The full text on this page is automatically extracted from the file linked above and may contain errors and inconsistencies.
clevelandfed.org/research/workpaper/index.cfm Working Paper 94 13 THE ANNUITIZATION OF AMERICANS' RESOURCES: A COHORT ANALYSIS by Alan J. Auerbach, Jagadeesh Gokhale, Laurence J. Kotlikoff, John Sabelhaus, and David N. Weil Alan J. Auerbach is a professor of economics at the University of California, Berkeley; Jagadeesh Gokhale is an economic advisor at the Federal Reserve Bank of Cleveland; Laurence J. Kotlikoff is a professor of economics at Boston University; John Sabelhaus is a staff economist at the Urban Institute, Washington, D.C.; and David N. Weil is a professor of economics at Brown University. Mssrs. Auerbach, KotlikofT, and Weil are also associates of the National Bureau of Economic Research, Washington, D.C. The authors thank Douglas Gale and Mark Schweitzer for helpful comments, Felicitie Bell for critical data on U. S. population projections, and Jean McIntire for excellent research assistance. They are also grateful to the National Institute on Aging for financial support. Working papers of the Federal Reserve Bank of Cleveland are preliminary materials circulated to stimulate discussion and critical comment. The views stated herein are those of the authors and not necessarily those of the Federal Reserve Bank of Cleveland or of the Board of Governors of the Federal Reserve System. November 1994 clevelandfed.org/research/workpaper/index.cfm Abstract This paper constructs a unique cohort data set to study the changes since 1960 in the share of Americans' resources that are annuitized. Understanding these changes is important because the larger this share, the more cohorts are likely to consume and the less they are likely to bequeath. Hence, the degree of annuitization affects national saving as well as the transmission of inequality over time. Our findings are striking. Although the annuitized share of resources of younger Americans declined slightly between 1960 and 1990, it increased dramatically for older Americans (those age 65 or more). It doubled for older men and quadrupled for older women. Since the elderly have much higher mortality probabilities than do the young, their degree of annuitization is much more important for aggregate bequests and saving. According to our estimates, aggregate U.S. bequests would now be almost 50 percent larger had the post-1960 increase in annuitization not occurred. In addition, U.S. national saving would likely be substantially larger than is currently the case. clevelandfed.org/research/workpaper/index.cfm I. Introduction This paper constructs a unique cohort data set to examine changes since 1960 in the share of Americansf resources that are annuitized (cannot be bequeathed). Understanding these changes is important. Generations whose resources are more annuitized will consume more and bequeath less to their children and others.' This has implications for national saving as well as the intergenerational transmission of inequality. Auerbach, Kotlikoff, and Weil (1992) report dramatic increases between 1962 and 1983 in the annuitization of elderly Americansf resources. Their study relies on two cross-section surveys, the 1962 and 1983 Surveys of Consumer Finances. The nature of these data forced the authors to impute many of the future annuity streams available to survey respondents, including labor earnings, Social Security benefits, and private pension income, and to exclude from the analysis the large medical annuities provided by Medicare and Medicaid. This study takes a different approach. Rather than estimate the annuities of individual households, it considers the annuities of individual cohorts alive between 1960 and 1990. Specifically, it uses cross-section surveys to distribute to cohorts annual aggregate flows of income reported in the National Income and Product Accounts (NIPA) and other sources. Although this approach cannot address the interesting intracohort distribution issues considered in Auerbach et al. (1993), it offers a potentially more accurate and comprehensive method of assessing the overall degree of annuitization among Americans. Our findings are striking. Across all American males, the annuitized share of resources remained roughly constant between 1960 and 1990. For '1n this paper, the word "generation" refers to persons of a given sex born in the same year. clevelandfed.org/research/workpaper/index.cfm females it rose from one-third to one-half. Moreover, among elderly Americans (age 65 and over), the annuitized resource share rose from 22 to 45 percent for men and from 12 to 47 percent for women. Without this increase in the degree of annuitization, U.S. aggregate bequests would be an estimated 47 percent larger. That is, if the government were to alter the structure of Social Security benefits so as to return the degree of annuitization to its 1960 level, aggregate bequests would be almost 50 percent larger than current levels. Although the precise i~pacte x the consq&i;,ptionof the elderly of their increased annuitization is unclear, it appears to be substantial. Indeed, it appears capable of explaining a significant fraction of the decline in U.S. national saving. Section I1 provides some background to this study. It defines annuitized and nonannuitized resources, considers some general indicators of the increase in annuitization, and discusses how increased annuitization can affect national saving. Section I11 outlines the methods used for estimating the annuitized and nonannuitized components of resources. Section IV describes our data sources. Section V presents our findings and explores their implications. Finally, Section VI summarizes the results and draws conclusions. 11. Background Annuities are income flows that are contingent upon their owner's survival. Examples include Social Security benefits, private and public pension benefits, government-provided health-care benefits, and labor earnings. Government transfer payments in the form of Social Security, Some annuities are contingent on other factors as well, such as the need for medical services. clevelandfed.org/research/workpaper/index.cfm Medicare, and Medicaid constitute a significant and growing portion of U.S. annuities. These transfer payments have grown from less than 1 percent of GDP in the mid-1950s to over 9 percent today. Pension benefits have also increased faster than GDP. While pensions totaled 1.7 percent of GDP in 1960, they now total more than 5.4 percent. Not all annuities are positive. Future tax payments are examples of negative annuities. In recent decades, U.S. taxes have also grown relative to GDP. Another factor that has lowered annuitization is the reduction in annuitized labor income associated with the trend toward early retirement. In 1960, 33.1 percent of elderly males (those age 65 or more) and 10.8 percent of elderly females participated in the labor force. The corresponding 1992 percentages are 16.1 and 8.3 percent. The public can also reduce its effective degree of annuitization by purchasing life insurance. As Yaari (1965) pointed out, the purchase insurance is equivalent to the sale of an annuity. of life Cohorts that do not offset increases in their annuitization through increased life insurance purchase are likely to bequeath less and consume more than would otherwise be the case. Davies (1981), Abel (1985), and Kotlikoff et al. (1986) present simulations of the effects of introducing annuities into life-cycle economies. In their models, agents have no bequest motive and, consequently, do not offset increased annuitization by buying more life insurance. The ability to transform their net worth (or to have it transformed) into annuities permits these agents to stop worrying about outliving their resources when they are old and to consume more. Each of these studies suggests that a significant increase in annuitization will be associated with a substantial decline in both national saving and aggregate bequests, as well as a significant increase in the relative consumption-ofthe elderly. clevelandfed.org/research/workpaper/index.cfm In the United States, the increase in annuitization, which is documented below, has coincided with both a dramatic decline in national saving and a dramatic increase in the relative consumption of the elderly. Since 1980, the U.S. net national saving rate has