An empirical – not a theoretical – disagreement with George, Larry and Eli

Last week George Selgin warned us (the Market Monetarists) about getting to excited about the recent actions of the Federal Reserve. Now fellow Free Banker Larry White raises a similar critique in a post on

Here is Larry:

“While saluting Sumner 2009…I favor an alternative view of 2012: the weak recovery today has more to do with difficulties of real adjustment. The nominal-problems-only diagnosis ignores real malinvestments during the housing boom that have permanently lowered our potential real GDP path. It also ignores the possibility that the “natural” rate of unemployment has been hiked by the extension of unemployment benefits. And it ignores the depressing effect of increased regime uncertainty.”

Larry’s point is certainly valid and Bill Woolsey has expressed a similar view:

“Targeting real variables is a potential disaster.  Expansionary monetary policy seeking an unfeasible  target for unemployment was the key error that generated the Great Inflation of the Seventies.  Employment or the employment/population ratio could have the same disastrous result.”

I have already in an earlier post addressed these issues. I agree with Bill (and George Selgin) that it potentially could be a disaster for central banks to target real variables and that is also why I think that an NGDP level target is much preferable to the rule the fed now seems to try to implementing. Both Larry and George think that the continued weak real GDP growth and high unemployment in the US might to a large extent be a result of supply side problems rather than as a result of demand side problems. Eli Dourado makes a similar point in a recent thoughtful blog post. Bill Woolsey has a good reply to Eli.

To me our disagreement is not theoretical – the disagreement is empirical. I fully agree that it is hard to separate supply shocks from demand shocks and that is exactly why central banks should not target real variable. However, the question is now how big the risk isthat the fed is likely to ease monetary policy excessively at the moment.

In my view it is hard to find much evidence that there has been a major supply shock to the US economy. Had there been a negative supply shock then one would have expected inflation to have increased and one would certainly not have expected wage growth to slow. The fact is that both wage growth and inflation have slowed significantly over the past four years. This is also what the markets are telling us – just look at long-term bond yields. I would not argue that there has not been a negative supply shock – I think there has been. For example higher minimum wages and increased regulation have likely reduced aggregate supply in the US economy, but in my view the negative demand shock is much more import. I am less inclined to the Austrian misallocation hypothesis as empirically significant.

A simple way to try to illustrate demand shocks versus supply shocks is to compare the development in real GDP with the development in prices. If the US primarily has been hit by a negative supply shock then we would have expect that real GDP to have dropped (relative to the pre-crisis trend) and prices should have increased (relative to the pre-crisis trend). On the other hand a negative demand shock will lead to a drop in both prices and real GDP (relative to pre-crisis trend).

The graph below shows the price level measured by the PCE core deflator – actual and the pre-crisis trend (log scale, 1993:1=100).

The next graph show the “price gap” which I define as the percentage difference between the actual price level and the pre-crisis trend.

Both graphs are clear – since the outbreak of the Great Recession in 2008 prices have grown slower than the pre-crisis trend (from 1993) and the price gap has therefore turned increasingly negative.

This in my view is a pretty clear indication that the demand shock “dominates” the possible supply shock.

Compare this with the early 1990s where prices grew faster than trend, while real GDP growth slowed. That was a clear negative supply shock.

That said, it is notable that the drop in prices (relative to the pre-crisis trend) has not been bigger when one compared it to how large the drop in real GDP has been, which could be an indication that White, Selgin and Dourado also have a point – there has probably also been some deterioration of the supply side of the US economy.

Monetary easing is still warranted

…but rules are more important than easing

I therefore think that monetary easing is still warranted in the US and I am not overly worried about the recent actions of the Federal Reserve will lead to bubbles or sharply higher inflation.

However, I have long stressed that I find it significantly more important that the fed introduce a proper rule based monetary policy – preferably a NGDP level target – than monetary policy is eased in a discretionary fashion.

Therefore, if I had the choice between significant discretionary monetary easing on the one hand and NGDP level targeting from the present level of NGDP (rather than from the pre-crisis trend level) I would certainly prefer the later. Nothing has been more harmful than the last four years of discretionary monetary policy in the US and the euro zone. To me the most important thing is that monetary policy is not distorting the workings of the price system and distort relative prices. Here I have been greatly inspired by Larry and George.

