Every leftist who has engaged with mutual aid knows this and has known this for decades. Glad someone is backing it up with research.

Means testing hurts programs and leads to worse outcomes.

I think this has been piloted in other places too, including the US, with similar results. Similarly, means testing and friends also just makes the program more expensive with no benefits. The problem is those who could distribute have goals that are not just help people. Even people who want to help often do so with biases that make it so the problem is just poor management (because your own relative success is always earned)

Cash grants are incompatible with the deeply ingrained values and norms of the Protestant work ethic especially as practiced in the US. Until this changes, cash grants to the poor are DoA.

Ah, $400 or a year’s income as a one-time grant. Pretty cheap to help people in Uganda. Does make the EA case a little since similar research for $1k every month (more than twice as much and not just a one-time grant) shows no such effects in the US.

Just back of the envelope, median household income is $80k in the US so let’s set that to $40k per individual and the census says 34 m in poverty so that’s about $1.2 T in increased transfers (I.e. in addition to current transfers) to mimic this as a universal tactic here. I wonder what happens to constrained resources like housing etc. and what portion of transfers transfer into disposable income if something like this were attempted.