PARTICIPATION, TRANSPARENCY AND ACCOUNTABILITY IN VILLAGE LAW IMPLEMENTATION Baseline Findings from the Sentinel Villages Study 2018 The World Bank – Local Solutions to Poverty, Jakarta, Indonesia This work is a product of the staff of the World Bank, through the Local Solutions to Poverty program, with external contributions. The findings, interpretations, and conclusions expressed in this work do not necessarily reflect the views of The World Bank, its Board of Executive Directors, 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 or acceptance of such boundaries. Rights and Permissions The material in this work is subject to copyright. Because the World Bank encourages dissemination of its knowledge, this work may be reproduced, in whole or in part, for noncommercial purposes as long as full attribution to this work is given. Nothing herein shall constitute or be considered to be a limitation upon or waiver of the privileges and immunities of The World Bank, all of which are specifically reserved. Graphic Designer: Ardhi Yudho PARTICIPATION, TRANSPARENCY AND ACCOUNTABILITY IN VILLAGE LAW IMPLEMENTATION Baseline Findings from the Sentinel Villages Study October 2015 – February 2016 Writer Leni Dharmawan Gregorius D.V. Pattinasarany Lily Hoo Editor Peter Milne iv Contents Contentsv Tables vi Figures vii Boxes viii Abbreviations and Acronyms ix Abstractxi 1. Background 1 2. Objectives and Research Questions 5 3. Methodology and Locations 7 3.1. Methodology 7 3.2. Site selection and sampling approach 8 4. Conceptual Framework 11 4.1. Participation 11 4.2. Transparency and accountability 13 5. Key Baseline Findings 15 5.1. Participation in village planning and project implementation 15 5.2. Transparency 21 5.3. Responsiveness and accountability 25 5.4. Role of village activists and the BPD 28 6. Key Takeaways 31 7. Looking Forward 33 References34 Annexes Annex 1 : Characteristics of invitees and noninvitees, and their attendance at village meetings 36 Annex 2 : Village and hamlet meeting attendees and non-attendees 37 Annex 3 : Cited reasons for non-attendance and probable conditions for future attendance 38 Annex 4 : Probit estimates of attending village meetings (marginal effects) 39 Annex 5 : Respondents' characteristics by their awareness of last year's village programs and fund use 40 Annex 6 : Desire to know village information 41 Annex 7 : Sampling weight approach 42 v PARTICIPATION, TRANSPARENCY AND ACCOUNTABILITY IN VILLAGE LAW IMPLEMENTATION Tables Table 1. Selected characteristics of the qualitative sites 8 Table 2. Participation in infrastructure construction activities 16 Table 3. Perceptions toward village heads and hamlet heads (%) 18 Table 4. Village heads' previous experience and villagers' participation 20 Table 5. BPD's profiles and villagers' participation 20 Table 6. Village heads' perceptions of information needed by villagers 22 Table 7. Types of information villagers most often requested 22 Table 8. Viilage heads' perceived information needed and types of information most often requested by villagers 23 Table 9. Subjects of interest and villagers' characteristics 23 Table 10. BPD members' previous experience and their direct election, and villagers' awareness 24 Table 11. Village heads' dissemination and villagers' awareness 25 Table 11. Top three problems and solutions: views of villagers 25 Table 13. Complaint handling by village governments 26 Table 14. Activists' participation in deliberative meetings 28 Table 15. Correlation on problems cited by villagers, village activists and hamlet heads 29 vi PARTICIPATION, TRANSPARENCY AND ACCOUNTABILITY IN VILLAGE LAW IMPLEMENTATION Figures Figure 1. Participation rate in village and hamlet level meetings by district 16 Figure 2. Type of engagement during village meetings (%) 19 Figure 3. Type of engagement during hamlet meetings (%) 19 Figure 4. Villagers' awareness of village programs and the use of village funds 21 Figure 5. Perception on importance of village programs 26 vii PARTICIPATION, TRANSPARENCY AND ACCOUNTABILITY IN VILLAGE LAW IMPLEMENTATION Boxes Box 1. The total number of respondents 9 Box 2. Accountability meeting in Ndona Village, Ngada 27 viii PARTICIPATION, TRANSPARENCY AND ACCOUNTABILITY IN VILLAGE LAW IMPLEMENTATION Abbreviations and Acronyms APB Desa : Anggaran Pendapatan Belanja Desa (Village budget) BPD : Badan Permusyawaratan Desa (village council) CDD : Community-Driven Development FGD : Focus Group Discussion KDP : Kecamatan Development Program (Program Pengembangan Kecamatan, PPK) LLI : Local Level Institutions Musdes : Musyawarah Desa (village deliberation forums) Musdus : Musyawarah Dusun (hamlet deliberation forums) Musrenbangdes : Musyawarah Perencanaan Pembangunan Desa (Village-level planning meetings) PNPM : Program Nasional Pemberdayaan Masyarakat (National Program for Community Empowerment) PSF : PNPM Support Facility RPJM Desa : Rencana Pembangunan Jangka Menengah Desa (Village Mid-Term Plan) RT : Rukun Tetangga (neighborhood unit) RKP Desa : Rencana Kerja Pemerintahan Desa (Village Annual Work Plan) RW : Rukun Warga (community unit) SV : Sentinel Villages (Village Government and Community Empowerment Study) VL : Village Law ix Abstract PARTICIPATION, TRANSPARENCY AND ACCOUNTABILITY IN VILLAGE LAW IMPLEMENTATION Abstract Can the Village Law improve socioeconomic an issue because, even when they were invited, conditions in Indonesia’s villages through better more than half of the villagers did not attend. participation and governance? This is the central Village heads also selectively invited members question that the Sentinel Villages study sets out of the elite, community leaders and activists, and to answer. The study observes the first two years those whom they thought would be willing to speak of Village Law implementation to assess villagers’ and able to provide inputs. Village councils, which participation, the transparency and accountability had not yet been formed in accordance with the of village governments, and the influence of good Village Law, did not demonstrate their potential to governance principles on village decision-making improve villagers’ engagement in decision-making processes for development investments. At the and control over village governments. However, start of Village Law implementation, participation village activists’ concerns about local issues were in village decision-making was still dominated by more in line with village households, and women elites and men, particularly at village-level meetings, activists were almost as vocal and active as men while sub-village- or even neighborhood-level activists. Encouraging participation in sub-village meetings were more accessible to women and meetings and promoting village activists to voice poorer people from the bottom 40 percent. Villagers villagers’ concerns in village meetings may be an tended not to participate largely because of the effective way of offsetting the dominance of village high opportunity costs and the perception that the governments and village elites. discussions only concerned village government and community leaders. Being invited was less of Keywords: village law, village governance xi 1. Background PARTICIPATION, TRANSPARENCY AND ACCOUNTABILITY IN VILLAGE LAW IMPLEMENTATION Background Law No. 6/2014 on Villages (“the Village Law”, approach often resulted in a mismatch between or VL) provides opportunities to improve village what was needed by the community and what was governance in Indonesia by incorporating good provided by the government, and villages had little if governance principles of community participation, any control over their own development as they had transparency and downward accountability, and few resources with which to manage development providing additional resources and autonomy themselves. In the new circumstances following the to villages. These principles have already been end of the New Order era, CDD projects piloted practised through community-driven development through the Kecamatan Development Program (CDD) projects for more than 15 years in villages (KDP) aimed to provide communities with the across the country. The principles are based on opportunity to address their own development the premise that empowering citizens to choose or needs. This was achieved by providing space demand the goods/services they need will improve for communities to meet and propose their own their wellbeing. priorities, and by providing the necessary funds and technical support to implement the proposals. To CDD projects first started in Indonesia after the end ensure that funds were received by communities of the New Order era. Under the New Order regime, in full and in good time, the projects had their own villages were tightly controlled by higher levels management and accountability mechanisms, of government that decided which development and did not rely on the existing systems of village projects they could have. This highly top-down government. 1 PARTICIPATION, TRANSPARENCY AND ACCOUNTABILITY IN VILLAGE LAW IMPLEMENTATION Following the end of the New Order era, for a brief period Law No. 22/1999 on Regional Government enabled villagers to hold the village government accountable, and hence be more responsive to the needs of the community, by establishing an elected village council to represent the community. However, Law No. 22/1999 was short-lived and its replacement, Law No. 32/2004, vastly reduced downward accountability mechanisms by allowing village heads to appoint village council members themselves, giving more power to the village heads with almost no other village institutional control. This is the context in which CDD projects—merged into an umbrella program called the National Program for Community Empowerment (Program Nasional Pemberdayaan Masyarakat, or PNPM) after 2007—operated during most of the projects’ lives. Because they had separate management and accountability systems, such projects had limited influence on how village government operated, despite having had a presence for more than a decade at the village level (Dharmawan, Dewayanti, & Nugraheni, 2014), (Syukri, Mawardi, & Akhmadi, 2013), (Woodhouse, 2012). Nonetheless, PNPM was successful in providing good quality and cost- effective village infrastructure, reducing poverty and improving access to services, with minimal leakages (PNPM Support Facility, 2014a), (Syukri, Akhmadi, Hastuti, Kartawijaya, & Kurniawan, 2014), (Syukri Mawardi, & Akhmadi, 2013), (Voss, 2013), (Voss, 2008). Some viewed PNPM as a cost-effective CDD tool that helped to shift more of the funds to the beneficiaries, rather than as a means of social transformation (Mansuri & Rao, 2013). The hope now is that, with the passage of the Village Law and the greater provision of funds and its stipulation of good governance practices, there will be a significant improvement in the quality and results of village development. 2 PARTICIPATION, TRANSPARENCY AND ACCOUNTABILITY IN VILLAGE LAW IMPLEMENTATION The VL’s multiple accountability mechanisms include returning power to the revitalized village council (Badan Permusyawaratan Desa, or BPD) as community representatives, instituting village deliberation forums (Musyawarah Desa, or Musdes) to enhance general community participation, and providing transparency on government operations and reporting to district governments. However, given that many village governments are now managing increasingly significant financial resources with only limited capacity in good governance principles, concerns have been raised over the potential misuse of funds, the misalignment of priority development needs between village governments and the communities they serve, and the increasing exclusion of marginalized groups from the development process. Hence, it has become important to observe how these good governance principles are being practiced, especially in the early years of VL implementation. This Sentinel Villages (SV) study started its baseline fieldwork with a qualitative component in 10 villages in September to October 2015, and the quantitative component in another 112 villages in the same districts in March to April 2016 (details on the methodology used are provided in Chapter 3). This report provides overview findings from the baseline study on participation, the transparency of information, and accountability mechanisms. For a more extensive report on the qualitative work, see Kurniawan, Sedyadi, Kartawijaya, Syukri, Bachtiar, Diningrat & Alifia (2017). 3 2. Objectives and Research Questions PARTICIPATION, TRANSPARENCY AND ACCOUNTABILITY IN VILLAGE LAW IMPLEMENTATION Objectives and Research Questions The study tracks VL implementation progress in the women, poor villagers and marginalized first three years of the law’s implementation (2015-2018) groups, informed about: (i) village government with the following objectives:1 planning processes; (ii) village government decisions; and (iii) the implementation of 1) To examine whether VL implementation is projects supported through village funds? Why following the stipulated principles of participation, or why not? transparency and accountability in village c. Is the community able to hold the village governance processes; government to account for the use of village 2) To observe whether VL implementation is leading to funds through mechanisms stipulated in the more responsive village government, as reflected VL, such as the BPD and Musdes? Why or why in the decisions that correspond to community not? priorities; and 2) Does implementing the principles of participation, 3) To examine whether the existence of local transparency and accountability lead to village institutions (such as the BPD and/or adat councils) fund allocation that corresponds with the and village activists (such as former PNPM actors) community's priorities? influence the implementation of the VL.2 a. Are village funds allocated according to priorities that reflect the needs of non-elites, To examine the implementation of the VL, the study including those of women, poor villagers and sets out to answer the following questions: marginalized groups? Why or why not? b. Do community members, including non-elites, 1) To what extent are villages implementing the women, the poor and marginalized groups, stipulated principles of participation, transparency perceive changes in their interactions with the and accountability? village government after VL implementation? a. Are planning and implementation processes open Why or why not? to non-elites, including women, poor villagers and 3) Do existing local institutions (such as the BPD marginalized groups? Why or why not? and/or adat councils) and village activists b. Are community members, including non-elites, (such as former PNPM actors) influence VL implementation? What role do they play, if any? 1 The study has now been extended to 2018, one year longer than originally planned.. 2 At the time of the fieldwork, regulations on adat villages had not yet been established and none of the sampled villages had formally been declared an adat village. In the qualitative study, adat was still strong in community life but did not play a significant role in village government. Adat groups were treated similarly to other community groups. One village was known to have a separate adat organization to manage its adat forest, which was distinct from the village government. This created the potential for conflict between the two organizations. Further results from the research will be reported in the end-line study in 2018. 5 3. Methodology and Locations PARTICIPATION, TRANSPARENCY AND ACCOUNTABILITY IN VILLAGE LAW IMPLEMENTATION Methodology and Locations 3.1. Methodology as well as to gain a better sense of how villagers interacted with each other and with the village The study employs both qualitative and quantitative government in daily life. Three types of FGDs were methods to examine changes in how participation, also conducted during the baseline fieldwork: (i) on transparency and accountability principles are village governance; (ii) on village institutions and being put into practice, what factors influence key actors; and (iii) on the responsiveness of village these practices, and how villagers and village governments. The first two FGDs were conducted governments both perceive these changes from separately for male and female participants, their own perspectives. The qualitative method is while the third FGD was mixed. Overall, over 400 used to obtain an in-depth understanding of the villagers participated in the FGDs across all the relationship between the various factors, while the study locations. In addition, various interviews were quantitative method is used to illustrate the patterns conducted with key informants from the district, of these practices, together with their origins and sub-district and village levels, including interviews the perceptions of them among the community. with marginalized groups, to gain information from Both components involve baseline and end-line various stakeholders involved in VL implementation fieldwork. During the fieldwork, the qualitative on their experiences, challenges and expectations. component required a field observer to be placed in each district to collect information on related issues The quantitative part of the study surveyed over of VL implementation. 4,000 respondents of both genders, including village heads, BPD heads and village activists in 112 The qualitative part of the study was conducted villages (see 3.2 for details). The respondents were using various data-collection techniques, such as a mixture of household heads and members, village direct observation, focus group discussions (FGDs) heads, hamlet heads, village council members/ and in-depth interviews. Direct observations were head, village activists, and health and education used to gain an understanding of the specific sector workers. The surveys asked questions characteristics of the villages, including their to gain an understanding of these respondents’ geographical conditions and natural resources, involvement in VL implementation, their experiences 7 PARTICIPATION, TRANSPARENCY AND ACCOUNTABILITY IN VILLAGE LAW IMPLEMENTATION and understanding of VL implementation, their datasets only go down to the sub-district level. satisfaction with the services and information, and Given these constraints, the study used locations their perceptions about their village's priority needs. from the Local Level Institutions (LLI) studies, another longitudinal study conducted in 1996 3.2. Site selection and (LLI1), 2000/01 (LLI2) and 2012 (LLI3). These studies sampling approach sought to identify the preconditions for, and constraints on, local capacity (defined as the ability The study was carried out in locations that provided to resolve common problems collectively) and the a variety of characteristics of rural Indonesia that extent to which state structures complemented could have an influence on village governance in or impeded communities’ problem-solving efforts order to obtain a collection of detailed case studies. that fit the criteria as described in the methodology However, these locations were not intended to be (poor and resource-rich, Java and off-Java, strong representative of the whole country. Instead, they and weak local institutions). These provinces were were limited to include resource-rich and resource- Jambi, Central Java and East Nusa Tenggara (NTT). poor provinces, Java and off-Java, and strong and The LLI datasets provide additional advantages for weak local (formal/state, community/adat/religion- the study site selection as: (i) they have data at three based) institutions. different points in time over the past two decades (LLI1 in 1996, LLI2 in 2001, and LLI3 in 2012) that There are few accessible datasets from which the reinforce the longitudinal nature of this proposed study could draw samples that included micro-data, study; and (ii) they are the only datasets available such as the level of participation in village-level that provide information on the key indicators activities, and perceptions of transparency and that this study looks into, namely participation, accountability of village governments. PNPM transparency and accountability at the village level. Table 1. Selected characteristics of the qualitative sites Distance to APB Desa 2015 District Villagea Population size Poverty rateb subdistrict (Rp) Ngada Ndona 15 km 1,378 24.27 579,177,912 Lekosoro 16 km 913 20.41 576,132,552 Wonogiri Kalikromo 2 km 2,785 27.13 803,827,000 Beral 8 km 3,366 8.48 1,104,514,000 Banyumas Deling 3 km 4,836 18.73 939,912,188 Karya Mukti 3 km 13,038 24.02 1,802,637,497 Batanghari Tiang Barajo 15 km 1,965 9.23 856,953,280 Kelok Sungai Besar 16 km 2,087 13.58 843,110,280 Merangin Jembatan Rajo 3 km 1,261 3.21 383,213,333 Seberang Sungai 4 km 755 42.54 375,451,431 Source: Village Profile and APB Desa. a) All are pseudonyms; b) For consistency across all sites data were obtained from SMERU’s poverty and livelihood data map of 2010 at http://www.indonesiapovertymap.org 8 PARTICIPATION, TRANSPARENCY AND ACCOUNTABILITY IN VILLAGE LAW IMPLEMENTATION Hence, the study revisited five LLI districts: two in and Ngada. For logistical reasons, four villages Jambi (Batanghari and Merangin), two in Central were visited in each sub-district. Adjustments were Java (Banyumas and Wonogiri) and one in NTT made to ensure that the number of villages per (Ngada). For the qualitative work, 10 villages were district was a multiple of four. The sub-districts, selected from a set of 20 villages visited during villages and hamlets—one in each village—were the third round of the LLI studies in 2012, using the selected randomly. Based on the most recent list following criteria: of households provided by the hamlet head, 20 households were selected randomly and in each • Variations in participation level in village household two adult respondents (a man and a development activities and perceptions of woman) were interviewed. transparency and accountability (from LLI dataset); Box 1. The total number of respondents was: • Variations in village capacity for collective action (from LLI dataset); and • 2,240 households, comprising 1,841 households • Variations in village potential, such as resources, represented by two respondents and 399 access to infrastructure and access to markets households by one respondent. (from Potensi Desa, or PODES). • 4,081 adult household member respondents, comprising 2,125 women and 1,956 men. For the quantitative component, initially the plan • 112 village heads. had been to pick 100 villages outside the qualitative • 112 BPD heads. study sites. The number of villages in each district • 112 hamlet heads. was determined by the proportion of total villages • 222 community activists, 224 health sector workers in each district. However, in order to maintain the and 192 education sector workers. proportions, 12 villages were added in Merangin 9 4. Conceptual Framework PARTICIPATION, TRANSPARENCY AND ACCOUNTABILITY IN VILLAGE LAW IMPLEMENTATION Conceptual Framework 4.1. Participation While there have been some exhaustively cited successes [such as participatory budgeting in Participation is one of the key principles in the VL. Porto Alegre, Brazil (Baiocchi, 2003), but for more Article 3 of the VL lists 13 principles as the basis toned-down praise, see Boulding & Wampler, 2010], for village management (pengaturan desa), one participatory projects often continue to favor elites of which is participation. Participation in decision- who are “wealthier, more educated, of higher social making has been recognized as an important status, male and more politically connected than aspect in development programs in Indonesia, non-participants” (Mansuri & Rao 2013:5), see also especially after years of experience in implementing Bandiera and Levy (2011), and Dasgupta and Beard community-based development programs such as (2007). Participation may also inflict financial and the Kecamatan Development Program (Program social costs on poor and marginalized groups, and Pemberdayaan Kecamatan, or PPK) and PNPM on women.3 Positive impacts for these groups are (Wetterberg, 2014). The underlying assumption is often limited or highly dependent on the context that opening decision-making processes to include (Joshi, 2014), (Mansuri & Rao, 2013), (O'Meally, 2013). a wide range of actors will lead to more broadly shared and sustainable development outcomes. Even in projects that make participation compulsory Particularly in those contexts where non-elites and protect the processes against the village have been previously excluded, the inclusion of government’s (and the local elite’s) intervention, the community’s voice is expected to improve the such as PNPM, participation quality varies and village government’s performance (Clearly, 2007), confirms some of the concerns raised earlier. Data (Narayan, 2002). from PNPM Rural show that women and the poor had considerable involvement in the program, with Recent reviews of participatory approaches, women making up 45 percent of those engaged, however, show that participation does not always and with 50 percent of participants categorized lead to better and more equitable outcomes. as poor. The poor were also heavily involved in 3 See Sambodho (a, forthcoming paper, as part of this study) 11 PARTICIPATION, TRANSPARENCY AND ACCOUNTABILITY IN VILLAGE LAW IMPLEMENTATION sub-project implementation (mainly construction This horizontal pressure is needed given that the work), with more than 70 percent of the workers for pressures from above are not always effective or PNPM Rural infrastructure drawn from the poorest available, and neither are the pressures from below, segments of the village (Pokja Pengendali PNPM, as is also found in villages in Indonesia (Wetterberg, 2012). However, there are also reports that women Jellema, & Dharmawan, 2014). and the poor rarely participated in decision-making, which remained dominated by local elites (Neil, Using this knowledge of participation, this baseline 2013). Meanwhile, marginalized groups usually study looks at who participates in the decision- remained excluded from participation (Syukri et al., making process and the implementation of village 2013), (AKATIGA, 2012), (AKATIGA, 2010), and there projects funded by the significant increase in village was limited citizen empowerment and ownership in budgets, how they participate, and their perceptions remote and marginal areas of the poorest regions of the benefits. In particular, the study discusses the (PSF, 2015), (Neil, 2013). Nonetheless, in terms of following areas: outcomes of the participation process, the majority of PNPM beneficiaries (around 90 percent)— • Musdes: Village deliberation forums figure regardless of gender or poverty level—agreed that prominently in the VL as a means of involving the they had benefited from the investments in PNPM community to consider strategic matters in village block grants (PNPM beneficiary surveys, 2012 and government (Article 54). Outcomes of these 2015). deliberations should be referenced by the village government in the execution of their duties Elite capture certainly exists but is limited, especially (elucidation of Article 54), making the Musdes an among the informal leaders (vis-à-vis those in formal important decision-making body. leadership positions in the village), and so are • Musyawarah Dusun (Musdus): Usually hamlet the welfare losses it creates (Alatas, et al., 2013). deliberation forums precede and feed into the Another study distinguishes further between elite village forums. This is the forum that is physically control (over decision-making) and elite capture (of closest to the villagers and is attended by close the benefits), and finds that elites behave differently neighbors, and we compare the Musdus with the in different contexts—“not all elites who had power Musdes. were corrupt” and, in cases where they controlled • Facilitators: Many past CDD projects were highly the decisions, the benefits still went to the most reliant on facilitators in order to ensure that the deserving groups (Dasgupta and Beard, 2007:244; participatory principles were followed, bringing see also World Bank, 2017). different groups of people together and bridging them to outside resources. The VL also provides There is a need to capitalize on informal leaders and strong support for facilitation, both from the village activists, who are generally (but by no means community, as well as externally.5 always) among the better off in the village.4 A village • Participation in project implementation: The VL governance project in Zimbabwe experimented in specifies that the village community should be utilizing these informal leaders to create horizontal involved in implementing development projects pressure, arguing that they create leadership included in the village’s annual work plan (Article competition and increase monitoring, among 81). others (Baldwin, Muyengwa, & Mvukiyehe, 2017). 4 Following a study on PNPM marginalized groups, we define village activists as those with knowledge of government projects and who use the knowledge to be involved in later projects. They are not village government officials but have close relations with government officials and they are not necessarily the village wealthy. These activists include the cadres of health centers, government-led women’s groups, and farmers’ groups (AKATIGA, 2010). 5 At the time of the data collection, most facilitators were not available for various reasons. We will collect data on them at the end-line survey and in between (qualitatively). 12 PARTICIPATION, TRANSPARENCY AND ACCOUNTABILITY IN VILLAGE LAW IMPLEMENTATION 4.2. Transparency and accountability, villagers need a support system accountability through facilitation, as recognized in the Village Law. Apart from the delayed empowerment of Transparency is intended to drive accountability, the BPD, facilitation is largely unavailable in many as citizens can use the disclosed information to villages. voice their concerns over budget discrepancies or unfulfilled development plans. While sharing The VL builds on PNPM’s transparency and information (on government decisions, budgets, or accountability mechanisms. In particular, it service standards, for example) is not enough on its stipulates community monitoring by ensuring that own to ensure that the state is complying with its the community can solicit and receive information stated priorities, it does nonetheless give citizens from the village government, as well as oversee the means to hold state actors to account (Grindle, activities related to governance, development 2007). Fox (2007) conceptualizes a spectrum implementation, guidance and community ranging from transparency to accountability. By empowerment (Article 68). In addition, there is participating in decision-making, citizens may gain a specific article regarding the right to monitor the right to question the state’s past performance, development plans and activities (Article 82). The which produces a soft version of accountability. same article also obligates the village government However, “answerability without consequences to report on planning and implementation of the falls short of accountability” (Fox, 2007:668). It is RPJM Desa and APB Desa at least annually through only when officials and providers face “sanctions the Musdes. with teeth” (Joshi 2014:26) for shortcomings in the fulfillment of their responsibilities that hard In addition, the VL goes beyond PNPM’s accountability is evident. In short, information accountability mechanisms by strengthening the disclosure is an important element to push for BPD. The VL and associated legislation consistently accountability, but accountability is effective only state that the BPD must be chosen democratically, when it is backed up by “sanctions with teeth”, when which is an important shift toward re-establishing appropriate. the body’s independence from the village head (as per Law No. 32/2004). Furthermore, the Such hard accountability has been proven to be BPD’s functions include overseeing and soliciting effective in Indonesia, once again referring to the information from the village government, proposing country’s major CDD project, PNPM, through its draft village regulations, channeling community internal and external financial monitoring that led aspirations, and following democratic principles and to a low corruption level of below 1 percent in its gender equity (Articles 55, 61-63). The BPD should overall disbursements (PSF, 2014b), (Woodhouse, also play an important role in village planning, as the 2012), (McLaughlin, Satu, & Hoppe, 2007). The VL organizer of the Musdes. aims to emulate this oversight for accountability through “a threefold accountability structure: With the incorporation of such good governance horizontally, to an empowered BPD; downward, principles (albeit limited, i.e., mostly upward to the public, through a newly introduced village accountability), which was not the case in the assembly; and upward, to the district government” past, it will be interesting to see whether or not (Antlov, Wetterberg, & Dharmawan, 2016). The village governments will implement these legal community can solicit and receive information from requirements (and how they do so), and to measure the village government to monitor its activities. the impacts on villagers’ levels of satisfaction and However, to be able to exercise the demand for wellbeing. 13 5. Key Baseline Findings PARTICIPATION, TRANSPARENCY AND ACCOUNTABILITY IN VILLAGE LAW IMPLEMENTATION Key Baseline Findings In this study, we hypothesize that village 5.1. Participation in village governance—including participation, transparency planning and project and accountability—will improve once the BPD implementation and facilitators are active in providing checks and balances, and promoting good governance, In broad terms, the baseline survey categorizes two referring to the findings of the LLI studies types of participation: participation in implementing (Wetterberg et al., 2014). The baseline data development activities/projects (usually in the form for this report were collected at the start of VL of contribution of labor) and participation in planning implementation, when some of the key regulations and budgeting processes (or decision-making) and mechanisms were not yet in place, such as in village and hamlet meetings. Participation in those on BPD selection and the BPD’s role and decision-making meetings was generally lower responsibilities. When the survey was conducted, and less inclusive than in implementing activities the BPD had not yet been modeled in accordance (especially contributing labor), which were more with the VL and, in many villages, facilitators were inclusive—by gender, welfare group and district. also not in place. In this context, the baseline study Construction activities, especially those in villagers’ was able to capture the conditions before the own neighborhoods or hamlets, were known to VL was fully implemented. We expect that, as VL 84.7 percent of the respondents.7 About two-thirds implementation continues, we should be able to see of them (66.4 percent) said that they had been differences in the end-line study (to be fielded in involved in these activities during the past two years early 2018) to answer all three research questions of (Table 2). Ngada and Wonogiri topped the list. The this study. In this baseline report, we mainly answer most common form of involvement was in labor the first research question and some of the third provision, either for free or for pay (85.7 percent), question, to present a portrait of governance in in which villagers, mostly men (97.4 percent men vs village development activities.6 70.4 percent women), work to build, improve and maintain local infrastructure, such as village roads, 6 See Kurniawan et al. (2017) on the qualitative baseline report for more wide-ranging discussions 7 All figures in this report are weighted (see technical notes in Annex 7). 6 See Kurniawan et al. (2016) on the qualitative baseline report for more wide-ranging discussions. 15 PARTICIPATION, TRANSPARENCY AND ACCOUNTABILITY IN VILLAGE LAW IMPLEMENTATION trenches and local irrigation channels. With prior in particular, were driven by specific policies/ collective agreement, they could also forgo their programs of the district governments. Ngada was wages, either partially or in full, as their contribution able to maintain its high participation rate through toward expanding the size of the project. Donating its PNPM-like district program, known as Pelangi money, the second-most-common form of Desa, which started before the end of PNPM and participation, was much lower (33.3 percent). So continues operating to date. In this sense, the PNPM were other forms of participation, such as donating model has never really disappeared in this district. building materials (3.1 percent) and providing land Batanghari followed a different path to encourage (2.5 percent). Labor appeared to be most readily participation. There the district government available form of contribution. provided funds in Alokasi Dana Desa (transfers from the district government) for transportation Participation in decision-making on development allowances for villages to hold village-level planning plans is observed, but to a lesser degree than meetings (Musrenbangdes, or Musdes) for up to 70 participation in labor. Participation, as the participants. qualitative study found, took place mostly during the compilation of the village’s mid-term plan (RPJM Desa), which was developed during the last year of Figure 1. Participation rate in village PNPM (2014) and hence was facilitated by PNPM and hamlet level meetings by district facilitators. The process started at the RT- or hamlet- level discussions and these were well attended. Ngada Once the RPJM Desa had been issued, villagers’ participation decreased, as discussions of the Batanghari annual plan were more limited to the village leaders (including the RT heads) and community figures. Of Merangin Village Hamlet our five districts, the survey shows that Ngada had Banyumas the highest participation rate, more than double that of Batanghari, which came second (Figure 1). Wonogiri The performances of Ngada and Batanghari, 0.0 20.0 40.0 60.0 80.0 Table 2. Participation in infrastructure construction activities Districts Variable All obs Merangin Batanghari Banyumas Wonogiri Ngada F-test Participated in local 66.4 28.7 18.4 65.7 79.6 84.4 121.2 ** infrastructure activities 3,365 602 377 1,005 852 529 Forms of villagers participation: -- Labor/work 85.7 84.8 96.6 80.6 91.1 96.5 3.9 ** -- Money 33.3 17.6 0.0 36.2 31.8 32.5 66.9 ** -- Materials 3.1 7.9 0.9 3.2 2.4 6.8 1.1 -- Land 2.5 8.4 3.4 3.7 0.5 1.4 1.9 2,065 166 67 683 700 449 Notes: **, * F-test on the equality of means across kabupaten is statistically significant at 5 and 10 per cent, respectively 16 PARTICIPATION, TRANSPARENCY AND ACCOUNTABILITY IN VILLAGE LAW IMPLEMENTATION In general, hamlet meetings were more popular large enough to accommodate all or a large number than village meetings, as the hamlet is in the of villagers; and (iii) whether villagers were too closest proximity to where villagers live. Around busy to participate. No specific efforts were made 27 percent of respondents attended hamlet-level to encourage poor or marginalized groups to meetings, while only 16 percent attended village- participate in such meetings. level meetings in the past year. Wonogiri stood out as the second-highest district (after Ngada), Only in some villages were invitations given to with 44.4 percent of the villagers participating all villagers through public announcements. In in hamlets meetings—four times higher than the Ngada, for example, meetings were mostly open attendance rate at village meetings. This district to the public: all village heads claimed that they has a long tradition of community gatherings held invited everybody to the village annual planning every 35 days, known as selapanan, to discuss and budget meetings. Usually, the invitations were various hamlet issues—often proceeded by Quran announced after Sunday mass in church. A similar recitals/prayers and arisan dusun (hamlet-level process took place in Merangin, as observed in one rotating funds). Men and women held separate village (Seberang Sungai) in the qualitative study. selapanan at different times. When excluding both Here, no formal invitation was issued, as upcoming Ngada and Wonogiri, which increased the overall meetings were announced over the mosque’s participation rate, the participation rate at hamlet speakers, serving as an open invitation to all. But meetings drops to only 15.9 percent. Similarly, unlike Ngada, only a few people came to meetings excluding Ngada drove down the participation rate and those who did were mostly men. People of village meetings to 14.2 percent. Only 7.9 percent generally said they were too busy working and of respondents stated that they had attended both were not willing to incur the opportunity costs of meetings. In Banyumas, where villages have a much attending. Open information by itself is not sufficient larger population size (Table 1), meetings at sub- to bring people to meetings. hamlet or neighborhood level (RT/RW), attracted more participants (as observed later), which might A closer look at the data indicates that those explain low attendance rates at village and hamlet being invited to village meetings, compared levels. with those who were not invited, comprised the better off in the community, those active At first, being invited appeared to be an important in organizations, and those who had a positive factor that led villagers to attend village meetings, opinion of the village government. They comprised as 81.7 percent of the attendees said they had 38 percent of the villagers. These participants were been invited. However, only 44 percent of village more likely to be men, currently working, of the heads sent out invitations to villagers to discuss majority ethnic group, active in local organizations the village annual work plan (RKP Desa). An even and/or political parties, and had higher educational lower proportion of village heads (36 percent) attainment (see Annex 1). In addition, participants invited villagers to the annual budget (APB Desa) perceived the village leadership positively—that discussion, which was considered to be more the village head, hamlet head and BPD head were technical and complicated. Invitations turned out reliable in planning the village development and to be given selectively, indicating that village capable of executing the plans. Interestingly, those meetings were not equally open to all villagers. who had submitted complaints or reported problems Village governments limited not only the number (9 percent of respondents)—not necessarily critical of invitees but also whom they invited, and justified of the village government—were also more likely their actions based on: (i) whether villagers were to be invited to meetings. The village heads already appropriately represented by their hamlet confirmed that they invited villagers who they felt heads, community and/or religious leaders in the actively voiced their concerns or provided input meetings; (ii) whether there was a meeting venue to the village government, actively participated in 17 PARTICIPATION, TRANSPARENCY AND ACCOUNTABILITY IN VILLAGE LAW IMPLEMENTATION village activities and provided assistance for such to differentiate households by welfare distribution events. This deliberate choice by village heads (see below). In addition, the study findings show that indicated that they valued input from those who had in comparison to participating respondents, the non- concerns over village affairs or were more willing to participating groups usually had a lower opinion of participate in village activities. The same reasoning the village government in three areas: (i) the village was confirmed by the qualitative study. The village government’s reliability in making and executing heads not only looked for input, but were also trying development plans; (ii) the village government’s to reduce opposition and avoid conflicts. reliability in providing access to information; and (iii) perceived efforts by the village government to Conversely, the poor, women, and those who resolve villagers’ daily problems and perceived were less active, were less likely to be invited handling of their complaints.9 Villagers will not be and to attend village meetings.8 These findings motivated to participate if they think that village were consistent with the results of the LLI3 survey government is not reliable or able to respond to in 2012, indicating that there was hardly any their needs (Sambodho, a, forthcoming). change between the two surveys (Wetterberg et Where this responsiveness is lacking participation is al., 2014). Our present survey data show (Annex 2) seen by villagers as being a poor use of their time. that attendance was also much less likely among women. Comparing attendance at village and Unlike village meetings, hamlet meetings seem hamlet meetings, the characteristics were similar to be more broadly attended by those who are except in their magnitude, indicating that hamlet wealthier and people from the bottom 40 percent. meetings seemed to be more “friendly” to people There was no significant difference of likelihood with no formal education. Household members from to participate in hamlet meetings across wealth the bottom 40 percent of the welfare distribution quintiles. Given the proximity of hamlets to villagers’ were also less likely to come to village meetings, places of residence, it was easier for villagers while attendance at hamlet meetings did not seem to participate in hamlet-level activities, including Table 3. Perceptions toward village heads and hamlet heads (%) District All obs Merangin Batanghari Banyumas Wonogiri Ngada F-test Felt that […] is reliable to make development plans -- Village Head 74.9 50.8 45.2 75.8 83.8 74.5 18.4 ** -- Hamlet Head 78.5 62.9 61.8 76.0 89.0 75.3 14.7 ** -- BPD 58.3 58.3 60.0 60.2 53.8 72.3 2.2 -- PNPM Kecamatan facilitators 49.0 30.6 23.1 54.9 47.9 50.7 25.2 ** Felt that […] is reliable to implement development -- Village Head 77.5 56.5 52.3 78.0 85.9 76.2 17.0 ** -- Hamlet Head 80.8 67.0 65.5 79.1 89.6 76.9 13.1 ** -- BPD 59.0 58.2 61.8 61.1 54.2 72.3 2.1 -- PNPM Kecamatan facilitators 48.8 30.9 22.2 54.7 47.6 50.2 25.3 ** Observations 4,081 891 456 1,155 989 590 Notes: all figures are in percentage; ** statistically significant at 5 percent; * statistically significant at 10 percent 8 A small fraction of villagers (158 respondents) came to village meetings without invitation. They represented 7 percent of the uninvited (2,390 respondents). 9 For more discussions on the non-participating villagers, see Sambodho (a, forthcoming). 18 PARTICIPATION, TRANSPARENCY AND ACCOUNTABILITY IN VILLAGE LAW IMPLEMENTATION meetings in the evening. This is an indication However, in general not being invited turned out that the hamlet could be the locus where most to be less of a barrier to meeting attendance. participation from villagers can be expected. In When asked to provide motivations for attending addition, more villagers viewed hamlet heads as future village meetings, most villagers said that being reliable as opposed to village heads in four they would attend such meetings upon invitation out of five districts (Table 3). This difference is (76 percent) but, as shown in Seberang Sungai statistically significant. and mentioned earlier, open invitations did not necessarily bring people to the meetings. In fact, From the perspective of villagers, they face both more than half of the invitees did not attend the external and internal barriers that prevent them meetings (Annex 1), showing the unpopularity of from participating in meetings. Not being invited such meetings. And 11.5 percent stated that they (an external barrier) was the most cited reason would attend if attendance were made mandatory (70.1 percent), followed by internal barriers: by village authorities. villagers felt the meetings were irrelevant to them (17.1 percent),10 or they were too busy to attend A closer examination of various other factors (17.1 percent) (see Annex 3 for details). Women faced that may have influenced villagers’ attendance more internal barriers. They were less likely to claim indicates there were some significant differences that they were not invited, but more likely to say between men and women. For men, welfare that they were too busy, or that the meetings were seemed to influence their attendance positively— irrelevant to them. This was the opposite among the the richer they were, the higher the likelihood poor (bottom 40 percent), who faced more external that they attended village meetings (Annex 4). barriers. They were more likely to say that they did For women, their domestic role of taking care of not go to meetings because they were not invited toddlers (age 0-4) contributed to their decisions to and less likely to claim that they were too busy. attend or not to attend village meetings. This was not the case for men. Figure 2. Type of engagement Figure 3. Type of engagement during village meetings (%) during hamlet meetings (%) Provided suggestion Voted for decision Expressed Expressed opinion opinion Voted for Provided decision suggestion Asked about Asked about program program Asked about Asked about the program targets budget Male Male Asked about the Female Asked about Female budget program targets 0 10 20 30 40 50 0 10 20 30 40 50 Some of the reasons respondents gave in considering the meetings irrelevant to them included: (i) meetings were village government matters, (ii) 10 meetings were for men, and (iii) meetings had no benefits for them. 19 PARTICIPATION, TRANSPARENCY AND ACCOUNTABILITY IN VILLAGE LAW IMPLEMENTATION In both village and hamlet meetings, men were Participation was statistically significantly higher more engaged than women in the discussions, in villages in which village heads and at least half but the gap is generally closer in hamlet meetings of their staff were former PNPM actors (Table 4). (except in voting) even when men still dominate Even just the fact that the village head had had the attendance. The three top types of engagement previous PNPM experience helped to increase participants cited were: (i) voting on decision- participation in village meetings. Also, if the village making (39.9 percent); (ii) providing suggestions head had participated in VL-related training, his/her (39.1 percent); and (iii) expressing opinions or villagers were 7.3 percent more likely to participate passing judgment (38.3 percent) (Figures 2 in village meetings. The length of tenure, of at and 3). It was men (as household heads) who least three years, was also positively associated usually attended the meetings to represent their with participation in hamlet meetings, as well as households. When women attended, they were involvement in local infrastructure work. often discouraged from speaking up or, if they did, they were often taunted about being “rebellious” In addition, the experience of the BPD—as or “troublesome”, as shown in the qualitative study the other branch of village leadership—seems in Wonogiri. Most of the time women attended the to encourage participation, especially on women-only neighborhood or hamlet meetings to infrastructure work. Having a head of the BPD discuss day-to-day topics such as arisan (rotating who was 50+ years old or who had been residing fund) and weekly praying/Quran reciting group. in the village for more than 40 years led to higher involvement by villagers in local infrastructure work The level of previous experience of the village (19 percent and 11.3 percent, respectively) as shown leadership helps to drive villagers’ participation.11 in Table 5. Table 4. Village heads’ previous experience and villagers’ participation Villagers All obs VH has 3+ yr tenure VH and at least half of Village Head: former PNPM Village Head: has attended participated in Village apparatus are former actor training(s) related to VL PNPM actors Yes No Difference Yes No Difference Yes No Difference Yes No Difference Village meetings 23.7 27.6 21.9 5.7 38.2 22.3 15.9 ** 31.7 21.3 10.4 ** 24.5 17.2 7.3 * Hamlet meetings 30.6 40.1 26.1 13.9 ** 43.7 29.3 14.3 34.9 29.3 5.6 31.2 25.8 5.3 Infrastructure work 60.3 67.6 56.8 10.8 ** 71.6 59.2 12.4 ** 61.8 59.8 1.9 60.8 55.8 4.9 Observations 112 36 76 10 102 27 85 100 12 Notes: all figures are in percentage; ** statistically significant at 5 percent; * statistically significant at 10 percent Table 5. BPD’s profiles and villagers’ participation Villagers participated in All obs BPD Head: age 50+ yrs BPD Head: has resided in village 40+ yrs Yes No Difference Yes No Difference Village meetings 23.7 23.1 24.3 -1.2 20.0 26.4 -6.3 Hamlet meetings 30.6 34.3 27.2 7.1 33.5 28.6 4.9 Infrastructure work 60.3 70.1 51.1 19.0 ** 66.9 55.6 11.3 ** Observations 112 53 59 47 65 Notes: all figures are in percentage; ** statistically significant at 5 percent; * statistically significant at 10 percent 11 We collapsed villagers’ participation at the village level and associate it with village head, village apparatus and BPD characteristics. The latter includes gender, age, education, tenure and experience with PNPM of the respective village governance actors, as well as whether the BPD was directly elected and active in undertaking its tasks and responsibilities. 20 PARTICIPATION, TRANSPARENCY AND ACCOUNTABILITY IN VILLAGE LAW IMPLEMENTATION 5.2. Transparency to know village programs/activities implemented outside their hamlet (Figure 4). This higher level of Overall, less than half of all villagers were aware of knowledge concerning their own hamlets is another village programs, while a much lower proportion indication of villagers’ interest, as also illustrated by (10 percent) claimed to have knowledge of the use their greater participation in hamlet-level meetings. of village funds, paralleling villagers’ involvement in planning and budget discussions. Ngada The characteristics of villagers who are more likely noticeably remained at the top in both cases (Figure to be aware of village programs and finances are 4). Villagers in Ngada claimed to have the highest similar to those who are more likely to participate knowledge, while respondents in Batanghari had in meetings. Villagers attending meetings were the least knowledge, although Batanghari (with more likely to have higher educational attainment, Merangin) had the second-highest participation rate be currently working, come from the ethnic majority at village meetings, where presumably information group, and be active in local organizations and was shared (Figure 1).12 These different directions political parties (Annex 5). They were also more of participation in deliberative meetings and of likely to attend village and/or hamlet meetings, knowledge about village programs and funds, express concerns to the village government, and particularly in Batanghari, may indicate different have a higher opinion of the village government’s levels of interest and the limitations of information- reliability in planning and implementing village sharing, as discussed later in this section. development activities. Similarly, those who were less likely to be aware of village programs and In contrast, but as expected, villagers knew more the use of village funds were women, members of about their hamlet activities than those outside female-headed households, and those in the bottom their hamlets (village activities). More than 80 40 percent of the welfare distribution. percent of survey respondents stated that they knew of, and participated in, local infrastructure Almost all village heads, however, claim to have activities funded by the village government in socialized their village plans and fund use, but their respective hamlets during the past two years. mainly to selected groups (e.g., the BPD, hamlet However, only 47.8 percent of respondents claimed and RT/RW heads, and other community leaders). The information actually disseminated to the public was much lower than claimed by village heads (99.4 vs 75 percent on village plans, and 96 vs 68.2 Figure 4. Villagers’ awareness of village percent on the use of village funds), as shown in programs and the use of village funds Table 6. Batanghari and Ngada took turns to top the list in actual dissemination, while Merangin and Ngada Wonogiri were bottom for information on village plans and use of village funds. What villagers ended Banyumas up receiving was even lower, as shown earlier in Wonogiri Figure 4, although the top three media that villagers Village program and village heads liked concurred (Table 7). Village Merangin Village funds use heads’ other preferences of sharing information in writing (through brochures and, at a much lower Batanghari rate, information boards) turned out to be less 0 20 40 60 80 100 popular with the villagers, which might influence the effectiveness of the dissemination. 12 In our survey, the extremely low rates were mostly in Batanghari and Merangin. There were four villages (all of them are in Merangin) where less than 10 percent of respondents knew about their village programs. In 14 villages respondents had zero knowledge of village fund use. Seven of these villages were in Merangin, five in Batanghari, and one village each in Banyumas and Wonogiri. 21 PARTICIPATION, TRANSPARENCY AND ACCOUNTABILITY IN VILLAGE LAW IMPLEMENTATION Villagers turned out to want to hear different kinds labor. Information on village finances ranked fourth of information. Village programs only ranked third for village heads to share (58.6 percent) and also in terms of the kind of information that villagers for villagers to want to know about (26 percent). wanted to hear (33.7 percent), as shown in Table Both wanted to share and to hear about the 8. The first and second kinds of information most same issues, but interest in village finances was sought after were aid programs (63.4 percent) and much lower, both in terms of wanting to share implementation activities (45.3 percent). These the information by village heads, but particularly kinds of information were in line with the information in terms of villagers wanting to learn about the that village heads wanted to share with villagers, information. Hence, the low level of villagers’ with a slightly different order. More than 75 percent awareness on village finances (Figure 4). of village heads stated that they would like villagers to have more information on the implementation Our qualitative study provided some insight into of activities (84 percent), village programs (78.1 the main issues concerning information. Village percent), and aid programs (76.9 percent). Village heads did not proactively disseminate information to heads wanted villagers to know about activity villagers, although neither did they prevent villagers implementation largely because this was when from obtaining it. Village heads claimed that they villagers were expected to contribute or share their were happy to share information should villagers Table 6. Village heads’ perceptions of information needed by villagers District All obs Merangin Batanghari Banyumas Wonogiri Ngada F-test VH claimed to have announced Village 99.4 97.3 100.0 100.0 100.0 100.0 1.0 Programs Observations 112 24 12 32 28 16 VH actually announced it to general public 75.0 51.8 100.0 85.1 64.3 87.2 10.8 ** Observations 111 23 12 32 28 16 VH claimed to have announced Village 96.0 90.2 90.3 100.0 95.4 100.0 1.3 Funds Use Observations 112 24 12 32 28 16 VH actually announced it to general public 68.2 49.5 83.9 82.4 46.1 86.7 4.9 ** Observations 107 21 11 32 27 16 Notes: all figures are in percentage; ** statistically significant at 5 percent; * statistically significant at 10 percent Table 7. Types of information villagers most often requested Information dissemination media: All Observations Merangin Batanghari Banyumas Wonogiri Ngada Villagers VH view Villagers VH view Villagers VH view Villagers VH view Villagers VH view Villagers VH view Villagers vs. Village Heads needs needs needs needs needs needs Special meetings (vilage, hamlet, 69.7 87.2 57.6 56.3 72.2 84.7 70.9 96.1 70.6 100 76.6 93.4 RT/RW) Special meetings (prayer group, 20.8 34.6 30.3 56.3 55.2 70.8 17.7 31.4 19.6 7.9 6.9 30.6 etc.) Information board at VH office 3.7 9.8 6.5 8.5 7.0 15.3 3.1 2.1 3.6 18.9 2.6 11.2 Village electronic media/website 1.6 2.6 0.2 5.8 0.0 0.0 1.3 4.1 3.4 0.0 1.0 0.0 Brochure/invitation/pamphlet 4.4 12.0 6.6 8.0 1.5 0.0 4.5 11.6 3.4 17.1 5.2 18.4 Community radio 0.0 0.7 0.1 0.0 0.0 0.0 0.0 2.1 0.0 0.0 0.4 0.0 Announcement VH Office/ 4.2 17.5 21.6 74.6 6.7 4.9 2.4 0.0 0.7 0.0 2.8 10.2 Mosque/Church Verbal from Village apparatus 24.9 32.8 24.0 26.8 28 75 21.5 27.9 30.8 19.3 38.2 49.0 Verbal from community/religious 1.0 5.4 0.8 5.8 0.2 25.7 1.2 0.0 0.6 7.9 0.7 0.0 leaders Observations 2,757 112 644 24 350 12 845 32 468 28 450 16 Notes: all figures are in percentage; ** statistically significant at 5 percent; * statistically significant at 10 percent 22 PARTICIPATION, TRANSPARENCY AND ACCOUNTABILITY IN VILLAGE LAW IMPLEMENTATION ask for it. However, no effort was made to institute The survey data also indicate that villagers any systematic mechanism to channel information themselves do not appear keen on obtaining to villagers. Village heads often mentioned that they information on village affairs. More than one-third relied upon, or more likely assumed, that hamlet of respondents—with Wonogiri at the top (54.6 or neighborhood heads disseminated information percent)—stated that they had no interest in learning on village development to villagers. However, the about village-related information (Table 9). These hamlet or neighborhood heads were not required respondents belonged to the same group as those to report information back to villagers. Only in one who were not participating in meetings and who village in Ngada were village heads required to had little awareness of village programs/finances, report construction plans and budget details to the namely women, those with lower educational community before infrastructure projects started, attainment, members of female-headed households, and this was more for the purpose of calculating and less well-endowed villagers. Those who were the number of man-days the community needed older and had lived longer in the village also had to provide to participate in the construction, as a greater likelihood of not wanting to know about opposed to simply informing the villagers. village-related information. Qualitative findings Table 8. Village heads’ perceived information needed and types of information most often requested by villagers Information dissemination All Observations Merangin Batanghari Banyumas Wonogiri Ngada media: Villagers vs. Village Villagers VH view Villagers VH view Villagers VH view Villagers VH view Villagers VH view Villagers VH view Heads needs needs needs needs needs needs Type of information: -- Village program 33.7 78.1 34.5 80.4 44.8 80.6 26.7 74.0 43.8 71.8 62.3 93.4 -- Village funds use/ village 26.0 58.6 31.5 77.7 33.9 41.0 25.1 62.8 21.6 33.2 43.7 74.0 financial condition -- Implementation of village 45.3 84.0 45.3 87.1 42.4 79.9 48.2 84.3 36.0 72.9 61.6 100.0 programs -- Assistance programs 63.4 76.9 77.3 82.1 90.3 70.1 63.8 81.8 49.0 58.2 74.6 93.4 -- National/religious festivities/ 6.3 30.2 12.0 47.8 2.2 19.4 8.0 39.3 0.7 12.9 6.6 18.4 events -- Other information 8.8 19.5 4.7 0 0.2 4.9 9.4 23.1 11.2 46.1 5.8 5.1 Notes: all figures are in percentage; ** statistically significant at 5 percent; * statistically significant at 10 percent Table 9. Subjects of interest and villagers’ characteristics Types of information villagers most Gender District All obs often requested Women Men Difference Merangin Batanghari Banyumas Wonogiri Ngada Villagers do not want any village 36.4 42.1 30.3 11.8 ** 28.9 22.3 28.0 54.6 54.6 information Observations 4,081 2,125 1,956 891 456 1,155 989 989 Among villagers who seek information, they are seeking information on: -- Village program 33.7 29.6 37.3 -7.6 ** 34.5 44.8 26.7 43.8 62.3 -- Village funds use/village financial 26.0 18.5 32.8 -14.4 ** 31.5 33.9 25.1 21.6 43.7 condition -- Implementation of village 45.3 36.4 53.3 -16.9 ** 45.3 42.4 48.2 36.0 61.6 programs -- Assistance programs 63.4 68.3 59.1 9.2 ** 77.3 90.3 63.8 49.0 74.6 -- National/religious festivities/events 6.3 6.4 6.3 0.0 12.0 2.2 8.0 0.7 6.6 -- Other information 8.8 8.9 8.6 0.3 4.7 0.2 9.4 11.2 5.8 Observations 2,757 1,310 1,447 644 350 845 468 450 Notes: all figures are in percentage; ** statistically significant at 5 percent; * statistically significant at 10 percent 23 PARTICIPATION, TRANSPARENCY AND ACCOUNTABILITY IN VILLAGE LAW IMPLEMENTATION also showed that villagers were often not keen to Similar to participation, the experience of village find out information about village development. leaderships is positively correlated with villagers’ As long as villagers knew what was being built in awareness of village programs and the use of their neighborhood or hamlet, they were satisfied. village funds. Villagers residing in villages with a Villagers also indicated that they did not want to village head who had three or more years of tenure know too much detail on village budgets, citing that reported 4 percent higher awareness of the use of this was “the business of the village government”. village funds than those living in villages where the village head had less than three years of tenure. In As in participation and awareness about addition, having a BPD head who was older also village programs and finances, men and the led to higher villagers’ awareness of funds use better-off are more likely to have an interest in (nearly 5 percent) and satisfaction with the provision soliciting information than women and the poor, of information (nearly 7 percent). Village heads aggravating the disadvantages of the latter and staff with previous experience of PNPM also (Annex 5). However, unlike in participation, those correlated with higher levels of satisfaction among who already had positive views of their village head villagers on the information provided by the village and hamlet head (being reliable in making plans government (Table 10). and executing them) tended to have little interest in learning more information about their village (Annex Village heads’ proactivity in disseminating 6). They may have already felt satisfied with their information appears to have no effect on villagers’ village leader’s performance and were unwilling to knowledge or awareness. Around 75 percent ask further questions. Meanwhile, the inclination village heads announced village programs to was different when they thought their BPD head the public, while 68 percent announced the use was reliable. This group wanted to learn more about of village funds (Table 6). However, these efforts village affairs. The different correlations between did not influence villagers’ awareness of village a positive view toward village and hamlet heads programs or the use of village funds. Furthermore, and low interest in village information on one hand, villagers’ level of satisfaction with the provision of and the a positive view toward BPD heads and high information in general (i.e., not only information interest in village information on the other hand, specific to village programs and/or the use of need to be explored further in the end-line study. village funds) was not influenced by village heads’ dissemination efforts (Table 11). Table 10. BPD members’ previous experience and their direct election, and villagers’ awareness All obs BPD Head: age 50+ yrs Village Head VH and at least half Village Head: is former PNPM actor of Village Apparatus tenure 3+ yrs are former PNPM actors Yes No Difference Yes No Difference Yes No Difference Yes No Difference Villagers are: Aware of village programs 49.0 54.1 44.2 9.9 ** 59.2 48.0 11.2 * 59.2 48.0 11.2 * 51.4 47.8 3.6 Aware of village funds use 11.3 13.7 9.0 4.6 * 14.4 11.0 3.5 11.0 3.5 3.5 14.0 10.0 4.0 * Satisfied with information from 43.6 47.1 40.3 6.8 * 53.5 42.6 10.9 ** 42.6 10.9 10.9 ** 45.3 42.8 2.5 Village Government Observations 112 53 59 10 102 10 102 36 76 Notes: all figures are in percentage; ** statistically significant at 5 percent; * statistically significant at 10 percent 24 PARTICIPATION, TRANSPARENCY AND ACCOUNTABILITY IN VILLAGE LAW IMPLEMENTATION In summary, a couple of factors influence villagers’ 5.3. Responsiveness and awareness of information. First, the characteristics accountability of both the village head and the villagers—village heads with specific experience are likely more In addition to planning and executing village proactive in sharing information, while villagers programs, village governments are also expected of higher socioeconomic status are more likely to to respond to other priority problems that villagers receive the information. Second, more respected claim to be facing. Problems that respondents village leaders (i.e., older leaders) lead to higher cited included inadequate roads/infrastructure, crop villagers’ participation and awareness. In terms of failure and high unemployment rates (Table 12). interest in the information, villagers show much In almost all cases, there were more respondents lower interest in village finance than village heads’ in Ngada than in other districts that viewed the claimed was disseminated on the same topic. village government as being helpful in attempting Table 11. Village heads’ dissemination and villagers’ awareness Village Head announced Village Head announced Village Programs to general public Village Funds Use to general public All Obs Yes No Difference All obs Yes No Difference Villagers are: Aware of village programs 49.2 50.8 44.5 6.3 Aware of village funds use 11.6 12.1 10.7 1.4 Satisfied with Information from 43.7 44.4 41.8 2.5 44.9 47.2 40 7.2 Village Government Observations 111 83 28 107 73 34 Notes: all figures are in percentage; ** statistically significant at 5 percent; * statistically significant at 10 percent Table 12. Top three problems and solutions: views of villagers District All obs Merangin Batanghari Banyumas Wonogiri Ngada F-test Problems/challenges faced by villagers Road 43.7 39.2 62.6 47.2 35.8 57.3 5.2 ** -- attempts by villagers to address the 78.3 56.8 55.6 80.9 83.5 82.3 4.5 ** problem -- attempts by village government to 73.8 56.4 59.9 74.5 80.5 75.3 2.4 * address the problem -- problem resolved/mostly resolved 37.8 28.7 13.2 42.4 39.6 16.0 6.8 ** Crop failure 40.7 41.1 65.4 37.1 39.5 78.0 11.6 ** -- attempts by villagers to address the 56.5 29.1 39.9 59.7 60.4 70.3 10.1 ** problem -- attempts by village government to 29.5 15.4 14.6 31.0 28.6 58.1 11.7 ** address the problem -- problem resolved/mostly resolved 28.4 21.7 5.9 30.9 32.5 21.4 10.9 ** High level of unemployment 32.0 39.5 25.7 37.6 21.5 34.6 2.6 * -- attempts by villagers to address the 33.7 20.6 31.2 32.7 43.4 36.1 2.9 ** problem -- attempts by village government to 18.1 6.0 3.5 18.5 22.9 32.5 25.6 ** address the problem -- problem resolved/mostly resolved 10.0 3.5 5.7 11.6 9.2 12.9 3.0 ** Notes: all figures are in percentage; ** statistically significant at 5 percent; * statistically significant at 10 percent 25 PARTICIPATION, TRANSPARENCY AND ACCOUNTABILITY IN VILLAGE LAW IMPLEMENTATION to resolve problems. Respondents also felt that Ngada also received the most complaints from programs prepared by the village government were villagers, which may indicate village government needed, especially in infrastructure. There was accessibility, aside from the problems that almost no disagreement that the programs prepared villagers experienced. Overall, about 9 percent of by village governments were greatly needed respondents submitted complaints, mostly verbally (Figure 5). (Table 13). More men than women complained and more than one-quarter of complaints failed Figure 5. Perception on importance of village programs 80% 70% 60% 50% Male Female 40% 30% Respondents who said 20% that village programs are “very needed” (%) 10% 0 Merangin Batanghari Banyumas Wonogiri Ngada Table 13. Complaint handling by village governments District All obs Merangin Batanghari Banyumas Wonogiri Ngada F-test Complaint and Others Submitted complaint/reported problems to village/hamlet, 9.0 9.3 4.4 9.9 6.6 23.9 6.5 ** through: -- meetings conducted by village government 17.2 22.8 11.4 13.2 23.0 23.7 1.3 -- meetings conducted by hamlet/ward head 30.8 7.8 2.6 30.7 48.9 13.7 8.8 ** -- community meetings (routine/special) 2.2 1.3 0.0 1.5 4.3 2.5 2.7 * -- verbally to village/hamlet officials 56.7 82.7 98.1 54.9 40.7 74.6 22.3 ** -- protest/demonstration 0.1 0.0 0.0 0.0 0.0 0.9 0.9 Village/Hamlet head response: -- facilitate problems with the authorities 7.3 13.1 9.5 3.6 10.8 14.9 1.6 -- conduct a complaint verification/examination of cases 3.6 11.5 0.0 1.5 4.9 5.8 5.1 ** -- deliberate with the community 12.3 12.5 23.1 3.4 31.9 15.1 5.3 ** -- dialogue with the parties involved to seek a settlement 6.1 4.8 6.1 5.3 3.4 20.5 0.8 -- propose the addition of public facilities to village govt 5.8 7.6 6.4 5.6 5.0 7.7 0.1 -- propose to improve road/bridge to facilitate citizen access 16.9 7.0 17.9 17.5 23.5 5.8 3.0 ** -- submit complaints/reports of citizens to the village govt 19.2 13.7 26.6 13.3 21.9 55.2 2.9 ** -- to bridge the villagers and village govt to direct dialogue 2.2 9.5 0.0 0.4 3.7 2.7 2.2 -- NO RESPONSE from village/hamlet 28.6 28.5 30.7 36.3 14.7 14.9 3.6 ** Villagers perception on whether the problem is resolved: -- fully or mostly resolved 24.5 19.8 5.2 22.8 30.5 28.6 9.4 ** -- only a small fraction is resolved 21.5 31.6 42.7 21.3 19.1 13.8 3.0 ** -- unresolved 25.4 20.2 21.4 19.6 35.6 42.8 4.9 ** -- unresolved, as the problem was not addressed 28.6 28.5 30.7 36.3 14.7 14.9 3.6 ** Notes: all figures are in percentage; ** statistically significant at 5 percent; * statistically significant at 10 percent 26 PARTICIPATION, TRANSPARENCY AND ACCOUNTABILITY IN VILLAGE LAW IMPLEMENTATION to receive a response, with Wonogiri and Ngada having the lowest “no response” rate. The Box 2. Accountability meeting in Ndona Village, Ngada responses included relaying the complaints to higher levels of government, repairing roads In Ngada, all villages are required to have a (usually) using the village budget, and deliberating one-day forum annually to discuss the Laporan with related villagers. It is interesting to note that Pertanggjungjawaban or LPJ (end-of-year although Ngada has lower "no response" and accountability report) and also at the end of the "unresolved" (problem was "not addressed") rates, village head’s term. The BPD organizes the forum there are also high percentage of "unresolved" on any day between December and March. The problems. This might indicate that many problems village head submits his/her report to the BPD at in Ngada are beyond the capacity of village least two weeks before the discussion. The BPD government to handle. invites all villagers and the kecamatan government to hear the village head reporting on all the In general, downward accountability mechanisms development activities of the related year, and hear as stipulated in the Village Law had not been the BPD’s comments/criticisms. Villagers are not put into practice at the time of the baseline data allowed to talk (comment), as this is a forum for the BPD to scrutinize the village head’s performance. collection. Village governments only provided reports to the district government (upward This forum provides a window for villagers to accountability) as part of the administrative observe the state of the relationship between the requirements needed to obtain Dana Desa (from village head and the BPD. The village head of the national government) and Alokasi Dana Desa Ndona said, “The BPD can comment on the LPJ but (from the district government). Only in Ngada, as they cannot reject it, because the report has already shown in the qualitative study, were villages still been submitted to the Inspectorate previously (and required to hold accountability meetings in which was not rejected). So there is no way for the BPD village heads presented their village implementation to reject it. In 2015, the BPD criticized that our own reports (LPJ) to the BPD (horizontal accountability). revenues were too small, and that many villagers The details of the accountability mechanisms varied had not paid their iuran (dues) to the desa, and I across villages. Some meetings allowed villagers seldom talked to villagers.” to provide comments on the reports while in others they were invited just to listen, as in one of the qualitative study sites (Box 2). Survey results show that 39.8 percent of village heads claimed that they had conducted such meetings (compared with 97.2 percent for village planning meetings). The village government and the BPD then disseminate the village head’s report to the villagers, but more as a “for-your-information- only” activity. Dissemination usually took place in a variety of community gatherings, such as in parties or following prayer meetings. Some questions and discussions did take place, but any follow-up was at the village government’s discretion. 27 PARTICIPATION, TRANSPARENCY AND ACCOUNTABILITY IN VILLAGE LAW IMPLEMENTATION 5.4. Role of village activists supervision and demanding accountability and the BPD from village governments. As the regulation to operationalize the Village Law was not issued until As discussed earlier, given the challenges or late 2016, we found that the BPD was still operating barriers to increasing the number and the range based on a carry-over from the former law that of groups of villagers participating directly in presented it as a part of the executive branch and deliberation meetings, this prompted us to allowed the village head to appoint BPD members. look at the potential for making use of other BPD members had little knowledge of their role “representatives”. The formal representatives, i.e., and responsibilities, including providing checks and the village council or the BPD, were perceived balances on the village government. as less reliable in planning and implementation compared with other leaders, such as village However, when BPD heads were perceived as heads and hamlet heads. Villagers found BPD reliable, villagers tended to participate more members to be less reliable in developing and in village and hamlet meetings (Annex 1). Also, implementing village plans (Table 3). Admittedly, this respected BPD heads (elders) increased villagers’ is not a BPD task as the “legislative branch”, but its awareness of village funds use (Table 8). These involvement in the work of the village government findings indicate that strengthening the BPD’s was still not well recognized or understood by roles and capacity will have a positive impact on villagers. governance, especially in encouraging “demand- side” participation and accountability push. The qualitative study results confirm that the BPD members were perceived as being less effective Village activists also show potential to become than the village government in assisting villagers in villagers’ representatives. Most of these village resolving village problems. Few villagers considered activists (around 76 percent) were invited to, and the BPD to be of great importance or close enough attended, village and hamlet meetings (Table 14). to their constituents.13 The qualitative results also Their perceived socioeconomic status was closer show that the BPD was not yet active in providing to that of the village leadership, placing them Table 14. Activists’ participation in deliberative meetings All obs Gender Female Male Difference Attended Village meetings 75.6 81.8 73.6 8.2 Observations 222 110 112 Among those attended meetings: -- Provided suggestion 90.8 87.1 92.1 -4.9 -- Passed judgment 80.3 63.8 86.3 -22.5 ** -- Asked about program 66.0 55.1 70.0 -14.9 ** -- Asked about program targets 52.8 37.2 58.5 -21.3 ** -- Asked about the budget 53.6 43.0 57.4 -14.3 * -- Voted for decision 62.7 63.2 62.5 0.8 Observations 178 91 87 Notes: Figures are in percentage; ** statistically significant at 5 percent; * statistically significant at 10 percent Note: Male village activists comprise: religious leaders (48 percent), youth groups (16 percent), the business sector (10 percent); while female village activists comprise: PKK/Dasa Wisma (47 percent), religious leaders (29 percent), businesses (5 percent). 13 See Sambodho (a, forthcoming) for more discussions on the capacity gap between the village government (especially the village head) and the BPD. 28 PARTICIPATION, TRANSPARENCY AND ACCOUNTABILITY IN VILLAGE LAW IMPLEMENTATION relatively on a par with the village leadership, although this status also poses risks of perpetuating elite domination. Gender-wise, the difference in participation rates between male and female village activists was not statistically significant. Female village activists were almost as engaged as their male counterparts in meetings, indicating that, unlike non-activist women, they were less reluctant to talk in public forums and were thus better able to help air the voices of their fellow female villagers.14 Most importantly, these village activists seem to share the general community’s concerns over village problems/priority needs. Our survey listed 25 problems for villagers, village and hamlet heads, the BPD and village activists to choose as their top priorities. The most-cited issues by villagers were: access to road, harvest failure and high unemployment (Table 12). The responses from the villagers, hamlet heads and village activists showed a significant positive correlation for road and harvest issues (Table 15). Village activists did not share similar concerns with villagers on high unemployment. Otherwise, they had similar views on what they considered to be the most urgent village problems. Table 15. Correlation on problems cited by villagers, village activists and hamlet heads Villagers - Activists Villagers - HHs All obs Female Male Road 0.5024* 0.4374* 0.3494* 0.3074* Crop failure 0.5592* 0.3504* 0.5235* 0.437* High level of unemployment 0.1688 0.1222 0.1586 0.3046* Notes: ** statistically significant at 5 percent; * statistically significant at 10 percent 14 See Sambodho (b, forthcoming) for more detailed discussions on village activists: their potential and the risks in representing villagers. 29 6. Key Takeaways PARTICIPATION, TRANSPARENCY AND ACCOUNTABILITY IN VILLAGE LAW IMPLEMENTATION Key Takeaways Deliberative meetings (e.g., to discuss annual plans, villages. About one-third of villagers stated outright that or the budget and accountability reports) at the they had no interest in any village-related information. They village level are not an inclusive process. Less than were interested in information that directly and immediately half village heads made any effort to invite villagers impacted them, such as information on aid programs and to attend such meetings. When they did, village project implementation when they might be expected governments decided who to invite. From the villagers’ to work. Village heads concurred and shared the same perspective, the probability of receiving an invitation interests in disseminating such information. to these meetings depended on the individual and his/ her household’s characteristics. Women and those in In contrast to village-level meetings, hamlet meetings are the bottom 40 percent of the welfare distribution were more popular. These meetings had a higher attendance less likely to receive an invitation. Conversely, those rate and were also more inclusive. Participants came from with higher educational attainment, those actively different social groups. The level of welfare of participants involved in local organizations and/or political parties, did not seem to influence their attendance, nor did concerned villagers (i.e., those having raised issues distance. More villagers viewed hamlet heads positively and or complaints previously) and those who viewed the villagers also knew more information about project activities village head positively were more likely to be invited. in their own hamlets. In some areas in Java, meetings might even begin at neighborhood/ward levels due to the large However, not being invited is not the main reason population size of the villages. that villagers do not go to meetings. More than half of all villagers did not attend meetings regardless While villagers appear not to be interested in village of being invited, indicating that the meetings were meetings, most village activists (over 75 percent) attend unpopular. Villagers considered the meetings not these meetings. These village activists generally belonged to be their concern and that they were the business to the same socioeconomic status as other village of village leaders, while villagers also assumed that leadership members, making their interaction relatively they were already being represented by their hamlet/ easy. The village activists’ issues of concern were largely neighborhood and community leaders. In addition, similar to those of most villagers. In addition, there was less attending meetings imposed social and financial costs difference in the level of engagement in village discussions on villagers, taking them away from their work and between female and male activists compared with non- domestic chores, particularly for women. activists, suggesting their potential in representing the villagers, both men and women. Villagers’ views of village leadership (village government and the BPD) and their previous Finally, district policy appears to influence the level of experience also influence their participation in villagers’ participation. Ngada consistently showed a meetings and awareness of information. Village higher level of participation and awareness of village affairs heads who had been in office for three years or and information. The district had its own CDD-type of more had a positive influence, as did village heads projects, mirroring PNPM. Similarly, at the sub-village level and their staff who had previous experience of CDD there was a long tradition of regular community gatherings projects (i.e., PNPM). In addition, BPD heads who were in Wonogiri, which increased participation at hamlet perceived as reliable and respected by villagers also meetings, making Wonogiri second to Ngada. Regular had a positive impact. prayer meetings often served as a forum for sub-village level discussions. Where these good practices will lead is a Similar to participation at village meetings, villagers point of considerable interest to observe in the remainder do not seem to be interested in information about of the study. the use of funds or development plans in their 31 7. Looking Forward PARTICIPATION, TRANSPARENCY AND ACCOUNTABILITY IN VILLAGE LAW IMPLEMENTATION Looking Forward Improving representation is a much-needed step, power to check the village government is absent given the faltering levels of public participation. (weakened village councils and customary leaders) Direct public participation across all groups, including between elections, abuses are more likely (Wetterberg et poor and marginalized groups, may not be realistic, al., 2014). given the constraints these groups face, ranging from particularly high opportunity costs, limited access to Hamlet deliberation meetings need to be strengthened related information and knowledge, and gender bias as the primary locus of public participation. Villagers as often occurs in patriarchal communities. Indirect were more involved in planning discussions at the hamlet participation or participation by representation level. Socioeconomic status was less of a barrier here and should therefore be strengthened in tandem with their knowledge of development in their own hamlet was direct participation, especially by women and the also higher than of their village. If village activists and the poor. For this reason, the BPD, which was largely BPD represent villagers or their constituents at village- disregarded at the start of VL implementation, should level discussions, villagers can then focus on hamlet be strengthened in the spirit of the VL. However, discussions, particularly during the development of the as this report shows, village activists are also an mid-term plan (RPJM Desa). The annual plan, derived from alternative, and potentially more effective, source the RPJM Desa, need not be discussed as extensively, for strengthening the community’s representatives. unless there are new proposals of unforeseen urgency. There are legitimate concerns that figures such as Meanwhile, the representatives should be more involved village activists may have interests that are closer to at these village-level discussions. However, it remains the village government’s and not attuned to those of important for these representatives to report back to the general villagers. Despite this risk, a recent study villagers later at their hamlet meetings. in Zimbabwe shows that figures such as the health workers, school committee members and leaders of Improving ways of communicating village development farmers’ groups, when their capacity was improved, activities and finances are key in ensuring that villagers turned out to be able to exert horizontal pressure to are provided with opportunities/channels to seek out counter the village government (Baldwin, Muyengwa such information and, at the same time, to raise the & Mvukiyehe, 2017). The study cites several reasons community’s awareness of the “publicness” of funds why this approach works: the long tradition of having that are managed by the village government. Villagers countervailing elites acting against the community showed limited interest in the management of village chiefs, and community leaders who are generally funds, and tended to assume that this was the business of young and less partisan, and who are also likely to the village government and did not involve them. The BPD benefit themselves from reforms. and village activists, in particular, could all play a more active role in finding the best channels to communicate Strengthening more community figures will information and in ensuring information can be accessed also expand checks and balances outside the as and when needed. With the new regulation on the formal institutions. Village councils and activists BPD (Permendagri No. 110/2016), it will be important to can serve as a countervailing power to the village begin concerted efforts to improve the BPD’s capacity to government. The last round of the LLI study in supervise village governments and ensure that downward Indonesia shows that democratic elections improved accountability occurs. At the same time, capacity building village governance—that village heads are more should also be provided to village activists, strengthening likely to work in the villagers’ interests, and maintain them while also not weakening or neglecting capacity participatory and transparency norms. However, the building in village government institutions. same study also shows that when countervailing 33 PARTICIPATION, TRANSPARENCY AND ACCOUNTABILITY IN VILLAGE LAW IMPLEMENTATION References AKATIGA. (2010). Marginalized Groups in PNPM Rural. Jakarta: PNPM Support Facility. AKATIGA. (2012). Evaluation of PNPM RESPEK: Village Infrastructure and Institutional Capacity. Jakarta: PNPM Support Facility. Alatas, V., Banerjee, A., Hanna, R., Olken, B., Purnamasari, R., & Wai-Poi, M. (2013). Does Elite Capture Matter? Local Elites and Targeted Welfare Programs in Indonesia. Cambridge, MA: NBER. Antlov, H., Wetterberg, A., & Dharmawan, L. (2016). Village Governance, Community Life, and the 2014 Village Law in Indonesia. Bulletin of Indonesian Economic Studies, 161-183. Baiocchi, G. (2003). Participation, Activism, and Politics: The Porto Alegre Experiment. In A. Fung, & E. O. Wright, Deepening Democracy: Institutional Innovations in Empowered Participatory Governance. London: Verso. Baldwin, K., Muyengwa, S., & Mvukiyehe, E. (2017). Reforming Village-Level Governance via Horizontal Pressure: Evidence from an Experiment in Zimbabwe. Washington, DC: The World Bank. Bandiera, O., & Levy, G. (2011). Diversity and the Power of the Elites in Democratic Societies: Evidence from Indonesia. Journal of Public Economics, 1322-1330. Boulding, C., & Wampler, B. (2010). Voice, Votes, and Resources: Evaluaating the Effect of Participatory Democracy on Well-being. World Development, 125-135. Clearly, M. R. (2007). Electoral Competition, Participation, and Government Responsiveness in Mexico. American Journal of Political Science, 283-299. Dasgupta, A., & Beard, V. A. (2007). Community Driven Development, Collective Action and Elite Capture in Indonesia. Development and Change, 229-249. Dharmawan, L. R. (2014). Studi Kelompok Masayarakat PNPM. Jakarta: TNP2K. Dharmawan, L., Dewayanti, R., & Nugraheni, I. (2014). Studi Keompok Masyarakat PNPM. Jakarta: TNP2K. Fox, J. (2007). The Uncertain Relationship between Transparency and Accountability. Development in Practice, 663-671. Grindle, M. S. (2007). Going Local: Decentralization, Democratization, and the Promise of Good Governance. Princeton: Princeton University Press. Joshi, A. (2014). Reading the Local Context: A Causal Chain Approach to Social Accountability. IDS Bulletin, 23-35. Kurniawan, A., Sedyadi, G. S., Kartawijaya, Syukri, M., Bachtiar, P., Diningrat, R. A., & Alifia, U. (2017). Studi Implementaasi Undang-Undang No. 6 Tahun 2014 tentang Desa. Jakarta: SMERU. Mansuri, G., & Rao, V. (2013). Localizing Development: Does Participation Work? Washington, DC: The World Bank. McLaughlin, K., Satu, A., & Hoppe, M. (2007). Kecamatan Development Program Qualitative Impact Evaluation. Jakarta: The World Bank. 34 PARTICIPATION, TRANSPARENCY AND ACCOUNTABILITY IN VILLAGE LAW IMPLEMENTATION Narayan, D. (. (2002). Empowerment and Poverty Reduction: A Sourcebook. Washington, DC: The World Bank. Neil, N. (2013). PNPM Mandiri Rural Infrastructure Technical Evaluation Report (Summary). Jakarta: PNPM Support Facility. O'Meally, S. (2013). Mapping Context for Social Accountability: A Resource Paper. Washington, DC: Social Development Department, The World Bank. Pokja Pengendali PNPM. (2012). PNPM Mandiri Information Kit 2012-2013. Jakarta: Sekretariat Pokja Pengendali PNPM Mandiri. PNPM Support Faility (PSF). (2012). PNPM Incident of Benefit Study: Overview findings from the Household Social Economic Survey 2012 (SUSETI). Jakarta: PNPM Support Facility PSF. (2014a). PNPM Governance Six-Monthly Update: February 2014. Jakarta: PNPM Support Facility. PSF. (2014b). Progress Report 2013. Jakarta: PNPM Support Facility. PSF. 2015. Beneficiary Assessment of PNPM/RESPEK. Jakarta: PNPM Support Facility. Sambhodo, P. (a, forthcoming). Understanding villagers' non-participation in village deliberative forums. Sambhodo, P. (b, forthcoming). Friend or foe? Village activists as a horizontal accountability and representation mechanism. Syukri, M., Mawardi, S., & Akhmadi. (2013). A Qualitative Study on the Impact of the PNPM-Rural in East Java, West Sumatra, and Southeast Sulawesi. Jakarta: SMERU. Syukri, M., Akhmadi, Hastuti, Kartawijaya, & Kurniawan, A. (2014). Qualitative Study of the Proliferation and Integration of Community Empowerment Programs in Central Java, West Nusa Tenggara and South Sulawesi. Jakarta: PNPM Support Facility. Voss, J. (2008). Impact Evaluation of the Second Phase of the Kecamatan Development Program in Indonesia. Jakarta: The World Bank and Decentralization Support Facility. Voss, J. (2013). PNPM rural Impact Evaluation. Jakarta: PNPM Support Facility. Wetterberg, A. (2014). Transitioning Community Driven Development Projects into Policy: From PNPM Mandiri to the Village Law. Wetterberg, A., Jellema, J. R., & Dharmawan, L. (2014). The Local Level Institutions Study 3: Overview Report. Jakarta: Coordinating Ministry for People's Welfare, TNP2K, and PNPM Support Facility. Woodhouse, A. (2012). Governance Review of PNPM Rural. Jakarta: PNPM Support Facility. World Bank. (2017). World Development Report 2017: Governance and the Law. Washington, DC: The World Bank. 35 36 Annex 1. Characteristics of invitees and non-invitees, and their attendance at village meetings All respondents Respondents who are invited Respondents who are not invited All All All Characteristics Invited Not Difference Attended Did not Difference Attended Did not Difference respondents respondents respondents invited attend attend Individual Female 51.8 52.2 51.6 0.5 52.2 37.5 59.6 -22.2 ** 51.6 52.7 51.6 1.1 Age (years) 47.0 47.8 46.6 1.2 47.8 46.1 48.6 -2.5 ** 46.6 46.4 46.6 -0.2 No formal education 24.5 20.3 27.1 -6.8 ** 20.3 12.0 24.5 -12.5 ** 27.1 21.8 27.4 -5.6 Completed Primary 38.9 37.3 39.9 -2.6 37.3 36.8 37.6 -0.8 39.9 36 40.1 -4.1 Completed Junior Secondary 17.4 18.7 16.6 2.1 18.7 18.4 18.9 -0.5 16.6 23.8 16.2 7.5 Completed Senior Secondary and beyond 19.2 23.7 16.4 7.3 ** 23.7 32.9 19.0 13.8 ** 16.4 18.4 16.3 2.1 Currently working 76.1 76.6 75.8 0.8 76.6 82.5 73.6 9 ** 75.8 79.8 75.6 4.2 Belongs to ethnic majority group 92.5 94.2 91.4 2.8 ** 94.2 89.9 96.4 -6.4 ** 91.4 94 91.3 2.7 Belongs to religion majority group 98.7 99.0 98.5 0.5 99.0 99.3 98.9 0.3 98.5 100 98.5 1.5 ** # yrs resides in village (years) 36.5 37.4 36.0 1.4 37.4 35.8 38.2 -2.5 36 34.7 36 -1.3 Active in local organization 87.0 89.0 85.8 3.2 * 89.0 96.1 85.4 10.7 ** 85.8 91.4 85.5 5.9 Active in political party 5.1 6.1 4.6 1.5 * 6.1 11.0 3.5 7.5 ** 4.6 10.4 4.3 6.1 Household Household headed by female 9.8 10.1 9.7 0.5 10.1 5.8 12.3 -6.5 ** 9.7 3.8 10 -6.2 ** Bottom-40 38.9 36.1 40.6 -4.5 * 36.1 31.5 38.4 -7 * 40.6 32.3 41 -8.7 Distance to Village Head Office (km) 0.8 0.8 0.9 -0.1 0.8 0.7 0.9 -0.2 0.9 1 0.9 0.2 Village Head resides in hamlet 27.1 29.2 25.8 3.4 29.2 31.3 28.1 3.1 25.8 26.8 25.7 1.1 Village apparatus reside in hamlet 78.7 75.6 80.5 -4.9 * 75.6 84.0 71.4 12.6 ** 80.5 70.5 81 -10.5 * Opinions and perception Submitted complains or reported 9 11.9 7.2 4.6 ** 11.9 23.9 5.8 18.1 ** 7.2 19.8 6.6 13.2 ** problems PARTICIPATION, TRANSPARENCY AND ACCOUNTABILITY IN VILLAGE LAW IMPLEMENTATION Village Head is reliable in planning 74.9 79.3 72.2 7.1 ** 79.3 76.8 80.5 -3.6 72.2 86.8 71.5 15.4 ** Hamlet Head is reliable in planning 78.5 81.0 76.9 4.2 ** 81.0 75.6 83.8 -8.1 ** 76.9 81.2 76.6 4.5 BPD Head is reliable in planning 58.3 65.2 54.1 11.0 ** 65.2 74.0 60.7 13.4 ** 54.1 67.4 53.5 13.9 ** Village Head is reliable in implementation 77.5 80.5 75.7 4.9 ** 80.5 79.4 81.1 -1.7 75.7 87.2 75.1 12.1 ** Hamlet Head is reliable in implementation 80.8 83.4 79.2 4.3 ** 83.4 79.3 85.5 -6.2 ** 79.2 85.9 78.8 7 * BPD Head is reliable in implementation 59 65.1 55.2 9.9 ** 65.1 74.7 60.2 14.5 ** 55.2 65.2 54.7 10.4 Observations 4,081 1,691 2,390 1,691 835 856 2,390 158 2,232 Notes: all figures are in percentage unless otherwise stated; ** statistically significant at 5 percent; * statistically significant at 10 percent Annex 2. Village and hamlet meeting attendees and non-attendees Village meetings Hamlet meetings All All Characteristics Attended Did not Difference Attended Did not Difference respondents respondents attend attend Individual Female 51.8 40.3 54.0 -13.7 ** 51.8 40.4 56.0 -15.6 ** Age (years) 47.0 46.2 47.2 -1.0 47.0 48.6 46.4 2.2 ** No formal education 24.5 13.8 26.5 -12.7 ** 24.5 23.9 24.7 -0.8 Completed Primary 38.9 36.7 39.4 -2.7 38.9 37.0 39.6 -2.7 Completed Junior Secondary 17.4 19.3 17.0 2.3 17.4 17.6 17.3 0.3 Completed Senior Secondary and beyond 19.2 30.2 17.1 13.1 ** 19.2 21.5 18.3 3.2 Currently working 76.1 82.0 75.0 7.0 ** 76.1 86.0 72.5 13.6 ** Belongs to ethnic majority group 92.5 90.7 92.8 -2.1 92.5 95.9 91.2 4.7 ** Belongs to religion majority group 98.7 99.4 98.6 0.8 * 98.7 98.4 98.9 -0.5 # yrs resides in village (years) 36.5 35.6 36.7 -1.1 36.5 38.3 35.9 2.4 ** Active in local organization 87.0 95.3 85.5 9.8 ** 87.0 96.5 83.6 12.9 ** Active in political party 5.1 10.9 4.1 6.9 ** 5.1 4.7 5.3 -0.7 Household Household headed by female 9.8 5.5 10.7 -5.2 ** 9.8 6.3 11.1 -4.9 ** Bottom-40 38.9 31.6 40.2 -8.6 ** 38.9 36.1 39.9 -3.8 Distance to Village Head Office (km) 0.8 0.8 0.9 -0.1 0.8 1.1 0.8 0.3 ** Village Head resides in hamlet 27.1 30.5 26.4 4.0 27.1 16.9 30.8 -13.9 ** Village apparatus reside in hamlet 78.7 81.5 78.1 3.4 78.7 67.3 82.8 -15.5 ** Opinions and perception Submitted complains or reported 9.0 23.1 6.4 16.7 ** 9.0 13.6 7.3 6.3 ** problems Village Head is reliable in planning 74.9 78.7 74.2 4.5 ** 74.9 80.5 72.8 7.7 ** Hamlet Head is reliable in planning 78.5 76.7 78.8 -2.1 78.5 83.5 76.6 6.9 ** BPD Head is reliable in planning 58.3 72.8 55.6 17.2 ** 58.3 66.5 55.4 11.1 ** Village Head is reliable in implementation 77.5 80.8 76.9 3.9 * 77.5 83.1 75.5 7.6 ** Hamlet Head is reliable in implementation 80.8 80.5 80.8 -0.3 80.8 85.6 79.0 6.6 ** BPD Head is reliable in implementation 59.0 72.9 56.4 16.6 ** 59.0 68.2 55.6 12.6 ** Observations 4,081 993 3,088 4,081 1,274 2,807 Notes: all figures are in percentage unless otherwise stated; ** statistically significant at 5 percent; * statistically significant at 10 percent PARTICIPATION, TRANSPARENCY AND ACCOUNTABILITY IN VILLAGE LAW IMPLEMENTATION 37 38 Annex 3. Cited reasons for non-attendance and probable conditions for future attendance Reasons for not attending village meetings Will attend future village meetings if All res- Not invited Busy Irrelevant/already represented Invited Meetings are made mandatory Characteristics pondents Cited Not cited Difference Cited Not cited Difference Cited Not cited Difference Cited Not cited Difference Cited Not cited Difference Individual Female 54.0 51.6 59.6 -8.1 ** 58.3 53.1 5.2 * 59.9 52.8 7.1 ** 51.3 62.6 -11.4 ** 60.1 53.2 6.9 ** Age (years) 47.2 46.6 48.6 -2.0 ** 45.4 47.6 -2.2 ** 47.4 47.1 0.3 46.8 48.6 -1.8 * 50.3 46.8 3.5 ** No formal education 26.5 27.4 24.5 2.8 21.8 27.5 -5.8 ** 22.9 27.3 -4.4 26.1 28.0 -2.0 29.3 26.2 3.2 Completed Primary 39.4 40.1 37.6 2.6 33.6 40.5 -6.9 ** 45.3 38.1 7.2 ** 40.5 35.8 4.7 * 35.7 39.8 -4.1 Completed Junior Secondary 17.0 16.2 18.9 -2.7 19.6 16.5 3.1 17.6 16.9 0.7 16.6 18.4 -1.8 17.4 17.0 0.4 Completed Senior Secondary and beyond 17.1 16.3 19.0 -2.7 25.0 15.5 9.6 ** 14.3 17.7 -3.4 16.9 17.8 -1.0 17.6 17.1 0.5 Currently working 75.0 75.6 73.6 2.1 81.4 73.7 7.7 ** 73.6 75.3 -1.7 76.8 69.4 7.3 ** 63.9 76.4 -12.5 ** Belongs to ethnic majority group 92.8 91.3 96.4 -5.0 ** 94.0 92.6 1.4 92.2 93.0 -0.8 91.4 97.4 -6.0 ** 94.5 92.6 1.9 Belongs to religion majority group 98.6 98.5 98.9 -0.5 98.3 98.7 -0.4 99.6 98.4 1.2 ** 98.6 98.7 -0.1 98.6 98.6 -0.1 # yrs resides in village (years) 36.7 36.0 38.2 -2.2 * 33.5 37.4 -3.8 ** 37.3 36.6 0.8 36.1 38.4 -2.3 40.9 36.1 4.8 ** Active in local organization 85.5 85.5 85.4 0.1 81.0 86.4 -5.4 ** 89.2 84.7 4.5 ** 86.3 83.1 3.1 * 82.9 85.8 -3.0 Active in political party 4.1 4.3 3.5 0.8 3.8 4.1 -0.4 2.7 4.4 -1.7 4.2 3.7 0.5 6.4 3.8 2.6 Household Household headed by female 10.7 10.0 12.3 -2.3 9.1 11.0 -1.9 12.6 10.3 2.4 10.4 11.6 -1.3 11.0 10.6 0.4 Bottom-40 40.2 41.0 38.4 2.6 33.7 41.6 -7.9 ** 42.4 39.8 2.6 39.2 43.3 -4.1 41.4 40.1 1.3 Distance to Village Head Office (km) 0.9 0.9 0.9 0.0 0.7 0.9 -0.2 * 0.9 0.8 0.1 0.8 1.0 -0.2 0.8 0.9 0.0 Village Head resides in hamlet 26.4 25.7 28.1 -2.4 31.6 25.4 6.3 26.9 26.3 0.6 25.5 29.5 -4.1 26.2 26.5 -0.2 Village apparatus reside in hamlet 78.1 81.0 71.4 9.6 ** 78.8 78.0 0.8 74.1 79.0 -4.9 * 80.2 71.6 8.6 ** 78.8 78.0 0.8 Opinions and perception Submitted complains or reported 6.4 6.6 5.8 0.8 6.4 6.4 0.1 6.6 6.3 0.3 6.6 5.8 0.8 8.3 6.1 2.2 problems Village Head is reliable in planning 74.2 71.5 80.5 -9.0 ** 79.8 73.0 6.8 ** 76.0 73.8 2.2 73.0 77.9 -5.0 ** 71.8 74.5 -2.7 Hamlet Head is reliable in planning 78.8 76.6 83.8 -7.1 ** 81.1 78.3 2.8 83.6 77.8 5.8 ** 78.7 79.0 -0.3 76.7 79.1 -2.4 BPD Head is reliable in planning 55.6 53.5 60.7 -7.2 * 61.0 54.5 6.5 * 59.4 54.8 4.6 * 55.5 55.9 -0.4 53.6 55.9 -2.3 Village Head is reliable in implementation 76.9 75.1 81.1 -6.0 ** 79.8 76.3 3.5 79.0 76.5 2.6 76.4 78.5 -2.1 73.8 77.3 -3.5 Hamlet Head is reliable in implementation 80.8 78.8 85.5 -6.7 ** 84.1 80.1 4.0 ** 84.5 80.1 4.5 ** 80.5 82.0 -1.5 81.7 80.7 1.0 BPD Head is reliable in implementation 56.4 54.7 60.2 -5.5 60.4 55.5 4.9 59.2 55.8 3.4 56.4 56.1 0.4 56.7 56.3 0.4 Observations 3,088 2,232 856 496 2,592 651 2,437 2,418 670 379 2,709 PARTICIPATION, TRANSPARENCY AND ACCOUNTABILITY IN VILLAGE LAW IMPLEMENTATION Notes: all figures are in percentage unless otherwise stated; ** statistically significant at 5 percent; * statistically significant at 10 percent Annex 4. Probit estimates of attending village meetings (marginal effects) All Observations Women Men Coef SE Coef SE Coef SE Individual Woman -0.051 (0.014) ** Age 0.007 (0.004) * 0.008 (0.004) * 0.011 (0.006) * Age^2 0.000 (0.000) ** 0.000 (0.000) ** 0.000 (0.000) * Completed Primary 0.044 (0.021) * 0.013 (0.027) 0.066 (0.029) ** Completed Junior Secondary 0.099 (0.029) ** 0.065 (0.037) * 0.115 (0.044) ** Completed Senior Secondary and beyond 0.127 (0.028) ** 0.043 (0.037) 0.193 (0.039) ** Work 0.007 (0.017) 0.014 (0.015) -0.011 (0.044) Belongs to major ethnic group 0.017 (0.024) 0.027 (0.026) 0.004 (0.037) Belongs to major religion group 0.111 (0.072) 0.111 (0.072) 0.112 (0.118) #yrs residing in village 0.001 (0.001) 0.000 (0.001) 0.001 (0.001) Active in local organization 0.126 (0.028) ** 0.113 (0.045) ** 0.142 (0.039) ** Active in political party 0.081 (0.022) ** 0.090 (0.054) 0.090 (0.025) ** Household Female-headed household -0.058 (0.020) ** -0.038 (0.025) -0.090 (0.071) HH members aged 0-4 -0.026 (0.015) * -0.050 (0.019) ** 0.000 (0.020) HH members aged 5-14 -0.013 (0.009) -0.014 (0.012) -0.018 (0.013) HH members aged 60+ -0.002 (0.012) -0.002 (0.016) 0.003 (0.015) Welfare 0.012 (0.005) ** 0.010 (0.006) 0.016 (0.008) * Welfare^2 0.001 (0.003) 0.000 (0.002) 0.002 (0.004) Distance to Village Head Office 0.002 (0.007) 0.001 (0.011) 0.002 (0.008) Enabling environment Village Head formerly PNPM actor 0.009 (0.025) -0.007 (0.025) 0.028 (0.035) BPD Head formerly PNPM actor 0.007 (0.018) 0.007 (0.021) 0.010 (0.023) Local Activists formerly PNPM actor 0.004 (0.032) -0.027 (0.036) 0.038 (0.050) Presence of PD/PLD 0.029 (0.028) 0.019 (0.028) 0.038 (0.036) Observations 4,081 2,125 1,956 Note: Probit specifications also control for kabupaten, but not shown; ** statistically significant at 5 percent; * statistically significant at 10 percent PARTICIPATION, TRANSPARENCY AND ACCOUNTABILITY IN VILLAGE LAW IMPLEMENTATION 39 PARTICIPATION, TRANSPARENCY AND ACCOUNTABILITY IN VILLAGE LAW IMPLEMENTATION Annex 5. Respondents' characteristics by their awareness of last year's village program and fund use Village programs Village funds use All Characteristics Aware Not Difference Aware Not Difference respondents aware aware Individual Female 51.8 42.2 60.6 -18.4 ** 36.2 53.6 -17.5 ** Age (years) 47.0 46.6 47.4 -0.9 46.6 47.1 -0.5 No formal education 24.5 19.0 29.6 -10.6 ** 11.3 26.0 -14.7 ** Completed Primary 38.9 40.5 37.5 3.0 35.1 39.4 -4.2 Completed Junior Secondary 17.4 18.4 16.5 2.0 23.9 16.6 7.2 ** Completed Senior Secondary and beyond 19.2 22.2 16.4 5.7 ** 29.7 18.0 11.7 ** Currently working 76.1 79.1 73.4 5.7 ** 81.9 75.4 6.4 * Belongs to ethnic majority group 92.5 94.4 90.7 3.7 ** 97.5 91.9 5.6 ** Belongs to religion majority group 98.7 98.7 98.8 -0.1 97.2 98.9 -1.8 # yrs resides in village (years) 36.5 36.4 36.6 -0.2 35.4 36.6 -1.2 Active in local organization 87.0 93.0 81.6 11.4 ** 96.1 86.0 10.1 ** Active in political party 5.1 7.6 2.9 4.8 ** 12.0 4.4 7.6 ** Household Household headed by female 9.8 9.2 10.4 -1.2 9.5 9.9 -0.4 Bottom-40 38.9 35.1 42.3 -7.3 ** 26.2 40.3 -14.1 ** Distance to Village Head Office (km) 0.8 0.8 0.9 -0.1 ** 0.9 0.8 0.0 Village Head resides in hamlet 27.1 27.7 26.5 1.2 27.2 27.1 0.2 Village apparatus reside in hamlet 78.7 78.9 78.4 0.5 75.1 79.1 -4.0 Opinions and perception Submitted complains or reported problems 9.0 13.7 4.8 8.9 ** 25.5 7.1 18.4 ** Village Head is reliable in planning 74.9 79.3 70.8 8.5 ** 81.7 74.1 7.6 ** Hamlet Head is reliable in planning 78.5 80.1 77.0 3.1 83.4 77.9 5.5 BPD Head is reliable in planning 58.3 65.7 51.6 14.1 ** 77.2 56.2 21.0 ** Village Head is reliable in implementation 77.5 81.4 73.9 7.5 ** 82.0 77.0 5.0 * Hamlet Head is reliable in implementation 80.8 83.3 78.5 4.8 ** 83.4 80.5 2.9 BPD Head is reliable in implementation 59.0 66.6 51.9 14.7 ** 77.5 56.8 20.7 ** Observations 4,081 1,919 2,162 446 3,635 Notes: all figures are in percentage unless otherwise stated; ** statistically significant at 5 percent; * statistically significant at 10 percent 40 PARTICIPATION, TRANSPARENCY AND ACCOUNTABILITY IN VILLAGE LAW IMPLEMENTATION Annex 6. Desire to know village information Does not All Wants to Characteristics want to Difference respondents know know Individual Female 51.8 47.2 60.0 -12.77 ** Age (years) 47.0 45.1 50.4 -5.36 ** No formal education 24.5 20.5 31.5 -10.98 ** Completed Primary 38.9 37.9 40.7 -2.84 Completed Junior Secondary 17.4 18.7 15.1 3.62 Completed Senior Secondary and beyond 19.2 22.9 12.7 10.19 ** Currently working 76.1 76.6 75.3 1.24 Belongs to ethnic majority group 92.5 92.0 93.4 -1.42 Belongs to religion majority group 98.7 99.1 98.1 1.07 ** # yrs resides in village (years) 36.5 34.5 40.0 -5.45 ** Active in local organization 87.0 88.8 84.0 4.81 ** Active in political party 5.1 6.5 2.7 3.78 ** Household Household headed by female 9.8 8.7 11.9 -3.21 ** Bottom-40 38.9 39.3 38.1 1.19 Distance to Village Head Office (km) 0.1 0.2 0.1 0.09 Village Head resides in hamlet 27.1 30.1 21.8 8.34 ** Village apparatus reside in hamlet 78.7 81.5 73.6 7.91 ** Opinions and perception Attended village meetings 15.7 20.3 7.7 12.60 ** Attended hamlet meetings 26.7 28.0 24.4 3.65 Submitted complains or reported problems 9.0 12.9 2.3 10.61 ** Village Head is reliable in planning 74.9 72.7 78.6 -5.86 ** Hamlet Head is reliable in planning 78.5 76.9 81.1 -4.21 ** BPD Head is reliable in planning 58.3 61.6 52.6 9.02 ** Village Head is reliable in implementation 80.8 79.5 83.0 -3.45 * Hamlet Head is reliable in implementation 77.5 75.4 81.2 -5.82 ** BPD Head is reliable in implementation 59.0 62.3 53.1 9.27 ** Observations 4,081 2,757 1,324 Notes: all figures are in percentage unless otherwise stated; ** statistically significant at 5 percent; 41 PARTICIPATION, TRANSPARENCY AND ACCOUNTABILITY IN VILLAGE LAW IMPLEMENTATION Annex Annex 7. Sampling 7. Sampling Weight Weight Approach Approach Annex 7. Sampling Weight Approach Annex Annex Annex 7. Sampling 7. Sampling 7. Sampling Weight Weight Weight Approach Approach Approach Based Based on the on the sampling sampling methodology methodology described described in Section in Section 3.2,3.2, we can we can thenthen construct construct sampling sampling plan plan Based tables the sampling on and tables and sampling sampling design methodology design weight described weight for for eachineach Sectionsampling 3.2, units. sampling units. we can then construct sampling Based onBased Based the on plan on thethe sampling tables sampling sampling and methodology methodology methodology sampling described design described described weight ineach for in Section Section in 3.2, Section sampling we3.2, can units.3.2, we we can then can then then construct construct construct sampling sampling sampling plan plan plan tables tables tables and 1.andand sampling Sampling sampling sampling design weightdesign design weightforweight 1. Sampling weight for an individual dataanweight for eachfor for individual eacheach data sampling sampling sampling units. units. units. 1.Sampling SamplingSamplingplan plan table weight table for for for anpersonperson individual selection selection data andand sampling sampling weight weight for person for person data data are are described described 1. 1. Sampling below. 1. Sampling below. Sampling weight weight weight for an for for an an individual individual individualdata data data and sampling weight for person data are described below. Sampling plan table for person selection Sampling Sampling Sampling plan table plan plan table table for forfor person person person selection selection selection and and samplingand sampling sampling weight weight weight for forfor person person person data data aredata are are described described described th below. below. below. Annex Annex Annex Table TableTable 7.1. 7.1. 7.1. Sampling Sampling Sampling scheme scheme scheme table table table forforfor persons persons persons selection selection selection inin in each each each selected selected selected district-i district-i th district-ith StageStage Sampling Sampling Stratum Universe Sample Sampling Weight unit Stratum Universe Sample in Sampling Weight th th Annex AnnexAnnex Table Table Table Stage 7.1. 7.1. 7.1. Sampling Sampling Sampling Sampling scheme scheme table table Stratum forfor persons persons Universe selection selection Sample in each each selected selected Sampling district-i district-i th Weight unit scheme table for persons selection in each selected district-i unit Stage StageStage Sampling1 Sampling 1 1 Sampling Sub-district Stratum Sub-district Sub-district Stratum Stratum - - Universe Universe - Universe ! !Sample Sample Sample Sampling Sampling ! Sampling Random RandomRandom Weight Weight Weight ! ! ! unit unit unit (A, j) (A, j) (A, j) ! ! Sub-district 1 2 Sub-district 1 Sub-district - - Random Random 1 2 Village 2 Village Village - - -! - ! ! !" !"! ! !" =!"Random ! 4 = 4 Random RandomRandom ! !" ! !!" !" = !" j) j) (A, j) (A,(A, (B, k) (B, k) ! = ! (B, k) !" !!" 4 4 2 Village 2 3 Village 2 Village Hamlet - - - = 4 = 4=Random4 Random Random !" !" - - Random !" !" !" !" 3 3 HamletHamlet -!" !" !"# !" !"# !" !"# = !"#1 = 1 Random Random !" = !" = = !"#!"# !"# k) k) (B, k) (B,(B, !"# = (C, (C, l) l) 4 !" = 4 4 (C, l) !" !" !"# !"# 1 1 3 3 3 Hamlet Hamlet Hamlet - - - = 1 = Random1 Random Random - !"# !"# =1 !"# !"# = !"# !"#$ 4 4 4 HouseholdHousehold - Random !"# ! "#= = 20Random !"# !"# Household - !"#$ !"# !"#$ !"# !"#$ !"#$ 20 Random !"# = = !"#$ !"#$ l) l) (C, l) (C,(C, !"#$ = (D, m) (D, m) 1 !"# = 1 !"#$ 20 201 (D, m) !"# !"# !"#$ 4 4 4 Household Household- Household - - !"#$!"#$ = !"#$ 20 !"#$ = 20 = Random20 Random Random !"#$ !"#$ !"#$ !"#$ 5 5 Person 5 PersonPerson Gender: Gender: Gender: !"#$ !"#$ !"#$ = !"#$ = = (D, m) (D, m)(D, m) (E, n) (E, n) !"#$ !"#$ !"#$ 20 20 20 (E, n) 5 5 5 Person Person Person Gender: Gender: Gender: Male Male (1) (1) Male (1) (!) ( ! ) !"#$% (!)!"#$% ( ! ) Stratified = 1Stratified Stratified (! ) (! ) (! ) (! ) !"#$% !"#$% = 1 !"#$% n) n) (E, n) (E,(E, !"#$% !"#$% Random Random Random !"#$% ( ! = ) = Male Male (1) Male(1) (1)(! ) (! ) (! ( !"#$% )!) (! ) (! ) =Stratified Stratified = 1 Stratified (! ) (! )(( ! ! ) ) !"#$% (! ) (!1) 1 !"#$% !"#$% !"#$% !"#$% = !"#$% 1!"#$% 1 !"#$% !"#$% !"#$% !"#$% Female Female (2) (! ) (! ) (! ) (Random !) Random Random = )( !"#$% =(! = ) (! ) (! ) Female !"#$% !"#$% !"#$% !"#$%= 1= 1 (!) (! (!!) 1 ) !"#$% 1 !"#$% !"#$% 1 (2) (2) !"#$% !"#$% !"#$% !"#$% = ( ! = ) Female Female Female (! ) (! ) (! )(! ) !"#$% (! ) (! ) =1 (! ) (! ) (( ! ! )) (! ) (! 1) 1 !"#$% !"#$% !"#$% !"#$% = !"#$% 1!"#$% =1 !"#$% !"#$% !"#$% !"#$% !"#$% (2) (2)(2) !"#$% = ) (! ) = = 1 (! ) (! 1 be 1 BasedBased on the the sampling on sampling plan plan table table above, above, the sex-specific the sex-specific sampling sampling weight !"#$% weight can !"#$% can !"#$% be calculated. calculated. Based on the sampling plan table above, the sex-specific sampling weight can be calculated. Based Based onBased Weightthe Weight Weighton on the sampling for the for for sampling male sampling male male plan respondents: plan table respondents: respondents: plantable above, table above, the above, the sex-specific the sex-specific sex-specific sampling sampling sampling weight weight can weight be can can be be calculated. calculated. calculated. ) (! ) ! !"# !"#$ !"#$% ! !" !"#$ ( ! ( ! ) ! ! ; ! ! ( ! ) !" Weight WeightWeight for male forformale male respondents: (!respondents: !"#$% ) respondents: = = !"#$% ! !"#$% ;! !"#$%= = ! × ××4 !" !"#×× !"#!"#$ × × (× !) !"#$% = = ! !" !"# !"#$ !"#$% ! !! !4 20 20 !"#$% 80 80 (! ) ! ! ! ! ! !" ( ! !"#$ ) ! ! ( ! ) ( ! )( ! !) ! ; !! ; ! ! ! !" !" !"#$ !"#$ ( ! )( !! )!" !"# ! !"!!"#$ !"# !"# !"#$ !"#$% !"#$% (!) = = ! ;! !"#$% !"#$% = !" =× × × × ×× !"#$× ×× (!)× ×!"#$% = = !"#$% !"#$% = where, = × !"#$% = !"#$% !"# !"# where, !"#$% ! !! ! !! !"#$% !!! ! 4 ! ! !"# 4 4 20 20 20 !"#$% 80! 80 80 ! ! ( ! ) (! !"#$% ) !"#$%is male is male weight weight in selected in selected household-household- m, m , selected selected hamlet- hamlet- l, selected l, selected village- village-k, k, where, where,where, where, (! ) (! ) selected selected sub-district- sub-district- j, and j, and selected selected district- district- i, i, (! ) !"#$%is is isismale male male weight weight weightin in selected inselected selected household-m, household- household- selected m , m , selected selected hamlet-l, i, selected l,hamlet- hamlet- , selected l, lselected village-k, k,village- village- k, k, ! ! is number of sub-districts in selected district-i, hamlet- !"#$% male !"#$% weight is numberin selected of sub-districts household- in m selected , selected district- selected village- ! selected selected ! is selected selected sub-district- number sub-district- sub-district-j, sub-district- is number of selected and ofj,selected ,andand jselectedj, sub-districts and selected selected selected sub-districts district- district-i, , district- district- iin selected , i, district- in iselected district- i, i, ! !is number ! !" is !" isis number number number of is of number of sub-districts of sub-districts sub-districts of sub-districtsvillages in selected in in is number of villages in selected sub-district-j and selected selected inin selected selected selected district- district- district-i,i district- sub-district- , i , i , j and selected district- district- i, i, !isnumber ! is is number is number number of is of number selected of of selected selected of selected selected sub-districts sub-districts sub-districts sub-districts villages in in selected in in selectedin selectedselected selected district- district- district- district-i, i sub-district- , i , i , j and selected district- i, = !" ! !" !" is number of selected villages in selected sub-district-j and selected district- i, !" 4= 4 is is is number number of of villages villages in in selectedselected sub-district- sub-district- j and j and selectedselected district-district- i , i , !" number is number ofis villages in selected of villages number of in selected hamlets sub-district-sub-district-j in selected j and and selected village-selected k, district- district-i, selected i, sub-district-j, and selected !" !" !"# !"# is number of hamlets in selected village-k, selected sub-district-j, and selected !" !" is!" number is is is number number of number selected district- ofof of selected selected villages selected villages villages villages in selectedin in in selected selected selected sub-district- sub-district-j sub-district- sub-district- j and and j and selected j selectedand selected selected district- district-i, i, district- district- !" = , i!" i, 4 = !" 4 =4 district- i, i, !"# !"# is!"# is number isis number number of number is hamlets number of of ofhamlets hamlets hamletsof in in selectedin in selected selectedselected hamlets village- village-k, village- k, village- in selected k, k selected selected , selected selected sub-district- sub-district-j,village- sub-district- sub-district- k , j, andj,sub-district- selected j, sub-district- and selected and selected selected j, and !"# !"# is number of selected hamlets in selected village- k , and selected selected j, and district- , district- district- idistrict-i, i , selectedi, selected district-i, !"# =district- i , !"# 1 = 1 !"# !is "#!"#numberis is is number number of selected number of of ofselected selected selected hamlets hamlets hamlets hamlets in selected in selected in in selectedselected village- village-k, k,village- village- selected selected k , selected selected k, sub-district-j, sub-district- sub-district- and sub-district- j, andj, and j, and selected selected selecteddistrict- selected district- district- i, !"# = district-i, i i, 1 , !"# !"# =1 = 1 39   39       39   39  39         42 PARTICIPATION, TRANSPARENCY AND ACCOUNTABILITY IN VILLAGE LAW IMPLEMENTATION !"#$ is number of households in selected hamlet-l, selected village-k, selected sub- !"#$ is !"#$ number district- isj, andnumber of selected households of households district- in i, selected in selected hamlet-hamlet- l, selected l, selected village-k village- , selected k, selected sub- sub- !"#$ !!"#$ is !"#$ is number is is number number number of of of of households households households households in in selected inin selected selected selected hamlet-l, hamlet- hamlet- hamlet- selected l, selected ,l, selected lvillage-k, selected village- selected village- village- kk , , selected selected k, subdistrict-j, selected sub-sub- sub- !"#$ is number district- of district-j, and selected district-i,j , and selected selected households district- i in , selected hamlet- l , selected village- k , selected district- district- district- is number ,ofand jselectedj , j , and and selected of selected selected selected district- district- district-i, i,in households i ,i , !"#$ is !"#$number sub-district- and j, selected and district-i, selected households district- !"#$ =in selected 20 selected hamlet-hamlet- l, selected l, selected village-k village- , selected k, selected ! is !"#$ number is isis number number number of of of selected ofselected selected selected selected households households households households in in selected in selected in selected selected hamlet-l, hamlet- hamlet- hamlet- l , l selected , l , selected village-k, selected kk selected village- village- village-k , selected , ,selected selected (!!"#$) !"#$ !"#$% sub-district- is sub-district- number sub-district-, and jof males j, and in selected selecteddistrict-household- district- i, !"#$ =i, 20 m !"#$ , = 20 selected selected hamlet- l , selected village- k , (! ) (! ) sub-district- sub-district- sub-district-j, j, and j and ,j , andand selected selected selected selecteddistrict- district-i, district- district- i i, !"#$ =,i , ! 20 !"#$ !"#$ = ! 20 !" !"#$% (!)is selected (! number ! !) ! !"#$% is number of males in sub-district- of males selected j,males and in selected selected household- district- household-im, , selected m , selected hamlet- hamlet-l, selected village-k, l , selected village- k, ! is !"#$% is number is is number number number of of of males of males males in in selected selected in selected selected household- household-m, household- household- m , selected m m , selected , selected selected hamlet-l, hamlet- hamlet- hamlet- selected l , selectedl ,l ,selected village-k,selected village- village- village- k , kk,, (! ) !"#$% !"#$% !"#$% selected is number selected sub-district- of sub-district- selected j, and s in j, and selected selected selected district- household- district- i, m , i,selected hamlet-l, selected village- (! ) (!selected ) selected selected selected sub-district- sub-district-j, sub-district-sub-district- j, and and j,j,and andselected selected selected selected district- district-i, district- district- i, i,i, (!) is !"#$% number is number of selected of selected s in selected s in selected household- household- ) m, selected hamlet- l, selected village-village- !"#$% m  , selected hamlet- selected l,selected ! ! ! (! !"#$% selected ! ,!"#$% ( k ! !"#$%is ) is is number number is numbersub-district- numberof of selected of ofselected selected j, s selected and sin selected selected in selected s sin selected inselected district- household-m, household- i, household- household- m !"#$% selected , = m selected m ( !,1,selected ) hamlet-l, selected hamlet- hamlet- hamlet- l , selectedl,lvillagek, selected ,selected village- village- village- (! ) k, selected k, selected selected sub-district- sub-district- sub-district-j, j, and and j, selectedand selected selected district- district-i, district- i,  !"#$% i , =   (!) !"#$% 1( !! ! ) ! = 1 k, selectedkk selected , ,selected sub-district- sub-district- sub-district- j, and j,j,and andselected selected selected district- district- district- i,i  , i,  !"#$% !! = !"#$% 1= ! 1! Weight for female respondents: !"#$% Weight Weight female forWeight for female forrespondents: female ! respondents: respondents: ! !" !"#$ ! !" !"# !"#$ !"#$% (! ) Weight Weight Weight for ) for !for (female female female !"#$% = ! !"!"# respondents: respondents: respondents: ! ;! = × ×!"# × × (! ) !"#$% = ! ( ! ) (! ) ( ! ) ! ! ! ! ;! ! ! ;! 4 !" ! !" 20 !"#$ !"#$ (!! ) !" !"# 80 !" !"#$ !"#!"#$% !"#$ (! (! ) !"#$% !)!! (! ) !"#$% =!"#$% (! !)! ! =! ! !"!"#!;= ! !"# ! ! !!× != !! !!!" × × ! !"#! ×!"!"×! "# × ( × ! ) !!"#$!"#$ × = !"#$% !) ! !!= !" ! ! !" !!"# ! ! !" ! !"# !"#$!"# ! !"#$!"#$ ! !"#$% !"#$% (! ) ! !"! ; !"#$ ( ! ) (! !"#$% !"#$% !!"#$% =!"#$% !!= !! !!! !"!"#!" ! =!"! !"# = !"# 4× ! ! × !4 × !"# ×!! 20 × !"# !"# × !× 20 !"#$% × !! = !"#$%= ! 80! 80! where, ! !! ! ! !!!! ! 4 !!! 4! 20 20 !"!"#$%!"#$% 80!80 !"! !! where, (!where, !"#$% ) where, is female weight in selected household-m, selected hamlet-l, selected village-k, where, (! ) where, where, (! ) is female weight !"#$% is (! ) selected (! !"#$% ! ! is sub-district- female ! ) female weight in jin weight selected selected , and inselectedselected household-m, household- household- district- m i,,selected selected m, hamlet-l, selected hamlet- selected hamlet- l, selected village-k, l, selected village-k village- , k, ! is female !"#$% isis female selected femaleweight weight weight sub-district-j, in in selected in selected and selected selected household- household- household- district-i, m , m m , , selected selected selected hamlet- hamlet- hamlet- l , selected selected l,l, selected village- village- village- k , kk,, (! )!"#$% !"#$% selected !"#$% is number selected sub-district- of femalessub-district- j, and in j, and selected selected selected district-district- household- i, m, selected i, hamlet- l , selected village- k , (! ) (! ) selectedis selected selected number of sub-district- sub-district- sub-district- females j, and j,j,and in selected andselected selected selected district- household-m, district- district-i, selected i,i, hamlet-l, selected village-k, !"#$% (! )is ( !!) !"#$% ! ! is number selected number of females sub-district- of females in j , andselected in selected selected household- district- household- i ,m, selected m, selected hamlet-hamlet- l, selected l, selected village-k village- , k, (! )!"#$% ! is !"#$% is numberis number number selected of ofof females sub-district-j, females females !"#$% selected sub-district-j, and selected district-i, in in selected and in selected selected selected household- household- household- district-i, m , m m selected , , selected selected hamlet- hamlet- hamlet- l, selected selected l,l,selected village- village- k, kk village- ,, !"#$% is selectednumber sub-district- of selected j, and females selected in district-i,household-m, selected hamlet-l, selected selected (! ) (! )selected is selected selected number of sub-district-sub-district- sub-district- selected j, and j,j,and females and selected in selected selected selected district- district- district-i, i,i, household-m, selected hamlet-l, selected !"#$% !"#$% (! )isvillage- !!)! ! is (number k number , of selected selected of selected females sub-district- females in j , and selected in selected selected household-district- household- i ,m ,(! selected ) m, selected = hamlet-hamlet- l, selected l, selected !"#$% ! !"#$% !"#$% is number village-k, is isnumber number selected of selected ofofselected sub-district-j, selected females females females and in selected inselected inselected selected district-i, household- household- household- m, selected !"#$% m m (! ,1 , selected )selected hamlet- hamlet- hamlet- l,l,selected l, selected selected (! ) village-k village- , selected k, selectedsub-district- sub-district- j, and selected j, and selected district-district- i, !"#$% i,= (! ) !"#$% 1 (! !)!! = 1 village- village- village- kk k, selected selected , ,selected sub-district- sub-district- sub-district- j, and j,j,and andselected selected selected district- district- district- i, !"#$% i,i,!"#$% != !"#$% 1= ! 1! The sampling The sampling design design weightweight at enumeration at enumeration area (ea) area (ea) levelare level are depicted depicted in in Annex Error! 7.1.Reference source The The sampling notsamplingfound. design . design weight weight at enumeration at enumeration area (ea) area level (ea) are level depicted are depicted in Error! in Error!Reference Reference The The Thesampling sampling source sampling not design found. design designweight . weight weight at at atenumeration enumeration enumeration area area area (ea) (ea)(ea)level level level are are aredepicted depicted depicted in in Error! Error! inError! Reference Reference Reference source not found.. source Annex source source Figure not7.1. notnot found.. found. found. . . Annex Annex Figure AnnexFigure 7.1. Figure 7.1.7.1. Annex Annex Annex FigureFigure Figure 7.1. 7.1.and Trimmed Weight - EA level 7.1. Design Design and Design Trimmed A. Cut-off Design and and Trimmed Weight Trimmed weight: Mean+1*SD - Weight -Weight EA level EA B. -weight: level Cut-off EA level Mean+1.5*SD DesignDesign and Trimmed and Trimmed Design and Trimmed Weight Weight --EA - EA level Weight EA level level 3,000 A. Cut-off A. Cut-off weight: weight: Mean+1*SD 3,000 Mean+1*SD B. Cut-off B. Cut-off weight: weight: Mean+1.5*SD Mean+1.5*SD A. Cut-off weight: Mean+1*SD A. Cut-o weight: Mean+1*SD 3,000A. Cut-off weight: Mean+1*SD B. Cut-off 3,000 B. B.Cut-off Cut-o weight: weight: weight:Mean+1.5*SD Mean+1.5*SD Mean+1.5*SD 3,000 2,000 3,000 2,000 3,000 3,000 3,000 3,000 3,000 3,000 2,000 2,000 2,000 2,000 1,000 1,000 2,000 2,000 2,000 2,000 2,000 2,000 1,000 0 1,000 1,000 0 1,000 1,000 1,000 1,000 1,000 1,000 1,000 1 20 0 40 60 80 100 112 1 20 0 40 60 80 100 112 0 0 0 0 Enumeration area 0 0 Enumeration area 10 20 1 40 20 60 40 80 60100 112 80 100 112 1 0 20 1 40 20 60 40 80 60100 112 80 100 112 1 11 2020 Enumeration Enumeration 20 40 40 6040 60 area 60 80 area80 80 100 100 112 100112 112 1 11 20 20 4020 Enumeration Enumeration 40 area 6040 60 60 80 area80 80 100 100 112 112 100 112 C. Cut-off weight: area area Enumeration Mean+1.8*SD Enumeration Enumeration area D. Cut-off weight:Enumeration area area Mean+2*SD Enumeration Enumeration area 3,000 C. Cut-off C. Cut-off weight: weight: Mean+1.8*SD3,000 Mean+1.8*SD D. Cut-off D. Cut-off weight: weight: Mean+2*SD Mean+2*SD 3,000 C. C. C. Cut-off Cut-off Cut-o weight: weight: weight:Mean+1.8*SD Mean+1.8*SD Mean+1.8*SD D. Cut-off 3,000D. Cut-off weight: weight: Mean+2*SD D. Cut-o weight: Mean+2*SD Mean+2*SD 3,000 2,000 3,000 2,000 3,000 3,000 3,000 3,000 3,000 3,000 2,000 2,000 2,000 2,000 1,000 1,000 2,000 2,000 2,000 2,000 2,000 2,000 1,000 0 1,000 1,000 0 1,000 1,000 1,000 1,000 1,000 1,000 1,000 1 20 0 40 60 80 100 112 1 20 0 40 60 80 100 112 0 0 0 01 Enumeration area 0 01Enumeration area 10 20 40 20 60 40 80 60 80 100 112 100 112 10 20 40 20 60 40 80 60 80 100 112 100 112 1 11 2020 Enumeration Enumeration 20 40 4040 60 area 60 60 80 area80 80 100 100 112 100112 112 1 11 20 20 Enumeration 20 Enumeration 40 40 area 6040 60 60 80 area80 80 100 100 112 112 100 112 area area Enumeration Enumeration Cut-off value Enumeration areaDesign weight Enumeration Enumeration Trimmed area area Enumeration weight area Cut-off value Cut-off value Design weight Design weight Trimmed weight Trimmed weight Cut-o value Cut-off Cut-off value value DesignDesign Design weight weight weight Trimmed Trimmed weight weight Trimmed weight 40     40   40     FG" 40   40         " 43 PARTICIPATION, TRANSPARENCY AND ACCOUNTABILITY IN VILLAGE LAW IMPLEMENTATION The variation of design weight is relatively high and this can cause the higher standard error and wider interval estimates. Trimming weight will reduce the standard error but may cause the bias depends The upon variation ofthe number design of weights weight are truncated. is relatively high andThe cut-off this canvalue cause is based on the the higher mean and standard error and the standard deviation of sampling weights. The graph above depicts the plot of wider interval estimates. Trimming weight will reduce the standard error but may cause the bias sampling depends upon the design weights and trimmedof number weights weights for are eachtruncated. The cut-off ea with 4 different value value is based of cut-off. Panelon the mean and A depicts the the standard deviation weight plot of sampling with the cut-off weights. while value of Mean+SD, The graph Panel B above the plot depicts Panel is Mean+1.5xSD, C is of sampling design weightsand Mean+1.8xSD, and trimmed Panel weights for The D is Mean+2xSD. each ea with Panel 4 different A shows value that some of cut-off. trimmed weightsPanel are A depicts the weight plot with the cut-off value of Mean+SD, while Panel B is Mean+1.5xSD, still higher than the cut-off value and the proportion of weight that are trimmed is higher Panel C is Mean+1.8xSD, and Panel D is Mean+2xSD. The Panel A shows that some compared to the other cut-off value, as well as Panels B and C. Based on these plots, the best trimmed weights are still higher than the cut-off value and the proportion of weight that are trimmed is higher option is the cut-off value of Mean+2xSD (Panel D) because all trimmed weights are below the compared to the other cut-off value, as well as Panels B and C. Based on these plots, the best cut-off value and the smallest proportion of weight that are trimmed. option is the cut-off value of Mean+2xSD (Panel D) because all trimmed weights are below the cut-off value and the smallest proportion of weight that are trimmed. The same evaluation is done when we calculate the person weights. For male weight, we pick thesame The cut-offevaluation is done when value of Mean+2.7xSD we and for calculate female the weight is person weights. Mean+2.0xSD Forthe to give male bestweight, we pick the cut-off value of Mean+2.7xSD and for female weight truncated sex-specific weights for person data. Annex Figure 7.2. is Mean+2.0xSD to give the best truncated sex-specific weights for person data. See Error! Reference source not found.. Annex Annex Figure Figure 7.2. 7.2. Design and Trimmed Weight - Person data Design and Trimmed Weight - Person data A. Cut-o weight: Mean+2.7*SD (Male) B. Cut-o weight: Mean+2.0*SD (Female) 6,000 6,000 Weight Weight 4,000 4,000 2,000 2,000 0 0 1 500 1,000 1,500 2,000 1 500 1,000 1,500 2,000 i-th Household i-th Household C. Design and trimmed weight D. Male to female weight ratio 10,000 5 Male samples 4 Weight ratio 1,000 3 2 100 1 10 0 10 100 1,000 10,000 1 500 1,000 1,500 2,000 Female samples i-th Household Design W2 Trimmed W2 F! " 44" PARTICIPATION, TRANSPARENCY AND ACCOUNTABILITY IN VILLAGE LAW IMPLEMENTATION 2. Sampling 2. Sampling weight for household data 2.weight weight fordata for household Sampling household data 2. Sampling weight for household data Sampling Sampling plan table plan for table for household household selection and sampling weight for household data are Sampling Sampling planplan table table for selection for household and sampling selection selection household and and sampling weight sampling weight for household weight data for household data for are are household data are described described below. below. described described below. below. th th Annex Annex Table Table 7.2. 7.2. Sampling Sampling schemescheme table table for for household household selection selection in eachin each selected selected district-i th th district-i Annex Annex Table Table 7.2. 7.2. Sampling Sampling scheme scheme table table for for household household selection selection in in each each selected selected district-i district-i Stage Stage Sampling Sampling Stage Sampling StratumStratum Universe Stratum Universe Sample Universe Sample Sampling Sample Sampling WeightWeight Sampling Weight unit Sampling unit unit Stage Stratum Universe Sample Sampling Weight unit 1 1Sub-district Sub-district - - ! ! Random Random ! ! 11 Sub-district Sub-district -- ! ! ! ! Random Random ! (A, j) (A, j) (A, j) (A, j) ! ! 2 2Village Village - - !" !" = 4 !" = Random 4 Random !" !"!" !" ! 2 2 Village Village -- !" !" !" = 4 Random = Random = !" !" (B, k) (B, k) (B, k) (B, k) !" 4 !" 4= !" 4 3 3HamletHamlet - - !"# = 1 = Random 1 Random !"# !"# !"# !"# 33 Hamlet Hamlet -- !"# !"# !"# !"# !"# = 1 Random Random = = !"# !"# (C, l) (C, l)(C,(C, l) l) !"# 1!"# 1 = 1 !"# 4 4HouseholdHousehold - - !"#$ !"#$ !"#$ = 20 =Random20 Random !"#$ !"#$ !"#$ !"#$ 44 Household Household -- !"#$ !"#$ !"#$ = 20 Random Random = =!"#$ !"#$ (D, m) (D, m) (D, (D, m) m) !"#$ 20 !"#$ 20 = !"#$ 20 forfor forWeight WeightWeight householdrespondents: household household Weight forrespondents: household respondents: respondents: ! ! !! !"! ! !" !"#$ ! !" !"#$ ! !"# !" !"# !"#$ !"#$ ! = !"#$ !"#$ = !"#$ = !"#$ × = !× × × !"# ! × !" ×= =!"#$ ! !" !"# !"#$ !"# ! !! !"#$ = !!! ! 4 ! = 4 ×20 ×!20 !"#$ "# × 80! = 80! ! !! 4 ! 20 80 ! where, where, where, where, is!"#$ household household ishousehold weight in selected hamlet- l, selected village- selected k,sub-district-j, sub-district- j, !"#$ is !"#$ is weight weight household in selected in selected weight in hamlet- l, selected hamlet-l, selected selected hamlet- village- village-k, k l, selected, selected selected village- sub-district- k , , selected jsub-district- j, and selected and selected district- and selected district- i . district-i. i . and selected district-i. Annex Annex Error! Error! Reference Reference source not found. depicts theof plot of sampling design weights and Annex Error! source Reference found. not source depicts not found.the depicts plot sampling the plot of design weights sampling andweights design and trimmedtrimmed weights weights of of selected selected household household in each inselected each selected ea with ea4 with 4 different different value value of of cut-off. cut-off. trimmed weights of selected household in each selected ea with 4 different value of cut-off. Panel Panel A depicts A depicts Panel the the weight weight A depicts the plotthe plot weight with with the cut-off plotcut-off with the value value of cut-off of Mean+SD, Mean+SD, while Panel value of Mean+SD, while Panel B is B is Panel while B is Mean+1.5xSD, Mean+1.5xSD, Panel C Panel C is Mean+1.8xSD,is Mean+1.8xSD, and and Panel D Panel D is Mean+2xSD. is Mean+2xSD. The The Panel A, Panel and CA, B, C A, B, Panel B, and Mean+1.5xSD, Panel C is Mean+1.8xSD, and Panel D is Mean+2xSD. The and C show show that somethat some trimmed weightsstillare still higher than the cut-off andvalue and the proportion of show trimmed that some weights trimmed are weights higher are than the still higher the cut-off the than value cut-off proportion value and the of proportion of weight weight that are that are trimmed trimmed is higheris compared higher compared to the to the other other value. cut-off cut-off Based value. Based on these weight that are trimmed is higher compared to the other cut-off value. Based on these plots, the onplots, plots, the thesethe option best is best option is the value theoption cut-off cut-offofvalue of Mean+2xSD Mean+2xSD (Panel (Panel D). (PanelD). best is the cut-off value of Mean+2xSD D). 42   42   42         45 Annex Figure 7.3. Design and Trimmed Weight - Household data PARTICIPATION, TRANSPARENCY AND A. Cut-off weight: ACCOUNTABILITY IN VILLAGE Mean+1*SD LAW B. Cut-off IMPLEMENTATION weight: Mean+1.5*SD 3,000 3,000 2,000 2,000 1,000 1,000 Annex Figure 7.3. depicts the plot of sampling design weights and trimmed weights 0 household in each selected ea with 4 different of selected 0 value of cut-off. 1 20 40 60 80 100 112 1 20 40 60 Panel A depicts the weight plot with the cut-off value of Mean+SD, Enumeration area while Panel80 Enumeration area B is 100 112 Mean+1.5xSD, Panel C is Mean+1.8xSD, and Panel D is Mean+2xSD. The Panel A, B, and C show that some C.trimmed weights Cut-off weight: are still higher than the cut-off Mean+1.8*SD value D. Cut-off and the weight: proportion of Mean+2*SD weight that are trimmed is higher compared to the3,000 3,000 other cut-off value. Based on these plots, the best option is the cut-off value of Mean+2xSD (Panel D). 2,000 2,000 1,000 1,000 Annex AnnexFigure 7.3. Figure 7.3. 0 0 20 Weight Design and Trimmed 1 40 and Design - Household 60 80 Trimmed100data 112 Weight 1 20 data - Household 40 60 80 100 112 Enumeration area Enumeration area A. Cut-o weight: Mean+1*SD B. Cut-o weight: Mean+1.5*SD 3,000 Cut-off value 3,000 weight Design Trimmed weight 2,000 2,000   1,000 1,000 3. Sampling weight for hamlet leader data 0 0   1 20 40 60 80 100 112 1 20 40 60 80 100 112 Enumeration area Enumeration area Sampling plan table for hamlet leader selection and sampling weight for hamlet leader data are described below. C. Cut-o weight: Mean+1.8*SD D. Cut-o weight: Mean+2*SD 3,000 3,000 th 2,000 Table 7.3. Sampling scheme table 2,000 Annex for hamlet leader selection in each selected district-i Stage 1,000 Sampling Stratum Universe 1,000 Sample Sampling Weight unit 0 0 1 Sub-district - 100 ! Random ! 1 20 40 60 80 112 1 20 ! 40 60 80 100 112 (A, j) Enumeration area Enumeration area ! 2 Village - !" !" = 4 Random !" !" (B, k)Cut-o value = Design weight Trimmed weight !" 4 3 Hamlet - !"# !"# = 1 Random !"# !"# " (C, l) = !"# 1 3. Sampling weight for hamlet leader data " 3. Sampling weight for hamlet leader data Samplingplan Sampling plan table forhamlet table for hamlet leader leader selection selection andand sampling sampling weight weight for hamlet for hamlet leader leader data are data are described below. described below. Annex Table 7.3. Sampling scheme table for hamlet leader selection in each selected district-ith th Annex Table 7.3. Sampling scheme table for hamlet leader selection in each selected district-i Stage Stage Sampling unit Sampling Stratum Stratum Universe Universe Sample Sample Sampling Sampling Weight Weight 43     1 unit Sub-district - Random 1 Sub-district (A, j) - !! !! Random !! 2 (A, j) Village - Random !! 2 Village (B, k) - ! !" ! !" ! ! Random ! !" ! !" (B, k) ! 3 Hamlet - Random ! !" ! 3 Hamlet (C, l) - ! !"# ! !"# ! ! Random ! !"# ! !"# (C, l) ! ! !"# ! 46 PARTICIPATION, TRANSPARENCY AND ACCOUNTABILITY IN VILLAGE LAW IMPLEMENTATION Weight for hamlet leader respondents: ht for hamlet leader Weight for hamlet leader respondents: respondents: Weight for hamlet leader ! ! ! respondents: ! !" ! !! !" !"!"# ! !" !"# !"# = !"# !"# == × ! × = ×! ×!"# = ! ! ! !"# = ! !! 4!! ! ! !"# !!! 44 !" ! 4 !"!!"# ! !"# ! ! ! !"# ! ! ! ! ! !"# ! !! ! ! ! ! ! !! e, where, where, !"# is hamlet where, leader !"# hamlet is weight is hamlet leader leader in selected weight weightvillage- in selected , selected in selected k village-k, village- selected sub-district- , selected j, and ksub-district-j, selected sub-district- and selected j, and selected district-i. is ! !"# district-i. hamlet district- i. leader weight in selected village-k, selected sub-district-j, and selected district-i. Annex Figure 7.4. depicts the plot of sampling design weights and x Error! Reference Annex trimmed Error!source weights of not Reference found. selected source hamlet not found. depicts leader the of sampling plotdepicts in each selected the plot ea withdesign 4of sampling weights different valuedesign and weights and of cut-off. med weights of selected trimmed Annex Panel weights Error! A depicts hamlet selected of weight Reference the leader in source plot hamlet each with not theselected leader found. cut-off ineaeach depicts value with 4 different selected the of Mean+0.8xSD,plot ofwith ea value sampling while 4Panel of B is value different cut-off. design of cut-off. weights and l A depicts the Panel trimmed A depicts weight weights plot of with the weight the selected cut-off plot hamlet with value leader the of in Mean+0.8xSD, cut-off each value selected Mean+SD, Panel C is Mean+1.25xSD, and Panel D is Mean+1.5xSD. The Panels A, B, C show of eawhile Mean+0.8xSD, with 4Panel B different while is valuePanel of B is cut-off. n+SD, Panel C is A Mean+SD, Panel Mean+1.25xSD, C is weight Panel the depicts and Panel Mean+1.25xSD, plot D is Mean+1.5xSD. and Panel with the cut-off Dvalue is The Mean+1.5xSD. Panels of A, B, The Mean+0.8xSD, Panels C show whileA, B, C show Panel B is that some trimmed weights are still higher than the cut-off value and the proportion of weight some trimmed that weights Mean+SD,some Panel are still trimmed C is higher weights are still than Mean+1.25xSD, cut-off the higher and Panel value than Dand the iscut-off value and the proportion Mean+1.5xSD. of the The proportion weight Panels A, B, ofCweight show are trimmed that that are trimmed is some higher are trimmed is compared ishigher to compared higher the compared other to the to other cut-off the value. cut-off other value. cut-off Based Based onvalue. these on plots, these on Based plots, the the best best these plots, that trimmed weights are still higher than the cut-off value and the proportion ofthe best weight option option n is the cut-off value isis the the of cut-off cut-off value value Mean+1.5xSD of Mean+1.5xSD of(Panel D). Mean+1.5xSD (Panel D). (Panel D). that are trimmed is higher compared to the other cut-off value. Based on these plots, the best option is the cut-off value of Mean+1.5xSD (Panel D). Annex Figure 7.4. x Figure 7.4. Annex Figure 7.4. Annex Figure 7.4. Designand Design Design and Trimmed Trimmed Weight - Hamlet and -Trimmed Weight Hamlet leader Weight leader -data data Hamlet leader data Design and Trimmed Weight - Hamlet leader data A. Cut-off weight: Mean+.8*SD B. Cut-off weight: Mean+1*SD A. Cut-off weight: Mean+.8*SD B. Cut-off weight: Mean+1*SD 00 200 200 A. Cut-o weight: Mean+.8*SD 200 B. Cut-o weight: Mean+1*SD 50 150 200 150 150 200 00 100 150 100 100 150 50 50 100 50 50 100 0 0 50 0 0 50 1 20 40 0 601 80 20 40 11260 100 80 1 100 20 112 40 0 601 80 20 40 11260 100 80 100 112 Enumeration area Enumeration area Enumeration area Enumeration area 1 20 40 60 80 100 112 1 20 40 60 80 100 112 Enumeration area Enumeration area C. Cut-off weight: Mean+1.25*SD C. Cut-off weight: Mean+1.25*SD D. Cut-off weight: Mean+1.5*SD D. Cut-off weight: Mean+1.5*SD 00 200 200 C. Cut-o weight: Mean+1.25*SD 200 D. Cut-o weight: Mean+1.5*SD 50 150 200 150 150 200 00 100 150 100 100 150 50 50 100 50 50 100 0 0 50 0 0 50 1 20 40 0 601 80 20 40 11260 100 80 1 100 20 112 40 0 601 80 20 40 11260 100 80 100 112 Enumeration area Enumeration area Enumeration area Enumeration area 1 20 40 60 80 100 112 1 20 40 60 80 100 112 Enumeration area Enumeration area Cut-off value Design weight Cut-off value Trimmed weight Design weight Trimmed weight Cut-o value Design weight Trimmed weight   "   "   "   "   " 44   44     FF " " 47 PARTICIPATION, TRANSPARENCY AND ACCOUNTABILITY IN VILLAGE LAW IMPLEMENTATION 4. Sampling for activist weight weight 4. Sampling data data for activist 4. Sampling 4. 4. weight Sampling for Sampling activist weight for data weight activist data data for activist Sampling plan table Sampling plan for male table forand female male activist activist female and activist selection and sampling selection and weight weight sampling for male and Sampling plan table for male and female selection and sampling weight for male and for male and mplingfemale plan activist table female Sampling for data maleare activist plan table anddescribed data female are fordescribed below. below. activist described male and selection and sampling weight for female activist selection and sampling weightmale andfor male and female activist data are below. male activist data are described below. female activist data are described below. th Annex Table Annex Annex 7.4. TableSampling Table 7.4. 7.4. scheme Sampling Sampling table table scheme scheme for activist table for for selection activist activist in each selection selection selected inin each each district-i selected selected district-i district-i th th th nex Table 7.4. Sampling Stage Sampling scheme table for activist Stratum Universeselection in each selected Sample district-i Sampling selected Weight th Annex Table Sampling Stage 7.4. Sampling scheme Stratum table for activist Universe selection Sample in each Sampling district-i Weight Stage Sampling unit Stratum Universe Sample Sampling Weight age Sampling unit Stratum unit Universe Sample Sampling Sampling Weight Stage Sampling Stratum Universe Sample Weight ! 1 unit 1Sub-district - ! ! ! Random ! 1 unit Sub-district Sub-district - - ! Random Random (A, j) (A, Sub-district (A, -j)j) ! Random ! ! 1 Sub-district -! ! ! Random ! ! 2 j) 2Village (A, Village 2 (A, j) Village - - - !" !" !" = 4 !" = 4 Random Random Random ! !" !"!" (B, k) (B, = ! = !" Village (B, - Random Village k) = 4 !" !" 4 k) !" !" !" 4 !" 2 -!" !" !" !" = 4 Random = 3 k) 3Activist (B, Activist 3 (B, k) !" 4 = Activist Gender: !" 4 Activist (F, l) (F,(F, l)l) Gender: Gender: 3 Activist (F, l) Male (m) (! ) (! ) (! ) Stratified (! ) )(!) (F, Gender: Male (m) (m) Stratified !"# 1(!) = (! (! ) l) Male Gender: !"# !"# = !"# 1 Stratified PPS, !"# !"# !"# !"# Male (m) Male ( ! !"#(m) ) ( !"# ! ) Stratified PPS,size=# ( ! PPS, ) voters ( ! ) = (! ) ( = !"# = 1 Stratified (! ) (! ) (! ) (! (! )() ) ! ) (! ) !"# PPS, = 1 size=# !"# = !"# !) (! (!size=# !"# !"# !"#$ !"#$ !"# PPS, ) (! ) (!) !"# !"#$ = !"#$ Female (f) Female (f) (f) (! ) !"# ( ! ) (! ) size=# = 1 ( ! ) voters !"# voters!"#$ (! !"#$ ! )) (!) ( ! )(! ) (! ) (! ) Female !"# !"# !"# = 1 size=# !"# !"#$ !"# !"#$ Female (f) Female (! ) !"# (f) (! ) !"# = 1 voters voters (! ) (! ) (! )=!"# (!) ( = !"# (! ) !"# (! ) !"# = 1 !"# ( ! ) !"# ( ! ) ( ! ) (! ) ! ) (! ) =!"#(!!"#$ !"# !"#$!"# !"#$ !"#$ (! ) (! ) ) !"# = !"# !"#$ !"#$ (! ) (! ) !"# (! ) !"#$ !"#$ Weight for male Weight Weight forfor activist: malemale activist: activist: (! ) (! ) eight for male activist: Weight for male activist: ! ! ! !"! !"# (!! )!" !"#! !" (! ) (! ) ( ! ) ! ;! ! ; ! !" !"# !"# !"#$ ! = !"#$ = !"#$!"!"#$ = (×!) × = ! × (!" ! (× )!"# (=!) ! ) (! ) = (! ) (! ) (! ) (! ) !;! !!!! ! !!! !"# ! ! 4 4 4 ! !"#$ = ( ) !!"#$ = × ! ;!× = !!" !"#$!"# !"#$ ! !"#$ !" 4 ! !"# !"#$ = 4 !"#$ = (! ) × 4 × (! ) = where, where, ! !! !"#$ ! !! ! !"#$ ! 4 ! !"#$ ( ! ) 4 ( !) !"#$ ! !"#$ (! ) ere, !"#$ where, is(! weight ) is for male weight for activist- male l in selected activist- l in village-village- selected k, selected k , selectedsub-district- j, and selected sub-district- j, and selected (! ) where, !"#$ !"#$ is weight for (!)district- is male i district- weight, activist- i, male l in selected village- k , selected sub-district- j , and selected (!) !"#$ (!) is weight forfor male activist-l activist- l in selected in selected village-k, village- selected k, selected sub-district-j, sub-district- and selectedj, and selected !"# i, district- number is district-i, is number of male activists in selected village- of male activists in selected village-k, selected sub-district- k , selected sub-district- j , and selected j, and selected (! ) !"# district- i, !"# is number ( ! !"# is ) of district- is male i , activists district- number number of of i male, male in selected activists activists in selected village- in selected k , village-k,selected village-selected sub-district- k, selected sub-district-j, j , and sub-district- selected and selectedj, and selected (! ) !"# i, district- is(! number ) is district-i, of number selected of male selected activists male in activists selected in village- selected k , village- selected k , sub-district- selected j, and j, and sub-district- (! ) !"# district-i, !"# is number (!) of !"# selected isis selected number number of of district- male selected activists ( ! , !"# male iselected ) = 1 (activists male !) in selected in selected activists village- in selected k village-k,, selected selected village- sub-district- sub-district-j, k, selected j , and and sub-district- j, and selected district- i, !"# = 1 selected ( !) district-i, selected ( ! ) district- is number ( ! ) selectedi , of = voters district- 1 i, for (! ) selected = for 1 male male activist- l in selected village-village-k, selected sub- sub- !"#$ !"#$is number is number !"# of voters of for voters !"#selected selected activist-l male in selectedactivist- l in selected village-k, selected subdistrict-j, k, selected (! ) !"#$ is number (!)district- of voters j, and selected district- is number for selected j, and of voters district- selected male for selected i, activist- district- , imale l in selected village- activist-l in selected village-k, selected sub- k , selected sub- (! ) !"#$ and selected district-i, district- !!" j , and is ( ! is ) selected total number is district- total total number district- j, and of of number voters selected i voters of , of of voters male district- male activists ofactivists i, male in activists in selected selected invillage-k,village- selected selectedk, selected village- sub-district- k, selected sub-district-j, j, sub-district- j, (! ) !!" is total (!number ) and selected and of isselected voters of district- male , ivoters activists ( !) ( !in ) selected ( !) .∀! in ( ! village- ) k , selected sub-district- j , !"# = ∀! total and number district-i, selected ofdistrict- iof , male =activists selected village-k, selected sub-district-j, !"#$ . !!" !!" !"#$ (! ) (!) !"# and selected district- and selected i , = !"#district-i, !"#$ ∀ ! (!. ) = (! ) . !"# ∀! !"#$ 45   45       45   45     48 PARTICIPATION, TRANSPARENCY AND ACCOUNTABILITY IN VILLAGE LAW IMPLEMENTATION Weight for female activist: Weight for female activist: (! ) (! ) ight for female Weight activist: for female activist: ! ! !" ! !" Weight for female activist: (! ) !"#$ = ! ;! = × (! ) × !"# =(! ) ! ! ! !"# !! ! !"#$ ! ! (! ) ! (! ) (! ! ! !" ! ( !! ! !) !! !!! !! !" !"# 4 ! !) !; !! ! !!! !" !"# !" !"#4 !"# ! ! (! ) !"#$ = ! (!"#$ !) ! = × ! !"#$ × ! ;! ! = !" ! !"# !! !"#$ !"#$ ! !" !"# where, ! !! !"#$ = ! ! ! ! 4!"#$ = ! 4! ! (! ) = ! ! ! ! (! ) (! ) ! × ! × (! ) ! ! ! !! ! ! !! !"#$ ! 4 !"#$ !"#$ 4 ! !"#$ !"#$ (! ) !"#$ ere, where, !"#$ is weight for female activist-l in selected village-k, selected sub-district-j, and (! ) where, where, !! ! !"#$ is weight ! (selected ! ) for is district- weightactivist- female i,for female l in selected activist-l village- in selected k, selected village- k, selected sub-district- sub-district-j, and j, and !"#$ is is weight for female activist-lactivist- l invillage-k, selected village- k, selected and sub-district-j, and !"#$ weight for female in selected selected sub-district-j, (! ) selected !"# is district- selected number district- i , of female activists in selected village-k, selected sub-district-j, and i, district-i, (! ) selected ! ! ! selected district-i, !"# is number ! (selected ! ) is is ofnumber district- number female i, of female activists of female activists in selected activists in selected in village- selected village-k, k, selected selected village- k, selected sub-district- sub-district-j, and sub-district-j, and j, and (! ) !"# !"# is number of female activists in selected village- k, selected sub-district-j, and selected !"# is number district- selected selected i, of district-i, female activists in selected village-k, selected sub-district-j, selected district-i, (! ) !! ! selected district-i, !"# is number !( ! ) of and is is number selected selected number of of selected female district- selected (! ) activists i, !"# female 1 = female female in selected activists activists in selected in selected village- k, selected village- k, selected sub-district- j, sub-district-j, !"# !"# is number of selected activists in village-k, selectedselected village-sub-district-j, k, selected sub-district-j, and selected ( ! ) district-i, ! ! ! and (! )selected is number and district- selected ofi,voters !"# = district- 1 selected for i, !( !) ! ! female activist-l in selected village-k, selected sub- !"#$ and selected district- i, !"# !"# = 1 (! ) ! ! ! is number of voters for selected female activist-l in selected village-k, selected subdistrict-j, is number !( district- is ! ) of voters j, number andfor selected of voters selected district- for iselected female , activist- female l in selected activist- l in selected village- k, selected village- k, selected sub- sub- !"#$ (! ) !"#$ !"#$ and number district-i, is selected of voters for selected female activist-l in selected village-k, selected sub- district- j,is andtotal district- number selected and j,district- of voters selected i, of female district- i, activists in selected village-k, selected sub-district- ! ! ! is district- j, and voters of district- selected i, !"# total number of female activists in selected village-k, selected sub-districtj, (! ) is total !number , )and isof total selected number voters of female district- of voters ( ! ) activists of ∀ female ( ! ) in selected activists in selected village- k, selected village- k, selected sub-district- sub-district- j! ofi,voters !"# = !"#$ . activists in selected village-k, selected sub-district- ! ( and selected district-i, !"# !"# !"# is total number of female (! ) (! ) ! ! ! !! ! j, and selectedj,district- and selectedi, =district-∀! i, !.( !) ! ! ! !( !) . j, and selected !"# district- i, !"#$ !"# !"# = ∀! !"#$ . !"#$ Reference Error! Figure Annex Annex source 7.5. depicts not of the plot found. sampling depicts design the plot of weights and sampling design weights and trimmed nex Error! Annex weights Reference trimmed Error! of weights selected Reference source activist not found. of selected in source activisteach not depicts in eachselected found. the plot selected ea depicts ea with of with the the theplot sampling best best option of design sampling option ofof weightscut-off design cut-offand value value for for and weights male Annex and Error! female Reference activists. The source cut-off not value found. for male depicts activist the plot weight of is sampling Mean+1.7xSD, design weights while for and med weights trimmed of selected male and weights female of selected activist in The activists. each activist selected cut-off in each value ea selected for with male the best activist ea with is option weight the of best option cut-off Mean+1.7xSD, value cut-off offor while for valuevalue for femaletrimmed activist weights weight of selected iscut-off Mean+1.6xSD. activist in Panel each selected C shows ea the plot with the of design bestandoption of trimmed cut-off of for weightwhile e and male female activists. female and female activist The activists. weight The value cut-off for is Mean+1.6xSD. male value Panel activistfor C shows male weight activist is the plot weight Mean+1.7xSD, of design and is Mean+1.7xSD, while trimmed for weight of while forfor male male ale activistand female weight and female is female activist activists. activist weight Mean+1.6xSD. samples Theand is Mean+1.6xSD. Panel cut-off C Panel shows value Dtheshows Panel for male the C shows plot of activist male the and design toweight female plot of design trimmed isactivist Mean+1.7xSD, and weight weight of ratio. trimmed weight of The female male ratios activist and are female very weight similar is activist Mean+1.6xSD. samples comparing and Panel design Panel D showsC shows the malethe toplot of female design activist and weighttrimmed ratio. weight of e and male female and activist female samples activist samples and Panel and and D shows Paneltrimmed the D shows male weight. to the male female to female activist ratio. weight ratio. weightactivist maleThe and ratios female are very activist samples and similar comparing designPanel D shows and trimmed the male to female activist weight ratio. weight. very ratios e ratios are The similar are very similar comparing comparing design and trimmeddesign and trimmed weight. weight. The ratios Annex Figure 7.5. are very similar comparing design and trimmed weight. Annex Figure 7.5. nex Figure 7.5. Annex Annex Figure Figure 7.5. 7.5. Design and Trimmed Weight - Activist data Design and Trimmed Weight - Activist data A.Design and Trimmed Cut-off weight: Design Weight Mean+1.7*SD and - Activist Trimmed (Male) Weight B. data - Activist Cut-off weight: data (Female) Mean+1.6*SD Design and Trimmed Weight - Activist data 150 50 A. Cut-off weight: Mean+1.7*SD (Male) A. Cut-o weight: Cut-off weight: Mean+1.6*SD B.(Male) Mean+1.7*SD (Female) B. Cut-o weight: Mean+1.6*SD (Female) 120 40 150 150 A. Cut-off weight: Mean+1.7*SD (Male) 50B. Cut-off weight: Mean+1.6*SD (Female) 50 Weight Weight 150 90 30 50 120 120 40 40 60 120 20 40 Weight Weight Weight Weight 90 90 30 30 Weight Weight 30 90 10 30 60 60 20 20 0 60 0 20 30 30 10 10 1 3020 40 60 80 100 112 1 10 20 40 60 80 100 112 0 0 Enumeration area 0 0 Enumeration area 0 0 1 20 40 60 1 80 20 100 40 112 60 80 1 100 20 112 40 601 80 20 100 40 11260 80 100 112 1 area Enumeration 20 40 60 area Enumeration 80 100 112 Enumeration1 20 area 40 60 area Enumeration 80 100 112 Enumeration C. Design and trimmed weightarea Enumeration area D. Male to female activist weight ratio Female activist samples 50 8 C. Design and trimmed D. Male to female activist weight and trimmed weight C. Design D. Maleweight ratio activist weight ratio to female 40 C. Design and trimmed weight 8 D. Male to female activist weight ratio Female activist samples Female activist samples 6 Weight ratio 50 50 8 Female activist samples 30 50 8 40 40 4 Weight ratio 6 Weight ratio 6 20 40 6 Weight ratio 30 30 2 10 30 4 4 20 20 4 0 20 2 0 2 10 10 2 20 0 10 30 60 90 120 150 140 60 80 100 112 0 0 Male activist samples 0 0 Enumeration area 0 0 0 30 60 0 90 30120 60150 90 1 120 20 150 40 60 1 80 20 100 112 40 60 80 100 112 0 samples Male activist 30 Male 60activist 90 samples 120 1 150 Enumeration 20 area 40 60 area Enumeration 80 100 112 Male activist Design samples weight Trimmed weight Enumeration area Design weight Trimmed weight Design weight Trimmed weight Design weight Trimmed weight 46     46   FV" 46   "   49 PARTICIPATION, TRANSPARENCY AND ACCOUNTABILITY IN VILLAGE LAW IMPLEMENTATION mpling weight for 5. Sampling weight village and BPD leader data for data 5. village BPD leader and weight Sampling for village and BPD leader data 5.Sampling plan table for Sampling plan weight and village plan BPDtable village forleaderfor village and BPDand selection BPD leader and leader selection data sampling and for weight and sampling village and weight for village and Sampling BPD leader table datafor arefor village described and BPD below. leader selection sampling weight for village and Sampling er data are described plan table below. village and BPD leader selection and sampling weight for village and BPD leader data are described below. BPD leader data are described below. Annex Annex ble 7.5. Sampling Table scheme Table 7.5. 7.5. Sampling Sampling table scheme scheme for village table and table BPD for and and for village village leader selection in BPD eachleader BPD leader selection selection selected in selected in each each selected Annex th th 7.5. Sampling scheme table for village and BPD leader selection in each selected Table district-i district-ith district-i ampling Stage Stratum Stage Sampling Universe Sampling unit Stratum Sample Stratum Universe Sampling Universe Sample Sample Weight Sampling Sampling Weight Weight Stage Sampling unit Stratum Universe Sample Sampling Weight nit unit ! Sub-district ub-district 1 1 - Sub-district 1 Sub-district -- ! ! Random Random (A, j) (A, j) ! - ! Random ! ! Random ! ! , j) ! ! j) (A, Village !" llage 2 2- Village -- = 4 !" Random 4 !" = Random Random !" ! 2 Village (B, k) (B, k) !" - !" !" !" = 4 !" = Random !" !" = !" , k) !" 4 = !" 4 (B, k) !" 4 Weight for village and BPD leader respondents: Weight BPD village andWeight leader village and BPD leader respondents: for respondents: for village and BPD leader respondents: ! ! !" ! !" ! !" = ! !! !" ! ! !" = × = ! !" ! !" = !" = × = = !" = × ! ! !" 4= 4! ! !! ! !" 4 4! !! ! !" 4 4 ! !! ! ! where, where, where, is weight for is weight for village!" !" isor isBPD leader weight for ofvillage or BPD selected village or BPD leader village- k in leader of of selected selected selected village-kj,in sub-district- village- k in selected sub-district-j, and and selected sub-district- j, and weight for village or selected district-i,BPD leader of selected village-k in selected sub-district-j, and selected district-i, selected district-i, selected district-i, erence sourceError! not Reference found. shows source the not found. weight shows for shows village andthe weight BPD for village leader data. and BPD leader data. There There Error! is noReference extremely Annex Figure source high 7.6. shows not weight found. the weightthat may for village the increase and BPD weight the for village standard leader data. error There and BPD of leader data. estimates, There therefore the weight emely high is no that may extremely increase high weight the standard that may error increase estimates, ofthe standard therefore error of the estimates, therefore the design isto ght does not nobe weight extremely trimmed. does not to be trimmed. high weight that may increase the standard error of estimates, therefore the design weight does not to be trimmed. design weight does not to be trimmed. AnnexFigure Annex Figure7.6. 7.6. Design weight - Village and BPD leader data Design weight - Village and BPD leader data 15 10 Weight 5 0 1 20 40 60 80 100 112 Enumeration area 47   47     47     50 PARTICIPATION, TRANSPARENCY AND ACCOUNTABILITY IN VILLAGE LAW IMPLEMENTATION 52