66035 EAST ASIA AND PACIFIC REGION Social Development Reports Measuring the Impact of Community-Driven Development Projects on Gender A Toolkit for the Poverty Reduction Fund, Lao PDR GAP funded gender equality as smart economics A World Bank Group Gender Action Plan Measuring the Impact of Community-Driven Development Projects on Gender A Toolkit for the Poverty Reduction Fund, Lao PDR © 2011 The International Bank for Reconstruction and Development / THE WORLD BANK 1818 H Street, NW Washington, DC 20433, U.S.A. Telephone: 202-473-1000 Internet: www.worldbank.org/eapenvironment/sea-asia E-mail: feedback@worldbank.org All rights reserved. October 2011 This volume is a product of the staff of the International Bank for Reconstruction and Development / The World Bank. The findings, interpretations, and conclusions expressed in this paper do not neces- sarily reflect the views of the Executive Directors of The World Bank or the governments they represent. The World Bank does not guarantee the accuracy of the data included in this work. The boundaries, colors, denominations, and other information shown on maps 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 accep- tance of denoted boundaries. R I GHTS A N D PE R MI SSI ON S The material in this publication is copyrighted. Copying and/or transmitting portions or all of this work without permission may be a violation of applicable law. The International Bank for Reconstruction and Development / The World Bank encourages dissemina- tion of its work and will normally promptly grant permission to reproduce portions of the work. For permission to photocopy or reprint any part of this work, please send a request with complete infor- mation to the Copyright Clearance Center, Inc., 222 Rosewood Drive, Danvers, MA 01923, USA, tele- phone 978-750-8400, fax 978-750-4470, www.copyright.com. All other queries on rights and licenses, including subsidiary rights, should be addressed to the Office of the Publisher, The World Bank, 1818 H Street NW, Washington, DC 20433, USA, fax 202-522-2422, e-mail pubrights@worldbank.org. Photo credits: Cover photo by the World Bank Task Team of this publication. iii Table of Contents Acronyms and Abbreviations . . . . . . v Acknowledgments . . . . . . vii Executive Summary . . . . . . ix InTroDuCTIon . . . . . . 1 Gender-Related Indicators . . . . . . 2 The PoverTy reDuCTIon FunD . . . . . . 5 PRF's Funding Cycle . . . . . . 5 PRF’s Approach to Gender . . . . . . 6 PRF’s Monitoring and Evaluation Framework . . . . . . 7 Considering Gender-Based Indicators for the PRF . . . . . . 8 WoMen’s oPPorTunITIes To ParTICIPaTe In DeCIsIon MakInG . . . . . . 11 Gender-Based Issues in the Political Sphere . . . . . . 11 PRF in the Political Sphere . . . . . . 12 Active Participation . . . . . . 12 Awareness and Capacity Building . . . . . . 14 Impact of Women’s Participation on Decision Making within PRF . . . . . . 15 Spillover Effects . . . . . . 17 Perceptions of Women’s Roles in Decision Making . . . . . . 19 WoMen’s eConoMIC Well-beInG . . . . . . 21 Gender-Based Issues in the Economic Sphere . . . . . . 21 PRF in the Economic Sphere . . . . . . 21 Economic Capital . . . . . . 21 Use of Time . . . . . . 23 Engagement in Entrepreneurship . . . . . . 25 Food Security . . . . . . 27 iv I m pa c t o f c o m m u n I t y- D r I v e n D e v e l o p m e n t p r o j e c t s o n G e n D e r WoMen’s aCCess To PublIC servICes . . . . . . 29 Gender-Based Issues in Access to Public Services . . . . . . 29 PRF Generated Access to Public Services . . . . . . 30 Access to Public Services . . . . . .30 Influencing Access to Health Services and Education . . . . . . 31 reCoMMenDaTIons For nexT sTePs . . . . . . 33 Ample Reasons and Opportuntinies for Measuring the PRF's Impact on Gender . . . . . . 33 Next Steps . . . . . . 34 reFerenCes . . . . . . 39 annexes . . . . . . 41 Annex 1. Methods Used in the Pilot Test . . . . . . 41 Annex 2. Questionnaire . . . . . . 43 Annex 3. The Virtues of Random Sampling . . . . . . 54 Annex 4. Community Participation at Various Stages of KDP (data collection form) . . . . . . 56 lIsT oF Tables Table 1. PRF menu of subprojects . . . . . . 6 Table 2. Public positions held in Lao PDR, by sex (2009) . . . . . . 12 Table 3. Percentage of women actively participating at VNPA meetings . . . . . . 13 Table 4. Proposed PRF subprojects by sex and cycle . . . . . . 16 Table 5. Approval of proposed PRF subprojects by sex and cycle . . . . . . 16 Table 6. Type of other village-level meetings attended . . . . . . 18 Table 7. Effects of PRF’s subprojects on household income . . . . . . 22 Table 8. Effects of PRF’s subprojects on household cost of living . . . . . . 22 Table 9. Effects of PRF’s infrastructure subprojects on the time spent on household tasks . . . . . . 24 Table 10. Persons making decisions to seek health care for women . . . . . . 31 Table 11. Summary of gender-based indicators . . . . . . 34 Table 12. Target and performance indicators for selected Millennium Development Goals . . . . . . 36 Table 13. Illustrative target and performance indicators for PRF . . . . . . 37 lIsT oF boxes Box 1. Method Used in the Pilot Test . . . . . . 3 Box 2. Evaluative criteria for the gender-based indicators . . . . . . 9 v acronyms and abbreviations CDD Community-driven development GoL Government of Lao People’s Democratic Republic GRID Gender Resource Information and Development Center LECS Lao expenditure and consumption survey MIS Management information system M&E Monitoring and evaluation PDR People’s Democratic Republic PRF Poverty Reduction Fund SPIM Subproject implementation VNPA Village needs and priority assessment Note: Unless otherwise noted, all dollars are U.S. dollars. vii acknowledgments T his initiative was made possible through support from the Poverty Reduction Fund (PRF)—an organization under the prime minister’s Office in Lao PDR—under the management of Sivixay Say- sanavongphet, executive director and supported by The World Bank Task Team was comprised of: Nina Bhatt and Gillian Brown (Task Team Leaders); Helle Buchhave, Anders Engvall, Elizabeth Mann, Richard Tobin and Emma Townsend-Gault (consultants). The Enterprise Development Cooperation provided sup- Dr. Bounkhouang Khambounheuang, Phonesavanh port for data collection through Mouthita Phonephet- Luangsouphom, Hatthachan Phimphanthavong, Pascal rath, Dr. Vivath Sauvaly and Anousack Somsanith. Rigaldiès, Syvilay Vorachak, and many of PRF’s regional and district staff. The initiative was funded by the World The following World Bank staff provided both imple- Bank’s Gender Action Plan and supported by the Swiss mentation and intellectual support: Patchamuthu Illan- Agency for Development and Cooperation. govan (Country Manager) and Sean Bradley, Patricia Fernandes, Markus Kostner, Stephanie Kuttner, Vilay- The Lao Women’s Union contributed to the initiative, vanh Phonepraseuth, Helene Carlsson Rex, Jamele with special support received from Dr. Douangsamone Rigolini, and Ingo Wiederhofer. Peer reviewers from the Dalavong, under guidance from Mme. Sisay Leuded- Bank included Kathleen G. Beegle, Janmejay Singh, and mounsone, the Union’s president. Nora Dudwick. Several community-driven development projects in Lao PDR support villagers by funding minor infrastructure schemes, like this water source in Oudomxay Province. Women now can fetch water in the center of the village. Photo © Helle Buchhave/World Bank ix executive summary C ommunity-driven development (CDD) projects seek to empower communities, reduce poverty, and improve economic and social conditions of the poor, typically in rural and remote areas. No less important, CDD also typically addresses two persistent 77 Assist villagers to develop public infrastructure and gain improved access to services; 77 Build capacity and empower villages in poor districts to manage their own development in a decentralized and transparent manner gender gaps: (1) women’s lack of voice in public decision 77 Strengthen local institutions to support partici- making, and (2) their poor access to services and markets. patory decision-making that includes women and the poor. Much of the development community finds CDD to be appealing, and its use is widespread and growing. None- PRF has successfully demonstrated its commitment to theless, the evidence to support the assumption that CDD increasing opportunities for women, but this commit- effectively promotes development and enhances women’s ment is not suitably matched with efforts to document opportunities can be strengthened. Although most devel- these successes. For this reason, PRF and the World opment projects have monitoring systems, these systems Bank collaborated in the identification and field testing often focus more on outputs and less on outcomes and of indicators that could usefully and appropriately con- impacts. Such systems do not provide policymakers, man- tribute to an improved understanding of PRF’s effects agers, and stakeholders with an understanding of the suc- on women in Lao PDR. The toolkit assesses eighteen cess or failure of their projects or whether the well-being indicators in terms of four evaluative criteria: validity, of the intended beneficiaries has improved. reliability, and ease and cost of collecting the data asso- ciated with the indicators. When applying these crite- To address this concern, the World Bank’s Environ- ria, several of the indicators fared well. They are recom- ment, Social, and Rural Unit initiated a series of studies mended for PRF’s ongoing use as are a series of steps in Southeast Asia to identify and assess outcome-based that PRF might consider should it decide to incorporate indicators that can be used to assess the gender-related the indicators into the PRF’s monitoring and evaluation consequences of CDD projects. This toolkit contributes processes. to the initiative through consideration of the Poverty Reduction Fund (PRF) in the Lao People’s Democratic No less important, PRF’s efforts to promote—and mea- Republic (Lao PDR). PRF, an autonomous organization sure—women’s opportunities are potentially instructive established in May 2002, seeks to: for CDD projects elsewhere. Villagers working at construction site in Oudomxay province. Photo © Helle Buchhave/World Bank 1 Introduction C ommunity-driven development (CDD) is an approach to poverty reduction in which the intended, community-level beneficiaries of development programs participate in and influence decisions about the allocation and investment of the gers calls for rigorous monitoring and evaluation, but these calls often neglect outcomes related to gender and social equity. As the World Bank, the Food and Agricul- ture Organization, and the International Fund for Agri- cultural Development agree, “little progress has been programs’ resources. As Wassenich and Whiteside made in measuring outcomes in these areas� (World (2004) comment, CDD operates on “the principles of Bank 2009). local empowerment, participatory governance, demand- responsiveness, administrative autonomy, greater down- A further characteristic of many monitoring systems ward accountability, and enhanced local capacity.� CDD is that they are designed to capture compliance and to projects share common objectives: to empower com- “count� outputs.1 Such systems focus on monitoring munities, reduce poverty, and improve economic and and assessing how well a project, program, or policy is social conditions of the poor, typically in rural and executed. These systems focus on outputs, which typi- remote areas. No less important, CDD also addresses cally measure the goods and services that are produced two persistent gender gaps: (1) women’s lack of voice or supplied. A drawback with this approach is that it in public decision making, and (2) their poor access to does not provide policy makers, managers, and stake- services and markets. holders with an understanding of the success or failure of the project or whether the well-being of the intended Given these attributes and perceptions that top-down beneficiaries has improved (Kusek and Rist 2004). decision making may be less effective than CDD, many clients and much of the development community find In contrast, results-based monitoring and evaluation CDD appealing. Its use is widespread and growing. (M&E) provides feedback on outcomes and the goals Nonetheless, the evidence to support the assumption of project actions. Results-based monitoring helps to that CDD effectively promotes development and gen- answer the “so what� question (Kusek and Rist 2004). der equality can be strengthened. As an illustration, Programs or policies may be implemented successfully, Mansuri and Rao (2003) concluded that “evidence on but have they produced the intended benefits? Success- the actual record of [CDD] initiatives still lags consider- ful implementation of poorly designed projects is of little ably behind the speed at which such projects are being value; doing the wrong thing well rarely rescues a poorly implemented and ‘scaled up.’� A more recent study designed project. The introduction of a results-based (World Bank, Food and Agriculture Organization, and M&E system takes decision makers an important step International Fund for Agricultural Development 2009) beyond traditional implementation-focused monitoring reached a similar conclusion, namely that “the docu- systems (that attend primarily to outputs) to assessing mentation and evaluation of decentralization and CDD whether and how desired outcomes are being achieved. on building accountability to rural women and trans- forming gender relations are extremely limited.