averaged 4.1 percent, compared with 9.1 percent in the 1950s and 1960s, and 8.5 percent in the 1970s. In this decade, the net national saving rate has averaged only 2.5 percent. A comparison of the 1960-61, 1972-73, 1984-86, and 1987-90 BLS Consumer Expenditure Surveys shows an equally remarkable rise in elderly persons' relative consumption. Figure 1 presents indices of average consumption by age for each of the four periods.3 For each period, the average consumption of 40-year-olds is norrnalized to 1. As the figure indicates, the age-consumption profiles for later years are tilted upward compared to those for earlier years, indicating a rise over time in the relative consumption of the elderly. Table 1 reports the ratios of average levels of consumption of 70-year-old males and females to those of 30year-old males and females for each of the four periods. It shows that 70year-olds in 1960 consumed about two-thirds the amount consumed by 30-yearolds in 1960, whereas their consumption now exceeds that of 30-year-olds. The increase in the annuitization of the elderly is certainly not the only, nor necessarily the most important, explanation for the increase in The source for this figure as well as for table 1 is Gokhale, Kotlikoff, and Sabelhaus (1994). Their study describes their method of allocating household consumption to the adults residing in the households interviewed in the various Consumer Expenditure Surveys. It also describes their methods of allocating by age and sex those components of household consumption expenditure included in the National Income and Product Accounts but excluded from the Consumer Expenditure Surveys. Examples of such components include imputed rent and medical care. The calculated average values of consumption by age and sex used in this figure and in table 1 are benchmarked on a component-by-component basis against the National Income Accounts totals of household expenditures for the various years in question. clevelandfed.org/research/workpaper/index.cfm their relative consumption and the concomitant decline in national saving. Indeed, much of the explanation for these outcomes appears to lie in the government's massive transfers to the elderly, which have raised their incomes relative to those of young people (see Boskin, Knetter, and Kotlikoff [I9851 and Gokhale, Kotlikoff, and Sabelhaus [1994]). A Sim~leModel of the Effects of ~nnuitization~ Analysis of the steady state of the following simple two-period lifecycle model clarifies the theoretical argument that connects increased annuitization to the decline in bequests and national saving: Agents live for two periods. They work full time when young (earning W) and consume C when Y young and C, when old. Population is stationary, and the size of each cohort is normalized to unity. (1). Each agent survives to old age with probability There is no private annuities market. However, the government provides annuities by imposing a tax of T on each cohort when young and returning this amount with interest to surviving members of the cohort when old. Since there are (1-p) survivors in each cohort, each survivor receives an annuity of T(l+r)/(l-p), where r is the real interest rate. If the tax, T, does not exhaust private saving, members who die prior to their last period of life will leave a bequest. Assuming bequests are divided equally among the young, the bequest received per young person is pB, where p is the fraction of each cohort that dies before reaching old age and B is the bequest made per decedent. At the beginning of any period (before anyone has died), total wealth in the economy, K , equals the sum of private wealth of the elderly plus the This model is also presented in Auerbach et al. (1992). clevelandfed.org/research/workpaper/index.cfm wealth held by the government. The wealth held by the government is just T, the aggregate tax payments of each generation. Private wealth of the elderly can be traced to their saving when young, W+pB-T-C Y' Total wealth is just this sum plus T, so For those leaving bequests, we have For those agents who survive to old age, consumption, Co, is given by where the first term on the right-hand side of ( 3 ) represents principal plus interest on private savings, and the second term is the government's annuity payment to survivors. We close the model by assuming that agents maximize an expected, time-separable, homothetic utility function over consumption when young and old, given by where a is the time preference parameter. Maximization of utility subject to the budget constraint given in ( 3 ) implies that consumption when old is proportional to consumption when young, i.e., clevelandfed.org/research/workpaper/index.cfm where the factor of proportionality, 0, depends on a, r, and p.5 The above equations imply According to equation (6), aggregate wealth is a decreasing function of T, the amount of saving which is annuitized by the government. The intuition for this result is clear from equations (1)-(3) (2), and (5). According to (1) and raising T lowers the steady-state level of bequests as well as the steady-state capital stock, ignoring induced changes in consumption when young. If consumption when young were to fall as much as inheritances received when young (pB), aggregate wealth would remain unchanged. But, according to equations (3) and ( 5 ) , consumption when young falls by less than pB for two reasons. First, the propensity to consume when young is less than unity. Second, the annuity provided by the government increases the amount each generation can afford to consume over its lifetime because it reduces undesired bequests.6 In our model, agents have no interest in leaving bequests and, therefore, no interest in purchasing life insurance. As Yaari (1965) first demonstrated, the purchase of term life insurance is equivalent to the sale of an annuity. In this model, we are assuming that one cannot purchase annuities at the margin from private insurance companies. Allowing for such purchases would change the value of 0. Note that the reduction in aggregate wealth arising here is not, as in Feldstein (1974), the result of the government's directly transferring resources from the young to the old, but rather the result of the government's indirectly helping the old to reduce their transfers to the young. clevelandfed.org/research/workpaper/index.cfm If we modified our model to include a bequest motive and the voluntary purchase of life insurance, we would find that government annuitization of the saving of the young would simply lead them to purchase more life insurance; i.e., the annuities purchased by the government would be immediately resold. 111. Estimating Annuitized and Nonannuitized Resources In this study, we calculate the amounts of nonannuitized and annuitized resources for all male and female adult cohorts for the years 1960-1990. The components of annuitized resources are the present values of future labor earnings (human wealth), Social Security benefits, private and government employee pension benefits, government health-care benefits, welfare benefits, other government transfers, and, entering as negative annuities, the present values of future taxes. Taxes include labor and capital income taxes, indirect taxes, payroll taxes, and property and other taxes. Nonannuitized resources refer to holdings of net wealth. The computation of cohorts' nonannuitized resources for each year between 1960 and 1990 involves distributing by age and sex each year's aggregate value of household net wealth. The computation of each annuitized resource component employs a common strategy. First, for each year, the national aggregate for a particular type of payment (or receipt) is distributed by age and sex according to the cross-section, age-sex relative profile that is applicable to that payment (or receipt). For example, aggregate 1965 Social Security benefits are distributed according to the age-sex relative profile for these benefits that prevailed in 1965. This yields estimates of the per capita amounts of the payment (or receipt) by age and sex for that year. The per capita annuity values for years after 1992 are estimated by either 1) clevelandfed.org/research/workpaper/index.cfm distributing projected aggregate payments or receipts according to the latest available cross-section relative profile or 2) assuming that age- and sexspecific per capita values equal their