Death and Taxes

Background

Many people consider the Social Security program to be the crown jewel in Franklin Roosevelt’s New Deal. Legislated in 1935, it began to pay benefits in 1940. For covered individuals reaching age 65, there were retirement pension and survivor benefits for dependent families of covered workers who died prior to reaching retirement age. In the mid-1950s, disability insurance was added. In the late 1950s, early retirement was offered to women at age 62 and extended to men in 1961. In the mid-1960s, Medicare began to provide retiree medical insurance. In the early 1970s, benefits under the program were automatically indexed so they could stay abreast of workers’ wage growth prior to retirement and to inflation once benefits were claimed.

While many policy makers considered Social Security a great achievement, there was a widespread understanding that it did not provide sufficient income to assure that people retiring from career employment could maintain preretirement standards of living. To help retirees do so, employer-sponsored pensions were encouraged through the use of tax incentives. During World War II, many private employers offered pensions (and health benefits) to their workers in lieu of pay increases, which were strictly restricted under wage and price controls then in effect. After the War, pensions became a negotiable item for unions and there was a literal stampede to set up plans in the late 1940s and early 1950s. Some spectacular business failures and other practices that resulted in workers losing pension benefits led Congress to adopt the Employee Retirement Income Security Act (ERISA) in 1974. Subsequent amendments to ERISA have further attempted to bolster pension regulations in order to assure that workers earning retirement benefits will ultimately receive them.

After the early retirement age was implemented in Social Security, unions began to negotiate for early retirement benefits. By the latter part of the 1970s, many employers were allowing workers to retire as early as their mid-50s under their pensions. It was common that such early retirement benefits were subsidized with less than full actuarial reductions taken from monthly pension checks relative to “normal retirement age” benefits, usually payable at age 65. As we approached the end of the 20th century, employer-sponsored retirement plans were paying out benefits, in aggregate, that were larger than annual Social Security benefit payments.

Since 2000, the golden age of retirement seems long past. The Center for Retirement Research at Boston College now publishes an annual National Retirement Risk Index, which “measures the percentage of working-age households that are at risk of being unable to maintain their pre-retirement standard of living in retirement.” In their latest issue of the index, they find that the baby boom generation and the one behind faces a more challenging retirement period than current retirees. Among working-age households, 51% are at risk of not having adequate resources to maintain their living standard in retirement. Including health expenditures in the picture increases those at risk to 61% and adding in long-term care prospects drives it up to 65%. The proposed solution: “Saving more and working longer.” ⁵ In the most recent Employee Benefit Research Institute Retirement Confidence Survey, more than two thirds of workers indicated that they expect they will have to continue working is some capacity even after they “retire.” ⁶

Understanding the New Economic Reality of Retirement

My father reached age 65 in 1974; I reached that age in 2011. When my parents, their siblings and contemporaries reached their late 50s and early 60s, they felt good about their retirement income security prospects. Social Security seemed like a good deal with its automatic inflation protection and their pensions seemed more secure than ever. So many in my generation wonder why our parents felt so good about their prospects at the age we are reaching today or, conversely, why our outlook looks so much gloomier than theirs as we progress into our 60s. Why have things seemingly gotten off track in the period of a single generation and what are the implications even further down the line for our children and theirs?

Part of the explanation for our different outlook is that our retirement system has become much more expensive for the current generation of retirees than it was for those retiring 30 or 40 years ago. Table 1 shows the costs that various generations of workers reaching age 65 in the past would have incurred in order to achieve a retirement income that would have roughly allowed them to maintain their preretirement living standard once they retired. For the sake of discussion, consider the case of a worker, call her Jane, who started to work at age 21 and worked steadily until she reached age 65 and retired. In developing Table 1 and the example, I assumed that Jane would have earned roughly the average wages paid to workers in the U.S. economy throughout her career. ⁷ If Jane had reached age 65 in 1955, she and her employer(s) would have paid 2.2% of her lifetime wages to finance her Social Security benefits. For the Jane who turned 65 in 1965, the lifetime contributions would have climbed to 3.6% of her lifetime earnings, showing the effects of the payroll tax rates that had started to climb in the early 1950s.

OPEN IN VIEWER

Table 1.

Cumulative Lifetime Employee Plus Employer Payroll Taxes as a Percentage of Cumulative Lifetime Earnings, Supplemental Retirement Savings Rate and Totals as a Percentage of Pay for Workers Retiring at Various Dates

Year reaching age 65	Lifetime payroll tax as percentage of lifetime earnings	Required supplemental savings rate	Total retirement cost per year
1955	2.1	4.6	6.7
1965	3.6	5.4	9.0
1975	5.9	5.9	11.8
1985	9.0	6.1	15.1
1995	9.9	6.7	16.6
2005	12.0	7.1	19.1
2011	13.1	7.5	20.6

Note. All amounts stated as a percentage of salary or wages.

Source. P. 241 of Schieber, S. J. (2012). The predictable surprise: The unraveling of the U.S. retirement system. Oxford, England: Oxford University Press.

In the mid-1960s, Medicare was added to the retirement package under the Social Security Act and it was partially financed by payroll taxes. Each subsequent Jane reaching age 65 in later years would have paid successively higher payroll taxes throughout her working life. The one who turned age 65 in 2011 would have incurred payroll taxes equal to 13.1% of her lifetime pay. The average cost rates at various 65th birthday years appear in the second column in Table 1. At almost every point in time, Jane’s Social Security benefit would have been around 40% to 45% of final pay when she retired, but the cost over her career would have depended on when she retired. The Jane who retired in 2011 would have paid over 6 times more than the one who retired in 1955.

The third column in Table 1 shows the additional saving rate that would have been required to provide Jane with benefits supplementing Social Security that would replace another 40% of her preretirement earnings. The column shows the percentage of Jane’s income required to finance her supplemental benefit. The Jane retiring in 1955—or her employer—would have to have saved 4.6% of her pay each year. For the one retiring in 2011, 7.5% of pay would have been required. The cost of the supplemental benefits rose over time because workers were retiring earlier toward the end of the 20th century than they had in the 1950s or 1960s and because of improving life expectancies at older ages. Because of these phenomena, the Jane retiring in 1990 or 2000 has fewer years over which to accumulate her savings and the savings must last over a longer retirement than for the Jane who retired in 1965 or 1975. Changes in the supplemental savings rates were much smaller than they were for Social Security because of the greater efficiency of funded versus pay-as-you-go systems.

