56339 Rural Investment Climate Assessment ! rs () i/ f ii \/'! , · 1_, : j '! ~-' Best Available Copy Jakarta Stock Exchange Building Tower II/12th FI. JI. Jend. Sudirman Kav. 52-53 Jakarta 12910 Tel: (6221) 5299-3000 Fax: (6221) 5299-3111 Website: www.worldbank.or.id THE VVORLD B/-\NK 1818 H Street N.W. Washington, D.C. 20433, U.sA Tel: (202) 458-1876 Fax: (202) 522-1557/1560 Email: feedback@worldbank.org Website: www.worldbank.org Printed in 2006. This volume is a product of staffofthe World Bank. Thefindings, interpretations, and conclusions expressed herein do not necessarily refiect the views of the Board of Executive Directors of the World Bank or the governments they represent. The World Bank does not guarantee the accuracy of the data included in this work. The boundaries, colors, denominations, and other information shown on any map in this work do not imply any judgment on the part of the World Bank concerning the legal status of any territory or the endorsement of acceptance of such boundaries. Rural Investment Climate Assessment Website: www.worldbank.org/id/rica Acknowledgements This case was written by Christian von Luebke and ·IS part of the Rural Investment Climate Assessment Case Studies. The author would like to thank the people of Klaten, Kebumen, Solok, and Pesisir Selatan for their kind assistance in this work. The Rural Investment Climate Assessment Case Studies were prepared by a team led by Stefan Nachuk, and consisting of Novi Anggraini, Novia Cici Anggraini, Jennifer Donohoe, Janes Imanuel Ginting, Julie Hackett, Gregorius Kelik Agus Endarso, Kyle Lemargie, Agni Paramita, Peter Rooney, and NunikYunarti. ., The Rural Investment Climate Assessment Case Study team extends its deep thanks to the people of the five localities where the case study research was undertaken: Serang, Malang, ~anggarai, and Tegal districts, and Wiradesa, sub·district, Pekalongan. Without their thoughtful ideas and generosity, these studies would not have been possible. The Rural Investment Climate Case Studies are part of a project led by Neil McCulloch, the Indonesian Rural Investment Climate Assessment or RICA. The RICA project is one output of World Bank's Indonesia Poverty Team (lNDOPOV) led by Jehan Arulpragasam. Overall guidance was provided by William Wallace, Lead Economist for Indonesia, Rahul Raturi, Rural Development Sector Manager for East Asia, Shobha Shetty, Senior Economist Rural Development and Andrew Steer, World Bank Country Director for Indonesia. Project management support was provided Roksana Khan and Juliana Wilson. Logistical,editorial, and production support was provided by Maulina Cahyaningrum, Juliana Wilson, Una Marliani, Indra Irnawan, Niltha Mathias, and Deviana Djalil. Financial support from the World Bank-Netherlands Trust Fund for Institutional Development and Capacity Building as well as from the DFID Poverty Reduction Partnership Trust Fund is gratefully acknowledged. :-~"'~I': ,,' ~~ol~'.;m,' <~: ~·x~: '~\J,;, ,';<0 f;c';'~r;)',:l'-,', o."d .. ,\~, 011 :,::.3' ;2fo;'0,1 ~"1J; ""~,,,,"":: "r~<['(2' ;':lr',. ~:1.;r ~,:,t:<'.t C.:,., "-, ~'~"I'd .'ci ... ,-);-j '8:::;,'_ :lr:l,"c:i~ The Rural Investment Climate Assessment (RICA) project was funded by the World Bank-Netherlands Trust Fund for Institutional Development and Capacity Building as well as the DFID Poverty Reduction Partnership Trust Fund. The project is designed to provide a better understanding of the rural investment climate and its linkages to rural poverty reduction. The initial phase of the RICA project has four components: the Rural Investment Climate Survey, National Conference on the Rural Investment Climate, and five case studies that complement the quantitative survey with qualitative research. The case studies were designed to increase understanding of the policy and institutional setting and how it infiuences the failure or success in implementation and outcomes. The RICA Case Studies were chosen around five themes where knowledge gaps exist or where good examples of innovation have been found. The knowledge gained from these case studies will be incorporated into the RICA Report. For more information on the RICA project, please refer to www.worldbank.org/id/rica . . \' ~ i . "B. KAB. MANGGARAI - . ,'-,r , Case Study 6: Political Eoonomy of Local Business Regulations fL" ,I" . ,- · d- , . . '" .. " ,. ,-HlDlngs on iDem iax{D'tuon em : ucensmg r'mCl.u::es 'from FOUl District Cases in Centra! jalla and West Sumatera Abstract: ., Empirical evidence to date suggests that business climates in Indonesia's districts have hardly improved since the rapid enactment of decentralization. Based on four district cases in Central Java and West Sumatera, our empirical findings confirm the critical state of government-induced business conditions. Three districts have seen rising administrative capture practices, unsatisfactory one-stop licensing services, and a loc al revenues management that follows vague 'target systems' rather than sound accounting standards. The alarming state of government- related business conditions is strongly related to institutional deficiencies, both on the national and district level; on the national level ambiguity of national tax laws and restricted supervision creates room for misconduct of local governments; on the district level inverted incentive structures and adverse bureaucratic conventions negatively affect administrative practices. Yet, despite institutional impediments, some districts appear to be more conducive for local business than others. The cross-district comparison indicates that variations of more or less business-friendly tax and licensing practices are strongly affected by the quality of district leadership and to a lesser extent by civic pressures. In fact, the explanatory strength of civic 'voice' and 'exit' functions - as theoretically advocated in the Fiscal Federalism and Public Choice literature - appears limited due to Indonesia's infant democracy and socia-cultural features. The author is grateful forva!uable suggestions and supportorAndrew Macintyre, Pe~e( Timmer, Yusaku Horiuchi, Runako Samota, Taco Bottemo, Sebastian Eckardt, Neil McCulloch, Hadi Susastro,Agung Prambudi, Raemond Aye, SyarifHidayat, Thee Kian Wie, /wan Azis, Raksaka Mahi, Suka- mdi, Blane Lewis, Peter Larmour, Hal Hill, Chris Manning, and Ross McLeod; and indebted to local students in Klaten, Kebumen, Solok and Pesisir for their great assistance in administrating business surveys. Decentralising governance' has becorne a worldwide phenomenon. More than 80 percent of all developing and transitional countries with populations of more than 5 million have started to reshape their governance structures by transferring power to sub-national government levels (Dillinger 1994:302). Among these countries, Indonesia's case represents a remarkable example of substantial government reforms in terms of scope and speed. After decades of authoritarian rule, the Indonesian government implemented administrative, political and fiscal decentralization reforms in 2001 after a brief preparation period of 18 months. Formerly concentrated in Jakarta, Indonesia has strongly dispersed its governance system to more than 440 district heads, who are responsible to autonomously manage local economies in terms of infrastructure, agriculture, education, industry and trade, investment, development planning, and manpower affairs. Undoubtedly, this'Big-Bang'transition (World Bank2003) of rapid administrative reforms and extensive democratic changes' has its costs. While theory suggests that shifting government closer to people and enhancing democratic rights will improve government performance,' Indonesia's evidence to date indicates that costs may initially outweigh benefits for district economies. Local business surveys report policy uncertainties, examples of new local taxes distorting interregional trade, and predatory bribe practices in license administration. In most districts private sector respondents perceive regulatory conditions either same or worse compared to the'Ordre Baru' era (KPPOD 2005, LPEM-FEUI 2002; REDI2003; Suhirman 2002; own interviews and surveys). One economic indication for the critical state of district business climates is the stagnant Aow of domestic and foreign investment. Once an 'Asian Tiger' and part ofthe'East Asian Miracle' (World Bank 1993), side-by-side with Thailand and Malaysia, Indonesia has fallen behind in terms of investment and economic growth. Although the Indonesian government has successfully improved macro-economic stability since the crisis, including sound inAation and interest rates (Hill and Aswicahyono 2004), investments remain stagnant and GOP forecasts hardly exceed 4 percent (World Bank 2004, IMF 2003, Castle 2004). The fact that foreign and domestic businesses are reluctant to invest, despite sound macro-economic conditions and promising resource endowments, signals governance-related impediments that - given the decentralized setting - are no longer concentrated in Jakarta only but dispersed to district bureaucracies throughout the country. Following Kaufman et 01. (1999), governance is defined as "the traditions and institutions by which authority in Q country is exercised - including the process by which governments are selected, monitored, and replaced, the capacity of the government to effectively formulate and implement sound policies, and the respect of citizens and the state for the institutions that govern economic and social interactions among them. Accord- h ingly, local or district governance refers to the same definitional understanding but is limited to actors and their interaction within a district's boundaries. 