Principal investigator: Omar Barroso Khodr
Authors: Stephan Grimmelikhuijsen, Marija Aleksovska, Judith van Erp, Sharon Gilad, Libby Maman, Tobias Bach, Moritz Kappler, Wouter Van Dooren, Rahel M. Schomaker, Heidi Houlberg Salomonsen
Original title: Does enforcement style influence citizen trust in regulatory agencies? An experiment in six countries
Location of the Intervention: Belgium, Denmark, Germany, Israel, Netherlands and Norway
Sample Size: 5.765 responses
Primary Variable of Interest: Trust in regulatory agencies
Type of Intervention: Political and regulatory sensitivity analysis
Methodology: Cross-border experiments; Questionnaires; Linear regression
Summary
Establishing and maintaining citizen trust is vital for the long-term effectiveness and viability of regulatory agencies. However, little empirical research has been conducted on the relationship between regulatory action and citizen trust. The article by Grimmelikhuijsen et al. (2025) addresses this gap by investigating the influence of different regulatory enforcement styles on citizen trust. The authors conducted a pre-registered and representative research experiment in six countries (n = 5.765): Belgium, Denmark, Germany, Israel, the Netherlands, and Norway. The study focuses on three key dimensions of regulatory enforcement style: formalism, coerciveness, and accommodation. The hypothesis was that a strict and punitive enforcement style, with minimal accommodation, would increase citizen trust. Surprisingly, no overall effect of regulatory enforcement on trust was found. However, particularly high levels of formalism (rigidity) and coerciveness (punitiveness) showed a small positive effect on trust. Furthermore, no discernible impact of an accommodative style of regulatory enforcement was observed. Further analysis revealed that the effects of enforcement style were not consistent across countries and regulatory domains. This suggests the need to reconsider the assumptions underlying law enforcement theory, as the findings indicate that public trust appears less conditioned by strict law enforcement than initially anticipated.
- Policy Problem
According to the authors, regulatory agencies face great difficulty in establishing and maintaining public trust, despite this trust being essential for their legitimacy and effectiveness. This challenge stems from two main factors: (1) risk regulation is inherently uncertain—the harms avoided are invisible, while regulatory failures are easily observable and amplified by the media; (2) although there is extensive literature on how institutional factors (transparency, accountability, independence) affect trust, very little is known about the impact of central regulatory decisions—specifically, law enforcement styles (enforcement and punishment)—on public trust.
In this way, the authors emphasize that this gap is critical, since enforcement actions are the most visible manifestation of regulatory activity and, therefore, potentially decisive for citizens' trust. The article thus seeks to answer whether different styles of application (formalism, coercion, and accommodation) affect citizens' trust in regulatory agencies.
- Policy Implementation Context
The context of implementing regulatory policies involves not only the concrete actions that regulators take in response to rule violations, but, above all, the nature of the interactions established with those being regulated during this process. Although law enforcement (using the English term) enforcementAlthough largely defined by legal mandates, regulatory agencies generally have considerable discretion to adopt different enforcement styles, which may vary in emphasis.
In this way, these styles are understood as institutionalized and consistent approaches of the agency, organized into three main dimensions: the degree of formalism (rigidity versus flexibility in the application of rules), the degree of coerciveness (punitiveness of sanctions, which can range from warnings and educational actions to fines and lawsuits), and the degree of accommodation (consideration or not of the regulated entity's perspective in decisions). These dimensions are central because they directly shape the relationship between the regulator and citizens, who, as trustees, expect competence, benevolence, and integrity from the agencies.
Finally, the implementation context, therefore, is not limited to the enforcement itself, but also encompasses how law enforcement decisions are communicated and perceived publicly, potentially influencing citizens' trust in regulatory agencies.
- Evaluation Details
The study adopted a cross-national vignette-based experiment to test the effect of regulatory enforcement style on citizen trust. The experiment was conducted between June 21 and July 6, 2021, on representative samples of citizens from six countries: Belgium, Denmark, Germany, Israel, the Netherlands, and Norway. Data collection was carried out through an online questionnaire conducted by the company. Kantar, which recruited respondents from its volunteer panels, respecting pre-defined quotas to ensure representativeness.
In this context, after consenting to participate, respondents provided basic demographic data, such as gender, age, and education level. They then viewed three vignettes—one for each regulatory domain analyzed (food safety, finance, and data protection)—presented in random order. After each vignette, participants reported their level of trust in the corresponding regulatory agency.
Furthermore, within each vignette, participants were randomly assigned to one of eight treatment conditions (combinations of high and low levels of the three dimensions of enforcement style—formalism, coerciveness, and accommodation) or to the control group, which contained no information about the enforcement of the regulation.
Additionally, the experiment included manipulation checks to ensure that participants perceived the treatments as expected. These checks were performed both in a previous pilot study (with approximately 100 respondents per country) and at the end of the main experiment, being applied only to the last vignette shown to avoid bias, respondent fatigue, and premature disclosure of the study's objectives.
Therefore, citizens' trust in regulatory agencies was measured using a multidimensional scale based on the ABI (ability, benevolence, integrity) model, which captures three dimensions of trust: competence (ability to perform tasks professionally), benevolence (consideration of the public interest), and integrity (the agency's honesty). A reduced version of a validated scale was used, with three items measured on a 7-point scale. The dependent variable "trust" was constructed as the average of the responses to the three items, showing high internal consistency (Cronbach's alpha between 0,88 and 0,94 across all countries and sectors).
In this way, to ensure data quality, responses completed in 90 seconds or less were excluded, according to a pre-established exclusion rule. After this filtering, 5.765 responses were retained from a total of 6.077 respondents. Robustness analyses that included the excluded respondents showed no substantial differences in the results, indicating a low risk of post-treatment bias. The project received ethical approval from all participating universities and adopted rigorous measures to protect the privacy of respondents.
