THE DESIGN AND USE OF PERFORMANCE INFORMATION IN INDONESIAN LOCAL GOVERNMENTS UNDER DIVERGING STAKEHOLDER PRESSURES

SUMMARY

Public sector managers in less developed countries are confronted with opposing forces. A lack of sufficient resources along with a tradition of corruption are obstacles for developing and using performance measurement systems. However, recent public sector reforms in less developed countries, including decentralisation and anti-corruption programmes, stimulate the development and use of such systems. On the basis of a framework, which distinguishes different types of stakeholders, each with particular performance interests, we analyse how public sector managers are coping with the two opposing forces, given the relative power positions and the interests of their stakeholders. On the basis of four cases studies of local government agencies in Bali (Indonesia), we found that with respect to the annual performance reports, managers in these agencies focus more on fulfilling the formal requirements regarding the format of these reports and on their timely submission than on their contents, which are all symptoms of a symbolic rather than functional use of performance information. However, the reports include information on inputs that is linked to similar information in short-term reports, which the managers use in a functional manner. These managers show a kind of juggling behaviour, in the sense that they partially try to serve conflicting performance interests. Copyright © 2012 John Wiley & Sons, Ltd.

INTRODUCTION

Traditionally, the design and use of performance measurement systems in government organisations in less developed countries (LDCs) were problematic, not only owing to a lack of resources but also because of, for instance, widespread corruption practices (Dean, 1988, p. 152; Mimba et al., 2007; cf. Hopper et al., 2003, pp. 16–17). However, recent reforms in these countries — including decentralisation and marketisation programmes — are expected to lead to major changes. The reforms tend to increase the demand for performance information because government organisations are held accountable for their performance and often also because they have to improve their performance (Mimba et al., 2007; see also Leeuw et al., 1998; Smoke, 2001, p. 27). How the managers of these organisations react to this increased demand for performance information, for example, by developing and using particular performance measurement systems, is an intriguing research theme.

However, although public sector performance measurement — and more generally, public sector management accounting — has become an important research domain (van Helden, 2005; Broadbent and Guthrie, 2008), a large majority of the articles in this domain published in academic journals are mainly focused on developed countries. Therefore, Hopper et al. (2009) argue that more research on management accounting systems in LDCs is needed. Our article is an attempt to fill this gap. Our objective is to increase our understanding of the role of performance information in Indonesian local government organisations. More precisely, we will investigate the use of performance information by the managers of a few local government agencies. We will also pay attention to the LDC-specific context in which they operate and the relative power positions as well as the performance interests of their stakeholders.

This article proceeds as follows. The next section introduces the context in which public sector organisations in LDCs operate, including its influence on the role of performance information. The subsequent section develops a theoretical framework that will help us to understand how public sector managers in LDCs design and use performance measurement systems under partly conflicting pressures of stakeholders and also given the specific context in which they operate. The next section addresses the selection of sites for our empirical investigations and discusses the research methods we used. The subsequent section presents the findings from our empirical study. The final section provides conclusions and reflections on our findings.

LESS DEVELOPED COUNTRY CONTEXT AND ITS IMPACT ON THE ROLE OF PERFORMANCE INFORMATION

Public sector organisations in LDCs are characterised by a low institutional capacity, which means that their ability to decide on and pursue their goals, to perform tasks and to improve performance is limited (World Bank, 2004). This low institutional capacity can be explained by the interplay of different factors, ranging from a lack of tradition regarding governmental institutions and sometimes a high degree of politicisation of decision making in the arena of bureaucrats to — more pragmatically — a lack of resources to build up institutions and to increase competences of government employees. Symptoms of a low institutional capacity are a weak regulatory practice, a low level of public accountability, administrative inefficiencies and a lack of facilities (Olowe-Okere and Tomkins, 1998, pp. 320–321; IMF, 2002; Manning, 2001, p. 303; Batley, 2004, p. 35; Wickramasinghe and Hopper, 2005, p. 475).

A low institutional capacity is a hindrance for the design and use of adequate accounting systems, including performance measurement systems. Dean (1988, p. 152), for example, observes that some LDCs lack good financial managers and a system of government audits (of sufficient quality), whereas their government accounting records are inaccurate and unreliable. Such circumstances will hinder the design and use of performance measurement systems, which build on more basic accounting systems. Another factor that hinders the design and use of performance measurement systems in LDCs is the high level of corruption (van Rijckeghem and Weder, 1997). The reason for this is that corrupt public sector officials and civil servants tend to make decisions mainly on the basis of an assessment of their personal gains rather than on an assessment of the implications for organisational performance (Mimba et al., 2007; cf. Hopper et al., 2003, p. 18). A final hindering factor is that for a long time neither the market nor stakeholders in LDCs exerted strong pressures on public sector managers to attain performance improvements (cf. Splettstoesser, 1998; Graves et al., 2002). Often, funding bodies were the only type of stakeholders that had a substantial influence on public sector organisations (Tillema et al., 2010, p. 209). Together, these factors imply that the role of performance information in LDCs used to be quite limited (Mimba et al., 2007; Hopper et al., 2003, pp. 16–17).

