Introduction
Since the seminal work by David Easton on systems analysis (Reference Easton1965; Reference Easton1975), the study of support for and trust in institutions in democratic systems has been key for political scientists. This is because trust in institutions is crucial for the stability of democratic regimes and, more generally, for the overall health of a democracy (Braithwaite and Levi, Reference Braithwaite and Levi1998; Hetherington, Reference Hetherington1998; Misztal, Reference Misztal1996; Warren, Reference Warren1999). Understandably, the vast majority of the scholarly literature has concentrated its efforts in studying institutional trust at the national level, given the key relevance of national governments in policymaking. However, in the last decades, central governments have devolved relevant shares of their competences to regional and local authorities (Jennings, Reference Jennings, Braithwaite and Levi1998; Hooghe et al., Reference Hooghe, Marks and Schakel2010): local governments have become so relevant (Weitz-Shapiro, Reference Weitz-Shapiro2008; Tang and Huhe, Reference Tang and Huhe2016) that several studies both in advanced and developing countries (Chanley, Rudolph and Rahn, Reference Chanley, Rudolph and Rahn2000; Christensen and Lægreid, Reference Christensen and Lægreid2005; Cleary and Stokes, Reference Cleary and Stokes2006; Cordova and Layton, Reference Cordova and Layton2016; Uslaner, Reference Uslaner2018; Zmerli and Van der Meer, Reference Zmerli and Van der Meer2017; Wolak, Reference Wolak2020) now acknowledge that trust in and support for local governments is at least as relevant as trust in national institutions.
Nevertheless, comparatively less attention has been devoted to trust in local institutions (Christensen and Lægreid, Reference Christensen and Lægreid2005; Rahn and Rudolph, Reference Rahn and Rudolph2005; Kelleher and Wolak, Reference Kelleher and Wolak2007; Cleary and Stokes, Reference Cleary and Stokes2006; Wolak, Reference Wolak2020). This is unfortunate, since the literature has underlined that there are good reasons to analyse trust in local and national institutions as two separate objects (starting from the fact that local and national trust follow different patterns, with the former being higher compared to the latter, Fitzgerald and Wolak, Reference Fitzgerald and Wolak2016).
In this respect, several contributions show that when differentiating between local and national trust, citizens focus on the performance of each institution (Mishler and Rose, Reference Mishler and Rose2001; Van Elsas, Reference van Elsas2015; Muñoz, Reference Muñoz, Zmerli and van der Meer2017). However, the literature has so far focussed mainly on the relationship between (objective) institutional performances (measured in terms of provision of local services) and trust (Stein et al., Reference Stein, Buck and Bjørna2021). In this paper, we advance a different and original perspective on the study of trust in local institutions which looks at the effect on local trust of the perceptions that citizens have about the quality of life in their neighbourhoods.
Borrowing from Huckfeldt’s (Reference Huckfeldt1986) intuition which suggests that the perceptions individuals have of their everyday life in their local space (i.e., their neighbourhoods) is crucial in shaping their political attitudes and behaviours, we conceive the quality of life in the neighbourhood as encompassing two conceptually distinguished factors: on the one hand, the perceived neighbourhood decay, which refers to the perception that individuals have about the physical space of the neighbourhood in which they live; and, on the other, the perceived periphery, which is related to how far essential public services are perceived to be from the individual’s neighbourhood. Both forms of neighbourhood discontent are then hypothesised to negatively affect trust in local institutions.
Although our locally based predictors are primarily hypothesised to affect trust in local institutions, we also elaborate on the (untested) idea that local level issues could also act as potential drivers of trust in national institutions. We deem this aspect to be a further innovative contribution of this paper. Indeed, while the existing literature has worked on the effect of national level variables on trust in local institutions, no paper has inquired so far whether and to what extent local level determinants affect trust in national institutions.
To test our framework, we rely on an individual level survey of Italian citizens provided by the Italian National Institute of Statistics-ISTAT (approximately 40.000 respondents). Employing a series of multilevel regression models, our results prove that both perceived periphery and neighbourhood decay are amongst the most powerful predictors of trust in local institutions (and, to a lesser extent, of national ones). Furthermore, we find that the relevance of these new predictors is not conditioned by economic, social, and cultural capital.
The sources of institutional trust
Political trust is generally defined as ‘a basic evaluative orientation towards the government founded on how well the government is operating according to people’s normative expectations’ (Hetherington, Reference Hetherington1998, p. 791). The literature that tries to explain the sources of trust and distrust is abundant (Levi and Stoker, Reference Levi and Stoker2000; Bertsou, Reference Bertsou2019) and it has provided very different explanations of trust in institutions. In this section, we summarize this literature in order to provide a general framework, which we then use as a point of departure for a more specific analysis of trust in institutions at both the local and national levels.
In their famous contribution, Mishler and Rose (Reference Mishler and Rose2001) distinguish between the cultural and institutional approach to the study of trust. Cultural approaches posit that the origin of political trust is exogenous to the political system and is mostly connected to the early socialization of individuals: ‘from birth, individuals learn to trust or distrust other people by experiencing how others in the culture treat them and how, in return, others react to their behavior’ (Mishler and Rose, Reference Mishler and Rose2001, p.34). This is a process that begins within families, and it is then extended to interactions and connections with other socializing agents (schools, pairs, etc.). As individual networks expand, people tend to increase their trust in other people (Brehm and Rahn, Reference Brehm and Rahn1997). This reciprocal trust is at the core of Putnam’s social capital (Reference Putnam, Leonardi and Nanetti1993, Reference Putnam2000) (i.e., voluntarism, membership in associations, extensive social networks, etc.). Social capital produces a civic culture whose values of interpersonal, social, and institutional trust are transmitted from generation to generation (Almond and Verba, Reference Almond and Verba1989; Putnam et al., Reference Putnam, Leonardi and Nanetti1993; Putnam Reference Putnam2000). So, for cultural approaches, interpersonal trust is positively connected with institutional trust.Footnote 1
The institutional approach, on the other hand, frames trust in institutions as politically endogenous (Mishler and Rose, Reference Mishler and Rose2001), meaning that it depends on the (perceived) capacity of the political system and its political institutions to perform in a satisfactory way. In other words, trust in institutions is generated within the political system and depends on the performance of political institutions.
