Hernandez et al. (Reference Hernandez, Melson-Silimon and Zickar2025) present a visionary case for expanding industrial and organizational (I-O) psychology to include nonhuman animal workers, but we urge caution regarding the assumption that canonical I-O constructs (e.g., job satisfaction, organizational commitment, or engagement) can be validly assessed in these populations. These constructs already rely on language, metacognition, and self-awareness—capacities that underpin our dependence on survey- and self-report assessment paradigms. Even within human populations, measurement depends on introspection and linguistic articulation—features not transferable to nonhuman workers. We discuss these measurement challenges in greater detail below.
Given these limitations, the challenge of applying these constructs to nonhuman animal workers becomes even more apparent. Thus, in extending our lens to animal workers, we confront a central tension: How can we ethically and scientifically assess experiences we cannot directly access? Rather than stretch construct validity beyond its breaking point, we propose shifting focus from measurement to justice, relying on principles of fairness, burden, and exchange to guide humane and equitable organizational practices for nonhuman workers. We argue that I-O psychology must ethically evolve to include nonhuman workers as legitimate subjects of its frameworks, protections, and advocacy.
The limits of construct measurement in nonhuman workers
The limitations of psychological measurement in humans are well documented, particularly when it comes to constructs that rely on introspection, language, and contextually bound behaviors. I-O psychology frequently operationalizes constructs such as job satisfaction, engagement, and motivation through self-report instruments that presuppose an individual’s reflective capacity and willingness to respond accurately. However, even in human populations, these assumptions often break down.
Common method biases such as social desirability, consistency motifs, acquiescence, and transient mood states systematically distort self-report data and can account for a substantial portion of observed variance in organizational constructs (Podsakoff et al., Reference Podsakoff, MacKenzie, Lee and Podsakoff2003). Meta-analytic evidence suggests that as much as 26% of the variance in attitude measures, such as job satisfaction, may be attributable to method effects alone, with similarly concerning levels for other constructs such as personality assessments (Podsakoff et al., Reference Podsakoff, MacKenzie, Lee and Podsakoff2003).
Moreover, as Flake et al. (Reference Flake, Pek and Hehman2017) emphasize, many psychological measures, especially in applied fields, lack rigorous construct validation. Structural and substantive validity evidence is often neglected, with coefficient alpha standing in as a proxy for psychometric rigor despite its well-documented limitations. This neglect includes failing to assess dimensionality, assuming tau-equivalence, and overlooking alternative reliability estimates (e.g., Cho & Kim, Reference Cho and Kim2015; McNeish, Reference McNeish2022; Sijtsma, Reference Sijtsma2009). The result is a landscape in which even widely used constructs are sometimes under validated, context sensitive, and overly reliant on assumptions about reflective capacity and linguistic articulation.
Psychological constructs are often latent, inferred from patterns in self-report or observed behavior, interpreted through a human cognitive-emotional lens. However, without language, self-awareness, or the ability to reflect on one’s internal states—features foundational to human psychological research—any extrapolation of such constructs to animals risks anthropomorphism (Urquiza-Haas & Kotrschal, Reference Urquiza-Haas and Kotrschal2015). As Broom (Reference Broom2010) notes, acknowledging that animals possess cognitive and emotional capacities is not equivalent to asserting that these states can be meaningfully or reliably measured using constructs built for humans.
These challenges are magnified when extended to nonhuman contexts. Canonical I-O constructs were developed within a framework that assumes a human-specific cognitive architecture, namely, the access to and communication of internal states through language. As such, their direct translation to nonhuman animals risks conceptual overreach. Simply put, acknowledging cognitive and emotional capacities in animals does not imply that we can meaningfully measure them. Behavioral proxies for constructs like fairness, stress, or motivation can be insightful but remain fundamentally different from the internal subjective experiences they are presumed to reflect.For example, task refusal or walking away might signal disengagement or dissatisfaction, but it could also reflect confusion, discomfort, or even playful interaction.
Attributing human-like psychological states to animals without access to their subjective experience risks projecting our own mental architecture onto other species (Urquiza-Haas & Kotrschal, Reference Urquiza-Haas and Kotrschal2015). As Keupp, lead author of a recent primate study, reminds us, “we cannot ask our nonhuman primate subjects how they perceive an experimental situation; we can only observe their behavior… [and] need to avoid falling into the trap of viewing animal behavior through a human lens” (Wylesol, Reference Wylesol2023, para. 18). Thus, whereas nonhuman workers may demonstrate sophisticated behavior, the measurement of psychological constructs must be approached with caution. Acknowledging sentience or cognition is not synonymous with possessing the methodological infrastructure to assess it meaningfully.
Justice as a viable and necessary alternative framework
Although traditional I-O measurement has relied heavily on self-reports and introspection, justice-based frameworks offer a promising alternative for nonhuman workers, one that shifts the focus from internal states to observable structures, behaviors, and outcomes. Organizational justice, a central construct within I-O psychology, provides a robust and well-established framework for evaluating fairness without depending on subjective self-assessments. The four-dimensional model of justice—distributive, procedural, interpersonal, and informational (Colquitt, Reference Colquitt2001)—captures key aspects of fairness, including how resources are allocated, how decisions are made, and how individuals are treated. Importantly, these justice dimensions can be assessed not only through perceptions but also via concrete organizational policies and practices. Adopting this approach would support more objective, enforceable, and policy-relevant interventions than those based solely on affective or introspective measures.
Nonhuman workers, although unable to participate in surveys or express their perceptions of fairness, are nonetheless subject to organizational systems of reward, burden, and control. They face constraints on their freedom, are assigned tasks of varying risk and complexity, and are removed from service or life itself when deemed no longer useful. These realities place ethical demands on the institutions that employ them. If we accept that nonhuman workers contribute labor, often at personal cost, it follows that they are also entitled to protections grounded in fairness and equity. This does not require speculative attribution of human-like cognition. Rather, it requires a principled commitment to equitable treatment, aligned with justice models that do not rely on introspective access.
