Determinants of commuter tolerance toward underage and unlicensed bus drivers: an empirical investigation using the structural and institutional tolerance of enforcement (SITE) framework

Abstract

Introduction:

Road crashes remain a major public health concern in Bangladesh. Official records from the Bangladesh Road Transport Authority documented 5,495 crashes and 5,024 deaths in 2023, followed by 5,856 crashes and 5,480 deaths in 2024. Dhaka Metropolitan Police records covering 2011–2022 further indicate that 45.24% of recorded crashes involving drivers younger than 18 years were fatal, although these records included multiple vehicle categories and did not isolate public buses. Despite this substantial safety burden, limited attention has been given to why commuters tolerate public transport services operated by underage or unlicensed drivers.

Methods:

This study developed and empirically assessed the Structural and Institutional Tolerance of Enforcement (SITE) framework, which integrates Institutional Theory, Social Norm Theory, and the Psychometric Paradigm of Risk Perception. The framework examined Public Tolerance in relation to Perceived Enforcement Weakness, Corruption Normalization, Economic Empathy, Social Norm Acceptance, Personal Risk Perception, and Regulatory Trust. Data were collected through structured face-to-face surveys with 320 adult commuters at six major transit hubs in Dhaka and analyzed using partial least squares structural equation modeling and direction-sensitive importance–performance map analysis.

Results:

The measurement model demonstrated satisfactory reliability and validity, while the structural model explained 45.5% of the variance in Public Tolerance. Perceived Enforcement Weakness, Corruption Normalization, Economic Empathy, and Social Norm Acceptance were positively associated with tolerance, whereas Personal Risk Perception and Regulatory Trust were negatively associated with it. Personal Risk Perception had the largest path coefficient in absolute magnitude, while Economic Empathy had the largest positive coefficient. The importance–performance analysis also showed that Personal Risk Perception had the largest total effect in absolute magnitude, whereas Regulatory Trust recorded the lowest performance score.

Discussion:

Overall, the findings indicate that commuter tolerance is associated with interconnected institutional, social, and psychological conditions. The study identifies evidence-informed policy domains involving risk communication, transparent licensing and enforcement, professional credential verification, operator accountability, and institutional trust-building in support of Sustainable Development Goals 3, 11, and 16.

1 Introduction

Road traffic injuries remain a major public health and development challenge. Approximately 1.19 million people die in road crashes worldwide each year. About 92% of these deaths occur in low- and middle-income countries, although these countries account for only approximately 60% of the global vehicle fleet (World Health Organization, 2023). The burden is particularly severe in rapidly motorizing cities, where increasing travel demand frequently coincides with inadequate infrastructure, limited regulatory capacity, and inconsistent enforcement (Pojani and Stead, 2017). Within this broader challenge, the operation of public transport vehicles by underage or unlicensed drivers represents a serious but insufficiently examined road safety concern.

Bangladesh provides an important setting in which to examine this issue. The Road Transport Act of 2018 establishes minimum age, educational, physical, mental, and competency requirements for obtaining a driving license. Applicants must be at least 18 years old to obtain a nonprofessional license and at least 21 years old to obtain a professional license (Government of the People’s Government of the People’s Republic of Bangladesh, 2018). These provisions legally prohibit individuals below the statutory professional licensing age, as well as other legally unqualified persons, from operating buses and comparable public transport vehicles. Nevertheless, previous studies have documented persistent noncompliance with licensing, vehicle fitness, and traffic safety regulations in Bangladesh’s transport sector (Solaiman, 2021; M. A. Islam and Bhuiyan, 2024; Saha et al., 2021). The reported presence of underage or unlicensed drivers therefore reflects a continuing gap between formal regulation and everyday transport practices.

Recent statistics demonstrate the severity of the national road safety burden. The Bangladesh Road Transport Authority recorded 5,495 road crashes, 5,024 deaths, and 7,495 injuries in 2023. Buses and minibuses accounted for 1,830 of the vehicles involved in these reported crashes (TBS Report, 2024). In 2024, the official totals increased to 5,856 crashes and 5,480 deaths (Adhikary, 2025). Independent monitoring by the Road Safety Foundation documented 7,584 crashes, 7,359 deaths, and 16,476 injuries nationwide in 2025, including 409 crashes, 219 deaths, and 511 injuries in Dhaka (United News of Bangladesh, 2026). Because the 2025 estimates were compiled from newspapers, online news portals, electronic media, and the foundation’s monitoring sources, they should be distinguished from the official BRTA series rather than interpreted as directly comparable annual observations.

Evidence specifically concerning driver age and licensing status remains limited. A recent study examined crash records maintained by the Dhaka Metropolitan Police between 2011 and 2022. Among recorded crashes involving drivers younger than 18 years, 45.24% were classified as fatal, 11.90% involved grievous injuries, and 19.05% involved simple injuries. Possession of a valid driving license was also associated with a 28.2-percentage-point lower predicted probability of a fatal crash outcome (Islam et al., 2026). These findings indicate that driver age and licensing status are associated with crash severity. However, the study included several vehicle categories and did not isolate public buses. Available annual reports also do not simultaneously classify bus crashes by driver age and licensing status. Consequently, a precise annual count of crashes involving underage or unlicensed bus drivers is not currently available. This data limitation demonstrates the need for more systematic reporting of driver qualifications in public transport crash records.

The persistence of legally noncompliant transport operations raises a question that extends beyond driver behavior and regulatory enforcement. Commuters may perceive that a driver is too young, learn that the driver lacks the required license, or otherwise become aware of possible legal ineligibility, yet still board the vehicle, continue using the service, or refrain from objecting. This study conceptualizes such responses as Public Tolerance. Public Tolerance refers to a commuter’s attitudinal willingness to accept, continue using, or overlook a public transport service when the driver is perceived, reported, or known to be underage or unlicensed. The construct does not measure whether commuters can independently verify a driver’s age or licensing status. It instead captures their willingness to accept the service after the potential legal ineligibility of the driver has become salient.

Public Tolerance does not necessarily indicate ignorance of the associated danger or approval of the legal violation. It may instead reflect adaptation to weak enforcement, normalized noncompliance, economic considerations, prevailing social expectations, or limited confidence that individual objection will produce meaningful change. Continued passenger use may help sustain demand for noncompliant services, while the absence of objection may reinforce perceptions that such practices are common or unavoidable. Commuters should therefore not be viewed solely as passive recipients of transport risk. Their attitudes and responses may either reproduce prevailing unsafe norms or contribute to demands for safer and legally compliant services.

Existing transport safety research has largely adopted either a driver-focused or an institution-focused perspective. Driver-focused studies have examined driving errors, violations, attitudes, risk-taking, and behavioral intentions (Reason et al., 1990; Armitage et al., 2022; Elliott, 2010). Institution-focused research has examined legislation, deterrence, enforcement practices, and regulatory effectiveness (Elvik, 2021; Bates et al., 2012). These research traditions provide important explanations of unsafe driving and regulatory compliance. However, they generally treat passengers as contextual elements rather than as actors whose attitudes and travel choices may sustain or challenge unsafe transport practices. Consequently, limited empirical attention has been given to the institutional, social, and psychological conditions associated with commuters’ willingness to accept services operated by legally unqualified drivers.

This study addresses this gap by developing the Structural and Institutional Tolerance of Enforcement (SITE) framework. The framework integrates Institutional Theory (Meyer and Rowan, 1977; DiMaggio and Powell, 1983), Social Norm Theory (Cialdini et al., 1990; Cialdini and Trost, 1998), and the Psychometric Paradigm of Risk Perception (Slovic, 1987). It explains Public Tolerance in relation to six antecedents: Perceived Enforcement Weakness, Corruption Normalization, Economic Empathy, Social Norm Acceptance, Personal Risk Perception, and Regulatory Trust. The theoretical foundations of these constructs and their hypothesized relationships with Public Tolerance are developed in Section 2.

The study contributes to transport safety research in three respects. First, it conceptualizes and measures commuter tolerance as an attitudinal outcome within a legally noncompliant public transport environment. Second, it integrates institutional, social, and individual psychological explanations within a unified theoretical framework. Third, it estimates the relative associations and explanatory importance of the proposed determinants, thereby identifying policy-relevant domains that warrant further assessment. The hypothesized relationships are examined using partial least squares structural equation modeling (PLS-SEM), while importance-performance map analysis (IPMA) is used to compare the relative importance and performance profiles of the antecedent constructs. Because the study uses cross-sectional data, these results are interpreted as theory-consistent associations rather than causal effects or demonstrated intervention impacts.

The study is also relevant to Bangladesh’s commitments under the Sustainable Development Goals. The operation of public transport vehicles by legally unqualified drivers may undermine progress toward reducing road traffic deaths under SDG 3, improving access to safe transport under SDG 11, and strengthening effective and accountable institutions under SDG 16. Examining the conditions associated with commuter tolerance may therefore inform the development of licensing and enforcement measures, risk communication strategies, operator accountability mechanisms, and initiatives intended to strengthen regulatory trust and challenge prevailing social norms.

Accordingly, the study addresses the following research questions:

RQ1: How are Perceived Enforcement Weakness, Corruption Normalization, and Regulatory Trust associated with commuter tolerance of underage and unlicensed bus drivers in Dhaka?

RQ2: How are Economic Empathy, Social Norm Acceptance, and Personal Risk Perception associated with commuter tolerance of underage and unlicensed bus drivers?

RQ3: Which institutional, social, and psychological determinants exhibit the strongest associations with Public Tolerance and the most policy-relevant importance-performance profiles?

The remainder of the paper is organized as follows. Section 2 reviews the relevant literature, presents the theoretical foundations of the SITE framework, and develops the research hypotheses. Section 3 describes the research design, sampling procedure, measurement instrument, and analytical methods. Section 4 reports the measurement and structural model results, together with the importance-performance analysis. Section 5 interprets the findings and discusses their theoretical and policy implications. Finally, Section 6 summarizes the study’s contributions, limitations, and directions for future research.

