Institutional work, accountability and corruption: evidence from a social business organization in Bangladesh

Intra-firm accountability is a cornerstone of effective organizational governance and long-term sustainability, particularly within financial institutions. However, achieving and maintaining such accountability remains a significant challenge, especially in developing countries where corruption is deeply embedded in organizational culture. Recent studies have emphasized that ineffective accountability frameworks can exacerbate organizational vulnerabilities and hinder governance reforms (Lassou et al., 2024; Heredia et al., 2023). In social business organizations, these challenges are even more pronounced due to the high stakes of financial operations and regulatory compliance. In Bangladesh, social businesses face persistent challenges in combating corruption, which threatens their credibility and operational efficiency. Despite the implementation of various accountability mechanisms, intra-firm corruption continues to plague many organizations, highlighting the need for a more nuanced understanding of how organizational policies, structures and employee behaviors influence accountability practices.

This study is motivated by the growing global emphasis on organizational transparency and accountability, driven by increasing regulatory pressures and heightened public expectations for ethical governance. Recent works have highlighted the importance of adopting innovative governance strategies to meet these rising demands (Beretta et al., 2025; Heredia et al., 2023). The Grameen Bank (2017) in Bangladesh provides a unique and timely context for exploring intra-firm accountability and corruption prevention in a social business organization. Grameen Bank, often recognized for its innovative practices and social mission, offers a compelling case to investigate how internal organizational structures can resist corruption and promote accountability. Given the evolving regulatory landscape and the global focus on governance reforms, there is an urgent need to understand how targeted interventions can enhance intra-firm accountability in financial institutions. This research addresses that need by focusing on actionable insights and grounded empirical evidence, contributing to the literature on organizational change and accountability (Pillay and Kluvers, 2014; Azim and Kluvers, 2019). By exploring the intersection of accountability, anti-corruption and institutional change, the paper advances critical perspectives on governance practices within complex organizational contexts.

The Grameen Bank established by Professor Muhammad Yunus in 1983 following an experimental initiative in 1976, represents a pioneering model in microfinance aimed at alleviating rural poverty in Bangladesh, particularly among women (Bayulgen, 2008). Its innovative small-loan model transformed rural financial inclusion and inspired similar initiatives worldwide. This impact was recognized globally when Grameen Bank and Professor Yunus received the Nobel Peace Prize in 2006. Today, Grameen Bank serves predominantly rural women, who constitute approximately 95% of its clientele, and has consistently maintained a loan recovery rate of around 98%, underscoring its operational effectiveness (Hossain et al., 2017).

The bank’s lending practices incorporate flexibility through mechanisms such as low-interest housing loans, voluntary savings schemes, group-based lending, mandatory insurance and rescheduling options during crises. Grameen Bank offers four primary loan categories: general, housing, technology and collective loans. Technological integration has further strengthened its operations, with nearly all branches computerized (Grameen Bank, 2019). Despite its success in poverty alleviation and women’s empowerment, recent literature raises concerns about operational inefficiencies and transparency deficits that may foster corruption within organizational processes. This study addresses these issues by examining the factors that shape employees’ perceptions of corruption across Grameen Bank’s regional branches.

The primary aim of this research is to investigate the key factors influencing intra-firm accountability and to assess the effectiveness of various corruption-prevention mechanisms within Grameen Bank. Recent studies have emphasized the growing need for organizations to adopt accountability practices that respond to external institutional pressures (Isett et al., 2022; Leca et al., 2009; Aleksandrov et al., 2025). Specifically, this research examines how organizational structures, monitoring mechanisms and punitive measures enhance transparency and reduce corruption within social businesses. By assessing the impact of these internal controls alongside employee-specific factors such as ethical training and professional experience, the study aims to offer practical solutions that align with contemporary governance expectations (Beretta et al., 2025).

To delineate the theoretical scope of this study, we conceptualize accountability not as a separate construct to be measured, but as the set of organizational practices (Lawrence and Suddaby, 2006; Lawrence et al., 2009) through which institutional norms are enacted. These practices correspond to the forms of institutional work undertaken by organizational actors − such as enforcing rules, implementing preventive mechanisms and providing ethics training. Through boundary work, practice work and relational work, employees contribute to the functioning of accountability mechanisms within the organization (Hampel et al., 2017; Gidley and Palmer, 2021). The effectiveness of these mechanisms is captured empirically through perceived transparency, which serves as the outcome variable in our model. Transparency, therefore, reflects employees’ assessments of how well accountability practices operate in practice to reduce corruption. Accordingly, accountability functions conceptually as the pathway linking institutional work to employees’ perceptions of corruption, even though it is not modeled as a distinct statistical mediator.

Although accountability is not measured as a distinct construct, it is captured indirectly through employee perceptions of enforcement actions (boundary work), preventive procedures and training (practice work) and relational dynamics such as trust and collaboration (relational work). These dimensions are operationalized through institutional work practices (enforcement, preventive procedures, training), while transparency reflects the perceived effectiveness of accountability structures. To be precise, in this study, accountability is conceptualized as a set of organizational practices such as enforcement procedures, preventive mechanisms and ethical training through which institutional norms are enacted. We acknowledge that while accountability itself is not directly measured, the institutional work variables included represent the organizational mechanisms that operationalize accountability. In this context, perceived transparency, as the dependent variable, reflects the outcome of these mechanisms rather than their content. Consequently, transparency captures employees’ assessments of the effectiveness of accountability practices in mitigating corruption. Thus, accountability functions as a conceptual mediator between institutional work and corruption perceptions, even though it does not appear as a separate statistical construct.

Accordingly, the analysis explores how different forms of institutional work − boundary, practice and relational − shape employees’ perceptions of accountability and, in turn, influence their assessment of corruption. This approach enables an examination of the link between institutional work and corruption, mediated by accountability-related practices, even though accountability itself is not modeled as a separate dependent or independent variable. To address these concerns, this study is guided by two central research questions (RQs):

These RQs are theoretically grounded in the institutional work framework, which emphasizes the micro-level actions, practices and relational dynamics through which actors sustain, disrupt or transform institutions. By examining the mechanisms through which employees and organizational leaders at Grameen Bank engage in boundary work, practice work and relational work, this study aims to shed light on the institutional conditions that either enable or constrain intra-organizational accountability and anti-corruption outcomes. The first question emphasizes the role of agency and intentional actions by organizational members in creating and maintaining accountability systems, a core focus of institutional work (Lawrence and Suddaby, 2006; Zietsma and Lawrence, 2010; Ghafran and Yasmin, 2025). The second question focuses on identifying the specific forms of institutional work that help reduce corruption and promote accountability practices (Akanga, 2017; Smets and Jarzabkowski, 2013; Hampel et al., 2017).

