Management Control for Microfinance: An Examination of the Belief System of A Malaysian Microfinance Provider

This paper examines how the belief system of a bank that provides microfinance services influences formal and informal control mechanisms and consequently shapes decisions and the strategic direction of microfinancing in Malaysia. Using institutional logics perspective as our theoretical lens, we conducted a case study in a Malaysian developmental financial institution (DFI) responsible for providing microfinancing. The results suggest that it is difficult to achieve a balance between economic and social considerations when the banking belief system is strongly rooted in the overall banking practices. This paper highlights the dominance of the banking logic over the social logic as reflected in the DFIA­s management control system. Specifically, it demonstrates how its belief system underpins its microfinance activities due to a focus on risk-return considerations, which aim to minimise non-performing loans and maximise commercial profits. This consequently affects clients who have obtained microfinance products and services. This paper also demonstrates how the hiring carriers of social logic do not appear to infuse the organisation with social logic, due to the vague and compartmentalised structure of the microfinance segment and a lack of long-term social goals in both external and internal monitoring systems.


INTRODUCTION
Microfinance institutions (MFIs) and other microfinance providers that provide financial services to financially excluded groups of people face tensions between their social and economic objectives and must motivate their employees to achieve excellence in both (Battilana & Dorado, 2010;Epstein & Yuthas, 2011). Balancing these dual objectives is problematic, as achieving social objectives can impact commercial objectives or vice versa. Lending to the poor is labour intensive, risky and incurs higher transaction costs (Battilana & Dorado, 2010). To compensate for this risk, microfinance loans require a higher return, which might lead to the prioritisation of economic over social considerations (Siti-Nabiha, Azhar, Mohd Isa, & Siti-Nazariah, 2018). Thus, many researchers doubt the longrun feasibility of achieving this double bottom line and suggest a potential tradeoff between social and financial objectives (Dehejia, Montgomery, & Morduch, 2012;Hermes, Lensink, & Meesters, 2011).
The tension between financial and social objectives is exacerbated by the difficulty of quantifying social measures, which may result in a reliance on financial measures to monitor performance and manage resources (Durden, 2008;Norris & O'Dwyer, 2004). Such action could lead to focusing on financial performance instead of social performance, which might break the link between the objectives of microfinancing and the way the process is managed internally. To resolve this tension, several authors have argued that MFIs should use management control practices that could assist MFIs to balance the demands of their social and financial goals, while at the same time ensuring a coherent translation of the organisation's strategy and the achievement of both its economic and social goals (Epstein & Yuthas, 2010;2011).
However, the issue that surrounds management control for MFIs, concerning balancing the needs of social and financial performance, might be problematic for microfinance providers, especially those with a strong commercial orientation, given that their belief system might influence microfinance-related decisions. In other words, if a microfinance provider's value is deeply rooted in banking or commercial orientation, then this would be reflected in its belief system, i.e., the control system that directs or restricts strategic decision-making activities, and it would also influence other organisational control mechanisms. Consequently, this would hinder the balance between social and commercial objectives that underpins microfinancing, and the social objectives of microfinance might not be achieved. However, empirical research on control systems in the context of microfinancing is limited with regard to the issue of the role of belief systems as a lever of control. As such, the purpose of this paper is to examine how the belief system of an organisation that provides microfinancing influences formal and informal control mechanisms and consequently shapes decisions related to key microfinance activities in the organisation, e.g., loans approval, collection, disbursement and risk management.
To address the research questions, we conducted a case study with a Malaysian developmental financial institution (DFI), a specialised financial institution with a specific mandate to promote key sectors of strategic importance, and which has also been entrusted with the additional responsibility of providing microfinance products and services. Microfinance in Malaysia is generally a government-mandated programme, and the first established microfinance provider was a government-backed NGO. Subsequently, in pursuit of an agenda of financial inclusion, DFIs were given the responsibility of offering microfinancing services. The microfinance segments of DFIs differ from those of other commercial institutions because their objectives include not only economic aspects but also social aspects.

LITERATURE REVIEW
Management control consists of an array of control mechanisms that are employed to pursue organisational objectives (Widener, 2007;Simons, 1995;Merchant & Van der Stede, 2012). In contrast to formal systems, which seek to control behaviour through written procedures and policies or codes of conduct, informal control systems comprise shared values and corporate culture, and are not based on measureable objectives or explicit measures (Evans & Tucker, 2015;Norris & O'Dwyer, 2004;Riccaboni & Leone, 2010). Informal control systems are regarded as a form of social control because they consist of less well-defined practices and embody the connections and communications between organisational members (Tucker, 2011). Simons' (1995 levers of control (LOC) framework views management control as the combination of four control systems; it has been argued to be relevant to balancing the trade-offs faced by organisations such as between short-term and long-term goals, the needs of the different constituents and the different objectives/pressures faced by organisations. It is therefore appropriate for this research, which focuses on microfinance providers. These four control systems comprise: (i) belief system (transmitting organisational core values); (ii) boundary system (setting limits on organisational behaviour); (iii) diagnostic system (monitoring critical performance variables); and (iv) interactive control system (encouraging debates and dialogues in the process of monitoring uncertainties) (Simons, 1995). Despite the critical and significant role of the belief system as a control lever in managing and maintaining organisational members' identification with the organisational core values, compared to other control levers it has been overlooked and it is not widely used in the empirical literature, especially with regard to its role in shaping other elements of a control system (Collier, 2005;Chenhall, Hall, & Smith, 2010;Evans & Tucker, 2015).

