# Paul Collier's Exodus explained: diasporas, trust, brain drain and the pace of migration

> In Exodus (2013), the Oxford development economist Paul Collier judges migration by its effects on three groups at once: migrants, the host society and the people left behind in origin countries. His diaspora model holds that a large, slowly absorbed diaspora speeds up further arrivals and slows integration, so host countries should set the pace to protect mutual regard and trust, and small poor countries can lose too many of their educated people. Economists such as Michael Clemens, Justin Sandefur and Dilip Ratha reply that his evidence on trust is thin and that he understates migrants' gains and remittances. For Spain, the useful part is the three-sided ledger, which fits circular and training partnerships with Morocco, Senegal and Latin America.

- URL: https://www.radicalcentre.eu/en/articles/paul-collier-exodus-explained
- Language: en
- Publisher: radical centre / centro radical
- Published: 2026-11-15
- Topic: Migration
- Author: radical centre
- Licence: CC BY 4.0

## Key points

- Exodus (Oxford University Press in the US; Allen Lane in the UK as Immigration and Multiculturalism in the 21st Century; Turner in Spain as Éxodo) asks how much migration is best and how fast.
- Collier's diaspora model: migration rises with the income gap and with the size of the diaspora already settled, and a large diaspora that absorbs slowly speeds up arrivals and slows integration.
- He worries that fast migration erodes mutual regard and trust, and that small poor countries such as Haiti lose too many educated people, while large ones such as India and China gain.
- Clemens and Sandefur call the book one-sided: the trust evidence he cites concerns race in US neighbourhoods, and migrants' wage gains run to several hundred per cent. Ratha adds that remittances exceed $400 billion a year.
- Spain's GECCO origin-hiring programme brought 25,767 workers from 17 countries in 2025, with seasonal contracts and a commitment to return; it is a natural place to apply Collier's three-sided test.

Paul Collier is the Oxford development economist best known for *The Bottom Billion*, and in 2013 he turned to migration with *Exodus*. The book came out in the United States with Oxford University Press as *Exodus: How Migration Is Changing Our World*, in Britain with Allen Lane as *Exodus: Immigration and Multiculturalism in the 21st Century*, and in Spain the same year with Turner as [*Éxodo: inmigrantes, emigrantes y países*](https://www.marcialpons.es/libros/exodo/9788415832171/). It is the counterweight to the book we explained yesterday, [Hein de Haas's *How Migration Really Works*](/en/articles/hein-de-haas-how-migration-really-works-explained), and reading the two side by side is a good way to see what the European argument about migration is really about.

Collier opens by complaining that migration is argued about with plenty of emotion and very little knowledge, as his reviewers note. His way out is to treat migration as a fact and ask a quantitative question: how much migration is best, and how fast should it happen. To answer it he keeps three ledgers open at once. One for the migrants themselves, one for the society that receives them and one for the people left behind in the country of origin. He puts the host population's interests first and insists that the other two ledgers count as well. [The Penguin edition](https://www.penguin.co.uk/books/133269/exodus-by-collier-paul/9780141042169) sums up his frustration with a debate stuck between hostility to migrants and the view that open borders are an economic and moral necessity.

## The diaspora model and mutual regard

The heart of the book is a simple model. Migration from a poor country to a rich one rises with two things: the income gap between them, and the size of the diaspora already settled in the rich country, because relatives and compatriots make the move cheaper and easier. The diaspora shrinks as its members are absorbed into the wider society. If absorption is slow and the diaspora is large, it keeps growing and pulls in more arrivals, and those arrivals integrate more slowly still. Collier wants a host country to keep inflows roughly in balance with the rate at which newcomers become part of the mainstream. He worries that without any restriction migration would accelerate sharply; the Lowy Institute's review notes his use of a [Gallup poll](https://www.lowyinstitute.org/the-interpreter/review-paul-collier-s-exodus) in which about 40% of people in the poorest countries said they would move to a rich one if they could.

What is at stake in the host society, for Collier, is mutual regard: the sympathy and sense of shared belonging that let strangers trust each other, pay taxes for each other and cooperate. He treats the nation as a legitimate unit and national identity as a force that holds welfare states together. He argues that the economic effects of moderate migration on natives are small, and that the social effects of rapid migration are where the risk lies. His policy package follows from that: ceilings on numbers, selection by skills, employability and cultural affinity, integration policies such as spreading newcomers across the country and supporting migrant pupils at school, and legalisation for irregular migrants already present, with a partial status.

Then the third ledger, the one that makes the book unusual. Migrants, Collier agrees, are the big economic winners. For origin countries the picture depends on size. Some emigration helps: it brings remittances and ideas, and many leaders in developing countries were educated abroad. Large countries such as India and China gain. Small poor countries can lose too many of their educated people, and Haiti is his warning case. Clemens and Sandefur sum up his benchmark as an emigration rate around Bangladesh's roughly 4% doing good and one around Haiti's roughly 10% doing harm.

## What the critics say

The sharpest reply came from Michael Clemens and Justin Sandefur of the Center for Global Development, in *Foreign Affairs* in early 2014. They call the book a one-sided polemic. Collier leans on Robert Putnam's finding that people in racially mixed US neighbourhoods trust each other less; they point out that Putnam's data are about race, which Collier runs together with culture and nationality, and that Putnam drew no anti-immigration conclusion from it. They cite evidence that crime is lower in the English and Welsh neighbourhoods with the largest immigrant inflows and that immigration raises property values in Spain and the United States. On Haiti they put the share of educated people abroad closer to 75% than 85%, note that about 80% of Haitians earning more than $10 a day live in the US, and say no country has been shown to grow by stopping people leaving. They also agree with him on several points: immigration does not generally displace native workers, remittances matter a great deal, and undocumented workers deserve a legal amnesty.

