Trust in political institutions

Trust in political institutions forms a fundamental pillar of democratic governance. Scholars broadly agree that a healthy level of trust facilitates civic engagement, ensures citizens’ willingness to accept policy outcomes, and encourages deference to authoritative decisions, even when those decisions are unpopular. David Easton explains that institutions need to attain and retain a reservoir of goodwill that allows them to withstand criticism and policy failures (Easton Reference Easton1965). Levi and Stoker (Reference Levi and Stoker2000) similarly stress that trust fosters cooperation between citizens and the state, particularly under conditions of uncertainty.

Building on these established foundations, scholars have shown how trust is shaped not only by institutional performance but also by broader social and process factors. Putnam (Reference Putnam2000) links “vertical” trust in government to “horizontal” social capital, while Luhmann (Reference Luhmann1979) emphasizes that trust reduces complexity, allowing individuals to interact with large institutions without constant oversight. Tyler (Reference Tyler2006) highlights perceptions of procedural fairness as central to perceived legitimacy, which in turn promotes citizen compliance with the law.

The relationship between citizens and elected officials further illustrates the strategic role of trust. Bianco (Reference Bianco1994) argues that constituents evaluate whether granting representatives discretion is likely to secure favorable outcomes, treating trust as a deliberate calculation rather than a vague attitude. This logic parallels the concept of “dynamic representation” (Stimson, MacKuen, and Erikson Reference Stimson, MacKuen and Erikson1995), in which officials adjust their positions in response to public preferences – or risk being replaced. Together, these perspectives underscore that representation is not static but shaped by continuous interaction, with trust providing elected officials latitude to lead and its withdrawal signaling the need for alignment.

Despite its importance, trust has declined sharply in recent decades. Across established democracies, confidence in institutions such as parliaments and parties has eroded (Valgarðsson et al. Reference Valgarðsson, Jennings, Stoker, Bunting, Devine, McKay and Klassen2025), raising concerns about democratic stability (Citrin and Stoker Reference Citrin and Stoker2018; Foa and Mounk Reference Foa and Mounk2017). Harold (Reference Harold2024) emphasizes that this crisis intersects with polarization, inequality, and disinformation, magnifying its impact on governance. Polarization stands out as a central driver in this case. Hetherington and Rudolph (Reference Hetherington and Rudolph2015) demonstrate that ideological and affective divides erode trust, as citizens grow less willing to support institutions when power rests with the opposing party. This “us versus them” mentality makes trust conditional and fragile, reducing the space for consensus (Hetherington Reference Hetherington1998). Relatedly, Bankert (2024) highlights the role of negative partisanship, in which animosity toward the other side – not attachment to one’s own – drives political identity. Unlike traditional partisanship, which can sometimes foster engagement, negative partisanship fuels hostility, skepticism about impartiality, and resistance to compromise, further undermining institutional legitimacy.

Media fragmentation amplifies these trends. The proliferation of ideologically siloed outlets allows selective exposure, reinforcing preexisting biases and creating echo chambers (Levendusky Reference Levendusky2013; Prior Reference Prior2007). Individuals increasingly consume content that confirms their beliefs, deepening polarization and eroding confidence in both media and the institutions portrayed through partisan lenses (Hammoud-Gallego et al. 2024). As Graber (Reference Graber2009) notes, the media not only covers politics but actively shapes how the public interprets it, often undermining shared foundations for trust. Economic anxieties provide an additional layer. Periods of instability, inequality, and perceived failures of governance heighten skepticism toward institutions. Roberts (Reference Roberts2017), for instance, shows that unmet economic expectations feed disillusionment, suggesting a performance-based dimension to trust. When institutions fail to deliver prosperity or security, confidence erodes even further.

Taken together, polarization, negative partisanship, media fragmentation, and economic anxieties create a formidable environment for sustaining trust. The result is a more skeptical public, reduced civic participation, and challenges for effective governance. Finally, several scholars have highlighted the dual nature of political trust. Intawan and Nicholson (Reference Intawan and Nicholson2018) argue that while explicit trust often appears low in surveys, an implicit, “gut-level” trust rooted in early socialization persists beneath the surface. This resilience helps explain why, despite widespread dissatisfaction, the political system retains public trust and support during crises (Huang et al. Reference Huang, Intawan and Nicholson2023).

