The article is interesting, particularly where he discusses what actual smart bettors do to disguise the fact that they are smart (why would they do this? Read the piece.) My reaction to the article was twofold:
- HOW CAN HE POSSIBLY BE SURPRISED BY THIS?
- Likely because he is a smart and principled person (read: who has good ego integration and impulse control) and his model does not take into account that he is nears one corner of the distribution of humans in these dimensions.
My claim is that any enterprise which traffics in the aforementioned activities a) necesssarily entails moral hazard and b) from a pareto principle argument, will get a major chunk of its profits from a slice of the population who are irresponsible with their drugs, guns, sex and gambling, harming themselves and others in the process. This seems trivially obvious. It also seems like this should be obvious to everyone else, in some form, even if lacking explicit reference to some of these concepts. (I may be near another corner of the distribution in the cynicism of my opinions about humans and our impulse control.)
You might object that we could avoid moral hazard by allowing only those who are competent to make good decisions in these arenas, participate in these activities - and indeed, we are trying to do this in a very low-resolution way by setting a legal drinking age, an age of consent for sex (low-resolution because there are certainly people above those ages who prove repeatedly they make bad decisions about when/how much to drink or when/how/with whom to have sex.) But this is exactly the opposite of what online gambling sites do. (If you win too much in brick-and-mortar casinos, they find ways to get you permanently out the door, and they don't care if they can figure out how you're doing it or whether it's legitimate. Again, as a smart and moral person Zvi may not have known much about how casinos operate.) Scott Alexander links to Zvi's piece, adding (emphasis mine):
I try to err on the side of liberty when it’s at all plausible, but I think Zvi makes a convincing case that this has destroyed too many lives for too little gain (it doesn’t even encourage people to be better at understanding risk and probability; the betting sites ban anyone who doesn’t seem like a rube).Like Scott, I too have tried to err on the side of liberty but every advance in information technology is asymmetric in that it incentivizes increasing accessibility to the taboo domains (sex, drugs, weapons, gambling) for everyone, including (especially!) the people who should least have access to those things. People are manipulated into addictive behaviors via the components of happiness: pleasure, meaning, and flow (examples are porn and junk food, cults, and video games respectively.) That's why this is an interesting example - because increasingly, technology incentivizes liberalization of these domains. If you're reading this article, you probably would have found life in 1950s America intolerable for many reasons, many of them reducing to rigid uniformity - that is, not having the ability to make risky out-of-bounds choices to better yourself. You just weren't allowed many degrees of freedom in how you chose to live your life. But for people with poor impulse control, it kept them on the rails, and we've created a world that's great for Zvi, but not for his customers. This leads to the confession of a fallen libertarian: we harm people by liberalizing taboo activities, and we can't pretend that extending liberty does not give more room for those at the bad end of the impulse control spectrum, much more room to go off the rails.
I don't know what the solution is - jokingly (?) a drone that follows each person, and shocks them if they do one of the taboo activities that they are particularly vulnerable to - the problem being that, if this is in a democracy, the people that need the most protection are least likely to know they should consent to it.