2 Theoretical framework and hypothesis development

2.1 Theoretical and comparative foundations of the SITE framework

This study develops the Structural and Institutional Tolerance of Enforcement (SITE) framework to examine the institutional, social, and psychological conditions associated with commuters’ acceptance of public transport services operated by underage or unlicensed drivers. Public Tolerance refers to a commuter’s willingness to board, continue using, or overlook such a service after perceiving, learning, or being informed that the driver may be legally or professionally ineligible to operate the vehicle. This response does not necessarily indicate approval of the violation. It may instead reflect adaptation to regulatory failure, prevailing social practices, economic considerations, or limited confidence that individual objection will produce meaningful change.

The regulatory problem examined in this study is not unique to Bangladesh, although its institutional and operational context differs from those of other countries. Bangladesh, Thailand, and Hungary all impose minimum-age, competency, and professional qualification requirements on passenger transport drivers. In Bangladesh, the Road Transport Act of 2018 establishes a minimum age of 21 years for obtaining a professional driving license and requires applicants to have completed at least the eighth grade, demonstrate physical and mental fitness, and pass the prescribed driving competency examination (Government of the People’s Republic of Bangladesh, 2018). These provisions formally prevent children and other legally unqualified individuals from operating buses and comparable public transport vehicles.

Thailand similarly applies differentiated eligibility requirements to public passenger transport drivers. Applicants for type 1, type 2, and type 3 driver licenses for scheduled, non-scheduled, or small-vehicle passenger transport must generally be at least 22 years old, whereas applicants for a type 4 license must be at least 25 years old. The licensing process also includes prescribed training, testing, and other eligibility requirements (Government of Thailand, 2026) At the European Union level, category D and DE licenses for buses and coaches generally require a minimum age of 24 years, although licensing from 21 years is permitted when the applicant holds the required certificate of professional competence (European Parliament and Council of the European Parliament and Council of the European Union, 2025). Hungary applies this broader European licensing structure through its national regulations. Category D and DE licenses generally apply from the age of 24, while they may be issued from the age of 21 subject to specified conditions and domestic-use restrictions until the driver reaches the general age threshold. Hungarian regulations also require an applicant for category D authorization to already hold category C driving entitlement (Government of Hungary, 2011). Although the specific legal provisions differ across these jurisdictions, each system recognizes that operating passenger transport vehicles requires qualifications beyond those required for ordinary private driving.

Evidence from Thailand and Europe further indicates that formal licensing requirements do not, by themselves, ensure safe passenger transport operations. Research from Thailand has shown that vehicle type and operator legal status are relevant to the safety performance of interprovincial bus services (Ratanawaraha et al., 2019). Other Thai evidence has identified driver behavior, crash characteristics, vehicle conditions, and environmental conditions as factors associated with bus-crash injury and fatality severity (Mahikul et al., 2022). European research has similarly associated unlicensed driving with elevated crash involvement and risky behaviors, including speeding and alcohol or drug impairment (Sagberg, 2018; Møller and Janstrup, 2021).

These international examples provide comparative regulatory and road-safety context rather than suggesting that Bangladesh, Thailand, and Hungary are institutionally equivalent. Direct research on commuter tolerance toward underage or unlicensed bus drivers remains limited across these settings. Nevertheless, the comparative evidence indicates that the central problem is not simply the absence of formal age and competency requirements, but the institutional and social conditions that may allow violations to persist despite established legal restrictions. This perspective strengthens the rationale for examining the conditions associated with commuter tolerance in Bangladesh.

The SITE framework integrates Institutional Theory (Meyer and Rowan, 1977; DiMaggio and Powell, 1983), Social Norm Theory (Cialdini et al., 1990; Cialdini and Trost, 1998), and the Psychometric Paradigm of Risk Perception (Slovic, 1987). Institutional Theory explains how persistent discrepancies between formal rules and routine practices can weaken the practical authority of regulation. Social Norm Theory explains how perceptions of commonly observed behavior shape judgments about what is normal or socially accepted. The Psychometric Paradigm explains how subjective assessments of danger, severity, controllability, and personal vulnerability influence responses to hazardous conditions.

These three perspectives constitute the core integrative foundations of the SITE framework. Deterrence Theory, Moral Disengagement Theory, Attribution Theory, and procedural justice and institutional legitimacy perspectives are used as complementary construct-level explanations. They refine the theoretical basis of the individual antecedents without changing the central institutional, social, and psychological architecture of the SITE framework.

The SITE framework does not treat its core theoretical perspectives as separate explanations that correspond independently to particular constructs. Instead, it proposes an interconnected process through which institutional experience may shape social interpretation and individual judgment. Repeated enforcement failure may signal that licensing requirements have limited practical consequences, while the normalization of corruption may explain why such requirements are routinely circumvented. Together, these institutional conditions may strengthen commuters’ perceptions that tolerance of underage and unlicensed driving is common, expected, or difficult to challenge.

This process clarifies the theoretical relationships among Perceived Enforcement Weakness (PEW), Corruption Normalization (CN), and Social Norm Acceptance (SNA). PEW and CN are theoretically expected to strengthen SNA because SNA captures the perception that tolerance is a prevailing response among other commuters. At the same time, weak enforcement and normalized corruption may undermine the competing injunctive norm that underage and unlicensed driving should be opposed. Commuters may therefore regard the practice as legally and morally unacceptable while simultaneously perceiving public tolerance of it as normal or unavoidable.

Economic Empathy (EE) may reinforce this normalization process by providing a moral justification for continued use of the service. Commuters may recognize the legal and safety implications of the practice but remain reluctant to object when the driver’s involvement is attributed to poverty, limited employment opportunities, or household economic pressure. In contrast, Personal Risk Perception (PRP) and Regulatory Trust (RT) may counteract normalization. Greater risk perception maintains awareness of the potential consequences of unsafe travel, whereas regulatory trust reinforces the legitimacy and public-safety purpose of licensing requirements.

These interconnected mechanisms provide a theoretical explanation of how institutional, social, and psychological processes may jointly relate to Public Tolerance within the SITE framework. However, the empirical model in the present study estimates only the direct relationships between the six antecedents and Public Tolerance. It does not test mediation, moderation, or statistical interaction effects. Accordingly, pathways such as PEW SNA TOL and CN SNA TOL are presented as theoretically plausible mechanisms rather than empirically established indirect relationships. Future research may extend the SITE framework by testing these mediating and conditional processes using longitudinal or experimental designs.

2.2 Institutional normalization: perceived enforcement weakness and corruption normalization

Perceived Enforcement Weakness refers to commuters’ assessments of whether transport regulators and law enforcement agencies consistently detect and sanction underage and unlicensed driving. The construct is grounded in Deterrence Theory, which proposes that compliance depends on the perceived certainty, severity, and timeliness of sanctions (Beccaria, 2016; Bentham, 1879). Contemporary deterrence research indicates that the perceived certainty of detection and punishment is often more influential than the severity of the penalty (Klepper and D. Nagin, 1989; D. S. Nagin, 1998; Pratt et al., 2017).

When commuters repeatedly observe apparently unqualified drivers operating without regulatory intervention, they may conclude that enforcement is unlikely or ineffective. This perception reduces the practical relevance of licensing requirements. It may also discourage reporting because an objection may appear unlikely to produce a meaningful response. Continued use of the service may therefore occur without explicit approval of the violation.

Institutional Theory provides a complementary explanation. Institutional decoupling occurs when formally adopted rules become separated from routine practice (Meyer and Rowan, 1977). Persistent non-enforcement creates a visible gap between statutory requirements and actual transport operations. Over time, this gap may weaken the social authority of licensing rules and make noncompliance appear ordinary. Commuters who perceive enforcement as weak are therefore expected to report greater Public Tolerance.

H1: Perceived Enforcement Weakness (PEW) is positively associated with Public Tolerance of underage and unlicensed bus drivers (TOL).

Corruption Normalization refers to the extent to which commuters perceive corrupt practices within the transport sector as routine, expected, and difficult to challenge. The construct draws on the social-cognitive foundations of moral self-regulation (Bandura, 1986), Moral Disengagement Theory (Bandura, 2017), and the organizational literature on corruption normalization (Ashforth and Anand, 2003). Moral disengagement explains how individuals suspend moral self-regulation when unethical conduct is interpreted as inevitable, widely accepted, or beyond personal control.

Ashforth and Anand (2003) identify institutionalization, rationalization, and socialization as processes through which corruption becomes embedded in organizational practice. Repeated exposure to broker-mediated licensing, informal payments, political influence, or selective enforcement may lead commuters to interpret violations as systemic rather than exceptional (Suykens, 2018; Hasan and Dávila, 2018; Jackman, 2019). Under these conditions, individual opposition may appear ineffective because responsibility is attributed to a broader institutional arrangement.

Corruption Normalization does not require approval of corrupt conduct. Instead, it may reduce the perceived value of challenging a practice regarded as routine and unavoidable. Commuters who believe that licensing requirements are regularly circumvented may therefore become more willing to overlook the presence of underage or unlicensed drivers.

H2: Corruption Normalization (CN) is positively associated with Public Tolerance of underage and unlicensed bus drivers (TOL).

2.3 Social and empathic acceptance: economic empathy and social norm acceptance

Economic Empathy refers to sympathetic judgments toward underage or unlicensed drivers whose participation in transport work is attributed to poverty, limited employment opportunities, or household economic pressure. The construct is informed by Attribution Theory (Weiner, 1985) and the empathy-altruism hypothesis (Batson et al., 1997).

Attribution Theory proposes that emotional and behavioral responses depend partly on how individuals interpret the causes of another person’s conduct. A violation attributed to controllable personal choice is more likely to produce blame. When the same behavior is attributed to external constraints, perceived personal responsibility may decline and sympathetic responses may increase (Weiner, 1985). Economic vulnerability may similarly influence public judgments of legally noncompliant work in informal labor settings (Vargas, 2016).

This process creates a moral conflict for commuters. They may recognize the safety implications of underage or unlicensed driving while also viewing the driver as economically vulnerable. Opposing the practice may then be perceived as threatening the livelihood of a disadvantaged worker. Economic Empathy may therefore coexist with safety concerns while reducing a commuter’s willingness to reject the service.

H3: Economic Empathy (EE) is positively associated with Public Tolerance of underage and unlicensed bus drivers (TOL).

