Wednesday, 31 October 2018

LSE Politics and Policy Blog: Thermostatic public opinion: why UK anti-immigrant sentiments rise and then fall



Below is a copy of my blog post published on the LSE Politics and Policy site on October 29th, 2018

The Britain of today, it has been said, is far more sceptical and hostile toward immigration than the ‘forward thinking’, multicultural ‘cool Britannia’ inherited by Tony Blair in the 1990s. This claim certainly may have held up to evidence leading up to the turn of the decade, but what if now, rather than ‘more hostile than ever’, the British public is in fact becoming more open and tolerant of immigrants and immigration?

While it is often remarked than an ever-growing dissatisfaction with immigration underpinned the mobilisation and success of the BNP and UKIP, what if — almost completely paradoxically — a recent collapse in anti-immigrant hostility was in fact caused by the rise of these two parties? Lastly, what if — despite assertions to the contrary — trends in British immigration opinions have very little at all to do with the Brexit vote, and Brexit came at a time when immigration attitudes were already rapidly softening?

This is exactly what my latest research into British public opinion on immigration suggests. As part of a wider project about the political experience and representation of ethnic minority and immigrant origin groups in Britain, I have collected and extensively analysed trends in public opinion toward immigration over the last four decades. Figure 1 documents change in public opinion toward immigration in Britain since the early 1980s according to an aggregated measurement of almost 200 different survey questions from six international and national-level survey sources.


Figure 1: Public opinion on immigration in Great Britain (%)
Public sentiment about immigration has swung significantly since the 1980s.
Source: Stimson Dyad-Ratios algorithm estimation on 173 public opinion marginals. 27 question item series. Estimated line has Eigenvalue score of 66%, mean of 55.5, and standard deviation of 6.1.



The data, coming from the British Election Study, the British Social Attitudes survey, the European and World Values Studies, the European Social Survey, and the Eurobarometer, are ‘blended’ together using the Stimson dyad-ratios algorithm. (For more information on this method of aggregation and its advantages over single-source or one-poll research and reporting, see Stimson (2018), Jennings (2009) and English (2018).) Public opinion in Figure 1 ranges from 0 (completely open and tolerant of immigration) to 1 (completely hostile and anti-immigration).

Beginning in 1980, just over half of Brits would be considered to be generally hostile or negative about immigration. This falls, rises, and falls again by about 0.1 points over the late 1980s and early 1990s, before a low point in 1994 is followed by a meteoric rise in anti-immigrant hostility right up until the mid-2010s. A peak in 2010 ‘levels off’ to a mean of around 0.6 in the 2000s before suddenly and sharply plummeting through until 2018. Right in the middle of this sharp decline in public hostility toward immigration is the Brexit vote. We can visualise this remarkable ‘cliff edge’ in anti-immigrant sentiments in another way via Figure 2. This uses exactly the same data as Figure 1 but the time period is restricted to between 2013-2017 and the data are analysed quarterly instead of annually.
Figure 2: Quarterly public opinion on immigration in Great Britain (%)
Anti-immigrant feeling has fallen since 2013.
Source: Dyad-Ratios algorithm estimation on 51 public opinion marginals. 12 question item series. Estimated line has Eigenvalue score of 92%, mean of 47.9, and standard deviation of 7.0.



The fall from the early 2010s high through to the end of the study period is no less dramatic in this graph than the previous one. Here, the aggregated index score falls from its initial peak in an almost linear fashion with only a slight rise at the end of 2014 providing any significant deviation. By the time that the referendum comes along in the third quarter of 2016, aggregate public hostility toward immigration has already declined — and will continue to do so afterwards — by almost 20%.

The two graphs provide strong evidence substantiating the claims made by many commentators and academics in recent months: that attitudes toward immigration have indeed been softening since the time of the EU referendum. However, it also provides strong evidence that this softening has been much larger and more dramatic than previously thought; and has been taking place since well before the 2016 vote — disputing the idea that the Brexit vote has any causal connection to the recent reversal of a long-term trend in rising hostility toward immigration — as suggested recently by Michael Gove among others.