The Belief System as a Lever of Control
The belief system is defined as the "explicit set of organizational definitions that senior managers communicate formally and reinforce systematically to provide basic values, purpose and direction for the organization" (Simons, 1995, p. 34). The belief system informs managers what the organisation stands for and guides them in what they can or cannot do. Moreover, belief controls promote performance while simultaneously (through the established values) acting as a boundary system by ensuring organisational compliance (Tessier & Otley, 2012). As such, the informal control system is inherently embedded in the belief system (Collier, 2005;Ferreira & Otley, 2009), as implied by Simons' suggestion that the belief system influences the boundary system and that both systems are aligned and compatible with the organisational culture, which constrains as well as enables actions. The belief system can be communicated through formal methods (established mission and vision statements, statements of purpose and credos) and informal ways (such as through the socialisation process). For this reason, the belief system is also viewed as a form of social control and it influences the other three forms of control (Tessier & Otley, 2012;Widener, 2007).
Empirical research has provided insights into the importance of belief systems for communicating core values of organisations (Collier, 2005;Chenhall et al., 2010) and for facilitating organisational change (Marginson, 2002). Consequently, they affect and are influenced by other control mechanisms. Collier (2005) found that in the absence of formal system-based control in an entrepreneurial firm, the belief system is used as a lever of control and it influences and is reflected in the social control of the firm. Evans and Tucker (2015) found that among the four LOC systems, the belief system has a greater influence on facilitating the organisational response to change, as corporate values guide the change agenda. Similarly, Marginson's (2002) research on the role of management control system (MCS) in the development of ideas found that the belief control system facilitates the generation and filtering of ideas and thus guides the change process in an organisation. The insights from Chenhall et al.'s (2010) study on MCSs in non-governmental organisations (NGOs) highlight the importance of the belief system for adding value to an organisation, as it shapes and influences employees' behaviours and manages the tension between the employees' values and the values of the organisation. Moreover, the belief system was used to communicate the organisation's core values to current and potential employees. Chenhall et al.'s (2010) study also demonstrated that the dominant belief system mitigated the need for a boundary system in the NGO. The employees' bonding, or the strong relationships between employees, were further supported and developed by the organisational belief system. The belief system could also be reinforced through the use of performance measures (Tuomela, 2005). Hence, the findings from previous empirical research show the influence of the belief system on other control mechanisms and how it provides internal coherence and consistency in the array of control mechanisms used, especially social control. Thus, informal control systems can stipulate responsibility or accountability on behalf of formal control systems (Simons, 2005).
However, there is a lack of empirical literature on the use of MCSs in microfinance settings, specifically on how the belief system influences key organisational activities and other controls. Hence, examining the belief system as a control lever is important for commercial microfinance providers such as DFIs, given that engagement with microfinance is significantly different from that with other financial services, in part due to the social values underpinning the microfinancing programmes. Thus, the issues are how the belief system of a commercially oriented DFI shapes the microfinance activities and other control mechanisms in the organisation, and how the belief system influences the use of social control in the organisation, specifically personnel control (Merchant & Van der Stede, 2012). Personnel control is practised through selection and placement, training, the provision of resources, including informal and formal information transfers, such as transfer of knowledge and experiences, and best practices among organisational members (Merchant & Van der Stede, 2012;pp. 88-90).
In this study, we also address the criticism that the Simons control framework focuses mainly on the use of control by top management (Berry, Coad, Harris, Otley, & Stringer, 2009;Merchant & Otley, 2007) as the belief system is viewed as the dominant values and beliefs of senior management, while employees are considered passive actors (Tessier & Otley, 2012). However, research has shown that the creativity of middle managers also impacts organisational survival (Marginson, 2002). Moreover, the significance of the belief system to the organisation can be evaluated based on how it guides the organisational reaction to change (Evans & Tucker, 2015). Thus, our focus is not limited to senior management but encompasses the reactions and actions of other employees, i.e., the microfinance officers at the branches, to determine how the belief system shapes the actions of operational employees in microfinance practices and activities.

