25 Aug 2010

Two Men Enter, One Man Leaves: Murphy vs. Rosnick on Social Security

Shameless Self-Promotion 8 Comments

Another Public Square debate for you. (Here’s my lead-off essay, and then you can click through to the others from there.) If you ever wanted to read four essays on whether we should count the trust fund as legitimate–and still not have an answer at the end–then this debate is for you.

My favorite exchange:


Consider a silly scenario: Suppose an intern working at the Social Security Administration accidentally tips the drawer holding the trust fund notes into a paper shredder. She looks on in horror as $2.6 trillion in bonds issued by the U.S. Treasury are destroyed.

Naturally the intern’s boss is furious. He doesn’t know what to do, since the Social Security system is now $2.6 trillion poorer. They have nothing to sell off to cover their operating deficits.

On the other hand, officials at the Treasury Department are elated. Tim Geithner himself calls up the intern and congratulates her on reducing the national debt by $2.6 trillion. The incoming federal tax receipts that would have previously been used to pay the interest and principal on those outstanding bonds can now be spent on something else (or returned to the taxpayers).

But while one set of government officials is distraught while another is delighted, from the point of view of taxpayers, it makes no difference. They’ll still have to make up the $2.6 trillion gap either way.

Our hypothetical scenario illustrates the silliness of counting Social Security’s holdings of Treasury securities as genuine assets.

Then Rosnick:

Another argument is that the trust fund is irrelevant in any case because taxpayers “still have to make up the $2.6 trillion gap” if the bonds were put through the shredder by a careless intern. On this we agree: The physical presence of the bonds means nothing. When was the last time a bond trader actually held an actual paper bond? Is there some poor bike messenger every business day tasked with the receipt and delivery of billions of dollars in stock certificates? What of the trillions of dollars in currency exchanged every trading day?

It is unfortunate that Murphy might be unaware—or worse, pretend not to know—that if one can establish ownership of a lost savings bond, the government will simply issue a replacement. On the other hand, a good fantasy might give ordinary workers the idea that the trust fund is a fragile thing—easily lost—and the best thing to do is to get used to the idea of getting along without it.

Point taken. And come to think of it, who ever heard of a talking pencil?! Man these free-market writers are either idiots or liars.

I am too lazy to verify this, but I think Rosnick makes a huge mistake when he says:

The Congressional Budget Office predicts that the projected seventy-five-year shortfall in Social Security is less than one-half of 1 percent of GDP over that time—the equivalent of $70 billion today. By comparison, extension of the Bush tax cuts (not including an AMT fix) would cost $116 billion in 2011 and $221 billion in 2012.

What he is implying in the above quotation–and which he explicitly says in the final essay–is that Congress could completely solve the Social Security crisis just by removing the Bush tax cuts for a single year.

As I say, I think this is completely wrong, but I’m too lazy to actually run the numbers. If you click on the CBO report, they certainly don’t say that the Social Security actuarial gap is only $70 billion in present dollars. No, as I point out in both of my essays, the government actuaries say Social Security has a PDV of minus $7.7 trillion.

So I think what Rosnick did is take the CBO’s projection of the shortfall (over the next 75 years) as a fraction of the economy (over the next 75 years). Then to give us an idea of what that number means, he expressed it in terms of current dollars.

If it were really true that $116 billion in 2011 could plug the Social Security gap (at least for the next 75 years), then it would be pretty weird for the Treasury to say the system currently had a $7.7 trillion PDV in unfunded liabilities, right? I mean, whatever happens out beyond year 75 is already being heavily discounted, so that’s presumably not serving as the lion’s share of the $7.7 trillion figure.

I always get a little tripped up with these types of things. For example, is Rosnick’s error (assuming it is one) dependent on the relation between the discount rate we use in these calculations, versus the average growth rate of real GDP? Discuss.

8 Responses to “Two Men Enter, One Man Leaves: Murphy vs. Rosnick on Social Security”

  1. Bruce Krasting says:

    If you wanted to “fix’ it today the number is (I believe) $5+ Trillion. That is the check that one would have to write to address the problem that will occur over the next 75 years.

    We don’t have to write that check. We don’t have the money to do that. There is no need to fix ss today.

    We have to fix it every year for the next 75. The future value of the fix will be much higher than the 5T. But we will bleed it every year forever.

    A better question to ask is, Should we fix it? If we do we should change the entire system? This one is going to kill us.

  2. Dan says:

    It seems like he must have just said the total shortfall was x and divided by 75. I bet he came up with his idea of the total debt and came out with $70 billion, then said see the Bush tax cuts easily pay for this over the next two years. I think he was trying to make the total debt seem small when you think about it per year and then used the tax cuts as an example of easy solutions that could be done to take care of the shortfall in the future. I really doubt he is going as deep with it as you are. He is pulling the Spicoli “Relax, all right? My old man is a television repairman, he’s got this ultimate set of tools. I can fix it.” line on you.

  3. English Bob says:

    Are you unaware—or worse, pretending not to know—that items have to be fed into a shredder, not just tipped in there?
    Do you see how your entire argument falls to pieces now?

  4. Bryan Rosander says:

    My summary of the main arguments:

    Robert Nosick – As long as the US government is paying its bonds to shareholders, it will also pay them to social security. If Warren Buffet and Bill Gates can count this as an asset, so can social security. It isn’t a gimmick because it really is an asset.

    Robert Murphy – The government has a trust fund which it claims will support Social Security. The trust fund doesn’t have any real assets, however, and is based entirely on IOUs supported by tax dollars. Therefore, if the government can’t match the IOUs, it needs to raise taxes and Social Security is really “insolvent”, even though it was the Government raising taxes, not Social Security.

    On this, I don’t think Nosick understands Murphy’s point that insolvency means raising taxes. I think he is treating it more like personal bankruptcy, where you become bankrupt when when you have actually stopped making consistent payments. I don’t know much about accounting, but if I was an investor in Social Security like any other investment property, I would be pulling out immediately, as would every other investor.

    That seems to be the real problem with Nosick’s arguments. Murphy defined his criteria for insolvency up front. I don’t think Nosick ever agreed to it or clearly set his own however.

    • bobmurphy says:

      Another crucial point is that even if we count the trust fund as a genuine stockpile of assets, SS is still insolvent.

      • Bryan Rosander says:

        His answer to that seems to be in this quote from the first essay and I don’t know what to make of it:

        “Then there is the question of the long-term shortfall in Social Security revenues. First, it bears noting that even if nothing is done and the trust fund runs out in 2039, the program will still pay 80–85 percent of promised benefits on the basis of payroll taxes alone. These future retirees will have been much more productive over their working lives and so are scheduled for much larger benefits than those received by today’s retirees. Even accounting for any projected shortfall, the lifetime benefits of those entering the workforce over the next ten years will be more than 46 percent larger than the lifetime benefits of current retirees.”

        What assumptions is he making here, are they valid, and is his math right?

        • bobmurphy says:

          His assumption is that if someone pays you 80 – 85% of what you are contractually owed, that it isn’t a default. That is not valid.

          • Bryan Rosander says:


            So, is he saying that Social Security defaults just a few years earlier, say, 2035, but still delivers payments at the same rate?

            I don’t get his point, “scheduled for much larger benefits than those received by today’s retirees.”

            Maybe I don’t understand Social Security, but I thought that everyone got paid at the same rate.