Britain keeps pretending that all migrants are fiscally alike. They are not.
Across the Atlantic, the United States has quantified those differences in a way British policymakers can only envy. In October, economist Daniel Di Martino at the Manhattan Institute published a 2025 update on the fiscal impact of immigration to the United States. He found that Indian immigrants are, on average, the most economically beneficial group in America, each (including their likely descendants) reducing the U.S. national debt by over $1.6 million over 30 years. Chinese immigrants rank second, cutting projected U.S. debt by more than $800,000 per person over 30 years, followed by Filipinos, who trim it by over $600,000, and Colombians by about $500,000.
In contrast, the average Salvadoran immigrant increases the U.S. debt by over $50,000, and the largest immigrant group, Mexicans, add about $10,000 to the national debt each over 30 years. Quite the difference.
We simply do not collect or publish the right data
The policy implication is obvious. In a system with a limited number of visas, holding all else equal, any sane American government would want more Indians and Chinese, and be very cautious about large inflows from Central America.
This is exactly the point economist Garett Jones, author of The Culture Transplant, has been making for years. His rule of thumb for a points-based system is simple: “more net revenue, more points”.
Could Britain apply the same logic? In principle, yes. In practice, no. We simply do not collect or publish the right data.
The few serious UK fiscal studies are painfully crude. Official fiscal impact data lumps migrants into just two buckets: European Economic Area (EEA) versus non-EEA. A report for the Migration Advisory Committee (which advises the UK government on immigration policy) estimates that in 2016/17 non-EEA migrants as a group imposed a net cost of roughly £9 billion on the public finances, while EEA migrants reduced government borrowing by about £4.7 billion.
Alas, this data is basically useless because it means that a highly paid software engineer from Israel and a low-skilled migrant from Somalia are both in the non-EEA category. One is highly likely to be a lifetime net taxpayer; the other is much more likely to rely on benefits. Yet, if this is truly the only data available, then from the Treasury’s perspective, these immigrants are identical.
We do have some hints, via ethnicity not nationality data, of which groups are high performers in the UK. Indian households are the highest earners in the country, and Chinese households are the least likely to receive any kind of state support: only 25 per cent of Chinese families received benefits between 2018-2021, compared to 54 per cent of White British families. Everything we know about earnings, benefits usage, and educational outcomes suggests that Indian and Chinese migrants in Britain are, on average, exactly the sort of “massive net revenue generators” Jones and Di Martino identify in the US. Yet unlike the US (or Germany or Denmark or the Netherlands), in Britain we have no official fiscal data by country of origin to prove it.
Why does country of origin matter so much? Because, as Jones argues in The Culture Transplant, migrants do not arrive as blank slates. They bring persistent patterns of behaviour and belief — about saving, trust, family and the state — that last for generations. A 2018 study of three generations of immigrants in the UK found that saving behaviour in Britain tracks with the saving norms of the country of origin, and that this relationship persists well into the third generation. And another study shows that ancestry composition can explain a large chunk (around 60 per cent) of global national income differences. History casts a long shadow.
But what about education level? After all, Di Martino’s report shows that bringing in a 30-year-old immigrant without a high school diploma costs the US federal government more than $130,000 over 30 years, while admitting a 30-year-old immigrant with a college degree generates a net fiscal benefit of about $1.6 million over the same time horizon. So, does that mean that we should focus on education level instead of country of origin?
Probably not, because even when education is held constant, origin still has significant explanatory power. This means that even if we only admitted college-educated migrants, even in that elite slice, people from different origins have different average attitudes to saving, redistribution, trust, and work. Those differences show up in how much they save, how much they draw on welfare, and what policies they support — and these behaviors persist for generations. Indeed, the reason Indian and Chinese immigrants tend to be over-represented among STEM graduates, is not random luck.
We are talking in terms of average trends and not universal truths. Nonetheless, policy should operate on probabilities
None of this means that any individual’s prosperity is fated by their nationality. Country of origin is by no means a perfect predictor of an immigrant’s fiscal impact. We are talking in terms of average trends and not universal truths. Nonetheless, policy should operate on probabilities. And the probability that a randomly chosen immigrant will found a high-growth company, or quietly pay six-figure sums in lifetime taxes, clearly varies by country and cultural background.
There is no good reason why Britain should be uniquely incapable of calculating those probabilities. We don’t need to slam the door shut, and we shouldn’t fling it wide open either. But if, as everyone tacitly accepts, this small island can only let in so many people, the least we can do is choose them wisely.
