14 August 2009

Harsanyi Out Of Touch On Social Security

David Harsanyi, a libertarian opinion writer at the Denver Post today bemoans the lack of a "private option" for Social Security, and complains about the bite of payroll taxes. Could there be a worse time to make this point?

Wrong On Rates Of Return

Harsanyi decries Social Security as "a failing program and pay into a private one that offers more than a 1 percent return (like a savings account or a tooth fairy, for instance)." Sort of like the stock market.

Returns for the ten-year period ending June 30, 2009 were negative for the S&P 500. Social Security was invented, of course, to serve as a counterbalance to the hardships suffered by retirees in the wake of the stock market crash of 1929, troubles that we experiencing again now.

Stock markets abroad show this two time U.S. experience is not unique. The U.K. stock market "returned just 1.05% a year between 1998 and 2008 in nominal terms, according to Barclays Capital. . . . the decade ending in 1974 saw a weaker return at 1.02% a year. Returns look even worse after inflation: the UK stock market is down 1.4% a year in real terms over the past decade." The Japanese stock market hasn't been a sweet deal either: "In October 2008 the Nikkei 225 stock index reached a 26-year low of 6994.90."

Harsanyi is also simply wrong in the numbers he cites.

He understates the rate of return that the typical person receives from Social Security, which varies considerably depending upon marital status and the number of income earners in a married couple. About 95% of Americans marry, most of them have two incomes, and that about half of those who do not marry are women. Yet, single men make a 1% rate of return from Social Security, in excess of inflation, and single women and married couples do considerably better.

The returns of Social Security are considerably more than the 1% nominal rate of return cited by Harasanyi. The average annualized rates of return on contributions even after inflation, for people retiring in 2030 are as follows:

Single male 1.00%
Single female 1.90%
One-earner couple 3.37%
Two-earner couple 2.29%

No decade since the U.S. permanently departed from the gold standard in 1971 has seen inflation at an annualized rate of less than 2.78%. Thus, a real rate of return of 1.00% earned by a single male translates to a nominal rate of return of at least 3.78%, and the nominal rate of return earned by the median person, who is in a two-earner couple, is more than 5%. (Admittedly, predicting future inflation rates is not an exact science. The margin between the yield on Treasury bonds with fixed nominal rates of return, and Treasury bonds with inflation adjusted rates of return, suggest that the market expects inflation rates about 1.25% and 2%. This would implies returns of 2.25% to 3.00% for single males and 3.54% to 4.29% for two earner couples who are the median beneficiaries.)

The best interest rate available from any bank in Colorado on certificates of deposit, which typically offer the best returns available on bank deposits and are equally risk free, today, is 3.10%.

Moreover, unlike returns from a 401(k), traditional IRA, private defined benefit pension plan, or non-tax preferenced private account, a large share of Social Security benefits are tax free when paid to the beneficiary.

Corporate Aaa rated bonds currently earn about 5.6% nominal returns, which is not significantly better after tax than the median rate of return on Social Security contributions. Thirty year municipal bonds, which are tax free and low risk (but not zero risk), pay a 4.85% nominal rate on 30 year bonds, and considerably less on shorter term bonds.

The return numbers for private accounts are also before the considerable fees and commissions that apply in any form of private account, but are absent from Social Security. The payroll tax generates more dollars of revenue relative to the administrative costs of the tax, than any other tax used by the federal government. The administration of Social Security benefits, likewise, is exceedingly efficient relative to the volume of funds provided and the number of beneficiaries involved. Many private investment companies refuse to work at all with accounts as small as that of many people starting to participate in the Social Security system.

For the average family, the return they receive on their Social Security contributions is not significantly worse than what they could receive on an investment with low risk in a private account invested prudently. Of course, it is also true that not everyone would invest their private accounts prudently, creating a burden on society should their investments be unwise.

Risk in Private Accounts

Social security isn't perfectly in balance. But, it isn't in an immediate crisis, either, despite the fact that its benefits are inflation proof. "Even without any changes, current benefits are expected to be fully payable on a timely basis until 2037."

This level of security is something that is increasingly hard to find in private accounts.