THEORETICAL FRAMEWORK
Many factors shape organisational practices, including organisational objectives and demands by certain groups of stakeholders to address specific issues. These factors provide a basis for discussing how such practices unfold within organisations and their relation to internal and external pressures. Indeed, such practices have recently been discussed quite extensively in the literature in light of the institutional logics approach (Lounsbury, 2007). This approach concerns how institutions, defined by Friedland and Alford (1991, pp. 242-243) as "supra organizational patterns of human activity by which individuals and organizations produce and reproduce their material subsistence and organize time and space", influence organisational practices (and/or belief systems), which unfold as direct and/or indirect effects of institutional pressures (Lounsbury & Boxenbaum, 2013). As Lounsbury (2008) argues, organisations inevitably have certain prevailing logics that underpin and/or shape organisational practices.
In this paper, we will demonstrate how the broad institutional features of microfinance (i.e., social orientation and poverty alleviation) relate to the specific nature of the DFI (whose main business concern is performing banking activities commercially) and the development of particular logics that influence the use of the MCS.

Dominance of the Banking Logic
Most organisations function within environments in which multiple institutional logics are present, and these diverse logics are reflected in the organisation's structures and practices (Greenwood, Raynard, Kodeih, Micelotta, & Lounsbury, 2011;Kraatz & Block, 2008). In some cases, "one logic can be so dominant that it eclipses other logics, rendering them immaterial to organizational functioning" (Besharov & Smith, 2014, p. 366). The (re)construction of a dominant logic is broadly presented as a means of identification, according to the content of the identity being constructed in terms of an identity label and its main attributes (Creed, Scully, & Austin, 2002;Suddaby & Greenwood, 2005). The dominant logic strategically contests other logics to achieve particular objectives (Pache & Santos, 2013).
Some institutional logics studies, such as those of Thornton and Ocasio (1999) and Lounsbury (2007), which focus on dominant logics, indicate that institutional actors are prone to change their behaviour accordingly. They argue that each institutional logic is attached to a particular subject, which subsequently creates some forms of imbalance and/or tension (see also Battilana & Dorado, 2010). For example, in the context of the banking industry, banking logic may be a dominant logic given that banks are concerned about their legitimacy and sustainability in offering banking products and services (Almandoz, 2012). Indeed, they adapt easily to their environment's banking requirements, which require them to be strongly rooted in banking regulations with an optimistic commercial perspective. This perspective, however, may be contested in the event where the banks (start to) infuse community-oriented programmes into their existing business model as they have to pursue some potentially competing institutional logics to meet conflicting demands (Marquis & Lounsbury, 2007), in which dominant logic(s) may outshine less dominant logic(s).
We argue that this is prevalent in the case of microfinancing offered by certain financial institutions as they address specific social needs that are driven by social logic. Social logic is structured around the predominant goal of providing services to address specific social needs. Economic resources are the means by which these organisations achieve commercial goals. Profit is viewed as a means to achieve the organisation's end goal and is thus reinvested in the organisation's social mission. The social logic prescribes control as the appropriate means of monitoring strategy and operations, with a great concern for people and their social needs. Battilana and Dorado (2010), who studied a microfinance institution, found that while some loan officers' work practices were influenced by the banking logic, others were influenced by the development (or social) logic. They also reported that the banking logic, which has a commercial nature, appeared to dominate banking operations and administrative procedures.
Thus, it is justifiable to examine whether dominant institutional logics are used within particular environments to accomplish certain institutional demands. Nevertheless, we need to acknowledge that organisational practices are constrained by the availability of resources and organisational capabilities and competencies (Battilana & Dorado, 2010;Pache & Santos, 2013;Chiwamit, Modell, & Yang, 2014). These factors can provide a basis for discussing how MCS is embedded and/or driven by institutional logics.

RESEARCH METHOD
A qualitative case study approach was used for this research. Data were collected over a period of seven months from March to September 2016. To understand the context of microfinancing and the role of DFIs, the first phase of this research constituted three hours of focus-group interviews with key senior managers of three DFIs responsible for the provision of microfinance services in Malaysia. The officers from the three DFIs had vast experience in microfinancing and were responsible for their DFIs' microfinance programmes. The senior executive from the headquarters' (HQ) microfinance department of Banco (disguised identity), the case site, was among those interviewed during the session.
Of the three DFIs involved in microfinancing, Banco was selected as the case site because of its microfinance segment's growth and profitability. At Banco, interviews were conducted at HQ's microfinance department, i.e., the department responsible for microfinancing. In the department, the Assistant Vice President of Microfinance (AVPM) and the senior executive of the microfinance department were interviewed because both had worked at Banco since the establishment of the microfinance department. As we seek to understand how the control system shapes the decisions made at the operational level, interviews were conducted with the branch managers and microfinance officers at the two branches with the highest disbursement record of microfinancing loans. The West Branch microfinance officer had occupied the job for only one year, thus we also interviewed the former microfinance officer of that branch to obtain a better understanding of the issues involved and to triangulate the findings. As shown in Table 1, we interviewed key officers who manage Banco's microfinance programme at different hierarchical levels. The Assistant Vice President of Accounting and Reporting at HQ was also interviewed to determine both the internal and external reporting and the monitoring of microfinance at Banco. Follow-up clarifications were obtained from these key officers, especially the senior manager of the microfinance department, through online communications.
