Sorrow and Pity of Another Liquidity Trap, by Brad DeLong: ...At the end of 2008, as the economy collapsed and the pace of net Treasury debt increases quintupled,... I presumed we had a little time for expansionary fiscal policy to boost the economy -- a year, maybe 18 months -- before the bond-market vigilantes would arrive. They would demand higher interest rates on Treasury bonds, which would begin seriously crowding out the benefits of fiscal stimulus. ...
But it didn’t happen in 2009. It didn’t happen in 2010. And it isn’t happening in 2011. There are no signs from asset prices that the market is betting heavily that it will happen in 2012. Looking at the yield curve, it appears the market intends to swallow every single bond that the Treasury will issue in the foreseeable future...
Although I worked for three years in the Clinton Treasury Department, and am a card-carrying member of the economist guild, I predicted none of this. Like most of my peers, I was wrong. Yet the most interesting thing is that I could have -- should have -- been right. I had read economist John Hicks; I just didn’t quite believe him.
Hicks ... was responsible for the ... version of the IS-LM model that formalized and elevated a key insight: that interest rates paid by creditworthy governments would remain low after a financial crisis. ...
This is the liquidity trap. In this situation we need deficit spending. ...
I had read Hicks. I even knew Hicks. But I thought that his era, the Great Depression, had passed. Sitting in my first graduate economics class in 1980, I listened to Marty Feldstein and Olivier Blanchard -- two of the smartest humans I am ever likely to see -- assure me that Hicks’s liquidity trap was a very special case, into which the economy was unlikely to wedge itself again. Yet it did.
On my shelf is a slim, turn-of-the-millennium volume by Paul Krugman titled “The Return of Depression Economics.” In it he argued that we mainstream economists had been too quick to ditch the insights of Hicks -- and of economists Walter Bagehot and Hyman Minsky. Krugman warned that their analysis was still relevant, and that if we dismissed it we would be sorry.
I am sorry.
I remember feeling pretty lonely calling for aggressive fiscal policy very early in the crisis.
On the bigger issue, I've argued that we weren't asking the right questions prior to the crash. We were worried about how to conduct monetary and fiscal policy in an economy that is mildly fluctuating around its long-run optimal path due to price and wage rigidities, and we used DSGE models to answer that question (the answer was that monetary policy alone would suffice, so there was no need to even worry about fiscal policy and the political problems that come with it, and it was largely ignored). Though there were notable exceptions, e.g. the literature on financial frictions and financial accelerators, for the most part researchers did not build models that examined the question of what might happen, and how best to respond, if there was a financial panic and a break down in financial intermediation. That was becasue "Hicks’s liquidity trap was a very special case, into which the economy was unlikely to wedge itself again" -- and the "central problem of depression-prevention has been solved," as Robert Lucas told us.
There will be a frenzy of research on this issue, just as there was after the Great Depression, and we will come to conclusions about what went wrong and what we can do better next time. Then we will move on as other important questions come to the forefront.
The question is, will we forget again? Will we stop asking these questions as the crisis fades in our collective memories? When the next crisis hits, as it will someday, will economists who reach into their tool bag of macroeconomic models come up empty except for those decades old DSGE models (or whatever we come up with that explains our recent experience) and even older IS-LM formulations? Will they, once again, be forced to choose between modern models that were not built to answer the questions policymakers need answers to, and older models that, despite their limitations, were? I suspect that they will.
[Paul Krugman follows up on Brad's column here.]