Between Revenues and Public Service Delivery SOEs and PSAs in Indonesia

This article examines the development of policies regarding the state-owned enterprises (SOEs) and public service agencies (PSAs) in Indonesia. In 2004, the government of Indonesia introduced PSAs—government agencies that were given large autonomy to manage their financial affairs.The rationale behind this autonomy is consistent with the New Public Management ideal: the creation of more market-oriented government institutions with the objective of increasing the efficiency and effectiveness of public service delivery. The PSA policy has increased state revenues significantly, yet the quality of services and accountability has not improved accordingly. A comparison with SOEs reveals that the restructuring of government agencies and SOEs took place before asupportiveframeworkwassetinplace.Wearguethattotackleinformalityandtosafe-guard the social functions of public services, the spearheads of efficiency and revenues in Indonesian bureaucratic reform policies require a strong foundation, consisting of regulatory and ideological components.


Introduction
In his first speech after being re-elected, Indonesian president Joko Widodo, popularly known as Jokowi, stressed the importance of having a mindset change within the bureaucracy. He was quoted as saying: Bapak, Ibu, brother and sister compatriots. All the people of Indonesia whom I love. Ladies and gentlemen, we all have to be aware that now we live in a very dynamic global environment. We know the characteristics of this global phenomenon: full of change, full of velocity, full of risk, full of complexity and full of surprises that are far from what we expected. Therefore, we have to find a new model, new method and new values when looking for solutions to every problem with innovation and we all have to be willing [to change] and we will force [people] to be willing [to change]. We have to abandon old methods, old patterns, both in managing institutions and in managing government. What is no longer effective, we will make effective. What is no longer efficient, we will make efficient!1 Efficiency, bureaucratic reform, and a change of mindset are essential ingredients in Jokowi's administration's recipe for economic success. New Public Management ideals are seen as the key to Indonesia's development,2 as government agencies must become business-and investor-friendly. The Grand Design Reformasi Birokrasi 2010-2015 (the Grand Design for Bureaucracy Reform 2010-2025), a programme launched by President Susilo Bambang Yudhoyono in 2010,3 has been taken up with more vigour under President Joko Widodo's administration (Jokowi, 2015-present). Badan Layanan Umum (BLU, Public Service Agencies; hereafter PSA s), government agencies that possess considerable freedom to manage their own economic affairs, fit well in this New Public Management ideology. The first 13 national PSA s were established in 2005 under the President Megawati's administration. This number has grown to 222 PSA s today (2019).
This article examines the development of these PSA s in Indonesia, in particular how it relates to restructuring policies pertaining to Badan Usaha Milik Negara (BUMN, state-owned enterprises; hereafter SOE s). Restructuring of SOE s has led to different levels of hybridization, which denotes a process in which SOE s' management adopts private sector management styles and blends public sector and private sector characteristics (Kikeri and Nellis 2002). The world-wide adoption of private sector management styles by SOE s in the 1980s was part of the same New Public Management ideology that inspires Indonesian bureaucratic reforms today.
The main argument against state control of SOE s, apart from the libertarian ideological point that state interference in the economy should be kept as low as possible (Potrafke 2010), is their alleged inefficiency (Musacchio and Flores-Macias 2009). This persistent idea of inefficiency of SOE s, however, is increasingly being contested by research that moves away from the state-owned and privately owned dichotomy and takes into account the many different types of hybrid SOE s. SOE s differ in the proportions of SOE s' private ownership, public service function, the ascribed strategic function, and level of competition with domestic and global players (Bruton et al. 2015). Many hybrid SOE s perform economically very well-sometimes better than privately owned companies (Inoue, Lazzarini, and Musacchio 2013;Goldeng, Grunfeld, and Benito 2008).4 Under the Jokowi presidency, this positive assessment of SOE s facilitated a new developmentalism in which SOE s play an important role (Warburton 2016).
Although SOE s do not dominate Indonesia's economy, SOE s are among the largest Indonesian companies. To illustrate this: in 2013, 69 % of the volume of the Indonesian companies in the Forbes Global 2000-the world's 2,000 largest companies according to Forbes-were SOE s (Kowalski et al. 2013). In recent Indonesian history, and especially during times of economic crises, the government of Indonesia introduced models for reforms aimed at tackling the informality, nepotism, and corruption that mar SOE s (Ilmar 2012;Moeljono 2004;Diah 2003).
The establishment of PSA s in Indonesia was inspired by the ideas of New Public Management, an ideology that sees the private sector as model for public sector management. New Public Management, modelled to 1980s' public sector reforms in the Organisation for Economic Cooperation and Development (OECD) countries, is characterized by several shifts in management policies ranging 'from policy making to management skills, from a stress on process to a stress on output, from orderly hierarchies to an intendedly more competitive basis for providing public services, from fared to variable pay, and from a uniform and inclusive public service to a variant structure with more emphasis on contract provision' (Hood 1995:95). New Public Management was also at the heart of the worldwide managerial reforms of SOE s (Larbi 1999), including in Indonesia (Moeljono 2004;Diah 2003), which means that explorative parallels between PSA s and SOE s can be drawn.
Despite the global appeal of New Public Management reforms, there is no conclusive evidence that application of a more corporate style of public management will lead to a better performance of public services in the long run. International reviews of literature on privatization of government services show that results in terms of cost-saving are mixed (Bel and Warner 2008;Boyne 1998) and consequences for consumers in terms of prices and efficiency and quality of service delivery are inconclusive and most often not included in assessments at all (Petersen, Helby, and Vrangbaek 2018). For instance, in the United States and Europe, partial contracting out to private-health-service delivery firms did not increase cost-effectiveness and efficiency of the health services delivered (Bel and Esteve 2019).
