Can Interest Groups Improve Democratic Representation? A Citizen Perspective

ABSTRACT This article aims to examine how citizens evaluate the contribution of interest groups to the functioning of representative democracies, focusing on several dimensions related to political representation and democratic governance. Drawing on an original online survey conducted to a representative sample of the population in Portugal, we investigate the correlates of citizens’ perceptions regarding interest groups, in particular the impact of individual-level factors related to political involvement, partisanship, party closeness and ideology. We find that associational membership and partisan identities positively impact citizens’ views, while higher levels of (perceived) proximity between political parties and groups increasenegative perceptions. Moreover, people displaying greater trust in interest groups are more likely to hold positive views of organised interests. Overall, political features shape citizens’ attitudes towards groups to a great extent, as people anchored to the political system are more prone to support the institutional role of organised interests and to evaluate their contribution to the functioning of democracies more positively.


Introduction
Many of the challenges contemporary democracies are facing are related to the crisis of intermediation.Political parties and the media are the key players that link political institutions and representatives to citizens.Both actors have lost ground in recent decades, especially due to long-term transformations in our societies, such as the process of disintermediation and the growing personalisation of politics.A vast number of empirical studies show that party organisations are ever less trustworthy and find it increasingly difficult to represent and mobilise voters (e.g., Ignazi, 2017;Katz & Mair, 2018).As for the media, the digitalisation of communication and the emergence of social media have reduced the role of traditional channels such as TV, radio, or newspapers (e.g., Chadwick, 2013;McNair, 2010).At the same time, the media's functions of increasing deliberation and playing a gate-keeping role are declining, increasing the fragmentation and horizontalisation of public opinion (Benoit & Billings, 2020).
These structural changes represent a window of opportunity for the role that interest groups can play in representative democracies. 1 Like political parties, interest groups help foster citizens' participation, transmitting key information to the public and interacting with the media to pursue their objectives, thus increasing their awareness and knowledge of important and complex policies (e.g., Kitschelt, 2012;Schattschneider, 1948).In addition, they intervene both directly and indirectly in the decision-making process, trying to shape policy outcomes (Rasmussen & Willems, 2021).Last but not least, they interact (through competition or collaboration) with political parties and other institutional actors (parliaments, MPs, government) in order to represent citizens' grievances and demands, as well as to fight for relevant resources to spend in the political process.Yet recent structural changes in the system of representationparticularly a less consensual and more impositional decision-making processseem to challenge the role of interest groups.As Richardson put it (2020, p. 461), 'the decline of mass parties might now be being followed in many systems by the decline of interest groups as channels of representation'.
This article aims to examine how citizens view the role of interest groups in the political system, particularly with regard to key aspects of democratic representation, such as participation, equality or accountability, among others.In contrast to the growing trend of empirical studies on interest groups, research based on public opinion surveys is scant and mostly concerns issue representation.This is a relevant aspect that has been largely ignored by the scholarly literature on political culture, which has been limited to the study of traditional institutional actors and their contribution to (diffuse or specific) political support.Moreover, the study of interest groups as actors of intermediation has been relegated to a marginal place compared to research on political parties (see Lisi & Oliveira, 2022).The overall endeavour of this article is to seek to partially recalibrate this imbalance and to illustrate how citizens evaluate the contribution of organised interests to the functioning of democratic representation.
Understanding public attitudes towards the role of interest groups as intermediary actors is relevant for a number of reasons.First, it is of the utmost importance to know whether citizens' attitudes towards these actors differ from other channels of democratic linkages, in particular political parties (see Beyers et al., 2008;Thomas, 2001).Second, it is important to understand whether these attitudes are linked to the political system more broadly, thus eroding or strengthening political trust and the legitimacy of government.Third, public opinion plays a key role in informing political norms, especially with regard to interest group regulation.As political elites are sensitive to public perceptions, citizens' attitudes may inform a better regulation or lead to the elaboration of relevant legislation on the matter, even where formal rules are absent, as in many European countries.Last but not least, it is important to know the drivers of citizens' opinions on organised interests, i.e., what affects positive or negative evaluations and how these vary across distinct attitudinal and behavioural characteristics at the individual level.
This study is unique in two ways.First, it seeks to redress the balance in the literature, which currently includes many studies focusing on the 'supply side', that is, the interest groups' ability to gain access and to interact with representative institutions, but few that examine the 'demand side', namely, what citizens request from these actors and how satisfied they are with the response.The 'double-face' role of intermediation played by organised groups is a key feature that needs to be examined in detail, not only considering how groups aggregate pre-existing voiceless concerns, but also how people echo the output and performance of these associations.This remains an important gap in the fastgrowing literature on interest groups (Pritoni & Vicentini, 2022).Second, this article relies on an original survey specifically designed to capture citizens' opinions on the role played by interest groups in the functioning of the political system.Hitherto, mass surveys have not tackled the democratic functions performed by interest groups, and the topic has been (marginally) addressed from the interest groups' perspective only (for example, through the Comparative Interest Group Survey).From this viewpoint, this work considers fresh data on a topic that has not yet been explored, in particular what citizens think about the role of interest groups in the overall political system, the link between organised interests and representative actors, as well as their potential contribution to better governance.
