What makes people trust or distrust politicians? Insights from open-ended survey questions

Abstract High political trust is often interpreted as a sign of good democratic health, and widespread distrust as a sign of democratic ill health. However, there is little knowledge about the basis on which people make assessments about whether to trust or distrust political actors. This article develops and applies a typology for political trust judgement. Through a content analysis of 1,105 open-ended survey questions about political trust and distrust, the study finds that people tend to assess (dis)trust of politicians based on whether they see the politicians as predictable, intrinsically committed, competent and responsive. Moreover, the study finds that citizens use different judgement bases to describe trust and distrust, suggesting that the two concepts are not pure negations of each other. While predictability is the most frequently reported basis for trusting politicians, a lack of intrinsic commitment is the most frequently reported basis for distrusting politicians. This article discusses the reasons for and implications of the apparently different bases for weighing judgements on trust and distrust.

Political trust is widely seen as a crucial ingredient of well-functioning representative democracies and is said to secure institutional stability (Klingemann and Fuchs 1995), build support for governmental policies (Hetherington 2005) and foster law-abidingness (Marien and Hooghe 2011).Accordingly, political scientists have researched the concept since the advent of modern survey research (Levi and Stoker 2000).Recently, in the wake of the Great Recession, the 2015 migration crisis, the incumbency of populist candidates such as Donald Trump, and global pandemics, the concept has received unprecedented attention.Today, there are widespread concerns among both the broader public and academic researchers that established democracies are facing a political crisis of trust (Citrin and Stoker 2018;Karić and Međedović 2021;Schraff 2021; Van Ham et al. 2017).
Understanding people's reasons for political trust or distrust is important for comprehending the political ramifications of the alleged decline in contemporary trust ( Van Ham et al. 2017) and, more generally, the conditions for democratic legitimacy.Recent scholarship concludes that the most compelling explanation for political trust comes from political performance (e.g.Citrin and Stoker 2018;Thomassen et al. 2017).Political trust has been consistently linked to corruption and procedural fairness (Grimes 2017), economic developments (Van der Meer 2017a), and satisfaction with the welfare state (Kumlin and Haugsgjerd 2017).The perceived trustworthiness of politicians has been attributed to assessments of their competence, commitment, predictability, integrity, benevolence, morality, authenticity and responsiveness (see Fisher et al. 2010;Denters 2002;Grimmelikhuijsen and Knies 2017;Van der Meer 2017b;Valgarðsson et al. 2021).
Despite the impressive volume and richness of the political trust literature, previous research has relied on a limited methodological toolbox.Theories of political trust are almost exclusively examined through close-ended survey questions in which citizens are asked to rate a predefined set of theoretically derived reasons for trust.Although superior in efficiency, there are well-known drawbacks to relying exclusively on close-ended questions (e.g.Schuman and Presser 1979).In particular, close-ended questions may not provide an adequate set of alternatives that are substantively meaningful to respondents; a more valid picture of respondents' preferences could be obtained if they had to formulate the answers themselves.
In this article, we contribute to the literature on political trust by analysing open-ended survey questions about the reasons for political trust and distrust.While the extensive use of close-ended questions has certainly provided valuable insights, we believe that an open-ended approach can add important qualitative nuances to and validate existing theories about the basis for political trust and distrust.Furthermore, our study contributes to the literature by exploring whether people use different bases of judgement to describe trust and distrust.While most previous research has treated trust and distrust in matters of degree, recent studies have argued that trust and distrust are distinctly different concepts that relate to different evaluative or judgemental processes.In other words, a lack of trust does not necessarily equate to distrust (Bertsou 2019).Despite increasing theoretical attention, few, if any, studies have empirically assessed the judgement basis for the distinction between trust and distrust.Thus, in this study, we compare respondents' own descriptions of why they trust or distrust politicians.
In the following sections, we first discuss how previous research has conceptualised political trust judgements.Based on this literature review, we develop a preliminary typology of judgement bases for trust and distrust.After describing the empirical data and our methodology, we apply our preliminary typology to analyse open-ended questions from a representative population survey on trust and distrust in Norwegian local politicians (N = 1,105).Based on the analysis, we propose a refined typology of the basis for (dis)trust judgement, and we explore the prevalence of different understandings of political trust and distrust in the Norwegian population.

