IT Skills Gap in Germany – a polemic

[You know that an article will be a fun experience when it starts with a triple disclaimer.]

[edit: The initial draft of this article came as more beligerent than it appeared at time of writing and some of the “haha-funny” moments on a second reading fell down flat. So I did some minor content edits – that while still preserving the polemic tone and the main arguments of this article, do not distract of the notion that we are all in this together].

Disclaimer #1: The views presented in this article are strictly my own and might change a day, a week or even a decade after the publication of this article.

Disclaimer #2: While I did talk to a lot of people before finally writing this article, it is not by any means corporate or academic research. It is just a guy from the Internet comparing notes with their extended network.

Disclaimer #3: I am not an HR professional nor an executive. I am speaking from the viewpoint of a foreign engineer that HAD to relocate to Germany. Furthermore, I am speaking from my personal viewpoint (see #1 again).

So, let’s start, shall we?

A lot of talk and conversation happen in the DACH region, especially in Germany, about a skills gap that threatens the economy. This gap is exasperated in the IT sector as well – horror stories that companies have open positions unfilled for months, if not years are plenty, executives formulate plans to address this gap, the government changes laws and makes special deals about skilled immigrants and academia makes research on possible ways to address this issue. Recently, I was in a roundtable where some academic research was presented on the subject, followed by an open conversation on ways to address this problem. My observations and ideas were vehemently opposed, so I am grabbing this soap box to elaborate on them further.

Let’s start with the big elephant sitting in the room and nobody is talking about. Take a deep breath and here we go: 

GERMANY IS NOT ATTRACTIVE FOR HIGHLY SKILLED FOREIGN TALENT.

Before the pitchfork and torches brigade appear, this is NOT a personal viewpoint – in March 2023 OECD published publicly research (OECD research here) about the most skilled workers friendly countries. Sweden is #2 and is the only EU country represented in the top 5 category for workers and #1 for entrepreneurs (again, the only EU entry in the top 5). So, we are onto something here.

The OECD paper is giving a set of metrics, based on which they reached their index. I will not go into discussing these, more than reasonable, metrics. Instead, I will give my impression on what works and what does not.

The first major barrier is the language barrier. German, as opposed to English, is not a lingua franca of the Western world. Picking up German can be difficult and is not a trivial exercise. So, expecting foreign workers to speak German from the “get go” is not realistic. “But, my dear blogger, we have an English speaking workplace, we are modern, we have adapted” I can hear some of you digress. While this might be the case for a number of workplaces, it leaves out a number of important factors, the first of which is that (gasp!) THERE IS LIFE OUTSIDE OF WORK. A person with a limited German vocabulary WILL struggle in day-to-day life, as English does not work everywhere, including the infamous public sector. To give an example, your doctor most likely will speak English but what about when you call to make an appointment and the person on the other side does not speak English, as is usually the case? What happens when you receive a formal letter, as is too common, in German only, with a non-binding English version provided, not even as a courtesy? I compare this to my experience living in the Netherlands where English was a given everywhere. The language barrier has another, more sinister, dimension though. The higher up you go in a German company’s hierarchy, the bigger the unmentioned expectation that you will be a German speaker is. While there are some commendable D&I initiatives in certain places, this applies to a number of places. This limitation in upwards mobility is discussed in online communities and deducts points from the overall attractiveness. Acquiring a new language, helpful as it might be and recommended as it is, can take a significant amount of time, time that a person might not have, owing to work obligations.

The second aspect that makes things unattractive are outdated at best, broken at worst HR processes. In Germany, it is common practice in a lot of sectors, including IT, to ask for a Zeugnis document. For those unaware of these documents, these are documents produced by HR when someone exits the company and, while on the surface appear positive, a once-secret coded language gives a ranking to the employee, with coded phrases and signs covering aspects that even intrude on personal details. Let’s catch our breath for a second, especially those that learn about this for the first time, and see how broken this system is. Reference checks and background checks are important, do not get me wrong and in certain jurisdictions are compulsory in order to achieve certain company certifications. However, taking the coded ranking given in these letters at face value means that you “outsource” your critical thinking to past employers. Getting a list of titles, employment dates and duties should be more than enough and everyone should be able to provide a list of references you can contact, if needed. In addition, who is more likely to get a glowing recommendation? John, the engineer who disagreed with management on budget allocations and, while caring for the company and their craft, moved on or Bob, that the company is happy to see them go and become “someone else’s problem”? Think about this for a second. Personally, I stopped asking for these documents years ago – a number of people with similar backgrounds to mine are doing the same – on the basis that if a place has hard requirements for these, then it is not the place for us. Asking people from abroad for similar documents is just plain nasty, so please DON’T.

