Fact-checking Kaput by Wolfgang Münchau

[Estimated time: 12 mins]

Wolfgang Münchau’s Kaput covers a lot of ground on a subject I knew almost nothing about. On my first read, I came away surprised at how much of a corrupt, dysfunctional basketcase Germany seemed to be. That reaction made me suspicious, so I fact-checked the book more than I normally would.

What I found was that Münchau is sloppy with his facts. Some of his errors are immaterial, but some are not. The first section below covers claims that seem to be plain wrong, or misleading at best. The second covers claims that are more arguable. On some of these, Münchau might be broadly right, but still left out context that I think a reader needs.

Wrong or misleading

Decline of German private banks

Münchau claims that the German banking sector has been going downhill since the global financial crisis in 2007. He writes:

Münchau is badly wrong here. By asset value in 2023, Deutsche Bank was the 27th largest bank, another German bank called DZ Bank was 53rd, and Commerzbank was 69th. I don’t know where Münchau got his 729th and 1,132nd rankings from, but it’s probably from a ranking of all listed companies, not just banks. So Münchau’s implication that Deutsche Bank has fallen around 700 places in rankings is wrong — it’s actually been just 22 places.

Münchau’s claim that, twenty years ago, there were 4 German banks in the top 20 rankings is also a little off. Looking at a 2003 ranking, there were only 2 German banks in the top 20, though 4 made the top 30.

STEM graduates

Münchau implies that there’s been a long downward trend in German STEM graduates:

I have no idea where Münchau gets this. In 2025, 35% of German graduates were from a STEM field, which was the highest share in the OECD. The OECD average by comparison was just 23%. Perhaps Germany’s share was lower than the UK’s in the particular year Münchau looked at, but (1) I don’t know why the UK, rather than the OECD average, is an appropriate benchmark; and (2) clearly Germany has not experienced a long downward trend in STEM graduates as Münchau implies.

AI bet on the wrong technology

Münchau suggests that the €3 billion set aside in 2018 under Germany’s AI strategy was “a bet on the wrong technology”, because it was based on outdated approaches like expert systems rather than the machine learning and deep learning technology dominant in AI today.

However, while the AI Strategy document mentions expert systems, it clearly also recognises the potential of machine learning. Moreover, it explicitly declines to bet on individual approaches (at p12):

And when it came to the actual funding, several of the flagship centres appear to be machine learning centres.

Germany’s “green card” immigration scheme

In 2000, the German government introduced a German “green card” immigration scheme, with the hope of attracting more migrants in the IT sector. The government set a target of 20,000 cards. Münchau claims that by 2003, only hundreds had been issued:

This claim is way off. This policy brief states that by July 2003, almost 15,000 green card permits had been issued. Another source says that the scheme brought “just fewer than 18,000 IT professionals” into Germany by December 2004.

Global survey on expat life

Münchau writes that:

Münchau’s claim here is sort of correct, but pretty misleading. He seems to be referring to the 2022 and 2023 versions of this InterNations Expat Insider Survey but, instead of directly citing those surveys, he cites a secondary source (“Manager-Magazin“). The full article is paywalled, but relying on secondary sources when the primary is easily available is just bad practice.

Now, Germany did come last in both years in the “Expat Essentials” sub-index in that survey, based on its scores for Digital Life, Admin Topics, Housing and Language. However, Germany did not come last in the overall survey, though it’s not far from the bottom (42nd out of 52 in 2022, and 49th out of 53 in 2023).

What’s of greater concern to me is what Münchau tries to use the survey to show. His argument is that Germany is failing to attract the high-skilled migrants it needs to diversify away from its existing economic model. But look at the InterNations top 10 in 2023: Mexico, Spain, Panama, Malaysia, Taiwan, Thailand, Costa Rica, the Philippines, Bahrain and Portugal. Many of these countries do well at attracting retirees and remote workers, because of their low cost of living and warm climate.

Meanwhile, Münchau neglects to point out that in 2022 and 2023, Germany ranked 12th and 13th (respectively) on the “Working Abroad” index, which measures things like career prospects, salary, and work satisfaction. Thailand, Costa Rica, Portugal and the Philippines, which rank highly in the overall survey, all rank poorly in this sub-index. Being comfortable for foreign retirees and remote workers is therefore an altogether different thing from being attractive to long-term skilled workers.

