Last week, finance professor Michael Pettis spoke at an event about a new paper he co-authored. From what I can tell, Pettis has had some influence recently on the U.S. trade policy debate, but I have found his views hard to decipher. He and I have interacted a bit on Twitter, but none of the exchanges provided much clarification. After listening to him at this event, though, I feel like I might have a better understanding of why it's so hard, as someone in the trade policy field, to figure out where he is coming from: It's because what he is saying doesn't have much to do with trade policy as it is generally understood.
That's not to say the issues he raises aren't interesting. However, they are not closely tied to the debate over protectionism vs. free trade, which is at the core of trade policy. (Unless I missed it, the term protectionism did not even come up at his event).
That's fine, of course. He doesn't have to be interested in protectionism. But that's what trade policy is about: The never ending battle between export and import interests over where we end up on the free trade-protectionism continuum.
Instead, his focus is on trade balances, and he focuses on them in a way that mostly excludes a discussion of protectionism. What he's talking about is interesting as a separate economic policy matter, but it doesn't help us much in thinking about the core trade policy issues.
Let me illustrate all this through some quotes from the event.
Pettis suggests early on that he is for "free trade," in some sense of that term: "The point of our booklet is that free trade is a great idea, but we are nowhere near that ideal. And so in that case, policies that are implemented in order to push us closer towards that ideal are actually free trade policies." That sounded like it might be going somewhere good, but then as he continues with this thought, he does not seem particularly concerned with protectionism. Instead, he focuses on trade balances: "Those are the kind of things that we should be doing. Ultimately, they mean reducing the deficits of the persistent deficit countries, which, by definition, means reducing the surpluses of the persistent surplus countries." Getting more specific, with regard to these surpluses, he elaborates as follows:
And I think one of the great genius of Keynes was that he recognized the problem is the surpluses, right? It's one of the big criticisms I have of the WTO. We define the problem very narrowly and very incorrectly, so we never really solve the problem. So we say tariffs are trade interventions, so you can't implement tariffs without going through all of these various hoops. And what Keynes would have argued, and certainly what we argued is, yeah, tariffs are trade intervention, repressing interest rates are trade intervention, subsidizing research is trade intervention, putting downward pressure on wages, breaking up unions as trade intervention, China's hukou system is trade intervention. All of these things effectively intervene in trade, because anything that changes the balance between savings and investment domestically is automatically a trade policy. And since it's impossible to identify all the different ways you can distort trade, all the different forms of industrial policy, we need to go back to Keynes's argument, which is, you see, whenever there's a problem, you see the problem in the trade imbalances. So go after the trade imbalances.
As part of his argument for going after surpluses, he offers the concept of "trade interventions," which, as he lays it out, includes a wide range of domestic policies. I don't doubt that policies such as the ones he mentions affect trade in some way. I just don't think it is very useful to try to create a category of domestic measures that affect trade. Just about all domestic measures will do so to some degree, so what do we gain by having this category? To me, calling something a trade intervention just doesn't tell you much. And it ignores the categorization that tells you a lot: Is a measure protectionist?
Getting back to the trade imbalances, which he says he wants to "go after," as I understand it, the view of most economists is that if you want to have more balanced trade, you would need to address the macroeconomic factors affecting levels of savings and investment, which are the actual cause of the imbalances. Here, Pettis tries to make the case that this is something he wants to address, but based on the example he offers related to Germany, I'm not convinced. In this regard, he had an exchange with Greg Ip of the Wall Street Journal, and it left me thinking that he may not actually be willing to deal with issues such as differences in savings rates between countries. Here was Greg's question:
You've identified a lot of distortions in the book that have led surplus countries to run very high saving rates, all associated with China, but there's a lot of countries that persist with large surpluses that do not have those distortions. Germany, most notably, arguably, the currency, part of a common currency, might fall into the category of undervalued currency, but none of the others apply. Germany is a very liberal economy, no restrictions on mobility, no restrictions on interest rates, etc. And even South Korea and Japan have evolved considerably to get rid of most of those distortions, yet still run surpluses. So I guess my question is, even if you moved to a world where the distortions you identify were eliminated, aren't there underlying marginal propensities to save and consume, that would result in the persistence of deficits, including in deficit countries like the United States, and doesn't that therefore suggest that there are things the United States must do in its own domestic policy framework to correct these imbalances?
