Wednesday, March 12, 2014

The President's Economic Report and the Right Counterfactual

A common mistakes observers make when assessing economic policy is that they fail to do the counterfactual of no policy. That is, they fail to consider what the economy would have been like in the absence of the policy. This often happens in critiques of monetary policy. It also happens with critiques of fiscal policy. The new Economic Report of President acknowledges this point:
Evaluating effects of fiscal policy in general, and the Recovery Act in particular, is challenging for several reasons... A key issue is that estimating effects entails comparing what actually happened with what might have happened (what economists call the “counterfactual”)[p.105-106].
The White House believes the proper counterfactual for the American Recovery and Reinvestment Act of 2009 is to imagine how much worse off the economy would be in its absence. I am not sure, though, that this is the right counterfactual. For we saw in 2013 that Fed policy offset to some extent the tightening of fiscal policy. This suggests that it might have done more in 2009 had there been no fiscal stimulus. Ramesh Ponnuru, presents this view in a recent Bloomberg article:
To the extent that the central bank has a target for inflation (or nominal spending), and has the power to hit that target, the Fed constrains the power of fiscal policy. If Congress tries to stimulate the economy during a slump, for example, the Fed will offset that stimulus by loosening money less. Some of this offsetting will actually be automatic, based on market expectations that the Fed will stay on target. It's likely that if Congress hadn't enacted a large stimulus, the Fed would have done more quantitative easing early on, lowered interest on reserves or taken some other expansionary step.
This is the 'monetary offset' view that has been made by Scott Sumner. It not really a new argument, but the recent emphasis has been that the Fed will still offset fiscal policy even at the zero lower bound and in a slump if it is committed to its target. It may not do so perfectly, but if the Fed chooses it can still dominate fiscal policy at business cycle horizons. This counterfactual thinking is contentious. John Aziz, for example, responded to the Ponnuru piece with this:
Ponnuru's argument isn't that the stimulus caused the Fed to tighten, but that it would have been looser had there been no fiscal stimulus... The hypothesis that the fiscal stimulus limited the Fed in any way is just not supported by the real world record of what the Fed did... [I]n a huge, once-in-a-generation slump, offsetting expansionary fiscal policy just isn't on the agenda.
Aziz, like many, find it incredulous that the Fed would hold back in 2009 because of the fiscal stimulus. Fed officials would never admit doing this directly, but they would  admit to responding to changes in inflation and employment. Consequently, if fiscal policy influenced inflation or employment it indirectly allowed the Fed to hold back some of its firepower. For a believer in fiscal stimulus like Aziz, the logic is inescapable if the Fed is truly committed to its objectives.

This begs the question of how committed was the Fed to it target?  Officially, the Fed has been a flexible inflation targeter (FIT) with a 2% personal consumption expenditure (PCE) inflation target since January, 2012. Many, including former Fed chair Bernanke, argue it has been a FIT for much longer. Many Fed officials prefer the core PCE inflation measure since it is a better indicator of underlying inflation trends. The question, then, is how committed has the Fed been to its inflation target? Has it been so committed that it was slow to go full throttle in 2009 because of fiscal stimulus?

Below is a figures that sheds some light on this question. It shows the timing of the Fed's QE programs and changes in the core PCE inflation rate. The figure indicates several possibilities. First, the FOMC runs QE programs when core inflation is under 2% and falling. It also suggest that Fed has had an effective 1%-2% target range for core inflation over the past five years. Had there been no fiscal stimulus and had this mattered to core inflation, this figure suggests that the Fed may have done more sooner.

So contrary to Aziz's claim, Ramesh's counterfactual seems perfectly reasonable. Over the past five years the Fed seemingly was committed to targeting a core inflation range of 1%-2%. Consequently, anything that affected core inflation, including fiscal stimulus, also indirectly affected the response of monetary policy.


  1. Clearly, fiscal stimulus did not prevent core inflation starting to fall. And even if it meant that core inflation fell less than it would have done otherwise, I'm not sure the Fed knew a) how big the fiscal effect was going to be or b) how much less activist it would have to be to offset any effect.

    Deling with a slump of unknown but potentially massive proportion, it doesn't surprise me that Ben Bernanke has been looking for fiscal help. If the change in prices otherwise could be -5% or -3%, the Fed is going to have to do a hell of a lot of work to get inflation to 2%. Bernanke has time and again made it clear he is happy to let fiscal policy do some of the lifting. Potential effects of fiscal policy will be a consideration, but it is much less bad to overshoot to 5% and have to pull back than undershoot to -1% — giving debt deflation much more of a chance to set in — due to trying to offset a fiscal expansion.

    Given that the Fed's balance sheet monetary policy was experimental (and thus in Bernanke's eyes risky) anyway, the Fed was going to be a little cautious however loose or tight fiscal policy was. Would QE2 and QE3 come quicker and bigger if Calvin Coolidge and Grover Cleveland were running the Treasury department? Let's assume that they would have. Would that have been enough to keep the economy out of deflation? Possibly not. Direct government purchases have a very direct transmission mechanism for getting money into people's pockets. Quantitative easing relies on quite a nebulous transmission mechanism. That has been helpful in the wake of the fiscal stimulus and recession ending, but we don't know how quickly or slowly the economy would have turned around if monetary policy was the only option available.