2 During the fast two years, for the first time in history, government heads and parliaments at various government levels have been elected directly by Indonesian people. Although many challenges lie ahead, national presidential elections and district-level elections have set an important democratic landmark ~ especially because they took place without major conflicts and violence (AspinaIl2005). 3 Some PubliC Choice and Fiscal Federalism arguments are discussed in the foHowing section. For a more extensive theoretical discussion on the benel1ts of decentralization refer to Azfar and Kahkonen (1999). Yet, empirical evidencetodate also ind',cates variation in the degree of business obstructions since the enactment of decentralization. Apparent in the newest nation-wide ran kings of district investment attractiveness, some districts exhibit more favourable administrative practises than others. According to the Regional AutonomyWatch (KPPOD 2003, 2004, 2005), a Jakarta-based economic think tank, whether local governments release distortionary regulations and extract rents through predatory license procedures varies significantly throughout the country. Several questions arise in this context, which will be addressed consecutively in this paper: (1) To what extent does the ernpirical case evidence from West Sumatera and Central Java affirm the increase of predatory licensing practices and distortionary taxation? ., (2) Which political econorny factors - based on the evidence of four district cases in West Surnatera and Central Java - help to explain the critical tendency of deteriorating rather than improving business climates in a setting of decentralised governance? (3) What explains the variance among districts; does government leadership, given by the will and power of a district head ('Supali'), or civic pressure, given by the degree to which citizens organize themselves to voice or utilize their exit options, have significant bearing on district business climates? 2. /Evidence on District Business Climates The Indonesian decentralization experience to date does little affirm the theoretically suggested scenarios of more favourable business environments. On the contrary, empirical evidence provides by and large a critical picture on how district governments serve local businesses. According to the Regional Autonomy Watch (KPPOD 2004, 2005), many of Indonesia's 440 district governments have produced unsatisfactory business regulations and continue to obstruct the private sector with predatory licence practices. Lewis affirms that"many of the new regional taxes and charges have been judged as inappropriate by one standard or another"(2003:16) and outlines that the central government cancelled 28 percent of assessed district regulations in the first decentralization year. In its Public Expenditure Review of 2003, the World Sank concludes that many new local taxes "are seen to be damaging the business environment"(2003: 30); and that"business undoubtedlyfeltthe pinch of decentralisation" in terms of the"plethora oftaxes, levies and fees that were issued by the regions under the new law on regional taxes" (2003: 21). - --. , '- "--~,.-- Figure 1: Location of the Four District Cases Our field research in four district cases in Central Java (Klaten and Kebumen) and West Sumatera (Solok and Pesisir Selatan Pesisir Selatan) - based Kebumen Klaten on a mixed-method o 1000 ,... 2DOO 3000 kilomders approach business including a survey with more than 500 local respondents and 60 in-depth stakeholder interviews - provides a similarly critical account. It shows that the majority of respondents are indifferent or disappointed with the general conditions since the enactment of decentralization. Asked on their overall perception to what extent district government policies have affected their activities since 2001, 294 businesspeople perceive no change (57 percent), 131 perceive deteriorating conditions (26 percent), and merely 88 (17 percent) feel that business conditions have improved (See Figure 2). Figure 2: Infiuence of new district policies infiuence on business activities since 2001 [.-------_...-._.-._-.-----------------_._... _._-_.----_....-- ------ .-.... __.... __ .. ! ::5 Perception Of District Policies Since 2001 I ~ ! I i J~ "l '" o'-__________~===r==========~~r=============~=======cJ L ___._____.___ ___________~_____ ~______~______~_~__ "._.___J Note: 1= bbstructive' 2= 'no change' 3= 'supportive:· Source: Survey data based on 513 randomly selected business respondents in (our districts Regarding different governance periods, respondents indicate that poliCies were more conducive for their local economy before decentralization (see Figure 3). According tQ our survey, 44 percent of local bUSinesses perceive district policies as 'very conducive' or 'conducive' for the local economy before decentralization; while only 29 percent and 27 percent feel the same regarding early and current decentralization respectively. · "c'.', I . .,m Figure 3: Perception of district policies during subsequent governance periods Beforepecentratisation Cunt1l2000) Early Decentralisation Stage (2001, -2003) current Decentralisation Stage (2004-2005) Note: 1= 'very conducive' 2= 'conducive'3= 'moderate' 4='unconducive'5='very unconducive; Asked to what extent policy and administration aspects represent an obstruction to their business, respondents declared 'policy uncertainty' (61 percent), 'unofficial license payments' (53 percent) and 'license administration time' (48 percent) as the three highest ranking obstacles, followed by obstructions of'unofficial payments during tax collection' (45 percent), , official tax expenses' (42 percent), 'official license costs' (37 percent), and 'security payments' (24 percent). Source: Own estimates based on survey data; Likert Scale ranges from 1(no obstacle) ro 5 (very strong obstacle) This general indication of dissatisfaction in the business communities needs further elaboration. The following section will discuss business climate problems in further detail- subsequently in terms of (1) predatory business licensing, (2) and distortionary taxation. ~8!;i'(d! r.n.~'-,ol":r d lO"~,·.1 :'c,;,;,ess ;~<'(:l.Ilatl')n:, F·r,Tt.~, <.in l'Ydi ~'j/-Jt::;" ~"f: ~i~fr:~'; ,g P(d'_!i~~, iW(I) fOlj~ D,;:,'(; c"';~\ If] (e;:tr~i ;~'J~ ~nD 1,',1",5, S~.Tl'a""a ,' Predatory Business Licensing I,~ ! " ' :"..·.........··..... ' ··.·· . ,") Predatory administration of business licenses' is not new phenomena for Indonesian business people. In fact, most interview respondents admit to have already experienced predatory practices during Suharto's'New Order: Yet, empirical studies to date suggest that decentralised governance has brought little irnprovernents.' Decentralized district bureaucracies continue to require a high numberoflicenses and burden local entrepreneurs with enormous administrative paperwork. The 'chain of desks' remains long and technical departments are reluctant to let go of monopolistic licence authorities. In most cases, government officials consider licenses as a revenue instrument - priced well above cost recovery - and rarely understand its steering function for the local economy (Ray 2003, Suhirman 2002). Overall, the complexity of licensing procedures in most districts appears not only redundant but also gives way for public rent seeking. The evidence of our district surveys provides a mixed picture of increasing capture practices and improving administration times. The comparison of administrative practices before decentralisation with those during the last two years reveals that three out of four local governments are perceived to have increased administrative capture. As illustrated in Table 2, the Likert-scaled perception data of the districts Klaten, Pesisir and Kebumen rise between 4 and 11 percentto values between 2.16 and 2.33. This indicates that bribes to receive licenses in theses districts are increasing and currently range around 10 percent. Solok, however, represents a positive exception; capture practices are perceived low and continuously decreasing since decentralization. 4 Predatory licensing is understood as welfare reductions from a political economy perspective. It is the arbitrary act of public officials charg- ing extensive illegal fees, which arise in the context ofpersonal motives, interests and local power structures. In case officials hold unrestricted monopoly positions ofa required service, 5 The list ofcritical accounts on district license procedures is long: (I) According to a study by LPEM-FEUI (2004), a leading economic research institute associated to the University of Indonesia, business/icensing procedures on the district level continue to bea burden for the private sector. The study illustrates that license requirements and the number of involved technical agencies (Dinas Teknis) are excessive. It concludes that such complex procedures weaken the position of license applicants and make illegal charges more likely. (2) An earlier study of LPEM-FEUI- published in 2002 and covering 1736 businesses in 60 districts/municipalities - illustrates that the majority of businesses (79 percent) pay additional costs when dealing with the local bureaucracy. (LPEM-FEUI 2002: 5). These additional costs amount to 9.7 percent of overafl business expenses in Java and up to 11.2 percent outside Java (2002: 7). (3) The Regional Autonomy Watch in its 'Regional Business Attractiveness' report of 2004, (KPPOD 2004: 70), estimates the incidence of admin- istrative corruption even higher. Based on a sample of 5.140 survey respondents, KPPOD declares that 85 percent of the interviewed entrepre- neurs pay illegal charges to public officials, while 15 percent declined to answer. On average these addition illegal fees are seen to increase the costs of licenses and user charges by 60 percent (4) The study on 'Private Sector Perceptions on Local Business Climates in the Era ofOecentralisation; affirms that 'illegal extractions' are the high- est ranking factor obstructing local businesses (Ray 2003: 13; REDI2003). (5) Suhirman (2002: 79-81), after studying bureaucratic practices of two districts in West Java, identifies the following tendencies: (a) In many cases licensing still involves unofficial charges (pungutan tidak resmi?, which raises uncertainty and creates a high-costeconomy; (b) govern- ment departments (Dinas) are in a monopolistic position to issue sector-specific licenses and hardly cooperate with each other ('ego sectoral?; (c) and the drafting process of license regulations does not involve entrepreneurs and other affected citizens. I .