Topic: Social Security

The Correct Perspective on Social Security Privatization

In today’s WSJ, William Shipman and Peter Ferrara have a column criticizing President Obama’s recent and vehement rejection of Social Security private accounts. I agree with Shipman and Ferrara — it’s rather shabby logic from a president of all Americans.

Shipman and Ferrara correctly note that Social Security privatization options provide participants with a choice — opt for private accounts or stay with the traditional system. In other words, people can choose their preferred risk set — political or market.  The lesson here is that there’s no avoiding risk.

Shipman and Ferrara suggest that all investments in private Social Security accounts do not have to be in stocks; people can choose bonds as well.  Better yet, they can hold the market basket of all stocks and bonds through low-cost index funds and hold some cash.  They can select the mix between these elements to optimize the risk-return trade-off given their abilities/preferences on the two. This investment strategy is transparent and easy to learn; it requires only a modicum of financial literacy.

However, I find their ”Joe the Plumber” example unpersuasive. Who cares if investing on the planet Mars yields 50 percent annual returns if we cannot do it unconditionally — that is, without incurring costs that would neutralize its higher-than-Social Security returns?  Those additional costs arise from having to borrow to pay existing Social Security beneficiaries their “promised” benefits, and from carrying market risks on personal account portfolios of Martian investments. 

Market risk represents a real cost, even if investments are for the long term.  The Shipman/Ferrara calculations take account of the recent financial crisis.  But they don’t take account of the potential for fat tails in the distribution of financial crises going forward.  The recent crisis could have been less severe.  But what if it had been more severe and had wiped out all savings for many more people?  Is there zero risk of such an outcome? A generalization on the basis of just one 40-year record of investment returns is inappropriate and insufficient for ruling out the importance of market risk.

In the authors’ defense, however, is the fact that the historical evidence of market returns is conditional on the existence of Social Security (and Medicare and the rest of the government’s panoply of welfare programs, regulations, etc.).  Without such broad and deep government interference in markets, the history of capital returns may have been different: returns may have been smaller (because the economy may have been better capitalized) but also more stable. And correlations between worker average wage growth and capital market returns may also have been smaller, yielding important diversification benefits from a privatized system of retirement saving.  But the bottom line is that we just don’t have adequate data of the correct type to make the “analytical” arguments that the authors attempt in their op-ed.  

Shipman and Ferrara (jointly and individually) have never explicated this latter argument clearly. They persist with their “higher-and-sexy-market-returns” argument in support of private Social Security accounts.  As such, I’m compelled to say that their argument continues to exhibit a real and serious deficiency.

On balance, however, when faced with two extremes — 1) political risk that the government will muck things up so badly that we and our children will suffer considerably reduced living standards, and 2) market risks that could devastate retirement savings because a recession/depression wipes out the value of lifetime savings — I would recommend an “interior solution” that straddles both worlds.  That is, continue a strictly limited government-run Social Security system and supplement it with a privatized element as many other countries have done, the UK and Australia being important examples. 

Some would say that we have such a system already, in the form of 401k, IRA and other tax-qualified saving plans. However, not all workers have access to 401k plans.  And the evidence is that despite those plans, national saving has declined considerably over the last three decades.  My analysis suggests that the reason for the decline in saving is the very existence of (supposedly) government-guaranteed Social Security (and Medicare) benefits that lull us into a false sense of security.  The key shortcoming is the lack of a system of universal Social Security personal accounts wherein a minimum amount of saving is mandatory (despite government mandates being bad in general). Such a system would provide a vehicle for the rich and the poor alike to partake of the wealth creation process that capital markets can and do provide. 

We’re not there today, and the correct direction from where we are is toward, not away, from Social Security personal accounts.

Topics:

Liberal Dogma on Social Security Redux

Liberal posturing on Social Security reform continues unabated – betraying nervousness that Obama’s Deficit Reduction Commission will recommend Social Security benefit cuts. 

