RESEARCH DOSSIER
Authority, Obedience, and Institutional Legitimacy
Abstract. Authority, obedience, and institutional legitimacy form an interdependent field of social, political, organizational, and legal psychology. This monograph investigates why people accept instructions, distinguish authorized roles from coercive commands, comply with institutions, challenge orders, or refuse participation. It reconstructs classical obedience experiments and their variants, contemporary ethically modified paradigms, procedural-justice and legitimacy research, organizational authority, professional responsibility, dissent, and the methodological difficulty of inferring behavioral compliance from attitudes. Original experiments, published reanalyses, randomized field studies, observational surveys, systematic evidence syntheses, and conceptual scholarship are compared by research design and inferential scope. The resulting synthesis identifies substantial evidence for context-sensitive instruction-following and associations between fair treatment and perceived legitimacy, while recognizing incomplete causal evidence for durable voluntary compliance and important variability across institutions. It extends the research agenda to algorithmic authority through explicitly testable hypotheses. The method is critical narrative synthesis, without preregistered systematic review or new human-participant data.
Keywords: authority; obedience; institutional legitimacy; procedural justice; compliance; responsibility; dissent; moral conviction; organizational justice; psychological safety; police legitimacy; social identification; professional ethics; algorithmic authority; causal inference
Introduction: The Psychological Problem of Institutional Authority
Institutional authority organizes human conduct through recognized offices, professional expertise, procedures, and claims to justified direction. A supervisor assigns work, a physician recommends treatment, a court issues a ruling, and a police officer gives an instruction. These episodes involve related yet separable psychological questions: whether the person identifies an authority, whether the particular demand appears legitimate, whether compliance is expected, and whether the recipient acts. An institution can be widely regarded as rightful while a specific order is rejected, and an order can be obeyed under threat by someone who rejects the institution. The research object is therefore a configuration of perceived authorization, interpretive judgment, social relationships, sanctions, and observable behavior. Legitimacy research connects perceived rightful power with cooperation and obligation, whereas experimental obedience studies examine decisions under an immediately communicated directive. The present monograph integrates both programs while preserving their differences in measurement and causal inference. Tyler (2006) Nagin & Telep (2017)
The guiding research questions concern conditions under which authority elicits compliance, the processes that make institutions appear worthy of deference, and the boundaries at which principled refusal becomes more likely. A complete explanation must consider the authority figure's credibility, the perceived social purpose of a task, competing moral claims, institutional procedures, local group norms, and practical consequences of disagreement. These components operate at different temporal scales. Instructions may alter a decision in seconds; reputational judgments accumulate through recurring encounters; durable legitimacy can reflect decades of institutional conduct and social learning. Equating these processes would obscure how a brief laboratory episode connects to legal compliance or workplace cooperation. Haslam et al. (2014) Sunshine & Tyler (2003)
This research is a critical narrative synthesis of historical studies, original experiments, archival reinterpretations, organizational investigations, surveys, field trials, methodological critiques, and relevant conceptual analyses. The cited literature is selected for its capacity to test or clarify major mechanisms, not through an exhaustive preregistered search. Findings are interpreted according to the evidence they actually provide: experimental manipulation may demonstrate a local causal influence, cross-sectional association describes covariation, and theoretical accounts articulate mechanisms needing direct tests. The historical case of Milgram is examined alongside ethically adapted research; legal and organizational findings are considered through their actual institutional populations. Such distinctions permit an integrated account of authority without attributing a universal obedience percentage or an unconditional legitimacy effect to humanity. Milgram (1963) Burger (2009) Nagin & Telep (2020)
Definitions: Power, Authority, Obedience, and Legitimacy
Power denotes a capacity to influence another's opportunities, actions, or outcomes, frequently through the control of resources or sanctions. Authority denotes an institutionally or socially recognized claim to direct conduct within a specified domain. Obedience is behavioral conformity to a perceived directive issued by an authority; it should be distinguished from agreement with advice, spontaneous cooperation, or conformity to peer expectations. Institutional legitimacy is a judgment that an institution or office has a rightful claim to exercise power. Such legitimacy may encompass felt obligation, moral alignment, procedural trustworthiness, and evaluations of institutional performance, depending on the theory and measurement model. These definitions establish that compliance can be produced by fear, habit, identification, instrumental calculation, perceived expertise, or acceptance of rightful authority. None of these pathways can be inferred merely from the same outward behavior. Tyler (2006) Tankebe (2013)
In empirical work, the explanatory construct and the measured indicator must be distinguished. An agreement item such as 'people should follow the law' can reflect a generalized normative proposition, while willingness to cooperate with a particular officer concerns a situated choice. A direct measure of instruction following records action under specified conditions; an attitude questionnaire records respondents' reports or beliefs. Procedural fairness may be measured through perceived respect, opportunity for voice, impartiality, or trustworthy motives. Organizational justice inventories additionally separate distributive outcomes from informational and interpersonal treatment. These operationalizations are useful precisely because they target different parts of the authority relationship. Measurement studies caution against collapsing theoretically distinct dimensions into one undifferentiated legitimacy score. Colquitt (2001) Tankebe (2013)
An authorized directive has at least three analytically distinct properties: the office or role from which it originates, the scope of powers attributed to that role, and the ethical or legal status of the requested action. A legitimate role does not automatically legitimize every possible instruction. Likewise, perceived expertise supports the credibility of factual guidance without independently settling whether the speaker has normative jurisdiction. Research on moral conviction demonstrates that individuals sometimes reject even institutionally recognized authority when they understand a demand to violate a binding value. The scientific task is to measure the degree to which people separately endorse the office, the process, and the act, then determine which judgments predict behavior in the observed situation. Skitka et al. (2009) Tyler (2003)
Historical Formation of the Obedience Research Problem
Obedience became a central experimental question when social psychologists sought to explain how ordinary people respond to demands they experience as troubling. The research tradition emerged amid twentieth-century concern with bureaucratic command, institutional responsibility, and participation in harmful collective actions. Stanley Milgram's 1963 article gave the problem a controlled behavioral paradigm: a participant assigned the role of teacher received instructions from an experimenter to continue administering increasingly severe apparent shocks to a learner. The learner was a confederate, and the shocks were simulated. The task created a conflict between the experimenter's instructions and the learner's expressed distress. This procedure established a memorable experimental setting; its historical significance comes from its capacity to make instruction, hesitation, and refusal observable within one encounter. Milgram (1963) Blass (1999)
The subsequent history cannot be compressed into one iconic statistic. Milgram conducted multiple conditions varying the apparent authority and its presence, the learner's proximity, and the participation of others. Later scholarship reevaluated the archive, the structure of the experimenter's prompts, and the social meanings participants attributed to the study. The reconstructed corpus of conditions reveals substantial variation in behavior, and competing interpretations emphasize different causal pathways: obedience to command, identification with scientific purpose, task involvement, and interactional negotiation. These interpretations have distinctive empirical implications, yet the archived studies were built within a historically particular setting and cannot independently settle the prevalence of obedience across all contemporary institutions. Haslam et al. (2014) Gibson (2013) Haslam & Reicher (2012)
Parallel developments in legal and organizational psychology widened the question from the laboratory experimenter to institutions exercising authority over repeated encounters. Studies of police and public cooperation began examining procedural fairness, perceived legitimacy, and feelings of obligation; organizational scholars elaborated distinctions among fair procedures, interpersonal respect, and outcomes. The resulting field investigates both the immediate compliance decision and the enduring relational environment in which institutions gain or lose credibility. These historical streams were motivated by overlapping concerns, but they employ different outcomes and samples. A coherent monograph therefore examines them through a common vocabulary while preserving their research designs rather than treating historical succession as proof of theoretical convergence. Sunshine & Tyler (2003) Colquitt et al. (2001) Tyler (2006)
The Institutional Ecology of Commands and Decisions
Instructions are interpreted within organized sequences of roles and routines. A participant hears a directive against the background of the institution's stated purpose; an employee receives a request through a reporting hierarchy; a patient interprets a medical recommendation in the context of expected professional competence. For each episode, the relevant ecology includes formal jurisdiction, immediate interpersonal conduct, organizational norms, time pressure, channels for clarification, and the practical ability to exit. A directive may be heard as a request to assist a legitimate task, an obligation attached to membership, or a coercive threat. These interpretations differ even when the sentence spoken by the authority figure is identical. The procedural-justice and organizational-safety literatures show how patterns of interpersonal treatment and voice can influence evaluations of institutional encounters. Edmondson (1999) Hough et al. (2010)
