A decade-long study of Ghana’s commercial banks has found that the strength of a bank’s board matters for audit quality, but only part of that story is written in the boardroom. According to research published in Discover Analytics, human capital efficiency—how productively a bank’s workforce converts staff costs into value—acts as a powerful moderator that can reverse or reinforce the links between board characteristics and the quality of external audits. The findings come from an analysis of twenty commercial banks operating in Ghana between 2014 and 2023, yielding 200 firm-year observations drawn from annual reports.
The research team, Samuel Anaba, Osman Issah and Mutala Zubeiru, set out to address a gap in the literature on Sub-Saharan African banking, where very little is known about the factors determining audit quality and the role of intangible organizational resources. Their motivation is grounded in a changed regulatory landscape: after the Bank of Ghana’s 2018 recapitalization directive, the number of licensed commercial banks fell from 36 to 23, making corporate governance and audit practices increasingly consequential for the sector’s stability. The study is, according to its authors, the first empirical test of the governance–audit quality relationship within Ghana’s banking sector.
The researchers measured audit quality using two proxies: the audit fee ratio, calculated as the percentage of audit fees against total revenue, and audit tenure, defined as the number of consecutive years the same audit firm has served the bank. Governance was captured through board size, board independence and board gender diversity, while human capital efficiency was computed using the human capital component of the VAIC framework—value added, defined as total revenue minus operating expenses, divided by total staff salaries and benefits. All continuous variables were winsorised at the first and 99th percentiles to limit the influence of extreme observations.
The baseline estimation used a fixed-effects estimator with bank-clustered robust standard errors, validated by a Hausman specification test, together with bank and year fixed effects. To address potential endogeneity arising from reverse causality and omitted variable bias, the team employed a two-step system GMM estimator as a robustness check, restricting lag depth to a maximum of two periods and collapsing the instrument set to avoid instrument proliferation. The authors acknowledge that with 20 banks observed over ten years, the panel sits at the boundary of standard recommendations for this estimator, a methodological caveat they treat explicitly.
The descriptive picture reveals a banking sector with boards averaging roughly nine members (mean 8.745, standard deviation 1.796, ranging from 5 to 13 directors). Board independence averaged 61.2 percent, spanning a wide range from 16.7 percent to 88.8 percent across institutions. Gender diversity on boards was low, averaging 18.4 percent—below the threshold prescribed by the Bank of Ghana’s 2018 Corporate Governance Directive—a figure the authors flag as relevant to interpreting the gender diversity results.
In the direct-effects models, board size showed a significant positive relationship with both audit quality proxies: the coefficient was 0.057 for the audit fee ratio and 0.059 for audit tenure, both significant at the five percent level. Board independence emerged as the strongest governance factor, with coefficients of 0.446 for the audit fee ratio (significant at the one percent level) and 0.541 for audit tenure (significant at the five percent level). The authors interpret these results through agency theory: larger and more independent boards monitor management more vigilantly and demand higher-quality external auditing.
Board gender diversity, by contrast, was negatively and significantly associated with both audit quality measures in the fixed-effects models, with coefficients of −0.815 for the audit fee ratio and −0.545 for audit tenure, both at the five percent level—a result that contradicted the study’s third hypothesis. The authors suggest this may reflect the relatively small proportion of women on Ghanaian bank boards, which sits below the critical mass that theory suggests is needed for diversity to enhance monitoring. Human capital efficiency itself was positively and significantly related to audit quality at the one percent level across all models, indicating it plays an essential role as an audit quality determinant in its own right.
The moderating results are the study’s central contribution. When interaction terms between governance variables and human capital efficiency were added, the adjusted R-squared rose from 41.3 percent to 42.1 percent for the audit fee ratio model and from 35.1 percent to 56.2 percent for the audit tenure model. All three interaction terms were positive and significant: board size multiplied by human capital efficiency yielded coefficients of 0.068 for fees and 0.027 for tenure; board independence multiplied by human capital efficiency yielded 0.072 and 0.045; and board gender diversity multiplied by human capital efficiency yielded 0.058 and 0.079, all at the five percent level. Notably, the positive interaction for gender diversity means that human capital efficiency transforms the negative direct link between board gender diversity and audit quality into a positive one.
The system GMM robustness checks broadly confirmed these patterns. Board size remained positively related to both proxies (coefficients of 0.023 and 0.067), board independence stayed positive, and the interaction terms for board size and board independence with human capital efficiency remained significant, with coefficients of 0.047 for board size on fees and 0.065 and 0.043 for board independence on fees and tenure respectively. Board gender diversity was insignificantly related to the audit quality proxies in the GMM models and did not support the third hypothesis. Diagnostic tests supported the instrument set: AR(2) statistics were insignificant in all four models (p-values between 0.265 and 0.297), Sargan–Hansen test p-values ranged from 0.258 to 0.297, and the instrument count was deliberately kept small at 25 to 28 relative to the 20-bank sample.
The authors draw policy implications for Ghana’s regulators, suggesting the Bank of Ghana’s Corporate Governance Directive could include minimum independent director representation of no less than 50 percent, that banks should invest in workforce efficiency to maximize audit quality gains from governance, and that authorities consider a voluntary target of 30 percent or more women’s board representation. They also note limitations: the sample of 20 licensed commercial banks limits generalization to other financial institutions, only the human capital component of intellectual capital was examined, and unexplored variables such as audit committee composition, CEO power and political connections may account for remaining variance in audit quality. The study’s theoretical contribution lies in extending agency theory with human capital theory and the resource-based view, showing that the governance–audit quality relationship depends on the intellectual capital a bank can bring to bear.
Subject of Research: The relationship between bank governance mechanisms, human capital efficiency, and audit quality in Ghanaian commercial banks
Article Title: Bank governance, human capital efficiency, and audit quality: evidence from commercial banks in Ghana
Article References: Anaba, S., Issah, O., & Zubeiru, M. (2026). Bank governance, human capital efficiency, and audit quality: evidence from commercial banks in Ghana. Discover Analytics, 4(1), Article 44. https://doi.org/10.1007/s44257-026-00095-0
Image Credits: AI Generated
DOI: 10.1007/s44257-026-00095-0
Keywords: audit quality, corporate governance, human capital efficiency, board of directors, Ghanaian banks, VAIC, system GMM, fixed effects, agency theory, board gender diversity, Bank of Ghana, Discover Analytics
News Source: Denise Maddox. (October 28, 2026). Human Capital Efficiency Shapes How Bank Boards Influence Audit Quality in Ghana. Scienmag.