The fourth column helps explain why many people today find the sort of retirement their parents and grandparents enjoyed beyond their grasp. That column sums the percentages of salary required to finance Social Security benefits and supplemental pensions for workers of various vintages. For a worker retiring in 1955, the combination of Social Security and a roughly matching supplemental benefit would cost slightly less than 7% of lifetime earnings. My father’s generation retiring at age 65 in the mid-1970s would have spent around 11.8% of their lifetime earnings to provide for retirement income of around 75% to 80% of their preretirement earnings. For my generation coming to retirement today, the cost has risen to around 20.6% to finance a similar preretirement income replacement. The differential in retirement costs between people reaching age 65 today compared to the mid-1970s over a roughly 40- to 45-year career would amount to between 3.5 and 4 years of additional earnings that workers coming to retirement today would have had devote to finance a retirement income of about three quarters of their preretirement earnings level.

What happens to the lifetime payroll tax contribution rates in the future depends on how policy makers address the underfunding in federal retirement programs under current law. The 2012 Social Security Trustees report estimates that the payroll tax rates needed to sustain Social Security benefits defined in current law by 2030 will be more than 16% of covered payroll compared to the existing payroll tax rate of 12.4%, which has been temporarily reduced to 10.4% the past 2 years. ⁸ If payroll taxes are simply increased to cover currently legislated benefits, our children and grandchildren will be required to surrender 2 to 3 additional years of their lifetime earnings compared to people retiring today.

The Changing Social Security Deal

Part of the problem with the growing cost of retirement, especially in terms of steadily growing payroll taxes, is that current and future generations are getting low returns on that investment in their retirement security. For the worker retiring in 1975 at age 65, Social Security was an extremely beneficial economic deal. A single male retiring in 1975 who had average earnings over his lifetime could expect to receive benefits actuarially worth $134,000, in 2009 dollars, more than the payroll tax contributions accumulated with interest that were paid on his lifetime earnings. For a single worker who steadily earned at the maximum level subject to the payroll tax, the net added benefit over contributions was $150,000. For single-earner couples, the net added benefits over contributions were $285,000 for a worker with lifetime average earnings and they were $325,000 for a worker whose wage was consistently at the taxable maximum under the payroll tax. With these sorts of surplus benefits, it is easy to understand why both policy makers and program participants thought Social Security was a great deal. ⁹

The reason that Social Security could pay such high benefits in the early decades of its operations was that it was operated on a pay-as-you-go basis. Originally, about 60% of all U.S. workers were paying into the system, but the only people who received benefits in 1940, when first benefits were paid, were people who had been younger than 65 years 1937 and had paid into the system during 1937 to 1939 and were age 65 in 1940. This was a miniscule portion of the elderly population at the time, but the early beneficiaries received benefits as though they had participated in the system over most of their career. So there were many workers supporting few retirees. Even though benefits were generous for those who received them, there were so many workers relative to beneficiaries that the payroll tax rate could be maintained at much lower levels than would have been needed to finance a funded retirement benefit. But because none of the contributions were “saved” to pay future pensions, workers’ contributions did not accumulate interest over workers’ careers to help support their retirement benefits when they claimed them. Over time, as the system matured, the number of beneficiaries relative to contributors grew and the beneficial deal eroded.

Dean Leimer, a research analyst at the Social Security Administration, developed an analysis in the mid-1990s that showed that ultimately the net gains from participating in the program would turn negative. His estimate was that each collective birth cohort—the group born in a single year—from those born in 1938 onward would get back less in lifetime benefits than the accumulated value of contributions credited with interest on their lifetime earnings. ¹⁰ In other words, Social Security was going to shift from being a great economic or actuarial deal to one that was more expensive than the value of benefits that could be provided through a funded system, even one where the funds were invested in government bonds. The folks born in 1938 turned 62 in 2000 so the economics of Social Security tilted from being a good economic deal to a bad one somewhere around the beginning of the new millennium.

The Office of the Actuary at the Social Security Administration recently developed an analysis that compares the expected lifetime benefit levels from Social Security under current law to their lifetime contributions accumulated at interest for workers born in 1949. They estimate that for a single male with medium earnings, one with lifetime earnings slightly above average, can expect to get back lifetime benefits worth about 67% of the accumulated value of his contributions including interest. For the single female, benefits increase to 76% of contributions because her longer life expectancy would increase her cumulative benefits somewhat compared to her male counterpart. For two-earner couples with medium earnings, the expected combined benefits are estimated to be 77% of their accumulated contributions. For single-earner couples, the program is still a good deal, expected to pay benefits equal to 136% of accumulated contributions. For workers who earned the maximum taxable earnings throughout their careers, single males can expect to get back benefits equal to 46% of lifetime contributions, single females and two-earner couples 52%, and one-earner couples 92%. ¹¹ While these estimates of the relationship between lifetime contributions and expected lifetime benefits for the 1949 birth cohort give a sense of the contemporary efficiency of Social Security, they do not provide perspective on how Social Security fits into the larger retirement portfolio of most workers and its relative costs and benefits.

To understand the total effects of public policies on typical workers’ costs and benefits in the retirement system, we also need to consider participation in tax-favored retirement plans. To show this, I used the lifetime earnings streams of four of the workers that the actuaries used in developing their analysis of costs and benefits under Social Security but calculated the values to show in dollar terms rather than the ratios shown in Table 1. The benefits and contributions in both systems are maximized by assuming that the workers survive with certainty until retirement age. Retirement benefits were calculated using Social Security’s benefit calculator. ¹² The benefits associated with disability or early survivor benefits paid when a worker dies prior to retirement eligibility and leaves behind juvenile children were not considered. The accumulated contributions and present value of benefits for the hypothetical workers are shown in Table 2.

OPEN IN VIEWER

Table 2.

Value of Lifetime Contributions and Present Value of Expected Social Security Benefits for Hypothetical Workers Born in 1949 Retiring at Age 65 by Earnings Level

Value at retirement date	Low earner	Medium earner	High earner	Maximum earner
Lifetime payroll taxes ($)	159,225	353,800	566,077	898,346
Present value of single male benefit ($)	166,175	273,049	365,057	402,884
Net lifetime gain for single male ($)	6,950	−80,751	−201,020	−495,462
Present value of single female benefit ($)	185,478	304,767	407,462	449,682
Net lifetime gain for single female ($)	26,253	−49,033	−158,615	−448,664
Present value of one-earner couple benefit ($)	330,524	554,229	730,113	789,968
Net lifetime gain for one-earner couple ($)	171,299	200,429	164,036	−108,378
Lifetime payroll taxes by two-earner couple ($)	318,450	707,600	1,132,154	1,796,692
Present value of two-earner couple benefit ($)	370,956	609,534	814,924	899,364
Net lifetime gain for two-earner couple ($)	52,506	−98,066	−317,230	−897,328

Source. P. 285 of Schieber, S. J. (2012). The predictable surprise: The unraveling of the U.S. retirement system. Oxford, England: Oxford University Press.

