As has often been argued here (link1, link2, link3), the ultimate reason for the euro crisis is the divergence of inflation rates since the foundation of the EMU. In all debtor countries, inflation rates have been significantly above average, while in typical creditor countries like Germany, Finland and Austria, inflation rates have been below average, as figure 1 shows. Since the establishment of a common central bank caused a convergence of nominal interest rates, diverging inflation rates imply diverging real interest rates. As a result, in a monetary union high inflation countries have an incentive to run up debts, while low inflation countries have an incentive to build up savings.
Figure 1 – Divergence of Prices in EMU Member States
– Click to enlarge –
The results are current account deficits and growing negative net investment positions in high inflation countries, and current account surpluses and growing positive net investment positions in low inflation countries. If debtors in high inflation countries use a part of their credits to buy nontradables, like real estate and services, excess demand in high inflation countries can result and – since only a part of the credits flows back in terms of demand for tradables – excess supply can result in low inflation countries. Consequently, the divergence of inflation rates can become a self-reinforcing process and trigger debt- and respectively credit-spirals. To my understanding, this is a good argument to fight diverging inflation rates between the member states of a currency union. However, as often in economics, one may question various aspects of the argument, like the empirical size of inflation and real interest rate differentials or hint to demand side, demographic or structural factors that can also cause current account disequilibria and so on.
However, from figure 2 follows an argument for the necessity to avoid diverging inflation rates between member states of a currency union, which is nearly compellent – at least if you accept that it is not possible to change the wage share of GDP in the long-run without causing severe economic and social problems. Figure 2 shows that the period of deterioration of competitiveness, i.e. the period of increasing unit labor costs, of EMU debtor countries compared to Germany from the foundation of the EMU in the year 1999 to the year 2007, before the outbreak of the crisis, was at the same time a period, where all of these debtor countries held their wage share of GDP nearly constant (at least compared to the „normal“ time series behavior of wage shares). After the year 2007 the wage share increased, because the recession caused GDP to fall, while wage contracts are typically relatively rigid in the short-term. Since the year 2009, the wage share is again falling back – mainly caused by dismissals.
Figure 2 – Divergence of Labor Costs and Constancy of Labor Income Share
– Click to enlarge –
But how was it possible that the wage share of GDP stayed constant while unit labor costs did grow that strong? A little algebra shows that an increase of unit labor costs is is not only possible, but even necessary to keep the wage share of GDP constant if a country suffers from high inflation rates. Constancy of the wage share between period t and t+1 implies the following equation to hold:
where wt equals the nominal yearly labor compensation per employee in period t, Lt equals the number of employees in period t, Yt equals real GDP and Pt equals the price level of GDP. From this equation follows directly the growth rate of the nominal wage rate that keeps the wage share of GDP constant between period t and t+1:
Consequently, the nominal wage rate has to grow by the inflation rate times the growth rate of labor productivity. Constancy of unit labor costs between period t and t+1 implies the following equation to hold:
Consequently, in order to keep unit labor costs constant, it is necessary to keep the growth rate of the nominal wage rate equal to the growth rate of real labor productivity – and neglect the inflation rate. As a consequence, if there is inflation, Pt+1 / Pt > 1, the growth rate of the nominal wage rate, which keeps the wage share of GDP constant, will necessarily lead to an increase of unit labor costs. This however means that in a monetary union, where some member states suffer from higher inflation rates than others, and where labor market institutions in all member states try to keep the wage share of GDP constant, the unit labor costs of high inflation countries must deteriorate compared to unit labor costs of low inflation countries.
Unfortunately this is a big problem in a monetary union, because, unlike in a system with flexible exchange rates, in a monetary union unit labor costs matter. In a system with flexible exchange rates, the purchasing power parity mechanism causes the nominal exchange rate of high inflation countries to depreciate against the currencies of low inflation countries – at least in the long-run before things get too ugly. Hence if unit labor costs in high inflation countries grow stronger than unit labor costs in low inflation countries, but the currency of high inflation countries depreciates exactly by the difference between inflation rates, e£$t↓ = P$t↓ / P£t↑, unit labor costs of high inflation countries measured in foreign country currencies will stay constant. And unit labor costs measured in foreign currencies is what matters for the „international competitiveness“ of a country. However in a monetary union, no such exchange rate mechanism exists. Therefore, if nominal unit labor costs in Ireland grow stronger than in Germany, the real (productivity adjusted) price for labor in Ireland compared to Germany grows – such that the production of goods and services in Ireland gets relatively more expansive compared to Germany.