Central Bank officers were also interviewed, as this bank plays a major role in DFI microfinancing, especially in the monitoring of the microfinance performance of DFIs. Similar interviews were also conducted with key officers from an NGO-based MFI in the country, in order to obtain insights into the background and development of the microfinance industry, and the role of DFIs in microfinancing, as well as assessing the performance of DFIs. The case study also draws data from a documentary review including organisational data and other public documents such as economic and Central Bank reports.
The data were analysed in several stages. The first stage involved intensive analysis of the economic and Central Bank reports to understand the characteristics of microfinance and to provide insight into the microfinance performance reporting requirements and the DFIs' microfinancing responsibilities. This is followed by the analysis of the internal documentary data such as Banco's organisational reports and information about their microfinance products and services, so as to gain an understanding of the background of the organisations, its key microfinance products and services and its strategies. Banco's annual reports since the introduction of its microfinance programmes were extensively reviewed to identify specific strategies, missions or visions pertaining to microfinancing, the client charter and microfinance performance. The second stage involved reviewing and analysing the interview transcripts several times to identify general themes arising from the data. The conceptualisation of the MCS guided us in determining the belief system of the organisation and how the belief system transverses organisational activities, specifically the DFIs' microfinance activities. In so doing, we specifically examined the similarities and differences in the data between the various data sources, i.e. between the microfinance officers at HQ and the bank and other officers. The data were categorised according to the key activities of the microfinancing service provision in terms of the financing approval process, the factors influencing the decisions, the internal practices such as the training for officers, etc. Then we mapped out the how the control system influences the loan collection process, the monitoring mechanisms and performance measures for the branch, HQ and microfinance department.
The third stage of the analysis involved comparing the evidence from the findings of the interview data and the documentary data to determine the contradictions between the formal documents and actual practice. At this stage, we reviewed the annual reports and organisational website again to determine Banco's stated core values and how the provision of microfinancing and the client charter are explained in the documents. We then compared this information with our interview data from the various sources. Hence, the interviews were triangulated with the documentary data to assess the gaps between official claims and systems and the practices implemented within the organisation. Thus, the credibility of our findings was enhanced by triangulating the data from the various sources across different organisational levels. In the final stage, we linked the categories of data together to explain how the belief system influences microfinance activities as well as other forms of control. Finally, the institutional logic perspective was used to explain our findings. Thus, the analytical concepts discussed in the framework were useful for providing explanations and answers to the research question (Siti-Nabiha, 2009). As such, the theory and our research questions provide a guide and framework for explaining the research findings.

FINDINGS
Banco was initially established in the 1950s to improve the socio-economic status of rural communities in Malaysia and to provide opportunities for their self-improvement. Apart from meeting the needs of large corporations, Banco also focuses on small and medium-sized enterprises (SMEs), including micro enterprises and rural industry players. Banco is one of the DFIs that are required by the Central Bank of Malaysia to provide microfinance services as stipulated in the 2009 Microfinance Institutional Framework. The framework provides that the selected DFIs must provide easy microfinancing for business purposes of up to RM50,000 (with no collateral), the fast disbursement of loans and convenient and widely accessible microfinance products. Previously, only NGO-based MFIs were involved in microfinancing, with a desire to alleviate poverty (Che-Zakiah, 2004). Banco offers microfinance services and continues to be strategically involved in ensuring the growth of this microfinance sector. Banco's decision to accept the national mandate to offer microfinance services follows one of its strategic thrusts, i.e. to ensure that there is an alignment between its strategic direction and the agendas of the government and other regulatory bodies, as stated in Banco's annual reports. Although its microfinancing portfolio is largely financed by the internal funds, the Central Bank has a "power" over the ways the microfinance programme is delivered in Banco.
The coercive power imposed by the Central Bank (being a supervisory and regulatory body for DFIs) has been regarded as a push factor for the growth of this sector. Such power is exercised by requiring Banco to submit reports and to undergo face-to-face meetings related to microfinance performance. This exercise has greatly influenced the management of the microfinance programme at Banco, as noted by the AVPM: Now the Central Bank monitors … So, we know [our performance] … If we don't achieve, they ask why. This is part of the financial inclusion agenda which we are subjected to.
