Issues of electoral turnout in national elections and other votes are often taken as indicators of the legitimacy or illegitimacy of the political outcomes which stem from such votes. While it is assumed that high levels of turnout produce ‘legitimate’ results, low levels of turnout are assumed to be more problematic and enable actors to question the democratic legitimacy of the election. Lutz and Marsh (Reference Lutz and Marsh2007: 539) argue that ‘the legitimacy of democracy in general and the outcomes of elections in particular are undermined when many citizens do not vote’. A range of recent literature has identified declining citizen participation in elections in a number of advanced liberal democracies as evidence of a ‘democratic deficit’ and a challenge to the legitimacy of elected institutions (see, for instance, Blondel et al. Reference Blondel, Sinnot and Svenson1998). Such theories assume that by holding free and fair elections the legitimacy of state institutions will be increased. For example, Scully et al. (Reference Scully, Jones and Trystan2004) argue that low turnout in the first Welsh Assembly elections in 2001 led commentators to question the legitimacy of the developed body. In order to overcome such deficits some authors have proposed compulsory voting (Hill Reference Hill2004; Lijphart Reference Lijphart1997), though this is opposed by Saunders (Reference Saunders2012), who argues that elections should be considered ‘democratic’ based upon the opportunities of citizens to vote and independent of the number of votes actually cast.
The purpose of this article is to question the links made in the existing literature between electoral turnout and legitimacy. We argue that elections with lower turnouts can come to be seen as more legitimate than elections with higher turnouts due to sociopolitical factors exogenous to the electoral process itself. Our argument is that the concept of legitimization, conceived as an empirical process through which the authority of an institution is discursively constructed and conferred, is more important in studying electoral outcomes than ‘actual legitimacy’, defined as an evaluative measure of the level of trust and confidence an institution should command, in day-to-day struggles over the meaning of electoral results and their subsequent acceptance or rejection. We argue that commentators interested in electoral mechanisms should focus on the process by which authority is conferred – legitimization – rather than election results (including turnout) as a validating standard for determining legitimacy. Put simply, ‘actual legitimacy’ as an abstract, absolute (binary) concept is less important than legitimization as a process, when conceptualizing the narratives surrounding elections. If we are interested in establishing the perceived democratic legitimacy of elections then we ought to focus on wider sociopolitical processes of legitimization and delegitimization that include discursive contests about the meaning of electoral statistics within a wider structural context of uneven power relations. Through comparing two UK ballots, we define the term ‘election(s)’ in a broad sense as a democratic ballot that legally binds result(s) or decisions upon an institution or institutional arrangement. The police and crime commissioner (PCC) elections chose candidates to serve as police and crime commissioners, affecting various police forces. Equally, trade union ballots affected institutions such as the trade unions themselves and schools as they legally facilitated disruption to the classroom.
Our article will proceed in four sections. First, we focus on the conceptual distinction between ‘actual legitimacy’ and the process of legitimization and argue that the concept of legitimization better captures the dynamic process through which legitimacy is conferred on election results. Then, using two case studies from the UK – the police and crime commissioner elections in November 2012 and the trade union strike ballots of two teaching unions in 2013 – we show how processes of legitimization and delegitimization can be more important in determining the evaluation and acceptance of results than pure turnout statistics. In the police and crime commissioner elections, low turnout was not a barrier to the institutionalization of the police and crime commissioner positions, whereas in the trade union ballot the decision to strike was severely contested, despite turnout being double that in the police and crime commissioner elections. In both cases, how state actors interpreted the meaning of the turnout figures was crucial in determining whether the institutional decisions and structures imposed by the elections were initially contested.
In exploring questions of turnout we view turnout as part of the election results (as opposed to processes) and question recent interpretations of voter (non)turnout. Here the turnout can be calculated either as a percentage of all those who are on the electoral register or as a percentage of the voting age or voting eligible population (Hill and Louth Reference Hill and Louth2004; McDonald and Popkin Reference McDonald and Popkin2001). Irrespective of how it is calculated, turnout can only be calculated after all the ballots have been counted (and the total divided by the appropriate population). This is distinguishable from the process of elections – which occurs prior to the votes being counted. Here we distinguish between the process of turning out to vote (on an individual level) and the turnout (on an aggregate level) of the election and argue that turnout figures should be viewed as a result of the ballot in the same way that the number of votes cast for a particular party is viewed.
