S&P’s Error

Paul Krugman explains:  I Heard It Through The Baseline.

Short version:  S&P used the wrong baseline when making its calculations.  Slightly longer explanation at the Krugman link above.  An even long one via the Treasury:   Just the Facts: S&P’s $2 Trillion Mistake, which includes:

S&P has said their decision to downgrade the U.S. was based in part on the fact that the Budget Control Act, which will reduce projected deficits by more than $2 trillion over the next 10 years, fell short of their $4 trillion expectation for deficit reduction. Clearly, in that context, S&P considers a $2 trillion change to projected deficits to be very significant. Yet, although S&P’s math error understated the deficit reduction in the Budget Control Act by $2 trillion, they found this same sum insignificant in this instance.

Back to Krugman:

So what we just saw was amateur hour. And these people are pronouncing on US credit-worthiness?

FILED UNDER: Deficit and Debt, Quick Takes, US Politics
Steven L. Taylor
About Steven L. Taylor
Steven L. Taylor is a Professor of Political Science and a College of Arts and Sciences Dean. His main areas of expertise include parties, elections, and the institutional design of democracies. His most recent book is the co-authored A Different Democracy: American Government in a 31-Country Perspective. He earned his Ph.D. from the University of Texas and his BA from the University of California, Irvine. He has been blogging since 2003 (originally at the now defunct Poliblog). Follow Steven on Twitter


  1. kevinbsnyder says:

    CBO also said several months ago the White House underestimated the debt by $2.7 Trillion …. Paul Krugman also says the “stimulus” was way too small …. Krugman is not credible, his Noble is worth the same as Obama’s ….

  2. @kevinbsnyder: But, of course, none of that is actually relevant to the question of whether S&P made a mistake or not.

    It is problematic to claim that one does not like information because of where the information came from. Either the information is correct or it isn’t, yes?

  3. Tano says:


    Your comment seems to be a nonsequitor. How does Krugman saying (consistently, from the beginning) that the stimulus was too small mean that he is not credible?

    Clearly the stimulus had an effect in terms of stopping the decline, but not nearly enough of an effect to kickstart a robust recovery. Exactly what you might expect if Krugman were right all along.

  4. Dean says:


    It will be interesting to see if McGraw-Hill’s stock takes a hit on Monday since one of their key business units has shown to the world it continues to be inept. And, the ineptness is not the downgrade, but making a $2 trillion error. I guess helping to cause the financial crisis by incorrectly rating bad mortgages wasn’t enough for the geniuses at Standards & Poor’s

    By the way, McGraw-Hill, on their website, is already trying to distance itself from the S&P decision.

  5. PD Shaw says:

    As Senator Dirksen probably never actually said: A trillion here; a trillion there, pretty soon you’re talking real money.

    Just a reminder that government monopolies private enterprise makes mistakes too.

  6. john personna says:

    This is from your treasury link:

    Specifically, CBO calculated that the Budget Control Act, including its discretionary caps, would save $2.1 trillion relative to a “baseline” in which current discretionary funding levels grow with inflation.

    S&P incorrectly added that same $2.1 trillion in deficit reduction to an entirely different “baseline” where discretionary funding levels grow with nominal GDP over the next 10 years. Relative to this alternative “baseline,” the Budget Control Act will save more than $4 trillion over ten years – or over $2 trillion more than S&P calculated. (The baseline in which discretionary spending grows with nominal GDP is substantially higher because CBO assumes that nominal GDP grows by just under 5 percent a year on average, while inflation is around 2.5 percent a year on average.

    That is a tricky couple paragraphs to unwind, but I think in essence they are both making guesses about “discretionary” spending over 10 years. This even though we don’t know what path such spending will take. The Budget Control Act of 2011:

    …. specifies an incentive for Congress to act. If Congress fails to produce a deficit reduction bill with at least $1.2 trillion in cuts, then Congress can grant a $1.2 trillion increase in the debt ceiling but this would trigger across-the-board cuts (“sequestration”) of spending equally split between defense and non-defense programs. The across-the-board cuts would apply to mandatory and discretionary spending in the years 2013 to 2021 and be in an amount equal to the difference between $1.2 trillion and the amount of deficit reduction enacted from the joint committee. The sequestration mechanism is the same as the Balanced Budget Act of 1997. There are exemptions—across the board cuts would apply to Medicare providers, but not to Social Security, Medicaid, Medicare beneficiaries, civil and military employee pay, or veterans.

    So what are they both doing, assuming the worst, and the automatic cuts, and that all the way out to 2021 congress will let sequestration determine discretionary spending?

    I find it hard to believe either side is able to calculate what really happens if or when the committee delivers their report. Stepping back to political uncertainty is reasonable.

