This week I'll tweet a lot about development, aid, and emigration from low-income countries.

First: a new estimate of the cost, in foreign aid, to deter one irregular migrant to Italy: US$1.8 *million*.

By @paulclist & Restelli in The World Economy—> https://doi.org/10.1111/twec.13017

I just want to clarify, that is not a typo. One point eight million dollars *per deterred migrant*.

That is the highest effect seen across various regression specifications testing for a relationship between bilateral aid and any fall in border apprehensions by Italy, 2009–2016.

The study doesn't just test for contemporaneous effects, but uses flexible tests of several lags. It is possible, in principle, that larger effects arrive later than the time-window studied.

How about new asylum applications in Italy? Here the study cannot put an aid-price on asylum applications deterred, because the measured association is *positive*.

Each new asylum application is associated with an *additional* US$162,000 of bilateral aid.

The paper also contains a critical review of the whole literature on aid and migration that, if you're interested in research-based policy on this subject, is required reading.

Other studies imply an *even higher* aid cost per deterred migrant.

Is this a case against aid? Not a bit, as the authors point out. First, it is not a case for anything, it is an observed fact. Second, they clarify their own interpretation: "In short, aid does not deter regular or irregular migration, so should be used for other purposes."

That is, this evidence is irrelevant to the crucial roles of aid, such as strengthening health systems in low-income countries during a deadly pandemic. But it is relevant to the political project of repurposing aid as primarily a tool to assist migration barrier enforcement.