Legitimacy and Legitimization
Elections can enable large numbers of people within society to affect political outcomes, creating the sense of legitimate processes by which policies are derived. Through notions of mass participation, ballots – certainly in a democratic sense – have often been closely linked with legitimacy. This leads to elections, turnout and legitimacy being presented in a simplistic relationship whereby the higher the number of votes cast, the greater legitimacy is conferred to the election. Lijphart (Reference Lijphart1997) argues that voting in America should be made compulsory to overcome its ‘unequal participation problem’. In such a discourse there exists an assumption that ‘the legitimacy of democracy in general and the outcomes of elections in particular are undermined when many citizens do not vote’ (Lutz and Marsh Reference Lutz and Marsh2007: 539). Franklin (Reference Franklin1999: 205) asserts ‘low electoral turnout is often considered to be bad for democracy, whether inherently or because it calls legitimacy into question or because low turnout implies lack of representation of certain groups and inegalitarian policies’. This link to electoral turnout and legitimacy is reinforced in public discourse. When advocating voter turnout, commentators argue, for example, that ‘we need to ask ourselves why high turnout is so significant. The answer is simple: legitimacy’ (Koas Reference Koas2012). In the context of low turnout in European elections, policy analysts argue that ‘there is great concern that the legitimacy of the EU is at stake should there be a further slide in voter turnout’ (McDougal and Mody Reference McDougal and Mody2014), while high turnout (88.6 per cent) in ballots such as the 2014 Scottish independence referendum are praised as ‘clearly’ having ‘a very high level of legitimacy’ (Lake Reference Lake2014). Such assumptions explicitly link turnout to legitimacy; a high turnout can legitimize an election (or an election result) whilst a low turnout can be used to question the validity of the election results (Brennan and Lomasky Reference Brennan and Lomasky1997; Franklin Reference Franklin2004).
While these arguments appear straightforward, the relationship between the legitimacy of a ballot and the legitimacy of its outcome (the results) are two distinct entities, as there is no cut and dried point at which legitimacy can be determined. How high does turnout have to be for an election result to be deemed legitimate? Is there a point at which turnout could be so low that an election cannot, under any circumstances, be deemed legitimate? Our concern here is not to attempt to resolve these issues (for such discussions see Saunders Reference Saunders2012) but to point out that ambiguities exist and that the practical interpretation of results can be critical in determining whether an election is seen as legitimate ‘in practice’. If turnout is high, it may be assumed that a greater number of citizens have participated in the process, hence lending the result greater authority. Alternatively, if turnout is low, then it is often argued that the legitimacy of the election is questionable since citizens do not ‘bother’ to vote (Kelley and McAllister Reference Kelley and McAllister1984: 463; Kelly Reference Kelly2014: 204) or recognize the election as being a worthwhile or authoritative exercise. As we shall show presently, however, ballots with higher turnout can come to be viewed by important actors as less legitimate than those with lower turnout, even when differences are relatively marginal and ambiguous. This can have worrying, even anti-democratic implications when combined with the relatively authoritative status of the ‘will of the people’ given to political agendas victorious in ballots deemed as legitimate. ‘Low’ turnout – however defined – can be used by those in socioeconomically powerful positions as a rhetorical trope to bring into question the legitimacy of a ballot where they disagree with the outcome (and policies hence authorized), even when that ballot may not have had exceptionally, or even relatively low, turnout. Examining ‘actual legitimacy’ as an abstracted value – for example, procedural integrity and other ‘minimal’ conditions proposed by Schumpeter (Reference Schumpeter1947: 269), or the quality of deliberation in electoral processes, as Dryzek (Reference Dryzek2001: 657) suggests – may, in these cases at least, distract us from the power-plays that serve to determine which results are seen concretely as ‘legitimate’ or not. A more promising venture, for those interested in the legitimacy of these kind of ballots at least, may be to deconstruct and interrogate how actors seek to legitimize certain ballots even where the turnout is not exceptionally low, and conversely how those ballots with very low turnout are defended by those whose interests it serves to retain their legitimacy.
In this article we propose to analyse the process by which elections are interpreted as being legitimate or not. Such interpretations take place in a much wider social, economic and political context in which certain actors may have a particular interest in an election being perceived as ‘legitimate’ or not. When actors such as high-ranking politicians or business leaders are able to dominate media outlets in the wake of an election and set the agenda of public discourse, they acquire a disproportionately high ability to ‘define’ the election with regard to its legitimacy. As such, the process through which elections gain legitimacy is highly contingent and embedded in a wider context of unequal power relations (Beetham Reference Beetham1991). The purpose of this article is to uncover how this process of interpretation affects the overall perceived legitimacy of the elections. It argues for a greater focus on the extent to which political actors with greater socioeconomic power and resources are able to define elections as legitimate or not, demonstrating that this process is relatively divorced from a ‘rational’ consideration of the election results and turnout figures themselves.
