Longreads

• A while ago there was a viral headline about how there was $400bn of unemployment fraud during the pandemic (I was skeptical). Shawn Donnan at Bloomberg profiles the company that released this number, ID.me. ID.me's incentives are not ideal: the worse they are at verifying people's identities, the more they can report that fraud is a frequent occurrence. There's a tricky balance with privatizing functions like identity verification: a private company may have a better shot at delivering a working product than fifty states independently building their own version. But it can also have an economic incentive to build the wrong product.

• Brendan Borrell at The Atlantic asks how many lives were lost because vaccine approval was delayed from September to November 2020. The piece doesn't come to a definitive answer, but does bracket a few reasonable estimates. It's hard to consider counterfactuals since we don't know how many people would have reversed their views on vaccines given an earlier release date. Trust is a resource institutions have, which they sometimes need to wager or spend in order to achieve some other goal. But the relevant odds are hard to estimate even after the fact.

• Gary Hoover on the origin of UPS (2018). Some fun parallels to current logistics stories: early UPS offered white-label delivery to retailers who would otherwise use their own fleet. As with Doordash and Uber Eats, a bigger operator can get higher utilization, leading to lower costs than delivery tied to specific companies. And, like the modern crop of delivery companies, UPS made ample use of equity as a compensation tool. (Via Snippet Finance.)

• Nilay Patel interviews Volkswagen CEO Herbert Diess. The most valuable part of this was that electric vehicles are more modular, with similar batteries and components. This means that brands will differentiate more through cosmetic changes and software.

• Derek Thompson in The Atlantic on new funding models for scientific research, backed by tech money. There's some history to financial windfalls getting recycled into basic research (Alfred Lee Loomis is a worthy example). New institutions tend to be better at zero-to-one scaling while established ones are better at one-to-N. But if the goal of the institution is to advance the frontiers of knowledge, zero-to-one counts a lot more.

Books

• Kishi and Japan: The Search for the Sun: Nobusuke Kishi is a very odd historical figure. He was a bureaucrat in the Japanese puppet state of Manchukuo, where his rule led to bloodcurdling abuses. He signed the declaration of war against the US, tried to manage the economy during the war, was imprisoned afterwards for war crimes—and then ran for office again after the war and served as Prime Minister. (The book opens with a chapter on Kishi's home life, with a brief cameo in which his grandson Shinzo shows up after school to watch cowboy shows on TV.) This book is basically a work of cold war propaganda; it was published in 1960 and somehow omits the brutal policies of the prewar years. Still, it's worth thinking about Kishi's career: it's useful to look at the last common ancestor of wildly different ideologies. Kishi's management of Manchukuo influenced the postwar Japanese economic model, which was deeply integrated with US capitalism—but that ideology itself was influenced by the same kind of thinking that led to Stalin. It's fortunate for the world that modern states inherited a much less brutal model of rapid industrialization. This tweet is an enviably good summary: if you've ever played a strategy game and wondered "what if world leaders treated human lives as just another resource, like iron ore or Vespene gas?" then Kishi is a look at how that goes.

Open Thread

• Drop in any links or thoughts that might be of interest to Diff readers.

Leave a comment