So how can we explain the rises and falls in aggregate anti-immigrant sentiment over the past 40 years? I argue that the findings evidence the existence of a ‘thermostaticcharacter to British public opinion, in that aggregate public opinion is responding to movements and changes in policies, discourses, and mobilisation regarding immigration.

Firstly, we can understand the rising hostility toward immigration across the late 1990s and 2000s as a ‘reaction’ to increasingly multicultural and liberalising diversity and immigration policies in the New Labour period. This was a period where ‘multiculturalism’ reigned supreme — the result of a movement which had been in place since the 1960s, when Britain (along with many other Anglophone nations) abandoned the assimilationist approach. Labour governments introduced meaty legislation aimed at protecting minority and disadvantaged individuals and communities from discrimination — culminating in the Equalities Act of 2010 — and famously allowed ‘open immigration’ from the 2004 EU accession countries (mostly from Eastern Europe) when other EU countries chose to enact temporary restrictions. As we can see from the rising public hostility toward immigration over the same period, the British public were signalling that this might have all been ‘too much’.

Riding this wave of anti-immigrant public opinion, the BNP gained unparalleled success for a British far-right party, culminating in 2008 and 2009. Between these two years, the party campaigned on an unashamedly and proud anti-immigration platform and won a London Assembly seat, two seats at the European Elections, and had around 50 local councillors dotted around the country. In 2014, UKIP won 163 council seats in the local elections, won the European Elections (taking a plurality of 28% of the vote and 24 MEPs), and were heading for their best result at a General Election the next year, (13% of the vote, 1 MP). But there was a sea change underway in immigration opinions.

After the success of the BNP and the rise of UKIP, David Cameron’s 2011 declaration of the ‘failure of multiculturalism’ and his ‘war’ on its proponents and policies, and a very public ‘battle’ between those in favour of immigration and those against it in mainstream politics, it appears that pro-immigrant sentiments grew in response.

If we understand public opinion as involved in a symbiotic, responsive, ‘thermostatic’ relationship with the political environment — where movements ‘too far’ in one direction by either one will be reciprocated by a movement in the opposite direction by the other — we can in turn understand the recent dramatic positive change in public opinion toward immigration as a response to the environment created by the BNP, Cameron’s ‘war’ on multiculturalism, and (latterly) the rise of UKIP. It, in turn, was ‘too much’.

This could involve a ‘crystallisation’ of latent pro-immigrant attitudes — people being stirred into taking an openly positive stance who were never comfortable with the restrictive position in the first place — and also the movement of ‘soft anti-immigrant’ view to ‘soft pro-immigration’ views. Both of these would be easily detectable in aggregate survey analysis.

And what does this mean in the context of Brexit? For one, it certainly does not mean that the Brexit vote itself was caused by a rising tide of anti-immigrant hostility, or that it was some culmination of a mass public movement against immigration. If we return to Figure 2, the Brexit vote came slap bang in the middle of the rapid decline in anti-immigrant sentiments from 2010 to 2017. For another, it also suggests that Brexit has had nothing to do with the recent ‘cooling off’ of hostility toward immigration. Rather, the ‘cooling off’ has been much longer and larger in trajectory than this, and the Brexit vote happened in the context of this wider movement.

Post-Election Piece - 2018 Locals

Below is a summary piece from the 2018 Local Elections

In all, not all that much has changed despite expectations that Labour might make some significant gains overnight. That said, there have been a fair few changes of seats within councils. At the time of writing, Labour are around 60 seats up on their 2014 performance, the Liberal Democrats up by around 50, the Greens just about in the positive, the Conservatives around 10 below their 2014 result, and UKIP down by almost 125 – losing in nearly every seat they contested at this election (as indeed they did in 2017).

According to the BBC’s ‘projected national share’, if the contest were to be held UK-wide then both of the Conservatives and Labour would have been polling at 35%, the Liberal Democrats 16%, and the other parties on 14%. On 2014, this constitutes a rise of 6% of the Conservatives, 4% for Labour, and 3% for the Lib Dems. A small swing for the Conservatives on 2014 (though a small swing to Labour from the 2017 General Election) making it a complete dead heat.

