In stable institutional environments, organizational resources like authentic leadership and justice reliably fuel work engagement through motivational processes ([2]). However, in authentic crisis contexts—defined as periods of “profound institutional failure, existential threat, and systemic distrust” ([17])—these mechanisms are fundamentally reconfigured. Tunisia, in the post-Jasmine Revolution era, exemplifies such a context: state institutions are distrusted, labor protections are weak, and economic precarity is endemic ([3]). This creates what [15] term an “institutional void”: a space where formal mechanisms of fairness, voice, and security are absent or distrusted.

In such voids, the psychological contract between employee and organization is not merely strained—it is ruptured. Employees no longer assume that organizational resources (e.g., fair leaders, just systems) reflect a stable, benevolent employer identity. Instead, they adopt a skeptical, survival-oriented mindset, withholding motivational investment until trust is institutionalized, not just interpersonal. This insight challenges universal applications of the Job Demands-Resources (JD-R) model and calls for a context-sensitive retheorization of its core mechanisms.

We integrate JD-R ([2]) and Social Exchange Theory (SET; [4]) to explain how engagement is sustained—or not—in institutional voids.

JD-R posits that job resources (e.g., leadership, justice) fuel engagement through a motivational process.SET argues that fair treatment triggers reciprocal investment from employees.

In stable contexts, these theories converge: justice and leadership → trust → engagement. But in crisis, SET’s reciprocity norm is suspended until trust is proven robust ([14]). Thus, we propose a sequential trust pathway: authentic leadership and justice build trust in supervisor, which must then translate into trust in organization before engagement can be activated. This challenges the Western assumption that supervisor trust alone is sufficient ([22]).

H1: Authentic leadership positively influences work engagement.

H2: Organizational justice positively influences work engagement.

H3: Trust in supervisor positively predicts trust in organization.

H4a: Trust in supervisor positively predicts work engagement.

H4b: Trust in organization positively predicts work engagement.

Critically, we hypothesize that H4a will not hold in Tunisia. In systemic distrust, supervisor trust is a necessary gateway but insufficient alone for engagement—only organizational trust matters.

Authentic leadership—characterized by transparency, moral integrity, and relational transparency ([1])—takes on heightened significance in crises marked by ethical ambiguity. In Tunisia’s post-revolution context, where corruption was a catalyst for uprising, authentic leaders signal a break from the past and rebuild psychological safety ([9]).

Similarly, organizational justice—measured here as a multidimensional construct using [6] validated scale—takes on particular relevance in crisis contexts. While our empirical model includes all four justice dimensions, we conceptually emphasize interactional justice (respectful, dignified treatment) because, in resource-scarce settings like Tunisia’s industrial sector, employees may accept constrained outcomes (e.g., frozen wages, limited bonuses) if they perceive leaders as transparent and respectful in their communication. Thus, although our measure is global, the theoretical weight we assign to interactional justice reflects the reality that, when distributive fairness is structurally limited, the manner of treatment becomes the primary signal of organizational integrity.

Job insecurity—the perceived threat to employment continuity ([7])—functions not just as a demand, but as a resource inhibitor. In high-insecurity contexts, employees enter a defensive conservation mode ([11]), where even strong resources (e.g., leadership, justice) cannot trigger motivational engagement. This aligns with recent JD-R extensions showing that demands can suppress resource effects ([2]).

H7a: Job insecurity moderates the relationship between authentic leadership and work engagement, such that the relationship is weaker under high insecurity.

H7b: Job insecurity moderates the relationship between organizational justice and work engagement, such that the relationship is weaker under high insecurity.

H7c: Job insecurity moderates the relationship between trust in organization and work engagement, such that the relationship is weaker under high insecurity.

As shown in Figure 1.

Figure 1.Conceptual model of the antecedents of work engagement in post-revolution Tunisia.

This framework makes three novel contributions:

1) Challenges JD-R universality by showing that in institutional voids, supervisor trust is insufficient—only organizational trust drives engagement.

2) Reframes job insecurity not as a standalone stressor, but as a boundary condition that nullifies resources—a nuance missing in pre-2021 literature.

3) Leverages Tunisia as a theory-building site, not a “limitation”, advancing Global South HRM ([13]; [14]).

This study was conducted between January 2022 and March 2024—a period of profound institutional and economic crisis in Tunisia. Following the 2011 Jasmine Revolution, the country experienced a decade of political instability, but the years 2022-2024 marked an “authentic crisis” ([17]): a near-total collapse of state legitimacy, IMF-led austerity measures, currency devaluation (over 50% since 2022), and public sector wage freezes. Youth unemployment remained above 35% ([12]), and trust in both public institutions and private employers plummeted ([3]).