Both cultural and institutional approaches have been articulated along two dimensions, i.e., macro and micro. Macro cultural approaches to institutional trust have mostly focussed on understanding how societies, characterised by different levels of interpersonal trust, also differ in terms of trust in institutions. However, while societies with a strong civic culture have, on average, higher levels of institutional trust, ‘socialization into a culturally homogenous society nonetheless allows substantial variation among individuals’ (Mishler and Rose, Reference Mishler and Rose2001, p.35). To account for the differences between individuals, cultural approaches have thus also focussed on the effect of individual socio-demographic characteristics on trust in institutions (i.e., micro cultural approaches to institutional trust). In this vein, age, social class, income, cognitive ability, and education have been widely analysed (Rohrschneider and Schmitt-Beck, Reference Rohrschneider and Schmitt-Beck2002; Campbell, Reference Campbell2004; Christensen and Lægreid, Reference Christensen and Lægreid2005; Anderson and Singer, Reference Anderson and Singer2008; Jennings et al., Reference Jennings, Stoker and Bowers2009; Hakhverdian and Mayne, Reference Hakhverdian and Mayne2012; Gustavsen et al., Reference Gustavsen, Pierre and Røiseland2017; Holmberg et al., Reference Holmberg, Lindberg and Svensson2017). The results of this literature proved to be mixed, meaning that the conventional wisdom for which highly sophisticated individuals with a high socio-economic-status trust more in institutions is true, although results vary a lot across countries and political contexts (Mishler and Rose, Reference Mishler and Rose2001; Dalton, Reference Dalton2004; Holmberg et al., Reference Holmberg, Lindberg and Svensson2017). Just to give an example, education is negatively related to institutional trust in corrupt societies, but the sign of the relation is reversed in clean societies (Hakhverdian and Mayne, Reference Hakhverdian and Mayne2012). Moreover, the literature has shown the importance of attitudes and psychological traits, such as internal efficacy, the big five personality traits, ideological orientations, and interest in politics (Newton, Reference Newton2001; McLaren, Reference McLaren2011; Holmberg et al., Reference Holmberg, Lindberg and Svensson2017; Ackermann et al., Reference Ackermann, Ackermann and Freitag.2019).
On a different note, macro institutional approaches have focussed on the effect on trust of those aggregate-level variables measuring the objective performance of institutions: GDP growth and levels of unemployment are the usual suspects, even though the significance of these variables and their direction is disputed (Chanley et al., Reference Chanley, Rudolph and Rahn2000; Mishler and Rose, Reference Mishler and Rose2001; Van der Meer, Reference van der Meer2010; Wolak, Reference Wolak2020). Other scholars have also focussed on the level of corruption, institutional transparency, and political scandals, showing that corruption and lack of transparency undermines institutional trust (Chanley et al., Reference Chanley, Rudolph and Rahn2000; Mishler and Rose, Reference Mishler and Rose2001; Andersen and Tverdova, Reference Anderson and Tverdova2003; Morris and Klesner, Reference Morris and Klesner2010; Van der Meer, Reference van der Meer2010; Hakhverdian and Mayne, Reference Hakhverdian and Mayne2012; Grimmelikhuijsen et al., Reference Grimmelikhuijsen, Porumbescu, Hong and Im2013).
The micro level approach, instead, has focussed on individual perceptions of country- or local-level performances, showing that subjective economic evaluations are more powerful predictors of institutional trust than objective ones and that sociotropic evaluations are more relevant than egotropic ones (Mishler and Rose, Reference Mishler and Rose2001; Dalton, Reference Dalton2004; Wroe, Reference Wroe2016). More recent studies have also shown that individual perceptions of economic insecurity (mainly due the negative effects of de-industrialization, technological change, and de-unionization, see Wroe, Reference Wroe2016) have a negative effect on institutional trust. Other works have looked at the perception of public services offered by the institutions. These studies concentrate mainly on the local level (Kelly and Swindell, Reference Kelly and Swindell2002; Bouckaert and van de Walle, Reference Bouckaert and van de Walle2003; Downe et al., Reference Downe, Cowell, Chen and Morgan2013; Ellinas and Lamprianou, Reference Ellinas and Lamprianou2014) and use satisfaction with services such as education, health, transportation, bureaucracy or local councillors’ ethical behaviour as explanatory variables for the support of institutions. Other indicators prove to be equally important, i.e., process and probity (Citrin and Stoker, Reference Citrin and Stoker2018). Here the idea is that when citizens deem the institutional processes to be fair and public officials are considered to be honest, institutional trust increases (Borre and Andersen, Reference Borre and Andersen1997). Overall, the literature shows that both subjectively- and objectively-measured performance matters (e.g., DeHoog et al., Reference DeHoog, Lowery and Lyons1990; Fitzgerald and Wolak, Reference Fitzgerald and Wolak2016). To the extent local and national institutions (are perceived to) deliver, support for these institutions increases as well.
