20 June 2016

Summary/Key points

The EU referendum remains finely balanced and therefore hard to predict – over the past couple of weeks Leave appeared to be in the ascendency, but the most recent polls released over the weekend suggest Remain has been able to make up the ground. Open Europe has therefore published a comprehensive guide of what to look out for on the night of June 23 as the results start coming in. Using publicly available data on EU voting attitudes from the British Election Study, as well as demographic data, we have been able to gauge how favourable individual voting areas are towards Brexit. This will allow us to place individual results in context.

We have also picked out 20 areas from across the country that we think are worth paying particular attention to as they could give us a good idea as to the broader national picture, and also because they represent key sub-plots in the referendum drama:

  • Leave is expected to perform well in less affluent areas populated by lower skilled and older voters, particularly in Eastern England, while Remain is expected to do well in cities, particularly those with a large proportion of university graduates. Closer than expected results in areas declaring early could give us an indication of how the national vote has gone.
  • Watch out for traditional Labour strongholds in Northern England, the Midlands and South Wales to see whether much of the core Labour vote will end up supporting Leave as is being widely forecast.
  • Significant regional variance is expected, especially between the different constituent parts of the UK. Scotland is expected to be heavily for Remain; a significant divergence with the result from England and/or the UK-wide result could lead to tensions in the future. In Northern Ireland, the degree of divergence between nationalist and unionist voters will also be noteworthy.
  • Areas such as Swindon and Nuneaton, where swing voters broke for the Conservatives in the 2015 General Election, will once again be important to watch.
  • Turnout will be important, especially in metropolitan areas where Remain are relying on a good turnout among young voters.
  • Expect to have a good idea about the result by 5am unless it remains incredibly close.

160620_OE_Key_areas itemprop=

How will the referendum count be organised?

The referendum count will be organised into 382 counting areas: the 380 local government authorities of England, Scotland and Wales as well as one each for Northern Ireland (although results will be broken down by its 18 parliamentary constituencies) and Gibraltar. These counting areas range hugely in terms of their electorates – Birmingham is by far the largest area with around 700,000 eligible voters, while the Isles of Scilly is the smallest with around 1,600.

Polls will open at 7am and close at 10pm (though it is worth remembering that as many as one in five votes may have already been cast by post). There will be no exit poll and the verification and counting process will begin immediately, with individual results being declared throughout the night. The national turnout figure will be updated based on individual results as they come in. Unlike in parliamentary elections where the margin of victory is immaterial, here every vote counts towards the national total.

Background – Where are the Remain and Leave strongholds?

There is a paucity of reliable polling data broken down by local authority areas – the best is the British Election Survey (BES) data from May 2015 compiled and kindly made public by Professor Chris Hanretty from the University of East Anglia. The BES is well regarded because of its large sample size compared to standard polls (around 30,000 respondents) and its perseverance in contacting hard-to-reach voters, thereby getting a more representative final result. The actual voting intention figures are of little use given how out of date they are, but they enable us to rank individual voting areas in terms of their underlying favourability towards Brexit relative to each other.

In reality, support for Remain and Leave will not have changed uniformly around the country over the past year, but the ranking will nonetheless give us a sense of whether Remain and Leave are over-performing or under-performing on the night in their relative geographical strongholds.

160620_OE_Top10_Leave_Remain itemprop=

We can clearly see that the East of England is very fertile ground for Leave while the strongest areas for Remain are in London, large Scottish cities such as Edinburgh and Glasgow, as well as affluent university towns such as Oxford and Cambridge. In contrast to Leave, the Remain strongholds are much larger in electoral weighting so getting these voters out will be crucial to offset the potential loss of a greater number of smaller areas. Both sides will be aiming to win their respective strongholds by comfortable margins. If this is not the case – e.g. if Leave perform strongly in London or Remain in Essex – it would suggest they are in trouble nationally.

We also have a second source of data – Sky News have compiled their own pro-Brexit ranking based on analysing surveys conducted between May 2015 and February 2016 to reveal the demographic groups most likely to vote to Leave, and then identifying the areas in which they are most prevalent.

There is considerable overlap between the two rankings but also some notable differences; the demographic data suggests the top 10 Remain areas are exclusively in London while the top 10 for Leave become a bit more diverse, featuring Copeland and Knowsley in the North West. These areas come in at 126 and 281 in the BES ranking respectively, underlining the degree of uncertainty surrounding the result.

