The Dimensions and Impact of Political Discontent in Britain

Political discontent remains a pressing issue for UK parliamentary democracy that needs to be better understood. We offer a range of theoretical perspectives on dimensions ofpolitical disaffection and seektomeasurethem with substantially newsurveymeasuresthatassesshowcitizensperceivetheperformanceandmotiv-ation of politicians. Our results indicate that the public’s critique of politics and politicians takes a number of different forms, which varies in predictable ways acrosssocialgroupsandaccordingtootherpoliticalandattitudinalmeasures.Main-stream parties are ensnared by political discontent, but other parties can be beneﬁciaries of it. We show that discontent is at least as important a driver of UKIP’s support as social, cultural, demographic or economic factors.


Introduction
Survey research conducted in Britain over more than a quarter of a century shows that citizens have been far from content with politics (Hansard Society, 2013; Lee and Young, 2013). In the 2015 General Election, non-voters were once again the largest single bloc of the electorate. Understanding discontent with the practice of politics then remains a pressing issue, and this article contributes to an emerging research agenda on political discontent by arguing for and demonstrating the value of understanding the different strands of discontent, by exploring its impact on different social groups and by identifying its political impact in terms of the parties that gain from it.
What we know about political disaffection is largely conditioned by how it has been measured by scholars at different points in time. Earlier work on 'diffuse support' for political authorities tended to align with the use of survey questions about government approval, trust and satisfaction with democracy, capturing generalised attitudes towards the political system (Easton, 1975). Some of these questions date back as far as the 1930s and the pioneering polling of Gallup in the USA and the British Institute for Public Opinion (which would later become Gallup UK). A lot is known, therefore, about the degree to which citizens hold positive or negative attitudes towards the performance of politics and (leading) politicians. Indeed, there is tremendous value in repeated survey questions as they allow analysts to explore longitudinal trends in public opinion. So, for example, it can be clearly established that there has been a long-term decline in trust in government (Lee and Young, 2013). But there is also a cost in that questions designed for one era may not be sensitive to the idiosyncratic features of democratic politics in a new context. There is also a danger that these longstanding survey questions may emphasise forms or styles of politics that are considered part of the problem but not the solution.
We argue that to understand the nature and shape of political disaffection we need to ask questions of the public beyond those most commonly asked in political surveys about voting intention, participation, trust in politics and politicians and judgements about the competence of leaders or parties. In an era where disaffection with politics is perceived to be widespread and embedded, we need to find ways of exploring in more depth its dimensions. We therefore have devised a range of theoretically grounded and new questions that seek insights about how citizens perceive the conduct of politics and politicians.
After exploring the nature of political discontent, we seek to determine the identity of its owners. Are there differences between social, economic and partisan groups in their beliefs about what politics can deliver and the integrity of its conduct? Finally, while established or incumbent parties are likely to be trapped in the treacle created by negativity towards politics there are political forces and parties that can gain from it. In the case of the UK that prize has primarily been taken in England at least by UKIP, which recorded 12.6% of the vote in the 2015 General Election. The article concludes by looking at the impact of negativity towards politics on UKIP's support. There is a fruitful debate (Evans and Mellon, 2015;Ford and Goodwin, 2015), inspired by Ford and Goodwin's (2014) seminal Revolt on the Right, on this question and we aim to contribute to it.
We begin our article through an exploration of a range of theoretical leads to help shape our understanding and to enable us to test the complexity of responses of public opinion shaped around the broad issue of political discontent. This process enables us to specify a selection of new survey questions that were presented to representative samples of the British population in 2013 and 2015. We then explore how different sections of society might gravitate to particular types of political disenchantment. With our theoretical map established, and informing the design of our survey questions, we consider the construction of our data in more detail before analysing the results.

