Week of July 23

Radiology, reach-outs, rights-based zoning, and reckoning.

Hello again! Apologies for the longer break between posts — I’ve made it to Japan and have been visiting relatives, talking with uncles in izakayas, and catching up on the work I wasn’t keeping up with in India. In between all that, I took a boat (recommend this!) to Hokkaido, known in the summer for its lavender fields, rolling hills, and many many different culinary specialities (I don’t think I saw a single food that wasn’t marked “Hokkaido Specialty” that whole week).

The lavender farm opened in the 1970s and is located near a major ski resort, and the area-sponsored additional farms and infrastructure around it are in part a particularly interesting and successful strategy for rounding out seasonal demand. I’d recommend a few days in the area if you’re ever planning a visit to Japan in the summer — it’s much cooler than the mainland! As always, please don’t hesitate to let me know any thoughts or things I missed.

Links

Japanese zoning is a real gift: you can operate, even in the most residential areas, a cafe, workshop, small boutique, etc by right in the first floor of a building you own. This leads to an abundance of tiny shops and businesses that makes taking the side streets or walking an extra station’s distance almost always worthwhile: today, I stumbled across a boutique clothing store that was selling a soy milk-yogurt drinks with cinnamon and fruits (was amazing on a hot day), the other day met a relative for coffee in a tiny cozy coffee-and-bar, etc.

An under-appreciated element of this is the degree of agglomeration that this enables: neighborhoods specialize in topics as niche as “old maps” (神保町) or “buddhist altars” (田原町, no fewer than 20 such stores within a two-block stretch!). Small block lengths that encourage varied but equally efficient routes help here as well, as Jane Jacobs famously observes. Despite the heat, more than any other city I’ve walked this summer, Tokyo feels like an unrivaled pedestrian experience.

Speaking of Jane Jacobs: Paws on the Street (what a great title!). If you take the dog on a walk a few times a day, you begin to notice things, make friends with others in the neighborhood, etc. Maybe dog-ownership-in-lieu-of-child isn’t such a bad thing.

Do you ever consider reaching out to someone, but hesitate because it’s been a while, you might not know them as well (maybe you’re only subscribed to their Substack) and they might not appreciate it? If you’re on the fence, do it! Responders appreciate the surprise more than initiators think, via a set of experiments out last week.

Physician skill is a big deal: it explains some 39% of variation in diagnoses for pneumonia, with potentially huge welfare implications. On face, it seems like automating radiology should be a policy priority and a great ML opportunity. More broadly, in more patient-facing roles, a large part of doctor success seems to be demeanor — it would be great to check that somewhat by being able to look at average accuracy rates for different doctors on these types of questions, but I’m not sure how possible that even is.

Hybrid work is here to stay: slides from Nick Bloom. I was pretty bearish on WFH for a long time, but I’m starting to become convinced. The survey results here are pretty striking and much more positive for WFH than I anticipated.

What’s the marginal cost of convincing people to get vaccines? Showing a 27-second clip of Trump endorsing them ends up costing only $1/vaccine! Watch the ad itself here.

I love biking, so I am biased, but some great results from an RCT giving bikes to girls in Zambia: saves 1hr commuting each day, -22% sexual harassment, -28% absenteeism.

Kibbutz reforms! Market-based reforms in Israeli kibbutz (a kind of communalist farming community) increased support for capitalists generally and decreased support for left-wing parties, but didn’t impact support for social services or safety nets for the less fortunate. People like markets, and markets don’t make you evil.

Not a lot of foreigners living in China for how large it is, but even given the liberalization since 2011 this huge spike in those from Myanmar is surprising. I don’t know anything about the diaspora in China but am curious to try and learn more the next time I (hopefully) get to visit!

Finally — there’s no shortage of online discourse these days on the “reckoning” taking place for Venture Capital as valuations come down to earth, but just how bad (or temporarily embarrassed but in fact good) were investments? It’s probably still too early to say much, but enjoy this paper anyway: “approximately half of the investments were predictably bad—based on information known at the time of investment, the predicted return of the investment was less than readily available outside options … the [total] cost of these poor investments is 1000 basis points, totalling over $900 million”