(Click on the image below to see a zoomable version of the map)

160620_Open_Europe_UK_Top10_Areas_Leave_Remain_FINAL_web_version itemprop=


How the night will unfold and what to look out for

The first declaration will most likely be Sunderland, as is normally the case in parliamentary elections, at around 12:30am (estimated declaration times are courtesy of the Electoral Commission). Sunderland, the 40th largest counting area with some 200,000 eligible voters, is an interesting first result as it is a safe Labour seat with considerable numbers of older, lower income voters without university degrees, all strong indicators of support for Brexit. UKIP finished second in each of Sunderland’s three parliamentary seats with around 20% of the vote, and there have been many recent reports ‘from the doorstep’ that traditional Labour voters in less affluent areas are declaring for Leave in large numbers, largely motivated by concerns over migration. Sunderland will be the first opportunity to verify these reports.

Sunderland comes in at 174 on the BES ranking and at 29 on the demographic ranking so we would expect Leave to win here if they are also on course to win nationally. The margin of the win and the turnout will also be important; a narrow victory for Leave might suggest they will struggle to breakthrough in less favourable areas while a victory by any margin would be very good news for Remain. The risk for Leave is that despite winning over such voters to their cause, they then fail to turn up on the day; turnout among lower income groups tends to be well below average.

From the North East we are likely to head to Wandsworth in inner London. This should be an easy Remain win – it scores 366 on the BES ranking and 371 on the demographic ranking. Again, the margin of victory will be interesting, and it should give us an indication of how inner London as a whole has voted. Anything other than a comfortable victory in such favourable territory would be very concerning for Remain. Soon after we should also get the result from the City of London – it comes in at 380 and 374 on the respective rankings, making it the most anti-Brexit area of the country, albeit one with fewer than 7,000 voters in total.

At around 1am we should get results from Swindon and Oldham. Swindon comes in at 78 and 105, making it the kind of area Leave should be aiming to win. That said, Swindon also contains many of the kind of Labour/Tory swing voters who have swung towards the Conservatives under David Cameron – and perhaps a good chunk of these middle-income, mid-30s voters will end up breaking for Remain on the day. In contrast, Oldham – which comes in at 136 and 104 on the respective rankings – is more of a safe Labour area but as in Sunderland, with the kind of demographic makeup that leans towards Leave.

At around this time we should also get the result from Gibraltar which we expect to be a big win for Remain, albeit with a total electorate of just over 20,000 up for grabs, despite its symbolic importance, it will not be very significant in terms of the overall result.

Between 12.30am and 1.30am we should get a few results from Northern Ireland, including from the four Belfast parliamentary seats which collectively contain over 250,000 voters. Polling data from Northern Ireland has been limited (and we have no BES data) but the data we do have suggests Remain is in the lead, potentially due to concerns Brexit would hit the area particularly badly and fears it would lead to the restoration of a hard border with the Republic. The results from Belfast will reveal whether this is indeed the case. Also, with two of the Belfast constituencies are held by nationalist parties, and two by unionist parties, it will be interesting to see how this correlates with referendum voting; will we see a stark split along nationalist and unionist lines?

From this point on more and more results should be coming in including several large Northern and North Western areas including Newcastle which comes in at 307 on the BES-based ranking and 148 on the demographic ranking at around 1am, followed by Wigan (130/54), Salford (243/102) and Stockport (289/212) at around 1.30am. These results should provide further clarity on whether the reports that traditional Labour areas in the North will break strongly for Leave are true or not. If so, expect this to be one of the dominant stories of the night, with potentially far-reaching implications beyond the referendum result itself.

At around 1.30am we should also get our first Scottish results – with both Stirling and Eilean Siar (the Western Isles/Outer Hebrides) due to declare. Stirling comes in at 368 and 324 in the respective rankings while Eilean Siar comes in at 365 on the BES-based ranking (no demographic data available). Both should be Remain wins, but the margin could reveal whether the polls predicting a comfortable sweep for Remain across the whole of Scotland were accurate or not. Dumfries and Galloway (296/139), due to declare at around 3:30, could be Leave’s best chance of an outright win in Scotland – failure to carry any areas north of the border would result in a faultline which the SNP will would not hesitate to exploit.

