Just a note: these categories are different from last month, and will likely be different from next month. They might stabilize eventually.
Academia
In-Kind Transfers As Insurance made me substantially update about something moderately important to me. The current trend in development circles, particularly the ones I’m fond of, is to emphasize cash transfers as a default and likely superior option to in-kind transfers. That is, when trying to help poor people, your default should be to give them money, not cows or medicine. We’ve done a lot of research, and if you give people money they very rarely spend it on marginal alcohol or drugs, and while there are fade-out effects in ten years for cash transfers, there are fade-out effects in ten years for everything. So unless you have a strong argument for why you know better than your recipients how they should spend money, you should just give them cash. Endless examples of recipients of bicycles or goats or what have you immediately trying to sell is a lesson that, unless you somehow have a substantial cost or knowledge advantage, you should probably just give people money.
This article argues against that, with a clever bit of theory and some persuasive data. The core idea is simple: one of the biggest priorities, if you’re very very poor, is not starving: income fluctuations create caloric consumption fluctuations, and going hungry isn’t great for you. For a variety of reasons, saving is hard. Normally, there is a social network: the rest of the village can pitch in and help out. But if something, whether drought or flood or plague, affects the whole village at once, then they’ll be overwhelmed. Worse, the same problem will tend to increase the price of food. As a result, your savings are worth less at the precise time that you really need to buy food. Unfortunately, governments struggle to detect these events in a process that isn’t susceptible to Goodharting.
In-kind transfers can ameliorate this, because they provide the same number of calories regardless of whether food is currently cheap or expensive. The paper goes into different details, and talks about some interesting data, but the underlying theory is a good point that I hadn’t taken seriously before (and haven’t seen others consider). The authors’ claim is a little strong: they say in-kind wins as long as there is positive covariance between marginal utility of goods as measured in non-money and the price of goods, and their assumption is true, but they don’t account for the discount factors that sometimes you didn’t give people the right thing and governmental distribution is often more expensive. As a result, I’m more sympathetic to Heifer and direct food distribution, but my attitude towards non-food in-kind transfers remains the same as it was.
The Curious Case of the De-ICD is a challenging ethical case. James has a heart condition that, in the opinion of all of his doctors, requires an ICD, and one was implanted. He is told that, if his ICD is removed, he is likely to die. It has already activated twice, probably saving his life both times. He does not not like how it looks when he’s having sex, and is of the opinion that the times it went off were accidents. The doctors argue with him for a number of months, desperately trying to convince him to change his mind, and covering all of their legal bases carefully so that they can’t be sued, either by his family or if he’s driving, suffers an episode, and kills or injures someone in the resulting crash. But he is a mentally competent adult and ultimately they have no grounds on which to refuse him. After removal, he suffers another attack. He happens to be close to a medical technician who is able to save his life, and he furthermore manages to beat expectations to not suffer brain damage. In the aftermath, he’s put back on the ICD, and gets in line for a heart transplant as a result of the damage.
Does James deserve a heart transplant? I’m much more sympathetic to “obesity is hard to change” than “young men making moronic life choices can take advantage of very stretched resources.”
Should the cardiologist have removed the ICD?
Paper identified from Thirty Randomly Selected Bioethics Papers, which is worth reading, via Kelsey Piper, who is also worth reading. This ties into another recommendation for the month:
Bioethics in the Times of Contagious Populism was interesting, because it assumes a bioethics that isn’t basically populist. I disagree with it: bioethics is the complaint that secretive elites are experimenting on people in unreasonable ways and need to be stopped via the maximum possible amount of obstruction to any attempt to improve or understand the human condition. However, it’s interesting reading to understand how bioethicists view themselves.
I’ve come around to the view that people in my sphere are more against bioethicists than they should be: yes bioethicists are directly responsible and actively advocate for policies that result in the death of millions, but similar levels of sclerosis also happen in areas where they’re not involved, from psychology to business, so while their beliefs are morally abhorrent I’m not as convinced as I used to be that they’re particularly causally relevant. I’m not trying to dismiss them as entirely irrelevant: I still think that the average bioethicist causes harm. Skim Thirty Randomly Selected Bioethics Papers: remember that these are from the people who fought to block human challenge trials. There’s a lot of good discussion, and on one hand it feels wrong to dismiss ethicists because they disagree with what seems like common sense to me.