� 1As Kusakabe (2005) concludes in a study of gender mainstreaming in South East Asia: “While monitoring the quality of women’s par- However undesirable this situation may be, it is not ticipation and taking steps to improve it might be a good strategy, in general the focus on participation remains quantitative. It is limited uncommon. The investment of large sums of money in to documenting the number of female and male participants in dif- development projects typically and increasingly trig- ferent events.� 2 I m pa c t o f c o m m u n I t y- D r I v e n D e v e l o p m e n t p r o j e c t s o n G e n D e r Outcomes typically focus on the application or use of in the provincial and district levels where the field outputs to provide benefits or improvements in people’s work occurred. In response to a request from PRF, lives (or the improved performance of organizations). its staff in the two provinces also participated in the Unless outcomes receive attention in evaluation, they data collection. may be ignored or neglected. In addition to PRF, the toolkit’s intended audience This toolkit addresses these concerns about the gender- includes task teams of other CDD projects in Lao PDR, based evidence gap. The toolkit identifies and assesses the Lao Women’s Union, and government officials. At indicators to measure, monitor, and evaluate the out- the same time, the toolkit is part of and complements a comes and consequences of CDD on women’s engage- regional initiative to develop and pilot relevant outcome ment in the Lao People’s Democratic Republic (PDR). indicators that can identify long-term impacts of CDD The objective is to create specific tools that enable the projects on women’s opportunities in East Asia. The government’s Poverty Reduction Fund (PRF) and other toolkit also contributes to a Bank-wide effort (2006b, CDD projects to improve their targeting and monitor- 2006c, and 2009) to improve the monitoring and evalu- ing of changes in women’s opportunities and actions, ation of gender-related impacts by gathering data disag- while also assessing ways in which the measurement gregated by gender, by developing indicators to mea- of the impact and consequences of CDD projects on sure results and impacts with respect to gender, and by women can be improved. The toolkit offers PRF’s man- ensuring that gender is included as a key variable in the agement options to strengthen its operations to support search for explanations of the outcomes and impacts of increased opportunities for women to engage in their development projects. local public and economic spheres and provides tan- gible suggestions on how to enhance PRF’s monitoring and evaluation. To do so, the toolkit assesses a series of GENDER-RELATED INDICATORS indicators in terms of their validity and reliability, con- Women can be important drivers of sustainable devel- siders the cost and ease of the necessary data collection, opment, and their increased involvement in this process and offers suggestions about how to integrate the indi- is increasingly included among the objectives of devel- cators into PRF’s M&E framework. opment projects, programs, and policies. Nonetheless, with CDD’s emphasis on meaningful participation in The indicators considered in this toolkit reflect an decision making, the approach has the potential to pro- inventory of national and PRF-generated gender data mote equality of opportunity for everyone. and an analysis of potential links between key gen- der gaps in Lao PDR and PRF’s objectives. In turn, to As the UN Millennium Task Force on Education and promote the toolkit’s value and to encourage PRF’s Gender Equality (cited in United Nations Develop- eventual ownership, the World Bank worked closely ment Fund for Women, UNIFEM 2008) emphasizes, with PRF to design and implement a field-based pilot gender equality has several dimensions, notably access test to assess the practical value of the indicators (see to resources, opportunities, and capabilities related to Box 1).2 The Bank team and PRF discussed the meth- education and health. Access involves representation odology for the field work, the data collection, the and participation in the decision-making processes practical aspects of the field work in two provinces, that affect people’s lives, as well as equality of opportu- and the desirability of coordination with PRF’s staff nity to use one’s capabilities to access economic assets, including income and employment.3 As the Task Force further noted, empowerment, or the ability to control 2However appealing indicators of outcomes may be conceptually, one’s own destiny, is also essential. Access to resources they are of little practical value unless they share several essential characteristics and can be used without difficulty by those respon- and equal capabilities are insufficient when women sible for monitoring and evaluation. For these reasons, pilot testing in do not also have the capacity to make decisions and the field is a critical step before developing and investing in a results to express and exercise their preferences so they can framework that identifies intended outcomes and impacts. Piloting is a way of learning what works and what does not. It also alerts manag- improve their lives. ers that there may be some indicators for which data do not exist or for which data are too costly, time consuming, or complex to obtain 3Gender equality, according to the Task Force, also requires secu- (Kusek and Rist 2004). Piloting represents a live test of a design and rity (as reflected in the elimination of violence against women and data-collection methodology, as well as the last step before questions girls), but this is not an area in which the PRF has any mandate or of interest can be addressed (Iarossi 2006). responsibility. I m pa c t o f c o m m u n I t y- D r I v e n D e v e l o p m e n t p r o j e c t s o n G e n D e r 3 box 1 Method used in the Pilot Test The field work for the piloting of the indicators was conducted in 2009 in villages in which PRF has been active since its first annual cycle, which began in February 2003. Six villages in two districts were included in the pilot test. To capture regional differences, three villages were randomly selected from PRF-targeted communities in the Sukhuma District of Champassak Province in southern Lao PDR and three villages in the Huamuang District in Huaphanh Province in northern Lao PDR. Thus the sampling frame consisted of: (a) one district in southern Lao and one in northern Lao and (b) three randomly selected villages in each district where PRF has been active since 2003 and remains active. Several methods were used to gather information about the PRF in the six villages. Individual Interviews Data were collected from 125 women, all of whom had all participated in PRF’s processes within their villages. Some of the women had also represented their villages in koumban- and district-level meetings. About 60 percent of the women were Lao-Tai, the ma- jority population in Lao PDR. The other 40 percent represented Hmong (26 percent), Xuay (14 percent), and “other� (1 percent). A majority of respondents were in the “middle-age bracket,� older than age 25 but not yet 46 years of age. About one-fifth were under age 25 and a similar portion were older than 45. All respondents were either married (98 percent) or widowed (2 percent). Almost all respondents had children; only five respondents were childless. More than a third of the women had five or more children. Finally, almost all respondents derived their primary income from agriculture, and almost a third derived some income from trade. Other sources of income, such as forestry or fishing, were of minor importance. This pattern is broadly representative of the situation in all rural areas in Lao PDR (National Statistics Center 2005). All the women were interviewed using an identical questionnaire (See Annex 2). The questionnaire was based on the forms PRF cur- rently uses. The questionnaire thus conforms, as much as possible, to forms PRF and the Department of Statistics use.* This makes it possible to add the questionnaire, or parts of it, to PRF’s existing monitoring and data-collection system. The sample of respondents is not intended to be the basis for generalization to all PRF villages (and the sample is too small to do so in a meaningful way). In contrast, the primary purpose of the individual interviews was to test the suitability of the indicators and to assess the cost and ease (or difficulty) of collecting data related to them. Focus Group Discussions In addition to the interviews with 125 women, focus-group discussions were organized in each village with women who had par- ticipated in PRF’s processes. The discussions gathered information about qualitative indicators of gender outcomes. To provide a means for data triangulation, the focus-group discussions offered the participants the opportunity to talk about a range of issues, challenges, observations, and comments on their lives as well as PRF’ s processes and perceived impacts. Interviews with Village Representatives Each of the six villages has two PRF representatives, one man and one woman, and each was interviewed. They represent their villages at the koumban-level meetings and have a central role in voicing village-level concerns and priorities for subprojects. The interviews focused on the nature of active participation of community members and village representatives and tested the relevance of the items in the questionnaires used in the individual interviews, which are discussed above. Case Studies Case studies were conducted in the six villages to assess situations in which PRF has affected women’s opportunities and their en- gagement in local political, economic, and social spheres. The study team used local knowledge to identify suitable participants for the case studies; women who participated actively in the villages’ economic, social, and political life were the main targets. The need to substantiate the relevance of the gender-based indicators for measuring PRF’s impact on women motivated the case studies. The studies offer narratives about women’s engagement in local decision making outside PRF. The case studies provided an opportunity to gather opinions from women active in the PRF process and their perspectives about how PRF is influencing their lives. * The Department of Statistics was formerly the National Statistics Center. 4 I m pa c t o f c o m m u n I t y- D r I v e n D e v e l o p m e n t p r o j e c t s o n G e n D e r If gender equality requires access to resources and inequitable and hinder the development the policies are opportunities and capabilities related to education and intended to promote. health, then it is equally essential that data be available to assess the status of and changes in these characteris- More sophisticated quantitative gender analysis goes tics. Data informing the results of development inter- beyond disaggregation to capture other meaningful ventions such as PRF can come from two sources—a information. This can include information on socioeco- monitoring system and an evaluation system (Kusek nomic outcomes of particular importance for women, and Rist 2001). Monitoring typically involves the peri- such as information on child care, reproductive health, odic collection of information to assess adherence or time devoted to collecting water or wood for fires. to time schedules, completion of required activities, and appraisal of progress. Monitoring entails mea- At the same time, some gender specialists (such as Rein- surement; what is measured is the progress toward harz 1992) argue that quantitative indicators by them- achieving an objective or desired outcome. In many selves insufficiently capture and characterize women’s instances, however, outcomes cannot be measured experiences. For this reason qualitative indicators are also directly. They must first be translated into indicators of value, but quantitative and qualitative indicators are that provide information about whether outcomes are complementary. Both are important for effective monitor- being achieved. ing and evaluation because they can cross-validate each other, while both can also illuminate different but related Gender-sensitive indicators have the special function of problems. For example, quantitative indicators commonly identifying gender-related changes in a community. The reveal the presence of a problem (such as unequal access to usefulness of such indicators lies in their ability to point health services), whereas qualitative indicators can high- to changes in the status and roles of women or men over light the causal relationships creating the problem (such time. The use of indicators and other relevant moni- as social perceptions serving to diminish the importance toring and evaluation techniques can lead to a better of female and maternal health problems). understanding of how results can be achieved, so using gender-sensitive indicators can also lead to enhanced What are appropriate gender-related indicators for Lao project planning and delivery. PDR and its Poverty Reduction Fund? Equally important, what quantitative and qualitative methods can assist in A rudimentary way of capturing gender differences the building of a database to assess PRF’s impact on wom- involves the collection of sex-disaggregated data. Such en’s actions and opportunities? After a brief introduction data capture differences—as well as similarities— to PRF’s operations, the chapters that follow address between males and females and can be used to con- these questions by assessing indicators that relate to (a) struct indicators highlighting biases or (in)equitable women’s opportunities to participate in decision making project outcomes. Sex-disaggregated data can thus be that affects their lives; (b) women’s economic well-being; used to identify policies unfavorable to women or men and (c) women’s access to essential public infrastructure, and to affect changes in policies or practices that are including health, education, and sanitation. 