PSA s are established to create a more private-sector-like, corporate governance within government agencies. In developing countries, one of the main obstacles to reach this management type is the informality of the bureaucratic culture (Schick 1998). This article demonstrates how in Indonesia structural changes pertaining to PSA s and SOE s have been implemented without creating sufficient checks and controls to tackle informality of the bureaucratic culture. The case of SOE s shows how, following restructuring and corporatization policies in the 1980s, high bureaucrats in sectoral ministries and SOE management continued to use SOE s as a 'cash-cow' (Moeljono 2004:8-9). The SOE case shows that without changing informality within SOE management, restructuring will not reduce the risks of collusion and corruption.
This article argues that the history of SOE reforms demonstrates that policies regarding PSA s should not be measured only through cost-efficiency and state revenues. Instead, accountability must be prioritized and include both financial reporting and reporting of the quality and accessibility of social and public service delivery. Such accountability must have an ideological foundation that underlines social functions of SOE s and counters informality.

2
Indonesian SOE s and Their History of Restructuring SOE s are created to run economic sectors on behalf of the state. Arguments for creating SOE s are concerns about national security, insufficient investment from the private sector in certain business areas, as well as redistribution of certain services so that they are accessible for citizens all over the country. In many cases, SOE s with a public utility function (railways, postal services, telecommunication, electricity) have (or had in the past) a natural monopoly, which denotes a situation in which the government develops a monopoly for SOE s in market segments where conditions are not profitable for private competitors to invest (Heath and Norman 2004;Lewin 1982). The Indonesian 1945 Constitution heavily mandates the state to steer the country's economic development. The main architect of the economic provisions in Indonesia's 1945 Constitution was Mohammad Hatta, the first vicepresident of the Republic of Indonesia (Kusuma 2009). Article 33 of the 1945 Constitution delineates the role of the Indonesian government in economic affairs. The original Article 33 consists of only three paragraphs and reads as follows: 1. The economy is to be organized as a common effort based on the family principle; 2. Those sectors of production that are strategic to the country and those that affect public necessities are under the powers of the State; 3. The land, the waters and the natural resources contained in it are under the powers of the State and shall be used to the greatest welfare of the people. The 1945 Constitution does not stipulate full state control over its economyit is not a socialist model. However, it does require state control over natural resources and over economic areas that are deemed strategic or affect the ability of people to meet their basic needs. The state's management of essential economic sectors must offer the greatest benefit for the people, a provision that resembles European social-democratic ideologies. In the words of Muhammad Hatta: Whereas the free market [ideology] does not want governments to interfere in the economy through regulations, our idea of a government-led economy (ekonomi terpimpin) has the opposite objective: the government must actively regulate the economy so that social justice can be achieved. If we let the economy be run by free-market forces, this means that we allow the strong to prey on the weak.
Hatta's supportive words for state regulation and 'welfare economics' (Hatta 1979: 29), do not clarify what exactly economic control by the state entails. It is clear, however, that Hatta had no preference for an SOE s-dominated economy, but for an economy that is run by private enterprises 'run in the family spirit' , such as cooperatives (Hatta 1979:35, 36). Hatta's ideas of a nationalist economy still have a strong following today.

SOE s and Their Economic Rationale under Sukarno
The more than two decades of Sukarno's so-called Orde Lama (Old Order) regime  can be divided into two time periods. In the first period of pragmatism (1945)(1946)(1947)(1948)(1949)(1950)(1951)(1952)(1953)(1954)(1955)(1956)(1957), Indonesia took a pragmatic approach towards foreign investors, a group that was dominated by the Dutch. As the former colonizers, the Dutch still held a large stake in the Indonesian economy. The second period (1958)(1959)(1960)(1961)(1962)(1963)(1964)(1965) is characterized by nationalization of Dutch companies and economic nationalism (Lindblad 2011). Upon the transfer of sovereignty by the Dutch in 1949, which followed the Declaration of Independence in 1945 and four years of independence struggle, Indonesia took over several SOE s established by the colonial administration of the Netherlands Indies, including the post office, the state railways, and various mining and plantation interests. Other enterprises were set up by the government, including Bank Negara Indonesia (BNI, State Bank of Indonesia), the Central Trading Company, and the Usaha Industri Indonesia (Usindo, Indonesian Industries), in order to compete with Dutch banks and trading companies. Private banks, railways, utilities, and estates were nationalized, based on mutual agreement; compensation was paid based on actual worth.
Indonesia did not only indebt itself through nationalizations; it also took on over $1.1 billion of public debt, more than any ex-colony in Asia (Dick et al. 2002:71). This meant that the newly founded republic started its independence with a large burden of repayments to the Netherlands and little room for investments in its own economy.
Nonetheless, Indonesia achieved a moderate positive economic growth in the 1950s, as the country was able to take advantage of a boom in the world economy. Foreign enterprises, but also Indonesian smallholders, played an important role in the economic recovery after World War II (Lindblad 2010).
Relations with the Netherlands, as the former colonizer, remained strained for the following reasons. First, in 1951 Indonesia changed its federal state model-preferred by the Netherlands-into a unitary state model. Second, Dutch companies were slow in implementing the promised Indonesianisasi (Indonesianization) of personnel and management (Lindblad 2011). Third, Indonesia and the Netherlands were in conflict over the fate of West Papua. West Papua was part of the Netherlands Indies but remained in Dutch hands after the Dutch sovereignty transfer in 1949.5 In 1958 Indonesia terminated diplomatic relations with the Netherlands and pursued a policy of confrontation towards its ex-colonial power. The situation ignited nationalistic and anti-Dutch sentiments and resulted in the nationalization of hundreds of Dutch private companies that operated in a variety of business areas, such as mining, plantations, transportation, postal services, and pawnshops. Management of these SOE s was often handed over to the military (Dick et al. 2002:187). In 1959 Sukarno's administration abolished parliament and concentrated the powers in the hands of the president, which marked the so-called Demokrasi Terpimpin (Guided Democracy). The Guided Democracy had an economic component, the 'Guided Economy' . In the eyes of Sukarno, Indonesia's economy should be subject to strict political control.
