By political geography we here refer to that branch of psephology that examines how the geographical distribution of votes can determine the results of an election. Although this does not to matter that much in proportional voting systems, it is often an important factor in elections conducted by first past the post (FPTP) where the art of winning elections becomes not so much a matter of how many people vote for you, but of whether the right people vote for you in the right places.
A case in point would be the General Election of 2005 where the Conservative Party outpolled their Labour rivals by 35.74% to 35.46% in England. However despite the fact that the Conservatives received some 65,000 more votes than their main rivals they only managed to win 194 seats, whilst the Labour Party romped home with a total of 286 seats. Of course, Labour did much better in both Scotland and Wales and so their overall share of the national vote was 35% as against 32% for the Conservatives, and thus the party won a fairly handsome majority in the House of Commons, despite recording the lowest share of the vote ever recorded for a winning party at a UK General Election.
Thus it is said that the Political Geography of Britain favours Labour, in that the party enjoys an inbuilt advantage over their Conservative rivals believed to be of a hundred and ten seats. In consequence it has been said that David Cameron's Conservatives need a swing of 7% to obtain even the barest majority; and since this is larger than any swing achieved in any previous modern election, it is by implication unlikely to happen, and a hung Parliament is therefore seen by many as the most likely outcome of the General Election of 2010.
There are a number of reasons why the political geography of Britain should favour Labour Party in 2005, which fall broadly under the headings of unequal constituency sizes, differential turnout, voting efficiency, and the third party effect.
Unequal constituency sizes
To put it simply, an analysis of the results of the 2005 election shows that the average Labour held constituency comprised 66,800 voters whilst a Conservative held constituency was about 10% larger and included 72,950 voters; a situation that arose as a result of the combination of demographic movement, and the over-representation of both Wales and Scotland.
As a general rule of thumb, poor people tend to vote Labour and rich people tend to vote Conservative. Thus those urban parliamentary constituencies with a preponderance of the unemployed and the low paid tend to be Labour seats, whilst those rural seats that are inhabited by well-paid professionals and their ilk tend to be held by the Conservative Party. It is a simple fact of life that as soon as a member of the urban poor gets some money together they tend to leave the Labour voting inner city and go and live in a nice Conservative suburb. As a result of this urban drift the electorate in Labour held seats tends to shrink over time, and so Labour held seats tend to have a smaller electorate than Conservative ones.
Whilst successive Boundary Commissions seek to counteract this trend by redrawing constituency boundaries as best they can, they do so only after the fact and with a considerable lag. The Commissions are only required to report every eight to twelve years so that, for example, the General Election of 2005 was fought on the basis of the review carried out in 1995 using data collected in 1990. As a result it has been estimated that Labour enjoyed an advantage of some 21 seats as a result. However the General Election of 2010 will be fought on the basis of a new boundary review which it is believed will reduce this advantage by at least a dozen seats.
In addition for various reasons which most people have now largely forgotten, both Scotland and Wales have in the recent past enjoyed the privilege of electing more members of the House of Commons than their respective populations would warrant. The over-representation of Scotland was considerably reduced after the establishment of the Scottish Parliament, and thus whilst Scotland sent 72 members to Westminster in 2001 this was reduced to only 59 in 2005; although in order to ensure exact equivalence with England this total should be only 55, so that Scotland remains slightly over-represented. The situation is however far more pronounced in Wales where there are 40 constituencies rather than the 32 that it 'should have' to be on an equivalent basis with England. Since both Scotland and Wales have been predominantly Labour supporting in the most recent past, this was worth about six seats to Labour at the 2005 election, most of which is down to Wales which is worth five seats on its own.
It should however be noted that the Conservatives actually polled more votes in Wales than Labour for the first time in history at the European elections of 2008, and were the Conservatives to come even close to repeating this result at a general election then much of that Labour advantage would disappear.
Turnout tends to be higher in Conservative seats than it is in Labour seats, so that all other things being equal the typical Conservative MP receives more votes than his or her Labour equivalent. This may well have nothing to do with political affiliations per se, as the propensity to vote increases with age and socio-economic status, which is to say that a middle class old age pensioner is far more likely to vote than a twenty something factory worker, and since the propensity to vote Conservative rather than Labour also increases with age and socio-economic status, it may simply be nothing more than co-incidence.
But nevertheless it has been argued that a Labour supporter is only 65% as likely to vote as a Conservative or Liberal Democrat supporter. The polling organisations certainly think so, as the published results of opinion polls are generally weighted to take account of the higher likelihood that a declared Conservative supporter will actually vote. And human nature being what it is, given the disinclination to vote of the typical Labour supporter, they are more likely to make the effort to vote in those marginal seats where their vote might make a difference than in safe seat.
Hence it is also known as the abstention effect, in the sense that it is suggested that Labour supporters are more likely to abstain from voting in so-called safe seats, and has given rise to the belief that there is a large pool of untapped Labour support that simply needs to be motivated to actually turn and vote. Although it is perfectly possible that the real explanation is the other way around; i.e. those individuals who are most disinclined to vote are simply more likely to claim that they support Labour than any other party.
The concept of voting efficiency relates to the efficiency in the distribution of a party's support, which is to say that under first past the post it is far more 'efficient' for a party to receive votes in those marginal constituencies where it might make a difference as opposed to registering votes in those constituencies where the result is already a foregone conclusion.