If Münchau had cited the InterNations survey only to show Germany’s poor rankings on Digital Infrastructure, Admin Topics, and Language, I probably wouldn’t have had any objection.¹I’ve omitted Housing here, because I think Germany’s low score here partly reflects it being a rich country: expats in poor countries can often earn several-fold a local’s income, which distorts the housing scores on surveys like this. But as it stands, Münchau’s portrayal is rather misleading.

Lacking in context or nuance

Reforms to conflict of interest rules

Münchau claims that it is common for senior ministers to become lobbyists after their careers end in Germany and cites a couple of examples, with Gerhard Schröder being the most egregious.

What he neglects to mention is that Germany tightened up its rules after 2015. It’s certainly debateable whether the current rules go far enough, but he should have at least mentioned it. Otherwise, the reader can come away thinking (as I did on my first reading) that Germany is more of a basketcase than it actually is.

Nationalisation of a steel company

Münchau writes:

Münchau doesn’t mention that the Salzgitter nationalisation was only temporary, which I found a little misleading. The company went public a few months later, and the state sold half its stake for a profit (source: Tagesspiegel). From what I can tell, Lower Saxony still owns about a quarter of the company.

The mention “lucrative orders from Russia” also lacks context. Münchau seems to be referring here to the contracts to build pipe sections for both Nord Stream 1 and 2, which its 50-50 joint venture Europipe won. Europipe won those contracts after a competitive tender, and Schröder chaired the committee that make the tender decisions. But Europipe was an obvious bidder for such contracts, because there simply aren’t that many companies that can build such pipes. I don’t mind Münchau raising questions about this — maybe there was foul play, I simply don’t know. My issue is just that he doesn’t give the reader the full context.

Exposure to trade

(This one is a bit more pedantic.) When discussing Germany’s current account surpluses, Münchau writes:

I checked the figures here and got slightly different results: 89% (Germany), 63% (world average), 27% (US), and 37% (China) (the UK figure was the same). But the figures are similar enough that I assume the discrepancies are due to later revisions.

The bigger issue here is that the foreign trade ratio isn’t a clean proxy for exposure to trade. Both exports and imports are gross flows while GDP only measures value added. Imported components that are re-exported are therefore counted twice in the numerator (once as an import, and once as an export), while only the domestic value added registers in the denominator. For example, Singapore’s foreign trade ratio is over 300%!

It’s therefore possible for a “less exposed” country to have a higher foreign trade ratio than a “more exposed country”. To take a simplified example, assume Country A and Country B both have a GDP of 100. Country A imports 40 in components, processes them, and re-exports them for 45. Its foreign trade ratio is 85, but a trade shutdown would cost it only 5. On the other hand, Country B could import 20 in consumer goods and export 20 in domestically grown produce. Its foreign trade ratio would be just 40 — less than half of Country A’s. Yet a trade shutdown would cause a 4x greater GDP hit in Country B than in Country A.

To clarify, it’s still possible that Germany is more exposed to trade than other comparable countries. I don’t know. My point is that the foreign trade ratio alone is a poor proxy for trade exposure.

Conclusion

This was an interesting exercise in reading a book on a subject I know almost nothing about. It’s very tempting in such cases to absorb an author’s framing wholesale. The case of Kaput shows what a mistake this would have been.

Münchau’s errors consistently made Germany look worse than it is. His overreliance on secondary sources (like newspapers) over primary sources seemed to drive much of the sloppiness. My overall impression was that he reached for evidence selectively, to fit a conclusion he’d already reached, rather than the other way around.

None of this proves that Münchau’s broad diagnosis is wrong. It’s entirely plausible to me that Germany’s economic model is outdated and that its political system is too captured by industry to change course. But I lost enough trust in Münchau over the course of this exercise that I’m just going to file Kaput away as just one opinionated guy’s account of Germany’s troubles for now. I will have to read more on Germany and Europe before I can form a better view myself.

Let me know what you think of this post in the comments below! If you have other books you’d recommend for learning more about Germany or European economic policy, please share them.

• 1
  I’ve omitted Housing here, because I think Germany’s low score here partly reflects it being a rich country: expats in poor countries can often earn several-fold a local’s income, which distorts the housing scores on surveys like this. ↩︎