This study also provides critical insights into how institutional work can be applied to understand how accountability mechanisms evolve and adapt to institutional demands, contributing to recent advances in the accounting and governance literature (Isett et al., 2022; Leca and Laguecir, 2023; Aleksandrov et al., 2025; Heredia et al., 2023). Institutional work provides a dynamic lens to understand how individuals and organizations actively shape, maintain and disrupt institutions through intentional actions and practices. Unlike traditional institutional theory, which primarily focuses on the structures and external pressures that shape organizational behavior, institutional work emphasizes the agency of actors within organizations (Lawrence and Suddaby, 2006; Leca et al., 2009). This perspective highlights how actors engage in boundary work, practice work and leadership activities to create, sustain or challenge institutional norms and practices (Zietsma and Lawrence, 2010; Kraatz, 2009). For example, in the context of social business organization, institutional work can reveal how organizational leaders and employees undertake efforts to establish accountability mechanisms and combat entrenched corruption (Suddaby and Viale, 2011; Willmott, 2011).

Recent studies have further expanded the concept by exploring how institutional complexity and relational models of institutional work impact organizational change (Smets and Jarzabkowski, 2013; Hampel et al., 2017). By adopting the institutional work framework, this study examines specific accountability-related practices undertaken by frontline employees, including enforcing rules, implementing preventive measures, delivering or receiving ethics and compliance training, as well as the discretionary efforts through which organizational members interpret, adapt and enact these mechanisms in daily operations. By analyzing how these concrete practices shape employees’ perceptions of integrity and organizational conduct, the study provides fresh insight into how institutional work drives the evolution of accountability structures within a social business context. To ensure consistency in terminology and alignment with literature, we clarify that Grameen Bank is considered a social business organization for this research. While Grameen Bank operates as a financial institution, its structure and mission of poverty alleviation, social development and social inclusion more closely align with the characteristics of a social business, particularly in the context of organizational accountability (Yunus et al., 2010). The study adopts a survey-based research methodology to collect primary data from employees of Grameen Bank. The survey was conducted over six months, targeting 250 employees across various levels of the organizational hierarchy. The survey instrument was informed by existing literature on organizational accountability and corruption (Bédécarrats et al., 2012; Luo, 2005), incorporating key indicators such as information flow, monitoring mechanisms and the organizational structure for corruption prevention. Responses were measured using a Likert scale and analyzed using probit and logit regression models to identify significant predictors of perceived transparency and corruption. This approach is consistent with recent empirical studies in accounting that explore governance and accountability through employee perceptions (Hauser, 2019; Wang et al., 2018).

The findings of this study illustrate that institutional work plays a crucial role in shaping intra-firm accountability practices. Strict monitoring mechanisms and punitive measures emerged as the most effective tools for reducing perceived corruption and enhancing transparency within social business organizations. These findings align with the growing body of research emphasizing the importance of boundary and practice work in maintaining accountability structures (Lawrence and Suddaby, 2006; Zietsma and Lawrence, 2010; Ghafran and Yasmin, 2025). Notably, ethics training and moral guidance showed limited impact on improving transparency, highlighting the complexity of addressing entrenched cultural practices within organizations (Willmott, 2011; Hampel et al., 2017). Employee experience also proved to be a significant factor, with more experienced employees reporting higher perceptions of corruption than their younger counterparts, indicating that long-standing exposure to organizational practices may shape perceptions of accountability in complex ways.

This study makes significant contributions to theory, practice and policy. By applying the institutional work framework, it extends the literature on intra-organizational accountability and corruption prevention, showing how micro-level practices − shaped through the everyday actions and decisions of organizational actors − both reinforce existing accountability norms (institutional stability) and enable the development of new practices or structures (institutional change). In doing so, the study highlights the dynamic interplay between individual agencies and broader institutional arrangements.

The findings offer new insight into how employees’ perceptions of transparency are shaped by the micro-level practices that underpin intra-firm accountability mechanisms − such as enforcement procedures, compliance routines and training − which collectively influence how organizational actors interpret and respond to corruption risks. Taken together, the results demonstrate that institutional work is expressed through tangible, everyday practices, particularly enforcement actions, preventive routines and ethical guidance, all of which employees experience as central components of intra-organizational accountability (Lawrence et al., 2011; Smets and Jarzabkowski, 2013).

It offers new insights into how boundary work and practice work can be leveraged to reinforce intra-firm accountability mechanisms, such as sanctioning procedures, training and compliance processes, which collectively shape employees’ perceptions of transparency and reduce the risk of corruption in social business organizations. The findings provide actionable recommendations for managers and practitioners in social business organization, highlighting the critical role of robust monitoring mechanisms and targeted anti-corruption measures. This study underscores the limited utility of generic ethical training and suggests a more context-specific approach to fostering organizational integrity (Beretta et al., 2025). The study highlights the need for policy reforms that prioritize employee experience and knowledge as a central element in designing accountability frameworks. Policymakers are encouraged to incorporate relational and practice-based insights from institutional work to strengthen governance and accountability mechanisms in social businesses (Heredia et al., 2023; Hampel et al., 2017).

This study, therefore, makes an incremental contribution to accounting and organizational change literature by demonstrating how different forms of institutional work operate as micro-level mechanisms through which accountability is enacted inside organizations. While prior studies have focused mainly on formal accountability structures or macro-institutional pressures, our analysis shows how boundary work, practice work and relational work interact to shape employees’ perceptions of transparency and corruption in a social business context. By providing quantitative evidence on the uneven effectiveness of these forms of institutional work, the paper advances understanding of how organizational actors sustain or transform accountability practices in environments characterized by institutional complexity and endemic corruption.

The remainder of this paper is organized as follows. Section 2 provides a comprehensive review of the relevant literature, focusing on organizational accountability, institutional work and corruption prevention strategies. Section 3 outlines the research methodology, detailing the survey design, data collection process and analytical approach. Section 4 presents the data. Section 5 presents findings. Section 6 discusses key findings and their implications in the context of contributions, limitations and future research directions. Section 7 concludes the paper.

Corruption is a widespread problem worldwide. It weakens democracy, damages economic development and reduces public trust in governments. It breaks the relationship between citizens and the state, increases inequality and makes public services less effective − especially in developing economies. Several studies have shown that weak governance systems often lead to continued corruption (Everett et al., 2007; Siddiquee, 2010; Smith, 2010). In places with little supervision or weak, inconsistent or politically influenced rules, corruption becomes part of how things are done in both national governments and organizations. This opens the door for mismanagement, favoritism and illegal enrichment.

