The Trade-Off Between Political Trust and Political Participation

Abstract

This article explores the possibility of a trade-off between citizens’ trust in government and their rates of political participation. Trust and political participation are both of interest in political science, but the relationship between the two remains unclear. Drawing on political trust and political participation literature, the article utilises data from rounds 1–9 of the European Social Survey to investigate whether a trade-off between political trust and political participation can be found. The results suggest that although greater trust in government is variably predictive of either higher or lower rates of political participation, rates of participation reach a peak at a particular level of trust, followed by a decline. This drop-off effect suggests that citizens may become complacent and disengaged as they reach high levels of trust. The trust-participation trade-off illustrates one way in which policy choices may have unexpected consequences which policymakers should consider.

Keywords

political trust political participation quantitative methods

Introduction

Do trusting citizens continue to actively engage with political action, or does higher trust in government lead to political complacency? A wealth of literature has highlighted the challenges faced by modern democracies with regard to maintaining a politically engaged citizenry (e.g. Bakardjieva Engelbrekt et al., 2019; Kaase, 1999; Sloam, 2016). Both low political trust and low levels of political participation behaviour by citizens have been identified as key challenges for contemporary states (Chen et al., 2021; Hooghe et al., 2017; Hooghe and Marien, 2013). In recent decades, there has been an increased focus in public governance on improving these key metrics (Koivula et al., 2022).

It is therefore of interest to examine the relationship between trust and political participation. A number of authors have found that higher political trust is predictive of higher rates of political activity, either in the form of voting (Koivula et al., 2022) or other forms of participation (Hooghe and Marien, 2013; Kaase, 1999). However, the exact nature and shape of this relationship has received limited attention in the literature beyond the associations between the two metrics (Devine, 2024).

This article focuses on political trust as a predictor of citizens’ rates of political participation. Drawing on political trust and political participation literature (James and Moseley, 2014; Porumbescu, 2017; Van Ryzin, 2007), the article argues that trust and participation may be in tension with one another. As argued by Chen et al. (2021), very low levels of trust may result in political apathy, and very high levels of political trust may lead to political complacency. This article proposes that there may be a trade-off between trust and political participation, whereby trust in government predicts higher rates of participation at lower levels of trust and lower rates of participation at higher levels.

Much of the available literature focuses on the existence and direction of an association between political trust and political participation, and there is indeed compelling evidence that trust and participation are related (Devine, 2024; Hooghe and Marien, 2013; Kaase, 1999). However, the findings on the association are mixed, with differing results for different types of political participation (Hooghe and Marien, 2013), level of institution (Chen et al., 2021), political context (Braun and Hutter, 2016) and citizens’ individual characteristics (Crepaz, 2017; Michalski et al., 2023). In addition, existing literature does not differentiate between effects at different levels of trust and often assumes an implicitly linear relationship between the two, which ignores the potential complexity of the relationship (Chen et al., 2021; Hooghe and Marien, 2013; Kaase, 1999). Thus, information on any variance in the relationship between different levels of trust is lost.

This article looks to explore the relationship between trust and political participation further by examining the effects of trust on participation at different levels of trust. The article uses data from the 2002–2018 waves of the European Social Survey (ESS). The results suggest that a trust-participation trade-off exists when examining generalised political trust and trust in the police and the legal system. Results when examining trust in national political actors or trust in international political actors were mixed. The effect was most clear when voting behaviour was included in the analysis. The effect was clearer for institutionalised participation than non-institutionalised participation. The results thus provide limited support for the trust-participation trade-off.

Literature Review

The evidence on the direction of any association between political trust and political participation is inconclusive – there are circumstances where higher trust has been found to be associated with higher rates of political participation, as well as circumstances where it has been found to predict lower rates of political participation. The most common explanation for this (e.g. Hooghe and Marien, 2013; Kaase, 1999) is that trust has different and opposite relationships with institutionalised and non-institutionalised political participation. However, the relationship may also vary at different levels of trust – at low levels of political trust, citizens may view political participation as futile, and at high levels of trust, they may view it as unnecessary (Chen et al., 2021). It would thus be expected that a decline in political participation – or a trust-participation trade-off – would be observed at high levels of trust. However, the existence of such a trade-off has not previously been directly investigated. Much of the literature on the relationship between trust and political participation assumes a linear relationship between the two (e.g. Hooghe and Marien, 2013; Kaase, 1999; Li and Yu, 2023; Stals et al., 2022), which does not allow for examination of effects at different levels of trust. This section will provide an overview of existing research on political trust and citizens’ rates of political participation.

Political Trust

Political trust has received a great deal of attention in political science, with much of the focus being on the trend of and reasons for declining trust among western democracies (e.g. Kaase, 1999; Marien, 2011; Norris, 2022). The importance of trust is intuitive – the political mandate of the government assumes a certain level of trust from its citizens, as does the state monopoly on violence (Bakardjieva Engelbrekt et al., 2019). As such, political trust can be viewed as the expectation that the state will not abuse its power, and that citizens have the opportunity to influence politics should they disagree with state action or inaction (Norris, 2022). Political trust is distinct from social trust or interpersonal trust, meaning trust in society or in other people (Kaase, 1999; Norris, 2022). Norris (2022) argues that politically conscious citizens should approach state actors with a degree of scepticism, and that ‘blind credulity’ or blind trust are as counterproductive as excessive cynicism for citizens when evaluating government performance. Both cynicism and credulity may be of concern if they result in a politically disengaged citizenry.