While both approaches to institutional trust (i.e., cultural and institutional) provide fruitful insights on the drivers of support for institutions, our contribution to the study of local (and national) trust takes the micro-level institutional approach as a point of departure, although, as we shall see, in fact, it goes beyond it. Local trust (as a form of specific support) has been mainly analysed considering local institutions as service providers and citizens as service consumers who evaluate institutional performances and, as a consequence, either trust or distrust the institutions (Kelly and Swindell, Reference Kelly and Swindell2002; Bouckaert and van de Walle, Reference Bouckaert and van de Walle2003; Gustavsen et al., Reference Gustavsen, Pierre and Røiseland2017; Christensen et al., Reference Christensen, Yamamoto and Aoyagi2020). However, local public services, in our understanding, represent just a narrow conception of local institutions’ performances. In this regard, the literature has given scant attention to one possible crucial aspect, that is the role played by broader perceptions of quality of life in local communities, something that we discuss in the next section.
Perceived physical environment as a determinant of institutional trust
As stated above, the literature on local trust has focussed mainly on the relationship between the provision of (local) services and trust (Stein et al., Reference Stein, Buck and Bjørna2021). Only very recently, McKay et al. (Reference McKay, Jennings and Stoker2021) introduced perceived social marginality and perceived economic deprivation in their analysis of political trust. We partly change the perspective, shifting our focus from individual economic status to the perceptions that citizens have about the quality of life in the neighbourhood. We intend the latter as a more encompassing phenomenon compared to the perceived quality of public services or the perceived social and economic deprivation. Under the label of quality of life, we include two concepts: first, the perception of well-being in the neighbourhood, referring to the perceptions about the quality of the physical space where individuals live (hereafter, neighbourhood decay). Second, the more service-oriented concept of periphery, i.e., how distant individuals perceive they are from basic public services (hereafter, periphery). Notice that our focus here is on the perceptions of individuals. Indeed, while objective measures of individual distance could be a good proxy for marginality, it is also important to highlight that they are not able to capture the quality of life in the neighbourhood. It is possible that individuals live relatively (and objectively) close to public services, but still the quality of life in the neighbourhood is not perceived in positive terms.
In both cases, the mechanism that might link perceived physical space in the neighbourhood and institutional trust is analogous. On the one hand, people perceiving the place they live as far-away from the ‘centre’, not just in terms of physical distance, but also in terms of public services provision, might feel abandoned by the local authorities and, if this effect goes beyond the local level, by the national authorities as well. The literature has already shown, in this regard, the importance of developing individual place attachment to revitalise the neighbourhood (Brown et al., Reference Brown, Perkins and Brown2003). On the other, if individuals perceive a low level of general well-being in the neighbourhood, they might consider the institutions as incapable of delivering results. In our understanding, the two mechanisms should be kept separate as they capture two distinct aspects of urban life, which do not necessarily coexist. A clear example here is the emergence of completely new neighbourhoods around shopping centres and malls, in a process of reconfiguration of public spaces that involves both advanced and developing countries. Separated from the city centres, malls are by now considered by many as new contemporary squares (Staeheli and Mitchell, Reference Staeheli and Mitchell2006), and it is exactly around these contemporary squares that new forms of urban aggregation emerge. Responding to the needs of providing a reassuring environment to consumers, these new urban spaces are kept clean and tidy (Staeheli and Mitchell, Reference Staeheli and Mitchell2006). Yet, the organisation of essential services takes time. And, as people are attracted by the initially low prices for housing and rents in these new neighbourhoods, these spaces become rapidly populated, but often not sufficiently served by public services. Thus, despite the absence of urban decay, citizens living in these places tend to be disconnected from city centres.
More in detail, for us periphery is the perceived distance of individuals from basic and essential services, such as hospitals, supermarkets, and drugstores. In this respect, while we cannot assume a priori that closer public services would be better than public services located far away from the place of living, having them close means at least that local communities try to connect the area and to provide a specific service. Also, while the extent to which an individual belongs to a periphery has been measured through the distance from the capital or to the ‘centre’ of a specific area (Wolak, Reference Wolak2020), we believe that a ‘subjective’ perception of periphery should better capture feelings of disconnection from real (or imagined) urban, political, and social centres. In this perspective, the perceived distance from basic services assumes the symbolic meaning of exclusion.
As stated above, the perceived distance from public services is not necessarily related to the well-being in the neighbourhood. Nonetheless, well-being in itself has been seldom analysed in the literature. Even though the recent literature has shown a connection between (perceived) local government performance and trust in local institutions (with citizens being capable of properly judging the performance of public services) (Van Ryzin, Reference Van Ryzin2007), these studies have focussed on specific public policy outcomes and their relationship with good governance (Bouckaert and van de Walle, Reference Bouckaert and van de Walle2003), rather than on the quality of the individuals’ surrounding environment. More recent contributions show the importance of the centre-periphery cleavage in shaping trust (Mitsch et al., Reference Mitsch, Lee and Morrow2021). Well-connected areas that took advantage of globalisation sharply differ from left-behind areas (post-industrial cities and suburban communities), where globalisation impacted negatively on the quality of life and social division from the ‘centre’ grew exponentially (Jennings and Stoker, 2017; Reference Jennings and Stoker2016): the reaction against the mainstream is evident in these areas (Rodríguez-Pose, Reference Rodríguez-Pose2018). Physically measured periphery, measured as the distance from the capital, is also negatively connected with trust in politicians, according to a recent study (Stein et al., Reference Stein, Buck and Bjørna2021).