Governments and organizations often use formal tools, such as audits, legal checks, compliance procedures and performance reviews, to combat corruption. However, these tools usually rely on idealized models of governance that assume the rule of law, impartial institutions and political fairness. In real life, however, their success depends on whether institutions are respected and trusted − especially things like independent judiciary, free anti-corruption agencies and a legitimate government (Boateng et al., 2024; Amanquandor, 2024). Without public trust, even well-designed accountability systems can fail − or be misused for political reasons.

Governments often launch high-profile anti-corruption reforms, such as integrity commissions, transparency laws, asset declarations or online governance tools, sometimes in response to donor demands or to improve their international image (Asif et al., 2024). While these moves may signal a public commitment to good governance, they are often undermined by political interference and elite resistance. In countries like Malaysia and Nigeria, anti-corruption bodies have been hampered through political overreach, lack of independent prosecution and biased enforcement (Siddiquee, 2010). This can lead to reforms being used to attack opponents while protecting allies − making inequality worse.

International anti-corruption models, often promoted by Western donors and financial institutions, are typically based on technical systems that overlook local culture, politics and institutions (Gadowska, 2010a, 2010b; Okogbule, 2006). These models focus on rules and procedures, such as checklists and performance indicators, but may not align with local needs or ways of working. This leads to what scholars’ call “isomorphic mimicry” − when organizations appear to be following good governance, but nothing really changes.

At the organizational level, corruption is rarely just about a few bad individuals. Instead, it is often built into daily routines, informal rules and social networks − especially in uncertain environments or places with limited resources (Heredia et al., 2023; Isett et al., 2022; Leca and Laguecir, 2023; Aleksandrov et al., 2025). Nepotism, clientelism, unofficial payments or poor resource allocation are often viewed as normal ways to cope with challenging conditions. People inside organizations may even believe these actions are necessary or morally right.

Culture also affects how corruption is understood. In societies with strong group loyalty and respect for hierarchy, people may value family or social obligations more than abstract rules about fairness (Boateng et al., 2024; Gorsira et al., 2018). In these settings, people may ignore formal regulations in favor of helping their networks, and corruption can be seen as a way to solve problems or survive.

There is also growing criticism of the global “audit culture,” especially in public sector management. Critics argue that audits, evaluations and compliance checks often serve to please external donors rather than improve transparency or learning within organizations (Everett et al., 2007). These formal checks can become routine box-ticking exercises that ignore deeper problems. They may even make corruption worse by pretending there is oversight when there is not.

Because of these challenges, accountability is increasingly understood as a mix of formal systems and informal relationships. It is not just about rules but also about trust, social behavior and strategies that work in specific cultural and institutional settings. This view highlights the importance of trust-building, inclusive methods and respecting local norms − while still encouraging responsible behavior.

In Bangladesh, the relationship between accountability and corruption is further complicated by politics, institutional weaknesses and a lack of independence in public administration (Siddiquee, 2003, 2010; Uddin and Hopper, 2003). As a result, despite numerous reforms, accountability systems often exist only on paper. Corruption remains widespread across both the government and business sectors, damaging public services, policy delivery and public trust in authorities.

A significant problem is the acceptance of informal practices, such as bribery, favoritism and the use of personal connections. These behaviors are common in rural areas, where government services are often weak, and people frequently rely on middlemen or personal networks to obtain what they need. These informal systems usually replace formal ones and give the impression of following rules, while powerful groups stay in control.

A lack of trust is another significant barrier. People often do not believe the legal system or regulatory bodies are fair. This distrust encourages the use of bribes or “speed money” to expedite tasks (Toyon, 2022). It also reflects long histories of impunity, political interference and selective rule enforcement. For example, corruption investigations may focus on political rivals while ignoring wrongdoing by allies, damaging the credibility of anti-corruption efforts.

There is also a clear gap between what external donors expect − such as audits, compliance checklists and performance targets − and what happens. These tools may help social business secure funding or meet reporting goals, but they rarely lead to lasting change (Amanquandor, 2024). This is called “ritualised accountability” − when processes are followed symbolically rather than meaningfully. It is made worse by weak capacity, vague roles and a lack of involvement from people on the ground.

Research from Indonesia, Kenya and Pakistan also shows that local elites often co-opt reform efforts for their own gain (Campbell and Danar, 2025; Asif et al., 2024). In Bangladesh, this is clear in the way local governments are run − loyalty to political parties often decides who gets jobs or funding. This makes it hard for anti-corruption programs to succeed, because they get caught up in existing power structures.

This situation shows the limits of standard anti-corruption models that rely on law reforms and technical processes. These approaches often assume neutral bureaucracies and fair courts, which do not always exist. In Bangladesh, political influence affects decisions on appointments and promotions.

Real change in Bangladesh requires more than technical fixes. It means dealing with informal systems, cultural norms and the everyday realities of how people work. It is necessary to understand how staff interpret unclear rules, how informal networks influence decisions and how compliance is carried out − or avoided − in practice. Donor strategies also need to change, shifting from imported models to solutions based on local knowledge and experience.

To deal with corruption in social business, especially in financial contexts, this study draws on institutional work. This approach views institutions not as fixed systems but as flexible structures shaped by the actions of people within them (Lawrence and Suddaby, 2006; Lawrence et al., 2011). It focuses on how individuals and groups actively maintain, change or build systems of rules and norms.

In the context of corruption, institutional work helps us see how people within organizations manage issues such as power, legitimacy and accountability − not just by following external rules, but by adapting to their realities.

There are three key types of institutional work:

1. Boundary work: This is about setting clear standards for behavior. It includes punishment for unethical behavior, strict monitoring and protection for whistle-blowers who report wrongdoing (Zietsma and Lawrence, 2010).

2. Practice work: This involves putting those rules into everyday use − through training, standard procedures and ethical leadership. But how successful this depends on how well it fits with the organization’s culture and staff beliefs. So, in this study, training and ethics (TE) are treated exclusively as a component of practice work, consistent with our earlier conceptual definitions.

3. Relational work: This focuses on building trust and shared understanding across departments and organizational levels (Smets and Jarzabkowski, 2013). In Bangladesh’s social business, where top-down approaches often fail, this kind of work is especially important.

These forms of work matter most in complex environments, where multiple pressures from donors, communities and political groups coexist (Kraatz, 2009). Staff in Bangladeshi social businesses often have to balance these competing demands. Institutional work helps explain how they manage this through practical decision-making and the construction of legitimacy.