-~ --<~ Table 2: District Evidence on License Procedures - Administration Time and l ril,n ,rp Source: Own estimates baser;J on survey data; *7ime Scale: 1= 1~7 days, 2=8-14 days, 3= /5-21 days, 4= 22-30 days, 5=more than one month; **Capture Scale: 1= none, 2:s: 10%,3= 11-20%,4=21-30%,5=31-50%,6= 51-70%, 7=71-100%, 83 100% ., Regarding administration time, all districts show improvements. Recent license procedures require less time than those before decentralization. Yet, the two district cases in Central Java - Klaten and Kebumen, show significantly higher Likert values suggesting that many businesses still wait 2 weeks and more. Once again, Solok receives the best evaluation within the group. An example on how licensing practices differ among districts is illustrated in Table 3. The comparison of administrative guidelines of West-Sumatran districts Solok and Pesisir Selatan shows that time and cost requirements for five basic business licenses can vary significantly.' Table 3: Administration Time and Official Prices for Basic Business Licenses Source: District documents and interviews with officials at technical departments (Perindag, 8ag. Ekonomi); 'One-stop' license services Csatu pintu' or 'satu atap') represent another illustrative example on how differently local business needs are served in the four district cases. The quality of these service stations ranges from very good to very poor. Based on interviews with 6 to 8 business people in each district and direct observations during unannounced visits to service stations, we can evaluate the performance levels of one-stop services (055) in terms of (1) transparency and access of license information, (2) authority to decide on license applications, (3) human capacity, (4) physical facilities, (5) data management, (6) and additional measures to curb bribe payments. 6 Rustiani (2003: 80) confirms the notion that license procedures vary strongly among focal governments. She supports her argument by assessing costs and time to receive a basic industry permit (Tando Daftar Industri' TOI) in four local governments. The variance is significant with costs ranging from 5.000 Rupiah (Yogyakarta) to 477.000 Rupiah (Bandung) and process durations ranging from 2 days (Yogyakarta) to 20 days (Medon). "- ":'".... '"" .,-', Source: Own evaluations based on interviews with business people and personal observations. Note: Scores range from 1-5: 1= 'very poor: 2= 'poor: 3= 'moderate: 4='strong: 5= 'very strong' The low score of Kia ten's OSS is a result ofits overall low standards. During a spo'ntaneous visit (see Figure 4) license information was not available and the technical knowledge of present staff in terms appeared very limited. The dark and dusty room and the old typewriters in the corners did not produce an impression of a professional needs- oriented service unit. As a consequence, most businesses report that they avoid the OSS altogether and approach technical departments for their licensing requirements. In contrast, Solok's OSS ('Satu Pintu Plus') represents a positive benchmark of service provision. Since 1996 services have been continuously improved and extended to serve the licensing needs of Solak's business community. Now the portfolio includes 25 different types of business licenses and civil certifications. It is recognized for its high standard of professionalism and transparency by business people and civil society observers alike. Each applicant receives an information brochure precisely determining required documents, costs and administration times for each license.' For remote business people Solok's government has extended the access to licensing services by the so-called"Plus"option, which allows handing in applications through local post offices. During an unannounced visit to Solak's 055, the four service counters were attended by professional and well trained staff. The atmosphere of the clean and bright room was welcoming. Offices are equipped with rnodern computer technology and colourful information boards.'Anti-Suap'stickers ('no grease money" - see Figure 4) are placed on each counter window - an extraordinary feature of Solak's efforts to curb adrninistrative corruption. 7 APDF-file of the brochure can be downloaded from Solak's government website (www.sofok.go.id) 12 I in'''' .'" ~ ,-,' ;~ ,'.,',},. ~,1'1 ",./·,1. ,.-; '-""" ',', :'~_'.". ;i ;'~I.,.,. " ioU',' ,,~';C'; ~,i I,e: ~I ;..:~' :."'L',--', ,:>~c ':-: 'I,._,I! ':, ;\' 'i "i".:,.""'"' ;,:,,'1 Figure4: Impressions of'One-Stop'Services in Solok, Kebumen, Pesisir Selatan and Klaten (I) Solok ., (2) Kebumen (3) Pesisir (4) Klacen Distortionary Taxes Distortionary taxes" and user charges, have been reported in various empirical studies. The Regional Autonomy Watch (,Komite Pemantauan Pelaksanaan Otonomi Daerah' - KPPOD), a Jakarta-based think tank assessing the impact of decentralisation on regional investment climates, declares that local governments have enacted an alerting arnount of distortionary regulations. Based on their assessment of B96 business-related tax regulations in 200 sample districts, KPPOD argues that 64 districts (32 percent) have released distortionary regulations, while 8 Distortionary taxation is understood in this paper as economic welfare reductions. Undoubtedly, most taxes distort the behavior of economic actors by changing relative prices (substitution effect) and by reducing available income and consumption (income effect) (Stiglitz 2000; James and Nobes 1998). Yet, some taxes cause more severe welfare reductions, referred to as 'deadweight loss' or ~xce5S burden' (Auerbach 1985), than others. Economic theory has shown that government tax practices are distortionary when obstructing trade flows (Caves et o/. 1993; Ohlin 1933; Ricardo 1817; Samuelson 1962). Especially the taxation of mobile trade goods with high price-elasticity results in high welfare losses (Ramsey 1927). only in 29 districts (14 percent) regulations can be considered business friendly (KPPOD 2004: 71). The national chamber of industry and commerce (KADIN) is even more critical. According to their assessment, 73 percent of 353 examined local district regulations are obstructing local economies (KADIN 2003). Besides general conAicts with national laws and tax duplications with national or provincial charges, district authorities are imposing imprudent charges on intra-regional trade, which distorts rather than supports local economic development. The four cases in our study reveal some problems with tax regulations in terms of their direct distortionary impact - but even more they indicate the enormous loss of local revenues due to administrative inaccuracy and negotiation practices. Regarding the first pOint of economic distortions, the following example from Pesisir illustrates that district bills can obstruct local economies. Pesisir's introduction of road maintenance levies (local regulation 13/2001 - 'Perda Retribusi Pemiliharaan Jalan') raises the concern of local and external observers alike. In fact, before their official enactment in 2001, road levies had already caused public uproar among farmers and fishermen. In 1999, two collection posts, one at the border to Padang and one at the local bus terminal in Sago, were burned down by locals and had to be rebuilt (Interview with businessman Pak A.). The Regional Autonomy Watch KPPOD recommends to abolish the road levy regulation as it obstructs regional trade dynamics. According to a internal evaluation of KPPOD, Pesisir's bill to charge commercial vehicles for road usage overlaps with existing provincial vehicle charges and petrol taxes. It represents a burden rather than a stimulation of inter-district trade. This point is supported by a study of the research institute sMERU, which demonstrates the negative impact of road and border levies on local businesses. It shows that cumulative road fees of an average orange transporter from North Sumatera to Jakarta can accumulate up to Rps 1 million and reduce trade profits up to 7 percent.' Even more alerting than the enactment of distortiOnary taxes, is the way how district governments collect taxes and administer local revenues (PAD). The evidence from all four district cases indicates that revenue collection is based on a 'target system: which is a result of rough estimations and personal negations." Modern accountability systems are missing in both the private and public sector. Consequently, tax subjects - be it restaurants, hotels, shops, or home industries - are directly approached by tax officers and inspected regarding their tax potential. Both parties engage in a bargaining process and agree on a tax target based on estimated turnover of the 9 This illustration of trade obstructions is found in SMERU (200 1:26). The study estimates that "the total amount oflevies (official and non-official) paid to transport oranges from Kabanjahe to Jakarta ranges from Rp268.500 to Rp 1.008.500. Paying the lowest amount would only be possible if the truck complies with its permitted capacity. Nevertheless, even when trucks comply with the regulations frequently drivers stiff have to pay levies." According to SMERU's ca/culations the levies of 16 truck weighing stations reduce total profits between 2 and 7 percent. 