respective values in 1992 or some later year except for an adjustment for productivity growth. Second, for each generation in a given year t (say, males born in 1966), the present value of all future per capita payments of a particular type (say, indirect tax payments) is computed by multiplying these future per capita payments by the generation's projected population in those years, discounting these values back to year t, and dividing the sum of the discounted values by the number of members of the generation alive in the base year. This method produces actuarially discounted present values of the particular receipt or payment for each generation alive in period t. As an example of this method for calculating the different components of annuitized resources, consider the estimate of human wealth (HW). Our formula , ~ , for human wealth in year t of a person of sex x born in year k , H W ~ ~ is where es,k stands for the average earnings in year s of a member of the generation born in year k and of sex x, pXSSk is the population in year s of the same generation, R=l/(l+r), where r is the rate of interest, and D is the maximumage of life. The calculationof es ,k is givenby clevelandfed.org/research/workpaper/index.cfm In these equations Es is aggregate labor earnings in year s, dxs ,k is the ratio in year s of the average earnings of the generation born in year k of sex x divided by the average earnings in year s of our reference group - those males who were age 40 in year s (i.e., those for whom k=s-40). The construction of relative profiles by age and sex, dxtSk, is described in equations (9) and (10): In equation (9), zxs,k is the weighted average (across cohort members indexed by i) of labor income. IVXsPk is the number of observations in year s of individuals of sex x born in year k , zxs,k,i is the wage and salary income of the ith individual of sex x in year s who was born in year k , and w ~ ~is ,the~ person weight of this observation. Equation (10) shows the calculation in year s of the average labor income of members of the generation with sex x who were born in year k , relative to that of contemporaneous 40-year-old males. , ~ clevelandfed.org/research/workpaper/index.cfm IV. Data Sources The national aggregates used in our calculations come from the National Income and Product Accounts (NIPA), the Federal Reserve System's Flow of Funds (FOF), The American Council of Life Insurance (ACLI), the U.S. Census Bureau's Current Population Survey (CPS), and the Survey of Current Business (SCB). The sources for cross-section relative profiles are the CPS, the Survey of Income and Program Participation (SIPP), the Consumer Expenditure Survey (CES), the Survey of Consumer Finances (SCF), the Social Security Administration's Annual Statistical Supplement (SSASS), and the Health Care Financing Administration (HCFA). The computations also use the historic and projected population counts of the Social Security Administration (SSA). 7 The following is a more detailed description of our data sources and projections : Labor Income Aggregate labor income between 1960 and 1992 is calculated as labor's share of NIPA-reported national income. For each of these years, labor's share of national income is calculated under the assumption that its share of proprietorship income is the same as its share of national income.8 Relative profiles of labor income by age and sex are calculated for each year between 1963 and 1992 using that year's CPS data on individual wages and SSA's projections are available through the year 2066. These projections were extended to the year 2200 by using SSA's mortality, fertility, and immigration assumptions for the year 2066. Labor income's share of national income is a, where a satisfies C + aPI = aNI. In this equation, C is compensation paid to employees, PI is proprietorship income, and NI is national income. The calculated values of a are very stable over the years 1960-1992, ranging between 0.76 and 0.82.- clevelandfed.org/research/workpaper/index.cfm salaries. Since these data are top-coded, we followed Shyrock and Seigel (1971) and estimated for each year the average wage and salary for top-coded observations. Their procedure uses the fact that the upper tail of the wage and salary distribution can be well approximated by the Pareto distribution. Values of average wages for top-coded observations were calculated separately for males and females. These values were assigned to all top-coded observations before computing relative wage profiles. The annual profiles were smoothed over age by using a seven-year moving average of wages and salaries. The 1963 profiles are used to distribute aggregate labor income for years prior to 1963, and the 1992 profiles are applied for years after 1992. Per capita labor income for years beyond 1992 is projected under the assumption that, except for an adjustment for growth, cohorts of a given age and sex earn the same average labor income in future years as cohorts of that age and sex earned in 1992. For example, males who are age 50 in 1993 are assumed to earn the same amount on average, apart from an adjustment for growth, as males who were age 50 in 1992. The growth adjustment is 0.75 percent per year. Thus, the projected average earnings of males age 50 in, say, 1994 equals the corresponding 1992 average for 50-year-old males, multiplied by (1.0075). Private and Government Emvlovee Pensions. Workers' Com~ensation.and Veterans' Benefits This category includes four types of income -- benefits from private pension plans, workers' compensation, veterans' benefits, and government employee pensions. Aggregate private pension benefits for the years 1960-1988 are the NIPA estimates reported in Park (1992). The NIPA estimates are based The small number of observations precluded separate estimation by age. clevelandfed.org/research/workpaper/index.cfm upon administrative reports and appear to be more reliable than other estimates that are based upon household surveys. Estimates of aggregate private pension benefits for 1989 through 1992 were derived by extrapolating the 1988 reported level of benefits using the 1984-1988 average annual growth rate of real aggregate private pension benefits. The aggregates for the other three types of benefits are reported in the SCB. The relative profiles of the four types of income are computed from the March CPS. This survey contains information on income from a variety of sources including company or union pensions, workers' compensation, veterans' benefits, government employee pensions, and receipts from annuities and other regular contributions. Retirement, disability, and survivor benefits are included for each type of income., Unfortunately, receipts from several sources of retirement income are aggregated into one variable in the CPS data. For example, in the 1980-88 data, private pension income is combined with income from government employee pensions (including federal, state, and local government pensions, as well as military retirement pensions). Fortunately, the CPS specifies for each obser- vation the different types of income that are being combined into the pension and other income variable. We use this information to identify, for each age and sex, those observations receiving only private pensions and those receiving only government employee pensions. Next, we calculate, again by age and sex, the average values of the two types of income. Finally, we compute the ratio of average private pension income to the sum of the averages of income from private and government employee pensions. The ratio of average government employee pensions to average pension receipts for this age-sex category is one minus the ratio of average private pensions to average pension clevelandfed.org/research/workpaper/index.cfm receipts. These two ratios are used to impute the values of private and government employee pensions for observations receiving both types of income. The computation of relative profiles for each year uses the age- and sexspecific cell averages of actual and imputed private and government employee pension income, smoothed across age using a seven-year moving average. Separate profiles were obtained for each of the four categories of income for each year between 1970 and 1992. The 1970 profiles were used to distribute the national aggregates of these payments in years prior to 1970. For years after 