Conceptions of political trust
Political trust denotes the belief that political actors and institutions will look after citizens' interests and values when it comes to making political decisions and taking political action (Hetherington 2005;Miller 1974;Newton 2007).Whether citizens trust their elected representatives and political institutions is thus a critical measure of the health of representative democracies and an essential prerequisite for the legitimacy of political systems (Easton 1965;Van Ham et al. 2017;Warren 2018).
Like any form of trust, political trust is relational (Hardin 1999).Specifically, it involves a relationship between a subject who trusts (i.e. a citizen) and a trustee, an object in which the subject places its trust (i.e. a political actor or institution).A common way to distinguish political (dis)trust from neighbouring concepts, such as political support and satisfaction or cynicism and alienation, is to understand that trust involves risk and vulnerability on the part of the subject (De Blok 2020;Slovic et al. 1991;Van der Meer 2017b).Fisher et al. (2010: 163), for example, describe how 'political trust involves granting political actors discretionary powers over the use of collective goods while recognising that this delegation of power comes at some risk to oneself ' .Since there is no way to provide certainty about the future behaviour of the elected representative, political trust requires a 'leap of faith': citizens must believe that the political representatives and the institutions they govern will act in their interests with the knowledge that if the representatives do not meet this expectation, the citizens are vulnerable to harm (Möllering 2006;Van De Walle & Six 2014).
Trust rests largely on people's assessments of the trustworthiness of trustees.Although there is some disagreement about how to conceptualise trust versus trustworthiness, trust is generally understood to be a characteristic of the one who trusts (the trustor has trust), whereas trustworthiness describes the characteristics of the one who should be trusted (the trustor's perceptions of the trustee; Grimmelikhuijsen and Knies 2017: 586).While trust is the intention to accept vulnerability to trustees based on positive expectations of their actions, trustworthiness represents the reasons on which these expectations are based (Colquitt et al. 2007;Levi and Stoker 2000;Mayer et al. 1995) and thereby constitutes the foundations on which trust is built (Mayer et al. 1995;Grimmelikhuijsen and Knies 2017).The assumption is that trustworthiness relies on citizens' evaluations of whether the politician will take the citizens' interests into account when making decisions (Hardin 2006;Van der Meer 2017b;Valgarðsson et al. 2021).Understanding how people assess trustworthiness -their reasons for trust and distrust -therefore helps further the understanding of what people emphasise when they say that they (dis)trust politicians or political institutions.

Reasons for trusting politicians
Scholars of political trust have explored and typologised the reasons for political trust differently.For example, Fisher et al. (2010) discern between strategic moral and deliberative trust judgements, whereas Grimmelikhuijsen and Knies (2017) discern between judgements based on perceived competence, benevolence and integrity.Valgarðsson et al. (2021) add the category of authenticity, claiming and demonstrating that politicians perceived as authentic inspire trust.In our analysis, we depart from the typology of Van der Meer (2010), which we believe captures the most important reasons for political trust proposed in the otherwise vast trust literature.Building on Kasperson et al. (1992), Van der Meer argues that political trust is a subjective and rational evaluation of a relationship along four dimensions: the degree to which the object of trust is considered competent, intrinsically committed, extrinsically committed and predictable.To Van der Meer's originally fourfold typology, we have added a fifth category, responsiveness, which was proposed as a judgement basis for political trust by Denters (2002) and Torcal (2014), among others.Responsiveness refers to the degree to which the object of trust is considered responsive to the needs, wishes and opinions of the citizens, and cannot, we argue, be adequately subsumed under any of the four other categories.
The five categories are developed theoretically below and serve as starting points for our analysis.In choosing to map all reasons into five categories, some nuances will inevitably be lost.People trust and distrust politicians for a vast array of reasons, a fact to which the many ways of categorising reasons for trust testify.Moreover, the boundaries between the different reasons for trust and distrust are not always clear-cut, and we will discuss how they overlap and to what extent they can be said to include adjacent categories proposed by other authors.
A first basis for political trust may be the assessment of the trustee's competence, which means that the subject of trust (the individual citizen) thinks that the object of trust (a political actor) has the ability and competence to perform according to the subject's expectations and interests (Tomankova 2019;Van der Meer 2010, 2017b;Warren 2018).Competence may be specific to certain domains, which means that trust is based on an evaluation of whether the one to be trusted has competence in the domain over which trust is being given (Levi and Stoker 2000).A competence assessment may be related to political actors' knowledge and expertise or may be inferred from the actors' successes and failures in terms of policy outcomes (Bertsou 2019;Tomankova 2019).
Intrinsic commitment implies that the objects of trust -the political actors -intend to and will act in the best interests of the citizens because they have an intrinsic need to do so, for instance, because they care for the citizens or share with them the same goals or the same morals, values and notions of what is right and fair (Bertsou 2019; Levi and Stoker 2000;Van der Meer 2010).Intrinsic commitment means that the political actor is benign towards the citizens, and benevolence is evaluated, for example, with respect to the alignment of policies with their own values and judgement (Tomankova 2019: 170).Hardin's (2002) concept of political trust as 'encapsulated interest' implies an intrinsic commitment.According to this understanding, trust is based on the belief that 'we trust you because we think you take our interests to heart and encapsulate our interests in your own' (Maloy 2009: 151).Warren (2018: 1) describes intrinsic commitment as motivation, implying that the one to be trusted is motivated to act in the citizens' interest 'and will do so without overseeing or monitoring' .
Trust based on an evaluation of extrinsic commitment means that the citizen is able to control what political actors do or otherwise hold them accountable, for example, through the threat of punishing untrustworthy behaviour by denying future support (Van der Meer 2010, 2017b).Extrinsic commitment may come about when there are procedures that enable publicity and empower monitoring.Political representatives may be trustworthy in the sense that their actions are publicly known in ways that are sufficient for citizens to make judgements that motivate their votes (Warren 2018).Like intrinsic commitment, extrinsic commitment may serve as a motivation for politicians to act in line with citizens' needs and wishes.In contrast to the other three bases for making judgements proposed by Van der Meer, extrinsic commitment does not characterise politicians themselves.Extrinsically committed politicians may lack the other characteristics associated with trustworthiness but may be trusted because they are kept in check by political institutions.It follows that extrinsic commitment is perhaps less relevant for trust in political actors than for trust in political institutions.
The fourth basis for making judgements about whether a political actor is to trust in Van der Meer's (2010) typology is predictability, the extent to which the object's behaviour is consistent and in line with promises or expected actions (see also Ruokonen 2013).Predictability is closely related to, and, we argue, includes, integrity (Mayer et al. 1995).As politicians of integrity will do what they think is right, external pressure is less likely to change their behaviour.Accordingly, citizens can more easily predict their actions based on what they are saying and can therefore trust them.Still, predictability extends beyond integrity, because someone might act in a predicable manner for reasons other than integrity.
Responsiveness implies that politicians listen, consider and react to citizens' political signals, viewpoints and interests (Denters 2002;Linde and Peters 2020;Manin et al. 1999;Torcal 2014;Torcal and Christmann 2021).This category differs from the others in that it refers directly to the relational bond between the truster and the trustee, as experienced by the truster, rather than to the characteristics of the individual politician or the system.The category of responsiveness captures responsive behaviour regardless of politicians' motivations for such behaviour.Linde and Peters (2020) show that peoples' perceptions of political responsiveness are positively related to support for the political system, and Torcal (2014) finds that political responsiveness is an important predictor of trust in political institutions.
While we call this category 'responsiveness' , similar concepts have been proposed under other names.For example, Valgarðsson et al. (2021: 858) suggest and show that 'authenticity' , referring to being 'in touch with the lives and outlooks of ordinary people' , is a criterion by which citizens judge political trust.Fisher et al. (2010) suggest that citizens make trust judgements based on what they call deliberative reasons, which implies that politicians make themselves available for deliberative processes and are willing to discuss their opinions and decisions with citizens or in the public sphere in which citizens participate as audiences.Crucial to 'responsiveness' as a judgement basis is that the evaluation of trustworthiness is made with reference to the (perceived potential for) interaction between the citizen and the political actor.While responsiveness shares some traits with Van der Meer's original categories, it is not a definitory or necessary feature of any of them.Although perhaps unlikely, politicians can in theory be judged competent, intrinsically committed, extrinsically committed or predictable without interacting with or listening to their voters.Therefore, while responsiveness could be an indication of some of the other categories, it cannot be fully subsumed under any of them.
The delineations between the categories are not clear-cut and may be intertwined in numerous ways.For example, intrinsic and extrinsic commitment (wanting to or having to act with citizens' interests at heart) may motivate politicians to meet other trust requirements, such as acting in a predictable or responsive manner.Competence, referring to the ability to act in a manner that ensures the citizens' interests, might also include the ability to act in line with, for example, the norms of predictability or responsiveness.To further complicate the analysis of judgement bases, the categories refer to different types of phenomena.While the two commitment categories describe motivations for norm compliance, predictability and responsiveness may be regarded both as norms in themselves and as specific types of behaviour, whereas competence is a personal attribute.The porous contours of the categories and their dissimilar natures make the categorisation of trust assessments a complex exercise.Moreover, although these criteria provide a useful point of departure for exploring the basis on which citizens make their trust judgements, they are only partially specified.Through a qualitative reading of people's descriptions of trust and distrust, we aim to supplement and nuance this typology by answering the following research question:

