The German Turnaround: The reform story revisited, part I

(in cooperation with Ulrich Fritsche)

With politicians from right and left in Germany discussing a possible new extension of the duration of unemployment benefits for the elderly, a new debate in Germany has started concerning the reasons of the recent upswing. The central question of the argument is: What role did the labour market reforms of the Schröder government play for the “Teutonic turnaround” as Michael Burda calls the recent development in his first post for this blog?

While the German mainstream is now mainly crediting the “Agenda 2010” reforms for the recent upswing along the lines of Michael Burda’s analysis (although, in contrast to Michael, at the time of the reform package being discussed in the Bundestag most of them criticised the bills as being not far-reaching enough to have a feelable impact on economic growth or unemployment), there are dissonant voices from abroad: FT columnist Wolfgang Munchau calls the idea that the reforms contributed to the recent upswing as “nonsense”. David Soskice from Oxford and Wendy Carlin from London also question this assumption in a very careful comparison between the German and the British experience (published as CEPR discussion paper).

The discussion on the sources and driving forces of the turnaround in Germany is of great interest. First, the German case will surely be referred to as a benchmark for further reform programs as it is one of the few natural experiments we have for labour market reforms in large economies. Second, the analysis of the origins of the German turnaround is of great interest to investors. If the turnaround was mainly due to the reforms, a partial scaling back of the Agenda 2010 as it seems to be in the cards now would mean a danger of Germany falling back into stagnation with corresponding risks for German assets.

In a mini-series of three posts, Ulrich Fritsche (another contributor to RGE Monitor’s European EconoMonitor) and me will confront the story of the Agenda 2010 being the jump starter for the German recovery with some recent empirical data on economic developments in Germany.

We will start today with a look at the German labour market. In a second part over the coming days we will disect the impressive gain in German stock valuations since 2003 and will ask how plausible it is that this gain comes from a reform dividend. Finally, in a third post, we will propose an alternative explanation for the German turnaround.

While the Agenda 2010 was a potpourri of unrelated small reforms in many areas (dropping the regulation of a number of small crafts; speeding up some income tax cuts already decided by one year; closing some tax loopholes), the main thrust was the labour market: The duration for the receipt of unemployment benefits for the elderly was cut and the unemployment assistance for those being unemployed for more than one year was completely scrapped, pushing long-term unemployed into the welfare system. For the long-term unemployed this meant not only significant cuts in benefits, but also that they had to run down their savings as the new benefit was means-tested (which the old unemploment assistance was not). In addition, rules for temporary work were relaxed. Another key element was the reform of the Federal Unemployment Agency (“Bundesanstalt für Arbeit”) supposedly to enable the agency to bring unemployed into work more easily.

For most economists, the reform of unemployment assistance was the most important (and among the population the most unpopular) element of the reform. According to supply-side analysis, the cut has increased the incentives for the long-term unemployed to seek a new employment. As most German economists believe than unemployment in Germany was mainly (if not exclusively) structural as opposed to cyclical, this is seen as being at the root of the turn-around in the labour market.

Empirically, however, there is a problem with this analysis: If the increased incentive for the long-term unemployed to seek employment were behind the turn-around of the labour market, unemployment among this group should have fallen most.

However, this is not the case – not by far. The number of unemployed in Germany has fallen by an impressive 1.1 million to 3.7 million over the past 24 months. Over the past year alone, the number of unemployed dropped by 700,000. Since the fall of 2005 (two years ago), unemployment among the short-term insured unemployed (those who did not experience any cuts in benefits) has fallen by almost half from 1.33 million to only little more than 700,000. In contrast, the number of long-term unemployed in the welfare system has fallen by a little more than 10 percent from a little more than 2.8 million to 2.5 million now. Moreover, the drop in the number of long-term unemployed reacted later to the upswing than that of the short-term unemployed. Figure 1 below taken from the labour agency’s recent monthly bulletin illustrates the point: The number of those receiving benefits for short-term unemployment (“SGB III”) has fallen much faster than those on welfare (“SGB II”).


Of course, the strong drop of short-term unemployment is what you would expect in a cyclical upswing. However, it is not what you would expect in an upswing induced by increased incentives of the long-term unemployed to seek employment, but rather what you expect in any cyclical expansion of aggregate demand.

However, theoretically there might be another channel which can help to explain how the labor market reforms might have stimulated the turnaround: The scrapping of unemployment assistance might have made unemployment more painful thus diminishing the wage bargaining (monopoly) power of trade unions in labor markets. This might correct some insider-outsider bias at work in labor markets, which for a long time might have led to excessive wage demands and destroyed jobs. The resulting wage restraint not only redistributes the earnings of the boom differently by increasing the profit share but helps to invest in labor-intensive technology. In fact, this is the argument the Kiel institute makes in its analysis of the upswing.

