Free money doesn't change propensity towards crime.

From At least five interesting things: Debunking the Debunkers edition (#73) by Noah Smith.

One more nail in the Universal Basic Income coffin.  

In other words, giving people basic income instead of traditional welfare doesn’t seem to make them less criminal, and doesn’t seem to make them safer from crime. (If anything, it made people slightly more criminal, though the result wasn’t statistically significant, so “not noticeably different from zero” is the safest takeaway here.)

In other words, basic income has taken yet another “L” here. We’d like to tell ourselves that poverty is the root of crime, but in the short term, that’s not the case — giving people more money doesn’t make them less criminal, at least in Finland. The root causes of crime are either longer-term economic factors, or deeper sociological factors.

Cash benefits still give out cash, which makes people less poor. But they don’t have a lot of the side benefits many of us had hoped for. A lot of what happens in society can’t easily be reduced to how much money people make.

Smith has been a UBI enthusiast but also acknowledges that the accumulating evidence does not support the initial hopes for the policy.

I had high hopes for the idea that just giving people cash would fix a lot of society’s problems. I still think a system of unconditional cash benefits would be simpler, fairer, and easier to navigate than many of our current welfare programs, and I still think it’s worth giving poor people money in order to make them less poor. But over the past few years, a bunch of new evidence has shown that the costs of cash giveaways are higher (in terms of incentivizing people to stop working), and the social benefits are much narrower, than boosters like myself had believed. Kelsey Piper had a great writeup of this disappointing evidence a few months ago, and I wrote up some thoughts in one of my earlier roundups.

Now we have another piece of evidence showing that cash benefits solve fewer problems than we’d like it to solve. Aaltonen, Kaila, and Nix (2025) study a recent basic income experiment in Finland.

From what I can see, almost all the UBI studies come out with two conclusions.  1) Giving people more money increases their income under most circumstances, and 2) Giving people more money does not change any of their negative social, criminal, health, or education habits.  Occasionally there is a study which finds something marginally positive but from my passive monitoring, it seems that the solid majority do not.

It feels like we are edging closer and closer to the conclusion that UBI is not a viable policy.  It does not change the trajectory of people's life outcomes for the better and usually at significant cost and also often with moral hazards.  Necessarily, most of these studies are relatively small with dozens up to a few thousand participants and usually for months or at most 2-3 years.  

There is a valid argument that anything temporary simply cannot change long term behaviors and decision-making.  But there is very little evidence to support making the plunge into a longterm, large-scale UBI.  And a lot of evidence against it.

The corollary to all these studies though is interesting.  The UBI studies repeatedly and rather consistently seem to indicate that life outcomes are the product of individual behaviors and decisions rather than income.  Access to basic income does not change criminality.