Generally speaking, scholars examining mass trust in representative institutions suggest a myriad of potential drivers. Early studies of the electoral connection emphasized that citizens’ confidence in government is shaped by whether political institutions appear responsive and familiar to them (Gay Reference Gay2002). More recent research extends this insight to the state and local levels. In U.S. city governments, Heideman (Reference Heideman2020) finds that racial congruence between citizens and officeholders, perceptions of economic well-being, and personal civic engagement all boost trust, whereas racial homogeneity and fear of crime dampen it. At the state level, Wolak (Reference Wolak2020) reports that proximity to the state capitol, one’s party in power, and higher public spending raise trust, while economic distress, racial diversity, and educational attainment depress it. Taken together, these studies on electoral institutions reinforce the basic logic that citizens are more trusting of such entities when governments look like them, serve their tangible interests, and demonstrate value to the citizenry.

Institutional context may further shape how citizens feel about representative institutions. Banda and Kirkland (2018) demonstrate that polarized state legislatures erode trust, especially among politically attentive citizens – a finding that complements classic work showing that direct contact with one’s representative fosters confidence (Parker and Parker Reference Parker and Parker1993). In a similar manner, confidence in state courts may be driven by a variety of influences, including personal political efficacy, social capital, feelings toward government generally, electoral structures, and personal background and experience considerations, among other factors (e.g., Cann and Yates Reference Cann and Yates2016, 2021). For instance, citizens who serve as jurors and are exposed to symbols of justice and procedure express higher trust, while litigants or observers of unequal treatment report lower trust (Benesh Reference Benesh2006; Cann and Yates Reference Cann and Yates2008; Dougherty et al. Reference Dougherty, Lindquist and Bradbury2006; Wenzel et al. Reference Wenzel, Bowler and Lanoue2003). These wide-ranging findings highlight a key tension in the trust literature – whether legitimacy flows mainly from institutional design and dynamics, or rather, from citizens’ experiential judgments and personal identity.

While much of the literature aggregates trust in the justice system broadly, this approach overlooks key differences among legal actors. For prosecutors – even in a climate of skepticism – a baseline of implicit trust may sustain institutional public support and policy viability. In the next section we outline the place of prosecutors in the larger criminal justice system and suggest why public trust in them can be thought of in similar terms to other relevant government actors in certain ways, yet fundamentally distinct in other important ways.

Prosecutors – Power, elections, and public trust

In the American criminal justice system, local prosecutors hold an unparalleled position of power, a reality shaped significantly by their unique electoral status. Unlike federal prosecutors who are appointed, the United States stands alone as the only country where local prosecutors are elected. This practice, largely established between 1832 and 1860, saw nearly three-quarters of states shift from appointed to elected prosecutorial officials (Ellis Reference Ellis2012: 1530). Today, this electoral method prevails in forty-five of fifty states at the local district level, with only a handful of states opting for various forms of appointment or centralizing power with the Attorney General.Footnote ² This choice of selection method reflects a delicate balancing act, akin to state judicial elections: a desire for both independence and accountability in a powerful legal office.

The selection method for state judicial judges has been shown to affect public opinion of the judges and their actions (Peterson et al. Reference Peterson, Hare and Wrighton2012), as well as perceptions of other court actors like the Supreme Court (Hansford et al. Reference Hansford, Lucero, Intawan and Robles2024) and Supreme Court nominees (Krewson and Owens Reference Krewson and Owens2022). Whether a state uses an elective or an appointive system can alter citizens’ perceptions of the partiality of state court judges (Peterson et al. Reference Peterson, Hare and Wrighton2012; Barwick and Dawkins Reference Barwick and Dawkins2020). In similar fashion, citizens’ views of local prosecutors may be impacted by elections (which are more pervasive than in judicial selection) as well as election structures (e.g., partisan vs. nonpartisan method).