Conceptualising political discontent in an age of anti-politics
Britain is not alone in experiencing widespread political disengagement and disaffection (on Britain see, for instance, Stoker, 2006;Hay, 2007;Allen and Birch, 2015 and on the wider comparative context, see Pharr and Putnam, 2000;Torcal and Montero, 2006;Norris, 2011;Papadopoulous, 2013). Many contemporary democracies (but not all) might be characterised as in the grip of an anti-politics reflected in two phenomena. The first is in behaviour and attitudes of citizens that are turning away from mainstream politics and parties. The second is in the way political elites and institutions are aware of the negativity of citizens and are seeking to ameliorate it or to exploit it in some form. The issue that we want to explore is less the existence of disaffection in this age of anti-politics and more its shape and contours. We want to establish not the rise of disaffection as such but more what it is that citizens see as the sources of their disillusionment.
We identify five conceptualisations of political discontent below drawn from different theoretical traditions that focus attention on different dynamics of a failing politics. One judgement might be that politicians simply lack the technical knowhow to drive policy success. The idea that technical or expert government might achieve more in terms of better outcomes than democratic government has a long history (Crick, 2000). The initial response in some countries-Italy and Greece-to the economic downturn was to experiment with forms of technocratic government. Do citizens think that politicians would be better replaced by technocrats? We test this line of thought through a question that asks whether politicians have the technical knowledge needed to solve the problems facing Britain.
Consistent with a public opinion embracing neo-liberal ideology and mixing it with a considerable dose of fatalism, it might be argued that governments can make no real difference to the economic and political challenges faced by societies. Such fatalism as Gamble (2000) notes completely undermines politics that rests on the understanding that collective action can make a difference. The loss of faith in politics would be complete if citizens believed that politicians can stand on the side-lines and wave their arms but that they are as likely to be as effective as King Canute allegedly was in turning back the incoming tide. There are echoes here of the debate about whether governments could stand against the forces of globalisation, with the hyper-globalists arguing that governments were merely pawns in the hands of multinational business and global economic forces beyond their control (for a review and critique of this literature, see Hay, 2007). In a strong neo-liberal formulation of this perspective (see Plant, 2004), government lacks the understanding and skills to steer the economy. The key to policy success rests on allowing freer markets and for private actors rather than government officials to take leadership. Some point to a decline over the last decade in citizens' sense of what is the responsibility of government, suggesting that neo-liberal ideology has done some work in Britain (Seyd, 2013). Do most citizens buy the depiction of politics as a more or less useless ornament? We test this idea by asking if politicians in government can make a difference to the major social and economic issues facing the country.
Maybe it is not a lack of technical or governing capacity that is the issue; is it that politicians lack the guts to take really tough decisions? Since the idea of democratic overload became popular in the 1970s, it has been suggested that a fundamental problem of democracy is that politicians find it hard to resist the demands of citizens and for fear of getting outbid by opponents are not prepared to take the hard decisions needed to address economic decline and instead are pressured into the easier option of pandering to inflated public expectations. King (1975) put forward arguments along these lines expressing fears about terminal government overload. The Thatcher premiership was seen by many as a response to the problem of overload, with its strong leadership as the antidote. Since then the fear of over-inflated public expectations overwhelming politics has surfaced again with Flinders (2012, p. 33) arguing that there is a strong 'link between unrealistic public expectations and why we hate politics' . Our test for this line of argument was to ask whether politicians possess the leadership to tell the public the truth about the tough decisions that need to be made.
Another potential failing of politics is its democratic myopia driven by the dynamics of electoral competition and the relentless logic of 24 hour media coverage. Nordhaus (1975, p. 188) presents one of the original substantial theoretical statements on this line of argument based on a stylised rational choice model of a representative democracy. It views politicians and voters driven by self-interested calculation. Voters will judge parties on their performance in delivering for them in the short term and incumbent politicians are therefore under irresistible pressure to deliver short-term gains or risk being voted out of office. As a result democracies 'will make decisions biased against future generations' (Nordhaus, 1975, p. 187). With respect to public spending, then, the pattern is prudence followed by a big giveaway. Moreover, this line of reasoning predicts that 'politicians should never be willing to impose short-term costs for future gain' (Jacobs, 2008, p. 201). If you add the short-term pressures created by the emergence of a 24-hour media cycle to the ongoing pressures of electoral completion, then it would appear that you have an even more advanced formula for a failing politics, at least in terms of its ability to address long-term and difficult decisions. The 'mediatisation' of politics is a core feature of contemporary democracies (Papadopoulous, 2013). The core role of the media in presenting contemporary politics is widely acknowledged (Mair, 2005(Mair, , 2013Street, 2011) and indeed understood by citizens themselves (Hansard Society, 2012). If politics is only viewed and communicated through the 24-hour media frenzy-with its emphasis on news management, spin and presentation of sound bites-it is little wonder that many citizens struggle to connect to it with any attempt to grapple with serious problems and issues. For this reason, mediatisation and short-termism are inextricably linked. We tested whether the public had concerns about the short-termism of politics and its media focus by asking if politicians are too focused on short-term chasing of headlines.