In 1960 Sukarno issued a government regulation in lieu of law pertaining to SOE s, with reference to Article 33 of the Constitution and as the 'structure for a guided economy towards the implementation of Indonesian socialism' .6 In this regulation, an SOE (state-owned enterprise or perusahaan negara) is defined as a production unit that has three functions: to provide public services, to generate public revenues, and to generate income. SOE s should pursue these three goals at once. The Penjelasan Umum (Explanatory Memorandum) to the regulation mentions that SOE s have to be self-sustainable and are not allowed to become a burden to the state budget. SOE s were seen as an instrument for social and economic development, as set forward in state economic-development plans. These plans were financed through state banks.7 Nine state-owned trading houses were made responsible for import of essential goods (about 70% of the total trade) and the bulk of 5 In the end, a referendum was organized in which West Papua voted to become part of the Republic of Indonesia. While the way the referendum was organized is not without controversy, the outcome was internationally accepted. exports. The envisioned role of SOE s in developing the economy did not mean that the entire economy should be run by the state. While international trade was nationalized, domestic commerce remained largely in private hands (Lindblad 2006;Dick et al. 2002:187). At this stage, the envisioned state-led development under the guidance of SOE s was not realized. In the early 1960s, the performance of the economy rapidly worsened. The problems experienced were lack of credit, lack of foreign investment, and bad management. The Indonesian government lacked resources and tax income, while the nationalization campaign, aggressive foreign policy towards Malaysia (1963Malaysia ( -1966, withdrawal from the United Nations (1965), and general prevalence of socialist and nationalist political discourses against 'imperialists' forces made Western foreign companies hesitant to invest in the economy. Management of SOE s, which for a large part had been handed over to the Indonesian military, ended in failure due to a lack of managing skills and experience (Sungkar 2008:98). As a result of national and international economic developments, the republic entered its twenty-yearold anniversary with a hyperinflation rate of 175 % and a crumbling economy (Hossain 2005:50).
In 1965, in a complex political and economic context, a coup attempted by communist elements of the military took place. The Indonesian military, under the supervision of General Soeharto, were quick to act and headed a crackdown of the Partai Komunis Indonesia (Indonesian Communist Party), which led to many casualties and the incarceration of many of its members and sympathizers. In 1966 Sukarno was pressured to sign the Surat Perintah 11 Maret (Supersemar, 11 March Executive Order), which transferred many of his presidential powers to General Soeharto. On 27 March 1968, General Soeharto ascended to the presidency.
In the first phase of the New Order (1968)(1969)(1970)(1971)(1972)(1973), the goal of economic stabilization was primary as Indonesia sought to attract domestic and foreign investment. A switch in Indonesia's economic and political orientation in fact had already taken place in 1967, the last year of Sukarno's presidency, when General Soeharto was effectively in control. Undang-Undang No. 1 Tahun 1967 tentang Penanaman Modal Asing (Law No. 1 of 1967 on Foreign Investment; hereafter the 1967 Foreign Investment Law) opened up the economy to foreign investment by offering tax benefits and free transfers of profits to those companies willing to invest in Indonesia's economy. The explanatory memorandum to the law describes the reliance of the economy on foreign capital as a temporary necessity and a prerequisite to develop the country's national economy, human resources, and know-how. In 1968, Undang-Undang Penanaman Modal Dalam Negeri (the 1968 Domestic Investment Law) offered similar incentives and guarantees to Indonesian entrepreneurs.
The 1967 Foreign Investment Law did not open up the entire economy as it restricted the access for foreign companies in a number of sectors that were deemed strategic by the New Order government. Areas that remained fully the domain of SOE s were harbours, electricity supplies, ferries, domestic flights, drinking water supply, public train services, atomic energy development, and mass media. In all these areas, with the exception of the mass media, only SOE s were allowed to operate (Robison 1988 Table 1): the perusahaan jawatan (public service company), the perusahaan umum (public utility company), and the perusahaan persero (limited liability company).
The three types of SOE s introduced during the New Order are still based on the three functions of the perusahaan negara in the old 1960 SOE s Law. However, SOE s do not have to fulfil the three functions at once. The 1969 SOE s Law defined SOE s as state enterprises with the function of providing public services, creating public benefit, or generating income. Limited liability companies were allowed to issue shares on the stock market. What all SOE s had in Public legal entity Public legal entity Private legal entity common during this time was that their management fell under the responsibility of the relevant sectoral ministry. Thus, despite the financial autonomy they had on paper, their daily management, including financial matters, continued to be subject to strong government control. During this era, SOE s continued to dominate the 'strategic' segments of the economy (Booth and McCawley 1981:64). The continuation of nationalist economic policies became more visible during the second phase of the New Order. The boom in oil prices in the 1970s increased state revenues and made the New Order government less dependent on private investments. This led to increased state intervention in the economy. Tax revenues from foreign oil companies operating in Indonesia rose from about one third of total government revenues before 1974 to more than two thirds by 1979. Government's expenditure rose from 9.3 % to 22.1% respectively (Dick et al. 2002:208). Higher state revenues combined with nationalist sentiments in society, such as those expressed during the 1974 Malari riots against the growing Japanese business influence, led to the introduction of restrictive measures on foreign investment and preferential treatment of indigenous firms and businessmen. From 1974 new foreign direct investment could enter Indonesia only in the form of joint ventures with Indonesian companies (Robison 1988).
The third phase of New Order economic policies started with an economic crisis. In the early 1980s, the price of oil fell dramatically. A loss of 14 % of the government's annual budget meant that cuts in the state budget were necessary, including cuts pertaining to SOE s (Diah 2003:188). Foreign invest-ment was partly deregulated again and a number of restrictions in the exportoriented segments of the market were removed. This resulted in a large disparity between economic policies on micro-and macroeconomic levels, as the domestic market remained highly regulated, restricted, and characterized by subsidies and preferential treatment for Indonesian companies (Dick et al. 2002:214; see also Rock 2003).