The prosecutor’s authority in the American criminal justice system is immense, largely due to broad statutory latitude and limited external checks on their discretion. State criminal codes often grant prosecutors extensive latitude, allowing them to decide not only whether to bring charges but also which specific provisions to apply, regardless of the range of conduct involved (Wright and Miller Reference Wright and Miller2010; Epps Reference Epps2021). Critically, judges and other courtroom actors have little effective means to challenge a prosecutor’s decision to decline charges entirely. Furthermore, most criminal cases are resolved through negotiated plea bargains orchestrated by the prosecutor, rather than by trial, effectively removing the question of guilt from a jury or judge. While judges retain the theoretical power to reject these bargains, they rarely do so (Hessick, Treul, et al. 2023, 43). This concentration of power means that any institutional check on prosecutors is largely channeled through the electoral process.

Consequently, the election of a prosecutor represents a profound delegation of power from citizens, with significant implications for public trust. Despite their critical role, these electoral contests are often characterized by low salience and a lack of competitiveness, especially in rural areas, where incumbents frequently face no challenger at all (Hessick and Morse Reference Hessick and Morse2020). These are typically low-information affairs, making it difficult for voters to assess past performance or hold incumbents accountable. The pool of potential challengers is shallow, partly due to the comparatively low pay for prosecutors relative to other attorney positions, and the uncomfortable dynamic of running against a sitting boss or future negotiating counterpart (Wright et al. Reference Wright, Yates and Hessick2021). The emergence of “progressive” challengers in recent years has, therefore, drawn considerable media attention precisely because it disrupts these traditional dynamics, demonstrating that even long-entrenched incumbents in large metro districts are facing increasing difficulty retaining office (Carson et al. Reference Carson, Cann, Yates and Wright2024; Wright et al. Reference Wright, Yates and Hessick2021). The degree to which these elections truly serve as an effective mechanism for accountability directly influences public perceptions of fairness and, by extension, trust in prosecutors and the broader legal system.

Recent scholarship also shows that the increasing nationalization of local elections is reshaping how citizens view prosecutors. Carson et al. (Reference Carson, Cann, Yates and Wright2024) argue that prosecutorial races – once regarded as low-information, locally bounded contests – are becoming increasingly tied to national partisan politics. As voters begin to interpret local prosecutor elections through a national lens, these contests may gain visibility, but heightened salience does not necessarily yield greater trust. Framing races in stark partisan terms risks importing national polarization into local justice, undermining confidence in prosecutorial neutrality. Thus, greater levels of nationalization raise visibility but also the stakes for public trust, making trust in local prosecutors more contingent on broader political currents than on local performance.

Although prosecutors wield immense discretionary power, political science has only recently treated them as trust-sensitive actors. Classic sociological work finds that prosecutors are more likely to allocate resources to high-profile cases with sympathetic victims – behaviors ostensibly designed to signal responsiveness to public safety concerns (Myers and Hagan Reference Myers and Hagan1978). Contemporary legal scholarship argues that prosecutors’ legitimacy now hinges on transparency. Vardsveen and Tyler (2022) contend that process-tracing tools – publicly explaining charging decisions, discovery practices, and plea-bargain rationales – can elevate trust by demystifying prosecutorial discretion. Yet Green (Reference Green2019b) cautions that media appearances alone rarely enhance trust; transparency must be coupled with checks on misinformation. Wright (Reference Wright2020) and Ouziel (Reference Ouziel2021) emphasize prosecutors’ local electoral incentives: because they depend financially and politically on county governments, their accountability remains rooted in voter oversight, but only if citizens have reliable information about office performance.

There is likely a good deal of overlap regarding how the public views and holds trust for different criminal justice government officials (e.g., enforcement, prosecution, and adjudication actors). While the dynamics of public trust among criminal justice officials may have much in common, we briefly outline here some reasons why the dynamics of public trust in local prosecutors might be distinct from the police and judges. First, while all three actors experience (on average) an electoral connection, it is most pervasive with local prosecutors. Only a handful of states do not elect local prosecutors whereas, as noted previously, judges are selected and retained by a varied patchwork of systems (e.g., appointment, merit selection, retention, etc.) – only one of which is competitive election. In similar vein, while some enforcement officials (e.g., sheriffs) are elected, other police actors (e.g., head of state highway patrol, municipal police chiefs, etc.) are typically appointed to office.

