Assessing the Impact of Social Culture and National Regulations on Women’s Engagement in Logistics Decision-Making in Uganda

2.1 Theoretical review

Sandra Bem’s Gender Schema Theory (1981) examines how children internalise gender-defined attributes, including preferences, abilities, and behaviours. Bem (1981) argued that children learn societal roles for males and females from their cultural environment, with boys and girls developing masculine and feminine traits at an early age. Starr and Zurbriggen (2017) supported this notion, highlighting that children form gender schemas influenced by socialisation experiences, which in turn affect their categorisation and behaviour. In Uganda, traditional socialisation practices reinforce entrenched gender schemas, where girls are often directed toward domestic roles and boys toward leadership positions in areas like logistics. This contributes to the underrepresentation of women in leadership roles. Gender schemas shape institutional contexts, influencing the challenges women face and their motivation to pursue corporate leadership positions (Lewellyn and Muller-Kahle, 2020).

For instance, Ugandan cultural practices associate decision-making in logistics with masculinity, further solidified by the types of toys given to children. To encourage more women in logistics, early childhood education and altering parental attitudes are vital. However, Gender Schema Theory does not fully address systemic issues, such as workplace discrimination and unequal access to resources, nor does it account for women who succeed in leadership settings despite these barriers. These factors are better explained by Institutional Theory, which emphasises broader systemic changes needed to enhance women’s decision-making participation.

Institutional Theory examines the social, political, and cultural contexts that shape women’s roles by identifying the formal and informal rules influencing these roles. It helps develop context-specific policy recommendations to enhance women’s engagement in decision-making. Initially developed by Meyer and Rowan in 1977, this sociological perspective investigates how organisations adopt and conform to social norms and values (Lewellyn and Muller-Kahle, 2020; Zaman et al., 2022). Organisations strive for legitimacy by responding to institutional pressures and aligning with societal expectations, particularly in competitive environments (Allemand et al., 2014; Saeed et al., 2022).

Institutional pressures encompass regulatory pressures stemming from laws and regulations, normative pressures arising from professional and cultural norms, and cognitive pressures reflecting shared beliefs about appropriate behaviour (DiMaggio & Powell, 1983; Scott, 2001). According to Silva (2021), these dimensions shape organisational structures by influencing responses to external demands and obligations. The Theory emphasises that organisations adopt practices not only for efficiency but also for legitimacy (Adedeji et al., 2020). However, it also highlights symbolic compliance, where organisations implement policies to align with societal norms without making substantial changes (Ozkan, 2025), which may not lead to meaningful opportunities for women.

To overcome challenges, institutional reforms must prioritise creating genuine opportunities for women in leadership. This includes enforcing gender equality regulations, promoting gender-sensitive practices, and shifting cultural norms to contest traditional gender roles. Addressing these barriers can enhance women’s engagement in decision-making roles, contributing to economic advancement and aligning with national objectives, such as Uganda Vision 2040 and the African Union Agenda 2063.

2.2 National regulations and women’s engagement in decision-making

Advocates for strong regulatory interventions argue that national laws and policies are essential for creating an environment that fosters women’s leadership. Quota systems, affirmative action policies, and gender equality laws have proven effective in enhancing women’s representation in political and corporate decision-making. Rwanda’s constitutional quota has resulted in one of the highest proportions of women in parliament globally (Kang, 2013). Uganda’s National Gender Policy (1997) and the Uganda Gender Policy (2007) also strive to dismantle barriers that obstruct women’s advancement (Bosha, 2014).

The inclusion of women in decision-making is essential for achieving gender equality and sustainable development (Akinwale, 2023; Gupta et al., 2024). Despite various countries adopting legal frameworks for women’s governance engagement, their effectiveness is often debated. Rwanda’s gender quotas demonstrate that legal mandates can increase women’s involvement; however, questions remain about whether they address deeper systemic barriers or merely serve as cosmetic solutions. Additionally, Reddy & Jadhav (2019) note that gender quotas have enhanced female representation on corporate boards in many regions; however, they often encounter resistance due to their politically charged nature. Grounded in Institutional Theory, Terjesen et al. (2015) argue that regulatory, normative, and cognitive pressures influence the adoption of quotas.

Besides, critics contend that quotas may not transform organisational cultures but rather fulfil legal requirements. Societal norms complicate this further; in supportive environments, quotas succeed, while in less supportive ones, they may appear tokenistic. Zhang (2020) emphasises that gender diversity’s success relies on legal and social support, and without cultural acceptance, mandatory measures may have minimal impact. While proponents like Kirsch (2018) assert that legal quotas are essential for increasing female corporate board engagement, sceptics argue that numerical parity does not equate to meaningful engagement. Further, Orazalin & Baydauletov (2020) suggest that quotas can promote diversity but require genuine organisational commitment to equity. Without addressing structural and cultural factors of gender inequality, quotas risk becoming symbolic rather than transformative.

