Why ‘incumbency advantage’ may decide the next election


by Don Paskini    
10:59 am - October 8th 2012

      Share on Tumblr

A lot has been written about how the constituency boundaries mean that Tories face a tough challenge to win the next election. For example, top pollster Peter Kellner argues, “For the Conservatives to win an overall majority, they need around a seven point lead in the popular vote.”

In contrast, the Mail on Sunday reports that Tory strategists ‘are confident that incumbency gives them an advantage’ as they plot how to win the next election.

Let’s have a look and find out who might be right.

The case that ‘incumbency advantage’ could make a large difference is that at the next election in a constituency which the Tories gained in 2010, the Tories will have benefited from having an MP building up their profile, recruiting volunteers, raising funds and campaigning for five years.

In addition, the Tories will have received over £150,000 per year in state funding for the MP’s salary, office costs, free postage of letters and a team of staff. Between 2010 and 2015, in the top fifty marginal constituencies which Labour is seeking to win from the Tories, the taxpayer will have provided in excess of £35 million to help the Tories defend their majorities. On top of this, these are all seats where Labour used to have an incumbency advantage (which will have helped them at the 2010 election), but where they are now forced to cope without these resources.

There are some numbers to test the ‘incumbency advantage’ hypothesis. Professor Philip Cowley reports that, ‘In 2010, Labour’s vote fell by an average of 7.4 percentage points in its seats that were not defended by the incumbent MP, more than two points higher than the equivalent statistic in seats where the incumbent stood again (–5.2).

The Conservative vote rose on average by 2.9 percentage points in Conservative held seats that were not being defended by an incumbent, but 4.1 points where the incumbent MP was still in place. And incumbent Conservative MPs who first won their seats in 2005 – and who thus had the opportunity to acquire a personal vote for the first time – saw their vote increase on average by 5.6 points.’

To estimate what effect this incumbency effect might have, let’s compare two scenarios. In one, we assume a close election, with some recovery for the government compared to the situation now, where Labour and the Tories each get 38% and the Lib Dems 15%. In the other scenario, we assume the same level of support nationally, but in addition apply a 2% swing to the incumbent in all seats where there is an MP restanding for the first time. This is a cautious estimate, when compared to the 5% swing to first time Tory incumbents which Joan Ryan found in her research on the 2010 election, but should give us an idea about whether this matters.

Using Electoral Calculus’ election predictor, the first scenario gives Labour 321 seats, 281 for the Tories and 23 for the Lib Dems.

But if we add in the effect of incumbency as above, then by my quick tally, we get 302 seats for the Tories, 298 for Labour and 25 for the Lib Dems. The whole of the supposed ‘bias’ in the electoral system towards Labour from the current boundaries disappears.

Let’s consider one further effect of incumbency. To get a majority, the Tories will be attempting to win seats from their coalition partners, the Lib Dems. Again, there is some evidence that Lib Dems benefit from having an incumbent MP. So let’s assume that the incumbency bonus above also applies to all Lib Dem MPs. Then we would end up with the following result:

Labour 298 Tory 294 Lib Dem 33

So, in summary:

Past evidence and local results since 2010 suggests that there is an ‘incumbency effect’, particularly for MPs elected for the first time in 2010.

The benefits of incumbency mean that it is harder for either Labour or the Tories to win an overall majority, as they have to battle against incumbent MPs who have large amounts of state funding to bolster their campaigns.

The Lib Dems are likely to do better than suggested by election predictors which don’t consider the benefits of incumbency.

    Share on Tumblr   submit to reddit  


About the author
Don Paskini is deputy-editor of LC. He also blogs at donpaskini. He is on twitter as @donpaskini
· Other posts by


Story Filed Under: Blog ,Conservative Party ,Labour party ,Libdems


Sorry, the comment form is closed at this time.


Reader comments


1. Albert Spangler

This is interesting, but has some odd implications. Would this not be a symptom of the whole Safe Seat problem?

Perhaps in solidly Labour/Tory areas, the opposition vote is less motivated to actually bother voting, while the incumbent vote actually has a reason to go out there and make a change? It may be that this is what happens in Brighton, where people actually think voting Green isn’t a waste and so more people are likely to do so.

What about areas with steady voter decline though? The number of Labour supporters voting in solid Labour areas was on the decrease for many years. Or would the argument be that this decline was actually slowed due to this incumbency effect?

This really serves to illustrate the idiocy of our electoral and poltical system.