Second, unlike judges, most prosecutors’ policy decisions (e.g., decisions to charge, negotiated pleas, etc.) primarily take place behind closed doors – largely away from public and media scrutiny. Similarly, while police-citizen interactions are quite often subjects of news stories (with accompanying video), this is typically not the case with the work of local prosecutors. This is not to say that local prosecutors have no public-media interface, but much of their work and actions are opaque and shielded from external scrutiny. Finally, prosecutors occupy a unique and curious position in the confluence of branches of government. Like the police they are part of the executive branch and enforcement community. However, unlike police they are also officers of the court and part of the same legal profession as the judges they appear before. Further, while they are of the same profession as judges, they are not members of the judicial branch – i.e., those who possess neither purse nor sword. Prosecutors are, in fact, part of the sword – and perhaps depend somewhat less on public support to maintain their policy viability.

In short, while scholars agree that trust matters for prosecutors, they disagree on whether it is better secured through public outreach, institutional reform, or traditional electoral accountability. At present, scholarship on public trust perceptions of local prosecutors largely resides in legal journals and focuses on normative transparency – offering limited systematic evidence about the mass attitudes that reforms aim to influence. Linking theories of political trust specifically to prosecutors, we clarify when and why citizens trust these often “least visible” yet consequential actors. Using original survey data, we isolate the roles of ideology and experience, orientations toward government, and social dynamics, and pinpoint where transparency or electoral accountability is most likely to strengthen – or erode – prosecutorial legitimacy.

Theoretical development

The literature on public trust in political institutions offers numerous potential influences, with the choice of likely determinants often dependent on the institution being examined, the relative political context, or the specific theoretical premise being advanced. In estimating a model of the drivers of an individual’s decisions or views (whether an elite decision maker or regular citizen), we must account for how they come to perceive and interpret the world around them as well as the milieu of events, stimuli, and interactions that occupy their attentions and life experiences (see, e.g., Benforado and Hanson 2007; Yates et al. Reference Yates, Moeller and Levey2013).

In assembling our conception of the relevant determinants of individuals’ trust in local prosecutors we focus on: 1) The background and attitudinal concerns that may shape their outlook (i.e., personal identity, ideology, or life experiences); 2) the institutional structures they experience – specifically, how they see prosecutorial hopefuls attain the office (i.e., partisan or nonpartisan election or other); 3) how they consider other important government and enforcement actors such as police or judges; and 4) how they experience and conceive their relationships within the larger social environment and government (i.e., measured in terms of political efficacy and social capital). We examine each of these factors in more detail below.

Data and analysis

We assess these hypotheses using data from the 2022 Cooperative Election Study (CES), fielded by YouGov during the 2022 election (September 29–November 8, 2022). The CES is a large-scale, nationally representative sample. Administered online by YouGov, the CES employs a matched probability sample design in which a probability sample is drawn from the target population and then matched to similar respondents within YouGov’s panel. Roughly half of the survey consists of CES “common content,” while the remainder is “team content” created by participating institutions; we draw on both the common content and Emory’s 1,000-respondent team module, which includes a core battery of institutional trust items measured on a consistent response scale. These trust questions cover multiple institutions – including actors central to the criminal justice system such as the court system, judges, the police, and prosecutors – allowing us to benchmark confidence in prosecutors against closely related institutions using comparable measurement and standard post-election metadata and weights for analysis. Because we rely on some questions that are posed only on the Emory team module, we have a maximum of 1,000 observations. Non-response to various questions in the analysis leaves us with 675 observations in our analysis. A table of summary statistics of each of our variables appears in the online appendix.

The dependent variable is trust in prosecutors, measured on a 5-point scale asking respondents to rate their trust in prosecutors (always, most of the time, some of the time, about half of the time, some of the time, and never). We assess relationships between this variable and our independent variables using an ordered logit model; results of the model appear in Table 1. The overall fit of the model is reasonably good (χ² = 414.93, p <.001 and pseudo-r² = .22). To aid in interpretation, Table 2 shows predicted probabilities of an outcome in each of the 5 categories of trust in prosecutors (5 is always trust and 1 is never trust). Those predicted probabilities are calculated for each variable at its minimum and maximum values while holding all other variables constant at their means (or for dichotomous variables, their mode). The online appendix includes a figure that shows the change in the predicted probability of falling in the top two categories varying one variable from its minimum to maximum values while holding all variables constant at their mean values.

