Friday, November 9, 2012

Demand and Competitiveness: Germany and the EU

I put up a post the other day about Enno Schroder's excellent work on accounting for changes in trade flows. Based on the comments, there's some confusion about the methodology. That's not surprising: It's not complicated, but it's also not a familiar way of looking at this stuff, either within or outside the economics profession. Maybe a numerical example will help?

Let's consider two trading partners, in this case Germany and the rest of the EU. (Among other things, having just two partners avoids the whole weighting issue.) The first line of the table below shows total demand in each -- that is, all private consumption, government consumption, and investment -- in billions of euros. (As usual, this is final demand -- transfers and intermediate goods are excluded.) So, for instance, in the year 2000 all spending by households, firms and governments in Germany totaled 2.04 trillion euros. The next two lines show the part of that expenditure that went to imports -- from the rest of the EU for Germany, from Germany for the rest of the EU, and from the rest of the world for both. The final two lines of each panel then show the share of total expenditure in each place that went to German and rest-of-EU goods respectively. The table looks at 2000 and 2009, a period of growing surpluses for Germany.

2000 2009

Germany Demand 2,041 2,258

Imports from EU 340 429

Imports from Rest of World 198 235

Germany Share 74% 71%

EU ex-Germany Share 17% 19%

EU ex-Germany Demand 9,179 11,633

Imports from Germany 387 501

Imports from Rest of World 795 998

Germany Share 4% 4%

EU ex-Germany Share 87% 87%

Ratio, Germany-EU Exports to Imports 1.14 1.17

EU Surplus, Percent of German GDP 2.27 3.02

So what do we see? In 2000, 74 cents out of every euro spent in Germany went for German goods and services, and 17 cents for goods and services from the rest of the EU. Nine years later, 71 cents out of each German euro went to German stuff, and 19 cents to stuff from the rest of the EU. German households, businesses and government agencies were buying more from the rest of Europe, and less from their own country. Meanwhile, the rest of Europe was spending 4 cents out of every euro on goods and services from Germany -- exactly the same fraction in 2009 as in 2000.

If Germans were buying more from the rest of the EU, and non-German Europeans were buying the same amount from Germany, how could it be that the German trade surplus with the rest of Europe increased? And by nearly one percent of German GDP, a significant amount? The answer is that total expenditure was rising much faster in the rest of Europe -- by 2.7 percent a year, compared with 1.1 percent a year in Germany. This is what it means to say that the growing German surplus is entirely accounted for by demand, and that Germany actually lost competitiveness over this period.

Again, these are not estimates, they are the actual numbers as reported by EuroStat. It is simply a matter of historical fact that Germans spent more of their income on goods from the rest of the EU, and less on German goods, in 2009 than in 2000, and that the rest of the EU spent the same fraction of its income on German goods in the two years. Obviously, this does not rule out the possibility that German goods were becoming cheaper relative to the rest of Europe's, if you postulate some other factor that would have reduced Germany's exports without a growing cost advantage. (This is not so easy, since Germany's exports are the sort of high-end manufactures which usually have a high income elasticity, i.e. for which demand is expected to rise over time.) And it is also compatible with a story where German export prices fell, but export demand is price-inelastic, so that lower prices did nothing to raise export earnings. But it is absolutely not compatible with a simple story where the most important driver of German trade imbalances is changing relative prices. For that story to work, the main factor in Germany's growing surpluses would have to have been expenditure switching from other countries' goods to Germany's. And that didn't happen.

NOTE: This my table, not Enno's. The data is from Eurostat, while he uses the Penn World Tables, and he does not look at intra-European trade specifically.