All this given, to build our theoretical argument we move from the intuition by Huckfeldt (Reference Huckfeldt1986), who stressed the critical relevance of neighbourhood contexts in forging individuals’ political attitudes and behaviours. In this respect, the empirical literature has abundantly demonstrated that neighbourhood perceptions are linked to a number of attitudes and behaviours (Ellaway et al., Reference Ellaway, Macintyre and Kearns2001), place attachment (Brown et al., Reference Brown, Perkins and Brown2003), and social capital (Kleinhans et al., Reference Kleinhans, Priemus and Engbersen2007) amongst others. Individual life trajectories within the neighbourhood are not built in a vacuum, but within a physical space which inevitably affects how individuals perceive the context in which they live and how they perceive themselves within this context.
On this backdrop, we argue that liveable areas, i.e., non-polluted areas, with low levels of crime, low traffic, clean spaces, new roads etc., should reinforce trust in institutions, as citizens can evaluate on a daily basis the responsiveness of the local community to an individual’s basic needs. Citizens living in such areas have the possibility to see with their eyes if the local community is able to deliver.
Thus, we posit that the perception of the surroundings is the closest proxy citizens have to evaluate the performance of local community officials, and thus a potential key driver of support for local government. As Pfeiffer and Cloutier (Reference Pfeiffer and Cloutier2016, pp. 270–272) show in their review essay, there is a growing body of literature suggesting that the physical characteristics of a neighbourhood (housing design, street connectivity, availability of public spaces, perceived personal security, green areas) may lead to more or less social engagement, and ultimately, happiness. According to Morris (Reference Morris2019), it is subjective well-being that is the main component that distinguishes suburbs and cities, other conditions being almost similar. Thus, while it is true that the ‘revenge of the places that don’t matter’ (Rodríguez-Pose, Reference Rodríguez-Pose2018) is related to policy outputs, such as income inequality, it is also true that it is fostered by subjective perceptions of the physical space in which individuals live, i.e., through the personal evaluation of being (not) part of the places excluded from public services and forgotten by local politics.
Our main assumption is that perceptions of marginality and abandonment might spur distrust in those institutions that, for proximity reasons, might be more easily identifiable as responsible for (and thus blamed for) such malaises. Furthermore, while the quality of life can reasonably affect trust in local institutions, we cannot exclude that these perceptions may also produce variations in trust in national institutions as well. In particular, as long as neighbourhood contexts contribute both to the socialization of individuals and to forge their political attitudes and behaviours, the experience and perceptions of such neighbourhoods might well be connected to more general attitudes towards national politics. After all, the neighbourhood defines the breeding ground where political and social interactions occur in the first place, and where community bonds are constructed (Putnam, Reference Putnam2000). To the extent these physical spaces are perceived as deteriorated, the quality of social interactions as well as the general levels of interpersonal and institutional trust might be negatively affected. For example, as Cohen and Dawson (Reference Cohen and Dawson1993) have shown, among African Americans, that social isolation produced by neighbourhood decay tends to alienate this group from politics, making them less likely to participate.
In this respect, our argument differs from the existing literature since we problematize the relationship between individuals’ perception of their surrounding and the evaluation of institutions’ performance. On the one hand, from a rational choice perspective, citizens reward or punish institutions, which are responsible for providing specific services and improve the quality of the neighbourhood. On the other hand, it might be equally true that the perception of the surroundings, in which the political socialization of the individuals occurs, is so pervasive that individuals translate this blaming mechanism to the national level as well. As a consequence, while the literature so far has looked at national level performances as partially responsible for the levels of trust locally, we want to explore whether the quality of life in the neighbourhood is responsible for national level trust. Or, to the opposite, all performances confined to the local level do not affect national trust, as individuals separate between the spheres of influence of the institutions and thus correctly blame/reward only the institution responsible for delivering a specific public service. We acknowledge that these mechanisms might not be the same across individuals (see e.g., Proszowska et al., Reference Proszowska, Jansen and Denters2023; Muñoz, Reference Muñoz, Zmerli and van der Meer2017): for example, less educated individuals could be more likely to blame both governmental levels for their perceived marginality, being basically less likely to clearly attribute responsibility to (and thus blame) one or the other level of government. However, our main goal is to check whether the quality of life in the neighbourhood (in its twofold specification, i.e., neighbourhood decay and periphery) affects local trust primarily (and national trust secondarily).
Different twins? Local and national trust in comparison
So far, our argument has been based on the implicit assumption that trust in local and national institutions are different twins. This assumption, however, needs to be justified. The literature analysing commonalities and differences between trust in local and national institutions can be organised along the lines of two different approaches. On the one hand, a strain of the literature treats trust as a ‘monolithic’ bloc: citizens do not have the cognitive skills to differentiate between local and national institutions (Sniderman, Reference Sniderman and Finifter1993; Denters, Reference Denters2002) and they express their attitudes towards the (perceived) most relevant institution, usually the national institution, and then translate this attitude towards other institutional levels (we refer to this approach as the institutional salience approach) (e.g., Proszowska et al., Reference Proszowska, Jansen and Denters2023).