To understand the computations, consider the case of the “medium earner” whose results are shown in the second column of numbers in the table. The calculation estimates that payroll taxes paid on this worker’s earnings accumulated at Social Security trust fund interest rates would be worth $353,800 just prior to retiring at age 65. At this earnings level, a single male’s lifetime value of benefits assuming he lives a normal life expectancy is $273,049. This suggests that he would be a net loser to the tune of $80,751 for being included in the program. For a single woman, on the other hand, the lifetime expected benefit would be $304,767 because she would be expected to draw benefits for approximately 3 more years than a man her age. Since her payroll taxes would be of equivalent value to those paid by her male counterpart, her net loss for participating in the program would be $49,033. Following on through the calculations, one-earner couples could expect to receive lifetime benefits of $554,229 and be net winners of roughly $200,000. Two-earner couples where both earned medium earnings would have accumulated lifetime payroll taxes of $707,600 and benefits of $609,534 to be net losers of around $98,000. If the ratios of benefits to payroll taxes were calculated, they would show a similar pattern to those developed by the actuaries although they would tend to be different because these lifetime payroll tax values only include the Old-Age and Survivors Insurance component of the total. The benefits paid to one-earner couples completely change the nature of the program for this select group of participants compared to everyone else.

When these results for the 1949 birth cohort are compared with individuals born in 1910 and reaching age 65 in 1975 as discussed above, the implications are profound. For the single male medium worker retiring in 1975, Social Security paid benefits worth $134,000 more in 2009 dollars than lifetime contributions plus interest on the earnings used to calculate the benefits, whereas the 1949 worker will get back $81,000 less than accumulated contributions plus interest. For the maximum earner, the swing was from $150,000 in surplus benefits over contributions for the 1975 retiree versus a net loss of nearly $500,000 for the fellow coming to retirement in 2014.

In other words, to provide this worker with a combined package of benefits at age 65 that will allow him to maintain his preretirement standard of living, it requires accumulated contributions with interest of a half million dollars over and above what he will actually receive in the form of retirement income. Where the retirement system 25 or 30 years ago was heavily subsidized through Social Security operations, it is now heavily taxed. This is the major reason that the payroll tax contribution rates shown in Table 1 have risen so markedly for subsequent generations of workers to finance relatively comparable replacement of preretirement earnings across time.

Employer Plan Regulation: Lofty Goals but Failed Execution

Another element of the data in Table 1, contributions for supplemental pensions or savings, needs further elaboration because our actual contribution patterns do not bear much resemblance to what is reflected in the table. In developing those estimates, I assumed that workers or their employers would start contributing to their plans on a steady basis at the beginning of their career and contribute consistently throughout the time leading up to retirement. I also assumed that the accumulating assets would generate a steady stream of positive returns over the career period. The reality of what has been happening in recent decades is somewhat inconsistent with the assumptions used in developing Table 1.

In 1974, Congress adopted the ERISA with the intent that private sector workers’ retirement benefits would be secured against fiduciary malfeasance, unfair manipulation by plan sponsors and funded as they were earned. In the case of defined contribution plans, contributions had to be made during the period in which the benefits were earned. In the case of defined benefit plans, funding was to be achieved over time, allowing for some reasonable variation around a long-term trend toward plan sponsors laying aside adequate funds over covered workers’ careers to assure that by the time they reached retirement there would be sufficient assets to cover their expected benefits over their remaining lifetime. In 1975, as ERISA was beginning to take effect, there were 3 times as many total participants in private defined benefit plans as defined contribution plans and there were 2.5 times as many active participants in the former compared to the latter. ¹³

Once the initial transition to ERISA was accomplished, the rates of growth in the numbers of plans accelerated considerably from the initial aftermath of the law’s passage. During the 1979 through 1982 period, the number of defined benefit plans grew at an annual rate of 10.9% per year and the number of defined contribution plans at 17.3% per year. The number of active participants in defined contribution plans roughly doubled over the period, about the same as the expansion in the number of plans. This suggests that the new plans sponsors were similar in size, on average, to the prior ones. Over the same period, the number of participants in defined benefit plans increased by about 9% while the number of plans had increased by about 70% over the period. The patterns of growth, in this case, suggest that the new plans were much smaller, on average, than prior plans. ¹⁴ A substantial part of the growth in coverage under the defined contribution plans, however, was supplemental coverage extended to workers already participating in defined benefit plans. The Section 415 provisions in ERISA allowed plans that had some workers whose defined benefit pensions were newly limited under the law to provide an additional defined contribution benefit.

ERISA introduced service requirements for participation and vesting in private retirement plans with the hope that these would expand the share of the workforce receiving benefits where employers were sponsoring plans. During the first 5 years of ERISA’s phase in, benefits from private defined benefit grew at a rate of 0.7% per year in inflation-adjusted terms and those being paid by defined contribution plans grew at a rate of 1.9% per year. From 1979 through 1982, they began to grow more rapidly, at a rate of 10.9% per year for the defined benefit plans and at 17.3% per year for defined contribution distributions. ¹⁵

Under ERISA’s early operation, workers’ private pensions appeared to be increasingly secure. One measure of a pension plan’s security is the extent to which accrued benefits—those earned by participants at a point in time—are secured with assets. If the aggregate accrued benefits are fully matched by assets appropriately valued, the plan is said to be fully funded. The percentage of large plans that were fully funded rose from 25% in 1977 to roughly 85% by 1986. ¹⁶ In other words, ERISA was successful in its early years in encouraging employers to fund pension obligations.

