Written evidence submitted by Professor Andrew Russell,

University of Manchester (VUK 25)

 

Executive Summary

 

 

 

 

 

 

 

 

Voter Engagement: Introduction

  1. Voter engagement is usually taken to mean turnout in elections, although a wider discussion of citizen activities would be helpful.

 

  1. An individual’s decision not to vote in a particular election may be the result of any of a number of contributing factors. These might include legal or eligibility restrictions, inconvenience, awareness issues and lack of political knowledge, antipathy or alienation and a recognition that it might be rational not to vote since the chances of affecting the outcome are often less than the costs of voting.  I have deliberately left the term ‘apathy’ out of this list of contributory factors.

 

  1. There have been many attempts to reduce the inconvenience of voting in recent years. With the exception of the roll-out of postal voting, few have had much impact on turnout. Political knowledge issues require long-term solutions but have the aim of initiatives such as the addition of Citizenship to the national curriculum. Rational non-voters might be countered by those who see voting as a civic duty but it is highly unlikely that the proportion of rational non-voters has significantly increased in recent years.

 

  1. The focus then, for the development of democracy in the UK has to be on those electors who do not vote out of lack of awareness, or as a result of voter antipathy or alienation.

 

  1. Turnout means the proportion of the registered electorate that actually cast a vote (including spoiled votes) but it is important to make a point about comparability here. Firstly turnout is a comparative measure - the proportion of eligible voters needs to increase in order for turnout to improve. Secondly turnout is normally restricted to the registered electorate and thus its utility is linked to the usefulness of the registered list of voters. In the UK that there are hard to reach groups that might be missing for the electoral register. Young males in urban (particularly inner-city) areas are generally thought to be among the most likely of all groups of voters to be missing for the electoral register; and yet we know that those young urban males on the electoral register are the most likely of all social groups not to vote anyway. It might take considerable effort to improve the accuracy of the electoral list and increase the number of young males in inner cities on the register - and this would worthwhile – but the overall effect on turnout might be negative (as the exercise would improve the accuracy of the potential voters without, in itself, improving the propensity of that group to participate in elections). Therefore it is probably better to use the VAP statistic which records turnout as a proportion of the Voting Age Population (eligible rather than registered voters).

 

  1. By using data from International IDEA turnout in the 2010 general election was 61.1% of VAP rather the 65.8% (and 57.6% in 2001 rather than 59.4%). As a useful comparator turnout in the 2012 US Presidential election would be 68% by the standard measure but 54.6% of the VAP.

 

Turnout does not equal legitimacy

  1. The link is often drawn between low turnout and a perceived lack of a mandate to govern. This argument is frequently used by the academics, media commentators and politicians alike. In the aftermath of the 2001 general election, Professor Martin Harrop claimed that as a result of the historically low turnout “elections may have suffered a permanent loss of authority”. By 2006 the Power Inquiry reported that “the fundamental feature of any representative democracy is that a government derives its legitimacy and mandate to govern by winning the consent of its community through periodic elections. An ongoing and serious decline in turnout could mean that British governments no longer have a mandate to govern” (p.33) In February 2014, the Mayor of London writing in the Daily Telegraph about the London Underground strike called it a “continuing scandal that this strike was triggered by 30 per cent of those RMT members balloted. In other words there were fully 70 per cent of members who were not in favour of this action…. We need a ballot threshold – so that at least 50 per cent of the relevant workforce has to take the trouble to vote, or else the ballot is void.” (Daily Telegraph, February 2, 2014) 

 

  1. Leaving aside the consequences of applying such a threshold to referendums, local, European, Mayoral, PCC elections, by-elections or even constituency based results in general elections, it is clear that the link between a mandate to govern and turnout is problematic.

 

  1. Should we conclude that the re-election of US President GW Bush in 2004 where turnout was 62.1% of the VAP gave him more authority than the election of President Obama in 2008 when turnout was 57.5%? Did the wafer-thin victory of Kennedy in 1960 (where the contest attracted a turnout of 63.1% of the VAP) make him a more legitimate President than Reagan who achieved landslide victories in presidential contests from a turnout of 59% of the VAP in 1980 and 1984? In the UK context, anyone making a simple connection between turnout and political legitimacy would have to conclude that the most legitimate part of the UK according to highest turnout has consistently been Northern Ireland.

 

  1. Democratic theorists instead argue that legitimacy is gained not from the proportion of electors who voted in a particular election but from the actions of an office-holder, or from the general acceptance of the outcome by society as a whole including the losing candidate or party.

 

  1. Nevertheless there is an everyday logic behind this argument that does seem to have explanatory force. Most of us tend to believe that high turnout is good; low turnout is bad, and it might be possible to draw a participation threshold beneath which we could claim that the electoral system does not hold the trust of the public. The real problem is that we have no idea where that threshold should be or how close to it we are.

