Learning to code in schools

by | 05.10.2020 | Software development |

Java as first foreign language

I admit it right at the beginning: I am in favour of children learning to code as early as possible. Perhaps as early as primary school, but at the latest in the middle school. And preferably Java, because that is currently most in demand by companies.

Are there people who see this differently? Yes.

Are there reasons against it? Yes.

Would it be a radical change, if schools were to teach content that is important for companies and the labour market? Yes.

Should the school as an institution strive for a balance between general education and preparation for life after school? Yes.

Wouldn’t it also make sense to offer more courses in English, Spanish, French, Arabic or Chinese in schools as our lives become more colourful and communication more important? Yes.

Shouldn’t the handling and checking of information1 and ethics on the internet and in general be taught more intensively? Yes.

And now?

Are you interested in the pros and cons of “Coding in schools?” Sorry, I do NOT want to write about pros and cons. Nor do I want to discuss whether children should really first learn Java or one of the approximately 350 known programming languages.2 I do not want to address the fact that according to the current study by Stack Overflow, most children in Germany do not produce their first line of code until they are 14 years old, with boys starting about 1.5 years earlier than girls.3 Nor do I want to explain why the focus should be on conveying a basic understanding of coding and not on the differences between the current Java version and the previous version.

I would like to go a small step further: Imagine that the Conference of Education Ministers decides to offer a subject “Coding with Java & Co.” from the third grade on (I know, not a very likely scenario). The advocates of “Learning to code at school” have prevailed. And what happens now? How can the decision be implemented?

Reality in Berlin schools

Anyone who wants to change, possibly even revolutionise the curriculum in German schools should take a look at the reality in German schools. I would like to try this out briefly for the federal state of Berlin as an example:

  • There is an acute shortage of teachers in Berlin. At the beginning of school in 2020, for example, 2,600 new teachers were sought.4 The demand is so great that it can no longer be met by lateral entrants. The trend is rising.
  • The number of 26 children per school class is a benchmark in Berlin. Of course, there are studies that say that classes with fewer pupils lead to better performance5 , but since there is an acute shortage of teachers, decreasing class sizes are not to be expected. On the contrary.
  • In the 2019/20 school year, a total of 325,525 pupils will attend Berlin’s public general education schools. The number of pupils is expected to continue to rise over the 10-year period, so that a total of 381,920 pupils are expected in the 2029/30 school year. This means that the predicted number of pupils will increase by a good 56,000 within 10 years. This corresponds to a Berlin-wide increase of around 17%.6 As there is already a shortage of school places, this trend is also rising.
  • The expansion and development of the digital IT infrastructure in Berlin’s schools is “faltering” (alternatively, one or two drastic words could be used here). For example, there is a fundamental need to catch up in setting up and improving digital networking in school buildings, equipping them with school servers, setting up WLANs, procuring display and interaction devices or equipping them with digital work equipment, especially for technical and scientific education or vocational training.7 School-bound mobile terminals – notebooks or tablet PCs – are of course also lacking across the board.
  • In summary: There are not enough teachers! The number of pupils per class is too high, on average and often at the top; at 156 of a total of 430 primary schools the number of pupils exceeds the permitted size of 26.8 And more than 6,200 Berlin pupils are learning in larger classes than recommended.9 There are too few places in the schools or too few schools. And the technical equipment is in need of improvement.

I suppose you feel the same way I do: I find the figures frightening! However, the challenges have long been known. And now the Conference of Education Ministers – or if we want to make it a number smaller in the sense of the article – the Berlin Senate has decided to offer the subject “Coding with Java and Co.” How can that work?