Left-wing voices also continue to repeat the mantra that introducing private Social Security accounts would be a bad idea. Ronald Brownstein’s recent recent column in the National Journal is a case in point. However, Brownstein’s readers may come away thinking that he believes breaking promises is a good idea.

Brownstein concedes that “Social Security indeed faces a long-term imbalance between expected revenue and promised benefits.” I consider this to be progress — at least relative to the erstwhile “there’s nothing wrong and nothing to fix” mantra adopted by liberal adherents of the status quo on Social Security.

Notice Brownstein’s use of the term “promised benefits.”  A promise implies a commitment and obligation to make good on future benefit payments.  But the solution that Mr. Brownstein points to is as follows:

Instead [of private accounts], Obama argued, the two parties could emulate the Reagan model and arrive at a sensible solution… [T]he program’s long-term shortfall could be eliminated just by trimming benefits for the top half of earners [JG note: breaking the Social Security benefit promise here], linking the retirement age to lengthening life spans [JG note: breaking the promise here too], and imposing a partial payroll tax on earnings above $250,000 [JG note: that is, promise more benefits by expanding the definition of covered earnings and increasing payroll taxes on high earners].”

But all that the last element may achieve is to stave of the program’s insolvency for a few more years. 

My comment:  Please don’t drag Reagan into this “solution.”  The 1983 reforms were implemented under the gun, at a time when there was no way out of Social Security’s imminent revenue shortfall. If President Reagan had enjoyed the luxury of a couple more years to plan changes to Social Security, he would have adopted a different approach, and be much better off today. According to broad market indexes such as the S&P 500, total returns averaged well above 10 percent per year during the 1980s and 90s – so, well above inflation. (The first decade of the 2000s yielded a negative 1 percent return.)

Finally, Brownstein writes:

[T]he gap between the system’s revenues and obligations, relatively speaking, isn’t that daunting–less than 1 percent of the economy’s expected output over the next 75 years. 

Does Mr. Brownstein really appreciate how large that is? In present value terms, the Social Security actuaries report that the present value of Social Security’s shortfall over the next 75-years equals $5.4 trillion. That’s one-third of current annual GDP. In other words we have to devote that sum to earning interest each year for 75 years to cover Social Security’s financial gap.

Alternatively, since payrolls equal only one-half of national output, it means that payroll taxes would have to increase by an average of about 2.0 percent per year if they are levied over all wage earners.  However, the tax increase is to be levied only on those earning $250,000 or more.  There are about 3 million U.S. taxpayers with incomes above $250,000, with average income of about $500,000. (I’m rounding up based on information for 2006 available here.)  That makes a tax base of $1.5 trillion. (Actually, this is likely to be too large because I’m counting total income, not taxable income, which would be much smaller.) Raising the equivalent amount of revenues from these high earners (who face the highest marginal income tax rates already and are likely to alter their work effort in response to still higher taxes) would imply increasing their average tax rates by almost 11 percentage points. Of course, because some of the adjustment will be through benefit cuts and indexing the retirement age to increasing longevity, the tax increases that must be levied on high earners would be smaller. 

But are those benefit cuts politically realistic? Americans already face a normal retirement age of 66, and it is scheduled to increase to 67 in little more than a decade. Extrapolating from the French response to increasing their pensionable age from just 60 to 62, Americans’ would probably end up opening a third war front to resist further increases in Social Security’s retirement ages – a “generational war” here at home.

So where do we go from here?  One answer may be to first introduce “add-on” personal accounts using the 2.0 percentage points of payrolls – the amount required to plug Social Security’s current shortfall. This would not be a “tax” as the funds would be invested in personal accounts – and it would enable low earners an opportunity to partake in the long-term wealth creation mechanism that they have heretofore been unable to exploit.  As I have argued here, if this amount is effectively saved and invested – by insuring that the government does not borrow and spend those savings – it would create space for a “carve-out” addition to the “add on” personal accounts, increasing retirement wealth even more.  Finally, with the stock markets relatively stable and current P/E ratios of broad market indexes close to historical averages, now would be the right time to begin such a reform program for Social Security. 