 

Falling levels of engagement:

  1. Strictly speaking, turnout in UK general elections is not falling. Indeed turnout has risen in the last two contests. However there has been a sharp decline in voter turnout in the last generation. Turnout in 2001 was the lowest since men and women were given the right to vote on equal terms, turnout in 2005 the second lowest and 2010 the third lowest (Figure 1). Furthermore the slight increase in turnout since 2001 has probably stemmed directly from the considerable efforts to make the act of voting easier for electors. In particular, the development of the practice of postal voting has improved turnout but public confidence in the integrity of postal voting has not been boosted by rare but high profile incidences of postal voting fraud, misrepresentation and personation.

 

Figure 1 – Percentage Turnout at General Elections as proportion of registered electorate 1929-2010

 

 

 

 

 

 

 

 

  1. There is clear evidence of turnout decline in the UK – and not just at parliamentary level. In the 1999 European Parliament elections turnout among UK voters was 23.1% - fewer votes were cast for the leading political party at that election than for the winning candidate in the Grand Final of the TV programme “Stars in their Eyes” the following weekend (although voters could only vote legally once in the EP election).  Turnout has fallen in elections to the Scottish Parliament and National Assembly for Wales since the inaugural contests of 1999 and has continued to be significantly lower than that in 1997 referendums that set-up the institutions in the first place.

 

Percentage Turnout at Scottish and Welsh Devolved Institutions 1997-2011

 

Turnout: Scottish Parliament

Turnout: National Assembly for Wales

1997 Referendum

60.2%

50.1%

1999 Election

58.1%

46.4%

2003 Election

49.4%

38.2%

2007 Election

51.7%

43.5%

2011 Election

50.4%

42.2%

 

  1. Declining turnout is not a distinctly British phenomenon; indeed it is a well-worn pattern in established democracies. Across much of the democratic world, voters have seemed less inclined to participate in elections than their counterparts in previous generations.

 

  1. Canadian turnout has fallen from 68% of the VAP in the parliamentary elections of 1988 to 54% in 2011, German turnout has fallen from over three-quarters of the VAP in the 1998 to barely two-thirds in 2013, voter turnout in Spain has fallen from 83% of the VAP in 1982 to 63% in 2011, from 85% in the Dutch parliamentary elections of 1981 to 71% in 2011, while French voter engagement in parliamentary elections has fallen from 70% in 1986 to 46% in 2012. Furthermore relative decline in electoral engagement is evident even in some countries that started off with fairly poor turnout; in Switzerland turnout in parliamentary elections has fallen from over 60% in the 1950s to around 40% of VAP this century.

 

  1. The image of the first parliamentary elections in the new South Africa in 1994 was potent, as voters queued for miles to cast their ballot for the first time; turnout was very high (85.5% of the VAP). In fact it took the new South Africa precisely one election to ‘normalise’; finding the level of voter disengagement familiar to most established democracies. The second parliamentary elections (1999) saw turnout fall to 63.9%, the third (2004) 56.8% and the fourth (2009) 56.6%.

 

  1. Even in those few established democracies where turnout has held up well – Scandinavian countries Sweden and Denmark still enjoy elections that draw participation of over four-fifths of the VAP -  the regular pattern of who participates – older and more affluent voters and who abstains – younger and more marginalised electors is upheld.

 

 

Constituency turnout

  1. However it is measured there are certain drivers to electoral engagement and disengagement which are important to remind us that some variation in turnout is inevitable – and even nations with compulsory voting experience variation in turnout according to social structure.

 

  1. At the aggregate level by combining ecological data collected from the census at parliamentary constituency level we see places in Britain that could be characterised by large proportions of people employed in professional or managerial occupations, of home owner occupiers, of those employed in agriculture and of those aged 75 and over are all associated with relatively high rates of voter turnout in all elections. Similarly constituencies with high rates of households with no car, and large rates of persons per hectare (each could be a measure of inner-city living or relative poverty or both) with younger populations, manual workers and private tenants and some minority ethnic communities are usually associated with lower rates of turnout.

 

  1. At the individual level resources (money, education, time) can impel or discourage participation in the electoral process with the more a citizen has the more likely it is that they will vote. Furthermore those voters with relatively high senses of civic duty, partisan identity or a willingness to simply have a say are more likely to vote than their counterparts with lower levels of these characteristics. Interestingly when we surveyed voters (rather than abstainers) in 2001 almost all groups claimed to have voted through a sense of civic duty – the exception being first time voters who when they did vote tended to emphasise the inclusive nature of being seen as part of society. Survey research has shown that partisanship and the sense of civic duty are less widely shared by UK citizens than in the recent past.

 

  1. Electoral context is also important. The closeness and significance of an election can affect turnout rates (so greater numbers voted in the ‘too close to call’ election of 1992 than in the ‘foregone conclusion’ of 1997), as can the ideological distance between political parties (when it might genuinely matter more which party wins, and when supporters of parties that are bound to lose still want to signal their partisanship). Finally, and most crucially in the British case, the marginality of the constituency has a significant impact on rates of turnout. Indeed some very safe constituencies like Liverpool Riverside and Manchester Central seem to have adopted low participation as part of their public profile.

