IWH study: Regional effects of recession in Germany due to Russian gas supply cut

This study comes to the conclusion that approx. 2.7 million (or 6.1%) jobs in Germany would be "affected" by the Russian gas supply cut starting in mid-April 2022, whereas Southern regions will be more affected than the Northern ones. This result has been repeatedly used as the "proof" that embargo will lead to "millions of lost jobs" and mass unemployment. However, this interpretation is wrong for the following reasons:

1. Not all "affected" jobs will be actually lost. Since gas supply shortage (if any) would be limited to just several weeks or months, both government and companies would take measures to preserve the workforce (furlough, reduced working hours etc.).

2. Many affected jobs do not exist (yet). The study uses "baseline scenario" as the reference, which assumes GDP growth of 2.7% in 2022 and 3.1% in 2023. In other words, most "affected" jobs would be "new" jobs not being created due to lack of growth, and not the existing jobs being lost due to recession.

3. The estimates are based on (now) obsolete data. The study assumes that the gas flow from Russia would stop in mid-April. In this case, it estimates that German gas storage will be approx. 25% full in June. In fact, storage level is at 50% as of June, 4th. The difference amounts to 60 TWh, or 6 bcm, which is almost exactly the missing gas volume according to the DIW study. Hence, if we ban Russian gas NOW, there will be likely no shortage at all, given that we implement all suggested saving and diversification measures.

General

The analysis of IWH is based on the results from a larger collaborative study ("Gemeinschaftsdiagnose", henceforth: upstream study). The goal of the IWH study was investigate the regional variation of these results and compute detailed estimates on the county ("Kreis") level. Given this setup, it is expected that aggregated results (for all of Germany) from both studies should be exactly the same.

Affected vs. lost jobs

The study reports the number of affected (German: "betroffen") or threatened (German: "gefährdet") jobs. This value has been computed as the estimated reduction in working hours (compared to the baseline scenario, see below) divided by the nominal working time per employee. Obviously, this is a very simplistic way to quantify the effects on the workforce. In real world, a 10% reduction in amount of work, especially a temporary one as modeled in this study, will not result in 10% employees laid off or 10% increase in unemployment rate. Alternatively, all employees can reduce working hours by 10%. The truth is between those two extreme, i.e. some jobs will be permanently lost but others will be preserved thanks to flexible working hours and government subsidies (Kurzarbeit). Indeed, the upstream study estimates that unemployment rate will be 0.2% (2022) and 1.1% (2023) higher in case of embargo, compared to the baseline scenario (no embargo). As we can see, those numbers are much lower than 6.1% arithmetically affected jobs reported in the IWH study.

Baseline: growth scenario vs. status quo

Another source of confusion is the reference point. The language in many articles implies that the numbers from IWH study represent the number of people who have a job now and would loose it as result of embargo. Some go even further and assume that none of those people would find another job ("unemployment rate will grow threefold"). However, the baseline scenario in the IWH study assumes economic growth and hence creation of new jobs in 2022 and 2023. In other words, part of the reported losses is "virtual", i.e. it represents a missed opportunity of future improvement, but not the worsening of today's situation. And indeed, according to the upstream study, the 2023 unemployment rate with embargo would be just 0.3 percent points higher than it actually was in 2021 (6.0% vs. 5.7%).