What do citizens think of interest groups as actors of political intermediation?What are citizens' perceptions of the role of interest groups regarding specific dimensions of the quality of democracy?And what are the key drivers of these opinions?The article addresses these questions by using original public opinion data collected from a representative sample of Portuguese citizens.This analysis provides a comprehensive portrait of people's opinions and attitudes on organised interests and sheds further light on an understudied aspect of political representation.The study uses Portugal as a typical case of several democratic shortcomings that undermine political representation and the quality of contemporary democracies.Indeed, Portugal has experienced one of the largest increases in institutional distrust in Europe in the twenty-first century and presents very low levels of political engagement (Pinto et al., 2013;Torcal, 2014).In addition, empirical research on interest groups has been largely neglected and there are no studies dealing with public opinion and the role of these actors of intermediation.
The article proceeds as follows.The next section provides an overview of studies dealing with public opinion on interest groups and outlines our theoretical framework and hypotheses.The third section describes the data used and the methods of analysis.We then examine public attitudes towards organised interests looking at the Portuguese case.Finally, we conclude with reflections on the meaning of our results for democratic representation and the challenges of contemporary democracies.

Navigating Uncharted Waters: Citizens' Opinions on Interest Groups and Democratic Representation
While interest groups and public opinion have long been examined in distinct scholarly communities, recent scholarship has partially redressed this gap by studying the interplay between interest groups and citizens' preferences (see, for instance, Beyers & Arras, 2021;Rasmussen & Reher, 2023;Terwel, 2010).Scholarship has not only started to investigate how groups try to shape public opinion, but also vice versa, that is, how citizens' preferences may influence the action and dynamics of groups.There is also fast-growing interest in the congruence between public opinion and group agenda, as well as the impact of this responsiveness on policy outputs (e.g., Rasmussen et al., 2014Rasmussen et al., , 2018)).Yet few works focus on citizens' attitudes towards the role of groups as actors of intermediation, thus limiting our understanding of an important aspect of democratic politics.
Both theoretical and empirical studies on interest groups highlight the ambivalent nature inherent to the action these actors play in democratic polities.On the one hand, they help give visibility to citizens' preferences and to achieve better policies.On the other, interest groups' activities are associated with the negative side of social capital, 2 such as corruption, inequality of citizens' representation and even economic decline (e.g., Baumgartner & Leech, 1998;Maloney et al., 2007;Nownes, 2013).
On the surface, this ambivalence seems to be related mainly to the heterogeneity of the interest group population.Indeed, in some cases, groups represent diffuse interests or ideas linked to broad segments of civil society, such as environment, minorities or humanitarian causes.By contrast, other groups represent the specific interests of a well-defined sector, such as professional associations or business groups.While this heterogeneity can certainly explain, at least partially, the ambivalence in public opinion, we do not know how individual factors shape attitudes towards interest groups, particularly regarding key aspects of democratic political representation.
What do individuals think about the role interest groups play in the political system?If we consider the literature on the relation between organised interests and public opinion from the citizens' perspective -, at least three main topics of empirical research can be identified.The first considers confidence in distinct types of civil society organisations.Since the implementation of the World Value Survey (WVS), extensive empirical works have examined the extent to which citizens trust in a great diversity of associations, such as trade unions, environmental associations or other cause groups (as well as major companies).Scholarship found that confidence in civil society organisations is generally lower than for governmental institutions (e.g., parliaments, parties, see Rose, 1984).Extant research also suggests that political attitudes (e.g., political interest, life or financial satisfaction) are important determinants of confidence in private organisations, whereas socio-demographic variables play a negligible role (Listhaug & Wiberg, 1995).Unfortunately, this literature only surveys one particular aspect of citizens' political attitudes towards organised interests, and it concerns only specific types of associations (mostly trade unions, e.g., Son, 2020), limiting the possibility to have a more comprehensive and systematic picture of the role of interest groups in the process of political representation.
A number of surveys have examined the involvement of citizens in different types of association, most of which addressed the patterns of political participation and engagement in distinct types of associations, investigating trends over time and cross-country differences (e.g., Deth, 1997;Morales, 2009).Through the analysis of individual-level data, extant research has examined active and passive forms of associational involvement, as well as its consequences in terms of social capital (see Morales & Geurts, 2007).Overall, this research emphasises the importance of participating in associations as a means of generating other positive attitudes and behaviours. 3 A third important strand of empirical research related to organised interests focuses on public attitudes towards the involvement of interest groups in the policy-making process.Some pioneering works investigated public attitudes towards organised interests based on election survey data (e.g., Nielsen, 1985;Willems, 2021).For example, a study on the Swedish case found wide acceptance of the role (economic) groups (should) have in policy-making and as interlocutors of institutional actors, whilst trade unions were perceived as the most influential group in policy-making (Nielsen, 1985).In addition, links between organisations and their supporters (so-called 'constituency engagement') seem to constrain broad public support.Recent studies have found that people perceive interest representation to be biased in favour of business interests and they show more negative opinions when this kind of actor is over-represented in policy-making compared to the over-representation of cause groups (e.g., Bernauer & Gampfer, 2013;Beyers & Arras, 2021).The fact that some groups are considered more powerful (i.e., with more resources) and, at the same time, less representative of the population explains to a great extent individual-level differences in the perceived legitimacy of the involvement of distinct types of groups in decision-making processes (see Rasmussen & Reher, 2023).
