But it seems that no entitlement is left un-raided: the legislation put forward by Senate Majority Leader Reid, which contains the raid on the Social Security trust fund, would also impose some accounting tomfoolery on Medicare. It’s well-known by now that Reid’s plan would increase the Medicare payroll tax to help offset the costs of the plan. What I didn’t know, though, is that these new taxes would first be laundered through the Medicare trust fund, creating an entirely fictitious improvement in Medicare’s financial health. The new taxes are credited to the Medicare trust fund, created an entitlement to new revenues from the rest of the budget. But the actual revenues will immediately be used to cover non-Medicare health costs. Looks like double-counting to me. The folks over at e21 explain.

A shortfall of 2 percent of payroll implies that if the 12.4 percent payroll tax were immediately and permanently increased by 2.0 percentage points (or benefits were cut across the board by 13 percent), the system would remain solvent through 75 years. In the 76^th year, however, it would once again become insolvent. This means that workers who paid a full career of tax increases wouldn’t receive their promised benefits. To make the system sustainably solvent would require an immediate increase in the payroll tax of 3.4 percentage points, to 15.6 percent of wages. That gives the scale of the problem.

So unless our view of the future changes radically, Social Security needs reform. But why should we fix Social Security today? Three reasons:

• First, fairness: Social Security’s treatment of different generations of Americans is declining, such that those who retire in the near future will receive much higher benefits relative to their taxes than those who retire later. For instance, this study shows that a typical couple retiring today will receive around a 2.3 percent rate of return from Social Security, while a typical couple retiring in 2050 will receive around a 1.7 percent return. Compounded over a full career of paying taxes, these differences amount to a lot. By acting today, we can lower returns a little for near-retirees so we don’t need to hit future retirees as hard.
• Second, efficiency: The necessary tax increases or benefit cuts I cited above are if we act today – and only if we act today. If we wait, the necessary changes will be larger. If we waited until the system became insolvent in 2037, we would need to increase taxes by around 3.9 percentage points, with further increases to come. It’s a standard finding in economics that the “deadweight loss” of a tax rises with the square of the tax rate. Whatever we’re going to do, it hurts the economy less if we do a little to every generation than to hit certain generations a lot harder.
• Third, uncertainty: people planning for retirement know that something will happen to Social Security, but they don’t know what, when, or to whom. The sooner we act, the sooner people can adjust their plans to account for those changes.

In a break from Democratic orthodoxy, House Majority Leader Steny Hoyer last week argued that Social Security reform discussions should begin in the fall. The Obama administration yesterday pledged to take on Social Security after they finish health care, but their record of standing up to the Democratic left, who strongly oppose even discussing reform, hasn’t been good to date. Another option, endorse by Hoyer, is a bipartisan entitlement reform commission. Legislation to form a commission has been sponsored in the House by Reps. Frank Wolf (D-VA) and Jim Cooper (D-TN) and in the Senate by Sens. Kent Conrad (D-ND) and Judd Gregg (R-NH). AEI held an event on the idea of an entitlement commission last week.

• Workers in their 30s are most likely to favor not paying into the Social Security system, with 59% responding they would rather not pay the taxes and not receive benefits.
• 51% of workers age 40-49 prefer to not participate.
• 44% of workers age 50 to 59 prefer to not participate in the Social Security program, and 39% of workers 50 and older would rather not participate.
• Even a significant amount of respondents who are nearing traditional retirement age would choose to stop paying Social Security taxes. One in three (33%) workers over the age of 60 said they would stop paying Social Security taxes even if it meant they would not receive any benefits.
• Almost half (47%) of Americans with a household income of less than $25,000 would choose to opt out of the system, and 48% of those making between $25,000 and $50,000 a year would as well.
• Slightly more than half (52%) of Americans making over $125,000 a year would choose to stop paying Social Security taxes and not receive the benefit.
• 57% of men age 40 to 49 would opt out of Social Security, while 45% of women in that age group would choose to opt out.
• 62% of men age 30 to 39 would opt out. Just over half (56%) of women age 30 to 39 would choose to opt out.

It reflects, in my view, not a confidence in markets or individual investment skills – both of which have been shown lacking in recent years — so much as a lack of confidence in the Social Security program, which has long faced significant deficits but which has also seen a similar lack of government action to correct them.

Now, for many of the workers surveyed opting out of Social Security would clearly be a bad idea. A worker nearing retirement today is almost sure to receive his full benefits, and could not hope to make up for them in the market. But again, this reflects a broad lack of confidence in the program, which may have transferred over due to government mismanagement of other sectors.

