Sunder Katwala (sundersays)

There is prettu strong evidence the UK would be attractive. I agree with your numbers. There is no net migration impact and no fiscal impact if flows are even & there is an uplift to net migration and modest fiscal + if there is a UK inflow. I think good to promote the reciprocal model

This "more to do to make levels sustainable" was yesterday. The government needs much better policymaking than this and a proper process of planning immigration policy, and taking more seriously its interaction with its other policy goals
bsky.app/profile/sund...

Sunder Katwala: What happens to the politics of immigration when numbers start falling? | Conservative Home The politics of immigration are also driven more by asylum than by overall visa numbers. If there is a visible lack of control over small boats in the Channel and asylum accommodation, significant fal...

For ConHome March 2025. The big falls in net migration back to 200-300k were known about pre-election, but there was little expectation they would drop below 100k/zero
conservativehome.com/2025/03/17/s...

Keir Starmer’s Labour government has not decided what it wants to say and do about immigration numbers once they start to come down. Beyond delivering its manifesto pledge of an overall reduction, the government has no public view of what the right level of immigration to Britain should be. Home secretary Yvette Cooper and chancellor Rachel Reeves have long opposed a new migration target. They have no crystal ball of the economic or geopolitical conditions of 2030 – and no control over emigration levels. The politics of immigration are also driven more by asylum than by overall visa numbers. If there is a visible lack of control over small boats in the Channel and asylum accommodation, significant falls in net migration may have little impact on public perceptions. Whatever level the current government achieves, the opposition parties will say it should have gone further. Badenoch wants a legal cap on net migration – though she has not yet adopted Robert Jenrick’s proposal that it must be set at 100,000 or lower. Reform leader Farage believes that ‘net zero’ would be a good level for immigration to the UK. Yet, for parties that aspire to govern, this means grappling with the dilemmas of control if the slogans are to be turned into policies.

"If there is a visible lack of control over small boats in the Channel and asylum accommodation, significant falls in net migration may have little impact on public perceptions".

My view a year ago was that big falls in numbers might not shift the politics much, but the govt needed to respond

The fiscal numbers depend on letting it rise.

The Home Office is working to drive it down further (seemingly without clocking where it is already heading) + is inventing fiscal numbers that aren't true on settlement when its policy is creating a black hole

bsky.app/profile/sund...

OBR keeping projections well above the actuality and the trend. Otherwise the Chancellor would have sacrificed most/all of her fiscal headroom if 2026=2027 net migration is 200k or 300k below this!

OBR have average 235k 2026-30.
Cost is about £7bn for each 100k cut.

obr.uk/box/the-impa...

Home Secretary told Select Committee on Feb 4th it is "high on any measure" - a vibe based on treating figure for the year to June 2025 (net 205,000) which was eight months ago as the current level, implicitly adopting May's missed net migration target of 100k - without realising she has met it!

Neither Keir Starmer nor Shabana Mahmood have tried to say what the sustainable level of net migration is

Home Secretary told select committee net migration remains "high by any measure" (Feb 8th) and (yesterday) "more to do" to bring it down.

There is a fiscal cost to eliminating it by mistake!

It is better for a democratic party to narrow the gap by increasing support from the groups you under-perform with than to narrow the gap by experiencing a general collapse in support

Ipsos 2024
www.ipsos.com/sites/defaul...

The white: minority ethnicity gap for the Conservatives had never been lower in the first half of 2016 (before the referendum) then rose by 2019, but fell in 2024 (when their white vote collapsed but their minority vote was a bit stickier)

The Labour gap is prob lowest right now!

Ipsos produced these 1996-2024 estimates from aggregate polls. It is not the same as having representative minority data across the period.

Strategic objective of right is to turn ethnic minorities (esp new UK-born cohorts) into normal swing voters released from Powell/Tebbit community allegiance to Labour. Minority trends in education, work look promising for that by 2015 - as long as the Tories don't turn anti-graduate + anti-youth!

Yes, fair point, though the "can I guess how you vote from your skin/faith" is not entirely affected by whether there were socio-economic drivers of why I could guess that. Cons by 2010 underperforming socioecon expectations across minority groups, but Indian/Chinese progress narrowing that a bit

Weren't necessarily planning to have a vote for lots of it (some bits may need to go in a bill, so that may delay). It will surprise people about how immigration rules work.