5 The Poverty Reduction Fund T he Poverty Reduction Fund (PRF) was estab- lished as an autonomous organization by prime ministerial decree in May 2002 and was initially supported by the World Bank in the form of a low-interest credit of approximately $19.5 million. early 2010.6 Of these subprojects, over 75 percent involved infrastructure with the remainder devoted to local capacity-building activities. As this distribution of resources suggests, PRF seeks to provide public infra- structure that serves everyone’s needs while strengthening The decree allows PRF to receive and use funds from the participation of traditionally disadvantaged groups in other sources.4 the Fund’s decision-making processes. Due to PRF’s size and coverage, it is among the most important community- PRF aims to create strong links between the local gov- driven initiatives in the country (World Bank 2008a). ernment and the aspirations of villagers with PRF staff at the district, provincial, and national levels. The Fund has three primary objectives: PRF’S FUNDING CYCLE PRF invests in an annual cycle of activities; the first cycle 77 Assist villagers to develop community public began in February 2003. The Fund promotes its principles infrastructure and gain improved access to and methods in all villages of the targeted districts by services encouraging residents to express their needs, instructing 77 Build capacity and empower villages in poor them how to prioritize their wishes at the village, koumban districts to manage their own public investment (a group of villages), and district levels, and then to orga- planning and subproject implementation in a nize themselves in implementing the investments. Each decentralized and transparent manner participating district receives an annual, formula-based 77 Strengthen local institutions to support partici- grant based primarily on its level of poverty. Seventy-five patory decision-making and conflict resolution percent of PRF’s budget is invested at the village level: in processes at the village, koumban, and district 2009, PRF provided benefits to over 900 poor villages. All levels, involving a broad range of villagers, in- investments proposed by villagers must be within the lim- cluding women and the poor. its of a menu of possible subprojects (Table 1) to ensure they conform to PRF’s objectives, and all investments are PRF launched its first subprojects in ten districts in limited to the equivalent of $30,000 per subproject. The three provinces in 2003 and subsequently expanded to mean amount per project for the three most recent cycles twenty-one districts in seven provinces in 2009–10.5 (covering 2007–10) was about $11,500. The PRF invested $21 million to support more more than 2,900 subprojects between February 2003 and For each subproject, a village signs a contract with PRF. The community is then responsible for construction, 4 PRF received additional funding in 2008 from the World Bank ($15 managing contractors, maintenance, and bookkeeping, million) and from the Swiss Agency for Development and Coopera- with support from local authorities as well as PRF’s dis- tion ($5 million). The additional funding covers three years, from 2009 through 2011. trict and provincial teams. 5 In this report PRF is considered to be the “project� and its activities and investments to be subprojects. Lao PDR has seventeen provinces 6 Between February 2003 and early 2010 (cycles I through VII), PRF and 142 districts. Of the latter, the government considers seventy-two built or renovated 571 schools, 420 roads, 155 irrigation schemes, 84 to be poor and forty-seven of these to be priority districts for imple- bridges, and 54 health facilities. It also provided 3,245 school packages mentation of its poverty reduction strategy. and 1,667 vocational training sessions (PRF 2010). 6 I m pa c t o f c o m m u n I t y- D r I v e n D e v e l o p m e n t p r o j e c t s o n G e n D e r PRF’S APPROACH TO GENDER to reflect the preferences that women had expressed in their single-sex meetings (PRF 2003). Four village rep- PRF seeks to promote equitable involvement of women resentatives were also selected to present their village’s in decision making and the implementation of its sub- priorities at koumban-level meetings, and two of the rep- projects. When it began its operations in 2003, the initial resentatives had to be women. focus was on ensuring women’s opportunity to express their preferences for subprojects. Less attention was A review of these procedures in 2007–08 led to several devoted to the outcomes of the preferences expressed. important changes that have increased women’s roles At the village level, men and women would participate in and opportunities. The single-sex meetings continue; separate meetings where they each identified three pro- each proposes three infrastructure subprojects and two posed priorities for the use of PRF funds. A subsequent priorities for training for consideration at the VNPA village needs priority assessment (VNPA) meeting of meetings. Participants at these meetings still select both genders then created a combined list of three prior- three of the proposed infrastructure subprojects, but ity subprojects for the village. None of the proposals had two of them must now be from the women’s list. Simi- larly, among the two proposals for training activities Table 1 PRF menu of subprojects selected at the VNPA meeting, one must come from the Subproject women’s proposals. Village representatives are also still Type Items Eligible Item Not Eligible selected to attend koumban-level meetings, and one of the representatives must be a woman. Representa- Access Small bridges, footpaths, New roads and road tion of a minority ethnic group is encouraged but not tracks, culverts, ramps, surfacing/sealing piers, road repairs, and required (PRF 2008b).7 The VNPA meetings (as well as upgrading the koumban-level meetings) must also be conducted in the native language of the villagers. Prior to the changes, Community Minihydro generators, Gasoline or diesel many of PRF’s informational materials were in Lao and electrical wiring, line extensions generators supply many of the village-, koumban-, and district-level meet- ings were conducted only in Lao as opposed to the lan- Primary health Health centers (buildings, Generators guages used in the villages. care facilities furniture, latrines, supplies and medicines, temporary allowances At the koumban level, there are additional opportu- for contracted nurses/ nities for women’s equitable participation. A team midwives in cash or of three koumban facilitators heads the socializa- kind, village medical kits, tion and planning process. At least one of the team’s training, scholarships, members must be a woman. The village representa- medical equipment) tives also elect four koumban representatives, who Domestic water Wells, gravity water Piped household water then represent the villages at district-level meetings. systems supply, latrines, etc. hookups Two of the four representatives must be women and Education Schools and nurseries Any supplies provided by one should belong to a minority ethnic group. Each (buildings, latrines, the government koumban procurement team has three elected mem- temporary allowances bers, and at least one should be a woman. The same for contracted teachers, composition holds for each koumban’s implementa- supplies, equipment, tion and maintenance team. 8 furniture, training, scholarships, textbooks, and musical instruments) 7 PRF’s emphasis on representation of the country’s ethnic minorities reflects the fact that more than 80 percent of the population of the Agricultural Weirs, ponds, canals, Electrical pumps forty-seven priority districts is from non Lao-Tai ethnic groups. In Infrastructures bunds, gates, spillways, contrast, the Lao-Tai population represents about two-thirds of the and other structures country’s overall population (PRF 2008b). Villagers also select three members of a village implementation and maintenance team, one of Markets, Buildings, drainage, Generators whom must be a woman. community wells, and furnishings 8 PRF encourages village representatives to choose at least one person halls for the procurement team and inspection and maintenance teams who belongs to a smaller ethnic group if the person has the appropri- Source: PRF (2008b). ate education. I m pa c t o f c o m m u n I t y- D r I v e n D e v e l o p m e n t p r o j e c t s o n G e n D e r 7 The mandatory inclusion of women in PRF’s decision- making has important parallels in other countries. More Women attending PRF meeting in Hua Moung than thirty countries require or “reserve� quotas for village. A key feature of women’s political representation (World Bank 2001), PRF’s approach to gender and some research has investigated whether these quo- is that men and women tas affect the choice of public policies and the allocation participate in separate of public resources. As an illustration, Chattopadhyay meetings, where they each and Duflo (2004a, 2004b) examined the distributional identify three proposed consequences of women’s representation in village-level priorities for the use of PRF councils, or panchayats, in India. funds. © Anders Engvall/ World Bank As implemented, the seats reserved for women in India were allocated randomly, which means that some pan- chayats had women members while others did not, at least during the period of the study. This procedure allowed Chattopadhyay and Duflo to compare decisions made in villages with and without women’s representa- tion in the panchayats. They found important differences. In the two Indian states included in the study, panchayats with women members invested more in goods that were relevant to the needs of local women and less in goods that were less relevant to the needs of women compared with panchayats that did not have women members. As aspects of the Fund’s activities, including the partici- the authors concluded, India’s reservation policy has pation of women and ethnic groups. For this reason, important effects on policy decisions at the local level, gender-disaggregated data are collected in each village. and these effects were consistent with women’s priorities. These data include the number of women and men in PRF villages; the number of each gender participating in This research reinforces the desirability of assessing the decision-making meetings and in training events; and gender-based outcomes of PRF’s procedures. In turn, the registration of female and male representatives and the Lao experience offers the potential to contribute to facilitators in PRF meetings at various administrative the growing literature on the policy consequences of levels. Subprojects selected by women-only and men- gender-sensitive policies. only groups are also entered into the database (PRF 2008b). PRF’S MONITORING AND The MIS provides data for measuring the progress and EVALUATION FRAMEWORK achievements during the planning and implementa- PRF has a comprehensive M&E framework with three tion of subprojects. It also stores quantitative informa- main objectives: (a) to ensure that the Fund proceeds tion collected through a series of surveys conducted according to PRF’s principles and procedures and that at various levels. In addition to reports generated for inputs are efficiently transformed into targeted outputs internal purposes, the Fund’s M&E system ensures and outcomes to achieve PRF’s development objectives; that decisions about future funding are based on (b) to document PRF’s experiences and provide guid- results and assures stakeholders that money is spent ance for policy making; and, (c) to support day-to-day appropriately. The MIS data also help inform external management. Provincial M&E staff compile and enter assessments. PRF also commissions thematic reports information into the Fund’s management information and assessments by external consultants all of which system (MIS). The national PRF office maintains a com- incorporate MIS-generated data for an analysis of plete, updated, and aggregated master database of data PRF’s impact. from all participating provinces. The existence of an M&E framework and a well-func- In addition to the responsibilities just noted, PRF’s M&E tioning MIS, both of which exist within the PRF, do not unit must also ensure that it captures and monitors all currently provide the information needed to fully assess 8 I m pa c t o f c o m m u n I t y- D r I v e n D e v e l o p m e n t p r o j e c t s o n G e n D e r the Fund’s contribution to gender equality.9 Likewise, the data currently available within the system do not address the full range of potential