UPDATE: There's another question, which no one asked but which you should always try to answer: Why does it matter? The truth is, a big reason I care about this is that I'm curious how capitalist economies work, and this stuff seems to shed some light on that, in terms of both the specific content  and the methodology. But more specifically:

First, seeing trade flows as driven by income as well as price fits better with a vision of economy that has many different possible states of rest. It fits better with a vision of economies evolving in historical time, rather than gravitating toward an equilibrium which is both natural and optimal. In this particular case, there is no reason to suppose that the relative growth rates consistent with full employment in each country are also the relative growth rates consistent with balanced trade. A world in which trade flows respond mainly to relative prices is a world where macropolicy doesn't pose any fundamentally different challenges in an open economy than in a closed one. Whatever mechanisms operated to ensure full employment continue to do so, and then the exchange rate adjusts to keep trade flows balanced (or appropriately unbalanced, for a country with a good reason to export or import capital.) Whereas when the main relationship is between income and trade, they cannot vary independently.

Second, there are important implications for policy. Krugman keeps saying that Germany needs higher relative prices, i.e., higher inflation. Even leaving aside the political difficulties with such a program, it makes sense on its own terms only if there is a fixed pool of European demand. To say that the only way you can have an adequate level of demand in Greece is for prices to fall relative to Germany, is to accept, on a European or global level, the structural theory of unemployment that Krugman rejects so firmly (and rightly) for the US. By contrast if competitiveness didn't cause the problem, we shouldn't assume competitiveness is involved in the solution. The historical evidence suggests that more rapid income growth in Germany will be sufficient to move its current account back to balance. The implications for domestic demand in Germany are the opposite in this case as in the relative-prices case: Fixing the current account problem means more jobs and orders for German workers and firms, not  higher inflation in Germany. [1]

So if you buy this story, you should be more pessimistic about a Greek exit from the euro -- since there's less reason to think that flexible exchange rates will lead to balanced trade -- but more optimistic about a solution within the euro.

I don't understand why, for economists like Krugman and Dean Baker, Keynesianism always seems to stop at the water's edge. Why does their analysis of international trade always implicitly [2] assume a world economy continually at full capacity, where a demand shortfall in one country or region implies excess demand somewhere else? They know perfectly well that the question of unemployment in one country cannot be reduced to the question of who is getting paid too much; why do they forget it as soon as exchange rates come into the picture? Perhaps it's for the same reasons -- whatever they are -- that so many economists who support all kinds of domestic regulation are ardent supporters of free trade, even though that's just laissez-faire at the global level. In the particular case of Krugman, I think part of the problem is that his own scholarly work is in trade. So when the conversation turns to trade he loses one of the biggest assets he brings to discussions of domestic policy -- a willingness to forget all the "progress" in economic theory over the past 30 or 40 years.

[1] A more reasonable version of the higher-prices-in-Germany claim is that Germany must be willing to accept higher inflation in order to raise demand. In some times and places this could certainly be true. But I don't think it is for Germany, given the evident slack in labor markets implied by stagnant wages. And in any case that's not what Krugman is saying -- for him, higher inflation is the solution, not an unfortunate side effect.

[2] Or sometimes explicitly -- e.g. this post has Germany sitting on a vertical aggregate supply curve.


  1. Your argument from tables of numerical data is much clearer and more persuasive than Enno’s graphs. (Though I’m slightly worried about the cherry-picking issue; maybe you could present data for other years so the reader can judge whether the cherry-picking distorts the conclusion or not.)

    People should think twice before resorting to graphic visualization of quantitative info; very often the listing of numbers combined with written exposition is a much better method.

  2. If you rule out a monetary solution (because Germany won't accept inflation), and you rule out a fiscal solution (because Germany doesn't want to be in a transfer union), and you rule out a breakup of the euro zone, then what is left? Only the "natural cure" of relative price changes. No?

    Now where Krugman's logic is a little strange - for a Keynsian - is that he attributes the original source of the problem to prices getting out of whack. It's sort of like saying that the Great Depression was caused by wages rising too much relative to goods prior to the depression, thus causing unemployment while wages and goods prices came back into balance.

    1. Oh, Max, you are making me sad. Why is the point not getting across?

      Inflation = change in relative prices. The whole argument of this post is that neither a rise in German prices, nor a indefinite fiscal transfers to finance large Southern deficits, are required. Faster German income growth, with relative prices unchanged, would be sufficient to move the German current account toward balance.

      I like your second paragraph.