Less government funding revealed that the performance of many SOE s was weak. The government attempted to increase the financial accountability of SOE s by stipulating dual supervision by the Kementerian Keuangan (Menkeu, Ministry of Finance; hereafter MoF) and the relevant sectoral ministry.9 SOE s were also made subject to a rightsizing policy. Rightsizing is defined by the BusinessDictionary as '[t]he process of a corporation reorganizing or restructuring their business by cost-cutting, reduction of workforce, or reorganizing upper-level management' .10 The dual supervision and rightsizing policies, however, did not lead to professionalism and accountability within SOE s (Diah 2003:187). On the contrary, SOE s continued to be a source of collusion between the central government, the military, and local bureaucracies (Rock 2003;Robison 1988;Liddle 1985).
The fact that the New Order reduced the number of SOE s from 822 in 1967 to 200 in 1989 (Diah 2003:187) is not indicative of less support for SOE s under the New Order. While the role of SOE s in the New Order's economy was limited-SOE s produced approximately 15% of GDP in 1991-SOE s' symbolism as carriers of a nationalist economy remained strong (Sungkar 2008:98). Even during the economic crisis of 1997 President Soeharto remained strongly opposed to privatization of SOE s (Moeljono 2004:14). Several large SOE s were allowed to dominate strategic sectors of the economy. These included Pertamina (an oil company), Garuda (the state airline), state commercial banks, and telecommunications companies.
Meanwhile, Indonesia's measures to tackle the economic crisis appeared successful. As an exporter of labour-intensive manufactured products, Indonesia was no longer dependent on oil prices. Oil palm also became an important export product to Indonesia. From 1988 to 1996, economic growth in Indonesia was high, between 7.3% and 9% annually. By the 1990s, Indonesia was coined hereafter Government Regulation 55/1990) with the stated purpose of increasing the autonomy of SOE s to manage their own financial affairs. Under the 'go public' policy, profitable SOE s were encouraged to look for private investors and to become hybrid SOE s: limited liability companies that operate in a competitive market and are partly in private hands. The government developed a strong preference for the limited liability company, a hybrid SOE with a profit maximalization rationale. As we will see, after President Soeharto stepped down, this preference for limited liability companies would be enshrined in law.

Further Liberalization during the Reformasi (1998-2004)
The fall of Soeharto's New Order regime in 1998 was triggered by the 1997 Asian crisis. The way that the Asian crisis hit Indonesia took most economists by surprise, as in 1996 Indonesia still recorded an economic growth of 7.8 %. Within two years' time, Indonesia's GDP decreased with more than 15 %. The fall of the rupiah, from IDR 2,400 per US dollar to IDR 17,000, tripled Indonesia's foreign debt (Sjöholm 1999). In 1997 President Soeharto reluctantly signed a US$ 43 billion loan agreement with the IMF. The crisis was detrimental for Soeharto's reputation as the Bapak Pembangunan (Father of Development), the architect of Indonesia's economic success. The economic crisis and the social and political unrest made Soeharto's position untenable. On 21 May 1998, President Soeharto resigned. The first years of the Reformasi, the time period of political and economic reforms following Soeharto's resignation (1998)(1999)(2000)(2001)(2002)(2003)(2004), saw economic reforms and decentralization of the government. On 19 October 1998, the Government of Indonesia's signed a Letter of Intent for economic reforms, as part of its loan agreement with the IMF. Just as in the early 1980s, a dramatic decrease of the state budget forced the government to look into the role of SOE s in Indonesia's economy. In 1998, with the aim to increase its grip on SOE s, the government decided to transfer the ineffective dual supervision of SOE s to a new Kementerian Badan Usaha Milik Negara (Ministry of State-Owned Entreprises; hereafter Ministry of SOE s).11 Within 18 months, the ministry restructured state banks and closed 16 of them, liberalized the agricultural sector, and restructured the state airline and the state electricity company. The Ministry of SOE s sold shares in five out of 12 companies intended to be privatized (Sungkar 2008:101).
In 2002, Article 33 of the Constitution was amended. The amended Article 33 still stipulates a state-controlled national economy for the welfare of the people, yet with the addition that the state is expected to balance several economic and social development goals. Hence Article 33 kept providing supporters of nationalist economic policies with constitutional legitimacy. For example, the prominent nationalist economist Sri Edhi Swasono argued that Indonesia's economic reforms during the Reformasi were contrary to Article 33 of the Constitution.12 During the Reformasi the Indonesian government was under pressure to reform its domestic economy; hence, restructured SOE s. Restructuring of SOE s sometimes had adverse social consequences. For example, in 2003 the World Bank warned that generation shortfalls were '[one of the] most critical issues facing [the country]' . Indonesia had requested funding to improve its electric generation branch. The World Bank insisted that Indonesia had to restructure the Indonesian electricity company's urban operations and privatize the generation companies first and provided funding for such restructuring. A long period of electricity outages in Indonesia's cities was taken for granted (Engel 2006 14 Fear for privatization of basic public utilities such as electricity and water facilitated nationalist resistance against the privatization of all types of SOE s (Nugraha 2004). Indeed, the political situation in Indonesia, where the military and politicians have a large stake in SOE s, makes privatization of SOE s a very delicate political matter. Managerial reforms of SOE s are difficult to achieve, let alone privatizations.

6
Restructuring of SOE s after 1998 As said, in 1998, under the transitional government of President Habibie (1998)(1999), the Ministry of SOE s was created. It took over the supervision of SOE s from 17 sectoral ministries. Tanri Abeng, who had been the main advisor of the former president Soeharto concerning SOE s, became the first minister of SOE s. A blueprint for SOE reform was developed, with the ultimate objective of transforming SOE s into professional super holdings operating in ten sectors: telecommunications and media; construction; energy; tourism; logistics; strategic industries; mining; forestry, paper, and publishing; the agro-industry and agro-processing sector; and financial services (Moeljono 2004:16). A restructuring scheme was developed, which would become the model of future SOErestructuring policies (see Figure 1).