Importantly, change does not just come from senior leaders. Middle managers, compliance staff, frontline workers and community links all play a role. Research shows that accountability systems work better when these people understand, trust and believe in them, and that this is helped by training and ethical support (Gorsira et al., 2018; Hope, 2016). Changing the overall workplace culture is often more effective than just introducing rules.

Recent research further underscores the relevance of institutional work for understanding how accounting actors actively shape accountability infrastructures across organizational and national contexts (Arora et al., 2023). Their findings highlight that institutional work is not merely technical or procedural but deeply relational and context-dependent, shaped by professional judgment, organizational power dynamics and local institutional conditions.

Finally, success in institutional work also depends on legitimacy − people must believe that accountability measures are fair and worth following. Relational work helps build this belief, as trust and shared purpose are key for voluntary cooperation and meaningful change.

In this study, punitive/boundary work, preventive–training/practice work and relational work are conceptualized as the organizational mechanisms through which accountability is enacted. These practices shape how employees interpret integrity, enforcement of rules and organizational conduct. Perceived transparency, therefore, represents the outcome of these accountability mechanisms rather than the mechanisms themselves. This conceptual pathway links institutional work to corruption-related perceptions through the functioning of accountability structures.

As discussed in this section, prior studies on accountability, corruption and institutional work provide important insights into the structural conditions, governance arrangements and cultural dynamics shaping organizational integrity, particularly in developing-country contexts such as Bangladesh. However, much of this literature remains either macro-oriented − focusing on formal reforms and institutional design − or predominantly qualitative, emphasizing narratives of change without systematically assessing the relative effectiveness of different accountability mechanisms. As a result, there is limited empirical evidence on how specific forms of institutional work (such as enforcement-driven boundary work, preventive practice work and ethics-oriented relational work) operate within organizations and how employees perceive their effectiveness in mitigating corruption.

This study addresses this gap by empirically examining how distinct institutional work practices are associated with perceived transparency as an outcome of accountability. The study is guided by theoretically grounded expectations drawn from prior literature − namely, that enforcement-based boundary work is likely to exert a stronger deterrent effect on corruption perceptions than preventive or training-based interventions, particularly in contexts characterized by entrenched informal norms. In doing so, the paper advances existing research by linking institutional work theory to empirical evidence on intra-organizational accountability, thereby clarifying how micro-level practices translate into differential governance outcomes.

The starting point of our empirical modeling is to define the broad spectrum of corruption, identify factors affecting accountability and examine the perceived level of corruption among its employees. We identify four types of transparency indicators: namely, the presence of transparent processes for loan approvals, the existence of an extensive control system and compliance with loan approval and staff promotion procedures. These four questions correspond to questions 1, 4, 7 and 14 of section B of the survey questionnaire (refer supplementary material). These individual categories are linked with the fraud triangle framework, i.e. incentive, opportunities and rationalization to formulate a composite variable representing overall corruption. The responses are obtained using a scale of 1–5, with 5 and 4 representing strongly agree and agree; and 3, 2 and 1 representing neutrality, disagree and strongly disagree, respectively. We then transform each of the four categories of transparency indicators into binary numbers, assigning 1 to all answers with agreement and 0 to the remaining three responses. The composite corruption variable is therefore a binary response variable indicating agreement (1) or disagreement (0) with the presence of transparency.

The survey instrument was explicitly informed by the Institutional Work framework, particularly the typology of boundary work, practice work and relational work (Lawrence and Suddaby, 2006; Zietsma and Lawrence, 2010; Smets and Jarzabkowski, 2013). These categories guided the construction and grouping of survey items into theoretically meaningful constructs. Specifically, boundary work was operationalized through items relating to punitive measures and enforcement actions, reflecting the efforts of organizational actors to establish and protect accountability boundaries. Practice work was captured through items on compliance with rules, ethics training and control systems − indicating the embedding of accountability in daily routines. Relational work was reflected in questions about trust, the accessibility of information and collaboration, which facilitate shared understanding and interpersonal cohesion. The alignment between these institutional work categories and specific survey questions is summarized in Table 1. This structured approach ensured that the survey design was theoretically grounded and that the quantitative analysis could be meaningfully interpreted through the lens of institutional work.

Although items B1, B4 and B7 appear in both the practice work construct and the composite transparency measure, they serve two analytically distinct purposes, consistent with prior research on accountability and institutional work (Luo, 2005; Hauser, 2019; Gorsira et al., 2018). Within the practice construction, these items capture the formal presence of preventive routines and compliance processes, indicating how organizational actors embed accountability mechanisms into everyday operational procedures. In contrast, within the transparency construct, the identical items reflect employees’ assessments of how effectively these routines function in practice. This distinction allows us to separate the existence of organizational practices (institutional work) from employees perceived outcomes of those practices (institutional effectiveness). The dual use of these items therefore aligns with established methodological approaches in the accountability literature, in which process indicators may serve simultaneously as measures of structural routines and as proxies for perceived institutional performance. This analytical separation is essential in contexts such as Grameen Bank, where formal processes may exist on paper but vary in their practical implementation across branches and employee groups.

Given the research questions outlined in the Introduction, focusing on how organizational actors engage in institutional work (RQ1) and which forms of institutional work are most effective in reducing corruption and enhancing accountability (RQ2), we identified several explanatory variables aligned with the institutional work framework. In this framework, the independent variables − such as enforcement actions, training and preventive mechanisms − are conceptualized as concrete manifestations of institutional work that collectively enact intra-organizational accountability practices. These variables reflect the operational efforts of social businesses to establish boundary work (e.g. rule enforcement), practice work (e.g. capacity building and preventive protocols) and relational work (e.g. fostering trust and collaboration). Although accountability is not measured directly as a standalone construct, these practices are widely recognized in literature as core components of accountability systems. The dependent variable, transparency, reflects the outcomes of accountability mechanisms, indicating how effectively institutional work practices translate into perceived organizational integrity and reduced corruption. Hence, it is essential to clarify that transparency does not represent the accountability mechanisms themselves; instead, it captures employees’ assessment of how well these mechanisms function in practice. As such, the analytical focus lies in examining how different forms of institutional work contribute to perceived transparency and reflect organizational accountability in complex institutional environments. Specifically, the variable AACOR (actions taken against corruption) captures aspects of boundary work, while PREV (preventive mechanisms) and TE (training and ethics) reflect elements of practice work and relational work. In addition, to account for micro-level heterogeneity and actor-specific attributes, we include employee-level characteristics (e.g. gender, age and experience) which may mediate how individuals perceive and respond to accountability mechanisms within the organization. This approach enables us to operationalize the different dimensions of institutional work and statistically assess their impact on perceived transparency and intra-firm accountability.