More recently, however, important changes are taking place in the institutional environment of public sector organisations in LDCs, now that particular stakeholders, including international donor agencies, require these organisations to implement public sector reforms (ADB, 1997; IMF, 2002; Hopper et al., 2003, p. 16; Smoke, 2001, p. 29). These reforms include decentralisation, marketisation, the introduction of direct elections and the implementation of anti-corruption programmes. The reforms imply that next to funding bodies, also other groups of stakeholders of public sector organisations are becoming powerful. For instance, direct elections and decentralisation imply that elected bodies, also at the local level, become more important, whereas marketisation increases the power positions of purchasers and clients. In addition, the reforms stimulate the demand for performance information; because stakeholders require organisations to be transparent about their performance, they hold them accountable for this performance and demand performance improvements (Leeuw et al., 1998; Sanderson, 2001; Smoke, 2001, p. 27).

On the basis of the aforementioned discussion, we can conclude that a tension exists between the increasing demand for performance information as a consequence of the public sector reforms and a hindrance to meet this demand coming from a low institutional capacity and a high level of corruption. This tension raises the question of how the increasing demand for performance information influences the design of public sector organisations' performance measurement systems and the use of the information produced by these systems by the organisations' managers. The following section develops a theoretical framework that can be used as a starting point for answering this question.

THEORETICAL FRAMEWORK

The previous section described how the recent public sector reforms in LDCs lead to a greater involvement of stakeholders in public sector organisations. These stakeholders pressure organisations to provide performance information and to make performance improvements. To analyse the consequences of the changing stakeholder pressures for the design and use of organisations' performance measurement system, we use a revised version of a neo-institutional framework developed by Brignall and Modell (2000) (see also Tillema et al., 2010). This framework assumes that each group of stakeholders has its own performance interests.

Powerful stakeholders and their performance interests

For local government agencies, which will be the subject of our empirical investigations (see the next section), we distinguish three groups of powerful stakeholders: funding bodies, statutory boards and purchasers.

Funding bodies, such as central government and donor agencies, provide a public sector organisation with financial resources. Although they may specify for which (groups of) tasks these resources can be used, they do not identify the quantity of services that should be delivered by the organisation (cf. Groot, 1999, pp. 357–360). This distinguishes funding bodies from purchasers, which will be defined in the succeeding paragraphs. Funding bodies represent the interests of tax payers, and are therefore interested in the ‘disciplined use of financial resources’ (cf. Brignall and Modell, 2000), and hence in performance indicators on inputs. However, given the prevalence of corruption in the public sector in LDCs, funding bodies are also interested in the extent to which public sector organisations are transparent about and account for their use of financial resources (cf. UNDP, 1997; Agere, 2000). Therefore an adequate ‘financial governance structure’ is also an important requirement for receiving funding from international donor agencies nowadays (ADB, 1997; UNDP, 1997).

Statutory boards, such as local councils and regulatory bodies, are organisations or bodies that have representative/democratic or supervisory functions in a broad sense. They can represent the interests of voting citizens, or they can assess the organisation's conformance with legislation and other formal requirements. Statutory boards emphasise performance aspects that are related to their representative or supervisory function, such as the degrees of responsiveness, participation and political conformance (cf. Andrews and Shah, 2003, p. 3.4; Boyne, 2002). These are aspects of the organisation's ‘political governance structure’.

Similar to funding bodies, purchasers provide a public sector organisation with financial resources. However, in contrast to funding bodies, purchasers pay a price per unit for the services delivered by the organisation, implying that there is a relationship between the amount of funding and the quantity of service delivery (cf. Groot, 1999, pp. 357–360). Furthermore, if purchasers are unsatisfied with the organisation's services, they may decide to switch to another supplier. Examples of purchasers are acquiring parties in situations of purchaser–provider splits or of competitive contracting. However, in local government agencies, also top-level managers within the same local government organisation (e.g. the head of a municipality) who contract for the provision of services with the agencies' managers can be regarded as purchasers. Usually, individual clients do not have sufficient power to influence an organisation, but purchasers can be expected to partly represent their interests. This implies that purchasers are interested in the ‘quantity and quality of services’, in addition to the ‘costs of services’ (cf. Brignall and Modell, 2000; Tillema, et al., 2010). These performance dimensions are reflected in performance indicators on outputs and on links between inputs and outputs, respectively.

