Life Expectancy and Public Policy

While differences among groups are important, so too are averages. Half of men are out of the labor force by age 64, half of women by age 62. On average in 2008, life expectancy for 20 year olds was an additional 59 years, or to age 79. That would mean that the average American works 42 or 44 years and is retired for 15 or 17 years (though in reality, most people are not in the labor force all the time). Crunch all these statistics together and you find that, on average, people typically spend roughly one-third of their adult lives in retirement if one accounts for time spent out of the labor force for child-bearing, for education after age 20, and in unemployment. If consumption is spread evenly over the adult life cycle, roughly one-third of lifetime consumption will occur in retirement.

Stanford economist John Shoven points out that most people live as couples. That changes the arithmetic. Males usually marry women somewhat younger and who have somewhat longer life expectancies than themselves. Shoven points out that roughly 30 years will elapse, on average, before both members of a couple consisting of a 62-year-old man married to a 60-year-old woman will die. Typically, they will have worked no more than 40 years. Shoven bluntly asserts, "You can't finance 30-year retirements with 40-year careers without saving behavior that is distinctly un-American." Whether one uses my arithmetic or Shoven's starker version, it is surely fair to ask whether people will be willing to divert from current consumption enough to both support ever-lengthening retirements and pay for the rest of what they want government to do.

Perhaps they will. After all, workers retire earlier in many other developed countries and receive pensions more generous than those in the United States, even though their life expectancies equal or exceed our own. Still, the reaction of U.S. elected officials to current and projected budget deficits suggests that the United States will not readily accept European-level taxes. Europeans are, to be sure, cutting back pension commitments, but they are doing so from levels much higher than those in the United States and facing elderly populations that, relative to total populations, are considerably larger than any anticipated in the United States. The Republican Party wants no tax increases whatsoever. Even most Democrats support permanent extension of most Bush-era tax cuts.

For this reason, supporters of the current social-insurance system -- even as they fight against any cuts at all -- must think about changes in Social Security, Medicare, and other elements of the social safety net that reduce spending in the least damaging ways and that may accomplish other goals. Prominent among such goals should be measures to put in place financial incentives to "nudge" those who can do so without undue hardship to work until later ages than they now do.

Social Security

Two fundamental facts make the prospect of any cuts in Social Security particularly galling. One, Social Security benefits were cut significantly by 1983 legislation. (The law enacting those cuts is still only partly implemented; more cuts are to come.) Two, U.S. benefits look downright parsimonious when compared against those offered in other developed nations.

The 1983 Social Security Amendments, largely modeled on recommendations of a commission appointed by President Ronald Reagan and Congress and chaired by Alan Greenspan, cut benefits two ways. They skipped a cost-of-living increase for current pensioners and built that reduction into the benefit formula. They also put in place what is perhaps the most widely misunderstood and misnamed legislative provision in all U.S. history, the benefit cut that was misleadingly called an increase in Social Security's so-called "normal retirement" age -- from 65 to 66 starting with those born in 1938 and to 67 starting with people born in 1955.

Despite the name, that change had next to nothing to do with when people normally retire or claim benefits. What Congress actually did was to raise the age at which unreduced benefits -- the amount generated by the Social Security benefit formula -- are paid. Congress left unchanged the age when people can and, typically, do first claim benefits, age 62. Retirees who claim benefits before the "full-benefits" age receive less than the full benefit. The reduction is 6.67 percent multiplied by how many years before the full-benefits age that people take their pensions. For that reason, raising the age at which unreduced benefits are paid by two years simply cuts benefits for all retirees by just over 13 percent.

The Social Security checks people receive have fallen still more for another reason. Before it mails checks, the Treasury Department subtracts each pensioner's premiums on Part B of Medicare. Medicare premiums have outpaced pension growth and are expected to continue to do so. Furthermore, taxation of Social Security benefits will also increase as called for by the 1983 amendments. For all these reasons, the ratio of Social Security net take-home pay to earnings has fallen. For a worker with average earnings claiming benefits at age 65, Social Security take-home pay, which was 39 percent of average lifetime earnings in 2002, will fall to 35 percent by 2015 and to 31 percent by 2030.