opportunities for informing CDD stakeholders about gender-related ben- efits or an explanation of why these benefits do or do not occur. If PRF is successful in promoting gender equality, then it is clearly desirable to monitor and document this success and to allow other CDD efforts to benefit from PRF’s experience. Furthermore, although the primary focus of the tool- kit is on how PRF can identify the gender-based con- sequences of its intervention, a secondary purpose is to enhance the internal capabilities of PRF’s M&E staff. As is often the case, organizations’ monitoring capabilities typically exceed their capabilities in evalu- ation. Consideration of the indicators discussed in this toolkit and the methods used to collect the neces- sary information for these indicators can provide an opportunity to improve PRF’s evaluation skills. PRF staff members have access to the data collected during the pilot testing; additional analyses of these data can improve PRF’s analytic skills. Several of the indicators present opportunities to use rigorous and state-of-the-art evaluation designs, including those associated with impact evaluations that involve treatment and comparison groups as well as data collected before and after PRF’s inter- ventions in a village. These designs may be of con- siderable interest to PRF. PRF’s Administrative Board has expressed a desire to know more about the Fund’s quantitative outcomes (Swiss Agency for Develop- ment and Cooperation and the World Bank 2009), and impact evaluations have been suggested as a way to respond to these desires. CONSIDERING GENDER-BASED INDICATORS FOR PRF For PRF’s monitoring system to remain effective and efficient, it must make wise choices about which indica- tors to use and which data to collect. Moreover, how- ever appealing gender-based indicators may be concep- tually, they will be of little practical value unless they Women digging canal for piped water source granted 9The Government of Lao PDR (GoL) recognizes the value of strength- by a community-driven development project in ening the integration of women into socioeconomic development. Sayaboury Province. Photo © Adam Rogers/UNCDF Promoting gender equality is a central national goal, as expressed in the government’s Five Year Socio-Economic Development Plan (2006–10) (GoL 2006). The plan also identifies PRF as a key vehicle for implementing the national poverty reduction strategy. I m pa c t o f c o m m u n I t y- D r I v e n D e v e l o p m e n t p r o j e c t s o n G e n D e r 9 share several essential characteristics and, ideally, are tested in the environment in which they will be applied box 2 evaluative criteria for gender-based indicators before they are incorporated into the Fund’s M&E sys- Indicators are deemed to be valid when the information they provide tems. These essential characteristics are reflected in is close to the reality being measured. As an illustration, enrollment four evaluative criteria: validity, reliability, ease of col- rates in primary schools are generally considered to be a valid indica- lection, and cost (Box 2) tor of the number of students in school. In contrast, the number of a company’s customers is not necessarily a valid indicator of the eco- The four evaluative criteria are the basis for the pilot nomic viability of that company. Validity can be enhanced through testing of the gender-based outcome indicators dis- triangulation, which involves the use of multiple sources of informa- cussed in the chapters that follow. For each of the indi- tion and data as well as through the use of methods that rely on both cators the goal is to provide a practical approach that quantitative and qualitative approaches to data collection. explains how to integrate gender-based outcome indi- An indicator is reliable when multiple uses of the same method of cators into PRF’s M&E system. This goal is addressed measurement yield the same or similar results, regardless of who is by (a) assessing each indicator in terms of the four using the indicator. Reliable indicators are of no value when the indi- evaluative criteria; (b) making recommendations about cators are not also valid. the suitability of the indicator; and, (c) offering some Ease of collection is largely self-explanatory. Collection of data for options and suggestions about data collection and who the indicator should not be unduly burdensome or require skills or should collect the data. expertise that exceed the data collectors’ abilities. Not unexpectedly, the pilot test revealed that some indi- Cost considerations are essential to ensure that the data needed for the indicators can be collected and used at a reasonable cost. cators are more (or less) desirable than others. Indeed, the pilot test revealed that some indicators should be dropped from further consideration due to concerns about their validity, reliability, or ease or costs of data some weaknesses or concerns as well, and a double ++ collection. Nonetheless, these indicators are discussed reflects strength or appeal with respect to the evalu- to ensure that readers understand why the indicators ative criterion. Regardless of an indicator’s seeming are not recommended for PRF’s use. appeal based on the four criteria, there may be other factors that increase or diminish that appeal. For For each indicator discussed below a rating is pro- this reason, although PRF is encouraged to consider vided for each of the four evaluative criteria. A rating adopting the highly rated indicators, the final decision of – indicates a problem or weakness with the evalu- about which indicators to adopt is best and appropri- ative criterion, a single + reflects some strengths but ately left to PRF. A total of 125 women in four villages in the provinces of Champassak and Huaphanh were interviewed for the pilot testing of gender indicators. They had all participated in the PRF process at local level and some had represented their villages in koumban or district meetings. Photo © Anders Engvall 11 Women’s opportunities to Participate in Decision Making A ccording to the Gender Resource Information and Development Center (GRID) of the Lao Women’s Union, women’s empowerment should provide opportunities and create appropriate conditions for women to build their own capacities and decision- women’s representation in the National Assembly is not matched at lower levels of government. In 2004, there were only three female district governors and two female vice-district governors throughout the country. The situ- ation changed little in the years that followed (Table 2). making abilities about their lives and communities. To be empowered, women must have access to basic services, Among government employees, about six of ten are have confidence in their ability to improve their lives, and male, but males occupy more than 80 percent of the participate in development activities. most senior positions (World Bank 2010). There were no female village heads in 2000. Less than one in fifty vice- For the purposes of this report, the discussion of the com- village heads were women in 2005 (GRID 2005), although ponents of gender equality well illustrates the compelling this situation did improve in later years (Table 2). links between the support for women’s participation in local decision making and PRF’s objective to include women in A livelihood study sponsored by the UN Development its decision-making processes. Does PRF effectively pro- Program and the National Economic Research Institute of mote these objectives? If so, what evidence is there to dem- Lao PDR (Alton and Ratthanavong 2004) in Luang Nam onstrate that PRF is doing so? To answer these questions, Tha Province identified similar gaps. The authors found in this chapter the toolkit derives and assesses indicators limited women’s representation in virtually all formal vil- related to political participation. In chapter 4, it looks at lage organizations other than the Lao Women’s Union and indicators for economic well-being; and in chapter 5, indi- the youth league. None of the eight villages in the study cators for women’s access to essential public infrastructure, had women represented at their development councils. including health, education, water, and sanitation. The Lao National Commission for the Advancement of GENDER-BASED ISSUES IN THE Women (2009) recently lamented the “very low partici- pation of women in all areas of public, political, and pro- POLITICAL SPHERE fessional life, including the realms of government, diplo- As in many countries, regardless of their level of devel- macy and public administration.� The committee also opment, women are underrepresented in Lao’s politi- noted its concern about the lack of women’s representa- cal sphere. Among the members of Lao PDR’s National tion in the judiciary and among the police and “the low Assembly elected in 2006 (for a term of five years), about number of women in senior management in general.� To one quarter were women, which is high by international address this situation, the committee recommended that standards but disproportionately low based on the pro- the government pursue policies to promote women’s “full portion of women in the country.10 More important, and equal participation� in decision making in all areas of public, political and professional life and to “empower vil- 10Lao PDR’s Decree on the Electoral Law for the National Assembly declares that it shall “include appropriate proportion of the represen- lage women so that they can participate equally in village tatives of the people of different strata, sex, and ethnic minorities� matters and serve as village chiefs.� (cited in Lao National Commission for the Advancement of Women 2005). Of the 115 members of the National Assembly elected in April 2006, 86 (or 75 percent) were men. The vice-president of the National Despite such advocacy on behalf of women, national Assembly in 2009 was female (Buchhave 2009). policies that attempt to mainstream gender into deci- 12 I m pa c t o f c o m m u n I t y- D r I v e n D e v e l o p m e n t p r o j e c t s o n G e n D e r Table 2 Public positions held in lao PDR, by sex (2009) Women Men Position Total Number % Number % Ministers and vice ministers 64 6 9.4 58 90.6 Provincial governors 17 0 0.0 17 100.0 Vice-provincial governors 34 1 2.9 33 97.1 District governors 143 2 1.4 141 98.6 Vice-district governors 192 4 2.1 188 97.9 Village head 8,726 148 1.7 8,578 98.3 Vice-village head 17,128 863 5.0 16,265 95.0 Source: Statistics on Local Administration 2010, Ministry of Home Affaires, Department of Local Administration, Division of Local Administration and Statistics. Here from World Bank, 2010 sion making have a tendency to “evaporate� at the lower Active Participation levels. In the case of Lao PDR, Kusakabe (2005) con- PRF monitors the number of women and men attend- cluded that training on gender awareness and a nation- ing the village needs priority assessment (VNPA) meet- wide women’s network did not ensure gender equality ings, but mere attendance is a poor indicator of women’s throughout the country. influence and thus an inadequate and insufficient mea- sure of outcomes (Narayan 2005; World Bank, Food and Agriculture Organization, and International Fund for Agricultural Development 2009). Moreover, when PRF IN THE POLITICAL SPHERE attendance is perceived to be compulsory—as is often The low levels of women’s participation in decision mak- the case with PRF (World Bank 2008a)—attendance is ing at the local level represent an issue that PRF’s proce- not a refined measure of influence. Attendance at VNPA dural requirements address. At the village and koumban meetings can be useful as a measure of outputs and levels PRF generates outputs related to political partici- compliance with PRF’s requirements, but active partici- pation: that is, meetings where villagers participate in pation is a more meaningful indicator of engagement. decision making and receive information about the deci- PRF recognizes the importance of active engage- sions made at these meetings. ment. According to PRF’s guidelines for facilitators of the VNPA meetings, participants should be encour- Both of these outputs create opportunities for PRF to aged to express their opinions to other participants influence the gender gap discussed above and suggest (PRF 2008b). As a result, measuring the proportion of several questions related to gender-related outcomes. women using this opportunity during VNPA meetings As an illustration, does women’s increased participation may reveal the level of active and meaningful participa- in PRF’s decision-making processes result in: tion. Similarly, forming an opinion on a village’s needs and expressing this opinion by voting or speaking at the 77 More active participation among women in meetings can be a proxy for how actively women engage other decision-making venues? in PRF’s decision-making processes. 77 The provision of public services that are priorities for women? Potential indicators: 77 Increased capacity among women to engage in Indicator P1: Percentage of women at VNPA decision making at the village, koumban, and meetings who speak or vote district levels? 