Why Liberals Should Back Social Security Reform

Liberals should not be fighting Social Security reform — they ought to be leading the charge for change, for a simple reason: The program is no longer progressive. Contrary to popular opinion, the structure of federal retirement programs today favors middle and high earners over less well-off retirees.

The notion of Social Security as progressive is based on the criterion of annual benefits received per dollar of contributions. By that measure, the program offers a more generous return to those who earn less and therefore contribute less.

But a more accurate measure of the program’s distributional impact is total lifetime benefits received by workers at different wage levels. By this standard, Social Security is not progressive because low-wage workers live on average a few years less than wealthier workers.

This gap in life expectancy is likely to increase over the next few decades as the costs of breakthrough medical therapies rise. In addition, the gap will have a greater impact on lifetime benefits as the normal retirement age edges up to 67 by 2027. This is why any further increases in the normal retirement age should include special protections for low-wage workers, especially those engaged in physical labor.

Another indication of Social Security’s lack of progressivity is how much of the U.S. wage base is subject to Social Security taxes. Historically, these taxes have applied to about 90 percent of wages. During the past two decades, however, that level has fallen and is now close to 85 percent — meaning that better-paid workers are not paying as much in Social Security taxes as they have been called on to in the past. In concrete terms, Social Security taxes apply to the first $106,800 of earnings. But if the amount were increased to the historical benchmark, the first $170,000 of a worker’s earnings would be taxed.

As a practical political matter, it is doubtful that Congress would impose the full Social Security tax — 12.4 percent (half paid by the employer and half by the employee) — on all wages up to $170,000 annually. Nevertheless, to make the program more progressive, Congress might impose a surcharge — for example, 1 or 2 percent, on all wages over the current ceiling, up to $1 million per year.

Moreover, the progressivity of Social Security should be looked at in the broader context of federal support for retirement savings. The federal government confers more than $100 billion in tax benefits each year on participants in employer-sponsored retirement programs. These benefits flow almost entirely to middle and high earners, who generally participate in their employers’ retirement plans. By contrast, the employers of many low-paid workers do not offer retirement plans — and if they do, more low-wage workers do not have the financial leeway to make the necessary contributions.

Because of the differential impact of these tax benefits, the overall federal approach to retirement is regressive. For Social Security to become both solvent and progressive, Congress should maintain existing benefits for low earners while slowing the growth of these benefits for middle and high earners.

Under the current schedule, because the growth in the initial level of benefits is linked to wages rather than inflation, the purchasing power of Social Security benefits is slated to rise by almost 50 percent over the next 50 years. In other words, Social Security checks will increase not only in absolute dollars but will also buy much more in goods and services.

Congress could therefore gradually slow the growth of Social Security benefits for middle and high earners while still allowing these benefits to rise in terms of absolute dollars and purchasing power. Lower-wage earners would receive everything they are now promised.

These benefits reforms, if applied gradually only to workers 59 and younger, would reduce by roughly two-thirds the $4.8 trillion long-term deficit of Social Security. And, as liberals ought to like, these reforms would make the program more progressive.