The AV Referendum: What went wrong?

By Professor John Curtice, University of Strathclyde

There are three possible explanations for why the Yes campaign for the AV referendum was unsuccessful:

• The ‘Yes’ campaign was “outgunned” by the ‘No’ campaign who were able to reach more voters;
• Partisanship – in a referendum people do not just vote on an issue, but on what the political parties have to say on the matter, and this worked to the Yes campaign’s disadvantage; and finally
• The messages used by both sides during the campaign, with the ‘Yes’ campaign’s messages proving ineffective, and the ‘No’ campaign’s messages resonating more strongly with voters.

With regards to the first argument there is contradictory evidence on how effective both campaigns were in reaching voters. Many voters reported some contact with the ‘Yes’ campaign and the ‘No’ campaign and most polls show them about level. It may be the case that the ‘No’ camp did reach more people during the campaign – or did so more effectively – but of this we cannot be sure.

Partisanship is potentially a key influence in a referendum, especially one where people do not fully understand the issue itself. By April 2011 support for the Liberal Democrats was far lower than it had been at the time of the May 2010 election. This factor was damaging for the ‘Yes’ campaign because the Liberal Democrats were the only major political party to be united in support of a ‘Yes’ vote.  Meanwhile, Labour’s standings had improved – so gaining the support of the Labour party had become even more important.

During the course of the campaign, the Conservative party successfully got their supporters on side but polls show that in contrast, the Liberal Democrats failed to do so, while, despite the position adopted by the Labor leader, Ed Miliband, Labour supporters shifted from being marginally in favour to being clearly against.

That leaves us with messaging. The initial research conducted for the Yes side in November 2010 showed that the public were not particularly concerned about how hard their MP worked – so the message that AV would “make MPs work harder” could not necessarily be relied upon to resonate strongly. The research also showed that those who were concerned about how hard their MP worked were actually no more likely to be in favour of AV than those who did not.

The surveys also found that not trusting politicians, or not feeling that elections make a difference, was not linked in people’s minds to AV or the impact it might have. Similarly, concerns regarding MPs expenses – concerns that in any event seemed to have dissipated compared with 2009 – did not appear to translate into support for AV.  These results should have acted as a warning that the Yes campaign’s messages would not be successful in persuading people to vote ‘yes’.

To make matters worse, according to research at Exeter University, the ‘No’ campaign’s messages against AV became more popular during the course of the campaign, whilst the ‘Yes’ campaign’s arguments did not. For example, the research showed that the ‘No’ campaign’s message that AV would be an expensive waste of money proved particularly popular and correlated strongly to how people voted on the day.

Conversely the Yes argument that AV would mean no more wasted votes became less popular and the argument that AV would restore trust in politicians also failed to resonate.

At the same time, research from the British Election Study, showed that, although far more people agreed than disagreed that AV would make MPs work harder, this did not translate into ‘Yes’ votes on 5 May. In this study, the strongest correlation between agreeing with a message and turning that agreement into a vote was the ‘No’ campaign’s message that First Past the Post is part of British tradition.

Findings from focus groups conducted for the Yes campaign after the referendum indicated that the ‘No’ campaign’s messages were far more effective – a view which was also put forward by ‘Yes’ supporters taking part in the groups.

The ‘Yes’ campaign was always facing an uphill battle in a political context over which it had no control and put simply, the Conservatives were highly effective, but Liberal Democrats and Labour were not. This was a reality that was out of the hands of the ‘Yes’ campaign.

The criticisms of AV put forward by the ‘No’ campaign were far more popular and proved more effective in shaping how people eventually voted on 5 May. The ‘Yes’ campaign’s key arguments were either lost or did not resonate with people in terms of why they should vote ‘yes’.

With the benefit of hindsight, research should have been carried out at the very beginning of the campaign to pre-empt the arguments the ‘No’ campaign would use. This would have left the ‘Yes’ campaign better prepared for the criticisms of AV that the ‘No’ campaign utilised so successfully.

Finally, although the research has shown that the ‘Yes’ campaign’s own messages were unsuccessful, it is unclear whether there were any messages in favour of AV that would have fared better. The fact was that AV was a difficult product to sell and none of the perhaps stronger arguments that could have been made about a proportional system could be employed to make a more powerful case for change.