Most Americans don't have access to private defined benefit pension plans at all (the private equivalent of Social Security), which are the real failing program. In a 2004 speech to the U.S. Chamber of Commerce, the Executive Director of the Pension Benefit Guarantee Corporation, the government version of the FDIC for private defined benefit pension plans explained:

The number of private sector defined benefit plans grew through the 1960s and '70s before reaching a peak of 112,000 in the mid-1980s. At that time, some 40 percent of Americans workers were covered by defined benefit plans.

Since then, there has been steady erosion. Over the past two decades, the number of defined benefit plans has fallen by 75 percent to just over 31,000 plans today. Moreover, just 1 in 5 workers—20 percent of the workforce—now participates in a private sector defined benefit plan.

People who don't have private defined benefit pension plans invest their funds directly in the securities market, which, as noted above, has a record a plummeting at some point during the relevant time period of a person's working career, with catastrophic effects if you are unlucky enough to need the money at the bottom of one of these crashes.

But, even those employees who, in theory, have a promised pension benefit that their employer is required by law to pre-fund at actuarial sound levels can't count of getting their promised benefits when they retire. The downside risk of private pension plans betting their beneficiary's retirement on the stock market has been realized:

In the 12-month period ending October 9, 2008, equities held by private defined benefit plans lost almost a trillion dollars ($.9 trillion).

For funding purposes, the aggregate funded status of defined benefit plans has fallen from 100% at the end of 2007 to 75% . . .

More than 50% of private defined benefit plans are less than 80% funded.

The aggregate contribution that employers will be required to make to such plans for 2009 could almost triple, from just over $50 billion to almost $150 billion.

The underfunding crisis matters because private pension plans, unlikely Social Security, are not fully backed by the full faith and credit of the United States government:

If a company fails while its pension plan is underfunded, the PBGC is obliged to pick up the plan and pay retired workers. But it may not pay the full pensions that workers thought they were going to get; in some cases, such as the steel companies and airlines that filed for bankruptcy in the 1990s and early 2000s, many workers who had retired early with their full promised pension saw their monthly checks cut by more than 25%. A 2008 PBGC study of 125 plans terminated between 1990 and 2005 found 16% of the 525,000 participants suffered an average benefit reduction of 28%. Of the 70,000 retirees of Bethlehem Steel, whose plan the PBGC took over on Apr. 30, 2003, about 11,000 lost benefits, typically $500 of their $2,050 average monthly check.

The PBGC itself is underfunded by $33.5 billion.

The Non-Problem Of Social Security Payroll Taxation

Harsanyi also proclaims:

How about those payroll taxes most of us pay to fund Medicare? Isn't it time that Washington instituted an opt-out clause so that future generations are able to select private options if they wish?

Yet, Americans are less concerned about taxes being too high than they have been at any time since 1956 (when the top marginal tax rate on incomes over $400,000 was 91%; the top marginal tax rate on incomes is now 35% and applies at a little more than $311,000 of income).

The Social Security tax base, in particular, has grown smaller:

[W]hen the maximum limitation was enacted, it caused 92 percent of earnings to be subject to social security tax on the employee. Over time, that percentage fell as real wages increased, and Congress amended the law and provided for cost-of-living increases to the specified maximum dollar amount. Yet, presently only 80 percent of wages are subject to social security taxation on the employee.

Of course, in any situation where there is a positive rate of return on your tax dollar to you personally, it is hard to call that a tax in the ordinary sense at all. Mandatory contributions to private accounts would take money out of your paycheck just as reliably as Social Security does, but without any reliable promise that there would be a positive rate of return.

The Poverty Issue

The other problem with a "private option" for Social Security is that Social Security does more than simply facilitating savings for retirement. It is also the most effective anti-poverty program every invented, which eliminates all sorts of other social problems that take government money to address, which poverty creates.

While people should set aside enough savings for their retirement and secure adequate permanent disability insurance without being required by tax law to do so, the reality is that many do not (and this was true before Social Security was adopted, indeed this was the main reason that Social Security was adopted in the first place):

Social Security provides more than half of the total income for almost 60 percent of beneficiaries. For almost 30 percent, it provides more than 90 percent of income. . . . The poverty rate among the elderly in 2000 was approximately 10 percent, down from a rate of 35.2 percent in 1959. Without Social Security, the poverty rate among the elderly would be 48 percent.

Low income workers do receive higher rates of return on their Social Security contributions, but we all get the payback of that higher rate of return because we don't have to spend money on more expensive to administered means tested programs to provide elderly workers who can't support themselves with the minimum level of support they need to survive.

