At 10.01pm on 7th May last year, the pollsters and many political observers were in a state of shock and confusion. The General Election exit poll had just overturned the common expectation of a hung parliament – and in some cases a hung parliament in which Labour were the largest party. Instead, all of a sudden the Conservatives were predicted to increase their number of seats, while Labour were to fall back. As it turned out in the end, of course, the Tories did even better, securing a majority, while Labour did even worse than the exit poll expected.

The reasons for that shock have been pored over at length. The pollsters got it wrong, but why? There has been extensive public debate about their panels not including certain demographics, and their models incorrectly weighting those they did reach, meaning that they missed the true national result, but that’s only part of the picture. The other factor was that the Conservatives had deployed a targeting strategy in marginal seats which made the concept of national swing far less relevant to the eventual result. I wrote about the inside story of that campaign a few weeks after the election, and the later row about whether the spending involved was legal of course is still underway.

In the EU referendum, the only thing that matters is the national totals for each side – therefore, there are no marginal seats as such. However, an effective targeting strategy still matters. With the polls (subject to the obvious caveats) showing a neck-and-neck race, the result may well come down to which side can first find and then successfully turn out its supporters. While the air war rumbles on, over the next week it is the ground war which really comes into its own.

So how are the two sides’ ground operations shaping up for the battle to Get Out The Vote?


The starting point

Vote LEaveThose hoping for a vote to leave the European Union began the campaign with a much larger operational mountain to climb than their opponents. Vote Leave, the official campaign, does not have the support of any of the main political parties – meaning it does not enjoy access to any staff, structure or previously collected data. As one Vote Leave staffer put it, while the No2AV campaign (on which many of its senior figures worked) “was able to piggyback on local Tory organisation”, they’ve “had to build this from scratch” – and constructing an effective national political campaigning organisation over the course of a few months is a big task.

The experience of the campaign team counts in their favour. Dominic Cummings and Matthew Elliott, the campaign director and chief executive, are most well-known, as are their their respective histories as a combative strategist and an experienced campaigner. Both are unusual in British politics in having fought and won referendums – the former against the North East Regional Assembly; the latter against Nick Clegg’s ill-fated Alternative Vote. Those were less high-profile battles, but having built up a successful referendum campaign before is valuable experience. By their side is Stephen Parkinson, former Special Adviser to the Home Secretary, who also worked on No2AV. Managing the ground operation are Nick Varley, a Conservative former PCC candidate and parliamentary staffer who has worked for Crosby Textor Fulbrook, and his deputy Tom Waterhouse, a former councillor who has also done time working for MPs and at CTF. At a regional level, the full-time co-ordinators are often Conservatives, but there are a number with UKIP backgrounds, too. The area, borough and constituency operations they oversee are managed by volunteers drawn from a range of partisan and non-partisan backgrounds.

The sometimes bitter division in the Eurosceptic side has certainly complicated the challenge for Vote Leave – UKIP threw its weight first behind Leave.EU and then behind Grassroots Out, meaning that the only large party on that side was denying Vote Leave the benefit of its data and campaign machine. As separately registered campaigning bodies, Vote Leave, Grassroots Out, Leave.EU and UKIP aren’t allowed by the Electoral Commission to directly co-ordinate their campaigning anyway, even if the personal and political differences were put aside, but in practice a lot of UKIP activists appear to have decided to join Vote Leave after it was designated the official campaign, which helped to avoid some duplication or conflict of effort.

The troops

As well as a good dose of UKIPers, the Leave campaign’s ground troops comprise a large number of Conservative activists, a smaller number of Labour Party members, a smattering of Greens and Lib Dems, and, interestingly, a large quantity of people entirely new to politics. Speaking to organisers on the ground in various parts of the country, each reports their surprise at the number of people who have never been motivated to get involved in politics before but see this fight as worth their time and effort. (As a local Vote Leave organiser myself, in Lambeth, we have had the same experience, and Emily Knight reported seeing it in Newham in her campaign diary.)

This of course brings challenges of its own – in an age of low political participation, any activist is a welcome addition, but those with no experience require training up. Nonetheless, it’s a nice problem to have – and certainly preferable to not having these volunteers at all. Those involved also report that these relatively green troops have proved open not just to leafleting but to taking on tasks key to the targeting operation, including canvassing.

It hasn’t all run smoothly. I gather some areas have experienced difficulties in encouraging UKIP volunteers to move away from the less targeted but higher-visibility tactics which are core to a smaller party’s campaigning, such as street stalls, and towards canvassing and data-gathering. Others report their surprise at finding some UKIP branches to be much smaller than they had expected. Both of these factors fit with the difficulty Farage and others experienced last year in translating energetic campaigning into victory in target marginals. The People’s Army are certainly enthusiastic and in many cases very experienced, and in some places they are proving very effective, but in terms of troop numbers or strategic approach there appear to be distinct shortfalls in some parts of the country.

Canvassing and data-gathering

In order to run a well-targeted Get Out The Vote operation, a campaign needs two things: troops and data. With the first in place, the Leave campaign still faced an uphill climb to identify its voters. After all, this is an issue which cuts across party lines, and they didn’t have the benefit of the data already gathered by the only anti-EU political party to use as a starting point (the FT reports that in March they only had the details of 43,000 known supporters).

They focused on three routes to gather enough information to make up the ground. The first was to leverage all their leafleting as a data-gathering route: ensuring their leaflets included freepost slips for recipients to declare their support.

The second was traditional canvassing, using consumer data fed into their custom-built VICS (Voter Identification and Contact System) database to try to guide local activists to the most promising streets (an essential function leading Tory Leave activists to knock on doors in core Labour areas they haven’t visited for years, though not always an accurate one) and then feeding that data back into the system to further sharpen that targeting process. The canvassing operation has been underway for some weeks, but for the last fortnight it has been the main focus of the Vote Leave grassroots operation, with organisers instructed to make it their top priority.

The third was online voter identification – something the Conservatives used to a degree in the run-up to 2015 by using Facebook adverts to drive voters to fill out online surveys on different issues. Here Vote Leave decided to be even more radical, borrowing an idea from the United States to launch, an online competition which offered a £50 million prize for correctly guessing the outcome of the Euro2016 football championship.

They won’t be drawn on exactly how many Leave voters they have identified through each of these routes, but I am told that those identified by the national campaign (ie through leaflet returns and online data-gathering) number “in the thousands” in every constituency. That is in addition to those found via traditional canvassing – and they intend to use VICS to crunch the numbers from that canvassing to pick out streets which are more likely to be Leavers, offering the potential to target GOTV operations at people whom the ground campaign has not yet spoken to. That, at least, is the plan.

Conservative activists will remember the problems experienced last year with Votesource, the Conservative Party’s database, developed in-house, which suffered various issues during the campaign including the loss of some voting intention data, before being struck by an outage during election day itself. VICS, too, has been developed in-house by Vote Leave, rather than bought off the shelf, but so far it seems to have impressed those using it – as Knight put it recently: “…it works and is incredibly reactive to results. I wish we had a system like this in the party, and I hope they will look at designing something similar.” Whether it can withstand the huge traffic involved in referendum day itself is yet to be seen.

The final week – Getting Out The Vote

Having built the machine, gathered the troops and identified apparently large numbers of voters to target, the final week is when all three must come together into a GOTV operation.

Given that their whole organisation is newly constructed, Vote Leave had a trial run of just such an exercise targeted at postal voters. As well as using a portion of their free mailshot to send a leaflet directly to registered postal voters, they used their ground operation to spend the Spring Bank Holiday weekend hand-delivering a named and addressed leaflet to postal voters as well, to ensure maximum Leave turnout. These didn’t go to every voter – instead, the campaign used VICS to identify 3,000 target postal voters in each constituency.

It’s fair to say the experience placed a lot of pressure on their voluntary arm – not least in the process of printing and sticking names and addresses on all those leaflets even before delivery could begin – and I know I’m not alone in having walked over 60 miles that weekend ensuring all of them got delivered. But for the most part the fledgling organisation seems to have withstood the test.

They’ve learned some lessons from that experience. For a start, there won’t be individual stickers to apply to the GOTV leaflets going out over the next week, allowing all resources to focus on contacting voters. Some local organisers I will be using the streets suggested by the VICS algorithm as targets but adding a dose of personal judgement, given the system still occasionally points them to streets which aren’t ideal Leave territory.

From Saturday to Wednesday, they will be repeating the leafleting exercise used on postal voters but on a much larger scale – and targeted specifically at those whom their data suggests are likely Leavers. Each constituency will be delivering up to 10,000 such leaflets. On the big day itself, the operation will move to dawn raids, followed by knocking up, on a model familiar to many ConservativeHome readers. Its effectiveness will rest on the accuracy (and stability) of VICS, the precise quantity of data collected in a few short months after starting with a blank sheet of paper, and the willingness of many relatively fresh ground troops to trust in the strategy and pursue tactics which are unfamiliar to them.


The starting point

stronger-inOn paper, Britain Stronger In Europe, the official pro-EU campaign, was starting from a stronger position than its opponents. It not only led in the polls, and enjoyed the support of the Prime Minister, but it enjoyed the formal backing of the Labour Party (despite Corbyn’s personal history of Euroscepticism), and the data that comes with it. While they still had a lot of work to do, and Labour remained a formally separate campaigning organisation, they certainly benefited from some local infrastructure to build their operation around, and were able to “piggyback” on Labour efforts in a way that Vote Leave could not.

It’s also worth noting that the Corbyn phenomenon offered a boost of its own – a variety of Labour activists and MPs who are far from enamoured with the new direction of their Party (Will Straw, Peter Mandelson, Chuka Umunna) saw the referendum as an opportunity to focus their efforts on something productive, and to distract from other problems looming on the horizon.

In terms of personnel, Straw is an able campaigner, though even he concedes he hasn’t run anything on this scale before. Interestingly, just like Vote Leave the pro-EU campaign have looked to the Conservative Party for their Head of Field Operations – between 2010 and 2015, Stuart Hand ran CCHQ’s training programme for Conservative Campaign Managers and he oversaw campaigning for battleground seats in the 2010 General Election. His deputy, Claire Hazelgrove, is a former Labour candidate who has worked on grassroots campaign operations in the charity sector as well as managing volunteers on GOTV operations in the 2012 Obama campaign. They’re bolstered by the advice of Stephen Gilbert, the former Deputy Chairman of the Conservative Party who played a key role in designing the 40/40 targeting strategy in the run-up to 2015.

The troops

The precise number of activists on each side is still unknown, and various guesstimates are hotly disputed. The broad breakdown of those active on the In side is a large number of Labour activists, a leavening of Lib Dems and some Conservatives, all joined by a surge of people new to politics similar to that experienced by Vote Leave. Students, given the broadly pro-EU position of many campuses and the formal backing of the NUS, also make up a decent share of those involved.

ConservativesIn, the Tory Remain group run by Charlotte Vere and Nick Herbert, is active – particularly through its touring battlebus – but many of its activists seem to be campaigning as part of Stronger In as well, using their distinct Party brand to focus on winning over their fellow Tories at more targeted events.

The same integration can’t quite be seen among Labour Remainers. A notable gap has opened in the last couple of weeks between the work of the official campaign and the work of LabourIn, the Opposition’s campaign vehicle. Apparently driven by a number of factors – stinging criticism of their leader’s lacklustre message, concern about the official campaign’s slide in the polls and most of all a morbid fear of a repeat of the Scottish disaster, in which they totally lost touch with their own core vote – Labour has increasingly separated out its campaign in terms of message, branding and activity. Party members have been sent prominently branded “LabourIn” posters to display, branches have been mobilised en masse and MPs are leading the charge in a way that their leader has notably failed to do.

That adds to the Remain side’s weight on the ground, particularly in London, where Labour members are disproportionately located, but it also risks disrupting their strategy (one pro-EU Tory activist reports a near absence of Labour activists in his local Stronger In operation as a result) and their message (some pro-EU Labour activists express concern that Labour’s arguments on the EU still spare time to attack Cameron, even as they are meant to be on the same side of the referendum).

There are a few reports of mis-steps within the grassroots operation. On initially joining Stronger In, a volunteer simply gives their email address and, optionally, their postcode. By contrast, the Vote Leave sign-up form includes a variety of tick-boxes for different types of activity the volunteer would be willing to take part in – leafleting, canvassing etc. It’s a small difference, but it created extra work for the pro-EU campaign in the early months. One volunteer recounts agreeing to take part in phone canvassing, only to find that instead of contacting voters their job was to phone signed up Stronger In supporters to find out what if any activity they might be interested in helping with. That is time spent talking to itself that a campaign cannot get back.

Canvassing and data-gathering

The pro-EU side definitely benefited from Labour’s support in the race for voter identification in the capital, at least. Anyone canvassing for Sadiq Khan was encouraged to add in a voting intention question on the EU referendum, a highly effective bit of piggybacking. Stronger In, too, have sought supporter data via leaflets and online gathering – though the latter has not been as eye-catching as the 50million campaign, they have been pushing targeted adverts on Facebook, aided by Craig Elder and Tom Edmonds, who performed the same service for the Conservatives in the 2015 election.

They’ve certainly been very active on the ground. An analysis of both campaign’s events pages by political scientists Matthew Goodwin and Caitlin Milazzo claims that Stronger In has held more campaign sessions than Vote Leave. The anti-EU campaign contest that they have held many more events than appeared on their website (or as Cummings put it, characteristically, “you are proper charlatans”), and thus that their number is understated, a claim which is by its nature hard to prove.

Whatever the truth of the totals, activity does not always equal data gathering – there are some signs that Stronger In’s campaigning has focused more on high-visibility leafleting (outside stations and on high streets) than on canvassing. Of the impressive 141 events they held yesterday around the country, for example, only 18 involved knocking on doors – the bulk of the rest involved leafleting in high-footfall public locations. Even in London, where they have a lot of troops, some of those on the ground express concern that there has been an insufficient focus on canvassing. While Goodwin and Milazzo counted the number of events, they did not count the type of activity involved – potentially missing a disparity in data-gathering.

Enjoying access to Labour’s historic relationship with its own voters is certainly useful, though perhaps not as much as some might have assumed a few months ago. In some areas, Labour MPs are growing increasingly worried that their own core vote is turning heavily to Leave. Stephen Bush reports one frontbencher saying: “The polls say it’ll be fine but every doorstep someone tells you to f*** off.” The Guardian‘s John Harris filmed a canvassing session in Stoke with Ruth Smeeth, a local MP who has already ditched both the Stronger In and LabourIn messages for a locally crafted leaflet about the pottery industry, which is painfully brutal to behold. At door after door, she finds her own voters declaring for Leave. I’m told LabourIn phone bankers calling known Labour voters outside London are met with a similarly negative response.

The problem isn’t just that Remain may be losing among Labour’s core, but that the functionally independent LabourIn campaign may continue to allocate its resources to trying to win back these lost voters rather than switching to target possible waverers in other areas. They certainly feel the need to keep on taking the abuse in order to try to avoid a Scotland-style crisis for their party in England, but that means the pro-EU campaign effectively loses the benefit of resources that might be better used elsewhere.

The final week – Getting Out The Vote

While the two campaigns have differed from the start in various ways, Stronger In’s intentions for the final week are very close to those of Vote Leave – a reminder that for all the bells and whistles which social media and big data can bring to politics, some things never change. They, too, will be pursuing target voters by leaflet, by dawn raids and by traditional knocking-up, as well as knocking-up in some areas the night before polling day. They may benefit from the extensive experience of the many LabourIn activists in exactly what’s involved in a classic polling day operation – perhaps offering them an edge.

The big question for Stronger In will come down to the quality of their targeting – have they managed to gather the data required to identify their key voters or did they spend too much time doing visible (rather than doorstep) campaigning? Do they now know exactly where their voters are, or did they assume for too long that Labour’s core vote would be on their side? And can they persuade the Labour Party to put its own interests aside and divert its resources away from trying to shore up its electoral future?