WPS6977 Policy Research Working Paper 6977 The Dynamics of Centralized Procurement Reform in a Decentralized State Evidence and Lessons from Indonesia Audrey Sacks Erman Rahman Joel Turkewitz Michael Buehler Imad Saleh Governance Global Practice Group July 2014 Policy Research Working Paper 6977 Abstract A central policy of the Government of Indonesia’s strategy local government procurement expenditures point to an for enhancing its country’s economic and social devel- overall decline in the volume of procurement, especially opment is to develop infrastructure and expand service in poorer districts. This paper uses qualitative case stud- delivery. Public procurement reform is a key component ies of procurement reform in six local governments and of this policy. Despite the decentralization of financial finds that local government leadership is associated with responsibility and authority to relatively autonomous the uptake of reform. There is little evidence to suggest that local level governments, procurement reform in Indone- procurement reform has been “demand”-led, since neither sia is a centrally-driven effort. In this paper, we examine the private sector nor Civil Society Organizations (CSOs) the extent to which procurement reform is translating have been active in advocating for procurement reform. into improvements in sub-national performance. Data on This paper is a product of the Governance Global Practice Group. It is part of a larger effort by the World Bank to provide open access to its research and make a contribution to development policy discussions around the world. Policy Research Working Papers are also posted on the Web at http://econ.worldbank.org. The authors may be contacted at jturkewitz@ worldbank.org. The Policy Research Working Paper Series disseminates the findings of work in progress to encourage the exchange of ideas about development issues. An objective of the series is to get the findings out quickly, even if the presentations are less than fully polished. The papers carry the names of the authors and should be cited accordingly. The findings, interpretations, and conclusions expressed in this paper are entirely those of the authors. They do not necessarily represent the views of the International Bank for Reconstruction and Development/World Bank and its affiliated organizations, or those of the Executive Directors of the World Bank or the governments they represent. Produced by the Research Support Team The Dynamics of Centralized Procurement Reform in a Decentralized State: Evidence and Lessons from Indonesia 1 Authors Audrey Sacks 2, Erman Rahman 3, Joel Turkewitz 4, Michael Buehler 5, and Imad Saleh 6 Keywords: Procurement Reform, Decentralization, E-Procurement, Indonesia JEL: H7, 018 1 This paper was part of a larger project on procurement reform ``Strengthening Reform and Improving Public Procurement Performance in the Africa (AFR) and East Asia and Pacific (EAP) Regions'' managed by the World Bank Institute's Open Government practice and funded by the Governance Partnership Facility Trust Fund. The Indonesia portion of this project was conducted in collaboration with the World Bank Office in Jakarta, Indonesia and The Asia Foundation. The data presented in this paper is based on information that was available and the situation that was prevailing at the time the study was undertaken. 2 Urban, Rural and Social Development – GP, World Bank 3 The Asia Foundation, Jakarta 4 Governance – GP, World Bank 5 Department of Politics and International Studies at the University of London’s School of Oriental and African Studies (SOAS) 6 Operational Services, Development Effectiveness, Latin America & Caribbean Region, World Bank 1 Introduction Policies designed to develop infrastructure and expand service delivery are key elements of the Government of Indonesia’s (GoI) strategy for the coun- try’s economic and social development. There is broad consensus among national officials and analysts that the expansion and improvement of the country’s transportation network is vital in order to expand internal trade and essential in order to improve the competitiveness of Indonesian manufac- turing. Increasing access to services such as education, health, and sanitation have a similar degree of prominence given their importance to improving the lives of people and communities and enabling individuals to reach their full potential. Indonesia is becoming one of the most decentralized countries in the world in terms of spending, with subnational governments spending 37 percent of total public funds (World Bank, 2007, xv). Local governments are heteroge- neous in terms of their wealth, administrative capacity, the quality of polit- ical leadership, and the preferences of constituents along with several other factors that are likely to impact local government efforts to improve citizens welfare. Despite the decentralization of financial responsibility and authority to relatively autonomous local level governments, procurement reform in In- donesia is a centrally-driven effort. The central government introduced new regulations, organizations, processes, and capacities to increase the efficiency and effectiveness of the procurement system. Given the vastly different condi- tions found across sub-national units, it is not surprising that early evidence suggests that the uptake and partial implementation of procurement reform is not translating into uniform improvements in procurement performance across local governments. Sub-national decision-making is not necessarily furthering Indonesia’s progress toward achieving national objectives. In fact, the decisions of a large portion of local governments with respect to how much to procure and what to procure appear to be at odds with the GoI’s developmental goals. Over the last few years, district governments have reduced the absolute and relative value of their procurement spending. Declining procurement signals a move away from capital investments in infrastructure and the purchasing of goods and services vital to expanding service delivery to spending on either administrative expenses or social protection. Existing disparities in infras- tructure quality and services across local governments appear to be widening, as poorer districts are particularly likely to be reducing their procurement 1 spending. Thus, the general trend of declining procurement expenditures is exacerbating widening disparities in economic and social well-being across the country. A review of the broader literature on procurement reform shows that very little attention has been paid to how procurement reforms are actually implemented in any setting and that studies are particularly sparse in rela- tion to the implementation of reforms outside of central agencies. Overall, policymakers tend to pay very little attention to the implementation of pro- curement reform in decentralized settings. The focus of donor agency and government efforts to monitor the implementation of procurement reform has been on compliance. An emphasis on enforcing compliance is particularly un- helpful in contexts like Indonesia, where local officials operate with a great deal of independence from central authorities. In the context of Indonesia, while LKPP is making an effort to support sub-national governments on pro- curement reform, the agency is not fully equipped to respond to the large number of districts and the scope of the reform. This study examines the implementation of procurement reform at the local level in Indonesia. It explores the pattern of the implementation of re- form across time and space, and empirically tests a set of initial hypotheses regarding factors that are likely to influence the uptake and implementa- tion of reform. Specifically, we test whether the quality of local government leadership and private sector voice influences local government decisions to adopt and implement procurement reform. Factors relating to the the efforts of civil society to provide an oversight role over procurement and the func- tioning and organization of the central government’s and local governments’ public financial management are also likely to contribute to the current pro- curement dynamics. This study is the most detailed analysis of local level procurement reform in Indonesia that has been undertaken to date and one of the most detailed analyses of local level procurement reform in any context. There are several limitations to this study including the recent nature of procurement reform in Indonesia, limited standardized and disaggregated data on local government procurement expenditures and performance and the small number of case studies that our research covers. Given these limitations, this paper should be viewed as an initial attempt to create a robust understanding of the dynamics of procurement reform in Indonesia. An additional contribution of this paper is the development of an analytical approach to procurement reform designed to provide policy makers with operational inputs. 2 The research presented in this paper is based on twelve months of desk and field work that was undertaken to provide the empirical basis for de- signing options for improving procurement performance in the short- and medium-term.1 Our case study narratives of public procurement reform in six districts and two provinces present a more detailed and nuanced picture of the dynamics underlying the adoption and sustainability of the reforms than this overview paper. The focus of other forthcoming analytical work (see Turkewitz et al 2013) is on the relationship between the uptake of procure- ment reform and subsequent improvements in procurement performance.2 Section 2 examines the main features of the public procurement system in Indonesia, including a discussion of its historical antecedents. Section 3 provides a justification for using Indonesia as a case study to examine sub- national variation in administrative reforms. Section 4 describes the study’s methodology and fieldwork. Based upon a scan of existing procurement data, Section 5 profiles procurement performance at the local level. Section 6 dis- cusses the features that we expect to be associated with the uptake and implementation of procurement reform at the sub-national level in Indone- sia and Sections 7 and 8 present the main findings of our case studies. We conclude the paper in Section 9 by offering a set of short-term policy rec- ommendations for the GoI and pressing empirical questions for policymakers within Indonesia and research on procurement more broadly. 2 Background on Procurement Reform in In- donesia Following the demise of the New Order regime in 1998, comprehensive pro- curement reforms were instituted as part of a broader effort to improve In- 1 This paper draws on background work that was commissioned by the World Bank to inform the overall project, “Strengthening Reform and Improving Public Procurement Performance in the AFR and EAP Regions” (Buehler, 2012; The Asia Foundation, 2012). The Asia Foundation case studies are a joint product of the World Bank and The Asia Foundation. 2 In this companion piece, we discuss the availability and limitations of of sub-national procurement data including: the scale of procurement; the quality of the contracting pro- cess; the quality and value of procurement outcomes; and the equity (including compliance) of the procurement process. 