A final line of argument about failing politics that is prominent and worth testing in terms of its grip on public opinion is the fear that the process of politics has become dominated by special or powerful interests. A general case that several countries are developing a post-democracy regime where the political class is more concerned with meeting the needs of big business rather than ordinary citizens is developed by Crouch (2004). Transparency International (2013) found in its 2013 survey that citizens in many countries do indeed fear that their government is run at the behest of a few big interests, although the pattern does vary between countries. Indeed one does not need to subscribe to the view that politics is in practice dominated by a cosy elite of politicians, business interests and other lobbyists but merely recognise that citizens-divorced from any real experience of how politics work are prone to fearing that they will be duped and cheated in a political world of strangers (Allen and Birch, 2015). One organising idea here is that many citizens in today's established democracies hold a 'stealth' view of democracy (Hibbing and Theiss-Morse, 2002). These citizens do not want to hold decision-makers to account for the details of their decisions, but rather want to be reassured in a general sense that decisions are being made in a non-self-serving way by those entrusted with power. What citizens want is not more direct involvement in decision-making but latent representation; an assurance that decisions are being taken on grounds of general interest and not at the behest of sectional interests. What they fear is that politics is being run at the behest of others, with feather nesting politicians making decisions in cosy alliance with powerful interests rather than in the interests of the general public. We tested this proposition by asking if politics is dominated by self-seeking politicians protecting the interests of the already rich and powerful in society.

The responses of sub-groups: four types of determinant
Specific points of discontent may be more or less popular among different sections of society. We here identify four key ways in which different subgroups may gravitate towards particular beliefs about the flawed character of politics and politicians.
First, we first expect that there will be life cycle effects that relate to evaluations of the competence and technical knowledge/capabilities of politicians as well as their ability to take long-term decisions. Over a lifetime, as individuals experience successive governments and periodic social and economic crises, political scandal and failures in policy delivery, the accumulation of negative signals will contribute to variations in attitudes across age groups: older people will be more sceptical about the practical/technical efficacy of politicians. To the extent that British politics exists in a permanent state of crisis and/or scandal (e.g. King, 1975;Moran, 2003;Runciman, 2013), older generations will tend to lose faith with the ability of politicians to get things done, having repeatedly seen the optimism of new beginnings dashed on the rocks of governing reality. For political discontentment, the accumulation of effects of the costs of governing (Green and Jennings, 2014) is likely to exist long after the passing of a particular government. Similar effects are expected both in relation to beliefs about the technical capacity and expertise of politicians and their propensity to take short-term decisions, with negative attitudes being brought increasingly into focus through the repeated experiences of voters as they acquire a long-term view on what politics does or not deliver.
Second, certain points of discontent with contemporary politics will likely stem from presumptions about what are either desirable or necessary characteristics of politicians. We expect that gender might impact upon expectations about the attributes of politicians and their actions in government, due to political socialisation and/or situational factors (Orum et al., 1974;Tedin et al., 1977). This would lead men and women to tend to hold differing views about shortcomings of the political class. Specifically, the bias of men towards cultural traits such as strength, authority, leadership or decisiveness would exacerbate negative attitudes concerning the leadership capabilities and technical competence of politicians.