  7. john personna says:

    The strength of CBO analysis is that it can compare two plans, say gas tax up 10 cents versus marginal rate up 1 percent, and tell you “all else being the same” the net difference. A or B.

    That’s what economics can do for you, really.

    The thing it can’t do is tell the future. It can’t tell you “given plan A, will we really get plan A?”

  8. @john personna:

    Specifically, CBO calculated that the Budget Control Act, including its discretionary caps, would save $2.1 trillion relative to a “baseline” in which current discretionary funding levels grow with inflation.

    S&P incorrectly added that same $2.1 trillion in deficit reduction to an entirely different “baseline” where discretionary funding levels grow with nominal GDP over the next 10 years.

    So the Treasury assumed discretionary spendingwould grows with inflation and S&P assumed it would grow with GDP. Both of those seem like completely reasonable models. So again, we have a legitimate difference in opinion over economic models being treated as an error.

    This is just the Treasury trying to spin things in it’s favor.

  9. john personna says:

    @Stormy Dragon:

    So the Treasury assumed discretionary spendingwould grows with inflation and S&P assumed it would grow with GDP. Both of those seem like completely reasonable models. So again, we have a legitimate difference in opinion over economic models being treated as an error.

    If they are both assuming sequestration (and not the best case committee-based cuts) it may work another way. Sequestration itself may dictate a baseline(?)

    [Update: a lot rides on sequestration really working, or alternately on the committee producing real, non-BS cuts]

  10. reality slave says:

    I didn’t like the design..which I told the engineers. It looked vulnerable to me. I was asked to provide an analysis, which due to its complexity I attempted. Turned out during the analysis I used the wrong baseline assumption for one of the formulas, which was caught by one of the engineers and rapidly my concerns were dismissed. One we went into development.

    Later, when we got it to the lab, under test the device went up in loud smoke and smelly fire.

    I turned to the engineers and said: “I still don’t like the design”.

    I suspect S&P’s main fault, after receiving blame for not convincing people of its accuracy by making a math error, will be that its downgrade warning, like mine, will turn out to be too understated.

  11. reality slave says:

    “So the Treasury assumed discretionary spending would grow with inflation and S&P assumed it would grow with GDP.”

    Historically, growth in discretionary spending exceeds both of those indices, so use of either baseline in growth assumption is flawed..

  12. john personna says:

    @reality slave:

    Excellent parable.

  13. @john personna:

    Looking at the actual law, not only is the baseline not specified, but it’s the OMB (not the CBO) who decides whether the baseline is being exceeded AND congress is free to change the baseline whenever they like:

    22 LIMITS.—
    23 ‘‘(1) CONCEPTS AND DEFINITIONS.—When the
    24 President submits the budget under section 1105 of
    25 title 31, United States Code, OMB shall calculate
    1 and the budget shall include adjustments to discre
    2 tionary spending limits (and those limits as cumula
    3 tively adjusted) for the budget year and each out
    4 year to reflect changes in concepts and definitions.
    5 Such changes shall equal the baseline levels of new
    6 budget authority and outlays using up-to-date con
    7 cepts and definitions, minus those levels using the
    8 concepts and definitions in effect before such
    9 changes. Such changes may only be made after con
    10 sultation with the Committees on Appropriations
    11 and the Budget of the House of Representatives and
    12 the Senate, and that consultation shall include writ
    13 ten communication to such committees that affords
    14 such committees the opportunity to comment before
    15 official action is taken with respect to such changes.

  14. @Stormy Dragon:

    Let me check that, the CBO determines if the baseline is being exceeded, but the OMB determines what the baseline is. The OMB also gets to dispute the CBO report, and if it does, congress gets to decide what the truth is “by consultation”.

  15. JohnMcC says:

    I have gone as far into the weeds as my brain will allow, and you fellows are going farther than I can. But if I may, I think S&P are making a specifically political, not an economic, judgement. The related thread on OTB begun by James Joyner directed me to Felix Salmon’s defense of S&P which seems compelling. I read between the lines that Krugman, Salmon and S&P are all expecting political intransigence to become a new normal. It’s what Mitch McConnell pretty specifically told Neil Cavuto: “…no president is going to be able to get the debt ceiling increased without a re-ignition of the same discussion…so we’ll be doing it all over.” (As quoted on ThinkProgress)

    And FWIW, S&P has pretty much destroyed the notion that we could fail to increase the debt ceiling, then use income to pay our bonded obligations on T-Bills, and pretend that’s not a default. Which was a pretty common theme from the TeaParty. S&P is saying BULLSH#T.