Table 1.

Ordered Logit Model of Trust in Prosecutors

Table 1. Long description

From the top, the table is divided into sections: Identity, Ideology, and Experience; Institutional Electoral Environment; Perceptions of Government; Personal Social and Efficacy Orientation; and Model Information. Under Identity, Ideology, and Experience, coefficients and standard errors are as follows: Minority negative 0.303 star, 0.176; Woman negative 0.351 star, 0.154; College negative 0.048, 0.156; Ideology negative 0.115 star, 0.048. Under Institutional Electoral Environment: Non-elected negative 0.449, 0.367; Nonpartisan negative 0.273, 0.239. Under Perceptions of Government: Trust in state courts 1.206 star, 0.107; Trust governor 0.219 star, 0.083; Trust state legislature 0.284 star, 0.105; Police make feel safe 0.340 star, 0.101. Under Personal Social and Efficacy Orientation: Social Capital 0.153 star, 0.088; Political Efficacy 0.311 star, 0.085. Model Information includes tau sub 1 1.594, 0.408; tau sub 2 4.242, 0.436; tau sub 3 6.269, 0.470; tau sub 4 10.023, 0.572. Likelihood ratio chi-squared 415.01, p less than .0001; Pseudo r-squared 0.22; n 675. A star indicates p less than .05, one-tailed.

* denotes p <.05, one-tailed

Table 2.

Predicted Probabilities from Ordered Logit Model

Table 2. Long description

The table contains nine variables, each with two rows for minimum and maximum values. Columns from left to right are Pr(y equals 5), Pr(y equals 4), Pr(y equals 3), Pr(y equals 2), and Pr(y equals 1), each with standard errors in parentheses. For Minority, at minimum: Pr(y equals 5) is 0.007 (0.002), Pr(y equals 4) is 0.22 (0.03), Pr(y equals 3) is 0.46 (0.03), Pr(y equals 2) is 0.28 (0.03), Pr(y equals 1) is 0.03 (0.01); at maximum: 0.005 (0.002), 0.17 (0.03), 0.44 (0.03), 0.34 (0.04), 0.04 (0.01). For Woman, at minimum: 0.01 (0.002), 0.28 (0.03), 0.47 (0.02), 0.22 (0.03), 0.02 (0.005); at maximum: 0.01 (0.002), 0.22 (0.03), 0.46 (0.03), 0.28 (0.03), 0.03 (0.007). For Ideology, at minimum: 0.01 (0.003), 0.28 (0.04), 0.47 (0.03), 0.22 (0.03), 0.02 (0.01); at maximum: 0.005 (0.002), 0.16 (0.03), 0.44 (0.03), 0.35 (0.04), 0.04 (0.01). For Trust in state courts, at minimum: 0.0006 (0.0003), 0.03 (0.01), 0.15 (0.03), 0.57 (0.03), 0.25 (0.05); at maximum: 0.07 (0.02), 0.70 (0.04), 0.19 (0.04), 0.03 (0.01), 0.003 (0.001). For Trust governor, at minimum: 0.004 (0.002), 0.16 (0.03), 0.43 (0.03), 0.36 (0.04), 0.05 (0.01); at maximum: 0.011 (0.004), 0.30 (0.05), 0.46 (0.03), 0.20 (0.04), 0.02 (0.01). For Trust state legislature, at minimum: 0.004 (0.002), 0.15 (0.03), 0.43 (0.03), 0.37 (0.05), 0.05 (0.01); at maximum: 0.013 (0.005), 0.35 (0.06), 0.45 (0.03), 0.17 (0.04), 0.02 (0.01). For Police make feel safe, at minimum: 0.003 (0.001), 0.12 (0.03), 0.39 (0.04), 0.42 (0.05), 0.06 (0.02); at maximum: 0.009 (0.003), 0.27 (0.03), 0.47 (0.02), 0.23 (0.03), 0.02 (0.01). For Social capital, at minimum: 0.005 (0.002), 0.17 (0.03), 0.44 (0.03), 0.34 (0.04), 0.04 (0.01); at maximum: 0.008 (0.003), 0.25 (0.03), 0.47 (0.02), 0.25 (0.03), 0.026 (0.006). For Political efficacy, at minimum: 0.003 (0.001), 0.13 (0.03), 0.41 (0.03), 0.39 (0.04), 0.05 (0.01); at maximum: 0.014 (0.005), 0.37 (0.06), 0.44 (0.04), 0.15 (0.03), 0.014 (0.004). Probabilities for y equals 3 are generally the highest across variables, with notable increases in Pr(y equals 5) and Pr(y equals 4) at maximum trust in state courts.