Third, we expect that an important strand of discontent with politics is likely to be structured by social class. Those enduring a disproportionate share of economic austerity-through the squeeze on living standards, employment insecurity, the withdrawal of social transfers and cutbacks in public services-will not necessarily take the view that politicians cannot make a difference to the problems facing the country, but rather will perceive this to be a distributive choice. Lower social grades are thus expected to view the political class as more interested in fingerpointing about who is responsible and as serving the interests of the already rich and powerful in society. In contrast, they might believe that politicians can make a difference, but that the failure to do so is down to the dominance of electoral and economic interests in decision-making.
Fourth, it is expected that partisanship will structure certain channels of discontent with politicians, as is observed in other political evaluations (e.g. Campbell et al., 1960;Evans and Andersen, 2006). This predisposes partisans to expressing negative views about politicians, in general, when their opponents are in office. Supporters of the party in government on the other hand are predisposed to view the political class favourably. We therefore expect that partisan effects will be observed in evaluations of the technical competence and efficacy of politicians and in perceived behaviours such as short-termism. We would also expect that supporters of outsider parties are more likely to express dissatisfaction with the political class as a whole, reflecting a deeper sense of discontent with the political system. Lastly, partisan effects might also be structured by changes in electoral preferences, so supporters who defect from the party in government within a particular time period might be more likely to express negative attitudes about the political class in general.

Survey design
To test these alternative explanations of public discontentment with contemporary politics, we developed a series of survey instruments designed to have face validity in measurement of specific aspects and perceptions of politics that shape this discontent, and also to enable analysis of the factors that influence the survey responses of individuals. The former offers a basis for a descriptive response to the question of what it is in general that people is think is wrong with politics in Britain today, whereas the latter can produce an inferential analysis of who in particular is most unhappy with politics and politicians. We thus explain the design of our measures, describe the responses to an online survey of the general public piloting the questions and present a multivariate regression analysis of demographic and political determinants of each of these dimensions of political discontent in Britain.
We start by formulating a series of statements about the capacity of politicians to respond to the problems facing Britain. Each statement corresponds to a particular explanation of the inadequacies of politics: (i) that politicians lack the technical expertise/capacity to deal with the complex problems facing the country, (ii) that politicians can make a difference to pressing social and economic problems, (iii) that politicians lack the leadership to take tough decisions and are too deferential to the demands of public opinion, (iv) that politicians are too tied to short-term decision-making and obsession with media headlines than long-term planning and (v) that politicians have been captured by special/sectoral interests to such an extent that politics no longer represents the interests of ordinary citizens. The question is formulated as a statement with which respondents may agree or disagree: 'Thinking about the problems facing Britain today, do you agree or disagree with the following statements?' A Likert-type scale is used for possible responses, using the options: 'strongly disagree', 'tend to disagree', 'neither', 'tend to agree' and 'strongly agree'. A mix of positive and negative statements is used to guard against acquiescence in survey responses, while the question order is randomised.

Q1. Politicians have the technical knowledge needed to solve the problems facing
Britain today. Q2. Politicians in government can make a difference to the major social and economic issues facing Britain.
Q3. Politicians possess the leadership to tell the public the truth about the tough decisions that need to be made. Q4. Politicians are too focused on short-term chasing of headlines. Q5. Politics is dominated by self-seeking politicians protecting the interests of the already rich and powerful in our society.
These survey questions were developed and trialled through an online survey conducted using the YouGov panel. Fieldwork was undertaken between 5 and 6 June 2013, and the total sample size was 1905 adults. The figures are weighted and are representative of all GB adults (aged 18+).

Descriptive statistics
We first present summary statistics of the survey. These are reported in Table 1. They reveal some of the complexity around citizens' disillusionment with politics. There is strong support for two critical statements about politics. First, there is overwhelming endorsement of the idea that politics is blighted by short-term pursuit of headlines (80% strongly agree or agree with this critique). Second, there is also a strong sense that politicians are self-serving and protect the interests of the rich and powerful (72% agree with this view). On other dimensions of political disillusionment the pattern is more complex. Only 20% think politicians have the technical knowhow to solve problems (with 52% in disagreement) suggesting that many citizens have doubts about elected politicians are capable of making good policy decisions. On the issue of whether politicians lack the leadership to be truthful with the public about the tough decisions that need to be taken, public opinion appears to be balanced (with 33% agreeing they have that capacity but 39% in disagreement). Despite the negativity about politics, people still believe that politicians in government can make a difference (just 13% expressed disagreement with this view). Add to this the finding that just 26% of people agree with the idea that politics is a waste of time and we have support for the general observation that it is the way politics is being conducted by politicians, rather than a perceived lack of potential value in politics or government, that is the dominant factor in disenchantment.