1992, real average pension benefits at a given age and sex are set equal to their 1992 values, adjusted for the assumed 0.75 percent rate of growth. Social Security Benefits Aggregate Social Security benefits between 1960 and 1992 are those reported by NIPA. Between 1993 and 2004 we use the Office of Management and Budget's (OMB) projections of Social Security benefits computed on a NIPA basis. Aggregate Social Security Old Age, Survivor, and Disability Insurance (OASDI) benefits after 2004 equal the 2004 aggregate adjusted for growth. The growth rates applied in this case are those embedded in a special Social Security Administration projection of total benefit payments for years after 2004. This projection incorporates Social Security's intermediate economic and demographic assumptions, with one exception: the productivity growth rate is assumed to equal 0.75 percent. The SSASS reports average benefits by age and sex by type of benefit as well as the total number of recipients in each age-sex category. These data were used to form population-weighted per capita OASDI benefit profiles by age clevelandfed.org/research/workpaper/index.cfm and sex. Relative p r o f i l e s of OASDI b e n e f i t s f o r each year from 1960 through 1990 were obtained from t h a t y e a r ' s SSASS. For years a f t e r 1990 we use t h e 1990 r e l a t i v e p r o f i l e of S o c i a l Security b e n e f i t s by age and sex. Medicare and Medicaid Benefits Aggregate Medicare and Medicaid payments from t h e inceptions of these programs through 1992 a r e those reported by NIPA. OMB provided us with unpublished p r o j e c t i o n s , on a NIPA b a s i s , of aggregate Medicare payments f o r t h e years 1993 through 2004. For t h e years between 2004 and 2030, we extrapo- l a t e d aggregate Medicare payments using HCFA's 2004-2030 p r o j e c t e d Medicare growth r a t e s . I n t h e case of Medicaid, we applied HCFA's p r o j e c t e d annual Medicaid growth r a t e s between 1993 and 2030 t o t h e 1992 aggregate NIPA value of Medicaid. Medicare and Medicaid payments beyond 2030 a r e assumed t o grow i n accordance with demographic change and our assumed p r o d u c t i v i t y growth rate. Relative p r o f i l e s of Medicare and Medicaid b e n e f i t s a r e based on HCFA d a t a on average b e n e f i t s by age and sex. a r e a v a i l a b l e only by five-year I n t h e case of Medicare, t h e d a t a age groups. U n e m p l o ~ e n tInsurance, Aid t o Families with Dependent Children, Food Stamps and General Welfare Benefits Aggregate values of these f e d e r a l , s t a t e , and l o c a l t r a n s f e r s a r e those reported by NIPA. Supplemental s e c u r i t y income, a s well a s t r a n s f e r s f o r employment and t r a i n i n g , a r e d i s t r i b u t e d according t o t h e r e l a t i v e p r o f i l e f o r AFDC. General welfare b e n e f i t s include f e d e r a l black-lung b e n e f i t s , s t a t e general a s s i s t a n c e , s t a t e energy a s s i s t a n c e , education b e n e f i t s , and other f e d e r a l , s t a t e , and l o c a l t r a n s f e r s . The aggregate amount of earned income clevelandfed.org/research/workpaper/index.cfm t a x c r e d i t s i s d i s t r i b u t e d according t o the r e l a t i v e p r o f i l e f o r food stamps. P r o f i l e s f o r unemployment insurance, food stamps, AFDC, and general welfare a r e computed from t h e 1983 SIPP. These r e l a t i v e p r o f i l e s a r e used, i n conjunction with y e a r - s p e c i f i c population counts by age and s e x , t o d i s t r i b u t e t h e i r r e s p e c t i v e aggregate expenditures by age and sex f o r a l l of t h e y e a r s between 1960 and 1992. For f u t u r e y e a r s , we assume t h a t t h e age- and sex- s p e c i f i c values of each of t h e s e d i f f e r e n t types of t r a n s f e r payments keep pace with p r o d u c t i v i t y growth. Labor Income Taxes Aggregate f e d e r a l , s t a t e , and l o c a l income t a x e s f o r 1960 through 1992 a r e those reported by NIPA. For 1992 through 2004, we use OMB's p r o j e c t i o n s of f e d e r a l income t a x revenues. S t a t e and l o c a l income t a x e s f o r 1993 through 2004 a r e p r o j e c t e d using O M B ' s GDP f o r e c a s t and assuming t h a t t h e same r a t i o of s t a t e and l o c a l income t a x e s t o GDP p r e v a i l s between 1993 and 2004 a s t h a t which p r e v a i l e d i n 1992. Aggregate l a b o r income taxes i n each year a r e c a l c u l a t e d a s t h e product of t o t a l f e d e r a l , s t a t e , and l o c a l income taxes and l a b o r ' s share of n a t i o n a l income. We d i s t r i b u t e aggregate l a b o r income taxes based on t h e CPS p r o f i l e s of l a b o r income described above. A f t e r 2004, we assume t h a t age- and sex- s p e c i f i c values of l a b o r income taxes keep pace with p r o d u c t i v i t y growth. P a v r o l l Taxes The NIPA r e p o r t s aggregate values of p a y r o l l t a x e s from 1960 through 1992. OMB provided us with p r o j e c t i o n s of aggregate f e d e r a l p a y r o l l taxes from 1993 through 2004. Aggregate s t a t e and l o c a l p a y r o l l taxes f o r 1993 clevelandfed.org/research/workpaper/index.cfm through 2004 were calculated based on OMB's projection of GDP between 1993 and 2004 and the assumption that the 1992 ratio of state and local payroll taxes to GDP prevails through 2004. Aggregate payroll taxes in the years 1960-2004 are distributed by age and sex according to 1963-1992 CPS profiles of covered earnings, where covered earnings refers to labor earnings that are subject to Social Security payroll taxes.lo Age- and sex-specific values of payroll taxes beyond 2004 are assumed to equal their 2004 values, adjusted for growth. Excise and Sales Taxes The NIPA is our source for aggregate excise tax (including property tax) and sales tax revenue from 1960 through 1992. For the period 1993-2004, we use OMB projections of federal excise and sales tax revenues. State and local excise and sales tax revenues between 1993 and 2004 are calculated using the 1992 ratio of these revenues to GDP and applying OMB's GDP forecasts through 2004. Relative age-sex profiles of excise and sales taxes were calculated from the 1960-61, 1972-73, 1984-86, and 1987-90 CES. Separate profiles were constructed for tobacco, alcohol, property taxes, and all other sales and excise taxes. The 1960-61 profiles were used for years prior to 1966. The 1972-73 profiles were used for the years 1967 through 1978. The 1984-86 profiles were used for the years 1979 through 1986, and the 1987-90 profiles were used for 1987 and beyond. Age- and sex-specific values of sales and lo Unfortunately, the data do not permit the calculation of separate profiles for state and local payroll taxes, which are not necessarily subject to earnings ceilings. However, non-Social Security payroll taxes are a small fraction of the total (less than 30 percent), so the bias associated with using Social Security covered earnings profiles is likely to be small. clevelandfed.org/research/workpaper/index.cfm excise taxes beyond 2004 are assumed to equal the 2004 values, adjusted for growth. Capital Income Taxes Aggregate capital income taxes between 1960 and 2004 are calculated as capital's share of national income multiplied by actual or projected values of aggregate federal, state, and local income tax revenues. Relative profiles of capital income taxes come from the 1962 and 1983 SCFs. These profiles are based upon weighted average net-worth holdings by age and sex, where the weights applied are SCF person weights. This procedure could be applied only to individuals age 80 or less because of limited data for older individuals. The profile of average net-worth holdings by age and sex was smoothed and extrapolated through age 90 using a fourth-order polynomial. Age- and sexspecific values of capital income taxes after 2004 are assumed to equal the 2004 values, adjusted for growth. Nonhuman Wealth Age- and sex-specific values of nonhuman wealth (NHW) in each year between 1960 and 1992 are constructed by distributing by age and sex each year's level of total private net wealth. Aggregate private net wealth for ~ relative profiles of wealth these years is reported in the F O F . ~The holdings by age and sex are calculated by using data from the 1963 and 1983 