A Multidimensional Conceptualization of Employee Safety Voice

Abstract

Previous conceptualizations of safety voice have described this construct as fundamentally unidimensional, compared to several other multidimensional models that have been outlined for the more general concept of employee voice. Drawing from previous studies in occupational safety, previous safety voice conceptualizations (challenge vs. defense of safety systems), dynamic safety capability, and wise proactivity, we outline an extended conceptualization of safety voice. Adopting a functionalist approach, we identified and described promotive, preventive, prohibitive, and hostile safety voice. For each type, we discuss the definition, its function, its implications, and examples. This multidimensional conceptualization provides a conceptual structure that supports more accurate measurement, intervention design, and discussions surrounding safety-related communication. It also clarifies when and how distinct forms of safety voice emerge, identifies levers for system and leadership design, and outlines actionable implications and testable propositions for advancing prevention, learning, and research.

1. Introduction

Safety at work is not only a matter of regulatory compliance or physical protection—it represents a psychological, organizational, and societal imperative intimately linked to human flourishing, reflective awareness, and long-term sustainable development [1,2,3,4]. From this perspective, workplace safety must be understood as a dynamic and regenerative process grounded in psychological resources, positive relational environments, and inclusive systems that foster well-being, empowerment, and active participation [1,2,3,4]. In 2023, over 2.6 million nonfatal occupational injuries were recorded in the US, with approximately one-third resulting in time away from work to allow for recovery [5]. These statistics highlight the urgency of adopting sustainable, human-centered approaches to workplace safety that account for the psychological mechanisms behind proactive safety behavior and provide employees with the necessary resources, voice, and environmental conditions not only to remain safe but also to thrive and contribute to collective well-being [6].

A lapse in safety experienced by employees in safety-oriented settings may constitute a salient and objectionable condition, prompting the motivation to speak up [7,8,9]. However, for safety-related communication to contribute to sustainable and healthy workplace environments [3], employees must feel psychologically safe and confident in their ability to raise concerns constructively and without fear of retaliation [1]. Although no precise statistics exist on how frequently employees raise safety concerns, research consistently indicates that silence is the more typical response when potential issues are perceived [6,10,11,12,13,14]. Employees are likely to frequently notice potential risks or opportunities for improvement; however, in the absence of an empowering culture rooted in sustainability values, such observations often remain unvoiced or unaddressed [1,15].

The term voice has a long history in organizational psychology, and other related disciplines (e.g., sociology, economics, political science, history, industrial relations, and labor studies [16]), For this reason, it is essential to begin with a clear definition of employee voice, commonly understood as the discretionary communication of ideas, suggestions, or concerns related to work issues, directed toward individuals with the authority or capacity to act, and intended to promote positive change [13]. Voice is often contrasted with silence, which involves the deliberate withholding of potentially valuable input or the choice to refrain from expressing one’s thoughts [17]. Safety researchers have adapted the construct into the safety realm, defining it as a behavior that seeks “to improve safety by identifying current limitations and possibilities for positive change” [18].

In one of the early studies, safety voice was presented as part of the broader framework of safety citizenship behaviors [19]. However, subsequent research has largely retained a unidimensional perspective [20]. In this model, safety voice is theorized as one dimension of safety citizenship behavior, alongside helping (supporting colleagues in fulfilling their safety responsibilities), stewardship (acting to protect other people from risky situations at work), and whistleblowing (reporting violations and deviations from safety standards and safety policies; [21]). However, research on employee voice has advanced several competing frameworks, including multidimensional models (see, e.g., [22,23]). We then propose a multidimensional conceptual model of safety voice to more accurately reflect the psychological and organizational dynamics underlying this critical behavior.

Several reviews, meta-analyses, and articles have advanced our understanding of antecedents, consequences, and correlates of voice behavior, contributing to a theoretically coherent network of psychological constructs [10,12,14,20,24]. These factors are commonly classified into individual, contextual, and motivational domains that can either encourage or inhibit voice behavior. The question of why an employee decides to speak up or remain silent has no easy solution, but it is nonetheless important for organizational scholars to try to answer [25]. Evidence suggests that employees are often reluctant to engage in voice, particularly in situations where speaking up may be perceived negatively [15]. Understanding and facilitating safety voice behavior, therefore, represents a key challenge for building psychologically sustainable, healthy, and development-oriented organizations. These are environments where employees not only feel physically safe but are also psychologically empowered to co-create safer, healthier systems that contribute to individual well-being and collective sustainability [1,3].

This is a conceptual, theory-integrating article that introduces a multidimensional framework of employee safety voice. Whereas extensive research has examined employee voice, prior studies have often remained confined within single disciplinary lenses and have rarely integrated insights across fields. In safety-critical contexts, this fragmentation is particularly limiting: safety voice has seldom been distinguished from safety silence, and the specific antecedents, boundary conditions, and outcomes of safety voice remain under-investigated. These gaps limit the development of coherent theoretical models and hinder both empirical investigation and practical interventions in safety management. Recent work (e.g., [20,26]) has acknowledged this issue and identified the need for multidimensional conceptualizations of safety-relevant employee communication. Rather than proposing yet another narrow model, this paper builds on the existing foundation by elaborating on one of the few multidimensional approaches already recognized in recent reviews [25]. Accordingly, this article has four main objectives:

• Reconcile divergent disciplinary definitions of employee voice and specify scope conditions for safety-relevant expression.
• Differentiate safety voice from safety silence, with particular attention to channels and boundary conditions in safety-critical work.
• Develop a multidimensional framework with testable propositions linking antecedents, channels, and outcomes of safety voice.
• Derive implications for research design, measurement, and organizational practice to inform both future empirical inquiry and applied safety management.

Guided by these objectives, the paper addresses two overarching research questions:

• How can safety voice be conceptualized as a multidimensional construct that meaningfully differentiates between distinct forms of safety-relevant expression?
• How do different forms of safety voice relate to safety silence, and what channels, antecedents, and boundary conditions shape their occurrence in safety-critical contexts?

By pursuing these objectives and questions, the paper contributes an integrative, theory-driven perspective that advances the conceptualization of safety-relevant employee voice and provides a foundation for systematic empirical investigation and practical application.

2. Approach to Conceptual Development

This article’s main goal is conceptual and theoretical development, not exhaustive cataloging or statistical aggregation. We therefore adopted a concept-centric, purposive synthesis, identifying and integrating foundational and safety-relevant sources on voice and silence to clarify construct boundaries, mechanisms, channels, and boundary conditions. We searched major scholarly databases and checked reference lists to ensure inclusion of seminal work and safety-specific contributions. We identified relevant articles in the extant literature by conducting a targeted search on Web of Science using the keyword “employee safety voice.” To ensure comprehensive coverage, we complemented this database search by scanning the reference sections of influential articles on safety voice. This approach allowed us to include papers that may not have been indexed in the database but are widely cited and foundational to the topic. By combining systematic keyword searching with backward citation tracking, we aimed to capture both core and peripheral contributions to the safety voice literature.

From each source, we extracted (a) definitions and scope conditions; (b) level(s) of analysis; (c) proposed mechanisms; (d) boundary conditions; and (e) implications for safety management. We organized these insights to: (1) reconcile disciplinary understandings of voice; (2) delineate safety voice from related constructs (including silence); and (3) derive falsifiable propositions specifying when and how different forms of safety voice emerge and with what outcomes. This approach emphasizes conceptual clarity and integrative theorizing, consistent with established guidance on theory contribution in management and safety science (e.g., [26,27,28]).

3. Conceptual Background: Voice, Silence, and Safety-Critical Contexts

3.1. Safety Voice and Proactive Behavior

To situate our work within the broader literature, it is important to acknowledge that employee safety voice has been conceptualized differently across disciplines, each emphasizing distinct mechanisms, purposes, and channels of employee expression. In Hirschman’s exit–voice–loyalty framework [9], voice functions as a remedial mechanism used to arrest decline when exit is costly or loyalty is high, emphasizing organizational or system recovery and feedback channels. In organizational behavior and industrial–organizational psychology, voice is typically defined as discretionary, change-oriented communication aimed at improving work processes and directed to those with authority, often differentiated into promotive and prohibitive forms [7,22,29]. In industrial and employment relations, voice denotes employees’ influence over decisions through direct mechanisms (e.g., consultation, suggestion schemes) and collective mechanisms (e.g., grievances, union representation, works councils), grounded in a pluralist view of the employment relationship [30,31]. Our safety voice framework is deliberately situated at the behavioral level, capturing how employees communicate safety-relevant concerns, while acknowledging that these behaviors are often channeled through institutional arrangements emphasized in economics and industrial/employment relations. This positioning clarifies the scope of the construct, reduces potential slippage, and helps explain when different safety voice types are likely to emerge through informal, formal, or collective channels.