Footnote On the Public Option

Of course, Harsanyi didn't write his column because the nation is in the midst of seriously considering private accounts in the Social Security system. He did so as an argument against a "public option" in health care reform, although it is not obvious why a libertarian should argue against more choices for consumers.

In that connection, it is worth noting that the population that the American health care system has the best record of providing health care to shown, the elderly, as illustrated for example, in its good outcomes from cancer treatment by international standards, does not just have a public option. For senior citizens, the American health care system is a single payer health care program. This program is essential, together with Social Security, in keeping America's senior citizens out of poverty, it has far lower administrative costs than private health insurance plans do, it is very nearly universal, and it has managed to operate without forcing health care providers to be government run and without significantly discouraging medical treatments for elderly patients.

About 78% of men, and 86% of women die at age sixty-six or later. This means that their health care in their final year of life, which makes up a large percentage of all health care costs, takes place while their primary health care provider is Medicare. A disproportionate share of the rest have Medicare or Medicaid as their primary health care provider, because parts of these programs provide health care to those who are disabled (who have shorter life expectancies than the general population).

The moral issues involved in balancing end of life care decisions against cost in the United States already arise overwhelmingly in the context of people who are in a single payer government health care program.

The British National Health Care system, for all the criticism it has received from American conservatives (often based on inaccurate claims such as the claim that physicist Stephen Hawking, who receives his care from the system, would have died sooner if he was cared for by the British National Health Care system), yesterday received resounding testimonials of support not only from the ruling Labor Party prime minister, but from the leader of the ranking leader of the opposition Conservative Party (sometimes called the Tories).

There is also no reason to believe that a non-profit health care plan would be more stingy about providing health care to those at the end of life than existing for profit health insurance companies, whose tendency to try to deprive sick beneficiaries of care is well documented.


Michael Malak said...

If the goal is to keep people out of poverty the Social Security benefits should be means-tested.

There is no more Social Security Trust Fund. Social Security benefits will start drawing from the general tax base starting in 2017.

It's time to throw off the charade that Social Security is a retirement account and make it fully a welfare system.

Doing so would eliminate the regressive payroll tax.

As a states-rights advocate, I think it should also be administered by the states, that will likely enter into reciprocity agreements on their own. The point is that a libertarian state could opt out.

Another problem with Social Security is its forced secularization. Before the New Deal, charity was charitable and provided by religious organizations.

When it comes to healthcare, it should also be administered by the states if at all. It should also be for basic healthcare, and not a way to funnel money to Big Pharma over patented drugs. There are a lot of other issues with healthcare as well, but these are the top issues when constrained to looking at issues similar to Social Security.

Andrew Oh-Willeke said...

The lesson of Social Security is that categorical programs are much cheaper to administer, much less fraud prone, and much more popular than means-tested ones.

Also, is it really regressive if the poor get a better deal from the program taken as a whole than they do if it didn't exist at all?

And, "states rights" doesn't make much sense in this situation. States are not being prevented from doing anything and the administrative costs of a well functioning program are surely lower if it is federalized. This makes sense at a federal level because it keeps the program simple. If one were to pre-fund the Social Security Trust fund with private assets, decentralizing the responsibility to state treasurers might make sense, but the rest of the program works much better at the federal level alone. Programs like worker's compensation, unemployment insurance and Medicaid run at the state level with elaborate qualification rules are comparative failures.

Aid to the elderly has never been exclusively or even significantly religious; historically and even today in places with weak safety nets, the responsibility was primarily with extended family. The need for Social Security has more to do with shrinking family size and geographic mobility of people's descendants andthe rise of a cash economy than it does with secularization.

State administered of health care benefit programs (e.g. Medicaid) have a poor record of success compared to purely federal programs (e.g. Medicare). Some of this is the means test v. categorical issue, some is not. The Germans and Canadians have good experience with having states administer social welfare programs under a federal framework, the Americans and French do not, although even there programs differ. There isn't any profound reason that this is so; it is simply a question of history and bureaucratic evolution.

Notably, there are no libertarian states. While states differ from each other a great deal on social issues and on the dimension of generalized governmental competence, on issues of governmental administration philosophy and big v. small government, there are no real material differences between the states in any systemic manner.