- Method
The aim of the study was to evaluate the effects of three dimensions of regulatory enforcement style — formalism, coerciveness, and accommodation — on citizens' trust in regulatory agencies in the food safety, finance, and data protection sectors. Initially, the researchers presented an overview of average trust levels by regulatory domain in each of the six countries analyzed (Belgium, Denmark, Germany, Israel, Netherlands, and Norway).
Thus, the descriptive results indicated that levels of trust in regulatory agencies were slightly above the neutral point on the scale (4 on a 7-point scale) in most countries, being highest in the Netherlands and Norway, and lowest in Israel. Furthermore, it was observed that the data protection and food safety sectors enjoyed higher levels of trust than the financial sector.
In this context, to test the general hypothesis that the mere existence of information about the application of regulations positively affects trust, the authors compared the trust of respondents assigned to the control group with that of those assigned to the treatment conditions.
Thus, the results showed that information about regulatory enforcement led to higher levels of trust in the financial regulator in the German sample and in the data protection regulator in the Israeli sample, but undermined trust in the data protection regulator in the Norwegian sample. A robust regression analysis with pooled data from all countries and sectors indicated that, overall, information about actions did not generate significant differences in respondents' trust assessments.
Next, the study specifically analyzed the effects of each dimension of enforcement style. Regression models revealed that both high coerciveness and high formalism (as opposed to low levels) led to greater citizen trust in regulatory agencies across all three sectors. However, these positive effects were small in magnitude: shifting from low to high levels of coerciveness resulted in an average increase of 0,14 points on the 7-point trust scale, while the effect of formalism was 0,13 points. The effect of accommodation was inconsistent across domains, reaching statistical significance only in the financial sector, where high accommodation increased trust by 0,08 points.
Additionally, to assess the robustness of the findings, the researchers disaggregated the results by country and sector. In the food safety sector, high formalism significantly increased confidence in the samples from Denmark and Israel; high coerciveness had a positive effect in Germany and Israel; and high accommodation reduced confidence only in Denmark.
Finally, in the financial sector, formalism did not affect trust in any country; coerciveness increased trust in Denmark, Norway, and the Netherlands; and accommodation increased trust in Norway, contradicting the initial hypothesis. In the data protection sector, formalism increased trust in Israel and the Netherlands; coerciveness had a positive effect only in Norway; and accommodation reduced trust in Israel but increased it in Germany.
- Main results
The study investigated how different styles of regulatory enforcement (formalism, coerciveness, and accommodation) affect citizens' trust in regulatory agencies in the food safety, finance, and data protection sectors across six countries. Overall, the results showed that simply informing citizens about the existence of enforcement actions had a minimal and inconsistent effect on trust, making it impossible to confirm the hypothesis that any form of regulatory enforcement increases trust. When specific dimensions were analyzed, formalism (rigidity in the application of rules) and coerciveness (punitiveness of sanctions) produced positive, albeit small, effects on citizens' trust.
In this context, accommodation (consideration of the regulated entity's perspective) yielded mixed and inconsistent results, sometimes increasing and sometimes decreasing trust, depending on the country and sector. Furthermore, the effects did not replicate robustly across different countries and regulatory domains, indicating that public trust is less influenced by punitive or rigorous law enforcement styles than the literature initially suggested. In short, the hypotheses received limited and partial support, suggesting the need to revise the theoretical assumptions about the relationship between enforcement and public trust.
- Lessons in Public Policy
Grimmelikhuijsen et al. offer important lessons for policymakers and managers of regulatory agencies. The main conclusion is that a heavy-handed or punitive approach to regulatory enforcement—characterized by high formalism and high coerciveness—can only increase citizen trust to a limited extent and is highly context-dependent. This means that regulators should not assume that tightening enforcement and sanctions will, by itself, be an effective and universal strategy for gaining public trust.
On the contrary, the results suggest that citizens may have more sophisticated and nuanced judgments than previously thought: they likely expect agencies to adapt their enforcement style to the specific circumstances of each case. For example, citizens may support formalistic and punitive stances toward repeat or intentional offenders, but prefer greater flexibility and leniency toward negligent first-time offenders. This lesson points to the need for agencies to adopt proportionate and context-sensitive enforcement styles, rather than generalized rigidity.
Furthermore, the study reveals that the effects of enforcement styles vary significantly between countries and regulatory sectors, possibly influenced by recent incidents or the institutional environment of each location. This indicates that there is no single formula or one that can be directly transferred between different contexts. Public managers should therefore consider national and sectoral specificities when designing their enforcement strategies, avoiding automatically replicating practices from other countries without adaptations.
Another relevant lesson is that agencies need to be mindful of how they communicate their actions to the public, since citizens' trust is shaped not only by the content of decisions, but also by the perception of competence, benevolence, and integrity of regulators.
Finally, the study suggests the need to replicate the experimental approach in jurisdictions with less robust institutional safeguard mechanisms and to deepen comparative qualitative research on the role of regulatory behavior in high- and low-trust contexts. In short, the public policy lessons indicate that blindly relying on punitive styles to build public trust may be a mistake; instead, flexibility, case-by-case adaptation, and attention to contextual peculiarities are recommended as more promising paths to effective and legitimate regulation.
Average confidence in regulators by country and sector, with 95% confidence intervals. The confidence scale in the figure ranges from 1 – very low confidence to 7 – very high confidence. Countries from left to right: Israel, Germany, Denmark, Belgium, Norway, and the Netherlands. Source: Grimmelikhuijsen et al. (2025).