3 donesia’s PFM system and the provision of public services and works.3 In 2003, the GoI adopted Presidential Decree No. 80/2003, which was later replaced by Presidential Regulation (Perpres) No. 54/2010 on Public Goods and Services Procurement. The reforms cover goods, works, and services that use public funds irrespective of the size or value. It also established reg- ulations for government procurers at all levels of the bureaucracy. Central components of the reform include: the establishment of a National Public Procurement Agency (LKPP-Lembaga Kebijakan Pengadaan Barang/ Jasa Pemerintah); the roll-out of e-procurement; mandatory certification for offi- cials involved in procurement; and the establishment of Procurement Units (ULP-Unit Layanan Pengadaan) that are responsible for conducting public procurement procedures on behalf of the government’s technical departments (SKPD- Satuan Kerja Pemerintah Daerah) and national offices. First, established at the end of 2007 under Presidential Decree No. 106/2007, the LKPP is a fully independent body reporting directly to the president and chaired by an appointee from the bureaucracy. The LKPP is tasked with the development of policies related to the public procurement of goods and services. It also aims to strengthen procurement practices within the govern- ment and provides advice and recommendations as well as dispute resolution (Gatra, December 26, 2007, 12). The LKPP does not undertake procurement operations directly nor does it have any purchasing or contracting function. It is solely in charge of formulating procurement policies and overseeing their implementation. Second, Presidential Decree No. 80/2003 requires all local bureaucrats involved in procurement to be procurement-accredited. To implement this requirement, the GoI set up a certification system. The rationale for requir- ing accreditation for procurement professionals was to try and professionalize procurement within the bureaucracy. As a legacy of the New Order, procure- ment professionals were limited in numbers and mostly employed in selective line ministries. There were no distinct career paths or salary incentives for procurement professionals. Bureaucrats joined procurement committees on an ad hoc basis and upon completion of the project, returned to their former positions. Consequently, institutional memory with regard to procurement procedures remained fragmented and inefficient. Permenpan No. 77 issued in 3 We understand the term “public procurement” to include all phases of the process — the determination of needs, creation of specifications, tender process, contracting, contract management and oversight up to the point of receiving and accepting the good, work, or service. 4 2012 has tried to address this issue by creating a career path for bureaucrats. Third, the decree requires government units at both the national and sub- national level to adopt an electronic procurement system (LSPE — Layanan Pengadaan Secara Elektronik) for the processing of transactions by 2012. Transactions include: the publication of contract opportunities; distribu- tion of bid documents; submission of bids; and the publishing of contract award winners. Adoption of e-procurement is viewed as essential to improv- ing procurement outcomes due to its potential for reducing processing times, increasing transparency, and reducing opportunities for collusive behavior among bidders and between the government and the private sector. Utilizing a common e-procurement platform also creates the possibility of promoting the standardization of procurement processes, regardless of the size, sub- stance, or location of the specific transaction. The system also includes an on-line complaints mechanism that allows users to submit information about perceived malfeasance during tender procedures. If e-procurement is suc- cessfully rolled-out, the LKPP estimates that the LPSE will lead to cost reductions of IDR 315 trillion (US$ 34 billion), which is around 30 percent of the state budget. Fourth, ULPs were established in an effort to standardize the organiza- tion of procurement across the public sector. Presidential Regulation 54/2010 mandated that ULPs be established at all level of government by 2014. ULPs are formal technical departments responsible for creating procurement plans, scheduling and running of tender competitions, preparing cost estimations, handling contracting with the winning bidder, and receiving and responding to complaints and bidder appeals. ULPs do not have a mandate to monitor the implementation of works and services. Movement to create ULPs has been relatively slow, especially at the district level, perhaps due to a regu- latory conflict regarding the number of permissible technical departments in local governments.4 Perpres 54/2010 also aimed to increase transparency and competition and to criminalize collusive procurement practices. Under the new regulatory framework, procuring entities are now required to announce their procure- ment plans, tender invitations and other related information in designated national newspapers and the national procurement website. Documents in- 4 Government Regulation No 41/2007 established numerical limits on technical depart- ments at the district level. Many district governments have reached their limit and have elected not to form ULPs, relying on the greater authority of a Government Regulation over a Presidential Regulation based on the Indonesian hierarchy of laws. 5 cluding criteria for tendering committees and procurement project leaders have been standardized. To try and increase competition, Presidential Decree No. 80/2003 abolished an earlier stipulation that only contractors located in the province in which the tender was issued were eligible to submit tender applications. The Decree also explicitly criminalizes collusive procurement practices. Article 22 of the Competition Law, for instance, prohibits bid rig- ging and collusive tenders. Likewise, the adoption of Presidential Decree No. 85/2006 on the 6th amendment of Presidential Decree No. 80/2003 includes various articles aimed at the prevention of bribery and other corrupt conduct in public procurement. On paper, Indonesia’s procurement regulations are largely consistent with international standards and practices. In practice, there is large sub-national variation in the uptake of procurement reform, as well as variation in pro- curement performance (see Section 5). As we discuss in the next section, Indonesia offers a justifiable context in which to better understand the fac- tors that are associated with the uptake and implementation of procurement reform. 3 The Advantages of the Indonesia Case For several reasons, post-Suharto Indonesia provides a justifiable case in which to develop an empirically robust method to identify features that are associated with the uptake of procurement reform and improved procurement outcomes at the sub-national level. First, Indonesia is undergoing procure- ment reform and there is a particularly compelling need to examine how the GoI might modify reform efforts to be able to achieve improvements in procurement performance. Second, extensive decentralization of financial authority to local govern- ments in Indonesia presents enormous opportunities for increasing the effec- tiveness of public spending since it creates the potential for improving gov- ernment accountability to citizens. Between 2004 and 2011, higher levels of fiscal grants have been transferred to Local Councils. At present, subnational governments obtain most of their revenues from general block grants from which they finance around 80 percent of all district expenditure. Earmarked grants have increased in past years but are relatively small in comparison with other transfers. Own-source revenue is low at the district level (World Bank, August 2012) and is expected to stay small compared to general block 6 grants (World Bank 2007: xxi). Further, there has been an increase in the number of service delivery and administrative functions that have been decentralized to local governments. Local governments can now decide freely on how they want to spend the Gen- eral Allocation Grants (DAU), which constitutes the bulk of local revenues. Also the funds from revenue sharing are untied and expenditure of these, like the DAU, is completely at the discretion of subnational governments (Lewis 2005: 291-317). There is a constitutional mandate that requires subnational governments to spend a minimum of 20 percent of their budget on education. Over time, the capacity and effectiveness of local councils in delivering public services to their communities has improved. However, changes in local governments’ responsibilities have occurred unevenly. At the local govern- ment level, there are also significant and persistent problems in the level of public spending and in the delivery of public infrastructure and services, which makes the need for improving procurement performance more press- ing. Indonesian local governments may be spending more but they often fail to channel these resources in directions that would meet local citizens’ needs (Skoufias et al., 2011). Third, Indonesia’s decentralization process has yielded significant varia- tion in local government performance across sub-national units. Thus, In- donesia provides a good case to examine the antecedents of the uptake of reform given the sub-national variation on several key dimensions of inter- est including the strength of local governments’ public financial management (PFM) systems5 , the quality of local governments’ leadership and the com- petitiveness of the private sector. Fourth, as we discuss in Section 5, there is also a wide amount of variation in the uptake of procurement reform initia- tives and in procurement volumes across sub-national units. 4 Methodological Approach We first commissioned a scan of procurement reform at the district and provincial levels in Indonesia to be able to examine variation in both the uptake of procurement reform initiatives and in spending patterns, as well as variation in the composition of procurement expenditures over time. As 5 Initially, we wanted to look at the role PFM systems play in procurement reform but were unable to do so due to a lack of qualitative and quantitative data on PFM in the districts in which we conducted the field work. 7 Table 1 illustrates, these initiatives are grouped into the following four dimen- sions: (1) process — the establishment of e-procurement; (2) organization — the establishment of ULPs; (3) capacity – the certification of officials in procurement; and (4) monitoring — CSO oversight and the establishment of complaint handling mechanisms. 