Authority relations contain both vertical and horizontal dependencies. Vertical authority concerns the recognized superior role, while horizontal dependencies include peers, beneficiaries, clients, professional colleagues, and affected third parties. A worker considering an unsafe order may anticipate an evaluation by a supervisor and simultaneously feel accountable to coworkers or customers. In Milgram's experimental setting, the relation to the experimenter competed with the moral salience of the learner. The empirical synthesis across experimental variations links obedience to features of both relations, including the authority's presence and the learner's proximity. This pattern suggests that a single trait labelled obedience is inadequate to describe the observable variation; situational relations alter what the demanded behavior signifies. Haslam et al. (2014) Milgram (1963)
Organizations also structure access to relevant information. When consequences are remote, responsibility is distributed, or objections are discouraged, a decision-maker may have difficulty recognizing the implications of compliance. Conversely, avenues for consultation and protected challenge can change what courses of action feel available. Research on employee voice examines why potentially useful concerns are expressed or withheld, while psychological-safety studies associate a shared climate for interpersonal risk-taking with learning behaviors in teams. These findings motivate a more precise research proposition: institutional authority becomes behaviorally consequential partly through its control of the decision environment, not only through conscious endorsement of a superior's personal status. The proposition requires study designs that measure the available alternatives and their perceived costs. Morrison (2014) Edmondson (1999)
Distinct Mechanisms of Compliance
Instrumental compliance occurs when a person responds to anticipated rewards, sanctions, or material consequences. Normative compliance arises when a rule is treated as proper within a valued social relationship or institution. Informational reliance occurs when the apparent authority is believed to possess superior knowledge. Identification with a shared collective purpose can encourage cooperation even when commands are not presented as threats. Habit and repeated role routines may reduce the amount of explicit deliberation devoted to ordinary instructions. These pathways may coexist in one decision. A survey question about perceived legitimacy, a record of disciplinary sanctions, and an observed response to a request each illuminate only part of this causal ensemble. Research on legitimacy and power makes these mechanisms distinguishable theoretical alternatives that require more than correlation for adjudication. Tyler (2006) Keltner et al. (2003)
Authority cues gain meaning from their context. Official titles and uniforms can signal institutional recognition; technical language may signal competence; consistency across representatives can create an impression of coordinated governance. Yet a cue's influence depends on whether recipients regard the associated institution and task as trustworthy. Milgram's variations and later reconstructions show that physical presence, coherence of the experimenter's directions, and the learner's relationship with the participant all covary with the likelihood of continued action. In a modern adapted experiment, the prestige of a university and a research rationale did not uniformly account for obedient responding, underscoring the need to test rather than assume the strength of particular cues. Haslam et al. (2014) Grzyb et al. (2025)
A robust account must include refusal as a positive behavioral outcome. Declining a request may follow moral conviction, skepticism about authority's jurisdiction, identification with a different group, fear of adverse consequences, or an accessible procedural channel for contesting an instruction. Some dissent is impulsive, and some is deliberative; some challenges improve institutional performance, while others undermine a legitimate function. The normative conflict model of dissent examines how group identification can motivate critical engagement when members believe prevailing group practice violates an important norm. Moral-conviction research further indicates limits to perceived legitimacy as a predictor of deference. Consequently, willingness to question authority and willingness to cooperate are not fixed opposites; both can emerge from identification with the institution's stated principles. Packer (2008) Skitka et al. (2009)
Milgram's Original Behavioral Experiment
Milgram's 1963 report describes forty men recruited for an ostensible investigation of learning. Each participant was assigned to administer an escalating sequence of apparent electric shocks after errors by a learner, whose distress and eventual silence were scripted elements of the situation. The participant was repeatedly encouraged by an experimenter to continue despite objections. The shocks were simulated, but the participant was led to believe they were real. The design's central behavioral measure was the highest shock level a participant would administer before stopping. Twenty-six of the forty reached the final marked voltage, the source of the often-repeated figure of 65 percent. That result describes one experimental condition and participant pool, not a timeless constant of human behavior. Milgram (1963)
The original article also documented visible conflict. Participants often hesitated, requested clarification, or showed signs of tension as the learner protested. Such episodes matter because continuation to a maximal setting cannot be equated with emotional indifference or a settled belief that the act was proper. The experimenter maintained the formal organization of the task, while the learner expressed direct interpersonal claims on the participant's judgment. These competing demands defined the situation. The outcome was therefore a sequence of negotiated actions under strong experimental framing. The report establishes that substantial numbers continued in this arrangement, while the observed distress resists any interpretation in which their behavior simply reflected uncomplicated desire to cause harm. Milgram (1963) Gibson (2013)
The experiment's interpretation must be constrained by its ethics and design. Participants faced deception and in some cases acute distress; modern research protections limit faithful reproduction of the historical procedure. Their behavior occurred in a university-sponsored laboratory and was structured by a single experimenter's intervention and a staged scientific objective. The original participant population was narrow, and the encounter involved an artificial time frame and reward structure. Ethical concerns are therefore also an inference issue: the unusual intensity of this design made certain responses observable but prevents unqualified translation to workplaces, democracies, or contemporary populations. Explanations that invoke institutional legitimacy need to specify precisely what evidence in the study measures recognition of rightful authority, as opposed to situational trust in the experiment or compliance with its immediate demands. Blass (1999) Burger (2009)
Variation Across the Milgram Conditions
The experimental series varied who issued instructions, how close the learner was, whether the participant acted directly, and whether competing social signals supported defiance. These variations are essential evidence because they convert a famous demonstration into a set of conditional observations. Haslam, Loughnan, and Perry reconstructed twenty-one conditions with a common maximum-voltage endpoint and identified 740 participants. Across those conditions, 323 reached the endpoint, an aggregate rate of 43.6 percent. Their analysis coded properties of the experimenter, participant, learner, and setting; several relational features were associated with continuation. The aggregate pools heterogeneous small conditions, and its meaning is the distribution of behavior across the experimental variants included, not a nationally representative estimate of obedience. Haslam et al. (2014)
The same synthesis found lower rates when an experimenter was distant, inconsistent, or less directive and when group pressure encouraged disobedience. More direct or intimate contact with the learner was also associated with lower rates, although individual condition comparisons have limited precision. These patterns show that the participant responded to a relational configuration rather than a single immutable authority stimulus. However, the authors reconstructed a historical series rather than randomly assigning a newly recruited common population across every coded feature. Some conditions bundled multiple changes, and the study's coding necessarily simplified complex interactions. Regression across variants helps identify plausible explanations but cannot fully isolate each mechanism from historically confounded experimental details. Haslam et al. (2014)
The experiments also illustrate the difference between a successful manipulation and its psychological interpretation. Moving a participant closer to the learner can increase empathic salience, alter perceived responsibility, and make the conflict harder to ignore; decreasing the experimenter's presence can reduce supervision, weaken perceived authority, or change expectations about the task. The same outcome shift is compatible with several intervening processes. An explanatory test would measure or manipulate those candidate mediators without compromising participants' welfare. Studies that report only an obedience percentage leave the mechanism underidentified. Contemporary research can advance by specifying competing psychological predictions before examining how authority presence, peer support, institutional purpose, and affected-person salience combine to shape a particular decision. Haslam et al. (2014) Haslam & Reicher (2012)
Ethically Modified Replications and Their Limits
Burger's 2009 investigation adapted Milgram's procedure to contemporary ethical expectations. Participants were screened, a confederate was used, and the experiment terminated at a point corresponding to 150 volts rather than exposing people to the historical full sequence. The stopping rule reduced risk and preserved a decision point at which the learner had made a clear objection. The study assessed whether participants were willing to continue past that checkpoint under controlled circumstances. Its finding of substantial willingness demonstrates the continuing relevance of authority-framed decisions, but the modified design measures a proximal behavioral choice, not actual completion of the older experiment's maximum-voltage sequence. The original 65 percent and the adapted continuation proportion therefore cannot be treated as directly interchangeable outcomes. Burger (2009)
