Sunday, January 29, 2012

Did the Federal Reserve cause the Great Recession?

[Fifth in what was originally supposed to be a short series]


Let's take a break from arguing about breakfast cereal and return to the more civilized topic of the cause of the Great Recession.

I spent a couple of hours over the weekend digging into market monetarism, which turns out to be quite the rabbit hole. I'm not sure that I fully understand it, but let me try to paraphrase the part I think I get:

1. The Fed does have the means to increase the money supply beyond setting interest rates to zero, and hence to escape liquidity traps. One thing many people don't realize is that the Fed does not really "set" interest rates directly, it buys and sells government bonds and other financial instruments in an attempt to move the credit markets towards a target. If it buys, it drives price up and yields down and vice versa. Or something like that. I may have dropped a sign somewhere. But the point is, the Fed acts (or at least tries to act) like a closed-loop controller for some economic parameter. That parameter has historically been the discount rate, but that's an arbitrary choice. Market monetarists argue (if I understand them correctly) that this is the wrong parameter to try to control, and that the right parameter is nominal GDP.

2. Because the Fed is trying to control the wrong parameter, it believes (or wants the world to think that it believes) that it has reached the limits of its control authority (in the strict control-theoretic sense) because interest rates are near zero and they can't go any lower than that. But this is wrong because real interest rates can (and in this case should) go below zero (at least relative to inflation, which is the measure that really matters), which will discourage savings, encourage spending, and kick-start the economy. The Fed's failure to do this has prolonged (you would say created) the Great Recession.

It's a very appealing theory, but it has several problems. For starters, it assumes efficient markets, and they aren't, but let's leave that aside for now. Second, as you yourself pointed out earlier in this thread when you discussed the Laffer curve, economies are complicated things, and simple stories about them, intuitively appealing as they may be, are likely to be wrong merely by virtue of their simplicity. But let's leave that aside as well. After last night's reading binge I'm not sure how much more complexity I can handle.

The bigger problem with market monetarism (or at least the market-monetarist theory of the G.R.) is the same as the problem with the theory that higher taxes on rich people causes unemployment: it makes intuitive sense, but it is not in fact supported by the data. You write:

I'll agree with you, that the subprime residential housing crash was the original spark. But it's like living in a dry forest, while somebody at a campfire plays with matches, and the professional firefighters that you've hired for this job are standing watch. Yeah, one of the matches got dropped in a bush. But this kind of minor, industry-local crisis happens all the time. That's why we have a US Fed, so that when a bush starts to spark, the professional firefighters who are standing right there, put it out.

Only, in 2008, they didn't. And they still haven't, for the last four years.


Inflation is obviously a spectrum. For the last four years, it has been far too low, averaging 0-1% instead of 2-4%.

Inflation in 2008-2011 was 3.8, -0.4, 1.6, and 3.2% respectively (BLS inflation data). The average of those four numbers is 2.05, just inside your target range. (Contrast this with the period 1930-1933, when the Fed really did fan the flames and we had four years of deflation ranging from -2.3% to -9.9%.)

Or, if you want to compute "average inflation" the right way, taking the effects of compounding into account, then the average rate over the last four years has been just over 1%. The trouble with that approach is that compounding is very sensitive to very small changes earlier in the sequence. If, for example, the 2009 figure had been zero instead of -0.4, then the resulting average would be 2%, again within your target range. So really, the difference between what you say you want the Fed to achieve and what it actually did achieve is less than half a percentage point in a single year (one that happened to follow a catastrophic stock market crash). Considering how hard it is to steer an economy, I'd say that by your standards the Fed has been doing a pretty admirable job.

No matter how you slice it, the difference between what the numbers are and what you say they should be is very small, at least compared to what it was during the Great Depression. So there are two possibilities. The first is that the market monetarist account of the G.R. is simply wrong. The second is that it's right, and the apparent discrepancy between the naive linear version of the theory and the actual numbers is accounted for by the underlying complexities of the economy.

Economies are complicated, full of non-linearities (even non-monotonicities), time lags, and hysteresis. So maybe the economy is exquisitely sensitive to differences in inflation in the 1-2% range, and less so in the negative 5-10% range. Or maybe the sensitivity depends on the particular conditions at the time. Under normal circumstances 1% inflation would be OK, but the particular conditions set up by the sub-prime crash set up a situation where 1% was disastrous while 2% would have been OK. But that begs the question: what exactly are those conditions, and how can you know if they pertain except in hindsight? And if you concede that the economy can get into exquisitely sensitive control regimes, where tiny changes in a crucial parameter can result in enormous differences in outcome, then how can you possibly know what would have happened if the Fed overshot by a little instead of undershot by a little? You can't have it both ways. Either small difference matter, or they don't. You can't arbitrarily say that small differences matter in one direction but not the other. You have to have a theory that explains why (at least if you want to be considered a legitimate scientific theory). And market monetarism doesn't as far as I can tell.

The argument that sub-prime did not cause the G.R. is kind of like the argument that it's not the fall that kills you, it's the sudden stop at the end. The economy jumped off the cliff when the banks loaned vast amounts of money to non-credit-worthy borrowers and left investors holding the bag. Once they did that, massive defaults and the resulting ripple effects throughout the economy were inevitable, it was just a question of when. 4% inflation might have changed that outcome, but we don't know and we can't know because we didn't do that experiment. We were -- and still are -- in uncharted economic waters, and God only knows how the system will respond to control inputs at this point.

Franky, I dont see how even 4% inflation could have changed the course of events much. That would not have been nearly enough to cause wages to rise enough for all those formerly non-credit-worthy borrowers to suddenly be able to afford to pay their mortgages. So you still would have had a wave of defaults, AAA-rated bonds becoming worthless, rapid deleveraging from unwinding all the credit default swaps that suddenly came due, financial institutions going bankrupt but for government bailouts, the stock market crash, foreclosures, yadda yadda yadda.

The only way I see that inflation would have gotten us out this mess is if it had been high enough to essentially reset the principal balance of all those sub-prime mortgages to affordable levels before the interest rates reset. I haven't done the math, but that would probably have required inflation closer to 100% than 4%.

But let us leave even all that aside for now. Let us assume that the market monetarists are 100% correct, and that all that is needed to get us out of the recession is a little more quantitative easing. That begs the obvious question: why hasn't the Fed done it? I have a theory, but this post is already too long so I'll save it for the next installment.


Don Geddis said...

Tag. You're it.

Don Geddis said...

BTW, Ron, you say: "I don't see how even 4% inflation could have changed the course of events much".

I just want to clarify that MM aren't directly seeking additional inflation. They are seeking steady growth (typically 5%) in nominal GDP.

In ordinary times, the annual change in NGDP is typically made up of about 2% productivity growth (produce more wealth using the same amount of resources and labor), 1% population growth (including immigration), and 2% inflation.

The point of MM is that the steady 5% NGDP growth is more important than a steady 2% inflation rate. Given a 5% NGDP growth, the actual mix of that growth between productivity, population, and inflation is not under control of the Fed, and not important to the theory.

But of course, everyone would much prefer 4% real growth with 1% inflation, instead of 1% real growth and 4% inflation.

I just want to clarify that "MM want more inflation" is not completely accurate.