Recent scholarship has also investigated how interest groups can shape public opinion by framing certain issues under specific arguments or rhetoric (Dür, 2019;Jungherr et al., 2021).This strand of research found that groups may not only make some topics more salient and meaningful, but that they also effectively influence public opinion.Moreover, the literature provides clear evidence of the capacity of organised interests to perform issue representation across different political settings.Seeking to complement recent studies that describe the dynamics of this 'supply side', this study focuses on the 'demand side', describing citizens' opinions and attitudes towards interest groups, and in particular how they view the role of these groups as actors of intermediation.Indeed, the opinions of ordinary people have rarely been subject to a systematic analysis in the studies of interest groups and their contribution for political representation.This lack of attention to citizens' opinions is puzzling, given that the public is not only a key source of legitimacy for civil society organisations, but can also influence the bargaining between political elites and groups, as well as the way groups' demands are channelled through policy-making processes.

Hypotheses
Given the diversity of interest groups and the multiple issues and instruments associated to their action, there is unlikely to be uniformity in public opinion.But what are the potential factors that shape distinct opinions?We propose several testable hypotheses regarding how citizens evaluate the role of organised civil society as intermediary actors and channels of democratic representation.In doing this, we go beyond simply describing aggregate levels of citizens' opinions towards interest groups by investigating how people perceive the role of interest groups and why this perception is positive or negative.
Firstly, we expect partisanship to matter.Previous research suggests that partisan identities influence attitudes towards political corruption and the role of special interests in politics (e.g., Bowler & Donovan, 2016;Gerber et al., 2010).People aligned with a political party are more likely to be integrated in the political system, to be more satisfied with the overall functioning of democracies and to better understand the role of organised interests for democratic representation.The opposition between 'in-group' and 'outgroup' members that lies behind the rationale of partisanship does not simply refer to parties but may also apply to group politics.In addition, it is highly likely that people close to political parties (i.e., members of political parties or simply party identifiers) will also be members of civil society organisations.As a consequence, we expect that citizens displaying partisan attachments will be more likely to show more positive attitudes towards interest groups (H1).
However, partisanship can also have a negative impact when there is a subordination of interest groups vis-à-vis political parties, that is, when organisations are seen mainly as an instrument for achieving partisan goals.Empirical research indicates that party dominance affects access, particularly influencing the way parties play the role of gatekeepers (see Binderkrantz et al., 2015;Willems, 2020).Party closeness often implies the emergence of conflicts in the political arena.This will increase the salience of party-group interactions, and citizens' attitudes will probably reflect the way the main media frame a specific issue.When the action of interest groups is perceived to be aligned with some political parties, the contribution of these groups to democratic governance will be more controversial and can be the source of disagreements because of its partisan alignments.Therefore, we posit that the more citizens believe interest groups are close to political parties, the less likely they are to support organisations as important actors for democracy (H2).
Our third hypothesis refers to the role of ideology in shaping public attitudes.Although ideology is closely correlated to partisanship, there are conceptual distinctions (Sharp & Lodge, 1985).Conventional scholarship tells us that unions tend to support (or to be associated with) parties on the left, while business more often supports parties on the right.In addition, those who identify with the left tend to be more concerned about equality (e.g., Fuchs & Klingemann, 1990).Indeed, left-wing voters generally seem more supportive of interest groups' involvement in policy-making, especially that of trade unions (Nielsen, 1985).Although left-wing voters may be worried about the undue influence of strong business groups, there are two strong arguments that lead us to hypothesise a positive link between a leftist ideology and more supportive views of the role of organised interests.The first is that people more to the left tend to display higher levels of political engagement and associational membership (e.g., Jakobsen & Listhaug, 2014;Quaranta, 2015).Indeed, political science research shows that in most contexts, leftist (and often trade-unionised) citizens are more likely to engage in political activism.This is likely to increase the knowledge of and the confidence in groups.The second reason is based on the fact that leftist citizens tend to display higher values of communitarianism and universalism (Piurko et al., 2011), two important aspects associated with pluralism.Consequently, we expect individuals who identify themselves as being on the ideological left to display more positive attitudes towards groups, whereas those on the ideological right will be less likely to consider groups as important actors for representation given their support for free market capitalism.Our third hypothesis claims that individuals who position themselves on the ideological left will be more likely to display positive attitudes towards interest groups (H3).
The final hypothesis focuses on the potential impact that associational membership may have on public attitudes.A well-developed strand of scholarly work shows that involvement in civil society organisations influences the set of attitudes and orientations that citizens hold (see, for example, Deth, 1997;Maloney et al., 2007).Associations have often been portrayed as 'schools of democracy' since they can promote positive feelings towards other social and political institutions (for example: political trust, legitimacy or interest in politics).By interacting with other trusting people, individuals are supposed to gain increasing feelings of confidence and efficacy, further strengthening positive views on political institutions.The opposite is also true, as individuals with higher levels of inter-personal trust are also more likely to join groups.Be as it may, most scholars maintain that group membership is beneficial for democratic values even when they recognise potential side-effects (Mahoney 2008).Our theoretical expectation is that the greater the social capital (more involvement in associations), the greater the political trust, bringing more supportive views regarding the contribution of organised interests for political representation.Therefore, our hypothesis is that individuals who are members of civil society organisations are more likely to display positive attitudes towards interest groups (H4).
These hypotheses will be tested through the analysis of original data, as explained in the following section.It is important to note the exploratory nature of this research.As the name implies, the main aim of this kind of study is to explore the research questions without providing a final and conclusive answer to existing problems.In other words, we are dealing with an issue that has not yet been clearly defined and our research aims to contribute to a better understanding of this problem.

