreover, as the percentage of overall retirement income derived from Retire with the sim ment Secu the time of the earnings and the tim ulated retiree rity Projection ex Mod penditures el,” EBR to determ I Issue Brief e the ine h indivi no.26 ow m 3, N dual reaches age 60. As a result, if a worker uch eac ovemb h fa er 2m 003 ily . unit would need to save today 16 The views expressed in this statement are solely those of Dallas L. Salisbury and should not be attributed to the Employee Benefit Research For additional detail on how these findings differ by assumptions for the use of housing equity to pay retirement different measure of the work force is examined: wage and salary workers ages 21–64, representing defined benefit plans decreases, f 2 emales desiring longevity insurance in the form of an annuity will face (as percentag Institute (EBRI), t leaves the wo eh of their current wages) to maintain e EBRI Education and Re rk force early (perhaps to take care search Fund, any of its progra a prespecified “ ms, officers, of ch trc ustees, sponsors, or other staff. The ildren or an a omfort level” g (i.e., confide ed parent), t Eh m n e previous p ce level) loyee Benefit http://socialsecurity.gov/pressoffice/factsheets/women.htm, last accessed 3/9/2006. expenses as well as alternative approaches to Social Security reform, see Jack L. VanDerhei and Craig Copeland, individuals who have a stronger connection to the work force and work for someone else. For this group, the disadvant Research Institute age of having to purchase products priced is a nonprofit, nonpartisan, education and research organization es using gende tablished in W r-distinct m ashington, DC in 1978. The ortality tables instead of testimony draws 3 earnings will not be artificially lowered due to overall wage growth in the country. Secondly, that they will be able to able to afford the simulated expenses for the remainder of the lifetime of the “Can Social Security p America Afford T rovides om bo orrow' th sus rv Re ivorsh tirees: ip Resu and dlts Fro isability b m thee EBRI-ERF nefits in additio Retire n tom retiremen ent Security Proj t benefits. See ection Model,” the sponsorship rate increases to 59.5 pe heavily from research publications of the Employee B rcent (59.0 enefit Research pe In rcent for mal stitute, but any eres and 60.1 rors or misinterpretatio percent for females) and ns are those of the witness. the implicit gender-neutral nature of the defined benefit annuity structure. Given the longer life family unit (i.e., death of second spouse i the calculation process typically n a fam allows several ily). These savings rates ar years of low (or zero) wage y e reported byears to be age cohort and EBRI Iss www.ssa. ugo e B vr fo ief r n add o.263 itio, No nal details. vember 2003. Jack VanDerhei, T Jack VanDerhei, T Jack VanDerhei, T Jack VanDerhei, T Jack VanDerhei, T Jack VanDerhei, T Jack VanDerhei, Te e e e e e emple University mple University mple University mple University mple University mple University mple University a a a a a a and EBRI Fellow, T nd EBRI Fellow, T nd EBRI Fellow, T nd EBRI Fellow, T nd EBRI Fellow, T nd EBRI Fellow, T nd EBRI Fellow, Te e e e e e estimony before stimony before stimony before stimony before stimony before stimony before stimony before Senate Aging Committee March 15, 2006 Senate Aging Committee March 15, 2006 Senate Aging Committee March 15, 2006 Senate Aging Committee March 15, 2006 Senate Aging Committee March 15, 2006 Senate Aging Committee March 15, 2006 Senate Aging Committee March 15, 2006 6 2 3 4 5 7 8 Jack VanDerhei, Temple University and EBRI Fellow, Testimony before Senate Aging Committee March 15, 2006 Jack VanDerhei, Temple University and EBRI Fellow, Testimony before Senate Aging Committee March 15, 2006 Figure 3 Figure 2 Figure Figure 7 6 Percentage of Wage and Salary Workers Ages 21–64 Who