Little noticed at the time legislators were finalizing ERISA, Congress also adopted the Congressional Budget Act in 1974, formalizing the concept of “tax expenditures” as a required element of the budget document that the president submits to Congress each year. Until then, the tax expenditures were a theoretical concept, but the Congressional Budget Act defined tax expenditures as “revenue losses attributable to provisions of the federal tax laws which allow special exclusion, exemption, or deduction from gross income or which provide a special credit, a preferential tax rate, or a deferral of tax liability.” ¹⁷ The notion that tax breaks cost the government revenue just as a spending program costs the government revenue had been formally legitimized in the Congressional Budget Act and so “tax expenditures” from tax breaks were now going to be calculated along with actual spending. Soon, the largest single category of tax expenditures was the costs of the tax preferences accorded employer-sponsored retirement plans.

There have been many criticisms of the concepts behind these measures and the methods used to estimate them. There are demographic issues, timing issues and even fundamental questions about what tax treatment would be on funds invested with posttax dollars. ¹⁸ The idea that the government was incurring an expenditure when employers made contributions to their employees retirement plans, often extra contributions under the new government-mandated funding requirement, did not sit well with plan sponsors and their outside professional advisors. It was clear that the matter was controversial as a conceptual idea and its direct use in consideration of public policy would be even more so. Thus, without a clear motivating event, the use of the concept was likely to lay dormant. That motivating event arose in the early 1980s.

Ronald Reagan campaigned for the Presidency during 1980 on a platform of reducing federal income taxes if he was elected. After taking office, he advocated and Congress passed the Economic Recovery Tax Act of 1981, which called for a 3-year phased reduction in tax rates. Having reduced the projected revenue from federal income taxes, his administration prepared its next fiscal budget calling for reductions in expenditures. Among them, the plan called for dramatic reductions in Social Security early retirement and disability benefits. The public was stunned by the proposals; those close to retirement age were particularly outraged. The reaction created a political consensus probably not matched again until the terrorists’ attacks against the United States on September 11, 2001. In response to the Reagan proposals, the Senate unanimously (96:0) adopted a sense of the Senate Resolution stating that “Congress shall not precipitously and unfairly penalize early retirees.” The author of the resolution was Bob Dole, the Republican Chairman of the Senate Finance Committee. ¹⁹ In the floor discussion of Dole’s proposal, not a single Republican rose to defend the President. ²⁰

In the House of Representatives, Representative Charles Rangel, a Democrat from New York, responded to the Reagan proposals by introducing legislation to reduce maximum pensions that could be provided on a tax-preferred basis. His rationale was that if we needed to cut back retirement benefits to help balance the budget, then we should cut them for “fat cats” rather than rank-and-file workers. His proposal ended up as part of the Tax Equity and Fiscal Responsibility Act of 1982. The rationale for cutting the tax-preferred retirement benefits that could be provided under the ERISA was that it would result in added federal tax revenues without raising tax rates. The thinking behind Rangel’s proposal persisted throughout the 1980s and into the early 1990s and motivated a variety of reductions in the benefit and funding limits on employer-sponsored retirement benefits. The alphabet soup of legislation with these provisions included the Deficit Reduction Act of 1983, the Tax Reform Act (TRA) of 1986, the Omnibus Budget Reconciliation Acts (OBRA) of 1987, 1990 and 1993.

The TRA86, OBRA87 and OBRA90 had particularly significant effects on the funding of private defined benefit plans. In the mid-1980s, policy makers became concerned about the phenomenon of plan sponsors terminating their defined benefit plans and taking a reversion of surplus assets. An element of TRA86 was a 10% excise tax over and above regular income tax obligations and any surplus assets that were taken out of a tax-qualified plan in such a reversion. The 1987 legislation changed the full funding limits for private defined benefit plans from an accrued benefit perspective to one based on current liabilities. The 1990 legislation increased the excise tax on reversions to 50%. In other words, if an employer took an asset reversion from a plan, it triggered a federal income tax obligation plus a 50% excise tax plus any state income taxes that the sponsor might be subject to depending on where they were located. The combination was a virtual expropriation of any excess assets if a plan was terminated.

The purpose of the 1987 change in the full funding limits was to delay pension funding in order to reduce employers’ tax deductible contributions in the short term and to increase federal income tax collections without increasing tax rates. When the OBRA87 funding limits took effect in 1988, the leading edge of the baby boomers reaching age 40. This law significantly delayed the funding of the baby boom generation’s defined benefit pension promises. It retarded pension funding in the immediate aftermath of its passage. In an analysis based on a 1986 survey that included 664 plans, each with more than 1,000 participants, developed while the legislation was under consideration, it was estimated that 40% of them would be facing contribution limits under the new proposal, compared with only 7% under prior limits. ²¹ In 1980, private defined benefit plan sponsors contributed $3,747—stated in 2010 dollars—per active participant in their plans. By 1990, private defined benefit plan sponsors contributed $1,466 per active worker in 2010 dollars. In 1999, per capita contributions were $1,732. From the time the OBRA87 was implemented through 2004, the per capita contributions to private defined benefit plans averaged 54% of what they had been in 1980 in inflation-adjusted dollars. ²²

Table 3 shows what happened to private pension funding levels following the 1986 legislation and the subsequent developments. Most of the data is based on a series of surveys that Watson Wyatt Worldwide (now part of Towers Watson) conducted. The survey results track the Department of Labor data in the second column in the table for the years that there is overlap but are more complete, so they are used here to show what happened to pension funding.

OPEN IN VIEWER

Table 3.

Funding of Current Liabilities in Private Defined Benefit Plans for Selected Years

		Watson Wyatt Actuarial Assumption Survey
Plan year	All Private plans median funding ratios	All plans median funding ratios	Percentage of plans overfunded	Median funding ratios of overfunded plans	Percentage of plans underfunded	Median funding ratios of underfunded plans
1987		1.45	83	1.57	17	0.82
1988		1.44	82	1.56	18	0.83
1989		1.40	85	1.50	15	0.86
1990	1.35	1.36	85	1.45	15	0.83
1991	1.30	1.38	85	1.45	15	0.88
1992	1.30	1.33	85	1.41	15	0.89
1993	1.24	1.27	82	1.35	18	0.88
1994	1.18	1.18	76	1.27	24	0.87
1995	1.13	1.15	70	1.25	30	0.88
1996	1.15	1.20	81	1.27	19	0.91
1997		1.23	84	1.30	16	0.93
1998		1.24	82	1.34	18	0.92
1999		1.25	82	1.32	18	0.93
2000		1.11	69	1.22	31	0.89
2001		0.99	48	1.22	52	0.86
2002		0.92	35	1.20	65	0.81
2003		0.90	33	1.11	67	0.82

Source. The funding ratios for all private plans are based on unpublished Department of Labor tabulations of plan disclosure data. The Watson Wyatt Worldwide data are taken from annual surveys of a sample of plans with more than 1,000 active participants. These surveys ascertained information on the actuarial assumptions used in valuing pension plans and their funding status from various years.