Methods and Data
Before explaining the methodology and data used in this study, some relevant features of interest groups in Portugal should be highlighted.As mentioned in the introduction, the study of interest groups in Portugal has been neglected, mostly because of the protagonist role played by political parties during democratisation and the consolidation of the new regime (Morlino, 1998;Schmitter, 1992).In addition, civic associations were clearly dependent on political elites and the main state institutions.As several studies have noted (see Magone, 2014;Razzuoli & Raimundo, 2019), one of the main legacies of the revolutionary process was the emergence of deep ideological and political cleavages, leading to a representation overlap between parties and certain groups.While the communist party (PCP [Portuguese Communist Party]) enjoyed clear hegemony over the labour movement, business associations aligned with the main governing parties (PS [Socialist party] and PSD [Social-Democratic Party]).On the other hand, the rightwing party CDS [Social-Democratic Centre] had closer collaboration with agricultural associations.Finally, the legal and political status of certain types of organisation under the Portuguese Constitution is also of relevance when interpreting citizens' opinions on the role of interest groups.From this viewpoint, it is worth highlighting the role attributed by the Constitution to trade unions and labour associations, as their participation in the decision-making process is mandatory on any matters concerning labour legislation and the development of economic and social plans.Agricultural associations are also mentioned as important players in this policy area; however, the Constitution makes no reference to employers' organisations, a fact that is associated to its leftist orientations and the principle of subordinating economic power to political power.These peculiarities are important not only because of the relevance that partygroup alignment may have on people's perceptions, but also due to the potential impact of ideology, especially for left-wing voters.
Individual-level data usually reports citizens' opinion on traditional institutions (parliaments, governments, parties, courts, etc.) or their effectiveness when participating in policy-making ('efficacy').The innovative aspect of the survey is the in-depth inquiry into citizens' attitudes and preferences on interest groups.The survey included many dimensions, some of which replicated or adapted questions asked in international public opinion studies, such as PARTIREP or the European Social Survey (for example in relation to member participation), or in national studies (e.g., Portuguese Election Study).In this work, we focus on original questions strictly related to political representation and the relationship between interest groups and the functioning of democratic regimes.
This study analyses key aspects of the functioning of political systems and the quality of representative democracies generally.In fact, the evaluation of intermediary actors through citizens' attitudes has been at the core of the analysis on democratic governance and political representation (e.g., Ferrín & Kriesi, 2016;Landman, 2008).Our approach follows the literature on public opinion studies and citizens' evaluation of the functioning of representative democracies, which emphasises the need to adopt a multidimensional perspective when surveying citizens' beliefs.From this viewpoint, our starting point is to recognise that individuals may have different things in mind when thinking about the way advanced democracies work.Therefore, our questionnaire includes the following question on the general role interest groups are expected to play in the political system: 'Some say that civil society organisations in Portugal are very necessary for the proper functioning of our political system.Other people think that these organisations are not at all necessary for the proper functioning of the political system.Using a scale between 1 (organisations are not necessary at all) to 5 (organisations are very necessary), where would you position yourself?' 4 This is a very similar question to 'support for political parties', an item regularly included in public opinion surveys.As a robustness check, we also include another question related to institutional support, namely whether individuals believe interest groups can positively or negatively contribute to the functioning of democracy. 5These are the main dependent variables in this study.
In addition to the questions related to 'institutional' support towards the role of interest groups, we also investigate more 'specific' aspects of group politics for political representation.In particular, we measure members' perceptions by their expressed level of agreement (on a 5-point scale) on a continuum with contrasting views of the functions played by organised interests.The statements were as follows: (a) Civil society organisations lead to increased corruption in the country vs. Civil society organisations lead to more accountable governments and help ensure greater transparency in public decisions (governmental accountability, accountability henceforth); (b) Civil society organisations do not encourage citizen participation vs. Civil society organisations contribute to increasing civil society participation (participation); (c) Civil society organisations contribute to increasing inequalities vs. Civil society organisations contribute to reducing inequalities in society (equality); (d) Civil society organisations make it difficult to understand public policies vs. Civil society organisations transmit important information about certain policies (policy-making); (e) Civil society organisations only serve to represent the interests of specific groups vs. Civil society organisations contribute to increasing pluralism and representing interests that are not represented in Parliament (pluralism); Moving to the operationalisation of independent variables, we include two distinct dichotomous variables for membership association and partisanship.These variables take the value 1 if the respondent belongs to an organisation or feels close to one party and 0 otherwise.For party closeness, we construct a composite index based on three items, namely citizens' perception of the closeness between parties and unions, employer and professional associations (ranging from 0 = no closeness to 1 = closeness for all three group categories).The coding of ideology is more challenging as a number of distinct options can be considered.After several preliminary tests, we decided to consider both the traditional left-right continuum on a 11-point scale (0 = extreme left; 10 = extreme right), and also its squared term.The rationale behind this is to account for the possible non-linear relation between citizens' attitudes and ideology, that is, the fact that more ideologically radical voters might hold distinct beliefs than citizens who position themselves in the centre of the ideological spectrum.Moreover, a study based on a survey in Portugal shows that citizens with centrist ideological positions present more distinct preferences with regard to various actors of political representation compared to more politicised individuals (Vieira & Wiesehomeier, 2013).