The ethical redesign also highlights a problem of selective inference. Early willingness to continue could predict later persistence, but later conflict and opportunities for refusal might change behavior. Participants entering a contemporary experiment may bring different expectations about research ethics, institutions, and staged tasks. Screening exclusions may change the recruited population, and the instructions retain cues indicating that the setting is a carefully supervised study. Inference should therefore remain anchored to the manipulated situation and operational outcome. A legitimate question concerns whether similar social mechanisms recur in other forms, rather than whether a numerical historical result has been replicated without qualification. These considerations should guide both power calculations and preregistered statements about what behavioral endpoint the study estimates. Burger (2009) Blass (1999)
Grzyb and colleagues' 2025 research extended the discussion by testing institutional and research-justification cues in an adapted obedience paradigm. The design compared contextual claims concerning institutional prestige and the stated purpose of the activity. The paper reported that those factors did not produce the straightforward differences predicted by simple prestige accounts under the specific conditions tested. This result narrows a theory that treats the famous university setting as a sufficient explanation, while remaining local to the studied task, sample, and manipulations. It does not show that institutional reputation can never affect compliance. Taken together with earlier variations, this evidence favors explicitly conditional hypotheses about how credibility, group purpose, sanctions, and relational accountability interact rather than assuming a universally dominant causal cue. Grzyb et al. (2025) Haslam et al. (2014)
Obedience as Interaction, Negotiation, and Identification
A command is not merely a stimulus delivered to a passive individual. Its meaning is established through talk, the social relationship, and the participant's interpretation of what the organization is attempting to achieve. Gibson's analysis of Milgram's recordings and experimental exchanges treats obedience as an interactional process in which participants negotiated, resisted, or reformulated the demand. The familiar sequence of the experimenter's prods acquired significance from how participants responded to them. Especially revealing is the limited effectiveness of an explicitly coercive prompt, which complicates a picture in which increasing force automatically increases compliance. This reading identifies rhetorical and conversational processes that are often lost when experiments are reduced to final shock values. Gibson (2013)
Haslam and Reicher propose that identification with an institution's valued collective purpose can produce engagement with authority. Under this account, participants may understand themselves as contributing to important science and may consequently view continuation as cooperation with a worthwhile project. This argument reorganizes the explanatory focus around what the authority stands for, whether participants identify with that objective, and how the affected person's welfare enters the moral appraisal. It also supplies testable predictions about the effects of shared identity and competing moral identification. The theory should be evaluated against evidence rather than substituted for a direct report of each participant's internal motives; archived behaviors are consistent with multiple cognitive and interpersonal accounts. Haslam & Reicher (2012) Haslam et al. (2015)
An immersive digital restaging examined the relation between identification with scientist and learner and willingness to administer apparent shocks in a mediated context. Its design created a different type of realism and measured experiences that could not be assessed in Milgram's original archival record in the same way. The findings support the relevance of social identification as an explanatory variable within that environment, while virtual simulation introduces its own differences in stakes and perceived harm. The strongest general inference is that obeying a recognized authority depends partly on how the actor constructs the action's meaning and role commitments. Research must still determine how these commitments interact with explicit coercion, situational uncertainty, and observable consequences in natural institutional settings. Haslam et al. (2015) Gibson (2013)
Responsibility, Agency, and the Limits of Delegation
Institutional hierarchies allocate decisions among roles, yet their psychological operation can blur the perceived location of responsibility. A person may act because an instruction appears to transfer decision authority to a supervisor or expert, while still recognizing that the action has consequences for another individual. The original obedience study makes this conflict visible: participants confronted a learner's distress while the experimenter claimed responsibility for continuing the task. To interpret a person's behavior, researchers need separate measures of perceived choice, causal contribution, role obligation, and moral blame. A response to the question 'Who was responsible?' after an event may reflect self-protection, actual beliefs, or socially learned narratives, and cannot by itself explain the earlier decision. Milgram (1963) Gibson (2013)
A useful distinction separates institutional accountability from moral delegation. Institutions can legitimately assign responsibilities to professionals, managers, or officials, and participants often need to rely on qualified guidance. Nevertheless, the recipient may retain opportunities to request reasons, verify scope, identify a conflicting obligation, or refuse. Which opportunities are perceived as realistic depends on the hierarchy and the potential costs of objection. Organizational voice research describes how anticipated reactions from superiors, workplace relationships, and shared norms shape whether concerns are raised. These findings suggest measuring the procedural accessibility of challenge alongside individual attitudes toward authority. A culture that permits clarification can sustain legitimate coordination while reducing the chance that a questionable directive proceeds without scrutiny. Morrison (2014) Edmondson (1999)
The distinction also matters for theoretical explanations. If obedience results chiefly from fear of punishment, reducing sanctions should change behavior even if legitimacy attitudes remain fixed. If it results chiefly from confidence in professional expertise, evidence that the official lacks competence should be consequential. If it results from identification with institutional values, showing that an order violates those values should encourage challenge among strongly identified members. If obedience results from habit, requiring an explicit reflective justification may disrupt automatic continuation. These are separable predictions, not settled universal findings. A scientific framework of responsibility must therefore treat agency as an empirically investigable set of perceptions and choices within a structured context, rather than relying solely on retrospective moral classification. Packer (2008) Skitka et al. (2009) Tyler (2006)
Legitimacy as a Psychological Judgment of Rightful Authority
Legitimacy concerns an institution's perceived entitlement to exercise authority and a corresponding evaluation of whether compliance is appropriate. Tyler's psychological approach distinguishes legitimacy-based cooperation from compliance driven only by surveillance or threat. Citizens may follow rules because they regard the institution and its procedures as proper, allowing organizations to coordinate behavior without monitoring every individual act. This is an explanatory proposition connecting normative judgments with collective order; it becomes empirical when legitimacy, felt duty, institutional behavior, and compliance are measured independently. The same person can endorse a legal system while disapproving of a particular officer or procedure. Research therefore requires a multilevel account that distinguishes broad institutional support, office-specific evaluation, and judgment of a specific act. Tyler (2006) Tyler (2003)
The empirical construct has competing definitions. Some instruments define legitimacy through perceived obligation to obey; others emphasize moral alignment, trust in institutions, legality, effectiveness, or procedural fairness. These dimensions may correlate, but a model that builds obedience into the definition of legitimacy risks using one form of stated compliance to predict another. Tankebe's analysis of police legitimacy argues for a more differentiated treatment of legality, distributive and procedural justice, and effectiveness, exposing conceptual disputes over which dimensions form the construct itself. Jackson and colleagues examine obligation and moral alignment as potentially distinct aspects of perceived legal authority. An adequate study reports item content and latent measurement structure rather than interpreting all favorable evaluations as the same psychological property. Tankebe (2013) Jackson et al. (2012)
Normative legitimacy is dynamic. A procedure experienced as fair can strengthen willingness to recognize an institution's claims, while arbitrary treatment may generate distrust or selective refusal. Yet favorable responses need not imply agreement with every outcome, and fairness assessments can be influenced by expectations, personal experiences, information about performance, and group relations. The appropriate research question is which component of perceived legitimacy changes after a specified encounter, whether that change endures, and whether any resulting behavioral difference extends beyond survey intentions. The evidence for these stages is unequal: cross-sectional associations are well documented, whereas durable effects on actual lawbreaking or long-term obedience require stronger longitudinal and experimental identification. Sunshine & Tyler (2003) Nagin & Telep (2017) Nagin & Telep (2020)
Procedural Justice: Voice, Neutrality, Respect, and Trustworthy Motives
Procedural justice concerns the perceived fairness of the process through which institutions make and implement decisions. The tradition emphasizes opportunities to present one's view, neutrality of decision-making, interpersonal dignity, and the apparent trustworthiness of decision makers. These qualities may communicate that the institution recognizes a person's standing even when the immediate outcome is unfavorable. This mechanism offers a psychologically concrete path from a single encounter to broader evaluations of authority. Police legitimacy studies and organizational research converge on the importance of treatment, although the constructs differ across settings and measurement instruments. To assess the proposed pathway, studies should distinguish the manipulation of a procedure from subsequent reports that the procedure was fair, and from any later cooperation or compliance behavior. Tyler (2003) Colquitt (2001)