Social Norm Acceptance refers to the perception that tolerating underage or unlicensed drivers is common and socially accepted among other commuters. The construct is grounded in descriptive Social Norm Theory (Cialdini et al., 1990; Cialdini and Trost, 1998). Descriptive norms communicate what people commonly do, whereas injunctive norms communicate what people believe ought to be done. These forms of social influence may differ. A commuter may privately reject an illegal practice while continuing to tolerate it after observing that other passengers do not object.

In uncertain situations, the behavior of others provides information about what is normal, feasible, or socially appropriate. Repeatedly observing passengers board a vehicle without questioning the driver’s qualifications may therefore provide social evidence that acceptance is the prevailing response. The absence of objection may be interpreted as apparent collective acceptance, even when many commuters privately disapprove. This process is consistent with pluralistic ignorance, in which individuals interpret the silence or conformity of others as evidence of genuine acceptance (Moojen et al., 2023; Miller, 2023).

Social influence is particularly relevant in public transport because travel decisions are often made in the presence of other passengers and under time pressure. Objecting may delay the journey, create conflict, or violate expectations of conformity. Road safety research has similarly shown that perceived norms are associated with responses to speeding, seatbelt non-use, and impaired driving (Cestac et al., 2011; Moan and Jostein Rise, 2011; Dun and Ali, 2018). Commuters who perceive tolerance as the prevailing social response are therefore expected to report greater Public Tolerance.

H4: Social Norm Acceptance (SNA) is positively associated with Public Tolerance of underage and unlicensed bus drivers (TOL).

Economic Empathy and Social Norm Acceptance represent related but distinct mechanisms. Economic Empathy provides an individual moral justification for acceptance, whereas Social Norm Acceptance provides social evidence that such acceptance is common. Together, these constructs provide a theoretical explanation for why commuters may continue using an unsafe service despite recognizing its legal and safety implications.

2.4 Personal risk perception and regulatory trust

Personal Risk Perception refers to a commuter’s subjective assessment of the likelihood and potential severity of harm associated with traveling in a vehicle operated by an underage or unlicensed driver. The construct is grounded in the Psychometric Paradigm of Risk Perception (Slovic, 1987). This perspective recognizes that responses to hazards depend not only on objective probabilities but also on perceived severity, controllability, familiarity, and personal vulnerability.

Higher perceived risk is generally associated with lower acceptance of hazardous activities (Slovic, 1987; Nordfjærn et al., 2011). Road safety studies similarly indicate that stronger perceptions of danger are associated with lower acceptance of speeding, impaired driving, and mobile phone use while driving (Logan et al., 2019; MacLeod et al., 2015; Nguyen-Phuoc et al., 2020). Commuters who anticipate a greater possibility of collision, injury, or death are therefore expected to report lower willingness to use services operated by legally unqualified drivers.

The present model examines the direct association between risk perception and tolerance. It does not model familiarity or habituation as a moderator. Personal Risk Perception is therefore theorized as a countervailing mechanism associated with lower acceptance of underage and unlicensed driving.

H5: Personal Risk Perception (PRP) is negatively associated with Public Tolerance of underage and unlicensed bus drivers (TOL).

Regulatory Trust refers to commuters’ confidence that transport regulators and law enforcement institutions are competent, legitimate, fair, and committed to public safety. The construct is grounded in procedural justice and institutional legitimacy theory (Tom R Tyler, 2006). These perspectives propose that legal rules carry greater normative authority when the institutions responsible for them are viewed as legitimate.

Regulatory Trust is conceptually distinct from Perceived Enforcement Weakness. PEW concerns assessments of current enforcement activity, such as whether authorities inspect licenses and sanction violations. RT concerns broader beliefs about institutional competence, integrity, fairness, and commitment to public welfare. A commuter may perceive enforcement as infrequent while still believing that the responsible institutions are legitimate. Conversely, visible enforcement may fail to generate trust when it is regarded as selective, corrupt, or procedurally unfair.

Trust in legitimate institutions may reinforce the normative authority of licensing requirements. Commuters who trust transport regulators are more likely to regard licensing standards as justified and directed toward public safety. They are therefore expected to be less willing to overlook violations of those standards. Evidence from taxation, policing, and environmental regulation indicates that institutional trust is associated with rule acceptance and compliance (Kirchler et al., 2008; Tyler and Huo, 2002; Winter and May, 2001). Transport research also indicates that trust in responsible authorities is associated with stronger acceptance of road safety policies and less permissive attitudes toward traffic violations (Huber and Wicki, 2021; Tankebe et al., 2020).

H6: Regulatory Trust (RT) is negatively associated with Public Tolerance of underage and unlicensed bus drivers (TOL).

Personal Risk Perception and Regulatory Trust are theorized to be negatively associated with Public Tolerance through conceptually distinct mechanisms. Risk perception emphasizes the potential personal consequences of unsafe travel, whereas regulatory trust reinforces the legitimacy of formal licensing requirements. Together, they may counteract the normalization associated with weak enforcement, corruption, social conformity, and empathic justification. The six hypothesized direct relationships are presented in Figure 1.

FIGURE 1

3 Methodology

3.1 Research design

This study employed a quantitative cross-sectional survey design to examine the factors associated with commuter tolerance of underage and unlicensed bus drivers in Dhaka, Bangladesh. A deductive approach was followed. The hypothesized relationships derived from the Structural and Institutional Tolerance of Enforcement framework were tested using primary data collected from adult public transport users. This design was appropriate because the study sought to estimate relationships among latent attitudinal constructs and to explain variation in Public Tolerance.

Although the cross-sectional design does not establish temporal or causal precedence, the directional specification of the structural model was determined a priori from the theoretical foundations of the SITE framework rather than derived from the observed data. Perceived Enforcement Weakness, Corruption Normalization, and Regulatory Trust represent relatively enduring evaluations of the institutional environment that develop through repeated experience with transport regulation and enforcement. Social Norm Acceptance reflects perceptions of prevailing commuter behavior, while Economic Empathy and Personal Risk Perception represent evaluative orientations toward the social and safety consequences of underage and unlicensed driving. Public Tolerance was therefore modeled as the more proximal attitudinal response to these institutional, social, and psychological evaluations.

This theoretically proposed ordering is consistent with Institutional Theory, Social Norm Theory, Attribution Theory, and the Psychometric Paradigm of Risk Perception, which conceptualize institutional experience, perceived social norms, causal attribution, and risk appraisal as antecedents of acceptance or behavioral response. The hypotheses and structural paths were specified before the empirical analysis, and the model was restricted to theoretically justified direct relationships. Nevertheless, alternative directional relationships and reciprocal effects cannot be excluded using cross-sectional data. Accordingly, the estimated path coefficients are interpreted as theory-consistent associations rather than causal effects.

The methodological process is presented in Figure 2. It comprised research design, target population specification, sampling, instrument development, pilot testing, data collection, response screening, measurement model assessment, structural model assessment, and importance-performance map analysis.

FIGURE 2

3.2 Target population and sampling strategy

The target population comprised commuters aged 18 years or older who regularly used bus services within Dhaka. To ensure relevant experience with the study context, eligible respondents were required to have used a bus or comparable public transport service at least once per week during the 3 months preceding the survey. This criterion ensured that participants had recent and repeated exposure to the transport environment under investigation (Campbell et al., 2020).

A purposive intercept sampling strategy was used, with respondents approached at selected public transport locations. Convenience sampling was applied within each location because a complete sampling frame of Dhaka bus commuters was unavailable. Data were collected at six major transit hubs: Mohakhali Bus Terminal, Gabtoli Bus Terminal, Sayedabad Bus Terminal, Gulistan Bus Stand, Mirpur 1 Bus Stand, and Farmgate Intersection. These locations were selected to capture commuters traveling through different parts of the city’s public transport network. The use of several collection sites increased sample heterogeneity, although the nonprobability design does not permit claims of statistical representativeness.

An a priori power analysis was conducted using G*Power 3.1 for linear multiple regression with six predictors (Faul et al., 2009). The analysis specified a medium effect size of , a significance level of , and statistical power of 0.95. The resulting minimum sample size was 146. The final sample of 320 respondents exceeded this minimum requirement and was considered adequate for estimating the proposed structural model. Figure 3 presents the G*Power output.

FIGURE 3

3.3 Instrument development and pilot testing

The questionnaire contained 35 items measuring seven reflective constructs, with five items assigned to each construct. Responses were recorded on a five-point Likert scale ranging from 1, strongly disagree, to 5, strongly agree. All items were contextualized to the operation of underage and unlicensed bus drivers in Dhaka.

The Perceived Enforcement Weakness items were informed by deterrence theory and related empirical measures (D. S. Nagin, 1998; Pratt et al., 2017). The Corruption Normalization items drew on the organizational literature concerning the institutionalization and normalization of corruption (Ashforth and Anand, 2003). Economic Empathy was operationalized using Attribution Theory and the empathy-altruism literature (Weiner, 1985; Batson et al., 1997). Social Norm Acceptance was measured using items informed by descriptive social norm research (Cialdini et al., 1990; Cialdini and Trost, 1998). Personal Risk Perception was based on the Psychometric Paradigm of Risk Perception and its application to road safety (Slovic, 1987). Regulatory Trust was informed by procedural justice and institutional legitimacy research (Tom R Tyler, 2006).

The five Public Tolerance items were developed specifically for the present study. They measured respondents’ willingness to board, continue using, accept, or overlook a transport service operated by an underage or unlicensed driver. The complete questionnaire and the source associated with each construct are provided in the Supplementary Material.

For the purposes of this study, an underage driver referred to a person below the statutory minimum age of 21 years required to obtain a professional driving license in Bangladesh. The term therefore denotes being below the legal age threshold for professional passenger transport operation and does not necessarily mean that the individual is below the age of legal adulthood. An unlicensed driver referred to a person who did not hold a valid professional driving license authorizing the operation of a public transport vehicle.

The survey did not require respondents to inspect identification documents or independently verify the age or licensing status of a specific driver. Instead, the measurement items assessed attitudes toward situations in which a driver’s legal ineligibility was perceived, reported, disclosed, or otherwise known to the respondent. The study therefore examined attitudinal tolerance toward perceived or disclosed underage and unlicensed driving rather than commuters’ ability to determine a driver’s legal status.