Not only are Social Security benefits growing more slowly than earnings, they are lower than benefits in most other developed countries. For average earners, U.S. old-age benefits in relation to earnings are 9 percent lower than those of Germany, 30 percent lower than those of France, and 51 percent lower than Denmark's. Compared to pensions in the 16 economically developed members of the Organization of Economic Cooperation and Development (OECD), U.S. benefits for average earners are 34 percent lower and rank fourteenth. If one takes account not only of the amount paid at a point in time, but also longevity, the ages at which pensions are first available, and other features of pensions systems, U.S. pensions are 40 percent below the OECD average in absolute value, despite our higher average incomes. This mocks allegations that U.S. benefits are lavish or unsustainable.

The structure of Social Security could be improved, however. Poverty among the elderly rises with age. People deplete their assets, pensions often decline or end when a spouse dies, and the capacity to work wanes. The age-related increase in poverty rates suggests that one currently popular idea for scaling back Social Security benefits -- the introduction of a technically more accurate cost-of-living adjustment -- is perverse. The consumer price index used to adjust Social Security benefits somewhat overstates inflation. In the name of accuracy, various commissions have endorsed replacing the current index with one that more accurately measures inflation. The effect of this index shift would compound from year to year, cutting benefits each year relative to those now paid by ever larger amounts the longer one receives benefits. Use of the alternative index would cut benefits for an 85-year-old retiree by about 7 percent. The impact on the long-term disabled could be even larger because they may receive benefits for even longer than do the elderly. This genuflection to statistical precision would reduce benefits most for the long-term disabled and the very old.

A better reform would be to boost Social Security benefits for those who have been on the rolls for a long time. For example, a benefit increase of $25 per month for each year on the rolls after age 75 would amount to a benefit increase of about 2 percent per year of the average retirement benefit now being awarded. By age 85, a 2 percent annual increase would boost benefits 20 percent. The increase would be largest for the very old and would be proportionately larger for those with low benefits and smaller for those with high benefits.

These data do not change the need eventually to close the gap between promised Social Security benefits and currently projected revenues. But they should influence how we do it. Since pension benefits are low compared to those of other countries and are falling relative to earnings, exclusive reliance on increased revenues is where discussions of how to close the projected long-term deficit should begin. Those revenues could come from increases in the current payroll-tax rate, from raising the cap on earnings subject to tax, or from extending the tax to currently exempt compensation. The payroll tax is now 4.2 percent for workers (though the rate is expected to go up to 6.2 percent in 2013 at this writing) and 6.2 percent for their employers. Earnings above $110,100 in 2012 are exempt, as is compensation channeled into the increasingly popular medical savings accounts, dependent care accounts, and transportation reimbursement plans. Gradually raising the fraction of earnings subject to tax from the current 84 percent of earnings to the historical target of 90 percent of earnings, boosting the payroll-tax rate from 6.2 to 7 percent, and taxing currently exempt cash compensation would fully close Social Security's projected long-term financing gap.

For progressives, restoring long-term funding balance with added revenues should be the starting point for negotiations. Realistically, however, some combination of net benefit reductions and revenue increases may be inevitable. It would be worth accepting modest and well-targeted net benefit cuts to end fears that Social Security is inadequately funded. The lessons of the 1983 amendments are instructive. The 1983 compromise undercut more radical proposals for restructuring the system. To be sure, a similar compromise today would not mollify those who are ideologically opposed to the very idea of Social Security. But it would end their capacity to mobilize a credible attack on the system because it would reassure the majority of voters who value Social Security but have come to fear for its financial viability. If projected trust-fund deficits persist, proposals to replace the system with private accounts or to otherwise change it fundamentally will continue to get a hearing. Only one such attack needs to succeed to undo the crown jewel of U.S. social insurance. Removing that chance is why progressives should be open to compromise.

But the acknowledgment that net benefit cuts may be inescapable does not settle whose benefits should be cut and how they should be cut, nor does it exclude targeted benefit increases. Increases in longevity have raised pension costs most on behalf of those with comparatively high education and earnings. Fairness dictates that if benefits are to be cut, the reductions should fall on benefits for those groups. The way to cut spending is not to cut annual benefits for everyone but to cut them selectively for early retirees with comparatively high earnings, thereby encouraging them to work longer and claim benefits later than in the past. A trend to later retirement is already under way. Labor-force participation among people over age 55 has been rising over the last couple of decades. Even so, 50 percent of new benefits awarded to retirees in 2011 were claimed at age 62, and 94 percent were awarded on or before the full-benefits age of 66.