Procedural fairness can affect interpretation of a decision independently of whether the outcome benefits the individual. An opportunity to explain one's position may improve perceived recognition; consistent standards can support impartiality judgments; respectful communication may reduce identity threat. Nonetheless, people also care about distributions, substantive accuracy, and the consequences of the decision. Organizational justice research separates distributive, procedural, interpersonal, and informational elements so that a favorable process is not assumed to compensate automatically for serious harms. Its meta-analytic literature finds relationships between justice dimensions and a range of workplace attitudes and behaviors, while demonstrating the importance of keeping outcome measures conceptually separate from fairness perceptions. Colquitt (2001) Colquitt et al. (2001)
A procedural-justice hypothesis can be stated causally: holding substantive outcome and legitimate jurisdiction constant, increasing respectful and impartial treatment should increase perceived fairness, and perhaps subsequent institutional legitimacy. A second, more demanding hypothesis concerns whether the change yields repeated voluntary compliance under circumstances where noncompliance is possible and detectable. The first may be studied in randomized encounters or vignettes; the second requires observed follow-up and controls for enforcement, baseline behavior, and selection. The distinction is crucial because reported support for an officer or institution can rise without a corresponding reduction in offending. Reviews of policing interventions explicitly caution that improvements in the citizen's encounter experience and changes in criminal behavior are different outcome domains. Mazerolle et al. (2013) Mazerolle et al. (2013) Nagin & Telep (2017)
Police-Citizen Encounters and Institutional Trust
Police encounters provide an important field setting because officials exercise legal authority in situations where citizens may experience uncertainty, coercive capacity, or unequal influence. Sunshine and Tyler used survey evidence to investigate how perceptions of procedural justice and legitimacy relate to public support for policing. The results helped establish an empirical research program treating fair treatment as psychologically meaningful beyond the perceived risk of punishment. Their measures document relationships among beliefs, trust, and declared willingness to cooperate. Because much evidence is observational, shared circumstances and prior attitudes may influence all measured perceptions. The appropriate interpretation is an established association consistent with a legitimacy mechanism, rather than proof that improving fairness in any single encounter necessarily changes later behavior. Sunshine & Tyler (2003)
Jackson and collaborators analyzed survey data on why citizens comply with law and how they evaluate legal institutions. Their account emphasizes the distinction between a felt obligation to obey and moral alignment with institutional principles. That distinction matters when people see an institution as lawful but dispute one of its practices, or when they trust a particular representative without endorsing the whole organization. Hough and colleagues likewise connected procedural justice, trust, and legitimacy as components of the relationship between institutions and the public. Such studies offer valuable construct and population-level evidence, but retrospective self-reports of compliance remain vulnerable to desirability effects, unobserved confounding, and common-method correlations. Jackson et al. (2012) Hough et al. (2010)
Institutional trust should be measured at several levels: confidence in the competence and integrity of a specific official, expectations about the institution's treatment of persons generally, and willingness to accept fair decisions in the future. These judgments can move in different directions. A person may view one encounter as respectful while remaining skeptical that institutional rules are applied consistently; another may support the institution broadly but regard a particular intervention as unlawful. Repeated-contact designs and administrative measures can help identify how experiences accumulate over time. They must also confront selection: people who encounter the police frequently may differ systematically from those who rarely do, and the circumstances eliciting intervention may themselves shape both treatment and subsequent behavior. Tankebe (2013) Nagin & Telep (2020)
Field Trials, Systematic Reviews, and the Causal Evidence Gap
Randomized field trials provide a stronger test of whether procedural changes alter citizens' immediate evaluations. Mazerolle, Antrobus, Bennett, and Tyler examined an intervention designed to improve the procedural-justice content of police contact. Random assignment is valuable because it helps reduce preexisting differences among encounter conditions, permitting more credible causal inference about the specific measured response. But randomization does not automatically establish broad effects on future offending, a different outcome that may occur rarely or require prolonged follow-up. Researchers must report how the intervention was implemented, how directly fairness was manipulated, whether participants perceived it, and which measured attitudes or behaviors subsequently changed. In addition, field trials are embedded in specific enforcement environments whose procedures may shape transportability. Mazerolle et al. (2013)
Mazerolle and colleagues' systematic review aggregated research on policing interventions intended to promote legitimacy and procedural fairness. The evidence generally supports improvements in immediate perceptions of procedural treatment and related attitudes, while individual outcomes vary in precision and design. Some pooled effects on perceived legitimacy are uncertain, and the range of policing settings is narrower than the institutional world to which conclusions are often generalized. This review should be read as an evaluation of specified interventions and outcomes rather than blanket certification of procedural justice as a durable solution to all problems of institutional cooperation. The difference between observed respectful treatment and a later behavioral effect remains central to determining the practical significance of the research. Mazerolle et al. (2013)
Nagin and Telep's critical reviews argue that the causal evidence linking procedural fairness or legitimacy to actual legal compliance is substantially weaker than often suggested by influential correlational models. They distinguish fair-treatment interventions, attitudes toward authorities, intentions to comply, and objective or self-reported behavior. The methodological problem is not that fair procedures lack value; it is that an evaluative benefit cannot substitute for direct evidence of a predicted behavioral consequence. Longitudinal confounding, reciprocal relationships, unclear timing, and the difficulty of observing offending all complicate the proposed chain. An evidence-based synthesis should therefore present procedural justice as an established domain of perceived institutional fairness and a promising but incompletely demonstrated causal route to durable compliance. Nagin & Telep (2017) Nagin & Telep (2020)
Legal Obligation, Sanctions, and Voluntary Cooperation
Law creates formally specified duties, while psychology asks what individuals believe those duties require and how such beliefs interact with incentives. A lawful instruction can be complied with because a person anticipates a sanction, recognizes a moral obligation, accepts an institution's jurisdiction, or wishes to avoid interpersonal conflict. These processes often produce the same immediate action, yet they make different predictions about behavior when monitoring is removed. Legitimacy models argue that recognized rightful authority can support voluntary regulation under relatively low levels of surveillance. An adequate test must distinguish actual enforcement exposure, subjective detection risk, moral approval, perceived procedural fairness, and commitment to the specific legal rule. Otherwise, compliance cannot be confidently attributed to the legitimacy of the institution. Tyler (2006) Nagin & Telep (2017)
The correspondence between formal legality and experienced legitimacy is imperfect. Citizens may perceive a legally proper action as arbitrary or disrespectful, and may trust an official who acts outside an institution's authorized procedures. That distinction is particularly important in forensic and legal psychology, where procedural rights and professional duties define the relevant lawful domain. Psychological surveys can explain how persons evaluate authority, but cannot themselves determine the legal validity of a decision or the normative soundness of a rule. Interpreting public attitudes alongside actual procedural records, legal standards, and independent behavioral outcomes can reveal which parts of the institutional encounter generate trust or resistance. Hough et al. (2010) Tankebe (2013)
Voluntary cooperation includes supplying information, reporting harm, participating in legitimate procedures, and maintaining rule-following when sanctions are unlikely. These are related but distinct outcomes. A police-community intervention might increase willingness to cooperate in a survey while leaving measured offending unchanged; alternatively, it may improve a specific form of reported cooperation without broadly transforming institutional trust. Careful design therefore specifies a primary behavior, relevant opportunity to perform it, realistic comparison condition, and period of observation. The strongest evaluations would combine independent administrative indicators with confidential reports and account for changes in enforcement intensity that alter what gets recorded. This methodological discipline supports actionable institutional research without treating favorable opinion as an adequate proxy for every form of conduct. Mazerolle et al. (2013) Nagin & Telep (2020)
Professional Authority in Health Care and Other Expert Systems