8 Table 1: Local Government Reform Initiatives Concept Reform Initiative Process Eprocurement Organization Procurement Service Units (ULPs) Capacity Mandatory certification of officials involved in tender committees Monitoring CSO oversight & complaint handling mechanisms Our indicators of the four initiatives include the following: 1. Eprocurement • Status of the implementation of e-procurement • Number and value of packages procured through e-procurement between 2010 and 2010 2. ULPs • Status of the establishment of ULPs 3. Mandatory certification of officials involved in tender committees • Description of initiatives to develop the capacity of GoI officials • Number of procurement accredited staff 4. CSO oversight and complaint handling mechanisms • Description of CSO or community monitoring activities • Status of the establishment of complaint handling mechanisms The scan also collected data on the following three budget items for 2007 to 2010: total expenditure; procurement expenditure, or the total allocated for goods and services and capital expenditures; and the contribution of unspent budget to total expenditure. We then commissioned structured case studies that are paired around three hypotheses relating to three factors that are associated with procure- ment performance: the quality of PFM systems for planning, budgeting, spending and oversight of public money; the quality of a local government’s 9 leadership; and the strength of the private sector (see Section 6.1). The case study narratives also brought to our attention two additional variables that contributed to the uptake of reform: the role of CSO voice; and inter-district competition and diffusion of knowledge. We discuss these alternatives expla- nations in Section 8. Fieldwork was carried out in four regencies — North Luwu, Sijunjung, Madiun, and Lamongan — and two cities, Makassar and Balikpapan. To select districts, we used data from a scan of procurement reform efforts at the sub-national level to create a database of all districts that have initiated procurement reform. From this sample, we selected pairs of cases that have different values on a key explanatory variable but similar values on control variables. From the subset of candidate districts, we selected cases that are relatively accessible and where either B-Trust or The Asia Foundation has contacts (see Table 2). Table 2: Case Selection Variables Low High Leadership Quality North Luwu Sijunjung Private Sector Competitiveness Madiun Lamongan Public Financial Management Makassar Balikpapan We selected two districts, North Luwu and Sijunjung, which differ on the quality of the leadership of their mayor or regent but have similar values on other explanatory and control variables (e.g. population size and whether the district is urban or rural). We measured leadership quality using data from the 2007 and the 2010/2011 Local Economic Governance Surveys (KPPOD and The Asia Foundation), which covered 12 and 20 provinces, respectively. These surveys probed business owners about their perceptions of the integrity and capacity of the Mayor/Regent. The survey asked respondents about whether they agreed that the: (1) mayor/regent takes strong action against corruption cases, (2) mayor/regent is not corrupt; and (3) the mayor/regent has strong leadership. We selected two cases, Madiun and Lamongan, which have different levels of private sector competitiveness but have similar values on other variables. We used the 2009 Regional Gross Domestic Product (RGDP) as an indicator of private sector competitiveness. For districts with missing 2009 RGDP values, we used 2008 data. 10 Finally, we selected two districts, Makassar and Balikpapan, which differ in the quality of their governments’ PFM systems but have similar values on other variables. As a measure of PFM quality, we use the average procure- ment budget (capital investment, goods and services) as a proportion of total expenditure for 2007 to 2011.6 The budget data for 2007 and 2008 is actual budget realization and the data for 2009 to 2011 is planned expenditure.7 Since the narratives of the case studies did not include sufficient detail on the relationship between the quality of Makassar’s and Balikpapan’s PFM systems and the uptake of reform, we chose to eliminate the discussion of this relationship from this paper. In a forthcoming paper (Turkewitz et al. 2013), we use quantitative, time series data on the quality of districts’ PFM sys- tems and procurement to illustrate the relationship between PFM systems, the uptake of reform and procurement performance. In each of the six localities, teams from Bandung Trust (B-Trust) and The Asia Foundation carried out fieldwork for a period of two weeks. The research teams used a snowball sampling approach to identify key individuals involved in procurement for the study. Using instruments, they conducted in-depth interviews and focus groups with key respondents from the local executive and legislative branches, government departments, the private sec- tor, journalists, academics and civil society. The teams collected government documents, data and information, as well as policy documents, newspaper clippings, reports and books. A short description of the local governments included in our case studies is below: • North Luwu is a remote district (population 450,000) of South Sulawesi that was a sub-district of Luwu district until 1999. Relatively large geographically, North Luwu has large forest reserves (55% of the district is zoned as protected forest) and natural resource opportunities that 6 Our measure of PFM quality average expenditure on capital investment and goods and services is an approximate but best available measure. 7 We considered using the proportion of “unspent budget” (SILPA) and the value of the administrative budget (personnel, interest, subsidy and contingency) as proxies, but decided not to use these as indicators for two reasons. Our first batch case studies found that a higher SILPA may indicate a higher level of competition in procurement rather than the district’s inability to manage its budget. The administrative budget was not used since it is already reflected by look at the average proportion of procurement in the district budget (i.e., a higher administrative budget corresponds to a lower procurement budget). 11 include timber, rattan, palm oil, chocolate and oranges. The district has a large migrant population from Java and Makassar, along with indigenous communities. The forest sector is the largest contributor to district revenues. • Sijunjung is a small rural district in West Sumatra. Established in 1949, it was first named Sawahlunto/ Sijunjung until Kota Sawahlunto became administratively separate in 1965. In 2008, the district fur- ther split into Sijunjung and Dharmasraya, leaving the former with a population of around only 200,000. With a geographic area of 3,180 km2 , Sijunjung is sparsely populated by Indonesian standards. Its per capita economy is around the Indonesian average: in 2008, the district’s RGDP of IDR 2.4 trillion (USD 269 million) averaged out at IDR 12.1 million/person. Its poverty rate of 18.1% is quite high for Sumatra. • Madiun is in East Java, located alongside the major transportation lines between Surabaya and Yogyakarta. With a population of 770,440, and a published poverty rate of about 6.5%, its economic base is mostly agricultural (37%) followed by the trade and service sectors. • Lamongan is quite a large district, with a population of 1.4 million. It is located near East Java’s capital city of Surabaya and its economic base is dominated by agriculture, fisheries and home industries. The population has a comparatively high level of poverty, with 16.5% of the people living in poverty. The construction sector has grown in the past decade due to the expansion of privately-funded industrial and business projects, and the development of public infrastructure including roads, bridges, irrigation, the port, airport and tourism-related projects. • Makassar is the provincial capital of South Sulawesi and is southern Sulawesi’s primary port. As of 2010, its population size was 1.34 mil- lion with an official poverty rate of 5.86%. Its economic base depends primarily on the service sector, followed by the restaurant and hotel business, transportation and communication, trading and finance. • Balikpapan is a municipality located in the natural resource-rich province of East Kalimantan. Its population of 621,862 mostly consists of second— and third-generation migrants from South Sulawesi, North Sumatra 12 and East Java who have established service sector businesses that sup- port the large-scale oil and gas, palm oil, coal and other mining op- erations in the districts neighboring the municipality. Balikpapan is a relatively affluent city with a low official poverty rate of only 2.5%. 5 Trends in Procurement Expenditure Overall, since 2007, average annual procurement spending at the district level has decreased, which has implications for local governments’ ability to provide essential public goods and services. From 2007 to 2011, real procure- ment expenditure (IDR Billion, 2011 price) at the district-level: declined in 65 percent of districts; increased in 13 percent of districts; and remained relatively constant in 22 percent of districts (see Figure 1). Decreases in the proportion of district spending that occurred via procurement were of a simi- lar magnitude, moving from 50% of expenditures in 2007 to 41% in 2011. By comparison, average annual procurement spending increased at the provin- cial level over the same period. Specifically, real procurement expenditure increased in 45 percent of provinces, declined in 15 percent of provinces, and, remained constant in 39 percent of provinces (see Figure 2). As Fig- ure 1 shows, districts in ‘Outer Island’ Indonesia spend more on procurement than provinces and districts in Bali and Java. The same is true for provinces. 5.1 Process, Organization and Capacity: Variation in the Uptake of E-Procurement, ULPs and Certifi- cation of Officials Following the passage of the 2003 Presidential Decree, national ministries and sub-national governments developed e-procurement systems. For example, the Ministry of Public Works and the Municipal Government of Surabaya developed an e-procurement system, which it subsequently promoted in a number of other districts. Since the 2010 Presidential Decree requires all sub- national governments and ministries to adopt the LPSE system, ministries, as well as local governments that initially adopted the Surabaya or another system, are supposed to be transitioning to the LPSE system. As of 2011, of Indonesia’s 497 districts, 21.92 percent instituted the government-promoted LPSE and an additional 55.13 percent instituted a par- tial e-procurement system, making e-procurement the most widely adopted 13 Average Growth in Real Procurement Expenditure (2007−2011) (IDR Billion, 2011 Price) missing under −0.15 −0.15 − −0.05 −0.05 − 0.05 0.05 − 0.15 0.15 − 0.25 over 0.25 Figure 1: As this figure illustrates, from 2007 to 2011, real procurement expen- diture (IDR Billion, 2011 price) declined in the majority of Indonesia’s districts. Specifically, real procurement declined in 65 percent of districts, increased in 13 percent of districts and, remained relatively constant in 22 percent of districts procurement reform initiative. The adoption rate of LPSE, which is still quite low, is significantly lower in Outer Island Indonesia (see Figure 3). A small number of provincial and district governments have adopted and utilize the LPSE system to accept bids and announce tenders. All five local govern- ments (LGs) in DI Yogyakarta province use the LPSE for bid submissions, in addition to announcing tenders. In West Java, 81% of the 26 LGs have adopted the LPSE while the remaining 19% are only using a partial e-procurement system for tender announcements. The proportion of LGs that have adopted the LPSE is also high in South Kalimantan (62%), Central Java (60%) and West Sumatra (58%). By comparison, the use of e-procurement is still very low in several