Nevertheless, Brieger et al. (2019) note that stronger emancipative forces lead to higher gender diversity on boards by challenging traditional norms emphasised by Zubeyr et al. (2013), who indicate that regulatory measures with enforcement mechanisms enhance women’s leadership participation, compelling organisations to diversify. The logistics sector could particularly benefit from frameworks promoting gender diversity.

In conclusion, national regulations are essential for fostering women’s engagement in decision-making, but their impact is influenced by socio-cultural and institutional contexts. A multifaceted approach combining legal frameworks with cultural transformation and organisational change is needed. Thus, the hypothesis;

2.3 Social culture and women’s engagement in decision-making

The relationship between social culture and women’s engagement in decision-making is a complex and evolving topic, marked by contrasting findings. While many studies highlight social culture as a significant barrier to women’s leadership, recent research indicates that the restrictive influence of culture may be overstated, particularly as societal norms shift and economic demands evolve (Adhikari & Kusakabe, 2022). This shift, facilitated by globalisation and educational advancements, has begun to advance gender diversity in organisational leadership.

However, Hofstede Insights (2023) and Carrasco et al. (2020) argue that deep-rooted cultural biases still hinder women’s ascendancy in leadership roles, especially in societies with entrenched traditional gender roles. Eagly and Heilman (2022) further emphasise that cultural expectations significantly affect women’s opportunities and the legitimacy of their leadership. In Uganda, traditional gender norms within a patriarchal framework limit women’s decision-making engagement (Namatovu & Kyejjusa, 2022), with rural communities perpetuating stereotypes that delegitimise women’s leadership aspirations.

Despite this, evidence suggests that cultural attitudes in Uganda are evolving due to increased educational attainment, urbanisation, and exposure to global gender equality norms (Mugisha et al., 2023). Religious and traditional practices exert a dual influence, with some serving as platforms for women’s empowerment (Tripp, 2019). Globalisation introduces new ideas promoting gender equality, yet resistance remains where local cultural values conflict with these norms (Kabeer & Subrahmanian, 2023).

In conclusion, understanding the intricate relationship between social culture and women’s decision-making engagement requires examining the interplay of these global and local influences. Strategies to enhance women’s leadership must address cultural barriers while leveraging emerging opportunities for gender-inclusive leadership in Uganda and similar contexts. Therefore, the hypothesis;

2.4 National laws and regulations mediate the relationship between social culture and women’s engagement in decision-making

The relationship between social culture and women’s engagement in decision-making is complex, often mediated by national laws that reinforce or challenge cultural norms. Social culture, which includes traditional gender roles, can hinder women’s leadership, especially in patriarchal contexts (Inglehart, 2021). However, institutional interventions, such as gender equality laws, can transform the influence of social culture on leadership opportunities.

Institutional Theory suggests that organisations adapt to external regulatory pressures (DiMaggio & Powell, 1983; Scott, 2014). In societies with restrictive norms, laws promoting gender equality can disrupt “normative inertia” and enhance women’s leadership visibility (Kabeer & Subrahmanian, 2023). Additionally, cross-national studies affirm that strong gender equality frameworks weaken the negative impact of cultural biases on women’s leadership (World Economic Forum, 2023; OECD, 2022), as evidenced by countries with comprehensive policies experiencing higher rates of women’s engagement (Carrasco et al., 2020).

Moreover, True & Parisi (2021) note that legal reforms can catalyse societal change, reshaping cultural attitudes toward women in leadership. However, the effectiveness of these laws relies on enforcement, political will, and societal acceptance (Htun & Weldon, 2018). In Uganda, for example, traditional norms have limited women’s participation, but legal reforms like the 2006 Equal Opportunities Act have started to challenge these barriers (Namatovu & Kyejjusa, 2022; Mugisha et al., 2023). Nevertheless, Adhikari and Kusakabe (2022) warn that without accountability and cultural sensitisation, laws may lead to mere “symbolic compliance” (Sundström et al., 2017), failing to foster meaningful change. In conclusion, national laws mediate the impact of social culture on women’s decision-making. When well-enforced and supported by broader initiatives, legal reforms can disrupt traditional barriers to women’s leadership. Thus, the hypothesis.

