I recently made the following point about David Leonhardt’s column on the German economy:

One thing I found odd: Leonhardt focuses on wage growth since 1985, yet it is widely recognized that wage restraint has been a big part of Germany’s success in maintaining employment levels, a phenomenon Michael Spence describes at length in his excellent book The Next Convergence. German offshoring to eastern Europe has also played a critical role, as Dalia Marin observed last year.

Matthias Rumpf asked a related question, and Leonhardt offered an interesting reply. First, Rumpf wrote:

For most observers in Germany, wage moderation that took place over the past 10 years (and which was endorsed by the trade unions) was one of the key factors behind the increase in competitiveness and the higher growth rates you are seeing now.

Then Leonhardt writes:

That’s a very good point. For space reasons, I did not delineate between two different time periods in the column, and it’s worth doing so here.

I’m not sure I’m convinced by this: if leaving out the fact of wage moderation gives a misleading picture of the Germany economy, one assumes other sections of the column could have been trimmed for space.

From roughly 1985 to 1995, German workers enjoyed considerably larger real hourly wage increases than American workers. From 1995 through the present, wages in the two countries have grown at a similar pace – growing initially and then slowing down, to a pace more recently not much faster than inflation. (These trends are visible in the second chart in a recent blog post.)

Why did German wages grow faster in the first period? Some statistical quirks involving the reunification of the country played a role. But inequality also played an important role. In the United States, much of the benefit of economic growth went to a small slice of the population at the top of the income distribution. In Germany, the gains were more broadly shared.

This suggests that rising inequality is an independent phenomenon rather than a reflection of underlying factors, e.g., uneven productivity gains across sectors and across firms within sectors. A larger share of U.S. productivity gains came from firms in knowledge-intensive service industries that embraced pay-for-productivity compensation schemes, in part because of relatively free labor markets that allowed for a trial-and-error approach to building organizational capital. One assumes that the changing demographic composition of workforce also played a role, a subject we’ll revisit below. Leonhardt tells us that the benefits of economic growth went to a small slice of the population in the U.S. It would be helpful to understand the sources of economic growth in the United States versus Germany, and the sources of income for top earners in both countries.  

Fortunately, we do have some data that sheds light on inequality in Germany from 1992 to 2001, the first period Leonhardt references, from scholars CEPR:

Compared to France and the US, the share of wage income at the top is quite small in Germany. In the US, in 1998 about 45% of all income accruing to the top 0.01% consisted of wage income; for the correspondinggroup in France the share was about 22%. Thus, our analysis adds a novel aspect to the comparison of Germany with the US and France. Previous research showed that, with respect to the concentration ofincome, Germany is a middle case between the highly concentrated US income distribution and the less concentrated French one. With respect to the income composition pattern, our analysis shows it is France which lies between the US and Germany. The German affluent rely much less on wages and salaries for their incomes than their counterparts in France and the US. However, the increasing wage share of the German affluent corresponds to recent developments in the US where increasing income inequality was largely driven by an increasing share of wage income in the top percentile of the income distribution. [Emphasis added]

To clarify, the German affluent were relying more heavily on capital income. Wages and salaries represented a smaller share of income, due in part to marginal tax rates and other work disincentives.

Moreover, as the CEPR report goes on to observe, inequality increased throughout this period in Germany:

Comparing the evolution of the mean and the median shows that income inequality has markedly increased.Whereas real mean income remained virtually constant between 1992 and 2001, median income fell by almost 25% in this period. As a result, the difference between the mean and the median increased by more than 60 percentage points. This is mainly related to an increasing number of people with very little or no market income, which pulls down the median. Evidence of a broad increase in income equality can be seen by looking at the extremes. In 2001, more than 40% of total market income accrued to the top decile,and its share increased by 7.3% between 1992 and 2001. At the same time, the income share going to the middle of the distribution declined. We suspect that this may be due to compositional effects, in particular the significant increase in unemployment in the period 1992 to 2001, especially in East Germany. [Emphasis added]

Leonhardt was definitely not wrong to say that inequality was higher in the U.S. than in Germany during this period. But it might have been helpful to note that at least part of the difference reflects the fact that the German affluent are, relative to the American and French affluent, a non-working class. Might Germany have been richer had its highest earners had work incentives? I think the answer is yes. Might it have also been more unequal? The answer is very clearly yes. Do we have any reason to believe that higher income gains at the top would have reduced incomes at the median? I don’t think we do. 

Leonhardt continues:

And do the similar wage gains in the second period mean that workers in both countries have done equally well? No, German workers have still done better, because German job growth has been stronger.

So the typical worker who remained employed in Germany from the mid-1990s through today has received a pay increase comparable to that of the typical worker who has remained employed in the United States. But the odds of an American worker’s being unemployed have risen substantially over the last 15 years. The odds of a German worker’s being unemployed have fallen.

There are a few questions I’d want answered before, e.g., are there meaningful differences in the age composition of the workforce of both countries? The Stanford Center on Longevity tells us that while the U.S. workforce has experienced moderate growth, the Germany workforce has shrunk over the period in question. This could have no significance at all in shaping the economic environment, but that seems unlikely. 

My guess is that Leonhardt believes that the U.S. financial crisis was endogenous, which is why he considers the comparison in employment levels a meaningful reflection of policy. I think that’s fair. Texas and Germany pursued wiser housing policies than the rest of the United States, and that, among many other things, contributed to a financial crisis that has had a powerfully negative impact on employment levels. I do wonder if the German labor market would be immune to a eurozone financial crisis, which may well happen in the near-term.   

All that said, I’m glad that Leonhardt has brought the German model to the attention of his readers, not least because it has often been praised by conservatives.