Overall, a rather disappointing night for Labour saw them struggle to make big gains, and also saw them lose control of Derby and Nuneaton & Bedworth councils. They did pick up a large number of seats in the capital and won overall control of Plymouth and Kirklees, but were unable to flip any of their target London boroughs in Westminster, Kensington and Chelsea, or Wandsworth, and missed out on taking overall control in North East Lincolnshire and Stockport. Labour did do very well in areas with significant amounts of young people (aged 18-34), with a rise of nearly 11% on 2014 recorded in the BBC ‘key wards’ where the young population was over 35% of all adults (according to the 2011 UK Census).

For the Conservatives, they will be disappointed to have lost control of Trafford and suffered a range of seat losses across London, but will be encouraged by results elsewhere in Peterborough, Great Yarmouth, Redditch, and Basildon where they gained control. According to the BBC key ward analysis, their main strength came in areas with high percentages of Leave voting, rising an average of 13 points in authorities voting to leave by 60% or more in the 2016 referendum, compared with just a 0.1% rise in areas with less than 20% Leave voting.

The Liberal Democrats did a bit better than expected, taking the Richmond-upon-Thames, South Cambridgeshire, and Kingston-upon-Thames councils from the Conservatives, and picking up seats in many areas across the country – particularly in Hull (nine in total there). The party performed especially well when it was competing in Conservative controlled councils, rising an average of around 5% in key authorities with majority-Conservative administrations (compared with around 1% in Labour-held councils). In many of these areas (including the above three councils) there was a substantial Remain vote in 2016.

For UKIP and the Greens, it was a tale of contrasting fortunes as they officially traded 4th and 5th places in English politics (in terms of seats). UKIP lost in every single contest they competed in bar three. Surprisingly, in one of these victories they unseated the Labour leader of Derby Council. The Greens have reasons to be cheerful in mounting a decent defence of what was a strong 2014 result for the party. They also made gains in places such as Trafford, Worcester, and across London. They were down substantially however in Norwich – a council where they have been losing seats for quite some time now having previously built a stronghold there.

To the local issued raised in my blog before the elections. In Sheffield, the trees saga played out exactly how many expected it to with the Labour administration losing five of its councillors to anti-felling candidates in the centre and west of the city. The Greens gained two seats (on top of defending their Nether Edge and Sharrow seat of Alison Teal) while the Liberal Democrats picked up three. Labour won a seat from UKIP, meaning their majority was cut by eight seats. A clear message was sent by many voters in the city regarding their dissatisfaction with the council’s tree felling priorities.
Meanwhile in Kensington and Chelsea, very few seats changed hands. Labour were up by around 7% across the borough and turnout surged by 6%, but there was only one gain for the Labour Party as they took the second seat in St. Helen’s – just over the road from Grenfell – by a comfortable margin. A small Grenfell effect then perhaps, but nothing at all seismic.  

The impact of the anti-Semitism row on the contests in Barnet and Bury produced some striking results. In Pilkington Park in Bury – the ultra-marginal ward with almost 25% Jewish population – the Conservatives won the seat from Labour with a 10% swing. They also saw a vote surge in Sedgley – where there is a 33% Jewish population – amounting to a 6% swing. In Salford, the Conservatives took the Kersal ward – over 40% Jewish –  from Labour on a big swing. In Barnet, Labour failed to make advances in crucial marginal wards such as Hale – with around 20% Jewish population – and West Hendon – around 10% Jewish population – which ultimately cost them control of the council. In terms of any fallout from Windrush, we did see boroughs with high density Black voters – and particularly Black Caribbean voters – tending to break heavily toward Labour, swinging away from the Conservatives. On average where there were 5% of more of the population from such backgrounds according to the 2011 census, Labour were up by nearly 10 points and the Conservatives just by 1.

Finally, in Swindon Labour were (as with many other targets) not in the end able to take the council. There was no evidence however of the voter identification trial supressing turnout; in fact, turnout was up overall by four points, which contrasts to no real overall rise in turnout on average across the country.

In all, this was ultimately a frustrating night for Labour considering their previously lofty ambitions and bold targets prior to the event. They quite clearly still have some convincing to do – particularly outside of London – that they are indeed a government in waiting. For the Conservatives, this was a much better night that was previously anticipated. For the Liberal Democrats and the Greens, there was cause for quiet and cautious optimism, while for UKIP the electoral wipe-out writing is all over the walls. And from Sheffield to Barnet there were fascinating local stories in play which added to the interest and showed once again how important neighbourhood factors can be in council elections.