In addition to the focus on the financial inclusion agenda, Banco faces pressures over profitability and sustainability. The reduction of development aid and a lack of governmental guarantees of funds raised have pressurised Banco to move towards commercialisation and profit orientation. At the same time, Banco needs to contribute to achieving the social economic agenda of the government. Such dual competing demands from the government have created tension, especially over running and managing the microfinance programme.
With limited experience in microfinance services, Banco started to develop its microfinance structures by hiring those with experience in microfinance and setting up a microfinance department at HQ. Banco continued to mimic the practices of some established MFIs by sending staff from different hierarchical levels, i.e., officers, branch managers and top management, for training and visits at those MFIs, in order to learn good practices that could be replicated in Banco. Consequently, Banco's microfinance model was massively adapted from their good practices. The defining characteristic of microfinancing, i.e., group, noncollateral lending with a regular repayment schedule, is practised at Banco. All microfinancing loans are offered for business purposes as per the MFI framework. However, in the last two years, due to issues in group-based lending, particularly group formation, Banco has also offered a new microfinance product: individualbased lending charged at a higher rate. Banco provides microfinance services through its existing branch network using the same tiers used for other financial services.
The above descriptions suggest that Banco has pushed really hard for the commercialisation and sustainability of the microfinance programme.
Interestingly, this appears to have further promoted the dominant banking logic which underpins its belief system and permeated the ways MCS is used to manage the microfinance programme.

Belief System Underpinned by the Banking Logic
The belief system in Banco is rooted strongly in the banking logic and it shapes the actions of members of the organisation and acts as a boundary system by providing guidelines for acceptable behaviour by organisation members. Although Banco's microfinance service differs from other banking services due to its nontraditional approach to lending, the belief system seems to have exerted a great deal of influence over the microfinancing that is being offered. Banco's belief system is clearly elucidated in its risk-return considerations and the minimisation of non-performing loans (NPLs). It shapes its decisions relating to: (i) rates/fees charged, (ii) loan recipients, (iii) loan assessment criteria, and (iv) loan collection processes.
The belief system is used to communicate the organisational core values through the guidelines for processing and loan approval, which also set the boundaries for microfinance activities, particularly those pertaining to loan activities and collections. This banking logic is shown in the targeting of microfinance clients with the main emphasis on loan quality, profitability and risk minimisation. This is clearly reflected in the statement by the former West Branch microfinance officer: [the branch manager] said it is not useful to have a lot of loans, but of low quality… if there [is] a lot of loans, but all do not pay, the bank will suffer losses.
Risk minimisation has shaped the loans approval process as loan are given to those that offer lower risks, i.e., mainly to those with business track records such as two years of operating experience. As AVPM commented: We focus on the not so poor [client]... he must be in business already for two years. He must have experience. The bank does not want to take the risk.
Consequently, Banco focuses on the breadth not the depth of outreach as it targets poor, not hard-core poor, clients.
The belief system, underpinned by the banking logic, shapes Banco's decisions regarding the value of loans and the approval process. This is shown in the decision to maintain loan quality and to ensure that branch NPL does not exceed the stated limit. The desire to maintain loan quality and minimise NPL resulted in action to ensure client's ability to pay by checking applicants' credit reports and reference systems, and conducting site visits to assess the condition and potential of the application business, especially if the client does not have formal records. To minimise risk and NPL, a lower value loan may be given to those clients who have one or two outstanding payments. The provision of loans is tightened when branch NPL exceeds the target limit, which means that even those with just a single outstanding payment will not be awarded microfinance loans. As a result, loans are provided to clients who have shown their ability to pay based on a good payment record of six months, together with a good business cash flow.
The focus on minimising NPL has a significant influence on the amount of credit approved by Banco. Banco usually gives only a smaller amount for the first loan. Only after the client has shown an ability to pay it back will the bank increase the loan amount. Even for subsequent loans, Banco will not give a large or lump-sum loan amount as clients do not have the ability to repay larger amounts, as indicated by the West Branch manager: We will not give a lump sum as we notice this is the root cause [of failure to repay]. Initially his payment is good. When we give for a second time and we give too much, he can't pay.
Clearly, Banco has placed much emphasis on the risk-return considerations in its microfinance programme. Apparently, such considerations have been reflected in the rates (or fees) charged for the loans disbursed. For loans derived from funding from government sources, Banco charges minimally or interestfree. Meanwhile, loans disbursed using Banco's internal funds are charged at higher rates to cover the operational costs and to absorb the default risk resulting from borrowers' inability to repay the loans, which have neither collaterals nor guarantors.
In view of the above, it can be argued that Banco's belief system is underpinned by the banking logic that appears to be dominant in managing the microfinance programme and which acts as a control level in guiding employees' actions and decisions. Succinctly, such a programme is managed through the use of similar MCS mechanisms that are not uncommon in commercial banking practices in Banco. This belief system acts as a control lever that is also shaped and reinforced by the other control mechanisms, specifically the performance measures used internally to monitor Banco's microfinance performance.