10 NYa, dengan melihat, kan jadi kira-kira{ .. J Misafnya warung A omzet kira-kira setiap hari berapa, minimal, jadi kita bikin yang paling sepi /ah. [... Sistem akuntansij saya pikir masih jauh, kita tawar-menawar aja masih seperti itu~ ''We are looking at estimated values [. ..}. What is the daily turnover of a small restaurant A for instance. We look at a minimal turnover - on a day that it is least frequented. [...} we are still far away from [accountancy systems). Wejust negotiare - thars the way things still are 11 (Interview at Kebumen's local revenue department). Similar statements have been made in revenue departments in Klaten, Solak and Pesisir. business, its size and ability to pay. The agreed tax target is then collected at the end of the each month. This target-based collection of local revenues is critical in two ways. Firstly, the missing accountability creates room for misconduct. Tax subjects may understate their actual turnover, while the tax collectors may abuse administrative powers to scale up tax targets and bargain for informal rents. As illustrated in Table 5, respondents report higher capture since decentralization in all districts but Solok. Interestingly, the reported capture fees rarely exceed 10 percent of official tax fees, which could either signal a relatively low level of misconduct or an understatement of respondents given their involvement in bilateral negotiations. Source: Own estimates based on survey data; CapwreScafe: 1= none, 2 s: 10%,3= 11-20%,4=27-30%,5=31-50%,6=51-70%,7=71-/00%,83 100% .. District tax offices calfect smaller charges - hotel, restaurant, advertising, small-scale mining taxes etc; ** District branches ofnational tax offices coffect income, value-added and property taxes (PPH,PPN,PBB). And secondly, the bargaining conventions and limited outreach of tax officials induces a significant loss of potential revenues and an uneven division of tax burdens. Many tax subjects are not approached by tax officials - either because of staff shortage or because businesses are considered remote. A selection of specific examples in Box 1 illustrates this notion. Inaccuracy is not only a problem in terms of revenue collection but also in terms of how revenues are administered among government units. Senior officials of local tax offices in all four districts acknowledged that there are vast.challenges ahead for their administration. On the one hand, government units ('Dinas, Kantor, 8agian') lack well trained staff and IT facilities. On the other hand, the prevailing standard of revenue 'targets' rather than clear accounts among departments increases this imprecision even further. Akin to conventions with local businesses, technical departments also pledge to collect and transfer revenues in their field of expertise based on a predetermined target level. One senior official argued that - given the low level of accountancy and transparency standards in many administrations - these target conventions may well provide counterproductive incentives for senior offiCials to capture revenue surpluses. For example, if the industry department (Perindag) pledges to collect Rps 150 million forthe fiscal year 2004, but happens to collect Rps 200 million, it is by no means certain that excess revenues will be delivered to the district revenue account entirely. In fact, they may account for 170 million, receive appreciation for exceeding their target, and capture the remaining surplus (interview with Pak w., senior official in Klaten). Box 1: Examples ofLow DistrictTax Effort (1) In the tile production sector, for instance, nanty around 25 out of 1000 producers are paying taxes~ The tax officials are not approaching 97.5 percent of tile businesses as they are either "far from a major road, too small, or they are scared of the fierce attitude of the owners ~(interview with businessman Pak H. in Kebumen). (2) ~Local officials only approach our hotel with their tax request while other hotels in the region are not requested to pay. {.. .} Such treatment is not {air,N (interview with hotel owner Pak. F. in Pesisir Selatan J. (3) "For example, jf a restaurant taxar Rps 100.000 is due, the owner pays 50.000 of which 25.000 are captured by the tax official and 25.000 ap- pear on the official receipt. [. .. Jthat is common (Interview with a businessman Bapak H. in Klaten). R (4) The owner ofa smaflseaside restaurant, very popular with local public officials, reports that she wasn't asked to pay taxes for years. Now, despite a monthly turnover ofapproximately Rps 10.000.000 (conservative estimation of the owner), and consequently a legal tax obligation of Rps 1.000.000, the owner negotiated a fixed monthly charge of Rps 25.000 with tax officials. Other restaurants in their proximity are not taxed at all (Interview with restaurant owner Ibu D. in Pesisir Selatan). (5) According to Solak's budget report (ABBD 2004), the revenue officers collected Rps 41 million from 87 tax subjects (restaurants) in the fiscal year 2003. This is equal to an overage tax collection of Rps 39.000 per restaurant each month. Thus, given a tax rate of 10 percent this implies an over- age turnover ofRps 390.000 monthly, or Rps 13.000 daify.lf we assume that an average meal in Solak ('Nasi Padang? costs around Rps 10.000, this would lead us to the conclUSion that an average restaurant sells approximately one meal a day. If this would be true, Solak's restaurants were bankrupt by now. More likely, an average restaurant sells more than ten meals a day, hence implying that actual collected taxes do not exceed 10 percent of the tax potential. (6) The owner ofa large restaurant, revealed that the restaurant's daily turnover ranges between Rps 2 and 5 million. A monthly turnover between Rps 60-150 million would imply tax payments ofRps 6-15 million. Yet, in fact, the restaurant has negotiated a monthly payment ofRps 250.000 with the local revenue office. In other words, payments range only between 1.6 and 4.1 percent of the legal tax obligation. (Interview with restau- rant owner Pak. A and his staff in Solak) Summary To recap, our survey indicates - akin to other empirical work to date - that licensing practices vary significantly among decentralized district government but remain in a critical state. In three out of four district cases decentralization has seen rising rather than decreasing capture practices. Time requirements for administrative procedures have decreased during the last years but remain extensive in Klaten and Kebumen. One-stop license services, with the ·exception of Solak, continue to operate on a level that is unsatisfactory to the business needs. Local taxes and user charges may evoke distortions for local businesses in specific cases, as the example of Pesisir's road maintenance levies shows. But even more alarming is the low standard of revenue collection and administration. The convention to estimate revenue targets - externally with businesses and internally among government units -rather than pursing higher accountancy standards gives way to further capture practices and induces high revenue losses for district budgets. Comparing the performance of each district, Solok's government demonstrates by far the best business climate in terms of its licence and tax administration. Not only are capture practices lower and decreasing since decentralization, but also has the district government taken its responsibility serious to serve local licensing needs efficiently. The comparison of the two districts in Central Java is not as clear cut as the one in West Surnatera. Business conditions in Klaten seem to be worse than in Kebumen - especially in terms of administrative capture and one-stop licensing services. Overall, governance-related business conditions in Pesisir, Klaten and Kebumen are perceived unsatisfactory by business communities and show little improvements since the enactment of decentralization. Following the argument of Bard han (1997), in order to understand different policy outcomes of local governments, we must comprehend the features of their .political economy. The following section sets out to do this by assessing institutional frameworks, prevailing bureaucratic conventions, democratic oversight, and governmental leadership. 