Tribal affiliations, political and government positions of directors and earnings quality in Kenya

This study examines the association between the informal and indigenous characteristics of company directors—specifically, their tribal affiliations—and the accounting quality of firms in Kenya, using earnings quality as a proxy. The significance of accounting quality, as assessed through earnings quality, is well-established in the literature (Dechow and Schrand, 2004). Prior international accounting research has examined the relationship between formal institutional features (Cai et al., 2014) and, to a lesser extent, informal ^[1] institutional features of company directors and firms' accounting quality (Nabar et al., 2007; Leventis et al., 2024).

Previous studies have primarily focused on broad elements such as national culture and religion affecting accounting quality (Callen et al., 2011; Kanagaretnam et al., 2011; Hussein et al., 2019; Tee and Rassiah, 2020; Alsaadi, 2021). However, these elements may not fully capture the essence and subtleties of informal institutions specific to a country (Leventis et al., 2024). In contrast, this study investigates tribal affiliations—an informal institutional characteristic indigenous to a country, Kenya—to assess their influence on accounting quality.

Prior management studies have observed the influence of kinship on business decisions (Smith et al., 2012; Zoogah, 2016). However, limited attention has been given to how such influences affect accounting quality. This research seeks to fill that gap by providing insights into the relationship between tribal affiliations and accounting quality in the Kenyan context. Kenya was selected as the research context due to the pervasive role of tribal affiliations in shaping both political dynamics since independence (Kasara, 2013) and corporate practices, particularly in the appointment of company directors, which frequently reflects tribal considerations (Ireri, 2016).

In politically entrenched countries, government policies often fail to serve all citizens equitably, particularly when such entrenchment is rooted in social class, religion, or tribal affiliations (Saint-Paul et al., 2016). Extant literature informs that politically associated businesses use tribal affiliations to manipulate government policies in their favour (Liedong, 2022). When such manipulations result in improved firm performance (Tawiah et al., 2022), they may reduce the perceived need for transparency. Conversely, when they are associated with poor performance, they may create incentives to manipulate financial reporting to conceal underperformance and deliberately obscure transparency. This study addresses this concern by examining the influence of directors' tribal affiliations on the earnings quality of firms in Kenya, after controlling for their political and governmental connections. Specifically, we raise the following research question: Do directors with tribal affiliations influence the accounting quality of firms listed on the Nairobi stock exchange?

In high power distance contexts like Kenya, directors with tribal affiliations who also hold or have held political or public office may face heightened scrutiny discouraging self-serving behaviour and promoting stakeholder accountability (Bratton and Kimenyi, 2008; Lankeu and Maket, 2012). Their public exposure and professional training may reinforce ethical governance. However, such dual roles may also foster complacency, weakening board oversight and financial reporting quality. This risk is amplified in high power distance societies (Hofstede, 1980), like Kenya, where authority is rarely questioned (Hofstede, 1980). Accordingly, we pose our second research question: Do politically affiliated tribal directors who hold or have held public office influence accounting quality differently from their counterparts who have not held such positions?

To answer these questions, the tribal identities, and political and government associations of 870 directors were hand-collected using sources that described their profiles. We measured these features at the director level. We find that tribal affiliations have an adverse relationship with accounting quality, suggesting that tribal kinships of directors influence managerial decisions unfavourably for corporate information users. However, this is partly reversed when directors with tribal affiliations are in or have held public offices, such as parliamentary membership or have held or are holding government offices. Hence, this study contributes to the overall discussion of the economic influences of informal institutions, such as the director's indigeneity, on the quality of corporate reporting.

Building on prior research into informal institutions and accounting outcomes, this study offers four distinct contributions. First, while earlier work linked tribalism to broad firm performance indicators such as Tobin's Q, ROA, or Net Profits (Tawiah et al., 2022; Zoogah, 2016), this study focuses on the relationship between tribal affiliations and accounting quality, using accrual-based earnings management as a proxy. Second, rather than limiting analysis to the tribal identity of the CEO or chairperson (Tawiah et al., 2022), or relying on small interview samples (Liedong, 2022), this study hand-collects data on all board members' tribal affiliations from multiple non-annual report sources. Third, it contributes theoretically by extending agency theory through the lens of ethical tribal values, challenging the stereotypical negative view that tribally affiliated managers always act in self-interest (Bosse and Phillips, 2016). Finally, the study considers external moderating factors, particularly directors' parliamentary or governmental positions, to explain variance in the effect of tribal affiliations on accounting quality.