Performance Measures and Reporting
The belief system, which emphasised the core value of risk-return consideration, was further reinforced by the external measures used to monitor Banco's microfinance programme and this consequently shaped the internal performance measurement and monitoring system. The Central Bank monitors the microfinancing performance of the bank and industry as a whole, based on commercial indicators. These indicators, which have to be reported by banking microfinance providers, consist of (i) outstanding balance; (ii) approval rate/ cumulative approvals; and (iii) disbursed amount and impaired financing ratio (NPL). These measures are cascaded down to HQ's microfinance department and to the branch as the measures related to microfinancing, i.e., loan disbursement, NPL, collection rate, loan growth, group formation and customer complaints, are commercial in nature and similar to those reported externally.
Moreover, the belief system based on banking logic is perpetuated through the measurement system used to evaluate branch performance, which consequently leads to a lack of emphasis on microfinancing at the branch. As the microfinancing portfolio is small, it has an immaterial impact on the branch's KPIs; hence, it is not surprising that microfinancing is viewed by a branch manager as a corporate social responsibility (CSR) activity rather than as an integral part of Banco's arms-length business transactions.
The social logic underpinning microfinance did not flourish in Banco due to the separation of microfinance from other banking activities. The way the microfinance is structured in Banco also reflects the belief system and ensures the domination of the banking logic. This is done through the "separation" of microfinancing from the overall banking activities, as reflected in the responsibility structure of the microfinance programme and the performance evaluation of those involved. Even though the performance of microfinancing is the responsibility of the HQ microfinance department, the department has limited flexibility for managing performance as the microfinance activities mainly occur at the operational levelin the branch under the responsibility of the microfinance officers, who report to their respective branch managers. Microfinance performance, to a certain extent, is positioned outside of the department's control, despite the fact that the department manager's rewards and appraisal are linked to this performance. Similarly, at the branch level, microfinance is also "separated" from other banking activities as microfinancing performance measures are the responsibility of microfinance officers, who are responsible for the entire financing process, from processing loans to the control of problematic accounts and loans recovery. This represents a departure from other banking products at the branch, where different units are responsible for each type of activity.
As such, the microfinance performance only affects the microfinance officer's individual performance, not the branch's performance. The performance of the microfinance officer is also geared towards minimising NPL with the collection effort, in particular, being an important evaluation criterion, as the West Branch manager explained: The main thing that I look [for in the performance of microfinance officers] is the collection.... At least 80% of the total collection, [the microfinance officers] must get.
Moreover, the non-achievement of performance targets by the microfinance officer would not have a significant impact on his performance if the officer also assisted in other activities at the branch. The branch manager conducts reviews with the microfinance officer related to externally measured and monitored microfinance indicators, as well as activities that support microfinance, and actions to address the NPL. The belief system is underpinned by the core values of risk-return consideration and minimising NPL, and this is also reflected in the emphasis on the monitoring of microfinance NPL, as it will impact the overall branch NPL.
The insignificance of the microfinance portfolio is also reflected in the reporting formats, as microfinancing is classified alongside "education and personal financing" and its profit is reported under "other income". The detailed reports on microfinance performance are prepared by the microfinance department and comprise information on the approval rate, collection rate, disbursement and NPL. The microfinance department also sends monthly performance reports to the branches to provide a performance overview of the branch and its peers. The department also issues reports on key areas of the microfinance programme to the Senior Vice President for Services and Products, which include information on the approval rate, collection rate, disbursement and NPL. Thus, the use of commercial indicators for evaluation and monitoring are cascaded throughout the organisational levels, which leads to further sedimentation of the dominant banking logic, which is reflected in the belief control system and which also shapes and influences other formal controls, i.e. the performance measurement and reporting system.

Social Control
Microfinancing clientele and activities are different from normal banking as microfinance clients, compared to clients of other products, are mainly disadvantaged groups with lower educational backgrounds and in most cases, they do not have proper financial records. As such, social control, specifically through the selection and placement of informal and formal information transfers, was utilised to ensure the sustainability of the programme. However, while the social control system is used to ensure that microfinance operations are smooth, it is also shaped by the Banco belief system, thus further perpetuating the belief system.
