Social Security’s Job

Source: The Urban Institute Ratio of Social Security benefits to Social Security taxes paid, by race or ethnicity and year.

Does Social Security need to be fixed?

As Democrats and Republicans grapple over how to reduce the government’s budget deficit in the face of rising costs for pensions and health care, whether Social Security should be touched remains one of the most controversial topics in American budgetary politics.

But something big is missing to the debate over the finances of what is still the largest component of the social safety net: an understanding of how well it does its job.

When you peek under the hood, it doesn’t always look so great. Indeed, this supposedly great redistributive program â€" which uses a broad tax on all workers to protect the elderly from poverty â€" exhibits some fairly stark regressive features.

One well-known regressive feature comes from the rule that benefits must be annuitized, paid out over time in monthly installments rather than as a lump sum. This means that richer people who tend to live longer will get a bigger benefit than poorer people with shorter life spans. Survivor benefits redistribute money from the singles â€" who don’t get the benefit â€" to the married, who do.

Eugene Steuerle, Karen Smith and Caleb Quakenbush of the Urban Institute in Washington just discovered another unsuspected regressive feature. Considering its transfers across the generations, Social Security redistributes money from minorities â€" blacks, Hispanics and others â€" to usually wealthier whites.

For every $100 paid into the system, white beneficiaries receive $113 in benefits, blacks receive $89 and Hispanics receive $58. This feature will become more pronounced over time. Over the next 10 years, whites will get $120 for each $100 they put in â€" on average. Blacks will get $91 and Hispanics $62.

The reason is straightforward: The black and Hispanic populations are generally younger. For every elderly pensioner drawing Social Security, there are more workers contributing payroll taxes.

Mr. Steuerle and his co-authors contend that this matters. “If one of Social Security’s goals is to provide greater relative protections to the most vulnerable, one must ask whether that was a desired or accidental outcome,” they write.

The program was indeed intended to be progressive â€" replacing a higher share of lifetime earnings for poorer workers. But as Mr. Steuerle notes, other features can annul this progressivity, and even turn Social Security into a regressive program.

“Social Security is intended to be redistributive,” he told me. “Very few people actually measure the extent to which it achieves this goal.”

This kind of finding may not justify any cuts to Social Security benefits. But it might suggest that other reforms â€" a bigger minimum benefit, for instance â€" might be in order.