On the other hand, other authors argue that citizens differentiate across different tiers of governments, although through different mechanisms (we refer to this approach as the differentiation approach). Within this approach, proximity theories, for example, maintain that closeness to institutions matters, as closer institutions are considered by citizens as more reliable and trustworthy (Proszowska et al., Reference Proszowska, Jansen and Denters2023; Denters and Klok, Reference Denters and Klok2013; Hansen, Reference Hansen2013). Re-elaborating the paradox of distance stated by Frederickson (Reference Frederickson1997), Christensen and Laegrid (Reference Christensen and Lægreid2005, p. 488) claim that ‘[w]hereas people trust government officials who are nearby, they believe the government officials who are far away are lazy, incompetent and probably dishonest’. According to others, instead, differentiation across institutional levels is produced by rational evaluations of citizens, who assess the performance of different institutions on their own merits (e.g., Mishler and Rose, Reference Mishler and Rose2001; Van Elsas, Reference van Elsas2015; Muñoz, Reference Muñoz, Zmerli and van der Meer2017). In this perspective, both national and local trust are forms of specific support, mostly determined by how well/bad the citizens evaluate the performance of different institutions. Wolak (Reference Wolak2020) for example, shows that trust in state government in the USA is associated with the character and performance of the state government, rather than being derived from trust in national institutions. Similarly, Steenvoorden and van der Meer (Reference Steenvoorden and van der Meer2021, p. 9), focussing on the Dutch case, concludes that ‘local political performance matters for local political support’, although a certain degree of overlap between trust in local and national institutions do in fact exist.
This review suggests that while the more general support of national institutions affects local trust, the latter is influenced by the performance of the local institutions, and as such, should be analysed separately from national trust. In addition, while the differentiation between local and national trust is theoretically sound, it also offers to us the chance to test whether and to what extent our indicators of the quality of life in neighbourhoods affect not only local trust but also national trust. This is, in our perspective, a further innovative aspect of this paper. Indeed, while the literature has mostly focussed on the effect of national trust on local trust, in fact, assessing how the performance of national institutions might affect trust in local institutions, we test here the potential inverse relationship, that is to say the impact of local determinants on trust in national institutions.
Hypotheses
Given our emphasis on the relevance of subjective perceptions of the places in which individuals live, our first hypothesis clearly connects local institutional trust to perceived neighbourhood conditions. As argued above, more intense feelings of peripherality and urban decay might be linked to the inefficiency of local institutions to respond to the citizens’ needs. We thus hypothesise that:
1.1 The higher the perceptions of periphery of the neighbourhood and (1.2) the higher the perceptions of the neighbourhood’s decay, the lower the trust in local institutions.
Although responsibility for neighbourhood decay and marginality can be more easily attributed to local governments (which are the closest level of government to local communities and thus more easily identifiable), it is not possible to exclude that the perception of the quality of life in the neighbourhood might be then reflected into more general feelings of distrust in the overall national system. As we argued above, as neighbourhood contexts contribute to the formation of political attitudes and behaviours, the experience and perceptions of such neighbourhoods are hypothesised to be connected also to more general attitudes towards national politics. For example, research on the surge of protest and populist votes has widely shown that (objective and subjective) conditions of deprivation at the local level are associated with better electoral performances of protest and populist parties, which often mobilised voters on an anti-systemic rhetoric (Schulte-Cloos and Leininger, Reference Schulte-Cloos and Leininger2022). On this basis, we hypothesise that:
1.3 The higher the perceptions of periphery of the neighbourhood and (1.4) the higher the perceptions of the neighbourhood’s decay, the lower the trust in national institutions.
While we posit that neighbourhood decay and periphery have a negative effect both on local and national trust, we also expect that our focal predictors should have stronger effects on local trust. This is because, based on our theoretical argument, both perceived neighbourhood decay and periphery should be primarily connected to individual experiences with local institutions.
Also, we advance further expectations about possible interactive mechanisms with other predictors. We take into consideration here whether individual resources mitigate the negative effects of the perceived environment. We do so by leveraging the CARs (i.e., capitals, assets and resources) approach to social stratification and borrowing from the simple intuition according to which individuals enter the social life with ‘different endowments, capacities and resources’ (Savage, Reference Savage2015, p. 49), which are inherited from their past and produce either constrictions or enhanced opportunities for their future. We assume that individual resources (and agency) might allow citizens to experience, in a differentiated way, the everyday complexities associated with a peripheral and/or decayed physical environment.
In line with Savage and colleagues (Savage et al., Reference Savage, Warde and Devine2005), we posit that three different types of individual resources (i.e., capitals) can differentiate the way in which citizens experience their physical environment and, as a consequence, their level of trust in political institutions. These are: economic, cultural, and social capital. Economic capital refers to the overall wealth of individuals and it broadly indicates how far individuals are from material needs (Savage, Reference Savage2015). Cultural capital broadly concerns cultural credentials that individuals can leverage in society. It refers to formal education, but also to tastes and interests, that are either socially approved or disapproved of (i.e., socially legitimated) (Savage, Reference Savage2015, p. 92). In particular, the concept of cultural capital encompasses broad cultural tastes and interests which are a revealing symbol of an individual’s position in society. As the literature does not agree on the role of education in reinforcing institutional trust, with a higher level of education both acting as a driver for trust in institutions but also for distrust in institutions, we rather investigate the role of cultural capital while controlling for the education level of individuals, given its great social equalizer role.
Finally, social capital relates to human and socially relevant relations that generally produce positive effects. Putnam et al., (Reference Putnam, Leonardi and Nanetti1993) and Putnam (Reference Putnam2000) refers to social capital as an asset fostered by civic engagement and extensive social networks that are able to produce positive effects both at personal and social level. It thus refers to the extension and density of social networks, which enormously affect individuals’ life chances. As Savage puts it (2015, p. 52) ‘contacts can accumulate over time and may be a resource which can be mobilized to gain information about jobs, accommodations, exciting opportunities and so forth’. Social connections are therefore resources that structure opportunities in society.