SCF. In estimating the relative profiles, components of wealth that are owned jointly by members of a multiperson household are divided equally among such Our aggregates are net of the FOF's estimate of the value of residential structures, plant, and equipment owned by nonprofit institutions. clevelandfed.org/research/workpaper/index.cfm members. The 1963 profiles are used for years prior to 1963, and the 1983 profiles are used for years after 1983. The profiles for intermediate years are constructed by linearly interpolating between the 1963 and 1983 profiles. The Term Value of Life Insurance Aggregate face values of life insurance for the years 1960 through 1992 are reported by the ACLI. The 1962 and 1983 SCF are used to distribute these amounts by age and sex. Fortunately, the SCFs report term as well as face values of life insurance. Consequently, we were able to calculate the ratio of term value to face value of life insurance on an age- and sex-specific basis for the years 1962 and 1983. Multiplying these ratios by our calculated age- and sex-specific face values of insurance produced age- and sex-specific term values of insurance for 1962 and 1983, and, after interpolating, for other years as well.12 V . Findings A. Changes i n t h e Cohort D i s t r i b u t i o n o f Resources Total Resources The total resources of a cohort is the sum of its human, nonhuman, and pension wealth, less its generational account. The generational account refers to the present value of a sex-specific generation's future tax payments net of the present value of its future receipts of transfer payments. Our calculations include all tax payments made to, and transfer payments received from, federal, state, and local governments. 12~otethat the cash value of life insurance is counted as part of nonhuman wealth. clevelandfed.org/research/workpaper/index.cfm Tables 2a and 2b contain per capita values of total real resources and resource components for male and female cohorts in 10-year age groups for the years 1960, 1970, 1980, and 1990. The tables also show the per capita resources of all cohorts age.20-89 and of older cohorts age 65-89. For the entire populations of males and females, total resources grew substantially over the three decades since 1960;but and for women. percent. they grew more rapidly for the elderly For males as a group, per capita resources rose by 39.6 For older males, they grew by 119.9 percent. For females as a group, per capita resources rose by 124.0 percent. For older females, they grew by 123.8 percent. Between 1960 and 1990 female per capita resources rose from 39 percent to 62 percent of male per capita resources. Some of the reported differences in resource growth across ages and sex are particularly striking. For example, males age 20-29 experienced only a 7.1 percent increase in their average resources over the 30 years, whereas males age 70-79 experienced a 125.5 percent increase, and females age 20-29 experienced a 153.5 percent increase. The relative growth in elderly Americans' resources appears primarily to reflect government intergenerational redistribution, coupled with improvements in their longevity. Between 1960 and 1990, the average generational account of older males fell from -$3,400 to -$80,200. The decline was even larger for older females. Their average generational account was -$6,600 in 1960, but in 1990 it was -$99,300. Over the same period, the generational accounts of younger cohorts rose dramatically. For example, the accounts of males age 2029 rose from $145,800 to $191,700 and those of females age 20-29 rose from $66,900 to $118,800. The components of generational accounts shown in tables 3a and 3b clearly indicate that changes in the relative values of the genera- clevelandfed.org/research/workpaper/index.cfm tional accounts of the old and the young primarily reflect increases in Social Security and government-provided health-care benefits (Medicare and Medicaid), on the one hand, and increases in labor income and payroll taxation, on the other. The relative growth in femalesf resources primarily reflects their increased participation in the labor force. According to tables 2a and 2b, in 1960 the human wealth of females age 20-29 was $148,900 per capita - just 29 percent of the corresponding male value of $521,500. In 1990, the per capita human wealth of females in this age range was $326,600 -- 56 percent of the 1990 male average of $581,800. / The Composition of Total Resources Tables 4a and 4b show the composition of total resources. For younger cohorts of both sexes, human wealth represents the bulk of resources. The reason is simply that most of their working years lie in the future. In fact, these cohortsf human wealth is larger than their total resources because the latter are calculated net of their positive generational accounts. In contrast, older cohortsf total resources are predominantly held in the form of nonhuman wealth. Over the three decades, the share of human wealth in total resources declined for all male cohorts over 40. The same is true for female cohorts age 50 and over. For the male population as a whole, the share of nonhuman wealth in total resources remained roughly constant, but it declined significantly for male cohorts over 65 years of age. The share of nonhuman wealth declined for the female population as a whole, and it declined significantly for women over 65. The decline from 85 to 52 percent in the share of nonhuman wealth for older females was greater than the decline from 74 to 53 percent for older males. clevelandfed.org/research/workpaper/index.cfm As tables 4a and 4b indicate, private pension wealth is a small but growing component of total resources. Its share of total resources increased from 10 to 18 percent for older males and from 5 to 9 percent for older females. The share of resources represented by generational accounts also changed significantly over the three decades. Generational accounts as a share of total resources increased for males age 39 or less, but declined significantly for females in the same age categories. For older males the excess, in present value, of future transfers over future taxes (the negative of the generational account) made up almost a quarter of total resources in 1990, compared with only 2 percent in 1960. The corresponding female figures are 37 percent in 1990 and 5 percent in 1960. Tables 5a and 5b express the components of generational accounts as shares of total resources. Among other things, they show that health benefits rose from an insignificant share of elderly Americans' resources in 1960 to 14 percent of older males' resources and 23 percent of older females' resources in 1990. B. Changes in Bequeathable and Annuitized Resources Tables 6a and 6b present the components of bequeathable resources nonhuman wealth plus the term value of life insurance - as - well as the difference between bequeathable resources and total resources. This difference is annuitized resources. Tables 7a and 7b report these components as a fraction of total resources. The degree of resource annuitization, R ~ , is computed as the ratio of annuitized to total resources, i.e.: (11) a R = l - TERM + NHW HW+NHW+PW-GA ' clevelandfed.org/research/workpaper/index.cfm where TERM stands for the average term value of life insurance, HW stands for average human wealth, NHW stands for average nonhuman wealth, PW stands for average private pension wealth, and GA stands for the generational account. Table 7a shows a significant increase in Ra for older males -- from 0.22 in 1960 for cohorts age 65 and older to 0.45 in 1990. For older female cohorts, the increase reported in table 7b is even larger -- from 0.12 in 1960 to 0.47 in 1990. This larger annuitized share of elderly persons' resources implies, of course, an equal and opposite decline in their share of bequeathable resources. The increased annuitization of older males is offset by the decreased annuitization of younger males. Because younger males outnumber older ones, overall male resource annuitization declines by a small amount. Specifically, Ra for males falls from 0.74 to 0.69. For females, however, the ratio of annuitized to total resources increases for all age cohorts. For the female population as a whole, Ra rises from 0.33 to 0.53 during this period. Sensitivity Analysis The calculations reported earlier assume a 0.75 percent rate of productivity