The remainder of this paper is organised as follows. Section 2 explains the history and relationship of tribal affiliations in Kenyan corporate boards and specifies the hypotheses. Section 3 describes the data and research methods, and Section 4 reports and discusses the results. Section 5 draws conclusions, identifies the implications, and discusses the study's limitations.

A tribe is a social category of people who identify with one another based on shared history, ancestry, ethnicity, religion, language or culture, or a mix of these features (Liedong, 2022), and they are bound by traditions that call for certain obligations from each member of the tribe (Bates, 1974). Within the tribes, some emerge superior by dominating the people of other tribes, motivated by the desire for security, social well-being (Tamarkin, 1973), and economic benefits (Tawiah et al., 2022).

Although tribal affiliations existed in Kenya from prehistoric times, the rivalry amongst tribes intensified after the British colonisation in 1888. To subjugate the population, the colonial rulers divided them into the privileged/ “ruling” tribes and the non-privileged/ “ruled” tribes (Mizuno and Okazawa, 2009). They then expropriated the resources of the “ruled” tribes through forced acquisitions and reallocations of their tribal lands by deploying and empowering the “ruling” tribes (Gatheru, 2005). The resentment of the oppressed tribes eventually led to the Mau Mau insurrection between 1952 and 1957, championed by the Kikuyu tribe (Ogot and Ochieng, 1995) culminating in Kenya's independence in 1963.

In the post-independence period, the Kikuyus emerged as the dominant tribe in politics, government, and business (Meisler, 1970). This exacerbated the tribal divisions and rivalries among the other tribes that comprise Kenya's population (Ferree et al., 2021). Eventually, presidents were elected, and governments were formed along tribal lines with presidents pursuing policies favouring their ethnic and tribal groups to exclude others, further aggravating tribal rivalries (Burgess et al., 2015).

Two competing tribes currently dominate Kenya's politics: the Kikuyu and the Luo, comprising 20% and 10% of the population, respectively. The Kikuyu, in political alliance with the Kalenjin, has ruled the country since its independence. All five presidents since the country's independence in 1963 [Jomo Kenyatta – Kikuyu (1963–1978), Daniel Arap Moi – Kalenjin (1978–2002), Mwai Kibaki - Kikuyu (2002–2012), and Uhuru Kenyatta – Kikuyu (son of Jomo Kenyatta, 2013–2022), William Ruto - Kalenjin (September 2022 to date)] have been from either the Kikuyu or Kalenjin tribes (Stiftung, 2018). The Luo have long criticised the polarised policies of successive governments and remain a strong opposition force.

Kenyans, based on their tribal affiliations, uphold moral codes, stemming from their belief in a super God, respect for ancestors, submission to good leadership, respect for the institution of marriage, and respect for individual property rights (Kenyatta, 1937; Wandibba, 2004). Kenyans prefer to collaborate with people who share these values (Meisler, 1970) and hence, the role played by tribes in terms of co-ethnic favouritism or antipathy is strong. (Bratton and Kimenyi, 2008). Tribal moral codes, therefore, influence how directors manage businesses, discharge accountabilities, and disseminate information.

The accounting quality of firms is shaped by both formal and informal institutional factors (Leventis et al., 2024). Formal institutions—including the rule of law, democratic governance, and protection of private property—serve as fundamental mechanisms for ensuring high-quality financial reporting. However, in developing economies, the effectiveness of these institutions is often limited, thereby increasing the influence of informal institutions such as cultural norms, religious beliefs, ethnic identities, and tribal affiliations (Smith et al., 2012). In such institutional environments, business decisions are frequently driven by localised social customs rather than rational, calculative approaches (Agyemang et al., 2021). Moreover, the prevalence of tribal affiliations has been found to negatively affect financial reporting quality, as evidenced by Baatwah et al. (2023).

Research examining the effects of formal and informal institutions on accounting quality has frequently relied on agency theory and resource dependence theory. Agency theory suggests that managers, when given the opportunity, may act opportunistically, using discretion to manage earnings in ways that serve their interests (Healy and Wahlen, 1999). However, a more nuanced view of agency theory (Bosse and Phillips, 2016) acknowledges that not all managers act selfishly, as their actions may be guided by social norms such as fairness and reciprocity (Chen et al., 2018).