  16. john personna says:


    I can kind of see S&P saying, “OK now that I understand your math and your assumptions, I REALLY hate it.”

    You have to go into the weeds to make sense of it. This is not simple accounting of currently legislated spending. This revolves on future unnamed cuts, or alternately that preposterous idea that congress would endure sequestration in the years 2013 to 2021 without changing the law and unburdening itself.

    2013 to 2021. Seriously.

  17. A voice from another precinct says:

    @Steven L. Taylor: If they can’t complain about the messenger, they will have nothing to say at all. This way what they have to say is not useful (except to their own biases) but preserves their “right to an opinion.” You are really expecting too much from the commentariat if you are expecting them to always be germane.

  18. A voice from another precinct says:

    @Tano: In the same way as Dr. Taylor, you don’t get it. Krugman is a liberal, therefore he is always wrong. Nothing that he says is the truth, nothing that he says is accurate, all of his prescriptions are wrong–even if others agree with him,. Even if it turns out that he is right–he is still wrong. He is incapable of getting anything right because he is a liberal and therefore, he is wrong.

    Note to kevinbsnyder: did I explain that for you ok, or should I have put in a few more “he is wrong”s?

  19. Ben Wolf says:

    The reason for the downgrade is irrelevant, but Krugman is right: S&P’s actions reveal a startling ignorance of monetary theory and a total lack of understanding anout what sovereign debt is. Unfortunately the popular wisdom repeats these fallacies about nine thousand times per day,including just about every post on the issue here at OTB.

    Sovereign nations by definition cannot run any risk of default unless they choose to. The debt ceiling battle was pure theatre.

  20. Ben Wolf says:

    By the way: S&P continues to be a totally corrupt clusterf*ck:


  21. @Ben Wolf:

    Sovereign nations by definition cannot run any risk of default unless they choose to. The debt ceiling battle was pure theatre.

    Well, I assume you mean that the US could have paid its interest payments (which is true) or it could have printed more money. However, even if that was done, there were the fiscal policy implications–i.e., an immediate reduction in government spending in the 40%+ range that got conflated into the term “default” in a lot of there conversations.

    And I would note that Mexico and Argentina (and other LatAm states) did a pretty good impression of defaulting in the early 1980s.

  22. john personna says:
  23. Ben Wolf says:

    @john personna: This is obtuse, John. And you’re a smart man. Your argument is a classic appeal to authority, a logical fallacy. What the Chinese “say” is irrelevant. Their ratings agency is entirely a political organization. It also happens to clash with millenia of empirical evidence that sovereign nations in control of their own monetary policy do not default unless they choose to default.

    The U.S. can retire any amount of debt at any time without shorting anyone. This is basic monetary theory. I dn’t care that everyone is pretending the United States is just like a corporation about to go into bankruptcy. It isn’t and it will never go into bankruptcy. It is not possible.

    The ratings agencies do not understand monetary theory. The Republicans do not understand monetary theory. Barrack Obama doesn’t understand it either. Anyone blathering on about how hard it will be for the U.S. to pay down its debt is revealing to you a stunning level of ignorance.

    Start here for a very basic understanding of why a sovereign entity does not default:

  24. I recommend downgrading S&P’s rating credibility to “junk” status…

  25. Additional Thoughts says:

    @Ben Wolf: You have commented on the fallacy “appeal to authority” contained in the link provided by “John Personna”, however, your argument in this regard is in itself a fallacy “appeal to motive”. You assert that because the Chinese have their own motivation, their reasoning cannot be true.

    I agree with you that sovereign nations only default by choice, and that nations cannot go bankrupt (by definition).

    However, to argue the semantics of default and bankruptcy while spewing out assumptions about the financial intelligence of Republicans and Barack Obama is completely unwarranted. Why not spew out the same about all the news agencies reporting about the US defaulting — Reuters, CNN, Fox, NBC, CBS, etc.? The real question is whether the being a creditor to the US goverment is a good idea.

    Before really evaluating S&Ps ratings, one must look at what they are intended to do. S&P’s ratings on credit risk are designed to evaluated whether or not an entity (issuer) is willing and able to “meet their financial obligations in full and on time”.

    You’re right — Any nation with boundless control of its own money supply can prevent default on any obligations requiring that type of money. However, it is incorrect to assume that all nations who print their own money are on an equal footing in regards to paying their obligations. If they were, all such nations would receive AAA ratings.

    The other part of the equation is still important: willingness to pay. In the US (and other countries), rapid devaluation of the currency (inflation) is not acceptable. As printing money can cause this condition, there is still a political tradeoff between paying debts on time and printing money. Because of this, countries that spend vastly more money than they acquire should and will have their credit ratings lowered.