Paul Krugman Writes About the Long-Term Budget Math of Social Security

Paul says:

Long-run budget math - Paul Krugman: Start with the current position. Last year, federal spending on Social Security, Medicare, and Medicaid was 8.5 percent of GDP, equally divided between Social Security and the health care programs. Dismal long-run projections, like those of the GAO, have this total rising by 10 percentage points of GDP by mid-century.

So, how much of this is a Social Security problem? Pundits like Tim Russert love to point out that in its early days Social Security had 16 workers paying in for every retiree receiving benefits. But this is irrelevant; looking forward, we’ll see the worker-beneficiary ratio fall from about 3 to 2 as the baby boomers retire. This will raise the percentage of GDP spent on Social Security from about 4 to 6 — that is, a rise of about 2 percentage points of GDP, which is a small fraction of the entitlements problem.... What’s more, Social Security has already been strengthened to deal with this rise. In 1983 the payroll tax was increased and adjustments made to the retirement age, so as to build up a trust fund.... This brings us to the claim that the trust fund doesn’t exist, because it’s invested in government bonds. The full explanation of why this is sophistry is here.

The bottom line is that Social Security is just not the major problem. Now, part of the projected rise in Medicare and Medicaid costs represents the effects of an aging population. But as a new report from the CBO explains, demography is only a minor factor — mainly it’s rising health care costs. What’s more, the proposed “solutions” for the Social Security problem have no relevance to the issue of rising Medicare costs.... The Beltway obsession with Social Security is a classic case of a little knowledge being a dangerous thing. People have picked up a few facts about demography, and think they understand the long run budget problem. They don’t.

PS: OK, from some communications I see that 2017 — the projected date at which payroll taxes no longer cover benefit payments — has raised its ugly head. But there is no interpretation under which 2017 matters. Social Security legally has its own dedicated funding; if you believe the government will honor the law, the surpluses the system is now running are building up a trust fund, which will finance the system for decades after 2017, and maybe forever. If you think the law will be ignored, then Social Security doesn’t really have its own budget — the payroll tax is just one of many taxes, and SS benefits are just one of many government costs. In that case the relationship between payroll taxes and benefits is irrelevant.

The only way to make 2017 matter is to change the rules midway: when SS runs surpluses they don’t count, but when it runs deficits they do.

Comments

[Ken]

Don Boudreaux, at CafeHayek notes:

"The only evidence that Krugman presents to support his case against the proposition that Social Security is headed for insolvency (unless it undergoes big changes) is simply that Medicare and Medicaid are headed for insolvency that's even worse."

[MaryLou]

Obviously Don Boudreaux didn't follow the links embedded in the NYT piece. In a separate (linked) essay, Krugman explains about the 'trust fund' and the concept of 'running out of a surplus' in 2017 in more depth. Joe Sixpack might not be a Beltway policy guru, but he's not going to fall for the con job necessary to make the case for a SS 'crisis'.

[John Morrison]

Joe Sixpack might fall for the con job, if the Peter Pundit gets into his expected mode and repeats the propaganda over and over and over again. I'm not sure how much Peter Pundit believes what he says, versus how much he lies or BSes. In any case, he can't be corrected.

Paraphrasing Upton Sinclair, it's nearly impossible to get Peter Pundit to understand something if his job depends on his not understanding it.

This is (of course) yet another counterexample disproving the idiotic assumption of perfect rationality and perfect information in economic behavior. Peter Pundit is behaving in a rational way maximizing his utility or income, by refusing to understand. He can only succeed if he is not uniformly viewed as a BSer, in which case others behave irrationally and/or lack information.

[bigTom]

Well since the trust funds will have been spent instead of invested, the post 2017 general fund will have to begin paying back with interest what it shouldn't have borrowed earlier. This debt is not really different from other government debt, which it can be argued has been growing unsustainably. So the real question is under the severe budget constraints that we think are likely will US government be forced to default on its obligations. And if it is forced into a partial default who gets precedence in the (bankrupcy sense), social-security recipients, or (mostly) foreign bond holders.

[Bruce Webb]

I left a comment at Krugman that should survive moderation. In case not:

No one and I mean no one dared suggest that the Special Treasuries in the Trust Funds were ' Phony IOUs' prior to the stretch of Social Security Reports that pushed Trust Fund Depletion from 2029 (1997 Report) to 2037 (2000 Report). This rate of improvement spooked opponents and so they flailed around and settled on Shortfall (Income excluding Interest lagging cost) instead of Depletion (Income including Interest lagging Cost) ignoring the fact that this just equated to abrogation of Full Faith and Credit of the United States. The argument for crisis at depletion makes logical sense though not exactly surviving encounter with actual numbers. The argument for crisis at shortfall (2017) requires tying your policy, politics and economic arguments into a pretzel. There is no numeric case and there is no plausible political case for defining Shortfall as crisis. The numbers don't run. I've got Data Tables and I am not afraid to use them.