3. Political Economy of Local Business Regulations Indonesia's rapid transition to decentralized governance, against the backdrop of slowly evolving institutional settings, resembles the pursuit of "building [a] ship while sailing" (Buente 2004). Initiated by a historical 'window of opportunity' (Turner.et al. 2003), the central government reduced authoritarian control and moved towards decentralization and democracy, transferring extensive administrative powers to 440 district governments. After more than thirty years of strong national administration under President Suharto, the sudden leap to more autonomous decision making has created long-term opportunities as well as short-term threats." While district governments are empowered to better respond to local needs, they also have ample space for misusing the enhanced discretion by pursuing private rather than common objectives. The degree to which governments establish favourable business conditions ultimately depends on their political economy - that is to what degree prevailing institutions succeed or fail to curb vested 10cal·lnterests. Undoubtedly, institutions are paramount for governance and development." Following North (1990: 3), they define the "[rules] that structure political, economic, and social interactions:' Applied to context of this study, these rules appear as formal institutions, such as Indonesia's nation laws or district regulations, and as informal 11 McLeod (2000, 2003) argues that the transition from Suharto's authoritarian regime to decentralisation creates more uncertainty for the business community, at least temporarily, by putting an end to Q well-established vertical system of benefit sharing. Key administrators under Suharto devoted their loyalty and resources to an all-informed president and - similar to a 'franchise concept' - received rewards and limited authority in return. A decentralised form ofadministrative corruption, which is no longer fine-tuned bya powerful oligarch who maximizes rentS over time by securing capture practices do not suppress investments and economic growth (Macfntyre 2001), may initially be more harmful to the business environment. 12 The New Institutional Economics and Public Choice literature demonstrates that effiCient allocation ofscarce resources can only take place if the rule of law and individual property rights are well developed within a society, and reinforced by its general value system (Olson 1996). Moreover, due to their effect on trust and certainty, institutions determine the degree of entrepreneurial risk taking and hence the impulses for economic growth (North 1990J.II/-defined institutions, on the other hand, give way to short-term and selkentered indiVidual objectives rather than long- term goals, thus endangering sustainable development. ;;(,)';"("cl' I:.:u,,(,n'y d Lr.;~'.1 3,1'''~.~)' i;.t'!;\J'~;'O"; ,;r,::J;"<'~' (;!' L0:-.o< T"",l!is" 3"(i L;a:~"I:~1 Ptd~t'Ci;': irNP r,W: Di,,'-'c, ,:",~, :n CE';~trJi J;l'd ~na '/,'",;: S·orroa,,,;'c. institutions, such as prevailing conventions and behavioural norms (Furubotn and Richter 1991) - determining the interplay of public and private actors in a district. The following paragraphs discuss examples of deficient formal and informal institutions that create inverted incentive structures. According to our district cases, poor oversight by local parliaments and endemic bribe conventions in the civil service are likely to induce adverse government conduct. Finally, in the latter part of this section, governmental leadership and civic pressures are assessed in their potential to counterbalance adverse conduct and improve business conditions. Deficient Institutions - The National Level The assessment of Indonesia's decentralization laws and their application reveals two problematic aspects for local business climates: (1) the vague wording of the regional tax law 34/2000, (2) and the insufficient capacity of national ministries to monitor new district tax regulations. The enhanced authority of local revenue generation is a core element of the Indonesian decentralisation concept. It is embodied in the 'law on fiscal decentralisation' (25/1999) and specified in the 'law on regional revenue generation' (3412000). While major taxes, such as income, value-added and property taxes remain with the centre, law 34 authorizes districts to administer smaller taxes - such as hotel, restaurant, advertising, parking and small-scale mining taxes. Moreover, it allows the creation of new revenue instruments, as long as they do not oppose the law. Yet, in this point, an institutional problem arises. The legal wording" - new taxes should not 'obstruct local economic development' or the 'common good' and should not be applied to 'goods with a high mobility' - is vague and provides insufficient guidance for district government. Moreover, with Law 34, the burden of proof lies with the central government. National ministries have to prove legal shortcomings of the districts ex post ('innocent until proven guilty'), as opposed to common practise that districts have to defend regulations ex ante ('guilty until proven innocent') (Ray 2003: 23). As a consequence, the Ministry of Finance (MOF) is struggling to evaluate an ever-increasing number of local tax regulations (World Bank 2003: 29) based on a vaguely specified law. This evaluation task is hardly achievable, given the capacity constraints and missing enforcement structures. Indeed, Lewis (2003b; 2003a) reports that the MOFwas only ableto review around 40 percent of the tax regulations of the fiscal year 2001. Moreover, national officials have no credible penalty instruments in case districts fail to revise rejected tax regulations. This deficiency is further accelerated by bureaucratic complexities at the national level. While the MOF can only recommend rejections, final decisions to annul district regulations have to be made by the Ministry of Home Affairs. This inter-ministerial separation of technical recommendations and decisions bears institutional uncertainty and further undermines the credibility of national oversight. 13 Pursuant to Article 2.4, district authorities can enact any new tax or user charge that fulfils the following criteria: (a) it does not represent a no+ tional or provincial tax; (b) it does not contradict the common good; (e) it has sufficient revenue potential; (d) it does not obstruct local economic development; (e) and it does not address goods essential for basic needs or (f) goods with high mobility. 18 I Overall, the imprecise legal wording of the district revenue law combined with the limited supervisory capacity and authority of the Ministry of Finance creates inverse incentives for district governments. Given these deficiencies, it is by no means certain that districts seriously take note of Law 34 when enacting new business regulations. On the contrary, the above mentioned examples of distortionary district taxes - obstructing regional trade or overlapping with existing charges - are to a certain extent accelerated by this institutional vagueness. Deficient Institutions - The District Level On the district level, two institutional features.strike to have a bearing on the political economy of local business regulations; one arising in the legislative and one in the executive sphere of government. First, local parliaments (DPRD) appear institutionally detached from their constituency given the way their members and budget are determined. And second, our case evidence suggests the occurrence of endemic bribing conventions in public administrations. The combination of these features raises the risk of administrative misconduct and poor legislative oversight, thus increasing the likelihood of business distortions. Between 1999 and 2004, the first DPRD members of the decentralization era were determined by closed party- lists not direct vote. District residents were able to vote for parties but not entitled to specify representatives directly." In the early stage of democratisation, in which parties are still in the search of distinct programmatic profiles and people in the process of understanding democratic rights, this resulted in a general detachment between parliaments and their constituencies. The survey in all for districts affirms a strong dissatisfaction of businesspeople towards their DPRD representatives (see Figure 5). According to survey results, 57.0 percent of the respondents in Kebumen, 66.7 percent in Klaten, 70.2 percent in Pesisir and 62.3 percent in Solak are either 'dissatisfied' or'very dissatisfied'with their local assembly." 14 As mentioned in Footnote 2, this has changed in 2004 with Indonesia's first direct election of DPRO representatives. For detailed descriptions on the political developments see Aspinall (2004). 15 This dissatisfaction can be connected to the high incidents of corruption of DPRO representatives throughout the country. Based on news cover- age, the Indonesian Corruption Watch (feW) has identified a list of more than 80 DPRO corruption cases in 25 provinces since the enactment of decentralization (for more derail refer to www.icw.orJd). Figure 5: Satisfaction of Businesses towards their Representation by Local Parliaments ,---------------------------------------------------------------------------------------------------------------, , , , , , , : , g tltl~~1~[f~~F~D~JiIz~1.ffj1i1~ ~iiili{~1~~~t~~~f~Jlfull~~~1~ml~~~~m i I i , , , , , , , o : , , , 1 2 3 4 5 1 2 3 4 5 : t _______________________________________________________________________________ ~ ______________________________ _ Note: 1= 'very dissatisfied,' 2= 'dissatisfied; 3= 'undecided: 4='satisfied; 5= 'very satisfied;' Source: Survey data based on 513 randomly selected business respondents in four districts Another critical aspect can be seen in DPRD budgets. Budget lines are linked to the amount of local revenue generation (PAD) without consideration of a district's economic growth. Pursuant to Article 14 of the Presidential Decree PP 110/2000 the budget for local parliament expenses, including salary payments and travel expenses, can account to up to 5 percent of a district's PAD. Considering that parliamentarians also enact local regulations on taxes and user charges, this decree gives parliamentarians an incentive to raise revenues without considering economic consequences. One way of delimiting distortionary taxation and encouraging DPRD to focus on long- term economic development, may be the setting of budget ceilings based on district GDP indictors (PDRB). Yet, to date, this has not been considered. The second feature of institutional deficiency arises in the public administration. Our case findings indicate a bureaucracy-specific convention that has negative bearing on district business climates: the 'indebtedness' of public servants (PNS) due to bribe payments to enter the bureaucracy system. Estimating bribe conventions scientifically in terms oftheirvalues or frequencies is difficult and raises a note of caution. Undoubtedly, given the limited scope of our study we are not in the position to generalize. Yet, in the context ofthe four cases, incidences of PNS bribing are reported frequently and need to be considered in the political economy of a district. 