Consistent with the broader employee voice literature, scholars have begun to transpose the employee voice construct into the safety realm [20,32]. This transposition is not merely conceptual but reflects a push to foster sustainable, health-oriented work environments where employees are empowered to contribute to collective well-being through active participation. When effectively supported, safety voice has the potential to improve work-related conditions, procedures, and environments, thereby facilitating a psychosocially sustainable approach to occupational safety, one that safeguards individual and public health while supporting long-term organizational success [1].

3.2. Defining the Construct

Having employees openly express their concerns about safety is especially relevant in high-risk industries, as it can prevent injuries and encourage organizational learning [21,33,34,35]. When examined through the lens of sustainability, these practices are understood not only as protective mechanisms but also as transformative processes. They address immediate harm-reduction needs while simultaneously promoting the long-term psychological and social sustainability of the work environment [1].

Striving for a more formal definition and operationalizable construct, Conchie et al. [18], p. 105 defined employee safety voice as a behavior that seeks “to improve safety by identifying current limitations and possibilities for positive change.” Other authors observed that safety voice behaviors can include raising safety concerns with a supervisor or union representative [35], speaking before a safety committee [36], reporting dangerous working conditions to public health authorities [37], and educating newcomers and coworkers on safer work techniques/procedures [38]. A common element across these examples is that the recipient of the message is perceived by the employee as having the authority or capacity to enact change [31,39]. It can be inferred from the previous examples that employee safety voice (a) is a form of communication driven by the desire to address unsafe working conditions with potentially harmful consequences, (b) flows through both formal and informal channels of communication, and (c) may be directed toward several targets such as supervisors, government agencies, and unions [31]. However, for safety voice to be effective, it has to overcome significant barriers such as fear of retaliation [40] and the normalization of safety violations [41]. For this reason, it is relevant for organizations to actively promote and sustain a culture where safety voice is not only encouraged but also normalized [18].

3.3. Similarities and Differences Between Employee Voice and Safety Voice

Although employee safety voice shares conceptual ground with employee voice [42], proactive problem-solving [43], and safety participation [44], it represents a distinct construct with unique implications. As Tucker et al. [31] argued, safety voice behaviors are often protected or even prescribed under legal frameworks in many Western countries—unlike other extra-role behaviors, such as general employee voice, which lack formal legal protection. While some scholars have observed potential overlap between safety voice and proactive behavior in high-risk organizational settings [45], Tucker et al. [31] emphasized a key distinction: proactive problem-solving is typically directed toward management-sanctioned goals, while safety voice may involve dissenting views (i.e., prohibitive safety voice) that challenge unsafe norms or power structures [46].

Safety voice conceptualizations draw from research on the broader constructs of employee voice and silence [31], often conceptually overlapping with the former [13] because both constructs encompass extra-role behaviors aimed at addressing perceived issues in the workplace [47]. Specifically, LePine and Van Dyne [42], p. 326 defined voice behavior as “constructive, change-oriented communication intended to improve the situation.” It is worth noting that the potential recipient can vary considerably, including supervisors, team members, union representatives, external stakeholders, or public health authorities. This versatility underscores its relevance for both micro- and macro-level sustainability interventions, bringing individual empowerment with institutional accountability. Table 1 shows four conceptual differences between employee voice and safety voice. Our goal is not to elevate safety voice above general employee voice, but to emphasize that safety voice, especially when analyzed through its multidimensional expressions, represents a crucial behavioral component of sustainable organizational development.

Table 1. Differences between Employee Voice and Safety Voice.

Characteristics	Employee Voice	Safety Voice
Consequences	Less severe	Severe
Conceptualization	Multidimensional	Often unidimensional
Mechanism	Informal	Formal and informal
Practical relevance	Variable	High

First, the consequences of safety voice, or lack thereof, can be significantly more severe (e.g., personal harm and organizational accidents) than those typically associated with general employee voice (e.g., lower performance, turnover). Moreover, research on employee voice often addresses issues that are comparatively less urgent and carry a lower degree of moral obligation [20]. Second, although the conceptualization of employee voice has been widely recognized as multidimensional, with more than a dozen typologies identified [23], the same multidimensional approach has not been systematically applied to safety voice. Incomplete or overly narrow definitions risk overlooking critical expressions of concern or dissent that are essential to building adaptive and resilient safety systems. Third, safety voice differs from general employee voice in terms of its institutional embeddedness. While employee voice is often informal and confined within organizational boundaries, safety voice is more likely to be supported by formal channels and external authorities, such as safety committees, unions, and public health bodies [31,48]. These structural mechanisms reflect a systemic orientation aligned with sustainability principles—particularly in fostering accountability, stakeholder engagement, and the protection of vulnerable populations. In contrast, employee voice is frequently rooted in interpersonal dynamics or discretionary behavior without formal protection or recognition.

Lastly, the practical relevance of safety voice is arguably higher in high-reliability organizations, because of the potential to avoid accidents and improve safety systems, compared to organizations where safety is not the main concern. In these contexts, safety voice can be considered not only a behavioral resource but also a proactive, protective function that enhances the psychological and physical sustainability of the work environment [45]. Thus, safety voice contributes to sustainable development at both micro and macro levels, supporting safe work environments and broader public safety outcomes.

3.4. Advantages and Functions for Organizations

Safety voice serves as a key behavioral driver of sustainable organizations. It promotes open communication, proactive risk management, and stakeholder engagement, all essential components of long-term organizational health. Previous literature has identified several advantages and functions of safety voice. For instance, employee engagement in safety voice has been found to predict higher rates of near-miss reporting and lower incidences of lost-time injuries [49]. Rather than signaling increased risk, this proactive reporting reflects a psychologically safe climate, one that supports open communication and early hazard identification, both critical for safety systems. This pattern, although counterintuitive, is in line with previous research showing that organizations with higher near-miss event reporting also had a stronger safety culture [50] because near-miss reporting may be understood as a positive and proactive way of communication aimed at anticipating safety issues [51]. Safety voice is associated with lower overall accident rates [32]. These findings are consistent with the notion that development within organizations depends not only on environmental or economic outcomes but also on socially and psychologically sustainable practices, such as enabling all members to contribute to safety outcomes. In this view, positive communication exchange among stakeholders becomes a foundation for resilient, inclusive systems, facilitating both accident prevention and the identification of effective safety programs.

A growing body of research provides empirical support for the idea that safety voice and silence are not uniform behaviors but are strongly shaped by organizational conditions. For example, Hofmann et al. [19] and Tucker et al. [31] show that employees are more likely to raise safety concerns when they perceive leadership openness and responsive reporting systems. Conversely, Morrison [13] demonstrates that anticipated social or career costs often suppress voice, leading to silence even when risks are recognized. Recent reviews further emphasize that silence is not simply the absence of voice but a differentiated construct with distinct antecedents and effects [20,25]. In our extended model, these insights are operationalized by distinguishing four forms of safety voice—promotive, preventive, prohibitive, and hostile—and five forms of silence—ineffectual, diffident, relational, underreporting, and job-based. Each form is influenced by specific contextual factors: just-culture climates and legal protections enable preventive and prohibitive voice; participatory structures and resource availability sustain promotive voice; while relational climates and authority gradients often condition the expression of prohibitive and hostile voice. By mapping how these conditions differentially affect both voice and silence, the framework provides a more fine-grained explanation of when, why, and how employees engage, or refrain from engaging, in safety communication. This sets the stage for Section 4, where we elaborate on the extended model of safety voice and silence in safety-critical contexts.

4. The Multidimensional Safety Voice Framework

Other scholars have extended Liang et al.’s [22] framework into a four-type model of voice [23], acknowledging the need to account for destructive and defensive expressions of voice [29,52,53]. Considering the breadth of behaviors that can be meaningfully classified as safety voice behavior, the promotive-prohibitive framework [22] appears to be insufficient when applied to occupational health. Other scholars have expanded the Liang et al. framework [22] into a model with four types of voice [23], noting that more than a dozen voice behaviors have been investigated and there was the need to include voice behaviors that are destructive in nature [22,23,29,52,53,54]. Given the wide range of behaviors that may reasonably be classified as safety voice, the original promotive–prohibitive dichotomy appears insufficient in this domain. Multidimensional conceptualizations are better equipped to capture the diverse expressions of safety-related communication and are thus more appropriate for informing sustainable development strategies.

4.1. Overarching Framework

Table 2 shows how the conceptualization of safety voice has evolved through time, along with some limitations that will be addressed. Despite notable advancements in the safety voice literature, several critical gaps persist. The vast majority of empirical work has relied on measurement instruments that primarily capture socially desirable aspects of voice behavior. To measure the construct, Hofmann et al. [19] adapted a scale that measures employee voice [54] resulting in items that reflect the underlying assumption that voice is conceptualized as a promotive behavior intended to challenge the organizational status quo. In their seminal contribution, Tucker et al. [31] conceptualized safety voice in a more definite fashion, yet it remained broadly consistent with earlier definitions [19,54], framing safety voice as communication intended to address and improve safety procedures or environments perceived as hazardous. Therefore, the overall focus is still heavily promotive, as the goal is to challenge unsafe working conditions and promote positive change. To measure safety voice, Tucker and colleagues [31,55] developed two similar scales, but the content of some items has a preventive focus (i.e., “remind coworkers to take precautions” and “I inform the union boss when I notice a potential driving hazard”). Although both promotive and preventive behaviors aim to ensure safe operations, they differ in their underlying mechanisms: promotive behaviors typically involve suggesting improvements or changes to existing safety norms, whereas preventive behaviors are oriented toward averting future risks or failures.

