Conor Burns fought the Eastleigh constituency in the general elections of 2001 and 2005. He is a former member of Southampton City Council. He is the Senior Honorary Vice President of Conservative Future.

Let me begin with a health warning for any LibDems tempted to read beyond the first paragraph of this article: Don’t.  It will raise your blood pressure.  I am a supporter of first-past-the-post. I always have been and almost certainly always will be.  I’ve yet to see a model of PR that does not in some way sever the constituency link or deliver almost perpetual power to a small party who act as eternal power broker. That power is corrosive to radical thinking and decisive action.

Our whole parliamentary system is based on the adversarial approach.  The cut and thrust of debate and a clash of beliefs which exposes ill thought out and unsound ideas. A system underpinned by the simple dynamic that if you don’t like the Government in office you can change it at the next general election.

Yet that is not as easy as it once was.  All Conservatives are rightly thrilled to see our Party above the all important 40% for the first time in far too long.  In any decade until the 1990s that would probably have seen us returned to office.  No longer.

The electoral expert Professor John Curtis of Strathclyde University (who I make no apology for citing frequently) now writes of "electoral bias."

He writes:

"To understand what happened we have to look at a separate feature of the electoral system, electoral bias.  By electoral bias we do not mean that one party acquires a higher or lower share of the seats won than it did of votes cast. Rather by electoral bias we mean that one of the two main parties gets a higher proportion of the seats for any given proportion of the votes than the other main party would get if it were to win the same proportion of the votes. In other words any exaggeration of votes into seats from which both main parties can benefit is not bias but simply the exaggerative quality of the electoral system.  Bias refers only to any exaggeration from which one party is able to benefit but not the other." [i]


In 1945 the House of Commons was increased from 615 to 640 members. It was also decided that there should be a routine redistribution in the life of every Parliament.  The Representation of the People Act 1948 provided for a 625 House and almost all constituencies were within 25% of the national quotas. This measure also abolished the remaining double member seats. When the first routine distribution was effected in 1955 (which raised the House to 630) 270 constituencies had their boundaries significantly altered.  This resulted in huge anger among MPs and led to the Redistribution of Seats Act 1958 providing that neutral Boundary Commissioners need only take action every 15 years.


Any ‘first past the post’ electoral system will, on occasion, throw up an ‘unfair’ result.  This is accepted by believers in this system because it outweighs the flaws in alternative systems. So it was in the 1950s. Labour accepted with relative tranquillity its defeat by the Conservatives in 1951 although it had 0.8% more of the national vote.  Ted Heath did use the fact that the Conservatives had gained more votes in 1974 as the basis for his abortive attempts to negotiate with Jeremy Thorpe.  Although he largely then went quiet on the subject he did not tire, even at the end of his life, of reminding people the he had in fact ‘won’ two general elections.

However these anomalies were small in comparison to what could now lie ahead.


In the last four general elections the voting system has delivered significant bias against the Conservative party. Labour gains an advantage to begin with from its seats in the inner cities and industrial areas with declining populations and lower turnouts. Conservatives suffer from doing well in rural and suburban areas.  It is a stark fact that Labour seats have fewer electors than Conservative ones – in 1997 5200 fewer, in 2001 6400 fewer and in 2005 6200 fewer.  (‘J Curtice ‘General Election 2001: repeat or revolution?’)

It is worth looking in some detail at how the situation has got markedly worse for the Conservative Party over recent elections.


In 1992 the ‘bias’ almost cost the Conservative Party its majority.
Again as Professor Curtice notes there would have been no Conservative
overall majority in 1992 if the swing to Labour had been a mere 0.5%
greater – even though this would have meant that Labour would still
have been over 6.5% behind the Conservatives. Three more stark
statistics spring from analysis of the 1992 result. Firstly Labour
would have become the largest party in the House of Commons even at 3%
behind the Conservatives in the popular vote.  Second Labour would have
secured a majority with only a 0.5% lead and thirdly if the votes had
been evenly split, Labour would have had a 38 seat lead over the
Conservatives.  The Conservatives secured an overall majority of only
21 despite having a lead of almost 8% in the popular vote over Labour.


In 1997 the position became even worse for the Conservative Party.
Leaving aside the political arguments about what led to the 1997
humiliation it is fascinating to get behind the figures. If the
Conservatives had the same swing against Labour on the 1997 results
that delivered Labour a majority of 179 the Conservative majority would
have been only 45.  If they had got the same number of votes in 1997
Labour would still have had 79 more seats than the Conservatives. To
match Labour in seat outcomes the Conservatives would have had to have
been 6.7% ahead in votes. Labour could still have secured an overall
majority even if they were 1.5% behind.  To achieve the same overall
majority the Conservatives required a 10% lead.  As Professor Curtice

"Never has the electoral system exhibited such a strong bias in
favour of one of the parties vis-a-viz the other. On this basis the
only post-war elections that would have seen the Conservatives win an
overall majority would have been those of 1983 and 1987." [ii]

also saw the gap in population in the average seats of the two main
parties widen to 6500. If one then looked at the gap in population and
voters actually turning out (turnout in Labour defending seats was
lower than in Conservative defending seats) that 6500 gap widens to
over 9000.