Not having a vote is a risk, as it means people have to consider doing things like resigning as PPS, ministerial roles

My understanding/memory of what Embes 2010 was that ethnic minorities were still as much sticking to Labour *despite* social class as much as because of social class - ie, it was a party identity issue similar to Scotland. Would expect this to change over time & Cameron tries hard to unblock it

That the 1964-83 minority voters are much more likely to be migrants as well as minorities makes that period of "immigration/race relations" being the same thing until we recognised they were not the same thing once people are born here a bit different

The Embes 2010 is a v high quality study. Earlier period anecdote-dependent on what is obvs true. Ipsos later constructed these 1990s-2020s estimates from aggregating polls. While 1997 is an unusual high for Labour, it seems v likely 1974/1984 quite like 2017-2019 but more so

In aggregate, ethnic minority citizens less monolithic than media thinks
- more socially cons, more pro-faith but more pro-migration
- more likely to experience child/family poverty
- more educ: more likely uni graduate, esp younger cohorts
- quite a lot more Reform-sceptic (1/6 may support Reform)

There is a big ethnicity effect in Labour voting and Conservative voting 1964-2010 (when minority party identification is high, later, in the first generation and partly the first UK-born generation, before it collapses in the second/third UK-born gen: which I take as an integration effect)

It would be obvious only if you don't listen to media and political commentary about ethnic minorities, but do things like talk to people from those backgrounds at work, college, church etc

But race (not being white) was a pretty strong predictor in England from 1974-2005/10: stronger than almost anything else, though 1/6 were voting on the right

A media blindspot in spotting that age, class, income, place will matter within minorities too

Some interesting differences (eg being a graduate sends you left if white British but correlated with more likely to vote right-of-centre if black/Asian (compared to the minority group aggregate)

In a binary referendum, white people were about 53-47 Leave, ethnic minorities were 2-1 (Asian) and 3-1 (Black) for remain, but on a lower turnout. These were mostly very reluctant agnostic remainers, with some anxious leavers: turnout above GE for white British, below GE for minorities

Would true (generational shifts) without the fragmentation, and the fragmentation makes it much more so.

Ethnic minority party *identification* still v high in 2010, when generally low (class/dealignment). But collapsed across generations after 2010. Fragmentation also results from weak party id

Labour may well want 400k Muslim votes (40% again) but weird discourse about a bloc vote determining its policy when 20% current share (40% general election share) of a 6% group within its 17% overall vote (33% general election share). In fragmented party system, ethnic vote gaps narrower than ever

Labour probably had something like 800,000 South Asian Muslim voters in 2019 (about 1 in 12 of its 10 million votes) but closer to 400-450k by July 2024 (about 1 in 20-25 of its 9.7 million voters) and may have 200-250k Muslims in its 4.5 million loyalists in current polls (a similar ratio).

Minorities report: the attitudes of Britain's ethnic minority population - UK in a changing Europe This report looks at the diversity of political opinion, social values and economic preferences not just between Britain’s white and non-white population, but between different ethnic and religious gr...

So the pattern of attitudes between minority groups *and within each of them* - by age, education, class, geography, worldview, faith identity, practice/secular worldview - is more contested than ever, yet the discourse gets simpler
ukandeu.ac.uk/reports/mino...

In 2026
> a 1/3 chance a white voter is Reform; 60% chance they are not. (Education, age, geography wd help)
> about 1/3 chance a black voter is Labour now, maybe 1/4 to 1/3 Green, about 1/5 on right
> Asian Muslims have competing Green/Indy options (50-60%), maybe 1/4 to 1/3 Labour, 1/10 on right

Firstly, 2024 sees Labour gets lowest ever minority share

< 50% of all minorities for the first time,
< 50% of Asian votes for first time
< 50% of Muslims for second time (with 2005)
but still the plurality choice
- > 50% of black voters

Secondly, Labour has now lost half of its voters mid-term

I think we would encourage a less polarised politics if, over this next generation, it got harder rather than easier to guess somebody’s party preferences from their ethnicity and faith alone. My “one nation test” for any party that aspires to govern is that no citizen should feel any tension between supporting that party and their colour or creed. For all of their diversity at the top table, the post-Cameron Conservatives risk going backwards again with black British and British Muslim citizens, in particular. Starmer’s challenge may be both to reconnect with British Muslims, beyond this conflict, and to rebuild trust with Britain’s Jews after Labour’s failures on anti-Semitism – seeing no contradiction between the two.

Is it getting harder or easier to guess somebody’s party preferences from their ethnicity + faith alone?

It might surprise most media commentators that it is in fact now harder to guess party vote from just having ethnicity/faith to go on in 2024-26 in Britain that at any point in last 50 years?