Dilip Ratha of the World Bank was blunter. He argues that Collier overestimates how many people actually want to move and underestimates the pain of leaving family behind, that he ignores the large migration flows between poor countries, and that remittances of more than $400 billion a year, over three times official aid, deserve more weight. Ratha also objects to Collier's fixed view of identity, drawing on Amartya Sen's point that people hold many identities at once. In the *Central and Eastern European Migration Review*, Georgiana Turculet questions the split between insiders and outsiders, and finds selection by cultural affinity morally troubling. De Haas, for his part, argues that restrictions discourage return and turn circular migrants into settlers, which bears directly on Collier's absorption arithmetic.

## Collier's ledger in Spain

Spain already runs a scheme that fits the three-sided test better than most. GECCO, the collective management of hiring in countries of origin, started in 2000. In 2025 it brought 25,767 workers to Spain from 17 countries, about 25% more than the year before, mostly for farm work in 21 provinces, and 84% of the international workers went to Huelva. Morocco provided 81% of the contracts, 92% of them for women; Colombia accounted for 13% and Honduras for 4%. Senegal, Gambia and Mauritania are on the list too, alongside twelve Latin American countries. The contracts are seasonal and carry a commitment to return. Wafira II, agreed by Spain, France and Morocco with EU support in December 2025, will accompany 3,000 Moroccan seasonal workers over 2026–2028, with training and help with reintegration at home. At European level, the Commission's Talent Partnerships with Morocco, Tunisia, Egypt, Pakistan and Bangladesh include vocational training, skills recognition and the reintegration of returning workers.

Run those channels through Collier's three ledgers and they score well on all of them. Migrants earn far more than at home under legal contracts. The host society chooses the pace and the sectors. The origin country gets wages, savings and skills back, and loses few of its doctors and engineers. Our position takes something from each side of the argument. The left is right that rich countries owe origin countries and migrants real obligations, so fair treatment, labour rights and development belong inside migration policy. The liberal and conservative tradition is right that a democracy may set the pace of arrivals to protect the trust and shared norms that welfare states and markets rely on. Collier's ledger is what joins them: Spain should publish, for every channel, its effect on the people who come, on Spanish towns and on the country left behind, and grow the circular and training partnerships, with Morocco, Senegal and Latin America, that send skills home and keep the way back open. That design answers some of Collier's critics too, because circular movement is exactly what de Haas says restriction destroys, and the theme carries into the [state in control of immigration](/en/articles/state-in-control-of-immigration-who-comes-in-how-when-leaves) we sketched yesterday.

## FAQ

### What is Paul Collier's Exodus about?

It looks at migration from poor to rich countries from three sides: the migrants, the societies that receive them and the countries they leave. Collier argues that migration is a fact and that the real policy question is how much and how fast. He concludes that moderate migration does good, and that rapid, sustained migration can strain trust in host societies and drain skills from small poor countries.

### What is Collier's diaspora model?

Migration grows with the income gap between countries and with the size of the diaspora already in the host country, because settled migrants make it easier for others to follow. If the diaspora is absorbed into the wider society slowly, it keeps growing and accelerates arrivals. Collier wants inflows kept in balance with the rate at which newcomers integrate.

### Does Collier think brain drain harms poor countries?

For small poor countries, yes. He cites Haiti, which he says has lost about 85% of its educated people, and suggests restrictions in such cases, while large countries such as India and China gain from some emigration. Clemens and Sandefur put Haiti's figure closer to 75% and argue that emigration is the main escape from poverty for Haitians.

### What are the main criticisms of Exodus?

Michael Clemens and Justin Sandefur, in Foreign Affairs, say many conclusions lack research support and that Collier mixes up race, culture and nationality. Dilip Ratha of the World Bank says he overestimates how many people want to move, understates remittances and other gains, and treats identity as fixed. Georgiana Turculet questions his insider and outsider categories and selection by cultural affinity.

## Sources

- [Paul Collier (2013), Exodus: Immigration and Multiculturalism in the 21st Century (review by Georgiana Turculet)](https://ceemr.uw.edu.pl/node/181), Central and Eastern European Migration Review (2014)
- [Let the People Go: The Problem with Strict Migration Limits](https://www.cgdev.org/publication/let-people-go-problem-strict-migration-limits), Center for Global Development (originally Foreign Affairs) (2013)
- [Collier's Exodus: Reckless Recommendations](https://blogs.worldbank.org/peoplemove/collier-s-exodus-reckless-recommendations), World Bank, People Move blog (2014)
- [Los programas GECCO permiten que 25.767 personas migrantes trabajen en España en 2025, un 25% más que el año pasado](https://www.lamoncloa.gob.es/serviciosdeprensa/notasprensa/inclusion/paginas/2025/291225-programas-gecco-migracion-circular.aspx), La Moncloa / Ministerio de Inclusión, Seguridad Social y Migraciones (2025)
- [Más de 25.000 personas de 17 países trabajaron en España en 2025 a través de los programas de migración circular](https://www.lamoncloa.gob.es/serviciosdeprensa/notasprensa/inclusion/paginas/2026/250626-programas-migracion-circular.aspx), La Moncloa / Ministerio de Inclusión, Seguridad Social y Migraciones (2026)
- [Talent Partnerships](https://home-affairs.ec.europa.eu/policies/migration-and-asylum/legal-migration-and-resettlement/talent-partnerships_en), European Commission, Migration and Home Affairs (2025)