With this barrier out of the way, there are still a lot of things that can be done better. Asking people for a copy of their degree, perhaps for compliance purposes, is fine – asking people that have entered the workforce years ago a transcript of their university courses is just plain idiotic. Unless the person on the other side is really, really desperate, they will do an 180 and move on. Adding barriers to the process that have nothing to do with the job at hand is another big no-no. IQ tests, logic tests are not an obstacle that will be overcome by candidates, it is an obstacle that will make people do a 180 again. Topics like job advertisements listing 20+ required skills (thus scaring a lot of candidates, including a lot of women, away), outdated technologies that belong in a museum still being adopted, applying to an English advertisment to get a reply in German etc are self-explanatory.

If someone re-reads the OECD paper they will see a favorable mention of Germany – in the quality of University graduates. Germany invests significantly in their third-degree education, which is quite accessible for most of the population and the results are there – papers from German universities achieve high indexes and ratings of German universities worldwide are quite high. However, here is the MAMMOTH that no-one is talking about:

IN ADDITION TO GERMANY NOT BEING ATTRACTIVE FOR FOREIGN SKILLED WORKERS, GERMANY SUFFERS FROM BRAIN DRAIN TOO

This is pure economics at play and it applies both to German-native and foreign employees and can be summed up in a simple, single sentence: On a global scale (and post-COVID global is closer than you think) competition for talent, German salaries just are not competitive. Germany has the second highest tax rate and social contribution rate in the EU, Belgium being the first. This is already putting companies at a disadvantage, as they have to pay more to be financially competitive. Noone is expecting German companies to be able to compete with FAANG compensation packages – few companies in the planet are – but, and I am going into full hypothesis mode here, owing to a hierarchical mindset, salaries are low in the global scale (hypothesis mode off, this is a fact). The Dutch solved this by giving skilled foreign workers tax breaks – here everyone pays a gradual tax that desperately needs adjustment, as a 42% tax bracket kicks in very, very early. So, engineers with choices are left essentially with two options: bite the proverbial bullet or (again, in the post COVID-19 world) work abroad or remote. Guess which choice an engineer will make given both options. “Ha! I got you there” – I can hear you calling out – “US companies did massive layoffs at the first sign of economic downturn, a German company would never do that since we are not “overpaying” you”. Well, according to the respected Handelsblatt, German companies do too (Handelsblatt article here – paywall) and a trip over LinkedIn will confirm this. So, dear executive that probably will never read this, start paying people competitively and dear government that will definetely never read this, start giving tax breaks or have a sane tax structure to begin with.

But wait, can’t we apply engineering principles to the problems above? AI coding is all the rage now and for sure, we can use it to augment the output of our engineers. And can’t we automate processes enough so humans are less and less needed? At the end of the day, aren’t all engineers equivalent?

Let’s start with the last question – if you are an executive that believes all engineers (or all knowledge workers for that matter) are interchangeable, read the next sentence carefully: Why not all are equivalent and why just getting “butts-in-seats” will not solve your problem has been proved to death, essentially since the 70s – these two startings points might commence the learning procedure (step 1) (step 2).

Automation is good. After all, in certain engineering disciplines, such as Site Reliability Engineering, automation is a noble endeavor, as it provides repeatable, debuggable outcomes and can help to minimize human error. However, when you apply this to product software engineering, the requirements can change so frequently – if not constantly – that maintaining automation might present a burden similar, if not greater than delivering the actual product itself. So, by all means, do use as much automation as you can but do not expect any silver bullets, even in contexts where automation is an optimal target.

“AI will save the day!” I hear the AI zealots shouting – “I can go to chatGPT or equivalent, type in a list of requirements and get working code out of it, hugely increasing productivity”. Here’s are a few gotchas: chatGPT and all other publicly accessible AI technologies are owned by US companies. These technologies “as-is” are a data-loss disaster (reference article). But this is only the tip of the iceberg. The main problem is that product companies live-or-die by their product, which technologically consists of many factors: innovation, speed and quality of execution and yes, even internal processes. By using AI, you are removing a significant factor – the quality of the people creating, operating and securing your product. The same AI that you are using, even if not leaking data left and right, is the same AI your competitors are using – removing any competitive advantage you might have. Perhaps, for internal IT applications this is fine, once you have seen a payroll app, you have seen them all (or not!) but for product based companies, this is plain suicide. So yes, AI is here and is not going anywhere but do not count on it to save the day – at least, not today. Oh yeah, and if you are using metrics like Lines Of Code to measure productivity, see repeat the procedure described above.

Coming to an end, why does all this matter to me? I am living and working in Germany, after all, by choice, the skills gap is indeed a material risk – it would be foolish to deny this – but the proposed measures, according to me and me alone, leave a lot to be desired.  Failing to address the skill gap and failing to adopt a product mentality will mean that German products will have to compete on cost rather than quality. We all know where this leads to and we are all in this together. Until next time folks!