In general, our expectation is that those who are endowed with a high economic, cultural, and social capital, even if living in a peripheral and decayed neighbourhood, can still leverage their material and immaterial resources to overcome systemic inefficiencies. To give few straightforward examples: lack of public transportation can be overcome by using private cars or paying for a cab; people with a higher level of sophistication might find it easier to have information to reduce the material and immaterial distance with the centre (cultural capital); people with high social capital, can rely on large networks of friends for personal (emotional) and material (economic) support; and they are more likely to be informed about chances and opportunities within the neighbourhood (but also outside it).
We thus formalise the following hypotheses:
2. Higher levels of (2.1) economic, (2.2) social, and (2.3) cultural capital reduce the negative impact of perceived neighbourhoods’ periphery and decay on institutional trust (both local and national).
Data and methods
Case study and data sources
We test our hypotheses focussing on a peculiar case study, Italy. This is a country with one of the lowest levels of trust in national institutions in Western Europe and the country where Putnam et al., (Reference Putnam, Leonardi and Nanetti1993) proposed his argument about social trust and on how democracy could perform better when individuals trust each other and participate in public life. Despite this work, there are no studies that focus on this context of the relation between the individual perception of the physical space and local (and national) institutional trust. We believe that, despite their limited generalisability, discussing our hypotheses in relation to the Italian case is interesting in itself. First, because of the lack of recent studies on the sources of variation in institutional trust in this country; and second, because of the originality of our argument. We point out here that while previous studies focussed on the perception of public services and objective distance from city centres to capture the malaise of individuals, we make a step forward. We posit that, on the one hand, the perception of the environment in which citizens live cannot be simply related to the (perceived) quality of services, as life in neighbourhoods implies a broader set of interactions with the physical space. On the other, we maintain that the perceived peripherality of the neighbourhood is something to be analysed separately from the perception of neighbourhood decay.
Given the novelty of the theoretical approach to the study of trust, fine-grained data are needed to test our hypotheses. From this point of view, Italy represents a unique opportunity, given the large amount of individual-level data measuring perceived marginality and decay of neighbourhoods in exceptionally large-N datasets which are provided by the Italian National Institute of Statistics (ISTAT). We rely here on survey data gathered under the project ‘Aspects of daily life’ in 2017. Footnote 2 This is one of the largest surveys available for Italy and it covers the resident population in private households, by interviewing through P.A.P.I. technique a range of 20.000–50.000 individuals. Our samples vary from 41.108 to 41.127 respondents.
Dependent variables
The dependent variables are two questions on local and national trust. For national trust, we focus on a question on trust in National Parliament (as a robustness check we replicated the model using as a dependent variable an additive index, which includes a question on the trust in the European Parliament besides the National Parliament, see below). For trust in local government, we rely on a specific question related to trust in local authorities (as robustness checks we replicated the model using as a dependent variable an additive index, which includes questions on the trust in provincial and regional governments, see the online Appendix, Section 1. The robustness checks are in line with the findings presented in the main text, see Appendix, Section 3).Footnote 3 . As mentioned earlier, there are good theoretical reasons for expecting local trust and national trust to behave differently: albeit a partial overlap is possible, the two concepts are not identical and the drivers for both sources of trust might be different according to the literature. To empirically show that local and national trust do not necessarily move together, we run two factor analyses in which we include all trust-related questions included in our survey. The first factor analysis includes five trust questions about national and subnational institutions: our factor analysis shows that the question on national and European parliaments (the first two questions) load in a different dimension, compared to the sub-national ones (regional, provincial, and local authorities), which are part of another dimension (for more details see the discussion of the results in the Appendix, Section 5, Figure 15A). The second factor analysis adds to the picture the questions about trust in non-governmental nationwide institutions, the judiciary and the police. This factor analysis shows that while justice loads – although with a relatively poor factor loading – in the second dimension, trust in police loads in a third dimension. Thus, even when including all political institutions, we detect two dimensions, one related to local trust, the other related to national trust (for more details see the Appendix, Section 5, Figure 16A). Thus, the first empirical test reassures us that it is worth distinguishing local trust and national trust.
Independent variables
Our main independent variables, which we labelled as neighbourhood decay and periphery, are two different additive indexes. To measure perceived periphery, we rely on additive indexes measuring how peripheric the place of living is according to respondents’ perceptions. The index is composed of seven questions (Cronbach Alpha 0.9). Respondents were asked to what extent they find it difficult reaching the following services: pharmacy, emergency room, kindergarten, local markets, postal office, municipal office, and police station. The levels range from ‘no difficulties at all’ to ‘a lot of difficulties’. As for the neighbourhood decay, the index is composed of eight questions (Cronbach Alpha 0.8). Respondents were asked to what extent the following aspects are present in the neighbourhood: uncleaned streets, air pollution, crime, bad street lighting, traffic, bad smell, noise, parking difficulties. The levels range from ‘not at all’ to ‘a lot’.
To control the effects of our main predictors on local and national trust, we include variables related to the different ‘capitals’ of respondents (i.e., economic, social, and cultural). Economic capital is measured as an additive index comprising of three questions (Cronbach Alpha 0.7) related to family income situation, the economic resources of the family, and the satisfaction of the personal economic situation (see the Appendix for the wording).
Secondly, cultural capital is built as an additive index comprising of three questions related to how frequently interviewees went to the cinema, theatre, museum/exhibition in the last year. These are ordinal variables, with five levels ranging from ‘never’ to ‘more than 12 times’. This measure is in line with our definition of cultural capital as a broader concept compared to formal education. However, we are also aware that education is most notably used to measure cultural credentials of individuals. Therefore, we always use education as a control variable.