In 1987, the median plan had 1.45 times current liabilities and 83% of all plans were overfunded. By 1990, 85% of plans were overfunded but the median funding ratio was down to 1.36, meaning that some of the better funded plans in 1987 had dropped the extent to which they were overfunded. The 1990s were unprecedented in terms of asset returns in the financial markets. At the end of 1990, the price-to-earnings ratio in the U.S. stock market was 15.8 and had risen to 44.2 at the end of 1999 as the market started to turn. ²³ At the end of 1990, private defined benefit pension trust funds held $900 billion. By the end of 2000, they held $2 trillion and 63% of private pension assets were invested in equities. ²⁴ Despite an absolutely booming financial market driving up the value of pension assets at unprecedented levels, the median plans steadily became less well funded relative to current liabilities. By 2000, when the stock markets began to dip in the United States the median plan held assets only 1.1 times their current liabilities—the obligations that had been earned to date without taking into consideration future increases in the obligations due to increasing salaries that are an extremely important element of final-average pay plans that were prevalent at the time. During the greatest period in the history of U.S. financial markets to be investing, 16% of private defined benefit plans had shifted from being overfunded to being underfunded relative to current liabilities.

Part of the reason that plans had become less well funded between 1987 and 2000 was the growth in liabilities, which were being driven up by three separate forces. The average ages and service tenures in private defined benefit plans rose relatively rapidly during the late 1980s and throughout the 1990s. One study of 51 large corporate sponsors of defined benefit pension plans in the 1990s found that average ages of workers covered by the plans increased by 1.4 years in a 5-year span while average tenures increased 0.9 years. The phenomenon was occurring in virtually all the plans. ²⁵ Given the basic arithmetic of benefit formulas in defined benefit plans, the first factor driving up plan costs was the increases in average tenures in plans that increased pension liabilities. The second force driving up pension liabilities was the aging of the workforce covered by plans.

When the funding limits were reduced in 1987, the oldest of the baby boomers were 41 years of age and the average age of the group was 32. With each passing year, they aged 1 year closer to making their benefit claims. We often hear about the power of compound interest and how an asset grows ever more rapidly the longer it is invested at some positive rate of return. We seldom consider the mirror image of that phenomenon—what might be called the power of compound discounting. In this case, the closer and closer a benefit obligation comes to being actually paid, the more rapidly its present value grows. By 2000, the oldest of the baby boomers were within a couple of years of qualifying for early retirement benefits in many plans and the failure to fund their benefits earlier in their careers meant not only prior contributions had to be made up but the foregone earnings on those contributions had to be made up as well. The third force in play during the period of the 1990s was falling interest rates, which further accentuated the growth of liabilities in traditional defined benefit plans.

By 2000, when the financial markets started their roughest decade since the Great Depression, pension plan sponsors were standing on the razor’s edge of having their pension plans adequately funded to meet current liabilities. By the end of the first quarter in 2003, the price-to-earnings ratios on U.S. stocks were less than half what they had been at the end of 1999. ²⁶ By the end of 2003, the median private pension plan had assets that would cover only 90% of current liabilities and 67% of the plans were underfunded and worse was yet to come. The stock market rebounded moderately and stabilized after the bottom it hit in early 2003 and stood at a price-to-earnings ratio of 27.4 at the end of June 2007 but then declined to 13.32 by the end of the first quarter of 2009. ²⁷ An extension of funding data in Table 3 is not available to track what happened to pension funding directly in line with what is shown in that table under ERISA but there are data tracking the Financial Accounting Disclosure reports of Fortune 1000 firms that continue to sponsor defined benefit plans in recent years. The median funding ratio for such plans in 2007 disclosures was 0.94, meaning asset values were equal to 94% of projected benefit obligation. By the end of 2008, this ratio dropped to 0.71, in 2009 it was back up to 0.75 and at the end of 2010 it stood at 0.78. ²⁸

The funding ratios do not begin to describe the carnage that underlay them. The softening economic situation early in the new century was greatly exacerbated by events on September 11, 2001. Some companies that could sustain growing pension funding requirements when revenues were strong were simply overwhelmed by their retirement plan “legacy liability” servicing costs when demand for their products and services declined. From January 1, 2000, through December 31, 2009, 11 out of the largest 14 private pension failures occurred that resulted in a claim being made against the Pension Benefit Guaranty Corporation (PBGC). These 11 claims, a total of $27 billion, represented 63% of the total claims against the PBGC’s single-employer insurance program over its total history of operations from 1975 through the end of 2009.

The marquee companies that succumbed in the steel industry were Bethlehem, National, LTV and Weirton; in the airline industry, it was TWA, US Airways, United and Delta; and in the auto industry, it was Delphi, the large parts supplier that had been spun off from General Motors some years prior to its 2009 bankruptcy. ²⁹ The claims from the auto sector would have been much larger had not Congress supplied bailout money for General Motors and Chrysler to survive their economic reorganization in bankruptcy and had not the Obama Administration played a major role in getting the holders of bonds from those two companies to forego legal claims so residual assets could be used to buy off retiree health claims by members in the unions with contracts with the companies.

In the midst of the evolving carnage, policy makers reawakened to the need to fund pensions as they were accruing and passed the Pension Protection Act of 2006. Whereas policy makers thought that it would be prudent to allow pension sponsors 30 years to amortize new unfunded liabilities that arose in their plans when they originally passed ERISA, now they would only allow sponsors 7 years to pay them off. This legislation also set new rules for calculating the pension funding obligations for plans by requiring that the interest rates used in estimating the liabilities be correlated to the time line of the claims that would be made by participants in the plans. The law clearly intended to provide better funding for accruing pension obligations than had been operable policy since the mid-1980s. But two decades over which pensions could have been prudently funded were lost, and it turned out that the need to quickly pay off massive unfunded liabilities in the midst of the worst recession since the Great Depression was not economically feasible, in many cases, or politically palatable. So various amendments to the 2006 funding legislation have allowed plans sponsors some added flexibility. Ultimately, however, the benefit claims have to be paid if the plans are sustained.

