4 ms·
I'll dispute the results. Actual study here: https://static1.squarespace.com/static/5fdc101bc3cfda2dcf0a2244/t/66a8cd0fc256a2303a8569fc/1722338580211/CGIR+LA+B
by CSMastermind 2y ago
I'll dispute the results.
Actual study here: https://static1.squarespace.com/static/5fdc101bc3cfda2dcf0a2244/t/66a8cd0fc256a2303a8569fc/1722338580211/CGIR+LA+BIGLEAP+Final+Report.pdf https://static1.squarespace.com/static/5fdc101bc3cfda2dcf0a2...
One very obvious problem is that they had two groups: one that received $1,000 a month and another control group that received nothing.
Their results are based on surveys completed by two groups. The incentive for filling out the survey was a $50 gift card.
Unsurprisingly the control group saw significantly more attrition in terms of survey response than vs the people being given $1,000 / month.
People who were even marginally better off in the control group were less likely to respond since the hassle of meeting someone in person to fill out a survey isn't worth the $50 to them. So the control group was essentially concentrated down to the respondents who were worse off vs the broader spectrum of the experimental group they were sampling.
In a brilliant example of motivated reasoning the people conducting the study saw this difference in attrition as a sign that the payments were having a positive effect as opposed to concluding that people are more likely to show up to fill out your survey if you've given them thousands of dollars because humans have a fundamental sense of reciprocity.
The actual study design itself is extremely questionable but even if you take the results at face value they were just comparing the people who got money with people who got nothing. As opposed to for instance comparing money with no strings attached to the efficacy of existing social welfare programs.
I'm unfamiliar with the Center for Guaranteed Income but they seem like a political advocacy group more than a serious research team.
- snapplebobapple 2y agoAnother group funded by that openai guy did the same study in dallas and chicago and found tge guys given 1k a month ended poorer than the control group after three years. Thats why you hear about secondary resul%s like the paid group going to the dentist more. The primary result is failure but they wanted to soin that study to support thwir goals
- snapplebobapple 2y agoa decent summary of the study I referenced above: https://www.youtube.com/watch?v=oyoMgGiWgJQ https://www.youtube.com/watch?v=oyoMgGiWgJQ