Political trust and distrust -two sides of the same coin?
If high levels of political trust require citizens to consider politicians as competent, intrinsically committed, extrinsically committed, predictable and responsive, does it follow that distrust is a function of politicians not living up to these standards?In other words, is distrust the mere negation of trust?Most previous research treats trust and distrust as a matter of degree, assuming that the absence of trust equals distrust.Recently, however, some studies have suggested that trust and distrust might well be distinctively different concepts that refer to different evaluative or judgemental processes (Bertsou 2019;Van De Walle and Six 2014).
It follows from the definition of political trust (e.g.Newton 2007) that a person who lacks political trust believes that political actors will not look after their interests.On the other hand, according to Bertsou (2019), distrust is something more than a mere lack of trust; a distrustful person believes that political actors may intentionally or unintentionally harm them.Bertsou defines political distrust as 'a relational attitude that reflects perceptions of untrustworthiness specific to the political system in its entirety or its components.The evaluative part of distrusting attitudes is distinctly negative and entails the expectation of harmful outcomes ' (2019: 220).Distrust, therefore, is something in its own right and something more than the absence of trust.
The trust-distrust distinction is largely uncharted empirical territory.Bertsou suggests that perceptions of untrustworthiness are likely based on evaluations of 'technical incompetence or failure; conduct that violated shared notions of right and fair and conduct that is incongruent with the citizens' best interests' (Bertsou 2019: 221).These evaluation bases involve two of the above-described judgement bases: competence and intrinsic commitment.Following Bertsou (2019), we will explore how the evaluations of competence and intrinsic commitment are involved in judgements of trust and distrust, respectively, extending the analytical framework to include the other three judgement bases described above.Our aim is to explore the bases on which such judgements of distrust are made and whether these bases differ from those on which people make trust judgements.Thus, we ask the following question: RQ 2: Does the basis for (dis)trust judgements depend on whether the respondents are asked to characterise trust or distrust?