Table 2. A Summary of the Evolution of Safety Voice Conceptualizations.

Study	Conceptualization	Key Attributes	Limitations
[19]	Initial theoretical definition as a safety citizenship behavior. Defined as a change-oriented extra-role behavior.	Altered a general employee voice scale to adapt it to safety voice.	No formal definition.
[31]	“Communication motivated toward changing perceived unsafe working conditions” (p. 320) “Can flow through formal and informal channels” (p. 320) “Can be directed towards numerous targets.” (p. 320)	Safety voice is conceptualized as a challenge to the current safety procedures.	Unidimensional construct. Only partially took into account that safety voice may also aim at protecting the system.
[18]	“Behaviors that seek to improve safety by identifying current limitations and possibilities for positive change.” (p. 105)	Safety voice is conceptualized as a multidimensional construct. Preventive focus is added to the definition.	The construct was treated as unidimensional due to multicollinearity.
[56]	A behavioral effort aimed at enhancing workplace safety, following mental simulation and anticipation of risk.	Safety voice as a part of a dual goal-regulation process: goal generation and goal striving	As the behavioral manifestation of goal-striving, the construct is still treated as unidimensional.

While this distinction is not explicitly addressed in their work, it is considered critical for advancing the theoretical and practical understanding of safety voice within the context of sustainable organizational development. Moreover, there are a few items in the 2008 scale that seem to have a prohibitive focus—i.e., they achieve the goal of improving safety by preventing or stopping someone from acting unsafely. Items such as “I tell my colleague who is doing something unsafe to stop” and “I report to my boss if my colleagues break any safety rule” appear to follow a prohibitive direction, rather than promotive or preventive.

The idea that safety voice may not be a completely unified construct was further developed by Conchie et al. [18], who measured safety voice through three different scales from Hofmann et al. [19]: namely, voice, whistleblowing, and initiating change. Although these dimensions were ultimately treated as a unidimensional construct in their analysis—primarily due to multicollinearity concerns—it is important to recognize that the selected scales capture distinct aspects of safety voice. As previously discussed, voices and initiating change scales primarily measure promotive behaviors, whereas the whistleblowing scale captures prohibitive behaviors, such as reporting unsafe acts. Furthermore, Conchie [57] again used a collapsed, single measure of safety voice, including both promotive and protective themes. Building upon this foundation, and in alignment with sustainable development goals emphasizing equity, participation, and psychological well-being, a more differentiated approach is proposed. Specifically, it is argued that safety voice should be explicitly categorized into four distinct dimensions. This proposition is supported by parallel literature on employee voice (e.g., [23]), employee silence (e.g., [53,58]), and, to a lesser extent, safety silence [59], all of which highlight the theoretical and practical value of adopting a multidimensional perspective over a unidimensional one. In a now seminal contribution on voice and silence, Van Dyne et al. [53] argued that the different reasons that could motivate employees to speak up or remain silent indicate the need for multidimensional accounts of such constructs. The same argument was taken up by safety scholars to justify the only multidimensional account of safety silence in the literature so far [59]. Although we agree with that line of reasoning, we offer two more reasons why safety voice is best conceptualized multidimensionally. First, categorizing safety voice into four distinct forms offers conceptual clarity that is crucial for effective, sustainable organizational interventions. In fact, the literature on voice has been historically plagued by definitional uncertainty [23]. Without clear definitions, the literature suffers from construct contamination (too broad) or deficiency (too narrow), thus hindering empirical progress and limiting practical relevance. In turn, empirical investigations have faced difficulties in operationalizing such behaviors. By splitting the construct into clearly defined dimensions, scholars and practitioners will be better equipped to recognize and classify safety voices. Second, focusing only on safety voice as a unitary construct does not take into account how organizations create conditions that enhance or inhibit such behaviors [60]. In fact, discretionary individual behavior is intrinsically embedded in the organizational environment in which it takes place, which in turn limits the choice to voice or remain silent. Additionally, research in this area has often underemphasized the role of external regulatory frameworks that either support or constrain opportunities for voice. By offering a multidimensional account of safety voices, we hope to enable scholars to empirically investigate how these and other structural determinants affect different types of safety voices.

4.2. Theoretical Background for an Extended Model

We aim to expand the current conceptualization of safety voice into a multidimensional account that better reflects the ethical, proactive, and systematic dimensions of sustainable workplaces. Building upon the Liang et al. [22] model of employee voice (i.e., the promotive-prohibitive framework), the framework is extended to incorporate two additional forms of safety voice that are particularly salient in today’s high-risk organizational contexts. Promotive voice reflects an optimistic, developmental stance where employees envision and pursue a safer, ideal future state through the expression of new ideas and suggestions for improving the organization. On the other hand, prohibitive voice taps into employees’ concerns about behavioral infringements, omissions, operational shortcuts, deviations from the organizational standards, and other arbitrary practices in the workplace that may pose risks to the organization. Thus, the primary focus of prohibitive safety voice is on halting or preventing harm before it occurs. Although this model has been the backbone of both employee voice and safety voice research in the last ten years, two main limitations should be acknowledged: first, it does not acknowledge hostile voices that may emerge within an organization; and second, it only makes a fairly simple distinction between voices that encourage change and those that hinder it. These limitations will be addressed in the following sections, along with the theoretical justification for expanding the model. Accordingly, the proposed framework introduces four distinct types of safety voice: it retains promotive voice, splits and reframes the traditional prohibitive voice, and adds a new category—hostile voice—to better reflect the complexity and diversity of safety-related communication in contemporary workplaces.

Research on employee voice has long recognized that not all types of voice are inherently positive [52], a perspective more recently integrated into a multidimensional framework by Maynes and Podsakoff [23]. In fact, these scholars noted that employees may engage in harsh criticisms of current organizational procedures and also engage in petty grievances. For the purposes of this contribution, we will focus on the former. Such voice behaviors may be viewed not as dysfunctional but as critical indicators of systemic strain, misalignment with employee values, or unsustainable work conditions. This perspective resonates with a recurring tension in the literature: while organizational scholars have emphasized the importance of reducing excessive demands on employees to protect well-being, economic perspectives frequently highlight the pressures organizations face in responding to intensifying market competition. Thus, organizations are likely to react by requiring higher performance from employees and, at the same time, lowering labor costs. The current worldwide economic conditions —characterized by the continuous erosion of profit margins and the increasing pressure on managers to get more from employees for less [61]—suggest that certain types of safety voice (e.g., promotive) are likely to decline, while safety voices that consist of complaints, criticisms, and demands are likely to rise because of higher job stress and work intensification [62]. As such, safety voice theory must account for the broader macroeconomic forces and structural conditions that shape employee behavior, including the emergence of organizational hostility. By acknowledging hostile safety voice as a product of unsustainable conditions, we aim to foster greater systemic awareness and organizational accountability. In the next section, we introduce the hostile safety voice to address this issue.

The promotive-prohibitive model [22] assumes that employee voice can only challenge the current organizational status quo, somewhat in contrast with older literature on the same topic. As will be discussed more in depth below, types of voice can easily be categorized into two broad clusters: safety voices that aim to challenge the status quo versus safety voices that aim to defend or reaffirm the status quo. In this regard, prohibitive voice has been meaningfully categorized in both clusters: as noted above, Liang et al. [22] theorized it as a positive, somewhat defensive type of voice, while Maynes and Podsakoff [23] categorized it as a negative, destructive type of voice. In the context of safety research, we believe that it is best to conceptualize two different voices, which we term prohibitive (conceptually closer to Maynes and Podsakoff’s dimension [23]) and preventive (conceptually closer to Liang et al.’s dimension [22]). We will discuss in detail why this difference is worthy of consideration later in this contribution.

As shown in Table 3, the proposed framework is organized around two core dimensions along which safety voice may vary. Our first dimension distinguishes safety voices that preserve the safety system and procedures (i.e., sustain and ultimately defend the current system) from those that challenge them (i.e., try to correct or question the current system). This dimension is rooted in both theoretical and empirical work: Hirschman [9] described voice as the verbalization of critical opinions; Gorden [52] stated that voice can either “correct” or “reaffirm” the status quo; and Van Dyne et al. [53] distinguished between voice that expresses support and those which challenges, while the Liang et al. [22] and the Maynes and Podsakoff [23] models also recognize this dimension. The self-regulatory focus framework [63] lends further theoretical support to this dimension, as it posits that safety behaviors may be driven by either a promotion focus (i.e., allowing the organization to achieve positive outcomes) or a prevention focus (i.e., allowing the organization to reduce potential risks). The second dimension of the framework introduces a temporal perspective, distinguishing between safety voice behaviors that are present-oriented—focused on immediate risks and short-term outcomes—and those that are future-oriented, emphasizing long-term consequences and strategic foresight. This dimension is rooted in the work on dynamic safety capability [64], wise proactivity [65,66], and the distinction between reactive and proactive change found in Nadler and Tushman’s work [67]. At the individual level, a future-oriented perspective enables greater proactivity and strategic anticipation of risks. Building on the concept of wise proactivity, it is posited that future-oriented safety voice reflects a balanced consideration of strategic, social-relational, and self-regulatory factors.