77 Increased participation in public decision-mak- P1: Validity ++ Reliability ++ Ease of collection + Cost of collection + ing forums unrelated to PRF’s processes? I m pa c t o f c o m m u n I t y- D r I v e n D e v e l o p m e n t p r o j e c t s o n G e n D e r 13 Indicator P2: Percentage of women at VNPA Table 3 Percentage of women actively participating at meetings who are active participants VNPa meetings P2: Validity ++ Reliability ++ Lao-Tai Women Ethnic Minorities Ease of collection + Cost of collection + (%) (%) Indicators P1 and P2 accurately capture the level of Only talked in front of group 14.7 6.1 participation among women attending village-level Only voted on subprojects 25.3 28.6 PRF meetings. The two indicators summarize the level Both talked and voted 46.7 53.1 of active participation at PRF meetings. During the pilot testing, women were asked whether they spoke in Neither talked nor voted 13.3 12.2 front of the other participants at the most recent VNPA N= 75 49 meeting they had attended and whether they had voted on the proposed subprojects. The pilot testing found the indicators to be fairly consistent when replicated, thus Recommendation confirming their reliability.11 The indicators are recommended for PRF’s use but with a revised process for data collection to improve If a woman either spoke or voted (but not both), she was reliability and the ease and cost of this collection. recorded as “actively participating.� About two-fifths of The indicator is an ideal measure of outcomes for the Lao-Tai women and about one-third of the minority gender-focused reports, reviews of participation pro- women surveyed belonged to this group of speakers or cesses, and for annual reports on the progress of PRF’s voters (Table 3). The women that both spoke and voted implementation. were deemed to be the most active. About half of all the women interviewed belonged to this group. The inac- At least four options exist for collecting the information tive women neither spoke nor voted. needed for the two indicators: The relative consistency between the Lao-Tai and 77 A PRF meeting facilitator’s assistant or village the minority women is both interesting and instruc- M&E assistant can observe, count, and record tive. Although it is not possible to generalize to all participation during VNPA meetings using a PRF villages, the data in Table 3 suggest that PRF has standard form.12 For purposes of comparison, successfully encouraged and increased the participa- the information about participation should be tion of ethnic minorities. In prior years, PRF (2008c) collected about the behavior of both males and had expressed concerns that language barriers and females.13 the lack of translation services had discouraged this 77 A facilitator, facilitator’s assistant, or village participation. PRF’s policy that requires VNPA meet- M&E assistant can ask participants (again, both ings to be conducted in the local language provides a males and females) as they leave VNPA meet- plausible explanation for the nearly equivalent levels ings whether they spoke or voted at the meeting of active participation. This is all the more notewor- and record the replies anonymously.14 To pro- thy because the minority women in the survey were mote the success of this approach, the facilitator much less likely to have completed upper primary would explain the process to the participants education than were the Lao-Tai women. In contrast, before they leave the meeting and assure them of the data do not reveal the extent of women’s partici- the anonymity of their responses. pation relative to men’s. 12 An example of a form used in Indonesia for a similar purpose can 11 The pilot testing did not take place when the village meetings oc- be found in Annex 4. curred, so women were asked to recall their participation. This situ- 13 Recording information about speaking and voting from all partici- ation required active prompting from the interviewers. The need to pants may require that two people be responsible for recording what prompt the respondents increased the cost of data collection because occurs. it increased the time required for the data collection. Moreover, the 14 If this option is chosen, PRF should consider the possibility that prompting may have encouraged some women to provide responses some women may be hesitant to reveal whether they spoke or voted they believed interviewers wanted. In addition, the interviewers had at the VNPA meetings. Likewise, the women may provide answers to assume that the women’s recall matched what had actually oc- that they perceive to be “correct� or expected by the person asking curred at the meetings. the questions. 14 I m pa c t o f c o m m u n I t y- D r I v e n D e v e l o p m e n t p r o j e c t s o n G e n D e r Women from the White Hmong ethnic group participating in a focus group discussion for this initiative in Lang Anh village. One of the key findings of the 2008 PRF evaluation was that village meetings had to be conducted in the native language of the villagers in order to better ensure active participation, particularly by women, since more ethnic women than men only speak their native language. Photo © Andres Engvall/World Bank 77 Secret balloting is a possibility. Before each Once the data are collected provincial-level PRF staff VNPA meeting each participant could be given a can enter the data into the Fund’s management infor- token, which could be as simple as a small stone, mation system. The data could then be analyzed at the to use when voting for or against proposed national level and included in PRF’s annual reports. subprojects. Males and females would receive tokens of different colors, so the results could be Awareness and Capacity Building disaggregated by gender. For women—and naturally also men—to be able to par- 77 The data can be collected by PRF staff (or a con- ticipate in the PRF process, it is necessary to understand sulting firm) in a random sample of villages each and be aware of how decisions are made. Conversely, if year. The questions can be included in question- participants fail to grasp the basics of PRF’s decision-mak- naires and would require about ten minutes per ing processes, achieving the Fund’s objectives regarding person to collect. The challenge with this approach ownership and participation would be challenging. Con- is that it relies on participants’ ability to recall sequently, women’s knowledge of the subprojects nomi- events, several of which may be considered insig- nated at the last VNPA meeting can serve as a proxy for nificant, that occurred many months in the past. awareness of PRF’s processes (World Bank 2005b). I m pa c t o f c o m m u n I t y- D r I v e n D e v e l o p m e n t p r o j e c t s o n G e n D e r 15 For capacity building, if women have developed knowl- whether they were aware of the priorities identified at edge about how to affect public decision making through the previous meeting. The easiest (and recommended) their participation in the PRF process, then qualitative approach would be to include collection of the infor- data can be collected to assess how lessons women have mation in PRF’s current data-collection procedures learned have been used to engage in decision making used for each VNPA meeting. outside the PRF operation. Once the data are collected, they would be entered into Potential indicator: the Fund’s MIS at the provincial level and analyzed at the national level. Finally, because the indicator repre- Indicator P3: Percentage of women who are aware of PRF’s decision-making processes sents a measure of outcomes related to capacity build- ing and participation, the indicator could appropriately P3: Validity + Reliability ++ be included in PRF’s annual reports. Ease of collection ++ Cost of collection ++ Impact of Women’s Participation on This indicator relies on a proxy measure to assess Decision Making within PRF women’s awareness of PRF’s decision-making pro- One of PRF’s objectives is to empower villagers, espe- cesses. Relying on a proxy measure invariably raises cially women, the poor, and minorities, to engage in concerns about an indicator’s validity because aware- decision making that affects their lives and the imple- ness of a decision is not the same as awareness of how mentation of the Fund’s subprojects. 15 Empower- the decision was reached. In contrast, the reliability of ment occurs when women can advance their interests indicator P3 is high. Women’s responses about which through their own choices and actions (Petesch, Smulo- subprojects were nominated can be compared with the vitz, and Walton 2005). For this reason it is desirable for actual list of nominated subprojects. This list, for every CDD programs like PRF to establish a mechanism to PRF village, is already included in PRF’s management monitor and report on women’s successes in achieving information system. Data on women’s recall of previ- their desired outcomes because of their participation in ous decisions can be collected at little cost at the start public decision making. of each PRF cycle. PRF presently collects data that reflect women’s suc- Among all the women interviewed approximately one cess in advancing their priorities at VNPA meetings. As month after the VNPA meetings had occurred, includ- shown in Table 4, 9 percent of the priorities identified by ing the ethnic minorities, just over 70 percent indi- women-only groups in PRF’s first cycle were endorsed cated they knew what subprojects the women’s group at the larger VNPA meetings compared with 14 percent in their village had nominated in the most recent of the priorities advanced by the men-only groups. In VNPA meeting they had attended. The responses indi- other words, men were more successful in advancing cate a high degree of awareness of decisions taken and their agendas than were women. Nonetheless, a far that most women are able to remember the outcomes higher percentage (77) of proposals that both men and of the decision-making process. The results thus sug- women had favored were approved as village priorities gest that women are knowledgeable about the deci- at these meetings. sion-making processes and are informed participants in these processes. By the fourth cycle, the success at VNPA meetings of proposed subprojects originating from women-only and Recommendation men-only groups had declined to 5 and 6 percent, while Data on indicator P3 should be collected routinely the success of proposals that both groups had advocated from both men and women. There are several ways had increased to 89 percent. Much changed during the these data can be collected. As in the pilot test, a ques- sixth cycle when PRF changed its procedures. As noted tionnaire could be used. With well-trained enumera- above, beginning in 2008–09, PRF required that at least tors, the use of a questionnaire would promote con- two of three infrastructure proposals selected at each sistent measurement because the same data-collection VNPA meeting must come from the women’s list. There methods would be used in each village. With another has been a large increase in the number of proposed alternative, PRF staff could collect the required infor- mation immediately before a VNPA meeting simply 15According to the World Bank (2000), empowerment is one of the by asking participants to raise their hands to indicate key priorities of development policy. 16 I m pa c t o f c o m m u n I t y- D r I v e n D e v e l o p m e n t p r o j e c t s o n G e n D e r Table 4 Proposed PRF subprojects by sex and cycle Cycle 1 Cycle 2 Cycle 3 Cycle 4 Cycle 5 Cycle 6 2003-04 2004-05 2005-06 2006-07 2007-08 2008-09 Subprojects proposed by: (%) (%) (%) (%) (%) (%) Women-only groups and approved at VNPA meetings 9 7 8 5 7 21 Men-only groups and approved at VNPA meetings 14 10 10 6 8 20 Both women-only and men-only groups and 77 83 82 89 85 59 approved at VNPA meetings Source: PRF, personal communication. Table 5 approval of proposed PRF subprojects by sex and cycle Cycle 1 Cycle 2 Cycle 3 Cycle 4 2003- 2004- 2005- 2006- 04 05 06 07 Subprojects proposed by: (%) (%) (%) (%) Women-only groups and converted into implemented subprojects 7 6 4 2 Men-only groups and converted into implemented subprojects 12 5 5 1 Both women-only and men-only groups and converted into implemented 81 90 91 97 subprojects Source: PRF (2008a). subprojects that differ between men and women (and a village may also have discussed and decided their priori- corresponding decrease in the percentage of proposed ties before the single-sex meetings. It is not possible to subprojects on which both men and women agreed in know whether and how well these decisions reflected the single-sex meetings). women’s preferences and priorities as opposed to deci- sions that may have been imposed on women in tradi- A more important issue concerns the successful con- tional ways—by men. version of proposed subprojects into implemented sub- projects. As shown in Table 5, the gender-based results An alternative explanation suggests that decisions are similar to those discussed above. As an illustration, about which priorities to advance to the VNPA and 7 percent of proposals originated from women-only koumban meetings may occasionally be made in ways meetings (and 12 percent of proposals from men-only that do not fully reflect the preferences of single-sex meetings) were converted into implemented subproj- meetings. Some evidence supports this explanation. A ects during PRF’s first cycle. These percentages declined review (World Bank 2008a) of the effectiveness of CDD in the following three cycles.16 The most successful pro- programs in Lao PDR found that some district officials posals were those that both men and women had sup- decided