Adopted from literature and modified by the researcher

3.1 Data analysis

This study utilised both descriptive and inferential statistical analyses. Descriptive analysis was conducted to produce summary statistics and frequency distributions, which aided in understanding the demographic and background characteristics of the respondents. In contrast, inferential analysis was employed to explore the relationships among the study variables. To achieve the research objectives, a two-step quantitative analytical approach was implemented using SmartPLS software. The first step involved evaluating the measurement model by testing for validity, reliability, indicator relevance, and collinearity, in accordance with the guidelines of Ramayah et al. (2018). The second step focused on assessing the structural model by analysing path coefficients, regression results, and model fit indices, as recommended by Hair (2014). The study employed Partial Least Squares Structural Equation Modelling (PLS-SEM), a second-generation statistical technique well-suited for evaluating both latent and manifest variables modelled in both formative and reflective ways (Hair et al., 2010). Given the inclusion of both reflective and formative higher-order constructs in the model, PLS-SEM was considered appropriate for testing the hypothesised relationships among the variables.

Ethical approval for this study was obtained from the Faculty of Graduate Studies and Research, Makerere University Business School (MUBS) under the Makerere University Research Ethics Framework. Approval was granted on 15 July 2023, through an official introduction and clearance letter, Reference No. MUBS/FGSR/2023/07/015, which authorized the researcher to conduct data collection among logistics companies, industry associations, and regulatory agencies in Uganda. Independent Institutional Review Board (IRB) approval was not required, as the study involved non-clinical survey and interview data obtained from adult professionals (aged 18 years and above) and posed minimal risk to participants. All participants were provided with a full explanation of the study’s purpose and voluntarily provided informed written consent prior to participation. Confidentiality and anonymity were strictly maintained throughout the research process.

3.2 Validity and reliability

In this study, the internal reliability of the questionnaire was evaluated using Cronbach’s alpha coefficients, calculated through the inter-item consistency method as advised by Cho and Kim (2015). As shown in Table III, all computed Cronbach’s alpha values surpassed the threshold of 0.70, indicating acceptable internal consistency and reliability of the measurement scales (Nunnally, 1967). To assess the validity of the instrument, content validity was initially established through consultations with experts and practitioners. Their insights contributed to the development of the Content Validity Index (CVI), which produced satisfactory results: 0.88 for national regulations and laws, 0.89 for social culture and 0.94 for Women’s Engagement in Decision-Making, all exceeding the acceptable benchmark (Natalio et al., 2014). Following this, convergent validity was evaluated by calculating the Average Variance Extracted (AVE) and Composite Reliability (CR), in accordance with the guidelines provided by Hair et al. (2010). As illustrated in Table III, all constructs exhibited composite reliability values above 0.70 and AVE values exceeding the 0.50 threshold ( Table I), affirming convergent validity (Henseler et al., 2015; Fornell & Larcker, 1981). Moreover, discriminant validity was assessed using the Heterotrait-Monotrait (HTMT) ratio of correlations. Table II demonstrates that all HTMT values were below the critical cut-off value of 0.90, thereby satisfying the criteria for discriminant validity (Yusoff et al., 2020). Consequently, the constructs utilised in this study were both reliable and valid, confirming that the instrument effectively measured the intended study variables.

3.2.1 Construct validity

In this study, construct validity was evaluated through the assessment of both convergent and discriminant validity, following the recommendations of Blumberg et al. (2014). Convergent validity focused on determining whether items intended to measure the same construct were highly correlated, thereby confirming internal consistency and unidimensionality. This was assessed using Confirmatory Factor Analysis (CFA), during which the Average Variance Extracted (AVE) was calculated for each latent construct. According to Hair et al. (2010), convergent validity was deemed adequate when AVE values exceeded the threshold of 0.50, and all item loadings were statistically significant and above 0.60. This indicates that each construct captured a sufficient proportion of variance from its associated indicators. Additionally, discriminant validity, which ensures that conceptually distinct constructs are empirically distinguishable, was assessed using both the Fornell–Larcker criterion and the Heterotrait-Monotrait (HTMT) ratio of correlations. HTMT values were compared against the conservative threshold of 0.85, in line with recent methodological guidance, to confirm the absence of construct overlap (Henseler et al., 2015). To establish comprehensive construct validity, the study first conducted an Exploratory Factor Analysis (EFA) to uncover the underlying structure of the measurement items, followed by CFA to confirm and validate that structure. The integration of EFA and CFA, along with rigorous validity assessments, ensured that the constructs utilised in this study were both theoretically grounded and empirically robust.

3.2.2 Exploratory Factor Analysis (EFA)

To evaluate construct validity, an Exploratory Factor Analysis (EFA) was performed before conducting a Confirmatory Factor Analysis (CFA). The purpose of the EFA was to empirically identify items that corresponded with the theoretical constructs, following the guidelines established by Hair et al. (2010). Only factors with eigenvalues greater than 1.0 were retained, and items needed to demonstrate factor loadings of at least 0.50 to be considered significant. Items exhibiting cross-loadings exceeding 0.50 across multiple factors were excluded to improve discriminant clarity. Principal Component Analysis (PCA) was used as the extraction method to enhance the interpretation of the factor structure. Furthermore, the Kaiser–Meyer–Olkin (KMO) measure of sampling adequacy and Bartlett’s Test of Sphericity were employed to confirm the appropriateness of the data for factor analysis and to verify the presence of significant inter-item correlations.