Table 3. Proposed Framework for Safety Voices.

Dimensions	Present-Oriented	Future-Oriented
Challenge	Hostile Safety Voice	Promotive Safety Voice
	Definition: Intentionally harsh or critical comments about safety practices or policies.	Definition: Sharing ideas or suggestions aimed at improving safety measures.
	Function: To voice dissatisfaction with current safety systems.	Function: To propose forward-looking improvements to enhance safety
	Implications for others: Often viewed as negative and emotionally charged.	Implications for others: Typically seen as constructive and beneficial.
	Intended target: Workplace, organization	Intended target: Task, team, workplace, organization
	Judicial status: Not protected or expected.	Judicial status: Not protected or required.
	Examples: Criticizing company safety protocols; making aggressive remarks about safety decisions	Examples: Recommending new safety tools; suggesting updates to safety procedures.
Defense	Prohibitive Safety Voice	Preventive Safety Voice
	Definition: Reporting deliberate violations or unsafe behaviors.	Definition: Communicating concerns about situational or unintentional risks (e.g., fatigue, equipment failure).
	Function: To call out human errors that threaten safety.	Function: To alert others to potential dangers and avoid future incidents.
	Implications for others: Intention may be positive, but it may be received as critical.	Implications for others: Generally seen as responsible and proactive.
	Intended target: Team	Intended target: Workplace, organization
	Judicial status: Protected and expected	Judicial status: Mostly protected and expected
	Examples: Stopping a coworker from unsafe conduct; reporting repeated breaches of rules.	Examples: Notifying supervisors of near-misses; reminding others to follow safety protocols.

Note. The x-axis represents the time perspective dimension (i.e., present-oriented vs. future-oriented), while the y-axis represents the disposition toward the safety systems (i.e., challenge vs. defense). Each quadrant represents a type of safety voice that is described in the main text, starting from the top, right-hand quadrant and proceeding clockwise.

Thus, a wise employee would consider “the interests of the broader system, the interests of others in the system, and one’s own interests” [65], p. 224. For instance, immediately stopping a colleague who is acting unsafely is contextually sound because it is in the organization’s interest to avoid accidents and injuries; it is other-focused because that behavior prevents a possible hazard and protects the colleague’s health; and it is personally sound because the employee would feel relieved that they avoided a possible hazard.

At the organizational level, attention to future change has been linked to the organization’s capacity to establish dynamic routines [68]. Planning and forward-looking activities require a deliberate orientation toward future conditions, allowing organizations to build the capabilities necessary to respond effectively to emerging and unforeseen challenges. As Griffin et al. [64] noted, future orientation is a core feature of dynamic safety capability, defined as a set of future-oriented activities that allow a given organization to make changes to its operations in order to improve safety. Thus, future orientation is not aimed at reducing uncertainty, but rather focuses on anticipating and adapting to future, uncertain possibilities. This is extremely relevant for safety-critical organizations, which have historically been perceived as reactive and compliance-driven rather than future-oriented. This concept highlights the need to identify and recognize the importance of developing and fostering a stronger future orientation around safety-related activities in high-reliability organizations, which should creatively adapt to dynamic environments. We positioned the present-oriented perspective on the other end of this dimension, drawing from the idea of reactive change [68]. This form of change typically occurs in response to adverse events —such as accidents—and is characterized by heightened vigilance and a concentrated focus on immediate conditions, often at the expense of long-term strategic planning. Additionally, short-term and present-oriented strategies, such as compliance procedures that ensure stability and predictability, come at the expense of building a capacity to react to change in the future [69].

Positioning the defense/challenge dimension on an axis and the time perspective dimension on the opposite one creates a two-by-two matrix that divides safety voices into four quadrants (see Table 3). We labeled the behaviors in the (a) challenge/future-oriented quadrant promotive safety voice; (b) defense/future-oriented quadrant preventive safety voice; (c) defense/present-oriented quadrant prohibitive safety voice; and (d) challenge/present-oriented quadrant hostile safety voice. Consistent with Maynes and Podsakoff’s argument [23], the proposed categories are not conceptualized as subdimensions of a single, overarching construct, but rather as distinct forms of safety voice behavior. Each category is a separate construct, and the behaviors within each safety voice type reflect the specific construct.

In sum, the expanded four-type model offers a more nuanced framework. It aligns with the psychology of sustainability framework by: (a) supporting inclusive participation; (b) promoting well-being through ethical organizational climates; (c) addressing systemic stressors and structural barriers to prevent avoidable harm; and (d) encouraging resilient, future-oriented practices in safety systems. This reconceptualization supports a shift from reactive to sustainable safety cultures, where every form of voice (constructive or critical) is a potential resource for sustainable development and long-term individual and organizational well-being.

4.3. Positioning the Extended Model

Having reviewed the evolution of safety voice conceptualizations, we now position our extended model in relation to prior frameworks to clarify how it builds upon, diverges from, and improves earlier approaches. Much prior research has conceptualized safety voice in unidimensional terms, for example, as the general act of speaking up about safety concerns [19,31]. Later work introduced a two-form distinction between promotive (future-oriented suggestions) and prohibitive (warnings about risks) voice [22]. While influential, this binary perspective remains limited, as it collapses diverse forms of safety-relevant communication into broad categories. Other multidimensional frameworks in organizational behavior (e.g., [23]), have expanded the scope of employee voice, but these models are not safety-specific and therefore insufficiently address the distinctive dynamics of safety-critical communications.

Our framework extends this foundation in three key ways. First, it adds two forms—preventive and hostile voice—that are rarely theorized explicitly but frequently observed in practice. Preventive voice captures future-oriented defense behaviors such as raising weak signals or reporting near misses before they escalate [49]. Hostile voice, in contrast, represents present- or future-oriented challenges expressed in a confrontational or critical tone. Both add analytical precision beyond the promotive–prohibitive distinction and align with recent calls for multidimensional conceptualizations of safety-relevant expression [20,25]. Second, our model integrates safety voice with safety silence. Prior studies often assumed “voice = absence of silence,” but our framework differentiates five silence types (ineffectual, diffident, relational, underreporting, job-based) and maps how each selectively suppresses or shapes specific voice forms. This creates a more realistic account of when and why employees refrain from speaking up, and how silence interacts with voice in safety-critical contexts. Third, the extended model enhances practical applicability by linking forms of voice to organizational design levers. Promotive voice aligns with participatory forums and idea-to-pilot pathways; preventive voice with low-friction near-miss systems; prohibitive voice with red-flag escalation protocols; and hostile voice with protected escalation channels under a just-culture climate [59]. By specifying outcomes, indicators, and enabling conditions, the model transforms abstract categories into actionable guidance for safety management (see Table 3). In sum, our contribution is not another descriptive typology but a theory-driven framework that advances conceptual clarity, integrates safety silence, and specifies practical levers for safety-critical organizations. This positions the extended model as an improvement over earlier approaches, while also offering a foundation for empirical testing through the propositions developed in Section 5.

4.4. Definition of Each Type of Safety Voice

4.4.1. Promotive Safety Voice

Promotive safety voice represents the future-oriented/change quadrant in our model. To be promotive means being constructive and proposing new ways to improve safety procedures, departing from what is currently being done (i.e., challenging the status quo, also found in [22,23,52,54]). We define promotive safety voice as the voluntary communication of new ideas or suggestions to functionally improve safety procedures. This includes discussing and proposing new ways to improve safety procedures and proposing how to fix problems in new ways. The overall function of promotive safety voice is to highlight how the organization can achieve its goals in the future by changing what it is currently doing. It follows that the target is twofold: it seeks to challenge both the formal safety systems established by the organization and the informal social dynamics maintained by its members. In addition, the voicer may be seen more positively [70], while others have argued that because promotive safety voice challenges the status quo, it may damage interpersonal relationships [71]. Promotive safety voice is thought to be an extra-role behavior, and empirical results support this claim [19]. As such, it is neither mandated nor protected under most employment laws, and employees typically perceive it as falling outside their formal job responsibilities. In other words, it is completely up to the employee to decide whether to speak up and give suggestions to meaningfully improve safety procedures or remain silent.

