Here is an exclusive extract in which the Croydon Central MP recounts the events of Polling Day, May 7 (and May 8), in which he just managed to retain the seat
Twelve months before, Labour had wiped the floor with us on polling day. On Twitter, their London regional director was boasting about how they planned to do the same again. They had set themselves the target of getting 500 activists from across south London onto the streets of Croydon Central.
After asking everyone we could think of – and, let’s be honest, begging – we were confident we would have over 300 people knocking up. This was far more than we had ever had before, but still 200 fewer than Labour were hoping to have. We needed a plan to bridge the gap by making better use of the people we had.
In the end, all campaigns come down to it. The final fifteen hours before the battle’s lost and won. And when the polls close at 10 p.m., you’re not finished. Within minutes, the count gets underway. In a political campaign, it’s the longest day.
The productivity challenge
To explain the plan we came up with, it’s helpful to understand what we normally do on polling day. By this stage of the campaign, we’ve given up trying to talk round opposition supporters or speak to people we haven’t been able to canvass yet. The focus now is on what the Americans call GOTV (get out the vote): ensuring the people who’ve told us they are likely to vote for us actually do so. In an ideal world, this involves four things.
First, we put ‘tellers’ outside every polling station. Their job is to ask everyone who comes to vote for their electoral roll number (which is on the polling card that most people bring with them) or their name and address if they don’t have their roll number.
Second, the information our tellers gather is relayed back to our campaign centre in that area, which is normally the home of one of our supporters. In a local election, we would have a campaign centre for each target ward; in a general election, we might have several campaign centres in each constituency. Once the data on who has voted has got to the campaign centre, it is cross-referenced against the list of people we are trying to get out to vote. If any of these people have voted, we cross them off the list.
Third, the person in charge of the campaign centre then sends out volunteers to call on our potential supporters who have not yet voted to encourage them to do so. This is known as ‘knocking up’. Those people who tell us they’ve voted, they’re going to vote later or they’re not going to vote can be crossed off our list. If no one is in, we leave a leaflet reminding them it’s polling day. And we keep knocking up until everyone is crossed off our list.
Normally, we only knock up people who have told us they are going to vote Conservative. However, if we get to early May and we haven’t found enough definite Conservatives to be confident of winning, we might add people who have told us they are probably going to vote Conservative or people that CCHQ’s voter model predicts are likely to be Conservative supporters.
The final thing we need is a small team of drivers to take potential supporters who are unwell or relatively immobile to their polling station.
This ideal operation requires a huge amount of manpower. If we’re short of people, we do what’s known as a ‘blind knock-up’ where we forget about the telling and use everyone we have to knock up. We’ll waste a bit of time calling on people who have already voted, but we’ll still cover more ground than if we used most of our manpower telling and therefore had only a handful of people knocking up.
The key point here is that while telling provides useful intelligence that can reduce the amount of knocking up we have to do, it does not on its own increase the Conservative vote. It is knocking on people’s doors, reminding them that it’s polling day, persuading those who still haven’t made up their minds to go our way and arranging lifts for those who need them that make a difference.
Even with a blind knock-up, though, precious time is often wasted. The knocking-up teams are usually given an hour or two’s work at a time and therefore have to keep travelling back and forth to the campaign centre to get more work. There is absolutely no reason why this should happen – the list of people to knock up isn’t being updated with information from tellers. It is simply force of habit.
Furthermore, anyone who has ever helped on polling day for a general election will tell you that most of the people you knock up were going to vote for you anyway. You have relatively few conversations where you feel that you changed someone’s mind. When it’s a council election, many people won’t know it’s polling day so the simple act of reminding them makes a big difference, but the wall-to-wall media coverage and avalanche of literature you get in a marginal constituency makes it pretty hard not to know when there’s a general election on.
To try to make maximum use of the people we had, we therefore made three changes to our normal polling day plan.
The first change was who we knocked up. Because we thought the result was going to be very close, we wanted to knock up key floating voters – the undecideds who preferred David Cameron to Ed Miliband as Prime Minister or who wanted me as their MP, Liberal Democrats who preferred us to Labour, Conservatives tempted by UKIP, and even solid UKIP supporters. If we knocked up all of these on top of our definite and likely Conservative voters, though, we’d have to call on over 25,000 people. We simply didn’t have the manpower for that.
We therefore decided not to knock up people who we had canvassed as strong Conservatives since May 2014 (i.e. since we had been specifically testing whether people were tempted by UKIP) and who had voted in that year’s elections. The theory was that if we were sure they were Conservatives and we knew they had voted even in a low-turnout election, we didn’t need to knock on their door – they would vote for us anyway. Put simply, we wanted our knocking-up teams to focus on speaking to people who might need some persuading.
The second change was not to use anyone as a teller who was willing to knock up. What mattered was talking to as many potential supporters as possible, not standing outside polling stations modelling blue rosettes. There were some people who either wouldn’t or physically couldn’t knock up – we would use them to tell in our best areas. In the rest of the constituency, however, we would do a blind knock-up.
The final change was to do away with people going back and forth to a campaign centre. We would divide the constituency into twenty-seven sectors and appoint an experienced activist as team leader in each sector. In the more Conservative parts of the constituency, a sector might equate to one polling district. In the less Conservative areas it might be two polling districts. Each team leader would hold the master list of our potential supporters in their sector and lead a team of volunteers knocking up those people. If we were telling in their polling districts, they would periodically collect data from the tellers and cross off anyone on their list who had voted.
With one exception, everyone would spend the whole day – apart from lunch, which we would organise at a central venue – out on the streets. The exception was Tim Pollard. He would liaise with the team leaders during the day to monitor what proportion of their list they had crossed off and reallocate resources between them accordingly. Those people who couldn’t join us for the whole day would be given Tim’s number to call when they were on their way. Rather than coming to a campaign centre and then being sent on somewhere to knock up (another waste of time), they would be directed straight to one of the team leaders.
I wouldn’t be assigned to any of the teams. Instead I would move from sector to sector accompanied by some close friends and family members, whose job was to try to keep me relaxed as well as helping with the knocking up.
This plan was the brainchild of Jason Cummings. I loved the idea of doing away with trooping back and forth to a campaign centre and only using people who would not knock up for telling, but I took some convincing not to call on Conservatives who had voted in the low-turnout elections twelve months before. In the end, Jason talked me round. The only change to the plan I made was to introduce a team of Indian activists knocking up Indian target voters and a team of Tamil activists knocking up Tamil target voters.
We had our plan: it was time for the big day.
Not the best start…
Some campaigns start polling day at 4 a.m. with a dawn raid. The idea is to deliver a ‘Good morning, it’s polling day’ leaflet through the door of all your potential supporters before anyone leaves home to encourage people to vote on their way to work.
While I’m as keen on flogging myself to death in an attempt to win elections as the next candidate – actually, more so – experience has taught me that dawn raids are counter-productive. Every time I’ve seen them tried, the result is all your key activists are dead on their feet by early evening. Right at that point at which your effort should be reaching a peak, people start giving up, exhausted. Far better, I felt, to let everyone have a lie-in and a nice breakfast and start knocking on doors at 10 a.m., so that they could go through to close of poll at 10 p.m. We therefore started the day with a breakfast briefing for our team leaders at the Harvester in Addington Village.
As I was parking my car, my mind was on other things and I managed to scrape the car next to me. I was furious with myself. Now, when I should be briefing our team leaders I was going to have to wander round the pub to find whose car I had damaged. Fortunately – at least from my point of view – it turned out to belong to Helen Pollard, Tim’s wife. She told me not to worry about it and we sorted it all out once the election was out of the way.
I was joined for the day by my brother Richard; two of my best friends, Barry and Mark (the rest are Labour voters and therefore not all that keen on canvassing for a Tory); and Scott Colvin, who used to work with me at Conservative Central Office.
We started on Woodmere Avenue and the roads off it in Ashburton ward, a strongly Conservative area with some UKIP leanings. The response was very encouraging.
Everyone we spoke to, whether they had been canvassed as Conservative supporters or floating voters, had voted for me or was going to vote for me apart from one person who had been canvassed as undecided but wanting me to continue as the MP. She said she had never voted Conservative in her life and couldn’t bring herself to do so this time.
At the end of the session, I checked my emails and found this from another elector we had canvassed as undecided but wanting me to continue as her MP:
I cast my vote promptly this morning about 8.45. I voted for you – I broke the pattern of a lifetime! I’m not a Tory, but I wanted you to remain my MP. You have done a great job over the last five years. I don’t know what will happen tonight when the votes are counted, but I just want to say thank you for all your hard work and for making a difference for Croydon.
You win some, you lose some…
Then it was off to Oaks Farm for a quick lunch and a debrief with the team leaders. In our strong areas, the response was very good – virtually everyone on our list was saying they were going to vote for me. In Addiscombe, Woodside and parts of the town centre, we were finding relatively few people at home. In Fieldway and New Addington, there was a huge amount of Labour activity, but it seemed surprisingly untargeted – cars driving round with loudspeakers reminding our supporters as well as theirs that it was polling day. In light of this, we decided to move resources out of our best areas, initially to those two wards and then later in the afternoon to Addiscombe and Woodside.
After lunch my team headed to Featherbed Lane and the roads off it in Addington. This was quite a good area for us, though not as strong as the Woodmere Avenue area because it was even more UKIP leaning. Again, the response was very encouraging. Then we went up to New Addington to join the push there. Here, the response was not as good, though still not bad. The Conservative/UKIP waverers were dividing 50/50 between us and UKIP and our canvassing wasn’t as accurate as elsewhere – some people who had been recorded as potential Conservatives turned out to be lifelong Labour voters.
We bumped into a team of Labour activists for the first time. Once again, the downside of them having so much outside help was clear – they looked like they didn’t know where they were and seemed to wander around fairly aimlessly.
By now it was late afternoon. We had large numbers of people working in New Addington, so my team moved to Regency Walk in Shirley, another area that was pretty good for us but with UKIP leanings. Richard and Scott had to leave; they were replaced by my wife Karen and Barry’s wife Lucia. The response here was very encouraging, more akin to the Woodmere Avenue and Featherbed Lane areas than New Addington.
At about 6 p.m., we sent out the final email bulletin of the campaign, a short video of Boris, which we had recorded on Tuesday night. The message was simple: ‘Folks, this is it. We are now in the last few hours of the election campaign. You can make a difference in Croydon Central and get your superb MP Gavin Barwell re-elected. Get out to vote now, folks, you know it makes sense.’
Then we were off to the Bingham Road area of Ashburton ward, an area I had long felt was a bellwether for the whole constituency. The floating voters here were dividing pretty evenly between us and Labour, but what was encouraging was the large number of Conservative activists we saw in neighbouring roads. In this crucial area, at this crucial time, we were outnumbering Labour on the ground.
By now it was about 8.30 p.m. We hit one last area – Grant Road and the surrounding roads in Addiscombe ward – then called stumps just before 9.30. The polls were still open for another half an hour, but most people were no longer opening their doors and those who did were starting to get annoyed at being disturbed at that time of night. We weren’t going to get anyone else out to vote.
And that was it.
For the last year, this campaign had dominated all my waking hours and disturbed my sleeping ones. Now it was over. In a few hours’ time, I would find out whether it had all been worth it.
The response on the doorstep had felt good. The people we’d canvassed as Conservatives had nearly all voted for us. So had the people we’d canvassed as Conservative/UKIP waverers. And we’d done well among the probable Conservatives and the undecideds who preferred Cameron to Miliband or wanted to keep me as their MP, particularly in the more Conservative parts of the constituency. I was confident that I would get more votes than I had in 2010. The question was: would that be enough?
The problem was we had no idea how well the Labour vote had turned out. When you’re telling, you can keep a running total of how many of your potential supporters have voted versus how many other people, giving you a pretty good idea of the result, at least in areas where it is a two-horse race. When you’re doing a blind knock-up, however – as we were in all but the most Conservative areas – you get very little feel for how your opponents’ votes are turning out. If the Liberal Democrat and Green votes completely collapsed, I could do quite a bit better than in 2010 and still lose.
I’d have a better idea when the exit poll was published just after 10 p.m.
A night at home with my mates
With the polls closed, the count could begin.
I enjoy polling day, but I hate counts. I always have done. When each ballot box arrives, the clerks count the number of ballot papers in the box, checking it matches the number of ballot papers issued at the relevant polling station to ensure that there’s been no fraud. At this point, they’re not counting the number of votes per candidate, just the total number of ballot papers. While this is going on, however, you can see a good sample of the ballot papers. Unless it’s very close, you can tell who’s won. The clerks then spend hours counting the number of votes per candidate. You stand around with very little to do, surrounded by people who are exhausted and either understandably upset, a bit gloaty or very stressed.
Take it from me, counts are truly awful occasions.
I had therefore formulated a cunning plan. I would stay at home, order a curry and watch the results with my closest friends – Barry, who had been with me all day, and Atul, Paul and Pete, who were Labour supporters and so obviously hadn’t, but wanted to be with me when the result came through. My campaign manager Ian would give me a call when he had anything to report (I suspected the first call was going to be, ‘It’s too close to call…’). I would only turn up at the count five minutes before the returning officer declared the result.
My two eldest children – Jack and Sam, who were twelve and nine – were desperate to come to the declaration. I wanted them there too if I won. After all, this might be their only chance to see me win – if the seat continued to move against us demographically the 2020 election would be an uphill struggle whatever the national picture. I didn’t want them to come if I lost, however – I thought they’d find it too upsetting. So Karen packed them off to bed, promising that we would wake them up if I won. We settled down in front of the TV and awaited the exit poll.
When it was announced, it suggested we were going to do better than I expected. The BBC were predicting that we would win 316 seats, 9 more than we had won in 2010 and just short of an overall majority. Labour were predicted to win 239, 19 fewer than they had won in 2010; the Liberal Democrats just 10, 47 fewer than in 2010. The poll also suggested that the SNP would win a remarkable fifty-eight of the fifty-nine seats in Scotland. My boss, Michael Gove, was visibly shocked. Paddy Ashdown said he would eat his hat if it proved accurate.
The actual results, once they started coming in, were even better for the Conservatives than the exit poll suggested. The moment I started to believe the exit poll was when Marcus Jones increased his majority in Nuneaton, a key West Midlands Conservative/ Labour marginal that we had only won by 2,000 votes in 2010. As well as doing well against Labour, we also seemed to be winning every Liberal Democrat seat in the south-west. Amazingly, it began to look like we would win an overall majority. If we were doing this well nationally, surely I would be OK?
My phone rang. It was Jason, who was at the count. They had done a tally when the ballot paper numbers were checked and they thought I had won by about 1,000. However, it was too close to be certain. That brought me down to earth with a bump. Then I saw on the news that Labour had gained Ealing Central & Acton, an outer London marginal very similar to mine. Whatever was happening nationally, Croydon Central was clearly going to be a nail-biter.
The rest of the night was a bit of a haze. I’m pretty sure the Ilford North declaration was before mine. This was another Labour gain in outer London – and in a seat that hadn’t been considered a marginal. It was clear that the one part of the country where Labour were doing well against us was outer London (they would also gain Brentford & Isleworth and Enfield North, another seat that was very similar to mine, but I don’t remember hearing these results on the night).
At about 4 a.m., Jason called again to say the count was nearly finished. It looked incredibly tight, but they thought I was just ahead. Then another call: it wasn’t finished at all. The returning officer’s provisional declaration was that I had won by 167 votes, but Labour had (understandably) asked for a recount. Another call about an hour later: it still looked very close. I suggested to Jason that he might not want to call again until he had some actual news to report (I think the exact language I used was rather more colourful – sorry, Jason).
The next ninety minutes were not ones I would care to repeat. At the risk of sounding selfish, if there was a swing to Labour and I was one of forty or fifty Conservative MPs to lose their seats, I could take that. To lose my seat on the same day the party was winning an overall majority, however, would be incredibly tough.
I phoned the returning officer, Nathan Elvery, and explained that I was only five minutes away and that I wanted to bring my eldest two children if I had won, but not if I had lost. I asked if he could give me an indication before I set off for the Fairfield Halls. He agreed to do so, something I will always be grateful for.
The fat lady sings
Time went by so slowly.
Shortly before 6 a.m., I got another call. This time it was Ian. He had the final result. I had won by 165 votes, one of the slimmest margins in the entire country. To put the margin in context, if just two people in each polling district had switched from Conservative to Labour, I would have lost.
The best moment of that night was waking Jack and Sam. As soon as they opened their eyes, they knew what it meant. I then gave my mum a call to tell her the result and to meet us at the Fairfield Halls and off we went.
When we got there, I went straight onto the stage for the declaration:
Barwell, Gavin (Conservative) 22,753 votes ELECTED
Jones, Sarah (Labour) 22,588 votes
Staveley, Peter (UKIP) 4,810 votes
Sutton, Esther (Green) 1,454 votes
Fearnley, James (Liberal Democrat) 1,152 votes
Ashley, April (Trade Union & Socialist Coalition) 127 votes
Camden, Martin (UK Progressive Democracy Party) 57 votes
Despite UKIP getting 9 per cent of the vote and despite all the difficult choices we’d had to take in government, I had got more than 3,000 more votes than in 2010. Across the country as a whole, we had increased our vote share by 0.7 percentage points. In London, it had gone up by just 0.3 points. My own share, however, had gone up by more than 3.5 points. I had got more votes than any Conservative candidate in Croydon Central since 1979.
Even that, though, had only been enough by a whisker. The Liberal Democrat vote had collapsed, which was no reflection on their candidate, James Fearnley, who is a real talent. Partially as a result of this, Sarah Jones had got nearly 6,000 more votes than her predecessor, Gerry Ryan, in 2010 and had increased the Labour share of the vote by more than 9 percentage points, substantially better than Labour had managed either in London overall (seven points) or across the country as a whole (1.5 percentage points). She had got more votes than any Labour candidate in Croydon Central since Geraint Davies (now the MP for Swansea West) in 1997.
The truth was it was a very good result for both main parties.
Sarah looked devastated, as I would have in her position. I could probably understand better than anyone what she must be feeling. To work hard and lose is tough, but to lose by such a small amount is even tougher because you inevitably think, ‘If only I had done…’ I’m sure it was small comfort, if any, but in my speech I paid tribute to her and her campaign team. I don’t share their values, but I admire anyone who gives up their time to fight for what they believe in.
I thanked those who had voted for me and particularly those who had given up their time to help me win. Conscious that I had only won by a handful of votes, however, I also spoke to those who had voted for Sarah. I reassured them that I would work with Steve Reed, Chris Philp and our Labour council to do what was best for the town and try to ensure that everyone benefited from the economic recovery.
To be honest, it wasn’t my best speech. I’d spent much more time working on the alternative version since that would have been the last thing I would say in public life. Still, it hopefully got across my sincere gratitude and desire to listen to and represent all my constituents, whoever they had voted for.
After the declaration there were hugs for some of the people who had made this victory possible and a round of media interviews – the national broadcasters, the Evening Standard and the local papers. Someone got some nice pictures of Karen, Jack, Sam and me, which made it onto London news bulletins. Their friends were very impressed they were on TV.
At about 7 a.m., it was time to go home. Karen was exhausted and went straight to bed. I couldn’t sleep, however. I wasn’t euphoric, just… awake. Which was lucky, because there was one final debate to be had. Jack had missed his bus to school and was strongly of the view that this meant he should have the day off. I drove him to school. Michael Gove would have been proud of me.
- How to Win a Marginal Seat – My Year Fighting for My Political Life by Gavin Barwell (Biteback, £12.99), will be published on March 17
- MP Barwell’s links to figure at centre of Tory expenses scandal
- Still Croydon’s only independent news source, and based in the heart of the borough: 1.97 million page views 2013-2015
- Inside Croydon: Named among best regional media campaigns, 2014
- If you have a news story about life in or around Croydon, a residents’ or business association or local event to publicise, please email us with full details at firstname.lastname@example.org