Would liberal policymakers and analysts take on this approach?  No prizes for guessing the answer.

Personal Accounts for Social Security an Election Killer — Not Quite

You can tell its election season because Democrats are once again attacking Republican’s for daring to propose reforms to Social Security.  These attacks come despite the fact that Social Security is already running a temporary deficit, and that deficit will turn permanent in just five years.  Overall, the amount the system has promised beyond what it can actually pay now totals $18.7 trillion.

But the latest Pew Poll suggests that attacking Republicans for wanting to “privatize” Social Security might not be such an effective tactic after all.  According to the poll, Americans support proposals to “allow workers younger than age 55 to invest a portion of their Social Security taxes in personal retirement accounts that would rise and fall with the markets” by 58 – 28 percent.   Younger voters supported personal accounts my an astounding 70-14 percent margin, but every age group except seniors was supportive.  Seniors split evenly.   Independents, widely believed to be the key to the upcoming election, supported personal accounts by 61-27, and even Democrats favored the idea by 50-36.

Maybe this will finally give the Republicans some courage on the issue.

A Birthday Gift from Paul Krugman

I turn 41 this summer (thank you for the condolences). Along with the well wishes of family and friends, I received an unexpected gift from NY Times writer Paul Krugman: this column in which he bashes people who are critical of Social Security in its current form or who worry about its ability to deliver expected benefits.

At first glance, the column hardly seems like a gift: it’s long on pointless insults, short on thoughtful discussion, and misleading. But it offers such a poor defense of the Social Security status quo that I suspect readers will be more skeptical of the program after seeing the column, not less. Hence, Krugman’s gift.

He writes:

Social Security has been running surpluses for the last quarter-century, banking those surpluses in a special account, the so-called trust fund. The program won’t have to turn to Congress for help or cut benefits until or unless the trust fund is exhausted, which the program’s actuaries don’t expect to happen until 2037 — and there’s a significant chance, according to their estimates, that that day will never come.

OK, 2037 — no worries. Except that, as I said, I turn 41 this summer, which means I’ll turn 67 and qualify for full Social Security benefits in mid-2036. The very next year, the Social Security trust fund will be exhausted, according to the “intermediate” scenario contained in the most recent Social Security Trustees Report, available here (see Section IV-B and Appendix E). The program will still pay out some benefits — but less than 3/4s of what it now promises. So what happens then? That’s not a good question if you’re my age or younger.

But suppose you’re not my age or younger. Suppose you’re 10 years older than me, and will have collected 10 years of benefits by 2037. Don’t feel smug — you’ll be asking “So what happens next?” when you’re 77. That’s not a good question at your age, either. 

In fairness to Krugman, the Trustees Report considers different Social Security cost scenarios, the most optimistic of which projects that the trust fund will not be fully exhausted over the 75-year period the report considers. Krugman says there’s “a significant chance” this will be the case, but my (admittedly quick) skim of the report suggests it’s more just “a chance.”

One quick aside about the 2037 exhaustion date: when Krugman wrote this column in 2005, the Trustees’ intermediate scenario projected that the trust fund would last until 2042. In five years’ time, that date has grown 10 years closer. Not good.

Krugman writes that, if the trust fund does run out, Social Security can maintain its benefits using money transferred into the program by Congress from elsewhere in the federal budget. In fact, Congress will have to direct money from elsewhere to Social Security much earlier than 2037. Under the Trustees’ intermediate scenario, beginning in 2018 the amount of money Social Security pays out in old-age benefits each year will be greater than what the public pension program takes in in payroll taxes. (The Disability Insurance component of Social Security is in even worse shape, according to the Trustees, such that the program overall will go in the red in 2015.) To cover the difference, Congress will have to begin paying off the treasury bonds that currently comprise Social Security’s trust fund in order to provide the promised benefits. (In fact, Congress will have to do that this year because, as a product of the recession, Social Security obligations are greater than revenues. Hopefully, the economy will rebound and give the program a few years’ respite before transfers become an annual necessity, though the pessimistic scenario predicts no such respite.)