4.4.2. Preventive Safety Voice

Preventive safety voice represents the future-oriented/defense quadrant. To prevent means to defend and sustain the current system. We define preventive safety voice as the voluntary communication of safety concerns regarding contingent and immediate work situations due to fatigue, mechanical failures, and other events not linked to human intentionality. This conceptualization draws upon Reason’s framework [72,73] of organizational accidents, which posits that human error is not the root cause of failure but rather a symptom that demands explanation. The true causes often lie in error-producing conditions embedded within the work environment. From this perspective, meaningful progress in safety arises from understanding how human behavior is intrinsically linked to task demands and operational settings, and from influencing those conditions to reduce risk. Preventive safety voice includes behaviors such as informing the supervisor about safety hazards and reminding colleagues to take precautions. As discussed above, the main function is to highlight potential organizational accidents linked to error-producing systems or cognitive failures. It follows that preventive safety voice, in contrast with promotive safety voice, is task-focused: the underlying principle guiding applied research should be that the concatenation of multiple vulnerabilities exists within the organization long before accidents happen and their interrelation with unsafe acts by human operators. Given its relevance, public health agencies in Europe and elsewhere have taken affirmative steps to set a “basic law” that outlines workers’ and employers’ obligations (e.g., European Council Framework Directive 1989/391). Reporting current unsafe working conditions is a duty of employees, making behaviors labeled as preventive safety voice protected under the law and expected of all workers. In line with a sustainability-informed perspective, preventive safety voice reflects a shared responsibility for maintaining safe, healthy work environments. This raises the question of whether employees perceive preventive safety voice as an extra-role behavior, similar to promotive safety voice as described in voice theory, or instead as an in-role obligation mandated by law. Clarifying this distinction could contribute to the advancement of safety voice conceptualizations, particularly in cross-cultural contexts where labor laws and expectations vary.

4.4.3. Prohibitive Safety Voice

Prohibitive safety voice represents the present-oriented/defense quadrant. To be prohibitive entails stopping others from engaging in unsafe behavior and calling out or reporting violations, making this form of voice applicable to actions such as intervening when a colleague is knowingly acting unsafely or reporting the intentional breach of safety rules and procedures. As shown in Table 3, prohibitive safety voice shares a common defensive function with preventive safety voice—both are oriented toward protecting and preserving the existing safety system, rather than challenging or seeking to reform it. However, they target different organizational hazards. As noted elsewhere [72,73], the overall direction an organization takes is a tradeoff between different goals that have to be met simultaneously. It follows that safety is not inherent in systems; it is rather one of these goals, and production rate, time constraints, dealing with resource shortages, and general workplace pressures may be others. In addition, individual perceptions such as hazard level, management commitment to safety, and perceived barriers (i.e., skepticism about the efficacy of safety measures, cavalier attitude) have been linked to unsafe work behavior [74]. The interplay of some (or all) of the aforementioned constructs may motivate voluntary unsafe behaviors in the workplace, which prohibitive safety voice attempts to tackle. This quadrant tries to formally acknowledge the reality of high-risk organizations, in which careless behaviors prevail in many routine tasks, despite managerial supervision [75,76]. We define it as the voluntary exposure of unsafe working practices due to intentional violations of safety rules/procedures by colleagues. In this sense, it is closely linked to whistleblowing (defined as “the disclosure by organization members of illegitimate practices under the control of their employers,” [77], p. 4), but it is broader, as it includes direct behaviors aimed at stopping the unsafe act. Both constructs share the same legal status; in fact, employees are generally required to report illegal activities in the workplace, but legal protection varies widely (e.g., US labor laws are considerably weaker than the European counterpart).

As stated previously, employees are mandated to report safety hazards and, if it is necessary and they are capable, may directly intervene. In line with research on preventive safety voice, empirical investigations of prohibitive safety voice should focus on the interplay of organizational priorities and goals and intentional/recurrent safety violations. Although prohibitive and preventive safety voices may appear similar and could be superficially construed as a single construct, we argue that maintaining this distinction is valuable for fostering long-term organizational resilience and psychological well-being. This distinction supports not only the technical sustainability of safety systems but also their psychological sustainability by promoting reflective and inclusive engagement with safety concerns [1]. Our rationale is grounded in established safety literature that differentiates between unintentional human errors in the interaction between human operators and technical components of the system (e.g., skills-based errors, rule mistakes, knowledge-based mistakes) and errors due to frontline employees’ procedural violations, which can be further distinguished in routine or situational violations [72,73]. Since effective risk management strategies in any organization depend imperatively on the organization’s capacity to establish a reporting culture, it is paramount to differentiate between the sources of error. Specifically, when organizational learning is the primary goal, the expectation is that a risk report will become a lesson learned, an analysis of the event will highlight the contributing factors, and interventions that aim at reducing the likelihood of recurrence will be implemented [78]. This is closely aligned with our definition of preventive safety voice, which has the function to attract organizational attention to potential sources of accidents associated with unintentional mistakes by human operators (i.e., cognitive failures like slips and lapses, using the correct procedure in the inappropriate situation). Preventive safety voice thus acts as an early-warning mechanism, contributing to a culture of anticipation and improvement, key tenets in sustainable organizational systems. Conversely, systems could be created for which the main goal is accountability; thus, the reporting of unsafe behaviors and procedural violations becomes more central. In such organizations, the focus is on the individual origin of the error, and often they place reports in human resources files that are then used for evaluation of performance. This is closely aligned with our definition of prohibitive safety voice, which aims to stop unsafe behaviors and report errors due to intentional procedural violations, omissions, and shortcuts. These violations can be situational (i.e., an occasional conflict between production and safety instances) or can happen routinely (i.e., a deliberate allocation of time resources that penalize safety standards). In both cases, situational and routine violations are intentional actions. We propose that, whereas the focus of preventive safety voice is to address risks and threats due to unintentional mistakes by individuals in the fulfillment of their work activities, the focus of prohibitive safety voice is to address deliberate violations enacted on an intentional basis.

4.4.4. Hostile Safety Voice

Hostile safety voice represents the present-oriented/change quadrant. Being hostile means being hurtful and extremely critical (i.e., challenging the status quo). We define hostile safety voice as the voluntary expression of opinions and ideas that are derogatory and hurtful with regard to safety procedures. This form of safety voice shares key characteristics with what has been termed destructive voice in prior literature [23,52], as both emphasize intense criticism of current organizational practices. Illustrative behaviors include disparaging remarks about an organization’s safety protocols or making harshly critical statements about existing safety policies, typically without constructive intent. The primary aim of hostile safety voice is to express dissatisfaction with organizational safety procedures. While the underlying concern may be legitimate, the manner in which it is conveyed—often through harsh criticism or disparaging remarks—can elicit negative interpretations due to the negative emotions associated with harsh criticism and disparaging expressions. Similarly to promotive safety voice, hostile safety voice behaviors are neither protected nor expected from an employment law perspective and are perhaps more likely to bring negative consequences to the voicer. These behaviors are commonly enumerated under the label of counterproductive work behavior [79]. These behaviors target mainly the organization as a whole, seen as inefficient or on the wrong path. This conceptualization of safety voice challenges the often-unchecked assumption in organizational literature that employees simply agree and internalize management’s priorities and perspectives about safety in the workplace [45]. This proposition is supported by ethnographic studies, which showed that employees may actively resist, ignore, or reframe top-down safety messages [80]. In line with the psychology of sustainability approach, a hostile voice can function as a signal of psychological misalignment, an expression of frustration when more constructive avenues for change are blocked. Though it may not always be effective, it should not be dismissed outright: it can raise critical awareness and challenges dominant narratives, potentially catalyzing reflective inquiry and structural reconsideration.

In sum, each form of safety voice contributes uniquely to safety and sustainability in organizations. Promotive voice advances transformational sustainability through innovation. Preventive voice ensures development sustainability through risk anticipation. Prohibitive voice reinforces normative and structural sustainability by protecting ethical standards. Hostile voice, while contentious, can provoke reflective sustainability by surfacing tensions and calling for deeper dialogue. By framing these behaviors in terms of sustainability psychology, organizations can better understand the diverse functions of safety voice and cultivate healthy environments in which speaking up is seen not as a liability, but as a vital asset for long-term success.