The government under President Abdurrahman Wahid (1999)(2000)(2001) formulated its own SOE-restructuring programme, which for a large part was based on the blueprint. This programme emphasized the public service and social functions of SOE s by setting three conditions: that the level of service is maintained; that the accessibility of services is maintained; and that the restructuring does not increase the economic burden on the state. The large coalition on which   (2002)(2003)(2004)(2005)(2006), aimed at partial (not full) privatization of SOE s. The stated objective was to create competitive and hybrid SOE s, but the plan was not as clear-cut and comprehensive as the 1998 blueprint had been. On 20 May 2003, the new law on SOE s was adopted.
The five goals of SOE s mentioned in Article 2 paragraph (1) of Law No. 19 of 2003 on SOE s relate to economic and social development. They are: contributing to the development of the national economy and increasing the income of the state; profit maximization; offering a public function by providing public goods and public services of high quality to fulfil the basic needs of the people; becoming pioneers in segments where private firms and corporations are not yet active; and supporting the initiatives of entrepreneurs from underprivileged groups, corporations, and local communities. A single SOE does not need to fulfil all these goals at once, which has raised concerns that policies will focus on state revenues and profit maximization only while ignoring social functions.15 The law seeks to increase efficiency of SOE s in two ways: through the corporatization of the management of SOE s and through (partial) privatization. The corporatization of management includes the implementation of the international standards of good corporate governance. After the good corporate governance standards have been implemented, the next step is (partial) privatization.
Privatization of SOE s, however, never really took off after the Reformasi (2004)(2005)(2006)(2007)(2008)(2009)(2010)(2011)(2012)(2013)(2014)(2015). Despite the stated preference for privatization in the SOE Law, in practice there was insufficient political support for a large-scale overhaul of SOE s. Privatization progressed slowly in this context of strong opposition. The main problem was that the civil and military bureaucrats who controlled the companies were not willing to give up their vested interests in SOE s (Sungkar 2008:111). Citizens' concerns on the issue of SOE s and the bureaucrats' vested interests sometimes aligned, while at other times they collided, for instance in the forestry and plantation sectors (Lucas and Warren 2003).
Privatization took the form of partial privatizations of the largest limited liability companies, with the government maintaining a majority of the shares (Kim 2018). Between 2004 and 2017, eight companies were partly privatized. The fall in the number of SOE s from 159 to 115 in this time period (see Table 2), is solely the result of mergers carried out with the objective of making SOE s more efficient and profitable before 'going public' . 15 See also 'Kembali mempersoalkan peran negara dalam UU BUMN' , Hukumonline.com, 10-2-2018. https://www.hukumonline.com/berita/baca/lt5a7eb750f044e/kembali -mempersoalkan-peran-negara-dalam-uu-bumn/ (accessed 23-10-2019). The rationale for the increased role of SOE s in Indonesia's economic infrastructure development is pragmatic. SOE s are the only entities that are likely to obtain sufficient capital to carry out such large-scale infrastructure projects. As Indonesia has insufficient tax revenues-it has the lowest tax-revenue ratio in the Asian Pacific region18 and does not manage to attract sufficient foreign direct investment to finance large-scale infrastructural projects-the government needs SOE s to both finance and carry out the projects.
According to the Indonesian government, the way for SOE s to generate more funds is by organizing them into holdings. Ministry secretary Imam Apriyanto has recently stated that privatization of SOE s is no longer on the agenda, as the main focus is on setting up holding programmes for infrastructure, housing, pharmaceuticals, and asset management. He continued: 'By doing this, we can speed up the process of getting extra funds without using the state budget.'19 Smaller cement, fertilizer, plantation, and forestry SOE s have been merged with this objective in mind.20 The budget allocated to SOE s is not always managed well. Since 2015, the Komisi Pemberantasan Korupsi (KPK; the Corruption Eradication Commission; hereafter KPK) has filed 241 corruption cases related to infrastructure projects.21 Corruption by SOE s is persistent in other sectors as well and involves some of the most renowned SOE s in Indonesia. The list of KPK suspects for the shared with shareholders' , pinterpolitik.com, 12-3-2018. https://pinterpolitik.com/indone sias-growing-state-enterprises-information-needs-shared-shareholders/ (accessed 23-10 -2019). Pending the elections of 2019, the government disregarded persistent managerial problems at SOE s. However, after it was clear that President Jokowi had won the elections, the re-appointed minister of finance, Sri Mulyani, scolded the SOE bosses for continued practices of corruption.23 Despite Jokowi's pleas for a revolusi mental (mental revolution) and more efficiency, today's restructuring of SOE s primarily takes place as a way to secure more funding for projects, while improving management and accountability appear to be a secondary concern. Efficiency and productivity are mainly assessed by measuring how much of the projects have been realized (3,000 km of roads, 40,000m of bridges, toll roads, railways, airports),24 as well as the tax and non-tax state income SOE s generate. In view of the weak performance of Jokowi's predecessors in developing the infrastructure, this output is indeed impressive; yet, managerial issues remain and require just as much attention.
When the blueprint for SOE s was developed in 1998, the rationale for creating holding companies was threefold: synergy between SOE s and economies of scale will reduce inefficiencies; centralized holdings will decrease the influence of politicians and sectoral ministries over SOE s' management; and the focus on holdings will reduce the supervisory load for the government. According to the blueprint, the processes of corporatization and partial privatization must precede the establishment of holding companies. The OECD maintains that President Jokowi's administration rushes the creation of holdings, without creating the required corporate and regulatory infrastructure first.25 Indonesia looks at neighbouring countries Singapore and Malaysia, which have successfully improved the performance of SOE s by organizing them into state-owned holdings. However, in these countries managerial reforms preceded the structural reforms. As Kim (2018:326) phrases it: 'It is not the SOE ownership structure that has automatically resulted in market-oriented behaviour of, for example, Singaporean SOE s, but rather the institutional and legal basis that have supported more autonomous corporate strategy-making mechanisms while improving oversight apparatus.' Indonesia now organizes the holdings first and attempts to improve the SOE s' checks and controls and accountability in the process. Policies pertaining to PSA s see a similar course of action: structural reforms are introduced while lacking the required strong regulatory framework.