In addition to these independent variables, political trust is an important dimension that needs to be considered when dealing with political attitudes.From a comparative perspective, the Portuguese consistently stand out for their distrust of political parties (and parliaments) vis-à-vis their European counterparts (see Belchior, 2015).Our study corroborated this trend as regards the trust placed in interest groups.On a scale of 1-5, where 1 means 'no trust' and 5 'you can trust most organisations', 36% of the respondents are in the middle of the scale, while about half of the sample (49% of respondents) consider that caution is paramount.This distrust is manifested above all in relation to employer associations and religious organisations (Figure 1).The average level of trust in trade unions is also very low.By contrast, environmental associations are considered the most reliable by about 40% of the sample and this is the only category with a mean above the average of the 5-point scale (3.1).There is also a relatively high level of trust in public interest groups such as consumer associations or social solidarity organisations (mean of 2.9 and 2.8, respectively).Although the overall picture shows that groups are neither trusted nor distrusted, some organisations are more trusted than others.Despite these differences, there is a strong correlation between trust in distinct organisations.In fact, a factor analysis performed on the eight items included in the survey shows that the battery forms one single solution (see Table A3 in the appendix).This allows us to create a composite index on the basis of eight different targets of citizens' trust.This index varies from 5 (no trust) to 18 (full trust in all types of association) and it is used in the multivariate analysis as an important control variable to be included in the full model. 6Finally, we also include satisfaction with democracy in the model as it has been found to be a relevant factor for predicting citizens' attitudes towards the political system (e.g., Ferrín & Kriesi, 2016), but also specifically to account for the trade unions' capacity to represent Portuguese people (Vieira & Wiesehomeier, 2013).
Starting from this operationalisation, the empirical analysis of this article is divided into two parts.The first is mainly descriptive and depicts citizens' perceptions of the role interest groups play in the political system.The second tests the hypotheses outlined in the previous section through multiple regression analysis.In particular, it tests whether citizens' evaluations are contingent on their socio-demographic characteristics, political profile and other relevant attitudinal traits.In addition to the key explanatory variables, we control for age, gender, education and religiosity, measured as the frequency of attendance of religious services.Full details on the wording of questions and coding are provided in the appendix (see Tables A4 and A6).Note that some independent variables could also be the consequences.For example, positive views on the role interest groups perform in representative democracies could also easily be a cause of trust in (private) organisations.Our goal is first and foremost to offer a comprehensive and systematic analysis of correlates in search of potential causes and consequences of citizens' support towards interest groups. 7 The data used in this study are based on an original survey conducted within the project 'GRIP' (Lisi et al. 2021).An online questionnaire was applied through the company NetQuest to a representative sample of the Portuguese population, with fieldwork taking place between 31 December 2020 and 11 January 2021.Overall, 1,516 responses to this online survey were collected from respondents aged 18 or over.The  2021). .Notes: (1) question wording: 'In general, do you think that most civil society organisations can be trusted or, on the contrary, do you think that care is little?' (2) 1: All care is little; 5: Most organisations can be trusted.
sample was selected using quotas, taking into account sociodemographic variables such as region, gender and age of the participants, in order to allow the construction of a representative sample of the characteristics of the Portuguese population.The sample is about 52% female and 48% male.There is a good balance between the different age groups, namely 12% are aged between 18 and 24 years, 13% between 25 and 34 years, 17% between 35 and 44 years, 17% between 45 and 54 years, and finally about 25% are over 65 years.Likewise, the sample is geographically balanced, despite the prevalence of Lisbon (28%) and the North of the country (37%). 8

Results
Before investigating the sources of public attitudes towards interest groups as actors of intermediation, some key findings on their role in the chain of democratic representation should be reported.
When we asked whether citizens feel represented by some civil society organisations, a slight majority (53%) considers that there are indeed groups that perform this function. 9 On the other hand, when we asked whether citizens believe that these associations are necessary for the functioning of representative democracies, a large proportion of respondents were ambivalent (45% position themselves in the intermediate category of the scale), and almost 37% have a clearly positive view (Figure 2a).This means that organised interests are largely seem as indispensable for democratic regimes, a result similar to what we can find for political parties (Teixeira et al., 2016).This finding is confirmed when we consider an alternative measure based on the evaluation of interest groups' contribution to democracy.Although half the respondents (53%) take a neutral position, about one third of our sample believe that organised interests make a positive contribution to democracy (Figure 2b).This result must be interpreted with caution because the choice of the mid-point may be closer to 'don't know' for some respondents. 10 It is also interesting to note that this question is closely linked to the citizens' opinions on the funding of organised interests.When asked whether these organisations should Notes: (1) Figure 2a, question wording: 'Some say that civil society organisations in Portugal are very necessary for the proper functioning of our political system.Other people think that these organisations are not at all necessary for the proper functioning of the political system.Using a scale between 1 (organisations are not necessary at all) to 5 (organisations are very necessary), where would you position yourself?';(2) Figure 2b, question wording: 'In your opinion, do you consider the general contribution of interest groups to democracy to be positive or negative?';Scale: from 1 (negative) to 5 (positive).
receive state support, a small majority (53%) of respondents disagree, and these citizens are also more likely to think that these groups are not necessary actors for the functioning of representative democracies. 11But opinions on the resources interest groups require to carry out their activities are divided into two almost identical groups, making this aspect worthy of further investigation.In addition, there is only a moderate overlap between institutional support of interest groups and public funding, as the correlation between these two dimensions is only 0.30 (Pearson coefficient, significant at 0.01 level).