Table 1 summarises the aforementioned discussion. The stakeholders mentioned in this table should be regarded as stereotypes. In practice, stakeholders may show similarities with more than one of these stereotypes (e.g. when they combine a role as funding body with a role as statutory board), or they may have a hybrid role (e.g. when they link their funding to particular performance indicators but do not pay a price per unit).

Table 1. Powerful stakeholders and their expected performance interests

Funding bodies	Statutory boards	Purchasers
* Not a specific type of performance indicators.
Expected	Expected	Expected
• Disciplined use of financial resources (inputs)	• Political governance structure*	• Quantity and quality of services (outputs)
• Financial governance structure*		• Costs of services (inputs versus outputs)

The design and use of performance measurement systems

On the basis of Table 1 and our understanding of the context of LDCs, we can derive expectations with respect to the design and use of performance measurement systems in local government agencies in LDCs. The previous section observed that traditionally, funding bodies were the only group of stakeholders in LDCs with a strong power position. Brignall and Modell (2000) argue that in situations where one stakeholder is dominant, his or her performance interests will be served by the organisation's manager. Performance information related to these interests cascades down the organisational hierarchy and is used functionally, implying that it gives guidance to decisions. On the basis of these arguments, it can be expected that the performance measurement systems of local government agencies in LDCs were dominated by the performance dimensions ‘disciplined use of financial resources’ and ‘financial governance structure’ and that performance information related to these dimensions was used functionally. However, it should also be noted that the low institutional capacity and the high level of corruption hindered the supply of performance information and — as far as it was available — stimulated a symbolic use of this information.

The recent public sector reforms lead to a situation in which funding bodies increasingly have to share their power with other stakeholders. More particularly, statutory boards are becoming more important owing to direct elections at the local government level combined with the implementation of decentralisation programmes, whereas the influence of purchasers is increasing owing to the marketisation element of the public sector reforms and the associated establishment of formal purchasing functions. Therefore, we expect that the performance measurement systems of local government agencies in LDCs now also reflect the performance dimensions ‘political governance structure’, ‘quantity and quality of services’ and ‘costs of services’. However, these performance dimensions may conflict with the performance dimensions that funding bodies are interested in. In such a situation, where two or more stakeholders with diverging interests are powerful, some forms of decoupling of performance measures between the top and lower levels of the organisation are likely, giving room for a symbolic use of performance information (Brignall and Modell, 2000). This implies that performance information has a role in legitimising the organisation to its stakeholders but is not used as an input for decision making.

However, in contrast to Brignall and Modell (2000), we expect that particularly for local government agencies, purchasers can play a mediating role between the two other groups of stakeholders. They are the top managers of the same government organisation, implying that they are held responsible for the performance of the agencies as well. Moreover, their performance interests partly overlap with those of the funding bodies and statutory boards. That is, funding bodies and purchasers share an interest in financial issues, whereas statutory boards and purchasers share an interest in service delivery. Hence, purchasers may be able to reach compromises between the different groups of stakeholders, creating the possibility of a functional use of performance information. Nevertheless, it should be acknowledged that although the reforms may improve the institutional capacity and lower the level of corruption, these changes are taking place only gradually. Therefore, we expect that much of the information produced by the broader performance measurement systems — and especially those parts that reflect the interests of the stakeholders who have become powerful recently — will be used symbolically rather than functionally.

Figure 1 summarises this discussion. It indicates that the current public sector reforms have two opposing effects on the design and use of performance measurement systems. Traditionally, funding bodies were the only powerful stakeholder, but a functional use of performance information related to the funding bodies' interests was hindered by the LDC-specific context. More recently, the public sector reforms reduce the influence of the LDC-specific context, but at the same time, they imply that other stakeholders (with diverging interests) become more powerful, which hinders a functional use of performance information. The specific circumstances will determine whether the reforms will increase or decrease the functional use of performance information.

Figure 1. The influence of contextual circumstances and stakeholders' power positions on the design and use of performance information.

The theoretical framework as presented in Figure 1 gives rise to the following research question: How do the (changing) relative power positions of stakeholders and their performance interests influence the design and use of performance measurement systems by public sector organisation organisations in LDCs, given the (changing) contextual circumstances these organisations are faced with?