As mentioned in the introduction section, we use three employee perception indicators, i.e.:

1. information flow about corruption;

2. monitoring mechanisms to prevent corruption; and

3. the organizational structure of the social business organization to avert corruption.

These indicators are aggregated to obtain the composite indicator for corruption prevention, which is then transformed into a binary variable. Action against corruption is represented by responses as to whether participating employees agree that the social business organization will take strong action if corruption is found. As in the above cases, all responses in the agreement were represented with 1, with remaining responses coded as 0, in the formation of the corresponding binary variables. The third explanatory variable measures the effect of training and ethics and is constructed as a dummy variable using the same procedure as above and aggregating three categories representing anti-corruption training, corruption detecting mechanisms and provision of ethical guidance to employees.

For modeling corruption, our binary dependent variable is modeled using a probit or logit model. The use of probit and logit models is justified given the well-known limitations of linear probability models that predict the dependent variable outside its range. The probit and logit models overcome the limitations of a linear probability model estimated using an ordinary least squares (OLS) regression that yields biased and inconsistent estimates along with heteroskedasticity (Horrace and Oaxaca, 2006). The probit and logit models correctly predict the dependent variable, which is bounded between 0 and 1. Consider the following model:

where we need to estimate the conditional probability that Y = 1 as a function of certain known explanatory variables. The functional form $f(.)$ is chosen in a way to ensure that y lies between 0 and 1 for all possible values of $β0+β1x1+β2x2+⋯+βKxK$ . For this study, we re-specify our model of corruption as follows:

where TRANS represents transparency as a converse measure of corruption, meaning higher transparency indicates lower corruption and vice versa. PREV represents the preventive measures against corruption taken by the social business organization. AACOR stands for actions taken against corruption, and TE stands for ethical training and guidance. Similar to the binary response variable, these explanatory variables are also expressed as dummy variables and in terms of the original response values obtained from the field survey. Additionally, we use three variables representing employees’ characteristics that give the following extended model:

where GENDER, AGE and EXPER represent a gender dummy, age group and experience group, respectively. Both age and experience variables are coded from 1 to 4, with one representing 20–30 and 4 representing 51–65. Experience is coded as 1 for less than 1 year’ experience, 2 for the experience of 1–5 years and going up to 4 for the experience of 11–15 years. The nonlinear transformation used for the function $f(.)$ is the logistic function for the logit model and the standard normal cumulative distribution function for the probit model. Both these models are derived from an underlying latent variable model. We can then write our corruption model as:

where $Xβ$ is a matrix representation of the explanatory variables and the parameters in equations (2) and (3). As per the standard procedure for the logit model, we express the log of the odds ratios measuring the probability of the existence of transparency (no corruption) to no transparency (corruption), as a linear function of the three explanatory variables and the parameters. For the probit model, we derive the following from equation (4):

Further, in line with standard practice, the parameters of both the logit and probit models are estimated using maximum likelihood.

This study uses primary data based on a field-level survey conducted over six months. Development of the survey questionnaire (refer supplementary material) was informed by economic and business theory, as well as the extant literature on organizational corruption (Trivunovic, 2011; Bédécarrats et al., 2012). A field-level survey was chosen as the preferred method because it enabled the collection of data and information from a broader range of respondents, resulting in a larger sample size. Besides, with a structured questionnaire-based survey (refer supplementary material) as compared to semistructured interviews of only selected employees, it was possible to minimize interview biases, thereby raising reliance on statistical and other analytical inferences. Questionnaires (refer supplementary material) were developed based on Luo’s institutional model (2005), which shows that corruption in an organization is shaped by procedures, structures and behaviors within the organizational environment (Azim and Kluvers, 2019). The model indicates how the task and institutional environments provide opportunities for entrenching corruption. Based on this concept the questionnaire (refer supplementary material) developed aims to measure:

• corruption and organizational environment;

• corruption and organizational behavior;

• corruption and organizational consequences; and

• corruption and organizational architecture.

These constructions are interrelated yet exhibit sequential issues that collectively constitute organizational explanations of corruption.

The survey was conducted in two stages. At the first stage, the researchers and their team distributed 454 survey questionnaires (refer supplementary material) to various employees of Grameen Bank. Probability sampling methods were used to randomly select employees from the entire staff list (Fowler, 2009). There were an initial 146 responses, which increased to 250 following reminders and repeated contacts, yielding a 55.06% response from the initial number of survey mail outs. Among the respondents, 144 (57.6%) were male, and 106 (42.4%) were female. Respondents were primarily young to middle-aged. Among them, 109 (43.6%) respondents were aged between 20 and 30, 123 (49.2%) were aged between 31 and 40, 13 (5.2%) were aged between 41 and 50 and only 5 (2%) were aged between 50 and 65. The majority of respondents, i.e. 123 (49.2%) representatives, had experience of 1–5 years, followed by 89 (35.6%) respondents with 6–10 years of experience. Only one respondent had more than 20 years of experience. Out of 250 respondents, the majority, 122 respondents (48.8%), worked in the branch office, followed by 74 (29.6%) who worked in the head office, 23 (9.2%) who worked in the Zonal office and 31 (12.4%) who worked in the area office.

The demographic breakdown indicates that the sample is broadly representative of Grameen Bank’s employee base, capturing a balanced mix of gender, age groups, experience levels and work locations. The high response rate among employees with 1–10 years of experience offers valuable insights into how younger and mid-level professionals perceive organizational corruption and accountability. Additionally, the diverse office locations ensure a comprehensive perspective on how corruption and accountability mechanisms operate at different organizational levels.

We first report descriptive statistics in Table 2. As shown in the table, most interviewees agree on the existence of transparency and corruption-prevention mechanisms, and on the actions taken to discipline corruption at the Bank. These measures address the opportunistic behaviors identified in the fraud triangle framework. Interestingly, about 60% of respondents disagree with the provision of anti-corruption training, ethical guidance and the existence of corruption-detection mechanisms at the Bank (Table 2). Also, as indicated by the columns on the right of the table, “agree” seems to be the modal response for the three variables, except for TE, which has the highest standard deviation of responses. The distribution of AGE shows that 43.6% of the respondents are in the 20–30 years age range (group 1) and 49.2% are aged 31–40 years (group 2). The remaining 5% of the respondents are aged between 51 and 65 years. Also, as seen from Table 2, the distribution of the variable representing experience reveals that an overwhelming majority of 49.2% and 35.6% of the respondents worked at Grameen Bank for 1–5 years and from 6 to 10 years, respectively.

