Though I am sympathetic to advocates of using chained CPI to set cost-of-living adjustments to Social Security benefits, Josh Barro raises a deeper issue:
[I]f Social Security is intended to be a program for ensuring that seniors continue to share in national prosperity after they retire, why shouldn’t their benefits rise along with national income, and why should older seniors draw smaller benefits than younger ones?
Social Security reform proposals, including the Simpson-Bowles proposal, often contemplate benefit enhancements for the very elderly, in effect to offset the stagnation of benefits that comes from price indexing. An alternative would simply be to annually adjust benefits in line with wage growth, at least for lower-income seniors. [Emphasis added]
The obvious rejoinder is that Social Security ought to be a program for ensuring that seniors can maintain a decent standard of living, which is one reason why I favor Andrew Biggs’ approach to reforming Social Security:
Biggs’s proposal, an early version of which appeared in the AEI contribution to the Peterson Solutions initiative, creates a universal flat defined benefit (a predictable, stable minimum benefit that would eliminate poverty among the elderly, regardless of lifetime earnings); universal defined contribution retirement accounts (this would shift the system away from pay-as-you-go, and it would tend to strengthen work incentives), and the accounts would be automatically annuitized on retirement; COLAs for the minimum benefit would be set on the basis of chain-weighted CPI; the payroll tax would be eliminated at age 62 to encourage continued labor force participation; and the Retirement Earnings Test (RET) would be eliminated.
The Biggs proposal would greatly benefit low lifetime earners relative to the current system while requiring that middle-income and affluent retirees rely more heavily on private savings in the form of universal, dead-simple accounts that receive favorable tax treatment. It is true, however, that older seniors would see the relative value of their benefits erode, and this will become more of an issue as longevity continues to increase.
But it is reasonable to question whether chained CPI is the right approach given that it will have very different impacts on seniors across the lifetime earnings distribution. Dylan Matthews has written a helpful post outlining alternatives to chained CPI, including progressive price indexing.
I also recommend a RealClearMarkets op-ed by Aspen Gorry and Sita Nataraj Slavov that address a number of other issues, including the incentive Social Security’s current structure creates to choose shorter careers and to discourage work on the part of secondary earners:
The second problem with the Social Security benefit structure is that it discourages second earners from working and discriminates against two-income families. That’s because many second earners don’t anticipate higher Social Security benefits from working more; rather, they expect to collect spousal and survivor benefits regardless of whether they work. Also, because single-earner couples do not pay additional payroll tax in exchange for the non-working spouse’s benefit, they get a better deal from Social Security compared to two-earner couples.
One way to address this problem is to replace spousal and survivor benefits with earnings sharing, in which each spouse gets credit for half the couple’s total earnings. Alan Gustman of Dartmouth College and Thomas Steinmeier of Texas Tech University show that this reform can substantially increase the labor force participation of older wives. Melissa Favreault and Eugene Steuerle of the Urban Institute demonstrate that such a reform would help equalize the treatment of one-earner and two-earner couples.
One hopes that an ambitious Social Security reform will be part of the policy conversation in 2013.