Election 2015 – Poll of poll of polls – 30 April 2015

Headlines – As you might expect, the predictions are finally falling into line. There are only 8 seats separating predictions for the Conservatives and 7 seats for Labour.

Even if you take the best Conservative prediction with the worst Labour (or vice versa), it fails to produce anything close to a majority. It barely produces workable coalition mathematics (as shown by the FT’s running poll of polls based on Election Forecast).

With a week to go until polling day, where does this leave us? There are three likely scenarios and two outliers:

1. The predictions are broadly correct

If the predictions set out below are correct, we won’t be much wiser on 8 May as we were on 7 May. Days, if not weeks, of wrangling over coalition forming will ensue. Politicians will be navigating constitutionally murky waters. Ultimately, a left of centre coalition or supply and confidence deal is the only one that the electoral maths will permit.

2. The predictions mask the ‘shy Tory’

There are two ways in which the Conservatives could emerge with either a slim majority or be large enough to be the dominant party in coalition. One is the psephological phenomena of the shy Tory.

This was most noticeable in 1992, when polls consistently indicated a Labour victory and, instead, the public delivered a Conservative majority. One of the reasons was a reluctance to tell pollsters that  you were planning to vote Conservative. Once inside the privacy of the polling booth, head ruled heart and millions put their cross next to the Tory candidate. It would only take the Conservatives getting about 2% more of the vote (and Labour correspondingly 2% less of the vote) for them to get close to, if not over, the 323 figure.

3. UKIP voters come back to the Conservatives

A third possibility is that a similar percentage of current UKIP supporters return to vote Conservative. Whether scared by the prospect of an SNP kingmaker or convinced that the Conservatives will deliver a referendum on the EU, it would only take a small shift to save a handful of essential Conservative marginal seats.

4. Miliband the Leader

As an outlier, it is also possible that Miliband pulls off an astonishing political feat by convincing a majority of voters that he is ready and right for power. A late surge in Labour votes, combined with an inbuilt bias in favour of the Labour Party, would overcome even catastrophic losses in Scotland.

If Labour emerge as the largest party, they’ll be in a much stronger negotiating position for coalition building. There is a huge difference between forming the only workable coalition and picking a coalition partner (e.g. between the SNP and the Liberal Democrats). If Labour gets 300 MPs, its coalition building options are massively increased. If it only gets c. 270 seats, it is pretty much dependent on the SNP.

5. Grand coalition

But who said a coalition has to include the small parties? Is a grand coalition between Labour and the Conservatives possible? Arguably, they have more in common than Labour and the SNP. The parties could agree to differ on a raft of less essential policies and just have a core government agenda to try and ensure stability and growth.

Could it be done? Grand coalitions have been formed between the CDU/CSU and SPD in Germany twice without any discernible crisis. They have also been a feature of Austrian and, more recently, Italian politics. Even the UK has had peacetime governments of national unity (Ramsay MacDonald and Stanley Baldwin’s National Governments of the 1930s).

About the poll of poll of polls

There are a number of great websites that are collating data on voting intentions for 7 May’s general election. This series intends to track these poll of polls to try and pick up shifts, subtle changes and patterns. It is a poll of poll of polls.

Note: Unlike the people behind the websites listed below, I am not an expert in statistics, odds and outcomes. I’m just very interested in psephology and elections. Please let me know if I’ve got something wrong and I’d be happy to correct where necessary.

Data taken at 13:00 on 30 April 2015 from:

• Election Forecast
• UK Polling Report
• Bet 2015

Any mistakes in interpretation of the data are mine. I’d be happy to discuss and, if necessary correct.