From Scotland we go to Wales with Denbighshire in the north and Merthyr Tydfil in the south, both of which are due to declare around 1.30am/1.45am. Both appear marginally more favourable to Remain based on the BES data – coming in at 248 and 240 respectively – but the demographic makeup could benefit Leave with Denbighshire coming in at 129 and Merthyr Tydfil in at 96. Recent polls suggest the vote is neck and neck in Wales and as in Labour’s northern heartlands, there are reports of traditional Labour voters breaking for Leave. It is notable for example that UKIP did well in last month’s Welsh elections, winning 7 seats on 13% of the regional vote (up 8.4% on 2011). The decline of coal mining and steel production has been a dominant theme in these areas – Merthyr Tydfil has one of the highest unemployment rates in Wales for example. Neath Port Talbot (298/66), due to declare at around 2am will also be a result to look out for – the Leave camp have sought to link the struggles of the area’s steelworks to EU membership, arguing that outside of the EU such industries could be better protected from global competition.

From around 2am we should get another burst of results from all around the country. Key results to look out for include Basildon, Harlow and Castle Point in Leave-friendly Essex – which come in at 15/34, 28/13 and 4/8 in the respective rankings. If Remain are having a good night, they could win in South Norfolk (208/96) while they will definitely be looking for solid wins in the likes of Oxford (376/302) and several London boroughs including Westminster (350/376), Lambeth (375/365) and Merton (343/353). Barking and Dagenham (158/255), due at around 2.30am could be one of Leave’s best prospects of winning in London along with Havering (25/344). Both areas have seen strong UKIP performances in parliamentary and local elections.

Another interesting result will be Nuneaton and Bedworth; Nuneaton is a key bellwether parliamentary seat, and Labour’s failure to win it under Ed Miliband was seen as emblematic of their failure to connect with swing voters. According to the BES-based ranking it is among the more sceptical parts of the country – coming in at 83 – but its demographic makeup appears to be marginally favourable to Remain, coming in at 203. This area will reveal whether David Cameron has been able to motivate the kind of swing voters that pushed the Tories over the line in 2015 to come out for Remain.

From 3am till 4am the results will be coming in thick and fast from all across the country. At around 3am we should get the result from West Oxfordshire (285/181), a relatively wealthy, strongly Tory area of the country which includes the Prime Minister’s parliamentary seat of Witney. It is therefore not only symbolically important, but it could be indicative of the strength or weakness of Remain in affluent, rural English shire areas. Cheltenham (318/211) in the South West and Harrogate (258/308) in Yorkshire, are due to be decaled at 3.30am and 4am respectively. Both are affluent spa towns currently represented by the Tories at Westminster but previously held by Liberal Democrats. Remain will be hoping to carry such areas by a comfortable margin. Leave voters in these areas may be more motivated by economic and sovereignty considerations rather than by immigration.

Just after 4am, we should have enough results in from around the country to establish a credible turnout figure and to identify clear regional trends. By around 4.30, most of the strongest Leave areas on paper will have declared, so if Leave do not hold the lead or even if it is very close, it may bode badly for them.

Meanwhile, results should also be coming in from the largest voting areas in the country – Birmingham (213/290), Leeds (291/202), Glasgow (374/340), Sheffield (272/161), Manchester (345/221), Bradford (135/216) and Liverpool (317/89) should all declare between 4am and 5am. Collectively, these cities contain around 3 million voters and Remain will be hoping for a solid turnout to boost their overall total. As their rankings show, these areas are mostly favourable for Remain, but if we see the same seepage of Labour voters to Leave as we might in other parts of the country, some of the results could be closer than expected.

Just after 5am, with the vast majority of the results in, it might be possible to forecast the result for one side or the other unless the result is absolutely neck and neck (the BBC called the Scottish referendum at 5.14am, almost an hour before the result was officially confirmed). If the result is still in the balance, it will come down to the remaining 30 or so areas which should be declared between 5.30am and 7am. These include Fenland (1/2), one of the strongest areas of the country for Leave, as well as a few large cities including Bristol (330/200), Leicester (275/305) and Nottingham (278/239) which ought to benefit Remain.

The final outcome will then be formally declared in Manchester by Electoral Commission Chair Jenny Watson, the Chief Counting Officer for the referendum.