Many pension sponsors have found operating these plans an impossible challenge given the combined forces of regulatory requirements and financial market volatility. In 1975, we noted there were 2.5 active participants in private defined benefit plans for each one in defined contribution plans. The number of active participants in the private defined benefit plans peaked at 30.1 million active participants shortly before the legislative onslaught of the 1980s. By 2009, the number of active participants in private defined benefit plans was down by 41% from its 1980 high-water level and there were 3.4 active participants in private defined contribution plans for each one still participating in a defined benefit plan.

The majority of workers now find themselves with only defined contribution plans to earn a supplemental retirement benefit to augment Social Security. Because of concerns about workers’ ability or willingness to use these plans effectively, various plan features have been reengineered in recent years to make them work more like defined benefit plans. Workers are now automatically enrolled and scheduled for default contribution rates with the option of declining to participate or to do so at the level the sponsor prescribes. Their monthly savings are often invested in default life cycle funds that take into account their age and time until retirement unless they choose to override the default and direct their own investments. Despite all these advancements, defined contribution participants have been as badly battered by the financial market machinations of the past dozen years as defined benefit sponsors and the angst the results have created were delineated in the opening discussion.

Recent Legislation to Deal With Retirement Income Security Problems

Earlier this year, Congress passed and President Obama signed the Moving Ahead for Progress in the 21st Century Act. This was supposedly a “highway bill” but included funding relief for private pension plan sponsors. The motivation for this new pension “funding relief” was that “our national economic policy to keep interest rates low has resulted in artificially high pension plan funding obligations . . . Similarly, the current-law requirements to amortize long-term liabilities over seven years have created unmanageable volatility in funding requirements.” ³⁰ The argument was that the Federal Reserve’s expansionist policies were distorting the interest rates being used to value pension obligations and that such rates did not reflect rates that might be “consistent with the long-term nature of pension liabilities.” ³¹ The prevailing low interest rates were driving up pension funding requirements and, because of the slackness in the economy, were threatening job creation and even the economic viability of some pension sponsors.

Under the new law, pension obligations can be calculated using a 25-year average of historical interest rates rather than current rates as required in prior law. Recognizing that adopting this would slow private pension funding, Congress increased the PGBC premium rates to cover the added potential risk of companies going bankrupt and putting unfunded obligations to the government’s insurance program. There are a number of reasons to be concerned about this legislation.

The first is that it addresses symptoms and not causes of pension underfunding and could exacerbate problems some plan sponsors are already facing. One does not have to read very far into the financial pages of the news media to find that many companies are sitting on record levels of cash holdings these days. The idea that companies in this situation are inhibited from offering someone a job because of a pension funding obligation that was earned yesterday or maybe 10 years ago lacks credibility. Indeed, there may be some companies with inadequate resources now to meet high pension funding burdens, but why adopt universal policies that may generally exacerbate pension funding problems to deal with a limited problem? The major inhibition to the funding of private defined benefit plans is the excise tax on any assets that revert to a plan sponsor when the plan’s liabilities are settled.

Richard Ippolito, the former chief economist at the PBGC, was concerned about the evolving trends in pension funding that developed during the 1990s as reflected in Table 3 above. Among other things, he knew that the prospects of default claims on the pension insurance program increased as pension funding levels fell. Ippolito developed an empirical analysis of funding patterns across nearly 20,000 plans. He looked at the effects of both the reversion taxes imposed on plans in 1986 and raised in 1990 and at the funding limitations imposed by the OBRA87. He concluded that the effects of the reversion tax were about 3 times those of the funding limit changes. ³² Ippolito estimated that if the reversion taxes had not been implemented, employers’ contributions to their pensions would have averaged 6.6% of covered pay per year beyond 1987. He estimated that the 10% reversion tax passed in 1986 reduced contributions by 1.5% of covered pay and that the higher rates imposed in the 1990 legislation reduced contribution rates by another 1.3%. His estimate was that the excise tax on reversions had reduced pension contributions by 42%. But this was his general conclusion. When he split the sample into groups by their levels of funding at the beginning of the period, he found that those with the greatest excess funding reduced their contributions by 60% while those at the lowest levels of funding reduced theirs by only 16%. ³³ Ippolito concluded, “Even in the face of historically high investment returns, plan sponsors succeeded in reducing their excess pension assets by 60 percent.” ³⁴ There are many private pension sponsors today with underfunded pensions who would be willing to fund up their plans fairly quickly if they did not run the risk of interest rates rebounding resulting in their plans becoming overfunded. If this were to occur, they would have no practical way to get back the excess assets in their plans. As a result, some plan sponsors have come up with public policy arguments for why current legal funding claims should be reduced.

Those pension sponsors who believe that interest rates may rebound might be right—but they might also be wrong. Many pundits today argue that our economy has remained so sluggish because of the combined adverse housing markets and financial sector problems that arose during the past decade. Part of it may also be related to slack demand because of an aging population here and elsewhere around the world. Japan is a much older society than ours, and while they seemed to be the economic behemoth of the free world in the late 1970s and 1980s, they are now well into their second “lost decade” of slack economic demand. The economic realignment going on in Europe right now is significantly related to an aging population and may slow recovery of many of the country economies as the full ramifications of evolving policies play out. When the central mass of a population is in its 20s or 30s and forming households, starting and raising families or in its 40s and buying larger homes, there is a lot of consuming of housing, appliances and all the other accoutrements of getting ahead. When that central mass ages into its 50s and 60s, consumption patterns tend to decline relative to income. Given what is going on in most of the developed world, we could have a soft economy for some time. Wouldn’t we be better off if pension sponsors could fund their existing obligations without risking loss of their assets for doing so if interest rates rebound than by playing Russian roulette if they don’t?

The real frosting on the recent legislation is that if policy makers slowed the funding of corporate pensions, it would raise their tax revenues in the short term. Once again, if private employers were to reduce tax deductible contributions to their pensions, more of their current income would be immediately taxable. This turned out to be a double bonanza on the revenue side because the increase in the PBGC premiums also counts as added federal revenue without Congress having to raise anyone’s tax rates. Never mind that no one has called the premium increase the pure and simple head tax that it is on most of the plan sponsors who pose little risk to the pension insurance program.

Future Policy Directions

After President Obama took office, he appointed a bipartisan National Commission on Fiscal Responsibility and Reform to identify policies to improve the federal government’s fiscal outlook in the medium term and to achieve fiscal sustainability over the long term. It was chaired by Erskine Bowles, former chief of staff in the Clinton Administration, and Alan Simpson, a former Republican Senator from Wyoming. This group is often referred to as Bowles–Simpson Commission. All but one of the other 13 members were members of Congress. ³⁵ The exception was Dr. Alice Rivlin, the first Director of the Congressional Budget Office, the Director of the Office of Management and Budget in the Clinton Administration and a former Vice Chairman of the Federal Reserve.