Finally, we include an additive index to measure the social capital, which is composed of fourteen questions (Cronbach Alpha 0.7). We provide a broad definition of social capital to tap all dimensions related to this multifaceted concept: the first dimension concerns political participation for which we include four variables. Respondents were asked whether they took part in such activities in the last 12 months. The second dimension relates to civil society participation, which includes questions about participation in different types of associations’ meetings, volunteering for one of these associations, and donating money. The third dimension is related to work participation, which includes questions about participation in trade unions’ meetings, professional associations, volunteering for a trade union.
Sociodemographic controls include age (cardinal), gender (1 = Female), and education. Education is an ordinal variable, going from 1 (‘Lower than high school’) to 4 (‘Tertiary’). We also include a variable for political interest, which is an additive index (Cronbach Alpha 0.75) of two variables measuring respectively the frequency of political information and political discussion (1 = ‘Everyday’; 6 = ‘Never’). We control for the political interest of respondents, since, as we specified also above, it is a relevant source of political trust (Catterberg and Moreno, Reference Catterberg and Moreno2006). To account for the overall economic condition of the respondent we add a control for employment status (1 = ‘Employed’). Finally, we also include two regional-level variables as further controls: the GDP per capita and the unemployment rate in 2017. Unfortunately, the dataset does not include provincial/municipality-level identifiers so we could not include more refined data on this aspect. We include these two regional-level variables to control for Type 1 causal mechanisms, for which, among other factors, trust is linked to the economic performance of a country (a region in this case).
All the descriptive statistics of our independent and dependent variables are presented in the Appendix, Section 2.
Modelling
We opt for a multivariate multilevel model, in which individuals are clustered in regions (20 + 1, see below). This N is not unreasonably low, even if we acknowledge its potential limitations when it comes to maximum likelihood estimation (Stegmueller, Reference Stegmueller2013). We include all twenty Italian regions plus another category, which includes ‘other’ responses. This is a very marginal category (N = 138, 0.3% of our sample): even when removing it, the results do not change. In this case, for regional-level variables we input the regional-level score for the GDP per capita and unemployment (for the category ‘other’, we input the national mean). We first create two different models, in which we change the dependent variable (local and national trust). Second, we estimate a set of interactive models in which we test whether economic, cultural, and social capitals modify the effects of both perceived neighbourhood decay and periphery. We report in the main text the Figures with the main effects, while full model specifications are available in the Appendix (Section 3, Tables 2A, 4A, 5A, 6A).
Discussion
Local and national trust explained
Figure 1, Panel A, shows that all variables identified by the literature as potential drivers of institutional trust are indeed significant in the expected direction, when it comes to local trust. The most relevant independent variable is the neighbourhood decay, which is negatively related to local trust (b = −0.16, P < 0.001); thus, the more people perceive that the quality of life of their neighbourhood is low, the higher is the distrust in local institutions. Also, our periphery index proves to be significant and negative, so the more people perceive that the local services provided by the public authority are far away from their place of living, the lower is the institutional trust (b = −0.019, P < 0.001). Local trust, thus, is apparently related to both socio-economic status (economic, cultural, and social capital) and to the way people perceive that the local institution is delivering for them. Overall, without rejecting the classical explanations for trust in local authorities, we also find confirmation for our hypotheses 1.1. and 1.2.
The determinants of local (Panel A) and national (Panel B) trust. Note: ***P < 0.001; **P < 0.01; *P < 0.05.
In addition, the economic capital is the variable with the second strongest predicting power in the model (b = 0.147, P < 0.001), but also the cultural (b = 0.015, P < 0.001) and social capital (b = 0.04, P < 0.001) contribute to local institutional trust. The only exception to this trend is education, which has a negative and significant effect (although it should be made clear once again that the effect of education can meaningfully go in both directions).
The fact that local and national trust do not follow the exact same logic is apparent when looking at Figure 1, Panel B, reporting the results of our regression analysis for national trust. Focussing on our main independent variables,Footnote 4 the results show that periphery is no longer significant, while neighbourhood decay is still significant at P < 0.001 level, albeit its effect decreases by two-thirds (but remains amongst the most powerful predictors). In this sense, respondents might perceive local authorities as responsible for providing the basic services and, thus, target the local institutions as the main ones that are responsible for (not) providing them. This blame/reward attribution does not translate plainly into the national institution, though. So, the local institutions are perceived as responsible for providing close-to-home basic public services, not the national ones. However, the quality of the neighbourhood in which respondents live does affect national trust. The results hold when switching our dependent variable from a single item to the two composite indexes for local and national trust (see Table 3A in the Appendix).
Our main independent variables, and in particular neighbourhood decay, were never tested before, so we advance a potential explanation, which is related to the fact that being socialized in a neighbourhood with a lower quality of life has a more profound impact on respondents’ evaluations of the political institutions as such, if compared with the perceived functioning of public services. The quality of the surroundings of respondents can be the very first ‘political’ evaluation that individuals make about the role of the institutions in general: regardless of the distance of basic services, living in a neighbourhood where the perceived problems make the quality of life lower might push respondents to lose confidence not just in what local authorities do, but in institutions in general. We thus find strong empirical confirmation of hypothesis 1.4., while we reject hypothesis 1.3.
