Thread by Alon (they/them)

twitter.com
1 Recommender
1 Mention
Recommend
Post
Save
Complete
Collect
Thread
1. Brief thread on some state capacity issues, coming from @KlingBlog's article here: nationalaffairs.com/publications/detail/designing-a-better-regulatory-state
@KlingBlog 2. So, first, as a historical nitpick, its story of property ambiguity has two mistakes, one random and one systemic.
3. The random error: as documented by Hernando de Soto, early-19c America did have conflicting property claims, regularized in court.
4. The systemic one is that the 18c view of government developed in the US, UK, and France in tandem - and UK/FR farmers lived in villages.
5. He says a 19c farmer would not encounter the commons much, but that was not true in Europe, and classical liberalism was trans-Atlantic.
6. Now, on to the actual issue - but I promise the systemic nitpick is relevant. Kling proposes a chief operating officer, or COO.
7. The COO is to direct various agencies - he mentions infrastructure but also the CDC - and have corporate management background.
8. Moreover, the COO should not last between presidents, and be firable at will - this isn't a permanent cabinet secretary.
9. In other words, Kling's idea is to have a CEO with management but no subject-matter or regulatory experience direct expert agencies.
10. This is common in traditional American business culture. Managers are supposed to know management; subject matter details are for peons.
11. It is a disaster. CEOs golden-parachute out of the companies they mismanage, and are often hired in the public sector, where they fail.
12. The US's frontier industries - tech and biotech - don't work that way. Bill Gates, the Google guys, and Mark Zuckerberg coded.
13. Noubar Afeyan has a biotech Ph.D. Biotech is trans-Atlantic; Uğur Şahin and Özlem Türeci are medical and research physicians.
14. In successful government - Kling speaks of infrastructure programs, so let's go with that - good leaders come from within the field.
15. While American political appointees wreck all that they touch, lifelong engineer Isabel Pardo de Vera Posada has built inexpensive rail.
16. For her role in low-cost rail building, Pardo was made president of Adif and then permanent secretary of the ministry of transport.
17. And here's the thing: exactly zero CEOs who would parachute to a COO position (say, the MBTA's Luis Ramirez) know who Pardo is.
18. Few American CEOs know who builds low-cost infra (e.g. SEurope) and how. Why would they? This knowledge isn't rewarded in US business.
19. The lack of reward is the important part. Individuals can know important things, but the systems surrounding them don't care.
20. Balaji Srinivasan was an early corona cassandra and, in spring 2020, spoke favorably of the centralized quarantines of Asia-Pacific.
21. And yet, the tech libertarians who listen to him when he incorrectly trumpets Bitcoin and Uber did not echo him on quarantines.
22. I know some people within that community support me regarding infrastructure costs, like Patrick Collison, but it's not as systemic.
23. There's enough support that the Transit Costs Project is funded, but the CEOs who'd parachute to Kling's COO role would do it wrong.
24. Tyler Cowen, cited positively by Kling re state capacity, said to my face he doesn't care about costs because the US Sunbelt has cars.
25. A good hint for what Kling gets wrong is the analysis he writes between the historical background and the COO proposal.
26. He speaks about the failures of US conservatives, who are anti-state, and liberals, who overlook the present model's flaws.
27. He then positions himself in the middle, as does every centrist or would-be centrist reformer in the US. It's part of the genre.
28. But usually core problems are not about a tug-of-war, but about accepted principles on both sides, even in the context of a cold war.
29. For a labor vs. bosses cold war example: in the US and Europe it's accepted by both sides that in a pinch, junior workers should suffer.
30. This leads to gerontocracy and difficulties of finding workers despite high wages, because junior workers have poor conditions.
31. So the solution is likely to involve not navel-gazing about internal US lib vs. con fights, but questioning shared US assumptions.
32. These shared assumptions are often so fundamental they are invisible to practitioners.
33. Some of those assumptions are good: that the PRC is a bad regime, that an independent judiciary and media are good, etc.
34. Politically, the origin of the alt-right disaster is in rejection of the shared assumption among Americans that immigration is good.
35. But some of those assumptions are wrong - and in this case the assumption is that apolitical civil service is impossible or undesirable.
36. So, Kling is proposing a political COO, who serves at the president's pleasure and has no continuity across parties or administrations.
37. But more successful examples don't do it this way. They - even we, in medium-cost Germany - have permanent civil service.
38. When a permanent secretary is appointed by politicians, as in Israel, they come from within the ranks of the bureaucracy.
39. This is akin to the appointment of top generals: the IDF chief of staff and US four-stars are appointed politically from the next rank.
40. For that matter, Thomas MacDonald was likewise so appointed: he was Iowa's road czar, and his success led to his federal appointment.
41. And in such examples, nobody cares how the candidate voted. Israel is politically divided but nobody asks this of the IDF heads.
42. (This is despite the fact that IDF heads tend to enter politics after leaving office: nobody asks on what side they'd enter.)
43. So no, a COO would not improve the bureaucracy. Successful bureaucracies globally do not work like that; they do the opposite.
44. The role of elected politicians is not to micromanage. When they and their appointees micromanage, the results are negative.
45. In NEurope, their role is to set macro-scale policy, aided by our large left-right differences on tax, labor, and similar issues.
46. We have a center-left coalition, so there's more funding for rail. If CDU had gotten more votes, funding would go to roads instead.
47. In effect, as with the military of every democracy, there's civilian control (over funding and priorities) but no political control.
48. This has knock-on effects - the US needs to transition to a culture in which civil servants are empowered to make swift decisions.
49. Today, they aren't: they are trained to stall and delay unless a political appointee wants something done.
50. This is not something a COO would solve; COO-type managers usually slow things down by demanding that civil servants ELI5 things.
51. Rather, what's required is more top-level support of expert civil servants to make autonomous decisions without micromanagement. /end
Mentions
See All