It is almost trite at this point to talk about the British voters’ opposition to immigration. Time and time again, polls have reflected strong opposition to the numbers of people coming to settle in the country. For the electorate, Brexit itself was mostly driven by this question, not trade, regulation or fishing rights.
Thanks to YouGov polling conducted in 2016, shortly after the Referendum result, and again earlier this year, we can now measure how this opposition took place on a per-nationality basis and compare it with the numbers of people who came subsequently with a different demographic skew.
To make a fair comparison, we rebased the polls to exclude those who responded ‘don’t know’ to any given country — this enables us to look only at respondents who expressed an opinion about immigration from a particular country of origin.
The results demonstrate that the bulk of post-2016 flows were of nationalities that voters had explicitly said they wanted reduced. Using the Home Office’s Migrant Journey dataset, which tracks people as they move through the immigration system, we can proxy flows by nationality and match them against the opinions expressed in 2016. Four of the five nationalities with the largest number of Migrant Journey starts between 2017 and 2025 were among those towards which there had already been substantial public opposition.
In fact, around two-thirds of the covered journeys came from countries where a majority of respondents expressing an opinion wanted fewer or no immigrants in 2016. This represents 2.6m migrant flows. A further 885,000 journeys came from countries where between 45% and 49% wanted immigration reduced.
There is an obvious objection to using the Migrant Journey data alone. A ‘journey’ does not necessarily represent a unique long-term immigrant, and the same individual can begin more than one journey depending on their interaction with the immigration system.
Fortunately, the same broad pattern appears in the Office for National Statistics’ estimates of long-term immigration.
Looking at the nationalities for which the latest ONS figures allow a direct comparison, India is again by far the largest flow, followed by Nigeria and China, with Pakistan also substantial. The precise numbers differ because the Home Office and ONS datasets measure different things, but the nationality ranking is remarkably similar.
Whether we look at immigration-system journeys or estimated long-term immigration, the same underlying picture emerges: some of the largest flows were from nationalities towards which substantial public opposition had already been recorded.
This also makes it difficult to argue that the recent ‘Boriswave’ made immigration from these countries unpopular after the fact. For several of the largest nationalities, it was already unpopular before the larger flows took place.
In 2016, after excluding those who expressed no opinion, 64.5% wanted fewer or no Pakistani migrants and 66.2% said the same of Nigerian migrants. The equivalent figure was 52.7% for India and 48.6% for China.
By 2026, opposition to Indian immigration had risen to 56.4% among those expressing a view, Chinese immigration to 50.6%, and Pakistani immigration to 66.7%. Nigerian immigration was the exception, falling slightly to 63.2%, although remaining very high.
Other nationalities moved more dramatically. Russia and Israel, for example, experienced large increases in opposition, although comparatively few migrants from either country arrived in Britain over the period examined. Subsequent geopolitical events may well have influenced attitudes towards both.
It also turns out that Brits in 2016 did not want to replace European workers with Commonwealth migrants — rebased for ‘don’t knows’, Poles went from 45% favourable up to 66% in the decade since the Referendum result. Simply put, they wanted lower immigration from Poland, and were even more opposed to migration from India, Pakistan, Nigeria etc. This will be explored in closer detail in an upcoming Cambridge Circus Research paper.
There is another interesting feature in the polling. Among the countries with the largest subsequent flows, the share answering ‘don’t know’ generally fell while explicit opposition increased. For India, ‘don’t know’ fell from 26% to 22%, while the raw share wanting fewer or no Indian migrants rose from 39% to 44%. Similar movements occurred for China, Pakistan and the United States.
There is an important caveat. These are separate cross-sectional polls rather than a panel following the same individuals, so we cannot say that particular undecided voters moved into opposition. But at the aggregate level the movement is nevertheless suggestive: opinions became more defined, and for four of the five largest subsequent flows they became more restrictive.
The public had been asked about individual nationalities, and in many cases had expressed clear opposition. Yet some of the largest flows that followed came from precisely those countries.
Across the 17 comparable non-EU nationalities examined here, around two-thirds of covered post-2016 journeys came from countries where a majority of those expressing an opinion had already wanted immigration reduced in 2016. Almost 88% came from countries where that figure was at least 45%.
The verdict is clear. The British people were explicitly asked what kinds of immigration they wanted less of and answered clearly. What followed bore remarkably little relationship to those preferences. The information was there; it was simply ignored.
Any government that wants to restore public confidence in the immigration system will have to correct this grave error.
This article was written by Hugh Percy, a researcher at Cambridge Circus Research. Have a pitch? Send it to submissions@pimlicojournal.co.uk.
If you enjoyed this article, please consider subscribing. If you are already subscribed, why not upgrade to a paid subscription?