There has been extensive discussion in recent literature on the most appropriate ways to operationalise and measure trust (e.g. Hooghe, 2011; Marien, 2011). As identified by Devine (2024), the multidimensional aspects of trust argued for by a number of authors (e.g. Grimmelikhuijsen and Knies, 2017; Hamm et al., 2023; Norris, 2022) have largely not been adapted into the measurements employed for trust in available survey data, which generally employ a more limited conceptualisation and indexed operationalisation of trust. This is in part due to practical limitations, as trust is often included in surveys as part of a broader battery of political questions, as is also the case in the ESS data used in this article. In addition, much of the available data on trust comes from European or North American research, potentially limiting the applicability of the data in different contexts (Devine 2024; Marien, 2011).

Political Participation

Political participation can be viewed as citizens making use of the opportunities afforded to them to affect political decision-making and can take a variety of forms. As identified by Hooghe and Marien (2013) and Kaase (1999), these are typically categorised as institutional, such as voting, or non-institutional, such a protesting. Institutional forms of political participation represent the standard forms of participation opportunities provided to citizens by the state, most typified by voting and campaign activities (Hooghe and Marien, 2013). Voting is the most common and most fundamental form of political activity available to citizens and has been extensively studied as an outcome of interest (e.g. Hooghe and Marien, 2013; Koivula et al., 2022; Li and Yu, 2023). Democratic systems rely on sufficient voter turnout to establish a political mandate for state actors, making voting rates crucially important. Other forms of institutional political participation involve activity directly relating to the official electoral process, such as contacting officials or participation in campaigns (Hooghe and Marien, 2013). These forms of participation can be considered secondary to voting in that they only indirectly relate to the political process.

Non-institutional forms of participation, by contrast, represent the forms of political participation that are not directly sanctioned by the state (Kaase, 1999). Non-institutional participation is closely related to political voice and exit, based on the ‘exit, voice and loyalty’ framework introduced by Hirschman (1970). In the context of a state, citizens are not able to opt out of the restrictions and obligations set on them by the state, making their opportunities for exit inherently limited (Hooghe et al., 2011). Thus, the ways in which they are able to express dissent or disagreement with the state are limited to forms of political participation that convey these sentiments, such as protesting, political organisation and protest voting. The opportunities citizens have for these activities and the rate at which they engage in them may be key in understanding their relationship to the state (De Blok and Kumlin, 2022; Nakamura and Kim, 2010).

The Relationship Between Political Trust and Political Participation

There has been some inconsistency in the literature on the precise relationship between political trust and political participation (Devine, 2024). Hooghe and Marien (2013) argue that high trust can be seen as a prerequisite for participation, but that low trust can also be a motivating factor for protest behaviour and other non-institutional forms of political participation. A common observation in the literature is that low political trust contributes to higher rates of non-institutionalised participation and lower rates of institutionalised participation (Braun and Hutter, 2016; Hooghe and Marien, 2013; Kaase, 1999). The evidence for the relationship is somewhat inconclusive, though the literature does largely support this view (Braun and Hutter, 2016; Li and Yu, 2023; Michalski et al., 2023). The effects of trust on participation may also vary by individual characteristics and trust level. Ruelens and Nicaise (2020) found that the individual profile of citizens predicted their patterns of trust in national and international political institutions. Similarly, Crepaz (2017) found that citizens with higher in-group trust were more likely to vote and that citizens with higher out-group trust were more likely to engage in non-institutional political activities. This supports the notion that political participation may shift from institutional to non-institutional forms of participation.

By contrast, the relationship between political trust and voting behaviour is quite consistent. Higher political trust has generally been found to predict higher voting rates (Devine, 2024; Hooghe and Marien, 2013; Kaase, 1999). However, this may to a degree be context dependent. Examining the Belgian case, Hooghe et al. (2011) found that in a system in which citizens are compelled to vote, distrusting voters are significantly more inclined to cast a blank or invalid vote. In a system in which voting is compelled, blank or invalid votes may be taken to represent self-selected non-participation by citizens.

There are a number of factors for which it may be necessary to account when examining the relationship between trust and political participation. Potentially relevant covariates include age (Sloam, 2016; Stals et al., 2022), social and generalised trust (Hooghe and Marien, 2013) experiences with public services (Holbein, 2016; James and Moseley, 2014) and the efficacy of public communication by state actors (Porumbescu, 2017). In addition, the overall political environment has been highlighted as important. Political engagement has been found to reflect local political culture (Sloam, 2016), suggesting that the relationship between trust and participation may be impacted by the political context. Similarly, perceived democratic quality may be relevant, with the effects of both mistrust and high trust on political participation being higher in political contexts where the quality of democratic institutions is considered high by citizens (Braun and Hutter, 2016; Camaj, 2014; Ceka, 2012). The influence of the political environment on trust has been found to be level-specific, meaning the trust citizens express in different institutions does not consistently carry over across institutions and levels of government (Huhe and Chen, 2024; Proszowska et al., 2022). However, trust in different levels of government has also been found to be highly associated (Chiru and Gerghina, 2012), meaning there is some spill-over effect.

The state of the literature on the relationship between political trust and political participation remains inconclusive. In a meta-analysis on the effects of trust on a variety of factors including political participation, Devine (2024) found that trust had no effect on informal political participation such as protesting and a small positive effect on voting rates. While a great deal of attention has gone into the existence or direction of the relationship between trust and different types of political participation, very little attention has been given to the shape of any relationship. This article looks to address that gap by utilising data from waves 1–9 of the ESS. The inconsistent and at times conflicting results from previous research underline the complexity of the relationship. The focus in this article will be on examining whether previous differing results can be reconciled by further insight into the shape of the association between trust and participation.