I have long thought that the most compelling argument for Social Security accounts was not higher returns through stocks, but the security that comes through personal ownership. (Unfortunately, the White House in 2005 focused on rate of return arguments, which are also factually dodgy.) Now, some will argue – not without evidence – that personal ownership of 401(k)s, homes and other assets did not protect Americans from the financial downturn. True enough. But I believe there is a strong, perhaps innate, preference for individual ownership that comes through even in difficult times. (And, as I’ve argued elsewhere, personal ownership has significant non-financial benefits to individuals and their communities.) This survey appears to buttress that view.

If and when President Obama chooses to address Social Security – and I have good reason to believe he will choose to do so – lawmakers on the right need to promote policies that address Social Security’s significant financial shortfalls while retaining and strengthening its safety net against poverty in old age. The key to this, however, is also building significant new retirement savings by individuals. It is only if personal saving rises that Social Security can be reformed to hold the line on costs and target its mission more closely toward low earners. If Social Security continues to act as a de facto retirement saving vehicle for middle and high earners, program growth and significant tax increases are inevitable.

The “it’s only a health care crisis” argument has something to it, so it’s wrong to dismiss it out of hand. Medicare and Medicaid are the largest drivers of long-term entitlement costs. However, it’s also worth considering these claims in a different light, since alternate viewpoints may lead to alternate conclusions. To start, here’s how Henry Aaron lays out the argument:

Here is what the projections indicate. Over the next four decades, government spending on all entitlement programs other than Medicare and Medicaid will increase negligibly as a share of national outputÑby only about 1 percentage point of GDP. That change is the difference between a projected increase of roughly 2-percentage points in the share of income going to pay social security benefits and a nearly 1-percentage point drop in spending on other non-health entitlements.

Meanwhile, Congressional Budget Office projections indicate that national health care spending will skyrocket, rising from 16 percent of gross domestic product to 37 percent by 2050. Aging of the baby-boomers explains some of the increaseÑthe old cost more to care for than do the youngÑbut not much. If population aging were all that is going on, national health care spending would rise by less than a quarter as much as current projections indicate.

Aaron then argues:

The challenge is to design ways to pay for and produce health care services, private and public, so that spending goes for services expected to produce benefits worth what they cost, are produced as efficiently as possible, and are available equitably to all Americans. Then, the nation must decide how best to pay for that careÑthrough premiums, out-of-pocket charges, and taxes.

Taken on its face, there’s not much to argue with here: the cause of the fiscal gap is so-called “excess cost growth” – that is, growth of per capita health spending in excess of growth of the economy – and to address it we need to fundamentally reform private and public sector health care provision. That’s pretty much what President Obama’s plans for health reform amount to, although there’s some reason to believe that the two would collapse together: that is, private sector health care would become increasingly untenable under the constraints imposed under Obama’s proposals, and more and more working-age Americans would resort to the government-provided health coverage alternatives that are a central part of Obama’s plans.

But here’s an alternate take on the sources of the overall entitlement “crisis,” based on some previous work I’ve done.

First, Aaron’s argument boosts the health share of total entitlement spending increases by netting out increases due to Social Security – around 2 percent of GDP, or around 10 percent of typical total federal spending – against projected reductions in other non-health entitlement spending. It’s arbitrary to net out certain spending reductions against Social Security instead of Medicare, but in any case CBO’s baselines often tend to project reductions in certain federal spending due to how the baselines are constructed, even if these reductions aren’t likely to come to pass. It’s pretty rare that any sector of federal spending declines by 1 percent of GDP, particularly entitlement spending. So I’m not sure I’d bank on that extra cash. In any case, through this netting out Aaron concludes that the whole action is in the health care sector.

Second, Aaron argues that the action in the health care sector is almost entirely in excess cost growth, not population aging. So if the entitlement crisis is really a health care crisis, and the health care crisis is really an excess cost growth crisis, then it makes sense to focus policy on curbing the growth of per capita health care costs, not on issues related to the aging of the population. Moreover, since health care cost inflation in Medicare is related to health care costs in the private sector, we therefore need to reform private sector health care. In this way, Obama’s proposed health care reforms – which do almost nothing to actually restrain Medicare’s growing costs – can be portrayed as entitlement reforms, while simultaneously giving the increased federal control over private sector health care that many on the left desire.

But there’s a bit of divide and conquer here. Almost all the increases in Social Security spending are due simply to population aging, plus part of the cost increases in Medicare and Medicaid. If we divide total increases in Social Security, Medicare and Medicaid spending between population aging and excess health care cost growth, it comes down to right around 50 percent for each. The chart below shows the annual evolution of cost increases. Population aging includes all cost increases for Social Security, plus Medicare/Medicaid cost increases due to the aging population. Excess cost growth includes costs increases due to the fact that individual health care costs are rising faster than the economy.