And yet bioethicists as a net position are really terrible. If I want to hike a mountain for fame and fortune, even at substantial risk to myself, I am not stopped. But if I want to advance science, or save lives? Now it is made illegal, and they cheer. If something is selfish, like donating a kidney to save a family member, the medical establishment supports you. But undirected kidney donation for purely altruistic reasons results in a multitude of roadblocks including psychiatric verification, because it would be wildly unethical to let a crazy person help others. If I want to torture people for a reality TV show, there is money. If I want to do it for the sake of understanding the human mind and helping others as part of a psychology experiment, it becomes almost impossible.
Tech Policy
Rule By Data: The End of Markets
The good: There’s a subtle and interesting argument which I’ll try to convey here about the nature of human rights and data, but I recommend reading the original. Data, to the author, is a tool for prediction, which enables control. If I know how you will respond to changes in your environment, I can control you. “The act of capturing data by Big Tech (or governments) gives them unchecked power to impose their economic or political preferences on the data producers.” The goal of data, in this framework, is to effectively undermine self-determination, and what the author (following a literature I’m not familiar with) calls self-authorship. I parse it as the ability to decide your own desires on your own terms.
This, in turn, is developed into a broader argument about data as a tool of control and soft coercion. Information is generated about you, very clever analysis done, and you become, not an equal party, but a controllable toy. At that point, we no longer have a nominally equal market, where people meet and bargain on at least superficial equivalency, but a system of hierarchy and control. “Most legal systems will not enforce a contract that has been entered into under duress. Using control over data to ensure that a target will enter into a transaction may not be the same as holding a gun to her head; however, it is arguably closer to it than the idea of a voluntary transaction between parties that contract to maximize their mutual benefits in the face of future uncertainty.” The solution proposed is also interesting, if complex: read the article for the full context. The article also, in my view correctly, concludes that vesting individuals with “data ownership” or the like is essentially meaningless, both because the marginal returns are so low and because the line of necessary information is so limited.
The bad: As is predictable at this point for American elites who derive their income from writing words, the author really dislikes “Big Tech”, and I think makes some statements that a more neutral person would have rightfully rejected. “Big Tech has found a way to retain ownership over data even as it sells the data again and again, and on terms that Big Tech controls. Money changes hands, yet only access to data and their predictive power is granted in return. If this is a market in the original sense of the word, it is a rather peculiar one.” There may be a subtlety I’m missing: I’m not an expert in legal analysis of economic frameworks, and perhaps something subtly different is meant, but it seems normal to me to pay for access to skydiving equipment and their fun possibilities without expecting ownership of part of the airplane.
I also think it understates the wildly different world the modern consumer lives in in terms of access to information about purchasing. Amazon will offer me five alternatives for anything I want, and other stores are just a Google away. True, there is no consumer-facing prediction model for goods and services, but there isn’t enough consumer demand for such things: transparent public prices have a massive impact. I’ve read papers in econ arguing that we should include an implicit discount factor on the inflation rate due to consumer prices falling for everything that can be bought online due to increased search capability. This isn’t exactly “the buyer on the other end of this chain is left in the dark.” I will agree that the situation isn’t equal, but it never was equal: the seller is (almost always) playing a repeated game, the buyer is (almost never) playing a repeated game. The exceptional cases, when buyers are sophisticated regular players and sellers do something rarely, are famous for screwing over sellers, whether it’s giving away mining rights for almost nothing or pawn shops buying gold for pennies on the dollar.
The neutral: read the article in conversation with Maneki Neko, Bruce Sterling’s vision of a data-driven future with a very different sort of economy. The latter serves as a reminder of what the promise of the future used to contain, and what a world with machines that favored human-human connections could look like. Also Maneki Neko is fantastic. I first heard it many years ago, at the basement of a since-shuttered science fiction convention, and have loved it since.
Why TLS1.3 isn’t in browsers yet is about technology and applications, but I’m interested in it for the analysis of ossification. Most people, most of the time, don’t actually look at the original specification. They adapt something that works, play with it a bit, test it against the known implementations, and then call it a day. So if none of the commonly used implementations implement a particular option, over time that option will die out. The important analogy to politics, I think, is that theoretical options that are used less than once a generation are essentially worthless except in times of extreme crisis. That’s not to say that they can’t ever be shoved open: “each Republic shall retain the right freely to secede from the U.S.S.R.” was completely meaningless, until it determined everything 17 years after it was enshrined.