20 I ,', . ," .. ". -.,-< »_.1_" Source: In-Depth Interviews with 15 respondents in each district, including: Businesses(6), Officials(3), Parliament (I), Academia(l), NGO(2), Media(2);arrows (t /J,) indicate perceptions thatupfront payments increased/decreased sfnce decentralization; 'Exist'indicates that a respondent affirms the existence af payments but daesn't provide estimates. Empirical evidence displays an irregular but widespread convention of upfront bribe payments by PNS candidates." As illustrated in Table 6, interviews in three out offour districts, affirm that a number of applicants still pay high premiums ',n order to be accepted in the public service. The strongest evidence appeared in Klaten. Here, 10 out of 15 respondents gave clear estimates. Based on these estimates, bribe incidents amount to Rps 64 million on average - ranging from Rps 5 to 100 million. In Pesisir and Kebumen bribe conventions are reported lower. Yet, with perceived averages of Rps 23 and 10 million, they still appear significan1.ln Klaten, and to a lesser degree in Pesisir, respondents report an increasing tendency of bribes since decentralization. Sol ok, once again, is a clear exception in this group. All interviewed respondents in Solok assert that bribe payments do not occur. While some report incidents in the past, the overall opinion is that PNS selections are transparent and 'clean'. An overview of selected statements on the degree of bribe payments is presented in Box 2. So why are sOrlJe citizens willing to pay enormous upfront investments in order to enter the public service? In orderto pay a bribe ofRps 50 million, for instance, an applicant may be required to take a bank loan or sell personal assets. Given that initial offiCial salaries in the district government rarely account to more than Rps 1 million, this investment is equivalent to the amount of more than 4 years of income. The reasons why some applicants are nevertheless Willing to make high investments are manifold. In general, demand of PNS positions highly exceeds supply. According to interviews in all four districts, an opening of 50-1 00 pOSitions is generally answered by thousands of applications. This excessive demand is based on economic as well as socio-cultural factors. Regarding high unemployment and a general risk aversion since the economic crisis, many people prefer working in the public rather than the private sector. A PNS position secures life-long employment, health insurance, and 16 Bribe incidences have a/so been reparted by Indonesian media. The Indanesian newspaper Kompas (2004), for instance, reparts af a PNS candi· date camplaining ta President S. B. Yudhayana far being asked Rps 40 millian ta enter the public service in Bekasi, West Java. '.(K~; ~\I; '0'1 ;:;".],.I:·J:".1r.~ f,l\~;·"n .,<,d L:::,~n:":1l:: r,.·.~l'(~·., [.;,):11 cOw, ;');,,'!,Cc ~j;C; n C''''''f~ '3"" Ji;,j ';V,-,,: :it';n<,:"rJ lr1 pension schemes. Moreover, short office hours allow for additional work on the side, thus compensating for the I relatively low wages. And, given the inheritance of Dutch colonialism and strong authoritarian regimes, society at large still considers government positions as the socially most prestigious employment. As one respondent put it: "[ ... JThe high social esteem of PNS positions are deeply rooted in the heads of our society since the Dutch colonization days.[ ... J Fathers rather give their daughters in the hands of a secured public servant than an uprising businessman"(lnterview with Businessman PakA in Pesisir Selatan}. Thus, the pursuit of economic security and social recognition accelerates the interest on PNS positions - in some cases to such an extent that applicants pursue a fast-track option of paying bribes to agents that can guarantee the desired position (Interview with Public Officials Wand R. in Klaten and Kebumen). Box 2: Perceptions on PNS Bribe Payments from Stakeholder Interviews H In order to become a PNS some people pay between Rps 40 and 60 million. {...}. Automatically - against this backdrop of putting in money upfront - one has to coordinate means in order to get a profit out of this [investment) cost. That's the law of economics, correct? {..] So what happens in the end? In the context of license {administration] one needs to receive money back. This becomes a burden to the local economy~ I (Interview with Senior Official W in Klaten) I i I"People in this districtthink thot becoming 0 PN5 is 0 'fortunate gilt of god' ('rejeki dori tuhon? They assume the position as public servontcon I ] be used to benefit one's family and personal incerests'[...JYet, given our weak law enforcement and surveillance, what else do we expect r ({nter- view with Businessman Pak H. in Kebumen) ! i ~Some PNS candidates pay between Rps 2S to 30 million depending on their degree. f . .] Some Teachers are asked for Rps 15 million [. . .] police I officers around Rps 35 million. [. . .]1 was asked myself to pay Rps 10 million in 1996/7 in order to get accepted into the public service. Butl I , declined.f... ]It is sad that people rarely choose to become an entrepreneur. (Interview with BUSinessman Pak A in Pesisir Selatan) i I-DUring the Order Boru times some people poid opproximotely Rps 2.5-/0 million to become public officiols. Now there are no poyments ony j/onger. Police candidates are still asked to pay Rps 30-40 million to their vertical superiors. [. .. ] I believe public servants in other districts in West I Sumatera pay Rps 20-40 million to be accepted# (Interview with Academic Pak U. in Solak) #Before [in the early stage of decentralization] there were rumours that a high-school head had to pay Rps 19 million. f .. .] Also, it doesn't seem unlikely that some people had to pay a certain amount to become PNS before. f .. .] It seems that brokers 'selling' PNS positions do not have to be i in Kebumen - they can be situated in Yogjakarta. [ ...] An interesting feature that was reported from another districts, is that the District Head i gave out 'PNS options' to local parliament fractions or other senior officials as a sign of'appreciation: They can then 'sell' this option to applying I PNS candidates willing to pay.N 1 (Interview with Senior Official/bu. R. in Kebumen) 1 ! N{ know of no extra fees for Incommg PNS candidates In Solak There used to be some - but under the leadership of Bupatt Gawawan Fauzi they appear no longer. { .J The best proof IS that even a member of his own family has not been excepted as PNS last yea~ f. ] The clean PNS selec- I tion process IS dlfectly linked to Solak's transparency reforms" (Interview NGO RepresentatIVe Pak E In Solak) L _____ Yet, this indebtedness of a number of public servants may critically affect the administrative conduct in a district. Assuming rational behavior, the extent of this debt has a bearing on license and tax practices: the more public officials are in debt the higher the likelihood they will pursue 'break even points' by capturing rents outside their L'" _. ".'- L···' _.-,- official salaries. This economic rationale combined with low levels of parliamentarian checks and balances may explain rising predatory licensing practices, which were illustrated in three out of four district cases above. To summarize, in the fifth year of Indonesia's rapid decentralization, institutional problems on the national and district level appear aggravated. The revenue law 34/2000 appears vague in its wording and limited ministerial supervision provides districts with Virtually unrestricted authority to release and administer local business regulations. Prov·lded the evidence of our case studies ·In West Sumatera and Central Java, parliamentarian checks and balances·are perceived as unsatisfactory by the business communities. Linking DPRD budgets to PAD levels does not secure critical oversight of-long-term economic impacts but raises interest in instant revenue generation. Simultaneously, three districts out offour display an alarming trend of bribe payments for incoming government officials. Against the backdrop of prevailing economic and socio-cultural conditions, it appears that some district citizens are making high upfront investments in order to obtain a secure and prestigious PNS position. As a result, a number of public servants are in debt and likely to pursue compensations in their administrative area - including the administration of tax and license procedures. Seen as a whole, these institutional deficiencies may well provoke the 'grabbing hand' (Shleifer and Vishny 1998) - a state in which officials misuse their discretion to capture rents and obstruct the business development. The argument is illustrated in Figure 6 Figure 6: Institutional DeficienCies on National and District Levels National Institutions § r--------~====i..!H~is~to~r~ic:al~/~s~Oc~ie~c~u~lt~u~ra~IF:a~ct~o:rs~ -cE Historical window ______________________--" C" ~ a EO. of Opportunity~ r Deficient National Institutions ~:2 _0. "- Wo + -"" .Q1 t:: :I:~ ~ Party Lists instead of Direct Elections (-2004) r~"-,;=-=t:=====~",;:~'!iii'!::§:=;==~=::::~Ui~:=~=::;;~ OPRD Budget linked to PNS Bribe Payments PAD not PDRB - DebtTrap- Poor Licensing (High Capture / Low Services) District Institutions ,,: ,_":;,, 0")-,' . - ";~_ , ...."' ''.'