Professional authority is grounded partly in specialized training, licensing, institutional accountability, and the practical need to coordinate complex work. A nurse receiving a physician's order, an aviation crew member responding to a captain, or a laboratory worker following a supervisor may possess relevant expertise of their own while occupying a formally subordinate role. The central psychological problem concerns how recipients balance deference to recognized competence with independent judgment about safety and procedural validity. Hofling and colleagues' historical study of nurse-physician interactions investigated compliance with an unexpected medication instruction in a hospital setting, illustrating how role expectations could shape immediate action. Contemporary interpretation must account for changes in professional rules, workplace safeguards, and institutional culture, rather than transferring a mid-twentieth-century result directly to current clinical practice. Hofling et al. (1966)
Professional obedience is rarely a simple choice between accepting and rejecting all instructions. Orders vary in urgency, clarity, scope, and the consequences of delay. Competence may be domain-specific: an expert in one task need not be entitled to decide every moral or operational issue. An employee can accept routine supervisory direction while requesting verification of an unusual prescription, unsafe procedure, or unauthorized request. Evaluating such conduct requires data about institutional protocols, training, available consultation, and the seriousness of potential harm. In studies involving medical decisions, researchers should distinguish adherence to accepted safety procedures, inappropriate override of qualified judgment, and the measured quality of care. A theory that counts every instance of refusal as undesirable would misclassify some professionally appropriate behavior. Morrison (2014) Colquitt (2001)
The central evidence limitation is historical and institutional transfer. Laboratory and early field studies identify mechanisms of role deference, yet clinical safety now involves multiple professional checks, documentation requirements, and contextual differences that may change the meaning of compliance. Research should use ethically approved scenarios, anonymized audit records, simulation, and organizational process measures to test how authority claims interact with practice standards. Assessment should include both response to an order and downstream verification activity, because a brief hesitation could protect a patient even when a final response is recorded as compliance. The resulting evidence would distinguish the psychological recognition of legitimate expertise from the mistaken inference that hierarchy alone makes every proposed action appropriate. Hofling et al. (1966) Edmondson (1999)
Organizational Justice and the Experience of Supervisory Power
Supervisory authority operates through evaluations, promotions, workload allocation, disciplinary procedures, and informal recognition. Organizational justice theory differentiates the fairness of distributions, the procedures used to reach them, the respect shown during their implementation, and the accuracy of accompanying explanations. Colquitt's measurement work established distinct organizational-justice dimensions, allowing investigators to evaluate which features of institutional treatment relate to employee attitudes and behavior. This differentiation is essential to the study of authority because an employee can approve of a supervisor's professional competence while regarding a particular decision process as unfair. The institution's capacity to command and its perceived rightfulness are therefore partly shaped through recurring ordinary acts of administration. Colquitt (2001)
The meta-analysis by Colquitt and colleagues synthesized organizational justice research across outcomes such as job satisfaction, organizational commitment, citizenship behavior, and evaluation of authority. A meta-analytic association provides information about the average covariation in studies meeting the review's criteria; it does not establish that altering one justice component will cause the same improvement in every organization. Shared response styles, institutional selection, unmeasured employment conditions, and reverse causality can contribute to observed effects. Nevertheless, the findings indicate that how organizations exercise authority is associated with attitudes and conduct relevant to coordination. The practical question is which organizational practices preserve substantive decision quality while making procedures clear and open to challenge. Colquitt et al. (2001)
Authority can be formally legitimate yet psychologically inaccessible. Employees may know that policies provide complaint channels while expecting retaliation or reputational damage if they use them. A fair-process intervention should therefore measure effective access to voice, perceived protection from reprisal, timing of decision explanation, and subsequent institutional response. An organization that invites input but systematically ignores it may generate different experiences from one that incorporates criticism into decisions. To distinguish symbolic consultation from effective participation, researchers can compare policy text with recorded interactions and follow-up behavior. The experimental manipulation should be described narrowly, and effects on trust, actual reporting, error correction, and staff retention evaluated separately. Morrison (2014) Edmondson (1999)
Psychological Safety, Employee Voice, and Institutional Learning
Psychological safety refers to shared perceptions that a work team permits interpersonal risks such as asking questions, acknowledging errors, and expressing concerns. Edmondson's study of work teams linked team psychological safety with learning behavior, offering a mechanism through which authoritative organizations can coordinate work while retaining the capacity for correction. The construct is relational and local: a team can be supportive of challenge even within an institution characterized by rigid formal authority, while another team may suppress objection despite permissive written rules. Evidence from team research suggests that the accessibility of voice changes what information reaches decision makers. These findings connect obedience research to the maintenance of institutional accuracy over time. Edmondson (1999)
Employee voice is purposeful expression of ideas, warnings, or concerns intended to improve an organization. Silence is not a direct index of agreement: individuals may withhold a concern because it seems futile, dangerous, socially costly, or outside their responsibilities. Morrison's review distinguishes motives for withholding and speaking up, and identifies supervisory openness, organizational climate, and interpersonal relations as relevant research domains. When studying authority, measuring observed silence without asking what employees knew and what opportunities existed for speech is insufficient. The researcher must also separate an employee's willingness to challenge a decision from the decision maker's subsequent response. A spoken warning may be ignored, implemented, or punished, each producing different future expectations about the institution. Morrison (2014)
Voice and obedience should be studied as a sequence rather than competing personality categories. A professional may challenge an instruction, receive clarification, and then comply; another may voice an objection and refuse because the revised instruction still violates a safety rule. These trajectories share early critical engagement but differ in the substantive demand and institutional response. Research designs that retain only an endpoint such as 'complied' erase the mechanisms distinguishing them. A more complete observational protocol records the availability of consultation, timing and content of warnings, procedural review, superior response, and objective outcome. This approach aligns legitimacy research with institutional learning, where the ability to correct errors may be as important as immediate instruction-following. Edmondson (1999) Morrison (2014)
Moral Conviction, Dissent, and Principled Refusal
Moral conviction can create limits on deference to authority when people regard a demanded action as violating a central principle. Skitka and collaborators examined moral and religious convictions as constraints on legitimacy's influence, emphasizing that perceived rightful power may lose predictive force when the decision concerns an intensely held moral commitment. This research reveals a boundary of simple legitimacy models: generalized respect for an institution does not guarantee willingness to follow every ruling. The boundary is not automatically equivalent to irrational defiance. The individual's judgment may derive from a stable value, a competing professional duty, or a perception that authority has exceeded its role. Empirical research must measure the relevant conviction before the specific compliance decision to reduce retrospective rationalization. Skitka et al. (2009)
Packer's normative-conflict model treats dissent as a possible expression of group membership and concern for collective standards. A member who identifies strongly with an institution may object precisely because a current practice violates values that the institution publicly claims to uphold. This framework generates different predictions from a theory that equates disagreement with low attachment. The character of dissent depends on whether criticism is aimed at an organization's foundational norms, current leadership, or a particular harmful instruction. Studies can distinguish identification, perceived normative conflict, anticipated sanctions, and observable voice to test which mechanism is active. Such distinctions are indispensable when evaluating whistleblowing, professional protest, or refusal to carry out an ethically suspect directive. Packer (2008)
Principled refusal is also an institutional design question. Workers and citizens need to understand which requests they are authorized to reject, what review paths exist, and whether disagreement can be expressed without disproportionate harm. Empirical research can compare clear appeal pathways, protected reporting arrangements, and situations where leaders invite scrutiny against settings where challenge is punished. Yet such interventions should be evaluated by both their immediate effect on speaking up and their impact on error correction, procedural fairness, and legitimate coordination. The relevant criterion is not maximum dissent or maximum obedience; it is an empirically demonstrated pattern of reasoned decision making under properly specified role duties and constraints. This framing places moral conviction and organizational learning within a common analysis of institutional authority. Morrison (2014) Skitka et al. (2009) Edmondson (1999)
Status, Coercive Capacity, and Institutional Hierarchy
Formal authority frequently intersects with unequal status, control of resources, and capacity to impose sanctions. Keltner and colleagues' approach-inhibition theory relates power to patterns of attention, motivation, and social behavior, providing hypotheses about how control over others can shape leaders' and subordinates' responses. This perspective is useful for distinguishing an institution's jurisdiction from the psychological impact of unequal dependence. A directive may be complied with because the speaker appears authorized, because disobedience threatens employment, or because the recipient anticipates unfavorable evaluation. These pathways require separate measurements. Institutional badges and titles may be visible cues, but their effect depends on the organizational context, established practices, and the perceived consequences of refusal. Keltner et al. (2003) Tyler (2006)
