Marc Andreessen lit off a tweet storm on this topic of increasing the number of tech start-ups, and a few of his ideas were about K12 education:

On the K-12 side, he calls for more: Montessori, project based learning, entrepreneurial themed schools, STEM summer camps, and high school internships at tech companies.

Marc Andreessen is an expert in start-ups; he is not an expert an education.

That being said, watching many successful founders might give him some insight into how certain educational experiences influence entrepreneurs.

Of course, causation is very difficult to improve with anecdotes. An entrepreneur might say: “that math summer camp is what launched me to be an entrepreneur,” but it’s difficult to prove that this is actually true.

In terms of evaluating Marc’s K12 proposals, I have never seen any research linking Marc’s proposed solutions to increases in entrepreneurship rates. While I’m open to the idea that these ideas could work, I start with some degree of skepticism.

As a somewhat expert in education and a non-expert in entrepreneurship, here are some musings:

1. The average age of a successful start-up entrepreneur is 40. This number is lower in the tech industry. The average Y combinator founders is 26.

2. Paul Graham notes that domain expertise is very important. This fits my intuition.

3. So if we want more start-ups, I think we should create educational systems and professional environments that maximize the number of individuals who attain domain expertise as quickly as possible, and then attempt to create the structural conditions (VC money, cultural status, immigration, tax policy, university environments benefit cushions, etc.) that will support entrepreneurs to take risks, even when they are in middle age.

Andreessen’s list gets at a lot of these structural factors, and offers many specific solutions of which I am no position to evaluate (such as taxes on capital gains).

But, again, I’ve seen no research that connects his K12 proposals to increases in entrepreneurship.

Moreover, there is some risk that his K12 proposals, if implemented poorly, could decrease rates of domain expertise (project based learning is hard to pull off and can lead to lower rates of academic attainment when not done well).

Lastly, it’s worth noting that only a very, very, very, very, small portion of the population will ever  launch a successful tech start-up (or attempt to launch a tech start-up that has any chance of succeeding at scale).

So while we should definitely experiment with programs that create pathways future entrepreneurs, we should keep in mind that most people will be working at organizations that they did not found, and will thus need mindsets and skill-sets that are different than those of founders (even if there is some overlap).