3.2.3 EFA of National level factors

According to Table IV, factor loadings indicate the degree to which each item correlates with its underlying latent construct. According to Hair et al. (2010), a loading of ≥0.60 is generally regarded as strong, while a threshold of 0.50 is typically acceptable (Hulland, 1999). In the case of the National Regulations construct, all items (NRL1–NRL10) displayed moderate to strong factor loadings, each surpassing the 0.50 threshold, thereby signifying acceptable representation of the construct. Likewise, for the Social Culture construct, all items (SC1–SC9) demonstrated high factor loadings exceeding 0.60, indicating excellent internal consistency and robust factor convergence. The Kaiser-Meyer-Olkin (KMO) measure of sampling adequacy was found to be 0.829, which falls within the acceptable range, indicating that the sample size is adequate for factor analysis (Kaiser, 1974). Furthermore, Bartlett’s Test of Sphericity yielded an Approximate Chi-Square value of 2486.087 with 171 degrees of freedom, achieving a significance level of p < 0.001. This statistically significant outcome confirms that the correlation matrix includes sufficient inter-item correlations, thereby validating the use of factor analysis.

3.2.4 EFA for women’s engagement in decision-making

According to Table V, all forty items assessing women’s engagement in decision-making displayed strong internal consistency and convergent validity, with factor loadings exceeding the acceptable threshold of 0.50. The four underlying dimensions collectively accounted for 63.826% of the total variance, surpassing the minimum recommended threshold and indicating that the items effectively capture the intended constructs. The Kaiser-Meyer-Olkin (KMO) Measure of Sampling Adequacy was 0.803, exceeding the recommended benchmark of 0.80 and confirming the appropriateness of the data for factor analysis (Kaiser, 1974). Furthermore, Bartlett’s Test of Sphericity produced an approximate Chi-Square value of 4700.207 with 780 degrees of freedom and a significance level of p < 0.001. This significant result demonstrates that the correlation matrix is not an identity matrix, indicating sufficient inter-item correlations to justify the use of factor analysis.

3.2.5 Confirmatory factor analysis (CFA)

CFA plays a vital role in assessing the psychometric properties of constructs, including aspects such as item reliability, internal consistency, content validity, convergent validity, and discriminant validity. In this study, the constructs identified through Exploratory Factor Analysis (EFA) were validated through CFA to develop measurement models for these dimensions, utilising the Partial Least Squares Structural Equation Modelling (PLS-SEM) approach.

3.2.6 CFA measurement model for National level factors

Following the Exploratory Factor Analysis (EFA), a total of 19 items initially exceeded the minimum factor loading threshold of 0.50. However, subsequent Confirmatory Factor Analysis (CFA) revealed that only six items (SC1, SC3, SC4, SC5, SC6, and SC7) out of the original nine for the Social culture construct were deemed statistically significant and retained. In the case of the National Regulations and Laws construct, eight out of the ten items that had succeeded in the EFA phase met the more rigorous CFA threshold of 0.60 and were included in the final model. Figure 2 displays the standardised loadings, illustrating the strength of the correlations between each observed variable (item) and its corresponding latent construct within the broader domain of national-level factors.

Figure 1. The conceptual model generated from literature review.

Adapted from (Ouedraogo, 2018; Kang, 2013; Kahn, 1990).

Figure 2. The latent construct National level factors with its indicators; National regulations and laws (NRL) with its observed items (NRL1–NRL8) and Social culture (SC) with its observed items with (SC 1-SC 7) standardized loadings above 0.60, confirming convergent validity.

3.2.7 CFA measurement model for women engagement in decision making

Following the Exploratory Factor Analysis (EFA), all 40 items surpassed an initial threshold of 0.5. However, upon conducting the Confirmatory Factor Analysis (CFA), only five items related to cognitive engagement (CE1, CE2, CE3 CE7 and CE8) were recognised as significant, similarly only five items for emotional engagement (EE3, EE4, EE6, EE7 and EE8) also achieved an elevated threshold of 0.6 and above. For behavioural engagement, the number of items was reduced from ten to four remaining with (BE4, BE7, BE8, BE9 and BE10). Lastly, regarding physical engagement, all 10 items initially passed the EFA, but only three remained after CFA (PE1, PE2 and PE3). The figure below ( Figure 3) illustrates the correlations between each observed variable (item) and its corresponding latent construct (Women’s engagement in decision making).