The Bowles–Simpson Commission released its final report, Moment of Truth, on December 1, 2010. ³⁶ The Commission’s charter indicated that if 14 members voted to endorse the package, that it would be submitted to the Congress as a comprehensive package for an up or down vote. In the end, it only received 11 favorable votes. Despite the fact that this effort was unsuccessful in meeting its charter, many people consider its recommendations as a reasonable platform for considering how we might modify current retirement policy to make it more sustainable for the future. The commission had recommendations that would affect both Social Security and employer-sponsored retirement plans.

In terms of Social Security, there were a number of recommendations to bring the system back into financing balance. First, the taxable maximum income would be gradually increased from covering around 83% of total earned income today to covering 90% by 2030 and then would continue to increase accordingly to maintain the 90% rate. Second, the minimum benefit for workers with at least 30 years of covered earnings would be raised to 125% of the federal poverty line. Third, the benefit formula would be adjusted to slow the rate of growth in initial benefits for high earners. Because wages tend to grow somewhat faster over time than prices, the purchasing power of initial benefits tends to grow over time. Beyond retirement, the benefits are then indexed to stay abreast of price inflation in order to maintain retirees’ purchasing power while receiving benefits. Fourth, the government price index used for adjusting retirees’ benefits for inflation would be changed from the consumer price index for all urban workers to the consumer price chained index for all urban consumers. Without getting into the technical differences in the two, this would slow benefit growth due to inflation by approximately 0.3 percentage points per year. Fifth, in recognition that life expectancies continue to climb, once the current increase in the normal retirement age reaches age 67 in 2027, both the early retirement age of 62 and the normal retirement age would be indexed with further improvements in life expectancy. On the basis of the Social Security actuaries’ estimates, this proposal suggests that the normal retirement age would reach age 68 by around 2050 and 69 by about 2075. The early retirement age would move up to 63 and 64 in lockstep. Finally, all state and local government employees hired after 2020 would be covered under Social Security. About 30% of them remain outside the system today.

While the Bowles–Simpson Commission reviewed Social Security policy carefully in their final report, the discussion about the remainder of the retirement system is extremely scant, but the recommendations could have significant ramifications. The issue is framed in the context of controlling tax expenditures, a major driving force that has motivated regulatory policies covering private employer-sponsored retirement plans over the past 30 years. The scope of the Bowles–Simpson Commission discussion on pension and retirement savings programs is “The new tax code must include provisions (in some cases permanent, in others temporary) for the following:” then as the last of five bulleted recommendations is added, “Retirement savings and pensions.” In a figure that follows from the commentary is the substance of their recommendation: “Consolidate retirement accounts; cap tax-preferred contributions to lower of $20,000 or 20% of income, expand saver’s credit.” ³⁷

At about the same time the Bowles–Simpson report was released, another bipartisan group, the Debt Reduction Task Force chaired by Dr. Alice Rivlin, the only person who served on both groups, and former Republican Senator Pete Domenici from New Mexico had an identical proposal for adjusting retirement pension and savings treatment under the tax code. The Domenici–Rivlin Task Force was a bit more expansive in explaining the rationale for their proposal than the Bowles–Simpson report:

The Task Force plan will let most individuals retain the ability to contribute enough to qualified retirement plans to accumulate enough tax-free assets to purchase an annuity that replaces a substantial share of their earnings in retirement. Individuals and employers combined will be able to contribute up to 20 percent of annual earnings to qualified plans, up to a maximum of $20,000 per year, indexed to inflation. However, qualified plans will no longer be a vehicle for wealthy individuals to convert a substantial share of their assets into tax-free retirement assets. In addition, to spur saving by rank-and-file workers, the plan will introduce an expanded and refundable savings credit for taxpayers in the 15 percent bracket. ³⁸

To explore the implications of current tax treatment of retirement saving under ERISA, I used the same four workers’ earnings histories as posited by the Social Security actuaries to develop Table 2. I considered each worker’s potential participation in a defined contribution plan over their career and estimated how the value of their lifetime accumulation would differ from what they would alternatively have accumulated if they had saved in a fully taxable savings account. In the tax-favored account, I assumed their contributions each year were made with pretax dollars and that the earnings on the account were tax free during the accumulation and then taxed at dispersal. In the nonqualified account, I assumed the contributions were made with posttax dollars, that the earnings on the account were taxable each year and that there was no tax obligation at dispersal. The difference in the two is an estimate of the maximum value of the tax preference for retirement savings.

I understand that there are many other ways workers can save and accumulate wealth much more favorably than a taxable savings account but I wanted to estimate the maximum possible benefit under the current system in order to analyze the implications of the proposals for the commissions’ recommendations. The results of my estimates are shown in Table 4. Under the assumptions that I used to develop these estimates, it is clear that the tax treatment of retirement savings under the federal tax code has potentially greater benefit for higher earners than those at lower levels. Despite the fact that tax policy may treat high earners’ retirement savings more favorably than lower earners’, I believe the Bowles–Simpson proposals and similar ones are flawed on equity grounds. I believe they fail to appreciate the starting point that workers face for participating in Social Security and the interaction of the two major components of the retirement system supported through the alternative legs of federal tax policy—payroll and income taxes.

OPEN IN VIEWER

Table 4.

Value of Lifetime Net Benefits from a Tax-Qualified Retirement Savings Account Compared to a Regular Savings Account for Hypothetical Workers Born in 1949 and Retiring at Age 65 by Earnings Level

	Low earner	Medium earner	High earner	Maximum earner
Gross retirement plan accumulation ($)	96,944	359,015	722,417	1,904,088
Assumed marginal tax rate (%)	15	15	25	28
Value of income tax liability at retirement ($)	14,542	53,852	180,604	533,145
Net pension distribution ($)	82,402	305,163	541,813	1,370,943
Accumulated value of savings if taxed as a taxable savings account ($)	65,217	241,521	389,799	937,339
Value of the tax preference versus a regular savings account ($)	17,185	63,642	152,014	433,604

Source. P. 287 of Schieber, S. J. (2012). The predictable surprise: The unraveling of the U.S. retirement system. Oxford, England: Oxford University Press.