In order to explore the processes through which legitimacy is interpreted by political actors, this article adopts the processes of legitimization and delegitimization over a binary concept of ‘actual legitimacy’. Here we do not reject Beetham’s broader point that legitimacy has a normative component; this article, rather than seeing legitimacy as being either present or absent, suggests that the process of legitimization allows for a more meaningful comparison between perceptions and understandings of elections. Here we do not contend that the concept of legitimization can be used interchangeably with legitimacy, nor can delegitimization with illegitimacy. Both refer to perceptions or narrations of ballots and are used subjectively to promote certain agendas.
Such concepts posit legitimacy as resulting from a social process in which it is constantly contested and fought over (Johnson et al. Reference Johnson, Dowd and Ridgeway2006). They find their roots in Max Weber’s (Reference Weber1956: 23) social theory, which proposed a sociological conception of legitimacy, defining it as ‘Legitimitaetsglaube’ – the belief of actors in the legitimacy of rules that govern them. Here, the concept of legitimacy is transformed from a normative standard into an empirically objective variable. If the relevant population (the electorate, general public, or more likely key opinion-formers such as elite politicians and the media) views a ballot as being legitimate, then its legitimacy has been achieved, and where it is not viewed as legitimate, legitimacy has not been achieved. This view, although enticingly simple, is also highly problematic; as Beetham argues, Weber’s conception:
misconceives the relationship between legitimacy and the beliefs that provide the justificatory basis for rules of power. It fails to recognise that, although prudential and normative reasons for obedience are indeed distinct, nevertheless, people’s interests can be harnessed to legitimacy through actions expressive of consent. Finally, it leaves the social scientist helpless in the critical task of analysing an erosion of legitimacy in power relations, by proposing a report on people’s ‘belief in legitimacy’ rather than identifying a developing discrepancy between rules of power and the norms that provide their justification. (Beetham Reference Beetham1991: 23)
To avoid these criticisms Beetham (Reference Beetham1991: 19–20) proposes an alternative approach linking legitimacy critically to power, which ‘can be non-legitimate in very different ways . . . signalled by three different terms: illegitimacy, legitimacy deficit and delegitimisation’. Illegitimacy refers to power which is ‘acquired in contravention of the rules’, a legitimacy deficit occurs when ‘the rules may find only weak justification in terms of accepted beliefs about the proper source of authority or ends of the government’ (Beetham Reference Beetham1991: 205). The latter concept interests us here. For Beetham (Reference Beetham1991: 205), delegitimization refers to the ‘withdrawal of consent’ for specific rules by political actors, ‘those whose consent is necessary to the confirmation of legitimacy’. Legitimacy is not an ‘all or nothing affair’, but is relative: ‘power can be said to be legitimate to the extent that . . . legitimacy may be eroded, contested or incomplete; and judgements about it are usually judgements of degree, rather than all or nothing’ (Beetham Reference Beetham1991: 19–20).
We concur with Beetham that analysing processes of legitimacy relates to examining this expression or withdrawal of ‘consent’. Such expression or withdrawal can be done either explicitly (actively rejecting the authoritative nature of a ballot or instigating legal procedures to overturn or challenge it) or, more likely, implicitly by expressing how the otherwise binding results of the ballot ought to be (or ought not to be) taken seriously, or by expressing the view that those who might wish to undermine the authority of a ballot would be (or would not be) justified in doing so. We propose to analyse the processes of legitimization and (de)legitimization (that is, attempts to withdraw consent or encourage others to withdraw their consent) as concepts here to highlight the highly dynamic and gradational nature of these processes of expression or withdrawal of consent, which we seek to capture below.
Following the above, our key empirical question is ‘how do (some) elections come to be seen as legitimate, or their authority consented to?’ That is, ‘how is the process of legitimization determined?’ Below we use two case studies to examine the processes of legitimization in the UK, the police and crime commissioner elections (November 2012) and the teaching trade union ballots for industrial action (spring 2013). Our analysis takes as its starting point that these elections could conceivably be deemed ‘legitimate’ in an absolute sense – that is, they were conducted under free and fair processes (this assumption is made on the basis of no contradictory evidence) and therefore any differences in the expressions or withdrawals of consent occur not because of self-evident failings of (abstractly defined) legitimacy, but more likely due to political interests in determining the authority of the results.
