Closing the Deficit: How Much Can Later Retirement Help?

4. Nudged, Pushed, or Mugged: Policies to Encourage Older Workers to Retire Later

72 4 1. Salhgren (2012), Dhaval, Rashad, and Spasojevic (2006), and Behncke (2009) reach similar results as do Coe and others (2010) and Rohwedder and Willis (2010). Other investigators have found that retirement is neutral or even beneficial to the retiree’s health. See Bound and Waidmann (2007), Kerwin Kofi (2002), and Coe and Lindeboom (2008). Nudged, Pushed, or Mugged: Policies to Encourage Older Workers to Retire Later henry j. aaron Policies to encourage people to defer retirement are increasingly attractive for several reasons. Life expectancy is rising. Projected increases in budget deficits are traceable largely to anticipated growth of spending on pension and health benefits for the elderly and disabled. Within the federal budget, social insurance trust funds have their own funding gaps. Increasing the labor supply of older people and those with impairments would ameliorate each of these challenges, provided that such increases can be achieved at reasonable cost. Delays in retirement may even be good for the health of those affected.1 The policy challenge is to identify services that will enable the elderly and the impaired to remain economically active and to design incentives for them to do so. The danger is that such policy suggestions, often advanced by incumbents of sedentary jobs, will inflict serious inconvenience or genuine hardship on those for whom continued work imposes serious burdens or is downright impossible. No social program precisely reaches all intended beneficiaries and no others. Typically, programs miss some putative targets and serve some unintended beneficiaries . Such incompleteness and inaccuracy arise both when new programs are introduced and when old programs are changed. However, inaccurate targeting is viewed differently for new and old programs. If a new program misses worthy ben- eficiaries and helps some who are not worthy, the inaccuracy is typically regarded as regrettable but inevitable. Conversely, withdrawing benefits from existing needy beneficiaries is often regarded as politically intolerable. This phenomenon has been observed in many contexts. Kahneman and Tversky built their theory of loss aversion around a related insight.2 Charles Schultze articulated a similar principle in his political version of the Hippocratic Oath—“do not be seen to do obvious harm.”3 Changes in programs that benefit the retired, impaired, or elderly are definitely subject to this problem. Yet because no set of characteristics neatly distinguishes various classes of beneficiaries—the disabled versus nondisabled, early retirees versus late retirees—even the most carefully designed benefit cuts will harm some needy beneficiaries, and even the most carefully designed benefit liberalization will help some unintended recipients. A series of tables showing the only slightly different characteristics of those who claim benefits at various ages appears in the appendix to this chapter.4 Two other factors further complicate the task of reducing the growth of social insurance spending. Life expectancy has increased more among some groups than among others and not at all for some. The well-educated and those with comparatively high earnings have enjoyed sizeable increases in life expectancy.5 Those with comparatively low education and earnings have experienced little or no increase in life expectancy. In designing incentives to encourage older workers to delay retirement, it is important to keep in mind that older Americans work more than do adults of the same age in most other developed countries. Movements in the U.S. labor supply have been the result of strongly offsetting trends among men and women (table 4-1). From 1950 through 1990 labor supply by men under age 65 fell slightly, but among men over age 65 it fell sharply. Overall, the proportion of men age 16 and over who were not working for pay or looking for work rose from 13.7 percent in 1950 to 23.6 percent in 1990. However , overall the labor supply increased as younger women flooded into paid work outside the home. Overall the share of Americans age 16 or older not in the labor force fell from 40.8 percent in 1950 to 33.5 percent in 1990. Starting around 1990, many of these trends shifted. Labor force participation by older men began to increase, even as the proportion of younger men in the labor force continued to fall. Decades of rapidly rising female labor force participation How to Encourage Older Workers to Retire Later 73 2. Kahneman and Tversky (1979). 3. Schultze (1977), p. 70. 4. Tables 4A-1 through 4A-7 are drawn from Aaron and Callan (2011) and document that the characteristics of those who retire from work at each...