-''''';;'' 5".'J '." "" {:~:-;, 'c',' , s:,;: (o;c" 'c, '~._C'.:,": ~': ~ -':.'j',\' :>" ,':.c,' Political Economy of Change Considering that all four district cases face similar institutional deficiencies - a vague law on regional revenues, a poor national supervision, and a poor parliamentarian oversight - an interesting question remains: why do some districts in our case study perform better than others? In more specific terms, why does Solok - and to a lesser extent Kebumen - display more conducive licensing and taxation practices than Pesisir and Klaten.Two possible explanations come to mind (see Figure 7). Figure 7: Counterbalancing Factors - Governmental Leadership and Civic Pressure Vision, Political Will Degree to which and Power of a citizens organize District Head themselves to to improve voice or utilize Administrative their exit options r--------~. I PNS Bribe Payments I - Debt Trap- 1--- Distortionary New Taxes ITrade Obstructions) Poor Business Climate First, following theoretical propositions of the Fiscal Federalism and Public Choice literature, civic pressures in terms of people's exit and voice options may play an important role of counterbalancing adverse adrninistrative conduct. And second, arising from Indonesia's specific historical and socio-cultural context, governmental leadership may have a significant bearing on the performance of district bureaucracies. The impact of both of these factors is explored in the following. (a) Civic Pressures The argument that decentralising governance will improve licensing and taxation practices by providing businesspeople with increased means to 'voice' against or 'exit' from governrnental rnisconduct, originates from .:v.: :J , ·. ::: .· , ,c ,"." '" .:'; .,: :;0;:.;':~:~, ~.'; . ;",:;;::'_:";':.~~ I ·. '-".-..::\~. '.' :-c':..,.;', :, ':'~,~ ~~'._,:,' __ :::,-:__..-,-,;'.. _:_-::~.:.. ,':<~::~'.t:'.: :':~~(\~:::::,'~}j~ theoretical discourse. Yet, while many Fiscal Federalism and Public Choice theorists regard civic pressures as paramount in the context of western hemispheres, the four Indonesian case studies display a divergence. In this early stage of democratization, socio-cultural circumstances and prevailing interest structures restrain the ability of local businesses to voice and exit. Theoretically, decentralization creates a closer proximity between decision makers and citizens enhancing responsiveness to local preferences and allocative efficiency (Oates 1972). This argument is often stipulated in two ways. First, following the Tiebout model)' decentralization stimulates an inter-district competition for mobile taxpayers among local governments ultimately producing more favourable tax-service ratios (Tiebout 1956). And ., second, applying Hirschman's propositions)' decentralization may put citizens in a better position not only to 'exit' a less favourable jurisdiction but also to 'voice' concerns politically (Hirschman 1970). Yet, both theoretic frameworks are based on assumptions that are inadequately satisfied in the Indonesian case. In our four case studies, however, Tiebout's assumption of high mobility is not confirmed. In the short and medium term, socio-cultural factors severely restrict businesses in their ability to move to other districts. For one, Indonesian businesspeople are strongly connected to family and community bonds. For another, Chinese businesspeople - who constitute the largest economic force in Klaten and Kebumen - are reluctant to move out offamiliar district environments in fear of more ethnic discrimination and bureaucratic complexities." The argument of high information access is also rarely met on the district level. District regulations on taxes as well as service provision standards are hardly accessible by local businesses; the transparency and dissemination efforts of district policies are limited. According to our study 52 percent of the 513 respondents state to be 'dissatisfied'or'strongly dissatisfied'with their access to government information on business regulations, while 17 The 7iebout model argues in favour of decentralised tax administrations based on a neocfassical framework with the assumption of perfect information and mobility. It argues that citizens will make use of their (ational choice and 'vote with their feet' for the best Qvailable service-tax combination within their reach. Put differently, in a decentralised setting, in which local jurisdictions offer different policy packages of services and taxes, it is up to civic customers to compare and move co the location which meets their preferences best. Following this free-market logic, 17ebout envisions a scenario, in which sub-national governments are forced to compete for critical tax-paying reSidents. The 'invisible feet' of citizens pressure local governments to utilize local tax revenues efficiently and provide adequate public services. 18 Hirschman's argument (1970) complements 77ebout's economic propositions by adding a political dimension. According to Hirschman, citizens have two powerful options to influence government policy. First, akin to Tiebout's logic, citizens can decide to 'exit' from public service provision; either by migrating to more attractive jurisdictions or by choosing private service providers within their jurisdiction. And second, in case exit op- tions are limited by low mobility or bureaucratic monopoly, citizens still have the option to 'voice'theirconcerns. People can activelyengage in the public sphere by expressing dissatisfaction directly (Hirschman 1970: 4). Byengaging themselves in public debates - be it through local electiOns, public hearings or participatory development planning - citizens improve policy environments rather than simply escaping from it (1970: 30). These propositions are well reflected in the good governance reports of development agencies (World Bank 200 1; Picciotto 1995; Rietbergen-Mc- Cracken 1995). Empirical findings affirm that civic voice, the active participation of/oca/ constituencies, is an important ingredient for successful decentralisation. 19 All in-depth interviews with Chinese businesspeople confirmed that moving into a new district implies high costs ofaccustoming to a new social environment and government bureaucracy. Chinese respondents clearly stated that they are generally anxious of discriminatory treatment and violence once they leave their home area. ~)')i'!;G;, ",.:0'-,:');",. ,~f ~X~: 'lu,;,v",;: Il'=;!,-;;Jr;s:~: ~;"ci"p or. L::;(·)I "i~X~;'0'1 ,:r..j: 1c"",n,;ng Yr2~{;(21 ,':sm ~Q~I D:~t~U: c.,~,:;; ,('on,.3! !~'12 ~rc '.'ie.lt )u:on'ei,:! only 14 percent declare to be 'satisfied' or 'very satisfied' (see Figure 8). Overall, it seems unrealistic to assume perfectly informed and mobile citizens while holding government conditions constant (Faguet, 1997: 9). In fact, the empirical cases indicate the contrary. Businesses appear constrained in their mobility and information while district governments are changing more drastically through shifting assignments and local elections. Figure 8: Overall Business Perceptions on Aspects of Civic Pressure . l~l ctih II~J Jfh Il~l ~ "" 1 1 '" 1 , , , I " information by ~he Qovernmen Inform ~byJocal media _Involvement of the riva l~l A i i , 1 Representation throuQh'pariiament , I!~l 11\. I!31Th i i i 1 Representation throuqh chamber i i i 1 j CooperatIon for folnt private sector voice I l~l~ 11;1~1 "'" '" Ii L~====:M===e=:=on=,=,=e~=o=ru=='O=V=O=ic=e=======D=e==,e=e=O=f,=,u=:='.=m=o=n::":b=u'='n:e=,,==e=o:'=e~__________ - - - - - - - - ' , . Note: Scores range from 1-5: 1= 'very dissatisfied,' 2= 'dissatisfied: 3= 'undecided; 4='satisfied: 5= 'very satisfied;' Source: Survey data based on 513 randomly selected business respondents in four districts Table 7: District-specific Business Perceptions on Aspects of Civic Pressure Score (percent of max. score 40 ) 21.53 (54) 19.52 2 0 .2 1 2 1.5 8 (49) (51 ) (54) Note: Scores range from /-5: 1= 'very dissatisfied,' 2= 'dissatisfied: 3= 'undecided; 4='satisfied: 5= 'very satisfied;' Source: Survey data based on 513 randomly selected business respondents in four districts 26 I ',;-, .'.::.,- ", ,-""., The 'civic voice' - as proposed by Hirschman - a state in which local businesses can criticize government shortcomings either directly or through their representatives, is yet to unfold. The empirical data of our four district cases (see Figure 8) affirms that business respondents are dissatisfied with (a) their involvement in new business regulations (52 versus 12 percent), (b) the representation through local parliament (64 versus 7 percent), (c) the representation the local chamber of commerce (55 versus 8 percent), (d) their joint efforts to voice with other businesspeople (52 versus 15 percent), (e) as well as their individual efforts to deliver their opinions to local governments (62 versus 11 percent). Local media coverage and the level of trust among businesspeople are the only two indicators that are not perceived as insufficient. Most interestingly, a cross-district comparison shows that these low perception values hardly vary between cases (see table 7). Low standard deviations in all aspects as well as similar district scores indicate that civic pressure is perceived equally low in Klaten, Kebumen, Solok and Pesisir. Reasons for low perception values are manifold: First, given the long history of authoritarian rule, the voice of Indonesia's "citizens is still in an early stage of development. Throughout the reign of ancient kingdoms, Dutch colonialism, Japanese invasion, Sukarno's 'GUidance; and Suharto's 'New Order; Indonesian people have continuously experienced authoritarian regimes (Mcintyre 2003, Dick 2002). Compared to its historical entirety, Indonesia's democratic transition