To avoid the problem stemming from diverging unit labor costs, two possibilities exist: Either the central bank of a monetary union takes care that the price level in all member states grows, at least in the medium term, with the same pace or workers in high inflation countries accept a decreasing wage share of GDP – and decreasing real wages, if the rate of inflation is larger than the growth rate of real labor productivity. Figure 3 displays the hypothetical wage share, necessary to keep the unit labor costs equal to their 1999 level. Compared to its actual wage share, Germany should have increased its wage share by 1,4% up to the crisis year 2007, in order to avoid the reduction of its unit labor costs that had actually taken place (and made the life of its partner states in the EMU even a little bit worse than it would had been anyway…). All debtor countries should have decreased their wage share (real wage rate) by the year 2007 dramatically, in order to keep their unit labor costs at their 1999 level: Ireland by 10,4% (12,7%), Greece by 7,6% (5,3%), Spain by 13,3% (26,7%), Italy by 6,9% (17,5%) and Portugal by 10,3% (15,8%). It is highly questionable, whether such a strong decrease of real income for employee households would have been socially acceptable and politically manageable. And in the presence of continuing inflation differentials between member states, this process of decreasing wage shares must go on and on!
Figure 3 – Hypothetical Wage Share of GDP and Real Wages Necessary to Keep Unit Labor Costs at Their 1999 Level
– Click to enlarge –
Therefore, I conclude that the only viable solution to avoid the problem of diverging unit labor costs in a monetary union, is an equalization of inflation rates by the central bank. The European Central Bank (ECB) must „redefine“ its understanding of „single monetary policy“ not as „uniform refinancing conditions“ for all commercial banks of the union but as „uniform inflation rates“ across all member states of the union. The refinancing conditions for the commercial banks must be differentiated accordingly. As I have argued here (based on proposals by Black (2010), Palley (2006) and Holz (2007), this could be done by differentiating either the main refinancing rate or the minimum reserve requirements according to the inflation rates in the countries to which commercial banks lend money. For credits flowing to countries with high inflation rates the main refinancing rate or the minimum reserve requirement must be correspondingly higher than for credits flowing to countries with low inflation rates, such that the inflation process becomes mean reverting, with the ECB target inflation rate as mean for all member states.
This adjustment of the ECB strategy seems to be necessary in order to keep the EMU together – after the current crisis is overcome. Another question is, what does the above analysis imply for the current crisis? I think the idea that the inflation process must become mean reverting with the same mean for all member states, can be applied to the current crisis too. Since the ECB has not managed to make the inflation process mean reverting so long, this must be made up now over the next couple of years. Accordingly, the ECB has to take care that up to now low inflation countries like Germany, Austria and Finland become high inflation countries, and up to now high inflation countries become low inflation countries until the diverging index time series in figure 1 converge. As the example of Ireland shows, this is possible. The price index of Ireland has already converged from above to the EMU average. Now countries like Germany, Austria and Finland must do the same – from below.
This may cause some „mental discomfort“ for a some Germans (Personally, I think that the discomfort cannot be much larger as the discomfort during second half of the last football match against Sweden. Hey, the first half was really quite good !). But that is not an economic problem. German savers have been privileged since the beginning of the EMU, while savers in the current crisis countries had to suffer from (to use a now quite fashionable term) „financial repression“ as figure 4 shows (explanation of data). Now this process has to be reverted in order to restore the symmetry. It is an economic necessity, but I think, from a quite common hold normative point of view, this kind of reciprocal treatment can also be regarded as fair.
Figure 4 – Average Yearly Real Returns in Germany compared to EMU Crisis Countries
– Click to enlarge –