Moving to more specific dimensions of the functioning of representative democracies, it is interesting to note the variation in citizens' opinions on the role of interest groups across distinct dimensions.Figure 3 reveals the average for (governmental) accountability, participation, equality, policy-making and pluralism.The results suggest that neither a clearly negative nor a positive view prevails for most of the dimensions considered in the survey.As an example, 1/3 of respondents believe that civil society organisations contribute to increasing corruption in the country and another 1/3 believe that they contribute to making political rulers more accountable and promoting transparency.In contrast, there are spheres of action where a positive or negative view clearly predominates.On the positive side, most individuals consider that civil society organisations contribute to the reduction of inequalities (39%) and recognise them as important vehicles for policymaking (34%).However, with regard to the role of representing citizens' interests, most respondents believe that these organisations serve mainly to represent the interests of specific groups (38%) and not to increase pluralism and interests not represented in parliament. 12As for other similar questions included in the survey, it is impossible to say whether respondents have sufficient information to support their opinion.It can be the case that those respondents choosing the mid-point are not only neutral vis-à-Figure Citizens' perceptions on distinct dimensions of representative democracies.Source: see Figure 1.
Notes: (1) Question wording: 'Thinking now about the role that these organisations can play, in general, in the political system, what is your opinion regarding the following statements?' (2) scale: 1 (negative opinions); 5: (positive opinions).vis the two opposite views, but they can also be unsure about the response or without a strong opinion on the matter (or even 'don't care').
To test the hypotheses identified above, Tables 1 and 2 present the results of the regression models, for each of the dependent variables, covering different aspects of perceptions of interest groups' role in the political system.We begin by analysing the general questions regarding institutional support towards organised interests, moving then to more specific aspects of democratic governance.For each dependent variable we present two models.The baseline model includes socio-demographic controls and associational membership, while the full model also considers the impact of key political attitudes traditionally associated with the evaluation of representative democracies.
Table 1 reports the results for citizens' opinions on the general role interest groups play in the political system.The results indicate that belonging to an association increases the probability of having a positive opinion on the functions played by organised interests.This finding goes in the expected direction, thus confirming hypothesis 1.In addition, socio-demographic variables are also important in explaining the variance in citizens' opinions on the role of groups for political representation.First, more educated people display more positive views regarding organised interests.Second, religiosity is also positively associated to positive attitudes.However, people with an occupation show more negative attitudes towards organised interests.This unexpected result may be due to the greater involvement of some inactive groups (students, pensioners or even unemployed) in associational life, as well as more time availability.Finally, when we look at institutional support (groups necessary to democracy) we can observe that older citizens are more critical of the role played by interest organisations.
When we add more specifications to the model, that is, including key political attitudes, the variance explained increases significantly (see model 2).The results indicate that all the main correlates are statistically significant and play an important role in explaining the variance in people's attitudes.In particular, both associational membership and partisan identities have a positive impact on the variable related to the contribution of interest groups to democracy (H1 and H4 confirmed).The closeness between parties and groups negatively affects citizens' opinions.This finding supports our second hypothesis (a stronger perception of the subordination of groups towards parties means a more negative evaluation).Ideology also achieves standard levels of statistical significance and presents a positive coefficient, which suggests that citizens with a rightist orientation are more likely to display a positive view of interest groups (H3 not confirmed).The squared term of ideology also has a significant impact, which indicates that positive views are concentrated mostly in centrist ideological positions.The data do not allow for a clear interpretation of this finding as it may be associated to a higher level of alienation and lack of interest of this centrist electorate or, alternatively, to higher levels of dissatisfaction with the political system (that is, displaying critical attitudes towards democratic outputs).
The impact of key independent variables is also confirmed after controlling for more relevant political attitudes.From this viewpoint, we find that both trust in organisations and satisfaction with democracy also have a relevant effect.Citizens with higher levels of trust and that are more satisfied with democratic outputs are more prone to have a positive view of the contribution of organised interests to the functioning of the political system.Lastly, people with higher levels of political interest tend to view the role of interest groups more positively.
The second dependent variable of institutional support (interest groups are necessary to democracy) serves as a robustness check; however, it is more difficult in this case to account for the variation in citizens' attitudes.Two main differences are worth noting.The first is that the relation between ideology and citizens' attitudes is no longer relevant, even when we consider its squared term.The second is that the impact of associational membership and partisanship declines in the full model specification.Nevertheless, the importance of the remaining correlatesparticularly party closeness, trust and democratic satisfactionto determining citizens' opinions is confirmed.
When we look at specific aspects of democratic governance, our findings confirm the idea that being a member of associations is strongly correlated with positive views regarding their role for political representation.In particular, members of an organisation are more likely to believe that groups are important to ensuring accountability, more participation, better policies and they may also improve the representation of the diversity of civil society's interests (Table 2).Older people are more sceptical about organised interests, regardless the specific dimension considered in the survey.In addition, citizens with a higher socio-economic status (income and, to a less extent, education) are more likely to display positive political attitudes towards groups, while occupation achieves statistical significance only in the case of pluralism.Again, the negative coefficient is due mainly to the more positive views of students, pensioners and maids.
