TCS Daily

What's Wrong with Social Security?

By Arnold Kling - April 28, 2004 12:00 AM

"Occasionally, I run into people who believe that no one in his right mind would design a retirement system like the one we have. Some of the details do seem far from satisfactory to me. However, looking at the big picture, this structure makes sense."

-- Peter Diamond, "Social Security," American Economic Association Presidential Address

Peter Diamond, this year's President of the American Economic Association, is a leading expert on the economics of Social Security. His address, published in the American Economic Review (March 2004) but not available on line, defends the basic features of the current system.

Diamond's prestige and authority certainly dwarf mine. I have no quarrel with his description of how Social Security works and of the key analytical issues involved. Nonetheless, I intend to dispute his conclusion that our Social Security system is basically sensible.

Paternalism Revisited

In What's Wrong with Paternalism? I wrote that the case against paternalism has three layers. Diamond glides right past the pure libertarian layer, which argues for giving people freedom of choice on principle. Instead, he says, "To my mind, the heart of the context for thinking about Social Security is that it substitutes for poor decision making and for missing insurance opportunities...Mandating savings makes sense if you think that many workers would not provide themselves a reasonable replacement rate."

To a libertarian, just because you believe that people save too little for their own good does not mean that "mandating savings makes sense." Even those of us who are far from pure in our libertarianism ought to at least pause and reflect about the issue of a mandatory program. Could we instead have a voluntary program, particularly now that workers have had a chance to observe the benefits of Social Security for their own parents? Could we encourage savings through incentives rather than by force? Could we just lead the horse to water, but if he chooses not to drink, leave him alone?

The second layer of the case against paternalism is utilitarian, meaning the loss of well-being that results from interfering with the market. In the case of Social Security, the tax and benefit structure lowers output by distorting decisions that individuals otherwise would make about labor, leisure, saving, and consumption.

Several years ago, economist Martin Feldstein estimated that these distortions cost the economy 5 percent of GDP per year. This sort of analysis is sensitive to modeling assumptions, and many economists would argue that the 5 percent figure is too high. However, Diamond does not even discuss the issue. He simply focuses on whether a typical individual ought to be happy or unhappy with the structure of the program, ignoring the aggregate loss of output. Since the utilitarian issues are the heart of Diamond's (or anyone's) economic analysis, I will focus more on his arguments below.

The final layer of my case was the Public Choice argument, which is that even if a philosopher-economist-king could know the optimal policy, we are governed by flawed human beings. Ironically, Diamond uses this anti-paternalistic argument to defend keeping Social Security's basic structure intact. He writes, "I think that much of the apparent appeal of radical reform lies in the implausible implicit assumption that such a reform will pass into legislation untouched by political hands...A major reason for my concern about radical reform is the potential for ill-advised design, driven by political ideology rather than a realistic assessment of likely outcomes."

Diamond has made a valid point. I sometimes think that the cliche "If it ain't broke, don't fix it..." could be amended to add, "...And even if it is broke, don't underestimate the ability of Congress to break it even worse."

Cutting Diamond

It is time to cut into the core of Diamond's argument in favor of Social Security. He makes a utilitarian case for Social Security's three salient features:

  • Social Security is mandatory, rather than voluntary.
  • Social Security recipients are forced to accept an annuity, without the option of taking their retirement benefits as a lump sum.
  • Current recipients' benefits come from current workers' tax payments, rather than from the recipients' own previous tax payments.

Diamond believes that in the absence of Social Security, workers would fail to save enough for retirement. He says that many have inadequate savings even with Social Security, therefore clearly they would be in trouble without it. He writes, "Excessive reliance on Social Security [is] the best evidence for evaluating whether workers make adequate preparation for retirement."

Diamond gives the following as a basis for his claim that the elderly rely excessively on Social Security.

"One-third of the elderly received at least 90 percent of their income from Social Security in 2001 [even though the] average new award for a retired worker in 2002 was just over $900 per month...a worker retiring in 2002 at age 62 (the modal retirement age)...received a benefit of roughly one-third of (wage-indexed) lifetime average would not want to retire on one-third to one-half of what you had earned on average in your lifetime."

Concerning Diamond's point that Social Security constitutes a large portion of the income of many elderly, it is worth noting that the assets owned by the elderly may not produce much in terms of reported income. A $250,000 house that is appreciating at an annual rate of 5 percent would be counted as not producing any income, even though the appreciation is worth $12,500 per year. (We also should count the rental value of the home--the value of housing services--as income.)

A $300,000 stock portfolio might produce reported dividend income of less than $3000 per year. Most of the income from stocks comes in the form of capital gains, which might not be reported in the study that Diamond was using. If capital gains and losses were included in the income measure on which he relies, then they reduced income in the year 2001 that he cites, because it was a down year for the market. It seems safe to say that one way or the other the estimate of income from stock market portfolios in 2001 was much lower than the average stock market income over the past thirty years.

The supposed inadequacy of savings is contra-indicated by Diamond's incidental mention that more people retire at age 62 than at any other age. If you had reached age 62 without adequate savings to live on, why would you retire? Your Social Security benefits are much lower at that age than would be the case if you worked another few years.