Case Studies
Throughout this article we shall examine two UK case studies: the police and crime commissioner elections and the ballots for strike action by the two teaching unions, the National Union of Teachers (NUT) and the National Association of Schoolmasters Union of Women Teachers (NASUWT), known collectively as the Teachers’ Unions. We examine both of these studies, exploring how legitimacy was generated, created or afforded to the ballots.
Police and Crime Commissioner Elections
Police and crime commissioner elections were held ‘in forty-one police force areas in England and Wales on 15 November 2012’. These elections ‘gave life to the new constitutional arrangements for governing the police’ (Lister Reference Lister2013: 239). Police and crime commissioners were established to provide a link between the public and the police, and to hold the police to account (Home Office 2012). For the first time in Britain, members of the police were elected by the public. One definition of the new roles read, ‘Commissioners will be responsible for appointing the chief constable of their force, setting out local policing priorities, reporting annually on progress, and setting out the force budget and community safety grants. The government says commissioners are not there to run local police forces but to hold them to account’ (BBC News 2013).
Police and crime commissioners were described by one minister as ‘at the heart of’ reforms to policing (Green Reference Green2013). The elections reportedly cost a total of £100 million (Beckford Reference Beckford2012). The Electoral Reform Society (2012) calculated that this equated to a cost of £14 for every vote – four times higher than the previous general election and three times higher than the European parliamentary elections in 2009. Other criticisms were linked to notions of low turnout: the suitability of the Electoral Commission’s publicity of the elections was questioned, with no leaflets being funded or distributed and the apparent over-reliance on the use of the internet to advertise the elections when ‘over a fifth of UK voters do not have internet connections’ (Dunleavy Reference Dunleavy2012).
The elections were noted for both low turnout and a high number of spoiled ballots, which have been described by some as a means of protesting against the elections (Renwick Reference Renwick2012; Travis Reference Travis2012). Turnout for the elections stood at 15 per cent (BBC News 2012a; Rogers and Burn-Murdoch Reference Rogers and Burn-Murdoch2012), ‘the lowest recorded level of participation at a peacetime non-local government election in the UK’ (Electoral Commission 2013: 6), and one polling station in Newport, South Wales, recorded no votes at all (Wade Reference Wade2012). The 15 per cent figure masked a relatively wide variation in turnout: Staffordshire had a turnout of 11.6 per cent and Northamptonshire 20 per cent. Even the highest turnout was lower than previous lows; the lowest turnout for European elections is 23 per cent (1999), local elections (England only) 31 per cent (2012) and a general election is 59.4 per cent (2001) (BBC News 2012b).
The unusually high numbers of spoiled ballots varied between 8.5 per cent (Bedfordshire) and 1.7 per cent (Humberside) (Rogers and Burn-Murdoch Reference Rogers and Burn-Murdoch2012). In comparison, the number of rejected ballots in the 2010 general election was just 0.3 per cent (Rallings and Thrasher Reference Rallings and Thrasher2010). The Electoral Commission’s report into the elections found that ‘a significant proportion of rejected ballot papers were deliberately spoiled by electors wanting to register their concerns about the elections, although it has not been possible to quantify the extent of this activity through available electoral data’ (Electoral Commission 2013: 8).
Despite this, ministers constantly defended the elections. Damien Green, policing minister, claimed, ‘the electorate was still warming to the new role’, Prime Minister David Cameron said that turnout was ‘always going to be low’ as the posts were new, and Home Secretary Theresa May claimed that low turnout would not affect the police and crime commissioners’ mandate (BBC News 2012a, 2012c; Mulholland and Wintour Reference Mulholland and Wintour2012).
The elections were further defended, on relative grounds. The police now, according to government ministers, had some legitimacy. David Cameron, quoted in the Daily Mail, acknowledged the low turnout but asked people to ‘remember, these Police and Crime Commissioners are replacing organisations that weren’t directly elected at all’ (Chorley Reference Chorley2012). Speaking in the House of Commons, Theresa May said, ‘I make no apology for introducing police and crime commissioners, who have a democratic mandate for the first time. For the first time, the public know that there is somebody who has been elected who is visible, accessible and accountable to them. PCCs have replaced invisible, unaccountable, unelected police authorities’ (Hansard 2012). Ministers argued that the process of the election was sufficient to deem the newly created police and crime commissioners as legitimate, certainly compared with the unelected structures of policing they replaced.