Hierarchy changes what information people volunteer and what errors remain visible. When subordinates expect punitive responses, supervisors may receive artificially favorable reports. Conversely, a leader who invites questions may still exercise decisive authority after relevant evidence has been reviewed. Such patterns complicate any attempt to classify institutions simply as hierarchical or egalitarian. The research problem is the interaction between role inequality, procedural accessibility, and accountability. Studies should measure actual sanctions, reported fear, observed feedback exchanges, and the availability of independent review rather than inferring coercion from position alone. The psychological-safety and employee-voice traditions offer empirical outcomes suited to such designs. Edmondson (1999) Morrison (2014)
There is also a temporal dimension. A single coercive encounter can produce immediate compliance while damaging trust; a consistently fair process may strengthen durable expectations of institutional reliability without changing behavior in every case. An institution can lose authority through misconduct even when its legal mandate persists. Organizational justice and political legitimacy offer parallel research questions about these trajectories, but their evidence bases often concern different settings and units of analysis. Future longitudinal work should examine repeated interactions, changes in perceived legitimacy, actual opportunities to challenge, and later responses to comparable orders. The inference of a stable authority effect requires a sequence of observations and attention to competing environmental explanations. Colquitt et al. (2001) Nagin & Telep (2020)
Authoritarian Dispositions and System Justification
Individual differences in deference may interact with institutional context, but a personality-scale score cannot substitute for observed behavior. Research on system justification examines how individuals may rationalize existing arrangements, including unequal or contested institutions. Jost, Banaji, and Nosek synthesize a tradition of evidence on motivation to defend the social status quo, placing legitimacy judgments within broader processes of social cognition. An individual's endorsement of established institutions may make some commands seem more justified; however, the relationship depends on what the individual believes about the institution and its role. The same disposition can plausibly support obedience to a favored authority and resistance to an authority perceived as violating the established order. Researchers should therefore state the target institution, political context, and predicted outcome rather than invoking a generalized trait as a complete causal explanation. Jost et al. (2004)
Social hierarchy also affects information access and the consequences of dissent. Power-related motivational models propose changes in attention and approach behavior for those who control resources, while subordinates may become especially attentive to threat and anticipated sanction. Such predictions concern statistical tendencies under specified conditions rather than an automatic transformation of every leader or follower. Situational obedience experiments show that changing relationships and task demands can affect the continuation rate even when broad background characteristics remain constant. A comprehensive account should compare contextual manipulations with measured individual differences and estimate interactions, rather than forcing a choice between personality and situation. A strong design tests whether legitimacy perceptions predict conduct after accounting for sanction threats, prior institution-related beliefs, and assignment to different authority conditions. Keltner et al. (2003) Haslam et al. (2014) Jost et al. (2004)
The distinction between ideological alignment and institutional legitimacy is essential. Approval of existing arrangements may reflect group affiliation, expected stability, moral values, perceived effectiveness, or fear of disruption. Each route implies different reactions to a new rule. A person supporting institutional order may challenge an official who acts outside established procedures; another person may tolerate perceived unfairness in order to preserve a valued system. Normative-conflict theory and moral-conviction research offer mechanisms for such qualified responses. Future studies should use independent behavioral tasks, competing authority claims, and longitudinal measures to determine whether self-reported status-quo approval predicts specific actions rather than only related attitudes. Jost et al. (2004) Packer (2008) Skitka et al. (2009)
Group Membership, Competing Loyalties, and Institutional Identity
Institutions represent groups as well as offices. A person receiving instructions may identify as a worker, researcher, professional, citizen, patient, or member of a moral community. These identities organize expectations about whose interests matter and what actions are appropriate. Social-identification accounts of Milgram propose that endorsement of the scientific enterprise helped render a troubling instruction meaningful to some participants. Group identification also creates pathways for criticism: a member can oppose a directive because it conflicts with values associated with the same institution. Therefore, institutional membership should be treated as an empirical moderator of how authority claims are interpreted, not as a direct measure of obedience. Haslam & Reicher (2012) Packer (2008)
Competing loyalties are most visible when the consequences of a directive fall on someone outside the immediate hierarchy. The learner in the Milgram paradigm was a person to whom participants could feel direct responsibility, while the experimenter embodied the requirements of the stated research task. Professional practice offers comparable structural conflicts when a supervisor's demand competes with a duty to a client, patient, subordinate, or public. The social-identity hypothesis predicts that emphasizing the welfare of the affected individual, or a shared value of care, may change how the instruction is construed. Yet the strength of these influences must be demonstrated with specific outcome measures. The historical obedience conditions and immersive adaptations provide partial evidence for this relational interpretation but do not identify a universal mediator. Haslam et al. (2014) Haslam et al. (2015)
Multilevel analysis is needed because identity operates in relationships and communities. An employee's willingness to challenge authority may depend on team norms even after controlling for personal conviction. A citizen's interpretation of police action may depend on prior experiences within a community and on perceived institutional treatment of fellow members. In experiments, researchers can vary the official's stated collective purpose and the availability of supportive peers; in longitudinal field studies, they can examine how identification changes after repeated encounters. Neither approach should assume that stronger identification always promotes compliance. Depending on the contested directive and its perceived relation to shared values, attachment can organize disciplined cooperation, principled dissent, or disengagement. Packer (2008) Sunshine & Tyler (2003) Morrison (2014)
Cultural Context, Institutional History, and Generalizability
Authority and legitimacy are interpreted through legal systems, social norms, historical experiences, and professional traditions. An institution regarded as trustworthy in one community may be associated with coercion or exclusion in another. The same visible instruction can consequently carry different assumptions about rightful jurisdiction, likely sanctions, and acceptable forms of challenge. Contemporary studies of policing and organizations often rely on populations situated within particular countries and institutional arrangements. Their quantitative associations should not be extrapolated across political regimes, workplaces, or professional fields without testing whether the measured constructs have comparable meanings. The relevant scientific question is transportability: which relations recur under altered procedures, populations, and histories, and which depend on localized social expectations. Jackson et al. (2012) Tankebe (2013)
Cross-context measurement requires more than translating questionnaire items. The term obligation may evoke respect, legal duty, loyalty, fear, or social conformity in different settings. Measures of procedural fairness may also differ in their salience where citizens expect opportunities for voice versus where such opportunities are rare. Researchers should test measurement equivalence, examine the meaning of items through qualitative interviews, and compare outcomes against observed institutional practices. These checks are especially important when evaluating whether a legitimacy scale measures the same latent construct across communities. It is also necessary to distinguish perception from institutional performance: individuals may share an evaluative belief while receiving objectively different treatment, and comparable treatment may generate different experiences. Tankebe (2013) Hough et al. (2010)
Institutional history adds a longitudinal dimension to legitimacy. Public confidence after repeated misconduct or reform may change gradually, with immediate encounters interpreted through accumulated collective experience. A one-time vignette can isolate a specific fairness cue but cannot represent all the histories that determine credibility in ordinary life. Repeated surveys, policy change analyses, and natural experiments can address longer time scales if confounding reforms and selection into contact are measured. Research should also document the circumstances under which people have genuine opportunities to refuse, appeal, or exit. Without that contextual information, low reported dissent may represent acceptance, constrained alternatives, or lack of access to a recognized complaint process. Nagin & Telep (2020) Tyler (2006)
Algorithmic Authority and Human–Artificial Institutional Relations
Artificial systems are increasingly considered for recommendation, triage, allocation, and decision support within institutions. Their psychological relevance to authority arises when humans attribute expertise or procedural credibility to algorithmic outputs, or when an institution delegates a decision stage to an automated process. An artificial recommendation can be authoritative in the user's eyes because it appears accurate, arrives through an official interface, or is treated by the organization as the default. These are hypotheses about human attribution and institutional design, not evidence that an artificial system possesses legal authority or moral responsibility by virtue of generating an instruction. The conceptual distinction between recognized jurisdiction and informational credibility remains applicable when authority cues are mediated by software. Tyler (2006) Colquitt (2001)