[pgl]

Paul's last sentence was his best. After all, changing the rule is precisely the GOP agenda since George W. Bush took office in 2001. See Dean Baker's latest takedown of the WaPo and his discussion of Fred Thompson's Soc. Sec. proposal as an example.

[save_the_rustbelt]

Take a look at the latest polls, the American people do not trust George Bush and do not trust Congress.

And why would they (trust them)?

Social Security is as solvent as the integrity of our politicians. That should scare us.

[Albrt]

Why would anybody think that the rules would NOT be changed midway, or at whatever other point is convenient? The rules have already been changed multiple times, resulting in a benefit formula that will work out well only for the cohort who need computers, flat screen TVs and other components of the CPI during retirement, rather than food, fuel and other commodities.

[Dirty Davey]

The 1983 Social Security fix was in effect a deal: "For the next couple of decades, we'll collect more than we need from the regressive Social Security taxes and use that to buy bonds to fund spending. Then for the following few decades we'll redeem those bonds and use the progressive income tax to prevent a shortfall in Social Security."

So of course there's a core group of Republicans opposed to progressive taxation (among other things) which would like nothing more than to default on that "deal"--to in effect abscond with the SS "Trust Fund" by putting off paying back the bonds indefinitely. It is entirely possible that this was the goal all along, and that the Republican agreement to the 1983 compromise was not in good faith.

To my mind, the simplest fix to the Social Security "problem" would be to allow the SS administration to sell the bonds they hold on the open market. That way there could be no question of trying to quietly ignore or default on the SS bonds without anyone noticing.

[Rob]

bigTom---No, no, a thousand times no. This is what will happen in 2017. The SSA will hand in a bond. The Treasury will write a check. The Treasury will pay for this check by auctioning off a new bond. The net debt position remains exactly the same. This same things happens all the time when a bond matures. Somehow no one runs around saying bonds maturing means we need raise taxes to pay for them.

[Slocum]

"Social Security legally has its own dedicated funding; if you believe the government will honor the law..."

My belief is that whether or not the government honors the law is irrelevant. The government surely will honor the legal fiction of the trust fund while making whatever adjustments necessary to benefit levels, retirement ages, means testing, or payroll taxes to assure manageable levels of 'withdrawals' from the 'trust fund'.

"This will raise the percentage of GDP spent on Social Security from about 4 to 6 — that is, a rise of about 2 percentage points of GDP..."

Just two itty bitty percentage points of GDP? Sounds like a 50% increase to me (and recall that the current overall budget deficit is less than 2% of GDP).

But in a larger sense, Krugman and I are on the same page because he thinks we should do nothing to save social security now, and I agree -- we can have that political fight when the pain of declining SS revenues and increased outlays gets bad enough. I certainly don't want to see higher taxes now to 'invest' more in the 'trust fund'.

[bakho]

Good point Rob. A minor difference is that the interest on SSTF goes directly into the SSTF account. Once the SSTF starts depleting, the interest actually leaving the government will increase.

So in 2007, around $110 Billion in interest payments were transferred to the SSTF. This $110 Billion increases the Total debt, but does not require the selling of bonds.

Net interest payments on about $5 Trillion dollars (in 2007) at 4.7% interest was about $235 Billion dollars. If the SSTF were instead paid down, the interest flowing out would be $345 Billion. It is the difference between having $339 Billion 'on budget' deficit instead of a $181 Billion 'off budget' deficit.

However you look at the problem, collecting enough revenue to pay off SS and Health care obligations could be done and the US tax rate still remain below that of most EU countries. It won't break the bank.

[kharris]

Rob is correct. The argument being made is for selective, intentional default. That is why we keep hearing the argument that Trust Fund assets are fictional, that the Fund holds "nothing but IOUs", as if somehow IOUs are a bad thing.

Think of the vocabulary of finance. We use the words "bond" and "trust" and "obligation" for a reason. There is no reason to expect that in the normal course of things, the Social Security Trust Fund will be unable to make full payment. It is the abnormal run of things that we need to worry about. It is intentional, selective default that must be the concern.