More depth to the analysis is provided by consideration of the pattern of responses across different sections of the population. Here we summarise the findings, again reported in Table 1. First, there is little difference in responses across social grade, gender or region (although Scottish respondents are slightly more negative about the character of politicians). In contrast, there is some stratification across age groups, with younger respondents tending to be less negative about politicians than their older counterparts; especially for the questions about    short-termism (Q4) and self-serving behaviour (Q5). This is interesting given that it is the young who are often identified as having lost faith with politics, disengaging from many formal channels of participation. These findings suggest that it is older respondents who are disillusioned with the character of contemporary politics, disappointed by its failures. This is borne out by the fact that the oldest group of respondents (60+) are most likely to agree that politicians can make a difference (Q2). The youngest age group (18 -24), on the other hand, most strongly disagrees with the view that politics is a waste of time (with just 16% agreeing with this proposition for Q2).
Partisan variations in responses are also revealing. Starting with current vote intention, supporters of the 'established bloc' of Conservatives, Liberal Democrats and Labour are the least likely to express negative attitudes. A general pattern across all the survey items is that the supporters of UKIP and other parties outside that bloc, in addition to those expressing no vote intention or an unwillingness to vote, tend to be more negative about the character of the political class. In the breakdowns of supporters of the main parties, there is some evidence that the degree of discontentment varies according to party vote preferences. For example, Conservative supporters are more likely to agree that politicians possess the leadership required to tell the public about tough decisions that have to be made. On the other hand, Labour supporters are more likely to view politicians as engaged in short-term pursuit of headlines (Q4), but UKIP supporters even more so (with 87% in agreement). Consistent with the traditional electoral base of the party, Labour supporters are also more likely to view politicians as protecting the interests of the rich and powerful in society, with 86% of its supporters expressing agreement with Q5.
Tellingly, partisan variations are less pronounced when responses are considered according to actual vote at the 2010 Election. Here, Liberal Democrat voters in 2010 are less likely to currently view politicians as having technical knowledge (Q1) than their Labour or Conservative counterparts. They are also quite similar to Labour voters from 2010 in terms of agreement with the view that politicians can make a difference (Q2), that they possess the leadership to tell the public the truth about tough decisions (Q3) and that they protect the interests of the rich and powerful (Q5). These findings suggest that part of political discontentment observed during the 2010 -2015 period was linked to the splintering of Liberal Democrat voters after 2010, with those who remained supporting the party in coalition tending to hold more positive views of politicians' technical knowledge, ability to make a difference, short-termist chasing of media headlines and protection of the interests of the rich and powerful. At the political centre, it appears that discontentment with politicians has been exacerbated by a sense of betrayal among a strand of probably left-leaning Liberal Democrat electoral support. How this dynamic might play out after their electoral wipe-out in the 2015 Election remains to be seen but it may be that a clearly left-wing leader of the Labour Party has and will pick up some support from left-leaning Liberal Democrats but alienate those with a more centrist orientation. Having provided an overview of variation of expressions of discontentment, we next consider the validity, consistency and reliability of our new survey measures.