Krugman is unconcerned by these transfers, dismissing those who worry about them as engaging in “three-card monte.” His column doesn’t acknowledge that these transfers would need to occur at a time when Congress will be scrambling to cover other growing costs: similar deficits in Medicare, obligations to the ever-growing federal debt, and Medicaid’s increasing burden on federal and state governments. I worry that future taxpayers will not be amenable to having so much of their tax money directed to retirees (who refused to reform Social Security when they could have done so at relatively lower cost) rather than to government services for current taxpayers.

Krugman ends the column criticizing the proposal to reduce Social Security’s cost by raising the age at which retirees become eligible for full benefits. As part of an adjustment that began in 2002, retirees must now wait until age 66 to receive full benefits; beginning in 2021, the age requirement will slowly be raised until it reaches 67 in 2027. (Retirees will still be able to take reduced benefits at 62.) Some have suggested raising the full-benefit age to 70. Krugman says that would be unfair:

America is becoming an increasingly unequal society — and the growing disparities extend to matters of life and death. Life expectancy at age 65 has risen a lot at the top of the income distribution, but much less for lower-income workers. And remember, the retirement age is already scheduled to rise under current law. So let’s beat back this unnecessary, unfair and — let’s not mince words — cruel attack on working Americans.

There is something to what Krugman says. From 1980 to 2000, life expectancy at birth for the poorest decile (i.e., 10%) of the U.S. population increased from 73.0 years to 74.7 years, while life expectancy for the wealthiest decile increased from 75.8 years to 79.2 years. (The disparity in life expectancy between the top and bottom decile groups does decline to 1.6 years at age 65, which is up from 0.3 years in 1980. H/T to Dr. Daniel Coyne at the Washington University in St. Louis School of Medicine.)

But inequality in Social Security benefits would exist whether the eligibility age is 65, 66, or 70. Because Americans are required to participate in Social Security, and because all Americans become eligible for full retirement benefits for the rest of their lives at a single threshold age, then the longer-lived wealthy will receive more in benefits than the shorter-lived poor no matter what that threshold is. This is the product of having a one-size-for-all public pension plan (with lousy benefits). The way to address this inequality problem is through Social Security choice.

As I wrote at the beginning, Krugman’s column should leave thoughtful and informed readers more concerned about Social Security, not less. He couldn’t have given me a better present.

The Washington Post Misleads Readers about Medicare & Social Security Funding

Here’s a poor, unsuccessful letter I submitted to the editor of The Washington Post:

The Post’s economic reporters need to convey to readers that the Medicare and Social Security “trust funds” contain zero funds [“Medicare Funds to Last 12 Years Longer than Earlier Forecast, Report Says,” August 6].

This is not up for dispute.  When those programs’ revenues exceed outlays, Congress puts the excess in general revenues and spends it.  Congress marks the event by leaving an IOU to itself in these “trust funds.”  Those IOUs are not “funds,” any more than an IOU that you write to yourself is money.  These so-called “trust funds” therefore have no bearing on the (in)solvency of Medicare and Social Security.

Yet every year, the trustees for these programs claim that they do, making the Medicare and Social Security trustees reports an annual, ritualized lie that the U.S. government broadcasts to the American people.

Properly educating reporters, editors, and politicians about the Medicare and Social Security “trust funds” is a decades-long project.

The Social Security and Medicare ‘Trust Funds’ Are a … What’s the Word?

Yesterday’s New York Times editorialized:

It’s the time of year when the trustees of Medicare and Social Security release their annual reports on the programs’ financial health. And that means Americans are likely to be bathed in a fog of political rhetoric that makes it hard to sort out fact from fiction.

