As a follow-up, this recent NBER paper on people who receive unexpected cash windfalls in Sweden rejected the idea that poverty was the main causal factor behind poor people committing crimes at a higher rate; their non-significant point estimate was in fact that wealth transfers increased crime (though this could easily be noise). It would be good if we could contrast it with the GiveDirectly result, to see if the different environment (random vs predictable, Sweden vs Kenya) made a significant difference.
As a follow-up, this recent NBER paper on people who receive unexpected cash windfalls in Sweden rejected the idea that poverty was the main causal factor behind poor people committing crimes at a higher rate; their non-significant point estimate was in fact that wealth transfers increased crime (though this could easily be noise). It would be good if we could contrast it with the GiveDirectly result, to see if the different environment (random vs predictable, Sweden vs Kenya) made a significant difference.