Validity, consistency and reliability of the measures
To assess whether these measures are capturing distinctive expressions of discontentment or are reducible to a prevailing sense of political alienation that underlies survey responses, we consider the dimensionality of the survey items. First, factor analysis is used to determine whether the survey measures are capturing a latent construct. We estimate factor loadings for the first dimension, to see the extent to which these expressions of discontent are reducible to a single underlying dimension. The results of factor analysis are reported in Table 2. We test the stability of our findings for the measures trialled in the YouGov survey using a follow-up survey, fielded in early 2015. 1 These are reported in the columns on the right-hand side of Table 2. Here we see that each of the survey items to some degree loads onto the first dimension-the direction of the factor loading is consistent with negative attitudes towards politics/politicians (i.e. the loading is positive for Q1, Q2 and Q3, which were positive statements about the capacity/character of politicians, whereas it is negative for Q4 and Q5, which were negative statements). The communality of each item varies from around 0.1 to 0.5, which suggests that these measures are only partly explained by the common factor. This value is equal to one minus the uniqueness of the item, suggesting that the differences are due to more than sampling error. While these measures may be picking up an underlying sense of alienation with politics, they appear to be capturing distinct expressions of discontent too. (We also estimated a single-factor measurement model using structural equation modelling, based on item-response theory, reported in online Supplementary data, Table S1, which confirms the findings from the factor analysis.) Finally, Chronbach's (1951) alpha is used to assess the reliability of the items, indicating the degree of correlation between each item and the underlying factor. This again reveals a substantial amount of shared variation, with a equal to around 0.5 or 0.6 for all items. Together, these results confirm the validity, reliability and consistency of these measures. They offer novel and distinctive measures of discontent that nevertheless are related to a prevailing expression of political alienation.
While these descriptive statistics and additional tests are indicative of the ways in which the public views the character of politicians, it is necessary to undertake 1 The follow-up survey was conducted using the Populus online panel. Fieldwork was undertaken between 25 February and 1 March 2015. The sample of 4000 respondents is weighted to be representative of the national population of all GB adults (aged 18+). individual-level analysis to ascertain the specific factors associated with each of these views about the capacity of politicians to respond to the problems facing Britain.

Predictors of political discontent
We next consider results from a series of ordinal logistic regressions of political discontent based on data from the follow-up survey (from February to March 2015). These are presented in a step-wise fashion in Table 3. We first estimate a model that includes demographic predictors only; next we add current vote intention; and then we add self-reported past vote (at the 2010 general election) as dummy variables; finally we include a number of other political attitudes that might be thought to fuel discontent (e.g. disapproval of equal opportunities for particular social groups, negative attitudes towards immigration and the EU and preferences for income equality). The base categories for factor variables are indicated in Table 2.
The results are largely consistent with patterns observed in the descriptive statistics, while providing several new and additional insights. First, there are significant differences by gender for the questions about whether politicians have technical knowhow (Q1), possess the leadership to tell the truth about the decisions that need to be taken (Q3) and the short-term chasing of headlines (Q4). For each, men are significantly more negative about the character of politicians; for Q1 and Q3 the odds ratio of less than one for male respondents indicates that men are less likely to agree that politicians have technical knowledge or possess the leadership to tell the truth about hard decisions, while for Q4 the odds ratio of greater than one indicates men are more likely to believe politicians are preoccupied with chasing media headlines and wins. Second, there are significant differences by age group, though these are not as clearly stratified as in the descriptive statistics. Older age groups tend to be more negative across most of the questions, especially in the 45-54, 55 -64 and 65+ bands. The only question where age is not a significant predictor of negativity is that relating to whether politicians can make a difference to the major social and economic issues facing the country (Q2), where the 65+ group is more likely to take a positive view (i.e. the odds ratio is greater than one for the 65+ group in all models). Third, there are also differences by social grade for most questions, although interestingly the direction of effect varies across statements about politicians. For the positive statements about their technical knowledge, whether they possess the leadership to tell the truth and shorttermism (Q1, Q3 and Q4) higher grade social groups (ABC1) are more likely to be in disagreement than their lower grade (C2DE) counterparts. This suggests that professional/middle class respondents are more sceptical about the capabilities of politicians than working class respondents. In contrast, for negative statements about self-serving behaviour and working in the interests of the rich and powerful  (Q5), working class respondents are more likely to be in agreement. This is unsurprising given a greater sense of divorce from those forces that might be expected among the working class.