The results for the full model are displayed in Table 3 and can be summarised as follows.First, the variance explained increases substantially when we add attitudinal variables to the multivariate regression.However, there are differences in the fitness of the model across the different dimensions considered in this study.While it does a good job of explaining variation in the case of accountability, the variance is quite low for policy-making.This can be due to the higher salience of corruption and other issues related to the responsibility of politicians, while citizens have more difficulties to evaluate (in abstract) the input of groups in the decision-making process.
Second, trust in interest organisations is a key variable for explaining citizens' views on the role of groups in democratic regimes.The greater the trust in organised interests, the more positive the attitudes towards the benefits that groups can bring to the functioning of representative democracies.The results are consistent across all dimensions surveyed in the study, namely accountability, participation, equality, policy-making and pluralism.
Third, attitudes towards parties are also correlated with citizens' views towards organised interests.In this respect, two mechanisms are worth noting.On the one hand, people closer to a political party are more likely to display positive attitudes towards organised interests in terms of accountability, participation and policy-making.This finding suggests that people relying on actors of intermediation believe in the beneficial effects of groups.In other words, groups seem to play a complementary role in political representation and are not seen as challenging the role of political parties, as some literature seems to hypothesise (Schmitter, 2001).On the other, the perceived closeness between groups and political parties negatively affects citizens' opinions.This means that organisations deemed to serve partisan goals are seen to be less beneficial to democratic governance.The finding may be interpreted in the sense that citizens highly value the independence of interest groups vis-à-vis parties when they think of their role in the political system.This is also confirmed if we look at the reluctance of civil society organisations to explicitly support specific parties, especially during electoral campaign or the discussion of state budget (Lisi and Silva, 2023).Fourth, satisfaction with the way representative democracies work is also associated to a positive evaluation of the role interest groups play in the channel of democratic representation.While plausible, caution is required in terms of causal relations as our analysis cannot unveil the mutual relationship behind the association of these two variables.Finally, although associational membership is less important than other variables, it still has a significant impact on equality.By contrast, ideology does not significantly impact citizens' perspectives of the role interest groups might play in the functioning of contemporary democracies.The impact of control variables confirms previous findings, especially in the case of age and (to a less extent) occupation.

Conclusion
Interest groups enjoy a special place in the circuit of political representation because they project a public image of their social role whilst bringing important aspects of the institutional arena to society.This function is associated with important facets of representative democracies, but this intermediation is different from the one performed by political parties or social movements.The study examined this topic, focusing on citizens' attitudes towards the role of interest groups and their overall contribution to the political system.The correlates of citizens' attitudes towards interest groups are then examined, based on the analysis of an original public opinion survey conducted to a representative sample of Portuguese population.By using individual level data, this study investigates the impact of key independent variables on perceptions of interest groups as actors of intermediation and their role for democratic governance.
The fact that a large proportion of people (approximately one third) does not have a strong opinion on the potential impact of interest groups on democratic governance may be related to the heterogenous nature of interest groups' action.While some organisations clearly have greater visibility and a more regular presence in the institutional arena, others are more concerned with the delivery of services or goods to their members, thus marginalising their institutional functions.However, citizens evaluate organised interests more positively when considering their important role in policymaking, especially as a source of information for decision-makers and public opinion, and for the potential reduction of inequalities.This is also an important contribution of interest groups because they are considered relevant tools for strengthening the voice of groups that find it more difficult to access the institutional arena.
However, the fact that a significant proportion of citizens does not display a strong opinion on the representational function of groups does not mean that they are not able to distinguish among distinct organisations.For example, people show less confidence in business and religious organisations, while they tend to have higher levels of trust in public interest groups.Moreover, business groups are considered to be more aligned with political parties than professional associations and, to a less extent, trade unions.At the same time, there is evidence that suggests that business associations are considered to play a more important role in the parliament than professional or union organisations. 13 As for the sources of citizens' attitudes towards interest groups, the study finds that political experiences are extremely important to account for the variation of opinions at the individual level.People with stronger partisan identities and that are more involved in the political sphere tend to display more positive attitudes towards interest groups as actors of intermediation.Moreover, higher levels of trust in organisations and satisfaction with democracy are associated with a better evaluation of the institutional role performed by civil society organisations.In other words, citizens who are more anchored in the national political system show a higher level of institutional support and a better evaluation of organised interests.However, citizens clearly view interest groups and parties as distinct actors for political representation and more negative opinions emerge when people believe that groups are subordinated to parties' goals.This indicates that autonomy is a crucial feature that groups need to preserve if they want to be recognised and supported as important actors for the functioning of representative democracies.The findings are also interesting insofar as they discard theories based on the importance of ideology.Indeed, our third hypothesis is the only one to be clearly rejected, while the remaining expectations are substantially confirmed regardless of the dependent variable used to evaluate citizens' opinions.Another interesting implication is the partial confirmation of cognitive mobilisation theories.While education does not play an important role in explaining people's views on the contribution of organised interest for democratic representation, political interest positively impacts citizens' opinions.
The constellation of attitudes investigated in this study poses relevant challenges for democratic governance and political intermediation.Politicians face a dilemma regarding the extent to which organised interests should participate in policy-making.On the one hand, governments are expected to grant organisations access to the institutional arena and to take the advice of distinct groups into consideration, but in so doing they risk being accused of giving the organisations too much power.On the other, policymakers may have more autonomy and may isolate themselves from the 'voices' of organised civil society, thus risking criticism for not being responsive enough.Public attitudes towards interest groups reflect this dilemma and the difficult equilibrium between the autonomy of representatives and their linkage with groups and citizens.
