Non-fiction‎ > ‎

Time to Hang Up First Past the Post

(Or how the 2010 General Election demonstrates the need for electoral reform in not many words and a few useful pictures)

© 2010 Christi Scarborough

Creative Commons License This work and the accompanying graphs and tables are licensed under a Creative Commons Attribution-Noncommercial-Share Alike 2.0 UK: England & Wales License.  If you wish to license any part of this work for commercial use, please contact me. 

Introduction


I am not here to wax philosophical about why proportional representation, the idea that the number of Members of Parliament should reflect the proportions of votes cast for a particular party, is a good idea.  Other people have done that much more eruditely and comprehensively than I am likely to be able to.   If you're interested in the philosophical arguments, the Electoral Reform Society's excellent PR Myths pamphlet is a great place to start.  What I am trying to accomplish here is to make a very abstract issue a little more concrete.  To this end, I've taken the voting figures from the UK General Election, held on Thursday 5th May 2010 and used them to illustrate exactly what the lack of proportional representation means for people living in Britain today.

Welcome to Reality Gap Britain


The Conservatives pretty much won the election.  They have 305 elected (and 1 un-re-elected due to a postponed election) MPs.  That's more than anyone else by a good 40 MPs and a clear mandate to run the country, isn't it?  Well, actually, no.  The Conservatives got 36.1% of the vote.  Or put another way 63.9% or almost two thirds of the country don't want the Tories to be running it. 

So how does this happen?  Our elections run under a system called first past the post.  It's the oldest democratic system in the world, and it's really starting to show its age.  The UK is divided into 650 parcels called constituencies of varying sizes, and these change from election to election.  If we attribute noble motives to the people who change them, then they change to reflect the movement of population throughout the country.  If we're a bit more cynical, we perhaps think that they change to ensure that the party in power has the best chance possible of staying that way.

The rules are simple.  Whoever gets the most votes within a constituency wins.  Everyone else loses.  Or to put it another way, all the people that voted for someone other than the leading candidate in each constituency might as well not have bothered.  Their votes have no effect on who is in power or runs the country.

Their votes are wasted.

A Proportional Revolution


There is another option though.  It is possible for the make up of the House of Commons to reflect the level of support throughout the country for each political party.  I'd like you to imagine we lived in a country where the number of MPs who sat in parliament were directly in proportion to the number of votes.  So a party who got 35% of the votes would get 35% of the seats.

This is Proportional Representation, and in the real world there are a number of clever ways to work out who gets selected and still maintain the link between MPs and local constituencies.  These won't be covered here today because I'm already on paragraph seven, and there haven't been any pretty pictures yet.  So without further ado, here's a graph showing the difference between the number of MPs we have after the election, and the number of MPs we would have had if our government had been chosen under PR.


This shows pretty clearly how First Past the Post gives an advantage to the larger, established parties over the newer, smaller parties.  The Tories and Labour each get about 70 seats more than they would under PR, whereas the next largest party, the Lib Dems, loses out to the tune of 100 seats.  If our electoral system reflected what the country had actually voted for, there would be a 150 Lib Dem MPs right now. 

 
As you can see, the UK Independence Party and the British National Party would both be entitled to seats under proportional representation.  It's not something I particularly relish, but it's kind of the point of democracy that if a group of bigoted idiots can persuade people into voting for them, then those people's voices should be heard.  Naturally freedom of speech means that we can dissent and mock as loud as we want, but almost a million people voted UKIP and over half a million for the BNP.  It's wrong to deny them representation just because what they say is harmful and divorced from fact.   Far better to let people know just how wrong and divorced from fact what they say is.  Denying 1.5 million voting adults a voice seems like the greater of two evils here.

What's Going On?


Our system essentially ignores the fact that in modern politics the vast majority of voters are voting for a party or their choice of Prime Minister rather than a local MP.  The whip system ensures that an MPs discretion in voting is severely limited and voters understand that to elect an MP is to further that party's legislative agenda rather than the MPs personal one.  There are both good and bad things about this, but it's hard to deny that this is the way modern UK politics works.

Note that this data is entirely agnostic as to what system we use for selecting who gets to be an MP under Proportional Representation.  Most real world implementations of PR contain various refinements to reflect geopolitical realities or to limit the number of minor parties.  Discussion of how to implement PR while maintaining a link between constituency and electorate is not covered here.  This is an exploration of how the UK parliament would be made up if the number of seats was proportional to the total number of votes cast country wide.

What a waste


Even supporters of the main two parties can lose out as a result of first past the post.  If you're a Conservative voter in the North East, or a Labour voter pretty much everywhere in the South of England outside of London, it's likely your voice is also being ignored.  The likelihood of this happening does vary from party to party.  Let's have a look at how likely a vote for each of the parties is to be worth nothing.


("Others" in this graph are parties with less than 250,000 votes)

As you can see, over 90% of Green Party votes go to waste.  But if we look at the Conservatives, we can see that almost a third of Tory voters get no say in the make up of their government.  Possibly they might not mind, as under first past the post, the votes of the 70% who did get heard more than make up for their loss, but however you look at it, with an electoral system that's ignoring so many of the people is it surprising that so many people don't vote because they regard it as a pointless waste of time?

So we can see that a lot of votes get wasted.  It's time to look at exactly how many votes in a UK general election really count for nothing.  Here are the number of votes that got someone elected in 2010, and the number of votes that had no effect for each of the big three UK parties.