RESEARCH SITES AND METHODS

The empirical part of our research is concerned with Indonesian local government organisations. Agencies within these organisations are our focal point. These agencies have a relatively high degree of autonomy, which implies that they represent the level of the local government organisation at which it is decided how the organisation will deal with the influences of different stakeholders. The agencies we selected are units that are responsible for delivering services to citizens. For conducting our case studies, we selected agencies in two local governments located in Bali Province: Denpasar City and Jembrana Regency. Denpasar city is located in an urban area, whereas Jembrana Regency is located in a rural area. This difference may have an impact on the role of performance information, as local governments in urban areas are usually better able to raise local revenues (which could be used to increase their institutional capacity) and have more educated citizens (which could imply that the local council — being the statutory board that represents these citizens — is more critical).

Within each local government, we selected two agencies or district bureaus, that is, the District Bureau of Public Works and the District Bureau of Health.1 These agencies were selected because the different nature of their activities might influence the role of performance information. More precisely, as public works agencies have more tangible outputs and more standardised services than health agencies, they are expected to be better able to assess their performance. The data for the case studies were collected through 20 interviews, a large number of performance reports and related regulations issued by the Indonesian central government as well as observations of several meetings in which these reports were discussed.

An important step in analysing our data was the classification of the performance interests of the powerful stakeholders. For this purpose, we tried to categorise all the remarks of the interviewed stakeholders regarding their performance interests by linking these remarks to the dimensions mentioned in Table 1. For remarks that did not fit into any of these dimensions, we developed new dimensions. Whenever a remark was difficult to categorise, the three researchers discussed it and took a decision. After the performance interests had been classified, the performance reports were analysed to see to what extent they included information that was related to the performance interests of the stakeholders. Finally, the data obtained through the interviews and the attendance of meetings were used to examine how this information was used by the district bureau managers.

EMPIRICAL FINDINGS

The core of this section is the explanation of the use of performance information by the four district bureau managers. On the basis of our theoretical framework, we expect that their use of information is influenced by the performance interests of their powerful stakeholders and by the LDC-specific context in which they operate. The LDC-specific context is expected to influence the managers' use of information not only directly but also indirectly through its effects on the supply of performance information. Therefore, this section first analyses the performance interests of the powerful stakeholders and the design of the performance measurement system in the four district bureaus before it analyses the use of performance information by the district bureau managers.

Powerful stakeholders and their performance interests

The Indonesian government has implemented several public sector reforms, including decentralisation, marketisation, direct elections and anti-corruption programmes. In the district bureaus we investigated, we found that these reforms have led to an increase in the number of powerful stakeholders. It appeared that all four bureaus show similarities in terms of their powerful stakeholders, the roles and performance interests of these stakeholders and the dominant stakeholder as perceived by the district bureau managers. The three most powerful stakeholders, ranked from more to less powerful, are the mayor/regent, the local council and the central government. The managers of all four district bureaus regard the mayor/regent as the dominant stakeholder.

In both local governments, the mayor/regent2 allocates funding to the district bureaus through the local government budget. As such, he or she is the linking pin between the political and administrative levels of the local government organisation. The mayor/regent also has a significant influence on the district bureaus' performance targets, and he or she evaluates their performance achievements. Moreover, he or she has the authority to replace a district bureau manager at any time. These circumstances imply that there is a link between the amount of funding allocated to a district bureau and the performance that it has to achieve. Therefore, the role of the mayor/regent is roughly similar to the role of the ‘purchaser’ in our theoretical framework. However, it should be noted that mayor/regent does not meet all the characteristics of purchasers in their purest form, as he or she does not pay a price per unit of services provided.

The second powerful stakeholder is the local council, and more particularly, specific commissions of this council. In the two local governments, the local council has three broad roles. First, it gives approval for the local budget, including the allocation of funding among district bureaus. Given this budgeting role, we classify the local council as a ‘funding body’. Second, it issues local regulations and approves the drafts of the regulations prepared by the mayor/regent. Third, it exercises direct supervision over the local government organisation, including its district bureaus. Given these legislative and supervisory roles, we classify the local council also as a ‘statutory board’.

