Economics and politics - comment and analysis
24. January 2021 I Heiner Flassbeck I Business Cycle Analysis, Economic Policy, General, Statistics

The riddle of German employment in the year of coronavirus

Cross-posted from Brussels Morning (

Last week, the Federal Statistical Office in Germany presented its first estimate for 2020’s gross domestic product (GDP). The statistical service forecasts a decline in economic output during the year of the coronavirus at 5%. In its press releases, the Office does not tire of emphasising that this is a less severe slump than 2009 — the financial crisis’s reference year — when GDP fell by 5.7%. Once again, statisticians confirm the federal government’s position, reiterating the same point it made last autumn.

Smaller crisis, bigger unemployment

But it is another excerpt of the press release that we should take note of:

“The 2020 economic output in 2020 created on average 44.8 million positions of employment in Germany. That was 477,000 people or 1.1% fewer than in 2019. Due to the Corona pandemic, this ended the 14-year increase in employment that had even outlasted the financial and economic crisis of 2008/2009”.

This outcome is astonishing. The economy shrunk less in 2020 than it did in 2009. Yet, the number of people in work has decreased in absolute terms in 2020 (by 1.1%), whereas it continued to rise in 2009 (by 0.2 %).

What happened to produce such a result, especially in those first few months of the coronavirus outbreak? According to the Federal Bureau of Statistics, the short but severe slump, triggered by the first lockdown, led to a decline in employment not yet overcome.

Growth and employment

One can illustrate the result more clearly by comparing the average annual results of the financial crisis years with those last year (Figure 1). To see the actual labour market effect, GDP has to be compared with the number of hours worked. It is obvious how closely the development of employment is related to the number we are used to calling “growth”.

According to official estimates, the last two recessions are completely different as regards to the labour market. In 2020, working hours and GDP growth was synchronised. In 2009, there was aso a massive slump in hours worked, dropping 2.8%. However, in 2009 the drop in growth and employment were not synchronised: as noted, growth was 5.7% while productivity dropped by just under 3% (5.7 minus 2.8).

Figure 1

Declining labour productivity in a recession means that companies are keeping more employees than they actually need to cope with a sharp drop in production. They may do this for a number of reasons. One is down to companies’ expectation that the recession will be short after which they will desperately need scarce labour, so they avoid layoffs and use only short-term work. There may also be technical reasons for having more workers temporarily on the job than is justified from the production side. In any case, it means a decline in unit profits for companies if wages remain largely unchanged but labour productivity falls significantly due to the crisis.

But then comes the fabulous year 2020, which seems to turn all these findings on their heads. Although GDP “only” falls by 5%, the number of hours worked falls by a whopping 4.8%, which means that the productivity effect in this crisis year is very close to zero. How can this be explained? Although at the beginning of the crisis companies could count on the loss of production lasting only a few weeks and they knew from the outset that they would receive generous support from the state to cushion their losses, they relied on laying off their employees to a far greater extent than ever before.

One has to imagine that in 2009, during the global financial crisis, the results of which no one could realistically assess, the number of people in employment still rose by 0.2%, whereas in 2020, it fell by 1.1%. Did the companies lay off employees at the drop of a hat during this crisis, even as they enjoyed numerous state safeguards ranging from short-term work to insolvency protection?

How to interpret the odd phenomenon?

Given the very different developments in the relationship between growth and employment in the crisis of 2009 and 2020, one has to ask whether the Federal Statistical Office really managed to capture the decline in value-added correctly last year. If the decline in output was, in fact, much greater, one could say that all is normal in the relationship between growth and employment.

Perhaps we should pay closer attention to the dynamic of specific sectors. There is evidence suggesting that conditions in catering, retail and transport sectors are not comparable with those in industry. In this crisis, many companies were affected by the collapse (or even a total loss) of demand, which, like restaurants and hotels may have experienced a much smaller collapse in 2009 or none at all.

The companies in these sectors certainly have a completely different employment structure than industry. There is presumably a much lower degree of employee commitment to the company here, because, for example, there are fewer collective agreements that protect against quick dismissal. In addition, the workforce is likely to be less well trained, in general, and, according to the entrepreneurs’ assessment, easy to replace at the end of the crisis.

This in turn means that the coronavirus crisis, far more than the economic downturns before it, has affected and continues to affect the most socially vulnerable groups. The new ongoing lockdown that began in early November in Germany is again centred on the hospitality sector. Consequently, one has to expect that this sector in particular, where short-term employment is prevalent, will not bring the desired results if politicians do not quickly succeed in keeping the businesses afloat. But that is precisely what is written in the stars, now that it is once again unclear what subsidies are available to companies in the sector.

Policy implications

Add to this the fact that there are still no clear political guidelines for the current month, and many businesses are still up in the air. A company’s insolvency means that employees lose their job and fall back to the lowest social protection net (Hartz IV) after only one year of unemployment. The negative political consequences are incalculable if German politics (including large parts of the opposition) do not finally recognise that Hartz IV is a completely inadequate response to the challenges of structural change and deep production crises. Fundamental reform here is the central key to building trust in democratic institutions.

Whichever narrative one adopts to account for the discrepancy between GDP and employment in Germany, there can be no doubt that the past year’s shock has had a huge negative impact on unemployment that will linger for a long time. It is unemployment rather than government debt that is the true cost of the unfolding crisis. For a region like Europe, which has already failed in the ten years after the financial crisis to realise an economic development that provides full employment that is an incredibly heavy mortgage. Germany should finally admit that its insistence on its mercantilist variant of permanent current account surpluses holds back a successful European strategy to exit the crisis.