Or to put it another way; faced with the choice of which party to vote for in any particular constituency, an individual elector might decide to vote for a party that loses, in which case it would be a wasted vote, or for a party that would otherwise have won anyway, in which case it would be a surplus vote, and only in certain cases would their vote be an effective vote in that it actually helped determine the outcome of the contest. Thus the extent to which any party succeeds in attracting effective votes it is said to possess voting efficiency, and it was the case that the Labour Party had a greater vote efficiency in 2005 than the Conservatives.
It is however worth noting that historically speaking the opposite was true in the period between 1950 and 1989, mainly because the Labour Party used to possess a significant number of very safe seats in which it accumulated a large number of surplus votes. However the post-war decline in mining and certain large scale manufacturing enterprises resulted in the gradual disappearance of such seats, effectively dispersing the Labour vote over a wider area. At the same time there is evidence that from 1992 onwards the Labour Party engaged in more focussed campaigning, and directed its resources at winning support in marginal seats. It was also the case that the Labour advantage from voting efficiency declined in 2005 compared to 2001 as a result (it is believed) of a more targeted campaign by the Conservative Party and may well decline further in 2010 thanks to the 'target seat campaign' which is funded and masterminded by Michael Ashcroft.
The Third Party effect
British general elections are often seen as a battle between the two competing main Labour and Conservative parties, and although this was the case in the years 1945 to 1970, we have since seen the re-emergence of the Liberal Party, subsequently the Liberal Democrats, as at least a semi-significant electoral force, together with the emergence of both the Scottish National Party and Plaid Cymru as serious challengers in their respective domains.
The appearance of a third party candidate into the fray does have an effect on the battle between Conservative and Labour, and as a general rule where third parties enter the contest and begin to gain a substantial proportion of the vote, then it is the Conservative Party that benefits. However where third parties do well enough to win seats, they do so at the expense of the Conservatives. Or to put it in simply in terms of recent British political experience, whilst a Liberal Democrat candidate might split the anti-Conservative vote and make it easier for the Conservative to win, those seats that the Liberal Democrats win tend to be previously Conservative held seats. As far as the UK General Election of 2005 was concerned, the Conservatives suffered more from the loss of seats they would otherwise have won, than they gained from the splitting of the vote, and so the net effect was worth nine seats to Labour. This was less than the previous election in 2001, where the net third party effect was worth fifteen seats to Labour.
At this point, the question of tactical voting naturally arises, particularly since anti-Conservative tactical voting has at least anecdotally been seen as one of reasons behind Labour's recent electoral successes. It is often assumed that the reason for this is that both the Liberal Democrats and Labour are 'progressive' parties, and that therefore their respective supporters see the other as the least bad option. However logic dictates that it would be in the best interest of Liberal Democrat supporters to vote tactically against whichever one of the two major parties they believed was their major competitor for the coveted No. 2 slot. In 1997, 2001 and 2005 that would have been the Conservative Party; in 2010 it would more likely be the Labour Party.
However the biggest problem with tactical voting is that it is almost impossible to quantify, since it is very difficult to disentangle the effects of tactical voting from the more general third-party effect or indeed voting efficiency. For example, if it was the case that the Labour Party had performed better than expected in one constituency by squeezing the Liberal Democrat vote, it is hard to say whether that arose as a result of tactical voting by Liberal Democrat supporters, or simply because Labour devoted more resources to campaigning in that constituency and had thus succeeded in persuading voters that it was the better option. It's therefore the case that many see tactical voting as simply one of the drivers behind voting efficiency.
In summary therefore, the outcome of a British General Election depends not just on how many votes each party wins but also on how those votes are distributed across the nation's constituencies, and indeed on how those constituencies were constructed in the first place. As far as the most recent British General Elections are concerned it is undeniable that the pattern of voting has been favourable to the Labour Party. However, one should not make the erroneous assumption that this Labour 'bias' in the British electoral system is somehow inbuilt and permanent. Rather it is simply the result of the combination of many factors which have changed in the past and will change again in the future; the political geography of Britain actually favoured the Conservative Party in 1950-1964, was broadly neutral in the years between 1966 to 1987, and has only favoured Labour since 1992.
But the existence of this bias explains the continued repetition in the media of the claims that the Conservative Party requires to achieve certain targets (such as a 10% lead over Labour, or to a vote share of over 40%) in order to form a majority. The truth is however, as far as the forthcoming General Election is concerned, that no single party actually needs to achieve any kind of swing, or to hit any particular target share of the nation's vote; all that victory requires is for 326 of that party's candidates to receive one more vote than any one of their competitors. It's as simple as that, and everything else is just post-hoc analysis.
NOTE: Of course, what we are considering here is the Political Geography of Britain, rather than the Political Geography of the United Kingdom of Great Britain and Northern Ireland, since the Political Geography of Northern Ireland is something else entirely.
- Peter Kellner, Reasons to be cheerful, Fabian Review, Autumn 2008
- UK General Election 2005:Regional distribution of seats and percentage vote
- Martin Baxter, Con-Lab Gap Analysis, 3 September 2006
- Anthony Wells, Electoral Bias
- Ron Johnston, David Rossiter, Charles Pattie, Disproportionality and bias in the results of the 2005 general election in Great Britain: evaluating the electoral system’s impact