Clearly, excess cost growth is the bigger factor over the long term. But in the short and medium term, meaning the next three or four decades, population aging is actually the larger cost driver for federal entitlements. Moreover, the larger role of excess health care cost growth in later years is more or less academic: if we haven’t fixed entitlements over the next several decades, they’ll “fix themselves” by collapsing. The short story is that over the foreseeable policy future, population aging – not structural factors in health care – is the biggest driver of rising entitlement costs.

Aaron and others are correct that we need to address health care cost inflation, for the reason that in many cases Americans aren’t getting full value for their money. But given the large role of population aging, there are other policies besides a potential federal takeover of the health care sector to consider: policies to keep Americans working longer, which effectively holds back aging by raising the ratio of workers to retirees; policies to support families, which may raise birth rates, and potentially even policies to increase immigration. None are magic bullets and all have downsides, but are worth at least considering as we confront large entitlement cost increases in coming years.

The key takeaway here is to look carefully at claims that only health care costs increases are behind the entitlement “crisis.” A good chunk of future entitlement deficits are simply due to the fact that America is getting older, and we can address that without a potential federal takeover of health care economy-wide. 

For a more detailed analysis, see this paper from the American Enterprise Institute.

President Obama is eager to seek a bipartisan solution to ensure the long-term solvency of Social Security, people who have spoken with him say, but he is running into opposition from his party’s left and from Democratic Congressional leaders who contend that his political capital would be better spent on health care and other priorities.

However, the pressure isn’t simply to focus Obama’s political capital where its returns may be highest. It is pressure against taking any action on Social Security, in particular any action that might imply any reduction in future benefit growth. This article from Dean Baker is pretty typical of the reaction from the left. (Also see this piece from Barbara Kennelly and this summary post from Jane Hamsher.) What I’m hearing, for what it’s worth, is that the administration is putting Social Security on the back burner for now.

Here’s my two cents: the fiscal gap is almost entirely caused by rising spending, not by lower tax receipts, and so within the reams of the politically feasible it seems that a good chunk of that gap will have to be solved by restraining spending growth. While Medicare is the 800 pound gorilla of the entitlement mess – I guess making Social Security the 200 pound chimpanzee who nevertheless carries a mean bite – Obama’s health plans don’t include any reasonable reductions in Medicare spending growth (or, for that matter, any health care spending growth) in the near future. Obama had to know that solving the long-term budget mess must include restraints on Social Security cost growth. And he also had to know that many in his own party would resist necessary steps to keep Social Security costs under control. The only question is whether Obama’s new style of politics would involve standing up to his own party. So far, it hasn’t.

The 1983 Social Security amendments are today often invoked as a process model for conducting politically difficult entitlement reforms. Success in reforming entitlements now depends not only on whether negotiators can summon a similar bipartisan spirit, but also on whether they are willing to define the problem as transparently as negotiators did in 1983, and on whether they avoid analytical shortcomings that have set back subsequent reform efforts.

Social Security reform will always require difficult compromises between individuals with strongly-held opposing views.  Today, negotiators face a daunting additional barrier: fierce disputes about the nature, size and immediacy of the problem to be solved. 

There is today a significant danger of misremembering and misapplying the lessons of 1983.  The 1983 effort succeeded in large part because participants from across the political spectrum had a shared understanding of Social Security’s financing shortfall.  Analytical methods were employed that prevented issues such as Trust Fund accounting from injecting confusion and discord into the discussion. 

Ironically, the 1983 amendments themselves, and subsequent accounting changes, have inadvertently destroyed this shared understanding, giving rise to pervasive misimpressions about Social Security’s current finances – and about the intended effects of the 1983 reforms themselves.  Specifically, negotiators in 1983 did not intend to run decades of surpluses to amass a large Social Security Trust Fund, nor would they have believed that doing so would effectively pre-fund benefits in deficit years now projected for 2017 and beyond. 

Click here to read the whole paper.

Word has it that President Obama intends to appoint a task force the week after next which will be charged with “reforming” Social Security. According to inside gossip, the task force will be led entirely by economists who were not able to see the $8 trillion housing bubble, the collapse of which is giving the country its sharpest downturn since the Great Depression.

Dean, as some may know, is not a big fan of Social Security reform.