Data and Methods

The study utilised data from rounds 1–9 of the ESS, collected between 2002 and 2018. These waves were selected due to the consistent items measuring political participation in these surveys. Waves 10 and 11 were excluded from the analysis due to missing the questions ‘Worked in political party or action group last 12 months’ and ‘Worked in another organisation or association last 12 months’. The trust and political participation items included were those included in every survey in waves 1–9. The pooled dataset included data from 38 different countries and had a total sample size of 430,870. All data analysis was performed using IBM SPSS 29.

The dataset included eight items measuring political participation (Table 1). Each of the questions were binary variables that measured whether or not the respondent had engaged in the given activity in the previous 12 months, with the exception of voting, which measured whether the participant had voted in the most recent national election. The variable for voting was recoded to exclude participants who were not eligible to vote in the given election.

Table 1.

Forms of Political Participation Included.

Contacted politician or government official last 12 months (institutional)

Worked in political party or action group last 12 months (institutional)

Voted last national election (institutional)

Worn or displayed campaign badge/sticker last 12 months (institutional)

Signed petition last 12 months (non-institutional)

Taken part in lawful public demonstration last 12 months (non-institutional)

Boycotted certain products last 12 months (non-institutional)

Worked in another organisation or association last 12 months (non-institutional)

As this article is interested in the effects of trust on the overall level of political activity by respondents, the eight items included in the study were recoded into four count variables measuring political participation. These recorded the number of activities the respondent had engaged in during the previous 12 months. These were the overall number of activities performed both excluding and including voting, the number of institutional activities performed and the number of non-institutional activities performed. The institutional and non-institutional coding for the forms of participation included is presented in Table 1 and descriptive statistics for the outcome variables are presented in Table 2.

Table 2.

Activity Descriptive Statistics.

	Activity count, without voting	Activity count, including voting	Institutional activity count	Non-institutional activity count
Valid N	430,030	430,634	430,612	429,771
Mean	0.8324	1.5409	0.9655	0.5766
Mode	0	1	1	0
Standard deviation	1.27907	1.42056	0.77646	0.91357
Variance	1.636	2.018	0.603	0.835

The different forms of political participation were recoded into institutional and non-institutional forms of participation using the categorisation employed by Hooghe and Marien (2013: 138) and Kaase (1999), with the addition of voting being categorised as institutional. The aggregated count variable was examined both excluding and including voting as voting has previously been identified as a distinct form of political participation from the other forms included (Devine, 2024). Voting also contributed a disproportionate amount of the overall political participation by respondents, potentially skewing results for the count variable – voting alone accounted for approximately half of the political activity in which citizens engaged.

Exploratory Factor Analysis

Each wave of the survey included seven items measuring political trust (Table 3). Each of these items was recorded on an 11-point ordinal scale of 0 (no trust at all) to 10 (complete trust). Three relevant types of trust were identified to have been included in the survey: trust in legal authority, trust in national political actors and trust in international political actors. These categories are similar to the distinctions made by Norris (2022) and consistent with previous analysis on ESS data by Marien (2011).

Table 3.

Factor Analysis Factor Loadings, Rotated and Unrotated. Factor loadings above 0.6 highlighted.

Item	Unrotated component matrix			Rotated component matrix
Item	Factor 1	Factor 2	Factor 3	Factor 1	Factor 2	Factor 3
Trust in the legal system	0.803	-0.313	0.267	0.424	0.225	0.764
Trust in the police	0.711	-0.355	0.496	0.209	0.201	0.891
Trust in politicians	0.87	-0.106	-0.371	0.874	0.265	0.265
Trust in the European Parliament	0.752	0.538	0.062	0.357	0.843	0.146
Trust in the United Nations	0.731	0.526	0.258	0.207	0.872	0.271
Trust in political parties	0.858	-0.059	-0.394	0.873	0.292	0.218
Trust in country’s parliament	0.849	-0.168	-0.167	0.73	0.256	0.422

Extraction method: principal component analysis.

Rotation method: Varimax rotation with Kaiser normalisation.

Exploratory factor analysis was performed to examine this factor structure. Eigenvalues and the percentage of variance explained by factor number is presented in Table 4. Based on the Kaiser criterion of a cut-off point of eigenvalues below 1, a one-factor solution would be the most appropriate (second factor eigenvalue = 0.833). However, there is also a notable elbow in both the eigenvalues and in the variance explained between components three (eigenvalue = 0.709) and four (eigenvalue = 0.341). A three-factor solution is also consistent with theoretical expectations. In the rotated factor solution, the trust items included in the survey were found to have loaded on the three factors in the manner described above, with no strong overlapping loadings. As such, the three-item factor solution was used for the study. The recoded trust categories and their Cronbach’s alphas are presented in Table 5.

Table 4.

Factor Analysis Eigenvalues and Variance Explained.

Component	Eigenvalue	Percent of variance explained
1	4.463	63.757
2	0.833	11.905
3	0.709	10.128
4	0.341	4.872
5	0.271	3.878
6	0.259	3.698
7	0.123	1.762

Table 5.

Trust Categories.

		Cronbach’s alpha
Trust in the police and the legal system	Trust in the legal system	0.798
Trust in the police and the legal system	Trust in the police	0.798
Trust in national political actors	Trust in politicians	0.911
	Trust in political parties
	Trust in country’s parliament
Trust in international political actors	Trust in the European Parliament	0.830
Trust in international political actors	Trust in the United Nations	0.830
All trust items		0.904

The seven trust items were recoded into three composite variables while maintaining the original 11-point ordinal measurement scale. In line with previous observations that a unidimensional measure of political trust may be most appropriate and is most commonly employed in the literature (Devine, 2024; Hooghe, 2011; Marien, 2011), a fourth composite variable comprising all seven trust items was created, also maintaining the 11-point ordinal scale.