Incidentally, the 1936 Soviet constitution is a delight. Articles 124 and 125 are as blatant a reference to the First Amendment of the US Constitution as you can get: speech, press, assembly, petition, religion. It promises full and equal rights to women, six years after the UK had equalized voting. It was one of the most liberal constitutions in the world.
Google security shut down a counterterrorism attack method, and we’re going to see this debate play out more and more over the next few years. What do you do for users, vs a friendly government? Bumble shut down searching via political preference because one woman was searching for Republican men and exposing them as being January 6th insurrectionists. That certainly protected the users who were being targeted! What about the other users, though?
Also worth noting that this was a “watering hole” attack, by which they mean “just attack everyone who comes by and sort out your actual targets afterwards.” The bigger question, of course, is to what extent Google is willing to be an arm of the US government, and to what extent the US government will tolerate Google being more powerful than most state governments.
Words, Words, Words
Utopian For Beginners is about Ithkuil, mostly. It gets a little weird, in the way of all good stories, but as an introduction to Ithkuil it is fascinating. We can ask questions like “When I look at a rainbow, I see it clearly divided into the colors I grok. Do Russians see the rainbow as including both of their blues?” But Ithkuil is the first language that lets me understand, directly and intuitively, how it could make me think better. Translating sentences into it is an exercise in thinking very very carefully about what, precisely, you mean. Actually speaking the language seems hellish: there are 58 phonemes. Six of the consonants come in unaspirated, aspirated, and ejective. It has seven tones, and stress is grammatically meaningful. Latin has six cases, which people sometimes complain about. Ithkuil has 96 cases, and no, that wasn’t a typo.
Ithkuil understands words and roots in a very different way from natural languages. I think it’s absolutely worth reading the story, and perhaps glancing through the lexicon. It’s mind-expanding in a good way if you’re at all fascinating by language, thought, and what both could be.
You’re Using The Wrong Dictionary is an impassioned defense of using an older and more artistic dictionary. It helped me understand Nabokov’s Good Readers and Good Writers a bit more deeply. Nabokov writes better in his third language than I will ever be in my first: I once complained to a friend that I wasn’t sure if an error in Lolita was intentional or not, the perils of trusting publishers. She explained that actually it was correct, and that dentists do use a weird spelling of a particular word, because she happened to be a professional copywriter who helped small professionals, including many dentists, set up websites. When Nabokov says that a reader should have a dictionary, he does not mean the modern thing, strong enough to trap a word yet so weak as to fail to catch it, but something more true to words. The Hacker News commentary from 2014 is also worth reading, with a few more interesting links and perspectives.
William Butler Yeats fought Crowley with magic. Just a little context on a poet whose The Second Coming was very popular during the previous presidency.
Book of Lamentations is an old story, but it’s one I reread from time to time, particularly as we talk about being “crazy” during a pandemic. It’s a literary analysis of the DSM-V.
Deirdre McCloskey has a moving piece on some of her experiences as a trans woman. I’m not sure how to recommend it, but it’s very well-written. One of the changes I’ve been noticing in myself over the past year is an increased emphasis on what is well or poorly written. In part, this reflects my desire to become a better producer: surrounding myself with good writing seems one of the surer paths to that.
Politics
More policing, less law enforcement is an ex-cop’s plea for an older version of their profession. The essential argument is that most of what people want isn’t law enforcement. They want the drunk guy starting a fight to get dragged out and sit for twelve hours while he sobers up. They don’t want him going to jail for five years for assault! They want to be sure that the woman in the apartment upstairs is OK: any judicial punishment of the man screaming at her is incidental to making sure that she’s safe. If tossing him in jail stops the child support payments he was making, that’s not necessarily an improvement.
But we’ve constructed a system in which the standard formal term for cop is LEO, or Law Enforcement Officer (still infinitely better than warfighter, the DoD’s monstrosity of a word for soldier). Enforcing the law on ordinary citizens is bad, because the law is an inflexible instrument, but that’s increasingly what other parts of the system (DAs, judges, elected officials) judge police officers and departments on. Arrest metrics! The plea is for understanding police as people who use their discretion, combined with the power of the law behind them, to solve social problems.