I have stressed similar points in numerous earlier posts:

NGDP level targeting – the true Free Market alternative (we try again)
NGDP targeting is not about ”stimulus”
NGDP targeting is not a Keynesian business cycle policy
Be right for the right reasons
Monetary policy can’t fix all problems
Boettke’s important Political Economy questions for Market Monetarists
NGDP level targeting – the true Free Market alternative
Lets concentrate on the policy framework
Boettke and Smith on why we are wasting our time
Scott Sumner and the Case against Currency Monopoly…or how to privatize the Fed

I think we Market Monetarists should be grateful to George, Larry and Eli for challenging us. We should never forget that targeting real variables is a very dangerous strategy for monetary policy and we should never put the need for “stimulus” over the need for a strictly ruled based monetary policy. And again I don’t think the disagreement is over theory or objectives, but rather over empirical issues.

As our disagreement primarily is empirical it would be interesting to hear what George, Larry and Eli think about the euro zone in this regard? Here it to me seem completely without question that the problem is nominal rather than real (even though the euro zone certainly has significant supply side problems, but they are unrelated to the crisis).


Update: David Beckworth and George Selgin joins/continue the debate.

Leave a comment


  1. Alex Salter

     /  September 24, 2012

    It’s not so much about supply vs. demand shocks, Lars, as it is about malinvestment. The argument against monetary stimulus says that market forces should be allowed to recalculate, deciding where best to put the resources that were misallocated during the boom. You might say that if 4+ years is long enough for wage and price expectations to get unstuck, then it ought to be long enough for recalculation and reallocation to take place. Maybe; I’m not sure myself if one implies the other. There’s also the question of regime uncertainty. If businesses are sitting on their cash because they’re worried about their future tax liabilities, for example, then additional monetary stimulus won’t accomplish all that much.

    • Alex,

      I most say I have a very hard time understanding why it would take 4+ years to get rid of the misallocation. The Austrian position that we are still in the “hangover phase” of the crisis to me makes little sense. Property prices and construction employment has long adjusted back to “pre-bubble” (if there was a bubble) levels. So it can only be an supply side argument.

      In terms of regime uncertainty I agree that there is “monetary regime uncertainty”. A clear rule based monetary framework will clearly reduce this uncertainty. I believe that the fed indeed has moved closer (but not far enough) to a rule based monetary policy after four years of discretionary failure. That is the important thing about the latest fed decision rather than the “stimulus”.

      See my early post on regime uncertainty here (a concept I love…):

  2. George Selgin

     /  September 25, 2012

    Lars, the sentiments you express here are very fine ones, whatever the empirics may say.

    My only quibble is with your statement, “I most say I have a very hard time understanding why it would take 4+ years to get rid of the misallocation.” Of course, the same may (and has) been said regarding the MM claim that a general excess demand for money has persisted for all those years. Both possibilities are no doubt hard to believe. But surely the structural misallocations toward which the malinvestment theory points, if indeed these have occurred, might be assumed to be harder to correct than mere money shortages, which might be corrected through mere abstinence from price and wage rate increases.

  3. Hallelujah! I thought that god had forsaken me.

    Lars, SUMNER AGREES. I have been accusing him of not admitting this publicly and privately and he’s been dead silent.

    Scott PREFERS make-up, tis true.

    but if he had to choose either or, he’d go with his baby because he KNOWS the rule is really what matters.

    The issue is that getting NGDPLT without makeup (or a throw away make-up) is easier to accomplish

    This puts you in direct contradiction with the DeKrugman “its a good way to get some more inflation” right now crowd.

    Here’s where it gets interesting:

    WHY would the DeKrugman crowd refuse this change to a rule based system, without make-up?

    IF the problem is AD, then very quickly NGDP plummets and we get QE4,5 etc.

    But IF the rule is in place, and we push past 4.5% now there is a policy choice: cut Federal Spending, and rates stay low.

    And maybe, just maybe, the adoption of the rule, convinces the market the Fed is out of the way, and less uncertainty leads to longer term commitments.

    Anyway, $10 at Applebees for whoever get Scott to admit he agrees with you in an either or choice.

  4. I wish I lived in America, because I’d love to try Applebees… Scott would admit that right away. In fact I’m pretty sure he already has, several times.

  5. Saturos, FIND IT stately explicitly. Without make-up, (which means rates go up soon) the large benefit of NGDPLT will occur, and the $10 is yours.

    Hint: you’ll know when it is stated explicitly because the DeKrugman crowd will go apeshit.

  6. Alex Salter

     /  September 26, 2012


    Reallocation is a two-step process. Step one is liquidation: Recovering as much as you can from the malinvested resources. Step two is redirection: Putting those resources back to where you think they ought to go. If you believe regime uncertainty is an important factor (and I certainly do), then it’s not a stretch to say that part one has been done for a while, but folks are holding off on part two. 4+ years of reallocation due to an incomplete step two doesn’t seem strange to me, really.


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: