UK election prediction: this week’s result won’t reflect the voters’ will

The UK general election could go either way. The one certainty is that the numbers of seats won won’t match the votes for each party. AAP/Newzulu/Stephen Chung

This week the “mother of parliaments” faces a general election in the UK. The level of voter support that individual parties are likely to get can now be predicted at about one third for each of the two traditional parties (Conservative and Labour) and one third divided between a range of other parties. This has been the finding of opinion polls for several years now.

What we can’t predict from these figures, because of the electoral system, is who would “win” the election and form a majority in the House of Commons.

Why is this so?

The electoral system in use for the 650 members of the House of Commons is the single-member electorate system, whereby each winner is decided by the “first past the post” method. This means that voters have to put a cross next to the name of one candidate only. Whichever candidate gets the highest number of crosses wins and is the sole representative of the constituency, even if well short of 50% support.

When this happens across a range of constituencies, it leads to distorted election results. For example, in 2005, the Labour Party under Prime Minister Tony Blair “won” the election with a large majority in Parliament, 356 seats, but only 35.3% of the votes.

Today, in many constituencies in the UK, there are three, sometimes four or even more candidates who have a real chance of winning. Consider a possible constituency where not only Conservative and Labour, but also Liberal-Democrats, UK Independence Party (UKIP) and the Greens each has solid support. Norwich South is an example, where five parties are over 12% in recent opinion polls. If candidates of those five parties are the only candidates on the ballot paper, the leading candidate can be elected with just over 20% support, as in this hypothetical example:

In this situation, despite the fact that 78% of votes were cast for other candidates, the Greens would win. Fortunately in Australia, preferential voting means that the Greens candidate could only go on to win if also preferred by sufficient voters whose first preference was for candidates receiving fewer votes.

Results like this hypothetical example, where a large of majority of voters do not support “the winner” are increasingly common in first-past-the-post systems such as in the UK, Canada and India, where there are also multiple parties. When this is repeated in many constituencies, as is often the pattern, parties with far less than 50% support can form a majority government.

In India, for example, the present government coalition received just 31% of the vote but won 282 of the 543 seats. In Britain, which has effectively been a multiparty state for over a generation, the Conservatives and Labour routinely get a far higher percentage of seats than votes.

UK opinion poll trends, April 2010 to present

Lines represent the opinion polling trend (using a 15-day moving average) and each dot an individual poll result. BLUE: Conservative. RED: Labour. PURPLE: UK Independence Party (UKIP). YELLOW: Liberal Democrats. GREEN: Green Party. Wikimedia Commons/Impru20, CC BY-SA

The fracturing of the vote in the UK means that in recent opinion polls, the lead is see-sawing between Labour and the Conservatives, averaging out to something like this:


Let’s suppose that there is a cluster of five constituencies, as a group voting exactly according to this pattern. And let’s suppose the votes for the parties in each of these constituencies are as in the table below. Although in these five constituencies the Conservatives lead the total votes by 1%, because of the support within each of these winner-take-all constituencies, they win one seat, Labour wins three and UKIP wins one.


What has happened here is that Labour has comfortably retained its two safer seats, but the Conservatives have lost votes to UKIP in their formerly safest seat by a narrow margin. (Exactly this happened in several recent by-elections.) Labour wins the seat listed as “marginal” by a narrow margin.

The election contest is wide open

It is certainly possible that results like this could be repeated in sufficient different parts of the country to mean that, for example, Labour could “win” the election despite having fewer votes than the Conservatives and less that one-third of the overall vote.

ABC election analyst Antony Green has observed that electoral results suggest the system may indeed work in Labor’s favour. Others, including celebrated US tipster Nate Silver, are predicting a Conservative victory, with the caveat that the outcome could be “incredibly messy”.

Regardless of who forms government, because of the electoral system, the House of Commons is extremely unlikely to represent the will of the British people, as expressed by actual votes cast. With exactly the same percentage of votes cast across the country, similar to the opinion polling cited above, there could easily be one of three results:

  1. A large majority of seats for Labour

  2. A large majority of seats for the Conservatives

  3. A hung parliament.

In the UK it is clear that no party has any chance of getting the support of anywhere near a majority of the voters, and yet one or other of those parties may yet win a substantial majority of seats.

Is it any wonder that UK voters are increasingly turned off from the political process?

The solution is to introduce 21st-century electoral arrangements that actually represent the voters’ will. The best system, known as single transferable vote proportional representation, has been operating for nearly a century in the UK’s closest neighbour, the Republic of Ireland, and was adopted for the Northern Ireland Assembly with cross-party support. Proportional representation for the UK would mean that:

  • Each of the parties that receives a substantial percentage of the vote would actually be represented in parliament (the first-past-the-post system favours parties with concentrated support in some electorates, rather than broadly distributed support).

  • The majority in parliament would represent the views of the majority of voters

  • Voters for substantial minority parties across the country would have their own representatives

  • There would be competition everywhere, leading to regional representation in line with local support levels.


The author thanks Geoffrey Goode and Bogey Musidlak, of the Proportional Representation Society of Australia, for their help in preparing this article.

You can read The Conversation’s comprehensive coverage of the UK general election here.