5. Safety Silence and Safety Voices

Thus far, we provided a theoretical rationale for an extended model and described four types of safety voices. Given that it is arguably unrealistic for employees to engage in all four types of safety voice at any given time, we rely on recent insights from the literature on employee silence to discuss how five known types of safety silence could potentially impact the proposed types of safety voices. After each section, we offer a few conceptual reflections aimed at stimulating future empirical research on the relationship between safety voices and safety silence. Silence is arguably more elusive than its voice counterpart and has generally been neglected by academic researchers [53]. Two main reasons can be provided as to why there has not been research on this organizational construct: first, it seems that silence is conceptualized as a mere absence of voice (and therefore a non-behavior). The absence of a behavior is not particularly noteworthy in organizations, and therefore, it does not attract attention. Second, and somewhat related to the first point, the absence of behavior is inherently more challenging to investigate than behaviors that are overt and measurable. The only recent empirical investigation on safety-related silence identified four motives behind the employees’ decision to remain silent [59], such as the fear of damaging existing interpersonal relationships, an organizational climate that is not supportive of safety voice, the belief that the potential safety hazard is unlikely to occur or it will not affect anyone negatively, and the inhibitive role of certain job characteristics (e.g., high workload, production pressure). Although that contribution fell short of offering a multidimensional conceptualization of safety silence, it nonetheless calls for it. It is reasonable to assume that employees are unlikely to engage in all four types of safety voice simultaneously or consistently; therefore, it is relevant to include a comparable conceptualization of safety silence that takes this into account, along with reasons that could inhibit safety voices and enhance safety silence. Additionally, the variety of reasons for employees to withhold information justifies the benefit of differentiating the types of safety-related silence, rather than combining them into a single construct. Last, we argue that investigating safety-related silence along with safety voice could be extremely useful for understanding the boundary conditions of the latter. We believe that a more in-depth understanding of what variables prevent safety voices from happening and what organizational barriers are in place is a worthy research agenda. Here, we provide a brief overview of five types of silence (i.e., ineffectual, diffident, relational, underreporting, and job-based) and we describe how these could affect the four types of safety voice described in the previous section. Table 4 depicts a schematic synthesis of the proposed forms of safety silence and their relationship with the four types of safety voices.

Table 4. Relationship between Safety Voices and Safety Silence.

Characteristics	Ineffectual Silence	Diffident Silence	Relational Silence	Underreporting	Job-Based Safety Silence
Theoretical Foundations	Expectancy Theory [81]	Conservation of Resources Theory (Hobfoll, 1989) [82]	Perceived Organizational Support [83]	Behavioral Reasoning Theory [84]	Job Design Theories [85]
Empirical Investigations	[17,58,59,86]	[42,58,59]	[15,58,59]	[87,88,89]	[59]
Safety Voices Impacted	Promotive and Hostile safety voices	Promotive safety voice	Promotive, Preventive, Prohibitive, and Hostile safety voices	Preventive and Prohibitive Safety Voices	Promotive, Preventive, and Prohibitive Safety Voices
Examples	Management did not appear interested in hearing about these types of issues; The safety procedures, as they are, do not put anyone at risk.	Not feeling confident enough to speak up; Feeling unsure what to say; feeling that it could lead to a negative perception of self.	Fear of retaliation for speaking up; Speaking up would make it difficult to work with colleagues; do not want to create tension with colleagues	Not reporting an accident, near miss, risk, or otherwise safety-relevant event.	Heavy time pressure at work; heavy workload; Not in the job description

5.1. Ineffectual Safety Silence

Previous literature defines ineffectual safety silence as the refusal to share relevant ideas, opinions, and information to functionally improve safety procedures. In more general terms, this type of silence is coherent with Morrison and Milliken [17] conceptualization of climate of silence, which is characterized, in part, by the shared belief that speaking up will not produce the desired change; Pinder and Harlos [86] acquiescent silence (implying a felt acceptance of organizational circumstances), and ineffectual silence (the belief that speaking up will not lead to positive outcomes; [58]). Drawing from employees’ neglect in organizations [9], this form of safety-related silence could be conceptualized as a form of inaction. This form of safety silence reflects employees who lack hope or expectation that the current status quo can be meaningfully changed or challenged. It aligns with Manapragada and Bruk-Lee’s [59] typology of issue-based and climate-based safety silence.

This form of safety silence may also represent those employees who believe that the safety procedures are already functional and do not need to be overhauled; therefore, they actively refuse to share ideas to improve them and would rather engage in other forms of safety voice that are more defensive of the status quo. The presence of this type of silence is more likely to affect promotive and hostile safety voice because employees believe that changing the current organizational and safety procedures will not be productive. Two established frameworks that may help in clarifying the relationship between this type of silence and safety voices are the expectancy theory [81] and denial of the need for change [90]. The former theory suggests that employee behavior can be seen as a function of whether they expect their behavior to effect change or achieve a predetermined goal. If they believe that their efforts (i.e., engaging in promotive safety voice) cannot effectively improve the work environment in ways that they deem satisfactory, it creates passivity and disengagement [91]. Considering such a scenario, employees have very few motivations to speak up, and therefore, they are likely to engage in safety silence. The latter takes an organizational change perspective: the first milestone for any change is a reliable diagnosis of what are the structural barriers that prevent the organization from reaching a desired goal [92]. This diagnosis will then be the basis for a meaningful change plan and a cogent case to make to the decision-makers. However, employees may resist changing safety procedures for a variety of reasons (e.g., self-interest, misunderstanding, differential assessment, and low tolerance to change; [93]). Notably, employees who perceive organizational change as unnecessary (i.e., have a differential assessment) may try to resist said change by either remaining silent (i.e., inaction) or stepping up in defense of the status quo [94]. Therefore, we would predict that employees who believe that change is not necessary could either engage in safety silence or preventive safety voice. Consequently, employees who believe that change is not necessary in the workplace could choose safety silence because, essentially, they are not motivated to engage in challenging forms of safety voice (i.e., promotive; hostile). However, they could still be motivated toward engaging in other defensive forms (i.e., preventive; prohibitive). Future research should investigate this theoretical proposition to clarify the boundary conditions and the underlying motivations that guide employees toward either behavior.

5.2. Diffident Safety Silence

Previous literature defined diffident silence as the refusal to share ideas, concerns, and opinions due to lack of self-confidence, self-doubt, and/or uncertainties. It is inherently linked to proactive motivation and role-breadth self-efficacy [95], defined as the capability to carry out a broader and more proactive set of tasks beyond their prescribed job duties. Research on this topic is rather scarce, but LePine and Van Dyne [42,54] noted that self-esteem is positively associated with externalizing behaviors, while neuroticism is negatively associated with employee voice. This type of silence is aimed at avoiding internally focused negative outcomes (e.g., embarrassing oneself) and was also identified by Manapragada and Bruk-Lee [59] in their study on safety silence. We suggest that one of the motives behind it may be that employees lack role-breadth self-efficacy. This involves a wide array of on-the-job competencies that employees may not feel they master, for instance, self-direction, interpersonal and group communication skills, coordination of the production line with other departments, and many others; therefore, they may choose to remain silent. We argue that this type of silence may inhibit safety voices that are considered as an extra-role behavior by employees. In fact, in-role behaviors are usually clearly defined for each position and by definition do not involve any degree of proactivity because they are required in the job description, while extra-role behaviors cannot be mandated by the organization, and it is (almost) completely up to employees to enact them. Likely, employees assess their personal resources and the potential to gain or lose resources when faced with a stressful situation (e.g., decide whether to engage in safety voices). Given that engaging in promotive safety voice requires a host of personal resources, employees may feel that (a) they do not currently have those resources, (b) they will lose resources (or there is a threat of loss) if they engage in safety voice, and (c) they will not be able to replenish the resources invested. Consequently, we could expect that they may remain silent either because they lack the proactive drive required for extra-role behaviors or because they feel they anticipate a loss of resources by engaging in such behaviors. However, this does not exclude that employee may still engage in in-role behaviors, such as defensive forms of safety voice. Future empirical research should test this proposition in more detail and elucidate the motivational or structural mechanisms underlying these behaviors.

5.3. Relational Silence

Relational silence has been investigated both as a general and safety-specific construct [15,58,59]. Employees often remain silent out of fear of damaging existing interpersonal relationships in the workplace, for instance, fear of retaliation from management or supervisors, or a desire to avoid conflict with colleagues. Arguably, this kind of safety-related silence poses a threat to all four types of safety voice we identified. Perceived organizational support [83] is a resource that may influence safety behaviors [96,97,98]: theory suggests that employers provide employees with both tangible (e.g., wages and other benefits) and intangible (e.g., caring, respect, support) resources. If employees feel they are being supported by their colleagues and organization at large in the form of positive, supportive experiences with other people in the organization, they are more likely to engage in extra-role behavior [99,100]. Relatedly, social exchange theory [101] may also explain the relationship between relational silence and safety voices. A basic tenet of the theory is reciprocal interdependence: an action by a party (e.g., providing a benefit) triggers an in-kind response by another (reciprocating by providing a benefit). An illustrative example can be found in the 1988 Clapham railway accident [102], where strained social relations and a lack of mutual trust between management and frontline employees contributed to employees engaging in safety silence, which was perceived as an appropriate response to a management team seen as distant and lacking commitment to safety. Thus, given the importance of social relationships in the workplace [85,103,104], we believe that strained social relationships (or the fear of straining them) may negatively affect all types of safety voices.