Negative Binomial Regression Analysis

This article focuses on the aggregated level of political participation by respondents. As such, the outcome variable of interest in the study was the number of political activities respondents had engaged in. As shown in Table 2, the outcome variable was unsuitable for Poisson regression due to overdispersion, with a variance exceeding the mean for three out of four outcome variables. As such, negative binomial regression was used as the data analysis method. Negative binomial regression can be used to model overdispersed count data (Lawless, 1987). The negative binomial regression model was calculated using robust standard errors.

Estimated marginal means were extracted from the regression model using an 83.4% confidence interval rather than the standard 95% interval. This was done to make the comparison of the confidence intervals more meaningful – when using an 83.4% confidence interval, when the confidence intervals of two means do not overlap, the means can be considered different with a confidence level of approximately 95% (Austin and Hux, 2002; Knol et al., 2011; Payton et al., 2003).

The estimated marginal means represent the expected mean number of activities respondents partook in at a given level of trust for each classification of political participation. When comparing groups, any other predictors included in the model are held constant. This allows the shape of the relationship between trust and political participation to be examined on an ordinal scale while controlling for extraneous variables.

The analysis included a number of control variables. The control variables included in the study are presented in Table 6. As the dataset consisted of a pooled dataset over a prolonged period, the ESS wave of the respondent was included as a control to account for clustering by time period. Citizens’ generalised social trust has previously been identified as a potentially relevant factor influencing the relationship between political trust and political participation (Hooghe and Marien, 2013), as has their interest in politics (Quaranta et al., 2021). These were also included as control variables in the study.

Table 6.

Control Variables Included in the Analysis.

ESS round

Country

Most people try to take advantage of you, or try to be fair (social trust 1)

Most of the time people helpful or mostly looking out for themselves (social trust 2)

Most people can be trusted or you can’t be too careful (social trust 3)

Interest in politics

Gender

Age

Education

Citizens’ descriptive characteristics have also been identified as potentially relevant in the relationship between political trust and political participation. The most consistently highlighted variables were the respondents’ country (e.g. Ceka, 2012; Chiru and Gerghina, 2012; Sloam, 2016) and demographic characteristics. For demographics, the variables used were the respondents’ age, gender and level of education. Each of these factors has been previously identified as potentially significant for the relationship between political trust and political participation in the literature (e.g. Fung et al., 2022; Hu, 2016).

Results

The focus in this section is on the estimated marginal means of the negative binomial regression models. As discussed in the ‘Data and Methods’ section, the particular area of focus in this article was whether a differential effect of trust on political participation was present at different levels of trust, potentially reconciling disparate results in the existing literature. The main focus of the results, therefore, is the overall shape of the relationship between trust and participation, rather than the existence or direction of any association. If political participation increases with political trust, peaks and declines, this would be indicative of a trust-participation trade-off. Each regression model presented in this section includes the control variables presented in Table 6. The estimated marginal means for the negative binomial regression analysis are plotted in Figures 1 –4. The estimated marginal means are also presented numerically in Supplemental Appendix A, aggregated variables expressed as binary variables in Supplemental Appendix B and individual means in Supplemental Appendix C. Visualisations for binary and individual variables are presented in Supplemental Appendix D.

Figure 1.

Negative Binomial Regression Results, Including Voting.

Figure 2.

Negative Binomial Regression Results, Not Including Voting.

Figure 3.

Negative Binomial Regression Results, Institutional Political Participation.

Figure 4.

Negative Binomial Regression Results, Non-Institutional Political Participation.

Political Participation Including Voting

When including voting, the results for all trust items trust in the police and the legal system and trust in national political actors indicate a trade-off between political trust and political participation. In each of these cases, the results are characterised by a sharp increase in predicted participation from trust = 0 to trust = 1, followed by a plateau and decline in predicted participation from trust = 6 onwards. In each case, predicted participation was lowest at trust = 10. For all trust items, predicted participation at trust = 10 was significantly lower than at any other level of trust. For trust in the police and the legal system, predicted participation at trust = 10 was significantly lower than predicted participation at any other level of trust apart from trust = 0. For trust in national political actors, predicted participation at trust = 10 was significantly lower than predicted participation at any other level of trust apart from trust = 9.

The results for trust in international political actors indicate no clear relationship. The lowest level of predicted participation was at trust = 0, though this difference is not significantly different from predicted participation at trust = 5 or trust = 10. Apart from trust = 0 and trust = 5, other predicted participation levels of trust were not significantly different from one another.

Political Participation Not Including Voting

When not including voting, the results for all trust items and trust in the police and the legal system indicate a trade-off between political trust and political participation. For both items, there was an increase in predicted participation from trust = 0 to trust = 1, followed by a decline in predicted participation as trust increased. The lowest level of predicted participation for both items was at trust = 10. For both items, predicted participation at trust = 10 was significantly lower than predicted participation at each other level of trust.

The results for trust in national political actors indicate a broadly linear negative relationship between trust and rate of participation when not including voting. A drop-off in predicted participation followed by a nonsignificant increase can be observed between trust = 4 and trust = 6. Predicted participation at trust = 10 was significantly lower than all levels of trust except trust = 9.