growth (g) and a discount rate (r) of 6 percent. We denote these as the base-case values for r and g. Table 8 examines the sensitivity of Ra to alternative interest rate assumptions. As the table shows, the conclusion that the resource annuitization of the elderly has increased dramatically since 1960 holds for values of r of 3, 6 , and 9 percent. Higher interest rates produce smaller values of annuitized resources, but they do so for each of the years considered. Hence, they do not have much clevelandfed.org/research/workpaper/index.cfm impact on the increase over time in the ratio of annuitized to total resources. For example, when r equals 9.0 percent, Ra in 1960 is 0.19 for elderly males and 0.08 for elderly females. These values are smaller than the respective base case (r=0.06 g=0.75) amounts 0.22 and 0.12. However, the growth in Ra for elderly males and females between 1960 and 1990 is about the same in this case as in the base case. For older males, Ra rises from 0.19 to 0.40, compared to 0.22 to 0.45 in the base case. For older females, Ra rises from 0.08 to 0.42, compared with 0.12 to 0.47 in the base case. As mentioned, projected health benefits are an important component of annuitized resources for the elderly and the middle-aged. Table 9 examines the degree to which Ra would be different under alternative assumptions regarding future government health-care policies. We consider three alternatives to the current policy (the base case). The first incorporates the administration's official revenue and expenditure projections for President Clinton's health reform proposal (columns 3 and 4 in table 9). Through the turn of the century, the President's plan entails essentially the same total level of spending on health care (if one includes the proposed new subsidies to early retirees, etc.) as under current policy. However, after the turn of the century, real government health-care spending is slated to grow no faster than the rate warranted by demographic change and growth in labor productivity. The second health-care policy alternative (columns 5 and 6 in table 9) modifies the projections arising under the President's plan by assuming that real Medicare spending will grow from 2000 through 2020 at a rate 2 percent higher than the plan foresees. The third health-care policy (columns 7 and 8 in table 9) limits growth in real government health-care spending to the clevelandfed.org/research/workpaper/index.cfm amount warranted by demographic change and labor productivity growth starting in 1994. As table 9 shows, incorporating the administration's official revenue and expenditure projections under President Clinton's health-care plan, with or without the extra 2 percent growth between 2000 and 2020, reduces the values of Ra only very slightly. For the elderly, the reductions in Ra are evident only for the 1990 figures. This is because for the elderly in 1980 and earlier, the benefit cuts will occur only in the distant future. For the entire population, however, reductions in Ra are seen as early as 1980 because, compared to the base case, the benefit cuts will have been fully phased in by the time these cohorts receive government health-care benefits.13 Compared with the base case, even a policy of stabilizing health-care spending beginning in 1994 does not significantly alter the degree of annuitization for the elderly. Under the base case, 47 percent of the resources of older females are annuitized in 1990, compared with 44 percent under the 1994 stabilization policy. For older males, the respective figures are 45 percent and 43 percent. Thus, the post-1960 increase in annuitization remains dramatic, despite the 1994 stabilization policy. C. Implications for Aggregate Bequests and National Saving As discussed earlier, cohorts with higher degrees of annuitization will, ceteris paribus, bequeath less and consume more. To assess the impact on aggregate bequests of changes since 1960 in Americans' degree of annuitiza1 3 ~ h e1990 annuitization ratio for females as a whole is slightly larger for the 2 percent faster health-care cost growth (0.52) than under the Clinton health reform scenario (0.51). This occurs because younger females receive substantially more in health-care benefits over their remaining lifetimes under the former scenario. clevelandfed.org/research/workpaper/index.cfm tion, we first estimate the total flow of bequests in 1990 for base-case values of r and g. We do so by multiplying the aggregate 1990 values of bequeathable wealth (net worth plus term life insurance) for individual male and female cohorts by their respective 1990 mortality probabilities. Summing the products over all cohorts yields an aggregate 1990 bequest flow of $245.1 billion. Next, we calculate 1990 aggregate bequests under the assumption that a cohort's bequeathable resources in 1990 equal its total resources in 1990 multiplied by its 1960 ratio of bequeathable resources to total resources. This produces a 1990 bequest flow of $360.8 billion for the base case. Thus, without the post-1960 increase in resource annuitization, aggregate 1990 bequests would have been an estimated 47.2 percent larger. Note that we hold the total resources of each cohort fixed in this counterfactual experiment. The $115.7 billion difference between these two bequest amounts constitutes the additional amount that generations alive in 1990 appear likely to have consumed as a consequence of this increased annuitization. This $115.7 billion figure is substantial: It represents 74 percent of total net national saving in 1990. Table 10 indicates that the percentage reduction in estimated 1990 bequests due to the increased resource annuitization would not be much affected by any of the three alternative future paths of health-care spending by the government; even if health-care spending were stabilized in 1994, the reduction would be a sizable 41 percent. Table 11 examines the sensitivity of the reduction in bequests under alternative assumptions for r. Large reductions are indicated for each interest-rate assumption. The smallest reduction in bequests is 40.8 percent, and the largest is 55.5 percent. A different question about the reliability of these findings involves our use of the random bequest method to estimate the flow of bequests. This clevelandfed.org/research/workpaper/index.cfm method assumes that the net worth and life insurance holdings of those people who actually die at age x at time t do not differ systematically with respect to their wealth holdings and life insurance from those who do not actually die at age x at time t. Admittedly, many of those who actually die at age x at time t may know ahead of time that they are about to die and spend down some of their assets through high living. In addition, those who actually die may incur particularly large uninsured medical expenses. But this bias in the random death method's calculation of bequests, whatever its size, is a bias that holds for each of our calculations of actual bequests in 1990, as well as the hypothetical bequests that would have prevailed in 1990 had the degree of annuitization been that of 1960. Indeed, if one assumes that in 1990 end-oflife uninsured medical expenses, as well as other end-of-life expenses, would have been the same had Americansr annuitization been that of 1960, our procedure underestimates the percentage decline in bequests.14,15 VI . Conclusion This paper combines a large array of micro and macro data to study changes since 1960 in the degree of annuitization of Americansr resources Although we find no increase in the annuitization of younger Americans, we find a dramatic increase in the degree of annuitization of older Americans. This finding is robust to alternative assumptions about interest and growth l4 The reason is that the difference in bequests is the same, but the level of actual 1990 bequests is smaller, producing a larger percentage change in bequests in the hypothetical exercise. 