I take the argument seriously, and I take it as an argument for abolition. Existing police departments aren’t trustworthy. They are too full of people who like using violence, and will use violence in a politicized way. That time the SFPD bombed the mayor’s house because their pay raise wasn’t going to be as big as they wanted. The contrasting cases of the capitol police, who took selfies with the January 6th insurrectionists. Remember that a priest was tear gassed so Trump could have a photo op holding a bible in front of a church! Politicized violence is unacceptable in a democracy, and existing police forces can’t be trusted. But, after they’re abolished, a set of people who can be called on for social problems, who do have the authority to lock someone up and can use a baton if necessary, has a compelling pitch. Just don’t think that their job is to enforce the law.
P.S. There’s a good reason I try to read people who disagree with me. Elsewhere on their blog, they link to a Pro Publica report. Pro Publica is, without a doubt, the publication I trust most. I like individuals, but institutionally? Pro Publica uber alles. Politically, I’d call them slightly left of center, which probably means that they’re left of center after adjusting for being a PhD student in LA. I mentioned the selfies above. Here’s the money quote from Pro Publica. “Watching hundreds of Parler videos shows that the disturbing ones that first surfaced publicly, of officers taking selfies with protesters and otherwise laying down for the attackers, offered a picture that was far from complete. The police visible in the videos fought tenaciously, and the resulting sense of betrayal in the crowd is palpable.”
“Far from complete”. My view today is more nuanced than it was the day of, unsurprisingly. I think that police officers did not take the threat seriously in the slightest. I think that part of why they didn’t take it seriously was a basic political alignment: pro-Trump protestors and pro-Trump officers. They may be harder to find now, but the images of anti-lockdown people screaming directly into the faces of masked police officers were striking, and I remember being genuinely impressed by the cops’ restraint and self-control. This is what officers should be like all the time, I thought. But it seems that the sort of restraint only happens for white conservatives.
Once the insurrectionists were in full-on assault mode and invading the capitol, the situation changes. And it was a fluid mess, with different fronts and different emphases, but there is more than enough evidence for someone looking for it that police officers fought bravely and one died honorably protecting the capitol and America’s political leadership.
Both are true. The racism and political attitudes of police officers and leadership led directly to the capitol being breached, and many police officers fought hard, and one died as well as any could hope to. The former is important! The former is fixable! I’m still a police abolitionist! But I’m going to keep trying to understand people who disagree with me. In many cases, I think it just comes down to “what do you think is more typical?” Here’s the same author talking about qualified immunity. To me, “cops stole over $200,000 from your rare coin collection” is completely unacceptable and a central case, though I’m more outraged over assaulting and arresting a Black citizen on their own doorstep whose only offense was swearing at a cop. To the author, it is bad, but the outrage obscures the technical nature of the case, and the important thing is that if someone drives drunk but gets into their home fast enough, “hot pursuit” may or may not cover entering their garage. But the drunk shouldn’t be able to claim money from the cop! That is completely unacceptable and the central case to the author.
Now, I’m not trying to argue “so how can we know anything?” One of us is closer to right, the other closer to wrong. QI is good or bad. But if you don’t think about media dynamics when understanding beliefs, you will misunderstand people you’re talking with.
My Journey From Free Market Ideologue to Strong Towns Advocate is getting recommended, in large part, because I like conversion stories. Whether it’s “I once was lost, but now am found” or a clear explanation of the specific ideas that played a role, there’s something deeply inspiring in people who change, not just their minds, but their whole worldview.
The Strong Towns analysis is an interesting one, and well-worth understanding. If you want the condensed form of the analysis, it’s in Part 4. Strong Towns attempts to understand why so many American small towns are in fiscal crisis. The conclusion is that that their infrastructure spending completely fails to take into account maintenance costs, on every level. In large part, though the story emphasizes this less, this is because towns just don’t get that much revenue: a small sales tax, a small property tax because people hate property taxes, there’s just not a lot of money. The author gets very invested in local business networks, locally owned stores, and the like, which I am more skeptical of, but the central claim that engineers and planners don’t think in terms of revenue and cost is terrifying.
Mind you, the author doesn’t come off smelling of roses either: in part 3 they note that two alternative methods of storm drainage have a factor of 20 cost difference, and ask if the surrounding property has a factor of 20 value difference, which is obviously silly. Houses are very expensive, and storm drains very cheap. You don’t have to increase the price of a house very much for even a very large increase in the price of a storm drain to pay off.