Karl Smith writes:
Then there is basically everyone else who is some version of a New Keynesian. This literature is much bigger than I am but the bedtime story that we are told in graduate school is that the modern models began with this guy. You might have heard of him: http://www.economics.harvard.edu/files/faculty/40_Small_Menu_Costs.pdf.
But how about John Stuart Mill (1829)? And what are George Akerlof and Janet Yellen (1985), "A Near-Rational Model of the Business Cycle", chopped liver?
Karl Smith goes on:
For these guys recessions are about an excess demand for some type of financial instrument. Some people see it as explicitly bonds. This means everyone is trying to save at the same time but this is impossible. Some people see it as explicitly money. This means everyone is trying to build up cash reserves at the same time but this is impossible. Some people see it as explicitly high credit instruments. This means everyone is trying to run away from risky investments at the same time but this leads to a cascade that makes moderately risky investments very risky, which in turn makes mildly risky investments very risky and so forth.
I tend to think that all three work together as an agglomeration because of a host of transaction cost and principle-agent problems. But that is beyond the scope of this post.
For the most part the central issue in this line of thinking is: how can swapping around financial instruments which are in some sense just pieces of paper have these huge effects on the real economy? For this to work the connection between the real economy and the financial economy has got to be sticky in some way. The bedtime story is that Greg offered one of the first consistent reasons why. By and large most economists are working with some type of New Keynesian model. The difference is the focus or the details. In terms of policy, however, I think political economy concerns dominate.
My thoughts tend this way: look this all about money (or bonds or credit) and so the focus of everything is the Fed. I was a stimulus skeptic. I could see the reasoning but I thought it better to focus all of our attention on monetary policy. I also suggested then and now that if fiscal stimulus must be done, that it should consist entirely of tax cuts or increases in direct assistance to the poor. However, this was for political economy reasons. Not model reasons. If someone asked whether I thought GDP was higher or lower as a result of the Obama stimulus, I would answer: higher. On the other hand, Brad and Paul like to focus on spending. I suspect this is in no small part because they think government spending is too low anyway. Why not kill two birds with one stone: build some roads and get some jobs.
A lot of the spending guys like to focus on the higher multiplier of spending when compared to tax cuts. I think this is to some extent a red-herring. We get more bang for the buck with spending but we can also move more bucks with tax cuts. For example, a complete payroll tax holiday in 2009 would have resulted in something like 900B in stimulus in a single year. With ARRA I think we got something like 300B and even then much of that was from the tax cut portion. Spending money just takes time. At the end of the day, your choice of New Keynesian instruments: monetary policy, government spending or tax cuts depends mostly on political economy concerns. That is, your view of the fundamental relationship between the government and the economy.
I would say that it's not primarily whether the government should be bigger or smaller.
It's political economy in the sense of which instrumentalities--congressional appropriations, IRS, Treasury, Federal Reserve--can actually implement successfuul programs.
And it is the exact shape of the financial excess demand--which will always be, empirically, a mixed matter--that determines which strategic governmental interventions will be most effective. Printing up more government bonds will fail when the root problem is excess demand for money and money demand is not very interest-elastic. Standard open market operations in short Treasuries will fail when the root problem is excess demand for bonds-as-savings-vehicles or for high-quality assets and money demand is very interest-elastic.