provinces throughout Eastern Indonesia. None of the five local governments in West Sulawesi has instituted e-procurement. Less than 15% of the LGs in Papua, West Papua, Maluku and North Maluku use e-procurement. On average, LGs with LPSE have higher procurement expenditures than those with a partial e-procurement system or those without e-procurement systems. Between 2007 and 2011, LGs using the LPSE also experienced a higher growth in procurement expenditure as a portion of the total budget compared to LGs, which have a partial or no e-procurement system. The ma- 14 2007-2011 Average Procurement Volume (IDR Billion, 2011 Price) by Proportion of Proc. to Total Expenditures and Trend of Proc. Expenditure 70% Average Annual Growth 2007-2011, Constant Price 60% Aceh 50% 40% W Papua Bengkulu 30% 2011 20% E Java DKI Jakarta W Java N Sumatra W Nusa Tenggara N Maluku 10% Bali W Sumatra W Kalimantan C Java S. Sulawesi Riau Islands 0% E. Kalimantan S Sumatra Riau Papua -10% 20% 25% 30% 35% 40% 45% 50% 55% 60% 65% Average Percentage of Procurement to Total Budget (2007-2011) Figure 2: As this figure illustrates, from 2007 to 2011, real procurement expen- diture (IDR Billion, 2011 price) increased or remained constant in all but five provinces. jority of LGs only use e-procurement to procure a small portion of their total budget for goods, services and capital investment. On average, the value of the packages procured through e-procurement is 11% of the budget allocated for procurement expenditure. Only 18 LGs are using e-procurement to pro- cure more than 20% of their procurement budgets. Three LGs, Kota Banda Aceh, Kota Tangerang, and Kota Sawahlunto, are using the LPSE system for more than 50% of their annual budget for procurement expenditure. The total value of packages tendered using e-procurement in 2011 was the highest in Kota Surabaya (IDR 843 billion), Kota Tangerang (IDR 291 billion) and Kota Pekanbaru (IDR 219 billion). As of 2011, ULPs have only been adopted in 26% (130) of LGs. As Figure 4 shows, districts with ULPs are concentrated in eight provinces: West Java, South Sulawesi, East Java, Central Java, West Sumatra, South Sumatra, Central Kalimantan, and Central Sulawesi (7 LGs). Meanwhile, none of the 63 LGs in Papua, Riau, Jambi, and Maluku have established a ULP. There is less variation across districts in terms of local capacity build- 15 Use of E−Procurement as of 2010/2011 No E−procurement Partial E−Procurement E−Procurement Figure 3: As of 2011, of Indonesia’s 497 districts, 21.92 percent instituted the government-promoted LPSE and an additional 55.13 percent instituted a partial e- procurement system, making e-procurement the most widely adopted procurement reform initiative. ing for public officials including certification and training in how to oper- ate and maintain e-procurement. From 2007 to 2011, the overall number of procurement-certified staff has increased at the district level across Indonesia. Almost two thirds of districts have more than 100 procurement-accredited staffs (see Figure 5). There is a strong association between the number of staff certified in procurement in a district and average procurement expen- diture (real terms) between 2007 and 2011; the bivariate correlation, or R2 , is 0.48. There is a strong correlation between the uptake of one component of procurement reform and the uptake of another. Local governments that have adopted the LPSE are both more likely to have implemented ULPs and more likely to have a higher number of staff certified in procurement. Of the LGs with the LPSE or partial e-procurement, 45% and 23%, respectively, also have a ULP. By comparison, of the LGs without e-procurement, only 18% have a ULP. The number of procurement-accredited staff is higher in LGs with the LPSE (µ = 222.29) than in LGs with partial (µ = 83.95) or 16 Establishment of a ULP as of 2010/2011 No ULP ULP Figure 4: ULPs have only been adopted in 26% of LGs, or 130 LGs. As Fig- ure 4 shows, districts with ULPs are concentrated in eight provinces: West Java, South Sulawesi, East Java, Central Java, West Sumatra, South Sumatra, Central Kalimantan, and Central Sulawesi. no e-procurement (µ = 142.99). LGs with ULPs are only likely to have a larger number of staff certified in procurement (µ = 174.88) compared to LGs without ULPs (µ = 83.95). The scan points to a strong positive association between the wealth of a district and the uptake of e-procurement, as well as the certification of staff involved in procurement. The average RGDP (2008, nominal terms) for districts with LPSE and without e-procurement is IDR 13,120,497 and IDR 2,490,203. The correlation, or R2 , between RGDP (2008, nominal terms) and the number of staff certified in procurement as of 2011 is 0.36. Although positive, the correlation between a district’s RGDP and the establishment of a ULP is very small (0.02). As the maps show, there are also geographic clusters of reform localities. As shown in Figure 3, Figure 4 and Figure 5 very few LGs throughout Outer Island Indonesia including Papua, West Papua, Maluku and North Maluku have implemented procurement reform. By comparison, of the 43 districts 17 Number of Staff Certified in Procurement as of 2010/2011 missing 0 to 50 50 to 100 100 to 150 150 to 200 200 to 250 over 250 Figure 5: From 2007 to 2011, the overall number of procurement-certified staff has increased at the district level across Indonesia. Almost two thirds of districts have more than 100 procurement-accredited staff. that have set up both the LPSE and ULP, the majority falls in West Java (25.6%), West Sumatra (19%) and Central Java (19%). Future research will examine the correlates of the spatial clusters of reform including districts’ distance to the early innovators, especially Surabaya, the presence of natural resources and construction booms and the competitiveness of local elections. Next, we turn to a discussion of the likely drivers of the adoption of procure- ment reform, specifically, leadership and private sector voice. 6 Theoretical Framework 6.1 Leadership and the Uptake of Reform A central focus of recent literature on policy reform is on leadership and how leadership contributes to positive macro-economic, political and orga- nizational outcomes and public goods provision (Ahlquist and Levi, 2011; 18 Grindle, 2007, 2004). For example, from an analysis of eight controlled case comparison in Indonesia, which includes 1,000 business surveys and 120 in- terviews, von Luebke (2009) finds that the quality of government leadership plays a central role in driving the quality of policy outcomes.8 Specifically, he finds that district heads with solid managerial skills and long-term career aspirations are more likely initiate broad-based reform and supervise bu- reaucratic performance. Similarly, we expect the quality of local government heads to affect whether they adopt procurement reforms and try to improve local governments’ procurement performance. There are a number of different arguments as to how leaders can posi- tively impact these outcomes. Research points to the role that leaders play in overcoming collective action and coordination problems within organiza- tions by reducing information asymmetries and building trust between agents (Ahlquist and Levi, 2011; Calvert, 1992). Other research shows how lead- ers can effect change in an organization through the institutionalization of a system of rewards and punishment that shifts how individuals work to- gether (Andreoni et al. 2003). Within an organization, when leaders treat subordinates with respect, subordinates respond with loyalty and greater productivity towards producing public goods (Cook, Hardin and Levi, 2005; MIller, 1992). The social choice tradition stipulates that leaders who know how to ma- nipulate the institutional situation can shape outcomes to their own benefit or can modify institutional design to suit their goals (Arrow, 1970). A leader who can control the agenda can also shift the equilibrium from any status quo to almost any desired outcome (Ahlquist and Levi, 2011). In other words, in the presence of multiple equilibria, leaders can help to shift the group decision to their preferred outcome (Ahlquist and Levi 2011). Consequently, leaders often serve a narrow constituency at the expense of others. Yet, institutional arrangements, specifically agenda control and the presence of multiple veto players including opposition parties and oversight institutions, can constrain leaders. The extent to which leaders are able to influence policies depends on institutional context — whether the country is, for example, a democracy or autocracy. Even within democracies, however, leaders do not necessarily serve the interests of the country. A larger literature documents how lead- 8 von Luebke (2009, 210) measures the quality of government leadership using n index consisting of perception based indicators of the district heads’ integrity, political will, communication skills, power to change and popularity levels. 19 ers can be harmful for countries (Weber, 1968). For example, leaders who adhere to ideology and ignore evidence-based policies and even constituents’ demands are particularly dangerous. When leaders ignore feedback and dis- courage learning, they are also likely to be harmful (Levi, 2006). Leadership often involves building coalitions within and across organiza- tions to be able to help implement a specific policy. There is some but limited research that shows that reforms are more likely to be successful if leaders across different hierarchical levels of authority come together as a coalition and assume their various responsibilities and tasks (Fligstein, 1997; Mazaheri et al., 2013). Most of the academic research on coalitions is mathematical and abstract, or it focuses on the politics of developed countries with well- organized parties, lobbies, and interest groups. There is very little academic literature on coalition building within the civil service and in developing countries (Humphreys, 2008; Maravall, 2010; Riker, 1962). Coalitions face a myriad of obstacles that undermine their formation and durability including: difficulties in bargaining enforcement, free rider problems and the existence of multiple potential sets of possible coalitional partners (Levi, 2011). The election of a new politician or the appointment of a civil servant may result in renegotiation among the coalition partners, which can potentially undermine the coalition. To summarize, we expect that the take-up of procurement reform will be closely associated with support from the executive and that support from the executive is associated with quality of leadership. 