Matt's Thoughts In Between - Issue #9

Who becomes an entrepreneur?

There's a big academic literature on entrepreneurship, much of it not particularly useful from a practitioner's perspective. This paper by Deepak Hegde and Justin Tumlinson, however, generated a lot of interest in the last couple of weeks, thanks to an excellent tweetstorm from Josh Wolfe. It's worth reading the whole thing.

The central finding is that people whose actual ability is higher than is signalled by their credentials (i.e. education) are more likely to become entrepreneurs. If you believe that employers will undervalue you relative to your assessment of your potential entrepreneurial earnings (or just that the system doesn't "get" you), you're more likely to go it alone.

There are many implications for venture capitalists who can identify talent that don't have traditional credentials, as Wolfe discusses. But I'm actually as interested in the other side of the equation: what if we can systematically increase the expected value of entrepreneurship for certain individuals? At the core of EF's thesis is the idea that for the most talented individuals, the potential upside from entrepreneurship is at an all-time high - and growing. The world will look very different if an increasing share of high-skill individuals assess the entrepreneurship/employment trade off and come down on the side of starting up.

Signals, signals everywhere...

Once you get into the idea of "signalling", you see it everywhere. Bryan Caplan has a new book, The Case Against Education, which argues that most of education spending is wasteful. Caplan says education is less about improving individual's abilities and more an expensive way to create a signal of intrinsic talent. Given this, he argues, we'd be better to focus on cheaper ways to demonstrate actual ability. It's controversial, particularly on the left, but Caplan has an interesting post arguing that this idea should actually be popular among liberals. (Counterpoint: evidence that education does significantly increase IQ)

The idea of credentials as an imperfect "signal" of ability is an interesting case of a much broader phenomenon. Robin Hanson and Kevin Simler have a fascinating new-ish book, The Elephant in the Brain, that argues that a huge amount of human behaviour can be explained if you ignore what people say their motives are and instead look at what signals they want their behaviour to send. The book is excellent, but if you prefer podcasts, Hanson discusses the ideas at length with Tyler Cowen and 80,000 Hours.

I strongly recommend getting to grips with this idea, even if you end up disagreeing. It's one of the most useful mental models I've come across in the last couple of years.

Can countries transcend physical space?

I wrote a couple of weeks ago about the idea of startup countries. Since then, Balaji Srinivasan, whose video "Silicon Valley's Ultimate Exit" I linked, has published a follow up video on the Network State, which is worth watching.

At the same time, though, I've become more skeptical that it will be as easy for new country-like-entities to transcend physical borders as the Sovereign Individual crowd sometimes thinks. The article I linked to last week about AI hardware prompted a lot of conversations that caused me to update my views. There is something important about the essential physicality of some resources (above all people, but also various pieces of infrastructure, including computation or especially the supply chains that enable it) in an increasingly virtual world. When the world is basically peaceful, physicality can largely be ignored; when it is not, the abstraction of virtuality perhaps breaks down.

For example, can the rich really opt out of society? If everyone of a certain level of wealth moved to a zero-tax state or moved their assets "onto the blockchain", would that actually be a defence against unrest? I'm not sure. It is hard to design a life that is not ultimately physically seizable in some way. Or, rather, if it is possible for a given individual, it certainly seems hard for a whole class - at least until the world looks very, very different from today.

Quick links

Big data will find you out: Amazing study using taxi data to suggest a very cosy relationship between the Fed and the largest banks
Buy your local paper: Trump did best in places where the local media had collapsed. Lots of cause/correlation conflation, but still interesting (I wonder if the same is true of Brexit in the UK?)
London as alien planet: Stunning video (3 mins) of London taken during the "harvest moon" last October
ISIS as bureaucracy: Striking tweetstorm on how much the terrorist group loved paperwork
AI needs D&D, not Go: Interesting short argument that the next stage of artificial intelligence needs to learn to play Dungeons and Dragons - or at least learn to role play

Your feedback

If you enjoyed this, I'd love you to share it. The easiest way is to forward it to someone you think would enjoy it. I’m always pleased to get your thoughts and suggestions - just hit reply.

Until next week,

Matt