Yorkshire Post: How the election played out across Yorkshire - according to University analyst.

Below is a copy of my piece for the Yorkshire Post, written shortly after the 2018 local elections and published on 4th May this year.

Yorkshire and the Humber was the setting for some fascinating local politics battlegrounds on Thursday, with Sheffield, Kirklees, and North East Lincolnshire all providing quite some interest. 

Dr Patrick English, a BBC elections expert and University of Sheffield post-doctoral researcher, analyses the local elections picture The region also saw the election of its first Mayor, with Dan Jarvis winning the contest on second preferences. In all, initial projections suggest that there has been around a 2.5% swing to the Conservatives from Labour in our region. 

The Conservatives had the best of it in terms of taking council seats, winning an additional 18. The Liberal Democrats made one net gain, and the Greens gained as many as they lost (two). Labour will be disappointed to find themselves down by nine and UKIP (rather expectedly) saw their regional seat total reduced by 11.


Despite some very interesting developments within councils, only one Yorkshire and the Humber authority changed hands; Kirklees went Labour from no overall control after the party took two seats from the Liberal Democrats. Labour’s vote share there rose by around 5% on 2014. 

Elsewhere, Labour maintained full control of Wakefield, Bradford, Hull, and Leeds councils, while the Conservatives comfortably held Harrogate (increasing their majority by 10) and Craven. Calderdale and North East Lincolnshire both remained in no overall control. 

North East Lincolnshire was a Labour target before the election, and they will have been disappointed not to take control of the council. Indeed, it was the Conservatives who made the most significant advances there, rising by 25% across all seats contested and taking seven of them. This was one of a number of Conservative advances from second positions in our region which also included successes in Leeds (up three seats) and Wakefield (up four). 

North East Lincolnshire was once upon a time a stronghold for UKIP, but on Thursday provided apt demonstration of their collapse across the country. While UKIP achieved 35.5% of the vote there in 2014, coming a clear 1st, this time around their vote share fell by around 30 points and there are now just three UKIP councillors on the authority.

In one of the most interesting stories across our region, the Greens and Liberal Democrats were looking to capitalise on the huge amounts of local anger and frustration over the controversial tree felling programme in Labour controlled Sheffield. 

The Greens will be pleased to have taken two seats from Labour and to have emphatically increased the majority of Councillor Alison Teal in Nether Edge and Sharrow – who was of course arrested and taken to court by her own council for protesting against felling in her ward – by nearly 1,500 votes. 

The Liberal Democrats also profited in Sheffield, taking three seats from their Labour rivals. 

Across the Steel City, the Greens were up by around 4% and the Liberal Democrats by around 3.5%. Interestingly, in every single ward making up Jared O’Mara’s Sheffield Hallam constituency, the Liberal Democrats won. The Liberal Democrats also had a very good night in Hull where they took nine seats away from Labour to substantially reduce their majority.

Britain Elects: Trees, Grenfell, Anti-Semitism, Windrush, and Voter IDs — All About the Locals

Below is a copy of my piece published with Britain Elects ahead of the 2018 Local Elections, on 30th April this year.

3rd May will see the first England-wide test of voter support for the country’s political parties since the General Election last June. Voters will head to the polls to elect councillors up and down the country, and in six of those areas there will also be mayoral elections. Though the general stories of Labour doing well in cities (particularly London) and with young voters and the Conservatives strengthening their grip on many pro-Brexit heartlands in the North and Midlands will most likely continue, there are many individual contests and battlegrounds with their own unique stories and nuances which may well buck — or exacerbate — the overall trends.


The General Story
Both the Conservatives and Labour look set to be up in terms of vote share on 2014. But while Labour are expected to make a fair amount of seat gains, the Conservatives look set to fall back. Recent polling over the last few days however suggests that the Labour gains may not be so large was anticipated perhaps a couple of weeks ago. Furthermore, though gains do seem likely, Labour might struggle to make a huge impact in terms of taking control (or removing opposition majority control) of local authorities this time around. As Professor Rob Ford pointed out on Saturday, many of the areas up for grabs on Thursday are already dominated by Labour councillors. They are also re-contesting a strong set of results produced by Ed Miliband’s Labour Party in 2014.