    1. I agree that QE involves a “nebulous transmission mechanism”. In addition, QE and indeed monetary policy as a whole, is distortionary. That is it feeds stimulus into the economy just via investment. At least it has that effect if monetary policy actually influences investment, and the evidence seems to be that interest rate adjustments at least have little effect on investment. See:


      Re “distortion” feeding stimulus into the economy just via investment, that makes as much sense as feeding stimulus into the economy only via car production and restaurant meals, which would be equally distortionary.

  2. What if the Fed's inflation target can be superceded by other targets ? What if , unlike most academic economists , they understand that debt/gdp can't grow without limit , and that high debt burdens slow growth ? Then they might like to see , at worst , a stable level of debt/gdp or , at best , a gradually declining burden. It sure looks like someone is paying attention to this :

    Blue is total nonfinancial debt/gdp , red is private only ( the graph label at the top is incorrect ). The private sector was actively deleveraging until ~ mid-2012. Total debt/gdp has been declining very modestly until about the same time , and may be showing signs of increasing again ( the trigger for tapering ? ).

    If fiscal stim contributed to private sector deleveraging , then you can make the argument that it should have continued. If overall debt/gdp is going to be held constant regardless , as appears to be the case , better to have the gov't bear an increasing share of the debt burden than the private sector until the private sector has a debt load that will not impact growth negatively. After all , the private sector can't choose the interest rate it pays on debt , like the Feds can.


    1. I do like the idea that “most academic economists” think that the debt/GDP can “grow without limit”. Can you quote me a single “academic economist” who as argued that the debt/GDP might rise to, or should be allowed to rise to 1,000%, never mind “without limit”?

      Re your phrase “high debt burdens slow growth” I assume you are trying to say that high debt reduces growth. As you doubtless know, the main proponents of that are Rogoff and Reinhart, and their work has been debunked.

    2. If you noticed I was mainly talking about private-sector debt , and you seem to be just one of the economists I was talking about. With the R&R blunder as your proof , you seem to have concluded that high public - or private - debt burdens won't impede growth , and that has been by far the most common attitude. Post-crisis more seem to be coming around , but very few had concerns before the crisis , and those that did were obsessed with public sector debt.

      Determining the levels where sectoral debt/gdp becomes problematic is a work in progress , but it's only recently that most of that work has begun. That's a failure on the part of the profession. The fact that some kind of guidance on this issue cannot yet be offered to policymakers is a travesty.

      The reason the R&R study , and their spreadsheet error , attracted so much attention was because it was on this uncommon topic , and in spite of the error I give them credit for being ahead of the pack they were ahead of the pack. Empirical , cross-country studies on debt vs growth were just not considered sexy by economists , I guess.

      As to naming an academic who dismisses the whole notion of debt burdens making a difference - how about Sumner ? He's famous for debt-denialism. Here's one of his posts I found with a quick Google search , but I'm sure you can find more :

      “Are you saying that high rates of debt don’t limit demand and can’t cause down turns (whatever causes them ) to be more severe ?”

      SS :Yes, I am saying if the Fed stabilizes NGDP, debt fluctuations don’t matter.


    3. 1. Sumner doesn’t refer to debt rising “without limit” which is what you referred to. He refers to it growing to 100% of GDP.

      2. You ask, “Are you saying that high rates of debt don’t limit demand and can’t cause down turns?” My answer is that there is no reason they should “limit demand” and I notice you don’t give any actual reasons. But that’s not surprising: no one has produced any convincing reasons as to why high debt should limit demand, far as I know.

      As to CHANGES in the amount of debt, that’s different. I agree with the conventional view that an EXPANSION in private debt will boost demand, while a decline will cut demand. The latter “cut demand” point is just a re-statement of Irving Fisher’s “debt deflation” idea.

      As to a stable and high level of public debt, I totally fail to see the problem, as long as the real or inflation adjusted rate of interest paid is about zero. The latter rate means the debt is not a real burden on taxpayers.

      And if creditors start demanding a higher rate, that’s no problem: just print money and pay them off (which is what we’ve been doing on an unprecedented scale via QE recently). And if there is any inflationary effect from that, then just raise taxes so as to counteract the inflationary effect.
      It’s all one huge non-problem.

      Unfortunately, my above points will be way beyond the comprehension of Rogoff, Reinhart and several other individuals at the Harvard economics department, who I regard as incompetent not just because of their spreadsheet error.

      Coincidentally I set out one of Rogoff’s bits of incompetence in a post I put online earlier today. See:

      But if you want more on his incompetence, then Google his name and the name of my blog.

  3. "My answer is that there is no reason they should “limit demand” and I notice you don’t give any actual reasons. But that’s not surprising: no one has produced any convincing reasons as to why high debt should limit demand, far as I know."

    I think that's very representative of the thinking in the academic community , as I said in my first comment here.

    Should we be concerned if we go from getting , say , $2 of additional ngdp from $1 of additional nominal debt , to getting only $1 from $1 ? How about getting only $.10 from $1 ? A penny from $1 ? What if we only got a hundredth of a penny's worth of additional ngdp from each additional $1 of nominal debt ?