Data and methods
Through an analysis of two open-ended survey questions, we investigated the basis on which a representative sample of Norwegian citizens trusts or distrusts local political representatives.Like other Northern European countries, Norway stands out as a high-trust country (e.g.Norris 2011; Torcal 2017).Studies have shown relatively stable (Segaard et al. 2020) or even increasing (OECD 2022) trust in politicians in Norway in recent decades.Within this context, however, there is considerable individual variation, and some groups stand out as having systematically lower trust than others (Listhaug 2005;Haugsgjerd and Segaard 2020;Segaard et al. 2020;OECD 2022).On a scale of 0-10, the average trust in local politicians was 5.5 (Haugsgjerd and Segaard 2020).Comparably, in this sample, the average trust in local politicians was 3.1, measured on a five-point scale (see Online appendix Table A1).Consequently, both high and low levels of trust and distrust can be examined in the Norwegian context.In the concluding section, we discuss the extent to which our findings can be generalised to other contexts.
There has been much debate in the literature about how to conceptualise and measure the object of trust (e.g.Fisher et al. 2010;Hooghe 2011).We build on the conventional assumption in the literature that political trust captures attitudes towards middle-range objects (e.g.political actors and institutions) in Norris ' (1999, 2011) multidimensional framework of political support.Specifically, we focus on citizens' trust in elected representatives in local government assemblies.
The survey was conducted in 2019 through the fifteenth wave of the Norwegian Citizen Panel, a survey panel of pre-recruited respondents who agreed to receive up to four online questionnaires per year.Within this wave of the survey, the section that included the trust questions was sent to 3,175 respondents, of which 1,105 answered, giving a response rate of 34.8%.These 1,105 respondents answered all questions, including the open-ended ones.As some of the answers to the open-ended questions did not make sense and were removed (accounted for below), we ended up with a net sample of 1,079 respondents, which gave a response rate of 34.0%.Young people and people with low levels of education were underrepresented in the sample, and the frequencies were therefore weighted to adjust for these biases (Skjervheim et al. 2019).The sample was split into two groups.Approximately half of the respondents (575) answered a question about what made them trust local politicians, and the other half (530) answered a question about distrust.The questions, intended to capture how people assess whether to trust or distrust politicians, were formulated as follows (translated from Norwegian): 1   • What is important for you to have trust in municipal politicians?
• What would you say contributes to distrust in municipal politicians?
Please write down the first that comes to mind.We want all types of answers -a couple of sentences or just a few words.
The wording of the questions requires a comment.While the trust question refers to what makes the respondent trust politicians, the distrust question asks the respondent to reflect more generally on the causes of distrust in politicians.Ideally, the questions should have been similar for the sake of comparability, but due to the different normative standings of trust and distrust in a political context, the two concepts required slightly different formulations.Notably, while political trust is perceived as normatively good, political distrust is perceived as normatively bad.We could therefore not formulate the question on distrust in the same way as for trust (i.e.'What is important for you in order to have distrust of local politicians?')as such a formulation would create the presumption that distrust is something desirable.While it would have been semantically possible to make the trust question mirror the distrust question (i.e.'What would you say contributes to trust in municipal politicians?'),this way of wording the question would, in a Norwegian context, likely have been perceived as abstract and artificial and would have invited vague answers.Thus, motivating the respondents to reflect on reasons for trust and distrust required somewhat differently worded questions.We should have different wordings in mind when interpreting the results.Specifically, the wording of the trust question may lead respondents to consider personal experience, while the wording of the distrust question may trigger more abstract reflections or reflections on what people in general think.However, the difference should not be exaggerated.Such general reflections on distrust may well be based on the respondent's own evaluative basis for distrust, and we believe that both questions were worded in a fashion that allowed respondents to think of a broad range of evaluation bases.That the 'natural' way of asking about distrust differs from the natural way of asking about trust may indicate that trust and distrust are not mere opposites but may be subject to different evaluative processes.
In order to assess the relationship between trust judgement types and trust level, the respondents were asked to assess their trust in the local politicians in their municipality on a five-point Likert scale ranging from no trust (1) to complete trust (5).This trust item was placed right before the open-ended questions about the reasons for (dis)trust.Both the close-ended and open-ended questions were posed immediately after questions about voting behaviour.There is no reason to suspect that the preceding questions interfered with the responses, and if so, the influence was uniform across the sample.
The responses to the questions were analysed in two steps.In the first step, two of the authors conducted a qualitative reading of all the answers with the ambition of refining and operationalising the theoretically defined judgement categories.This first reading revealed that most judgement reasons occurring in the 1,105 responses could meaningfully be subsumed under one of the five original judgement categories.Furthermore, a set of subcategories was identified under each of the five overarching judgement categories, and an elaborate coding scheme was developed.Twenty-six responses (2%) could not be categorised into one of the five categories.These included responses such as 'I don't know' , 'I am happy with everything' , 'democracy disappears' and responses that did not make sense in this context, including long and unintelligible rants.These were removed from the analysis, and the number of responses in the analysis was consequently reduced to 1,079.
In the second step, three of the authors coded the entire sample using the agreed-upon coding scheme.Each response was coded as 1 if a judgement basis occurred and as 0 if one did not occur.Many respondents mentioned different reasons for (dis)trust.One single response could therefore be coded 1 for several different judgement bases and for several subcategories within each judgement basis.A complete coding scheme and descriptive statistics are provided in the Online appendix (Table A2).
The respondents' written answers were the units of analysis.The percentages presented in the analysis refer to the percentage of replies on which a judgement basis occurred.We have reported occurrences in which at least two of the three coders agreed on the coding.Interrater agreement coefficients (Cohen/Conger's Kappa) were calculated and are reported in the Online appendix (Table A3).
The answers varied in length from one word to an entire paragraph.The average length of the responses was 11 words (10.5 words for describing trust and 11.4 for describing distrust).Most responses were rather short and were typically off-the-cuff answers.Our data, therefore, say something about what comes quickly to mind when people are asked about their reasons for trust and distrust.While such easily accessible evaluation bases say something about how people make trust assessments, the brevity of the answers and the situations in which the answers were given (in a small text box on a computer or a portable device) represent a limitation of our material.To account for evaluative processes with more nuance, other methods, such as in-depth interviews or experimental situations, would be required.