The descriptive analysis reveals essential patterns. Most respondents perceive a positive presence of transparency and disciplinary mechanisms to prevent corruption. A substantial number of employees highlighted inadequate anti-corruption training and guidance, representing a critical area for improvement. Younger employees (20–40 years) and those with 1–10 years of experience accounted for the majority of respondents, suggesting that their perceptions may strongly influence the overall results of this study.

This section presents the findings from the linear probability model (OLS) and the probit and logit models, which analyses the impact of key organizational factors − punitive actions, corruption prevention measures and anti-corruption training − on achieving greater transparency, a proxy for reduced corruption. These findings are interpreted through the lens of institutional work, focusing on how organizational actors shape and sustain accountability mechanisms and which forms of institutional work are most effective in combating corruption.

As an essential step to ensure the reliability of our model estimation, we ensure that the key explanatory variables, namely, PREV, AACOR and TE, along with the employee characteristics (GEND, AGE and EXP), are not highly correlated; that is, the models do not suffer from multicollinearity. To demonstrate this, we estimate the correlation coefficient matrix including all six explanatory variables. These results are reported in Table 3. As Table 3 (panel A) shows, the correlation coefficients range between 0.01 and 0.63 in absolute terms, indicating the absence of any multicollinearity among the variables. This is also substantiated by the computed variance inflation factor (VIF) reported in Table 3, panel B, which is well below the conventional threshold of 10, ranging from 1 to 1.84, with a mean of 1.31, indicating that multicollinearity is not a concern in our model.

To ensure the reliability and validity of the survey instruments, we pre-tested the questionnaire (refer supplementary material) with a small group of respondents to refine and clarify questions, eliminating ambiguity. Internal consistency was assessed through inter-item correlations and exploratory factor checks. Inter-item correlations were positive and within conceptually appropriate ranges, indicating that items designed to measure each construct behaved coherently with no negative or near-zero relationships. Exploratory factor analysis further confirmed that items loaded primarily on their intended factors (AACOR, PREV, TE, Transparency) with no cross-loadings above 0.30. Although Cronbach’s alpha was computed for key constructs, the estimates were not uniformly high (0.52), likely due to the small number of items per scale and the assumption that tau-equivalence does not hold uniformly. As suggested in some recent papers (Dunn, Baguley, and Brunsden, 2014; McNeish, 2018), Cronbach’s alpha may not always provide an accurate estimate of internal reliability under these conditions; instead, item-level consistency and factor structure offer a more robust assessment of internal reliability.

To assess the presence of standard method bias, we conducted Harman’s single-factor test using an unrotated factor analysis. The analysis revealed that the first factor accounted for approximately 18.6% of the total variance, well below the 50% threshold commonly used to indicate problematic levels of standard method variance (Chang et al., 2010). Thus, we find no evidence of substantial standard method bias in our data.

Table 4 reports the estimated linear, probit and logit models. As per standard econometric practice, the linear probability model is estimated by OLS, with heteroskedasticity-robust standard errors reported. In contrast, the probit and logit models are calculated using a maximum likelihood (LM) procedure. The estimates of the linear probability model reveal that among the three explanatory factors, action taken against corruption or punitive measures have a significant positive effect on corruption reduction or transparency. To be precise, the estimates show that with an action taken against corruption, the probability of transparency or corruption reduction is 18%. Compared to this, preventive measures and the provision of anti-corruption and ethical training imply a lower probability of corruption reduction. However, unlike the linear probability model, the coefficient estimates of the logit and probit models do not represent the magnitude of change in the probability of transparency or no corruption in response to actions against corruption or other explanatory variables. Instead, as the estimates imply, we can conclude that more actions or punitive measures against corruption are likely to significantly raise transparency or reduce corruption in the social business organization. The action taken against corruption (AACOR) emerged as the most significant explanatory variable across all models. The linear probability model shows that punitive measures increase the probability of transparency − or reduction in corruption − by 18%. This finding reflects the importance of boundary work performed by organizational actors who enforce accountability by defining and maintaining organizational norms through decisive actions against misconduct (Zietsma and Lawrence, 2010; Hampel et al., 2017). The significance of AACOR across the logit and probit models further highlights how actors engage in practice work, embedding accountability mechanisms into the organizational routine to sustain transparency.

The regression results suggest that punitive measures (AACOR) are the most effective institutional work mechanism for combating corruption. While corruption prevention measures (PREV) and anti-corruption training (TE) did not show statistically significant effects in the linear probability model, their presence reflects ongoing relational and practice work efforts aimed at shaping organizational culture. The limited effectiveness of these measures may indicate the need for more tailored and context-specific approaches to institutional work, integrating relational strategies that bridge the gap between formal policies and actual employee practices (Smets and Jarzabkowski, 2013).

The inclusion of employee-specific characteristics such as age, gender and work experience provides further insight into how micro-level practices and actor-driven responses influence accountability perceptions. The extended models revealed that younger employees and those with fewer years of experience reported higher levels of perceived transparency. This suggests that organizational actors at different career stages may experience institutional work differently, with early-career employees potentially more receptive to newly established accountability mechanisms.

These findings emphasize that boundary work (e.g. punitive measures) is crucial for defining the limits of acceptable behavior and deterring corruption. However, practice work and relational work − such as anti-corruption training and cross-departmental collaboration − require further refinement to increase their impact. The results are aligned with previous studies highlighting the importance of adaptive institutional work in addressing complex challenges like corruption in social businesses (Isett et al., 2022; Leca and Laguecir, 2023; Aleksandrov et al., 2025).

The survey included several questions about employees’ personal characteristics at Grameen Bank, aiming to explore how these attributes relate to their perceptions of corruption and organizational transparency. Table 5 presents estimate from models that incorporate individual characteristics, such as age, gender and experience, offering insights into how these factors influence employees’ engagement with and responses to institutional accountability mechanisms.

The estimates confirm that actions taken against corruption (AACOR) remain the most significant factor in increasing transparency and reducing corruption. The linear probability model indicates that punitive measures reduce the likelihood of corruption by 12%, reinforcing the importance of boundary work in shaping institutional norms and ensuring compliance. The probit and logit models support this finding, revealing that actions against corruption consistently raise the probability of transparency across all specifications.