The third powerful stakeholder is central government, and more particularly, the Minister of Finance, the State Minister for Administrative Reforms, the Minister of Home Affairs and the Minister of Health (for the two District Bureaus of Health) or the Minister of Public Works (for the two District Bureaus of Public Works). The Minister of Finance has an important role in establishing the amount of funding from the central level to local governments, whereas the State Minister for Administrative Reforms and the Minister of Home Affairs have an essential role in evaluating the performance of all local governments and in defining the performance reports that have to be submitted. The Ministers of Health and of Public Works play indirect roles in monitoring the performance of all district bureaus that undertake activities in their policy area. However, although central government has various roles, its main role is providing funding to local governments. The revenues of local governments are mainly sourced from this funding, implying that also the funding for the activities of the district bureaus is largely provided by the central government. As these funding relationships are not complemented by requirements regarding the quantity of services to be provided, central government is classified as a ‘funding body’ rather than a ‘purchaser’.

Our research showed that the powerful stakeholders of the four district bureaus are interested in a broader spectrum of performance dimensions than we expected based on their roles and on the links between roles and performance interests in our theoretical framework. Central government — as one of the funding bodies of the district bureaus — also shows an interest in ‘outcomes of services’ and ‘value for money’, which were not included in our theoretical framework. These performance dimensions are related to performance indicators on outcomes and on links between outcomes and inputs, respectively. The other funding bodies — that is, the local councils in the two municipalities — have no additional performance interests. However, in their role as statutory boards, the two local councils are also interested in the ‘quantity and quality of services’, which our theoretical framework exclusively attributed to purchasers. The mayor and the regent, being the purchasers of the district bureaus, also pay attention to the ‘disciplined use of financial resources’ and the ‘political governance structure’, and the mayor also to the ‘financial governance structure’, which our theoretical framework all attributed to other stakeholders. In addition, the mayor and the regent pay attention to ‘outcomes of services’, which were not included in our framework. In accordance with our expectations, the mayor and the regent also play a mediating role towards the other stakeholders. Table 2 summarises the actual performance interests of the powerful stakeholders.

Table 2. Powerful stakeholders and their actual performance interests

Funding bodies (here, central government and local council)	Statutory boards (here, local council)	Purchasers (here, mayor/regent)
* Not a specific type of performance indicators. † Only for the District Bureau of Public Works in Denpasar. ‡ Only for the District Bureaus of Health.
Expected	Expected	Expected
• Disciplined use of financial resources (inputs)	• Political governance structure*	• Quantity and quality of services (outputs)
• Financial governance structure*		• Costs of services (inputs versus outputs)
Additional	Additional	Additional
• Outcomes of services (outcomes)‡	• Quantity and quality of services (outputs)	• Disciplined use of financial resources (inputs)
• Value for money (outcomes versus inputs)		• Outcomes of services (outcomes)‡
		• Financial governance structure*, †
		• Political governance structure*, ‡

We also found that central government's interest in ‘outcomes of services’ as well as the mayor's and regent's interest in ‘outcomes of services’ and the ‘political governance structure’ was restricted to the District Bureaus of Health. This can be explained by differences in the district bureaus' activities. Whereas the District Bureaus of Public Works mostly have activities with tangible outputs (e.g. road construction projects), the District Bureaus of Health mostly provide services with less tangible outputs (e.g. drinking water quality assessments). Given its more tangible outputs, it is relatively easy to assess the performance of the District Bureaus of Public Works. Furthermore, in the District Bureaus of Public Works, both the mayor and the regent pay more attention to the financial performance dimensions because in both local governments, these bureaus receive a significantly larger amount of funding than the District Bureaus of Health.

The broader spectrum of performance interests of all three powerful stakeholders can be expected to imply that the performance measurement systems of the four district bureaus include a more diverse set of performance indicators, including indicators on ‘outcomes of services’ and ‘value for money’. It can also be expected to have implications for the use of performance information by the district bureaus' managers. As the overlap in performance interests between the different stakeholders is larger than expected, the potential for conflicts of interest is smaller, which could stimulate a functional use of performance information.

The design of the performance measurement system

As the mayor/regent is regarded as the dominant stakeholder, we expect that his performance interests are reflected in the performance measurement systems of the district bureaus. This would imply that these systems include performance indicators on inputs, outputs, outcomes and links between inputs and outputs. To examine whether this is the case, we will now describe and analyse the main contents of the various types of performance reports that are available in each of the four district bureaus. The contents of these reports are largely determined by the Indonesian central government, which — since the decentralisation — has issued extensive regulation that prescribes which reports local governments should produce and what the contents of these reports should be.

