Polls, Swings and Votes


Finally, the worst-kept secret is no more.  We have a General Election on Thursday 6th May.  Politicians have been avidly watching the polls assessing how the public might be reacting to their policy announcements and the cut and thrust of debate.

The first General Elections that really stuck in my memory were those of February and October 1974, with the earlier of these producing a hung Parliament.  There were a number of possible alliances that could be formed with the situation of no overall control providing disproportionate influence to the smaller political parties.  It was a time of economic and political instability.

Sadly the prospect of a hung Parliament is heightened today by the effect of some dubious constituency boundaries.  These lines on the map require the Conservative Party to have a much higher voting lead over Labour to form an equivalent majority government.  This was illustrated in the 1992 General Election when the Conservatives polled a still-record 14 million votes but formed a government with a majority of just 21 seats.

I also remember the manual swingometer device used in previous elections.  This crude pendulum apparatus would show how many seats the two main parties would gain or lose based on different swings in public opinion.

A swing is based upon the change in voting share.  If there is a constituency at the last election where Labour polled 60% of the votes, the Conservatives 30% and the Liberal Democrats 10%, we can calculate the swing produced by a new result.  Where the actual result is Conservative 45% (up 15%), Labour 35% (down 25%) and Liberal Democrat 20% (up 10%), the swing from Labour to Conservative is given by ( 15 - (-25) divided by 2 meaning a swing of 20%.  The political parties can use this data in targeting constituencies for special attention.  Not many campaign helpers would admit to relying on these methods alone.  There is too much risk of an individual being labelled geek of the week.

We can be too mathematical.  Elections are determined by people and not by our attempts to pigeon-hole electors through statistical analysis.  Hopefully our politicians will also switch on some consultation and even someinformed intuition to decide what issues to target and what policies to promote.

Then there are factors that no mathematical effort can pick up.  These include the all-important attributes and experience of the individual candidates that we have to choose from.


Councillor Bob Lanzer, Leader of Crawley Borough Council

7th April 2010