Analysis
Starting with the five theoretically defined judgement categories, we analysed the responses to 1,079 open-ended survey questions with the aim of exploring the bases on which people judge whether they can trust or distrust political representatives.Three of the five categories were divided into distinct subcategories.The frequency with which the five judgement bases occurred in the entire sample and for those being posed the question of trust and distrust, respectively, are displayed in Table 1, as are the subcategories.Because the respondents could mention several reasons in their answers, the sum of the percentages exceeded 100%.note: all the differences between trust and distrust are statistically significant (p < .05).
Table 1 shows that overall, predictability was the most commonly used judgement basis for trust and distrust, followed by intrinsic commitment.In other words, politicians seemed to be judged mainly on whether they were true to their words and whether their intentions were perceived as good.A nonnegligible proportion of respondents also mentioned the representatives' competence as a reason for trusting or distrusting politicians, while being responsive was mentioned by 12% of the respondents.Few mentioned extrinsic commitment as a reason for political trust or distrust.
In 30% of responses, reference was made to more than one of five main trust judgement categories.There were no clear patterns as to which categories tended to appear together.A factor analysis showed that none of the factors represented a common underlying dimension, which supports the theoretical argument that the categories are conceptually independent (see Tables A4-A6 in the Online appendix).
While our first research question concerned (dis)trust judgements in general, our second question concerned whether people judged political trust and distrust on different bases.As illustrated in Table 1, there were indeed significant differences in how often the different reasons for trust and distrust occurred in the material.Most notable are the differences in the two most frequent (dis)trust reasons: predictability and intrinsic commitment.Predictability was mentioned by more than half of the respondents describing trust but only by a little under a third of respondents describing distrust.For intrinsic commitment, the pattern was the opposite: it was used far more often to explain what created distrust.
As shown in Table 2, the frequency of reasons given for trusting or distrusting politicians remained significantly different after controlling for the personal characteristics of the respondents and for their level of trust.
This analysis indicates that people who were trusting of politicians were more likely to mention intrinsic commitment as a judgement basis.People distrustful of politicians were more likely to mention competence.Moreover, people with a higher level of education were more likely to mention intrinsic commitment.There were no significant differences between men and women or between different age groups.In the following, we discuss each of these categories in turn, unpacking their meaning and nuances and demonstrating how they are used differently to describe trust and distrust.Providing illustrating examples, our discussion centres on how the empirical data can inform and develop the theoretical conception of the five judgement bases presented in Table 1.

Predictability
Predictability, referring to the extent to which politicians' behaviour is in line with promises or expected actions (Ruokonen 2013; Van der Meer 2010) was by far the most cited basis for trust judgements and the second most cited basis for distrust judgements in the data set.It was also the easiest judgement base to identify, as a number of seemingly ready-made 'standard' phrases were used to convey the importance of predictability.
When describing reasons for trust, these included references to whether the politicians were reliable: whether they could, in the words of the respondents, be counted on to be 'true to their word' , 'keep their word' or 'stick to the party programme' .Predictability seemed to be the most easily accessible notion of how politicians should perform to be worthy of trust; they should simply act as they speak.When used to describe the reasons for distrust, predictability was mostly referred to in terms of what it was not.Respondents distrusted politicians who 'did not keep their word' or 'did not fulfil election promises' .Also frequently mentioned was an apparently taken-for-granted notion that politicians had a 'big mouth' and hence that their promises could not be trusted.They tended to 'promise more than they can keep' , and there was 'much talk, little action' .Unpredictability was also sometimes described by words or phrases assigned exclusively to unpredictable behaviour.For example, politicians were seen as untrustworthy in an unpredictable sense because they were 'wavering' or 'change [d] their mind all the time' .The absence of these negative behaviours, which conveyed reasons for distrust, was also used to describe trust.Some respondents, for example, trusted politicians who 'did not waver' .
Overall, predictability and unpredictability as reasons for trust and distrust seemed to be conceived of as negations: trustworthy politicians kept their promises, whereas untrustworthy politicians did not; untrustworthy politicians wavered, whereas trustworthy politicians did not.While, as we shall see, other categories could be divided into distinct subcategories, this was not the case for predictability.Rather, the simple concept of being true to one's word and doing as one said was phrased in different ways.