The defence of the elections should not come as a surprise; the idea of electing police and crime commissioners was widely seen as the ‘flagship policing and criminal justice policy of the Coalition Government’ (Lister Reference Lister2013: 239). This policy commitment, along with the amounts of money being spent in a time of austerity, made it hard for ministers to reject the ballots. The results were not only politically important in terms of who was elected in November, but could be seen as integral to the coalition’s policy on law and order – a policy area often traditionally associated as being a strength of the Conservative Party, though lost to Labour under Blair’s ‘Tough on crime, tough on the causes of crime’ narrative (Green Reference Green2009: 526–7).
Trade Union Ballots
Prior to conducting strike action, trade unions are required to ballot their members. This requirement dates back to the 1984 Employment Act, introduced by the Thatcher government to weaken the powers of trade unions and prevent trade union militancy (Blanchflower and Freeman Reference Blanchflower and Freeman1993: 21). In March 2013 two teaching unions, the National Union of Teachers and the Teachers’ Union – who between them account for 90 per cent of teachers – announced ‘an escalation of their industrial action’ to include rolling strike action from the summer. This followed a ballot whereby 82 per cent of those balloted voted to strike (Burns Reference Burns2013).
Politicians and media commentators did not question the conduct of this ballot. Instead, attention focused on the turnout figures. The issue of turnout was played up by government ministers. Education Secretary Michael Gove and his staff calculated that turnout was 27 per cent. In reporting the figures, ‘Gove concluded only 32% of NASUWT members and 22% of NUT balloted voted for strike action, less than a quarter of all teachers’ (Wintour Reference Wintour2013). This figure, ‘less than a quarter of teachers’, was publicized by departmental spokespersons in debates in the media (Burns Reference Burns2013; Levy Reference Levy2013; Sky News 2013). The presentation of the results in such a manner demonstrates a clear bias. Here the percentage of teachers voting for strike action was taken as a percentage of all teachers eligible to vote, not just those who took part in the ballot. Here those who did not partake in the ballot, 70 per cent of teachers, were presented as being against the strike. Through this, two different accounts of the ballots occurred: for the trade unions 82 per cent of teachers voted to strike (this was the results of the ballots cast), for the government just one-quarter of teachers voted to strike (equating teachers who did not vote with those who voted against strike action).
The unions further claimed they had the support of parents and the wider public in opposing reforms. At the National Union of Teachers annual conference in Liverpool the results of a YouGov/NUT survey were released which stated, amongst other responses:
-
∙ After nearly three years in power, only a tiny minority of parents and teachers (8% of parents and 5% of teachers) think that this Government has made a positive impact on the education system. Over two-fifths (44%) believe it has had a negative impact.
-
∙ A majority of parents said they trusted head teachers (59%) and teachers (58%) to deliver their child’s education compared to only 6% who said they trusted the Secretary of State for Education.
-
∙ Less than a fifth (19%) of parents and just 7% of teachers think that the Government’s academies and free schools programme is taking education in the right direction. (NUT 2013)
The NASUWT for its part noted on its website:
The NASUWT industrial action is not aimed at disrupting pupils. Therefore where members have volunteered to take extra-curricular activities the action will not prevent members continuing to participate.
NASUWT’s rationale for calling an industrial action is that we want to ensure the provision of high-quality education. The action short of strike action is pupil, parent and public friendly – nothing in the action short of strike action will damage children’s education. The action short of strike action removes the tasks and burdens from teachers which distract them from teaching and learning. . . . In addition, strike action will be utilised sparingly only where it is necessary and appropriate to do so. (NASUWT 2014)
As an article in The Guardian (Adams Reference Adams2013) notes, the public’s response to the strikes (or at least what were perceived to be the causes of the strikes) were mixed. It quotes a Department of Education survey which claimed, ‘In a recent poll, 61% of respondents supported linking teachers’ pay to performance, and 70% either opposed the strikes or believed that teachers should not be allowed to strike at all.’ Yet the paper also notes, ‘Research published this month revealed that the UK public thought teachers should be paid about 15% more than their current salaries.’
Each side solely blamed the other for the breakdown of talks which could have alleviated the strikes. Gove, quoted in The Guardian (Walker Reference Walker2013), claimed he was ‘happy to talk any time, any place, anywhere to the trades unions’. Whilst such a comment may not be surprising, the reasons offered may appear so. Gove was prepared to enter discussions not to try and avoid the strikes but, he continues, ‘to try and get [the unions] to see the errors of their ways’. For their part, the unions ‘accused Mr. Gove of “megaphone diplomacy”, using public platforms to denounce them, and industrial action rather than engaging to discuss the issues’ (The Telegraph 2013a).