I have no independent knowledge of what an Obama administration task force on Social Security reform would do or look like. It might be modeled after the Clinton administration’s working group on Social Security, headed by Jeff Liebman who has returned to a post at OMB in the Obama administration. (This paper by Liebman, Doug Elmendorf and David Wilcox discusses some of that process.) If the task force is purely internal and aimed only at providing policy recommendations for President Obama, the Clinton administration model isn’t a bad one to follow, and the Obama administration staffers available to work on it – Liebman, Jason Furman, Peter Orszag and others – are clearly up to the job.

If, however, Obama wishes for a task force to build political momentum toward reform, he’ll want a broader group. Based on my experience on the staff of President Bush’s 2001 reform commission and work with the White House on reform in 1995, here are a few quick thoughts:

First, to build any political momentum a working group should represent the reasonable spectrum of views on Social Security. Something like the Diamond-Orszag reform proposal probably represents a reasonable bound on the left, as it retains the current program in pretty much its current form while making it sustainably solvent almost entirely through increased taxes. Something like the President’s Commission’s “Model Two” is a reasonable bound on the right, since it likewise makes the program sustainably solvent entirely by reducing benefits. There are folks on both the left and right whose views make it such that any reform compromise would be almost impossible – on the left, because many believe Social Security doesn’t even face a deficit, and on the right because they believe that simply borrowing a few trillion dollars and investing it in personal accounts will fix the whole problem. Neither view is correct. It may be possible to get some folks further left on board by proposing auto-correction policies that only come into effect to the degree the system faces deficits; I discuss one option in this paper. If they’re right that Social Security doesn’t face any shortfall, nothing will change.

One problem with President Bush’s 2001 Commission was that it didn’t represent the reasonable spectrum of beliefs on Social Security reform. This didn’t make it a dishonest commission; like President Roosevelt’s Committee on Economic Security, it was designed to put flesh on the bones laid out by the President. In this case, the Commission was tasked with designing a reform plan that included personal accounts and excluded tax increases. That said, a commission only builds political capital toward enacting reform if it’s seen as building a consensus through a process in which all views have been heard. In both the 2001 Commission and the later 2005 reform drive, Democrats didn’t feel they were part of the process. They clearly will be a central part of the process this time, but the goal will now be to include Republicans. Just as Republicans shouldn’t reflexively oppose any Obama administration reform plans for political reasons, so Democrats shouldn’t seek to exclude Republicans from the process.

Second, a reform task force should include a variety of different players, including members of government, both legislative and executive, representatives of outside interest groups, and experts who can provide technical advice and help ensure the integrity of the reforms decided upon. The 2001 Bush Commission didn’t include any sitting Members of Congress and only a small fraction of commissioners had the technical expertise needed to make the plans the best they could be. A broader group would be helpful.

Third, any task force or commission needs time. The 2001 Commission ran roughly from May through December of that year and had to conduct a number of public hearings. This was simply too much to do in too little time, and as a result the plans were fairly bare bones. There is plenty else on the policy agenda at the moment, so there’s no reason not to give a working group a year or more to put things together.

Finally, on a non-commission related topic, one comment by Dean is worth noting. He says,

“The vast majority of baby boomers will be approaching retirement with little other than their Social Security and Medicare to support them.”

While this is a pretty compelling talking point, it really just isn’t so and any reform plans should take the facts into account. The table below from this paper shows that Social Security accounts for around 39 percent of total income for new retirees in 2005. It may be a little higher now, but probably not that much. Moreover, to the degree we want people to have diversified sources of retirement income – and most Democrats at least talk the talk on that – raising taxes to pay higher Social Security benefits will only tend to squeeze out other forms of savings and make tomorrow’s seniors more reliant on Social Security. That can’t be a good thing.

At the same time, it’s worth checking into how Social Security’s investments have done over time. Surplus taxes paid into Social Security are invested in the Old Age, Survivors and Disability Trust Funds (OASDI), which hold special-issue government bonds whose interest rates are based on average Treasury bond interest rates at the time. The idea here is investments which provide safe, if modest, returns for the long-term.

But not many people have considered how modest. Effective annual interest rates on the trust funds are available through the Social Security actuaries’ web site (see here). To calculate real returns I subtracted the annual rate of growth of the consumer price index (CPI), available here. A couple charts tell an interesting story. First is a fairly conventional comparison: how did the trust funds’ returns compare to a mixed portfolio of 50 percent stocks and 50 percent bonds? The first chart shows average annual returns by decade and shows a couple interesting things. First, the mixed portfolio returns exceeded the trust fund’s returns in all decades except for the truncated 2000-2008 period, by an average of around 2.9 percent. Second, both the stock-bond portfolio and the trust funds lost money in two decades, although only the trust funds had a truly terrible decade, losing 3.5 percent annually during the 1940s.