Undoubtedly, there has been impressive “wage restraint” in Germany. However, the claim that it was induced by the reforms is rather weak as it has started much earlier – and long before the announcement of any reform package. The following figures illustrate this point. First in figure 2, we plotted nominal wage increases per employee of the total economy (sorry for not using hours but these data are not consistently available for our reference groups) for Germany vs. EU 15 (Source: AMECO). To smooth the fluctuating data and extract a trend, we used a regular Hodrick-Prescott filter (to avoid end-point problems, the data were extended using the forecast of the DG ECFIN). There is clear evidence that wage increases have been falling behind the EU 15 average for a decade.


The same conclusion can be drawn if you look at the nominal wage per employee of the total economy calculated as an index relative to reference groups (figure 3) — the source again being the AMECO data base.


What matters of course for competitiveness, however, is the nominal wage increase relative to productivity as measured by unit labor costs. Figure 4 shows the relative nominal unit labor cost index of several countries relative to the Euro area average (1998 = 100), the calculations are explained in a paper by Dullien and Fritsche (2006). Again, it is impressive to see how Germany steadily improved its competiveness over 10 years in a row — whereas other countries in the Southern cone of the Euro area show a steady decline over the same time span. But there is no evidence whatsoever that this trend accelerated after March 2003 when the Agenda 2010 was introduced (or after 2005 when it came into effect).


Thus, in our opinion, the empirical case is rather weak that the Agenda 2010 reforms of the Schröder government actually have triggered the recent turnaround in the German economy.

Read next: Why the development in German stock prices does not tell much about a German reform dividend

This post was co-posted at Eurozone Watch.

7 Responses to "The German Turnaround: The reform story revisited, part I"

  1. Nouriel Roubini   October 15, 2007 at 11:00 am

    Sebastian and Ulrich: this is a most excellent and interesting contribution to this debate. It seems that the issue of German structural reform is one of the hottest policy issues in Germany and Europe: Burda, Munchau, FT, you folks and many others are debating it. I look forward to reading part 2 and 3 of your arguments. Nouriel

  2. Anonymous   October 15, 2007 at 11:23 am

    Indeed very interesting! In fact, it would be very interesting to get to know why no one saw the “Teutonic turnaround” coming. Even the IMF repeatedly underestimated the strength of the German recovery. It seems that the German economy is very difficult to understand from the outside.

  3. Guest   October 15, 2007 at 12:12 pm

    Is there a trade-off between external competitiveness and domestic purchasing power? Could German wage restraint beyond the productivity threshold actually be the cause for its anemic growth performance over the past decade?

  4. Guest   October 15, 2007 at 2:24 pm

    “Thus, in our opinion, the empirical case is rather weak that the Agenda 2010 reforms of the Schröder government actually have triggered the recent turnaround in the German economy.”

    So I presume you agree with Munchau more than you agree with Burda. Does this mean that the reforms were useless and that further reforms are not necessary?

  5. Sebastian Dullien   October 15, 2007 at 3:23 pm

    That the reforms of the Agenda 2010 have not triggered the upswing does not mean that Germany does not need any reforms. The health system is a mess which would badly need a reform. Parts of the education system surely need an overhaul. Competition policy in the energy sector could be stricter. The system of federal fiscal redistribution is still a mess. As for any OECD country, you could easily construct a wish list of reforms. If they are necessary or sufficient for faster growth in the short or medium term, however, is a completely different question. I see the biggest problem in the badly balanced nature of the upswing, with consumption hardly growing and the current account surplus reaching 5 percent of GDP.

  6. Guest   October 15, 2007 at 8:06 pm

    Very good points and interesting analysis. Wage restraint may have started before such structural reforms. But this does not prove necessarily that labor market reforms did not contribute to sustain such wage restraint and thus competitiveness restoration. Those reforms were well overdue when they were started and ensured continued wage moderation even after the economy recovered and labor markets tightened once the unemployment rate started to fall.

  7. Ulrich Fritsche   October 16, 2007 at 1:59 am

    Very good point. Indeed, it is one argument that the wage moderation was sustained by the labor market reforms and in that sense helped to create conditions favourable for the turnaround. However, having followed the German debate for a while, I am a bit sceptical if there is a strong connection (if there is one at all). Take as an example the turnaround in 1999/2000 — the two years with high growth rates in the decade before the structural reforms. Even in that period, unit labor costs did not explode but remained very moderate. This indicates that other forces were at work before. Important explanations behind the wage moderation are in my interpretation: the abolishment of wage agreements between trade unions and employer’s associations on a large scale in the East (and to some extent in the West) which gave signals and weakened the trade union position, the loss of members on the side of the trade unions, a lasting slump for a couple of years — and last but not least a change in the arguments used in the trade unions. On the other hand, the wage moderation process in an stagnatory environment led to very low inflation — for some years significantly below the Euro area average — which in itself (assuming a short-run Phillips curve) is costly. Internal demand was quite weak for a couple of years — which enforced the wage moderation effect.