The Artificial Era provides an additional research domain: the relation between human judgment and institutional commands delivered through nonbiological systems. The questions are empirically precise. Does presenting identical advice as generated by a trained human expert or an artificial decision system change perceived legitimacy? Does a stated opportunity to challenge an algorithmic recommendation increase procedural fairness? Does disclosure of model uncertainty alter behavioral reliance? How do users allocate responsibility when a supervisor frames a machine output as binding? To answer these questions, a study must manipulate source attribution or decision rules while holding relevant content constant, record actual reliance or override, and separately measure trust, duty, and perceived fairness. Claims about artificial agency should be operationalized rather than inferred from human anthropomorphic language. Tankebe (2013) Nagin & Telep (2017)
Institutional accountability remains a key dependent variable. If a human worker cannot meaningfully contest a model output, the procedure may be experienced as closed even when the output is technically accurate. If the user can inspect reasons, submit counterevidence, and obtain qualified human review, the decision environment presents different opportunities for justified compliance and refusal. These propositions can be examined through preregistered randomized vignette experiments and field evaluations with transparent outcome measures. A complete research design would assess downstream correction of errors, uneven effects across affected groups, and whether perceived fairness corresponds to substantive institutional safeguards. This domain connects established legitimacy theory with Psychology of Human–Artificial Relations while maintaining the evidential status of these extensions as proposed research rather than established psychological laws. Colquitt et al. (2001) Morrison (2014)
Measurement: From Attitudes to Observed Behavior
The study of institutional authority uses several outcome types: perceived legitimacy, generalized duty to obey, support for an institution, willingness to cooperate, simulated instruction-following, administrative compliance, and the actual exercise of dissent. Each outcome answers a different question. A scale measuring generalized obligation cannot demonstrate the rate at which individuals will follow an ethically troubling order, just as a laboratory decision to continue a scripted task cannot demonstrate support for the legal system. Measurement should therefore follow a hierarchy of explicit construct definitions, validated items or observed indicators, and prespecified behavioral settings. Colquitt's scale work and Tankebe's analysis of police legitimacy illustrate why multidimensional constructs require structural tests and carefully selected indicators. Colquitt (2001) Tankebe (2013)
Psychometric assessment should include reliability, factor structure, stability over relevant intervals, and equivalence across institutions and populations. Wording can confound perceived authority with desired conduct; an item asking whether an institution ought to be obeyed may partly encode the outcome later treated as explained by legitimacy. Researchers should measure perceived fairness separately from obligation, and the sanction climate separately from both. Survey designs can then test whether distinct dimensions have independent predictive power while allowing for shared response tendencies and retrospective reconstruction of experiences. Where possible, observations of specific encounters and independent records should complement self-reports. The analytic model should identify whether the proposed effect occurs within persons following an encounter or merely between individuals who differ in their general trust. Jackson et al. (2012) Nagin & Telep (2020)
Behavioral measures present further difficulties. A high rate of compliance may reflect lack of meaningful opportunity to dissent, while low compliance may reflect unclear instructions rather than resistance to authority. Coding should preserve whether the demand was understood, whether alternatives were offered, whether consequences were salient, and whether the recipient received clarification. For ethically adapted obedience studies, the exact stopping point and severity of the apparent task must be specified. For field studies, institutional records may reflect detection and reporting practices rather than the underlying rate of misconduct. These distinctions prevent a measurement artifact from being mistaken for a psychological mechanism. Burger (2009) Nagin & Telep (2017)
Experimental, Longitudinal, and Field Research Designs
Randomized experiments can isolate local effects of authority cues, stated institutional purposes, opportunities for voice, or the apparent source of a directive. In low-risk vignettes, participants can evaluate identical instructions delivered under different role labels; in simulations, researchers may record whether participants request clarification or choose an available safe alternative. Ethical protocols should ensure realistic informed consent, minimize distress, permit withdrawal, and avoid recreating the harmful pressures of historical obedience paradigms. These adaptations change the endpoint, so an ethical trial's findings must be reported as the observed simulated decision rather than a literal replication of historical maximal-shock obedience. Registered hypotheses and manipulation checks improve interpretability, but do not eliminate differences between simulated and consequential choices. Burger (2009) Grzyb et al. (2025)
Longitudinal studies can examine whether experiences of fair or arbitrary treatment predict subsequent legitimacy, behavior, and institutional engagement. Repeated measurement helps order events and evaluate change within the same person, but temporal precedence alone does not remove time-varying confounding. Individuals who enter frequent contact with officials may differ from those who do not; institutions may alter treatment in response to prior behavior; and external events can shift trust simultaneously. Stronger designs might exploit randomized contact procedures, exogenous policy changes, or well-characterized natural experiments while assessing whether the underlying assumptions are plausible. Administrative records should be interpreted alongside the reporting system that generates them. Sunshine & Tyler (2003) Nagin & Telep (2020)
Field trials can test policy-relevant outcomes under ordinary institutional pressures. The procedural-justice policing trials demonstrate the possibility of manipulating encounter practices, while organizational field designs can evaluate protected reporting or structured consultation. The chief methodological requirement is to predefine the relevant unit of randomization, intervention fidelity, primary outcome, and follow-up interval. Cluster assignment may be needed where teams or offices share norms; spillovers should be anticipated rather than treated as independent individual responses. Investigators should preserve uncertainty, report attrition and measurement limitations, and examine whether improvements in immediate perceptions translate into observable conduct. Evidence on one part of the hypothesized process should not be silently generalized to the entire institutional causal chain. Mazerolle et al. (2013) Edmondson (1999) Nagin & Telep (2017)
Critical Evidence Appraisal and Conflicting Interpretations
The strongest direct evidence for immediate instruction-following comes from tightly specified behavioral paradigms and their experimental variations. Milgram's original observation is historically decisive within its condition, while the synthesis of variants demonstrates meaningful sensitivity to relations among experimenter, participant, and learner. Modern ethically modified experiments offer further behavioral evidence with changed endpoints and protections. These studies do not support a single cross-context obedience rate, and the historical paradigms contain interpretive ambiguity concerning motives. Negotiation, moral conflict, scientific identification, and perceived pressure can all be relevant without being equally well established as mediators in every case. An adequate appraisal separates reliable observation of action from theories about the actor's reasons. Milgram (1963) Haslam et al. (2014) Burger (2009) Gibson (2013)
Research on legitimacy and procedural justice provides substantial evidence that experiences of treatment are linked to perceptions of rightful authority and reported cooperation. Construct validation and randomized encounter interventions improve confidence in some of these links. The causal connection from fair treatment to durable voluntary legal compliance is more uncertain. Systematic reviews and critiques highlight variation in interventions, measurement, and follow-up, along with the risk of extrapolating from attitude outcomes to offending behavior. These qualifications matter for both scientific precision and institutional policy: a procedure can be worth adopting for reasons of dignity and fairness even when a particular downstream compliance effect has not yet been established. Colquitt et al. (2001) Mazerolle et al. (2013) Nagin & Telep (2017) Nagin & Telep (2020)
Some theories concentrate on coercive command; others on collective identification or normative endorsement. Evidence should distinguish the observable predictions of each. Coercion models predict sensitivity to threatened costs and reduced alternatives; identification accounts predict sensitivity to the perceived collective mission; legitimacy accounts predict cooperation conditional on judgments of justified institutional authority; moral-conviction accounts predict boundary conditions for deference. These approaches can also interact. Their explanatory contribution is best established by studies that manipulate one factor while measuring alternatives and a relevant behavioral outcome, then replicate across institutional settings. A theoretically compelling interpretation of an archival event is a source of hypotheses, but its evidential status differs from a successful prospective test of a precisely specified mechanism. Haslam & Reicher (2012) Skitka et al. (2009) Tyler (2006)
Unresolved Problems and a Cumulative Research Agenda
