Mid-term precedents: can the Conservatives win?

This page first posted 6 September 2013, revised 1 January 2015

The next general election is less than six months away. Current opinion polls (as of December 2014) give Labour a lead of 3% over the Conservatives, which is enough to make them the largest party in the House of Commons. But opinion polls can change in the run-up to the general election. Generally governments tend to gain popularity in the two years before an election, as shown by the following graphic:

Runups to General Elections

Poll Sources: Ipsos-MORI (1979-2001); Electoral Calculus (2001-date).

Each election since 1983 is represented by a two-segment line which shows the state of the major parties in the run-up to the election. The line shows where their poll support was at (a) two years before the election (start of the line), (b) one year before the election (shown by a small dot), and (c) the election result itself (marked with a labelled dot). The average opinion poll support for 2013 and 2014 (in each case measured from January to December) is marked with a blue and a gray dot joined by a one-segment line.

In every election since 1983, the Labour party has lost ground over the two years before the election. Generally, the Conservatives have gained ground over the same periods, but this has not been true at the two most recent elections. Some of these elections may have lessons for today, in terms of judging whether the Conservatives will recover their lost ground or whether Labour will cement their lead.

The election of 1983 was particularly notable with a swing of 10% to the Conservatives, converting a Labour lead of 7% in 1981 into a Conservative lead of 15% on polling day. During that period the Conservatives benefitted from three major events, namely a recovering economy, the Falklands conflict (April-June 1982), and the split of the SDP from Labour (March 1981).

Intriguingly, it is harder to explain the 6% swing to the Conservatives at the 1987 election. The swing was heavily concentrated in the short period from January to April in 1987 itself. This was before both the election campaign and Nigel Lawson's tax cutting budget on 29 April. The swing was also clearly after the year-long miners' strike which finished in March 1985. During the January to April period, there was some good economic news as unemployment was clearly coming down from its historic highs and British Airways was successfully privatised. But it remains hard to identify particular drivers for the noticeable change in public support for the major parties over that brief period.

Also worthy of comment is John Major's performance in 1992 after he had taken over from Margaret Thatcher as Conservative Prime Minister in November 1990. He (and his soap box) also managed to achieve a swing of over 10% at the subsequent election. This reversed Labour's large lead of 14% in 1990 into a Conservative poll victory by 8%. Much of this change was achieved by 1991, so was probably driven by the act of replacing Margaret Thatcher, whose popularity had declined by then, and his replacement of the widely-disliked community charge ("poll tax"). The Conservative achievement is perhaps not as impressive as it might appear, since about half the swing was due to pollster error. In 1992, the pollsters collectively had an error equivalent to a 5% swing (see our Track Record for 1992 for more details). So the net "real" swing from 1990 to 1992 was only about 5%.

The three victories by Labour under Tony Blair in 1997, 2001 and 2005 showed less dramatic movement during the two year run-up to polling day. Under Tony Blair, the Labour party enjoyed substantial poll leads which were successfully translated into real ballot-box votes. In both 1997 and 2001 there was some slippage in support from Labour to both the Conservatives and the Liberal Democrats, with a swing of about 8% each time. The election of 2005 was remarkable, if that's the right word, for having almost no observable change in poll support during the run-up period.

Gordon Brown's Labour government faced challenging economic and political cicumstances in the run up to the 2010 election, but managed to keep Labour's support approximately stable. The main movement in that period was due to Nick Clegg's Liberal Democrats who enjoyed a 6% swing to them and away from the Conservatives. This time, the Conservatives lost rather than gained support during the two years before the election. Some Conservative commentators blame this loss of support on an insufficiently right-wing platform, but it seems more likely that 2008 was a temporary high-point for the Conservatives since it co-incided with the worst of the economic crisis and a political low-point for Gordon Brown and the Labour government. The Conservatives then lost some ground as the economy moved away from the depths of the crisis, but with the Liberal Democrats benefitting from that more than Labour.

Taking an average of the run-up period changes over the four elections which had an incumbent Conservative government, the average change in support is: Con +8%, Lab -10%, Lib +1%, which is a swing of 9% to the Conservatives. Applying that change to the Summer 2013 poll figures would give support figures of Con 37.7, Lab 26.9, Lib 11.5, which is enough for a Conservative majority of over 40 seats.

However, if we use the average 6% swing taken over all seven elections, then the projected support figures would be Con 33.9, Lab 29.2, Lib 12.4, which gives a hung parliament with the two big parties approximately neck and neck.

And even that may be over-optimistic for the Conservatives. The key factors which can be associated with Conservative gains in election run-ups are:

But only one of these looks likely to happen in the next two years. That is the economic recovery, which is probably going to be visible, if not overwhelming. The Conservatives' misfortune is to inherit the economy after a major financial crisis, after which recovery is typically weak and slow. But they might expect some electoral dividends from even a weak recovery between now and 2015.

However, the new split in British politics is not on the left as it was with Labour and SDP during the 1980s. Rather it is now on the right, between the Conservatives and UKIP. Although UKIP are unlikely to win many seats, they take votes away from the Conservatives and make it harder from them to win a general election. Even if UKIP never break through as a long-term challenger to the Conservatives, the European Elections in 2014 will give them a potential springboard into the subsequent general election.

The third factor is probably not applicable in this parliament since the the current Prime Minister is not particularly unpopular and is 7% more popular than his party according to a recent poll (YouGov 30-31 Aug 2013). In the other direction Ed Miliband could be more vulnerable in this regard, since he is less popular than his own party, despite its success in the polls under his leadership.

Finally, successful foreign conflicts have been relatively rare in recent years and moreover the public mood is not currently well disposed to military campaigns. In retrospect the Falklands conflict of 1982 appears as a notable but unique event, which is unlikely to be replicated in the next two years.

Overall this analysis shows that it is certainly possible for the Conservatives to recover enough ground to get another hung parliament, but it is still very difficult for them to win an absolute majority in the House of Commons. They would need some events like a good economic recovery and/or the collapse of UKIP to achieve the swing of 7% which they need for a majority.

Appendix : Data

YearCon %Lab %Lib %Con-Lab %
198037.747.113.0-9.4
198130.937.529.6-6.6
198241.331.125.810.2
9 Jun 198342.427.625.414.8
198440.937.419.93.6
198535.736.126.3-0.5
198635.238.724.0-3.5
11 Jun 198743.331.523.111.8
198846.138.48.17.7
198940.443.06.5-2.6
199035.449.18.8-13.7
199140.341.714.8-1.4
9 Apr 199242.835.218.37.6
199330.144.721.5-14.6
199425.551.319.0-25.8
199527.155.713.4-28.6
199628.853.913.0-25.0
1 May 199730.743.216.8-12.5
199827.453.614.0-26.2
199928.052.413.7-24.4
200033.145.615.4-12.5
7 Jun 200131.740.718.3-9.0
200229.844.819.1-15.0
200333.337.022.4-3.7
200433.035.421.7-2.5
5 May 200532.335.222.1-2.9
200636.833.918.72.9
200737.035.316.61.8
200842.329.917.012.4
200940.527.118.413.5
6 May 201036.129.023.07.1
201135.640.111.3-4.5
201233.341.110.7-7.8
201330.338.210.2-7.9
201431.835.08.7-3.2

Poll Sources: Ipsos-MORI (1979-2001); Electoral Calculus (2001-date).

The table shows the opinion poll averages for each non-election year, and the general election result for each election year.

Other commentary on this chart

Two political commentators have also written their own interesting analyses of this chart. They have different points of view, but each is worth reading.


Return to analysis articles index, or home page.