That said, there are a few places in which, if Labour are indeed having a good night, they can expect to move into the driving seat and would have cause to celebrate. They will be keeping a close eye on London where, after a strong General Election performance last year, they have some highly ambitious targets — including taking the Conservative controlled boroughs of Wandsworth and Westminster. There may well be further success for the party in other Councils such as Amber Valley, Plymouth, and perhaps North East Lincolnshire and Newcastle-Under-Lyme.

The Conservatives seem set for a quietly disappointing showing, but given the current political climate (and the usual painful nature of local election nights for governing parties) they may be fairly happy with a result of anything around 100 seat losses. They might, if they are able to continue their strong pickup of ex-UKIP voters, even have something to cheer about in places such as Thurrock, Basildon, and Great Yarmouth (seeking overall control in each).

According to forecasts, the Liberal Democrats are expected to, once again, have a frustrating night and make only a handful of limited advances. The may be looking to Maidstone (seeking overall control) and Stockport (seeking to displace Labour) for some good news. London may provide some further encouragement, with the prospect of regaining substantial ground in Richmond (where they were pipped to the post in 2017 by the born-again-Conservative Zac Goldsmith) and Kingston upon Thames on the cards.

UKIP look set for another terrible night. They will, after a rapid two-year decline, fall behind the Greens in terms of council seats in England on Thursday; all but a dreadfully poor night for the Greens, which is not expected, should see them cement 4th place on the English local representation leader board.

Away from the general patterns, there are four stories worth picking out where local factors and controversies could provide some very interesting results and trends.

Trees in Sheffield

Sheffield City Council: Labour control, 30 seat majority, 28 seats up.

Firstly, a significant test of the extent to which voters ‘think local’ in council elections when a scandal is unfolding and ongoing around them. The Labour Administration in Sheffield is currently under huge amounts of scrutiny over their ongoing controversial tree felling programme. Such has been the strength of opposition and controversy that the events have hit national and international headlines. In 2012 the Council contracted a private company, Amey, to cut down over half of the city’s 36,000 trees over a 25-year period as part of a ‘highways improvement’ programme worth £2.2 billion. Needless to say, this has not gone down well with many of the local residents from a city which prides itself on its green credentials. Rallies and marches against the destruction of health street trees have attracted thousands of residents, young and old, and the support of some very high-profile supporters — including Michael Gove, Chris Packham, and Jarvis Cocker.

Opposition parties — namely the Liberal Democrats and Greens — will be seeking to capitalise on local anger and frustration. Intriguingly, in six wards their campaigns will be assisted by a non-partisan campaigning group which was founded this year with the specific aim of putting boots on the ground to unseat Labour incumbents in marginal wards — the “It’s Our City” campaign. Equally, the non-partisan “Sheffield Trees Action Groups” (or STAG) campaign, over 10,000 members strong, are also actively supporting the removal of the current Administration. Labour are mounting a total of 19 defences in Sheffield, meaning that they could theoretically lose control of the council in the face of this widespread opposition. This is however very unlikely, with Labour defending some very friendly (as well as some not so friendly) seats.

A further layer of interest in Sheffield comes from the sole Green seat being defended this year in the city. The current incumbent there is Councillor Alison Teal — one of four Greens on the council. Cllr Teal has a majority of just eight votes, and has been right at the heart of the tree campaign — as well as one of its most vocal critics, she was arrested and then taken to court by her own Council for protesting against a tree felling in her ward (Nether Edge and Sharrow). She was subsequently cleared of any and all wrongdoing. Nether Edge has been the centre of many of the most intense and controversial episodes of the tree felling saga, and so results in Cllr Teal’s leafy city-centre seat as well as other tree felling flash-points of Broomhill and Sharrow Vale, Ecclesall, Gleadless Valley, and the Crookes and Crosspool and Dore and Totley wards (both at the heart of Jared O’Mara’s Sheffield Hallam constituency) will definitely be worth keeping an eye on for an anti-Labour backlash in a city which turned all red only last year.