In addition to formal communication, managers communicate the dominant belief system informally through the socialisation process, internal networking, informal meetings and discussions over achieving its banking objectives. Hence, the belief system in Banco influences the way that microfinance activities are structured and managed. Initially, when Banco started its microfinance programme, DFI recruited a well-trained microfinance officer from an NGO-based MFI to head the department that is responsible for managing the microfinance programme and overseeing branch activity in delivering microfinance products. Both the head of the department and the deputy had longserved the NGO-based MFI and they appear to possess wide experience and skills in microfinancing. Subsequently, it can be seen that the microfinance department does have some influence on the selection of microfinance officers. This selection is based on certain characteristics, such as the ability to communicate with the 'elderly and disadvantaged groups'. The same characteristics are also required by the branch managers, who expect their microfinance officers to be well-known in the community and to practise close client engagement through frequent visits with the customer, usually consisting of weekly visits for collection or twiceweekly visits for problematic accounts. Nevertheless, the social control is shaped by the belief system of ensuring loan quality and minimisation of NPL, as closer customer interactions with microfinance clients are crucial for maximising collection and minimising NPL. Thus, the emphasis of the selection requirements is to obtain microfinance officers who are able to engage with the client in order to achieve the key objectives of ensuring collection and minimising NPL, in alignment with the belief system in Banco.

Formal and Informal Information Transfer
The sustainability of the microfinance programme is also leveraged through formal and informal control systems in Banco, which are also shaped by the Banco belief system. The external and internal networking and information transfer are central in its decision-making processes related to loan approval and collection processes. Apart from using some procedures to guide loan evaluation, the subjectivity of decision making and closer interactions with clients also require tacit knowledge. Together with an objective assessment based on Banco's guidelines, loan approval requires subjective assessments because financial records alone are seen as inaccurate predictors of a good payment record.
This transfer of information, especially of tacit knowledge, is achieved in various ways. One way is for the officers to visit other branches that offer microfinancing in order to become familiar with microfinancing by observing and learning how the work is done. Formal training by the microfinance department is another means of accomplishing the socialising process and transfer of knowledge. Information transfer is accomplished by formal training at the microfinance department at HQ, which usually covers subjective evaluation, as well as sharing the experiences of the branches, including several days of field visits to wellperforming branches to learn from them. The performance-reporting process also leads to information sharing on best practices in microfinance by Banco. The HQ microfinance department sends the monthly microfinance performance report to the branches. The microfinance officers usually seek to align their activities with those of other branches, and this has indirectly created competition among the microfinance officers over best practices and performance, which are ultimately based on commercial indicators, thus further perpetuating and reinforcing the belief system at Banco.

DISCUSSION AND CONCLUSION
This paper has explained the role of the belief system in influencing formal and informal control mechanisms and in shaping the microfinancing decisions in Banco. It can be argued that the power of a control system lies in how it is used to balance the competing demands faced by an organisation (Heinicke, Guenther, & Widener, 2016;Simons, 1995). In microfinancing, the tension or competing demands faced by microfinance providers are between economic and social considerations (Siti-Nabiha et al., 2018). In the case of Banco, there has been a push for this DFI to support the government's call to accommodate the needs of the poor, and to operate its microfinance business profitably. Its aims appear to be ensuring the programme's sustainability and the DFI's economic survival.
Generally, DFIs in Malaysia have moved towards commercialisation due to the government's pressure to be self-sustaining and to not be too dependent on the state's financial assistance. However, DFIs have to fulfil specific social roles, a fact that is reflected in the coercive pressure on the organisations to offer microfinancing. The social objectives of financial inclusion and poverty alleviation underpin microfinance programmes. The labour-intensive nature and higher transaction costs of microfinancing may have impinged DFIs' profitability and/or may have given rise to opportunity costs, which results in losing more profitable banking products as time and resources are utilised for microfinancing. It would be more profitable for banks to engage in normal financing; as one DFI officer said, "microfinancing in volume is high, but in financial terms, it's low". Due to the coercive pressure to provide microfinancing, these organisations must address the tension between their social objective of providing greater access to financially excluded individuals and/or groups and their commercial objective of being profitable banking institutions. It has been argued that MCSs can be used to balance this tension. However, our findings show that it is difficult to achieve a balance between economic and social considerations, given that the former are often prioritised over the latter due to the dominance of the banking logic, which represents its belief system.
The findings also show that the inherent value or dominant logic in an organisation influences its MCS (Heinicke et al., 2016;Chenhall et al., 2010). Profit orientation, which is sedimented and internalised in the organisation over the years, is reflected in the banking logic and shapes both the formal and informal control systems used. Consequently, the organisation's practices correspond to the coercive pressure to provide microfinancing, which is in accordance with its dominant logic. The use of such control systems preserves this dominant core so that microfinancing decisions are aligned with the banking culture; i.e., profitability and the minimisation of business risk. Similar to insights from previous research such as Collier (2005) and Chenhall et al. (2010), the findings of this study also highlight the key roles of the belief system and social control in perpetuating the banking logic. The belief system and social control are inherently consistent and complement each other to achieve the objective of the microfinancing programme at the bank. Although not explicitly and formally stated, we have observed that the underlying objective driving the case organisation's microfinancing is to ensure a sustainable microfinance programme which operates with self-sufficiency and is able to generate sufficient profits for future growth. The control systems used, i.e., the belief system and social control, aim to achieve this objective.
