Last Night's Event: Thanks to a commenter who found the video here. I don’t know if this is official or sanctioned, although it looks like it; I also don’t know why the Times tech people didn’t alert me. But anyway, it seems that those who missed the discussion can watch it via that link.
A word about the content: I continue to find Carmen Reinhart’s fatalist view puzzling. She agrees with me that we’re facing a demand-side problem--but insists that this problem can’t be solved quickly, that we need to go through many years of painful deleveraging that leave millions of potentially productive workers idle.
I agree that this is probably what will happen, given the political realities. But surely this is a huge failure of policy, not something we should accept as inevitable. It’s truly bizarre, if you ask me, to say that our economy suffers from too little spending, and that nothing can or should be done to increase that spending.
As I have said it before, I find it easiest to interpret Carmen's argument in Karl Smith's version of the Hicksian combination of the Wicksellian flow-of-funds and the Fisherian velocity-of-money models, the IS-LM model. From Fisher we get the idea that the economy will rapidly head for and then sit for a time at its short-run equilibrium where the level of spending PY and the safe short-term nominal interest rate i are such as to make people want to hold the liquidity--the money--that the economy has for them to hold:
And from Wicksell we get the idea that the economy will head for and then sit for time at its short-run equilibrium in which real incomes and production Y are such that savers are happy holding all the new bonds and loans that businesses and the government are taking out:
Add short-term sticky prices to this (or add some other rule for relating changes in nominal spending to changes in real income and output), and you have a short-run model on which everyone who has thought the issues through ought to agree on as a starting point.
Karl's wrinkle is to say that the business sector and the Treasury (unless it has the cooperation of the central bank) can't just issue all the bonds that they want to. They have to find financial intermediaries to hold them. And those financial intermediaries operate in a Stiglitzian credit rationing regime in which perceived quality matters, and promising to pay a higher interest rate may not reassure those you are trying to sell your bonds to. Hence Karl replaces planned bond issues by business and government I + (G - T) with a "bank lending" function that tells how many bonds financial intermediaries are willing to accept as a function of the short-run cost of funds to well-capitalized intermediaries i, the expected inflation rate π, and the ratio of the riskiness of the bonds being offered to the risk tolerance of financial intermediaries ρ:
Now let's shift into the world in which Carmen Reinhart lives. In that world, you are at the zero lower bound (and thus money demand in the LM curve is infinitely interest elastic). In that world, government credit is shaky so that shifting the composition of borrowing from the private sector toward the government sector by having the government issue more bonds and thus crowd-out private bonds does not diminish the riskiness of the outstanding bond stock. Thus, in CarmenWorld, the only government policies that can possibly work to stimulate the economy are those that (a) boost the risk tolerance of the private sector, and or (b) lower the perceived riskiness of the outstanding (private and public) bond stock.
Thus in CarmenWorld stimulative economic policy is reduced to falling on our knees and saying:
COME, O CONFIDENCE FAIRY, COME!!
I don't think that we live in CarmenWorld--for one thing, it seems to me that the Treasuries of the U.S., Japan, Germany, the U.K., Finland, and other countries that control their own currencies are still able to take a lot of private-sector risk onto their own balance sheets and fund a lot of extra spending and so shift the portfolio of outstanding bonds in a way that would substantially diminish its perceived riskiness.
But Carmen thinks differently. And she is very Notstupid, as Mycroft/Michelle the HOLMES 4 computer would say...