Pettis answered as follows:
Well, see, I would disagree that those economies don't have distortions. They have all sorts of distortions. You can argue that the euro is a huge distortion for all of the economies of Europe, including Germany, because the euro is an undervalued currency for Germany, it may not be for other countries. Remember, inflation in Germany -- after the adoption of the euro -- inflation in some countries of Europe was much lower than inflation in other countries. So what that means is that the real value of the currency in those other countries went up, and the real value of the currency in Germany went down. Why does an undervalued currency matter? Because an undervalued currency, like everything else in economics, is simply a transfer. An undervalued currency hurts importers and it helps exporters, and households are all net importers, and the manufacturing tends to be exporters.
But Germany is even a clearer case, because before 2003 German savings rates were much lower and Germany ran a deficit. We forget that, but in the 1990s Germany was running current account deficits. So what happened in 2003, 2004, 2005 that shifted Germany into, at one point, the highest surpluses in the world, maybe in history? And we all know what it was. It was the Hartz labor reforms. ... before the labor reforms, as German GDP went up, German wages went up just as quickly. After the reforms, as German GDP went up a little bit more slowly, German wages flatlined. So notice what happened in Germany. The household share of GDP after the labor reforms declined, right? And as the household share declined, not surprisingly, the consumption share declined. And as the consumption share declined, by definition, the savings share soared. And everyone started talking about those thrifty Germans. If you look at household savings, it didn't change at all. The reason savings went up is because wages went down and business profits went up. And business profits are savings, right? So the famous savings of the German people was really caused by a labor reform that transferred income from German workers to German businesses. German business profits went up 50% in a couple of years right. So I would argue that that was a policy issue.
I think Greg nails a key problem with Pettis's argument here, and I don't think Pettis gave a convincing response. In particular, Pettis doesn't seem willing to deal with the basic point about "underlying propensities." Cultural factors are real, and I'm not sure why he is so reluctant to acknowledge this. Germany has been famous for its thriftiness for a long time, not just since 2003-2005, something that comes out clearly in the data. With regard to the 1990s, what I think Pettis may be missing here is the impact of German reunification, which led to increases in spending, but I'd have to look into this more. Regardless, saying that Germany started being thrifty in 2003 seems clearly wrong. So, this isn't very credible as it relates to Germany, and it ignores Greg's point about U.S. savings rates as well. Pettis is very focused on reducing the surplus in surplus countries, but I'm not sure he's willing to support the changes in deficit countries that would be required.
Also on the Germany vs. U.S. savings issue, a new essay from Andreas Freytag and Phil Levy has some interesting insights:
The question of why savings are low can often be answered by examining an economy’s demographic structure (Box 3).
Box 3
Demographic trends and savings
The individual savings rate tends to be a function of age. Most people are net savers during their years in employment or self-employment (i.e., between 15 and 64 years). During their early, formative years before joining the workforce, they live on other people’s savings (e.g., parents, grandparents, or credit institutions) to pay for education. Once their working years are over, they dissave to finance their retirement. This implies that a younger population will save less than an aging population.
In the United States, which has a relatively younger population than other Western nations, people tend to save a smaller share of their income. On the same token, a young and growing population needs capital. Thus, the business sector tends to invest at a higher level. The resulting shortage of savings relative to investment demand drives the current account deficit.
Germany, in contrast, is more of an aging society. The supply of labor and human capital is in relative decline, while the savings rate remains high, at least as long as most people remain below retirement age. As members of an aging society, Germans invest a larger part of their savings abroad, creating net capital outflows and a trade surplus. The differing trade balances between the United States and Germany, therefore, have nothing to do with competitiveness or unfair trade practices.
Like the regular US current account deficit, the German current account surplus is the rational result of underlying economic and demographic factors.
As a final, and mostly unrelated, point, Pettis has this to say about Japan, Germany, and South Korea and their surpluses:
If you look at Japan, if you look at Germany, South Korea, if you look at the surplus economies, their manufacturing share of GDP is higher than the global average, and if you look at the advanced deficit economies, it's lower than the global average.
It is certainly true that Japan, Germany, and South Korea manufacture a lot. It's worth noting here, however, that this is an unsurprising result given their relative lack of natural resources and farm land. If they had oil, for example, they would probably spend more time on extracting that oil and less time on manufacturing.
Summing up, there are a whole bunch of trade policy fights going on right now, and the issue of trade balances is often thrown into these fights in unhelpful ways. Some people seem to want to see the trade balance as a scorecard of who is winning and losing in international economic relations, but that's not what it is. Trade balances are mostly driven by macroeconomic factors, and my sense is that international coordination on those issues will be extremely difficult, with governments of both surplus and deficit countries not very eager to make changes.