Prerequisites for learning to code in schools

Let me tell you a secret: it cannot work. Not if the solutions of the past, which obviously already have little effect in the present, are simply carried forward. A “keep it up” will not work. Even if there were sufficient finances, modern working tools and schools were networked, it will not work. For one simple reason: there are not enough teachers who can teach Java or other programming languages. The teachers who are hired today have studied German or English, can teach mathematics or music and know the history of the Prague defenestration. But they have no idea about runtime polymorphism and multidimensional arrays in Java. I’ll take up this thought again in a moment, but before I do, I’d like to mention three prerequisites that are important if “coding in schools” is to become reality:

  1. There must be sufficient financial resources. Separate pots with budgets that are not at the expense of other budgets.
  2. The technical equipment – networking + working equipment – must correspond to the status quo of companies. A computer that is still started with a hand crank is not a good working tool.
  3. The willingness to really want to change something and to break new ground.

The first two points are relatively self-explanatory. But point 3 is a revolution. A revolution in several parts:

  • Schools must learn to go their own way. No more delegation of responsibility. No more reasons why things do not work. Enough with the goal of offering everything uniformly at every school across the board.
  • Schools have to orientate themselves towards companies. What do companies do when they want to develop new products or services? They develop pretotypes, prototypes or minimum viable products. They work with Design Thinking or Design Sprints.10 They test, obtain feedback and use the findings for subsequent development. And if a company does not have its own experts to conduct workshops, open spaces or bar camps, they bring in outside experts for a limited period of time.
  • Schools need to cooperate with universities. What do universities want from first-year students? What skills should students already have when they start their studies? Curricula need to be developed here, but not down to the last detail, but gradually. If “Coding with Java & Co.” starts as a pilot project in some third grades, it is not important at the beginning to know what content the students will be taught 5 years later in grade 8.
  • There is a shortage of teachers, but there is no shortage of software developers (even if companies are constantly and non-stop looking for new developers). In Germany there are about 901,000 software developers.11 It is not important whether teachers know Java or not. There are more than enough experts who know it. And from this pool of experts, honorary staff must be recruited, similar to what is customary at universities or universities of cooperative education in numerous disciplines.

Doesn’t sound impossible all in all, does it?


“Coding with Java and Co.” is an utopia. I personally find it desirable. And also realisable. With small steps. Locally. On site at individual schools. Deliberately declared as a pilot project with the goals of imparting programming skills and gaining new knowledge about how to teach these skills.

Of course, such a procedure takes a lot of time. Maximum 11 years (from the third to the 13th grade). 11 years, during which the contents are refined and expanded step by step and the transfer to subsequent grades and other schools can take place. In which cooperation with companies can grow, the recruitment of expert can be expanded and cooperation with universities can be deepened. All it takes are financial resources, a functioning technology, the willingness to change and a small revolution.

Are there people who see things differently? Yes. And how do you see it?




[1] Stephanie Selmer has written an excellent article about the one talent we need in the digital future.
[2] On Wikipedia you will find a list of programming languages without any claim to completeness.
[3] Entwickler*innen 2020 – Motive und Zahlen
[4] Berlin braucht 2600 neue Lehrkräfte
[5] Studie zu Schulklassengrößen
[6] Bericht: Blickpunkt Schule 2019/2020
[7] DigitalPakt Schule 2019 bis 2024
[8] Berlins Grundschulklassen sind zu groß
[9] Mehr als 6200 Berliner Schüler lernen in größeren Klassen als vorgeschrieben
[10] Design Sprint vs. Design Thinking
[11] Wie viele Softwareentwickler gibt es in Deutschland?

Michael Schenkel has published additional posts in the t2informatik Blog, including

t2informatik Blog: Hidden German Podcast Champions

Hidden German Podcast Champions

t2informatik Blog: The organisation rebel - after all, a good idea?

The organisation rebel – after all, a good idea?

t2informatik Blog: VUCA, so what?!

VUCA, so what?!

Michael Schenkel
Michael Schenkel

Head of Marketing, t2informatik GmbH

Michael Schenkel holds a degree in business administration (BA) and is passionate about marketing. He likes to blog about project management, requirements engineering and marketing. And he is happy to meet you for a cup of coffee and a piece of cake.