Figure 3. The latent construct Women engagement in decision making with its indicators; Behavioural engagement (BE) with its observed items (BE4, BE7, BE8, BE9, BE10), Emotional engagement (EE) with its observed items with (EE3, EE4, EE5, EE6, EE7, EE8) Cognitive engagement (CE) with its observed items (CE1, CE2, CE3, CE7, CE8) and Physical engagement (PE) with its observed items (CE1, PE2, PE3) with standardized loadings above 0.60, confirming convergent validity.

4.1 Descriptive statistics

The data presented in Table VI indicates that among 298 participants, logistics companies constitute the largest group at 50.7%, followed by industry associations at 35.6% and regulatory bodies at 13.8%. This diverse representation offers valuable insights into the dynamics of the sector and its regulatory environments, which are essential for understanding women’s roles in decision-making processes within logistics. In terms of ownership, local institutions lead at 38.7%, followed by government ownership at 22.5%, foreign ownership at 21.5%, and family ownership at 7.4%. This distribution highlights a significant influence from indigenous institutions and the potential for international collaboration. The duration of operation among institutions reveals that 39% have been established for over 15 years, indicating a level of stability and knowledge retention. Conversely, the presence of newer institutions (established within the last 1-5 years) suggests that the market remains open to innovation.

Regarding workforce size, 32.9% of respondents are from organisations with 50–100 employees, while 31.5% come from firms with fewer than 50 employees. Larger organisations, comprising 100–500 employees at 15.1% and those with over 500 employees at 20.5%, tend to exhibit more formal structures, which may influence the decision-making roles of women. The age distribution shows that 52.7% of respondents are under 35, reflecting a youthful workforce in logistics. Those aged 35-45 years account for 38.6%, which is crucial for evaluating women’s participation in decision-making. Only 8.7% are over 45, indicating limited representation of older professionals, which may pose challenges for women seeking advancement into leadership positions. Finally, the educational attainment data reveal that the majority of respondents (56.7%) hold a bachelor’s degree, illustrating that the logistics and related sectors are largely comprised of academically qualified professionals. Additionally, 22.8% of participants possess a master’s degree, suggesting a significant proportion of advanced professionals with the qualifications typically sought for leadership and policy-level roles.

4.2 Bivariate correlations

Table VII shows there is a small but significant negative correlation (Pearson correlation = -0.182, p = 0.002) between national regulations and social culture, indicating that as national regulations strengthen, certain aspects of traditional social culture may weaken; meanwhile, a moderate positive correlation (Pearson Correlation = 0.377, p = 0.000) suggests that stronger national regulations are associated with higher levels of women’s engagement in decision-making, whereas a small to moderate negative correlation (Pearson correlation = -0.224, p = 0.000) highlights that traditional social cultural norms negatively affect women’s ability to engage in decision-making processes. This overall dynamic suggests a tension between the push for formal modernisation through supportive national regulations and the resistance posed by prevailing cultural norms.

4.3 Structural Equation Modeling (SEM)

Structural Equation Modeling (SEM) is a sophisticated multivariate statistical technique that merges elements of factor analysis and multiple regression to evaluate structural relationships among both observed and latent variables. In this study, SEM was employed for hypothesis testing due to its strong ability to assess both direct and indirect effects while also accounting for measurement error. Its use was further validated by its effectiveness in conducting mediation analyses, which allows for a more in-depth exploration of the mechanisms underlying the relationships between variables. As noted by Hair et al. (2010), SEM enables the simultaneous estimation of multiple regression equations, providing a comprehensive and precise understanding of the intricate interrelationships among the constructs being studied.

The structural model ( Figure 4) was assessed to examine the hypothesised relationships among constructs using SmartPLS 4.0. Results indicated that the model achieved satisfactory explanatory power and predictive relevance, as evidenced by acceptable R², Q², and fit indices (SRMR = 0.075). Direct path analyses revealed that national-level factors exert a significant positive influence on women’s engagement in decision-making, supporting the hypothesised relationship (0.027, 0.003). Correspondingly, mediation analysis further confirmed that national regulations and laws significantly mediate the relationship between social culture and women’s engagement in decision-making (β = 0.174, t = 4.025, p < 0.001), indicating that progressive regulatory frameworks can offset restrictive cultural norms and promote gender inclusivity in leadership.

Figure 4. Structural model for the relationships among Social Culture, National Regulations and Laws, and Women’s Engagement in Decision-Making (WEDM). The figure presents the hypothesised paths tested using PLS-SEM, showing both direct and mediated relationships. Arrows indicate the direction and strength of influence among constructs. Path coefficients reveal that NRL → WEDM is significant, SC → WEDM is not significant, and SC → NRL → WEDM demonstrates a mediated effect.

