How to Make a Mess of a Monetary Union, and of Analyzing It Too

Servaas Storm means well. He is alarmed that the eurozone’s official strategy of “internal devaluation” might do more harm than good by unnecessarily forcing countries that have lost their competitiveness into deflation (see here, here, and here). This is a very real concern indeed and Storm should be applauded for raging against the colossal folly that is wrecking Europe. Unfortunately, Storm goes astray in seemingly dismissing any role for unit labor cost competitiveness and German wage moderation in causing the still unresolved eurozone crisis in the first place.

Referring to bits and pieces of evidence derived from mostly partial-equilibrium empirics of one type or another, Storm fails to notice that no coherent macroeconomic analysis of the eurozone crisis emerges unless German wage moderation gets assigned a prominent role in the play. At the heart of the whole confusion is Storm’s attempt to attribute to those who emphasize German wage moderation as a key causal factor in the eurozone crisis the view that “expenditure switching” would explain 100 percent of the eurozone’s internal current account imbalances (and related balance sheet troubles). This would be a very peculiar view indeed – and I am not aware of anyone who actually holds it. Certainly the proponents of the “wage moderation hypothesis” that I know, including those who responded to Storm’s “critical analysis” (see here and here, and also this author), definitely do not hold this view. Effectively, Storm set up a straw man that he then defeats with flying colors; not realizing that his arguments are self-defeating and make a mess of the whole analysis of monetary union.

Storm discusses many of the key ingredients providing the material for a coherent macroeconomic analysis of the eurozone crisis but ends up getting matters upside down. Start with his explicit acknowledgement that wage moderation undermined domestic demand in Germany, which, in turn, translated into an easier monetary stance for the eurozone as a whole than would have been the case if German wage trends had been properly aligned with the ECB’s stability norm (i.e. HICP inflation of “below, but close to 2 percent” for the eurozone as a whole). Reading Storm’s own words on this issue makes it clear that he wishes to delegate cost competitiveness to the status of complete irrelevance: “German wage moderation mattered a lot, not through its supposed impact on cost competitiveness, but via its negative impacts on (wage-led) German growth and inflation, which in turn prompted the ECB to lower the interest rate in the first place.” Supposedly, in Storm’s view, the cost competitiveness factor is associated with expenditure switching and, given that Storm attributes to his opponents the view that expenditure switching would explain 100 percent of current account imbalances, he takes the opposite view that cost competitiveness explains nothing at all. In his mind, the income effect alone explains it all.

The trouble with this argument is that you can’t really have the one without the other. Wage moderation could hardly have dampened domestic demand in Germany, easing financial conditions across the eurozone, without simultaneously also boosting Germany’s cost competitiveness. Proponents of the “wage moderation hypothesis” do not hold the view that cost competitiveness and expenditures switching explain 100 percent of any intra-eurozone current account imbalances. Their argument is that German wage moderation provided an important exogenous cause of those imbalances. The so-called “consensus narrative” (see here), which Storm subscribes to, denies this, and in fact fails to even mention German wage moderation – a peculiar oversight that prompted Bofinger’s critique (see here and also here)[1]. A related important (legacy) matter is that it will hardly be possible to rebalance the eurozone without, in one way or another, also reversing out-of-kilter competitiveness positions (on which more below).

If there is agreement, as there appears to be, that German wage moderation suffocated German domestic demand with important area-wide ramifications through the ECB’s monetary stance, what was the supposed role of capital flows? Storm is not very clear on this issue either. On the one hand, he seems to suggest that gross capital flows caused the divergence in competitiveness positions. On the other hand, he seems to suggest that the ECB’s “one-size-fits-none” monetary stance fired up credit creation and asset bubbles in the euro crisis countries. These two propositions are not the same thing, nor do they rule out a prominent exogenous causal role for German wage moderation in the play.

Let’s begin with what Storm has to say about gross capital flows as such:

“the impact of trade flows on the exchange rate is—generally—overwhelmed by the impact of cross-border gross financial flows, which (importantly) are mostly unrelated to trade (Akyüz 2014; Bortz 2016). This holds true for the Eurozone as well: billions of euro lent by German banks to (financial) firms in Ireland and Spain, and by French banks to (financial) enterprises in Greece, Ireland, Portugal and Spain were unrelated to the financing of trade (see O’Connell 2015). It is exactly these gross financial flows from the Eurozone core to the periphery, mostly coming from powerful too-big-to-fail banks (O’Connell 2015), which played the central role in bringing about the imbalances and destabilizing the zone, a role recognized by Professor Bofinger (2016) and by the so-called Consensus Narrative (2015)—but left unmentioned and un-analyzed by Flassbeck and Lapavitsas.”

