By Tim Smith
At the 2010 election there was a considerable pro-Labour bias in the electoral system. Despite a 7.3% lead in votes over Labour in Great Britain, larger than had been achieved in outright victory by Ted Heath in 1970 and Margaret Thatcher in 1979, the Conservatives finished 19 seats short of an overall majority. The bias in the electoral system was estimated to have been 54 seats in Labour’s favour, such that if the two parties had achieved the same vote share, Labour would have won 54 more seats than the Conservatives. This bias in the system was actually considerably smaller than that estimated at the 2005 election, (111 seats). Labour had won an overall majority of 66 seats on a vote share of just over 3% more than the Conservatives in Great Britain. They had won 93 more seats than the Conservatives in England despite polling 70,000 fewer votes.
In the 2010–2015 parliament the Conservatives attempted, but failed, to complete a widespread boundary review in order to equalize the size of the constituencies in the UK, after the proposals were blocked in the House of Lords. A number of commentators have stressed the effects that the differing sizes of constituencies, known as malapportionment, has on the bias in the system because Labour constituencies tend to have fewer voters than Conservative ones. The reason given for this is that in the post-War period, voters have tended to move out of inner city (Labour) areas towards rural and suburban (Conservative) areas, leading to an increasing number of voters in Conservative seats relative to Labour seats, until each periodic boundary review corrects matters. As the following decomposition analysis of the two-party bias shows, malapportionment is in fact a relatively small part of the problem for the Conservatives.
The table shows that at the 1987 election there was a small pro-Conservative bias of 6 seats, but the bias swung in Labour’s favour at subsequent elections peaking at the 2001 election at 131 seats and then falling back to 54 seats at the 2010 election.
The first category, size, at 18 seats, was not the biggest factor at any of the elections. The ‘between nations’ component is largely due to the over-representation of Wales, which unlike Scotland, has not seen its quota of seats brought down since devolution. The bias from within the nations, due to differing seat sizes, dropped from a pro-Labour bias of 20 in 2005, to 9 in 2010 as the Fifth Review of the Boundary Commission for England & Wales took effect. Looking ahead to the 2015 election the consensus is that the size bias will increase slightly as the existing boundaries will be five years more out-dated than they were in 2010. There are two additional unknowns, firstly, the effect of individual voter registration, and second, the large last minute surge in registration which was estimated to have been approximately half a million on election day.
The biggest single reason for the bias in the electoral system is the effect of differential abstentions. At all elections in the table, but especially since 2001, there has been a considerably lower turnout in safe Labour seats, than in Conservative ones. The abstention bias dropped slightly at the 2010 election because the turnout in Labour constituencies rose slightly more than in Conservative ones, but was still 7 percentage points lower. There is little strong evidence that there is a going to be a significant change in differential turnout in England. However, polls are predicting that the turnout in Scotland will be much higher than it was in 2010. Despite the SNP surge, Scotland is still likely to see Labour beat the Conservatives in votes, and so an increase in turnout could reduce this component of the bias.
At the 2001 and 2005 elections, Labour votes were extremely efficiently distributed relative to the Conservative votes. At the 2005 election, Labour’s 35% of the English vote was far more unevenly spread than the Conservatives same total. Labour did abysmally badly in a large number of Conservative-Liberal contests where they had no chance of winning, but therefore wasting fewer votes than the Tories, who performed better, in relative terms, in their weakest areas. Labour also benefitted from incumbency advantage from MPs first elected in 1997 in former Conservative areas, many of whom were able to dig in and resist the modest Conservative swing in both the 2001 and 2005 elections. One of the Conservatives’ best achievements in 2010 was to obviate this efficiency advantage by Labour. They were assisted by the high number of incumbent retirements and possibly by better targeting of seats as well as Ashcroft money. At this election, the large number of first term Conservative incumbents defending marginal seats may mean they gain a disproportionate boost in exactly the right places, and its possible that the efficiency advantage could swing in their favour for the first time since 1992.
Third Party Votes/Victories
‘Third party votes’ shows the effect on the bias from wasted third party votes. Throughout the observed period this has favoured the Conservatives because in a large proportion of their seats the runner-up has been the Alliance/Liberal Democrats. The third party wins component shows the effects of wasted major party votes in seats where they lose to third parties, primarily the Lib Dems. At the 1987 election, The Alliance took a quarter of the vote but only 22 seats, and with a huge number of second places in Conservative areas, the net was a pro-Conservative bias of 9. Since the 2001 election the net of the two has been in Labour’s favour since the number of Lib Dem and other minor party wins in Conservative areas has more than offset their advantage from wasted third party votes.
In England in 2015 there is likely to be a big drop in both Liberal Democrat votes and seats, and so the net impact on the bias is uncertain. Wasted Lib Dem votes are likely to be replaced by wasted Green and UKIP ones, but with these two most unlikely to achieve many wins, there is considerable additional uncertainty. However, in Scotland with at least 30 Labour seats likely to be lost to the SNP, there will be a large change to the third party wins component against the Labour Party, and this could well mean that the net of third party votes and seats moves sharply in the Conservatives’ favour.
As this post shows, there are considerable uncertainties how the components of bias in the system are going pan out at this election. However, there is good reason to expect a fall in the pro-Labour bias, especially due to a change in efficiency of votes to seats caused by the large number of first term incumbent Conservatives and the predicted SNP earthquake in Scotland. For the first time in nearly 30 years, one cannot even rule out that on the same share of the vote the Conservatives could win more seats than Labour.
Tim Smith is a part time Politics PhD student at Nottingham University LDXTES@nottingham.ac.uk