4.3.1 Structural model evaluation results (coefficient of determination (r ²)/percentage of variance explained, Q squared)

Table VIII shows that the model demonstrates exceptional explanatory power, with national-level factors explained at 99.8% (R-square = 0.998, Adjusted R-square = 0.998) and women’s engagement in decision-making at 97.1% (R-square = 0.971, Adjusted R-square = 0.970), indicating strong predictive relevance and model accuracy, as evidenced by Q-squared values of 0.997 and 0.969, respectively, while the minimal difference between R-square and Adjusted R-square suggests robust performance without overfitting.

4.3.2 Model fit model estimates

Table IX shows that the model fit. It was evaluated through key indices from the PLS-SEM output, revealing a standardised root mean square residual (SRMR) of 0.075, which is below the recommended threshold of 0.08, indicating a good fit to the data; the negligible discrepancy between the saturated model (3.308) and estimated model (3.311) for d_ULS supports this conclusion, despite unavailable values for d_G and NFI, with an infinite Chi-square value suggesting that the model still adequately represents the empirical data for structural analysis.

4.3.3 Testing for relationships

The results from testing the direct paths between national-regulations and laws, social culture, and women’s engagement in decision-making are presented and elaborated below.

Testing for direct relationships

The results in Table X indicate that National regulations and laws (NRL) have a statistically significant positive effect on Women’s engagement in decision-making (WEDM), with a standardised path coefficient of 0.036, a t-value of 3.046, a p-value of 0.002, and a 95% bias-corrected confidence interval of 0.014 to 0.059, supporting hypothesis H1 and highlighting the importance of strong legal frameworks in enhancing women’s engagement in leadership roles within the logistics sector as illustrated in Figure 5.

Figure 5. Illustrates the standardized factor loadings linking NRL to WEDM, demonstrating a positive and significant association.

Arrows indicate the direction and strength of relationships between observed indicators and latent constructs. All retained loadings exceed 0.60, confirming construct validity and a good model fit.

The analysis in Table X and Figure 6 reveals that the path coefficient from Social Culture (SC) to Women’s Engagement in Decision-Making (WEDM) is β = 0.002 with a t-value of 0.073 and a p-value of 0.942, indicating no statistically significant effect, as evidenced by the 95% bias-corrected confidence interval ranging from -0.013 to 0.028, which includes zero; thus, hypothesis H2 is not supported, suggesting that social culture does not directly influence women’s engagement in decision-making, possibly indicating that its effects are mediated by other factors or exerted indirectly, warranting further exploration.

Figure 6. Illustrates the standardized factor loadings linking social culture (SC) to women’s engagement in decision making (WEDM), showing that social culture has no significant direct effect on women’s engagement in decision-making.

Arrows represent the direction and magnitude of factor loadings, all of which exceed 0.60, confirming construct reliability and validity.

Testing for indirect relationships

Social and National regulations and laws

The analysis in Table X and Figure 7 indicates that social culture has a statistically significant negative effect on national regulations and laws, with a path coefficient of β = -0.265, a T-value of 4.687, and a P-value of 0.000, suggesting that entrenched socio-cultural norms and traditional gender roles significantly hinder the development and implementation of policies supporting women’s engagement in decision-making, thereby illustrating how cultural resistance poses a substantial barrier to regulatory reform aimed at promoting gender equality.

Figure 7. Illustrates the negative and statistically significant path from SC to NRL, indicating that restrictive social-cultural norms weaken the development and implementation of gender-supportive regulations.

Arrows represent the standardized path coefficients and direction of influence between the constructs.

According to the mediation testing guidelines established by Baron and Kenny (1986), the process involves estimating three regression equations: first, regressing the mediator on the independent variable; second, regressing the dependent variable on the independent variable; and third, regressing the dependent variable on both the independent variable and the mediator. For mediation to be established, three conditions must be met: (i) the independent variable must significantly influence the mediator, where social culture acted as a negative influences on national regulations and laws; (ii) the independent variable must significantly impact the dependent variable where in this case, social culture did not influence women’s engagement in decision-making; and (iii) the mediator must significantly influence the dependent variable when included in the model, which, in this case, shows that national regulations and laws did influence women’s engagement in decision-making. Baron and Kenny suggest that the strongest evidence for mediation occurs when there is a significant indirect effect without a direct effect, a scenario they define as full mediation. Conversely, when both the indirect and direct effects are significant, it reflects partial mediation. In this context, the mediation variable (national regulations and laws) demonstrates full mediation between social culture and women’s engagement in decision-making.

The findings in Table X and Figure 8 indicate that social culture indirectly affects women’s engagement in decision-making through national regulations supporting Hypothesis 3 as evidenced by a statistically significant standardised indirect effect (β = -0.008, t-value = 2.070, p-value = 0.039). This finding suggests that social culture influences women’s engagement in decision-making indirectly by shaping national regulations. Specifically, restrictive social-cultural norms weaken regulatory frameworks, which in turn diminish opportunities for women’s participation in leadership roles. While the effect size is small (β = -0.008), it is statistically meaningful and highlights the importance of addressing cultural barriers to strengthen policy effectiveness and promote gender equity.