It is undisputed that there was a lot of lending in the run-up of the crisis by German and French banks; in particular, to financial firms in euro crisis countries. It is rather odd that Storm views the impact of capital flows on exchange rates under regimes of flexible exchange rates as equivalent to the impact of capital flows on wage trends inside a monetary union. In general, financial asset prices, such as exchange rates, are the fastest moving prices while money wages are the slowest moving prices in economies; as emphasized by Keynes in “The Economic Consequences of Mr. Churchill.” Storm appears to push economics back almost one hundred years here into the pre-1925 era of proto-macroeconomics. That does not count as progress I am afraid. And there is also no need to construct any tight direct links between capital flows in general and trade finance in particular or to suggest any need for a one-to-one correspondence between German banks’ lending to euro crisis countries and trade imbalances. German banks could have piled into US Treasuries while Spain’s financial sector could have relied on refinancing via Belgium and even Singapore, with other countries acting as in-betweens, for instance. That’s not the point. It is however of great interest for its own sake that German banks did actually build up huge direct exposures to euro crisis countries under the euro. For this fact might well have influenced the German authorities’ handling of the euro crisis.

The role of capital flows in general was to arbitrage financial conditions across the currency union. Europe’s earlier common market project had promised to turn the large European market into a level playing field, including Europe’s common financial market. The common currency project featured the promise to be the monetary unifier of Europe. As the ECB came to emphasize, once the crisis and financial (re-)fragmentation across Europe’s common financial market disrupted the “singleness” of its monetary policy, the idea is that its uniform monetary stance translates into largely uniform financial conditions across the currency union (fulfilling the level-playing-field promise).

The trouble is that its uniform monetary stance and uniform financial conditions may not actually suit a currency union that is subject to strong internal divergences. This is of course the very subject of “optimum currency area” theory, with its emphasis on “asymmetric shocks.” Traditionally, optimum currency area theory focused on exogenous asymmetric shocks and the need to somehow rebalance the shocked currency union. The wage moderation hypothesis argues that German wage moderation triggered the much-feared event of an asymmetric shock that became self-reinforcing (endogenous, in a sense). For one thing, fiscal austerity reinforced domestic demand stagnation in stagnant Germany (aka “sick man of the euro” at the time) as Germany struggled with the SGP for five years in a row in the 2000s. For another, the ECB’s “one-size-has-to-fit-all” stance came to fit no one – reinforcing stagnation in Germany while firing up bubbles in Spain and Ireland. The “consensus narrative” implicitly treats German wage moderation, if anything, as a consequence of divergences somehow caused by capital flows. As highlighted in Bofinger’s critique, the wage moderation hypothesis, while not at all denying the importance of capital flows, views German wage moderation as an important exogenous cause behind those very divergences. Storm acknowledges German wage moderation and the ECB’s part – but he also likes to agree with the “consensus narrative” and ends up blaming it all on ominous capital flows. You can’t really have it both ways I am afraid. If German wage moderation mattered for Germany, it also mattered for the eurozone as a whole. After all, Germany is too big a chunk of the eurozone to be irrelevant to its overall destiny.

Storm also makes a related point here about timing: wage inflation only accelerated in euro crisis countries above the ECB’s stability norm at a late stage of the bubble buildup. Yes, fine, as we said, money wages are slow-moving prices. German wage moderation started around 1996 – the original exogenous shock. It became a massive problem and driver of self-reinforcing divergences with the onset of the “global slowdown” of 2001. Rather than refuting, Storm’s point about timing only reinforces the validity of the wage moderation hypothesis.

So how about Storm’s other point about capital flows and how these supposedly relate to credit and asset bubbles in euro crisis countries? According to Storm: “Cheap credit in the South created unsustainable asset bubbles and facilitated untenable debt accumulation which fed into higher growth, lower unemployment and higher wages—but (totally in line with market rates of return) all concentrated in the non-dynamic and often non-tradable sectors of their economies.” At issue here is the supposed connection between the lending by German and French banks to financial firms in euro crisis countries and any asset bubbles in these countries. What mainstream economists tend to have in mind here is the age-old but flawed idea that “deposits make loans.” Deeply ingrained with another equally flawed idea – the loanable funds theory of interest – German saving thereby supposedly financed investment in the eurozone south. It is no surprise that flawed theories yield nonsense propositions. What is true is that the expansionary drive of European banks under the euro – as also reflected in exploding gross capital flows – were key to keeping interest rates (as anchored at the short end by the ECB) low across the eurozone and supporting asset prices in general. With unhindered arbitrage and absent exchange rate risk across the currency union, banking systems’ balance sheets expanded relative to the economy; banks’ asset purchases and lending grew faster than GDP.