since 1998 resembles a 'blink of an eye: Hence, it should not come to a surprise that democratic awareness and people's effortto voice is still in an early stage. Second, interviews indicate that community values and socio-cultural norms continue to play an important role on the district level; in Central Java, strong paternalistic structures persist in the realm of Jogjakarta's and Solo's sultanates; in West Sumatera, Minang Kerbau communities continue to follow matrilineal inheritance conventions. What both cultural spheres have in common, however, is that collective structures and communal consensus (musyawara) leave little room forthe'individual voice' common to Western hemispheres. Third, the ability to voice depends on social and economic status. The district evidence suggests that those who have a high social recognition - be it through generosity, communal achievements, or family lineage - are more likely to be heard; and those who are have higher economic assets are more likely to risk public exposure. Fourth, another restrictive factor is the 'mobilizing' of public voice. Anecdotal evidence suggests cases in which complaints and demonstrations are not based on free personal decisions but externally induced by well- paying local elites. In Kebumen, for instance, demonstrations of small shopkeepers against the building of a new shopping mall appear to be stimulated by the owner of an existing shopping mall in fear of competition (Interview with parliamentarian P. and journalist w.) In Solok, a large drinking water producer CAqua') withdrew investment plans after local Minang Kerbau leaders mobilized their communities to demonstrate for higher land and water compensations (Interview with senior official D. and parliamentarian 0.). ·,c_,:,,;,~~, ~,:~,-.C:;':,. G~ ~~~' e",:w", :·~"~:';JT.'~ ", ~L'1'::"~" .:..-. :_~n' ~:'_''':'( ,~~'l; ,,="',(1,} ~",,':;'( " f'-;';'", ~C;_( :;-,,<.1 (."", ':-. ': ,:w;;' ~~<_ ,Yr; l ,3: ;1J",1'~;o Fifth, local chambers of commerce (KADIN) appear dysfunctional. The interviews with KADIN heads in all four districts revealed alarming conditions .. For decades KADIN has been an umbrella organisation for construction and procurement businesses. Yet, with the enactment of a new procurement legislation (Kepres 18/80/2003) these businesses no longer need to be organized under KADIN nor do they require KADIN's authorization for their projects. Unfortunately, local chambers have not yet rose to the occasion and enlarged their outreach and service portfolio other business sectors. As a consequence, membership numbers have been continuously declining over the last years - in most cases ranging below 100. Operational budgets are very low and offices were often situated small run-down buildings. On the whole, given its critical Circumstances, KADIN appears unable to represent the voice of district businesses . ., And finally, the strength of private sector voice has to be seen against the backdrop of prevailing economic interest and power structures. In the four case districts, a large part of the local economy revolves around public infrastructure projects. Since government projects are lucrative and involve little risk, they attract many large entrepreneurs. In general, the construction of public involves three groups of actors: (1) senior government officials and project leaders (Pimpro), (2) assigned econornic representatives of the district parliament (DPRDI Komisi B), (3) and private construction businesses (Pemborong/Gapensi). Based on anecdotal evidence - akin to the literature on 'Iron Triangles' (Lowi 1969; McConnell 1966; Michels 1962) - bureaucrats, parliamentarians and construction companies establish a triangular collusion system. The benefits of invalid procurernents, by up-marking input costs or down-marking output quality - are shared. Concrete examples of such collusion agreements are reported in all four case studies. Box 3 sumrnarizes these staternents. Undoubtedly, the extent of collusion triangles has a direct bearing on the voice of district businesses. The more large entrepreneurs are involved in collusion agreements the less they will be able to genuinely criticize the government. '." ·· ~'~,::"" ,'("- '-"~'_"I ,~,-;;. c', ':"'.';',; . ,,,,~,,~:, ,_ \ ~S.i' .-~,_.-" ,,. "r.) .. 1,."'1,,"';' ,.~;,,'~). ,;r, ",' .r: C"._~ '::'-;<"", ,'-"\.~' _.~",~, -; :,,,,'1' ~l' 1~-,',l Box 3: Statements on Corruption in the Public Tender Business (1) I am currently involved in a public tendered infrastructure project myself. According to my calculations 30 percent of the total value of Rp 204 million [approx. USS 20.000] is captured: 15 percent by the department ofpublic works; and another 15 percent for the focal parliament. Initially , wanted to decline this project - but after ofl - whot can you do? (Interview with Entrepreneur B in District 1) (2) Endemic corruption in the public tender business has significant effect on the quality of streets in this district. In fact, only 50-40 percent of tendered budget is actually used for its purpose. The rest is divided amongst actors involved in a public-private collUSion system. The construction business winning the tender usually keeps between 10 and 20 percent of the tender budget; Political Parties and Parliamentarians abstract approximately 10 percent for their electoral campaigns; the government (Technical Project Leader and District Head) receive around 15 percent; the construction association Gapensi asks for another 5 percent; and security payments to police and local hoodlums (Preman) amount to approximately 10 percent; (Interview with Entrepreneur R. in District 2) (3) The government has made some efforts to make public tendering more transparent-In theory these steps are good, but their implementation in our current reality will rake time. We need a penalty system - a blacklist of corrupt construction companies. Moreover, as long as the rural economy is weak and people are struggling to provide basiC needs and schooling for their children, the nght for greater transparency is only limitedly successful. Overall, tender practices are better than before. Yet, there is still room for corruptjon. Either people know each other and make up a nctive 'competitive tender:' or the tender itself is legitimate but the winning applicant colludes with others during the actual implementation phase to cover up lower qualities and personal gains. Quality cuts still remain and make up to 30 percent of budget values. (Interview with NGO Representatives E. and A in District 3) (4) This year {2005J the purchase of school tables in the amount of Rp 1.5 billion {USS 150.000 Jwas openly tendered. 20 companies joined the tender. Yet, all tender applications but one were faked. It was a setup where 19 applicants were paid off for quietly withdrawing their application. This so-called 'Fake Tender' (Tender Behong-behongan) happens frequently. Eventualfy, since the winning application is based on up-scaled prices, the gain can be shared amongst the ones involved. In this case the real costs were around Rps 900 million, thus Rps 600 million {USS 6O.000J are divided. The involved entrepreneur receives around 70-100 million; 200 million go to government departments; the rest is diVided amongst local parliament and other powerful payers (InteNiew with Entrepreneur Y. in District 4) Note: District numbers are random to prevent back-tracing and secure the confidentiality ofrespondents. In sum, the current state of Indonesia's districts clearly diverges from Tiebout's and Hirschman's ideals. Assumptions of high mobility, high information access, high civic engagement and participation cannot be affirmed by the case evidence. Five years of decentralized democratic governance - in a country that experienced centuries of authoritarian rule - do not suffice yet to produce highly informed, mobile and vocal citizens. The genuine civic voice - which Hirschman envisioned - seems currently constrained by prevailing sOcio-cultural structures and vested interest structures. Hence, provided the evidence so far, the variance of more or less condUCive business conditions, displayed in the first part of this paper, cannot be sufficiently explained by means of different levels of civic pressure. (b) Governmental Leadership Leadership seems to have a stronger bearing on business conditions ..Our empirical findings show that better or worse licensing and taxation practices correlate to respective levels of governmental leadership. The ability to ?\~'I!:lCI' ~',:0f\0.-J1/ C)! ,.O~:J: ~l':,,,(oss F,(og.,IJ";,J,~' FI~d'~'C;\ C'f) ,-,)C,,: r,,',·:WCPl o'1J L;C,"S ~ :o;,),r,,,,''-' ;' ..';1.' :,' , ", ',::>-._] ",.':,.}.:', I 45 "1.' ';"9;, (,'1 c,.,,:.-,: " :,,;',- I ;"'-; c .'",:''''' r-"~,~t',: ", X;l ,;"",- ,";,,'.,.: ',;I;,': '" ">'C".r~: !,,"'j ~,;;'; ':,,,-,' ',~I",~:"" Weber, M., 1991. From Max Weber: Essays in Sociology, Routledge, London. World Bank, 1993. The East Asian miracie: economic growth and public policy, Oxford University Press, New York, N.Y. World Bank, 2001. Decentralization and Governance: Does Decentralization Improve Public Service Delivery?, PREM Notes, No 51, Poverty Reduction and Economic Management Network, Washington DC. World Bank, 2003. Decentralizing Indonesia -A Regional PubliC Expenditure Review, East-Asia Poverty Reduction and Economic Management Unit, Jakarta. World Bank, 2004. 'Indonesia at a Glance; www.worldbank.org/data/countrydata/aag/idn aag.pdf. (11. June 2004) . ., Yin, R.K., 2003. Case study research: design and methods, 3rd, Sage Publications, Thousand Oaks, California. Zhang, T. and Zou, H.-f., 1998. 'Fiscal Decentralization, Public Spending and ECSlnomic Growth in China; Journal of Public Economics, 67:221-40. 46 I