Mr. King's larger topic is the "failure" of macroeconomics since the Great Recession. He has a lot of interesting things to say about that, but he also addresses why the current recovery has been relatively slow and weak.
There are a number of theories:
- Larry Summers' Secular Stagnation model, by which interest rates can't go below zero.
- Robert Gordon's low-productivity model, which maintains the "miracle century" is over and done with and won't come again anytime soon.
- It's all the Fed's fault, which for malign and inexplicable reasons has kept interest rates very low for far too long.
These are not really mutually exclusive, but the long-term decline in global interest rates since 1990 requires an explanation. And Mr. King offers a good one.
One of the shortcomings of modern macro (in his view) is that the computer models all assume a single sector. This is best explained using the cartoon version: The economy consists of a GDP factory that manufactures GDP. Consumers buy the GDP, and if they buy all of it then we have full employment and a thriving economy. If, on the other hand, consumers only purchase some of the GDP, then there will be layoffs, declining production, deflation, and all the other ills that afflict a recessionary economy. The disease is generally described as "low aggregate demand", which just means consumers don't want to buy all the GDP. The cure is to increase aggregate demand, typically by the government buying up the excess GDP through deficit spending, and thereby putting folks back to work again.
A common criticism is that the economy is much more complicated than that--it's not just a GDP factory. Instead you have technology, and retail, and automobiles, etc. Some of these sectors may be healthy and others not so much. There's no such thing as "aggregate demand", but rather demand for the products of these individual sectors.
The computer guys are aware of this shortcoming, and so they inserted multiple sectors into their models and ran the programs that way. What they discovered is that it didn't make much difference--for all intents and purposes you got the same answer whether or not you use a multi-sector model or a single-sector model, i.e., a GDP factory. (I'm assuming that the sectors they tried are things like Industrials, Transports, Retail, etc., but I don't know that for sure.)
So now Mr. King comes along and suggests that the single-sector model is fundamentally wrong, and that at least two sectors are required. But rather than divvying up the economy as market indicators do, he instead makes the fundamental distinction between tradable goods and non-tradable goods.
The difference is that foreigners can buy tradable goods, whereas they can't easily buy non-tradable goods. So airplanes, movies, software, and computer chips can be exported abroad, and those are examples of tradable goods. Conversely, haircuts, taxi rides, restaurant meals, and doctors' visit are not readily tradable. (At the margin everything becomes fuzzy: foreign tourists can eat in our restaurants and foreign students can study at our universities, but for the most part restaurants and colleges are in the non-tradable sector.)
If the economy is split between tradable and non-tradable sectors, then there is some optimal distribution of capital and labor investment that optimizes total GDP. That is, the optimal proportion of capital is invested in tradable industries with the remainder in non-tradable industries. And similarly for labor.
This idea can be represented graphically (see below). The blue arrow (labelled a von Neumann ray) represents economic growth--the longer the arrow the faster the economy is growing. The direction of the arrow tells you the optimal proportions of investment in tradable (T) and non-tradable (NT) industries. Choose any different direction and the arrow will be shorter (i.e., suboptimal). Taken literally (which is not the intention) the diagram indicates that the bulk of the investment is in tradable industries since the arrow is closer to the T axis. (Put mathematically, the blue arrow is a vector in a 2-dimensional vector-space, with the coefficients representing the fraction of investment in either T or NT.)
 |
| Diagram taken from King, Mervyn: NBER Reporter (2017: 3, pp. 1-10) |
Of course there is no guarantee that the economy is optimal--if resources are misallocated between the two sectors then it won't be. A tell that there is misallocation is if there is a permanent trade imbalance such as what exists today. In today's world, China and (especially) Germany have been running huge trade surpluses, while the US and the UK have been running substantial trade deficits. These imbalances have persisted for decades (though in the case of China things are beginning to change).
This means that the US is producing too few tradable goods, and conversely producing too many non-tradable goods. Schematically, it implies that we've gone off the rails somewhere around point A on the graph and have been following the suboptimal, dashed-line curve to point B. This, in Mr. King's opinion, is the reason for the sub-par growth of the economy since the Great Recession.
It also explains low interest rates. A big trade surplus (such as for Germany) results in a comparably large capital surplus for the United States. This typically shows up in German and Chinese purchases of US government debt, i.e., treasury bills, resulting in very low interest rates.
Before the Great Recession we were able to muddle through. Since the Recession, however, the situation has become unsustainable and we (along with the rest of the world) are taking a hit to growth. The solution is to rebalance investment between the two sectors and follow the red line from B back to point C, on the optimal von Neumann ray.
For Germany the situation is opposite. Their economy also deviates from optimum, but to the left of the blue arrow rather than to the right. They are producing too many tradable goods and as a result required to export capital around the world. US government debt is the least of their problems; more troublesome are the large loans made to countries such as Greece.
My thoughts:
1) If Mr. King is correct, then huge investments in new infrastructure are precisely the wrong thing to do. After all, there is nothing less tradable than a new highway, and since we're already over-invested in NT, this just doubles down on a bad idea. I've never been a fan of infrastructure spending, which I regard mostly as wasteful pork. Mr. King simply confirms my opinion.
2) Among the things I most appreciate about the Republican tax plan is that it disses higher education. Of course that's mostly political--higher ed is certainly no friend of conservative politics, and for that reason alone they deserve to be defunded. But more, I think this country invests waaay too much in education generally, and higher ed specifically. And given that education is mostly an NT industry, Mr. King's thesis reinforces the point. Disinvesting from higher ed at the margin is a very good idea, and an excellent reason to support the Republican tax plan.
3) Mr. Trump's fixation on the trade deficit may not be quite as cockamamie as it sounds (though it's still pretty cockamamie). Most economists think that the trade deficit doesn't matter, and that's certainly what I used to think. To be sure, year to year fluctuations don't matter, but when the deficits are long term and endemic, there clearly is a problem. In his own inarticulate way Mr. Trump is addressing an important issue. Though I do wish he'd phrase it in less zero-sum terms.
In short, I learned something from Mr. King. If anybody thinks I misunderstand or have misstated his views, please let me know.
Further Reading:
- Book Review: The Rise & Fall of American Growth
- Book Review: Walmart
- Your Future Job: Building a Career in the New Normal
No comments:
Post a Comment