The results for trust in international political actors indicate no clear relationship. Predicted participation was significantly lower at trust = 5 than at other levels of trust. Other predicted rates of participation were broadly not significantly different from one another, with predicted participation at no level except trust = 5 being significantly different from predicted participation at trust = 0 or trust = 10.

Institutional Political Participation

When examining institutional political participation, the results for all trust items indicate a trade-off between political trust and political participation. The results for this item are characterised by an increase in predicted participation from a low at trust = 0 until trust = 4, followed by a plateau and a decrease in predicted participation from trust = 8 to trust = 10. Predicted participation at trust = 0 was significantly lower than predicted participation at any other level of trust. Predicted participation at trust = 10 was significantly lower than predicted participation at trust = 3 to trust = 8.

The results for trust in national political actors indicate a broadly linear positive relationship between trust and rates of participation, with a nonsignificant drop-off in predicted participation at high levels of trust. Predicted trust increased from a low at trust = 0 to a peak at trust = 8, followed by a notable but nonsignificant decline at trust = 9 and trust = 10. Predicted participation at trust = 0 was significantly lower than predicted participation at any other level of trust.

The results for trust in international political actors indicate no clear relationship. Predicted participation at trust = 0 was significantly lower than predicted participation at any other level of trust. Other predicted rates of participation were not significantly different from one another.

Non-Institutional Political Participation

When examining non-institutional political participation, the results for all trust items and trust in the police and the legal system indicate a trade-off between political trust and political participation. In both cases, predicted participation increased sharply from trust = 0 to trust = 1, followed by a linear decline with a slightly larger drop-off in predicted participation from trust = 9 to trust = 10. For both items, predicted participation at trust = 10 was significantly lower than predicted participation at each other level of trust.

The results for trust in national political actors indicate a broadly linear negative relationship between trust and rates of participation. Predicted participation decreased in a linear fashion from a peak at trust = 0, with the lowest level of predicted participation at trust = 10. Predicted participation at trust = 10 was significantly lower than predicted participation at each other level of trust.

The results for trust in international political actors could be interpreted as supporting either a drop-off in predicted participation or a linear negative relationship due to high error rates. There was a nonsignificant increase in predicted participation from trust = 0 to trust = 1, followed by a largely linear decline in predicted participation with the exception of particularly low predicted participation at trust = 5. The lowest level of predicted participation was lowest at trust = 10, with an estimated rate significantly lower than predicted participation at trust = 0–4 and trust = 6.

Discussion

The results provide limited support for the trust-participation trade-off. When examining all trust items and trust in the police and the legal system, rates of political participation increased from very low levels of trust, peaked and declined at high levels of trust. The trade-off was most apparent when all forms of political participation were included in the analysis and when examining institutionalised forms of political participation, though a trade-off was also apparent when not including voting and when examining non-institutional forms of political participation. Results were more mixed when examining trust in national political actors and trust in international political actors. The results suggest that the overall effect of citizens’ political trust on their rate of political participation varies by trust level, potentially reconciling some of the inconsistency in the existing literature (e.g. Chen et al., 2021; Devine, 2024; Hooghe and Marien, 2013; Stals et al., 2022). Notably, the trade-offs present in the data, particularly when not including voting in the analysis, are sufficiently small that they may not result in nonsignificant results in linear analyses. The results thus partially explain the inconclusive results found by Devine (2024).

For each form of political participation examined, there was either no relationship with trust in international political actors or a relationship consistent with other forms of trust but with greater levels of uncertainty. This is consistent with the observation by Hooghe (2011) that citizens may not differentiate between different political institutions as effectively as is often assumed. It may be that citizens had a greater degree of uncertainty in their trust evaluations of international institutions, resulting in spill-over effects in trust and greater uncertainty in the effects of trust on participation (Chiru and Gerghina, 2012; Huhe and Chen, 2024; Proszowska et al., 2022). Between the institution-based operationalisations of trust, trust in the police and the legal system had the most consistent trade-off with political participation. This is notable theoretically in that these institutions represent the state security apparatus. Thus, citizens with high trust in these particular institutions may represent citizens exhibiting complacency over the state monopoly on violence (Norris, 2022). The presence of the trust-participation trade-off for trust in these institutions is also notable empirically since, as shown in Table 7, trust in these institutions most closely conformed with a normal distribution of the trust variables examined here, creating more possibilities for the trust-participation trade-off to manifest.

Table 7.

Recoded Predictor Distributions.

Trust level	All trust items		Trust in police and the legal system		Trust in national political actors		Trust in international political actors
Trust level	N	Percent	N	Percent	N	Percent	N	Percent
0 (no trust at all)	12,705	2.9	15,771	3.7	43,358	10.1	26,195	6.1
1	23,706	5.5	12,991	3	38,950	9	17,336	4
2	37,792	8.8	20,254	4.7	49,430	11.5	24,942	5.8
3	54,337	12.6	33,678	7.8	56,725	13.2	36,996	8.6
4	67,795	15.7	41,693	9.7	60,256	14	45,280	10.5
5	80,903	18.8	61,034	14.2	68,978	16	75,948	17.6
6	72,030	16.7	59,210	13.7	51,575	12	60,108	14
7	50,571	11.7	67,524	15.7	35,645	8.3	53,845	12.5
8	21,638	5	63,037	14.6	14,720	3.4	35,905	8.3
9	5296	1.2	36,282	8.4	4105	1	14,361	3.3
10 (complete trust)	1741	0.4	15,615	3.6	1877	0.4	7307	1.7
Total valid N	428,514	99.5	427,089	99.1	425,619	98.8	398,223	92.4
Missing	2356	0.5	3781	0.9	5251	1.2	32,647	7.6

When including voting, there was a clear trust-participation trade-off for all trust items, trust in the police and the legal system and trust in national political actors. Notably, predicted participation had low points at both trust = 0 and trust = 10. This conflicts with previous observations of a linear association between political trust and political participation (Braun and Hutter, 2016; Hooghe and Marien, 2013; Kaase, 1999). Furthermore, that the trade-off was most apparent when examining the aggregated variables for trust and participation suggests that the effect is not specific to a particular form of participation, but manifests when examining the overall level of political participation. This supports the view that overall political participation may shift from one modality to another (Crepaz, 2017).