The ballots were presented as being ideological. Speaking at the Conservative Party conference, Gove claimed the trade unionists were ‘letting their ideology hold back our children’ and denounced the strike as being driven by ‘ideologically motivated’ union leaders (Burns Reference Burns2013). Such sentiment seems to overlook completely the process of balloting, implying that the decision to strike was not the will of the majority, but imposed by a few leaders. This links to questions of legitimacy. Gove seemed to conclude that the decision to strike was imposed by a section of the trade union that somehow went against the majority.
The government has further sought to use these strikes – along with others – to argue for a change in the law regarding the requirements for strike ballots. This proposal suggests that votes in favour of industrial action become valid, and legal, only when they achieve a turnout of 50 per cent of members – a proposal that trade unionists say would make legal strikes ‘close to impossible’ (BBC News 2014).
This demonstrates a clear distinction between ‘actual legitimacy’ and processes of legitimization. In this relationship, those outside the process of voting (government ministers) sought to delegitimize the ballot and the results. This article situates the government responses to the trade union election within its wider economy policy. Such a view places outcomes at the centre of the reasoning for the government’s response. Strikes have a negative impact on governments: not only do they directly affect the government as they are the employers of teachers; they also present a negative image in the minds of the public, who are affected by strike action. Fearing the strike and its consequences (a loss of votes at the next general election), the government sought to condemn the strike action – and one means of doing this was to delegitimize the result of the ballot.
It is not unreasonable to assume that, had the outcome of the ballot been different, the rhetoric surrounding the ballot would have altered. Had the trade unions overwhelmingly rejected strike action, it would be far-fetched to suggest that the government would question the legitimacy of the ballot as such a result would have been optimum for the government; issues or questions regarding the turnout would not have been raised in the same manner. Here the process of the ballot becomes of secondary importance to the impact of strike action. Linked to this are attempts to delegitimize the ballots – a political decision which seeks to discredit the ballots, not for the process under which they occurred, but for the (perceived) outcomes.
Whilst both the police and crime commissioner elections and the trade union strike ballots were considered to have taken place under due processes such as secret voting, accurate representation of results and offering voters a choice between competition alternatives, they were not treated in the same light or judged by the same criteria.
Discussion
The two elections were presented very differently based on turnout figures. In the case of the trade union strike ballot and despite four-fifths of those who voted being in favour of a strike, the low turnout (27 per cent) was used to delegitimize the vote. In doing so the government presented elections as outcomes, rather than processes – itself an ideological decision (remember that Gove criticized the unions for undertaking ‘ideological’ industrial action), which in turn alluded to an alternative definition of industrial democracy. Not only did this misrepresent the nature of the trade union election, it also stood in clear contrast to the government’s response to the police and crime commissioner elections, which were presented as neutral, ignoring the fact that the decision to hold the elections in the first instance was heavily value laden.
Here context is important. The police and crime commissioner elections were a cornerstone of the government’s police strategy; the decision to take industrial action went against the government’s policies for reforming pay and pensions. In this regard the government attempted to maximize the effects of the strike, claiming that the unions were responsible for widespread disruption and in doing so blamed the unions solely for instigating the strike. This was further added to by unfavourable headlines such as ‘Blackboard militants: The threatened teachers’ strike is an anachronism and jeopardises children’s education’ (The Telegraph 2013b). Two interrelated questions arising therefore are: ‘how important in the analysis of the election are those who instigate the election?’ and ‘how are the processes of (de)legitimization carried out?
Were the police and crime commissioner elections deemed legitimate because they were instigated by government? Would they have been legitimate, irrespective of turnout? The government saw these elections as a process, arguing and maintaining that, if duly followed, then the elections would be legitimate.
Was the legitimacy of the trade union election questioned, not because of the number of people who voted, but because of the way in which they voted? Had trade unionists, on a turnout of 20 per cent, voted overwhelmingly against strike action, would government ministers have been so quick to seek to delegitimize the ballot? Such a scenario would seem highly unlikely as the ballot result would not disrupt or question the government’s macro-economy strategy. The concerns of government – similar to those of employers – is primarily related to trade unions that are embroiled in disputes or which threaten to hinder the day-to-day running of the company/firm/school/hospital etc. or affect public services, threatening incumbent politicians’ chances of re-election.