5.4. Accident Underreporting

We argue that employees may also choose not to report potential safety hazards due to dormant risks or error-producing safety procedures (conceptually similar to issue-based safety silence in [59]), inhibiting preventive safety voice. Likewise, employees may not be willing to report intentional safety violations and accidents [105]. We consider accident underreporting, defined as the employee’s failure to report a workplace accident to their organization [87], as a form of safety silence. Employees who see a safety violation or identify a dormant risk are faced with the question “Do I report, or do I not report?” When they refrain from taking action, they become silent observers, which is an undesirable behavior in organizations striving for long-term learning, ethical accountability, and sustainable human development. Inhibiting preventive and prohibitive safety voices could be a byproduct of organizational strategy. Efforts to increase production rates and manage the organization’s external perception as safe could potentially adversely affect safety voices and foster accident underreporting for the sake of avoiding market share loss due to being labeled as unsafe, public embarrassment, and avoiding regulatory bodies’ scrutiny, which could lead to legal and financial liability [106]. Behavioral reasoning theory [84] suggests that employees evaluate context-specific reasons both for and against engaging in a particular behavior. For instance, when the reasons to remain silent, such as those tied to organizational strategy, outweigh the reasons to speak up, including a personal inclination toward reporting, employees are more likely to withhold their concerns. One common rationale is the perceived incompatibility between production and safety [107], with productivity often prioritized over safety considerations [108]. Moreover, performance evaluation systems emphasizing output over safety may further discourage voice. From a sustainability-informed perspective, these systems reflect short-term, efficiency-driven logics that are incompatible with long-term human and organizational sustainability. Last, employees’ reluctance to report safety risks and error-enabling procedures may also stem from employees’ lack of awareness of the potential issues in the workplace. In turn, this highlights the potential role of safety knowledge and safety training in fostering preventive and prohibitive safety voices, while contextually inhibiting safety silence. Although this theoretical proposition has not been tested yet, it is coherent with the notion that safety knowledge predicts safety behaviors [109]. We expect accident underreporting to affect both preventive and prohibitive safety voices: in fact, those constructs could be conceptualized as existing on a continuum between complete silence (e.g., choosing not to report an accident) and voice (e.g., reporting a colleague’s safety violation). Further empirical research is needed to explore what stands between those extremes: we believe that qualitative research designs offer a powerful tool to examine this question. Between these poles lie a range of context-sensitive strategies, such as informal, restorative responses. For instance, King [110] found that nurses witnessing a violation often preferred private conversations over formal escalation, highlighting the relational, ethical, and contextual dynamics that shape voice. These responses align with sustainability psychology’s emphasis on restorative practices, human-centered decision-making, and dignity-based leadership.

5.5. Job-Based Safety Silence

Job-based safety silence has been described as the choice to remain silent regarding safety issues because of specific job characteristics (e.g., job design, job duties, and job responsibilities) that inhibit employees from raising safety concerns. This silence is particularly relevant to sustainability discourse, as it reveals how the structure of work can either foster or suppress ethical agency, well-being, and participatory change—pillars of sustainable organizational development. This perspective aligns with longstanding insights from classical job design theories, which emphasize that job structure influences performance, employee well-being, and other key organizational outcomes [85]. Employees may not speak up because of several job demands, such as work overload, production pressures to deliver specific targets, or because they believe that speaking up is not part of their job duties [59,111]. In this light, job-based safety silence is not merely an operational challenge, it is a systemic barrier to psychological empowerment and inclusive change. Sustainability-oriented organizations must therefore view job redesign as a key intervention to encourage safety voice. Sustainable work design implies a co-creation process between management and employees, aligning performance goals with growth, learning, and participation. Participative approaches to work design have been shown to enhance effectiveness and ownership [112] and may prove instrumental in fostering promotive, preventive, and prohibitive safety voices. Work design can both stimulate and hinder proactivity [85]: for example, one study found that employees facing high workload show reduced levels of proactivity [113], while another showed that proactivity levels actually increase [114]. Likewise, job dissatisfaction was found to stimulate voice [115], and emotional exhaustion seems to lead to higher levels of organizational citizenship behaviors [116]. Thus, from previous research, it is not clear how job-based safety silence would impact promotive safety voice. These papers highlight the need for future research to understand under which conditions job design promotes or hinders safety voices and job-based safety silence. On a general note, job-based safety silence is likely to influence prohibitive, promotive, and hostile safety voices. First, employees may believe that signaling and reacting to colleagues’ intentional safety violations is not in their role description, and thus, they just passively carry out narrowly defined tasks and remain silent. Finally, the impact of job-based safety silence on hostile safety voice should be investigated. Arguably, several job demands are associated with higher perceived stress [117], and employees may choose to vehemently voice their frustration as a coping mechanism (i.e., engage in hostile safety voice) or they may choose to remain silent and use other coping strategies, such as crafting their job in order to reduce stress [118].

6. Propositions for Further Empirical Testing

Building on our multidimensional framework, we advance the following testable propositions to guide future empirical research:

• Higher levels of preventive voice will predict greater near-miss reporting and faster closure of corrective actions (preventive voice → early detection).
• Prohibitive voice will be negatively associated with the persistence of hazards and the reliance on workarounds (prohibitive voice → risk reduction).
• Promotive voice will foster the adoption of safety improvements, mediated by participatory structures and resource availability (promotive voice → innovation adoption).
• The effect of a hostile voice on learning will be positive under just-culture climates but negative under punitive climates (hostile voice × climate).
• Underreporting and job-based silence will suppress preventive and prohibitive voice, while diffident and relational silence will suppress promotive and hostile voice (silence profiles → voice expression).
• Leadership, team climate, and legal protections will moderate the effects of voice forms on safety outcomes (boundary conditions).
• Team-level safety voice will predict team-level safety performance (group-level voice effects).
• Transitions from silence to voice (and vice versa) will vary by form and context, shaping both short-term safety outcomes and long-term learning (voice–silence dynamics).

7. Discussion

Our multidimensional conceptualization of safety voice offers both theoretical and practical value for managers, occupational psychologists, and practitioners seeking to foster ethical, inclusive, and resilient workplaces. Evidence shows that many existing organizational reporting systems remain ineffective [119,120], often failing to safeguard employee well-being or support long-term organizational learning. By integrating a multidimensional perspective, our framework can strengthen safety management and communication strategies in ways that are both ethically responsible and psychologically sustainable. By articulating four distinct forms of safety voice, the model provides a more precise conceptual structure to guide measurement, intervention design, and organizational change. This clarity enables researchers and practitioners to move beyond “one-size-fits-all” approaches, acknowledging that voice behaviors differ in intent, form, and impact—all of which are essential for cultivating sustainable safety cultures. More accurate definitions and measures can support evidence-based tools that empower employees and extend participation, particularly among those who might otherwise remain marginalized or silent. Embedding this multidimensional perspective into daily practice may also help normalize diverse safety expressions, making it easier for employees to speak up authentically, a crucial yet demanding step for fostering collective safety and psychological well-being [120,121]. Crucially, this approach recognizes that constructive or promotive voice is not the only meaningful contribution: acknowledging multiple voice types can reduce conformity pressures and build a more nuanced culture of shared responsibility. In sustainable organizations, such an orientation promotes the early detection of risks, timely corrective action, and the protection not only of individuals and teams but also of the long-term resilience of the enterprise. In sum, this theory-integrating contribution advances the literature by (a) synthesizing insights across disciplines and clarifying the behavioral scope of safety voice; (b) differentiating safety voice from safety silence while specifying channels and boundary conditions; and (c) outlining testable propositions linking antecedents, mechanisms, and outcomes in safety-critical contexts. The framework thus offers conceptual guidance for measurement and intervention while pointing toward empirical avenues that remain underexplored.

7.1. Practical Implications for Different Forms of Voice

Our model offers practical guidance on how organizations can sustain and channel the different ways employees speak up about safety. Take promotive voice (future × challenge) as an example: here, employees bring forward ideas for improvement and innovation. To nurture this form, organizations can create participatory spaces (e.g., design reviews, learning labs, or structured “idea-to-pilot” pathways) where suggestions are taken seriously and backed by modest resources. Dedicating time to improvement work and setting transparent criteria for pilot approval sends a powerful message: constructive challenge is not only welcome but legitimate. Progress can then be tracked not just by counting ideas, but by looking at how many make it to implementation and whether employees feel safe to “challenge up”.

Preventive voice (future × defense) is about catching weak signals before they grow into serious risks. This requires systems that make it easy to report near misses, quick pre-job risk reviews, and clear stop-the-line authority. Crucially, responsiveness matters: employees need to see that their input is acknowledged and acted upon. Indicators such as the diversity of near-miss reports, closure times, and recurrence rates can indicate whether safety systems are functioning effectively (resulting in higher perceived safety climate [122]). Prohibitive voice (present × defense) happens in the moment, when employees step in to stop unsafe practices. Organizations can normalize this behavior by introducing simple red-flag protocols, clarifying micro-escalation steps (who to notify, when, and how), and training supervisors to manage authority gradients without defensiveness. Success can be evaluated by the frequency of “halt-to-fix” episodes, the elimination of precursors, and the quality of after-action learning. Finally, hostile voice (present/future × challenge) is the most controversial. It often emerges when conventional channels are blocked and frustrations build. Yet, instead of dismissing it, organizations can channel it constructively through protected escalation mechanisms—ombuds offices, ethics hotlines—and just-culture reviews that separate blame from learning. Metrics here might include resolution rates, absence of retaliation, and whether escalations lead to meaningful organizational change.