which projects one PRF village would receive ported in their separate meetings. before the single-sex and VNPA meetings had occurred. The same report also noted an instance in which local Several possible explanations exist for this convergence; officials had selected projects already included in the unfortunately, there is no way to discern which expla- district’s development plan, while another report (PRF nation is the best. For example, convergence may have 2007) found that one-third of approved subprojects occurred because men and women shared similar pri- were not necessarily part of needs identified at VNPA orities and aspirations for their villages. The adults in a meetings but had been added at the koumban or district level. PRF’s revised procedures, noted above, attempt 16 Comparable data subsequent cycles are not available. to address this situation. District administrative offi- I m pa c t o f c o m m u n I t y- D r I v e n D e v e l o p m e n t p r o j e c t s o n G e n D e r 17 cials are no longer allowed to attend either the VNPA pate in public decision making external to PRF. In other or koumban-level meetings. words, are there spillover effects associated with PRF? Narratives and case studies from other CDDs indicate Potential indicators: that there may be such spillovers (e.g., Department of Indicator P4: Percentage of infrastructure Social Welfare and Development 2009), so a plausible subprojects solely and initially hypothesis is that women who have participated in PRF selected at each single-sex meeting subprojects are more likely to participate in decision at the village level and (a) forwarded to the koumban level; (b) selected at making external to PRF (World Bank 2008b). the koumban level; and, (c) approved for funding at the district level Potential indicators: P4: Validity ++ Reliability ++ Indicator P6: Percentage of female PRF Ease of collection ++ Cost of collection ++ participants who participate in non-PRF public decision-making Indicator P5: Percentage of infrastructure institutions as a result of their subprojects that both single-sex groups experience with PRF selected and that are (a) forwarded to the koumban level and (b) selected at P6: Validity – Reliability ++ the koumban level; and, (c) approved for Ease of collection ++ Cost of collection ++ funding at the district level Indicator P7: Percentage of female leaders in local P5: Validity ++ Reliability ++ non-PRF public decision-making Ease of collection ++ Cost of collection ++ institutions as a result of their experience with PRF Collecting data for these indicators should be rela- P7: Validity – Reliability ++ tively easy and will provide an important measure of Ease of collection ++ Cost of collection ++ women’s influence. PRF already records all subproj- ects that are nominated by the women’s and men’s Indicators P6 and P7 serve as proxies for the impact of groups at the VNPA meetings (and should continue PRF on women’s engagement and leadership in local to do so), so it should be able to track the outcomes decision making other than PRF. The indicators have of proposed subproject as they move through PRF’s reasonably acceptable validity only if one is willing to selection system. assume that participation in PRF’s processes encour- ages or facilitates participation and leadership in other Recommendation decision-making institutions. This is not an unreason- The indicators are highly recommended as standard able assumption.17 indicators for regular data collection and for report- ing on gender outcomes. The indicators will allow During the face-to-face interviews women were first PRF’s managers and other stakeholders to assess the asked about their participation (but not their leader- relative influence of men and women in PRF’s deci- ship) in “any other group making decisions for the peo- sion-making processes and provide lessons regard- ple in the village.� Respondents were then offered exam- ing opportunities to increase awareness of these ples to illustrate what was meant. Respondents were processes, as well as women’s potential to influence expected to answer “yes� or “no.� These simple choices future decisions. and the unlikely presence of preferred responses sug- gest that indicator P6 is acceptably reliable. Reliability In addition, PRF might also consider using the data gen- was also strengthened because respondents were next erated for these indicators to determine whether and asked to identify the other decision-making groups how men’s and women’s preferences for PRF’s interven- with which they had participated. Interviewers were tions might differ. The results may have value for the way PRF operates, as well as for the gender targeting of 18The possibility exists, of course, that some women participate in other public service providers in Lao PDR. non-PRF decision making without having participated in the PRF’s activities. In other instances, participation in some village-level meet- ings may predate the PRF’s presence. If these situations exist, the indi- Spillover Effects cator’s validity as a measure of the PRF’s influence would be compro- A critical indicator of PRF’s influence on women’s politi- mised—unless participation rates of PRF attendees were considerably cal engagement is the extent to which women partici- higher than those for non-PRF attendees. 18 I m pa c t o f c o m m u n I t y- D r I v e n D e v e l o p m e n t p r o j e c t s o n G e n D e r instructed to probe for several alternative “engagement� A third tested indicator related to women participating opportunities. in public decision making external to PRF includes: Indicator P8: Lessons learned by participating in As the pilot test found, the information needed for this PRF used in non-PRF decision making indicator can be collected at low cost in a few minutes P8: Validity + Reliability ++ using only two questions. Ease of collection – Cost of collection – The pilot test found a high level of participation at non- This qualitative indicator assesses the degree to which PRF meetings. Eighty-six percent of ethnic minority women participating in PRF’s processes are able to apply women had attended other decision-making forums, the skills acquired in PRF meetings in other decision- compared with 92 percent of the majority Lao-Tai making venues. women. Understandably, however, this participation cannot be attributed solely to experience with PRF. The pilot test used focus group discussions to exam- ine the extent to which women’s experiences with Table 6 shows the type of village-level meetings the PRF enhanced their participatory skills elsewhere. women attended. Village meetings were the most Participants in these discussions identified several common. The popularity of these meetings might beneficial effects, but the women also suggested reflect their inclusive nature. All villagers are eligi- that they had much to learn. They requested that ble to participate in such meetings, but also strongly training targeted at participants in PRF meetings be encouraged to participate. Similarly, the Lao Wom- increased. These findings were consistent among all en’s Union is intentionally inclusive. Other meetings participants in the focus group discussions, including might be more exclusive in nature (and may meet the ethnic minorities. less frequently), such as village water groups and school boards. Groups like local school boards are The disadvantage of the indicator is that it requires a intentionally small, may not encourage or facilitate high degree of participation from facilitators of the participation (or have special provisions to encour- focus groups and training of the facilitators. Processing age women’s active participation), and voting may be of the information is also time consuming. Nonetheless, limited to elected members. the pilot testing showed that it is possible to collect data for the indicator using local consultants experienced in the collection of qualitative data. Table 6 Type of other village-level meetings attended Recommendation Lao-Tai Ethnic Indicators P6 and P7 provide interesting information Type of meeting Women (%) Minorities (%) about women’s political engagement but not necessar- Village meeting 94.2 93.0 ily about PRF’s role in promoting this engagement or leadership. As the data in Table 6 indicate, participation Lao Women’s Union 55.1 44.2 beyond village meetings and the Lao Women’s Union is Youth Union 15.9 9.3 negligible more than five years after PRF’s first funding cycle. If participation is negligible, then levels of wom- School board 5.8 7.0 en’s leadership will be even lower. Equally important, in Village credit group 5.8 0.0 the absence of comparison with women not involved with PRF, it is not possible to attribute participation Village water group 0.0 2.3 or leadership in other village groups to participation Lao Front for Development and 1.4 2.3 in PRF.18 Indeed, the opposite situation may be just as Reconstruction likely. Participation in other village meetings may be Labor union 2.9 0.0 Other 5.8 2.3 18The challenge of attribution is further compounded by the fact that participants and nonparticipants are self-selected rather than ran- N= 69 43 domly assigned to one group or the other. Participants may be predis- posed to engage in public decision making without the need for any Note: Percentages sum to more than 100 because multiple responses were possible. experience or success in PRF’s processes. I m pa c t o f c o m m u n I t y- D r I v e n D e v e l o p m e n t p r o j e c t s o n G e n D e r 19 the explanation for women’s attendance at PRF meet- If PRF’s interventions do affect women’s and men’s atti- ings. For these reasons, PRF’s collection of data for these tudes, then it could benefit by assessing their attitudes indicators is not recommended unless it is possible to and subsequently changing its processes and imple- initiate an evaluation that compares levels of women’s mentation of its subprojects in ways that reflect these leadership in PRF villages with these levels in villages attitudes, while maintaining PRF’s progender approach that have not participated in PRF. This approach would to development. Unfortunately, measuring attitudes is represent an impact evaluation, which can be expensive always problematic, especially when those assessing the and methodologically complex. attitudes share a socially desired preference and when that preference may not be shared by the groups from With regard to indicator P8, rather than using focus which the information is collected. For this reason, the groups to collect information for this indicator, tar- indicator is not as valid as is desirable. In contrast, using geted interviews may be more appropriate as well as a simple survey instrument with standard questions less costly. Given the value of the information collected could provide acceptable levels of reliability and allow in the pilot, use of this indicator is recommended for collection of the needed data relatively easily and at a inclusion in thematic impact assessments of PRF, but moderate cost. not as part of its regular monitoring process. The assess- ments can be outsourced to a local consulting firm. Recommendation PRF should consider the periodic use of a short sur- Perceptions of Women’s Roles vey instrument with four or five closed-ended ques- in Decision Making tions related to perspectives on PRF’s gender-based If PRF influences and promotes women’s opportuni- rules.21 As an illustration, questions could assess (a) ties to affect decision making in their villages, then one how strongly men and women favor or oppose women’s can reasonably expect that women’s and men’s attitudes involvement in village-level decision making, (b) per- about gender equity will change as a result, although ceptions of PRF’s decision-making processes; and (c) not necessarily in preferred ways. Men have long been opinions about the value of subprojects that women accustomed to making key decisions in PRF’s villages propose in their single-sex meetings. (and continue to do so), but the increased influence of women may be seen as altering traditional and widely To enhance the value of the survey, it would first be accepted norms and mores about appropriate gender pretested to ensure the appropriateness of the ques- roles. If men perceive the changes to be undesirable or tions. In addition, when PRF adds villages and districts as diminishing their own importance, they may resist or to its agenda, relevant portions of the village survey undermine PRF’s progender approach to development. could also be used to gather baseline data, thus permit- Conversely, men may also approve of the changes and ting subsequent assessment of the changes that have act to support them. occurred over time. Potential indicator: 19 This indicator was not included in the field test for the current ini- Indicator P9: Perceptions of women’s roles in tiative. Instead the issue captured in this indicator was included in the 2008 assessment of PRF (World Bank 2008a). PRF suggested the decision making19 desirability of including the indicator in the toolkit. P9: Validity + Reliability ++ 20 Closed-ended questions limit the choices respondents can choose, such as “yes� or “no� or “I strongly agree, I somewhat agree, I have no Ease of collection ++ Cost of collection + opinion, I somewhat disagree, I strongly disagree.