Several problems deserve priority. The first concerns the internal composition of legitimacy: its relation to felt obligation, moral alignment, fairness, legality, and effectiveness varies by institution and measurement strategy. The second concerns the distinction between self-reported intention and actual behavior, especially when researchers propose that respectful treatment reduces later violations. The third concerns competing motives underlying a compliant action. To resolve these issues, research programs should use explicit construct diagrams, independent indicators of institutional practice, and preregistered competing models, with data sufficient to examine whether mechanisms change over time and across social groups. Longitudinal and experimental results should be synthesized without substituting statistical significance for effect magnitude or practical relevance. Tankebe (2013) Jackson et al. (2012) Nagin & Telep (2020)
A fourth problem is the construction of safe but meaningful obedience research. Historical paradigms elicit strong behavior under conditions that would be difficult to reproduce ethically, whereas modern simulations often reduce the consequences of action. New paradigms can investigate verified professional instructions, conflicting informational advice, moral concern, and structured opportunities for refusal without exposing participants to severe distress. Such work should compare stated intentions with safe but consequential choices when possible. It should also record pauses, questions, procedural appeals, and negotiated modifications rather than coding only yes or no. The extensive variation observed in historical studies implies that interaction sequences contain explanatory information not captured by a single terminal measure. Haslam et al. (2014) Gibson (2013) Grzyb et al. (2025)
A fifth problem is causal transport across institutions and technological settings. Findings on police contact cannot automatically establish effects in hospitals, workplaces, courts, or human–artificial decision systems. Each setting differs in formal jurisdiction, stakes, prior trust, communication, review rights, and available alternatives. A cumulative research infrastructure would compare common measures while allowing institution-specific constructs, record policy context and participant characteristics, and test whether effect heterogeneity has substantive explanations. Data sharing should protect affected persons and preserve meaningful institutional privacy. Especially for artificial decision support, observed reliance, actual override, perceived legitimacy, and responsibility allocation should be measured separately. This program would turn the psychology of authority into a comparative science of institutional decision relations. Hofling et al. (1966) Morrison (2014) Tyler (2006)
Conclusion
Authority, obedience, and institutional legitimacy constitute an integrated psychological research field centered on how people interpret and act upon demands made through recognized roles and institutions. The evidence establishes that obedience is sensitive to the structure of the situation: the presence and consistency of the authority figure, the social meaning of the task, relations with affected persons, and opportunities for disagreement influence outcomes in classic experimental settings. The famous maximal-shock result belongs to a particular experiment, while the larger set of historical conditions shows substantial variability. Ethically modified contemporary research extends the subject through new designs with distinct endpoints. These findings support a relational, context-sensitive account of instruction-following that explains the same outward behavior through potentially different social and motivational processes. Milgram (1963) Haslam et al. (2014) Burger (2009)
Institutional legitimacy emerges through assessments of rightful jurisdiction, fairness, social purpose, and professional conduct. The procedural-justice tradition demonstrates that respectful and impartial treatment matters for citizens' and employees' perceptions of institutions. Yet perceptions, declarations of obligation, cooperation, and objectively measured compliance occupy different evidential levels. The causal path from a fair encounter to sustained voluntary compliance remains an important unresolved question. Institutions additionally depend on forms of reasoned dissent that expose errors, enforce professional responsibilities, and preserve accountability. Psychological safety, moral conviction, and organizational voice show that legitimate coordination can include the capacity to challenge particular instructions. The appropriate scientific endpoint is a precise understanding of when, why, and with what consequences authority is accepted, revised, or refused. Nagin & Telep (2017) Edmondson (1999) Skitka et al. (2009)
Future research should combine experimentally identified mechanisms, longitudinal institutional evidence, validated multidimensional measures, and cross-context comparisons. It should examine both the acts of officials and the experiences of those subject to their decisions, including individuals whose ability to refuse is materially constrained. As artificial systems mediate recommendations and institutional choices, the same framework can investigate how humans attribute epistemic and normative authority to nonbiological systems and whether procedures support meaningful accountability. These extensions require explicit operational definitions and new empirical testing. The enduring contribution of the research tradition is a set of questions and methods for analyzing human decision-making within institutions: what gives an instruction force, what makes that force appear rightful, what preserves independent judgment, and how institutional practices shape the conditions of responsible action. Tyler (2006) Morrison (2014) Tankebe (2013)
References
- Blass, T. (1999). The Milgram paradigm after 35 years: Some things we now know about obedience to authority. Journal of Applied Social Psychology, 29(5), 955–978. https://doi.org/10.1111/j.1559-1816.1999.tb00134.x
- Burger, J. M. (2009). Replicating Milgram: Would people still obey today? American Psychologist, 64(1), 1–11. https://doi.org/10.1037/a0010932
- Colquitt, J. A. (2001). On the dimensionality of organizational justice: A construct validation of a measure. Journal of Applied Psychology, 86(3), 386–400. https://doi.org/10.1037/0021-9010.86.3.386
- Colquitt, J. A., Conlon, D. E., Wesson, M. J., Porter, C. O. L. H., & Ng, K. Y. (2001). Justice at the millennium: A meta-analytic review of 25 years of organizational justice research. Journal of Applied Psychology, 86(3), 425–445. https://doi.org/10.1037/0021-9010.86.3.425
- Edmondson, A. (1999). Psychological safety and learning behavior in work teams. Administrative Science Quarterly, 44(2), 350–383. https://doi.org/10.2307/2666999
- Gibson, S. (2013). Milgram's obedience experiments: A rhetorical analysis. British Journal of Social Psychology, 52(2), 290–309. https://doi.org/10.1111/j.2044-8309.2011.02070.x
- Grzyb, T., Dolinski, D., Sudoł-Malisz, M., Kulik, G., & Mielczarek, Ł. (2025). Doubting the power of prestige: Obedience to authority beyond institutional and research justifications. Scientific Reports, 15, Article 24911. https://doi.org/10.1038/s41598-025-10331-4
- Haslam, N., Loughnan, S., & Perry, G. (2014). Meta-Milgram: An empirical synthesis of the obedience experiments. PLOS ONE, 9(4), e93927. https://doi.org/10.1371/journal.pone.0093927
- Haslam, S. A., & Reicher, S. D. (2012). Contesting the “nature” of conformity: What Milgram and Zimbardo's studies really show. PLOS Biology, 10(11), e1001426. https://doi.org/10.1371/journal.pbio.1001426
- Haslam, S. A., Reicher, S. D., & Millard, K. (2015). Shock treatment: Using immersive digital realism to restage and re-examine Milgram's “obedience to authority” research. PLOS ONE, 10(3), e109015. https://doi.org/10.1371/journal.pone.0109015
- Hofling, C. K., Brotzman, E., Dalrymple, S., Graves, N., & Pierce, C. M. (1966). An experimental study in nurse-physician relationships. The Journal of Nervous and Mental Disease, 143(2), 171–180. https://doi.org/10.1097/00005053-196608000-00008
- Hough, M., Jackson, J., Bradford, B., Myhill, A., & Quinton, P. (2010). Procedural justice, trust, and institutional legitimacy. Policing: A Journal of Policy and Practice, 4(3), 203–210. https://doi.org/10.1093/police/paq027
- Jackson, J., Bradford, B., Hough, M., Myhill, A., Quinton, P., & Tyler, T. R. (2012). Why do people comply with the law? Legitimacy and the influence of legal institutions. British Journal of Criminology, 52(6), 1051–1071. https://doi.org/10.1093/bjc/azs032
- Jost, J. T., Banaji, M. R., & Nosek, B. A. (2004). A decade of system justification theory: Accumulated evidence of conscious and unconscious bolstering of the status quo. Political Psychology, 25(6), 881–919. https://doi.org/10.1111/j.1467-9221.2004.00402.x
- Keltner, D., Gruenfeld, D. H., & Anderson, C. (2003). Power, approach, and inhibition. Psychological Review, 110(2), 265–284. https://doi.org/10.1037/0033-295X.110.2.265
- Mazerolle, L., Antrobus, E., Bennett, S., & Tyler, T. R. (2013). Shaping citizen perceptions of police legitimacy: A randomized field trial of procedural justice. Criminology, 51(1), 33–63. https://doi.org/10.1111/j.1745-9125.2012.00289.x
- Mazerolle, L., Bennett, S., Davis, J., Sargeant, E., & Manning, M. (2013). Legitimacy in policing: A systematic review. Campbell Systematic Reviews, 9(1), 1–147. https://doi.org/10.4073/csr.2013.1
- Milgram, S. (1963). Behavioral study of obedience. The Journal of Abnormal and Social Psychology, 67(4), 371–378. https://doi.org/10.1037/h0040525
- Morrison, E. W. (2014). Employee voice and silence. Annual Review of Organizational Psychology and Organizational Behavior, 1, 173–197. https://doi.org/10.1146/annurev-orgpsych-031413-091328
- Nagin, D. S., & Telep, C. W. (2017). Procedural justice and legal compliance. Annual Review of Law and Social Science, 13, 5–28. https://doi.org/10.1146/annurev-lawsocsci-110316-113310
- Nagin, D. S., & Telep, C. W. (2020). Procedural justice and legal compliance: A revisionist perspective. Criminology & Public Policy, 19(3), 761–786. https://doi.org/10.1111/1745-9133.12499
- Packer, D. J. (2008). On being both with us and against us: A normative conflict model of dissent in social groups. Personality and Social Psychology Review, 12(1), 50–72. https://doi.org/10.1177/1088868307309606
- Skitka, L. J., Bauman, C. W., & Lytle, B. L. (2009). Limits on legitimacy: Moral and religious convictions as constraints on deference to authority. Journal of Personality and Social Psychology, 97(4), 567–578. https://doi.org/10.1037/a0015998
- Sunshine, J., & Tyler, T. R. (2003). The role of procedural justice and legitimacy in shaping public support for policing. Law & Society Review, 37(3), 513–548. https://doi.org/10.1111/1540-5893.3703002
- Tankebe, J. (2013). Viewing things differently: The dimensions of public perceptions of police legitimacy. Criminology, 51(1), 103–135. https://doi.org/10.1111/j.1745-9125.2012.00291.x
- Tyler, T. R. (2003). Procedural justice, legitimacy, and the effective rule of law. Crime and Justice, 30, 283–357. https://doi.org/10.1086/652233
- Tyler, T. R. (2006). Psychological perspectives on legitimacy and legitimation. Annual Review of Psychology, 57, 375–400. https://doi.org/10.1146/annurev.psych.57.102904.190038