The questionnaire was prepared in English and Bangla. The Bangla version was developed through forward translation and back-translation to preserve semantic and conceptual equivalence between the two language versions. Differences identified during the translation process were reviewed and reconciled before pilot testing. Prior to the main survey, the instrument was pilot tested with 30 commuters to assess item clarity, contextual relevance, response-scale comprehension, questionnaire administration, and preliminary internal consistency. Cronbach’s alpha values ranged from 0.781 to 0.842, indicating satisfactory preliminary reliability across the seven constructs.

3.4 Data collection and response screening

Data were collected from May to September 2025 through structured face-to-face questionnaire administration. This mode was selected to include commuters with different educational backgrounds and levels of digital access. The exclusive use of an online questionnaire could have disproportionately excluded individuals with limited access to or familiarity with digital survey platforms (Hargittai, 2010).

Trained research assistants affiliated with Bangladeshi universities administered the questionnaires at the six selected locations. Where necessary, the research assistants read the questions aloud and recorded the respondents’ answers. This approach improved accessibility for participants with limited reading ability while preserving the original item wording, instructions, and response scale.

A total of 392 questionnaires were administered. Responses were screened before analysis using prespecified quality criteria. Seventy-two questionnaires were excluded because of incomplete responses , straight-line response patterns , failure to satisfy the eligibility criteria , or logical inconsistencies across related responses . The final analytical sample comprised 320 valid questionnaires. The retained sample represented 81.6% of the questionnaires administered.

3.5 Sample characteristics

Table 1 summarizes the demographic and travel characteristics of the respondents. The sample included 206 male respondents (64.4%) and 114 female respondents (35.6%). Respondents aged 18–24 years accounted for 37.2% of the sample, while those aged 25–34 years accounted for 36.6%. Together, these two groups represented 73.8% of the respondents.

Regarding educational attainment, 41.9% held a bachelor’s degree, 23.4% had completed the Higher Secondary Certificate, and 17.8% had completed the Secondary School Certificate. Respondents in service or salaried employment formed the largest occupational group at 38.4%, followed by students at 24.1% and business owners at 20.9%. More than half of the respondents (53.1%) used public transport daily. In addition, 43.1% reported having witnessed a crash that they perceived as involving an underage driver.

3.6 Assessment of common method bias

Because all constructs were measured using responses obtained from the same participants through a single questionnaire, procedural and statistical measures were applied to assess potential common method bias (CMB) (Cooper et al., 2020; Podsakoff et al., 2003). Procedurally, respondent anonymity was assured, no identifying information was collected, and the Public Tolerance items were positioned after the antecedent constructs to provide limited psychological separation between the predictor and criterion measures. The same item wording, instructions, and response scale were maintained when the questionnaire was read aloud to respondents.

Harman’s single-factor test was conducted as an initial diagnostic using all 35 measurement items. The first unrotated component accounted for 39.03% of the total variance, which was below the commonly applied threshold of 50% (Fuller et al., 2016). This result indicated that no single component accounted for the majority of the variance among the measurement items. Full collinearity was also assessed by treating each latent construct as an endogenous variable and regressing it on the remaining constructs. All full-collinearity variance inflation factor values were below the conservative threshold of 3.3, providing no indication of substantial common method-related collinearity.

Given the limited diagnostic sensitivity of Harman’s single-factor test, a supplementary unmeasured latent method construct (ULMC) analysis was conducted (Chin et al., 2012). A common method factor was specified across all measurement indicators and constrained to be orthogonal to the substantive constructs. The average variance attributable to the method factor was 2.19%, compared with 61.32% attributable to the substantive constructs, resulting in a method-to-substantive variance ratio of 3.58%. The inclusion of the method factor did not materially alter the directions, magnitudes, or inferential conclusions of the structural relationships.

Taken together, Harman’s single-factor test, the full-collinearity assessment, and the ULMC sensitivity analysis did not indicate a dominant common method pattern. Nevertheless, these findings were interpreted cautiously because no post hoc statistical procedure can establish the complete absence of method bias. In addition, the questionnaire did not include a dedicated marker variable, and the ability of the ULMC procedure to detect and control CMB in PLS-SEM has recognized limitations (Podsakoff et al., 2003; Chin et al., 2012).

3.7 PLS-SEM analytical procedure

The hypothesized relationships were examined using partial least squares structural equation modeling (PLS-SEM) in SmartPLS 4 (Ringle et al., 2024). PLS-SEM was selected because the study emphasized explanation of variation in Public Tolerance, introduced a new integrative theoretical framework, and focused on the explained variance of the endogenous construct (Hair et al., 2019; Hair and Alamer, 2022). The method also permits the simultaneous assessment of reflective measurement quality and the structural relationships among latent constructs.

The analysis followed two sequential stages. First, the reflective measurement model was evaluated to establish indicator reliability, internal consistency, convergent validity, and discriminant validity. Indicator reliability was assessed using standardized outer loadings, with values of 0.708 or higher considered satisfactory (Hair et al., 2019). Internal consistency was examined using Cronbach’s alpha, , and composite reliability. Values between 0.70 and 0.95 were regarded as acceptable. Convergent validity was assessed using average variance extracted (AVE), with a minimum criterion of 0.50 (Fornell and Larcker, 1981; J. Hair and Alamer, 2022).

Discriminant validity was assessed primarily using the heterotrait–monotrait ratio of correlations (HTMT). Values below 0.85 were interpreted as evidence of adequate conceptual distinction among the constructs (Henseler et al., 2015). The Fornell–Larcker criterion was also reported as a supplementary assessment. Under this criterion, the square root of each construct’s AVE should exceed its correlations with the remaining constructs (Fornell and Larcker, 1981).

Second, the structural model was assessed after satisfactory measurement quality had been established. Predictor collinearity was examined using inner variance inflation factor values. The hypothesized relationships were evaluated using standardized path coefficients, statistics, values, and bias-corrected confidence intervals. Statistical inference was based on nonparametric bootstrapping with 5,000 resamples (Hair et al., 2019; Hair and Alamer, 2022). The estimated coefficients were interpreted as theory-consistent associations rather than causal effects.

The model’s explanatory capacity was evaluated using the coefficient of determination, . The relative contribution of each predictor was examined using the effect size, , while predictive relevance was assessed using (Shmueli et al., 2019). The standardized root mean square residual, squared Euclidean distance, geodesic distance, and normed fit index were reported as supplementary model diagnostics. These indices were interpreted cautiously because global fit criteria are not the primary basis for evaluating a PLS-SEM model.

Finally, importance-performance map analysis (IPMA) was conducted with Public Tolerance as the target construct. IPMA compares each antecedent’s total effect, representing importance, with its average latent variable performance on a scale from 0 to 100 (Sarstedt et al., 2024). Because Public Tolerance represents an undesirable outcome and the antecedents were expected to have effects in different directions, the IPMA results were interpreted directionally rather than through a uniform high-importance, low-performance rule.

The absolute magnitude of each total effect was used to compare the relative importance of the antecedents, while the sign of the effect was retained to determine the appropriate direction of interpretation. For antecedents positively associated with Public Tolerance, higher performance indicates a stronger tolerance-promoting condition that may need to be reduced, disrupted, or decoupled from acceptance. For antecedents negatively associated with Public Tolerance, lower performance indicates a potentially underdeveloped protective condition that may warrant strengthening. The IPMA was therefore used as an exploratory policy-prioritization tool rather than as evidence that changing a construct would causally produce a corresponding reduction in Public Tolerance.

4 Results

4.1 Measurement model assessment

The reflective measurement model was assessed before examining the structural relationships. The assessment considered indicator reliability, internal consistency, convergent validity, and discriminant validity (Cheung et al., 2024). Figure 4 presents the standardized outer loadings.

FIGURE 4

All 35 indicators exceeded the recommended loading threshold of 0.708 (J. Hair and Alamer, 2022). The loadings ranged from 0.730 for SNA5 to 0.820 for TOL4, supporting indicator reliability. As reported in Table 2, Cronbach’s alpha values ranged from 0.830 to 0.858, while composite reliability values ranged from 0.879 to 0.898. All values were above the recommended minimum of 0.70 and below 0.95, indicating satisfactory internal consistency (Hair et al., 2021; Hair and Alamer, 2022).

The average variance extracted values ranged from 0.593 to 0.638. Each value exceeded the 0.50 criterion, supporting convergent validity (Cheung et al., 2024). The complete set of indicator loadings is reported in Table 3.

Discriminant validity was assessed using the Fornell–Larcker criterion and the heterotrait–monotrait ratio of correlations (HTMT). Under the Fornell–Larcker criterion, the square root of each construct’s AVE exceeded its correlations with the remaining constructs, as shown in Table 4 (Fornell and Larcker, 1981).

As shown in Table 5, the HTMT values ranged from 0.110 to 0.513. All values were below the conservative threshold of 0.85, providing further support for discriminant validity (Henseler et al., 2015). The highest HTMT value was observed between Personal Risk Perception and Public Tolerance.

Overall, the measurement model met the specified reliability and validity criteria. The structural relationships were therefore examined in the next stage.

4.2 Structural model assessment

Collinearity among the six antecedent constructs was assessed using inner variance inflation factor values. The values ranged from 1.169 to 1.376, well below the conservative threshold of 3.3 (Hair et al., 2019;Roy et al., 2018). These results indicate that collinearity was unlikely to materially distort the estimated path coefficients.

Supplementary model diagnostics are reported in Table 6. The standardized root mean square residual (SRMR) was 0.048, which was below the commonly applied threshold of 0.08 (Cheah et al., 2026). The normed fit index (NFI) was 0.842. The squared Euclidean distance , geodesic distance , and statistic are reported descriptively. These indices were treated as supplementary information rather than as primary criteria for evaluating the PLS-SEM model.

The hypothesized relationships were examined using bootstrapping with 5,000 resamples. Table 7 reports the standardized path coefficients, values, values, VIF values, and effect sizes.

All six hypothesized associations were statistically supported. Perceived Enforcement Weakness was positively associated with Public Tolerance (, , ), supporting H1. Corruption Normalization was also positively associated with Public Tolerance (, , ), supporting H2.

Economic Empathy had the largest positive path coefficient (, , ), supporting H3. Social Norm Acceptance was positively associated with Public Tolerance (, , ), supporting H4.