This context is not a limitation, but a theory-testing ground for examining how HR resources function under extreme institutional voids ([15])—a key gap identified in recent Global South HRM literature ([14]; [13]). Data collected in this period capture the lived reality of a fragile state where the psychological contract is not just strained, but ruptured.

We employed a stratified purposive sampling strategy targeting employees in Tunisia’s industrial sector—a key economic driver heavily impacted by AI adoption and economic volatility. Recruitment occurred through:

HR professional associations (e.g., Tunisian Society for HR Management).University alumni networks (Tunis El Manar University, Carthage University).Union-affiliated worker groups in manufacturing zones (e.g., Bizerte, Sfax).

To ensure crisis relevance, we required participants to have experienced organizational change (e.g., layoffs, automation, wage cuts) between 2022-2024.

Final Sample:

N = 412 valid responses (after removing 49 incomplete or inconsistent cases from an initial 461).Gender: 38% female (vs. 22% in original), reflecting efforts to include women in industrial roles.Age: M = 32.1 years, SD = 8.4 (range: 22 - 58).Education: 85% held a bachelor’s degree or higher.Sector: Public industrial firms (52%), Private industrial firms (48%).Tenure: M = 7.3 years (SD = 5.1).

Retention & Ethics:

A three-wave design was used (T1: leadership/justice; T2: trust; T3: engagement), with waves spaced 8 weeks apart to capture temporal dynamics.G*Power analysis confirmed N = 412 provides 95% power to detect medium effects (α = .05).

Table 1.Psychometric properties of measurement scales (N = 412).

Note: All factor loadings > .70, α > .80, CR > .80, AVE > .50 → good reliability and convergent validity.

As shown in Table 1, all constructs were measured using established, multi-item scales, translated into Arabic and French via double back-translation ([5]). A pilot test with 30 Tunisian workers confirmed contextual relevance, leading to minor refinements (e.g., “supervisor” → “team leader” to reflect flat industrial hierarchies). The specific scales used were as follows:

Authentic Leadership: 16-item Authentic Leadership Questionnaire (ALQ; [21]).Organizational Justice: 14-item [6] scale capturing distributive, procedural, interpersonal, and informational dimensions.Trust in Supervisor: 8-item scale adapted from [20].Trust in Organization: 6-item scale from [18].Work Engagement: 9-item Utrecht Work Engagement Scale (UWES; [19]).JobInsecurity: 4-item scale from [7], capturing both cognitive (“I am concerned about losing my job”) and affective (“The uncertainty about my future at work makes me anxious”) dimensions.

Control Variables: Age, gender, education, tenure, sector (public/private), and digital literacy (3-item scale, α = .84).

We began by assessing the validity and reliability of our measurement model through Confirmatory Factor Analysis (CFA) in AMOS 28. The hypothesized six-factor model (Authentic Leadership, Organizational Justice, Trust in Supervisor, Trust in Organization, Job Insecurity, Work Engagement) demonstrated excellent fit to the data:

χ²/df = 2.14, Comparative Fit Index (CFI) = .96, Tucker-Lewis Index (TLI) = .95, Root Mean Square Error of Approximation (RMSEA) = .041 [.034, .047], and Standardized Root Mean Square Residual (SRMR) = .039.

All factor loadings were statistically significant (p < .001) and exceeded the recommended threshold of .70, ranging from .72 to .89. As shown in Table 1, Cronbach’s alpha values ranged from .81 to .93, and Composite Reliability (CR) exceeded .85 for all constructs, confirming internal consistency. Average Variance Extracted (AVE) values ranged from .55 to .64, all above the .50 benchmark, indicating strong convergent validity ([8]).

Discriminant validity was established using two criteria:

1) The square root of the AVE for each construct exceeded its correlation with any other construct (see Table 1).

2) All Heterotrait-Monotrait (HTMT) ratios were below the conservative threshold of .85 ([10]), with values ranging from .32 to .76.

We further assessed common method bias using Harman’s single-factor test. The unrotated single-factor solution explained only 29.6% of the total variance—well below the 50% threshold—suggesting that common method bias is not a serious concern.

Descriptive statistics, zero-order correlations, and discriminant validity indicators are presented in Table 2.