 

  1. In 2001, 96 of the 100 lowest turnout seats were held by Labour, in 2010 (when there were far fewer Labour seats) 88 of the 100 low turnout constituencies were won by Labour. These seats usually shared other characteristics, most were highly urban, relatively impoverished, and in electoral terms, fairly to ultra-safe seats!

 

Efficient campaigning depresses turnout

  1. One of the most significant factors that might be driving low turnout is the development of political campaigning. It is a mistake to think that political parties have a vested interest in increasing turnout. Rather they have a particular interest in differential turnout – that is they want their voters to come out and their opponents to stay at home.

 

  1. Furthermore, party strategists are not concerned with piling up ‘wasted’ votes in safe seats; preferring to win constituencies with efficient margins so that precious campaign resources might be distributed to other winnable seats.

 

  1. Sophisticated campaign technology now means that parties pride themselves on knowing as much about their base and fringe support as supermarkets do about the commercial preferences of their customers.

 

  1. The focal point for most elections then is not the typical – or median - voter but those voters that can determine the outcome of the contest. With relatively few marginal seats parties will – naturally – focus their attention on target voters in target seats rather than core support. The sheer efficiency of modern campaigns maximises the effects of swing voter strategies at the expense of one of the historic features of electoral contests – they ability to make the general public feel involved with a national event.

 

  1. The focus on the relatively small number of voters that determine the outcome of elections has undermined the socialisation effect that the relatively mundane act of voting can have. The further society gets from the communal identity encouraged by the 1940s and 1950s, the harder it becomes to make turnout appeals through a sense of civic duty. This is exacerbated if electors see insufficient evidence that their vote is valued.

 

 

  1. Large hustings and national rail tours by party leaders are a thing of the past (and might be cynically received if they were revived) but voters are now much less likely to receive a personal visit by an election candidate during a campaign than previous generations. The practice of ‘knocking up’ is still alive and well but it belongs in an armoury of precision targeting by party strategists rather than the blanket bombing that old inefficient campaigns encouraged. Yet one of the advantages of the old inefficient ways may have been to encourage large swathes of people to feel that the election had something to do with their lives.

 

Non-voting activities

  1. Of course it is not just participation in elections that should concern us. We can point to the wide range of activities that citizens still claim to do. The Power Inquiry (2006) noted the extent of voluntary activities in modern Britain. It is instructive however that the 2010 contest with its uncertain outcome and intensive campaign (the first Leaders’ debate was seen or heard by 61% of all voters according to the British Election Study) failed to radically improve turnout rates.

 

  1. In 1983 half of British workforce was unionised, by 2001 unionisation had fallen to one-third (PM’s Strategy Unit, 2002); in 1972 28% of men belonged to Working Mens or Social clubs by 1999 it was down to 18% (Li et al., 2001); in 1972 there were 442,000 members of the National Federation of Women’s Institutes by 2002 there were only 240,000 (Haezewindt, 2004).

 

  1. British society now seems sluggish to respond to stimuli about electoral engagement and the stimuli reach fewer people.  Formal membership of political parties is in serious decline. In the 1950s there were more than £3m members of the Conservative party alone, now combined membership of the three main British political parties is under 400,000,. It does not matter if many were drawn into membership to play sport, attend functions or meet potential partners, chances are they as party members they were likely to participate in politics and elections. As late as 1983 nearly 4% of the electorate was a member of one of the main political parties, by 2010 this had been reduced to 1%. Formal politics has become a minority sport and one of the inevitable consequences is that engagement in democracy has declined too.

 

Conclusion: Ever-shrinking circle of concern

  1. All evidence suggests that low turnout begets lower turnout. As evident feelings of civic duty recede, as formal membership of political parties continues to decline, the number of cues that might have been expected to automatically deliver voters to the ballot box has diminished too. Voting is habit-forming but so is abstention. As electors realise that the sky does not fall in if they fail to vote they are more likely to not vote in future elections. As we all become accustomed to low participation rates in elections, the lack of mobilisation factors now available to candidates and parties has created an ongoing crisis of turnout.

 

  1. In the end it is tempting to subscribe the UK electorate to Nina Eliasoph’s telling idiom about US society, we are witnessing an “ever-shrinking circle of concern”. Interest in politics is not normally distributed. It is possible for those of us who are interested in politics to be better informed than any previous generation about the world of parliament and politics. We can watch the BBC Parliament channel and round the clock coverage on rolling news networks, see how our elected representatives voted on a collection of themes via www.theyworkforyou.com and access as much information about politics wide and narrow as possible. However, we can also avoid politics by watching the movie channels, the music stations, the sports networks, by not reading newspapers or by not reading the ‘politics’ sections of the newspapers’ websites. Facing up to the challenge of the ever-shrinking circle of concern is at the heart of the battle for engaging voters in the UK.

23 February 2014