I’m confused about equity. Yesterday post from Zvi reminded me about my confusion. Some feel like they’re fitting together on this topic. Also I don’t want to think to much about this right now and I feel like it’l be easier for me to stop thinking about this if I write things down.
So I’m confused. None of this is conclusive, it’s just writing down my confusions on the topic.
So. Long time ago (maybe in late 2014?) I remember watching a presentation (I’m not going to look for it) where presenter claimed (IIRC) that if you take developed countries then those with less inequality are doing better than those with less (income?) inequality (measured with Gini coefficient?) are doing better on a number of outcomes. And that this proves that inequality is bad.
So (IIRC) I thought that I don’t think this works, it just looks like they were measuring number of poor people in a roundabout way. Note: I never actually checked if doing something like controlling for a number of poor people actually makes this go away.
The thing is that (actions that make poor people less poor) overlap with (actions that reduce inequality) but they are not the same set of actions.
Toy example. Lets say we have 10 people. 9 of them are earning 1 unit of money / year, 10th is earning 100 units of money / year (Gini coefficient ~0.82). Now we can:
- Increase income of 9 people from 1 to 10, and income of the 10th person from 100 to 2 000. This increases inequality (Gini ~0.86) but also increases income of the poorest people in the example.
- Increase income of all people by 9. This increases income of the poorest people and reduces inequality (Gini ~ 0.45).
- Reduce income of 9 people from 1 to 0.9, reduce income of the 10th person to 50. This reduces income of the poorest people but also reduces inequality (Gini ~0.76).
So I though that maybe we want to make Pareto improvements (with regards to income)? But no, I don’t think that’s it. I think that having people be too rich might be causing problems (basically because people can and do use their wealth to make things worse for others in order to make their position more secure (I feel very uncomfortable about giving specific examples; I’ll maybe vaguely gesture at this, this).
So yeah. I don’t know what to do. People being poor is definitely a problem. Something like too much inequality is likely to cause problems. It still feels to me like:
- people being poor is directly causing problems. So I’m pretty sure people being less poor is pretty good.
- inequality is causing problems more indirectly. So maybe it might be better to change something else (or maybe not, again IDK).
BTW. I recently heard that markets don’t behave like an agent, but like aggregation of agents making pareto improvements. Also that things that aren’t maximizing expected utility can be dutch booked. I’m curious what’s up with that. I think I’ve never checked formal theorems that are often informally expressed like that (also that sequences here sound much less firm than what I often read). So:
- If I ever plan to rely on any of those things (or on Kelly criterion which I read pop-version-of often but never checked math details) I really need to check math to be sure I’m relying on math, not on some vague tale roleplaying math.
- Is it possible to dutch-book market that’s doing pareto optimization? Did anyone write down a nice example.
cashoyat 