However, the inclusion or non-inclusion of voting had a notable effect on the results. When not including voting, the results for all trust items and trust in the police and the legal system indicated a trust-participation trade-off. Most notably, although a trust-participation trade-off was apparent, only the increase in rates of predicted participation from trust = 0 to trust = 1 was inconsistent with a linear decline in participation as trust increased. The presence of a trade-off at lower levels of trust when not including voting is inconsistent with previous findings of differing effects from institutional and non-institutional participation (Devine, 2024; Hooghe and Marien, 2013; Kaase, 1999). Rather, the results suggest that there is a trade-off between trust aggregated participation, which is contributed to greatly by voting rates. When including voting, the trade-off occurs at either end of the trust scale. This suggests that citizens may be particularly likely to exhibit political disengagement in their voting behaviour at very high and very low levels of trust and conflicts with previous results that more trusting citizens are slightly more likely to vote (Devine, 2024).

When not including voting, the results for trust in national political actors indicated a linear negative relationship between trust and rates and participation. This suggests that citizens become less engaged in political activities other than voting as they exhibit higher trust in local political authorities. Interestingly, the inclusion of both institutional and non-institutional forms of participation did not alter this relationship, which runs counter to expectations form the literature (Hooghe and Marien, 2013; Kaase, 1999). Indeed, the results for non-institutional forms of participation were very similar to the results when including all forms of political participation other than voting. This suggests that it is voting specifically, rather than the institutional affiliation of the activity, that is important.

When examining institutional forms of participation, there was a clear trust-participation trade-off for all trust items. The results for trust in the police and the legal system and trust in national political actors broadly conformed with the results for all trust items, but differences between predicted rates of participation at high levels of trust were nonsignificant. The results for institutional forms of participation were closely aligned with the results for the aggregated measure of political participation. This again affirms the importance of voting specifically as a key measure of participation. The results again conflict with previous observations of higher political trust predicting higher rates of institutional political participation (Devine, 2024; Hooghe and Marien, 2013; Li and Yu, 2023). Where a clear relationship can be seen, it conforms more closely with a trust-participation trade-off, suggesting that the effects of trust on institutional political participation are inconsistent across trust levels.

Conclusion

This article explored the relationship between citizens’ political trust and their rates of political participation. The results provide important nuance to the common view of a linear relationship between political trust and political participation (Hooghe and Marien, 2013; Kaase, 1999; Li and Yu, 2023; Stals et al., 2022). When political trust and political participation were both aggregated, there was evidence of a positive relationship between trust and participation at low levels of trust and a negative relationship at high levels of trust. This effect was also present when examining trust in the police and the legal system, but results for trust in national political actors were less clear while the results for trust in international political actors suggested no clear relationship between trust and political participation. The trade-off was more clearly present when examining institutional political participation, particularly voting behaviour.

The key limitations of this study are artefacts of the use of secondary data from a survey not specifically designed to examine the relationship between trust and political participation. Future research may choose to utilise data more intentionally operationalising political trust and political participation and including specifically chosen control variables. The primary limitation of the study was the use of pooled data that aggregated effects across a number of contexts. Although this was accounted for by including survey wave and country as control variables, it is possible that the relationship would appear different in particular political contexts given the previously identified inconsistency in existing research results (Devine, 2024). As is commonly the case in the literature (Devine, 2024; Marien, 2011), the data used here only covers European states. As such, the results may not be generalisable outside Europe.

The analyses presented in this article may also be limited by the scales used to operationalise trust and participation. For trust, the scale is limited by its ordinal – rather than continuous – nature, as well as the practically necessary bounds of no trust and complete trust. For participation, the analyses are limited by the relatively small number of actions (8) contributing to the measure of political participation behaviour by respondents, with any other forms of political participation not being taken into consideration.

The primary implication of this article is that there may be additional complexity in the relationship between political trust and political participation originating from variable effects at different levels of trust. The trust-participation trade-off found in this article provides a plausible explanation for the inconsistent findings on the relationship between political trust and political participation in previous literature (Braun and Hutter, 2016; Li and Yu, 2023; Michalski et al., 2023). In addition to variation between different types of political participation (Hooghe and Marien, 2013; Kaase, 1999) and different levels of government (Chen et al., 2021), the relationship between trust and participation also varies between different levels of trust. Given the explorative nature of the analysis in this article, the results do not constitute a definitive test of nonlinearity in the relationship between political trust and political participation. Future research should look to examine the trust-participation trade-off in additional detail through a more robust test for nonlinearity than has been performed in this article. The key for performing such a test will be in collecting data which can reliably be treated as continuous and which has a sufficiently normal response distribution to allow potential nonlinearities to manifest.