We define our proposed justice-based framework as an approach that evaluates work using observable structural features—such as burden, exchange, constraints, and distribution of risk—rather than relying on introspective self-reports. This reframing enables I-O psychologists to assess organizational practices through testable and operationalizable criteria (Colquitt, Reference Colquitt2001; Greenberg, Reference Greenberg and Cropanzano1993), aligning with evidence-based traditions that emphasize observable, modifiable workplace features (Briner & Rousseau, Reference Briner and Rousseau2011). By shifting the methodological emphasis from inferring inaccessible subjective states to examining structural conditions, this framework strengthens the scientific rigor. It enables clearer operational definitions, systematic comparison across contexts, and falsifiable hypotheses about the impact of organizational practices on well-being and fairness (Cropanzano et al., Reference Cropanzano, Massaro and Becker2017). Thus, a justice-based approach does not dilute empirical standards; rather, it enhances them by grounding our assessments in observable reality while expanding ethical consideration to include nonhuman workers.
Another reason this shift is necessary is that attempting to measure internal psychological states,such as “motivation” or “engagement,” in nonhuman workers creates an illusion of ethical responsibility. We appear to care because we “measured,” but we often failed to act meaningfully on the findings. For instance, behavioral signs of stress or fatigue in search-and-rescue dogs are frequently overlooked unless their performance declines. This disconnect between observation and response highlights the limits of a measurement-driven approach. A survey by Plous (Reference Plous1996) found that although 85.9% of psychologists supported protecting the psychological well-being of primates used in human-benefitting research, only 42.8% were confident these animals were treated humanely. Rather than focusing on how animals feel, an inaccessible domain, as I-O psychologists, we should prioritize tangible protections. Shifting attention from inferred internal states to enforceable rights marks a more actionable and ethically sound approach to nonhuman labor justice.
Recommendations for equitable organizational practices for nonhuman workers
Principles such as equal burden for equal benefit or freedom from unnecessary harm reflect the fairness logic that our justice-based framework can systematically evaluate. These ideas parallel Rawls’s (Reference Rawls1971) view that inequalities are permissible only if they result in compensatory benefits to the least advantaged. Although Rawls’s framework was developed for human societies, its emphasis on fairness as a condition of cooperation resonates with the treatment of animal workers whose participation is similarly involuntary and whose welfare is often structurally subordinated to institutional goals. As Bowd and Shapiro (Reference Bowd and Shapiro1993, p.136) put it,“to base ethical decisions on species membership alone, […] is as arbitrary as relying on skin color or gender in hiring decisions.”
Real-world examples illustrate how justice-based reasoning already informs practice, even if implicitly. Consider policies surrounding K9 retirement benefits, where some jurisdictions provide pensions, adoption subsidies, or continued medical care, an acknowledgment of contribution and a mechanism of redress. Likewise, the design of humane workload rotation schedules, species-appropriate rest time, and protective gear in dangerous roles (e.g., military or police work) reflects principles of procedural and distributive justice (Colquitt, Reference Colquitt2001). Legal interventions that regulate the use of animals in entertainment or labor (e.g., pulling carts, sniffing explosives) often hinge on the idea that certain burdens must be limited, even if the animal cannot object.
Recent scholarship has begun to explicitly frame animal labor as a matter of interspecies justice, grounded not in cognitive equivalence but in structural participation. Blattner et al. (Reference Blattner, Coulter and Kymlicka2020) argue for a rights-based approach that treats animal workers as members of cooperative systems entitled to protections against exploitation. Their edited volume explores how concepts such as free and unforced labor, safe working conditions, and retirement security, long staples of human labor policy, can be meaningfully adapted to animal contexts. These arguments resonate with Rawlsian concerns about reciprocity and about ensuring that no participant in a cooperative scheme is used merely as an end (Rawls, Reference Rawls1971).
As the next step, the U.S. Animal Welfare Act (AWA, 2018) could be expanded to include all nonhuman workers, such as birds, research rats and mice, breeding livestock, poultry in production systems, and working horses. Although these animals are currently excluded from federal protections, they engage in sustained and often demanding labor in service of human interests. Their omission represents a significant ethical gap. Recognizing them as workers involves acknowledging not only their utility but also their vulnerability to harm, coercion, and exploitation. Expanding AWA coverage would establish a baseline of legal protection through mandatory welfare assessments, facility inspections, and strict restrictions on harmful practices. Such a shift would move away from flawed psychological proxies and toward an actionable, rights-based framework for nonhuman labor justice. If these animals are essential enough to be employed, they are essential enough to be protected.
Conclusion
As the scope of I-O psychology expands to consider the labor and contributions of nonhuman workers, its ethical frameworks must evolve accordingly. The limitations of traditional psychological measurement, especially those reliant on self-report, language, and introspective awareness, highlight the risk of overextending human-centric constructs in contexts where they are neither valid nor meaningful. Rather than force-fit constructs like job satisfaction or engagement onto populations lacking access to introspective reporting, a justice-based approach offers a more principled and actionable alternative. By grounding ethical obligations in structural fairness, through equitable burden sharing, procedural safeguards, and humane working conditions, I-O psychology can uphold methodologically rigor while extending moral consideration to previously excluded populations. A shift from measurement to justice does not weaken scientific integrity. It strengthens it by aligning our methods with the lived realities of those we seek to understand and protect. By grounding assessments in structural features, it ensures our approaches remain both ethically inclusive and empirically robust.
Competing interests
The author(s) declare none.
