Wednesday, April 11, 2012

Is There Really An Aggregate Demand Problem?

Probably the most important debate on U.S. monetary policy is whether there really is an aggregate demand shortfall and, as a result, a negative output gap.  If the answer is yes, then the Fed should be doing more. If no, then monetary policy should show restraint.  Some Fed officials like St. Louis Fed President James Bullard and Minneapolis Fed President Narayana Kocherlakota believe there is no significant aggregate demand or domestic output gap problem and thus want to tighten monetary policy soon.  Fed chairman Ben Bernanke, on the other hand, believes there is still ample slack in resource markets and is open to further monetary stimulus.  And then there are the Market Monetarists and other like-minded commentators in the blogosphere who think there is a significant aggregate demand problems since the Fed has been effectively tight since mid-2008.  Who is right?

The answer has always been clear to me. But for those still unsure let me direct you to some evidence that makes a very convincing case that there has been and continues to be an serious aggregate demand shortfall.  The evidence comes from the NFIB's survey of Small Business Economic Trends.  This survey, among other things, provides a question to the small firms of this nation that is relevant to this debate: "What is the single most important problem facing your firm?"  There are nine answers firms can choose, but below I focus on just four of them and their recent trends:

The first thing to note from this figure is that  concerns about sales or demand became the most important problem.  It alone explodes in 2008.   Regulation, taxes, and labor quality (sorry  Kocherlakota , no labor market mismatch problems here) did not sharply intensify.  Labor quality concerns actually fall and though regulation eventually starts to gradually rise, concerns about demand were and continue to be the foremost among small firms.

There is more to the poor sales data.  It closely tracks the unemployment rate and the output gap as seen below.  These findings undermine Kocherlakota and Bullard's claims that the CBO's output gap is overstated.  For if the output gap were truly smaller and there were no serious aggregate demand shortfall, then one would not expect to see the relationships below.

This evidence in conjunction with that of downward wage rigidity excess money demand, and the Fed handling the housing recession just fine for two years should remove any doubt about there an aggregate demand problem.  The real debate is how best to respond to this problem.

P.S.  See Mike Konczal's earlier post using the NFIB data.


  1. David, the agents that populate a multisector RBC model, subject to a sectoral shock that propagates intersectorally, would answer in the same way as people do in these surveys.

    I discuss this in a bit more detail here:

    So, I do not think you have the slam dunk evidence you think you have.

  2. David:

    Yes, there is an identification problem with survey data, but when viewed with the other evidence to which I link, I think it is reasonable to take a strong stand on what the data are saying.

    If we view the problem from one of your favorite perspectives, the breakdown of commitment and thus safe assets, we can still tell an AD shortfall story. Monetary and fiscal policy failed to replace lost private safe assets with sufficient numbers of public safe assets. Since this shortage of safe assets amounts to a shortage of money-type assets we have an excess money demand problem and thus an insufficient aggregate demand problem too.

    A question: how do you reconcile that labor quality in the survey is not an issue for these firms? Doesn't this run contrary to labor market mismatch theories?

  3. David, thanks for your inspiring posts. Where did you get the NFIB time series for poor sales? I can only find it back to 2000.

    Thanks, Chris

  4. Sure, but why would I want to call that a breakdown in AD? Why not a breakdown in AS?

    We should try to be more precise about what the nature of the problem is, rather than attaching antiquated labels to its symptoms.

    And as for the labor market survey, there is a lot of other anecdotal evidence that suggests otherwise. I talk about some of it here:

  5. With regard to sticky wages--some occupations have less sticky wages than others and it should be expected that those with the least sticky wages should be among the first to recover.

  6. David:

    Yes, the breakdown in commitment and private safe asset production can be viewed a real shock. But it could also be viewed as AD shock in the sense that the Fed failed to keep NGDP expectations stable in 2008 leading up to financial crash later that year. Had they kept NGDP expectations better anchored it is less likely the commitment shock would have occurred.

    The same point can be made about the Great Depression. Had the Fed been more aggressive early on and kept NGDP expectations stabilized the "real shocks" to financial intermediation (i.e. the banking panics) would have been far less. FDR's recovery of 1933-1936 was sparked by doing that just that. He talked up restoring the price level to its pre-crash level and then backed up the talk by devaluing the gold content of the dollar.

  7. Chris,

    I had to contact the folks at NFIB. They give it out upon request. If you contact me at the email address below I will be glad to forward you my copy of the data.

  8. "....the agents that populate a multisector RBC model, subject to a sectoral shock that propagates intersectorally, would answer in the same way as people do in these surveys."


    Real people running real businesses should be disregarded as to thier views on the economy , because our supply-side model says so. Virtual RBC people are smarter than the flesh-and-blood kind.

    What was the "sectoral shock" that elicited the following mistaken belief in a shortfall in demand ? :

    -- Weak consumer demand, rising labor costs and high fuel costs top CFOs’ lists of concerns. They
    say that additional capital spending and hiring cuts will follow if consumer demand weakens

    -- The level of pessimism about the U.S. economy is the highest in more than five years.

    Source : September 2006 ( !! ) Duke University/CFO Magazine Business Outlook survey

    CFO's of the largest firms in the country were aware of the artificially inflated demand associated with the housing bubble , years before it popped. That's why the recovery from the 2001 recession was so weak re: domestic employment and investment growth.

    Here's the CFO results going back to 1996 :

    Look them over. Consumer demand was either a nagging or a dominant concern throughout the 2000's , because CFO's knew the debt bubble-sustained demand couldn't last.

  9. Superb blogging.

    It's the money stupid. Not enough of it out there.

  10. David,

    Do you have (or can you send me) a paper that lays out a mathematical model which supports your proposition? I am just curious to know the theoretical basis for the strong statement you are making.


  11. David,

    Try this article for starters:

  12. David,

    Imagine the Fed had been following a NGDP level target prior to the financial crisis and it was a very credible policy (i.e. the public believed the Fed would do whatever was necessary to ensure the NGDP level target was hit). If this were in place, how could a systemic commitment shock ever emerge?

  13. David,

    OK, I have read the paper by Koenig. Good theoretical paper, couched within the context of a 2-period model.

    Now, let's get serious.

    What is the empirical evidence supporting the degree of nominal rigidity that would imply NGDP targeting right now would help get us out of this "slump?"

    I can see how such a policy might help dampen future recessions. But we are now 4 years out...there are presumably a new set of nominal contracts written up under a different set of price level expectations...people renegotiate...etc.

    Note: I am not saying you are wrong in any of this. I am asking you for the evidence that supports your very strong statements.