Figure 8. Presents the mediated pathway where SC indirectly influences WEDM through NRL.

Arrows indicate the direction and strength of standardized path coefficients, confirming a significant indirect effect (β = –0.008, p = 0.039). This model demonstrates full mediation, where supportive national regulations offset the restrictive impact of social-cultural norms on women’s leadership engagement.

5.1 Implications to the study

Practical implications

The findings elucidate significant practical implications for policymakers, organisational leaders, and development practitioners who are striving to enhance women’s engagement in decision-making within the logistics sector and other male-dominated industries.

Firstly, the demonstrable positive impact of national regulations and laws on women’s engagement in decision-making underscores the imperative for robust legal frameworks that explicitly advocate for gender equality. It is critical that policymakers prioritise the design, implementation, and stringent enforcement of gender-responsive legislation, including anti-discrimination policies, mandates for equal opportunities, maternity protections, and affirmative action regulations. Nonetheless, the modest magnitude of the observed effect suggests that legislation in isolation is insufficient; hence, regulatory reforms must be bolstered by systematic monitoring, enforcement mechanisms, and institutional accountability structures to ensure that they translate into actionable practices rather than remaining mere symbolic gestures.

Secondly, the non-significant direct effect of social culture indicates that institutional mechanisms can serve as effective buffers against the potentially constraining influence of traditional gender norms. For organisations operating within the logistics sector, this finding accentuates the necessity of embedding gender equity into formal organisational policies, recruitment and promotion practices, leadership development programs, and performance evaluation systems. Structured interventions such as gender-sensitive leadership training, transparent career advancement pathways, and anti-bias hiring practices can serve to mitigate subtle cultural biases that might otherwise obstruct women’s engagement in leadership roles.

Thirdly, the results advocate for the implementation of integrated, multi-level strategies. Legal reforms should be complemented by organisational change initiatives and broader societal efforts directed at shifting cultural attitudes regarding gender and leadership. Stakeholders, including governmental bodies, logistics firms, industry associations, and civil society organisations, are urged to collaborate in promoting public awareness campaigns, educating communities about gender equality, and advocating for cultural transformations that normalise women’s participation in leadership roles across all sectors.

Lastly, considering the indirect influence of social culture, there exists a compelling rationale for institutions to conduct periodic gender audits to evaluate the latent barriers that women may encounter in decision-making structures, despite the presence of formal legal protections. These internal assessments can guide ongoing improvements aimed at fostering genuinely inclusive environments.

In summation, these findings advocate for a holistic approach that synergises legal frameworks, institutional practices, cultural shifts, and individual empowerment strategies to sustainably advance women’s leadership and decision-making engagement, particularly within sectors where historical gender disparities have been entrenched.

Theoretical implications

The findings of this investigation bear significant theoretical ramifications, particularly when interpreted through the frameworks of Institutional Theory and Gender Schema Theory.

Primarily, the evidence supporting the influence of national regulations and laws on women’s engagement in decision-making substantiates pivotal propositions of Institutional Theory. As posited by DiMaggio and Powell (1983), formal institutions exert influence on organisational behaviour and social practices through mechanisms including coercive, mimetic, and normative pressures. In this context, the substantial positive correlation between national legal frameworks and women’s leadership involvement underscores the role of coercive mechanisms such as gender equality legislation and anti-discrimination statutes in prompting organisations, particularly in traditionally male-dominated fields like logistics, to embrace more inclusive leadership structures. This finding corroborates the assertion that formal institutional arrangements serve as potent tools for disrupting entrenched gender hierarchies and fostering normative transformations towards gender equity (Scott, 2014). Nevertheless, the observed modest effect size indicates that while institutional pressures are instrumental, they necessitate reinforcement by complementary cultural and organisational dynamics to effectuate substantial and lasting change. This nuance is consistent with an extended perspective in neo-institutional Theory, acknowledging that institutional reforms often require broader societal backing and internal organisational adaptations to fully manifest in practice (Greenwood et al., 2017).

Second, the non-significant direct effect of social culture on women’s decision-making engagement provides critical insights through the lens of Gender Schema Theory. Proposed by Bem (1981), Gender Schema Theory asserts that societal norms and cultural beliefs regarding gender roles establish cognitive structures or schemas that shape individuals’ perceptions and behaviours towards gender-appropriate roles. Historically, social culture has been regarded as a significant determinant of women’s leadership opportunities, with prevailing gender schemas reinforcing male dominance in decision-making domains. However, the findings suggest that the direct effect of social culture may be attenuated or mediated by the presence of robust formal institutional mechanisms, such as comprehensive legal frameworks and organisational policies. This observation aligns with evolving perspectives within gender theory that highlight the malleability of gender schemas and their potential disruption in the presence of formal structures and progressive interventions (Ridgeway & Correll, 2004).

