Next month’s elections to the European parliament raise two big
questions: how badly will Labour perform; and will the BNP make a
breakthrough by winning seats?
Last time, 2004, Labour did very
badly. It won just 23 per cent of the vote – its worst in any
Britain-wide election since 1918 – and only 19 of the 75 seats elected
on the British mainland. This time things look bleak: Labour stands
around seven points lower in the polls than it did five years ago; and
six fewer seats are up for grabs. We might be appalled, but should not
be surprised, if Labour’s vote share falls below 20 per cent, and its
contingent at Strasbourg falls to 14-15.
I would not bet too
much against such an outcome; but there are two reasons why things
might not be so bad. The first is that last year’s result in London
demonstrates the risk of predicting the results of elections conducted
by proportional representation. Labour stood lower in the polls last
May than at the time of the London elections in 2004; and, of course,
Ken Livingstone lost to Boris Johnson. But in the elections to London’s
Assembly, Labour’s share of the vote was slightly up. Indeed, Labour
gained one seat.
Could something similar happen in next
month’s European elections? The answer depends not only on Labour’s
share of the vote, but on the size and precise distribution of the
minor party votes. This brings us to the second reason for caution in
making any predictions. In 2004, UKIP won 16 per cent of the vote and
12 seats. They are unlikely to win as many votes or seats this time.
Could their decline help to save Labour from catastrophe?
To
make sense of the prospects, we need to look at the way we elect MEPs –
members of the European parliament. They are elected by the D’Hondt
system of proportional representation, region by region. The key
calculation is the threshold for each region – the vote a party must
obtain to win a seat. The precise figure depends on the exact
distribution of votes as well as the number of seats. In the south
-east, which elects 10 MEPs, a party will need around 7-8 per cent of
the vote to win a seat; in the north-east (three MEPs) the threshold is
more than double that.
There will almost certainly be enough
UKIP and BNP voters to win seats in most regions – if they were one
party instead of two. Last time, remember, UKIP had Robert Kilroy-Silk
and Joan Collins to guarantee them publicity, while negotiations over
the intended EU constitution made Europe unpopular, and Michael
Howard’s Conservative party had little public appeal. UKIP won 16 per
cent of the vote and 12 seats. This time, I would expect UKIP’s vote to
fall sharply, with many voters returning to the Conservative party. The
BNP’s vote is harder to predict. It is possible they will win up to
four seats (London, the north-west, Yorkshire & Humberside and the
West Midlands) – but equally possible that they will fall just short in
each region and win nothing.
For example, consider the West
Midlands. If the UKIP/BNP total is 18 per cent and the vote divides
equally, neither party wins any seats. If the vote divides two-to-one,
then the larger of the two parties wins one seat. In that example, then
an uneven division of the nationalist vote maximises their strength at
Strasbourg. On the other hand, if their combined vote matches the 23
per cent they won last time, then an even split could give them two
seats – while an uneven division keeps them at only one. On such tiny
variations depend the fate of the minor parties and, consequently, the
number of seats available to the big parties.
If we assume that
the Lib Dems, Greens, SNP and Plaid Cymru retain the 17 seats they won
last time, then 52 seats will be up for grabs for Labour, Conservative
and the nationalists. If UKIP and the BNP together divide the far-right
vote fairly evenly, and their combined vote is below 15 per cent
(compared with 21 per cent five years ago) they might win no more than
six seats between them. That would leave 46 to be shared between Tory
and Labour – the same as last time. My rough guess is that these would
divide Conservative 29-30 (up 2-3), Labour 16-17 (down 2-3). It would
be bad, but no disgrace, for Labour to lose two or three seats in an
election where the total across Britain is down six, from 75 to 69.
On the other hand, if the nationalist vote divides more efficiently,
and climbs towards 20 per cent, then UKIP and the BNP might win a
combined total of 12 or more seats. Labour could then well slip to just
14, or even fewer.
Peter Kellner
is President of YouGov