Friday, February 13, 2009

My Economic Model -- Fall 2008 and Now

After four or five months, now is a good time to take stock of my economic model. I see four important areas to consider:

  1. employment and labor usage
  2. productivity, GDP, and spending
  3. investment, and its decomposition between residential and non-residential
  4. the impotence of public policy

I have significantly revised my thinking on employment, but in the other three areas it seems my model does quite well.

EMPLOYMENT. I initially thought that the adverse wealth effects experienced in 2007 and 2008 would soon induce people to work more. I still believe that will happen, but not in the next couple of months. I started to change my thinking a bit already in November when I realized that mortgage modification was in essence paying people lots of money not to work. Other bad incentives like this have come along. I have not yet done the hard work required to quantify the impact of these incentives, but at this point I don't see how that could reduce employment by 3 or 4 million. So I have more to do here if I am going to be able to accurately predict when the employment situation turns around. My only consolation is that most economists and commentators are still barking up the wrong tree -- that spending and a credit crunch are the root problems.

PRODUCTIVITY. Whatever is happening to employment, I continue to believe that productivity will advance. And it has. So, even if employment continues down, GDP and spending will not be dragged down that much. Thus, I am more confident than ever with my prediction that real GDP will stay above $11 trillion. There is still a good chance that, when this recession is over, it will be considered mild by GDP standards. It is impossible to have a second Great Depression with advancing productivity.

INVESTMENT. I predicted earlier that residential investment would continue to decline. It has. I am staying with that prediction through May or June. I predicted that nonresidential investment will increase. It has for most of this recession, but not for Q4. So maybe this category is more like other recessions than I thought. But still not bad, let alone disastrous. Going forward, I predict that non-residential investment will do fine, and will be one of the first activities to improve when we move out of this recession.

PUBLIC POLICY IMPOTENCE. I was right that the bank bailout would no nothing but disappoint and embarrass the Congressmen who voted for it. The "fiscal stimulus" will be more wasteful, although probably less embarrassing politically because it has so many disparate pieces. Most of the pieces will eventually be acknowledged as wastes. But by (blind squirrel) luck a couple of small pieces will look pretty good, and those will get the spotlight from our politicians.

In summary, I am thinking about the neoclassical growth model with a labor market distortion, but close to efficiency on all of the other margins. For the reasons cited above, I am now giving the labor market distortion piece more emphasis than I did last fall.


Low spending is a symptom in this recession, not a cause. The fundamental problem is somewhere in the labor market (I don't yet know exactly what, although mortgage modification is part of it). That is why total spending is doing much better than employment and hours.


[update: Donald P. Morgan left a comment with links to a lot of interesting results. I haven't digested them yet, but I highlight them so readers can take a look themselves.]


1 comment:

Candide said...

Professor Mulligan,

I've been enjoying your blog lately. Your thesis that the recession reflects a labor crunch more than a credit crunch is interesting. I agree that "earnings" from qualifying for mortgage modifications might contract labor supply.

Perhaps another disincentive is the Bankrutpcy Abuse Prevention and Consumer Protection Act of 2005. BAPCPA reduces the supply of bankrutpcy protection, particularly for filers with better than median (for their state) means. The means test determines who gets to file Ch. 7 and have their unsecured debt (including credit cards) discharged, and how much filers under Ch. 13 owe unsecured creditors. The means test effectively taxes labor effort by borrowers on the verge of bankrutpcy.

Here are three little bits of evidence in support of the hypothesis that BAPCPA discourages labor supply by subprime (i.e., prone to bankruptcy) borrowers/workers.
1) students in my special project class at NYU estimated diff-in-diff regressions showing that unemployment increased more, post BAPCPA, in states with higher exemptions (where bankruptcy "demand" is higher).

2) initial claims for unemployment surged just before BAPCPA took effect in Oct. 15, 2005. Part of the surge reflects Katrina, but not all of it, as the surge more than reversed right after BAPCPA (the katrina effects were presumably a bit more persisent).
Coincidentally (or is it?), the surge to file unemployment coincided very closely with the surge in Ch. 7 filings just before BAPCPA took effect. Co-authors and I show the rush to file Ch. 7 was higher in high exemptions states.
http://www.newyorkfed.org/
research/staff_reports/sr279.pdf

3) Co-authors and I also have a paper showing the surge in foreclosures was higher in high exemptions states, and we explain why that is predictable given how BACPA made credit card debt harder to escape under bankrutpcy
http://www.newyorkfed.org/
research/staff_reports/sr358.html

Actually, I am not sure how 3) relates to the labor contraction hypothesis. We call the paper in 3) "Seismic Effects of the Bankruptcy Reform" and your hypothesis, and tidbits 1) and 2) make me wonder if another of the seismic effecs was labor supply contraction by subprime borrowers.

I plan to test of the surve in unemployment claims varied by state bankrutpcy exemptions. If so, it would surely support your thesis. I will let you know the results

--Don