Does a career grant actually stop brain drain?

A talented postdoc from a smaller country takes a two-year contract in Boston, then a better one in Zurich, and somewhere along the way stops planning the trip home. You have probably watched this happen to someone. Countries that train researchers keep losing them to countries that pay better and cite better, and everyone from the European Commission down has spent decades trying to reverse the flow. The obvious lever is money: hand the returning scientist a grant, and maybe they stay.

Does that lever actually work? A new paper in Quantitative Science Studies is, as far as I can tell, the first study to check the question quantitatively, and the short answer is: a little, and mostly for the people you would least expect.

The paper is Marco Seeber and colleagues, “The effect of public funding on long-term relocation of experienced researchers”, published in 2025. They studied the Marie Curie Career Integration Grant, or CIG, a scheme under the European Union’s Seventh Framework Programme that was built to counter the European brain drain to third countries and to help experienced researchers settle for the long term inside an EU member or associated country. It is exactly the kind of “come back and stay” incentive that governments love to announce and rarely measure.

Here is the finding, stated the way the authors state it. Getting the grant was associated with a 9.4% higher chance of long-term relocation in the host institution, and an 8.2% higher chance in the host country. Not nothing. Not enormous.

How did they measure it? They took the applicants to three CIG calls, from 2011, 2012, and 2013, reconstructed each person’s career from Scopus records and their CVs, and then compared where the funded applicants ended up against the applicants who applied and did not get the money. So the comparison is not grantees versus the general population; it is grantees versus their own rejected peers, which is a much fairer test.

The part I find most interesting is who the grant helped. The effect was concentrated in the applicants who had the fewest other options: people without a tenured position, scientists in the soft sciences, people who were not returning to their home country, and people moving to institutions that are not highly ranked. For a well-funded scientist heading to a top lab, the grant changed little, because that person already had other money and other offers. For someone with none of that, it moved the needle. A subsidy matters most to the person who cannot easily replace it. That is almost boring to say out loud, but it is good to see it hold up in the data.

And one honest negative that the authors report: the grant showed no relationship with landing a tenured position, and none with scientific productivity. It helped people stay in place. On this evidence it did not make them more successful once they stayed.

Now the caveat, because this is where I get twitchy. This is a comparison, not an experiment. Nobody flipped a coin to decide who got funded. The people who won CIG grants won because a committee judged their proposals stronger, and stronger proposals tend to come from people who were already more likely to build a stable career wherever they landed. The authors are careful with their verbs, “related to,” “associated with,” and so should we be. A 9.4% gap between grantees and rejected applicants could be the grant, or it could be the same qualities that got them the grant in the first place. Matching and careful career reconstruction narrow that gap. They do not close it.

It is also one programme, run by one funder, over a handful of years, on one continent. Whether an Israeli or Canadian or Brazilian version of the same idea would buy the same 8% is an open question, and this paper cannot answer it.

So why should you care? If you’re an early or mid-career researcher weighing whether a reintegration grant is worth the application effort, this gives you a realistic number instead of a brochure promise. The grant nudges your odds. It doesn’t rewrite your career, and it’ll do the most for you precisely if you are the applicant with no fallback. If you care more broadly about where science concentrates, the sobering reading is that even a targeted, reasonably designed grant buys single-digit percentage points against the pull of richer, better-cited places. The gravity is strong.

I would like to see the randomized version of this, if any funder is ever brave enough to run one. Until then, single digits is what we’ve. Would you move countries for an 8% nudge, or does that just confirm you were leaning that way already?