Our interpretation begins with factors related to identity, ideology, and experience when evaluating trust. Consistent with our expectations, racial and gender identity are both significant predictors of prosecutorial trust. This confirms our theoretical expectations that historically marginalized groups will perceive prosecutors in a more negative light. Substantively, the predicted probabilities in Table 2 indicate that being a minority member decreases the likelihood of being in the “always” or “most of the time” categories by about .05. For women, the probability of being in one of those top two categories decreases by about .07 relative to a man. These patterns underscore how lived experience and group identity deeply shape perceptions of prosecutorial legitimacy.

Ideology exerts a clear and negative effect: more conservative respondents are less trusting of prosecutors than liberals. This is consistent with our theorizing regarding reactions to the progressive prosecutor movement, which has the potential to decrease conservative support for prosecutors. The predicted probabilities demonstrate this vividly: the chance of falling into the “always” or “most of the time” categories of trust drops from .29 in the most liberal category of ideology to .17 among the most conservative, while the likelihood of falling in the lower categories rises accordingly. In contrast, college education does not exert any significant effect, suggesting that higher educational attainment and the experiences typically associated with it do not consistently translate into greater confidence in prosecutors.

Our second set of hypotheses concerned whether the method of prosecutorial selection – partisan election, nonpartisan election, or appointment – shapes public trust. Our results do not provide support for this claim. Neither residing in a state with nonpartisan elections nor one with appointed prosecutors significantly increases trust relative to partisan systems. This null finding suggests that institutional design, at least in the form of selection mechanism, plays a weaker role in shaping public opinion than do personal and attitudinal factors.

While method of selection doesn’t influence prosecutorial trust, orientations toward other government institutions do exert strong effects. Trust in state courts is the most powerful among these; this makes sense given the working relationship between courts and prosecutors. Moving from the minimum to the maximum of court trust raises the probability of “always trusting prosecutors” from virtually zero to .07, while boosting the probability of “trusting most of the time” from .03 to .70. Similarly, trust in governors and state legislatures exert somewhat smaller (but statistically significant) positive effects, suggesting that confidence in prosecutors is influenced not only by trust in courts but also within broader patterns of trust in state political institutions. Perceptions of police effectiveness and value also matter: respondents who report feeling safe because of the police are significantly more trusting of prosecutors, with the probability of “trust most of the time” more than doubling between the lowest and highest categories of this variable.

Finally, our measures of social capital and political efficacy both positively and significantly predict prosecutorial trust. Individuals who feel more connected to their communities and those who believe they can influence government are more likely to express confidence in prosecutors. The predicted probabilities highlight meaningful, though smaller, shifts: moving from the lowest to highest levels of social capital raises the probability of falling into one of the two highest categories of trust from .18 to .26 percent, while moving from lowest to highest political efficacy increases it from .13 to .38. These findings suggest that citizens’ sense of empowerment and embeddedness in civic networks translates into greater trust in prosecutorial authority.

Taken together, these results confirm that prosecutorial trust is not reducible to institutional design alone. Instead, such trust emerges at the intersection of personal identity, ideological orientation, confidence in other institutions, and broader social dynamics. The strongest drivers of prosecutorial trust appear to be trust in courts and other state institutions, ideological leanings, and feelings of political efficacy. Meanwhile, institutional features such as election method do not show measurable effects, indicating that reforms focusing solely on selection procedures may not substantially alter mass attitudes toward prosecutors.