Among the individual characteristics, employee experience stands out as a critical factor in shaping perceptions of transparency. The probit and logit estimates reveal a surprising pattern: employees with longer tenure at the Bank are more likely to perceive or engage in corrupt practices. This finding suggests that institutional work practices may become less effective over time, as experienced employees develop more profound familiarity with organizational processes and may exploit these for personal gain. It reflects a potential breakdown of practice work and a need for enhanced relational strategies to ensure that accountability mechanisms remain relevant and resilient over time (Zietsma and Lawrence, 2010; Smets and Jarzabkowski, 2013).

Additionally, this result highlights the importance of adaptive institutional work in sustaining accountability. Long-tenured employees may benefit from periodic re-engagement with the Bank’s accountability frameworks and targeted interventions to reinforce ethical practices. Without such ongoing efforts, the risk of normalization of deviant practices may increase, diluting the effectiveness of institutional work aimed at promoting transparency.

These findings underscore the complex relationship between individual characteristics and institutional practices. While boundary work remains crucial in defining and enforcing accountability norms, sustained practice work and relational strategies are essential to prevent the erosion of transparency over time.

The findings from the models with and without employee-specific characteristics are reported in Tables 4 and 5, providing a comprehensive analysis of factors influencing intra-firm accountability and perceptions of corruption at Grameen Bank. The results overwhelmingly highlight the significance of actions or punitive measures (AACOR) as a key institutional mechanism for combating corruption. This finding aligns with the notion of boundary work in institutional work theory, where actors enforce accountability norms by clearly defining and sanctioning unacceptable behaviors (Zietsma and Lawrence, 2010; Lawrence and Suddaby, 2006). Actions against corruption consistently emerge as the most impactful factor across all models, reinforcing the effectiveness of enforcement-driven strategies in promoting transparency.

Additionally, employee experience plays a significant role in shaping perceptions of the Bank’s culture of corruption. The probit and logit estimates suggest that longer-tenured employees are more likely to perceive corruption, suggesting a potential challenge to sustaining institutional accountability practices over time. This result underscores the importance of continuous practice work and relational strategies to ensure that accountability mechanisms remain resilient and relevant. Without adaptive responses, experienced employees may develop strategies to navigate and potentially exploit organizational processes, diminishing the effectiveness of existing institutional work practices.

As a robustness check, the models were re-estimated using untransformed independent variables (measured on a 1–5 scale) rather than the binary transformations used in the initial estimate, and the results are presented in Tables 6 and 7. As the models’ estimates indicate, the results remained generally consistent, confirming the stability of the findings and reinforcing the conclusion that boundary work and targeted institutional interventions are critical for enhancing transparency and reducing corruption at Grameen Bank.

These findings directly address the two research questions guiding this study. In relation to RQ1, how organizational actors engage in institutional work to shape and sustain intra-firm accountability mechanisms, the results highlight that actors participate most effectively through boundary work, namely through enforcement and punitive responses to misconduct (AACOR). This form of institutional work creates explicit behavioral norms and reinforces the organization’s ethical expectations. However, the limited impact of training and preventive mechanisms (TE and PREV) suggests that practice work and relational work − which require embedding formal mechanisms into day-to-day routines and building trust − are either underdeveloped or poorly implemented in this context. One plausible explanation is that in organizational contexts characterized by entrenched informal norms and weak enforcement environments, formal training and preventive mechanisms often fail to produce meaningful behavioral change because employees perceive them as symbolic rather than consequential. Prior research on accountability in developing-country institutions suggests that when corruption becomes pervasive and enforcement lacks credibility, training-based interventions tend to have limited effectiveness unless supported by strong, consistent sanctioning and monitoring systems (Siddiqui, 2010; Rahman and Kapur, 2020).

Regarding RQ2, what types of institutional work are most effective in combating corruption and fostering accountability? The findings show that boundary work (as captured by punitive action) is the most statistically significant and consistently effective intervention. In contrast, institutional work rooted in preventive or ethical training efforts lacks significance, suggesting either conceptual gaps in their design or resistance in practice. These insights demonstrate that organizational accountability mechanisms must be adaptive and context-specific, particularly for long-tenured employees whose perceptions may reflect either institutional fatigue or exposure to normalized deviance.

This study sets out to investigate two interrelated research questions: (RQ1) how organizational actors engage in institutional work to shape and sustain intra-firm accountability mechanisms, and (RQ2) what types of institutional work are most effective in combating corruption and fostering a culture of accountability in a social business organization. The findings provide empirically grounded insights into the micro-level practices and actor-driven strategies that underpin organizational accountability at Grameen Bank.

In response to RQ1, the study shows that organizational actors most visibly engage in boundary work through the enforcement of strict punitive measures against corruption. The AACOR variable − reflecting employee perceptions of disciplinary action − consistently emerges as the strongest predictor of perceived transparency across all statistical models. This indicates that institutional norms are primarily upheld through deterrence-based accountability mechanisms. By contrast, the comparatively weak performance of PREV and TE suggests that practice work (training, routine compliance) and relational work (trust-building, ethical support) are not yet fully embedded in the organization’s governance systems, limiting the depth and long-term sustainability of accountability efforts at operational levels.

Addressing RQ2, the findings further demonstrate that boundary work is the most effective form of institutional work in this context, as visible enforcement reinforces the credibility of institutional norms. The limited influence of training and preventive strategies implies that formalistic or generic interventions may not sufficiently engage the behavioral and cultural dimensions of corruption. Additionally, the tendency of longer-tenured employees to report higher perceptions of corruption highlights the need for adaptive institutional work, particularly ongoing relational engagement, to counter potential normalization of unethical practices and to strengthen accountability frameworks over time.

This finding highlights the importance of adaptive institutional work, in which accountability mechanisms are continually refined to address emerging challenges and prevent the erosion of transparency over time. Targeted interventions that re-engage experienced employees and emphasize the evolving nature of accountability practices could mitigate this risk.

This study makes several contributions to theory, practice and policy by empirically investigating how organizational actors engage in institutional work to shape intra-firm accountability (RQ1) and by identifying which forms of institutional work are most effective in combating corruption and fostering transparency within a social business organization (RQ2).

In response to RQ1, the study advances the institutional work literature by demonstrating how organizational actors, particularly in resource-constrained and politically complex contexts, actively perform boundary work to define and enforce institutional norms. This work is manifested through punitive actions that reinforce behavioral expectations and organizational accountability. While prior literature has explored the role of actors in creating or disrupting institutions (Lawrence and Suddaby, 2006; Smets and Jarzabkowski, 2013), our study contributes by emphasizing the maintenance of accountability systems via deterrence-based enforcement in environments where corruption risks are endemic.