7.2. Addressing Silence

Silence should not be dismissed as the mere absence of voice. It is, in fact, a complex and multifaceted phenomenon, rooted in different motives and organizational contexts. Recognizing these nuances is essential because each form of silence requires a different response. Consider ineffectual silence, that is, the familiar sense that “nothing ever changes.” Employees hold back because they do not expect their input to make a difference. Here, the most powerful antidote is closure: organizations need to demonstrate that reporting leads to tangible outcomes. Publicizing successful fixes and providing visible feedback signals that employee contributions matter.

Diffident silence stems from self-doubt, low confidence, or fear of negative consequences. Employees may want to speak up but worry about being judged. Countering this form of silence requires building psychological safety: implementing robust anti-retaliation policies, ensuring leaders actively model appreciation for contributions, and cultivating a culture that frames errors and questions as opportunities for learning.

A further dimension is relational silence, which occurs when employees refrain from speaking to maintain or protect interpersonal ties. To address this, organizations can introduce mediated challenge processes and depersonalized reporting channels, alongside training in conflict-sensitive dialogue, so that difficult issues can be raised without endangering relationships. Underreporting silence reflects a more practical barrier: bureaucracy. Complex forms, redundant processes, and fear of punishment all discourage reporting. The solution here is straightforward: simplify. Integrate reporting into daily workflows, eliminate unnecessary steps, and make it clear that reporting is non-punitive. The easier and safer it is to report, the more likely employees will do so.

Finally, job-based silence reminds us that sometimes employees stay quiet simply because they lack the time or energy to speak up. When workload pressure leaves no room for reflection, silence becomes the default. The way forward is to rethink job design: create spaces in the workflow for reflection and ensure that productivity demands do not crowd out opportunities to raise concerns. Speaking up should never feel like a performance cost. In short, silence is not uniform—it takes many forms, each with its own causes and remedies. And crucially, silence interacts with voice: ineffectual silence often blocks promotive ideas; diffident silence inhibits prohibitive warnings; relational silence dampens preventive signals; and job-based silence restricts the scope for hostile escalation when other channels fail. By tailoring organizational responses to these patterns, organizations can transform silence from a barrier into a stepping stone—opening space for the full spectrum of safety voice behaviors that sustain resilient and inclusive workplaces.

7.3. Broader Organizational Recommendations

When we consider the system as a whole, a few guiding principles emerge for fostering just culture and safety climate [59,123,124]. Organizations should regularly map their unit-level voice–silence profiles so that they can identify where interventions are most urgently needed, whether reinforcing near-miss reporting in one unit or opening new channels for promotive ideas in another. Second, organizations need to match channels to forms of voice. Not all input belongs in the same place. Preventive and prohibitive signals should feed into rapid-response systems where immediate action can be taken; promotive suggestions thrive in participatory forums where ideas can be tested and refined; and hostile voice, often the loudest but least welcomed, requires protected escalation mechanisms that allow frustrations to be expressed safely and transformed into learning. Third, organizations must close the loop. Silence in response to a report undermines trust. Timely acknowledgment, visible follow-up, and clear resolution timelines reassure employees that speaking up leads somewhere. When employees see problems addressed, the belief that “nothing ever changes” begins to fade. Fourth, it is crucial to protect and enable employees. Speaking up should not feel like taking a personal risk. Embedding just-culture principles, guaranteeing anti-retaliation protections, and aligning incentives so that workloads do not penalize time spent on safety matters, ensures that employees are free to contribute without hidden costs. Finally, investments in leadership are needed. Supervisors and managers are the first line of response when voice occurs, and their reactions can either encourage or suppress it. They need to know when to contain and mitigate urgent risks, when to explore and experiment with new ideas, and how to de-escalate heated exchanges into constructive dialogue. In short, leaders set the tone for whether safety voice becomes a sustainable practice or a fleeting initiative. Taken together, these principles make clear that sustaining voice is not about isolated tools or policies, but about aligning diagnosis, channels, feedback loops, protections, and leadership into a coherent system that makes it both safe—and worthwhile—for people to speak up.

7.4. Limitations

This article is a theoretical synthesis and carries several limitations. First, boundaries between forms of voice and silence may blur in practice (e.g., preventive vs. prohibitive, diffident vs. relational). Second, measurement tools have not yet caught up: most scales remain unidimensional or conflate forms, and are vulnerable to self-report and social desirability bias. Third, our examples draw primarily from safety-critical contexts; generalizability to low-hazard or highly dynamic environments remains to be tested, and regulatory differences may moderate effects. Fourth, our treatment of “hostile voice” is conceptual and raises normative concerns; careful operationalization and safeguards are required. Fifth, we outline temporal and network dynamics but do not model them empirically; transitions between silence and voice are inferred rather than demonstrated. Finally, the boundary conditions identified (leadership, climate, job design, legal frameworks, interpersonal dynamics) are illustrative rather than exhaustive.

7.5. Directions for Future Research

Future research can take this framework further by exploring several themes. A first area concerns norms and socialization. Teams often develop implicit rules about which forms of voice are acceptable and which are discouraged [125]. For instance, some organizational cultures may reward promotive suggestions but frown upon prohibitive warnings. Empirical studies could shed light on how these norms evolve, and on which forms of voice are systematically enabled or suppressed across different organizational and national contexts.

A second direction involves the form, tone, and timing of voice. Not all voice sounds the same. Prohibitive voice around an immediate hazard may be urgent and assertive, while promotive ideas for procedural improvements are often delivered more diplomatically. Future research could examine whether these variations reflect contextual adaptation, levels of psychological safety, or strategic self-regulation. Timing also matters: some employees speak up immediately, while others delay until they feel conditions are safe.

Third, researchers should examine more closely the costs and contingencies of speaking up. Employees often refrain because they anticipate psychological, social, or career-related costs. Yet sometimes even small encouragements, such as supportive colleagues, a positive mood, or a receptive leader, can lower the threshold for voice [13]. Future studies could investigate how perceived risks and efficacy differ across voice forms, and which contextual conditions reduce barriers to engagement.

A fourth avenue is to broaden the scope of outcomes. Safety voice research has historically emphasized promotive forms, but these risks overshadow the contribution of preventive, prohibitive, or even hostile voice. Comparative studies could test how each form uniquely contributes to resilience and organizational learning, and whether relationships are moderated by variables such as team climate or leadership style.

Fifth, future research should consider macro-level conditions. Engagement in voice is shaped not only by what happens within teams but also by broader institutional and economic contexts. Labor protections, regulatory regimes, national cultures, and even economic cycles influence which forms of voice are viable [61,126]. For example, during economic downturns, organizations may tolerate only mandatory reporting, while in periods of economic expansion, they may invite innovative suggestions. Future studies could map how these wider contexts condition voice–silence profiles across countries and sectors.

Finally, there are clear opportunities for methodological innovation. Field experiments could test whether matching reporting channels to specific forms of voice improves safety performance. Experience-sampling methods and sequence analyses (e.g., hidden Markov models) could capture how silence evolves into voice around incidents. Multilevel modeling with repeated measures could reveal how leadership behaviors and team climate moderate these processes. Comparative cross-country studies could illuminate institutional contingencies, while digital tools (e.g., anonymous reporting systems, AI-based triage) deserve scrutiny for their impact on preventive and prohibitive voice, closure times, and retaliation risks. In short, moving forward requires both conceptual breadth and methodological depth: exploring the when, how, and why of different forms of safety voice, and embedding them within the broader organizational and societal ecosystems that shape sustainable work practices.

8. Conclusions

Our multidimensional framework of employee safety voice includes four forms (i.e., promotive, preventive, prohibitive, and hostile), each defined by a specific combination of temporal orientation (present vs. future) and stance toward safety systems (defense vs. challenge). Differentiating these forms enhances conceptual clarity and provides a stronger foundation for developing precise measures and targeted interventions. We also integrate five types of safety silence (i.e., ineffectual, diffident, relational, underreporting, and job-based) to illustrate that silence and voice are not opposite ends of a continuum, but rather distinct responses shaped by motivational and contextual factors. Different types of employee silence can inhibit or enable specific forms of voice, underscoring the importance of boundary conditions such as organizational climate, leadership practices, job design, and regulatory frameworks. From a practical standpoint, the framework guides organizations to diagnose voice–silence profiles, align reporting channels with specific voice forms, and ensure that employee input translates into learning through timely closure. The overarching message is clear: organizations must match systems and leadership responses to the type of voice at stake, thereby fostering safety cultures that are not only more resilient but also more sustainable and human-centered. While this synthesis is conceptual and some distinctions between forms may blur in practice, the model lays essential groundwork for future empirical research. At the same time, it offers actionable insights for organizations striving to increase their employees’ participation in safety management processes.