Intrinsic commitment
Intrinsic commitment implies that politicians wish to act in the best interests of citizens because they have citizens' well-being at heart (Van der Meer 2010).Assessment of politicians' trustworthiness is thus based on the intent, purpose and thought behind actions, and not the actions alone.As demonstrated in Table 1, intrinsic commitment was the second most cited reason for trust in politicians, and lack thereof was the most cited reason for distrust.Three distinct aspects of intrinsic commitmentintentions, morals and attitude -could be identified in the answers (see Online appendix, Table A2 for frequency of subcategories).Good intentions or motivations were cited as a reason for trust and bad intentions as a reason for distrust.Regarded as trustworthy were, for example, selfless politicians who 'wholeheartedly work [ed] for the good of the citizens'; deemed untrustworthy were politicians who did 'anything to gain power' or 'were motivated by personal gain' .A wrong intention that generated distrust was defined as more than the mere absence of a good intention: simply disregarding citizens' interests was seldom cited as a reason for distrust, and to be deemed untrustworthy, politicians needed to be actively motivated by the wrong reasons, such as their own personal gain.
A second subcategory of intrinsic commitment referred to the politicians' morals; that is, whether they complied with what was deemed morally correct behaviour by societal standards.Breaching moral standards were cited as reasons for distrust, whereas complying with moral standards was seldom cited as a reason for trust.Bad morals causing distrust were, for example, described as operating with 'double standards' or as politicians 'not abiding by the rules and laws they, themselves, had made' , such as 'when their personal life totally collides with their political principles' .Also frequently cited as reasons for distrust, which imply a breach of societal morals, were 'comradery' and the more severe 'corruption' , which not only breach morals but also legal requirements.
A third subcategory referred to whether the politicians had an appropriate attitude.While displaying an appropriate attitude could arguably be seen as a form of competence, it was in the answers coupled with a commitment to doing good for the citizens, as in 'leaving grudges or personal conflicts aside for the sake of the greater good' .This category included references to the willingness to cooperate or have a good tone with political opponents to get things done, as illustrated in this quote from one of the respondents: 'If the other political parties have better ideas that will actually be good for the citizens, I wish politicians would cooperate instead of clinging to their own policies' .Mostly, however, references to attitude and tone were used to describe the reasons for distrust.Politicians were distrusted, for example, because of 'hateful speech and negative characterisations of opponents' or 'fighting and brawling in the public debate' .
Interestingly, intrinsic commitment seemed to be assessed along slightly different axes when trust was under consideration compared to distrust.First, the object of intrinsic commitment varied between trust and distrust assessments.Trustworthy politicians were believed to be intrinsically committed to the good of the people, while untrustworthy politicians were believed to be intrinsically committed to the good of themselves.Second, merely disregarding the good of citizens was not sufficient to be deemed untrustworthy.The absence of good intentions, good morals and good attitude was also not enough to be distrusted.In the sense of intrinsic commitment, being untrustworthy required actively bad intentions, bad morals and a bad attitude.Trust and distrust, therefore, appeared to be more than mere negations when judged on the basis of intrinsic commitment.

Competence
Competence as a basis for political trust implies that citizens regard politicians as having sufficient competence to perform according to the subject's expectations and interests (Van der Meer 2010, 2017b).Such assessments may be related to perceptions of the political actors' competence as well as the perceived successes and failures of policy outcomes (Bertsou 2019; Tomankova 2019).Three subcategories were identified, the first one referring to politicians' abilities to perform their tasks as elected representatives.Abilities included such things as whether the representatives had a relevant education, sufficient experience or knowledge about the political system, specific policy issues or the local context.Abilities also referred to overall intelligence and aptness, negatively framed in the words of this respondent, who mistrusted politicians who were 'incompetent -lacking the ability to understand complex issues' .A second category concerned the representatives' ability to act -to make decisions and to see the decisions through to implementation, including the ability to deliver tangible results.The negative rendition of this category was that, in the words of the respondents, for example, politicians were 'unable to make up their mind' or 'never got anything done' .A third category subsumed under the competence heading was responsibility, referring to whether the representatives acted responsibly -whether they 'looked far ahead' , 'acted in a rational manner' or had the 'ability to make priorities' .Distrust ensues if politicians, for example, make 'economically unsustainable and flawed decisions' .
By and large, all three subcategories seemed to describe clean opposites: inability was described as the negation of ability, failure to implement as the negation of implementation and irresponsibility as the negation of responsibility.When competence was the basis for judgement, trust and distrust seemed to be assessed along the same lines.

Responsiveness
A number of statements concerned the perceived responsiveness of politicians, referring to whether people believe that politicians consider and react to their political signals (Denters 2002;Manin et al. 1999;Torcal and Christmann 2021).Statements in this category pointed to, for instance, whether representatives listened to citizens, whether they took citizens' opinions and interests into account in their policymaking and whether politicians were available if the citizens had something they wanted to impart.Most references to the importance of responsiveness were about the act of listening, which was identified as a distinct subcategory.Trustworthy politicians should 'listen to their citizens' , whereas distrust occurs when politicians 'did not listen' .Listening also entailed being accessible and available to the citizens, as in 'easy to talk to when you run into them' .Classified as a second subcategory were references made to the visibility of politicians: to whether they were visible and present in the local community and in policymaking.'Visible and present' politicians inspired trust; 'invisible' or 'withdrawn' politicians who did not participate in the public debate inspired distrust.A third subcategory refers to the comprehensibility of politicians.Incomprehensible politicians using 'political language' or saying 'blah blah' were perceived as unresponsive and untrustworthy, whereas politicians communicating 'in a language that people understand' were perceived as responsive and trustworthy.Finally, a fourth subcategory of responsiveness included references to sociodemographic representativeness, addressing the relationship between citizens and their representatives.Unlike the other references to responsiveness, these references were not about how the representatives acted but about how their ability to be responsive was a product of who they were.Politicians were deemed more trustworthy and better able to understand the wants and needs of their constituents if they were, for example, born in the municipality or not too young or old.They were deemed untrustworthy if they were 'out of touch with ordinary people' or what in Norwegian are called 'political broilers' -young politicians with no experience in the 'real world' .
While references to responsiveness were made more often when describing reasons for trust than for distrust, responsiveness appeared to be defined along the same axis.Responsive politicians inspiring trust listened and were visible, comprehensible and socially representative; unresponsive politicians did not listen and were invisible, incomprehensible and socially unrepresentative.

