Making sense of the polls

Making sense of the polls

United Kingdom Election Preview

The United Kingdom election will be held on the 7 May. There are a total of 650 House of Commons seats, with members for each seat elected by the simple first past the post (FPTP) system, in which the candidate with more votes than any other candidate wins the seat. Of the 650 seats, 533 are in England (82% of total seats), 59 in Scotland (9%), 40 in Wales (6%) and 18 in Northern Ireland (3%).

Since the 1920’s, UK elections have been contested by two major parties: Conservative and Labour. The Liberal Democrats (Lib Dems) have been a third force, but their share of seats has been far lower than their vote share under the FPTP system. From 1931 until 2005 inclusive, either Labour or Conservative has won an outright majority of the Commons seats, with one exception in February 1974.

At the 2010 election, the Conservatives won 306 of the 650 seats, to Labour’s 258 and 57 for the Lib Dems. The Lib Dems formed a coalition with the Conservatives, and this coalition has governed Britain since the 2010 election. The 2010 vote shares were 36.4% for the Tories, 29.0% for Labour and 23.0% for the Lib Dems, but the Conservatives won 47.1% of the seats, Labour 39.7% and the Lib Dems only 8.8%.

Had the 2010 election been contested on just the 533 English seats, the Conservatives would have won an overall majority. England gave the Tories 297 seats to 191 for Labour and 43 Lib Dems. However, the Tories won only nine of the 117 non-English seats, winning eight out of 40 in Wales, one out of 59 in Scotland and none of the 18 Northern Ireland seats, which the major UK parties do not contest.

Since the 2010 election, Lib Dem support has collapsed as many of their former supporters came from the left, and were angry with the Lib Dems for joining a Tory government. To some extent, the collapse in Lib Dem support has been countered by the rise in the anti-immigration United Kingdom Independence Party (UKIP), so the total major party vote has only increased slightly on 2010.

At the 2010 election, Labour took 41 of Scotland’s 59 seats, to 11 Lib Dems and six for the Scottish National Party (SNP). Labour’s Scottish vote was 42.0%, with the SNP closest on 19.9% and the Lib Dems on 18.9%. However, recent polls have the SNP poised for a crushing win in Scotland; the latest YouGov Scotland poll has the SNP leading Labour by a huge 49-25, with 18% for the Tories and just 4% for the Lib Dems. If this poll were replicated at the election, the SNP would probably win at least 50 of the 59 Scottish seats.

The result of the looming Labour wipeout in Scotland is that Labour’s chances of winning an overall majority are dim. The current UK Polling Report poll average has Labour just ahead with 34%, the Tories on 33%, UKIP on 15%, the Lib Dems on 9% and the Greens on 5%. Based on uniform swing projections, UK Polling Report is giving Labour 308 seats, to 250 for the Tories, 53 for the SNP and a few Others, 21 for the Lib Dems and 18 for Northern Ireland parties. Under this scenario, Labour would form government with the support of either the SNP or the Lib Dems.

It seems strange that Labour could have a 58 seat lead over the Tories from only a 1% popular vote lead. However, Labour’s vote crash in Scotland means they are getting bigger swings elsewhere to stay level on overall voting intentions; most polls are showing about a 5% average net swing from Conservative to Labour in England and Wales.

Another reason for Labour’s strong seat performance is electoral bias. The electoral boundaries to be used at this election are based on electoral rolls as at 2000 - 15 years ago. As the Labour north of England is growing more slowly than the Conservative south, the lack of up to date boundaries helps Labour.

However, I do not think that Labour would win 308 seats on current polling because of the “sophomore surge” effect. The current projection assumes swings are uniform, but Labour will need to regain many seats they lost to the Tories in 2010. These seats were formerly held by Labour members, but are now held by Tories, thus giving the Tories a double personal vote effect in those seats. As a result, these seats are likely to be better for the Tories in swing terms than England and Wales overall.

I think this sophomore surge effect means we should subtract about 20 seats from Labour, and give them to the Tories. That still means Labour would have 288 seats to 270 for the Tories, and that Labour would need support from the SNP to govern.

Scotland only returned one Tory MP in 2010, and the SNP is a left wing party, so if Labour and the SNP combined are a majority after this election, Labour’s Ed Miliband is very likely to be PM. On current polling, a Labour/SNP alliance is the most likely government following the election. There are forecasts that assume that either the current polls are wrong, or that they will move towards the Tories in the last few weeks. My opinion is that you should not assume that this election will follow past elections where the Tories recovered; every election is different.

New South Wales election update

Counting has been completed in all NSW lower house seats, and the Coalition won 54 of the 93 seats, to 34 for Labor, three Greens and two Independents; that gives the Coalition an overall majority of 15. In Lismore, it appeared that the Greens were ahead on election night, but pre-poll votes heavily favoured the Nationals, who won the seat by a 52.9-47.1 margin.

The contest for the final upper house seat has become more interesting. My initial outlook overlooked the “Other” votes; these are not included in the initial party totals, but are assessed for formality later in the process. Many of these “Other” votes are formal below the line votes, and the Coalition is doing very badly on these votes compared to their overall upper house vote. The initial totals include only those who put a “1” in their chosen group’s square above the line.

The upper house check count is currently 61% complete. With many below the line votes now included, Antony Green is projecting final totals of Coalition 9.396 quotas, Labor 6.86, Greens 2.18, Shooters 0.84, Christians 0.62, No Land Tax 0.412 and Animal Justice 0.387. That means 9 Coalition seats out of 21, 7 Labor, 2 Greens, 1 Shooter and 1 Christian, with the final seat a three-way fight between the Coalition, No Land Tax and Animal Justice. Animal Justice may benefit from the distribution of Greens and other left wing micro party preferences, but most upper house preferences exhaust.

If the Coalition only wins nine seats, they would still have 20 of 42 total seats after this election, and would be able to pass legislation opposed by Labor and the Greens with the support of either the Christians or the Shooters. However, the left could benefit from an Animal Justice win in 2019, when the upper house half elected in 2011 will be up for election.

Facts matter. Your tax-deductible donation helps deliver fact-based journalism.