[Dirty Davey]

"It is intentional, selective default that must be the concern."

Hence my argument that the best way to prevent the looting of SS is to allow the open public sale of the bonds in the trust fund. If the SS bonds are indistinguishable from any other bonds, selective default is impossible.

[Brooks]

Both right and left are making a fundamental analytical/conceptual error. Social security "solvency" is completely IRRELEVANT to our policy choices (whether or not to reduce benefits and/or elegibility, raise SS taxes, both or neither). We have a dangerously high OVERALL long-term fiscal imbalance. ANY reduction in projected spending and ANY increase in projected revenues will mitigate that imbalance. The fact that we carve out of total revenues a particular amount to dedicate to a particular program -- thus making it "solvent" (or not) -- is meaningless from the standpoint of policy choices. See elaboration and illustrations at http://swordscrossed.org/node/1720

[dissent]

I am very annoyed with Obama for picking up on Bush's and the Republicans attack on social security.

I think the structure of the tax is deliberate: the New Deal policy folks knew the viciousness of America's elites and put social security out of harm's way by not taxing these elites (much) for it. This is a masterful piece of political handiwork.

Raise the tax on upper crusties, and social security will be hammered by them, maybe not right away, but it will be hammered.

Obama is a fool not to recognize this. Where has he been the last 15 years? Has he not seen the character of these people?

[howard]

Brooks, social security is funded by a dedicated tax. you eliminate or reduce the spending on social security, you reduce the income as well to offset.

ergo, social security has nothing at all to do with long-term fiscal imbalances.

in the closest case i've seen the supreme court rule - the highway trust fund - the ruling made it clear that the money isn't fungible, that a special purpose tax must be spent on its special purpose.

[Brooks]

Howard,

I think you're missing my point. The fiscal imbalance is the difference between projected TOTAL expenses and TOTAL Revenues, so ANY reduction in total expenses (including a reduction in projected SS spending if other expenses are held constant or increase by a lesser amount), if TOTAL revenues are kept constant (say, by reducing the SS FICA tax and increasing other taxes in a revenue-neutral fashion), then the fiscal imbalance is reduced (and without defaulting on any bonds/legal obligations, assuming we spend on SS what we are currently legally obligated to spend on SS).

See?

And I'm not saying we SHOULD do that. I'm just saying that SS "solvency" is irrelevant to the decision of whether or not to do so. Long-term "solvency" is just a function of how much we continue to carve out of total revenues and dedicate to SS, and we can adjust that up or down as we see fit. So it shouldn't be a consideration as we weigh our policy options. Do you see that?

[bakho]

A shortfall if there is one is less than 1% of GDP. There is nothing about SS that is not affordable.

[Slocum]

"If the SS bonds are indistinguishable from any other bonds, selective default is impossible."

No, no, no. What makes SS bonds different is the fact that the same government that redeems the bonds is the very same that sets the policies that determine how many bonds to redeem.

SO...if Congress determines it can't afford to redeem $XX billion worth of bonds in 2018, it doesn't have to default, it only has to adjust the cost of living increase or the retirement age or increase the payroll cap or introduce means testing of benefits -- there are several parameters.

The U.S. government won't ever default on the bonds in the trust fund. But it can choose to redeem them (or not) on whatever schedule it likes.

[Brooks]

Rob,

You write:
"This is what will happen in 2017. The SSA will hand in a bond. The Treasury will write a check. The Treasury will pay for this check by auctioning off a new bond. The net debt position remains exactly the same. This same things happens all the time when a bond matures. Somehow no one runs around saying bonds maturing means we need raise taxes to pay for them."

It sounds like what you are saying is that funds will be transfered (intra-governmentally) from general funds to SS, SS will spend that amount, and the government will then borrow to replace that amount. Is that what you're saying? If so, while this may not increase the total debt (i.e., including debt "owed" by one part of the government to another, which, from a macro standpoint, is like one of of the pockets of our pants owing the other pocket money), it WOULD increase the part of the debt that matters: debt held by the public. Right? And this would exacerbate any overall fiscal imbalance, right?

[urban legend]

What is the difference between (a) not redeeming bonds that the government is legally obligated to redeem because they are needed for Social Security to pay legally mandated benefits, and (b) defaulting? Either one is not paying on obligations backed by the full faith and credit of the United States of America. Either one would cause an enormous international financial crisis, and would be permanent political disaster for anyone who tried it.