6.2 Private Sector Voice Demand side pressures for local governance procurement reforms can po- tentially emanate from business firms and associations. For several reasons, businesses should have a stake in seeing procurement performance improve. In order to grow, firms need adequate infrastructure, such as roads, electric- ity networks, ports, and education systems that can produce skilled employ- ees. Businesses should have an incentive to pressure governments to improve the quality of procured goods, especially roads. Further, in contexts where bidding is susceptible to collusive practices, business that are not part of powerful cartels are likely to have an incentive to push local governments to improve the competitiveness and transparency of the tendering process. Yet, like CSOs, businesses face a collective action problem that can im- pede their ability to demand better procurement performance from govern- 20 ment. In Indonesia, there are different types of firms including small firms consisting of shop-keepers and craftsmen of various ethnic origins and repre- senting a large part of civil society. There is also a small group of large indige- nous firms that generate revenue from construction, procurement, resource extraction and manufacturing (Luebke, McCulloch and Patunru, 2009). Both groups are unlikely to play a role in demanding governance reforms. Because of its large and dispersed structure, the first type of firm faces the classic collective action problem. Business associations help the businesses to solve this collective action problem. The incentives for the latter group, large in- digenous firms, to criticize government and demand administrative reforms are very low given that these firms would not want to risk losing access to lucrative government contracts from which they currently benefit (Luebke, McCulloch and Patunru, 2009). Business associations help to solve collective action problems within the private sector. A large literature highlights the role business association’s play in facilitating rent seeking through lobbying and cartelization (Olson, 1982). Scholars also emphasize that business associations often impede the development of the introduction of new technologies (Pyle, 2006, 493). Other literature emphasizes that in weak states, business associations often press for clean and effective government administration and provide its members with services that promote economic growth and create social benefits (Doner and Schneider, 2000). As the literature on the developmental state in East Asia argues, when business associations regularly organize consultations with government and mediate between its members and government, they can en- hance economic development (Campos and Gonzalez, 1999). Further, busi- ness associations frequently disseminate information about new technologies and the reliability of potential trade partners (World Development Report, 2002). In localities that are economically competitive and where business asso- ciations represent a variety of types of firms, we expect business associations to play a role in demanding greater procurement performance. 21 7 Main Findings from the Case Studies on Drivers of the Uptake and Implementation of Reform 7.1 Leadership (Sijunjung vs. North Luwu) In Sijunjung, local government leadership, which consisted of a unique com- bination of three types of positions and skill sets, played a crucial role in local governments’ adoption and implementation of procurement reform. The first type, mid-level civil servants, created networks within the local government bureaucracy and between local governments to garner support and momen- tum for the reform. The second position, a civil servant, had the credible technical expertise to be able to implement the reform. The third type, a senior civil servant, had ties to the Executive and helped to secure the Exec- utive’s consent to proceed with the reform. By comparison in North Luwu, the district’s Executive (Bupati) drove the reforms and the coalitions of civil servants that implemented the reforms were substantially narrower than in Sjinjung. Ultimately, these two leadership styles had implications for the longer-term sustainability of the reforms. In Sijunjung, a small rural district, the Executive initiated procurement reform and a team consisting of a senior civil servant, mid-career civil ser- vants and IT specialists were responsible for implementing the reform. Our respondents traced procurement reform to early 2007, when the combination of an audit finding, alleged collusion and corruption in the tender process, the poor quality of completed procurement projects and a high proportion of unfinished contracts motivated the Bupati to initiate procurement reform. At the same time, there was too few procurement certified staff (as of 2006, there were only 30 staff) to handle the volume of procurement. In response to these factors, the Bupati established a ULP in 2007 and asked the ULP Secretariat to develop e-procurement. A team led by the Head of the ULP Secretariat, two mid-career Development Administration staff and a public official from the Inforkom Office with an IT background travelled to Surabaya to learn more about e-procurement. These local officials were critical to implementing the Executive’s directives. The two Development Administrative staff and the official from the Inforkom Office were also tech- savvy.9 9 It is also worth noting that of the 20 procurement-accredited staff from Sjinugng’s 22 Following this trip, although Surabaya offered to assist Sijunjung with the development of their e-procurement system, the ULP decided to develop its own e-procurement software. By 2008, the ULP and the Inforkom team had developed and piloted a new e-procurement system, which was even- tually replaced by the LPSE system in 2009. In addition to the ULP and e-procurement, Sijunjung undertook other procurement reform initiatives. To increase the number and capacity of procurement-accredited staff, the ULP encouraged SKPDs to use their own budget to train and certify addi- tional staff. Between 2005 and 2011, the number of procurement-accredited staff increased from 15 to 165. Sijunjung stands out in that it updates and publishes a list of standard unit costs for construction on a quarterly basis. The Executive issues the updated rates through a decree. Subsequent to the 2010 elections, continued executive and civil service support for reform has contributed to the sustainability of reform established in Sijunjung. We observed a similar pattern of reform in Lamongan, where three pro- active civil servants initiated procurement reform. Of the three civil servants, the first was a prominent young official, who was the Section Head for a Development Unit that was responsible for monitoring and assisting all local government technical units (SKPDs) in procurement practices. The Bupati instructed this civil servant to “make Lamongan well-known for reform.” The second member of the coalition was the head of a sub-directorate under the first civil servant’s supervision. The third was a former public works official nearing retirement, who was also close to the Bupati10 and was motivated to “make his mark” by implementing reform. Several factors drove this team of three to initiate the reform. First, there was a limited number of procurement-accredited staff in the LG. As of 2008, there was fewer than 50 accredited staff. Second, there were grow- ing concerns that corruption among vendors and political interference was undermining the competitiveness and transparency of procurement. Third, the team was inspired by the success of neighboring Surabaya, whose ULP and e-procurement appeared to be an attractive solution to the procurement problems in the district. They believed that the ULP would reduce the num- ber of requisite procurement-accredited staff, and that e-procurement would reduce physical interaction with bidders thus reducing corruption. Public Works Department, a quarter are women and all are trained engineers. Interviews revealed their enthusiasm for procurement and their deep understanding of procurement issues. 10 We use the terms Executive and Bupati interchangeably. 23 To implement the reforms, the team identified allies in Public Works, Agriculture, Planning and Social Affairs (all Echelon 3 (middle management)- level officials and below), and together they were trained by the LKPP in Jakarta. Subsequently, this larger group attempted to persuade their supe- riors, the SKPD heads, and solicit support from various departments and parliamentarians about the benefits of procurement reform. They faced lim- ited resistance from government officials and the close relationship between one of the civil servants and the powerful Bupati made other government officials reluctant to try and block progress. The Bupati issued an executive order to establish the ULP and LPSE in mid-2010. Unlike other districts, the ULP in Sijunjung did not take over respon- sibility for conducting procurement from the SKPDs. Instead, the ULP plays facilitating roles including scheduling procurement, announcing the ten- ders, copying and distributing tender documents, coordinating tender meet- ings, and receiving the bids from tenderers. More importantly, the ULP provides procurement-accredited staff to conduct procurement under each SKPD. Since the ULP is not authorized to conduct procurement, the local government of Sijunjung insists that this is not the same body as the one requested under Perpres No. 54/2010 on Public Procurement, and there- fore refers to it as a Procurement Service Secretariat (Sekretariat Layanan Pengadaan or SLP) and not a ULP. By contrast, in North Luwu, a small, remote, recently established dis- trict, the Executive drove the reforms and the coalition of civil servants that implemented the reforms was narrower than in Sijunjung or in Lamongan. Subsequent to his election, North Luwu’s Executive Leader started investing in IT and communications in a high-profile effort to “create North Luwu from nothing” and differentiate it from other districts. The district developed fast communication towers and Internet connections and in response to this heavy investment in IT, e-procurement was discussed as early as 2008. In 2007-08, a series of sensational stories about government corruption dominated local newspaper headlines. Against this backdrop, in 2008, the Executive Leader initiated procure- ment reform by first meeting with all SKPD heads and expressing interest in adopting Surabaya’s model of e-government as a way to combat corrup- tion. The Bupati then ordered the head of the development administration section to visit Surabaya together with him, members of the provincial leg- islative council (DPRD) and the SKPD heads. During this visit, the Execu- tive Leader signed a Memorandum of Understanding (MoU) with the Mayor 24 (Walikota) of Surabaya to support e-procurement in North Luwu. Upon his return to North Luwu, he issued a Bupati decree on e-procurement and estab- lished a ULP. Beginning in 2009, the ULP processed and used e-procurement for all procurement packages for all technical departments (SKPDs) with a value per package of IDR 100 million (USD 11,111) or higher. From 2010, the ULP administered all procurement packages, including direct appointments, using the LPSE system.11 A year later, following the launch of the national policy on LPSE, North Luwu migrated to the national LPSE system, and started using it to handle bidding for any type of project with a value of at least IDR 100 million (USD 11,100). Similar to the Mayors of Balikpapan and Madiun, the Executive tried to mitigate the local private sector’s resistance to e-procurement. The Executive confronted this resistance through open dialogue with the business associations promising training for e-procurement. The Bupati strategically communicated to businesses that by restricting training to local business, the district was conferring upon them an advantage. The local government also provided local vendors with on-line hot spots to submit tenders. The Bu- pati’s dialogue with local businesses and provision of training and hot spots preempted or at least lessened private sectors’ resistance to e-procurement and procurement reform, more broadly. In December 2009, the Bupati resigned from his post and returned to a position in the central government. Some in the successor government have publicly questioned the value of the e-procurement system while others are increasingly vocal about their objections to new practices. Limited high level support, combined with antagonism from mid-level executing agency officials and contractors, has stalled the progress of reform. To summarize, our case studies did not find that the executive played a critical role in either blocking or facilitating procurement reform. Rather, a critical factor for the adoption and sustainability for procurement reforms in Sijunjung was the presence of a leadership team consisting of a unique combi- nation of skill sets and positions. The reform team consisted of members who were tech-savvy, mid-career civil servants who were able to build coalitions to help gain support for the reform within the bureaucracy and a senior-level civil servant who had influence over the executive. The Executive’s support for the reform in Sijunjung was a necessary but insufficient factor as we saw from the case of North Luwu, where the Executive’s reforms were undone by 11 Each SKPD is responsible for making direct appointments decisions. 