� Woman harvesting. Sayaboury Province. Photo © Adam Rogers / UNCDF 21 Women’s Economic Well-being GENDER-BASED ISSUES IN THE 77 Has access to PRF’s infrastructure subprojects or ECONOMIC SPHERE training affected women’s ability to create small businesses? Agriculture and rearing of livestock are the main 77 Is increased participation by women in PRF pro- sources of rural peoples’ livelihoods and well-being in cesses affecting the number of women engaging Lao PDR. Agriculture is typically the most important in entrepreneurial activities? source of a rural community’s income. Poverty in the Lao context often results from a poor harvest or the How these questions can be answered is addressed in shortage or lack of rice and livestock. Secondary causes the sections that follow. of poverty include a lack of (or limited) arable land and water, agricultural pests, livestock diseases, sickness, Economic Capital and poor health. Other causes involve the lack of (or PRF’s subprojects can affect women’s overall economic limited) roads and accessibility to remote and moun- situation and opportunities as entrepreneurs. Economic tainous areas; lack of (or limited) clean drinking water conditions can expand as infrastructure improves access and poor sanitation, diseases such as malaria and diar- to economic, labor, production, and financial markets rhea; and limited or no access to health and education for isolated communities. Training can improve the facilities. Poverty is especially pronounced among rural quality of locally produced handicrafts and provide populations and ethnic minorities, notably women skills in marketing, business development, and count- (GRID 2005; World Bank 2006a). ing of profits and losses. PRF’s subprojects can create Poverty in Lao PDR thus has several dimensions. To a nearly unlimited agenda of economic opportunities. escape from this poverty women must have the oppor- Although assessing the impacts of PRF’s subprojects on tunity to do so, but at least two barriers constrain these opportunities: (1) the use of their time, and (2) the incomes, livelihoods, and opportunities for engaging in responsibilities they have as women. markets is challenging and perhaps better left for com- prehensive evaluations or national socioeconomic sur- veys, it may be possible to use simple quantitative indi- PRF IN THE ECONOMIC SPHERE cators to summarize the perceived impacts of economic PRF neither monitors women’s participation in eco- infrastructure on household incomes and expenditures. nomic activities related to PRF’s training or infrastruc- ture subprojects nor does it profile men’s and women’s Potential indicator: participation in labor or economic markets in PRF vil- Indicator E1: Perceived economic improvements lages. This is unfortunate because the omission pre- due to increased access to economic infrastructure cludes attention to several important outcome-related questions, including: E1: Validity ++ Reliability ++ Ease of collection ++ Cost of collection ++ 77 Are women gaining economic capital as a conse- quence of increased access to basic infrastructure? Poor people typically know when their economic situ- 77 Are women gaining time as a consequence of ation has improved, so asking them whether that situ- increased access to basic infrastructure? ation has improved provides suitably valid information 22 I m pa c t o f c o m m u n I t y- D r I v e n D e v e l o p m e n t p r o j e c t s o n G e n D e r (although not necessarily about the explanation for the presumed benefits far exceed the costs. Moreover, when changed situation). To promote reliability, significant incomes have increased faster than the cost of living, effort was made during the pilot test to ensure that the these costs may be less burdensome than they other- data collection on the household-level impacts of PRF’s wise would be. subprojects was based on a consistent method across communities. Enumerators were instructed to use Recommendation specific and standard examples from the respondents’ Indicator E1 should be included in the routine data everyday life and to probe for both negative and posi- collected as part of PRF’s monitoring. PRF staff can tive effects. use surveys to collect the information on household incomes and costs of living from a sample of villages The pilot test shed some light on the utilization of PRF and respondents each year. Collecting the information subprojects and their economic impact on households will require about ten minutes per individual. and communities. All respondents were asked whether their household had used (or accessed) PRF subprojects Once the information is collected, it can be entered into in their villages and all but a few did so. Those that did the MIS at the provincial level and then analyzed at the were then asked whether the subproject had an effect national level. The results can be reported in gender- on their income, including nonmonetary income (Table focused reports as well as in PRF’s periodic reports. The 7). Of those using or accessing the subprojects, most reports would ideally be linked to recent economic data reported a positive effect. at the village or district level. Women were next asked whether their village’s PRF At least one important concern exists about this indica- subproject had affected their household’s cost of liv- tor. Although respondents are likely to know whether ing. Among all respondents to this question, a majority their economic situation has improved, their ability to reported that their cost of living had either decreased attribute the change to PRF will be problematic if there or had no effect (see Table 8). The decreased cost of liv- are other development projects in their village or dis- ing was especially notable among ethnic minorities. In trict. These other projects may explain some or much contrast, nearly one in five women reported that the of the perceived change in the respondents’ economic PRF subproject had increased these costs. In developed situation. In the absence of valid comparisons with economies, with high standards of living, such increases villages in which PRF does not operate, the Fund will are rarely perceived favorably. In Lao PDR, however, an not be able to determine whether its subprojects are increased cost of living is not necessarily undesirable. responsible for any improvements. The ideal approach In some cases, costs may increase due to the expenses would compare perceptions in similar villages with and associated with attendance at a newly constructed, without PRF’s interventions. In the language of evalu- PRF-supported school or because previously unavail- ation, PRF villages would be in the “treatment group,� able medicines have been purchased after a PRF-funded bridge improved access to health care. In both cases, the TablE 8 Effects of PRF’s subprojects on household cost of living TablE 7 Effects of PRF’s subprojects on household income Lao-Tai Ethnic Women (%) Minorities (%) Lao-Tai Ethnic Women (%) Minorities (%) Increased cost of 24.2 10.9 living Increased income 80.3 63.0 Decreased cost of 32.9 63.0 Decreased income 0.0 4.3 living No effect 19.7 32.6 No effect 42.9 26.1 N= 71` 46 N= 70 46 Note: Includes only those respondents who indicated that their families Note: Includes only those respondents who indicated that their families had used or accessed a PRF subproject in their village, but excludes had used or accessed the PRF subproject in their village, but excludes responses coded as not applicable. responses coded as not applicable. I m pa c t o f c o m m u n I t y- D r I v e n D e v e l o p m e n t p r o j e c t s o n G e n D e r 23 and villages without PRF would be in a “control� or security (Lao National Commission for the Advance- “comparison� group. All else being equal, it would then be ment of Women 2005). Women spend about twice as possible to assess perceived changes in villagers’ economic much time collecting firewood and fetching water than situation and to identify how much of the change could be do men and more than four times the labor hours on attributed to PRF. housework (World Bank 2010). As UNIFEM (2008) has noted, although unpaid household work and child care Use of Time are essential, women’s reproductive and child-rearing Three in four people and five in six poor people live roles mean that women “will probably still do more in rural areas in Lao PDR (Ministry of Planning and family care than men . . . and be more likely to choose Investment 2010), and about two-thirds of the popula- part-time work in order to balance work and family.� tion work in the agricultural sector (National Statistics Center 2004). Paid employment outside the household Given all the tasks for which women have some respon- exists in large numbers primarily in Vientiane, but sibility, they have less time for sleep, leisure, and for is almost totally absent in rural areas. Without paid education than do men, at least according to the data employment, access to desired goods and services is available for Lao PDR.22 Women’s daily obligations often limited and occasionally even impossible. For this restrict their access to education, paid employment, reason, it is useful to examine women’s opportunities to participation in village decision making, and opportuni- engage in financially remunerative tasks. ties to increase their incomes (UNIFEM 2008). In short, time is an economic asset as well as a potential outcome Such opportunities are often related to the time avail- of PRF’s investment. able to pursue these tasks. “For poor women and girls,� UNIFEM (2008) has observed, “lack of time is perhaps A possible measure of such outcomes for PRF’s impact the most crippling form of poverty because it contrib- on women’s opportunities in the economic sphere utes to their lack of capability in almost all other dimen- could thus be measured by the amount of time released sions.� Without adequate time for themselves, women to women due to specific infrastructure subprojects, are deprived of opportunities for education, health ser- including wells, access roads, or walking paths (Buch- vices, and employment.21 have 2009). Such outcome monitoring would ideally be supported by evaluations that examine the use of Data from the 2007–08 Lao Expenditure and Consump- the released time, thus validating whether the time is tion Survey reveal that women spend about 0.3 hours merely transferred from one onerous task to another per day on weaving, sewing, and other handicrafts or whether it is applied to more economically produc- compared to 0.1 hours for men. Handicrafts contrib- tive uses. ute a significant proportion of cash income to house- holds in rural areas (GRID 2005). Women also spend Potential indicator: more time on their own businesses—0.7 hours per day Indicator E2: Increased availability of time for compared to 0.4 hours per day for men. Combined with women for nonhousehold tasks their handicraft work, this takes up much of women’s E2: Validity + Reliability ++ total income-generating activity hours. Ease of collection ++ Cost of collection ++ These numbers suggest that women’s opportunities to This indicator focuses on the impact on women’s increase their earnings are constrained. They have less household chores as a result of increased access to the time per day to devote to income-generating activities improved access to infrastructure that PRF provides. than do men and have more household and child-care Indicator E2 accurately and reliably captures an impor- obligations than do men. As an illustration, rural women tant measure of PRF’s economic effects, especially for in Lao PDR are responsible for their families’ food women and their opportunities to engage in economi- 21In an analysis of gender inequality, poverty, and human develop- ment in South East Asia, Francisco (2007) argued that women’s time 22 A livelihood case study (Alton and Rattanavong 2004) of minority spent on nonmarkets activities takes away time for them to partici- villages in Luang Nam Tha Province revealed differences in labor divi- pate in market activities, and concluded that gender dynamics at the sions between males and females and concluded that girls and women household level reinforces the economic, social, and political arrange- have more chores during the day than boys and men, both in terms of ments in the community and the society as a whole. numbers and time needed. 