The Indonesian central government requires local governments to develop and prepare four types of accountability reports. The most important, in the sense of comprehensive, report is the so-called LAKIP (Laporan Akuntanbilitas Kinerja Instansi Pemerintah) report (Indonesian Government, 2003). The managers of the district bureaus have to submit an annual LAKIP report to their mayor/regent. Each local government combines the LAKIP reports from its district bureaus into one report, which is submitted to the central government (i.e. the State Minister for Administrative Reforms). In 1999, the central government issued guidelines for the LAKIP report, which require that this report provides information about the local government's strategic planning, including its vision and mission, its goals and objectives, and targets for and achievements on performance indicators.

In addition to the four accountability reports, the managers of the four district bureaus use particular managerial reports for planning and control purposes. These reports differ in various respects from the accountability reports.

• First, the managerial reports are prepared on a monthly or quarterly basis, whereas the accountability reports have an annual basis.
• Second, to some extent, the managerial reports use different types of performance indicators than the accountability reports. Both types of reports have in common that they include performance indicators on inputs, but where the accountability reports include also indicators on outputs and outcomes, the managerial reports contain indicators on throughputs. This indicates that in terms of the types of performance indicators included, there is a substantial extent of decoupling between both types of reports.3 Table 3 gives a few examples of performance indicators used. It also illustrates that there can be links between the contents of — on the one hand — throughput indicators and — on the other hand — output and outcome indicators.
• Third, the managerial reports often contain more specific performance information than the accountability reports. For instance, the information in the quarterly financial report and the quarterly report on the use of specific grants from central government is much more detailed than the related information in the LAKIP report, which contains only information about the total amount of spending. In the case of the District Bureaus of Public Works, the managerial reports also include two weekly and quarterly reports on the ‘progress of physical completion and finance’ (Laporan Kemajuan Fisik Dan Keuangan), as required by the Minister of Public Works. The reports mention in detail the funding sources for individual programmes and projects, the amount of funding that has been used by these programmes and projects and the percentage of their completion. Some managerial reports include rather detailed nonfinancial indicators. For example, in the District Bureau of Health in Denpasar, they include indicators on training staff members to enable them to detect problems related to the growth of children.

Table 3. Examples of performance indicators in the accountability and managerial reports of the District Bureaus of Health and Public Works

	Accountability report		Managerial report
	Health	Public Works	Health	Public Works
Input	Total funding of drinking water quality assessment	Total funding of new irrigation systems	Funding of drinking water quality assessment divided into general purpose, specific and own-source funding	Funding of new irrigation systems divided into general purpose, specific and own-source funding
Throughput	No indicator	No indicator	Number of training sessions for staff on drinking water quality assessment	Percentage of completion of newly planned irrigation systems
Output	Number of inspections of drinking water facilities	Number of new irrigation systems	No indicator	No indicator
Outcome	Increased knowledge of staff of drinking water quality	Extent of stability of new irrigation systems	No indicator	No indicator

The aforementioned description shows that the design of the performance measurement systems of the four district bureaus is quite sophisticated, in the sense that — at least formally — it starts with the district bureau's strategy, translates this into different types of performance indicators (ranging from inputs to outcomes) and formulates targets for these indicators. Moreover, the systems produce various accountability and managerial reports, which can be used to check whether the performance targets have been achieved. Together, the performance reports provide the information on inputs, outputs and outcomes that is required by the dominant stakeholder, that is, the mayor/regent. However, the more sophisticated information he or she requires about the links between inputs and outputs is missing.

Our investigations also revealed that the district bureaus find it difficult to produce adequate performance information. Their difficulties can be explained by external and internal factors. The external factors include deadlines for submitting reports that are close to one another, as well as frequent changes in the reporting requirements issued by the central government. All four district bureaus are faced with this problem. Important internal factors are technical limitations (e.g. limited facilities, insufficient data) and a small number of staff members who possess the proper skills (e.g. to process numerical data and to create graphics using Microsoft Office applications). The latter limitations vary in some respects between the four district bureaus; that is, the district bureaus in Denpasar (urban area, with a relatively complex service delivery) are confronted with an insufficient availability of data, whereas those in Jembrana (rural area, with a relatively simple service delivery) have to cope with facilities that are far too limited.

Because of the difficulties in producing performance reports, all four district bureaus focus more on fulfilling the formal requirements regarding the format of the reports and on submitting these reports in time than on their contents. This has led to a low quality of the contents of the reports, which particularly applies to the output and outcome indicators in the accountability reports.