The evidence presented here, and our interpretation of it, must be considered in the light of some caveats.The first, and most important, is to disentangle the relation of causality between citizens' political attitudes and their evaluation of the role of interest groups.We urge scholars to investigate this relation through more sophisticated research design, for example by relying on survey experiments.Furthermore, additional work is needed to establish more precise measures of perceptions of interest groups' political role and, if possible, to better distinguish between distinct types of organisation.Our results could be affected by perceptions of corruption or the evaluation of government performance.These are some of the factors that can also shape citizens' opinions on interest groups.Lastly, given the ambivalence displayed by most respondents, a more in-depth investigation of the meaning of neutral or indifferent responses would be fruitful, namely whether this relates to citizens' apathy, lack of information or the fact that people have distinct types of organisation in mind.As such, quantitative or qualitative research may help to address this intellectual challenge, specifically the interplay between the role of interest groups and other important dimensionsfor example, political cleavages, intra-organisational aspects or agenda-setting.
These things considered, this work supports the idea that interest groups, like political parties, are seen as key actors of political intermediation, and that the evaluation of representative democracies should also consider the role these actors play in different arenas of democratic governance.In terms of generalisability of the main results, this case study suggests an optimistic interpretation of the potential role of organised interests for political representation.Given the low levels of associational membership and political confidence of Portuguese citizens, it is likely that more positive views will be found in other contexts characterised by more widespread and higher levels of social capital.In a nutshell, interest groups may still aim to fill the representation gap left by political parties.Therefore, the findings of this study suggest that investigating public attitudes towards organised interests is a fruitful pursuit if we wish to make sense of the nature and dynamics of political representation, especially beyond election periods.Further comparative research is required to examine these issues in detail, namely the extent to which citizens' opinions vary across distinct political and institutional contexts, and to explore whether these findings can be generalised to other countries in Europe and beyond.

Notes
1. We define interest groups as 'an association of individuals or organisations, usually formally organised, that on the basis of one or more shared concerns, attempts to influence public policy' (Thomas, 2022).The terms 'interest group', 'interest organisations', and 'organised interests' are used interchangeably in this text.The notion of civil society includes a wider universe of groups, which are not necessarily based on an organisational structure (e.g., social movements) and/or do not aim to influence public policies (e.g., civic associations).See also footnote 2 for a discussion on the implication of using distinct terms.2. Following Putnam's seminal study, we can define social capital as 'trust, norms, and networks that facilitate cooperation and civic action' (Putnam, 1993, p. 167).3. A related field of research has focused on citizens' opinions regarding the involvement of interest groups in election campaigns.However, this strand of research draws essentially on the American experience and remains out of the scope of this article.4. Given that several terms can be used as synonyms for interest groups, one question asked whether the respondent's understanding of each of these terms was positive or negative.We used the following terms: pressure groups, interest groups, organised interests, civil society organisations and social partners.As scholarship also acknowledges (e.g., Beyers et al., 2008), there is no consensus on the best definition of interest groups and empirically we can find different terms associated with this concept.Our data show that whereas the term 'pressure group' received the most negative evaluation, social partners was the most positive (Table A1 in the Appendix).Our survey uses the most 'neutral' terms, namely interest groups or civil society organisations.From a theoretical point of view, this choice still captures the basic features of interest groups, namely the fact that they have (1) a formal organisation; (2) a membership (which is voluntary) and (3) public relevance (this was specified in the introductory question, see Table A1 in the Appendix).We also believe that this strategy avoids an important problem that lies behind Portuguese political culture (as in other Southern European countries), namely the fact that the terms 'interest groups', 'organised interests' or 'pressure groups' very often have negative connotations in public opinion, and are wrongly associated with (or considered equivalent) to corruption. 5.The question was the following: 'In your opinion, do you consider the general contribution of interest groups to democracy to be positive or negative?'

Figure 1 .
Figure 1.Trust in different types of organised interests, Source: Lisi et al. (2021). .Notes: (1) question wording: 'In general, do you think that most civil society organisations can be trusted or, on the contrary, do you think that care is little?' (2) 1: All care is little; 5: Most organisations can be trusted.

Figure 2 .
Figure2.Perceptions on the role of interest groups in the political system, Source:Lisi et al. (2021).Notes: (1) Figure2a, question wording: 'Some say that civil society organisations in Portugal are very necessary for the proper functioning of our political system.Other people think that these organisations are not at all necessary for the proper functioning of the political system.Using a scale between 1 (organisations are not necessary at all) to 5 (organisations are very necessary), where would you position yourself?';(2) Figure2b, question wording: 'In your opinion, do you consider the general contribution of interest groups to democracy to be positive or negative?';Scale: from 1 (negative) to 5 (positive).

Table 1 .
Predicting citizens' institutional support towards interest groups.Table presents the results of linear regression models (OLS).Question wording and variable coding are listed in the appendix.Significance levels: *p < 0.1, **p < 0.05, ***p < 0.01.

Table 2 .
Predicting citizen attitudes towards the role of interest groups for democratic governance (OLS regression).

Table 3 .
Predicting citizen attitudes towards the role of interest groups for democratic governance: full model (OLS regression).Table presents the results of linear regression models (OLS).Question wording and variable coding are listed in the appendix.Significance levels: *p < 0.1; **p < 0.05; ***p < 0.01.