Personal Risk Perception had the largest absolute path coefficient and was negatively associated with Public Tolerance (, , ), supporting H5. Regulatory Trust was also negatively associated with Public Tolerance (, , ), supporting H6. These coefficients represent statistical associations within the proposed model and should not be interpreted as causal effects. Figure 5 presents the structural model.

FIGURE 5

The model produced an value of 0.455 and an adjusted value of 0.445 for Public Tolerance. Thus, the six antecedent constructs jointly accounted for 45.5% of the observed variance in Public Tolerance.

The largest effect size was observed for Personal Risk Perception , followed by Economic Empathy , Regulatory Trust , Social Norm Acceptance , Corruption Normalization , and Perceived Enforcement Weakness . Based on the conventional benchmarks of 0.02, 0.15, and 0.35 for small, medium, and large effects, respectively, all six predictors exhibited small individual effect sizes. Their combined explanatory contribution was reflected in the model’s value of 0.455.

The reported value for Public Tolerance was 0.422, which was above zero. The corresponding RMSE and MAE values were 0.765 and 0.632, respectively. These results indicate predictive relevance for the reported target measure (Wynne W Chin, 2009; Shmueli et al., 2019).

4.3 Importance–performance map analysis

Importance–performance map analysis (IPMA) was conducted with Public Tolerance as the target construct (Sarstedt et al., 2024; Ringle and Sarstedt, 2016). Importance was represented by the signed total effect of each antecedent on Public Tolerance, while performance was expressed on a scale from 0 to 100. Figure 6 presents the corresponding importance–performance map for Public Tolerance.

FIGURE 6

Because Public Tolerance represents an undesirable outcome and the antecedents were associated with it in different directions, the IPMA results were interpreted directionally. For positively associated antecedents, higher performance represents a stronger tolerance-promoting condition. For negatively associated antecedents, lower performance represents a potentially underdeveloped tolerance-reducing condition. The analysis was therefore used to identify policy-relevant patterns rather than to infer the causal effects of changing individual constructs.

As shown in Table 8, Personal Risk Perception had the largest total effect in absolute magnitude and a performance score of 60.496. Its negative effect indicates that higher Personal Risk Perception was associated with lower Public Tolerance. Regulatory Trust also had a negative total effect and recorded the lowest performance score, at 34.272. This combination indicates that Regulatory Trust was negatively associated with tolerance but was comparatively weak among the surveyed respondents.

Among the antecedents positively associated with Public Tolerance, Economic Empathy had the largest total effect (0.197), followed by Corruption Normalization (0.173), Social Norm Acceptance (0.171), and Perceived Enforcement Weakness (0.141). Their performance scores ranged from 61.825 to 65.063, indicating comparatively elevated levels of these tolerance-promoting conditions. The substantive and policy implications of these direction-sensitive patterns are examined in the Discussion section.

5 Discussion

5.1 Overall findings

This study examined the institutional, social, and psychological conditions associated with commuter tolerance of underage and unlicensed bus drivers in Dhaka. The SITE framework explained 45.5% of the variance in Public Tolerance, and all six hypothesized associations were statistically supported. Although the individual effect sizes were small, their combined explanatory contribution indicates that commuter tolerance is associated with several complementary conditions rather than a single dominant factor. The findings therefore reinforce the value of examining tolerance as a multidimensional attitudinal response embedded within a broader institutional and social environment.

Personal Risk Perception had the largest path coefficient in absolute magnitude (, ). Commuters who perceived greater personal danger reported lower tolerance of underage and unlicensed driving. This finding is consistent with the Psychometric Paradigm of Risk Perception, which emphasizes the importance of perceived vulnerability, severity, and controllability in responses to hazardous situations (Slovic, 1987). Greater awareness of the potential for collision, injury, or death may counter the routine acceptance of legally unqualified drivers. However, the cross-sectional design does not establish whether stronger risk perception reduces tolerance or whether commuters who are already less tolerant perceive the practice as more dangerous.

Regulatory Trust was also negatively associated with Public Tolerance (, ). Its comparatively low performance score in the direction-sensitive importance–performance map indicates that confidence in transport regulatory institutions was relatively limited among the surveyed commuters. This pattern is consistent with procedural justice and institutional legitimacy theory (Tom R Tyler, 2006). Licensing requirements may carry greater normative authority when the institutions responsible for implementing them are perceived as competent, fair, legitimate, and committed to public safety. Conversely, limited institutional trust may weaken expectations that reporting or opposing a violation will lead to a credible or meaningful regulatory response. The observed association should nevertheless be interpreted as attitudinal rather than causal.

Economic Empathy had the largest positive path coefficient (, ). Commuters who attributed underage or unlicensed driving to poverty, limited employment opportunities, or household economic pressure reported greater tolerance. This result is consistent with the attributional proposition that judgments of responsibility depend partly on how the causes of a violation are interpreted (Weiner, 1985). When legally noncompliant driving is attributed to external economic constraints, commuters may assign less personal blame to the driver and become more reluctant to reject the service. Economic Empathy should not, however, be interpreted as indifference to road safety. Instead, it reflects a moral tension between concern for passenger safety and concern for economically vulnerable workers.

Corruption Normalization (, ) and Social Norm Acceptance (, ) exhibited similar positive associations with Public Tolerance. When informal payments, selective enforcement, or the circumvention of licensing requirements are perceived as routine, formal regulations may lose practical authority in everyday travel decisions (Ashforth and Anand, 2003). Repeated observation of other passengers using the same services without objection may further strengthen the perception that tolerance is common, expected, or unavoidable (Cialdini et al., 1990; Cialdini and Trost, 1998). These findings are consistent with the distinction between private disapproval and publicly conforming behavior. Commuters may regard underage or unlicensed driving as legally and morally unacceptable while continuing to use the service because acceptance appears to be the prevailing response among other passengers.

Perceived Enforcement Weakness had the smallest path coefficient, although its positive association with Public Tolerance remained statistically supported (, ). Weak enforcement may therefore function as a broader enabling condition rather than as a sufficient explanation of tolerance. When commuters perceive that license inspections are infrequent, sanctions are inconsistently applied, or complaints are unlikely to produce a response, the practical significance of formal licensing requirements may decline. Nevertheless, the comparatively small coefficient indicates that perceived enforcement weakness operates alongside more immediate considerations, including economic sympathy, perceived social acceptance, personal risk, and confidence in regulatory institutions.

Taken together, the findings are consistent with the SITE framework’s proposition that commuter tolerance is associated with interconnected institutional, social, and psychological conditions. Weak enforcement and normalized corruption may create an environment in which noncompliance appears routine, while social acceptance and economic empathy may provide social and moral justification for continued use of the service. Personal Risk Perception and Regulatory Trust may operate as countervailing conditions associated with lower tolerance. However, the present model estimated only direct relationships. It did not empirically test whether Social Norm Acceptance mediates the associations of Perceived Enforcement Weakness or Corruption Normalization with Public Tolerance, nor did it test moderation or interaction effects. These interconnected processes should therefore be interpreted as theoretically plausible mechanisms rather than established indirect pathways.

5.2 Theoretical contributions

This study contributes to transport safety research by conceptualizing Public Tolerance as an empirically measurable attitudinal outcome. Existing research has predominantly examined unsafe public transport through driver behavior, crash severity, vehicle conditions, regulatory compliance, and enforcement effectiveness. The present study extends this literature by focusing on commuters who continue to use transport services after perceiving, learning, or being informed that a driver may be underage or unlicensed. Commuters are therefore conceptualized not only as recipients of transport risk but also as participants whose attitudes and travel decisions may be associated with the continuation or rejection of legally noncompliant operating practices.

A second contribution is the development of the SITE framework, which integrates Institutional Theory, Social Norm Theory, and the Psychometric Paradigm of Risk Perception within a unified explanatory structure. Institutional Theory explains how perceived enforcement weakness and normalized corruption may reduce the practical authority of formal licensing requirements. Social Norm Theory explains how repeated public acceptance may be interpreted as evidence that tolerance is common or socially expected. The Psychometric Paradigm explains why stronger perceptions of personal danger are associated with lower tolerance. The framework further incorporates Economic Empathy and Regulatory Trust as complementary construct-level explanations, linking causal attribution and institutional legitimacy to commuter acceptance.

The findings also clarify how these theoretical perspectives may be connected rather than operating as independent or parallel explanations. Perceived Enforcement Weakness and Corruption Normalization may contribute to an institutional environment in which licensing violations appear routine, persistent, or difficult to challenge. Social Norm Acceptance may reflect how such institutional conditions are interpreted through observations of other commuters’ behavior. Economic Empathy may provide a moral justification for continued acceptance when noncompliant driving is attributed to poverty or limited employment opportunities. In contrast, Personal Risk Perception and Regulatory Trust represent countervailing conditions associated with lower tolerance by emphasizing personal safety and the legitimacy of formal licensing requirements.

This theoretical integration extends existing transport safety perspectives by linking institutional experience, perceived social norms, moral attribution, risk appraisal, and institutional legitimacy within the same framework. It also distinguishes between conditions that are positively associated with Public Tolerance and those that may constrain it. The SITE framework therefore provides a foundation for examining how institutional, social, and psychological processes may jointly relate to public acceptance of unsafe and legally noncompliant transport practices.

These interconnected mechanisms remain theoretical because the empirical model estimated only direct associations. It did not test mediation, moderation, or statistical interaction effects. Accordingly, pathways such as PEW SNA TOL and CN SNA TOL should be interpreted as theoretically plausible extensions rather than empirically established indirect relationships. Future studies could test these pathways using longitudinal, experimental, or repeated-measures designs capable of assessing temporal ordering, reciprocal relationships, and causal mechanisms.

5.3 Policy implications

The findings identify four connected policy domains that warrant consideration: risk communication and credential verification, institutional legitimacy and enforcement, operator accountability and economic protection, and social norm change. These directions should be interpreted as evidence-informed options derived from the observed structural associations and direction-sensitive importance–performance patterns. They are not interventions whose effectiveness was tested in the present study.

Personal Risk Perception exhibited the largest path coefficient in absolute magnitude. This finding suggests that clear and credible communication about the safety implications of underage and unlicensed driving may be relevant to reducing public tolerance. Transport authorities could disseminate verified information about crashes, injuries, licensing violations, and the safety purpose of professional licensing requirements. Such communication should avoid sensationalism and should distinguish documented evidence from unverified claims about individual drivers.