In African economies, where market structures and regulatory frameworks remain underdeveloped (Peng, 2003), firms often rely on informal institutional mechanisms, such as tribal networks, to navigate operational challenges and enhance their prospects for success. Resource dependence theory asserts that an organisation's survival is contingent upon its capacity to interact with and obtain critical resources from its external environment (Pfeffer and Salancik, 1978). Empirical research focussing on the African context indicates that high levels of tribalism are strongly linked to in-group favouritism and nepotism, which, in turn, contribute to suboptimal resource allocation (Hearn et al., 2024). Consequently, in developing economies such as Kenya, where resource distribution decisions are frequently shaped by tribal affiliations, establishing strategic connections with the politically dominant tribe becomes a crucial determinant of business success (Aburime, 2009).

The research questions formulated in this study are grounded in both agency theory and resource dependence theory. The first research question draws upon agency theory, which posits that managers with tribal affiliations may uphold tribal moral values and act in the best interests of principals, stakeholders, and shareholders. Empirical evidence within the African context supports this notion, demonstrating that the appointment of indigenous directors has a positive impact on both financial and environmental performance (Tawiah et al., 2024). From a resource-dependence perspective, it is argued that directors may strategically utilise their tribal affiliations to influence government policies in their favour, secure regulatory approvals, and gain access to critical resources. While these actions may improve firm performance, they can also reduce the need for transparency or, in some cases, enable accounting manipulations to gain government benefits.

From the agency-theory perspective, the second research question considers whether public scrutiny of politically and governmentally connected directors might deter them from acting against stakeholder interests. Given that these officials are typically trained professionals who are familiar with ethical standards and concerned about their reputation, they may be less likely to act against stakeholder interests. From a resource-dependence perspective, political and government connections could either diminish the need for high-quality accounting due to improved firm performance or increase the likelihood of accounting manipulations. Additionally, in a country like Kenya, characterised by high power distance, directors with higher status may face less scrutiny, leading to lax financial reporting practices.

Extant research indicates that firms in developed markets, where formal institutional arrangements are stronger, have fewer opportunities to manipulate earnings (Cai et al., 2014). In contrast, firms in countries with weaker corporate governance, less robust regulations, lower levels of minority shareholder protection (Djankov et al., 2008), higher ownership concentration, and the influence of informal institutions native to the business environment (Claessens and Yurtoglu, 2013) are more prone to earnings manipulation.

In countries with underdeveloped market and regulatory institutions, institutional voids often result in weak enforcement and oversight mechanisms (Peng, 2003). In such contexts, firms frequently adopt non-market strategies such as political clientelism, tribal affiliations, and corruption, to ensure access to critical resources. In ethnically divided nations, politico-tribal connections serve as key enablers for firms to secure favourable treatment. For instance, in Nigeria, Tawiah et al. (2022) finds that political connections enhance firm value only when tribal alignment exists with the ruling party, while non-alignment results in diminished firm performance. Noticeably, tribal affiliation can outweigh political connections, even when the associated political party is out of power, the CEO's tribal identity continues to yield performance benefits. Liedong (2022) draws similar conclusions in the Ghanaian context, arguing that tribal alignment enhances firm prospects, whereas misalignments become liabilities.

The implications of politico-tribal connections for corporate reporting are nuanced. While some studies report improvements in non-financial disclosure quality (Osemeke and Osemeke, 2017; Shaheen and Luo, 2023) others highlight adverse effects. These include political rent-seeking through donations, which may incentivise earnings manipulation and degrade financial reporting quality (Robinson and Verdier, 2013). Despite tribal conflicts constraining competition and productivity (Oyedele and Firat, 2018), empirical studies on their impact on accounting quality remain scarce.

In Kenya, traditional loyalty wields significant influence, often imposing barriers to trade (Oyedele and Firat, 2018), impacting productivity, sales, and net profits . Consequently, corruption tends to permeate through inequitable resource allocations and unfair competitive advantages to a select few, facilitated by bribery of government officials and tribal leaders (Ufere et al., 2020). In this context, a firm's future viability hinges on its capacity to secure essential resources from the external environment (Pfeffer and Salancik, 1978), prompting firms to obscure their management quality and the accuracy of reported earnings (Athanasouli and Goujard, 2015). It is, therefore, reasonable to argue that cultural constructs such as tribal values would adversely impact accounting choices in countries like Kenya, where tribal identities play a strong role in business decisions. Hence, we posit:

Extant literature points out that politically associated firms have easier access to equity and debt markets (Johnson and Mitton, 2003); pay lower taxes (Faccio, 2010); gain regulatory protection (Kroszner and Stratmann, 1998); and have ready access to contracts, both domestically (Goldman et al., 2009) and abroad (Faccio, 2010; Chaney et al., 2011). Hence, from a resource-dependence perspective, directors belonging to the tribe in power have a diminished need to manage earnings to acquire resources.

