Matt's Thoughts In Between - Issue #64

Forwarded this email? Subscribe here. Enjoy this email? Forward it to a friend.

Why educated women are earning less

The New York Times published a fascinating piece on why the gap between the earnings of the most educated women and the most educated men is growing. The core thesis is that the returns to working very long hours in all consuming jobs (e.g. jobs like banking and law) have increased, but women are much more likely to reduce their hours than men if they're one of a mixed-sex couple with kids. This reduces their earnings and allows their male partners to earn much more. In other words:

Women don’t step back from work because they have rich husbands. They have rich husbands because they step back from work

What’s important about this trend is that it has more or less cancelled out the impact on the gender pay gap of higher levels of education among women. And, because highly educated men are now much more likely to marry highly educated women, it’s had a disproportionate effect on this group.

Some argue that this means that long hours have large negative externalities and should be taxed more (and there's evidence that the benefits of long hours are limited even for employers). I’m not sure about that. For one, the evidence on how hard the hardest workers actually work is mixed (and a lot of coverage of, e.g., extraordinary hours in Chinese tech companies seems to be thinly disguised Orientalism). But it represents an amazing opportunity for employers: there is lots of (underpaid) extraordinary talent out there, disproportionately female, that is ready and available, if only you can design your high performance jobs to not require 60 hours a week.

How (not) to invent the future

One of the most interesting books I read last year - though I disagreed with most of it - was Alex Williams and Nick Srnicek’s Inventing the Future (ITF). Williams and Srnicek come from the anti-capitalist left and ITF is a manifesto-meets-playbook for how to create a post-work Communist utopia. They are "left accelerationists" - i.e. they believe that the Left should embrace, not resist, automation as the path to reducing the need for work.

Scott Alexander, perhaps the most consistently interesting person on the internet, has a long, probing and largely negative review of ITF here. He spends a lot of time on one of its central arguments: that the Left should invest a lot of time and energy to embed itself in, and ultimately dominate, influential institutions, like universities and think tanks. They argue that this is how neoliberalism triumphed, through the efforts of Hayek and the Mont Pelerin Society.

Alexander makes an interesting critique: did neoliberalism win because it had a great strategy or because it was a good idea? He notes that it attracted an inordinate amount of talent. I wonder if there's a general framework for predicting which ideas will attract the most talented and ambitious people. I suspect it's ideas that offer disproportionate leverage and prestige to individual actors (Alexander gives Effective Altruism as an example). Perhaps that's the central challenge of trying to “invent the future” from the far Left today: is modern socialism really compatible with individual ambition?

Why books - and other things - don't work

Andy Matushak published an important and provocative essay on "Why Books Don’t Work". His argument is that most people struggle to retain the content of books, not because of any failing on their part, but because books are based on a flawed implicit model of how we learn - i.e. “people absorb knowledge by reading sentences”. Matushak has deep ideas on how we might improve on books. Indeed, he’s put them into practice in Quantum Country, an online learning journey through quantum computing, co-authored with Michael Nielsen.

His key insight, though, applies more broadly than books: “mediums can be designed, not just inherited”. That is, just because a medium has been handed down over generations doesn’t mean that it's optimal. This idea should resonate with entrepreneurs. My own experience co-founding Entrepreneur First has been one of questioning the established mechanisms for building tech companies. Are they fundamental or just contingent? In Matushak’s terms, how do we make an implicit model of company building explicit?

The difficult meta question, though, is which are the types of human activity where design wins (à la Matushak) and which are the type where evolution wins (à la Chesterton’s fence or Seeing Like à State)? When we can be sure that we truly understand the implicit model we’re changing and really are in an inadequate equilibrium? That, I think, is difficult to answer - though redesigning books seems a safe place to start!

Quick Links

Not-quite-killer robots. Interesting potential sighting of low level AI warfare in the wild
Fun free time activities. How long would it take to find two identical packets of Skittles? That burning question is finally answered.
The SEO curse. 19th century fiction meets PageRank.
The best case for optimism. What's happened to infant mortality since 1950?
Shameless plug. Conversation between me and Ben Reinhardt on the Ideas Machine podcast on investing in talent, building deep tech companies and lots more

Your feedback

Thanks for reading TiB. We had another record week for new readers last week. I'd love you to forward this to someone who might enjoy it. And feel free to hit Reply if you have any comments.

Until next week,

Matt