Another contra-indication is the dismal reception in the marketplace for reverse annuity mortgages. For years, mortgage lenders have been drooling over the prospect of making mortgage loans to low-income elderly. The small number of takers suggests that the elderly are not as desperate for cash as the lenders were hoping.

I should make it clear that I believe that the phenomenon of elderly people who have not provisioned themselves adequately for retirement is real. But the question is how widespread is the problem -- are we talking about 5 percent of the population or 50 percent? Diamond failed to provide convincing evidence that the larger number is a better estimate than the smaller one.

Annuities vs. Lump Sums

A subtle but critical element of Diamond's case against drastic Social Security reform is his argument that it "annuitizes" the wealth of the elderly. He observes that in the absence of Social Security, a much smaller portion of wealth would be annuitized.

What does it mean to annuitize? If you have $300,000 in assets at age 70 and you do not know how long you have to live, then you do not know how much you can safely spend each year. If you over-estimate your lifespan, you will spend too little. If you under-estimate your lifespan, you will spend too much. Spend too little and you leave a large estate when you could have lived more comfortably. Spend too much and you leave a smaller estate than you had intended, or you may even wind up destitute in your later years.

The solution to this problem of not knowing how long you have to live is to purchase an annuity. A life insurance company will take your $300,000 and convert it to an annual payment, based on its actuarial tables. If you live longer than expected, the insurance company eats the loss. Otherwise, the insurance company makes a profit.

Because Social Security payments are made as an annuity, the government acts like a gigantic life insurance company. However, your choice to purchase an annuity is involuntary -- unlike some private pension plans, you do not have the option of taking all of your Social Security benefits as a lump sum. With a private retirement account, such as an IRA, the retiree chooses how money is "distributed" (taken out) each year. With Social Security, you have no such choice.

Why is it so important to force people to take annuities? Diamond again makes a paternalistic argument. "Despite the advantages of annuities, we see only a small fraction of people doing voluntary annuitization...[What we observe is] hard to reconcile with sensible decisions by households and the equilibrium industry response we would expect in the presence of sensible demands."

I doubt that elderly people who maintain their assets in lump sum format rather than as annuities are as irrational as Diamond and others suggest. In the real world, the elderly face many more sources of uncertainty than just their date of death. For most other sources of uncertainty, a lump sum provides better protection than an annuity.

For example, take the risk of a sudden jump in expenses, due to a medical problem or other crisis. If you have a lump sum, you can handle the spike in cash needs without having to take out a loan.

Another example is the risk that your children will need financial assistance. That risk probably declines as the years go by. When you first retire, the risk that your children will need help may be high, and having a lump sum to give to them (or bequeath to them in case you die early) is valuable. If instead of keeping a lump sum you have bought an annuity, then you are not in a good position to help your children when the risk that they will need assistance is the highest.

By no means should we dismiss the possibility that many people make insufficient use of annuities. Calculations of probability and assessments of risk are very difficult to make, so it is likely that many people get them wrong. However, as someone who tries to make such decisions using all of my knowledge of economics and statistics, I find myself unpersuaded to annuitize my wealth or to recommend annuitization to friends and relatives.

The annuity issue matters because it affects the balance between government and private provision of retirement income. If you do not believe that forced annuities are a major benefit, then the private sector can take care of much more of the retirement saving process. Even if you believed that people should be forced to save more, you could allow them to use private vehicles, along the lines of IRA's and 401(K) plans.

With mandatory savings going into personal retirement accounts instead of Social Security, people would have the freedom to decide the rate at which to draw down their savings, rather than being forced to take an annuity. Diamond sees such freedom as a bad thing, and such an important bad thing that it justifies replacing the private sector with the government.

Burdening the Future

Social Security recipients are not recouping their own tax dollars. Instead, their tax dollars went to pay for previous recipients. As Diamond puts it, "Everyone is aware of the decision to pay earlier cohorts of retirees benefits far larger than could have been financed by the taxes they paid and the interest that could have been earned on them...the legacy of the early generosity of Social Security shows up in assets that are not there."

As I pointed out in The Ultimate Lockbox, this underfunding (Diamond's term) opens up Social Security for all sorts of Congressional mischief. Unlike a fully-funded system, Congress can increase promised benefits without raising taxes, passing the burden on to the future.

Diamond's argument for keeping the system underfunded is that it's too late to change now. He views the underfunding as a one-time sunk cost, and "spreading the implied cost over the indefinite future (not fully funding Social Security) makes sense and incomplete funding contributes to risk sharing across cohorts."

A point that I made in a The Ultimate Lockbox is that you can still spread the cost of the original underfunding into the indefinite future, by using debt to fund the transition to full funding. The problem with underfunding as I see it is that it is not just a one-time sunk cost. On the contrary, it is an ongoing problem that gets exacerbated every time Congress is tempted to provide additional benefits.

Professor Diamond believes that Social Security is approximately well-structured, and that any major reform risks making it worse. I respect his judgment, and I have tried to understand and appreciate his arguments. But after thinking about his address, I come away still believing that the current system is unnecessarily intrusive and costly. I think that those who are arguing for radical reform have a very good economic and philosophical case.

Arnold Kling recently wrote for TCS about What's Wrong with Paternalism?


TCS Daily Archives