In African countries, where national politics are dominated by tribal belief systems (Michalopoulos and Papaioannou, 2013), political parties are formed, and political appointments are made along tribal lines (Abbink, 2011). Directors' ability to exert political influence depends on whether they belong to the same tribe as those in power (Nyambura and Hull, 2008). For example, corporate executives who share a tribal kinship with those in power can successfully lobby for political decisions (Beugré and Offodile, 2001).

Political associations are forged by firms' directors through affiliations with current members of the parliament (MPs) or ex-members of the parliament (ex-MPs). Political networks, however, promote clientelism, favouring businesses in exchange for votes or monetary benefits (Robinson and Verdier, 2013). Hence, political associations have a “rent,” which firms pay through political donations or by providing board seats to politically associated personalities. When such “rents” need to be hidden from the books of accounts, firms can resort to earnings management activities, resulting in poor earnings quality.

Alternatively, we argue that tribally affiliated directors who hold or have held public office, such as parliamentary membership, will be aware of the consequences of adverse public scrutiny. There is a growing tendency in Kenya to scrutinise the actions of public officials, particularly the parliamentary members (Hope, 2015). Such public scrutiny is likely to mitigate the tendencies of parliamentary-affiliated directors to associate with poor corporate governance and questionable accounting practices. This may also be reinforced by the moral and ethical values enshrined in tribal values. Likewise, we hypothesise that:

The appointment of current or former government officials to corporate boards—often termed the “revolving door”—is a mechanism through which firms derive the benefits of political connections. Lester et al. (2008) report that by 2008, 53% of corporate boards in the U.S. included such individuals. These appointments are believed to confer four main advantages: strategic advice, enhanced communication channels with policymakers, access to government resources, and increased board legitimacy. However, the empirical evidence on their effects is mixed and highly dependent on the performance metrices chosen (e.g. ROA, stock returns, government contracts). While some studies report favourable outcomes—such as higher market value and public contract acquisition, others show null or even adverse effects (Nehls et al., 2024). For instance, in Indonesia, such appointments are linked to lower stock price synchronicity, suggesting better firm-specific information (Fatawijaya et al., 2025). In China, the benefits persist only while political ties remain intact (Chen et al., 2020).

In Kenya, politically connected directors, especially those sharing tribal affiliations with ruling elites, secure easier access to state resources (Beugré and Offodile, 2001). We argue that such access diminishes incentives for earnings manipulation. Moreover, these directors are often better educated, economically secure, and guided by tribal values that emphasise integrity and ethical conduct. These attributes are likely to foster greater professionalism and reduce the likelihood of opportunistic behaviour. Accountability reforms in Kenya's public sector (Cohen, 1993; Hope, 2012) further reinforce these dynamics. We therefore hypothesise that:

This study draws on director-level data for 870 of the 960 directors from 64 of the 66 (97%) publicly listed firms on the Nairobi Securities Exchange (NSE) in 2019. We manually collected and analysed director profiles to identify tribal, political, and government affiliations. A single-year dataset was employed for three key reasons. First, our unit of analysis is the individual director, not the board or firm, given our focus on the association between a director's tribal affiliation and earnings quality. This negates the need for panel data to estimate average effects. Second, board composition remained largely unchanged post-2019 due to the pandemic, limiting additional variability. Third, extending the analysis beyond 2019 would introduce noise in the earnings quality measure due to COVID-19-induced macroeconomic shocks. Kenya’s 2020economic contraction (The Kenya Institute for Public Policy Research and Analysis, 2020), for instance, would confound any observed associations, undermining interpretability under crisis conditions.

To ensure validity, the sampled directors' affiliations were checked through multiple sources (company websites, annual reports, newspapers, and the Thomson Reuters database).

Tribal connections were coded using binary numbers. The variable MAJ_TRIBE refers to the Kikuyu and Kalenjin tribes, which greatly influence Kenyan politics and society. Directors belonging to either of these tribes were coded as “1”, and “0” otherwise. We also separately identified directors from tribes other than Kikuyu and Kalenjin, referring to them as minority tribes (MIN_TRIBE). Directors from European, Asian, and other nationalities were coded as zero.