25 his successor. This finding highlights the need for a better understanding of leader- ship and coalition building within the civil service and how coalitions can contribute to reforms. Most of the academic research on coalitions is math- ematical and abstract, or it focuses on the politics of developed countries with well-organized parties, lobbies, and interest groups (Levi, 2011). There is very little academic literature on coalition building within the civil ser- vice and in developing countries (Humphreys, 2008; Maravall, 2010; Riker, 1962). Coalitions face a myriad of obstacles that undermine their formation and durability including: difficulties in bargaining enforcement, free rider problems and the existence of multiple potential sets of possible coalitional partners (Levi, 2011). The election of a new politician or the appointment of a civil servant may result in renegotiation among the coalition partners, which can potentially undermine the coalition. The case studies of Sijun- jung and Lamongan raise several questions including how the coalition in the former district was able to overcome a number of obstacles while the efforts of the coalition in Lamongan proved to be unsustainable. How were the commitments of the coalition members to advocate for and implement procurement reform made? How did the coalition members enforce these commitments? What conditions helped to keep the coalitions together over a long enough time for the reform to be implemented? In pursuit of procure- ment reform, what kinds of bargains did the coalition members make with one another and with politicians and members of the private sector? 7.2 Private Sector Competitiveness (Lamongan vs. Ma- diun) In localities that are economically competitive and where business associa- tions represent a variety of types of firms, we expected to observe business associations pressing for procurement reform. We also expected business associations to provide welfare-enhancing services to its members including information about new technology. By comparison, in localities where the market is thin and dominated by a small number of large indigenous firms, business associations are unlikely to try and block reform. Our findings pro- vide support for the latter but not for the former hypothesis. Lamongan is a large district, with a population of 1.4 million and agri- culture, fisheries and home industries dominate its economic base. In the 26 past decade, the construction sector has grown due to the expansion of privately-funded industrial and business projects, and the development of public infrastructure including roads, bridges, irrigation, the port, airport and tourism-related projects. As we learned from the case study, the presence of “premanistic”12 businesses with strong links to corrupt practices in collu- sion with elected representatives of various parties and government officials have undermined the competitiveness of the private sector and preempted demands for procurement reform. In fact, with the backing of the Bupati, the presence of a coalition of three civil servants who advocated for and implemented reform was ulti- mately not strong enough to overcome the private sector and politicians’ resistance to reform. The coalition initiated procurement reform during the Mayor’s second term for three reasons. First, there was a limited number of procurement-accredited staff in the LG. As of 2008, there was fewer than 50 accredited staff. Second, there were growing concerns that corruption among vendors and political interference was undermining the competitiveness and transparency of procurement. Third, the team was inspired by the success of neighboring Surabaya, whose ULP and e-procurement appeared to be an at- tractive solution to the procurement problems in the district. They believed that the ULP would reduce the number of requisite procurement-accredited staff, and that e-procurement would reduce physical interaction with bidders thus reducing corruption. During his second term in 2010, the Bupati issued an executive order to establish a ULP and LPSE. Under the Bupati’s successor, the three team members that drove procurement reform have been transferred to different positions that do not involve procurement-related tasks. The civil servant who was assigned to head the ULP denied that the ULP even exists in La- mongan. He also claimed that the district is no longer using e-procurement because the Presidential Instruction does not require the LPSE to be oper- ational until 2014. There are also allegations of political interference and corruption in the procurement process. Parliamentarians from PKB, PDIP, PAN and Golkar openly admit to having connections with contractors (or are contractors themselves), and to using their government position to en- sure certain vendors win government business (The Asia Foundation, 2011, 6) Rather than resisting procurement reform, the private sector in Madiun 12 Preman is a term for a gangster who uses violence and intimidation for extortion. 27 is using procurement reform to advance their particularistic interests. Ma- diun’s government initiated procurement reform in 2011 to comply with the national directive requiring e-procurement to be but in place by 2012. Ma- diun’s Development Administration unit selected procurement-accredited of- ficials from across different government units to be trained by the LKPP, and installed the LKPP’s e-procurement software in the IT system of the Local Development Planning Board (Bappeda). The government then dis- seminated information about e-procurement to all government implementing units, as well as to eight professional units whose members are vendors that bid on public projects. Local contractors used the launch of the LPSE as an opportunity to raise their concerns about what they view as unfair competition from large contractors from outside the district, who already have experience with using e-procurement in their home districts. The main concern of local contractors is the growing number of awards to competitors from Surabaya and other parts of Central and East Java, particularly for large-value projects (> IDR 1 billion/USD 110,000). Currently, business owners are lobbying the local government for mea- sures that would give preference to local vendors. For example, some private sector actors suggested adding selection criteria that would give more weight to local vendors, particularly on bids valued at below IDR 1 billion (USD 110,000). If vendors are local, they argue, the government will be able to monitor contractor performance more easily. The local government appears sympathetic to these arguments, and is considering introducing protectionist measures. In none of the other cases did we find that the private sector played a positive role in advocating for procurement reform. Of our six cases, Ba- likpapan has the most vibrant economy and the lowest official poverty rate (2.5%).13 The presence of lucrative private sector opportunities makes par- ticipation in public sector activities a less attractive option for local firms. Given the vibrancy of the construction industry in Balikpapan, the opportu- nity costs for large, local firms to participate in public procurement are high. A local Chamber of Commerce (Kadin) official pointed out that over 1,500 construction-related contractors are registered in the municipality, while only 300 public works projects are tendered annually. Because the private sector 13 The major industry is the service sector, which supports the large-scale oil, gas, palm oil, coal, and other mining operations in the districts neighboring the municipality. 28 has few stakes in the outcome of procurement reform, they neither resisted nor advocated for reform. Informants, NGOs, the business community, the media and the government agreed that with the advent of e-procurement, there has been little resistance to procurement reform and collusion and cor- ruption in public procurement has disappeared. 8 Other Findings from the Case Studies 8.1 CSO Voice There is very little evidence from the case study narratives to suggest that procurement reform has been “demand” led since CSO’s have not been very active in advocating for procurement reform. Yet, CSO’s as in Balikpapan and Madiun, are beginning to experiment with different ways to engage on procurement primarily through monitoring contract performance. These ef- forts have demonstrated the proof of concept — that CSOs can undertake monitoring, uncover problems, and find ways to bring those problems to the notice of government officials in ways that have effect. These is currently lit- tle sense of how CSO can scale up their procurement monitoring efforts and collaborate with others actors (public and private) in ways that will enhance procurement outcomes. A large literature emphasizes that local elections and decentralization empower civil societies to demand better government performance (Adsera, Boix and Payne, 2003). In some contexts, CSOs play an important role in advocating for and influencing government reforms. In the Philippines, for example, Procurement Watch Inc. (PWI) played a central role in advocating for the passage of Philippines’ procurement reform (Campos and Syquia, 2005). The ability of CSOs to influence policymaking depends on a variety of factors including the informational and monetary resources a movement can obtain and employ (Kitschelt, 1986). In contexts, where CSOs have both the requisite resources and access to the political sphere, they are likely to be able to influence procurement reform. In districts that have a vibrant CSO sector, we expected to observe CSOs advocating for local governments to adopt procurement reform as part of a broader effort to push for greater fiscal transparency. Only in Sijunjung did we observe a CSO indirectly contributing to the lo- cal government’s uptake of procurement reform. Although CSOs in Sijunjung 29 are quite limited14 , we identified one active NGO in Sijunjung, the Institute of Independent Monitoring