Tuesday, February 01, 2011

MIT Symposium Highlights Macroeconomic Muddle

MIT had a celebration of their 150th birthday, a Symposia on Economics and Finance: From Theory to Practice to Policy. Here's Bob Hall's solution to our problems (see around 50:00):
What we really like is for people to perceive that now is a great time to buy stuff instead of later, well there's nothing like inflation to get that mentality going. Normally we don't like to see that, we want to keep inflation under control because it's a sign of an overheated economy, well, we'd like a bit of that overheated economy now...

When the government buys more stuff, every model agrees that stimulates the economy...any economist who suggests that it doesn't stimulate the economy hasn't looked at a wide range of models, all of which agree...

Hey, the models all agree! In other words, if you look at the aggregate data, all we have here is a lack of demand. Just replace the missing private sector investment with government purchases, or by confusing investors and hoping they mistake nominal price increases for real demand, and all will be well. Most of the macroeconomists suggested that all we need to do is double down on the stimulus--only the political will is lacking for this obvious solution.

What we had was a misallocation--too much housing--and so now must move labor and capital to other areas, which creates temporary unemployment. To try to cover this up via having the government spend more on backfilling teacher's pensions, or have everyone buy an unsustainable amount of everything, would not solve this problem faster. Notice they don't spend any time discussing what government would spend this stimulus on because it doesn't matter to them. It's fun to think a solution to a hangover is more of a different type of alcohol, but I've tried it, and it doesn't work.

Unfortunately Keynesians don't have any intuition for this, because everything's all just 'aggregate demand', not housing, technology, energy, etc. Aggregation leads to simplifcation, but clearly it has a cost, and I think any macro theory that ignores the fact that an economy is a network of firms and individuals is pointless.

Hayek's early work on business cycles focused on misalignments in the structure of production. Alas, this was basically impossible to formalize. Keyenes's model, meanwhile, was adopted into the Hansen-Hicks synthesis that looked a lot like the simple Supply/Demand equations economists were used to, so everything seemed copacetic. A bad idea, in a tractable model, has a long life, because everyone forgets about all the hand waiving assumptions that underlie such models.

Joshua Bell is a famous violinist, but when he did an experiment, and played his $3.5MM violin at a train station, he made only $32. Reputation matters. In Bell's case, passersby did not realize he was truly a gifted violinist. In this case, while this was supposedly our best and brightest giving insight on the big issue of the day, the audience was treated to standard diagnoses and recommendations you hear everywhere on the left: we had a collapse in demand, so the government needs to spend more, or trick people into spending more via money illusion. If one wanted a better anecdote for William F Buckley's famous remark that he would prefer the first 100 names in the Boston phonebook to the Harvard faculty, I can't think of one.

As Cicero said, the purpose of wisdom is to know the good, in which case these people would have done better with the proverbial dollar fifty in late charges at the public library, rather than attend the esteemed establishment they were celebrating.

On the other hand, Google's chief economist, Hal Varian noted that Yahoo!, Microsoft and the rest are all hiring economists to design services based on game theory and other microeconomic specialties. This is in contrast to banks, that basically all went from having a large economic staffs in the 1970s, but a flak PR guy today for CNBC interviews. Bottom line: if you are going to do economics, don't do macro.


Anonymous said...

Re: Hayek, are you aware of any attempt at modeling a multisector economy with varying production times in different industries? When I was a student learning Sargent/Prescott style macroeconomics I tried to get somewhere with this many times but the math gets tedious. Fischer Black had some work that went part way with sector specific capital, but left off production times that varied by sector, which was central to Hayek's argument.

Re: All the models agreeing that increasing G always increases AD, I'm surprised anyone would say this. Even Keynesians should know that in some models crowding out takes place.

A Yeh-Chi Chen said...

Whoa, like other branches of economics are particularly better? Do you think looking for an obscure regulatory change in search of a significant t stat to add to the pile of t stats on the schooling-earnings relationship is that much better?

And anyway even if Macro is the worst of the fields currently, that doesn't mean one shouldn't try to make it better.

Patrick R. Sullivan said...

' William F Buckley's famous remark that he would prefer the first 100 names in the Boston phonebook to the Harvard faculty'

Congratulations on getting that anecdote right. Almost no one does.

Brad F. said...

But everyone loves macro! They love to be told what is going to happen. While shoveling snow my neighbor asked me yesterday:

what are you guys trading, Nasdaq?
No, Eurodollars, Interest rates
Where are we headed?
I don't know
If you had to be long or short, what would it be?

People love the casual prediction, so with so much demand, macro isn't going anywhere.

But if I were an academic economist I'd focus on the fun little stuff, like Dan Ariely or Satoshi Kanazawa... not that that behavioral economics is any more scientific.

Charles said...

Macroeconomics deals with abstractions that are difficult to define and even more difficult to measure. Relationships between these measurements are hypothesized for theory and statistically estimated for models. Then people assume that relationships that hold in one time period can be extrapolated into the future. They also hypothesize that a relationship between measurements will hold if the government manipulates an endogenous variable.

It's all a kind of pseudoscience like alchemy. Sometimes it seems to work.