Moreover, the study reveals the relative underperformance of practice work and relational work, which, although conceptually central to sustaining institutional legitimacy, remain underutilized or ineffective in this context. This finding refines existing theory by showing that institutional work is not uniformly enacted across its different dimensions and that certain types (e.g. relational work) may be contingent on organizational culture, trust dynamics and historical legacies of corruption. In this way, the paper contributes to the literature by illustrating the uneven institutionalization of accountability mechanisms within social enterprises operating in developing countries.

In relation to RQ2, the study provides empirical evidence that among the different forms of institutional work, boundary work − as captured through disciplinary mechanisms − is the most statistically robust and practically effective. This confirms the theoretical proposition that institutional stability, particularly in the face of corruption, may rely more on normative enforcement than on symbolic or procedural interventions. The findings challenge the assumption that ethical training or preventive structures necessarily translate into improved governance, particularly when such interventions are not context-specific or actively internalized by employees. Accordingly, the paper contributes a nuanced theorization of institutional work effectiveness, emphasizing that success is not only a function of formal design but also of organizational receptivity, actor engagement and cultural embedding.

From a policy perspective, the study highlights the limitations of compliance-driven anti-corruption reforms that rely on standardized training modules or externally imposed preventive controls. The evidence suggests that without robust enforcement mechanisms and actor buy-in, such reforms may fail to achieve sustained behavioral change. Policymakers designing accountability systems for social businesses, particularly in developing countries, should therefore adopt a layered approach, combining visible enforcement (boundary work) with adaptive and locally grounded practices (relational and practice work).

For managers and practitioners, the findings underscore the central role of visible enforcement in establishing credible norms of behavior. However, punitive measures alone are insufficient for building a resilient accountability culture. The limited impact of training and preventive mechanisms points to the need to reimagine ethics programs, moving beyond generic content toward more interactive, scenario-based training that reflects the real dilemmas staff face.

Practitioners should also consider investing in relational work: mechanisms that build horizontal trust among employees and vertical trust between staff and leadership. Such efforts could include mentoring programs, participatory audit practices or staff-led integrity forums. These relational strategies may not yield immediate statistical results but are essential for embedding accountability within the social fabric of the organization.

While this study provides important insights into the role of institutional work in shaping intra-firm accountability and combating corruption in social business organizations, it is not without its limitations. Acknowledging these limitations helps to contextualize the findings and offers avenues for future research. This study focuses on a single social business organization, Grameen Bank, in a developing country context. Although this provides rich, context-specific insights, the findings may not be generalizable to other social business organizations or sectors, particularly in different cultural or regulatory environments. Future research could adopt a comparative case study approach, analyzing how institutional work differs across multiple organizations and regions to provide a broader understanding of accountability practices. The study relies on employee perceptions to measure transparency and corruption, which may be influenced by personal biases and organizational dynamics. While perception-based data is valuable for understanding how employees experience accountability mechanisms, future research could incorporate objective measures such as external audits, financial performance indicators or documented instances of detected corruption to triangulate findings and enhance robustness. Although the study highlights the significance of boundary work and practice work, it offers only limited exploration of relational work and leadership activities in fostering accountability. Future research could focus on leadership practices and the role of key actors in mobilizing relational work, particularly in building trust and bridging institutional complexities (Hampel et al., 2017). The findings reveal that employee experience significantly influences perceptions of corruption, yet the underlying reasons for this pattern remain unclear. Future research could adopt qualitative methods, such as interviews or ethnographic studies, to explore how micro-level dynamics, such as social networks, informal practices and power relations, affect accountability and corruption perceptions.

This study examines how organizational actors actively shape and maintain intra-firm accountability mechanisms in social business organizations, with a focus on identifying the most effective strategies for combating corruption. Using institutional work theory, it highlights the critical role of agency-driven practices, including boundary work, practice work and relational work, in enhancing accountability and transparency.

The findings provide strong evidence that punitive measures and enforcement mechanisms are the most effective institutional practices for promoting transparency and reducing corruption within Grameen Bank. These measures, rooted in boundary work, help define organizational norms and deter misconduct, ensuring accountability remains a central organizational priority. However, the limited impact of corruption prevention measures and anti-corruption training suggests that while formal mechanisms exist, they are not always perceived as effective by employees. This underscores the importance of practice work and relational strategies to embed these mechanisms into the organization’s day-to-day functioning.

Employee-specific characteristics, particularly experience, reveal additional complexities in the unfolding of institutional work overtime. The study found that longer-tenured employees are more likely to perceive corruption, suggesting the potential erosion of accountability practices or the normalization of deviant behavior. This result calls for more adaptive institutional work and continuous engagement with experienced employees to sustain accountability and foster an ethical organizational culture.

From a theoretical perspective, this study contributes to the institutional work literature by demonstrating how micro-level practices influence the stability and evolution of accountability mechanisms. It expands understanding of how intentional actions by organizational actors shape institutional change in complex, resource-constrained environments. For practitioners and policymakers, the findings offer actionable insights into strengthening accountability frameworks by combining enforcement-driven strategies with more context-specific, relational approaches to training and engagement.

The study highlights that maintaining accountability and combating corruption requires a combination of vigorous enforcement, continuous refinement of practices and active employee engagement. Future research can expand on these findings by conducting longitudinal and multi-context studies to examine how institutional work evolves across different settings and how various forms of institutional work interact to drive sustained organizational change. By demonstrating how specific micro-level practices − enforcement actions, compliance routines and ethics-related training − shape employees’ perceptions of transparency, this study clarifies the mechanisms through which institutional work influences organizational accountability.

Taken together, these findings build on existing research in accounting and organizational change by providing empirical insights into how institutional work influences the development of intra-organizational accountability. The study shows that not all types of institutional work have the same impact: boundary work strongly and directly affects transparency, whereas practice work and relational work require deeper contextualization to be effective. This adds nuance to current discussions on how accountability systems either adapt or fail to adapt to shifting institutional pressures in developing-country settings. By illustrating how micro-level practices shape trajectories of organizational change, the paper offers a meaningful and incremental contribution to institutional work research within the accounting discipline.

The questionnaire and methodology for this study was approved by the Human Research Ethics committee of the Swinburne University of Technology, Melbourne, Australia (Ethics approval number: SUHREC 2010/268).

All authors certify that they have no affiliations with or involvement in any organization or entity with any financial interest or nonfinancial interest in the subject matter or materials discussed in this manuscript.

The supplementary material for this article can be found online.