Following Faccio (2006), directors' political associations were recognised if they were parliamentary members themselves or associated with current or ex-members of parliament (PARL). Given that a sitting MP will wield more political clout if he also holds significant parliamentary roles such as memberships in committees, we scored each director one (1) point for being or associated with an MP and one additional point each for each parliamentary position held. Directors who had no parliamentary connections scored zero.

Directors who are or were government officials themselves (GOVT) were scored with one point for their government appointments or government-affiliated agencies. As in previous cases, a score was given for each government position a director held. Directors who were not government officials scored zero.

We use earnings quality as the proxy for accounting quality. Earnings quality was measured using the Modified Jones model (Dechow et al., 1995), which estimates discretionary accruals that potentially impact earnings using the following equations.

where TA_ijt = total accruals in year t for firm i; Δ Current Assets ijt = change in current assets between years t and t-1; Δ Cash_ijt = change between years t and t-1; Δ Current Liabilities_ijt = change between years t and t-1; ΔSTD_ijt = change between years t and t-1 for debt included in current liabilities. ; and DAE_ijt = depreciation and amortisation expense in year t; NA_ijt = normal accruals in year t; ΔREV_ijt = change between years t and t-1revenues; ΔAR_ijt = change between years t and t-1 accounts receivables; PPE_ijt = gross fixed assets, plant, and equipment in year t; A_ijt − 1 = total assets in year t − 1; and εit = error term in year t. All the variables are scaled by total assets and the subscripts i, j, and t index firms, industries, and years, respectively. The absolute value of Ɛijt is the discretionary accruals, which is our accounting quality measure (EQUALi) (Subramanyam, 1996). The higher the (EQUAL_i), the lower the accounting quality. To control for kurtosis, EQUAL_i is transformed to its natural logarithmic (LnEQUAL_i) value. Prior studies on political associations, culture, family ownership and state ownership have used it as a proxy of earnings management or accounting quality (Hashmi et al., 2018).

To infer the direct influences of tribal affiliations and political associations on accounting quality, the following model was estimated with PARL and GOVT as political association variables and $MAJ_TRIBE$ and MIN_TRIBE as tribal affiliation variables:

Since accounting quality is inferred through discretionary accruals (DA), the presence of higher DA would evidence lower accounting quality (LnEQUAL). We use the natural log of the absolute value of DA to compute LnEQUAL because we are estimating the effects of our independent variables on the magnitude of earnings quality. While a positive DA indicates an income-increasing earnings management, a negative DA could be a reversal and an outcome of a previous upward earnings management or an income-decreasing earnings management like the big-bath. $EQUAL_dummy$ is a control for the direction of the DA. Estimation of coefficients $β1 and β2$ of model (2) will serve to accept or reject hypothesis H1. It is expected that the coefficients $β1′andβ2$ (tribal affiliations of directors) will take positive values to support the argument that tribal affiliation will negatively influence the dependent variable.

To ascertain whether political associations and government positions of directors moderate the negative impact of tribal affiliations on accounting quality (H2 and H3), we employed the Hayes Process I model regression. This process separated politically associated and government positions of directors from those who are not associated with studying the moderating effects (if any) of major versus minor tribal affiliations on the dependent variable, LnEQUAL.

All directors belonging to the major tribes need not have political or governmental connections and the same applies to directors belonging to minority tribes. Given that tribal values enshrine moral and ethical principles that are revered by Kenyan society, we believe that tribal affiliations moderate the impact political associations and government positions would have on accounting quality.

In Model (2), we also control for firm-specific, economic, and governance characteristics. We use return on assets (ROA) and leverage (LEV) to control for performance and financial risk impacts. Firms with higher profitability are less motivated to engage in earnings management (Kothari et al., 2005) and highly leveraged firms are scrutinised by lenders to deter them from managing earnings (Gross et al., 2016). Since growing firms need to attract additional resources, they would make better disclosure of earnings. Therefore, we control for market-to-book ratio (M2B) and asset growth (GROW). Similarly, the AGE of the company and auditors' reputation (BIG4) are deterrents to activities that could impair accounting quality. In line with previous studies, we expect all the above variables to be negatively associated with LnEQUAL.

We also controlled for the impacts of governance variables on LnEQUALi. The independent, executive memberships of directors in audit committees (AUDCOM) and the independent, non-executive memberships of directors in board nominating committees (NOMCOM) (Ahmad and Duellman, 2007), indicate the independence of directors in disapproving earnings management practices. Prior research has shown that boards with higher percentages of female directors have superior earnings quality (Gavious et al., 2012). Therefore, we expect all the above variables to be negatively associated with LnEQUAL.

We also controlled for directors' industry (IND_EXP) and financial (FIN_EXP) expertise (Menozzi et al., 2011). Since directors' expertise can be used to both deter or promote earnings manipulation, our expectation in this regard was neutral. Additionally, industry variations were also controlled. The detailed definitions of variables used in this study are presented in  Appendix 1.

The descriptive statistics for this study are presented in Table 1. The sample companies have ages ranging between 1 and 68 years, with most companies trading since the 1990s. The mean value of LnEQUAL is −3.58 with a standard deviation of 1.44.

Descriptive statistics confirm that the Kikuyu and Kalenjin tribes dominate Kenyan politics and businesses. Represented by the variable MAJ_TRIBE, on average, 39% of the directors sampled, belonged to either the Kikuyu or Kalenjin tribes. Directors from MIN_TRIBE made up 27% and the remainder came from European, Asian, and other origins.

The mean values of political association variables PARL (M = 0.47), and GOVT (M = 0.51) indicated that a greater portion of political associations amongst directors. The performance ratios ROA (M = 0.02, SD = 0.12) and Leverage (M = 3.59, SD = 4.99) corresponded well with previous studies relating to Kenya (Outa and Waweru, 2016). In addition, 94% of firms had Big 4 auditors (M = 0.94, SD = 0.24) and the control variables for the audit committee (M = 0.26, SD = 0.44) and board nominating committee (M = 0.25, SD = 0.44) suggested that the numbers of executive and non-executive directors in both committees were comparatively lower. Most companies had substantial industry (M = 0.59, SD = 0.49) and financial expertise (M = 0.55, SD = 0.50) and female directors (M = 0.21, SD = 41).

The bivariate Pearson correlation coefficients for the main variables are presented in Table 2. The dependent variable LnEQUAL significantly correlates with all the main test variables, except with GOVT and MIN_TRIBE. The positive and significant correlation indicates that tribal affiliations of majority tribes negatively influence accounting quality (r = 0.083). The MAJ_TRIBE variable is also significantly correlated with PARL (r = 0.106).

The lack of a significant correlation between LnEQUAL and GOVT suggests that there are limits to the political benefits received through this channel, despite GOVT being significantly correlated with PARL (r = 0.395). The dependent variable's significant and positive association with (PARL), is consistent with prior studies that found a negative impact of political associations on earnings quality (Faccio, 2010; Chaney et al., 2011).

Control variables correlate well with the main variables to justify their inclusion in the models, and none of the correlation coefficients has outliers or high VIFs that cause concern for multicollinearity.

The results of the multivariate regression analysis (Model 2) are presented in Table 3.

As hypothesised in H1, in Model (2), both MAJ_TRIBE and MIN_TRIBE returned positive and significant coefficients (β = 0.397, p < 0.01; β = 0.286, p < 0.1, respectively) indicating their negative influences on accounting quality. The economic impact of this finding suggests that an appointment of a director belonging to the Kikuyu and/or Kalenjin tribes reduces accounting quality by 39.7% while a director belonging to a tribe other than Kikuyu or Kalenjin reduces accounting quality by 28.6%, albeit at a lower level of significance.

The political affiliations of directors, particularly through current and former members of parliament (PARL), show a significant negative impact on earnings, as indicated by the positive and significant unstandardised coefficient (β = 0.182, p < 0.01). This finding suggests that accounting quality declines by 18.2% for each director with political ties to parliamentarians, supporting hypothesis H2. A possible explanation for this outcome is that politically connected firms benefit from implicit guarantees for resources, reducing their incentive to maintain high-quality financial reporting for capital markets. This diminished motivation to ensure transparency may contribute to the observed decline in accounting quality (Harymawan and Nowland, 2016).

GOVT returned a significant negative coefficient, indicating an improvement in accounting quality. This may be because firms with government ties have easier access to resources, reducing the need for earnings manipulation. Additionally, from an agency theory perspective, government officials, well-versed in laws and regulations, may enforce stricter oversight, enhancing financial reporting vigilance. In Kenya, the Companies Act 2015replaced the 1948 Act, strengthening accounting and auditing regulations (IFAC, 2019). The mandatory IFRS adoption and revised corporate governance code in 2015 may have further encouraged government-connected directors to enhance financial scrutiny (Outa and Waweru, 2016).

The regression results for the control variables ROA and Leverage are statistically significant and align with the hypothesised directions (β = −2.584, p < 0.01; β = −0.078, p < 0.01, respectively). In this study, M2B and GROWTH were employed as proxies for market mechanisms, given their role in moderating earnings management (Liedong et al., 2020). Both variables yielded statistically significant coefficients (β = −0.082, p < 0.01; β = 1.153, p < 0.05, respectively).

The positive relationship between M2B and LnEQUAL is consistent with theoretical expectations and aligns with prior findings (Abdul Wahab et al., 2020). However, the negative coefficient for GROWTH contradicts our initial hypothesis. A plausible explanation is that GROWTH, defined as the increase in net assets representing long-term expansion, may not align with LnEQUAL, which captures a more short-term perspective.

Additionally, AGE and Audit Quality produced statistically significant coefficients with the expected signs. Among the governance variables, only Gender exhibited a significant effect, though in the opposite direction to expectations (β = 0.242, p < 0.05). This outcome may be attributed to the characteristics of emerging capital markets, where female directors are frequently close family members of controlling shareholders and may lack the authority to assert independent perspectives (Muttakin et al., 2015).

To test whether directors' political associations or government positions moderate (H2 and H3) the relationships between tribal affiliation and accounting quality (LnEQUAL), The following interaction model was employed.

We used Hayes's Process 1 to identify the moderating effects of PARL and GOVT and the results of this regression are presented in Table 4.

Results are shown in Table 4 Panel A and Panel B. Panel A refers to the regression results for LnEQUAL as the outcome variable, TRIBE as the independent variable, PARL as the moderating variable and the interaction between TRIBE and PARL as an exogenous variable. The variable TRIBE is denoted by one (1) for major tribal affiliations of directors and zero (0) for affiliation with minor tribes. The variable PARL is binary, denoted by one (1) for the presence of parliamentary connections and zero (0) otherwise. The model summary indicates that a 2.08% change in accounting quality is being accounted for by the exogenous variables: TRIBE, PARL, and the interactions between the. The interaction effect in Panel A (TRIBE*PARL), which is of primary interest, is negatively statistically significant (coefficient = −0.256; p = 0.0306). Hence, we conclude that parliamentary connections of directors significantly negatively moderate the impact tribal affiliations have on accounting quality. The magnitudes of effect sizes reported under the conditional effects of the focal predictor (TRIBE) at values of the moderator (PARL) suggest that the above “weakening” happens both, when directors have no parliamentary connections, and when directors have parliamentary connections evidenced by an effect size of 0.4024 with a p-value of 0.000 and 0.146 with a p-value of 0.081, respectively. Likewise, parliamentary connections of directors moderate the impact on accounting quality more when they are not affiliated with major tribes than when they are.

The findings confirm the moderating effects of directors' parliamentary connections on both with and without major tribal affiliations in support of hypothesis 2. As argued earlier, the political visibility of tribally affiliated directors (both major and minor) who are members of parliament, could provide a reputational incentive for them to support better corporate governance and, thereby, more transparent reporting practices. Additionally, political scrutiny of being public office holders would encourage better adherence to tribal values. Also, being a member of a major tribe and holding public office allows directors to access resources through contracts and private knowledge of government policies, removing the need to manage earnings to attract resources.

Data presented in Panel B of Table 4 relates to the regression results for LnEQUAL as the outcome variable, TRIBE as the independent variable, GOVT as the moderating variable, and the interaction between TRIBE and GOVT as an exogenous variable. The variable GOVT is binary, denoted by one (1) for the presence of governmental connections and zero (0) otherwise. Measurements of other variables are the same as those reported earlier. The model summary indicates that a 1.42% change in accounting quality is accounted for by the exogenous variables: GOVT, TRIBE and the interaction term. The interaction is statistically significant at 1% (β = -0.4629; p = 0.0025). Accordingly, we conclude that directors' government connections significantly negatively moderate the impact of tribal affiliations on accounting quality, which suggests that the negative effect of governmental connections on accounting quality (β = 0.3969; p-value = 0.0002) is weakened in the presence of Government connections. However, the conditional effects of the focal predictor (GOVT) at values of the moderator indicate that only the non-association with the major tribes is statistically significant as the value zero falls within the boundaries of LLCI (0.2214) and ULCI (0.5724). In other words, when major tribal affiliations are not present, government connections of directors negatively moderate accounting quality. This result cannot be compared with the scenario where major tribal affiliations are present due to the statistical insignificance of the effect-size.