Fourth, turning to vote intention (as of March/February 2015) there are significant differences across the survey questions between the base category of Conservative supporters and those who intend to vote for most other parties-or who do not intend to vote at all (or are uncertain about vote choice). Differences in the effect size vary, though a few points are worth emphasising. Supporters of Labour, UKIP and other parties, and those uncertain who they will vote for or not intending to vote, tend to hold more negative attitudes towards politicians than their Conservative counterparts. Liberal Democrat supporters are also more negative in their views, but the effects are substantially weaker than other parties. Indeed, Liberal Democrat supporters are just as likely to believe politicians can make a difference as Conservative voters (Q2), most likely due to their status as coalition partners. Combined with negative effects for the supporters of Labour, UKIP and other parties, this suggests a more positive outlook about 'making a difference' is related to whether the party that an individual supports is in government.
Fifth, there are again telling findings with regard to past vote at the 2010 Election. Here, Liberal Democrat voters in 2010 are significantly more likely to agree that there is too much short-term chasing of headlines (Q4) or self-serving behaviour (Q5). That these effects all remain significant after controlling for current vote intention suggests that a portion of public resentment with the political class was a reaction of voters who defected from the Liberal Democrats after the party entered into its experiment in coalition politics. Lastly, some attitudinal predictors are associated with particular expressions of discontent, but the pattern is by no means clear or consistent. Disapproval of EU membership is associated with a lower likelihood of believing that politicians have technical knowledge, can make a difference or possess the leadership to tell the truth about hard decisions that need to be made (with the odds ratio equal to less than one for Q1, Q2 and Q3) and greater likelihood of agreement that politicians are short-termist and selfserving (with the odds ratio equal to more than one for Q4 and Q5). Respondents who are less concerned about income inequality are more likely to disagree that politicians can make a difference (Q2), but are also more inclined to disagree that politicians are too focused on chasing headlines (Q4) and protecting the interests of the already rich and powerful (Q5). Belief that immigration is bad for the economy is associated with fatalism about whether politicians can make a difference (i.e. the odds of agreement are lower for Q2), but also reduces the likelihood of agreement that politicians are too focused on short-term chasing of headlines. While concerns that equal opportunities have gone too far for ethnic minorities increase the odds of believing politicians are short-termist and self-seeking (Q4 and Q5), attitudes on equal opportunities for women reduce the odds for we offered. We argued that older generations would be more likely to offer technical/short-termist critiques of politicians because their longer experience of politics in practice would generate repeated and accumulative disappointment. We found strong support for these effects and more generally some support for age differences impacting in relation to all the potential responses to the connections between austerity and political disillusionment. Given the different ways that men and women might judge the behaviour of politicians, we argued that men might be inclined to greater doubts about the leadership of politicians. Our findings suggest men are indeed more likely to view politicians as lacking the leadership qualities to speak hard truths and were even more negative about perceived short-termism and media orientation of politicians. We also expected individuals from lower social groups to be more negative about politicians not paying attention to their interests and instead viewing the political class as protecting the interests of the already rich and powerful in society. We found strong support for both propositions but also that lower social groups were less likely to express the view that politicians lacked technical skills and leadership to tackle hard choices. Finally, we argued that partisanship might make a difference and in particular that supporters of the party in government would be more forgiving about the qualities and character of political elites. Again we found support for this expectation, with supporters of the Conservative Party the least negative generally about the performance of political elites whereas UKIP supporters rival those who are undecided or do not intend to vote for top slot in terms of negativity. The finding that UKIP supporters are the most negative towards politics of those who are committed voters spurred us to a final piece of analysis to examine whether suggest that any underlying latent dimension of discontent in public opinion does not fully capture the many repertoires of disaffection.

Discussion
At the outset of this article, we identified several questions that have all been addressed in our analysis. The first general question asked what it is that citizens object to about politics and the answer it appears is politicians and their behaviour rather than the political system. Our findings reveal the predicament of democratic politics in Britain today. The most intense points of citizen disillusionment with the political class resides in perceptions of its flawed character, in particular its fixation with headlines and protection of its own interests and those of the already rich and powerful. These beliefs are less to do with failings of the processes of politics, and more to do with the attributes of its protagonists in response to the pressures of outside interests.
