Written evidence submitted by Dr Stephen Barber, Reader in Public Policy,

London South Bank University (VUK 12)

1. Voter turnout at UK general elections has fallen since its peak in 1950 when 83.9% of the electorate cast their votes, reaching a nadir in 2001 when just a 59.4% turnout was recorded.  Despite a recovery in 2010, it is noteworthy that no individual constituency managed to equal even the overall average turnout in any of the four general elections 1950-59. Chart 1 below provides the figures for the 17 general elections 1950-2010. Given that the 1992 turnout was the highest since February 1974 at 77.1%, it is reasonable to identify the two decades since this election as being the period when voter engagement has been in decline and it is here that the Committee might concentrate its inquiry more acutely. 

Chart 1: UK General Election Turnout 1950-2010

(House of Commons Research Papers 01/37, 01/54, 05/33 & 10/36.)

2. Viewing this longitudinal data, there is some evidence to suggest that turnout improves at those elections which ‘matter’.  That is the more closely fought elections which produce or threaten to produce a change of government such as February 1974, 1979, 1992, 1997 and 2010, enjoy a rise in voter engagement.  By contrast, the lowest turnout recorded during the period, 2001, was broadly considered a foregone conclusion and, indeed, resulted in the difference of a single net seat between the incumbent and principal opposition.
3. This suspicion by the electorate that their vote does not matter perhaps goes to the heart of the structural weaknesses with our Westminster model and first-past-the-post electoral system.  In any general election, the battleground seats which are likely to change party perhaps amount for no more than 15% of constituencies. Indeed, the Electoral Commission analysed campaign monies spent by political parties in different constituencies in their report Penny for your Vote? (Terry, 2013).  They found that there were as few as 85 marginal seats and safe seats receive much less attention by parties than those which are winnable. At its most stark it meant that voters in the safe Labour seat of Bootle were valued 22 times less than the three-way marginal of Luton South.  In monetary terms this meant £3.07 spent per vote in Luton compared to just 14p in Bootle.  Consequently, some aspect of poor voter engagement must relate to the relative attractiveness of different voters to parties.  For the vast majority of the electorate in most constituencies, there is little incentive for parties to prioritise engagement, especially during a general election campaign.
4. The unequal nature of individual votes is further compounded when one considers the raw headline numbers.  When one divides the votes cast across the country with the number of members returned to parliament, it is possible to illustrate the point. In the 2001 general election which saw turnout fall to an historic low, it took on average just under 26,000 votes to elect a Labour MP but over 50,000 to elect a Conservative and just shy of 93,000 votes to elect a Liberal Democrat.  The position improved between the two bigger parties in 2010 but it can only breed a sense of disenfranchisement amongst many voters.  Here the Conservatives needed 35,000 votes, Labour 33,000 and the Lib Dems, who increased their share of the vote but lost seats, 120,000. Meanwhile, the Greens who secured their first parliamentary seat claimed 265,000 votes across the UK.
5. Electoral reform remains one method of addressing this structural weakness despite the failure to secure agreement in the 2011 referendum (where, incidentally, turnout was 42%).  It might be observed that there was little enthusiasm for the Alternative Vote system even amongst reformers and the campaign became dominated by short-term tactical positioning of the parties rather than representing a discussion about Britain's constitutional future.  The referendum is unlikely to have ended the debate and might even prove to be something of a 'stalking horse' which lays a marker for future reform.  The steady increase in votes for candidates other than Conservative and Labour, which has risen from 4% in 1955 to some 36% in 2010 could provide some impetus here and it would be desirable if reform proposals emerged from Parliament independently of party and from an evidence base. 
6. Notwithstanding the above, it is important not to overstate poor voter engagement purely as a structural problem.  Charts 2 and 3 below detail the turnout in the 10 most marginal and the 10 safest sets at the 2010 general election respectively. The 10 most marginal seats enjoyed an average turnout approximately the same as the overall turnout in the poll: around 65%.  If one might have expected this to be above trend in the most closely fought constituencies, what is more surprising is that average turnout in the 10 safest seats is only very slightly lower. Some of this can be explained by Labour voters staying at home rather than switching allegiance in the marginals and of course the safe seats include the constituencies of the prime minister and leader of the opposition.  But the numbers serve as a reminder that there is not necessarily a direct correlation in constituencies.

