Report Rebuffs Arguments for Privatizing Social Security

A recent working paper published by the National Bureau of Economic Research critically analyzes previous attempts to explain conversions of government pension provision systems, or social security systems, to personal accounts.

Common rationales include: higher returns; improvements in domestic financial institutional development; improving labor supply and retirement incentives; and hedging demographic changes. In each case, the report authors contend, a politically-stable and transparent government could have achieved similar results within the traditional system.  

In A Matter of Trust: Understanding Worldwide Public Pension Conversions, Kent Smetters and Walter E. Theseira point out that the traditional system also tends to have lower transaction costs, less adverse selection, and other costs. 

Looking at the United States Social Security system by generation over time, the report noted that effective annualized rates of return have declined dramatically. The composite average U.S. worker who began to collect a benefit at age 65 in 1941 received a 36.5% effective annual rate of return on his pension contributions. In sharp contrast, a person born today into the mature U.S. Social Security system is projected to receive less than a 2% effective rate of return. 

But, the report authors argue, the declining rate of return stems directly from pay-as-you-go financing itself: money flowing into the pension system gets distributed immediately as benefit payments to retirees. Early retirees receive a windfall that is matched in present value by payments made by future generations. Since these windfalls have already been consumed, there is no potential for recovering them. Increased returns only come from current generations sacrificing some of their consumption today, which they can do without privatization. 

Looking at those countries that have converted their systems, the study found private credit indeed expanded in each country following reform -- except for Mexico and Argentina, both of which suffered currency and other financial crises. This evidence would seem to buttress the case for the hypothesis that domestic capital market development was a major influence in reform. Many reforming Latin American and Eastern European countries also instituted restrictions that prevented workers from investing their money abroad, also seemingly consistent with this hypothesis. 

But, the report authors point to the fact that the post-reform rise in private credit tended to be sharper in Western European countries despite the smaller size of their pension reforms relative to the Latin American and Eastern European nations. Since most Western European countries already had sophisticated financial markets prior to reform, much of their increase presumably stemmed purely from increases in the amount of capital, since pension reforms likely had little impact on the relative reliance on equity financing. This result suggests that the increase in private credit in Latin American and Eastern European nations might also have been driven by the saving effect rather than by direct financial institution development.
In addition, except Mexico and Argentina, the availability of private credit began to increase in most Latin American and Eastern European countries before the introduction of personal accounts. (Since the personal accounts are unlikely to have a large effect for several years, the effective pre-reform trend is even longer.) The introduction of personal accounts does not seem to increase the pre-reform trend in most countries, suggesting that other factors might also be playing an important role in the expansion of private credit, the authors concluded.

The NBER working paper notes that payroll taxes in a traditional social security system are distorting for several reasons:

  • workers are forced to invest into a pay-as-you-go public pension “asset” that pays a rate of return below what they could have earned in the private sector; 
  • redistribution produces larger rates of returns for lower-income workers along with smaller returns for higher-income workers; 
  • borrowing constrained workers might not wish to make any contributions at younger ages, even at a market rate of return; 
  • particular features of the program, such as a spousal benefit, cause additional distortions; and
  • specific rules encourage early retirement.

Private accounts are argued to ameliorate these tax-like disincentives to labor supply. However, the report authors contend each of these distortions can be just as easily addressed with the traditional system through reform. Inferior rates of return are a natural outcome of pay-as-you-go financing or redistribution; similarly, tax-induced distortions caused by redistribution are simply the cost of redistribution, which would also exist in a private pension system that maintained the same level of redistribution. Borrowing constraints could be lessened by starting payroll tax rates at a low level early in life and increasing them with age in order to collect the same present value of income from each worker on average. The other distortions are consequences of the rules specific to different public pension systems and can also be addressed within traditional pension systems. 

Finally the report authors contend that demographic concerns are unable to explain conversions, at least outside of considerations tied to political trust. They found that higher-income countries tend to face the largest predicted declines in worker-retiree ratios; but, higher income countries are also those with the smallest system reforms. Among actual converting countries, there appears to be an inverse correlation between demographic problems and the size of actual reform. The direct empirical evidence, therefore, does not seem that supportive at a high level, the report said.  

In addition, the authors claim that in theory, a traditional system could effectively deal with changes in fertility rates by accumulating large reserves when the worker-retiree ratio is large and then spending down the reserves as the ratio decreases. (Since 1983, the United States has been building up its Social Security “trust fund” in an attempt to buffer future costs associated with the retirement of baby boomers.)  

If these reserves were truly saved, then taxes would remain roughly flat throughout the entire period, as collections would initially exceed costs and then fall below. A personal account system would do no better in hedging demographic risk.
The report authors said one central theme emerges from the study: the public pension conversions reflect a fundamental mistrust in the ability of the government to provide secure retirement resources.