The Creation of Public Service Agencies (2004-now)
In 2004 within the sphere of the government which provides public services in the form of goods or services, is non-profit, based on the principles of efficiency and productivity' .27 The main difference with other government agencies is that PSA s have a larger autonomy to manage their budget and are allowed to conduct commercial activities to supplement their budget (Waluyo 2018). Further, according to Article 68 of the 2004 State Treasury Law, PSA s are formed to enhance public services for the welfare and intelligence of society; have assets that are state-owned, not separated from the state budget, and managed and utilized to carry out the activities of the PSA; and fall under dual supervision: financial supervision by the MoF and technical supervision by the sectoral ministry. PSA s' management form, especially their dual supervision and relative freedom to manage their financial affairs, is very similar to the managerial organization of the public service companies prior to their placement under the Ministry of SOE s (see Table 1). This dual management by the sectoral ministry and the MoF could not tackle persistent managerial issues pertaining to SOE s and, as we argue, will face the same issues with PSA s as long as it lacks a strong regulatory framework which is based on a clear ideological foundation. PSA s have substantial influence on the tariffs they offer. The management of PSA s proposes the height of the tariffs (the Minister of Finance's approval is required). According to the website of PSA s, the PSA management must consider four aspects in implementing their financial autonomy: continuity and improvement of services; the purchasing power within society; the principles of justice and fairness; and healthy competition.28 Financial considerations are dominant. The continuity and improvement of public services are monitored through the financial reports of PSA s, whereas the quality of services is not mentioned. Purchasing power is defined as the ability and willingness of the people to pay for services. The principles of justice and fairness are to mean that the economic backgrounds of the service receivers must be considered in tariff setting and that tariffs must be consistent with government regulations. The principle of healthy competition means that PSA s must charge tariffs that are similar to those charged by private players in similar market circumstancessuch tariffs are considered as proper (wajar). This means that, with the exception of applicable government standards, market prices are leading.
Over the past 15 years, the government of Indonesia has established 222 PSA s in 22 ministries (see Figure 2). More than 90% of the PSA s are active in the health and education sectors (see Figure 3). The other 10 % manage indus- trial estates and financial services for cooperatives, small businesses, as well as micro-and medium-sized enterprises. Figure 3 shows that a large majority (90%) of PSA s operate in the health and educational sectors. Government spending in the health sector (IDR 123.1 trillion, or 5 % of the 2019 state budget) and education sector (IDR 492.5, or 20% of the 2019 state budget) is substantial in absolute terms29 but still low compared to GDP.30 The government presents the establishment of PSA s as a success, stressing the positive effects on the state's budget. In 2018, revenues generated by PSA s contributed IDR 55.4 trillion to the state budget, more than the projected IDR 43,3 trillion, up from IDR 29.7 trillion in 2014. The largest contributor, with IDR 11.5 trillion, was the Badan Pengelola Dana Perkebunan Kelapa Sawit (the Office for the Management of Palm Oil Plantations Funds).31 The performance of PSA s in other areas, such as improving accountability and public services, is much weaker. For instance, in the education sector, both the Kementerian Riset dan Teknologi (Kemenristek, Ministry of Research and Technology) and the universities, when confronted with the difficulty of ensuring poor students' access to higher education, ignore the problems underprivileged 29 Kementerian Keuangan Republik Indonesia, 'Informasi APBN 2019: APBN untuk mendorong investasi dan daya saing melalui pembangunan sumber daya manusia' . https://www .kemenkeu.go.id/media/11213/buku-informasi-apbn-2019.pdf (accessed 23-10-2019). 30 For 2017 education public expenditure to GDP was 3.6 %, while healthcare public expenditure to GDP was 2. students experience in accessing those funds;32 instead, they point to special grants for underprivileged students and cross-subsidization that are intended to keep public services accessible.33 The case of the national health insurance below will illustrate how difficult it is to prevent conflicts of interests when large government funds are involved. The case demonstrates how informality-related problems persisted after the management of state insurance funds was transferred from an SOE to a PSA, from a profit-oriented to a non-profit organization.

9
Restructuring of the Universal State Health Insurance The reforms in the health insurance system took place in the context of the constitutional amendments of 2002, especially Article 34 (2) of the Constitution, which establishes the right to social security. In 2004 a new social security law was introduced to regulate universal coverage of its national public health insurances. A public health insurance scheme for the poor (Asuransi Kesehatan Miskin, Askeskin) was rapidly developed in order to increase health insurance coverage rate in Indonesia. The SOE PT Askes was given the responsibility to manage the Askeskin programme, renamed the Public Health Insurance ( Jaminan Kesehatan Masyarakat or Jamkesmas) in 2008. Under the Askeskin and Jamkesmas programmes the coverage rate of health insurances in Indonesia increased significantly, from 20% in 2005 to 49 % in 2014 (Rokx et al. 2009). Despite this success, the programmes experienced many problems, including a lack of funds, conflict of interests, lack of supervision, and fraud (Pisani, Kok, and Nugroho 2017).
public health insurance scheme was given to the Badan Penyelenggaraan Jaminan Social (BPJS, Body for the Organization of Social Security), an independent government body that operates directly under the Kantor Kepresidenan (Office of the President), as the commercial function of the SOE PT Askes was deemed to be in conflict with the social service function it had. From a social perspective, the introduction of BPJS has been a success, as the number of Indonesians who are insured had increased from 145 million persons in 2014 (coverage of 49% of the population), of whom 86 million were subsidized at the start of the programme (Hidayat et al. 2015), to 203 million persons (coverage of 76%) by 2018, 120 million of whom were subsidized (the Indonesian Health Systems Group 2018). However, management of the BPJS schemes remains problematic. By 2018 BPJS had a negative balance of IDR 23 trillion and required a financial injection of IDR 15.6 trillion. Besides a hasty introduction, conflicts of interests between stakeholders and insufficient supervision continued, facilitating fraud.34 An earlier study found that PSA s did not improve service delivery in the health sector (Damhuri 2011). This means that the managerial problems attributed to the profit-oriented SOE PT Askes persisted under the non-profit PSA BPJS. Changing the outward structure of Indonesia's state health insurance management did not solve managerial problems that are related to informality of public management and lack of controlling mechanisms.