Grenfell Tower Fallout in Kensington and Chelsea

Royal Borough of Kensington and Chelsea: Conservative control, 24 seat majority, all seats up.

The story here is of course the political fallout from the tragedy of the Grenfell Tower fire. With the Public Inquiry ongoing, former Kensington and Chelsea Council leader Nicholas Paget-Brown has already quit and the huge amount of anger, frustration, and shame felt by many residents toward their local leaders is expected to make it a tough night for many Conservatives defending their seats. Indeed, Jeremy Corbyn has made clear his ambition to take the Council away from the Conservatives. That would require them to overturn a 24-seat majority.

Unlikely as that seems, as with the rest of London all of Kensington and Chelsea Council’s seats (50) are up for election, which does increase the possibility of a change of overall control. Furthermore, Labour did pull of quite the shock and take the Kensington constituency from the Conservatives, by a margin of 20 votes, in 2017.

Taking the council may be much more difficult however; by and large, there were not that many wards where Labour were all that close to the Conservatives in 2016, particularly towards the Thames in the South of the borough. Grenfell Tower sits in the Notting Dale ward, where Labour already hold all three council seats by a healthy margin over the Conservatives. Neighbouring wards may though provide the best opportunity for Labour to work toward reducing the Conservative majority. To the north is St. Helen’s ward, where the two parties were neck and neck in 2016 — taking one seat each. That one Conservative seat just over the road from Grenfell would be an obvious first target. Meanwhile, to the South the Conservatives comfortably won both seats in the Norland ward in 2016 but might find themselves in trouble there too if voter anger spills over. Elsewhere, less than 25% swings — not unthinkable given the circumstances — would be required in Chelsea Riverside, Earl’s Court, Pembridge, and Holland in order for Labour to take a potential 10 further seats. If Labour took every Conservative seat in each of the above six wards, they would indeed take control of the Council. Stranger things have happened.

Jewish Populations Barnet and Bury

Borough of Barnet Council: Conservative control, 2 seat majority, 21 seats up
Borough of Bury Council: Labour control, 8 seat majority, 17 seats up

Labour may well perform substantially under-par on Thursday in Councils such as Barnet and Bury where a number of wards are home to substantial Jewish populations. In fact, the damage and rift caused between Labour and many parts of the Jewish community as the anti-Semitism row continues could cost the party control of Barnet, where a two-seat gain would see them replace the ruling Conservative administration. Taking marginal wards such as Childs Hill and Hale could be crucial to Labour taking overall control, but the seats’ substantial Jewish populations (around 20% according to the 2011 Census) may well stop Labour doing so if the backlash is significant enough.

In Bury, Labour are already in fairly strong control of the council and are defending in some very strong Labour areas. The impact of a reaction to the anti-Semitism row therefore may not be so consequential here. That said, look out for wards such as the ultra-marginal Pilkington Park (Labour less than 1% ahead in 2014, approximately 25% Jewish) and the much more safer seats of Sedgley (approximately one-third Jewish) and St. Mary’s (around 10% Jewish) as potential backlash lightning rods. Relatedly, though Labour are generally already well in control of areas with substantial ethnic minority voter populations, in the wake of the Windrush scandal — which only yesterday claimed the scalp of the Home Secretary — the Conservatives may see below-par performances even by their expected London standards in high-density black voter boroughs such as Lambeth and Lewisham.

Voter ID in Swindon

Swindon Borough Council: Conservative control, 4 seat majority, 19 seats up

Swindon is one of the five areas in which a pilot voter identification scheme will be taking place on Thursday. What makes Swindon stand out is that it is very much a Labour target, and they need only four additional seats to take it (two to become the largest party). It will be interesting to see if turnout is significant affected by the trial, and how many stories come out of registered voters being turned away from polling stations due to a lack of suitable identification. With control of the council in the balance and Labour, then this pilot and its timing could create some controversy if it goes badly. Keep an eye out for results and turnout stories coming in from the Covingham & Dorcan, Haydon Wick, Lydiard & Freshbrook, and Shaw wards, where Labour stand a good chance of making the gains necessary to take control of the Council.