The conditional effects of the predictor based on data presented in Panel B of Table 4 suggest that the accounting quality is lower when directors are neither associated with the government nor the Kikuyu or Kalenjin tribes. These results suggest that government connections moderate the negative impact on accounting quality when directors are affiliated with the minor tribes. Regarding directors from the major tribes and with or without government associations, no inferences can be made as this conditional effect is not statistically significant (p-value = 0.5468).

The data presented in Table 4 reaffirms the moderating effects of parliamentary and governmental directors' positions on accounting quality, especially when affiliated with minor tribes. The intensity of public scrutiny on persons holding public office appears to be higher when they belong to minority tribes than to majority tribes. We accept H2 in the context of both minority and majority tribal affiliations and H3 in the context of only minority tribal affiliations.

To assess the stability of the regression model we recomputed the dependent variable using the Jones (1991) model, which estimates discretionary accruals that potentially impact earnings. Prior studies on political associations, culture, family and state ownership have used this as a proxy for earnings management (Hashmi et al., 2018). The regression results were qualitatively like those reported in Table 3.

Additionally, we also used a subsample consisting of only directors who had major tribal affiliations to test whether the coefficients for major variables retain their directions and significance. This subsample contained 340 directors (39% of the total). The regression results were similar to those reported for the total sample.

This study aimed to investigate how tribal affiliations of directors influence the accounting quality of companies in Kenya, a nation with a history of tribalism influencing both politics and business. We examined three key independent variables: the tribal affiliations of directors (whether aligned with majority or minority tribes), their parliamentary affiliations, and their governmental positions. When analysed individually, the results indicate that both tribal and parliamentary affiliations of directors are associated with a decline in accounting quality, echoing the findings by Baatwah et al. (2023) in the context of GCC countries. Conversely, when directors hold government positions, accounting quality improves.

Additionally, we find that the adverse effect of tribal affiliations on accounting quality is mitigated when directors with tribal connections also hold parliamentary positions, with this moderating influence being more pronounced for directors from minority tribes. Similarly, the negative impact of tribal affiliations is further attenuated when directors hold governmental positions, again with a stronger mitigating effect observed among directors affiliated with minority tribes.

The findings of this study offer critical insights into theory, practice, and society. Specifically, it sheds light on the underexplored role of a non-market institution—namely, the tribal affiliations of directors—in shaping organisational resource dependency. These affiliations, while facilitating access to valuable resources, weaken the disciplinary mechanisms of agency relationships, thereby compromising accounting quality. The study thus advances theoretical understanding of the influence of informal institutions by highlighting the dual impact of tribal ties: they function as both enablers of resource mobilisation and as potential impediments to the effectiveness of accountability structures that underpin high-quality financial reporting in traditional agency frameworks.

Secondly, from a practical standpoint, the findings highlight a critical need for regulators and policymakers in developing economies to acknowledge the embedded influence of informal institutions like tribal networks on corporate governance and accounting practices. Such institutions are prevalent in countries with weak formal institutions and mitigate the effects of those institutions. Accordingly, tailored regulatory interventions and context-sensitive policy guidelines are essential to uphold the integrity of the agency relationship between managers and shareholders within these socio-culturally complex environments.

From a socio-cultural standpoint, non-market influences—particularly the tribal affiliations of directors—ought to be treated as integral components when evaluating the quality and credibility of accounting information. What might be perceived as deviations or irregularities from an Anglo-Saxon cultural lens may constitute legitimate and socially sanctioned forms of accountability within these jurisdictions. This study thus underscores the importance of contextualising accounting norms, especially in environments where tribal identity supersedes ethnic or national affiliations (Liedong, 2022). In such settings, entrenched social structures present formidable obstacles to the uncritical transplantation of Anglo-Saxon accountability models, necessitating culturally informed and context-sensitive interpretations of governance and reporting practices.

While this study provides important insights into the influence of tribal affiliations on accounting quality, its findings are derived from the Kenyan context and may therefore be subject to contextual limitations in terms of generalisability. Moreover, the analysis is based on data from a single year. However, given the comprehensive coverage of all firms listed on the Nairobi Securities Exchange (NSE) and the relative stability of directors' characteristics across time, the reliance on a single-year dataset is unlikely to materially undermine the validity or reliability of the study's conclusions.

Future research could explore whether mixed-tribe boards enhance accounting quality by reducing in-group bias and impacting accounting quality. Comparative studies across strong and weak institutional environments also could reveal how tribal affiliations interact with formal controls to influence the robustness and integrity of financial reporting.