Incumbency matters in elections. This is especially true for the Liberal Democrats whose ability to hold on against the national tide is well known. With poll ratings ranging from 8 to just 12%, this tendency to cling on could be more vital than it has been at any election since 1979 if the Liberal Democrats are to avoid disaster at the next election.
Research by this author and others has suggested that Liberal Democrat MPs have gained an incumbency advantage of around 5-8%, whilst Conservatives and Labour incumbents have enjoyed advantages of around 1-2% and 1.5%-2.5% respectively. This incumbency advantage is personal to the MP rather than incumbent party and it accrues almost entirely during the member’s first term in the House of Commons.
The effect of this is that when an incumbent MP stands down, their party in that constituency underperforms, relative to the results in other seats, as the retiring MP’s personal vote is lost. But the candidate of the same party manages to win the election and stands four or five years later, the party will then be expected to outperform, as this new MP, re-standing for the first time, builds up their own following in the constituency. In literature on the US House of Representatives, this is known as the retirement slump and the sophomore surge – or ‘slurge’.
For the Liberal Democrats a retirement slump of 5-8% of the vote is bad news indeed. At the 2010 election, out of the seven seats where a Liberal Democrat MP stood down, four were lost to the Conservatives. (These losses were actually larger than the party’s net loss of three seats at the election, as net of retirement effects the party managed to gain one seat).
As of July 2014, eight Liberal Democrat MPs have already announced that they are standing down. Unlike in 2010, it looks unlikely that there is going to be any swing towards the Liberal Democrats, which means retirement-related losses can be ill afforded. On the other hand, the good news for the Lib Dems is that ten of their MPs (not counting the by -lection in Eastleigh), are incumbents fighting for re-election for the first time and so could be expected to experience the so-called sophomore surge.
Using the incumbency values previously calculated, one can make estimates of what the effects that these retirements and first term surges will be. It is important to note that Liberal Democrat incumbency could decrease at the next election if voters choose to punish individual MPs for voting for unpopular measures such as tuition fees. Equally, however, it could increase if it turns out that voters wish to punish the Liberal Democrats as a party, but not take it out on a local MP they consider to have done a good job locally. For now, let’s just assume the incumbency bonus stays the same, using the average for the last three elections. For seats in Scotland and Wales, due to lower number of cases, and the four party system, the data is calculated using the average of all elections back to 1983.
The two scatter diagrams below show the 57 incumbent Lib Dem MPs’ majorities over their nearest rival at the 2010 election. The chart on the left plots the relative safety of each of the Liberal Democrat MPs in 2010. The vertical axis shows the majority over their nearest rival, whilst the horizontal axis shows the Lib Dem share of the vote. The colour of the marker shows the party in second place; and the further towards the top-right, the safer the incumbent. The majority of the Lib Dem incumbents find themselves up against Conservatives as the second place challengers. Just two incumbents face Nationalist candidates in second place. The markers show the seats from the extremely precarious Simon Wright on 29% of the vote (a majority of under 1% over Labour), on the far left, to the ultra-safe Alastair Carmichael in Orkney and Shetland on over 60% of the vote (a majority of over 50% ahead).
The second chart shows what the results would be with a zero swing, but with the effects of the retirement slumps and first time surges built in. The bad news for the Lib Dems is clear: right away, three seats would be lost to the Conservatives: Berwick-upon-Tweed, Mid Dorset & Poole North and Somerton & Frome, whilst Brent Central would be lost to Labour. There is also some sorting within the constituencies. The retirement of Malcolm Bruce puts his Gordon constituency at risk with reducing the vote share to less than 30% and the SNP in a close second place. However, two seats narrowly won from Labour in 2010, Norwich South and Bradford East, move into safer territory, as their MPs would be expected to receive a first time surge.
In other words, even before there is any swing against the party, they are four seats down, purely due to retirements.
The effect of having ten first term incumbents is that some of the party’s seats will become easier to defend. The distribution is also important since seven of the ten seats are in constituencies with majorities under 5% of the vote over the nearest opposition party. This will mean that if there is a swing against the party, the number of seats beyond those already expected to go due to retirement effects might well be lower than would otherwise have been expected. The adjusted majority of the 20th safest seat the Lib Dems have on paper is 7.34%. Taking incumbency into account the 20th safest (including the 4 that would be lost) would be larger at 8.28%. Therefore with a theoretical swing of approximately 4% against them, the Liberal Democrats end up having more seats once incumbency is taken into account than they would on paper.
Liberal Democrat MPs have much larger incumbency advantages than those of other parties and this is likely to have important effects at the next election. The balance of the effects improves the worse the party does: a tiny swing towards the Liberal Democrats might see them losing seats, all other things being equal, whilst a large swing against them might be partially mitigated by having so many first term incumbents.
Those looking to target the most marginal seats – political parties, academics and journalists – might well choose the wrong targets if they fail to take incumbency changes into account. Wells, for example, may appear to be an extremely attractive target for the Conservatives, but should Tessa Munt experience a sophomore surge similar to those achieved by first term incumbents in the previous three elections, the seat could be a very much tougher nut to crack. The other implication is that with eight retirements already, the party might do well to persuade as many of the rest of its incumbents to continue.
Tim Smith is a PhD student in the School of Politics and International Relations at the University of Nottingham.