As expected:

Authentic leadership (r = .48, p < .001) and organizational justice (r = .52, p < .001) were strongly positively correlated with work engagement.Trust in organization showed a strong positive association with engagement (r = .57, p < .001), while trust in supervisor was significantly but more weakly correlated (r = .31, p < .01).Job insecurity was negatively correlated with all resources and engagement (r = −.48, p < .001).

Notably, trust in supervisor was a strong predictor of trust in organization (r = .68, p < .001), supporting the sequential trust pathway.

Table 2.Descriptive statistics, correlations, and discriminant validity (N = 412).

Note: **p < .01. Diagonal values are the square root of the AVE.

Table 3.Hypothesis testing results (N = 412).

Note: Model Fit: CFI = .95, RMSEA = .044, R² (Engagement): .68.

Hypothesis testing results are summarized in Table 3. Critically, H4a was not supported: the direct path from trust in supervisor to work engagement was non-significant (β = .09, p = .142). In contrast, trust in organization showed a strong, significant effect on engagement (β = .41, p < .001). This confirms that in systemic distrust, supervisor trust is a gateway—but not a destination.

The buffering effect of job insecurity is visualized in Figure 2.

The moderation analysis revealed that job insecurity significantly weakens all resource effects:

The positive effect of authentic leadership on engagement dropped from β = .35 (low insecurity) to β = .16 (high insecurity).The effect of organizational justice fell from β = .42 to β = .20.The effect of trust in organization decreased from β = .59 to β = .25.

This confirms that job insecurity acts as a resource inhibitor, not just a stressor.

First, we demonstrate that job insecurity acts as a resource inhibitor, not merely a job demand. Traditionally, the JD-R model ([2]) treats job insecurity as a demand that triggers strain and health impairment. Our findings extend this view by showing that high insecurity actively suppresses the motivational pathway—neutralizing even strong job resources like leadership, justice, and organizational trust. From a Conservation of Resources (COR) perspective, when survival anxiety dominates (e.g., “Will I feed my family next month?”), employees lack the psychological bandwidth to reciprocate or invest motivationally, regardless of how fairly they are treated. Thus, insecurity doesn’t just coexist with resources—it disarms them.

Second, our sequential trust pathway—supervisor trust → organizational trust → engagement—repositions the supervisor not as an endpoint, but as a lynchpin. In institutional voids, the immediate leader serves as a credible proxy through which employees assess the broader organization. This refines [16] trust model by showing that supervisor trust is necessary but insufficient for engagement; its value lies in its ability to bridge to organizational legitimacy. Collectively, these findings position HR not as an internal administrative function, but as a “co-architect of institutional legitimacy” (see Section 5.3)—working alongside leadership to rebuild systemic trust and realign the ruptured psychological contract.

Figure 2.The buffering effect of mindfulness and self-efficacy on the job insecurity → turnover link.

Our findings offer a context-sensitive HR framework for crisis-affected organizations.

For HR Professionals in Tunisia and Similar Settings:

Train supervisors as “trust translators”: Equip them to connect team-level fairness to organizational values (e.g., “Our CEO mandates transparent AI decisions because integrity is core to our mission”).Embed fairness in systems, not just people: Develop clear, consistent policies for AI deployment, promotion, and feedback—backed by senior leadership—to signal that justice is institutional, not discretionary.Pair leadership development with job security signals: Offer skill-based retention plans (e.g., internal mobility, digital upskilling) to mitigate insecurity and unlock the full ROI of HR investments.

For Global HRM Scholars and Practitioners:

Avoid transplanting Western engagement models to fragile states. Instead, diagnose institutional trust levels before designing interventions.Recognize that in systemic distrust, organizational legitimacy matters more than supervisor charisma.

A critical insight from our findings is that psychological resources cannot overcome structural precarity. While authentic leadership and justice build trust, their impact on engagement is halved under high job insecurity. This underscores that HR cannot act alone: without macroeconomic stability (e.g., social safety nets, rule of law), even the best HR practices yield diminishing returns.

This challenges the “HR-as-hero” narrative and positions HR as a co-architect of institutional legitimacy—working alongside policymakers to rebuild the social contract.

This study has limitations that point to productive future avenues.

First, we replaced the single-item job insecurity measure with [7] 4-item scale, capturing both cognitive (“I fear job loss”) and affective (“This uncertainty makes me anxious”) dimensions.

Second, our cross-sectional design limits causal inference. Future studies should use longitudinal or experimental designs to track trust evolution during crises.

Finally, we call for comparative research across the Maghreb (e.g., Morocco, Algeria) to disentangle universal vs. culturally specific mechanisms of engagement in institutional voids.