Here’s the bottom line…

The Times then proceeded to bathe its readers in fog:

According to the reports, the date of insolvency for Medicare’s hospital fund was pushed back, from 2017 to 2029, because of cost-saving measures in health reform. As for Social Security, without any changes, it will be able to pay full benefits until 2037 and partial benefits after that, the same estimate as in last year’s report, despite temporary setbacks from the recession…

A lot of attention will be paid to the finding in the Social Security report that payouts will exceed revenues in 2010 and 2011…That doesn’t endanger benefits, because any shortfall can be covered by the trust fund.

No.  It.  Can’t.  Because there are no funds in the Social Security “trust fund.”  There are no funds in the Medicare “trust fund.”  As Fortune magazine’s senior editor-at-large Allan Sloan explains in today’s Washington Post, those “trust funds” contain nothing but “funny money.”

In a 2006 blog post titled, “Sometimes, Governments Lie,” I offered the following proposition:

If the government knows that there are no assets in the Social Security and Medicare “trust funds,” and yet projects the interest earned on those non-assets and the date on which those non-assets will be exhausted, then the government is lying.

That still seems correct to me: the whole idea of the Social Security and Medicare “trust funds” is a lie.  An institutionalized, ritualized lie that the U.S. government tells the American people. Perpetuated by both political parties, and others with an interest in hiding the reality of these programs’ unfunded liabilities from voters.  One that many journalists uncritically repeat.

The State of Social Security: Maybe a Little Better, Maybe a Little Worse?

The Social Security Trustees released their annual report yesterday, showing a small improvement in the system’s finances over the long-term.  That’s rather surprising given that the recent recession has reduced the program’s revenues and brought forward the date when the program begins to drain money from the general budget — from 2016 last year to 2015 in the new report.  The Trust Fund exhaustion date is 2037, the same as it was in last year’s report. 

The new health care law is likely to increase the program’s revenues as employers reduce payroll-tax-free health insurance coverage and offset the reduction in employee compensation through higher wages that would be subject to payroll taxes.  This sets up a competition between the health care law–induced increase in Social Security revenues and declines in revenues and increases in outlays for other reasons — a sluggish economy, improving longevity, the addition of another year at the end of the 75-year projection horizon, and changes in economic and demographic data, assumptions, and methods.

The positive revenue effect of the health care law (14 basis points) more than offsets the negative effects of all of the other factors (6 basis points) on the system’s long-range actuarial balance. That yields a total improvement of the program’s actuarial balance from –2.00 percent of taxable payroll to –1.92 percent.  In next year’s report, however, this year’s “legislative” effects may be folded into changes from technical adjustments and incoming data. We may never know whether today’s assumptions on the revenue effects of the health care law are correct or not. 

It could be that those assumptions are too large, especially if Congress postpones the tax on Cadillac health care plans because of pressure from unions. It could also be too small if many employers decide to eliminate health insurance coverage and opt to pay the less costly penalty.  On balance, I’ve concluded that, faced with such wide uncertainty about future outcomes, the Social Security trustees have chosen to be relatively conservative in their estimates of the health care law’s revenue effect. 

Another curious item is that the program’s long-range imbalance increased from $15.1 trillion to $16.1 trillion. However, the report states that “the near-term negative effects on employment of the slightly deeper recession than assumed last year are offset by higher than expected real growth in the average earnings level” (Section D: Projections of Future Financial Status).  As a result, the program’s total (infinite-horizon) imbalance ratio declines from 3.4 percent in 2009 to 3.3 percent today.

Note that a deeper recession and higher unemployment than was assumed last year does not necessarily justify a correspondingly faster recovery, with unchanged long-term equilibrium unemployment and earnings growth rates.  The trustees are discounting the possibility that the unemployment rate may remain higher than was assumed last year and that, therefore, earnings may not rebound any faster compared to last year’s assumptions.  It appears that that incoming data on unemployment and GDP growth played little if any role in informing assumptions about future earnings growth rates. 

Finally, it should be noted that this year there were no public trustees to oversee and modulate the report as it was being produced.

Topics: