___ 1602 POLICY RESEARCH WORKING PAPER 1602 Program-Based Pollution Despite the absence of a reliable regulatory framework Control Management and credible monitoring and enforcement capabilities, a The Indonesian PROKASIH Program group of plants participating in the -Clean River' Program (PROKASIHJ did undertake Shakeb Afsah efforts to control their Benoit Laplante emissions. The program also Nabiel Makarim provides information that regulators can use to differentiate polluters willing to practice pollution controf from those less willing to do so. The World Bank Policy Research Department Environment and Infrastructure Division May 1996 POLICY RESEARCH WORKING PAPER 1602 Summary findings In 1989, Indonesia's Minister for Population and the group of plants did make efforts to control pollution Environment introduced its "Clean River" Program, emissions. A further and important contribution of the better known as PROKASIH. The program's purpose is PROKASIH program has been to identify the plants to improve water quality by reducing emissions from the willing to practice pollution control and those unwilling most important sources of water pollition in Indonesia. to do so. This information should prove useful in the Though participation in the prograin is not entirely allocation of monitoring resources. A program-based voluntary, compliance with tne terms of the agreement approach targeted at a specific subset of polluters can signed bv the plants is nor legallv binding, and to a large increase the incentives for pollution control. extent, is voluntarv. It is clear from this experiment that regulators must Both total biological oxygen demand (BOD) discharges establish a system for pollution control that analyzes and pollution intensitv (emissions per unit of output) environniental performance of plants reliably. Regulators fronm PROKASIH plants fell significantly; during the must confront such issues as self-reporting, information, period analyzed by Afsah, Laplante, and Makarim. But ;nspections, compliance assessment, and others. A the performance of plants varied widely and the general program-based approach can pave the way to setting in improvemenit in BOD discharges was achieved through place a reliable compliance management svstem. It can the efforts of a few plants. also provide the foundations of a river basin Despite the absence of a reliable regulatory framework environmental management svstem. and credible monitoring and enforcement capabilities, a This paper-a product ot the Fnvironmient and Infrastructure Division, Policy Research Department- is part of a larger effort in the department to pronmote clearer understanding of environmental policy issues. The study was funded by the Bank's Research Support Budget under the research project "The Economics of Industrial Pollution Control in Developing Countries" (RPO 680-20). Copies of the paper are available free from the World Bank, 1818 H Street NW, Washington, DC 20433. Please contact Angela Williams, room N10-02 1, telephone 202-473-7176, fax 202-522-3230, Internet address awilliams@worldbank.org. May 1996. (4 1 pages) The Policy Research WXorking 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 sho4uld he used and cited accordingly. The findings, interpretations, and conclusions are the authors' own and should not he attributed to the World Bank, its Executive Board of Directors, or any of its member countries. P'ri ,Juce.1 hb the P'olicy Rcscarch Dissemiination Ceiiter Program-Based Pollution Control Management: The Indonesian PROKASIH Program by Shakeb Afsah* Benoit Laplante* and Nabiel Makarim World Bank, Policy Research Department, Environment and Infrastructure (PRDEI), Washington, D.C. ** BAPEDAL (Environmental Impact Management Agency), Jakarta, Indonesia. Acknowledgment We are grateful to everyone in the PROKASIH team who have allocated time and effort to answer our numerous questions, and assemble the data necessary to conduct this analysis. We also thank John Patterson, David Wheeler, David Witzel, and Muhammad Zakaria for valuable comments and suggestions. Usual disclaimers apply. Executive Summary The implementation of environmental objectives is particularly demanding for environmental regulators of developing countries. Policy makers may indeed express concerns about diverting resources to pollution control when poverty, illiteracy and infant mortality are still major problems. Hence, most developing countries typically do not allocate the resources necessary to establish comprehensive and extensive systems of pollution control. This may explain that while regulation does frequently exist in developing countries, the monitoring of the regulated community and the enforcement of environmental standards are often extremely weak. As a result, incentives to comply with environmental standards and to control pollution emissions remain generally very small. It does not follow however that environmental regulators of developing countries have no options and should remain powerless in face of deteriorating environmental quality as a result of excessive emissions of pollution. Given the fast rate of industrialization and urbanization experienced by Indonesia, the environmental and health costs imposed by the increasing release of pollution was expected to grow rapidly despite the presence of environmental regulations both at the national and provincial levels. The fact is that the behavior of the regulated industries was not closely monitored, and enforcement of the environmental standards was, for most purposes, non-existent. The Ministry for Population and the Environment had limited resources to monitor or regulate industrial pollution, and Governors of provinces had no incentives to do so. Though reliable data is not available, it is widely believed that industrial plants simply ignored (or were unaware of) the environmental regulation. As a result, the Ministry decided in 1989 to focus its limited resources on implementing a program-based approach for controlling the discharge of industrial pollution in waterways. On June 19, 1989, the Ministry introduced its "Clean River" Program, better known as PROKASIH. Upon its establishment in 1990, the Environmental Impact Management Agency (BAPEDAL) chose to use the PROKASIH program to introduce control of industrial pollution of Indonesia's rivers and to begin implementation of the Water Pollution Control Regulations (PP20/1990), and related Ministerial Decree on Effluent Discharge (KEPMEN 03/199 1). The purpose of the program is to improve water quality by seeking pollution reduction from the most important sources of water pollution in Indonesia. Though participation in the program is not voluntary per se, a particular characteristic of the agreement signed by the plant is that it is not legally binding. Hence, once the agreement has been signed, compliance with the terms of the agreement is to a very large extent voluntary. We show that total BOD discharges from PROKASIH plants fell significantly over the period of analysis. However, we also show that this aggregate result hides considerable differences in the performance of plants. In particular, the reduction in total BOD discharges has been achieved through an improvement in the environmental performance of a small number of plants. A plant's performance can be explained by both i a change in its scale of activity and by a change in its level of emissions per unit of output (pollution intensity). We thus also look at changes in pollution intensity by PROKASIH plants. We show that pollution intensity changed significantly as a result of PROKASIH. Our analysis of the PROKASIH experience suggests that there does exist a group of plants that have exerted effort to control pollution emissions despite the absence of a reliable regulatory framework and credible monitoring and enforcement capability. A significant contribution of a program like PROKASIH is to delineate plants willing to exert pollution control effort from those less inclined to do so. This division should provide useful information for BAPEDAL, and set the stage for further and more focused intervention if needed. Moreover, the desire to control and monitor closely the environmental performance of a limited number of plants, confronts the regulator to the need of setting and implementing a system by which performance is going to be measured and analyzed reliably. Hence, another significant contribution of a program like PROKASIH is that it forces the regulator to confront issues of implementation of the objectives of the program, and more broadly, of the objectives of environmental regulations. Issues of self-reporting, information, inspections, compliance assessment, etc. must be dealt with. ii 1. Introduction Two issues in environmental economics have attracted most of the attention and research effort: the control of pollution emissions and the valuation of the costs and benefits of reducing those emissions.' With respect to the control of pollution emissions, most of the environmental policy debate has centered around the comparison of command-and-control (CAC) and economic instruments (emission charges, tradable permits, subsidies).2 Recent experiments with economic instruments have revealed that a combination of both CAC and econornic instruments is most likely to be efficient.3It remains the case however that the design and implementation of these approaches (or of a mix of them) is highly resource-intensive and impose stringent requirements on the regulator. The implementation of environmental objectives is particularly demanding for environmental regulators of developing countries. Policy makers may indeed express concerns about diverting resources to pollution control when poverty, illiteracy and infant mortality are still major problems. Hence, most developing countries typically do not I Cropper and Oates (1992) provide a survey of each of these issues. 2 Bohm and Russell (1985) summarizes the relative advantages of each approach. 3 For example, economic instruments are not easily tailored to location-specific environmental damages. In such circumstances, interventions of a CAC nature may complement the use of economic instruments. See Hahn (1989), Hahn and Hester (1989) and OECD (1989, 1991) for more details. Moreover, the use of an emission charge to meet a given target requires information that is not readily available. Baumol and Oates (1971) suggests a simple iterative process to achieve a given level of emissions reduction in the absence of information on marginal abatement costs. However, this process ignores that firms undertake significant investment when facing a given charge, and that this investment may not be optimal once the charge is changed. There could therefore be large costs associated with changing the tax rate. Moreover, the process proposed by Baumol and Oates ignores that firms engage in strategic behavior vis-a-vis the regulator. The welfare properties associated with this strategic interaction has been recently analyzed by Karp and Livernois (1994). 1 allocate the resources necessary to establish comprehensive and extensive systems of pollution control. This may explain that while regulation does frequently exist in developing countries, the monitoring of the regulated community and the enforcement of environmental standards are often extremely weak.4 As a result, incentives to comply with environmental standards and to control pollution emissions remain generally very small.5 It does not follow however that environmental regulators of developing countries have no options and should remain powerless in face of deteriorating environmental quality as a result of excessive emissions of pollution. On the basis of the Indonesian experience, we argue in this paper that a program-based approach targeted at a specific subset of polluters can increase the incentives for pollution control, and pave the way to setting in place a reliable compliance management system. Indonesia has achieved remarkable economic success over the last twenty-five 6 years: per capita income increased annually at a rate of 4.5% after 1970. This success 4 It should not be implied that monitoring and enforcement issues have been solved in the United States (and more generally in developed countries). Russell (1990) writes: "What is missing is a commitment of resources to checking up on whether those covered by the law and regulations are doing (or not doing) what is required of (or forbidden to) them." (p. 243). In a recent study, the General Accounting Office (1993) concludes that the EPA cannot ensure the accuracy of the pollution data reported by polluters. In Quebec, while 59 pulp and paper plants were in operation during the period 1985-1990, there has been a total of only 54 sampling inspections by the Ministry of the Environment (Laplante and Rilstone (1995)). 5 See O'Connor (1994) for more details. 6 Per capita income was US$50 in the late 1960s. It is now estimated to be US$650. Poverty fell drastically (from 70 million individuals to approximately 27 million); life expectancy rose from 41 years in 1960 to 61 years in 1990; primary school enrollment nearly tripled and secondary school enrollment increased 8-fold. 2 was achieved through rapid industrialization: while manufacturing represented 13% of GDP in the 1970s, it represented 23% of GDP in the 1980s. This rapid development had serious adverse impacts on the environment, especially in Java where 75% of the total Indonesian industrial activity is located. In particular, the quality of surface water has become a major source of concern.8 Given the fast rate of industrialization and urbanization, the environmental and health costs imposed by the increasing release of pollution was expected to grow rapidly despite the presence of environmental regulations both at the national and provincial levels. The fact is that the behavior of the regulated industries was not closely monitored, and enforcement of the environmental standards was, for most purposes, non-existent. The Ministry for Population and the Environment had limited resources to monitor or regulate industrial pollution, and Governors of provinces had no incentives to do so. Though reliable data is not available, it is widely believed that industrial plants simply ignored (or were unaware of) the environmental regulation. As a result, the Ministry decided in 1989 to focus its limited resources on implementing a program-based approach for controlling the discharge of industrial pollution in waterways. On June 19, 1989, the Ministry introduced its "Clean River" Program, better known as PROKASIH. Upon its establishment in 1990, the Environmental Impact Management Agency (BAPEDAL) chose to use the PROKASIH 7 Total industrial output has increased 8-fold since 1970. This development was accompanied with a rapid increase of Indonesia's urban population from 15% to 30% of total population. This was particularly true on the Island of Java, which accounts for 60% of the Indonesian population with a population density that is among the highest in the world. For more details on Indonesia's economic development, see World Bank (1994). Water is estimated to be an important factor of disease in Indonesia as most water sources are considered unsafe to drink. The benefits of solely reducing the diarrhea-related mortality by 50% were estimated to be in the order of US$300 million in 1990 (World Bank, 1994). This number ignores the gains from reducing the effects of water pollution 3 program to introduce control of industrial pollution of Indonesia's rivers and to begin implementation of the Water Pollution Control Regulations (PP20/1990), and related Ministerial Decree on Effluent Discharge (KEPMEN 03/1991).9 The purpose of the program is to improve water quality by seeking pollution reduction from the most important sources of water pollution in Indonesia. Though participation in the program is not voluntary per se, a particular characteristic of the agreement signed by the plant is that it is not legally binding. Hence, once the agreement has been signed, compliance with the terns of the agreement is to a very large extent voluntary. Recent literature has pointed out the potential role of other policy tools to induce greater pollution control effort from industrial units. In particular, programs based on "voluntary" participation and programs based on the provision of information to various stakeholders increasingly attract attention.1° There is some evidence from the US experience that emissions can be reduced through voluntary programs. Such an experience is the 33/50 Program, initiated by the US Environmental Protection Agency (EPA) to reduce releases of 17 toxic chemicals by 33% and 50% by the end of 1992 and on morbidity. Air pollution and toxic wastes are also important issues. However, in this paper we focus solely on water pollution. 9 Because of its limited resources, BAPEDAL created a program called JAGATIRTA whose purpose is to respond to complaints raised by local communities. The program is therefore extremely focused, and a follow-up is made only on those complaints judged by BAPEDAL to be significant. On the role and impact of information provision programs, see Hamilton (1995), Kennedy, Laplante and Maxwell (1994), Laplante (1995), Laplante and Lanoie (1994), and Muoghalu et al. (1990). 4 1995 respectively (hence the name 33/50). Participation in the program is voluntary, and commitments to achieve reductions are not enforceable by law. As of February 1992, more than 700 plants had committed their participation."' Early evidence suggests that for the period 1988-1992, toxic emissions fell by 40% (7 points above the target). After more than 20 years of CAC, the EPA claims that its 33/50 Program is an effective alternative to traditional regulation. The purpose of this paper is to identify the impact of PROKASIH on both total discharges of biological oxygen demand (BOD) as well as on the pollution intensity (pollution per unit of output) of participating plants. Looking at changes in BOD load is justified by the regulator's concern over ambient quality. Indeed, ambient quality is primarily affected by the total load of emissions discharged in receiving waters. Changes in total discharges is therefore of relevance to the regulator. We show that total BOD discharges from PROKASIH plants 2fell significantly over the period of analysis. However, we also show that this aggregate result hides considerable differences in the performance of plants. In particular, the reduction in total BOD discharges has been achieved through an improvement in the environmental performance of a small number of plants. Arora and Cason (1995) analyze the characteristics of the plants participating in the 33/50 program. In particular, they show that large plants are most likely to participate. 12 We use the expression "PROKASIH plants" to identify plants that are participating in the program. 5 A plant's performance can be explained by both a change in its scale of activity and by a change in its level of emissions per unit of output. We thus also look at changes in pollution intensity by PROKASIH plants. We show that the pollution intensity of PROKASIH plants fell significantly over the period of analysis. These results suggest that environmental regulators of developing countries, despite a lack of resources, can proceed forward to control pollution emissions and achieve significant results through a program- based approach targeted at a specific subset of polluters. In the next section, we discuss PROKASIH in more detail, and describe the dataset that has been used to perform the analysis. In Section 3, we analyze changes in BOD discharges by PROKASIH plants while in Section 4, we examine changes in pollution intensity. In Section 5, we discuss in more detail the role of a pollution control program of the PROKASIH nature. We conclude in Section 6. 2. The PROKASIH Program and the dataset (a) The PROKASIH Program The primary objective of PROKASIH is to prevent further decline in river quality. The program is based on pollution reduction agreements co-signed by provincial Vice- Governors, BAPEDAL, and participating plants. In 1989, 8 provinces were participating in the PROKASIH program. This number increased to 13 provinces in 1994. 13Vice- Governors serve as local coordinators. In each province, an implementation team (called 13 Four new provinces became PROKASIH provinces in 1995/96: Bali, Sulawesi Selatan, Manado, and Jambi. 6 PROKASIH team) has been constituted with representatives from various institutions: public works, regional development planning board (BAPPEDA), health department, laboratories, environmental study centers, etc. Both BAPEDAL and the provincial governments provide financial resources to the PROKASIH team. The responsibilities of the PROKASIH team include the following: * Identification and selection of industrial units that are significant polluters; * Measurement of the quality of polluters' effluents and water ambient quality; * Data collection and reporting to BAPEDAL In order to achieve the objectives of PROKASIH, priority is given to specific rivers, or portion of rivers where concerns over water quality are most serious, and by seeking pollution reductions from the largest polluters along the chosen rivers. It is the responsibility of the provincial PROKASIH team to choose both the rivers and the polluters. In 1994, 1405 establishments were participating in PROKASIH. Given the importance of industrial plants in the program (industrial plants account for 90% of participating establishments), we focus solely on these plants in this paper. The number of PROKASIH plants varies considerably across provinces. As shown in Table 1, Jawa Barat and DKI Jakarta represent by far the largest number of participating plants, with approximately 75% of the total number in 1994. Jawa Barat itself covers 56% of the total number of plants, and exhibits a substantial increase in 7 participation since 1990. Though both Jawa Barat and D.K.I. Jakarta appear to be very active in terms of enrolling plants in the Table 1: Number of PROKASIH Plants Per Province Province 1990 1991 1992 1993 1994 Jawa Barat 7 7 100 | 326 | 5S29 _ 723 723 DKI Jaka_rta 96 | 193 T 220 T 228 228 Jawa Tengah 44 44 44 64 64 Jawa_Timur 39 T 45 T 45 T 45 45 Lamnpung 9 T_30 T 30 T 34 34 Sumatera Selatan 33 33 33 33 33 Kalimantan Timur 30 30 30 30 31 Sumatera Utara 30 30 30 30 30 Kalimantan Selatan 0 0 0 0 20 Riau 0 18 18 19 19 D.I. Aceh 0 14 14 14 17 D.I. Yogyakarta 0 0 0 0 16 Kalimantan Barat 0 15 15 15 15 TOTAL 381 778 1008 1235 1275 PROKASIH program, it is interesting to note that they are also the two provinces where the budget of the PROKASIH team per plant is the smallest. As will be shown in the next section, this may explain that none of the PROKASIH plants in Jawa Barat and D.K.I. Jakarta survive our selection criteria for inclusion in our final sample of analysis. In particular, plants in D.K.I. Jakarta and Jawa Barat report their emissions at a frequency that is insufficient to estimate reliably their pollution profile.14 (b) The dataset Our dataset has been constructed in the following way. First, we identified plants that became PROKASIH plants in 1990, and for which there is at least one measurement per year, for each of the years covering the period 1990-94. A similar exercise was performed for the plants that became PROKASIH plants in 1991. Second, we develop a 14 Appendix I describes the process by which BAPEDAL collects information from PROKASIH plants. 8 strict set of guidelines that allowed us to identify and reject observations that could not be technically explained.15 As described in Table 2, the final dataset for the period 1990-94 covers 100 plants located in 6 provinces along 24 rivers, and for which a total of 2819 observations are available (this represents an average of 5.6 observations per plant per year over the time period); the 1991-94 dataset covers 55 plants located in 5 provinces along 10 rivers, and for which 937 observations are available (for an average of 4.5 observations per plant per year). The number of plants in our dataset (100 and 55) may appear small relative to the total number of PROKASIH plants participating in the program in 1990 and 1991 (381 and 778 respectively) . However, as noted above, observe in Table 2 that none of the plants in D.K.I. Jakarta and Jawa Barat are included in our dataset since the data from those provinces is too sparse and unreliable. If one excludes PROKASIH plants from those two provinces, the total number of PROKASIH plants in the remaining 11 provinces in 1990 is 185 (Table 1). Our 1990-94 dataset therefore covers more than 50% of the participating plants. Similarly, excluding D.K.I. Jakarta and Jawa Barat, 74 plants became PROKASIH plants in 1991. Our 1991-94 dataset therefore covers approximately 75% of those plants. The coverage of our dataset per province is described in Table 3. 15 Thlese guidelines are available upon request. 9 Table 2: Description of Datasets Period of analysis Provinces Rivers 1990-94 1991-94 Number of Number of Number of | Number of new plants observations new plants | observations Jawa Tengah Anyar 2 46 | Jawa Tengah Bengawan Solo 2 46 Jawa Tengah Kaligarang 2 44 Jawa Tengah Ngringo 6 134 Jawa Tengah Palur 1 21 Jawa Tengah Pengo 4 92 Jawa Tengah Pepe 1 23 Jawa Tengah Premulung 1 25 Jawa Tengah Sroyo 4 89 Jawa Timur Kali Brantas 14 612 Jawa Timur Kali Lesti 3 132 Jawa Timur Kali Porong 2 100 Jawa Timur Kali Surabaya 10 518 Jawa Timur Kanal Mangetan 1 51 -- -- Jawa Timur Kali Mediun -- -- 6 140 Kalimantan Timur Mahakan 8 147 -- -- Lampung Way Pangubuan 4 80 Lampung Way Seputih 1 20 -- -- Lampung Way Pegadungan -- -- 5 50 Lampung Way Sekampung -- | 6 81 Lampung Way Terusan -- 2 24 Lampung Way Tul. Bawang -- -- 5 59 Sumatera Selatan Kramasan I 15 -- -- Sumatera Selatan Musi 16 257 Sumatera Selatan Ogan 2 33 Sumatera Utara Asahan 5 129 Sumatera Utara Deli 6 143 Sumatera Utara Merbau 2 54 Sumatera Utara Semayang 2 52 -- -- D.I. Aceh Langsa -- -- 3 37 D.I.Aceh Tamiang 3 39 Kalimantan Barat Kapuas 5 119 Kalimantan Barat Kapuas Kecil 5 119 Riau Siak 15 269 10 Table 3: Coverage of Datasets per Province 1990- 1994 _ 1991 - 1994 Number of Number of Number of Number of Prokasih Prokasih % new Prokasih % PROVINCE plants in plants in coverage Prokasih plants in coverage 1990 dataset plants in dataset of new - 1991 plants Jawa Barat 100 0 0 226 0 0 DKI Jakarta 96 0 0 97 0 0 Jawa Tengah 44 23 52.3 0 0 Jawa Timur 39 30 77.0 6 6 100 Lampung 9 5 55.5 21 18 85.7 Sumatera Selatan 33 19 57.5 0 0 - Kalimantan Timur 30 8 26.6 0 0 - Sumatera Utara 30 15 50.0 0 0 - Kalimantan Selatan 0 0 - 0 0 - Riau 0 0 18 15 83.3 D.I. Aceh 0 0 - 14 6 42.8 D.I. Yogyakarta 0 0 0 0 - Kalimantan Barat 0 0 - 15 10 66.6 Both datasets will be used to assess the overall trend in pollution load and pollution intensity of industrial sources participating in the PROKASIH program. In the next section, we examine the aggregate changes in BOD load by PROKASIH plants for each of the rivers along which PROKASIH plants are located. In Section 4, we disaggregate this result to examine the response of individual plants following their participation in PROKASIH. 11 3. Changes in aggregate BOD load In this section, our interest is to analyze the trend in aggregate BOD load by PROKASIH plants. For this purpose, let Ci, be a measure of BOD concentration of plant i's effluent in year t, and C,, be the average BOD concentration of plant i's effluent in year t. Similarly, let Fit be a measure of flow rate by plant i in year t, and Pit be the average daily flow rate for plant i in year t. Let BODit be the BOD load of plant i in year t measured in kg/day. Then, BODit is given by:20 BOD= Ct FD 1000 Finally, let N1 be the number of PROKASIH plants discharging in river j. Let BODjt be the total BOD load by PROKASIH plants, in river j in year t, measured in kg/day. Then, BODjt is simply: BODJ, = IBOD,r i=l We examine the evolution of the variable BODj, for both the 1990-94 and 1991-94 dataset, along each of the PROKASIH river. Results are presented in Table 4 and Table 5 for each of the dataset, and then grouped into three categories in Figure 1, 2 and 3 according to whether aggregate BOD discharges increase, decrease or are uncertain along each of the river. As can be observed, the evolution of total BOD load is of a similar nature for both datasets: there are strong indications that total BOD discharges from PROKASIH plants have been significantly reduced in 18 of the 34 rivers in our dataset. However, in 9 rivers, BOD discharges by PROKASIH plants have reached higher levels in 1994 than in 1990 or 1991. In the aggregate, total BOD discharges fell significantly 20 (mg / liter) *(cubic meter / day) = (mg / liter) * (1000 liters / day) = kg / day. 12 Table 4: Percentage Ch ange of BOD Load by River (1990-94) Province River 1990 % change % change kg/day 1990-91 | 1990-94 Jawa Tengah Anyar 70 62 141 Jawa Tengah Bengawan Solo 1519 -80 29 Jawa Tengah Kaligarang 14 79 277 Jawa Tengah Ngringo 1542 21 -57 Jawa Tengah Palur 52 -9 1 Jawa Tengah Pengo 436 -27 11 Jawa Tengah Pepe 89 -58 -65 Jawa Tengah Premulung 826 31 65 Jawa Tengah Sroyo 255 143 -73 Jawa Timur Kali Brantas 5842 -32 -3 Jawa Timur Kali Lesti 8624 -49 -59 Jawa Timur Kali Porong 3771 346 51 Jawa Timur Kali Surabaya 10913 -8 -18 Jawa Timur Kanal Mangetan 8575 26 -63 Kalimantan Timur Mahakan 400 -18 -7 Lampung Way Pangubuan 2784 -39 -59 Lampung Way Seputih 709 -16 -94 Sumatera Selatan Kramasan 1599 -66 -58 Sumatera Selatan Musi 10406 -44 -53 Sumatera Selatan Ogan 2391 -52 -35 Sumatera Utara Asahan 3022 -38 -71 Sumatera Utara Deli 405 -48 -49 Sumatera Utara Merbau 586 56 -62 Sumatera Utara Semayang 247 -5 -80 over the period of analysis. In Table 4, while aggregate BOD discharges by the 100 plants were 65 077 kg/day in 1990, these discharges fell to 41 846 kg/day in 1994, a decline of 36.25%; in Table 5, total discharges fell from 106 147 kg/day in 1991 to 59 489 in 1994, a reduction of 44%. That such reductions were achieved by PROKASIH plants is certainly of clear interest to BAPEDAL. However, of potentially more significance for BAPEDAL, there is a clear indication, in Figure 1, that BOD load is on an upward trend in 1993 and 1994. Though additional data will 13 Table 5: Percentage C ange of BOD Loaby River (1991 Province River 1991 % change % change _ kg/day 1991-92 1991-94 Jawa Timur Kali Mediun 743 -10 -41 Lampung Way Pegadungan 22670 -59 -98 Lampung Way Sekampung 3463 85 473 Lampung Way Terusan 717 166 124 Lampung Way Tul. Bawang 18627 -28 -61 D.I. Aceh Langsa 729 -34 -47 D.I.Aceh Tamiang 4993 -90 -87 Kalimantan Barat Kapuas 2594 153 306 Kalimantan Barat Kapuas Kecil 9024 -83 -48 Riau Siak 42587 -67 -68 reveal whether or not this trend persists, it is worth pointing out that a number of factors may explain this development. First, it should be noted that total BOD discharges increase, ceteris paribus, with actual production. The important economic growth rate currently experienced by Indonesia may therefore explain this increase in BOD discharges. It is important to note that these increases could take place despite greater pollution control efforts by PROKASIH plants. This is why we examine, in the next section, BOD emissions per unit of output (pollution intensity), instead of total BOD discharges. Second, for some rivers, the number of participating plants in our dataset is small. Finally, this upward trend could also indicate that there may be a limit to the ability of a program of the nature of PROKASIH (in which agreements to reduce emissions are not legally binding) to induce persistent pollution control efforts on the part of the firms. This may be particularly the case in a situation where industrial growth is rapid, and where the enforcement of the environmental regulation has traditionally been, and to a large extent remains, lacking.2' 21 As mentioned earlier, currently most enforcement actions are undertaken through JAGATIRTA following complaints of local communities. The number of such actions is very limited. 14 Figure 1: PROKASIH Rivers Where Average BOD Load per Day Declined 1990-94 andl991-94 (Kilogram of BOD per day) 0.1. ACEH--TANIANB JAVA TENGAH--PEPE JANA TENGAH--SROYO JAVA TINLU--KALI HAOIUIN JANA TD4UR--KALI SUABAYA 6000 200 600 goo 1+ I o0 4000 00 600 50 400 50 2000 200 200 0 0- 0- 0 0- 90 91 92 93 94 90 91 92 93 94 90919293 90 91 92 93 94 90 91 92 93 94 JANA 1DtRl--EANAL NANGETAN KALINANTAN BAAAT--KAPUAS KECILLAMPUNG--WAY PANGLIUAN LANPLIN6--AY PEGADULI1AN LAIIPUN6--AY SEPUTIH 10000 1000 0 3000 30000 800 0 2000 20000 600 5000 5000 400 1000 10000 200 0- 0 0- 00 O 90919 90 9v,O9-92 93 gb 991 92 99 94 it 912 93; 0 9 91 9399 LAMRPN&--MAY TILANG BAVAHN R I A U--SIAK SUKA7ERA SELAIAN--KRANASAN SUIAIERA SLATAN-0GAN SLMATERA UTAR4--ASAHAN 20000 40000 500 3000 3000 15000 1000 2000 2000 10000 20000 5000 500 1000 1000 0 0I I I I 0 0 0 9091 92 93 94 O 90 92 93 94 90 91 92 93 94 90 91 92 93 94 90 91 92 93 94 SUNATERA UTkRA--DELI SUIATERA UTARA--NERBAU SUNATERA UTARA--SENAYANG 400 1000 300 300 200 200 500 100 100 0 0 0 90 91 92 93 94 90 91 92 93 94 90 gi 92 93 94 15 Figure 2: PROKASIH Rivers Where Average BOD Load per day Increased 1990.94 and 1991-94 (Kilogram of BOD per day) JAWA TENGAH--ANYAR JAWA TENGAH--BENGAWAN SOLO JAWA TENGAH--KALIGARANG 400 - 2000 100 - 300- I 1500 - 200 - 1000 50 0 f)~~~~~oi 01 I o0 91 92 93 94 90 91 92 93 94 90 9'1 92 93 94 JAWA TENGAH--PENGO JAWA TENGAH--PREMULUNG JAWA TIMUR--KALI PORONG 600 - 1500 1 20000 - 400 - 1000 - 15000 1 20 100001- 200- _________ /5001- 5000 so°° __/' __' 0 0 I1 I I I I1 f 90 91 92 93 94 90 91 92 93 94 90 91 92 93 94 KALIMANTAN BARAT--KAPUAS LAMPUNG--WAY SEKAMPUNG LAMPUNG--WAY TERUSAN 20000 - 20000 - 2000- 15000o- -i iso 1500 - iaooo ioooo / 00oo0- 5000- 5000 -f 500- 0- 91 92 93 94 90 91 92 93 94 90 91 92 93 94 Figure 3: PROKASIH Rivers Where BOD Load per day Trend is Uncertain 1990-94/1991-94 (Kilogram of BOD per day) D.I. ACEH--LANGSA JAWA TENGAH--NGRINGO JAWA TENGAH--PALUR 800- 2000- 1o0- 600 - 1500 - 60- 400- 1000- \40- 200- 500- - 20_- 0 - 0 -1 ~ ~~~~~I I I I 0 1 I I I I 90 91 92 93 94 90 91 92 93 94 90 91 92 93 94 JAWA TIMUR--KALI BRANTAS JAWA TIMUR--KALI LESTI KALIMANTAN TIMUP--MAHAKAM 6000 - 10000 - 600 1 4000- 5000 \ 400 2000 - 200 - 90 91 92 93 94 90 91 92 93 94 90 91 92 93 94 SUMATERA SELATAN--MUSI 15000 - 10000 -\ 5000 - __ 90 91 92 93 94 16 In the next section, we analyze the data at the plant level,, and seek to identify the individual contribution of the plant to the overall BOD reduction measured above.22 We also examine changes in pollution intensity and suggest, through counterfactual analysis, that the impact of PROKASIH may be larger than measured above. 4. Changes in BOD load and pollution intensity at the plant level In this section, we analyze the pattern of plant-level responses, using pooled data across rivers from the 1990-94 dataset.23 First, we examine changes in BOD load, and then analyze changes in pollution intensity. (a) BOD load Before looking in the details of the plant-level responses, it is interesting to rank the PROKASIH plants in terms of their individual contribution to total BOD discharges. In order to do so, we have computed an index that is similar in nature to the Lorenz curve developed in 24 industrial organization. Let BOD90 be the total BOD load, in 1990, by the 100 plants of the 1990-94 dataset: 22 It should be noted at this point that no attempt is made in this paper to obtain a measure of the relationship between ambient concentration at any given point along the rivers and BOD discharges of individual plants; nor is an attempt made at measuring in dollars the impact of BOD discharges in any given river by any given plant. If the PROKASIH plants in our dataset were the only sources of BOD discharges, this exercise could be performed, albeit obtaining a dollar measure of marginal damages would still be problematic. However, (1) we have eliminated a significant number of PROKASIH plants due to a lack of data; (2) PROKASIH plants are not the only industrial sources of BOD discharges along a river; and (3) one must also account for non-industrial sources of BOD discharges. These are the object of ongoing research. 23 For the purpose of this section, we are solely using the 1990-94 dataset since it offers a longer period of observations. This is particularly important for the analysis of pollution intensity since we have output data up only to 1993. The 1991-94 dataset would thus offer only 3 years of observation. 24 The Lorenz curve shows the percentage of total industry sales accounted for by any given fraction of the firns of the industry, with the firms ranked in decreasing order of market share. 17 ZQ BOD9 = ,BODj,. i=, Then rank the plants such that BOD190 > BOD2 > > BOD,w. The curve in Figure 4 represents the ratio IBOD / BODg accounted for by the fraction n / 100 of the largest plants i=1 in the dataset. When computed for the entire dataset (n = 100), the ratio is equal to 1. In Figure 4, the x-axis represents the cumulative proportion of plants, while the y-axis represents the cumulative proportion of total BOD accounted for by these plants. If each plant were contributing equally to aggregate discharges (BOD190 = BOD290 = ... = BOD,oo9o), the computation of the ratio described above would yield a straight line diagonal in Figure 4. The black curve indicates very clearly however that this contribution is far from being uniform: 50% of total BOD discharges is accounted for by less than 10% of the plants; 20% of the plants accounts for approximately 75% of total BOD discharges in 1990. Without any doubt, this suggests that most of PROKASIH's impact on total BOD discharges crucially depends on the behavior of a relatively small number of plants. A large reduction in BOD discharges by the largest 10% of the plants would significantly reduce pollution emissions. On the other hand, if 50% of the plants at the bottom end of the distribution were abating their emissions from current levels to zero, total discharges would fell by less than 5%. There are various ways by which one can analyze the contribution of each individual plant to changes in total BOD load during the period of observation, and identify the extent to which these changes are driven by the behavior of a relatively small number of plants. In order to account for the large variation in the BOD discharges across plants, an interesting way 18 Figure 4: Distribution of BOD Load- 1990 100% 90% __==__ 80% 0 -a70% 0 *j60% 50% 1_58 9 ~ 0% 0% 10% 20% 30% 40% 50% 60% 7(0% 80% 90% 100% Cumulative Share of Number of Plants to analyze the contribution of individual plant to changes in total BOD load is to express the change in total BOD discharge between 1990 and any given subsequent year t (%ABODt), as a weighted sum ofthe changes by each individual plant (%ABODit), with the weight being the plant's contribution to total BOD discharge in 1990 (si). The percentage change in total BOD discharges between 1990 and any subsequent year t is given by: %ABODt = BODt - BOD9 -A D = BOD90 After some manipulation, this can be rewritten as: 19 [BOD, -BODi BODj1 'L BOD,9O BOD9 j which becomes: % ABODt = Xs* % ABOD t Hence a large variation in the BOD load of a plant which accounts for only a small portion of total BOD discharges will have only a small effect on total BOD load. Conversely, large changes from plants accounting for a large share of total BOD load in 1990 will have a significant impact on the total measure. To illustrate, we have arranged the contribution of individual plants in descending order, from the largest positive contribution to changes in total BOD discharges (i.e. firms increasing their BOD load) to the largest negative contribution (firms decreasing their BOD load).25 Results appear in Figure 5. From the figure, it appears clearly that many plants did not meet the terms of their pollution reduction agreements. Indeed, over the period 1990-91 and 1990-94, approximately 10 plants increased their BOD load significantly. The figure also suggests an extremely skewed distribution of plant contributions to total changes in BOD discharges. In particular, the bulk of the reduction in BOD load is explained by less than 20 plants. Most strikingly, more than 65% of the plants had a negligible impact on the change in BOD load. 25 In other words, we have ranked the plants so that si %ABOD,t > s2%ABOD2t> 000 >SN%ABODNt. It shall be understood that the plant with the largest positive contribution is not necessarily the plant that has increased the most its BOD discharges since our index accounts for the plant's share in total BOD discharges in 1990. 20 Figure 5: Change in BOD Load Contribution-Share by Plant 40% 40% 35% -.- 1990-91 - 1- 1990-9 35% 30% 235% 25% 302% o 20% - 20% o , 15% 15% , 10% \ 10% -D 0% 10 | .. 5. % 0 0. ~ ~ ,% .5% >oo PROKASIH Plants (100) N5 - -10% --0 E-15% * -15% E -20% '. -20% -25% - -25% -30% -30% .35% * -35% 40% -40% As a result of these changes, the distribution of BOD load across the 100 plants of our dataset changes significantly from 1990 to 1994. This is illustrated in Figure 6 where the dotted line represents plants' share of total BOD load in 1990, ranked from the plant with the largest share (in 1990) to the plant with the smallest share.26 In 1990, the plant with the largest share explained approximately 13% of the total BOD load of the plants in our dataset; this same plant explains 7.5% of total BOD load in 1994. In general, observe that the "pollution share" of individual plants fell significantly between 1990 and 1994. However, some plants exhibit a very sharp increase. For example, plant number 9 went 26 For the purpose of clarity, we have truncated the figure at n = 64. The individual share of the plants not represented in Figure 6 is close to zero in both 1990 and 1994. 21 Figure 6: BOD Load Share Trend by Plant 14% ............ - - - - - - 13% '. -*-Sharel990 * Share1994 13% - . ........ ................ 1 2 % .. ' .... .......... ... .. . . . .. . ..... .. .. . . ............... . . . . . D~ 8%h. co% .l .. ..... .. .... .. . . . . . . . .. .. . . . . . .. .......... . co 6 % .... .. . . . . .. . .. . .. .. .. . . . . .. ... . . . .... .. .. . . . . . .. ...... . ........ 0) (U 4% !. .. .. 2% ± A . .. . ... [~~~ 'I1~~~~ PROKASIH Plants from 3% of BOD share in 1990 to almost 11% in 1994. Such plants would obviously be prime targets for further intervention if so desired. We have noted in Section 3 that BOD load by PROKASIH plants fell significantly from 1990 to 1994. The current results however strongly indicate that the overall reduction in BOD discharges is actually driven by a small number of plants in the dataset. Indeed, some plants have increased their discharges and a large proportion of the plants did not have any significant impact on BOD load (either because of their small size or because indeed their BOD load has not changed). Hence, though one may claim that PROKASIH has been successful at reducing total BOD load in the PROKASIH rivers, a closer examination reveals that the plants' response to PROKASIH varies considerably across plants. 22 Thus far, we have simply compared the BOD discharges in 1990 to the discharges in subsequent years. Since the regulator is primarily concerned with ambient quality of the receiving waters, changes in BOD load is clearly of relevance. Such a comparison ignores however that were it not for the program, BOD discharges could have been much higher than those observed. Indeed, total BOD discharges are a function of both the plant's scale of activity (ceteris paribus, the higher the level of output, the higher the level of discharges), and the BOD intensity (BOD load / total output). Changes in total BOD discharges will reflect changes in both of these parameters. Therefore, an important indicator of the impact of PROKASIH is changes in pollution intensity: constant BOD discharge from a rapidly-growing plant (which can happen only if pollution intensity falls) is clearly a sign of environmental progress. (b) Changes in pollution intensity In this section, we first want to examine changes in pollution intensity, and on the basis of these changes, provide a counterfactual analysis indicating that the impact of PROKASIH on BOD discharges may be larger than measured above. In order to perform this exercise, we must first identify plants for which production data is available and then calculate pollution intensity for each of the years of interest. As noted before, for the purpose of this analysis we are forced to use 1993 as the end year instead of 1994 since we have access to production data only up to 1993. From our 1990-94 dataset, we are able to identify 73 plants in the Statistical data base for which production data is available. For each of those plants we have calculated pollution intensity for each of the year over the period 1990-93, and then normalized the median 1990 pollution intensity to 100. Results are depicted in Figure 7, and are stunning. Despite the increase from 1992 to 1993, the median 23 Figure 7: Changes in Median Value of BOD Intensity (Index 1990=100) 0~~~~ S 1990 1991 1992 1993 BOD intensity fell by approximately 55% between 1990 and 1993. This is by all means very significant. It strongly suggests that considerable pollution control effort can be generated from a subset of plants even in circumstances where resources devoted to monitoring and enforcement activities are lacking. That PROKASIH had an impact on pollution intensity can be supported, to a certain extent, by looking at plants' investment in primary and secondary effluent treatments after PROKASIH was launched in June of 1989. In Figure 8, it appears clearly that installations of wastewater treatment systems increased significantly since 1989. The lack of data does not allow us to link unambiguously this activity to the introduction of PROKASIH. However, such activity is consistent with what we have observed in terms of reduction of pollution load and pollution intensity over the period 1990-94. 24 Figure 8: Installation of New Waste Water Treatment System 18- 16-U Primary Treatment 0 Secondary Treatm:en:t 14 - 4PROKASIH LAUNCHED IN JUNE 1 * * 1 12 _ _ _ 6 -.. 4 -I IE 1980 1982 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 Year of Installation The change in pollution intensity, as observed above, suggests that we may have underestimated the impact of PROKASIH on total BOD discharges. Indeed, let us assume that without PROKASIH, pollution intensity in 1993 would have been the pollution intensity observed in 1990. Then it is easy to show the extent by which we have thus far under-estimated the impact of PROKASIH. This is illustrated in Figure 9. In the figure, the curve labeled BODo (or BODI) represents every combination of intensity and output that yield a BOD load equal to BODo (or BODI). Note that BOD, > BODo. Let us suppose a plant in 1990 with a pollution intensity Io and an output Qo thus yielding a BOD load equals 25 Figure 9: Counterfactual Analysis In ten sity I Qo QI Output to BODo. This is represented by the surface OIoAQo. Let us suppose that this same plant increases its output to Q, in 1993 while decreasing its pollution intensity to II, yielding a pollution intensity equal to BOD1. This is represented by the area 0IIBQ1. If we simply compare BOD1 to BODO, we would conclude, as we did in the previous section, that BOD load increased despite the presence of PROKASIH. However, if we take into account that the impact of PROKASIH on pollution intensity, then this conclusion is wrong. Indeed, if pollution intensity in 1993 had remained at the 1990 level, that is Io, then given the output level Ql, this plant would have produced a total amount of BOD represented by the area OIoCQ1. Hence, the impact of PROKASIH is actually to reduce BOD load by the shaded area IIIOCB. In other words, though we may observe an increase in BOD load in 1993 (from BODo to BODI), without PROKASIH 26 actual BOD discharges would have been even higher than BODI. The impact of PROKASIH can therefore be measured by the difference between "what would have been", and "what is". It is possible to obtain an estimate of the amount by which we underestimate the impact of PROKASIH by looking solely at those plants that have shown a decrease in pollution intensity between 1990 and 1993.27 Of those 73 plants, 40 exhibited a reduction in pollution intensity, and 33 an increase. If we simply calculate the BOD load in 1990 and 1993 for those 40 plants, we observe a reduction of 21 596 kg per day. However, assuming that these plants would not have exhibited a decrease in pollution intensity, this reduction becomes 65 687 kg per day. This therefore indicates that we underestimate the impact of PROKASIH by a considerable margin. Though this is obviously a gross estimate of what may really be the impact of PROKASIH, the purpose of the exercise performed above is to indicate that the simple comparison of BOD load in any two given years is likely to yield a distorted image of that impact. 5. Discussion Our analysis of the PROKASIH experience suggests that there does exist a group of plants that have exerted effort to control pollution emissions despite the absence of a reliable regulatory framework and credible monitoring and enforcement capability. However, our analysis also shows that there also exists plants which despite their participation into 27 For the plants whose pollution intensity has increased in 1993, it becomes hazardous to calculate the extent of the under-estimation. Indeed, if the impact of PROKASIH is to reduce pollution intensity, then we can only conclude that in those circumstances, the intensity would have been even higher than the one observed. However, it would be difficult to identify what would have been that pollution intensity. We therefore prefer simply to ignore these situations and work solely with the plants that have shown a decrease in pollution intensity. 27 PROKASIH, have typically not exerted such effort. One could therefore conclude that pollution control programs of the nature of PROKASIH can only have a limited impact on the emissions of pollutants, and on environmental quality. Though our analysis may indeed support such a conclusion, it does not follow that such programs have no role to play. First, without PROKASIH, it is likely that few or none of the plants would have reduced their BOD load and/or intensity. Moreover, a significant contribution of a program like PROKASIH is to delineate plants willing to exert pollution control effort from those less inclined to do so. This division should provide useful information for BAPEDAL, and set the stage for further and more focused intervention if needed. Finally, the desire to control and monitor closely the environmental performance of a limited number of plants, confronts the regulator to the need of setting and implementing a system by which performance is going to be measured and analyzed reliably. It forces the regulator to confront issues of implementation of the objectives of the program, and more broadly, of the objectives of environmental regulations. Issues of self- reporting, information, inspections, compliance assessment, etc. must be dealt with. Such a compliance system is lacking in most developing countries, and to a large extent, is still lacking in Indonesia. PROKASIH has now reached a point in its development where a certain number of issues have to be dealt with, whether BAPEDAL wishes to focus its effort on current PROKASIH plants or to expand the number of participating plants. Important issues include, among others, the reliability of the data collected by the PROKASIH teams in each of the provinces; the 28 possibility for these teams and BAPEDAL to process and analyze the information; and the monitoring capability of PROKASIH teams. It is important to point out once again that the analysis performed above relies heavily on information provided by provincial PROKASIH teams to BAPEDAL. Though it is possible, as we have done so, to evaluate the quality and reliability of the data reported, the frequency at which PROKASIH teams collect information about the pollution content of the polluters' effluents remains very low. It shall be remembered that out of the 778 plants that have joined the program in 1990 and 1991, only 155 provided sufficient reliable data on which to base our analysis. There is clearly an important work to be done to improve plants' self-reporting as well as PROKASIH teams' data collection system. This is particularly a source of concerns given the possibility of mistakes in the sampling and analysis of the plant's effluent. Similar mistakes can obviously be performed by PROKASIH teams themselves when sampling a plant's effluents. Finally, given the limited monitoring and enforcement capacities of BAPEDAL, the expected costs of under-reporting true emissions levels, or simply avoiding self-reporting may be small. BAPEDAL must face these issues to preserve and augment the integrity of its PROKASIH program. Plants participating in PROKASIH must be expected to submit measures of the quality of their effluents at regular and frequent intervals; these measures must be performed according to a given set of rules to minimize the possibility of sampling errors; the way in which these measures are reported to PROKASIH teams must be standardized so as to minimize the costs of information processing; and PROKASIH teams must have the resources necessary to perform sufficient sampling and analysis to validate the plants' self-reports. To 29 summarize, a compliance management system must be set to collect data from the plants participating in PROKASIH, and to process, analyze and validate the data thus collected. This must be done without significantly increasing the costs for plants which participate in PROKASIH. The reliability of such a management system is crucial to establish PROKASIH's credibility and to allow PROKASIH to achieve fully its role and impact. PROKASIH forces BAPEDAL to confront those issues. Once this framework is in place, BAPEDAL will be in a position to expand its program to other plants. It will also be in position to use a broader mix of instruments aimed at controlling industrial pollution, such as pollution charges. BAPEDAL has recently adopted a program known as PROPER PROKASIH. The purpose of this program is to announce publicly the environmental performance of plants, and in particular to indicate, through a color scheme, how the plant deviates from the environmental standards defined in KEPO3/MENKLH/11/1991. The program was introduced partly as a response to the trend observed in the 1993 and 1994 total discharges of PROKASIH plants. The viability and reliability of this program crucially depends on BAPEDAL's ability to set in place a comprehensive compliance management system. 6. Conclusion In this paper, we have analyzed the impact of Indonesia's PROKASIH program on BOD discharges. We have shown that total BOD discharges from these plants have significantly been reduced since the introduction of PROKASIH. We have also shown that if it were not for PROKASIH, total BOD discharges would most likely have been considerably higher than the 30 levels observed in 1990. However, we have also shown that this overall performance is the result of a very heterogeneous response by a small number of individual plants. Indeed, less than 25% of the plants accounted for the observed reduction in total BOD load; most of the plants did not have any significant impact on overall reduction of BOD discharges. The Indonesian experience suggests that a program like PROKASlH can be a feasible and cost-effective strategy in the initial stages of the development of a comprehensive framework of public intervention to improve environmental quality. They can lead to significant pollution reduction within a relatively short period of time, and at the same time set into motion the development of a compliance management system that is necessary to implement any program aimed at reducing industrial discharges to improve environmental quality. A correct measure of the full impact of PROKASIH remains to be developed. In particular, as pointed out above, a large number of plants have failed to report their emissions. This needs to be improved. Moreover, though the ultimate objective of PROKASIH is to improve the ambient quality of important rivers in Indonesia (or to prevent their further deterioration), we are unable at this point in time to link changes in emissions by PROKASIH plants to changes in environmental quality. The location of PROKASIH and non-PROKASlH plants along every PROKASlH river is known. However, data on discharges of non-PROKASlH plants as well as of non-industrial facilities is clearly insufficient to isolate the impact of discharges by PROKASIH plants. The location of monitoring stations would also have to be modified for this impact to be accounted for. However, given the objective of PROKASIH, its 31 long-term sustainability may very well depend on its ability to demonstrate that ambient quality is improving as a result of the program. Finally, we have not analyzed the characteristics of the plants that have participated in the program (vs. not participated), and of those that have reduced their BOD load (vs. increased). These issues remain the object of on-going research. 32 - APPENDICES 1, 2 and 3 33 Appendix-1 Table 1: Budget of Prokasih Team (rnillions of Rupiah) Province 1990/91 199192 1992/93 1993/94 1994/95 Reo3na] Natio ToU Regoonal Natona Tota RegionW Nabonal Total Regiona Nationl Total Regional Natonal TotU D.I. Aceh -- -- 45.0 45.0 43.0 48.4 91.4 50.5 50.2 100.7 42.0 40.5 82.5 Sumatera Utara 80.0 80.0 90.0 90.0 82.0 58.0 140.0 80.0 52.8 132.8 80.0 27.5 107.5 Riau -- -- 55.0 = 55.0 38.2 47.8 86.0 49.5 40.0 89.5 52.8 38.2 91.0 Sumatera Selatan =- -- 25.0 25.0 100.0 = 100.0 93.0 47.0 140.0 96.0 39.0 135.0 Lampung 50.0 50.0 60.0 60.0 40.0 40.0 90.0 87.0 177.0 102.0 72.5 174.5 D.K.I. Jakarta 265.0 265.0 150.0 150.0 250.0 250.0 250.0 -- 250.0 350.0 25.0 375.0 Jawa Barat 175.0 175.0 375.0 375.0 280.0 280.0 350.0 350.0 300.0 32.0 332.0 Jawa Tengah 65.0 = 65.0 80.0 = 80.0 98.0 = 98.0 130.0 = 130.0 110.0 33.0 143.0 D.I. Yogyakarta -- -- -- -- 90.0 90.0 120.0 120.0 155.0 34.0 189.0 Jawa Timur 755.8 755.8 280.2 280.2 200.0 -- 200.0 250.0 20.0 270.0 250.0 35.0 285.0 Kalimantan Barat -- -- 150.0 = 150.0 75.0 40.0 115.0 50.0 40.0 90.0 85.0 30.0 115.0 Kalimantan Selatan -- -- -- -- -- -- -- -- -- -- 40.0 46.0 86.0 Kalimantan Timur 25.0 = 25.0 45.0 = 45.0 50.0 45.0 95.0 50.0 45.0 95.0 100.0 36.0 136.0 1 Started participation in PROKASIH in 91/92. 2 Started participation in PROKASIH in 94/95. However, the province had established a PROKASIH team in 92/93. 3 In Jawa Timur, the industrial sector must finance the analysis of the samples collected by the PROKASIH team. The budget devoted by the industrial sector for this analysis is as follows (millions of Rupiah): 180 (92/93); 180 (93/94); and 250 (94/95). Another government agency (Parum Jasa Tinta) has also financed a number of monitoring activities. Its budget for 94/95 was 200 millions rupiah. 4 Started participation in PROKASIH in 94/95. 34 Appendix-1 Table 2: Number of PROKASIH Establishments per Cate or Category 1990 1991 1992 1993 1994 Dept. store 0 0 0 5 5 Hospital 0 0 22 34 42 Hotel 0 0 19 41 50 Industry 381 778 1008 1235 1275 Laundry 0 1 3 4 4 Supermarket 0 0 0 1 1 Warehouse 0 0 1 2 2 Workshop 0 2 5 26 26 TOTAL 381 781 1057 1348 1405 Table 3: Average Employment per PROKASIH Plant per Province (1994) Province Number of Total number Employment Employment PROKASIH of plants' per per plant in plant PROKASIH province .______________ ______________ ______________ plant Jawa Barat 723 4833 267 515 D.K.I. Jakarta 228 2289 517 254 Jawa Tengah 64 2915 905 156 Jawa Timur 45 4195 1729 182 Lampung 34 215 468 155 Sumatera 33 270 789 206 Selatan _ _ _ _ _ Kalimantan 31 129 1296 437 Timur _ _ Sumatera Utara 30 1058 613 187 Kalimantan 20 166 909 264 Selatan _ _ _ Riau 19 223 1700 296 D.I. Aceh 17 97 446 153 D.I. Yogyakarta 16 250 549 122 Kalimantan 15 171 618 229 B arat _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ __35 35 Appendix-1 Table 4: Average Budget per PROKASIH Plant (1994) Province Number of plants Total budget Average budget per ________________________ (millions of rupiah) plart (millions) Jawa Barat 723 332 0.46 DKI Jakarta 228 375 1.64 Jawa Tengah 64 143 2.23 Jawa Timur 45 285 6.33 Lampung 34 175 5.15 Sumatera Selatan 33 135 4.10 Kalimantan Timur 31 136 4.38 Sumatera Utara 30 107 3.56 Kalimantan Selatan 20 86 4.30 Riau 19 91 4.79 D.I. Aceh 17 82 4.82 D.I. Yogyakarta 16 189 11.81 Kalimantan Barat 15 115 7.66 Figure 1: Coverage of PROKASIH Plants by KEPIMEN/03/1991 in 1994 |DIn KEP/MEN/03/1991 *Nolhi KEP/MEN/03(1991 397 400 350 328 300 250 B~~~~~~ 2M~~~~~~~~~~~~~~~~1 100 49 8 10 13 7 13 2 13 4 8is 1 14 20 11 23,7 e [. I - .50 F.6, FT~ FI F3 36 Appendix-1 Table 5: Potential Errors in Sampling and Analysis Designing a sampling strategy * Taking samples at locations or times that do not accurately represent the quality of the effluent being sampled. Collecting samples * Using equipment made of inappropriate material that may react with samples and contaminate them. * Using sampling equipment that is not decontaminated between sampling episodes. Handling, preserving, and transporting samples * Improperly refrigerating or holding for too long unstable samples. * Using improper procedures for transporting samples that may result in mismarked or lost samples. Preparing and analyzing samples in the laboratory * Calibrating instruments improperly. * Using incorrect analytical methods to test samples. Interpreting data * Transposing numbers. * Using incorrect formulas. * Misplacing decimal point. 37 Appendix-2 Information Collection by BAPEDAL Figure 1: Information production and collection l BAPEDAL Indicates samrpling and analysis - - - Indicates flow of information _ - - PROKASIH - - - - - - Team s |Plants with 1/\ lants without| |L ab orator Laborato ryl Designated Laboratories Independent . ~~~~~~~~~~Laboratories Upon agreeing with BAPEDAL on a target level of pollution, it is understood that plants will periodically measure the pollution concentration of their effluents, and flow rates (m3 of water/day).tPlants without laboratory facilities must have their samples analysed by independent laboratories or by PROKASIH-designated laboratories; plants with laboratory facilities can also use these independent and designated laboratories. PROKASIH teams collect information from two different sources. First, they have access to the data collected by the plants themselves (whether analysed by independent, designated or plants' laboratories). This is in some sense similar to a system of self-reporting, the difference being that PROKASIH teams must visit the plants and collect the information. Second, they can themselves perform a sampling and analysis of the plant's effluents. The information collected from these two sources is then transferred to BAPEDAL. Reduction in the emissions of total suspended solids are also part of the agreement. However, at the current moment, only BOD data have been systematically collected and computerized. 38 PROKASIH Plants and Sampling Points Brantas River, Jawa Timur I v PT. PERAWANG | P.EWAG |PT. PANCA S. Tapung Kanan PERKASAI ER EKA BINA \\r. NUSANTIA PT. INDAH KIAT | mn | \\ | PT.SUNION PT. RICRY P. PERTIMA SIK PT'. RANTAU PRIMA WUAYA SAKT IIrP KAMPARIWODI 0@ S S T 0@ 0 H{ULU / | S. Tapung Kiri) PTIP n I PT. SOLA TANDUN |GRATIAP. |pm. F SIAKRAYA BUATAN TIMBER TAPUNG GALUH Pr BANGKINANG | Sampling Stations CD Pekan Baru [i Perawang ( Slak Sri Indrapura 39 REFERENCES Arora, S. and T.N. Cason (1994), An Experiment in Voluntary Environmental Regulation: Participation in EPA's 33/50 Program, Journal of Environmental Economics and Management, 28, 3, 271-286. Baumol, W.J. and W.E. Oates (1971), "The use of standards and pricing for the protection of the environment", Scandinavian Journal of Economics, 73, 1, 42-54. Bohm, P., and C.F. Russell (1985), "Comparative analysis of alternative policy instruments", in Kneese, A.V. and J.L. Sweeney (eds), Handbook of Natural Resource and Energy Economics, Volume 1, North-Holland, Amsterdam: 395- 460. Cropper, M. L. and W.E. Oates (1992), "Environmental economics: a survey", Journal of Economic Literature, 30, 2, 675-740. General Accounting Office (1993), Environmental Enforcement: EPA Cannot Ensure the Accuracy of Self-Reported of Compliance Monitoring Data, Report to the Chairman, Committee on Government Affairs, U.S. Senate, GAO/RCED-93-21. Hahn, R.W. (1989), "Economic prescriptions for environemntal problems: how the patient followed the doctor's orders", Journal of Economic Perspectives, 3, 2, 95- 114. Hahn, R.W. and G. L. Hester (1989), "Marketable permits: lessons for theory and practice", Ecology Law Quarterly, 16, 2, 361-406. Hamilton, J.T. (1995), "Pollution as news: media and stock market reactions to the Toxics Release Inventory data", Journal of Environmental Economics and Management, 28, 1, 98-113. Karp, L. and J. Livemois (1994), "Using automatic tax changes to control pollution emissions", Journal of Environmental Economics and Management, 27, 1, 38-48. Kennedy, P.W., Laplante, B. and J. Maxwell (1994), "Pollution policy: the role for publicly provided information", Journal of Environmental Economics and Management, 26, 31-43. Laplante, B. (1995), "It's Not Easy Being Green: the Politics of Canada's Green Plan: a Comment", Canadian Public Policy, forthcoming. Laplante, B. and P. Lanoie (1994), "The Market Response to Environmental Incidents in Canada: a Theoretical and Empirical Analysis", Southern Economic Journal, 60, 3, 657-672. 40 Laplante, B. and P. Rilstone (1995), "Environmental inspections and emissions of the pulp and paper industry in Quebec", Journal of Environmental Economics and Management, forthcoming. Muoghalu, M.I., Robison, H.D. and J.L. Glascock (1990), "Hazardous waste lawsuits, stockholder returns, and deterrence", Southern Economic Journal, 357-70. O'Connor, D. (1994), Managing the Environment with Rapid Industrialisation: Lessons from the East Asian Experience, Development Centre Studies, OECD, Paris. Organization for Economic Cooperation and Development (1989), The Application of Economic Instruments for Environmental Protection, Paris. ---------- (1991), Recommendation of the Council on the Use of Economic Instruments in Environmental Policy, Paris. Russell, C.S. (1990), "Monitoring and enforcement", in P.R. Portney (eds), Public Policies for Environmental Protection, Resources for the Future, Washington, D.C. World Bank (1994), Indonesia: Environment and Development, A World Bank Country Study, Washington, D.C. 41 Policy Research Working Paper Series Contact Title Author Date for paper WPS1585 Public Finances and Economic Luca Barbone March 1996 C. Pelegrin Transition Hana Polackova 85087 WPS1586 Did External Barriers Cause the Azita Amjadi March 1996 S. Lipscomb Marginalization of Sub-Saharan Ulrich Reincke 33718 Africa in World Trade? Alexander Yeats WPS1587 Payments and Finance Problems Constantine April 1996 M. de la Puente in the Commonwealth of Michalopoulos 31206 Independent States WPS1588 Social Insurance in the Transition Deborah Mabbett April 1996 L Bely to a Market Economy: Theoretical 36280 Issues with Application to Moldova WPS1589 The Analysis of Emerging Policy Sudarshan Gooptu April 1996 R Vo Issues in Development Finance 31047 WPS1590 The Latvian Banking Crisis: Alex Fleming April 1996 S. Coffey Lessons Learned Samuel Talley 32535 WPS1591 Competition in Network Industries Michael Klein April 1996 S Vivas 82809 WPS1592 The Stockmarkets as a Source of Cherian Samuel April 199E C. Samuel Finance: A Comparison of U.S and 30802 Indian Firms WPS1593 Rural Finance for Growth and Poverty Saeed Quresh! April 1996 C Anbiah Alleviation Ijaz Nabi 81275 Rashid Faruqee WPS1594 Why Have Some Indian States Done Gaurav Datt April 1996 P Sader Better Than Others at Reducing Martin Ravallion 33902 Rural Poverty? WPS1595 In Search of Owners: Lessons of Cheryl W. Gray April 1996 M. Geller Experience with Privatization and 31393 Corporate Governance in Transition Economies WPS1596 Agrarian Structure in Poland: The Johan van Zyl April 1996 M Williams Myth of Large-Farm Superiority Bill R. Miller 87297 Andrew Parker WPS1597 Egypt and the Uruguay Round Bernard Hoekman May 1996 F. Hatab Arvind Subramanian 35835 WPS1598 Logistical Constraints on Francois Amiot May 1996 L. Lewis International Trade in the Maghreb Ovadia Salama 37369 Policy Research Working Paper Series Contact Title Author Date for paper WPS1599 Economic Implications for Turkey Glenn W. Harrison May 1996 M. Patena of a Customs Union with the Thomas F. Rutherford 39515 European Union David G. Tarr vVPS1600 Is Commodity-Dependence Nanae Yabuki May 1996 G. llogon Pessimism Justified? Critical Takamasa Akiyama 33732 Factors and Government Policies that Characterize Dynamic Commodity Sectors WPS1601 The Domestic Benefits of Tropical Kenneth M. Chomitz May 1996 PRDEI Forests: A Critical Review Kanta Kumari 80513 Emphasizing Hydrological Functions WPS1602 Program-Based Pollution Control Shakeb Afsah May 1996 A. Williams Management: The Indonesian Benoit Laplante 37176 PROKASIH Program Nabiel Makarim