What is the market failure?

The standard mantra of the government economist, invoked whenever she confronts a complicated public policy proposal, is “what is the market failure?” This follows from a celebrated result of economic theory: a free, competitive and perfectly functioning market, left to its own devices, will generate a socially efficient outcome. This theorem is central to how economists evaluate social policy, not because economists oppose government intervention, but because it provides a clear criterion for evaluating when and how such interventions should occur. It distresses me, therefore, that I have failed to hear this phrase invoked by any of those in Washington currently engaged in the on-going debate over Social Security reform.

From an economic perspective, there are three vital roles for government: to help free markets to operate; to intervene when those markets operate improperly or fail to exist, and to address issues of fairness, since there is no guarantee that free markets will generate outcomes that we, as a collective society, find desirable. Social security was certainly designed to make our society more palatable, by reducing poverty among the elderly, but also to address an important market failure: Social Security provides insurance against living too long. What is sold in private markets as life insurance is, in fact, death insurance since it pays off to your survivors should you die unexpectedly soon. What happens, however, if you live longer than you expected?

While private markets could in theory provide insurance to mimic that of Social Security – selling annuities to the young that provide an indefinite stream benefits on retirement – it is not altogether clear that they can do so as efficiently or comprehensively as the current system does. The complaint of those on the right arguing for privatization is that Social Security offers a rate of return that is, on average, lower than that which one could get by investing privately in financial assets. There are several reasons for this, including the fact that we are continuing to pay off the "free ride" that the first generation of elderly (the "greatest generation") received. Regardless, however, the argument that Social Security offers a "poor rate of return" is very much akin to saying that purchasing car insurance offers, on average, a very poor rate of return. Since the goal of private firms, unlike that of the government, is self-enrichment, it is hard to believe that financial markets would offer a more generous equivalent benefit, particularly if they were unable to exploit the same cost advantages from economies of scale.

The “privatizers” do have a point, however, even if they seem to have difficulty identifying it themselves in their talking points. Since Social Security was enacted in 1935, life expectancies have increased from 62 to 77, while the age at which retirees have been able to draw benefits has until recently been fixed at 65. As a result, the benefits the government is committed to paying have increased dramatically, with the rise in the average number of years in which those benefits are received.

The increase in benefits has necessitated several increases in the payroll tax, which is now close to 13% of earned income. (Nominally, employers pay half of this tax, but most economists agree the incidence of the full tax is born almost entirely by workers through lower wages). Lower after-tax earnings and higher guaranteed retirement benefits may contribute to the low U.S. saving rate. In effect, Social Security has slowly grown into more than just an anti-poverty insurance scheme, becoming increasingly a state pension scheme.

With this in mind, the dimensions of the debate are in fact quite clear: Democrats are arguing that, since Social Security was intended as an insurance system, a pay-as-you go system is preferable to self-insuring. Privatizing would defeat the entire purpose of Social Security. Republicans counter that, as a (de facto) pension system, not privatizing makes no sense since the “risk-adjusted rate of return” on retirement assets should be greater if those assets include private equity and corporate debt in addition to federal government bonds.

So which side do I think is correct? Well, I was quoted in a newspaper as opposing privatization because I was “to the right of George Bush.” While I strongly support facilitating access among the poor to tax-deferred saving vehicles, I do not favor the “big government” mandated saving program Bush proposes. Rather, I would maintain the original insurance (or “safety-net”) principle of Social Security, and close the funding gap by raising the retirement age (what I perceive as the source of the funding gap) rather than through higher payroll taxes or scaling back growth in benefit formulas. I oppose higher payroll tax rates because they are regressive, and I believe labor is already too highly taxed relative to other income sources (e.g. inherited wealth). Raising the cap on the taxable income above $90,000 will provide too little revenue. And slowing benefit growth through price- rather than wage-indexing will simply lead to the relative impoverishment of seniors in the years to come. (A thought experiment: imagine if current benefit payments were based on a 1935 standard of living….)

There are always, of course, trade-offs involved with any policy choice. Raising the age at which benefits are paid is likely to place the greatest cost on the most vulnerable. To see this, note that while 65 is often referred to as “the retirement age,” actual retirement ages differ substantially among the population. The fact that some individuals may choose to (or be forced to) retire earlier than others is an important consideration for assessing both the equity and efficiency of the system. Any reform that raises the age at which benefits are paid is likely to penalize the most vulnerable – those with the shortest life expectancy, least accumulated wealth, and/or most physically challenging occupations. My proposal would therefore include wage and occupational indexing for the retirement age. There may be a justification for gender and race indexing as well, although I like to hope that any such current disparities in life expectancy will have been eliminated by 2040….

It should be noted that although I use the “market failure” criterion to justify the current insurance-based pay-as-you go system over the proposed private account pension system, a creative economist could make arguments to favor private accounts from an efficiency standpoint. Here are some questions you might ask:

1) Does Social Security distort the incentive of households to save, because the complicated benefit formula creates a disconnect in individuals' minds between the OASDI tax payments they make and the future benefits they receive? If so, is there a way of addressing that problem without fundamentally changing the system?

2) The historical popularity of many Saving &Loan’s “Christmas accounts” (a non-liquid, zero rate of return asset) is often cited as evidence that people highly value commitment devices that keep them from spending their money. In other words, a role may exist for government in forcing people to save. Of course, this argument would require the Bush Administration make a 180-degree reversal on their standard party line that “people can spend their money better than the government” and embrace the sort of “big government paternalism” they often accuse Democrats of espousing.