Restructuring Higher Education
In the education sector, the other sector in which most PSA s operate, the Reformasi era led to calls for autonomy. Under the New Order regime, institutions of higher education were placed under strict bureaucratic control, both financially and ideologically (Rosser 2018 first universities with sufficient capacity to manage their own finances were given the regular status of Badan Layanan Umum, (BLU, hereafter BLU-PSA).
As the most prestigious universities received the PSA status of BHMN and the second-best, the status of BLU, the formal status of a university became associated with its social status: the more financial autonomy, the more prestigious the university. The most prestigious universities increased their tuition fees significantly after they had been granted the freedom to do so. Because they were the model universities, other universities followed suit, not only BLU-PSA s but also those institutes for higher education that were still fully under control of the Kementerian Pendidikan dan Budaya (Ministry of Education and Culture) (Susanti 2011).
In 2009, the government adopted Undang-Undang No. 9 Tahun 2009 tentang Badan Hukum Pendidikan (Law No. 9 of 2009 on Education Legal Entity; hereafter the BHP Law). The underlying philosophy of the law was that academic freedom must be matched with managerial and financial autonomy in order to improve the quality of education. The universities which had the status of BHMN based on a presidential decree-and therefore had to operate in an unclear legal framework-now were granted the status of BHP in accordance with the new law and received a high degree of autonomy (Ngo and Meek 2019). It was envisioned in the law that in the end all institutes of higher learningincluding private ones-would take the form of a BHP.
The new law met with strong resistance from within the education sectorespecially from private institutions-and from activists, who feared that the costs for higher education would rise further if the educational sector would be run based on market forces (Rosser 2018). The BHP Law was sent to the Constitutional Court for review. In 2010 the Constitutional Court annulled the law, as it ruled that the government disregarded its constitutional responsibility to ensure accessibility of education by giving a large financial responsibility to the universities and their overarching body. The Constitutional Court especially considered as problematic the fact that normal bankruptcy procedures would apply to state institutes of higher learning.35 Following the ruling, universities that had received the BHP status retained their former status as BHMN.
Undang A Badan Hukum university has a stronger mandate to charge relatively high tuition fees, as it mainly looks at the financial capacity of its targeted students and no longer at the purchasing power of society as a whole (see Figure 3). In 2016 state universities without managerial autonomy were still in the majority: 11 universities had the status of PT-BH; 24, the status of BLU; and 85, the status of satuan kerja, which means that they fully fall under the management of the Ministry of Education and Culture.36 Greater autonomy has resulted in a more diversified income for the universities. Whereas state universities in the 1990s were almost entirely dependent on state funding, today only 20% of the budget of Indonesia's state universities comes from state funding, while 60% comes from tuition fees and an additional 20% from service activities (Ngo and Meek 2019). However, the financial autonomy has not yet produced the intended large improvement in the quality of education and research in Indonesia. As the World Bank pointed out in its report on education financing in Indonesia (World Bank 2013), increased funding has not improved the organization of education within state universities. The quality of education in Indonesia still lags behind other countries in Southeast Asia (Rosser 2018). Unclear or complex regulations have not disappeared, corruption issues persist, and state supervision is weak (Komisi Pemberantasan Korupsi 2019).
The World Bank, in its review of the project 'Managing Higher Education for Relevance and Efficiency' , which ran from 2006 to 2012, still recommended 'more scope for revenue generation and charging appropriate tuition, although increases in tuition have been resisted to some extent by some sectors of the public' (World Bank 2015:4). However, as early as its 2013 report on financing education, the World Bank was warning that fees had become so high that they 36 See might come at the cost of the accessibility of tertiary education for the poor (World Bank 2013). Since this warning in 2013, tuition fees at Indonesia's universities have risen 15% annually; they are expected to double in the next five years and quadruple in the next ten years.37 The significantly higher amount of non-state funding has not resulted in improvements to universities' main functions, namely providing appropriate salaries for university staff and funding for research activities. Consequently, researchers in Indonesia 'often turn to other work to supplement their income' , mainly short-term projects (Ngo and Meek 2019:23). The government's claim of efficiency and effectiveness is difficult to maintain as long as the ultimate outcome of all the extra funding generated by tuition fees is neither better education and research nor improved management and accountability (Rosser 2018:19-20). Moreover, the new system creates a new problem: it impairs on the accessibility of higher education for the poorer segment of society, as access by underprivileged students to cross-subsidization and student grants schemes remains low.

Issues with New Public Management in Indonesia
The introduction of PSA s was intended to create more efficient and accountable government institutions. The cases of the health insurance and higher education do not indicate that restructuring has increased efficiency. On the contrary, in his analysis of Indonesia's PSA system, Jin-Wook Choi explains how the autonomy and flexibility of PSA s in financial management has decreased the accountability of PSA s. This resulted in a lack of governmental supervision over how PSA s' personnel and financial affairs are managed. For example, in 2014, 70 staff members at the Direktorat Badan Layanan Umum (Directorate of PSA s) at the MoF, many without specific qualifications for this task, had to monitor and evaluate more than 130 PSA s (Choi 2016:120). Performance improvement is the raison d'être of PSA s. A PSA is awarded a high degree of flexibility and autonomy in personnel and financial management under the premise that it will provide public services in a more efficient and effective manner. The success story of PSA s presented by the government, however, originates from a narrow focus on the state budget and non-tax 37 Yohanes Enggar Harususilo, 'Inflasi pendidikan tinggi, berapa biaya kuliah 5 dan 10 tahun lagi? ' , Kompas.com, 9-10-2018. https://edukasi.kompas.com/read/2018/10/09/13261841/infl asi-pendidikan-tinggi-berapa-biaya-kuliah-5-dan-10-tahun-lagi (accessed 23-10-2019).
revenues. Moreover, here are indications that at least some of the reported successes in terms of revenues are the result of inadequate financial reports.