To explain my point, the summary results from Table 2 are combined with those from Table 4 in Table 5. Here the combined net lifetime benefits from Social Security minus payroll taxes are added to the net lifetime benefits of the tax preference accorded retirement savings relative to a taxable savings account. These results suggest that after summing the net results of the two systems, higher earners are still net investors in the overall retirement system over their lifetimes to a considerable extent. For example, the high earner is one whose earnings are around 80% of the taxable maximum and single males, single females and two-earner couples are net losers or investors, depending on how you want to characterize it, in the combined systems. Their net investment is much smaller than in Social Security alone but their net benefit is also much smaller than in the tax-qualified system alone. Maximum earners do considerably worse. If there is one group clearly winning the retirement lottery here, it is single-earner couples. It seems there is an equity issue here overlooked by Bowles–Simpson.

OPEN IN VIEWER

Table 5.

Combined Value Social Security Net Benefit Gains Versus Contributions and the Lifetime Net Benefits From a Tax-Qualified Retirement Savings for Hypothetical Workers Born in 1949 and Retiring at Age 65 by Earnings Level

Value at retirement date	Low earner	Medium earner	High earner	Maximum earner
Net lifetime Social Security gains ($)
Single male	6,950	−80,751	−201,020	−495,462
Single female	26,253	−49,033	−158,615	−448,664
One-earner couple	171,299	200,429	164,036	−108,378
Two-earner couple	52,506	−98,066	−317,230	−897,328
Value of the tax preference in tax-favored retirement plan ($)
One-earner couple	17,185	63,642	152,014	433,604
Two-earners couple	34,371	127,284	304,028	867,209
Combined value of Social Security gains plus pension surplus from tax preference ($)
Single males	24,135	−17,109	−49,006	−61,858
Single females	43,438	14,609	−6,601	−15,060
One-earner couple	188,484	264,071	316,050	325,226
Two-earner couple	86,877	29,218	−13,203	−30,119

Source. P. 289 of Schieber, S. J. (2012). The predictable surprise: The unraveling of the U.S. retirement system. Oxford, England: Oxford University Press.

If you consider the implications of the combined proposals in the Bowles–Simpson report it would raise the level of earnings subject to the Social Security payroll tax from around $110,000 to around $185,000 under the current earnings distribution. Someone subject to this over their full working career would be facing a 60% to 70% payroll tax increase over their lifetime. For such an earner, the benefit for someone retiring at age 65 in 2050 would be about two thirds of what it would be under current law. ³⁹ Maximum earners retiring under the system today get back around half the value of lifetime contributions on covered earnings, unless they are a single-earner couple, and this will be worse under current law in the future because of past increases in the payroll tax rate that are still driving up lifetime averages.

This proposal would make that situation much worse. At the same time, the tax-qualified savings proposal in the package would reduce the tax-favored saving for anyone earning over $100,000 per year saving at the proposed percentage limit. Relative to current law, the benefits for workers earning between $100,000 and $200,000 could be reduced by as much as half. Does it really make sense to make so much of the needed adjustment to the retirement system on the back of this group that may be considered reasonably successful but hardly considered super rich by most measures used in our society?

One of the reasons this combination of proposals could be included in a package of this sort is that we tend to think about various aspects of the retirement system in isolation. But when the retiree goes to the grocery store and pays the clerk for the bag of goods that has just been rung up, no one cares whether the money came from Social Security, a pension check or one’s own savings. Much of what is in the Bowles–Simpson package makes sense, and some of it becomes even more justifiable if we think about retirement policy in a broader context as a system as suggested in Table 5 instead of separate disconnected elements. It is clear that we have to do something and do it soon or face much worse consequences down the road. But we must be careful about being overly punitive to a small segment of society.

A larger question that these policy recommendations raise is whether or not the current generation in retirement and those about to retire should be left relatively unscathed by the policy changes to the retirement system. The only changes in the Bowles–Simpson package that would affect current and near retirees is the shift to the alternative consumer price index for inflation adjustment of Social Security benefits in payment status. But this recommendation was made on the basis that the alternative price index more accurately reflects the effects of inflation. Congressman Paul Ryan, the Republican Party’s nominee for vice president in 2012, has crafted a Social Security reform proposal that his political opponents will characterize as out to destroy the system because, among other things, it includes an option for workers to put some of their payroll taxes in individual accounts. I have advocated such accounts myself in the past but believe they have become such a political lightning rod that they are a barrier to serious political conversations about how we can reform the system to make it sustainable. But even his proposal has been sharply criticized because it largely gives a pass to those nearing retirement who can afford to help pay for needed reforms. ⁴⁰ Vice President Joe Biden on the trail in the 2012 campaign has been willing to go even further than Congressman Ryan when asked about dealing with Social Security. He recently said, “Hey, by the way, let’s talk about Social Security. Number one, I guarantee you, flat guarantee you, there will be no changes in Social Security. I flat guarantee you.” ⁴¹

A strong case can be made that those who have retired in recent years and about to retire have played by the rules and “paid for” their benefits (and then some), but the prospects for future generations suggest they are going to get back even less for each dollar they put into Social Security than those nearing retirement today. The relatively low returns that Social Security is paying now and in the future is the result of the extremely generous benefits that were paid to earlier generations of retirees relative to what they paid into the system.

That largesse is history and those funds cannot be reclaimed. Given that the decisions causing much of our current dilemma are long past and cannot be rescinded, doesn’t it make sense that we share the cost burden of those past decisions as broadly as possible? The evidence suggests that the elderly have lower poverty rates than most other broad demographic segments of our society. How can we not ask the better off among those of us who are older to help bear some of this cost when we know that if we do not share in the burden, it will be thrust totally on our children and grandchildren who, in many cases, are suffering much more from recent economic developments than those of us nearing the end of our working days? The last thing we should do to help skate through our current budget woes is cut off saving opportunities for future generations.

George Bernard Shaw said, “We are made wise not by the recollection of our past, but by the responsibility of our future.” What my father had paid for his retirement package when he retired at age 65 in 1974 is really irrelevant. In fact, what I paid for mine if I had retired in 2011 is equally irrelevant. What is important are the trends that have demonstrated themselves between those points and what they portend between now and when my son reaches age 65 in 2041 or his son reaches that age in 2074. Any fair reading of the trends suggests that if we simply let them play out, our children and grandchildren will have to surrender most of, if not all, of the added rewards that earlier generations have received for improving productivity. Is killing the American Dream the inheritance we want to leave them?