The findings also highlight the potential value of making professional driver credentials easier to verify. Visible credential displays, vehicle-linked license information, or secure digital verification systems could enable passengers and enforcement personnel to confirm professional eligibility without relying on physical appearance. Any verification mechanism would need to protect personal data, prevent fraudulent display, and remain accessible to commuters with different levels of digital access.

Regulatory Trust warrants particular attention because it was negatively associated with Public Tolerance and recorded the lowest performance score in the IPMA. This pattern suggests that confidence in the institutions responsible for licensing and enforcement may represent an underdeveloped protective condition. Greater transparency in driver testing, license issuance and renewal, vehicle inspection, enforcement decisions, and complaint handling may help strengthen institutional credibility. Public reporting of enforcement activities and timely follow-up on commuter complaints may also demonstrate that licensing requirements are applied consistently.

These measures may be relevant not only to Regulatory Trust but also to Perceived Enforcement Weakness and Corruption Normalization. Transparent procedures, auditable licensing records, consistent sanctions, and independent oversight may reduce perceptions that violations are ignored, selectively punished, or routinely resolved through informal arrangements. However, the present findings do not establish that any specific administrative reform would directly reduce Public Tolerance. Such measures would require pilot implementation and formal evaluation.

The positive association between Economic Empathy and Public Tolerance requires a careful and nonstigmatizing policy response. Communication should not portray economically vulnerable young people as the sole source of the problem. Being below the professional licensing age of 21 does not necessarily mean that a driver is legally a child. Where minors are involved, child-protection concerns are directly relevant; where drivers are aged 18 to 20, the issue concerns lawful employment, professional eligibility, and passenger safety.

Policy responses should therefore direct primary accountability toward vehicle owners, employers, route operators, and transport organizations that permit legally ineligible individuals to operate passenger vehicles. Educational support, vocational training, lawful employment pathways, and appropriate social assistance may also provide safer alternatives for economically vulnerable young people. This approach recognizes economic hardship without allowing it to justify unsafe or legally noncompliant passenger transport operations.

The positive associations of Corruption Normalization and Social Norm Acceptance further suggest that enforcement alone may be insufficient. Consistent regulatory action can communicate that licensing violations are neither inevitable nor socially acceptable. Public communication should reinforce the expectation that passenger vehicles must be operated by properly qualified and licensed drivers. Reporting mechanisms should also be simple, accessible, and connected to visible institutional responses. Encouraging commuters to report violations without providing credible follow-up may reinforce perceptions of enforcement weakness and further undermine Regulatory Trust. Table 9 consolidates these policy directions into four connected domains. The recommendations are grouped by their principal theoretical and empirical relevance rather than treating each SITE construct as an isolated intervention target.

Figure 7 presents a conceptual framework for translating the SITE findings into policy domains. It links the principal constructs with institutional, communicative, social, and welfare-oriented responses. The framework should be interpreted as an organization of policy-relevant options rather than evidence that a particular intervention will produce a specified change in commuter tolerance.

FIGURE 7

These policy directions are relevant to Bangladesh’s commitments to reduce road traffic harm under SDG 3, improve access to safe transport under SDG 11, and strengthen effective and accountable institutions under SDG 16. Implementation should nevertheless be preceded by consultation with commuters, drivers, operators, regulators, law enforcement agencies, and youth and labor representatives. Any adopted measures should subsequently be evaluated using behavioral outcomes, administrative records, and longitudinal or experimental designs. Given the cross-sectional nonprobability survey design, the current findings should inform policy development and testing rather than be interpreted as evidence of intervention effectiveness.

6 Conclusion

This study examined the institutional, social, and psychological conditions associated with commuter tolerance of underage and unlicensed bus drivers in Dhaka, Bangladesh. It developed and empirically assessed the Structural and Institutional Tolerance of Enforcement (SITE) framework using survey data from 320 adult commuters. All six hypothesized associations were statistically supported, and the model explained 45.5% of the variance in Public Tolerance. The findings indicate that commuter tolerance is associated with several complementary conditions rather than a single dominant factor.

Personal Risk Perception had the largest path coefficient in absolute magnitude and was negatively associated with Public Tolerance. Commuters who perceived greater personal danger reported less willingness to accept services operated by underage or unlicensed drivers. Economic Empathy had the largest positive association, indicating that sympathy toward drivers facing poverty, limited employment opportunities, or household economic pressure may coexist with greater acceptance of legally noncompliant transport practices. Regulatory Trust was also negatively associated with tolerance and recorded the lowest performance score in the direction-sensitive importance–performance analysis, indicating comparatively limited confidence in the institutions responsible for licensing and enforcement.

The study contributes to transport safety research by conceptualizing commuter tolerance as an empirically measurable attitudinal outcome rather than treating passengers solely as recipients of transport risk. The SITE framework further demonstrates the value of integrating institutional experience, perceived social norms, economic attribution, risk appraisal, and institutional legitimacy within a unified explanatory structure. The findings identify policy-relevant domains involving risk communication, transparent licensing and enforcement, professional credential verification, operator accountability, economic protection, and social norm change. These directions are relevant to Bangladesh’s efforts to reduce road traffic harm under SDG 3, improve access to safe transport under SDG 11, and strengthen effective and accountable institutions under SDG 16. However, they should be treated as areas for policy development and evaluation rather than interventions whose effectiveness was established by the present study.

Several limitations should be considered when interpreting the findings. First, the cross-sectional design does not establish the temporal or causal ordering of the observed relationships. The estimated paths should therefore be interpreted as theory-consistent associations rather than causal effects. Second, the purposive intercept sampling strategy limits statistical generalization beyond the surveyed commuters and data-collection locations.

Third, the use of self-reported attitudinal data may have introduced social desirability and recall bias despite the anonymity and procedural safeguards applied during data collection. Respondents may have understated their tolerance of legally noncompliant driving or overstated their Personal Risk Perception to present themselves as safety conscious. Because these response tendencies may affect the covariance among the study constructs in different directions, the magnitude and direction of the resulting bias cannot be determined from the present data. The estimated relationships should therefore be interpreted cautiously. Future studies should incorporate validated social desirability measures, indirect questioning techniques, and behavioral or observational indicators of boarding, reporting, and objection.

Fourth, the study did not independently verify the age or licensing status of drivers encountered by the respondents. In actual travel settings, physical appearance alone cannot establish whether a driver is below the statutory professional licensing age, while licensing status may not be directly observable to passengers. Respondents may therefore have interpreted driver eligibility based on appearance, prior knowledge, information from others, or the situations presented in the measurement items. The findings should consequently be interpreted as attitudes toward perceived or disclosed underage and unlicensed driving rather than as evidence that commuters can accurately identify legally ineligible drivers. Future research could address this limitation using standardized vignettes, verified licensing records, or experimental scenarios in which age and licensing information is explicitly provided.

Fifth, Public Tolerance was measured as an attitudinal construct. The study did not observe actual behaviors such as refusing to board, reporting a driver, or confronting an operator. Stated tolerance may not correspond directly with behavior under real travel constraints, including limited service availability, time pressure, financial considerations, or concerns about personal conflict. Finally, the study focused on commuters in Dhaka and may not fully capture transport conditions, enforcement practices, service availability, or public attitudes in other urban and rural areas of Bangladesh.

Future research should assess the SITE framework using longitudinal, experimental, and quasi-experimental designs to clarify temporal ordering and evaluate changes following institutional or policy interventions. Behavioral measures should be incorporated to determine whether stated tolerance predicts actual boarding, reporting, objection, or service-switching behavior. Qualitative interviews with commuters, drivers, vehicle owners, operators, regulators, and law enforcement officials could also provide a deeper understanding of how economic vulnerability, corruption, social expectations, and institutional distrust shape everyday transport decisions.

Further studies should test the framework across different cities, transport modes, and national settings. Cross-national comparisons involving South and Southeast Asian countries may help distinguish context-specific conditions from more general mechanisms. Future models could also examine theoretically proposed indirect pathways, including PEW SNA TOL and CN SNA TOL, as well as possible moderating roles of Regulatory Trust, Personal Risk Perception, and perceived behavioral control. Intervention studies could subsequently evaluate whether risk communication, professional license verification, consistent enforcement, social norm messaging, complaint-response mechanisms, and operator-focused accountability measures are associated with lower tolerance and safer commuter responses.

References

• 1

  AdhikaryT. S. (2025). Fatalities Rise as Initiatives Ineffective. Dhaka, Bangladesh: The Daily Star, published by Mediaworld Ltd.Available online at: https://www.thedailystar.net/news/bangladesh/news/fatalities-rise-initiatives-ineffective-3798781 (Accessed July 27, 2026).

  • Google Scholar
  • View reference in article

• 2

  ArmitageS.RodwellD.LewisI. (2022). Applying an extended theory of planned behaviour to understand influences on safe driving intentions and behaviours. Transp. Res. Part F Traffic Psychol. Behav.90, 347–364. 10.1016/j.trf.2022.09.009

  • CrossRef
  • Google Scholar
  • View reference in article

• 3

  AshforthB. E.AnandV. (2003). The normalization of corruption in organizations. Res. Organizational Behavior25, 1–52. 10.1016/S0191-3085(03)25001-2

  • CrossRef
  • Google Scholar
  • View reference in article

• 4

  BanduraA. (1986). Social Foundations of Thought and Action: A Social Cognitive Theory. Englewood Cliffs, NJ: Prentice-Hall, 617. Available online at: https://books.google.com/books?id=k6a3AAAAIAAJ (Accessed August 28, 2026).

  • Google Scholar
  • View reference in article

• 5

  BanduraA. (2017).“Moral disengagement in the perpetration of inhumanities,” in Recent Developments in Criminological Theory (London, United Kingdom: Routledge), 135–152. 10.4324/9781315089089-12

  • CrossRef
  • Google Scholar
  • View reference in article

• 6

  BatesL.SooleD.WatsonB. (2012). “The effectiveness of traffic policing in reducing traffic crashes,” in Policing and Security in Practice: Challenges and Achievements (Palgrave Macmillan), 90–109. 10.1057/9781137007780_6

  • CrossRef
  • Google Scholar
  • View reference in article

• 7

  BatsonC. D.PolycarpouM. P.Harmon-JonesE.ImhoffH. J.MitchenerE. C.BednarL. L.et al (1997). Empathy and attitudes: can feeling for a member of a stigmatized group improve feelings toward the group?J. Personality Soc. Psychol.72 (1), 105–118. 10.1037/0022-3514.72.1.105

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 8

  BeccariaC. (2016). On crimes and punishments. New Brunswick, NJ: Transaction Publishers. Available online at: https://books.google.com/books?id=DJcejslITNcC (Accessed August 28, 2026).