15see, for example, Scheiner and Weil (1992) for evidence of decumulation of housing wealth just prior to death. clevelandfed.org/research/workpaper/index.cfm rates, as well as to various possible courses of future U.S. health-care policy. The increase in the annuitization of the elderly reflects increases in their receipt of Social Security and health transfers, coupled with their failure to increase their purchase of life insurance. Since the elderly have much higher mortality probabilities, their degree of annuitization is critical to the flow of bequests. According to our base-case estimates, holding fixed the total resources of each cohort, current aggregate U.S. bequests would be roughly 50 percent larger if these resources, particularly those of older Americans, were annuitized to the same degree as they were in 1960. addition, U.S. national saving would likely be substantially larger. In clevelandfed.org/research/workpaper/index.cfm References Abel, Andrew B., "Precautionary Savings and Accidental Bequests," American Economic Review, vol. 75, no. 4, 1985, pp. 777-91. Auerbach, Alan J., Jagadeesh Gokhale, and Laurence Kotlikoff, "Generational Accounting: A Meaningful Alternative to Deficit Accounting," Tax Policy and the Economy, ed. by David Bradford, National Bureau of Economic Research, vol. 5, 1991, pp. 55-110. Auerbach, Alan J., Laurence J. Kotlikoff, David N. Weil, "The Increasing Annuitization of the Elderly - Estimates and Implications for Intergenerational Transfers, Inequality, and National Saving," National Bureau of Economic Research working paper no. 4182, October 1992. Boskin, Michael, Michael Knetter, and Laurence J. Kotlikoff, "Changes in the Age Distribution of Income in the United States, 1968-1984," mimeo, Center for Economic Policy Research, Stanford University, October 1985. Davies, James, "Uncertain Lifetimes, Consumption and Dissaving in Retirement," Journal of Political Economy, vol. 89, 1981, pp. 561-77. Feldstein, Martin S., "Social Security, Induced Retirement, and Aggregate Capital Accumulation," Journal of Political Economv, vol. 82, 1974, pp. 90526. Gokhale, Jagadeesh, Laurence J. Kotlikoff, and John Sabelhaus, "Understanding the Postwar Decline in U.S. Saving: A Cohort Analysis," Boston University, Department of Economics, mimeo 1994. Kotlikoff, Laurence J., John Shoven, and Avia Spivak, "The Impact of Annuity Insurance on Savings and Inequality,''Journal of Labor Economics, vol. 4, no. 3, pt. 2, 1986, pp. 5183-207. Park, Thae S., "Total Private Pension Benefit Payments, 1950-88, in Trends in Pensions. 1992, ed. by John A. Turner and Daniel J. Beller, U.S. Department of Labor, Pension and Welfare Benefits Administration, 1992, pp. 271-83. Scheiner, Louise, and David N. Weil, "The Housing Wealth of the Aged," NBER Working Paper No. 4115, July 1992. Shyrock, Henry, S., and Jacob S. Seigel and Associates, The Methods and Materials of Demography, U.S. Department of Commerce, Bureau of the Census, vol. 1, 1971, pp. 365-66. Yaari, Menahem, E. "Uncertain Lifetime, Life Insurance, and the Theory of the Consumer," Review of Economic Studies, vol. 32, April 1965, pp. 137-50. clevelandfed.org/research/workpaper/index.cfm clevelandfed.org/research/workpaper/index.cfm Table 1 Consumption of the Elderly Relative to the Young Comparison 1960-61 1972-73 1984-86 1987-90 Male 70/Male 30 .672 .802 1.135 1.247 Female 70/Male 30 .667 .798 1.045 1.112 Male 70/Female 30 .664 .763 1.059 1.202 Female 70/Female 30 -659 .760 .975 1.072 Source: Authors' calculations. clevelandfed.org/research/workpaper/index.cfm Table 2a Total Resources and Resource Components -- Male Cohorts 1960-90 (Population Weighted Averages i n Thousands of 1992 D o l l a r s ) Age Group: 20-29 30-39 40-49 50-59 60-69 70-79 80-89 20-89 65-89 410.4 427.3 416.2 439.5 419.8 477.1 492.2 504.1 374.5 456.6 498.9 540.4 286.3 384.1 454.6 522.7 204.9 281.9 368.3 459.1 152.9 205.9 271.9 344.8 105.7 122.3 187.0 185.5 339.9 398.5 432.0 474.5 160.3 212.0 285.3 352.6 7.1 20.0 44.3 82.6 124.1 125.5 75.5 39.6 119.9 521.5 557.4 548.2 581.8 476.7 548.7 561.7 581.5 358.1 419.5 437.1 472.3 202.7 240.2 248.9 258.8 65.6 70.2 66.9 69.7 13.4 11.9 10.7 14.2 5.9 4.2 3.2 5.2 332.7 374.9 389.1 413.7 21.6 19.5 18.3 21.8 11.6 22.0 31.9 27.7 6.3 6.0 -11.9 24.3 0.1 12.0 12.9 17.1 18.1 44.4 49.1 62.6 61.6 81.8 100.3 122.1 131.8 112.2 143.5 170.0 198.2 126.5 162.5 191.0 225.0 119.2 142.8 177.1 190.6 92.6 84.1 128.5 89.6 71.3 85.3 98.3 108.2 119.1 141.5 175.4 187.1 50.8 38.7 61.1 76.6 77.9 59.9 -3.2 51.8 57.1 22.7 26.6 27.9 31.4 30.2 37.8 41.7 46.1 32.0 51.4 61.6 68.7 24.4 50.6 80.0 95.3 21.3 37.1 70.1 94.2 14.9 25.0 36.1 57.7 6.9 14.1 20.7 27.0 25.7 38.4 49.2 58.5 16.2 26.5 42.9 63.5 38.3 52.6 114.7 290.6 342.3 287.2 291.3 127.6 291.4 TOTAL RESOURCES 1960 1970 1980 1990 % Increase 1990/ 1960 HUMAN WEALTH 1960 1970 1980 1990 % Increase 1990/1960 NON-HUMAN WEALTH 1960 1970 1980 1990 % Increase 1990/1960 PENSION WEALTH 1960 1970 1980 1990 X Increase 1990/ 1960 GENERATIONAL ACCOUNT Source: Authors1 calculations. clevelandfed.org/research/workpaper/index.cfm Table 2b Total Resources and Resource Components - Female Cohorts 1960-90 (Population Weighted Averages i n Thousads of 1992 D o l l a r s ) Age Group: 20-29 30-39 40-49 50-59 60-69 70-79 80-89 20-89 65-89 96.2 163.4 204.7 243.9 112.3 171.5 237.0 289.4 154.8 199.6 274.3 323.9 169.8 222.6 291.2 349.5 147.2 219.3 281.5 362.8 115.1 161.7 229.2 278.4 89.3 83.8 166.9 129.5 131.0 185.7 244.5 293.5 120.1 163.6 229.9 268.7 153.5 157.7 109.2 105.8 146.5 141.9 45.0 124.0 123.8 148.9 227.0 286.4 326.6 121.1 173.3 251.0 301.1 98.9 124.5 168.2 229.5 60.5 72.4 82.0 111.8 17.3 20.9 20.8 25.3 2.9 2.9 3.0 4.4 0.5 0.7 0.8 1.3 87.6 121.0 163.1 198.3 5.2 5.1 5.0 6.5 119.3 148.6 132.1 84.8 46.2 51.7 160.0 126.4 23.1 7.2 13.1 11.9 25.0 42.9 56.9 62.5 79.9 91.9 104.6 134.9 132.2 118.1 139.1 175.0 183.2 113.8 147.1 172.0 203.9 99.9 114.5 146.6 147.3 88.4 56.7 116.6 36.0 72.1 85.4 99.9 108.5 101.9 113.7 146.8 138.7 247.2 86.2 43.9 55.1 79.2 47.4 -59.3 50.5 36.1 7.1 8.2 9.7 11.1 9.3 12.0 13.8 15.8 9.8 16.0 20.4 23.1 11.0 16.8 26.8 32.8 10.0 16.9 24.4 34.5 5.6 10.7 16.5 23.2 2.3 5.2 9.7 14.8 8.9 12.9 17.1 21.0 6.4 11.5 17.0 24.2 56.3 69.9 135.7 198.2 245.0 314.3 543.5 134.8 276.6 TOTAL RESOURCES 1960 1970 1980 1990 % Increase 1990/1960 HUMAN WEALTH 1960 1970 1980 1990 % Increase 1990/1960 NOW-HUMAN WEALTH 1960 1970 1980 1990 % Increase 1990/1960 PENSION WEALTH 1960 1970 1980 1990 % Increase 1990/1960 GENERATIONAL ACCOUNT Source: Authors1 calculations. clevelandfed.org/research/workpaper/index.cfm Table 3a The Components of Generational Accounts - Male Cohorts 1960-90 (Population Ueighted Averages i n Thousands of 1992 D o l l a r s ) 30-39 40-49 50-59 60-69 70-79 80-89 20-89 65-89 53.6 67.5 73.6 78.8 39.1 51.0 56.7 62.9 21.4 28.6 32.4 33.8 6.8 8.2 8.8 9.0 1.4 1.4 1.4 1.8 0.6 0.5 0.4 0.7 37.2 46.1 51.1 55.9 2.2 2.3 2.4 2.8 86.5 40.1 60.4 75.1 83.9 27.3 43.0 55.8 66.5 14.6 23.7 30.7 35.8 4.5 6.8 8.1 9.4 0.9 1.2 1.3 1.9 0.3 0.4 0.4 0.7 28.2 42.0 52.4 59.7 1.5 1.9 2.2 2.9 INDIRECT TAXES 1960 45.5 1970 49.7 1980 51.2 1990 57.2 43.6 49.4 51.7 57.3 36.2 41.8 44.7 50.9 25.1 31.2 34.4 39.1 14.9 19.5 23.1 27.1 8.0 11.1 13.8 17.0 4.4 6.6 8.3 9.1 33.3 38.3 41.4 46.7 9.4 12.5 15.3 18.4 26.2 27.0 29.9 33.6 31.1 31.6 35.6 40.6 30.5 31.2 34.8 40.1 23.9 26.1 28.7. 31.4 14.6 17.3 20.6 18.4 7.4 8.4 13.3 6.7 24.9 25.5 28.0 31.6 16.5 18.5 21.7 20.2 26.9 42.1 51.0 60.7 33.6 52.0 72.1 83.7 26.0 40.6 55.0 69.2 11.3 24.5 35.4 44.5 20.6 28.6 34.2 39.4 28.0 42.6 59.0 70.8 6.4 17.5 33.0 59.0 3.4 14.2 27.3 49.0 1.2 9.9 19.9 34.4 8.9 17.2 27.4 41.3 4.0 14.8 28.3 50.4 1.7 3.2 3.9 4.4 0.9 2.3 3.0 3.3 0.4 1.2 1.8 1.9 4.3 6.2 6.7 7.2 1.0 2.4 3.0 3.4 Age Group: 20-29 LABOR INCOME TAXES 1960 60.2 1970 1980 1990 70.0 72.8 80.0 PAYROLL TAXES 1960 50.3 1970 69.1 1980 77.7 1990 CAPITAL INCOME TAXES 1960 18.4 1970 1980 1990 19.1 21.3 24.4 SOCIAL SECURITY BENEFITS (OASDI) 1960 11.2 15.9 22.4 1970 1980 1990 13.0 14.4 16.0 19.7 22.1 24.4 28.4 34.0 38.1 HEALTH BENEFITS (Medicare and Medicaid) 1960 10.8 10.3 9.4 8.6 1970 1980 1990 16.7 22.8 30.8 UELFARE BENEFITS 1960 6.6 1970 8.5 1980 8.8 1990 9.5 18.4 26.4 35.3 18.1 30.2 43.2 5.7 7.7 8.0 8.6 4.4 6.3 6.6 7.2 Source: Authors1 calculations. 17.5 31.7 52.6 3.0. 