That does not mean though that German saving financed unproductive home properties in Spain. Spain’s cajas are quite capable of extending credit on their own (creating deposits out of thin air by doing so). Domestic credit creation was the primary driver of mortgage lending to households. For instance, Spain’s cajas did a lot of lending at the time because funding conditions seemed attractive and willing borrowers plentiful. Without ready cajas and willing mortgage borrowers, German banks can pile into Spanish government debt until they are blue in the face without ever channeling any German saving to Spain (because the German income and saving will not arise in the first place without the Spanish spending necessary to make that happen). Spain’s national financial system was amply “liquid” to readily meet any funding needs arising from local mortgage lending. External funding needs arose as a by-product of domestic credit creation as incomes and wages and, hence, import growth accelerated and trade imbalances surged. Under flexible exchange rates gross capital flows can easily be the prime movers of net capital flows (and corresponding competitiveness and trade imbalances) by pushing around exchange rates in a destabilizing fashion. Things are somewhat different inside a currency union as there are no exchange rates to push around – highlighting the pivotal role of relative unit labor cost (trends)! To repeat, at issue is whether German wage moderation was a mere consequence of divergences caused by capital flows, as implicitly assumed by the “consensus narrative,” or an exogenous cause of those very divergences, as held by the wage moderation hypothesis. Both the facts and the logic of currency union (as spelled out by Bofinger as well as Flassbeck and Lapavitsas) clearly support the latter view – as do parts of Storm’s own analysis, even if he ends up contradicting himself on numerous occasions.

Finally, let us then turn to what Storm sees as the real issue: “All this talk about labor-cost competitiveness diverts attention away from the real problem of the Eurozone: the common currency and monetary unification have led to a centrifugal process of structural divergence in terms of structures of production, employment and trade (as explained in my earlier notes). This centrifugal process has been fueled and strengthened not just by the surge in cross-border capital flows following the introduction of the euro, but also by the common currency itself as well as by the centralized and uniform interest rate policy of the ECB.” To begin with, note that once the correct connections are drawn between German wage moderation on the one hand and the ECB’s “one-size-fits-none” policy stance and capital flows on the other, the “centrifugal process of structural divergence in terms of structures of production, employment and trade” that is Storm’s main concern becomes part of the endogenous asymmetric shock ultimately triggered by German wage moderation. Without wage moderation suffocating domestic demand in Germany, these developments would not have arisen. This is why the logic of currency union requires national unit-labor cost trends to stay aligned with the ECB’s stability norm (again, as spelled out clearly by Bofinger and Flassbeck and Lapavitsas).

This does not mean that national unit-labor cost trends should stay aligned with the ECB’s stability norm under any conditions. If other factors cause competitiveness positions to diverge, national unit labor costs trends will (temporarily) need to diverge from the ECB’s stability norm to maintain balance. Storm argues that Germany has benefited from China’s rise (owing to its strength in capital goods) while the eurozone south has suffered (owing to heightened competition in their export products). Storm also argues that Germany benefited (more than its eurozone partners) from establishing new regional/global value chains that have integrated some of the new EU members (in the car industry, in particular). Germany thereby further strengthened its position in high value-added, higher tech products – while its eurozone partners (excelling in building too many new homes) fell further behind in matters of technology use and leadership. Storm suggests that EU fiscal policies should be framed to foster convergence and catching up in these matters.

These are all very valid concerns. The point is that they do not refute but strengthen the wage moderation hypothesis. For they imply that Germany experienced a positive terms-of-trade shock under the euro. Accordingly, to maintain balance inside the currency union, German wages should have grown faster than the average, not slower. Once again Storm’s argument underscores that German wage moderation provided a massively destabilizing force.

Clearly his proclamations that myths about labor cost competitiveness should be ditched are way off the mark then. It will not be possible to rebalance the currency union without also rebalancing intra-union competitiveness positions (see here). Contrary to what Storm seems to believe, it is rather irrelevant how large or small any direct spillovers from German spending in the eurozone south may be. The real issue is that eurozone members have to be able to live and prosper with one common monetary policy and one common exchange rate.

Unfortunately, they are currently still far away from that position. Perhaps the euro exchange rate may be right for Greece and Portugal today. Clearly it is totally out of whack for Germany. The view that unit labor cost competitiveness matters for intra-area balance does not imply that bilateral trade flows must balance. What matters for each member country is balancing trade multilaterally. It is this requirement that rules out – at least in the absence of a transfer union – the viability of stark intra-area imbalances in competitiveness positions; as were caused by German wage moderation. Storm is right to attack the deflationary route of attempted rebalancing currently pursued for causing vast and unnecessary collateral damages. But proponents of the wage moderation hypothesis have never suggested that this was the right way to go anyway.

[1] Peter Bofinger has recently reminded me of his “minority view” on German wage moderation as laid down in the 2004 “wise men” report. (See here, pages 716-732.)