The magnitude of medium-term consequences of a crisis is largely determined by labour market developments. Differences in European and US responses to the GDP decline have partly followed historical patterns, though unemployment has risen less than expected in Europe. In Sweden, the COVID-19 crisis carries traces of both the 1990s and global financial crises. The rapid policy response is reminiscent of the financial crisis, but labour-intensive service sectors are now also affected as in the 1990s.
This past spring, the main challenge to economists was to try to predict the size of GDP declines that lockdowns and other pandemic-related disruptions would cause. We are now entering a phase where labour market developments are of growing importance. More concretely, this phase will determine how big the damage will be further ahead.
There are many factors that determine how deep the downturn will be and how its pattern will diverge from earlier crises. How severe the underlying imbalances are, and how rapid and forceful the economic policy response is, are of great importance to the general depth and duration of the GDP downturn. How unemployment reacts to a given GDP downturn also varies significantly between different countries and crises. For example, the subsequent upturn may be smoother because many people have completely left the labour market, or because productivity falls when companies retain employees while awaiting a recovery.
Rapid responses and minor imbalances
Early in the COVID-19 pandemic, it was possible to identify some positive factors compared to earlier crises in modern times. Relief measures could be launched rapidly, without decision makers having to worry very much about negative side effects, for example in the form of distorted incentive structures. Underlying imbalances also seemed minor. Both the early 1990s crisis and the global financial crisis are usually classified as “balance sheet recessions”. In the prevailing situation, there were instead reasons to be relatively hopeful about the strength of the recovery and the effectiveness of stimulus measures, once restrictions began to ease.
Unemployment is harder to predict than GDP
Looking back, we can certainly say that forecasts of GDP growth during 2020 have been reasonably stable since May, in particularly for large advanced economies. Meanwhile the downturn in regions like the Nordics and Baltics now appears to be milder than previously feared. Despite this, the unemployment outlook has undergone major revisions, reflecting new factors changing the relationship between GDP and unemployment.
Major disparities between US and Europe
The differences between how labour markets work in the US and Europe are a recurring theme when discussing recession dynamics. In the most intensive phase of the global financial crisis, the unemployment in the US rose by about 5 percentage points. The upturn in the euro area during the same period was about half as large, although GDP fell more than in the US. Thereafter the unemployment declined in the US, while in the euro area unemployment was more persistent before the euro crisis pushed unemployment higher again.
The differences in GDP declines we are now seeing between the US and Europe are strongly reminiscent of the pattern from the financial crisis, but the differences in unemployment today are substantially larger. In little more than a month, US unemployment soared from 4 to 14 per cent. Since then it has fallen to 10 per cent in July. In the euro area, the upturn so far has been a very modest 0.6 percentage points.
Job numbers stay high, but hours worked plunge
The ECB has estimated that in May, half of the people in short-term work programmes were not working at all. In Germany, the number of hours worked in manufacturing in June was down 19 per cent from a year earlier, but the number of employees fell by only 3 per cent. In France, unemployment would triple if people in programs were included among the unemployed. However, the significance of these differences will gradually fade, and our forecast of unemployment further ahead will be dominated by more underlying factors. In Europe, unemployment will probably keep climbing as these programmes end and as companies adapt to a situation of long-lasting relatively weak demand in many sectors.
The consequences of the coronavirus crisis are also raising new questions about how labour markets are affected and what categories of people are most affected. One important dividing line concerns the connectivity of different jobs and whether people have the option of working from home, or whether their occupation requires a physical presence. The option of working from home often coincides with higher incomes, which means that weak groups in the labour market are hard hit to an even greater degree than in normal recessions.
Swedish productivity surged during the 1990s crisis
It is not unusual for crises to trigger or at least coincide with major structural shifts in the economy that affect the labour market during the recovery. Sweden’s 1990s crisis included several such phenomena. During the first phase of the crisis in 1990-92, manufacturers were pressed by a combination of international recession and an overvalued currency, for as long as Sweden could manage to defend its fixed exchange rate system. This forced them to enact tough efficiency-raising measures that marked the beginning of a period of high productivity growth in the entire economy. In the short term, this intensified job losses. The financial crisis showed a completely different pattern, with falling productivity and a modest downturn in hours worked, mainly attributable to manufacturing. The aid of economic policy stimulus reduced the secondary effects on the labour-intensive service sector.
Danger of long-term damage
Both productivity and participation rate trends underscore the importance of distinguishing between short- and long-term perspectives. When economic policy makers try to slow the upturn in unemployment, they must accept that production per hour worked may fall, or that university students will take a bit longer to complete their education, with lower labour supply as a consequence. The challenge to policy makers today is to avoid letting these side effects become permanent, for example because perpetually low interest rates reduce pressure for change in the economy and enable “zombie companies” to survive.
The current Swedish labour market situation has features in common with both the 1990s crisis and the financial crisis. As in the financial crisis, economic policy ammunition was available and was also quickly used. However, as in the 1990s this crisis is having a broad impact on the entire labour market, including the service sector.