Participated in an Percentage of Wage and Salary Workers Ages 21–64 Who Participated in an Figure 5 Figure 8 Figure 9 Composition of Estimated Retirement Wealth for Males at Social Security Composition of Estimated Retirement Wealth for Males at Social Security Employment-Based Retirement Plan, by Annual Earnings and Gender, 2004 Employment-Based Retirement Plan, by Work Status and Gender, 2004 Percentage of Added Compensation That Must Be Saved Annually Until Percentage of Added Compensation That Must Be Saved Annually Until Average Employee Contribution Rate to Salary Reduction Plans, Normal Retirement Age Under Baseline Assumptions, by Birth Cohort Normal Retirement Age Under Baseline Assumptions, by Birth Cohort 70% 80% Retirement For a 75% Chance of Covering Basic Retirement Expenses Retirement For a 90% Chance of Covering Basic Retirement Expenses Nonagricultural Wage and Salary Workers Age 16 and Over 100% 74.9% 100% (assumes current Social Security and housing equity is never liquidated) (assumes current Social Security and housing equity is never liquidated) 72.8% 72.4% 7.8% 90% a 58.2% a 70% 90% 60% 65.0% 25% 0.25 65.2% 55.4% Male Male 80% 80% 60% 7.6% Female Female 50% Male 70% 70% 52.9% 20% 0.2 51.1% Female 50% 60% 7.4% 60% 40% 50% 35.4% 50% 15% 0.15 40% 35.0% 36.0% 7.2% 30.1% single female single female 40% 30% 40% 27.6% single male single male 30% 10% 0.1 30% 30% 23.5% 20.2% 7.0% 22.9% 20% 20% 15.4% 20% 20% 15.5% 14.5% 12.5% 0.05 5% 12.1% 6.8% 10% 10% 9.5% 9.7% 10% 10% 0% 0% 6.6% 0% 0 0% 0% 1993 1998 2003 123 123412341234 412341 234123 123412341234 412341 234 Less Than $5,000 $5,000–$9,999 $10,000–$14,999 $15,000–$19,999 $20,000–$29,999 $30,000–$39,999 $40,000–$49,999 $50,000 or More Full-Time, Full-Year Worker Full-Time, Part-Year Worker Part-Time, Full-Year Worker Part-Time, Part-Year Worker female-db female-dc female-IRA male-db male-dc male-IRA Per Year 1936–1940 1936–1940 1941–1945 1941–1945 1946–1950 1946–1950 1951–1955 1951–1955 1956–1960 1956–1960 1961–1965 1961–1965 Source: Employee Benefit Research Institute estimates of the April 1993 Current Population Survey employee benefit supplement and the 1996 Work Status Annual Earnings Birth Cohort/Income Quartiles Birth Cohort/Income Quartile Source: Employee Benefit Research Institute, Retirement Income Projection Model. a a and 2001 Panel of the Survey of Income and Program Participation Topical Module 7. Source: Employee Benefit Research Institute, Retirement Income Projection Model. Source: EBRI-ERF Retirement Security Projection Model. Source: EBRI-ERF Retirement Security Projection Model. 25% = 25% or more. 25% = 25% or more. Source: Employee Benefit Research Institute estimates from 2005 March Current Population Survey. Source: Employee Benefit Research Institute estimates from 2005 March Current Population Survey. Jack VanDerhei, Temple University and EBRI Fellow, Testimony before Senate Aging Committee March 15, 2006 Jack VanDerhei, Temple University and EBRI Fellow, Testimony before Senate Aging Committee March 15, 2006 Jack VanDerhei, Temple University and EBRI Fellow, Testimony before Senate Aging Committee March 15, 2006 Jack VanDerhei, Temple University and EBRI Fellow, Testimony before Senate Aging Committee March 15, 2006 Jack VanDerhei, Temple University and EBRI Fellow, Testimony before Senate Aging Committee March 15, 2006 Jack VanDerhei, Temple University and EBRI Fellow, Testimony before Senate Aging Committee March 15, 2006 Jack VanDerhei, Temple University and EBRI Fellow, Testimony before Senate Aging Committee March 15, 2006 1936 1936 1937 1937 1938 1938 1939 1939 1940 1940 9 1 1 4 9 1 1 4 1942 1942 1943 1943 1944 1944 1945 1945 1946 1946 1947 1947 9 8 1 9 4 8 1 4 1949 1949 1950 1950 11 99 55 11 11 99 55 22 1953 1953 1954 1954 1955 1955 1956 1956 1957 1957 1958 1958 1959 1959 1960 1960 1961 9 1 1 6 1962 1962 1963 1963 9 4 1 6 1964