The case studies presented here demonstrate that elections can come to be viewed as legitimate or illegitimate depending upon outcomes, rather than processes. In the two cases explored here the government’s narration of events differed widely. The government sought to legitimize the police and crime commissioner elections and delegitimize the trade union election, despite the latter receiving a higher level of turnout.
Differences also emerged in the narratives of what those not voting represented. In the case of police and crime commissioner elections it was argued that people did not vote because of the timing of the elections or the fact that they were new – according to this narrative the four-fifths of people who did not cast a vote did not question the elections. If the electorate had been aware of the election or more familiar with the new roles being created they would have voted in the election. The ability to vote was therefore seen to be legitimizing in the case of the police and crime commissioner elections.
In the case of the trade union ballot those not voting were presented as being against the strike. The strike, according to this narrative, commanded only the support of one-quarter of teachers and was thus viewed as being undemocratic. The same balloting process which, according to the government, legitimized the police and crime commissioner ballots was used to question the legitimacy of the trade union election. Such narratives of the ballots, and especially of those not voting, were not constructed by those who failed to turn out to vote (for whatever reasons), but were constructed by the government. This demonstrates the separation of ‘actual legitimacy’ from the process of legitimization. Those who could offer the ballots ‘actual legitimacy’, through voting, were not the same people as those who could legitimize the ballots through the creation of certain narratives. Here the government can be seen as operating at the centre of the legitimization process; by creating such a discourse the government sought to promote its own agenda and in doing so shifted the focus away from the processes to the outcomes of ballots.
Throughout the process of legitimization, questions of turnout are important. Does the existence of a democratic ballot ensure legitimacy irrespective of the number of people who vote in it? If no, then at what turnout level does the ballot become legitimate? Is a turnout of 20 per cent, 30 per cent, 50 per cent sufficient to achieve legitimacy? The answers to such questions depend largely on the theory of non-voters: are they opposed to changes in policies? Are they apathetic? Or just not represented in the options offered? Should a non-vote be equal (in terms of value) to a vote cast when seeking the democratic will or analysing the election results? In the case of the trade union ballots non-voters were presented as being opposed to strike action, whereas in the case of the police and crime commissioner elections non-voters were seen as unaware or uneducated regarding the elections or nature of elections.
Numerous studies have suggested reasons why people do not vote, such as socioeconomic group, education, apathy and lack of resources (to get to polling stations) (Aldrich Reference Aldrich1993; Kavanagh and Cowley Reference Kavanagh and Cowley2010: 409–10; Lutz and Marsh Reference Lutz and Marsh2007: 540). Here the reasons for such non-voting are irrelevant; instead, what are important are the reasons, or reasons offered, as to why people did not vote and the importance of a non-vote in terms of the election result. Important to the notion and process of legitimization are the action(s) of non-voters – and the interpretations of those actions. The narratives behind individuals not voting often become more important than the action of non-voting. Instead we suggest that importance should be given to the effects of non-voting on the result of the ballot. If someone wishes to vote against policy A they may be able to select a ‘Not A’ option. Voting ‘Not A’ clearly makes it more difficult for policy A to be selected (each vote for ‘Not A’ requires additional votes for policy A should it be elected). A person not voting, however, does not affect the results in the same way – if they do not exercise their right to vote, their action(s) require no more votes for policy A to be selected – not voting clearly has different repercussions from voting ‘Not A’.
Equating non-voters with those who oppose changes instigated by ballots not only offers a one-dimensional approach to the reasons why voters do not participate in ballots but also exaggerates the effect voting can have on political outcomes. Ballots offer voters a choice between predetermined outcomes, they are not a broad engagement with the public in an attempt to aggregate or calculate ‘the will of the people’. Ballots can offer voters either the choice between two or more candidates who, if elected, will represent them through implementing preselected (often party) policies in a predefined institution, or a choice between two (or more) competing policies – direct democracy. In both instances the outcomes are limited; in the first example the outcome will be one of N candidates who have been put forward, in the second it will be one of N policies. In neither is the option to express a complex viewpoint offered. There is no space on ballot papers to offer conditional answers. In the case of the representative, voters must choose individuals to represent them and carry out a manifesto the voters have little or no control over or input into. In the case of direct democracy, votes must be in favour of policy A, B etc. There is no space to suggest an alternative. For example, the trade union ballot asked members if they wished to undertake industrial action (that is, to strike) or not – there existed little scope for the union member who wished to take industrial action short of strike action to communicate his/her feelings on the matter accurately in such a ballot. Here, although choice is offered, it is only a limited choice.