and Community Empowerment (LP3I)15 , which indirectly helped to catalyze procurement reform efforts from the Bupati. LP3I has five staff, as well as village-level volunteers, who monitor the im- plementation of government projects, collect community complaints about public services, and report their findings to the responsible SKPD. If it is not satisfied with the SKPD’s response, LP3I helps communities to lodge formal complaints. In 2008, while monitoring the tender of a local hospi- tal, LP3I raised concerns that the tender process was unfair and a contract was awarded on the basis of political connections. We were told that these allegations prompted the Bupati to take steps towards procurement reform. In addition to influencing the uptake of procurement reform, increasingly, CSOs in Indonesia and elsewhere are experimenting with third party mon- itoring of contracts as a means to improve procurement performance and service delivery. In some contexts, CSOs monitor the tendering process but more commonly, CSOs are involved in grassroots monitoring of procurement contracts (Olken, 2007).16 In Madiun and Balikpapan, while CSOs did not play a role in the uptake of procurement reform, they are actively monitoring procurement contracts and thus, contributing to the implementation of procurement reform. We, however, do not have any evidence as to whether these or monitoring ef- 14 There are no independent local newspapers, and the local correspondents for the three provincial level newspapers — Singgalang, Padang Express and Haluan — are government officials who work for the Inforkom office. Of the three local radio stations, the local gov- ernment owns two and a local vocational school owns the third. None of these media outlets monitor local government performance. A case involving the persecution of collu- sion among a contractor, the Head of the Irrigation Unit and the Head of the Procurement Committee relied on evidence an NGO, Wahana Komunikasi Rakyat (WKR) compiled. 15 LP3I, which was established in 2001, is part of the Regional Poverty Reduction Co- ordination Team (TKPKD). 16 We know little about how CSOs and communities are able to overcome several re- lated challenges to monitor procurement performance. For instance, accessing detailed information about contracts and developing the skills to be able to interpret and utilize this information is challenging. The evidence on the effectiveness of community-based monitoring of publicly procured service delivery is mixed. In his study, Olken (2007) ex- amines the impact of audit and grassroots monitoring of roads in Indonesia found that the expectation of external government audits was more effective than grassroots monitoring at preventing corruption in road construction. Grassroots monitoring reduced the theft of wages but not the theft of project materials. Devarajan, Khemani and Walton (March 2011) review other studies on grassroots monitoring). 30 forts in other districts are contributing to better procurement performance. In Madiun, the NGO, WKR, which was established in 2009 before the dis- trict initiated procurement reform in 2011, is focused on monitoring public projects, particularly infrastructure. With a staff that includes two engi- neers, as well as personnel with backgrounds in law, journalism and IT, in the past two years, WKR has monitored a number of public works projects. WKR’s approach has been to use its ties to local media to pressure con- tractors and/or the government to fix any problems it discovered before the project is completed. Moving forward, WKR plans to only release its find- ings following the completion of a project and then use evidence of poor practices to prosecute or blacklist and punish poorly performing contractors and officials. During the rollout of e-procurement reform in Balikpapan in 2007, Trans- parency International Indonesia (TII) supported the development of a local umbrella CSO, “Community Forum for Transparency” (FORMAT). After having received training from TI Indonesia and Indonesia Corruption Watch (ICW), FORMAT has begun to monitor eleven high-value procurement con- tracts. FORMAT selected these projects from a longer list of procurement projects they obtained from the ULP.17 FORMAT mobilizes volunteers from communities that are located near the project site to monitor the work and to submit bi-weekly reports designed to flag concerns. These volunteers analyze the contract, discuss the technical specifications with external contractors to ensure they understand the technical specifications of the tender, and identify indicators to best measure compliance with the contract. If the volunteers discover problems, an investigation team is then sent in to collect evidence of the infraction. Before releasing their final report to the public, they allow the government to comment on their findings. 8.2 Inter-District Diffusion of Knowledge and Compe- tition The literature on new institutionalism suggests that the head of organizations tend to adopt models from similar organizations if they believe that the orga- 17 Of the 11 projects selected, the team has only obtained technical specifications for the Public Works projects, despite making written requests to the Departments of Education and Health and the Mayor. They hope to use the Freedom of Information Act (FOIA), but a Provincial Information Commission has yet to be established. 31 nization is likely to enhance its organization’s legitimacy (Meyer and Rowan, 1977). Consequently, this form of normative isomorphism leads organiza- tions to adopt surprisingly homogenous organizational forms (Dimaggio and Powell, 1983; DiMaggio and Powell, 1991). In Indonesia, we observed that early adopters of e-procurement modeled their e-procurement systems after Surabaya’s as part of an effort to bolster their district’s prestige. Officials from Makassar, North Luwu and Balikpapan each reported that visits to Surabaya and subsequent low-cost technical assistance for e-procurement from Surabaya was central to their successful initiation of procurement re- form. Balikpapan’s interest in procurement reform first emerged when a number of mid-level Balikpapan officials visited Surabaya in the 2007-08 pe- riod and were impressed with the Surabaya’s nascent model of e-procurement. For a one-time royalty fee of IDR 10,000,000 (USD 1,100), Surabaya provided Balikpapan with software and technical support and trained Balikpapan of- ficials in how to use the software. Officials in Sijunjung, Lamongan and Madiun were also initially motivated by the example of Surabaya to imple- ment e-procurement. However, for various reasons they did not utilize its technical assistance. In many of the cases, an additional factor that con- tributed to the adoption of procurement reform was a desire on the part of the Executive to try and make his district well known throughout Indonesia. In Sijunjung, the Executive strove for his district to be among one of the first to implement e-procurement reform as he saw early adoption as a source of pride. Lamongan’s Bupati’s vision of trying to make Lamongan famous was consistent with the efforts of his civil servants to implement reform. More importantly, LKPP’s offer to help develop Lamongan into a regional hub for procurement training in East Java offered the Executive a chance to make his district shine among other districts in East Java. From the outset, North Luwu’s Executive was vocal about his ambitions to make North Luwu a model for successful decentralized governance. In 2003, the Mayor received his first leadership award from the State Ministry for State Administration and Bureaucracy Reform (Menpan), followed by numerous other recognitions and awards in subsequent years. The Bupati also started investing in IT and communications in a high-profile effort to “create North Luwu from nothing” and differentiate it from other districts. Communication towers and faster Internet connections were developed in the district, and in response to this heavy investment in IT, the Bupati started to explore e-procurement as early as 2008. Subsequent to a visit to Surabaya, the Bupati issued a decree to establish a ULP and adopted e-procurement. 32 In Makassar, the Mayor tried to bring visibility to his district by modi- fying the central government’s procurement policies and procedures. For ex- ample, while Makassar’s current e-procurement system comes straight from the LKPP, in 2012 they plan to revise the standard operating procedures (SOP). Makassar has not yet established a ULP claiming that the LKPP’s “off the shelf” ULP structure is too complex and needs to be adapted in order to work in the local context of Makassar.18 9 Conclusion The research discussed in this paper is a first step to create a rigorous ana- lytical approach to examining implementation of procurement reform. The overall decline in the volume of procurement at the local level raises several questions and challenges a national objective of improving infrastructure. As discussed in Section 5, since 2007, there has been a decrease in average annual procurement spending at the district level. From 2007 to 2011, real procurement expenditure (IDR Billion, 2011 price) has declined in 65 per- cent of districts. During the same time period, there was also a decrease in the proportion of district spending that occurred via procurement from 50% of expenditures in 2007 to 41% in 2011. The clusters of districts that have experienced a decline in procurement expenditures over time point to grow- ing disparities in procurement performance between rich and poor districts. The decline of procurement expenditures in poorer districts is of particu- lar concern as poorer districts have a greater need for public services and commercial opportunities afforded by public procurement. As this paper shows, there is a significant amount of variation in the uptake and implementation of procurement reform in Indonesia. Data on the uptake of procurement reform across Indonesia point to a strong association between the wealth of a district and the adoption of procurement reform. The geographic clusters of districts that have adopted procurement reform suggest that procurement reform is lagging in poorer districts. Further, across local governments we find a strong correlation between the uptake of one type of reform and the uptake of another type of reform. For example, the number 18 The Municipality of Makassar is also proud of its self-designed ombudsman program. When a Presidential Instruction required all cities to implement a centrally designed om- budsman system, Makassar appealed to the Constitutional Court and won the right to retain its own system. 