The second general question asked about differences between groups in their negativity towards politics. Here our findings can be used to assess the hypotheses (8)

Continued
UKIP are indeed the main party political beneficiaries of the public's sense that the political class has failed in its performance and its character. Specifically we test the extent to which each of the strands of political alienation can account for UKIP support. To do this, we present a simple comparison between a model of UKIP support based on social predictors, drawing on Ford and Goodwin (2014), and a series of models that include each of the measures of political discontentment, first individually and then collectively, and finally a model based on scores from factor analysis of the measures (see Table 4).
The social predictors of UKIP support identified by Ford and Goodwin (2014, p. 160) are individuals who are white, male, over-54, working class (social grades C2, D and E) and left school aged 16 or before. Our survey includes each of these demographic measures and thus represents a good baseline for comparison. We estimate a logistic regression where the dichotomous dependent variable is whether or not an individual intends to vote for UKIP or not. Our social predictors are also dichotomous: whether or not an individual is male, whether they fall into a particular age group (25 -34, 35 -44, 45 -54, 55 -64 and 65+, with 18 -24 as the reference category), whether they self-report as one of social grades C2, D or E, whether they finished school aged 16 or earlier, whether their household income is low (equal or less than £21,000 per year), whether they read right-wing tabloids and whether or not they live in a particular region of Britain (the Midlands, North, South or East of England, Wales and Scotland, with London as the reference category). Our antipolitics predictors are of course scaled on a range from 1 to 5, where 1 indicates strong disagreement with the statement about politics and politicians and 5 indicates strong agreement.
Consistent with the findings of Ford and Goodwin, we observe strong significant effects of social predictors on the probability of voting for UKIP (Model 1). Being aged 55 -64 or 65+ mean that the odds of the dependent variable, intention to vote UKIP, are around two and a half times higher. Similarly, being male, working class and a tabloid reader all lead to an increase of around 1.5 in the odds of support for UKIP. Turning to our new measures of political discontent, we see that a number of these are significant predictors of UKIP support (the effect of each measure is first tested in turn in Models 2-6). Here we can see that belief that politicians have technical knowledge to solve the problems facing Britain today, can make a difference and possess the leadership required to take tough decisions, each lead to a decrease in the odds of intending to vote UKIP. For the former measure, the odds ratio of 0.8 indicates that a one-unit increase in agreement with the statement that politicians have technical knowhow leads to a 20% decrease in the likelihood of intention to vote for UKIP. The effect size is comparable for the other measures too. Similar but larger effects are observed for the short-termist and self-serving critiques of the political class, which increase the propensity to support UKIP by around 50%. In the combined model (Model 7), these remain significant predictors of vote choice, as does scepticism regarding whether politicians can make a difference. If scores from our earlier factor analysis, which revealed some collinearity in the measures, are included instead (Model 8), the results confirm that discontent increases the odds of supporting UKIP by more than half. Importantly, compared with social predictors these measures of political discontent improve the explained variance in vote choice, with McFadden's Pseudo R 2 increasing from 0.07 to 0.09 in the full model and the model based on a single underlying factor.

Conclusion
Our results show that there is a complexity to the spread and nature of citizens' negativity that no single explanation is able to do justice to, no matter how compelling it may appear. There are plainly strongly supported and different strands to the public's critique of politics and there are substantial differences between social groups as to which elements of the critique are more important. Negativity is not a simple product of the rise of a social group that finds itself neglected by the political class as some suggest (Ford and Goodwin, 2014), given that our evidence points to a widespread negativity among all sections of society. Second, negativity is unlikely to be solely the product of a track record of British government incompetence as implied by Crewe and King (2013), given that it appears British citizens still retain some considerable faith in government and the technical failing of government was one of the lesser favoured elements of the critique of politics. Negativity is both driven by a complex of factors and the mix of factors varies between social groups.
For those interested in the reform of politics, the insights of this research may be of value. Our findings suggest that is the behaviour of politicians that is the main cause of concern rather than the processes of the political system or belief in government as such. The challenge of course remains how to get political elites to behave differently and to be perceived as having done so.

Supplementary data
Supplementary material available at Parliamentary Affairs online.

Conflict of interest
The surveys that are used in this research were funded by the University of Southampton and the University of Canberra, Australia. Will Jennings and Gerry Stoker