Voting is an action. If I wish to vote I must do something: either register and complete a postal ballot or go to my local polling station between specific times. If I am presented with a choice over A or ‘Not A’ – a decision to strike or a decision not to strike – I cannot offer an alternative action (industrial action short of a strike, for example). Here the introduction of a ‘Not A’ option, as opposed to an option B, is important. ‘Not A’ suggests that the two choices on the ballot paper are linked; having choices A and B suggests that the two entities are separate from each other. In the first instance (‘A’ vs. ‘Not A’ ballot), if I hold opinion B I am not able to register it. Instead I am asked whether or not I favour option A. If two candidates stand in an election, each is unlikely to stand on a platform of not being the other – candidate B is not likely to promote herself as a ‘Not A’ candidate, but will instead offer her own policies. The introduction of more candidates makes it less likely that a ‘Not A’ candidate would stand.
Such options are by no means novel, as Hill notes:
Ballots could offer options that record disaffection or could contain an open category: ‘protest vote’, with a blank space for respondents to write their own comments. Formal voters could also use the space to register comments and protest. A compulsory voting bill sponsored by British Labour MPs in November 2001 contained the provision that voters would be able to ‘register their abstention on a new slot on the ballot paper’. Alternatively voters could be offered the not unprecedented opportunity to vote for ‘None of the Above’. For example, Russian presidential elections offer voters the option: ‘Against All’. In the 26 March 2000 election, 1.88 per cent of the electorate chose this category. The British Electoral Commission of 2001 also contemplated the possibility of providing an entry on ballot papers for ‘positive abstention’ whereby voters could vote for ‘none of the above’. (Hill Reference Hill2004: 90–1)
Such problems reflect the need to incorporate wider issues of non-voting and limitations of voters’ choice into an analysis of elections. This could also be extended to elections in non-democratic or pseudo-democratic systems whereby large numbers of electorates turn out to vote even though the degree of choice offered can be questioned.
In this article we have explored two similar ballots in recent British politics. Through doing so we suggest that the process of democratic ballots can be insufficient to acquire ‘actual legitimacy’. Instead we highlight the importance of perceptions and narratives, rather than binary absolute conceptions, of legitimacy. In the case of the trade union ballots it is not inconceivable to suggest that the discourse surrounding the ballots and wider processes of legitimization would have been different had the ballots produced different results. Through doing so we introduced and explored the processes of legitimization and delegitimization, processes which are distinct from legitimacy itself. The process of legitimization is inherently unequal and can extend beyond those directly involved in the ballot. Whilst we could only study two ballots here, an insufficient amount to warrant a concise theory of the process of legitimization, we have sought to move the debate beyond questions of turnout.
By separating ‘actual legitimacy’ as an abstract concept and legitimization as a process we have demonstrated how agents external to ballots can seek to legitimize or delegitimize elections. The process of legitimization extends beyond that of elections – it extends beyond the campaign or election day and the counting of results. Could actors then seek to legitimize other ballots in a similar way? If so, how? Is it impossible to imagine a situation whereby fraud and exclusion of the electorate are rife, yet the ballot is considered legitimate? Equally could an election which has been conducted democratically and sees high turnouts (or even a 100 per cent turnout) be subjected to attempts to delegitimize it? Whilst neither ballot nor election studied in this article has resembled either scenario, this study has demonstrated that either could, theoretically, become a reality. Whilst such scenarios seem remote, further empirical study would be required to gauge which agents are able to influence the process of legitimization. Furthermore, do the processes of legitimization and their use in such ballots question the role of ballots in democracy or in the democratic process? The impact of those able to (de)legitimize events – in particular the ability to do so without being involved in the ballot directly – based upon outcomes suggests that the normative assumptions of democratic ballots as being positive may need to be reconsidered or amended.
By distinguishing between legitimacy as a normative concept and legitimization as an empirical process we question the extent to which turnout is definitive in the relationship between ballots and legitimacy. We have suggested that (de)legitimization is highly subjective and can be a deliberate goal of individuals or groups, in order to achieve wider political outcomes. Further study is required to assess how and why certain agents are able to (de)legitimize ballots which they have little or no influence in, though some answers to such questions have been alluded to here.
Acknowledgements
Matthew Wood acknowledges funding from the ESRC Future Research Leaders grant ES/L010925/1.