    2. Krugman is saying that the point of inflation is to accelerate relative price changes, but that's not what I'm saying. I'm just saying that there's no way to get the PIIGS out of the hole without overheating the German economy. They should accept this (there are worse problems to have!), but they don't.

    3. Max-

      I am not convinced that Germany could not have significantly faster income growth without inflation. Although it's true that your version of the argument is more reasonable than Krugman's, as I acknowledged in the first footnote to the post.

    4. Just to amplify the point -- German per capita GDP today is basically no higher than it was in 2007. Over the preceding 20 years, per capita growth averaged around 1.8 percent. So just getting back to trend -- which would almost certainly not be inflationary -- would imply a GDP of about 7.5 percent higher than actual, which given current import propensities, would erase almost all of Germany's surplus. And if you think, as I do, that Germany's longer term growth has been below potential, then the case against overheating is even stronger.

  3. Nicely done Josh--this is wonderful as an explanation. But I'd add something that is important here which you've alluded to and which I really liked about Enno's paper. This kind of explanation actually confirms a Post-Keynesian reading of the determinants of trade balances and has less place for price effects (Enno's regs seem to confirm this). Its not like there are no formal models out there that could explain this, but you'd have to take the idea that prices matter less than demand seriously then.

    1. Yes, some intellectual history should really be added to this story. Someone in comments to the previous post mentioned Kaldor, which is right; and of course Joan Robinson said all this a long time ago (as with almost anything worth saying in macro).

  4. What's the point of talking about "German income" as if it's some configurable parameter rather than mere statistic? German government doesn't control the wages in German private sector, and the Eurozone is one economy. Is a worker in Mulhouse paid more than an identical one in Freiburg, across the border?

  5. From a fast googling on italian sites (a financial site for polish wages and yahoo answers for italian ones) I get that Italian fiat workers are paid 1200-1600€ monthly while Polish fiat workers only 580€.
    Hence I would say that big differences in wages are quite possible among different European countries.

  6. But Poland is far away, and not even in the Euro zone.

    Anyway, it seems ridiculous to me to seriously discuss the Greek, Spanish, Italian, Portuguese economic issues as a function of ECB actions, German politics and stuff like that. These countries are fucked, internally. They don't have decent governments, decent institutions, and ECB and Germany can't do anything to help with that. Can't help a junkie by giving him more money. I know, this is a very orthodox and conservative view, but hey, what can I say, sometimes those just make more sense.

  7. @data
    Living in Italy, I think that your diagnosis is greatly exaggerated.
    It seems to me that the poor economy shows more evidently the corruption in the political system: for example, in a poor economy, businesses have to rely more on "pork" to go on, and this makes the corruption more evident. However I don't think that institutions in GIPSI countries are inherently more f@@@@d up than in other European countries (this is the reason I'm sceptic of moralizing movements like Grillo's).

    Since internal devaluation causes a fall of income, it can lead to a rebalancing of the trade balance even if competitiveness isn't that important IMHO; the problem is that it is a negative rebalancing that dumps the problem on the countries that export to the one that is internally devaluating.

    1. Since internal devaluation causes a fall of income, it can lead to a rebalancing of the trade balance even if competitiveness isn't that important

      I'm not so confident of that. The short term effects of a devaluation depend on a lot of things -- the composition of imports and exports, the respective pass-through rates of exchange rate changes (including what currency goods are priced in), and the different price elasticities. let's take an abstracted version of Greece, with exports of only tourism, and imports of only oil. Since tourism is priced in local currency, the short-run effect for home-country exporters is zero -- sooner or later lower foreign-currency prices of Greek vacations will increase sales, but this will take time. Meanwhile, the importers of oil -- priced of course in foreign currency -- immediately face higher local-currency prices. So in the short term, the income effects of the devaluation are unambiguously negative. In the long run, tourism demand *may* be price-elastic enough to produce a positive effect, but in the short run, yes, in the case of Greece a devaluation probably would improve the trade balance by reducing incomes. But if you want to go that route, you should not -- as Krugman does -- present devaluation as an alternative to austerity. It's just austerity in an even more disruptive form.