Chart 2: 2010 General Election: 10 most marginal seats

Chart 3: 2010 General Election: 10 safest seats

(Adapted from Hough (2013))

7. It is instructive, therefore, to survey the seats with the lowest turnout.  Chart 4 lists the 10 constituencies in 2010 which suffered the weakest voter engagement, averaging under 50% turnout.  Furthermore, the worst performing constituency has since shown further weakness. In a subsequent by-election in 2012, Manchester Central returned Lucy Powell to Parliament on an 18.2% turnout: the lowest since 1945. Thirsk and Malton can perhaps be discounted since the low turnout is probably accounted for by the countermanded vote following the death of the UKIP candidate John Boakes.  Elsewhere, most of these constituencies saw a slight increase in turnout while Leeds Central, which had recorded the previous by-election low of 19.1% in 1999, actually saw some improvement.

Chart 4: Ten Constituencies With Lowest Turnout

8. Upon initial inspection, some generalisations can be made about these seats. What they would appear have in common is that they are all safe, unfailingly returning candidates from the same political party (mostly Labour) and consistent with the arguments put above. These are not constituencies which parties tend to prioritise. In addition to this, however, many of these areas can be said to be economically deprived, suffering above average unemployment and social housing.  As such, it might be possible to argue that low voter engagement and turnout is the combination of structural weaknesses in the electoral system combined with more of a sense of social or economic disengagement as parties concentrate on attracting the median voter in other, marginal, seats.
9. For completeness, a list of the 10 constituencies with the highest turnout is listed in Chart 5 below.  Again, some broad generalisations can be made in comparison to Chart 4.  There are many more closely fought elections here, notably between the Conservatives and the Liberal Democrats making it more worthwhile to vote.  Added to this, these seats tend to be more affluent parts of the country suggesting greater economic and social enfranchisement.

Chart 5: Ten Constituencies With Highest Turnout

10. A further point about political engagement during a time which has seen the parties move closer together ideologically and where there is much diminished social or class identification with parties from the public is that the nature of political discourse breeds contempt. For example, the positions of the government and opposition over deficit reduction are near identical when reduced to raw numbers and yet the debate has been characterised by point-scoring and carefully targeted policy proposals aimed at achieving electoral advantage. There are simply very few incentives for politicians to be entirely honest and those who do not engage in the Dutch auction of promises can expect to be punished at the ballot box. Aided by a partisan media, it is all ultimately off-putting for an already disinterested electorate.  Not entirely a constitutional issue but voter engagement would be improved were the level of debate more intelligent and the public better informed. 

Recommendations

11. Below are some modest recommendations the Committee might consider:

a)     Put electoral reform back on the agenda.  An independent, intelligent, evidenced, public debate about the merits and demerits of alternative systems set in terms of engagement can only be a positive discussion.

b)     Notwithstanding the evidence from Thirsk & Malton where election re-runs produce a lower turnout, one might consider innovations such as a rule which triggers a second election in constituencies where the share of the vote falls below 50%. It might at least incentivise parties to work harder to engage the public.

c)     The Boundary Commission could be given greater authority and backing to adjust the composition of constituencies on the grounds that turnout is low with the aim of increasing electoral competition for those seats.

d)     More could be done to engage the public in the work of Parliament between elections, not least the positive work of Select Committees. The internet and social media surely have a place here.   What would be important here is not simply to inform voters of what is going on but to draw the public in to difficult decisions and debates to better understand the nature of policy choice and the trade-offs they entail.  The Speaker’s Commission and ideas of ‘eParliament’ are relevant here and the Committee’s work could feed into that investigation.

e)     The House of Lords Select Committee on Communications is currently investigating the General Election Debates.  Since this seems set to become an important feature of future elections, there is an opportunity in potentially operationalizing the negotiation process to ensure the interests of the electorate can be represented so as to open the discussion into one of three active stakeholder sets (parties, broadcasters, voters).  In doing so, the wider democratic interests might be represented in the negotiations and the subsequent format of the debates.  Democratic and voter interests might be considered as both explicit and tacit.  That is explicit interests might reflect voters’ concerns and questions as one might ordinarily expect: jobs, public services, crime, etc.  But if the mechanisms are to enhance democracy, tacit interests must also be considered.  That is, voters might be better informed about the complexities of political decisions, and better understand the trade-offs required, as a result of these debates. 

18 February 2014

References

David Hough (2013), ‘Marginal Seats’, Standard Note: SN/SG/06549

Terry, C. (2013), Penny for your vote? Counting the cost of an unfair electoral system, Electoral Commission.