The use of the performance measurement system

The four district bureau managers claim that when using performance information, they mostly follow the interests of the dominant stakeholder, that is, the mayor/regent. In addition, they explain that they use the annual performance reports (e.g. the LAKIP report) only for accountability purposes, whereas they use the short-term performance reports (e.g. monthly and quarterly reports) for managerial purposes. This implies that information about outcome and output indicators, which is only included in the annual reports, is used symbolically rather than functionally. Therefore, the actual use of information by the district bureau managers is only partly consistent with our initial expectation that the presence of a dominant stakeholder leads to a functional use of the performance information that is related his or her performance interests. The dominant stakeholder of all four district bureaus has an interest in performance in terms of inputs (‘disciplined use of financial resources’), outputs (‘quantity and quality of services’), links between inputs and outputs (‘costs of services’) and outcomes (‘outcomes of services’). However, only the performance information that is related to his interest in inputs is integrated throughout the organisation and is used functionally. Apparently, the other performance information is only used to legitimatise the district bureau. Nevertheless, this is not to say that the managers focus solely on inputs. Instead, they also pay attention to throughput indicators, which — as was illustrated in Table 3 — can be loosely coupled with output and outcome indicators.

Although the district bureau managers mostly accommodate the performance interests of the dominant stakeholder, they also pay attention to some of the interests addressed by the other two powerful stakeholders, that is, the local council and the central government. This especially concerns the ‘financial governance structure’. Owing to tight supervisions from the local council and audits conducted by central government, the district bureau managers feel a strong pressure to spend resources according to the associated guidelines. The other performance dimension that is largely specific for the other stakeholders (i.e. ‘value for money’) comes back in neither the accountability reports nor the managerial reports.

The empirical results show that whenever gaps between performance targets and actual performance arise, the district bureau managers try to provide explanations for why they did not fully accomplish the performance targets. The mayor/regent will be inclined to accept such gaps if he or she feels that the factors that explain them are beyond the district bureau manager's control. Whenever conflicts between stakeholders emerge, the mayor/regent and/or the manager try to come to political agreements with, for instance, the local council. This could be labelled as ‘juggling behaviour’. It implies that conflicts between stakeholders are not primarily solved by decoupling operations and performance information but by finding compromises between these stakeholders' interests to serve the diverging interests in a balanced way. The regent of Jembrana, for example, allocated the funding for overlaying roads equally to the five subdistricts, although on the basis of technical priorities, the funding should have been allocated to only three subdistricts. His decision was meant to accommodate the interests of different members of the local council, who act primary as representatives of their own subdistrict. Moreover, in response to a significant decrease in the specific-purpose funds for public works by the central government, the regent provided additional funding to the district bureau of Public Works to achieve the targeted length of overlaying of roads. Because of this additional funding, he or she quite easily received support from the members of the local council for the local budget revision.

In a more general sense, our research findings resonate with more recent insights into the use of neo-institutional theory for explaining performance measurement practices. As Modell (2009, p. 283) indicates in his review of the literature about this theme, loose coupling is only one of the wider range of organisational responses to institutional pressures. Another response is finding compromises to accommodate for partly conflicting interests, which is similar to the concept of juggling behaviour in our study. This also distances our findings from a purely rational view on performance measurement because relative power positions of stakeholders are conductive to the way these systems are embedded in organisational practices.

CONCLUSIONS AND REFLECTIONS

Our research, which is based on four case studies of Indonesian local government agencies (i.e. district bureaus), gives rise to the following observations and explanations concerning the design of performance measurement systems and their use by the agencies' managers. In the agencies we investigated, the public sector reforms have increased the power positions of stakeholders, but so far, one stakeholder is still dominant. The lack of institutional capacity as part of the LDC-specific context turned out to be an obstacle for the design of fully elaborated high-quality performance measurement systems. With respect to the performance reports that were prepared for accountability purposes, the managers of the agencies focused primarily on fulfilling the formal requirements regarding the format of the reports and their timely submission and largely ignored the quality of the reports' contents. These reports included performance indicators on inputs, outputs and outcomes but not on throughputs. The performance reports that were developed for managerial purposes contained performance indicators that the managers considered relevant. The reports comprised indicators on inputs and throughputs but did not encompass performance indicators on outputs and outcomes. This implies that output and outcome indicators did not cascade down from the top to lower layers of the organisation. Therefore, we can conclude that the accountability reports, except for the parts that deal with performance indicators on inputs, were used mainly symbolically. This was further affirmed by the low quality of these reports, as perceived by the district bureau managers themselves. Figure 2 summarises these findings.

Figure 2. Empirical results on the use of performance information in the context of one dominant stakeholder.