[Barkley Rosser]

Actually, what will happen in 2017 is probably that the fund will still be running a surplus and will continue to do so forever, if left alone. The projections that have it going into deficit then (and going "insolvent" later) have proven to be ridiculously pessimistic, but somehow they never get adjusted. Just the dates of the various supposed "crises," which keep getting moved further into the future.

[Dirty Davey]

I wrote:

"If the SS bonds are indistinguishable from any other bonds, selective default is impossible."

Slocum wrote:

"No, no, no. What makes SS bonds different is the fact that the same government that redeems the bonds is the very same that sets the policies that determine how many bonds to redeem."

My point is that if the SS fund were able to sell the bonds, then once they are sold the government no longer can choose whether or not to redeem them.

[urban legend]

". . . .the same government that redeems the bonds is the very same that sets the policies that determine how many bonds to redeem"

But this is not a matter of "policy." The obligations on those bonds are written into law. When the benefits are due, the benefits are due. They are paid, a bond is retired or "redeemed" from the Trust Fund account, and a charge is entered against the general fund of the U.S.A. Besides the fact that a default on its debt obligations would cause a major international financial crisis, anyone who would even hint at changing the law to do so would be committing permanent political suicide. It ain't gonna happen.

[Michael Sullivan]

"nothing but IOUs"

Yes, I love how people toss around that phrase as if there were any financial asset of consequence that is not at its root "nothing but an IOU", including dollars, euros and stocks.

The whole economy is predicated on the *huge* efficiency gain inherent in being able to treat certain special kinds of "IOUs" as having relatively stable tradable values.

If we really believed that IOUs were essentially worthless irrespective of their form and guarantor, and we put that belief into practice, we'd be living in the stone age.

When I see economists using that phrase to describe the SSTF, I really have to wonder about their intellectual honesty.

[Neil Bates]

It would be so easy (technically, not politically) to just take the cap off the SS-taxable income. That would keep Social Security solvent for decades - but the defenders of plutocracy will obstruct. Why really pains me, why is Hillary against taking the cap off?

[wood turtle]

I see no problems ahead.

[urbanlegend]

Neil -- It's not defenders of plutocracy who object to raising the cap, it's people -- like FDR himself -- who strongly believe in more progressive taxation for the general fund, but are trying to protect Social Security. Making sure the strong connection is maintained between lifetime withholdings and benefits -- which means almost everyone has a meaningful stake in the program -- is the best protection against political attack by the plutocrats as another welfare program that can get cut when the going gets tough. The cap is absolutely essential to continuation of Social Security as social insurance. People usually on the right side of things need to ge a better understanding of that: why FDR and Democrats ever since, otherwise supporters of a progressive tax structure, have been absolutely devoted to maintaining a cap on both earnings subject to withholding and benefits.

[Barkley Rosser]

Neil,

The problem has been discussed in other posts on this. It is fine for Obama to propose just raising the cap. Maybe that in itself would not be a bad thing. But that is not going to happen.

Why not? Republicans in the Senate will be utterly opposed to any tax increase. They will filibuster it, so 60 votes will be needed for the cloture vote to halt the filibuster. Despite their problems, I do not see the Republicans going under 40 in the Senate. So, some of them will have to be brought on board. They will demand other things, certainly future benefit cuts, and also probably some form of privatization.

Given the flimsiness of the projections of problems, do we need to do all that, especially now? Wait a decade and see how things go. It is more likely than the opposite that we will see fica tax cuts because of the ever-accumulating surplus on the SSTF by 2017 than that we will be worrying about what combo of tax increases and benefit cuts we will need to do in order to get SS out of what medicare is experiencing right now, that is, some of its expenses being paid by general revenues.

[Brooks]

Who says that a tax increase to provide incremental funds for SS must come from payroll taxes? Why not just add an additional 1% or 3% or whatever onto capital gains taxes and dedicate that to SS (i.e., revenues going to the "trust fund"). Given that SS taxes paid aren't really going into some savings account (individual or collective) and are used on a pay-go basis anyway, why should we assume that we are stuck inside the box of the current general structure for revenues?

[Brooks]

My point above was NOT to advocate that we do so (fund with incremental cap gains taxation), but merely to point out that people are really stuck in "inside-the-box" thinking re: SS financing and policy options. We really need to start focusing on what matters: projected TOTAL revenues vs. TOTAL expenses, what economic trade-offs are involved with alternative policies, and what trade-offs fit our values and priorities.

Get outside the box, folks! We can choose how much to spend on SS, how much to spend overall, how much (if at all) via a tax dedicated to SS, and how much to tax overall.