GiveDirectly and the direct transfer RCTs in Africa/Third World countries don’t answer the question about First World poverty because almost everyone, including the industrious and drug-free and high functioning people, in those countries is dirt-poor; in the First World, there is a much stronger correlation of pathology and poverty. To give an example, the direct transfers in Africa work because people there really are in poverty traps where $100 can make a big difference in letting them buy a cow or a motorcycle, and this is why the direct transfer RCTs show benefits; no one in America will show big benefits from a few transfers of $100 because poor people there have problems which can’t be solved by some cash.
The upcoming YC-funded experiment will help test the generalizability of basic income results, and the original American experiments decades ago suggest that a basic income wouldn’t cause lots of self-destructive behavior (or at least, wouldn’t make things noticeably worse), but on the other hand, the natural experiments of lotteries in the USA and elsewhere like Sweden show minimal benefits to random shocks of wealth (which could’ve been invested for income). So I wouldn’t be totally pessimistic, but I also wouldn’t be surprised if BI experiments in the USA do worse than one would predict from the earlier GiveDirectly results.
I hadn’t thought of that, good point. It still rings of the best example I have, but maybe not by as much. I have zero experience with actual people dying on actual streets so I use what I’ve got.
Yeah, I hope if experiments are done they’re done well. A half-baked experiment could easily do more harm than good.
GiveDirectly and the direct transfer RCTs in Africa/Third World countries don’t answer the question about First World poverty because almost everyone, including the industrious and drug-free and high functioning people, in those countries is dirt-poor; in the First World, there is a much stronger correlation of pathology and poverty. To give an example, the direct transfers in Africa work because people there really are in poverty traps where $100 can make a big difference in letting them buy a cow or a motorcycle, and this is why the direct transfer RCTs show benefits; no one in America will show big benefits from a few transfers of $100 because poor people there have problems which can’t be solved by some cash.
The upcoming YC-funded experiment will help test the generalizability of basic income results, and the original American experiments decades ago suggest that a basic income wouldn’t cause lots of self-destructive behavior (or at least, wouldn’t make things noticeably worse), but on the other hand, the natural experiments of lotteries in the USA and elsewhere like Sweden show minimal benefits to random shocks of wealth (which could’ve been invested for income). So I wouldn’t be totally pessimistic, but I also wouldn’t be surprised if BI experiments in the USA do worse than one would predict from the earlier GiveDirectly results.
I hadn’t thought of that, good point. It still rings of the best example I have, but maybe not by as much. I have zero experience with actual people dying on actual streets so I use what I’ve got.
Yeah, I hope if experiments are done they’re done well. A half-baked experiment could easily do more harm than good.