The findings in this article provide three key considerations for future operationalisations of political trust and political participation, in addition to the previous observation that voting behaviour should be considered a distinct modality of political participation (Hooghe and Marien, 2013; Kaase, 1999). The first is that the trust-participation trade-off was most apparent when both variables were aggregated, indicating that the trends in trust and participation may be more than the sum of their parts. The second is that voting was the form of political participation in which respondents most often engaged and which had the greatest impact on the results. This underscores the importance of considering voting a distinct form of political participation. The third is that the relationship between political participation and trust in the police and the legal system – or in the state security apparatus – exhibits a clearer drop-off in participation than in the case of national and international organisations of governance. Maintaining these distinctions in the operationalisation of trust and participation in future research will allow researchers to verify the extent to which these measures should indeed be considered different from the other trust and participation measures discussed in this article.

The key implication for policymakers from this article is that although high political trust is in the interest of states as political actors, it may come at the cost of the citizens’ willingness and propensity to participate in political matters. As such, it may be desirable for policymakers to avoid fostering blind or uncritical trust, as this may hamper citizen engagement. In addition, the finding of the trust-participation trade-off supports the argument of Norris (2022) that politically conscious citizens are more likely to be politically active and more likely to approach state action with a degree of scepticism. As such, it may be the case that very high levels of trust in government are in fact not desirable even for the state, so as to maintain state accountability to citizens. The challenge for the state is in providing citizens with a political environment that engenders their trust in a way that does not discourage political engagement.

Supplemental Material

sj-docx-1-psw-10.1177_14789299241304611 – Supplemental material for The Trade-Off Between Political Trust and Political Participation

Supplemental material, sj-docx-1-psw-10.1177_14789299241304611 for The Trade-Off Between Political Trust and Political Participation by Erasmus Häggblom in Political Studies Review

Footnotes

Acknowledgements

The author would like to thank Bjorn Kleizen, Wouter Van Dooren and the anonymous reviewers for their helpful comments and feedback on the article.

Data availability

The data used in this article are available in the ESS data portal, at .

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship and/or publication of this article: This work was supported by the Research Foundation – Flanders (Fonds voor Wetenschappelijk Onderzoek – Vlaanderen) (grant number G047822N).

ORCID iD

Erasmus Häggblom

Supplementary information

Additional supplementary information may be found with the online version of this article.

Appendix A: Regression results for aggregated variables, estimated marginal means.

Aggregated items estimated marginal means by trust level, activity count including voting.

Aggregated items estimated marginal means by trust level, activity count not including voting.

Aggregated items estimated marginal means by trust level, institutional political participation count.

Aggregated items estimated marginal means by trust level, non-institutional political participation count.

Appendix B: Binary regression analysis results.

Aggregated items estimated marginal means by trust level, binary activities including voting.

Aggregated items estimated marginal means by trust level, binary activities not including voting.

Aggregated items estimated marginal means by trust level, institutional political participation binary.

Aggregated items estimated marginal means by trust level, non-institutional political participation binary.

Appendix C: Regression results for individual variables, estimated marginal means.

Estimated marginal means by trust level, worn or displayed campaign badge or sticker last 12 months.

Estimated marginal means by trust level, boycotted certain products last 12 months.

Estimated marginal means by trust level, contacted politician or government official last 12 months.

Estimated marginal means by trust level, taken part in lawful public demonstration last 12 months.

Estimated marginal means by trust level, signed petition last 12 months.

Estimated marginal means by trust level, voted last national election.

Estimated marginal means by trust level, worked in another organisation or association last 12 months.

Estimated marginal means by trust level, worked in political party or action group last 12 months.

Appendix D: Binary and individual regression analysis results graphed.

Figure 5: Aggregated item results, binary.

Figure 6: Institutional and non-institutional participation results, binary.

Figure 7: Logistic regression results, all trust items.

Figure 8: Logistic regression results, Police and legal system trust.

Figure 9: Logistic regression results, National political trust.

Figure 10: Logistic regression results, international political trust.

Author Biography

Erasmus Häggblom is a doctoral researcher at the University of Antwerp, Belgium. His research focuses on the effects of performance information on how citizens view and engage with public services.

References

Austin

Hux

(2002) A Brief Note on Overlapping Confidence Intervals. Journal of Vascular Surgery 36 (1): 194–195.

Bakardjieva Engelbrekt

Bremberg

Michalski

, et al. (2019) Trust in the European Union: What Is It and How Does It Matter? In: Bakardjieva Engelbrekt

Bremberg

Michalski

, et al. (eds) Trust in the European Union in Challenging Times: Interdisciplinary European Studies. Uppsala: Palgrave Macmillan, pp.1–22.

Braun

Hutter

(2016) Political Trust, Extra-Representational Participation and the Openness of Political Systems. International Political Science Review 37 (2): 151–165.

Ceka

(2012) The Perils of Political Competition: Explaining Participation and Trust in Political Parties in Eastern Europe. Comparative Political Studies 46 (12): 1610–1635.

Chen

Paolino

Mason

(2021) Who Protests and Why: Hierarchical Government Trust and Protest Participation in China. Journal of East Asian Studies 21 (3): 499–513.

Chiru

Gerghina

(2012) Does the Confidence in the EU Spill Over to the National Level? A Longitudinal Analysis of Political Trust in Central Europe. Perspectives on European Politics and Society 13 (2): 226–245.

Crepaz

(2017) What’s Trust Got To Do With It? The Effects of In-Group and Out-Group Trust on Conventional and Unconventional Political Participation. Social Science Quarterly 98 (1): 261–281.

De Blok

Kumlin

(2022) Losers’ Consent in Changing Welfare States: Output Dissatisfaction, Experienced Voice and Political Distrust. Political Studies 70 (4): 867–886.

Devine

(2024) Does Political Trust Matter? A Meta-Analysis on the Consequences of Trust. Political Behavior 46: 2241–2262.