      Raise people's gas bills as a way to collect taxes (as the Greek state tried), or raise people's gas bills by devaluing -- either way you do reduce imports, yes: By making people poorer.

  8. You're the first Italian I know to deny that the Italian political system is extremely corrupt. And I don't hear the same mantra from Germans, or even French.

    1. Ha ha!
      I was waiting for it!
      I also believe that Italian politics is corrupt, but I don't think that it is a problem of institutions.
      For example, roughly twenty years ago we had the "tangentopoli" scandal, where a lot of politicians were jailed because they routinely asked for bribes. However a big part of those bribes was used by political parties to fund their campaign expenses, because Italian law was very strict about political donations of privates. The solution to this was, other than jailing certainly corrupt politocians, to make donations to political parties legal, actually making lobbing legal (previously it wasn't).
      I see this as very wrong headed, because the problem of corruption is that a politician that ask for bribes will make bad laws, not that the politician himself is getting some dirty money (that is also a problem but less relevant). More recently some scadals happened were some politician used some party money (public funds given from the state for campaining) for private expenses like buying a new house. This is also a blatant example of corruption but the great answer was to propose a law that eliminates completely public funding for political parties, that from my point of view is the worst idea. Most Italians like a lot this because they see corruption as the politician pocketing money, not as "some guy is giving money to a politician so that he can f@@@ us all".
      So I generally distrust those ideas about corruption because they focus on a relatively minor problem and don't see the real corruption, that is usually half legal (see lobbying above), and I have no reason to think that other countries are much better on this second kind of corruption (though on the first kind yes, we are recordmen).
      The problem is that, with an economy that is stagnating since like twenty years, those kind of "parasitic" behaviors became dominant - but this is an economic problem, not an institutional one IMHO.
      (Plus, why "even" French?)

  9. Please stop quoting P Krugman? He is a tool of TPTB. You just give him a platform.

    All politics involves rule bending. Hence all such dealing is "corrupt" to some one, legitimately. That is why Parliaments are usually exempt from laws that apply to non-members? Judiciary, on the other hand ....! The executive resorts to lying, as it usually has no such exemptions.

    1. I'm flattered by the idea that this little blog could add at all to Krugman's stature.

      Of course you are right that in general, one should try to promote people/ideas that are wroth reading, and avoid the temptation of arguing with stupid people. But in the world of liberal economists, Krugman is one of the good ones -- I don't mind adding my little inch to his giant plinth.

  10. The French like to complain, but there is (anecdotally, at least) no such wide-spread cynicism as among the Italians.

    It doesn't matter how they spend their bribes; what matters is that they act for the benefit of their sponsors, and against common interests. Obviously some level of corruption is inevitable, but, intuitively, there must be a point where even a small increment will cause the whole thing to become shaky and unstable.

  11. There is a reason we use data, Data, and don't just rely on anecdotes.

    The question is what explains the trade imbalances in Europe. A story about inefficiency, corruption, etc. could only do that if we saw a systematic shift in expenditure away from the (presumably) corrupt countries, to the clean ones. And that just hasn't happened -- except, interestingly, in the case of Italy, the one crisis country that has lost competitiveness. I don't have a good theory to explain that, but whatever it is, it can't explain the broader crisis since we don't see anything similar for Greece, Spain, Portugal.

    In the bigger picture, Italy has *always* been famously corrupt. Yet for several decades after World War II, it had the fastest growth rates in Europe. Economics is not a morality tale.

  12. Hey, who said anything about morality?

    And I don't think fast growth rates during a boom period contradict anything here, when government is in the pocket of big business, without any consideration, any long-term planning on the national scale. The wilder the boom, the stronger the bust.

  13. If any country were to leave the euro, how would that affect the adjustment process for the rest of peripheral Europe?

  14. And have you read Michael Pettis ( He seems to be the smartest economist out there, with a prejudice towards both Hyman Minsky and economic history, and everyone from Martin Wolf to Paul Krugman cites him. I think his views on Europe overlap a great deal with yours.