  • Google Scholar
  • View reference in article

• 9

  BenthamJ. (1879). An Introduction to the Principles of Morals and Legislation. Oxford: Clarendon Press. Available online at: https://books.google.com/books?id=NhksAAAAIAAJ (Accessed August 28, 2026).

  • Google Scholar
  • View reference in article

• 10

  CampbellS.GreenwoodM.PriorS.ShearerT.WalkemK.YoungS.et al (2020). Purposive sampling: complex or simple? Research case examples. J. Research Nurs.25, 8–661. 10.1177/1744987120927206

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 11

  CestacJ.ParanF.DelhommeP. (2011). Young drivers’ sensation seeking, subjective norms, and perceived behavioral control and their roles in predicting speeding intention: how risk-taking motivations evolve with gender and driving experience. Saf. Sci.49 (3), 424–432. 10.1016/j.ssci.2010.10.007

  • CrossRef
  • Google Scholar
  • View reference in article

• 12

  CheahJ.-H.SarstedtM.HairJ. F.RingleC. M. (2026). Consistent partial least squares structural equation modeling using SmartPLS. Struct. Equ. Model. A Multidiscip. J., 1–14. 10.1080/10705511.2026.2633754

  • CrossRef
  • Google Scholar
  • View reference in article

• 13

  CheungG. W.Cooper-ThomasH. D.LauR. S.WangL. C. (2024). Reporting reliability, convergent and discriminant validity with structural equation modeling: a review and best-practice recommendations. Asia Pacific Journal Management41 (2), 745–783. 10.1007/s10490-023-09871-y

  • CrossRef
  • Google Scholar
  • View reference in article

• 14

  ChinW. W. (2009). “How to write up and report PLS analyses,” in Handbook of Partial Least Squares: Concepts, Methods and Applications (Springer), 655–690. 10.1007/978-3-540-32827-8_29

  • CrossRef
  • Google Scholar
  • View reference in article

• 15

  ChinW. W.ThatcherJ. B.WrightR. T. (2012). Assessing common method bias: problems with the ULMC technique. MIS Q.36 (3), 1003–1019. 10.2307/41703491

  • CrossRef
  • Google Scholar
  • View reference in article

• 16

  CialdiniR. B.TrostM. R. (1998). “Social influence: social norms, conformity and compliance,” in The Handbook of Social Psychology. 4th ed (New York: McGraw-Hill), 2, 151–192. Available online at: https://www.semanticscholar.org/paper/Social-influence%3A-Social-norms%2C-conformity-and-Cialdini-Trost/bc4d09459f298901ebb6894652319c9be3c3b8b2.

  • Google Scholar
  • View reference in article

• 17

  CialdiniR. B.RenoR. R.KallgrenC. A. (1990). A focus theory of normative conduct: recycling the concept of norms to reduce littering in public places. J. Personality Social Psychology58 (6), 1015–1026. 10.1037/0022-3514.58.6.1015

  • CrossRef
  • Google Scholar
  • View reference in article

• 18

  CooperB.EvaN.Zarea FazlelahiF.NewmanA.LeeA.ObschonkaM. (2020). Addressing common method variance and endogeneity in vocational behavior research: a review of the literature and suggestions for future research. J. Vocational Behavior121, 103472. 10.1016/j.jvb.2020.103472

  • CrossRef
  • Google Scholar

• 19

  DiMaggioP. J.PowellW. W. (1983). The iron cage revisited: institutional isomorphism and collective rationality in organizational fields. Am. Sociol. Rev.48 (2), 147–160. 10.2307/2095101

  • CrossRef
  • Google Scholar
  • View reference in article

• 20

  DunS.AliA. Z. (2018). “Seatbelts don’t save lives”: discovering and targeting the attitudes and behaviors of young Arab male drivers. Accid. Analysis and Prev.121, 185–193. 10.1016/j.aap.2018.09.003

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 21

  ElliottM. A. (2010). Predicting motorcyclists’ intentions to speed: effects of selected cognitions from the theory of planned behaviour, self-identity and social identity. Accid. Analysis and Prev.42 (2), 718–725. 10.1016/j.aap.2009.10.021

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 22

  ElvikR. (2021). “Legislation, enforcement and education for traffic safety: a brief review of the current state of knowledge,” in Transport and Safety: Systems, Approaches, and Implementation, 67–83. 10.1007/978-981-16-1115-5_4

  • CrossRef
  • Google Scholar
  • View reference in article

• 23

  European Parliament and Council of the European Union (2025). Directive (EU) 2025/2205 of the European parliament and of the council of 22 October 2025 on driving licences. Official J. Eur. Union. Available online at: https://eur-lex.europa.eu/eli/dir/2025/2205/oj.

  • Google Scholar
  • View reference in article

• 24

  FaulF.ErdfelderE.BuchnerA.LangA. G. (2009). Statistical power analyses using g*power 3.1: tests for correlation and regression analyses. Behav. Res. Methods41 (4), 1149–1160. 10.3758/BRM.41.4.1149

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 25

  FornellC.LarckerD. F. (1981). Evaluating structural equation models with unobservable variables and measurement error. J. Marketing Research18 (1), 39–50. 10.1177/002224378101800104

  • CrossRef
  • Google Scholar
  • View reference in article

• 26

  FullerC. M.SimmeringM. J.AtincG.AtincY.BabinB. J. (2016). Common methods variance detection in business research. J. Business Research69 (8), 3192–3198. 10.1016/j.jbusres.2015.12.008

  • CrossRef
  • Google Scholar
  • View reference in article

• 27

  Government of Hungary (2011). 326/2011. (XII. 28.) Government Decree on Road Traffic Administration and the Issuance and Withdrawal of Road Traffic Documents. Budapest, Hungary: Ministry of Justice, National Legislation Database. Available online at: https://njt.jog.gov.hu/jogszabaly/2011-326-20-22.88 (Accessed July 27, 2026).

  • Google Scholar
  • View reference in article

• 28

  Government of Thailand (2026). Application for a driver license under the land transport act B.E. 2522. Government Information Center. Official public service information in Thai. Available online at: https://info.go.th/procedure/94e56d1c-17fb-4d28-bff2-985410285314/view (Accessed July 27, 2026).

  • Google Scholar
  • View reference in article

• 29

  Government of the People’s Republic of Bangladesh (2018). Road transport act, 2018. Bangladesh laws, ministry of law, justice and parliamentary affairs. Act no. 47 of 2018. Available online at: https://bdlaws.minlaw.gov.bd/act-print-1262.html.

  • Google Scholar
  • View reference in article

• 30

  HairJ.AlamerA. (2022). Partial least squares structural equation modeling (PLS-SEM) in second language and education research: guidelines using an applied example. Res. Methods Applied Linguistics1 (3), 100027. 10.1016/j.rmal.2022.100027

  • CrossRef
  • Google Scholar
  • View reference in article

• 31

  HairJ. F.HultG. T. M.RingleC. M.SarstedtM.DanksN. P.RayS., (2021). “Evaluation of reflective measurement models,” in Partial Least Squares Structural Equation Modeling (PLS-SEM) Using R: A Workbook (Springer), 75–90. 10.1007/978-3-030-80519-7_4

  • CrossRef
  • Google Scholar
  • View reference in article

• 32

  HairJ. F.RisherJ. J.SarstedtM.RingleC. M. (2019). When to use and how to report the results of PLS-SEM. Eur. Bus. Rev.31 (1), 2–24. 10.1108/EBR-11-2018-0203

  • CrossRef
  • Google Scholar
  • View reference in article

• 33

  HargittaiE. (2010). Digital Na (T) ives? Variation in internet skills and uses among members of the “net generation”. Sociol. Inquiry80 (1), 92–113. 10.1111/j.1475-682X.2009.00317.x

  • CrossRef
  • Google Scholar
  • View reference in article

• 34

  HasanMd M. U.DávilaJ. D. (2018). The politics of (im) mobility: Rickshaw bans in Dhaka, Bangladesh. J. Transp. Geogr.70, 246–255. 10.1016/j.jtrangeo.2018.06.002

  • CrossRef
  • Google Scholar
  • View reference in article

• 35

  HenselerJ.RingleC. M.SarstedtM. (2015). A new criterion for assessing discriminant validity in variance-based structural equation modeling. J. Academy Marketing Science43 (1), 115–135. 10.1007/s11747-014-0403-8

  • CrossRef
  • Google Scholar
  • View reference in article

• 36

  HuberR. A.WickiM. (2021). What explains citizen support for transport policy? The roles of policy design, trust in government and proximity among Swiss citizens. Energy Res. and Soc. Sci.75, 101973. 10.1016/j.erss.2021.101973

  • CrossRef
  • Google Scholar
  • View reference in article

• 37

  IslamMd A.BhuiyanMd A.-F. (2024). Factors affecting citizen safety of urban transportation service in Bangladesh: the case of Pabna municipality. Heliyon10.2, e0728. 10.1016/j.heliyon.2024.e24697

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 38

  IslamN.ChowdhuryN. A.ZamanM. A.HuqA. S.RaselS. F. M. (2026). An econometric analysis in identifying behavioral and demographic factors associated with road crash severity in Bangladesh: evidence from the Dhaka metropolitan City. PLOS ONE21 (1), e0340607. 10.1371/journal.pone.0340607

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 39

  JackmanD. (2019). Violent intermediaries and political order in Bangladesh. Eur. J. Dev. Res.31 (4), 705–723. 10.1057/s41287-018-0178-8

  • CrossRef
  • Google Scholar

• 40

  KirchlerE.HoelzlE.WahlI. (2008). Enforced versus voluntary tax compliance: the “slippery slope” framework. J. Econ. Psychol.29 (2), 210–225. 10.1016/j.joep.2007.05.004