33 of procurement-accredited staff is higher in districts that have adopted e- procurement. ULPs are more likely to be established in districts that have adopted full or semi e-procurement. To summarize, the strong association between the wealth of a district, its level of procurement expenditure and the uptake of reform raises a concern about whether different approaches to procurement may exacerbate disparities in development opportunities across Indonesia. Further, as demonstrated by the high degree of the incompleteness of pro- curement reform across space, our analysis shows that the implementation of procurement reform should not be viewed in binary terms, or in “yes” or “no” terms. As of 2011, whereas only about one-fifth of districts have adopted the LPSE system, a little over half of districts use a semi e-procurement system. The utilization of e-procurement is still limited as only 11 percent of the total district’s procurement budget was e-procured. Across our case stud- ies, there is also variation in the role and functions the ULPs are carrying out.19 In Sjinjung, for example, the ULP is not actually conducting procure- ment transactions; rather, it is merely facilitating procurement transactions across the local government’s SKPDs. By comparison, the ULPs in Balikpa- pan, North Luwu and Lamongan were established to conduct procurement of goods, services and works on behalf of the local government’s SKPDs. Our analysis suggests that the wealth of a district and local government leadership are associated with the uptake of reform. There are three policy implications that emerge from these findings. First, in order to find ways to increase procurement volumes in districts that have experienced a steep de- cline in expenditures over time, it is important to determine the factors that influence decisions regarding how much money to dedicate to procurement. Second, in the short term, there is a need to focus on approaches to improv- ing procurement in poorer districts and those where effective leadership does not exist. These districts may require the assistance of provinces to help with procurement, as well as help with identifying alternative ways of developing local-level infrastructure. LKPP could also play a more active role in ca- pacity building by benchmarking practices and outcomes across sub-national entities. Third, our analysis of the implementation of procurement reform has po- 19 In this paper, we are only examining the status of local governments implementation of procurement reform. As such, we describe the variation in the functions ULPs are carrying out. We are not comparing the effectiveness of the different administrative models. 34 tential implications for multi-stakeholder strategies to improve procurement performance. For example, our analysis of the six case studies indicates the interests of a portion of the private sector to maintain existing procurement practices. Some members of the private have concerns that procurement re- form will lead to greater transparency and competitiveness that may under- mine their ability to win tenders. While ultimately unsuccessful in blocking the reforms, members of the private sector in Sjinjung, Madiun, Makassar and North Luwu, used the vehicle of contractors’ associations to try and block the uptake and implementation of procurement reform. This finding should be taken into consideration multi-stakeholder initiatives, such as the Con- struction Sector Transparency Initiative (CoST), that sometimes presuppose active support from the private sector. The decrease in procurement expenditures that has taken place across the majority of Indonesia’s districts across time is worrying and raises several questions about the causes of the decline.20 In future research, we plan to examine the political economy dynamics behind the composition of local government expenditures including procurement expenditures and address several questions including the following. How much variation exists across space and time in the composition of local government expenditures? What are the trade-offs to a local government’s decision to allocate a greater or less of a share of its budget for procurement versus administrative/civil service or social safety net expenses? To what extent do electoral incentives and the threat of central government oversight including audits explain variation in the composition of local governments’ expenditure over time and space? In addition to examining the factors that influence decision-making on the level of procurement expenditures, in future work, we plan to examine other factors that might be associated with the uptake of reform including the tim- ing of the central government’s release of budgets to local governments, the strength of auditing and the sectoral distribution of expenditures. There is also an additional need to examine variables that are likely to affect procure- 20 There are several potential explanations for the relatively low level of capital invest- ments. For example, central block grants are calculated based on the size of the local government apparatus. This formula creates incentives for local administrators to increase the size of their bureaucracy and disincentives to spend their funds on the procurement of goods, services and works. Central government expenditures on capital investments at the local level tend to crowd out local government expenditures (World Bank, August 2012). Neither of these explains the variation in the composition of local government expenditures or the decline in expenditures on capital investments. 35 ment performance including CSO monitoring and the size of contracts. The focus of our on-going and planned follow-up work is on developing useful ways to measure procurement performance. Developing measures of procurement performance is essential to be able to understand the relationship between the uptake of reform and improved outcomes — a topic discussed in a com- panion piece by Turkewitz et al. 2013. In the companion piece, we identify a set of performance-based indicators of procurement and examine the re- lationship between the uptake and implementation of procurement reform and subsequent performance. Specifically, across the six case studies, the companion piece examines indicators of the level of competition, competence and capacity, transparency of procurement-related information, efficiency, and compliance. 36 References Adsera, Alicia, Carles Boix and Mark Payne. 2003. “Are you being served? Political accountability and quality of government.” Journal of Law, Eco- nomics, and organization 19(2):445–490. Ahlquist, John and Margaret Levi. 2011. “Leadership: What it means, what it does, and what we want to know about it.” Annual Review of Political Science 14:1–24. Arrow, K.J. 1970. Social choice and individual values. Vol. 12 Yale university press. Buehler, Michael. 2012. “Public Procurement Reform in Indonesia: The His- torical Institutional Context and Lessons Learned from Analytical Work.” Background Paper for the World Bank project, Strengthening Reform and Improving Public Procurement Performance in the Africa and East Asia Pacific Regions. Washington D.C., World Bank. Calvert, Randall. 1992. “Leadership and its basis in problems of social coor- dination.” International Political Science Review 13(1):7–24. Campos, Jose Edgardo and Joaquin Gonzalez. 1999. “Deliberation coun- cils, government-business-citizen partnerships, and public policy-making: Cases from Singapore, Malaysia, and Canada.” Asia Pacific Journal of Management 16(3):429–448. Campos, Jose Edgardo and Jose Luis Syquia. 2005. Managing the politics of reform: overhauling the legal infrastructure of public procurement in the Philippines. Number 70 Washington D.C.: World Bank Publications. Cook, Karen S., Russell Hardin and Margaret Levi. 2005. Cooperation with- out trust? New York: Russell Sage Foundation. Devarajan, Shantayanan, Stuti Khemani and Michael Walton. March 2011. Civil society, public action and accountability in Africa. Technical report Paper Policy Research Working Paper No. 5733, The World Bank Wash- ington, D.C.: . Dimaggio, Paul and Walter Powell. 1983. “The Iron Cage Revisited: Institu- tional Isomorphism and Collective Rationality in Organizational Fields.” American Sociological Review 48(2):147–160. 37 DiMaggio, Paul and Walter W. Powell. 1991. Introduction. In The New Insti- tutionalism in Organizational Analysis, ed. Paul DiMaggio and Walter W. Powell. Chicago and London: University of Chicago Press. Doner, Richard F. and Ben Ross Schneider. 2000. “Business associations and economic development: why some associations contribute more than others.” Business and Politics 2(3):261–288. Fligstein, Neil. 1997. “Social skill and institutional theory .” American Be- havioral Scientist 40(4):397405. Gatra. December 26, 2007. “Lembaga baru pemantau tender: pemerin- tah membentuk Lembaga Kebijakan Pengadaan Barang dan Jasa Publik.” Gatra . Grindle, Meri. 2007. “Good enough governance revisited.” Development pol- icy review 25(5):533–574. Grindle, M.S. 2004. “Good enough governance: poverty reduction and reform in developing countries.” Governance 17(4):525–548. Humphreys, Macartan. 2008. “Coalitions.” Annu. Rev. Polit. Sci. 11:351– 386. Kitschelt, Herbert. 1986. “Political Opportunity Structures and Political Protest.” British Journal of Politics 16(1):57–85. Levi, Margaret. 2006. “Why We Need A New Theory of Government.” Per- spectives on Politics 4:5–19. Levi, Margaret. 2011. “Coalitions.” Memos Prepared for the World Bank Institute of the World Bank. Luebke, Christian Von, Neil McCulloch and Arianto Patunru. 2009. “Het- erodox Reform Symbioses: The Political Economy of Investment Climate Reforms in Solo, Indonesia*.” Asian Economic Journal 23(3):269–296. Maravall, Jose Maria. 2010. “Accountability in coalition governments.” An- nual Review of Political Science 13:81–100. 38 Mazaheri, Nimah, Edouard Al-Dahdah, Sandeep Poundrik and Soujanya Chodavarapu. 2013. “Leadership and Institutional Change in the Public Provision of Transportation Infrastructure: An Analysis of India’s Bihar.” The Journal of Development Studies 49(1):19–35. Meyer, John W. and Brian Rowan. 1977. “Institutionalized Organizations: Formal structure as Myth and Ceremony.” American Journal of Sociology 83(2):340–363. MIller, Gary. 1992. Managerial Dilemmas. New York: Cambrid. Olken, Benjamin. 2007. “Monitoring Corruption: Evidence from a Field Experiment in Indonesia.” Journal of Political Economy 115(2):200–249. Olson, Mancur. 1982. The rise and decline of nations: Economic growth, stagflation, and social rigidities. Yale University Press. Pyle, William. 2006. “Collective action and post-communist enterprise: The economic logic of Russia’s business associations.” Europe-Asia Studies 58(4):491–521. Riker, William. 1962. The theory of political coalitions. Vol. 578 Yale Uni- versity Press New Haven. Skoufias, Emmanuel, Ambar Narayan, Basah Dasgupta and Kai Kaiser. 2011. “Electoral Accountability, Fiscal Decentralization and Service Delivery in Indonesia.” World Bank Polic 5614. The Asia Foundation. 2011. “Draft First Batch of Case Studies of Procure- ment Reform at Local Level.” 2011. The Asia Foundation. 2012. “Case Studies of Procurement Practices at the sub-National Level in Indonesia.” Background Paper for the World Bank project, Strengthening Reform and Improving Public Procurement Per- formance in the Africa and East Asia Pacific Regions. Washington D.C., World Bank. von Luebke, Christian. 2009. “The political economy of local governance: findings from an Indonesian field study.” Bulletin of Indonesian Economic Studies 45(2):201–230. 39 Weber, Max. 1968. Economy and Society. Berkeley: University of California Press. World Bank. 2007. Spending for Development: Making the Most of Indone- sias New Opportunities. Jakarta: World Bank. World Bank. August 2012. Indonesia Sub-National Public Expenditure Re- view: Optimizing Subnational Performance for Better Services and Faster Growth. Jakarta, Indonesia: . World Development Report. 2002. Building Institutions for Markets. New York: Oxford Univeristy Press. 40