Individual capitals as moderators of neighbourhood decay and periphery
In this section we test hypotheses 2.1., 2,2., and 2.3., for which economic, cultural, and social capitals moderate the effects of our main independent variables on local and national trust. We tested these moderating effects by interacting the three measures for individuals’ capitals with our indicators of neighbourhood decay and periphery. The results of these analyses are visible in Figures 2, 3, and 4, where we plot the estimated coefficients of neighbourhood decay and periphery on local/national trust at different levels of economic, cultural, and social capital (fully specified models are available in Tables 4A, 5A, 6A of the Appendix). Overall, counter to our hypotheses, we observe that, the negative effect of neighbourhood decay on both local and national trust increases when economic and social capital increase as well. The only exception is related to the moderating effect of cultural capital. In this case we find (consistently this time with our hypotheses) that when cultural capital increases, the negative effect of decay on local trust decreases, while it remains almost the same when national trust is considered. This might suggest that, contrary to our expectations, people who are endowed with richer capitals might be more affected by neighbourhood decay because they are more equipped to recognize the inability of both local and national institutions to deliver a better quality of life. We find a similar, yet not identical, trend when considering perceived periphery. With the exception of the moderating effect of economic capital, we observe that those who do have higher cultural and social capital are more negatively affected by perceived periphery compared to those who have low cultural and social capital.
Effects of Neighbourhood decay and Periphery by levels of Economic capital on Local Trust (top-right and top-left plots) and on National Trust (bottom-right and bottom-left plots).
Effects of Neighbourhood decay and Periphery by levels of Social capital on Local Trust (top-right and top-left plots) and on National Trust (bottom-right and bottom-left plots).
Effects of Neighbourhood decay and Periphery by levels of Cultural capital on Local Trust (top-right and top-left plots) and on National Trust (bottom-right and bottom-left plots).
All this given, these interaction effects proved to be by no means statistically significant, with the only (partial) exception of the cultural capital which is significant at P < 0.01 for neighbourhood decay and at P < 0.05 for periphery at the local level (the results are robust when using the two indexes of local and national trust as dependent variables, see Appendix, Section 4, Figures from 3A to 14A).Footnote 5 Thus, the results go against our hypotheses 2.1., 2,2. and partially against 2.3., meaning that individual assets and resources of the respondents do not mitigate the effect of the perceived neighbourhood decay and periphery on their trust in local and national institutions. In other words, no matter how wealthy, and well-connected people are, where they think they live is in itself a good indicator of the level of trust they have in institutions.
Conclusion
This paper proposed a different theoretical perspective for interpreting institutional trust in local (and national) institutions among citizens. We focussed on the effect of an under-analysed factor, i.e., perceived quality of life. Based on our theoretical argument, we distinguished two aspects of quality of life, which, although related, should be treated separately, as our findings show. On the one hand the perceived neighbourhood decay, which is aimed at measuring how individuals evaluate the environment in which they live and are socialized; on the other, individuals’ perceived periphery, which is related to the perceived distance of citizens from services.
We believe that this paper provided a useful contribution for three different reasons. Firstly, it empirically offers new indicators that go beyond classic subjective policy-output perceptions of citizens in evaluating local institutional trust: in this regard, our indicators move in the intersection of cultural and institutional approaches, highlighting the importance of keeping both into consideration when analysing institutional trust. Our results show also that local and national trust are at least partially different ‘objects’: even though the perceived quality of the surrounding is amongst the most relevant factors that is associated with trust in local and national institutions, we also show that output-oriented indicators related to the peripherality of the neighbourhood have an effect on the local trust only. Finally, and related to the previous point, it provided evidence that the quality of the surrounding is so relevant that it is not conditioned by the embeddedness of the single citizen in the society.
Trying to reconnect these findings with the existing literature on the differences between national and local trust, it seems to us that the institutional salience and differentiation approaches have both their own merits. In line with the differentiation approach, the evaluation of local public services is associated with trust in local institutions, but it does not affect national trust (i.e., citizens rationally evaluate local institutions on their own merits). However, we also do find important overlapping between the two types of trust, as suggested by the institutional salience approach: aside from socio-demographic variables, neighbourhood decay has an effect on trust at both levels, thus suggesting that the quality of life in the neighbourhood is more than a mere evaluation of a (local) public service and targets a more pervasive malaise. One of the important advancements of our paper lies exactly on this finding: while the differentiation approach has focussed exclusively on the impact of national level evaluations on local trust, our paper shows that local evaluations matter for national trust as well, meaning that where you live explains how much you trust local and national institutions.
We believe that our theoretical framework and our empirical findings can be a useful tool to increase our knowledge on the sources of institutional trust: similar operationalizations of our variables in other contexts might confirm (or disconfirm) what we have found for Italy. We don’t know whether the effects of quality of life on local trust vary depending on the institutional setting of the country. Italy is in between a federal structure and a fully centralized one, with a certain degree of autonomy granted to local governments, but also with a quite strong central state. One might hypothesize that the effect of our key predictors on trust in local institutions could be stronger in federal states (or in states with strong local governments), as in these contexts citizens could more easily identify local authorities as responsible for their own malaise. Conversely, the same variables could have a stronger effect on trust in national institutions in cases of strongly centralized states. Although we have no answer for now to these questions, we believe that our contribution represents a starting point to develop further research. Finally, these findings might be of interest for policy-makers and policy-analysts: in an era of democratic malaise, our paper shows that one way (not the only one) to improve citizens’ trust in political institutions (be these local or national) is to take particular care of the context in which citizens live. This is because institutions which deliver at the local level make citizens more trustful about them
Supplementary material
To view supplementary material for this article, please visit https://doi.org/10.1017/S1755773923000140.
Acknowledgement
The authors are grateful to Leonardo Morlino for his useful comments on a preliminary version of the paper. The authors also wish to thank the three anonymous reviewers and the journal editors for their comments and support.
Financial support
Davide Vittori has received funding from the European Research Council (ERC) under the European Union’s Horizon 2020 research and innovation programme (grant agreement No. 773023).