  • CrossRef
  • Google Scholar

• 41

  KlepperS.NaginD. (1989). The deterrent effect of perceived certainty and severity of punishment revisited. Criminology27 (4), 721–746. 10.1111/j.1745-9125.1989.tb01052.x

  • CrossRef
  • Google Scholar
  • View reference in article

• 42

  LoganE.KayeS.-A.LewisI. (2019). The influence of the revised reinforcement sensitivity theory on risk perception and intentions to speed in young male and female drivers. Accid. Analysis and Prev.132, 105291. 10.1016/j.aap.2019.105291

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 43

  MacLeodK. E.Karriker-JaffeK. J.RaglandD. R.SatarianoW. A.Kelley-BakerT.LaceyJ. H. (2015). Acceptance of drinking and driving and alcohol-involved driving crashes in California. Accid. Analysis and Prev.81, 134–142. 10.1016/j.aap.2015.04.035

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 44

  MahikulW.AiyasuwanO.ThanartthanaboonP.ChancharoenW.AchararitP.SirisombatT.et al (2022). Factors affecting bus accident severity in Thailand: a multinomial logit model. PLOS ONE17 (11), e0277318. 10.1371/journal.pone.0277318

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 45

  MeyerJ. W.RowanB. (1977). Institutionalized organizations: formal structure as myth and ceremony. Am. J. Sociol.83 (2), 340–363. 10.1086/226550

  • CrossRef
  • Google Scholar
  • View reference in article

• 46

  MillerD. T. (2023). A century of pluralistic ignorance: what we have learned about its origins, forms, and consequences. Front. Soc. Psychol.1, 1260896. 10.3389/frsps.2023.1260896

  • CrossRef
  • Google Scholar
  • View reference in article

• 47

  MoanI. S.Jostein Rise (2011). Predicting intentions not to “drink and drive” using an extended version of the theory of planned behaviour. Accid. Analysis and Prev.43 (4), 1378–1384. 10.1016/j.aap.2011.02.012

  • CrossRef
  • Google Scholar
  • View reference in article

• 48

  MøllerM.JanstrupK. H. (2021). Crash involvement among unlicensed 17 year old drivers before and after licensing at 17 was allowed. Accid. Analysis and Prev.156, 106109. 10.1016/j.aap.2021.106109

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 49

  MoojenR.GillebaartM.de RidderD. (2023). Using misperceived social norms as a license: does pluralistic ignorance trigger complacency in the food environment?Soc. Influ.18 (1), 2251643. 10.1080/15534510.2023.2251643

  • CrossRef
  • Google Scholar
  • View reference in article

• 50

  NaginD. S. (1998). Criminal deterrence research at the outset of the twenty-first century. Crime Justice23, 1–42. 10.1086/449268

  • CrossRef
  • Google Scholar
  • View reference in article

• 51

  Nguyen-PhuocD. Q.Oviedo-TrespalaciosO.SuD. N.De GruyterC.NguyenT. (2020). Mobile phone use among car drivers and motorcycle riders: the effect of problematic mobile phone use, attitudes, beliefs and perceived risk. Accid. Analysis and Prev.143, 105592. 10.1016/j.aap.2020.105592

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 52

  NordfjærnT.JørgensenS.RundmoT. (2011). A cross-cultural comparison of road traffic risk perceptions, attitudes towards traffic safety and driver behaviour. J. Risk Res.14 (6), 657–684. 10.1080/13669877.2010.547259

  • CrossRef
  • Google Scholar
  • View reference in article

• 53

  PodsakoffP. M.MacKenzieS. B.LeeJ. Y.PodsakoffN. P. (2003). Common method biases in behavioral research: a critical review of the literature and recommended remedies. J. Appl. Psychol.88 (5), 879–903. 10.1037/0021-9010.88.5.879

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 54

  PojaniD.SteadD. (2017). “The urban transport crisis in emerging economies: an introduction,” in The Urban Transport Crisis in Emerging Economies. the Urban Book Series (Springer), 1–10. 10.1007/978-3-319-43851-1_1

  • CrossRef
  • Google Scholar
  • View reference in article

• 55

  PrattT. C.CullenF. T.BlevinsK. R.DaigleL. E.MadensenT. D. (2017). “The empirical status of deterrence theory: a meta-analysis,” in Taking Stock (London, United Kingdom: Routledge), 367–395. 10.4324/9781315130620-14

  • CrossRef
  • Google Scholar
  • View reference in article

• 56

  RatanawarahaA.ChalermpongS.SamnakornP. (2019). How vehicle types and operator’s legal status affect safety of interprovincial buses in Thailand. Eng. J.23 (1), 35–47. 10.4186/ej.2019.23.1.35

  • CrossRef
  • Google Scholar
  • View reference in article

• 57

  ReasonJ.MansteadA.StradlingS.BaxterJ.CampbellK. (1990). Errors and violations on the roads: a real distinction?Ergonomics33, 1315–1332. 10.1080/00140139008925335

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 58

  RingleC. M.SarstedtM. (2016). Gain more insight from your PLS-SEM results: the importance-performance map analysis. Industrial Management and Data Systems116 (9), 1865–1886. 10.1108/IMDS-10-2015-0449

  • CrossRef
  • Google Scholar

• 59

  RingleC. M.WendeS.Jan-Michael Becker (2024). Smartpls 4. Smartpls Gmbh. Bönningstedt, Germany. Available online at: https://www.smartpls.com.

  • Google Scholar
  • View reference in article

• 60

  RoyS. K.BalajiM.QuaziA.QuaddusM. (2018). Predictors of customer acceptance of and resistance to smart technologies in the retail sector. J. Retail. Consumer Serv.42, 147–160. 10.1016/j.jretconser.2018.02.005

  • CrossRef
  • Google Scholar
  • View reference in article

• 61

  SagbergF. (2018). Characteristics of fatal road crashes involving unlicensed drivers or riders: implications for countermeasures. Accid. Analysis and Prev.117, 270–275. 10.1016/j.aap.2018.04.025

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 62

  SahaB.FatmiM. R.RahmanMd M. (2021). Traffic crashes in Dhaka, Bangladesh: analysing crashes involving unconventional modes, pedestrians and public transit. Int. J. Inj. Control Saf. Promot.28 (3), 347–359. 10.1080/17457300.2021.1928230

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 63

  SarstedtM.RichterN. F.HauffS.RingleC. M. (2024). Combined importance–performance map analysis (cIPMA) in partial least squares structural equation modeling (PLS–SEM): a SmartPLS 4 tutorial. J. Mark. Anal.12 (4), 746–760. 10.1057/s41270-024-00325-y

  • CrossRef
  • Google Scholar
  • View reference in article

• 64

  ShmueliG.SarstedtM.HairJ. F.CheahJ. H.TingH.VaithilingamS.et al (2019). Predictive model assessment in PLS-SEM: guidelines for using PLSpredict. Eur. Journal Marketing53 (11), 2322–2347. 10.1108/EJM-02-2019-0189

  • CrossRef
  • Google Scholar
  • View reference in article

• 65

  SlovicP. (1987). Perception of risk. Science236, 280–285. 10.1126/science.3563507

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 66

  SolaimanS. M. (2021). Corporate vehicular manslaughter provisions in the Bangladesh road transport act 2018: a textual comparison with their equivalents in Australia. Crim. Law Forum32, 125–173. 10.1007/s10609-020-09404-y

  • CrossRef
  • Google Scholar
  • View reference in article

• 67

  SuykensB. (2018). The mimicry of the state as a state practice: the regulation of rickshaw licenses in Dhaka (Bangladesh). Crit. Asian Stud.50 (3), 422–441. 10.1080/14672715.2018.1491319

  • CrossRef
  • Google Scholar
  • View reference in article

• 68

  TankebeJ.BoakyeK. E.Agaawena AmagnyaM. (2020). Traffic violations and cooperative intentions among drivers: the role of corruption and fairness. Polic. Soc.30 (9), 1081–1096. 10.1080/10439463.2019.1636795

  • CrossRef
  • Google Scholar
  • View reference in article

• 69

  TBS Report (2024). 5,024 killed in road accidents in 2023: BRTA. Bus. Stand.Available online at: https://www.tbsnews.net/bangladesh/5024-killed-road-accidents-2023-brta-775982.

  • Google Scholar
  • View reference in article

• 70

  TylerT. R. (2006). Psychological perspectives on legitimacy and legitimation. Annu. Rev. Psychol.57 (1), 375–400. 10.1146/annurev.psych.57.102904.190038

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 71

  TylerT. R.HuoY. J. (2002). Trust in the Law: Encouraging Public Cooperation with the Police and Courts. New York: Russell Sage Foundation. Available online at: https://www.russellsage.org/publications/book/trust-law.

  • Google Scholar
  • View reference in article

• 72

  United News of Bangladesh (2026). 7,359 people killed in road accidents in 2025: road safety foundation’s count almost static. UNB. Available online at: https://unb.com.bd/category/Bangladesh/7359-people-killed-in-road-accidents-in-2025-road-safety-foundations-count-almost-static/177075 (Accessed July 27, 2026).

  • Google Scholar

• 73

  VargasF. (2016). “Legal and social norms for development: why legal reform of the informal economy failed to influence vulnerable groups in developing countries.” in Social and Legal Norms: Towards a Socio-Legal Understanding of Normativity (Abingdon, United Kingdom: Routledge), 231–246. Available online at: https://www.taylorfrancis.com/chapters/edit/10.4324/9781315609416-18/legal-social-norms-development-legal-reform-informal-economy-failed-influence-vulnerable-groups-developing-countries-ana-maria-vargas-falla.

  • Google Scholar

• 74

  WeinerB. (1985). An attributional theory of achievement motivation and emotion. Psychol. Rev.92 (4), 548–573. 10.1037/0033-295X.92.4.548

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 75

  WinterS. C.MayP. J. (2001). Motivation for compliance with environmental regulations. J. Policy Analysis Manag.20 (4), 675–698. 10.1002/pam.1023

  • CrossRef
  • Google Scholar

• 76

  World Health Organization (2023). Global Status Report on Road Safety 2023. Geneva, Switzerland: World Health Organization. Available online at: https://www.who.int/publications/i/item/9789240086517.

  • Google Scholar