4.7 5.3 5.8 clevelandfed.org/research/workpaper/index.cfm Table 3b The Components of Generational Accounts -- Female Cohorts 1960-90 (Population Weighted Averages i n Thousands of 1992 Dollars) 30-39 40-49 50-59 60-69 70-79 80-89 20-89 65-89 17.5 28.7 38.0 44.7 13.9 21.5 32.9 40.7 10.9 15.2 21.8 30.5 6.4 8.6 10.6 14.6 1.8 2.5 2.7 3.3 0.3 0.3 0.4 0.6 0.1 0.1 0.1 0.2 9.9 15.0 21.4 26.8 0.5 0.6 0.7 0.8 18.1 31.6 43.0 51.2 13.7 23.4 37.0 46.7 10.2 16.1 24.4 35.2 5.6 8.9 11.6 16.9 1.5 2.5 2.9 3.8 0.2 0.3 0.4 0.7 0.1 0.1 0.1 0.2 9.7 16.2 24.1 30.7 0.4 0.6 0.7 1.0 INDIRECT TAXES 1960 44.7 1970 48.8 1980 50.4 1990 56.0 41.8 47.6 50.4 56.2 34.4 40.0 43.1 49.0 24.6 30.3 33.2 37.8 14.7 19.9 23.5 26.7 8.3 11.4 14.5 16.9 4.3 6.6 8.6 9.8 31.3 35.8 38.7 43.6 9.2 12.5 15.2 17.4 27.7 28.4 29.6 34.6 32.1 32.9 36.2 39.1 31.2 30.7 34.6 37.6 23.5 24.7 25.6 28.1 12.6 15.7 16.9 13.3 7.3 7.8 12.3 3.5 25.5 25.3 27.0 30.0 14.8 16.6 17.8 14.6 33.0 44.8 52.0 59.6 36.5 57.2 71.9 82.9 22.3 40.9 57.0 68.8 7.6 22.2 34.7 45.3 22.5 30.5 35.5 40.2 24.9 42.5 57.1 67.9 Age Group: 20-29 LABOR INCOME TAXES 1960 1970 1980 1990 PAYROLL TAXES 1960 1970 1980 1990 CAPITAL INCOME TAXES 1960 1970 1980 1990 19.9 19.2 21.7 26.7 SOCIAL SECURITY BENEFITS (OASDI) 1960 1970 1980 1990 11.2 12.4 13.5 14.9 16.5 19.4 21.0 22.9 24.6 29.3 33.4 36.3 HEALTH BENEFITS (Medicare and Medicaid) 1960 1970 1980 1990 12.4 17.2 21.8 27.4 13.4 21.4 28.8 36.2 13.6 23.9 37.0 49.8 13.1 24.8 41.9 65.0 9.9 24.6 44.5 75.1 5.2 19.1 36.3 63.8 1.8 12.5 24.9 44.1 11.9 21.4 32.6 47.6 5.9 19.5 36.4 62.9 6.2 9.4 9.8 11.7 3.6 5.6 5.8 6.9 2.0 3.2 3.6 4.1 1.1 2.1 2.6 3.0 0.6 1.5 2.0 2.3 0.3 1.0 1.4 1.6 4.5 6.8 7.6 8.9 0.7 1.6 2.0 2.3 WELFARE BENEFITS 1960 1970 1980 1990 9.7 13.9 14.5 17.5 Source: Authorsr calculations. clevelandfed.org/research/workpaper/index.cfm Table 4a The Composition of Total Resources Male Cohorts 1960-90 Age Group: 20-29 30-39 40-49 HUMAN WEALTH NON-HUMAN WEALTH PENSION WEALTH GENERATIONAL ACCOUNT Source: Authorsr calculations. 50-59 60-69 70-79 80-89 clevelandfed.org/research/workpaper/index.cfm Table 4b The Composition of Total Resources Female Cohorts 1960-90 Age Group: 20-29 30-39 40-49 HUMAN WEALTH NOW-HUMAN WEALTH PENSION WEALTH GENERATIONAL ACCOUNT Source: Authors1 calculations. 50-59 60-69 70-79 80-89 clevelandfed.org/research/workpaper/index.cfm Table 5a The Components of Generational Accounts As a Share of Total Resources Male Cohorts 1960-90 20-29 30-39 40-49 50-59 60-69 70-79 80-89 20-89 65-89 LABOR INCOME TAXES 1960 0.15 1970 0.16 1980 0.17 1990 0.18 0.13 0.14 0.15 0.16 0.10 0.11 0.11 0.12 0.07 0.07 0.07 0.06 0.03 0.03 0.02 0.02 0.01 0.01 0.01 0.01 0.01 0.00 0.00 0.00 0.11 0.12 0.12 0.12 0.01 0.01 0.01 0.01 0.12 0.16 0.19 0.20 0.10 0.13 0.15 0.17 0.07 0.09 0.11 0.12 0.05 0.06 0.07 0.07 0.02 0.02 0.02 0.02 0.01 0.01 0.00 0.01 0.00 0.00 0.00 0.00 0.08 0.11 0.12 0.13 0.01 0.01 0.01 0.01 INDIRECT TAXES 1960 0.11 1970 0.12 1980 0.12 1990 0.13 0.10 0.10 0.11 0.11 0.10 0.09 0.09 0.09 0.09 0.08 0.08 0.07 0.07 0.07 0.06 0.06 0.05 0.05 0.05 0.05 0.04 0.05 0.04 0.05 0.10 0.10 0.10 0.10 0.06 0.06 0.05 0.05 0.06 0.06 0.06 0.07 0.08 0.07 0.07 0.08 0.11 0.08 0.08 0.08 0.12 0.09 0.08 0.07 0.10 0.08 0.08 0.05 0.07 0.07 0.07 0.04 0.07 0.06 0.06 0.07 0.10 0.09 0.08 0.06 0.09 0.11 0.11 0.12 0.16 0.18 0.20 0.18 0.17 0.20 0.20 0.20 0.11 0.20 0.19 0.24 0.06 0.07 0.08 0.08 0.17 0.20 0.21 0.20 0.02 0.07 0.10 0.14 0.01 0.08 0.11 0.19 0.03 0.04 0.06 0.09 0.02 0.07 0.10 0.14 0.01 0.01 0.01 0.01 0.00 0.01 0.01 0.01 0.01 0.02 0.02 0.02 0.01 0.01 0.01 0.01 Age Group: PAYROLL TAXES 1960 1970 1980 1990 CAPITAL INCOME TAXES 1960 0.04 1970 0.04 1980 1990 0.05 0.06 SOCIAL SECURITY BENEFITS (OASDI) 1960 0.03 0.04 0.06 1970 1980 1990 0.03 0.03 0.04 0.04 0.04 0.05 0.06 0.07 0.07 HEALTH BENEFITS (Medicare and Medicaid) 1960 0.03 0.02 0.03 0.03 1970 0.04 0.04 0.04 0.05 1980 0.05 0.05 0.06 0.07 1990 0.07 WELFARE BENEFITS 1960 0.02 1970 1980 1990 0.02 0.02 0.02 0.07 0.08 0.10 0.03 0.06 0.09 0.13 0.01 0.02 0.02 0.02 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 Source: Authors1 calculations. clevelandfed.org/research/workpaper/index.cfm Table 5b The Components of Generational Accounts As a Share of Total Resources Female Cohorts 1960-90 Age Group: 20-29 30-39 40-49 50-59 60-69 70-79 80-89 20-89 65-89 0.18 0.18 0.19 0.18 0.12 0.13 0.14 0.14 0.07 0.08 0.08 0.09 0.04 0.04 0.04 0.04 0.01 0.01 0.01 0.01 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.08 0.08 0.09 0.09 0.00 0.00 0.00 0.00 0.19 0.19 0.21 0.21 0.12 0.14 0.16 0.16 0.07 0.08 0.09 0.11 0.03 0.04 0.04 0.05 0.01 0.01 0.01 0.01 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.07 0.09 0.10 0.10 0.00 0.00 0.00 0.00 0.46 0.30 0.25 0.23 0.37 0.28 0.21 0.19 0.22 0.20 0.16 0.15 0.14 0.14 0.11 0.11 0.10 0.09 0.08 0.07 0.07 0.07 0.06 0.06 0.05 0.08 0.05 0.08 0.24 0.19 0.16 0.15 0.08 0.08 0.07 0.06 0.25 0.17 0.12 0.12 0.21 0.16 0.13 0.12 0.18 0.14 0.12 0.11 0.16 0.11 0.09 0.08 0.11 0.10 0.07 0.05 0.08 0.09 0.07 0.03 0.19 0.14 0.11 0.10 0.12 0.10 0.08 0.05 0.19 0.20 0.18 0.17 0.25 0.26 0.26 0.23 0.19 0.25 0.25 0.25 0.08 0.27 0.21 0.35 0.17 0.16 0.15 0.14 0.21 0.26 0.25 0.25 LABOR INCOME TAXES 1960 1970 1980 1990 PAYROLL TAXES 1960 1970 1980 1990 INDIRECT TAXES 1960 1970 1980 1990 CAPITAL INCOME TAXES 1960 1970 1980 1990 0.21 0.12 0.11 0.11 SOCIAL SECURITY BENEFITS (OASDI) 1960 1970 1980 1990 0.12 0.08 0.07 0.06 0.15 0.11 0.09 0.08 0.16 0.15 0.12 0.11 HEALTH BENEFITS (Medicare and Medicaid) 1960 1970 1980 1990 0.13 0.11 0.11 0.11 0.12 0.12 0.12 0.13 0.09 0.12 0.13 0.15 0.08 0.11 0.14 0.19 0.07 0.11 0.16 0.21 0.04 0.12 0.16 0.23 0.02 0.15 0.15 0.34 0.09 0.12 0.13 0.16 0.05 0.12 0.16 0.23 0.06 0.05 0.04 0.04 0.02 0.03 0.02 0.02 0.04 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.00 0.01 0.01 0.01 0.03 0.04 0.03 0.03 0.01 0.01 0.01 0.01 UELFARE BENEFITS 1960 1970 1980 1990 0.10 0.08 0.07 0.07 Source: Authors1 calculations. clevelandfed.org/research/workpaper/index.cfm Table 6a The Components o f Bequeathable Resources - Male Cohorts 1960-90 (Population Weighted Averages i n Thousands o f 1992 Dollars) Age Group: 20-29 30-39 40-49 50-59 60-69 70-79 80-89 20-89 TOTAL RESOURCES NON-HUMAN WEALTH TERM VALUE OF LIFE INSURANCE BEQUEATHABLE WEALTH (Non-Human Wealth Plus Term Value o f L i f e 1nsurance) ANNUITIZED WEALTH (Total Wealth Minus Bequeathable Wealth) Source: Authors1 calculations. 65-89 clevelandfed.org/research/workpaper/index.cfm Table 6b The Cunponents of Bequeathable Resources - Female Cohorts 1960-90 (Population Weighted Averages i n Thousands of 1992 Dollars) Age Group: 20-29 30-39 40-49 50-59 60-69 70-79 80-89 20-89 TOTAL RESOURCES NON-HUMAN WEALTH TERM VALUE OF LIFE INSURANCE BEQUEATHABLE WEALTH (Non-Hunan Wealth Plus Term Value of L i f e Insurance) ANNUITIZED WEALTH (Total Wealth Minus Bequeathable Wealth) Source: Authorsf calculations. 65-89 clevelandfed.org/research/workpaper/index.cfm Table 7a Bequeathable and Annuitired Resources As a Share of Total Resources Male Cohorts 1960-90 Age Group: 20-29 30-39 40-49 50-59 60-69 70-79 80-89 20-89 TOTAL RESWRCES NON-HUMAN WEALTH TERM VALUE OF LIFE INSURANCE BEQUEATHABLE WEALTH (Non-Human Wea 1th P Lus Term Value of L i f e Insurance) ANNUITIZED WEALTH (Total Wealth Minus Bequeathable Wealth) Source: Authors1 calculations. 65-89 clevelandfed.org/research/workpaper/index.cfm Table 7b Bequeathable and Annuitized Resources As a Share o f Total Resources Female Cohorts 1960-90 Age Group: 20-29 30-39 40-49 50-59 60-69 70-79 80-89 20-89 TOTAL RESOURCES NON-HUMAN WEALTH TERM VALUE OF LIFE INSURANCE BEQUEATHABLE WEALTH (Non-Human Wealth P Lus Term Value o f L i f e Insurance) ANNUITIZED WEALTH (Total Wealth Minus Bequeathable Wealth) Source: Authors1 calculations. 65-89 clevelandfed.org/research/workpaper/index.cfm Table 8 Ratio of Annuitized t o Total Resources under Different Interest Rate ( r ) Assunptions Males Fema 1es Source: Authors1 calculations clevelandfed.org/research/workpaper/index.cfm Table 9 Ratios of Annuitized to Total Resources under Alternative Health Care Spending Outcomes Year Base Case Clinton Health Health Reform Reform Uith 2% Faster Stabilizing Health Care Cost Grouth Spending a f t e r 1994 Fema 1es Source: Authors1 calculations. clevelandfed.org/research/workpaper/index.cfm Table 10 Annual Bequest Flow i n 1990 Based upon Annui t i z a t i o n Ratios for 1990 and 1960 Alternative Health Care Spending Outcanes (Billions of 1992 Dollars) Year 1960 ratio 1990 ratio percent difference Base Case Clinton Health Health Reform Reform Uith 2% Faster Cost Growth Stabilizing Health Care Spending a f t e r 1994 360.8 356.6 357.2 345.4 245.1 245.1 245.1 245.1 47.2 45.5 45.8 40.9 Source: Authors1 calculations. clevelandfed.org/research/workpaper/index.cfm Table 11 Annual 1990 Bequest Flow Based upon Annui t i z a t i o n Ratios f o r 1990 and 1960 Alternative I n t e r e s t Rate ( r ) Assunptions ( B i l l i o n s of 1992 dollars) Bequest Flows 1960 ratio 381 .O 360.8 345.0 1990 ratio 245.1 245.1 245.1 percent difference 55.5 Source: Authors1 calculations.