Research into the financial reports of two state hospitals in Central Java found that the salaries of civil servants that are directly paid by the ministry were not included in financial reports. PSA s that employ civil servants therefore reported a much too positive financial performance, as if their employees worked for free (Ambariani 2014). Similar problems with disregarded employment costs are mentioned by Choi (Choi 2016).
A research report about PSA s in Indonesia by a research consortium led by the Korean Ministry of Strategy and Finance and commissioned by the government of Indonesia confirms that financial autonomy has not led to an improvement of the performance of PSA s (Ministry of Strategy and Finance Republic of Korea 2015). The report identifies multiple problems in Indonesia's PSA system, which can be subsumed under three main categories: governance structure; financial management; and performance management. With regard to governance structure, the report mentioned three problems. First, current regulations neither define the detailed roles and responsibilities of the MoF, sectoral ministries, and PSA s adequately, nor do they specify procedures and mechanisms regarding the management of PSA s. Second, there is a shortage of manpower and capacity within the Directorate of PSA s. Thirdly, the PSA s' governance lacks consistent and coherent managerial principles. Different types of PSA s with different forms of financial autonomy exist, which complicates supervision.
With regard to financial management, the ambiguity of budget composition causes problems. First, civil servants' salaries, operational expenditure, and the capital expenditure of the PSA are taken from the national budget and are not managed by the PSA s themselves. Ideally, PSA s manage all expenses related to their operations and they certainly must be included in financial reports. Second, there is no well-structured performance appraisal system, meaning that the MoF is not able to control inefficient spending of PSA s. Third, increasing expenditure of PSA s is not followed by improved performance. While many PSA s succeed in increasing their budget, they tend to increase unnecessary and redundant inputs, causing inefficiency in budget spending.
With regard to performance management, the report identified some key problems. First, performance appraisal in Indonesia is not linked to the PSA s' budget. High-performing PSA s are not rewarded and low-performing PSA s not punished through a budget reduction. There is a lack of incentives for PSA s to improve performance. Second, the assessment model for PSA s' financial performance is inadequate. The assessed KPI s do not represent all the dimensions of operation and management of a PSA.
In short, the report cautions that autonomy of public agencies must not lead to weak supervision. When financial autonomy is not met with accountability, this increased freedom can be misused by the management. Additionally, we argue that the problematic performance of PSA s must also be traced back to the unclear ideological foundation of PSA s. According to its legal definition, a PSA should not prioritize profit but public service. In practice, there is a pattern in which the government measures a PSA's success by calculating the revenues that it has generated, without looking at its overall performance as service delivery provider, and without assessing adverse consequences. PSA s mainly move in the education and health sectors, where the state has a constitutional responsibility to maintain accessibility. Autonomy has denoted an increase in service fees, which are partly paid by government subsidies-thus potentially impacting the sustainability, quality, and accessibility of subsidized state health care and education services for the poor and the lower middle class.

Conclusion
The comparison of SOE s and PSA s has enabled us to identify patterns concerning public sector restructuring in Indonesia. Restructuring in the line of New Public Management was considered necessary because of a lack of state budget due to low tax revenues. On paper New Public Management reforms require a formal corporate culture (corporatization). However, as changing an informal management culture is proven difficult and politically sensitive, new structures were generally introduced before corporatization and formalization processes had taken root. The case of SOE s demonstrates how policies aiming to achieve their partial privatization took place in a political context in which support for a nationalist economy remained strong and restructuring aimed at privatization mostly took place in a context of insufficient political support for privatization itself. As a result of political compromises, the new structures lacked sufficient control mechanisms and a clear division of tasks among the relevant government institutions. Thus, within these public sector institutions, an environment was created in which a culture of informality could continue. New structures created the illusion of reform, while in reality core issues that lay at the root of the informality of their management remained untouched and accountability remained weak.
The success of restructuring in terms of efficiency depends on how one defines efficiency. PSA s are efficient from the perspective of the state budget, as they manage to create much more revenue for the state than before the agency was introduced. The opposite conclusion is just as plausible: PSA s are inefficient as they do not improve public services, despite the much-higher revenues they generate. Moreover, reforms have not led to improved accountability.
Because restructuring took place without first creating the necessary regulatory and ideological framework and without changing the culture of informality, the implementation of restructuring depends strongly on the personal qualities of those in the highest ranks of the government. The persuasive power of President Jokowi, the MoF's Sri Mulyani, and the then minister of SOE s Rini Soemarno and others were key in getting bureaucratic reforms through the system during the first term of Jokowi's government. Patrimonialismnot a change of mindset or a revolusi mental within the bureaucracy-is the foundation of current bureaucratic reforms in Indonesia (Rosser 2018;Gaus, Sultan and Basri 2017). Today, there are signs that within Jokowi's broad coalition, opposition against bureaucracy reforms is rising,38 which may lead to less determination in implementing reforms.
We believe that the persuasive power of key ministers in President Jokowi's administration should be used to provide a clear justification for the establishment of PSA s. A 'clear justification' in our perspective refers to a clear philosophical foundation in which the social functions of PSA s and SOE s are underlined. Reforms in education and health sectors cannot be solely based on patrimonialism to be sustainable; they need strong financial control mechanisms and an accountability that values financial and social objectives equally. Quality of management and services should be just as important as revenues, and accessibility of services as important as bureaucratic reforms. Such an ideological foundation is essential to bringing about change in the bureaucratic culture-a real revolusi mental.