3 ms·

This study is not a study - they created a mathematical model that has a bunch of behavioural assumptions, given those assumptions reducing inequality improves

by olliej 5y ago

This study is not a study - they created a mathematical model that has a bunch of behavioural assumptions, given those assumptions reducing inequality improves it. It seems completely plausible that a different set of assumptions could create the opposite. A study would mean looking at actual people, and say directly reducing the inequality. Obviously such a study would be infeasible (time, money, ethics, etc), but that doesn’t mean giving this kind of “study” any credence is actually meaningful. Don’t get me wrong, there’s huge amounts of evidence int the negative outcomes in environments of extreme inequality, but bullshit papers like this aren’t going to do anything except boost the authors H index rating.

ModernMech 5y ago
> A study would mean looking at actual people, and say directly reducing the inequality. Obviously such a study would be infeasible Hence models. Your complaint boils down to "This is a model of reality, and not reality therefore it's meaningless." This is a very shallow criticism. Yes, all models come with assumptions. If you have a criticism, it should be about specific assumptions, and not the fact that this is a model.
- notahacker 5y ago
  Some models incorporate real world data. Some models draw analytical conclusions from [slight modifications] to relationships widely agreed to be true. And some are pretty purely abstract. With this one you can largely guess the conclusion from the stated assumptions despite the complexity of the algebra but... there isn't much reason to believe the stated assumptions approximate reality. Yes, if you model political polarization as economic interactions in which people gain more from riskier trades with people from a different race or political party, pick their political party according to these economic aspirations, and are risk averse in proportion to their income, then the logic of those assumptions is that inequality will make people more likely to sort themselves into parties which match their races and consider those important to their decisions. But I'm not sure that's really how US racial politics works, or that the complexity of the model's dynamics make it more rather than less plausible. There are better reasons to believe reducing inequality can reduce political polarisation than this model, as well as better reasons to believe it won't make much difference

by olliej 5y ago

ModernMech 5y ago

> A study would mean looking at actual people, and say directly reducing the inequality. Obviously such a study would be infeasible Hence models. Your complaint boils down to "This is a model of reality, and not reality therefore it's meaningless." This is a very shallow criticism. Yes, all models come with assumptions. If you have a criticism, it should be about specific assumptions, and not the fact that this is a model.

notahacker 5y ago
Some models incorporate real world data. Some models draw analytical conclusions from [slight modifications] to relationships widely agreed to be true. And some are pretty purely abstract. With this one you can largely guess the conclusion from the stated assumptions despite the complexity of the algebra but... there isn't much reason to believe the stated assumptions approximate reality. Yes, if you model political polarization as economic interactions in which people gain more from riskier trades with people from a different race or political party, pick their political party according to these economic aspirations, and are risk averse in proportion to their income, then the logic of those assumptions is that inequality will make people more likely to sort themselves into parties which match their races and consider those important to their decisions. But I'm not sure that's really how US racial politics works, or that the complexity of the model's dynamics make it more rather than less plausible. There are better reasons to believe reducing inequality can reduce political polarisation than this model, as well as better reasons to believe it won't make much difference