"Personally, I think morality play deserves a much larger place in economics than it currently has"

From the always interesting Interfluidity:

"Krugman's "hangover theory", revisited.

I'm trying to write something hard, and failing. I'll keep trying. But this is easy, and I cannot resist. Paul Krugman is once again attacking "hangover theorists", the idea that booms of a certain kind inevitably beget recessions. I do not buy the traditional Austrian story of hangovers — that misallocations and depletions of capital (including human capital) necessarily take time to undo. But I think that now and in his original piece, Krugman is far too quick in his dismissal of the idea that there must be something about some booms that makes subsequent recessions pretty hard to avoid( I AGREE ). Krugmans writes that "[a] recession happens when, for whatever reason, a large part of the private sector tries to increase its cash reserves at the same time( MOVE TO SAFETY. NOT FLIGHT. )." It is rather surprising, isn't it, that "whatever reason" almost always happens subsequent to years of unusual prosperity? Choose your poison — if you don't like the Austrians, go with Hyman Minsky — but if we don't acknowledge the relationship between some kinds of booms and the bad times that follow, we'll have a hard time preventing those bad times( GOOD POINT ).

Krugman is absolutely correct to inveigh against the "morality play" that sometimes seeps into the Austrian rhetoric surrounding recessions. Personally, I think morality play deserves a much larger place in economics than it currently has( I AGREE. IT'S CALLED POLITICAL ECONOMY ), but a fable in which masses of innocents suffer to absolve the sins of the reckless wealthy is hardly moral(GOOD POINT ). The "hangover theory" is best described as an immorality play, which we are watching unfold before our eyes this every moment as financial assets are relentlessly supported while the value of a pair of hands is let to plummet.

However, recessions and depressions do follow booms, and there are reasons for that. Austrians have their vices, but a vice of Keynesians is to underestimate the role of information. Krugman points out that the hangover theory...

doesn’t explain why there isn’t mass unemployment when bubbles are growing as well as shrinking — why didn’t we need high unemployment elsewhere to get those people into the nail-pounding-in-Nevada business?

The obvious answer is that when there is a boom, entrepreneurs know into what sector resources must be reallocated, and pull already employed workers from existing jobs into the new big thing. During a bust, from a God's eye view, the same process must occur: resources must be shifted out of some sectors and into others( RIGHT NOW, OUT OF HOUSING ). But entrepreneurs are only human. They do not know to where resources might be productively employed, only that they cannot be productively employed where they are. This is the asymmetry, I think, that explains mass unemployment during busts.( TRUE. PEOPLE ARE MOVING INTO NEW JOBS )

Krugman also points out that the hangover theory...

doesn’t explain why recessions reduce unemployment across the board, not just in industries that were bloated by a bubble.

I think that this gets to the point about why it is that only certain kinds of booms lead to great and terrible busts. Industries rise and fall all the time, in good times as well as bad. In the 1980s, there was a great boom in the recording industry owing to the advent of compact discs. The boom eventually went bust, but mass unemployment did not ensue. Hangovers result not from booms in and of themselves, but from booms which result in unhealthy concentration of the aggregate investment portfolio( I WOULD SAY RUNS LEAD TO THE REALLY AWFUL DOWNTURNS ). US capital, viewed as a whole, was overly concentrated in housing and construction this decade. China's capital has been overly concentrated in exports and construction( THERE'S NO REAL "OVERLY" ). Traditional portfolio theory views the menu of investments as fixed, and suggests that investors diversify among them. But in the aggregate, there is only one portfolio extant at any point in time. The art of "macro portfolio theory" is to control the evolution of that portfolio so that it remains reasonably efficient. The easy answers don't work: Micro portfolio choices don't necessarily compose into a dynamically sane macro portfolio( TRUE ). We have reason to be skeptical of very heavy-handed industrial policy. So we have work to do.

I'll end with an intuition: I think that there's a trade-off between microlevel diversification and macro-efficiency. Barry Bosworth warned that "diversification devalues knowledge". One reason that micro portfolio choices fail to compose is because it is often sensible for investors to "buy the market". Every individual has a unique information set, and ideally we would want all that decentralized knowledge "priced into the market" independently of the judgments of others. However, each individual knows that her own information is profoundly uncertain and incomplete, and that the market represents an aggregation of the judgments of millions of others. So, as passive-investment types have been telling us for more than a decade, it may be optimal for individuals to ignore their own information and defer to the judgement of the market-ex-me. (This is a kind of "information cascade".) But, each person who defers to the market increases the concentration of investment decision making, and decreases the breath of information that is priced into the market( I NEED MORE ABOUT THIS ). If the aggregate portfolio is disproportionately by the decisions of a relatively small group of people, there is no reason to suspect its quality would be better than that decided upon by a bureaucracy of planners( THERE MIGHT BE ). There is reason to suspect, in fact, that it would be worse, because at least the planners know they should at least pretend to serve a broad public interest, while private decisionmakers might quite legitimately think they're just trying to get a piece of next year's bonus pool.( THIS IS TOO GENERAL A STATEMENT )

In sum, I think there is a tension between micro diversification and macro diversification. If we want to maintain a well-diversified aggregate portfolio, it may be necessary to restrict the degree to which the portfolio of firms and individuals can be diversified. This implies forcing individuals to bear more risk than they would otherwise choose, in order to reduce systemic risk( THIS I AGREE WITH ). We might be better off by letting individuals shed risk via some form of social insurance( I AGREE ) while forcing investment choices to be sharp, than by encouraging people to blur the information they present in their portfolio choices in order to diversify and hedge."

The goal is to have a system that stops runs without guaranteeing everything. Bagehot's Principle's do just that.