Extrinsic commitment
Extrinsic commitment refers to whether citizens have the means to control politicians and hold them accountable through, for instance, procedures that enable publicity and empower monitoring (Van der Meer 2010, 2017b).In our study, a few respondents judged politicians based on their extrinsic commitment.The way the questions were posed -concerning what generates trust in the elected representatives rather than in institutionsprobably did not encourage people to come up with answers relating to extrinsic commitment.Still, the importance of transparency was emphasised by a number of respondents.When used as an argument for distrust, references were made to nontransparency and to non-transparent practices, for example, when decisions were made 'in the backroom' or 'shielded from public scrutiny' .When used to describe trust, respondents frequently mentioned the lack of nontransparency, for example, the importance of decisions 'not being made in the backroom' .However, some also worded transparency in a positive way by, for example, pointing to the importance of 'the press monitoring politicians' , which generated trust.Overall, however, extrinsic commitment seemed to be conceived along the same axis when assessing both trust and distrust.

Discussion and conclusion
In this article, we have explored the bases on which people make judgements about whether to trust or distrust politicians.Contrary to previous research, which almost exclusively used close-ended survey questions to capture people's reasons for (dis)trust, we analysed two open-ended questions about reasons for trust and distrust from a representative Norwegian population survey.Our theoretical point of departure was a five-fold typology of judgement bases for political trust that discerned between competence, intrinsic commitment, predictability, responsiveness and extrinsic commitment.
The first main conclusion of the article is that these five judgement bases capture the vast majority of people's (dis)trust judgements and hence provide a useful point of departure for future analysis of the reasons for political trust and distrust.Predictability and intrinsic commitment were the judgement bases most frequently applied, but all five bases were present.The least referred to judgement base was extrinsic commitment.We believe that this category, while valuable in research on trust in political institutions, is less suited to studies of trust in political actors.External control mechanisms may come to mind easily when evaluating trust in institutions but seem to be less relevant when assessing trust in political actors, who are evaluated mainly on the basis of their actions, perceived intentions and personal qualities, and rarely according to whether they are kept in check by political institutions.This illustrates the more general point that the basis of judgement is likely to vary depending on the object of the evaluation.
Moreover, the analysis revealed that the five overarching and somewhat abstract judgement bases could be divided into more concrete and tangible subcategories.The subcategories provide a means of analysing and identifying the different trust judgement bases and serve as operationalisations of the main typology.Furthermore, they contribute to nuancing the content, revealing different aspects of the main categories and demonstrating the different ways in which the overarching judgement bases are applied in practice.
Importantly, the analysis shows that people differ according to what they emphasise when assessing the trustworthiness of politicians.A certain behaviour or attitude may inspire trust or distrust in one voter but leave another voter unmoved.Furthermore, some of the judgement bases are incompatible in the sense that a behaviour that inspires trust in one voter inspires distrust in another.For example, to some citizens, being clear, principled and following the party programme are trust-generating practices, and those who make compromises are accused of being turncoats and breaking their promises.In contrast, others regard the willingness to listen, cooperate and find compromises as the proper attitude for politicians, while those who maintain their principles are criticised for 'riding their hobby horses' .Some respondents indicated listening to voters as the main basis of trust, while others discarded this as irresponsible populism.Because citizens give weight to different and sometimes conflicting judgement bases, it seems impossible to gain trust from the entire electorate.
Our second main conclusion is that citizens give different weights to the five judgement bases when evaluating trust and distrust.The main tendency is that while trust is most often evaluated with reference to predictability, distrust is most often evaluated with reference to intrinsic commitment.To put it simply, when people evaluate whether they should trust a politician, they first check whether the politician acts in accordance with her words; when they evaluate whether they should distrust a politician, they assess her intentions.
Although differently weighted, for four of the five judgement bases, trust and distrust seemed to be assessed along the same axes.Mostly, predictability spurred trust, while unpredictability spurred distrust.Likewise, competence, responsiveness and extrinsic commitment generated trust; incompetence, unresponsiveness and lack of extrinsic commitment generated distrust.For these four judgement bases, behaviours and intentions generating trust appeared to be the default -they were somewhat 'positively' conceived of -and distrust was associated with a lack of positive behaviours and intentions.Intrinsic commitment diverged from the other four judgement bases in that trust and distrust were not, to the same extent, assessed along the same axis, and judgements of distrust were based on assessments other than the mere lack of intrinsic commitment.For a politician to be distrusted, it was not sufficient to be intrinsically uncommitted to the interests of the citizens; untrustworthy politicians were described as being actively committed to themselves and their own interests.For intrinsic commitment, therefore, the contents of the axis poles did not mirror each other.
An important theoretical implication of our findings is that trust and distrust seem to be more than pure negations.First, because people give different weights to the different judgement bases when evaluating trust and distrust, trust assessments along some axes would affect trust more than distrust or vice versa.For instance, being predictable seems to build trust more than being unpredictable seems to breed distrust.Second, one of the important trust judgement bases, intrinsic commitment, does not seem to be entirely bipolar, as trust and distrust are judged based on evaluation of intrinsic commitment towards different objects -people and oneself, respectively.Thus, this asymmetry suggests that the cognitive processes involved in making judgements about trust may differ from those involved in making judgements about distrust.While the shared evaluation basis indicates that trust and distrust are not necessarily different concepts, the evaluative processes through which people assess whether to trust or distrust politicians may give precedence to different evaluative criteria.