5.2 Study contributions

This study addresses a critical research gap by examining the factors influencing women’s engagement in decision-making, specifically within the logistics sector, which is traditionally male-dominated and under-researched regarding gender equity issues. By demonstrating the significant positive role of national regulations and laws (NRL) and the non-significant direct effect of social culture (SC) on women’s engagement, this research provides context-specific empirical evidence, particularly relevant to developing countries like Uganda.

Utilising Institutional Theory and Gender Schema Theory, this study extends theoretical debates by illustrating how formal institutional structures, such as national laws and regulations, can act as transformative forces that regulate behaviour and reshape cognitive gender schemas. The finding that social culture does not directly influence women’s leadership engagement but may operate through institutional mechanisms supports evolving theories that advocate for multi-level models incorporating both institutional and cultural dimensions to understand gender dynamics. This nuanced understanding advances the intersection between institutional frameworks and gender cognitive structures, enriching both feminist institutionalism and gender leadership theories.

From an empirical perspective, this study validates the argument that strong legal frameworks are necessary but insufficient for achieving gender equity. The modest path coefficient suggests that legal reforms must be complemented by organisational and cultural initiatives to provoke substantial change. Additionally, this research introduces evidence that in highly regulated sectors, formal organisational policies can buffer the direct impact of traditional social culture on women’s leadership opportunities, inspiring future empirical inquiries into sector-specific dynamics.

Employing a quantitative approach with structural equation modelling (SEM) to analyse direct relationships and mediation pathways demonstrates the utility of sophisticated statistical techniques in understanding complex gender-related dynamics within organisations. The rigorous testing of hypotheses using bootstrapped confidence intervals, bias correction, and multi-path analysis contributes to improving methodological standards for future gender studies in leadership research.

5.3 Key takeaways/recommendations for practice and policy

Governments and regulatory bodies must not only enact gender equality measures such as anti-discrimination laws and gender quotas but also ensure that these laws are rigorously enforced and monitored, thereby translating legal rights into tangible benefits. In addition to compliance with legal requirements, organisations should foster cultural change by implementing gender sensitivity training, anti-bias initiatives, and mentorship programs to effectively dismantle informal barriers that hinder women’s advancement.

Collaborative partnerships among government entities, the private sector, academia, and civil society organisations are essential for advocating gender-inclusive reforms, sharing best practices, and maintaining momentum for change Institutions ought to conduct gender audits and impact assessments to monitor progress, identify ongoing barriers, and refine interventions aimed at promoting women’s leadership in the logistics sector.

5.4 Limitations of the study and directions for future research

The study concentrated specifically on the logistics sector in Uganda, a developing country context. While this provides valuable insights relevant to the sector, it also restricts the generalizability of the findings to other industries. Gender dynamics can vary significantly across sectors due to differing organisational cultures and regulatory environments. Future research should consider examining multiple sectors, such as finance, healthcare, education, and technology and diverse geographical contexts to assess how national regulations and cultural factors influence women’s leadership engagement in various settings.

Additionally, the use of a cross-sectional research design limits the ability to establish causality. The identified relationships between national regulations, social culture, and women’s engagement are associative rather than inherently causal. To gain a deeper understanding of causal relationships and the evolving impact of social culture and institutional reforms, future studies should employ longitudinal designs that track changes over time.

The data for both independent and dependent variables were collected using the same self-reported survey instrument within a quantitative framework, which raises the possibility of common method bias, even though efforts were made to mitigate this issue. Future research could benefit from mixed-methods approaches to examine how social culture evolves and how women navigate or challenge cultural barriers in their leadership journeys.

Moreover, the study focused on national regulations and laws as mediators; therefore, further research should investigate additional mediators such as organisational climate and leadership development programs and moderators, including individual resilience and industry gender composition. These factors could clarify or strengthen the relationship between national regulations, culture, and women’s engagement in decision-making.

Author contributions statement

Following the Contributor Roles Taxonomy (CRediT), the authors made the following specific contributions to this work: Nantongo Nabiira: Conceptualization; Methodology; Data curation; Formal analysis; Investigation; Visualization; Writing – original draft; Writing – review & editing; Project administration; Joseph M. Ntayi: Supervision; Validation; Writing – review & editing; Sheila Namagembe: Supervision; Methodological guidance; Writing – review & editing and Marcia Mkansi: Supervision; Conceptual refinement; Writing – review & editing. All authors have read and approved the final manuscript and agree to be accountable for all aspects of the work.