In 2001 for Conservatives and Labour to have the same number of
seats after a general election the Conservatives would have needed an
8.3% lead in votes. For a majority the Conservatives would need to be
11.5% ahead. If each party had got the same number of votes Labour
would still have had a majority of 79 seats and have had 140 more MPs.


Labour won just 36.2% of the vote and a majority of 66. Never before
had any party won an overall majority on such a low share of the vote.
On the four previous occasions when no party won as much as 40% only
once did the system deliver a majority to the largest party.  That was
as far back as 1923.

The Conservative Party only increased its vote share by a miserly
0.5% – less than half the increase it managed in 2001.  Its vote share
remained lower than at any time since 1857.

If Labour and the Conservatives had the same vote share Labour would
still have won 111 more seats. If the Conservatives had the same lead
over Labour as Labour did over the Conservatives, Labour would still
have had 57 more seats. Only a Conservative lead of 6.4% would have
resulted in equality of seats and the Conservatives would have required
a lead of 11.8% before gaining an overall majority.

Population movements between 2001-2005 meant that the electorate
grew by 1000 votes in an average Conservative seat and fell by 500 in
the average Labour one. Turnout was lower in Labour seats than
Conservative ones (58.6% to 65.2%) equating to a difference of 6200
electors, widening to 8500 in terms of votes. [iii]

There are two very obvious examples of bias within the single member
plurality electoral system.  One party may distribute its vote more
effectively than the other.  For example one could win many of its
seats with small majorities and win more seats while the other party
stacks up large majorities in a smaller number of seats.  Under our
system the only votes that count are those responsible for returning a
member to Westminster.  Votes over and above the 2nd place candidate
are ‘wasted’ as are votes cast in seats that are not won. Another bias
arises from seat size.  However as Professor John Curtis notes:

a party will win more seats if its votes are concentrated more in
smaller constituencies than in larger ones.  Less obvious is that there
is more than one way in which a constituency can be ‘smaller’ than
another.  One is indeed that it has fewer registered voters. But
another way it can be smaller is that fewer of its registered voters
turn out to vote." [iv]


Now, all of this may seem to some like nerd heaven for those wearing
the thickest lined political anoraks. Yet it is profoundly worrying and
could have serious implications for our democracy.  Consider what would
happen if Labour were elected to a majority in the House of Commons
having failed to win the popular vote in the whole of the UK – let
alone just in England.

Conservatives must grapple with this problem before it becomes a
crisis. If we leave it too late, and have no constructive proposals to
sort the problem we will abandon the argument to the LibDems and the
electorate may be sucked into the sticky embrace of PR.  Real
reflective Conservatives have over the years sought reform before
issues became crises. We must do the same.

I do not pretend for one moment to have all the answers but it is
clear that the terms of reference for the Boundary Commission must be
changed – or to coin a phrase – modernised.

In the internet age and with trends and statistics at levels of
detailed analysis not dreamt of in past decades here are some

The Commission should take into account house building numbers,
previous growth and trends when deciding on seat composition. I am not
suggesting for one moment that these should be the decisive factors in
deciding on a new boundary but they could inform thinking.

Reviews should occur in every Parliament so that the boundaries are
fresh and not already a decade out to date. For example I attended the
Boundary review in Hampshire which made the recommendations for the new
boundaries.  In my former seat of Eastleigh they looked at 2000 data.
The changed seat will be fought possibly as late as 2010.  In that time
there has been fast amounts of infill development and Greenfield
development.  Its not acceptable that seats are based on such out of
date data.

It may be possible (although I accept very controversial) that the
Commission could look at turn out at previous elections in deciding
composition to tackle Professor Curtis’s point about seats being even
further apart in numbers when voting turnout is taken into account.

And lastly we should reduce further the seats in Scotland and
Wales.  Welsh constituencies will continue to have on average 56000
electors compared to 70,000 for England and 65,000 for Scotland. Most
MPs will freely acknowledge that a lot of their time is spent dealing
with constituents’ issues on education and health matters.  Westminster
MPs from Scotland do not have that responsibility any longer.  In the
age of email, web and mobile phone and the largest budgets ever awarded
to MPs they can surely discharge their more limited role to a larger
number than their English equivalents.

It is an instinct at the core of every Conservative through the
decades to embrace limited reform when the possibility of more radical
action looms.  We should do so again.  A party out of power for a
sustained period hungers to return to office.  It is a good and
laudable ambition.  The first past the post system has served Britain
well over the years and can do so into the future.  I want to see it
survive and sustained.  But I worry profoundly that if we do not seek
to address its imbalances the first steps on the path to PR could soon
be trodden.


[i] Prof Curtis on result in 1992
[ii] Curtis Appendix 2: The Results Analysed p315
The British General Election of 2005 Kavanagh and Butler Appendix 2:
The Results Analysed John Curtice, Stephen Fisher and Michael Steed.
[iv] Curtis 1992 Appendix 2: The Results Analysed p349