In using the managerial reports, the agencies' managers primarily accommodated the performance interests of the dominant stakeholder, that is, the mayor/regent. However, this does not mean that the performance interests of the other powerful stakeholders (i.e. the local council and central government) were ignored. Our investigations led to three observations in this respect. First, the mayor/regent served broader interests than was expected on the basis of his purchaser role. The resulting overlap in performance interests implied that the interests of the local council and the central government were served to a relatively great extent, despite the managers' focus on the interests of the mayor/regent. Second, the district bureau managers also served some interests that were related to their own responsibilities and were not emphasised by the mayor/regent. This especially concerned the ‘financial governance structure’, which is subject to tight supervision by the local council and audits conducted by central government. Finally, in case of conflicts between performance interests, the mayor/regent and the district bureau managers were inclined to search for compromises and thus partially served these diverging interests. We have labelled this search for compromises as ‘juggling behaviour’.

Our theoretical framework has been helpful in explaining the design and use of performance measurement systems in local government agencies in LDCs. Consistent with our framework, we observed hindrances in designing fully elaborated high-quality performance measurements systems owing to the low institutional capacity that characterises the public sector in LDCs. We also saw that in situations with one dominant stakeholder, performance measurement systems can be used in a functional manner (i.e. to support decision making). However, in the district bureaus we investigated, this only concerned the parts that deal with inputs; all parts that deal with outputs and outcomes were used symbolically. In terms of our framework, the latter implies that the impact of the LDC-specific context on the use of performance information on outputs and outcomes outweighs the impact of the existence of a dominant stakeholder.

Our study shows a high degree of similarity among the four local government district bureaus in Indonesia we studied. This can be explained by the regulations issued by the Indonesian central government, which have to be followed by the local governments and which — in terms of New Institutional Sociology — lead to a substantial extent of coercive isomorphism (DiMaggio and Powell, 1983). Despite this similarity among the district bureaus, we also observed slight context-specific deviations, which regarded, for instance, differences between the District Bureaus of Health and the District Bureaus of Public Works (see also Modell, 2009, p. 289). Moreover, finding context-specific compromises to accommodate for partly conflicting interests has been found as one of the typical responses to pressures exerted by different stakeholders (see also Modell, 2009, p. 283). The high degree of similarity among the four district bureaus also had as a consequence that we only came across situations of one dominant stakeholder, although with less distinctive performance interests than our framework assumed. This hints at two venues for future research. One venue needs to assess whether our expectations regarding situations where two or more stakeholders are powerful are valid and whether such situations give rise to a mainly symbolic use of performance information. The other needs to investigate situations of a dominant stakeholder with distinctive performance interests in comparison with other stakeholders, which is expected to gives rise to a predominantly functional use of performance information.

On a more general level, our paper raises the question of which lessons can be learnt with respect to the design and use of public sector performance measurement systems in LDCs. The Indonesian central government has developed guidelines that require local governments to design and use rather sophisticated performance measurement systems. However, our case studies showed that in practice, organisations are not always able to produce (sophisticated) information of sufficient quality, whereas managers do not seem to need this information for planning and control purposes. Moreover, our interviews revealed that also the mayor and the regent do not comprehensively take notice of the various performance reports. Instead, they focus on the ‘sensitive’ parts of these reports, that is, the parts that point to situations in which targets are not reached or that are related to recent debates in the council. For other issues, the mayor and the regent rely on the district bureau managers, who are responsible for the full content of these performance reports.

Future research has to support public sector organisations in LDCs in developing performance measurement systems that are useful for managerial and accountability purposes and that fit with the organisations' low institutional capacity. Additionally, these systems have to enable organisations to reduce problems associated with a high level of corruption. The performance interests of powerful stakeholders could be the starting point of this research. However, it should be recognised that the types of stakeholders can differ between local government agencies and, for example, state-owned enterprises. In addition, it should be acknowledged that the context in which the performance measurement systems will function may vary, as LDCs are not a homogeneous group, but may differ in terms of their degree and distribution of poverty, their culture, their (colonial) history, and their political and economic systems (cf. Hopper et al., 2009, pp. 476–478). Although much work has to be carried out in the largely unexplored field of public sector performance measurement in LDCs, our study can be regarded as starting point because it provides both theoretical clues for future studies and preliminary findings for the specific context of Indonesian local government agencies.

• 1

  In Jembrana, the full name of this district bureau is the District Bureau of Health and Social Welfare.

• 2

  The highest official in Denpasar (urban area) is called the mayor, but the person who holds a similar position in Jembrana (rural area) is called the regent.

• 3

  In the District Bureau of Public Works in Jembrana, the accountability reports also include performance indicators on benefits and impacts, in addition to those on inputs, outputs and outcomes.