The second chart shows a running average return on the trust funds, beginning in 1940. The return value for each year represents the average of returns from 1940 through that years. Here’s something I found pretty interesting: from the program’s inception through 1986, the average annual return on the trust funds was negative. To repeat, through the first four and one half decades, the trust fund’s investments lost money on average each year. Following 1986 the running average of annual returns was positive, but barely so: even extending through 2008, the average annual return on trust fund investments, adjusted for inflation, was only 1.38 percent above inflation. These returns are safe, to be sure, but far lower than the 4.4 percent real annual return on the stock-bond portfolio.

So here’s a question: if the trust fund’s returns have been so low, how did Social Security manage to pay such high benefit returns to early retirees? (The benefit return is a function of taxes paid and benefits received, with the trust fund’s investment return having an only indirect effect on benefits.) We’ve talked here several times about the high returns paid to early retirees; here’s a chart showing average annual returns paid to beneficiaries. The answer is that while a sustainable Social Security program would have built up a significant trust fund balance over time to help pay future benefits, the trust fund balance was kept very low and the extra funds paid out as benefits. When Social Security was begun, the idea was for it to become a “funded program” carrying a large trust fund balance. Congress soon acted to delay scheduled tax increases and move up the payment of benefits, in addition to making benefits more generous. (Lesson: past Congresses were pretty much like present ones in terms of catering to current voters over future ones.) High benefits were paid, at the expense of the trust fund balance that could help the system fund itself in perpetuity. This was, in effect, like eating your seed corn: things look good in the short-term, but you don’t have the means necessary to keep things going for the long run.

Spitzer begins by simply highlighting that “Since Jan. 1, 2005, the year President Bush proposed the idea, the Dow Jones industrial average has dropped from 10,783 to around 8,000, a drop of more than 25 percent.” Ok, but in the Bush plan older workers accounts would automatically switch to bonds beginning at age 55, meaning that most people retiring today would have had much less exposure to that falling market than younger workers.

Then Spitzer says that he will quantify the losses to today’s retirees, “abeit roughly.” Spitzer should have said albeit very roughly, since Spitzer relies for his numbers on a Fortune article by Allan Sloan that Sloan himself later retracted due to technical misunderstandings of how personal account plans would have worked.

In this recent post, I gave the results of a much more detailed simulation of personal accounts using historical market returns. The short story is that there is no historical period in which a worker holding an account for a full career would have failed to significantly increase his total Social Security benefits, including workers retiring today. Moreover, even workers who held an account for only a few years before retiring under today’s market conditions would have experienced only tiny declines in total Social Security benefits.

Finally, Spitzer says, “as Paul Krugman has pointed out, the would-be privatizers make incredibleÑeven impossibleÑassumptions about the likely performance of the market to justify their claim that private accounts would outdo the current system.” The basic argument here is that as a slow-growing workforce reduces long-term GDP growth, stock returns must be lower as well.

But here’s the problem: first, it’s not the “privatizers” who make assumptions regarding future stock returns, it’s the non-partisan actuaries at Social Security and economists at the Congressional Budget Office. Second, as I showed in this blog post using historical data from sixteen countries over a 100-year period, the correlation between GDP growth and stock returns is statistically very weak.

There are plenty of good arguments to make against adding personal accounts to Social Security. Spitzer should have spent more time actually making them.

To date, I’ve not seen similar reports on President Obama’s process for vetting his senior appointments, which has included:

• New Mexico Governor Bill Richardson pulling out of his nomination to be Commerce Secretary due to an FBI investigation of so-called “pay to play” practices in which recipients of state contracts were expected to raise funds for elected officials.
• Treasury Secretary Timothy Geithner, who was confirmed to the spot despite not having paid over $34,000 in Social Security taxes during four years in which he worked at the International Monetary Fund. Geithner, like other American IMF employees, received and signed multiple statements indicating knowledge of his requirement to pay payroll taxes, for which in any case he received additional compensating pay from the IMF. (Meaning that, even had he paid his taxes, he would have paid no net taxes.)
• Health and Human Services nominee Tom Daschle, the former South Dakota Senator and Senate Majority Leader who withdrew his nomination after revelations that he had failed to pay over $146,000 in back taxes, and that he had delayed revealing this failure for a number of months.
• Chief performance officer designee Nancy Killefer backed out of the job after revelations that the District of Columbia placed a lien on her house after her failure to pay unemployment taxes for household help.

If one of their children gets pregnant out of wedlock, however, I’m sure the press’s full investigative resources will be brought to bear.