In contrast to previous research showing the impact of the belief system and informal control system in channelling change in an organisation (Evans & Tucker, 2015;Marginson, 2002), our findings have analysed the roles of the belief system and social control in ensuring stability and in perpetuating the banking logic. Microfinance practices underpinned by social logic do not lead to changes or the infusion of social logic into the organisation. Moreover, the lack of an interactive control system and the lack of emphasis on a diagnostic control system reveal a loose control system in terms of achieving microfinancing targets and objectives in the organisation. Similar to Tuomela (2005), our findings have showed the way the belief system was reinforced through internal measures used to assess performance. More crucially, those measures are based on those being monitored by the regulatory bodies and/or stakeholders, who place a great deal of emphasis on commercial performance. Meanwhile, social performance is merely measured by the growth of the microfinancing segment. An increased focus on external monitoring and assessing the social development of the programme might increase the infusion of the social logics into the organisation. Such an assessment might lead to real tension between social and economic performance and, consequently, the leveraging of the control system to address the two competing objectives at Banco.
The structure of microfinance also reflects the perpetuation of the banking logic in the organisation. Although microfinance is treated like any other banking product from a commercial viewpoint, its activities are separate from other banking activities in the organisation. The microfinance officer is involved in all aspects of microfinancing activities, in contrast to other products, which utilise specific functions and officers for specific activities. As a clear illustration of the compartmentalisation of microfinancing, the responsibility for its growth mainly rests with the microfinance department at HQ. There is a lack of emphasis on microfinancing at the branches because microfinance targets represent a small percentage of branch targets. Indeed, one of the branch managers noted that microfinancing is viewed as a CSR activity and not a core activity of the bank. This position is in contrast to the view of the microfinance department at HQ, which is that microfinancing is part of the core activities of the bank, or as commented by the head of the microfinance department at HQ: "…being a DFI, we cannot neglect microfinance as it is a part of our existence and core business". However, this view has not been well embraced by Banco's branches. Thus, it is not surprising that the performance of microfinance officers at the branch level is loosely tied to microfinancing targets as part of Banco's belief system. Furthermore, the officers at the branch are bound by the belief system of minimising business risk and considering profit, due to pressure to meet the respective branches' banking targets.
Clearly, the belief system in Banco is strongly rooted in the dominant banking logic. Interestingly, hiring carriers of the social logic with broad experience in microfinancing (Battilana & Dorado, 2010) has not altered the belief system of the organisation or led to significant changes in the organisation. The established dominant belief system, underpinned by the banking logic and the compartmentalisation of microfinancing in the organisation, has hindered the emergence of a sub-culture; responsibility and accountability mainly rest at the microfinance department at HQ. The carriers of social logic used informal social control in the organisation, but with a focus on ensuring profitable microfinance growth. Knowledge transfer and information sharing on skills in dealing with clients, assessing client potential, hiring and the socialisation process, ultimately allowed sustainable growth and released the bank from dependence on government funds for its microfinance activities. Indeed, the socialisation process, which is performed to ensure that the commercial objective is achieved, clearly resembles the banking logic. This banking logic exhibits a short-term focus on results rather than on long-term goals of social development (Akanga, 2017).
In conclusion, there appears to be a significant development in the roles of belief control and social control in the organisation; both forms of control seek to ensure stability and to perpetuate the banking culture in the organisation. Both control systems are used to ensure that the organisational goals are achieved. Even hiring carriers of social logic has failed to alter the goals and control system used due to (i) the vague structure of microfinance; (ii) the compartmentalisation of microfinancing; and (iii) a lack of long-term social development goals in the internal and external monitoring of the microfinance programme. The social logic is evident in the mimicry of ceremonial microfinancing practices of other established financial institutions. Regardless of this mimicry, the banking logic appears to be dominant in Banco as the belief system is more closely aligned with banking practices that emphasise profitability and efficient handling of business transactions. In addition, the microfinance client charter resembles those of existing commercially oriented banking products. Thus, it can be argued that the Banco belief system reinforces the commercial orientation and banking culture, as reflected in risk-return considerations and interest rates and/or charges with little consideration of the social objective and/or logic.
Social control is implemented to ensure better client engagement and the transfer of tacit knowledge to achieve profitable growth. Although the provision of microfinance by Banco appears to build on social logic, we have observed little dominance of this logic in the organisation. We found that while the social logic is used for its primary mission of complementing the government's mandate, banking logic is used for the success criteria for this mission. Moreover, the provision of services was externally imposed on the bank and did not develop internally. The social logic is only implicated in services such as advisory activities for clients, but even then, the objective is to ensure business success as measured by loan repayments and profit targets. Thus, this DFI is said to have dealt with both the social logic that reinforces the governmental agenda to help the poor, and the banking logic that conforms to commercial banking practices, so as to generate sufficient profits to enable it to meet its financial objectives.