Yes, you read that right - five and a half million of the six and a half million or so people who voted for the Liberal Democrats this election had no effect on its outcome.  That's about one ninth, or more than 10%, of the 45 million people allowed to vote in the UK.   But there's another seven million Labour and Conservative voters denied their say too.

In fact, 52.8% or more than half of the 29.5 million votes cast in the UK election had no effect on the process.  In 2010 more people were unrepresented by first past the post than were able to make their voice heard.  That can't be right.

Here's now it worked out for the smaller parties.  This graph only includes parties who had at least 1 MP elected.


Now earlier I said that first past the post favours majority parties.  In this chart you can see a slightly odd phenomenon, which is that the two main Northern Irish parties, Sinn Fein and the Democratic Unionist Party, seem to have done reasonably well out of first past the post.  This is because in many ways the Northern Irish constituencies operate as a microcosm.  The main UK parties don't stand in Northern Ireland as they tend to in Scotland and Wales.  It probably won't surprise you to learn that the third largest Northern Irish party, the Ulster Conservatives and Unionists, got over 100,000 votes, a little under two thirds of the votes of the DUP and Sinn Fein, and didn't manage to win a single seat.

Winners and Losers


At this point, I'd like to make a brief detour to look at the biggest winners and losers geographically as a result of first past the post.  Here are the ten most unrepresented constituencies in the UK.

Constituency Wasted Votes Total Votes Turnout Percent Wasted Votes
Norwich South 33591 47551 65% 70.6%
Brighton Pavilion 35596 51834 70% 68.7%
Argyll and Bute 30915 45207 67% 68.4%
Oldham East and Saddleworth 30334 44520 61% 68.1%
Great Grimsby 22177 32954 54% 67.3%
Hampstead and Kilburn 35490 52822 66% 67.2%
Birmingham Hall Green 32688 48727 64% 67.1%
Derby North 30184 45080 63% 67.0%
Ynys Mon 22954 34444 69% 66.6%
Ashfield 31957 48196 62% 66.3%

In Norwich South, less than 30% of those who voted got the MP or party they asked for.  Interestingly, Norwich South is a highly marginal Lib Dem constituency.  First past the post is fickle and inconsistent - while in general it works against minor parties like the Lib Dems, here it means that they hold a seat despite over two thirds of the constituency not wanting a Lib Dem MP.

Constituency Wasted Votes Total Votes Turnout Percent Wasted Votes
Liverpool Walton 9626 34335 55% 28.0%
Belfast West 9293 32133 54% 28.9%
Knowsley 13008 44658 56% 29.1%
East Ham 14902 50373 56% 29.6%
Glasgow North East 9309 29409 49% 31.7%
Coatbridge, Chryston and Bellshill 13907 41635 59% 33.4%
Bootle 13851 41277 58% 33.6%
Kirkcaldy and Cowdenbeath 16243 45802 62% 35.5%
Liverpool West Derby 12831 35784 57% 35.9%
Down North 12300 33481 55% 36.7%

You might think from the data for constituencies with least disenfranchisement that there are a few places that are served very well by first past the post, as these are constituencies where almost all of the voters are in favour of the same party.  These are all very safe seats, and it is interesting to note that the turnout in these constituencies are all lower than the national average of 65.1%.  This may be because the vote against the majority party is smaller than it would otherwise be, as people perceive there is simply no point in voting against the dominant party.  Perhaps they aren't being served as well by first past the post as it might appear at first.

Summing it all up


Here's a chart that shows us where all of those wasted votes went.


I offer no conclusions because this isn't an essay, just an attempt to give you the information you need to make up your own mind.  Obviously I'm a believer in PR and I hope that by reading this you too will come to see the merits of it, but that's for you to make your mind up.

If you would like to know more about Proportional Representation and what you can do to support it, please take a look at Take Back Parliament.

Notes


This will only be of interest if you want to know the details of exactly how the data was prepared. An analysis is only as good as the analyst and the data available, so there is room for error and results should be verifiable. 

This data is publicly available but hard to acquire in analysable form.  I am therefore indebted to the Guardian for making the data available.  No official results have been published at the time of writing, and the Guardian and the BBC have slightly different figures for votes cast, so there may be minor inaccuracies in this data.

For the purposes of this analysis, the one constituency that was not voted on in this election due to the death of a candidate has been removed from the data set.

The calculation of number of seats under Proportional Representation was intended to be illustrative and give the closest mathematical match to the number of votes cast.  Different mechanisms for calculating seats under PR are possible and no implication is intended as to the merits or otherwise of this particular one.  One postponed constituency, the Speaker's constituency, and votes in the Speaker's constituency were removed from the calculation, leaving a total of 648 seats to be allocated proportionally.  Figures for numbers of MPs are rounded down to the nearest whole MP.  The number of "left over" votes for each party is then calculated as follows:

Total number of votes cast for this party - ( number of seats allocated to this party * total number of votes cast / 648 )

The remaining seats were then allocated one each to the parties with the largest number of "left over" votes.

A copy of the full results used to generate this article is included below.  It is in Excel .xls format.   There is also a Visual Basic function used to calculate rankings in each constituency (very slowly).  This has not actually been used for generating this data, so it is safe to leave macros disabled.  It is included as I am hoping at some point to be able to expand the analysis to consider what percentage of the vote if would have been necessary for each major party to get in order to obtain a majority, but that's a project for another day.   Re-use of this file is permitted under the same terms as re-use of this document.  Be warned, the file is pretty large.  I have no idea why.
Ĉ
Christi Scarborough,
13 May 2010, 16:36
Comments