A political implication of these findings is that the way politicians are presented through the media or the way they present themselves matters differently in terms of building trust and breeding distrust.If, for example, the media, in its watchdog function, directs its attention towards whether politicians are looking out for themselves (intrinsic commitment), this might have a greater effect on the level of distrust than if the media directs its attention towards whether politicians keep their promises (predictability).Likewise, a politician who wants to be trusted might gain more by emphasising how she keeps her promises than she would by emphasising that she cares about the welfare of all citizens.
The analysis of what groups of citizens were most inclined to use the different judgement bases indicated a difference between those who were trusting and distrusting of politicians.Whereas the trustful were more likely to mention intrinsic commitment as a judgement basis, the distrustful were more likely to mention competence.A possible interpretation of this finding is that competence represents a sort of minimum requirement for trust.People can only afford to care about the other bases for trust if they believe that politicians are indeed capable of managing the community's common resources in the citizens' best interests.In the same vein of reasoning, intrinsic commitment might be thought of as a judgement basis that is relevant only when all other trust requirements are met.Consequently, people who take for granted that politicians are competent, predictable, extrinsically committed, and responsive might assess the politicians' internal commitment.
The analysis also indicated a difference in judgement between different social groups: those with higher levels of education were more inclined than those with lower levels of education to make (dis)trust judgements based on the evaluation of intrinsic commitment.Politicians who want to earn the trust of voters with a high level of education would therefore do well to emphasise their intrinsic commitment, whereas this strategy would work less well with voters with a lower level of education.
The different wordings of the trust and distrust question require further reflection.While the trust question was worded in a way that invited the respondents to provide personal reasons for trust, the distrust question was worded in a way that invited the respondents to provide (their beliefs about) more general reasons for distrust.If interpreted in this way by the respondents, the observed differences between trust and distrust judgements might be understood differently.For example, we found that the most frequently cited reason for trust was predictability, whereas the most frequently cited reason for distrust was intrinsic commitment.If, as a consequence of the differently devised formulations, the trust question was perceived as being about one's own and the distrust question about others' trust evaluation basis, this result would signify that people perceive others' trust judgements to be more driven by beliefs about politicians' motivations and intentions, while they perceive their own judgements to be more driven by results and especially by whether they received what they were promised.In other words, others are concerned with the politicians' intentions, while I am concerned with whether I get what I was promised.However, since we cannot know whether this was the case, we will have to leave it for future studies to investigate.Notably, asking explicitly about one's own as compared to (beliefs about) others' reasons for trust and distrust might contribute to clarifying this point.
Finally, to what extent can our findings be generalised to other contexts beyond Norwegian local politics?The fact that our survey questions concerned (dis)trust in local politicians may have implications for whether the results can be generalised to political actors on other levels.On the one hand, unlike national politicians, most local politicians do their political work in their spare time, and their messages are seldom polished by communication advisers.Accordingly, they are less likely to be criticised for losing touch with ordinary people or not being true to themselves.On the other hand, local politicians often make decisions that may directly affect their friends and neighbours.Therefore, there is reason to believe that they, to a larger extent, are subject to questions of impartiality and conflicts of interest (intrinsic commitment).Likewise, country-specific factors may certainly influence the results, as people's judgement bases presumably depend on the political context.For instance, that a judgement basis is not mentioned by a respondent does not necessarily mean that it is deemed unimportant; it could mean that it is taken for granted in a specific context.When people say that they trust politicians because they are competent, other motivations or behaviours may be implied: They assume that politicians are intrinsically committed and will keep their word if they can.Likewise, competence would perhaps be mentioned more often in contexts where the media focus is on the incompetence of politicians, corruption (categorised as morals) would perhaps be mentioned more often in contexts where corruption is widespread, responsiveness in a context with particularly unresponsive politicians, and so on.
In other words, the conceptualisations of trust and distrust may be affected by political institutions and the political culture within a given political system.These factors may condition the expectations of how politicians should behave.Moreover, voters have -to a greater or lesser degree -their own personal experiences with the political institutions, which will influence how they assess politicians.We can assume that the context of Norwegian local politics has influenced the incidence of different judgement bases.A high-trust country with relatively little corruption and relatively well-functioning political institutions, together with the smaller distance between citizens and politicians (for better and for worse) that characterises local politics, will affect the balance between trustful and distrustful citizens -and probably also the distribution between the different bases of trust assessment.Our main point, however, is that there is no reason to believe that the typology itself is context specific.On the contrary, it can be a useful tool for comparisons of how trust (or distrust) is developed in different political systems.
A task for future research, therefore, would be to assess the relative contribution of the different judgement bases in different political contexts -in other countries as well as at the national level.Moreover, future research might make further contributions by studying more closely the importance of different judgement bases for trust and distrust.This could, for example, be accomplished through conjoint experiments where people rate the degree to which they trust and distrust politicians with different combinations of predictability, intrinsic and extrinsic commitment, competence and responsiveness.

RQ 1 :
On what basis do people judge whether they trust or distrust politicians?

Table 1 .
prevalence of different judgement bases for trust and distrust (in percent).

Table 2 .
logit analysis of different judgement bases for trust and distrust (odds ratio).