10.

ESS Rounds 1-9: European Social Survey Rounds 1-9 Data (2002–2018) Sikt – Norwegian Agency for Shared Services in Education and Research, Norway – Data Archive and Distributor of ESS Data for ESS ERIC. Available at: https://ess.sikt.no/en/ (accessed 30 August 2024).

11.

Fung

Liu

Chung

(2022) Bowling Alone in Taiwan? Political Trust and Civic Participation of Taiwanese and Their Appraisal of Liberal Democracy and Personal Wellbeing. Social Indicators Research 159: 1085–1102.

12.

Grimmelikhuijsen

Knies

(2017) Validating a Scale for Citizen Trust in Government Organizations. International Review of Administrative Sciences 83 (3): 583–601.

13.

Hamm

Carrera

Van Fossen

, et al. (2023) Conceptualizing Trust and Distrust as Alternative Stable States: Lessons from the Flint Water Crisis. Ecology and Society 28 (3): 14.

14.

Holbein

(2016) Left Behind? Citizen Responsiveness to Government Performance Information. The American Political Science Review 110 (2): 353–368.

15.

Hooghe

(2011) Why There Is Basically Only One Form of Political Trust. The British Journal of Politics and International Relations 13: 269–275.

16.

Hooghe

Marien

(2013) A Comparative Analysis of the Relation Between Political Trust and Forms of Political Participation in Europe. European Societies 15 (1): 131–152.

17.

Hooghe

Marien

Oser

(2017) Great Expectations: The Effect of Democratic Ideals on Political Trust in European Democracies. Contemporary Politics 23 (2): 214–230.

18.

Hooghe

Marien

Pauwels

(2011) Where Do Distrusting Voters Turn to If There Is No Viable Exit or Voice Option? The Impact of Political Trust on Electoral Behaviour in the Belgian Regional Elections of June, 2009. International Journal of Comparative Politics 46 (2): 245–273.

19.

(2016) The Impact of Chinese Political Efficacy and Political Participation on Their Trust in the Police. The Journal of Chinese Sociology 3 (3): 1–17.

20.

Huhe

Chen

(2024) Stratified Political Trust in a Nondemocratic Society: Magnitude, Forms, and Sources of Political Trust in Urban China. Journal of Chinese Political Science 29: 55–75.

21.

James

Moseley

(2014) Does Performance Information about Public Services Affect Citizens’ Perception, Satisfaction, and Voice Behaviour? Field Experiments with Absolute and Relative Performance Information. Public Administration 92 (2): 493–511.

22.

Kaase

(1999) Interpersonal Trust, Political Trust and Non-Institutionalised Political Participation in Western Europe. West European Politics 22 (2): 1–21.

23.

Knol

Pestman

Grobbee

(2011) The (Mis)use of Overlap of Confidence Intervals to Assess Effect Modification. European Journal of Epidemiology 26 (4): 253–254.

24.

Koivula

Malinen

Saarinen

(2022) The Voice of Distrust? The Relationship between Political Trust, Online Political Participation and Voting. Journal of Trust Research 11 (1): 59–74.

25.

Lawless

(1987) Negative Binomial and Mixed Poisson Regression. The Canadian Journal of Statistics 15 (3): 209–225.

26.

(2023) Trust, Political Knowledge and Institutionalized Political Participation: Evidence From China. Current Psychology 42: 15019–15029.

27.

Marien

(2011) Measuring Political Trust Across Time and Space. In: Hooghe

Zmerli

(eds) Political Trust: Why Context Matters. Colchester: ECPR Press, pp.13–46.

28.

Michalski

Marchlewska

Furman

, et al. (2023) To Advise and Scrutinize the Government? Two Types of Political Knowledge, Political Trust and Unconventional Participation. Current Psychology 42: 12104–12117.

29.

Nakamura

Kim

(2010) Public Trust in Government in Japan and South Korea: Does the Rise of Critical Citizens Matter? Public Administration Review 70 (5): 801–810.

30.

Norris

(2022) In Praise of Scepticism: Trust but Verify. Oxford: Oxford University Press.

31.

Payton

Greenstone

Schenker

(2003) Overlapping Confidence Intervals or Standard Error Intervals: What Do They Mean in Terms of Statistical Significance? Journal of Insect Science 3 (1): 1–6.

32.

Porumbescu

(2017) Linking Transparency to Trust in Government and Voice. American Review of Public Administration 47 (5): 520–537.

33.

Proszowska

Jansen

Denters

(2022) On Their Own Turf? The Level Specificity of Political Trust in Multilevel Political Systems. Acta Politica 57: 510–528.

34.

Quaranta

Cancela

Martín

, et al. (2021) Trust, Satisfaction and Political Engagement during Economic Crisis: Young Citizens in Southern Europe. Southern European Society and Politics 26 (2): 153–179.

35.

Ruelens

Nicaise

(2020) Investigating a Typology of Trust Orientations Towards National and European Institutions: A Person-Centered Approach. Social Science Research 87: 102414.

36.

Sloam

(2016) Diversity and Voice: The Political Participation of Young People in the European Union. The British Journal of Politics and International Relations 18 (3): 521–537.

37.

Stals

Isac

Claes

(2022) Political Trust in Early Adolescence and Its Association with Intended Political Participation: A Cross-Sectional Study Situated in Flanders. YOUNG Editorial Group 30 (4): 377–399.

38.

Van Ryzin

(2007) Pieces of a Puzzle: Linking Government Performance, Citizen Satisfaction, and Trust. Public Performance and Management Review 30 (4): 521–535.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

2.43 MB