2. Barney

What effect would the blocking of the bill to decrease the Commons seats have on the Boundary Commission proposals, which were based on the smaller number of seats, and for which the English one is about to publish another proposal (Oct 16th)? Does this mean there will be no changes anywhere, or will the Commissions be told to get something new done before the 2015 election?

3. Green Christian

@2 Barney

The effect of blocking the boundary changes will be that the 2015 election will be fought on the same boundaries as 2010. There simply isn’t time to properly do another boundary review before the election.

4. ex-Labour voter

Yes, but incumbency can be a double edged sword.

I once had a terrible Labour MP. The more I knew about him, the less I liked him. He was so bad that I cheered when he lost his seat.

Also, people have problems and can write to their MP about them. However, a well-entrenched candidate can act in a similar way. Problems can be picked up by knocking on doors or by leaflets with complaints forms on them.

And so the problem you describe can be offset.

5. Planeshift

” Again, there is some evidence that Lib Dems benefit from having an incumbent MP”

Will this necessarily be the case now though? As the lib dems can’t benefit from protest votes.

6. Green Christian

@5 Planeshift

The Lib Dems seem to be doing a lot better in local elections in places where they have an MP or where they run the local council than in places where they don’t (in many of these places, either Labour or the Tories are non-existent – making them the only opposition). I think, therefore, they will benefit from positive incumbency results in a lot of seats (though not all of them – Nick Clegg is likely underperform the rest of his party). The incumbency effect will almost certainly see them retain more seats than many people expect them to.

7. Natacha

The implication of this MAY be that Labour gets quite considerably more votes than tyhe Tories but fewer seats.

8. Natacha

However I would reckon this pales into insignificance compared with the Tories having every major newspaper bar two in their side and all major TV and Radio networks (including the BBC), as well as vastly greater funds than all other parties combined.

9. Jack C

“The implication of this MAY be that Labour gets quite considerably more votes than tyhe Tories but fewer seats”.

Not according to the OP.

a) No incumbency effect: Labour 321 seats, 281 for the Tories (on the same share of the vote)

b) Incumbency effect applied (inc Lib Dems): Labour 298 Tory 294

As regards media bias, Labour has had the lion’s share of media support in recent times, and may win back “deserters” prior to 2015. That’s the more positive approach anyway.

10. Jack C

From the OP:
“The whole of the supposed ‘bias’ in the electoral system towards Labour from the current boundaries disappears”

This is a bum note in an otherwise very interesting OP.

“Supposed” bias? Actual and significant according to the previous paragraph. Whilst this bias may be covered by a different one (incumbency) in 2015, imagine what the pro-Labour effect was in 2010.

11. MarkAustin

@1. Albert Spangler

A comment on this. There is considerable evidence for differential voting patterns of Tory and labour voters in safe seats. In general if a seat is safe (either way) some Labourr voters don’t see the need to vote, as the result is, in essence, pre-determined, whilst Tory voters will turn out anyway. This effect does not appear in marginals, where voting patterns of Tory and Labour voters are the same. This effect accounts for about half the Tory-Labour imbalance in votes vs seats.

12. Chaise Guevara

@ 8 Natacha

“However I would reckon this pales into insignificance compared with the Tories having every major newspaper bar two in their side”

At least three tend to oppose the Tories: Guardian, Indy, Mirror. And two of the most influential (Sun and Times) are small-c conservative but vary on who they support in each election.

“and all major TV and Radio networks (including the BBC)”

The BBC is not on their side, and I really don’t think Channel 4 is Tory.


Reactions: Twitter, blogs
  1. Jason Brickley

    Why ‘incumbency advantage’ may decide the next election http://t.co/A8yaexLo

  2. Phil BC

    Great post by @donpaskini on the electoral effects of incumbency: http://t.co/CMC7UwVP

  3. leftlinks

    Liberal Conspiracy – Why ‘incumbency advantage’ may decide the next election http://t.co/a0poEA1Y

  4. Don Paskini

    @MikeSmithsonPB @anthonyjwells would be interested in your thoughts on attempting to measure the 'incumbency effect' http://t.co/BWeFue8q

  5. Noxi

    RT @libcon: Why 'incumbency advantage' may decide the next election http://t.co/TN29ScJN

  6. UK Polling Report

    [...] Paskini has linked to some of the attempts to measure incumbency bonus in a piece for Liberal Conspiracy her…, but the effect of incumbency is probably about 1.2% – 2% of the vote (or double that for [...]

  7. Mark Thompson

    Why ‘incumbency advantage’ may decide the next election http://t.co/ay0rB6qN (via Instapaper)





Sorry, the comment form is closed at this time.