24 I m pa c t o f c o m m u n I t y- D r I v e n D e v e l o p m e n t p r o j e c t s o n G e n D e r cally productive activities. No less important, the data drinking water. This finding suggests the desirability of for this indicator are easily collected and require only a more nuanced questions about the use of women’s time few minutes. with respect to PRF’s subprojects. Some of PRF’s sub- projects may decrease the time women spend on house- In the pilot test, women who had confirmed their use hold chores while others can increase the time devoted of or access to the PRF subproject in their villages were to such chores. asked whether the subproject had changed the time they spend on daily household tasks; they were also Recommendation given specific examples, such as cooking or collecting The indicator is recommended as a suitable proxy for water or firewood. The results, shown in Table 9, are women’s opportunities to engage in economically pro- interesting and perhaps even counterintuitive. On the ductive activities. If a decision is made to collect data on one hand, one-third of the Lao-Tai women and nearly this indicator, several options are available. three of five ethnic minority women reported that they spent less time on household chores as a result of the First, PRF staff can use a questionnaire to collect the PRF subproject. data in a few minutes per respondent. Second, the data can be estimated by subproject implementation On the other hand, about one-third of all women groups and recorded in the first—and a subsequent or reported that the time spent on household chores had last—subproject implementation form (SPIM), which is increased despite the PRF subproject. To understand already in use. The information would be collected only why this situation occurred, the pilot test included sev- on selected subprojects, notably those where collection eral focus group discussions. These discussions indi- of data is simple and the positive impact on time use for cated that educational subprojects can affect the time women and men is assumed and relevant. Such subproj- devoted to household chores. When older children ects could include those that provide access to water attend school, the burden on their mothers can increase or other infrastructure, including roads, foot paths, and because they must assume responsibility for taking care bridges. Using this option would not add much time or of infants, gathering firewood, or other tasks formerly cost to the data-collection process. handled by their older children. The second option would permit comparison before Other research conducted in Lao PDR (World Bank and after the completion of a relevant subproject. For 2008a) has found that women experience significant example, in the first SPIM, before the subproject is time-use savings when infrastructure projects provide initiated, information would be recorded on the esti- mated time (e.g., minutes/round trip access) for villag- ers to access water or markets and then entered into PRF’s MIS. Comparable data would then be collected TablE 9 Effects of PRF’s infrastructure a few months after the subproject’s completion. This subprojects on the time spent on before-and-after comparison would represent a form household tasks of impact evaluation despite the absence of an explicit Lao-Tai Ethnic counterfactual or comparison group. When there is no Women (%) Minorities (%) other plausible explanation for an outcome, such as Increased time 37.3 33.3 decreased time to collect water, the counterfactual is spent on daily implicit (White 2009). household tasks Decreased time 32.8 57.8 A third option exists. Monitoring the effectiveness of spent on daily gender-based policies does not always require that household tasks experts be imported into a village. As UNIFEM (2008) recommends, data should be collected through meth- No effect 30.0 8.9 ods that are gender responsive and that reflect women’s N= 67 45 concerns. One means of doing so would involve village women in data collection as enumerators with the assis- Note: Includes only those respondents who indicated that their families had used or accessed the PRF subproject in their village, but excludes tance of PRF staff. As an example, after being trained, responses coded as not applicable. one or two village women could be hired for one day I m pa c t o f c o m m u n I t y- D r I v e n D e v e l o p m e n t p r o j e c t s o n G e n D e r 25 a month to observe and record the amount of time a sample of women spend collecting firewood, fetching water, or preparing meals. Hiring local women would represent an extension of PRF’s current practice of hir- ing villagers to assist with the construction of PRF infra- structure subprojects while also developing local capac- ity, one of PRF’s implicit objectives. This approach risks diminished reliability, but the trade-off would be easily justified. Nepal’s experience with female enumerators and supervisors in its 2001 census demonstrates the viability and success of such an approach (UNIFEM 2008). Before this option is con- sidered, however, it should be field tested to assess the feasibility of engaging women in this task. Finally, this indicator shares the same concern associ- ated with indicator E1, namely the issue of attribution. The solution is the same, namely the collection of data Female stall owner being interviewed in Hua in villages in which PRF does not operate. Meaung village, where PRF is active. This pilot M&E initiative found that information on women’s Engagement in Entrepreneurship entrepreneurship was relevant for assessing PRF’s subprojects encourage development through the PRF’s impact on women’s economic activity. provision of infrastructure, such as electricity or roads that improve access to markets. In addition, however, PRF also promotes development through the provision had grown). Only 12 percent of minority women did so of vocational training designed to encourage new eco- in comparison with 45.3 percent of Lao-Tai women. nomic activities in villages. Income-generating activi- ties received little attention from PRF in its early years, Despite these attributes, attributing changes in eco- but villagers repeatedly requested assistance with these nomic activity to PRF is problematic—unless there is activities, and PRF now provides that assistance. If this a comparison group. Unlike the previous indicator, in assistance is successful, then there should be an increase which there was no plausible alternative explanation in the number of women-operated businesses (or an for the reduced time required to complete household increase in the volume of existing businesses) in PRF tasks, the situation is different for operating a business. villages after the infrastructure was built or the train- There are many reasons why poor women in PRF vil- ing provided. A related indicator would assess the per- lages might start a small business, including economic ceived benefits that PRF’s subprojects create for women necessity independent of anything PRF might do. entrepreneurs. Recommendation Potential indicator: Information on women’s entrepreneurship is relevant Indicator E3: Percentage of women in PRF villages for assessing PRF’s impact on women’s economic activ- operating small businesses ity, although the number of questions about business E3: Validity ++ Reliability ++ activities could usefully be increased. Additional ques- Ease of collection ++ Cost of collection + tions might inquire about the kind of businesses in which women engage (or would like to engage), how their If PRF increases women’s opportunities to engage in incomes have changed as a result of their businesses, entrepreneurial activities, there should be an increase and whether they have hired anyone to help them with in the number of women operating small businesses in the business. In addition, because some women already PRF villages. To assess this assumption, respondents in operate small businesses, it would be useful to ascer- the pilot test were asked whether they “operate a busi- tain how their incomes may have increased as a result ness on a regular basis� (not including sales of crops they of PRF’s subprojects. 26 I m pa c t o f c o m m u n I t y- D r I v e n D e v e l o p m e n t p r o j e c t s o n G e n D e r As just noted, however, the information about business indicator, which focused on whether women operated a activity should be placed in context. If PRF chooses to business, indicator E4 focuses on PRF’s effects on exist- assess changes in women’s economic activities, infor- ing businesswomen. mation should also be collected on economic activity in villages in which PRF does not operate.23 Doing so Among the number of women operating businesses, would permit a reasonable assessment of PRF’s impacts there was virtually universal agreement among them on economic activity but will also increase the overall that PRF subprojects had several positive impacts. 24 cost of data collection. As an illustration, among respondents who said that PRF had an impact on their businesses, all agreed Collecting data in villages not involved with PRF will that their costs for the goods they sold had decreased, increase costs, but there are benefits to PRF doing so. while the access to these goods had increased as had One objective of the Bank’s collaboration with PRF is the sales of these goods. Almost all of these respon- to strengthen the Fund’s capabilities to monitor and dents also agreed that the number of their customers evaluate. Comparing outcomes in villages with and had increased as a result of PRF’s subproject in their without PRF’s subprojects provides an opportunity for village. the Fund’s M&E staff to learn about and implement impact evaluations. Such evaluations are typically the These results suggest that these direct indicators of most effective means to identify and assess outcomes positive benefits are both valid and reliable. In turn, that can be reliably attributed to a project’s interven- the information about these benefits is easily collected tion. Familiarity with the methods of impact evaluation by asking the women about the effects of PRF subproj- is highly desirable at a time when attention to results- ects on their business operations. Attribution remains based management is increasingly important. a concern because the costs of goods acquired and sold (as well as the number of customers) can be due to many Potential indicator: factors, only one of which is PRF. PRF is likely to be a Indicator E4: Benefits for women’s plausible explanation for the positive impacts identified, entrepreneurship but not necessarily the sole or even primary explanation for these impacts. Assessing the relative role of PRF in E4: Validity ++ Reliability ++ Ease of collection ++ Cost of collection + creating positive benefits would require comparison of the situation with and without PRF but would provide still another opportunity to consider and implement an The impact of PRF’s subprojects on female entrepre- impact evaluation. neurs is important for understanding the Fund’s effects on poverty reduction and the economic opportunities Recommendation created for women. Is it the case that these subproj- Indicator E4 provides relevant information about the ects have positive impacts on women’s entrepreneur- economic benefits of PRF. Although information for ial activities? To answer this question, respondents in the indicator probably should not be collected rou- the pilot test who said they operated a business were tinely, such as for monitoring PRF’s implementation, the asked whether and how PRF subproject in their villages indicator can be of considerable value in demonstrat- had affected their business. In contrast to the previous ing PRF’s overall value, assuming that future results are similar to those noted above. 23Under ideal circumstances, villages would be assigned randomly to treatment and comparison groups. The former group would be the Nonetheless, one change in the information collected PRF villages; villages in the comparison group would not receive any might be desirable. PRF seeks to improve women’s eco- of the PRF’s interventions. If the number of villages in each of the two groups was sufficiently large, random assignment would produce nomic situation, but there is also good reason to collect two groups that would be nearly identical in terms of the charac- comparable data from men. PRF might provide benefits teristics deemed to be potential explanations for levels of economic for women, but it may also be the case that more of activity. Given that PRF is several years old and that villages in PRF have certain characteristics (such as high levels of poverty and high percentages of ethnic minorities), random assignment is not likely 24If respondents said that the PRF project in their village had not af- to be possible. As an alternative, a comparison group could include fected their business, it made no sense to ask them how the subproj- nonparticipating villages that share with the participating villages the ect had affected their business. As a consequence, a small portion of key characteristics believed to be related to different kinds of activity, female entrepreneurs in the PRF villages did not identify any positive including economic, political, and social. effects of the PRF subproject in their villages. I m pa c t o f c o m m u n I t y- D r I v e n D e v e l o p m e n t p r o j e c t s o n G e n D e r 27 these benefits accrue to males (and for a larger num- to establish businesses can be of value to PRF. PRF has ber of males) than for females. If that situation occurs, an implicit theory of change, but this theory should be then women’s economic situation will have improved, informed and amended based on information gained but their economic status or well being relative to men from the intended beneficiaries. If data are to be col- will have declined.25 A 10 percent increase for females lected about the support women need to start a small might be noteworthy, but it becomes much less so business, the data-collection process should be simpli- if males have experienced a 25 percent increase (or if fied, perhaps by including consideration of this issue in males receive 75 percent of a subproject’s benefits and one of PRF’s periodic thematic studies. The necessary females the remainder). In sum, determining changes in data would be collected only once and used to inform females’ relative well-being will not be possible without PRF’s decision making. data from both males and females. Food Security Potential indicator: Food insecurity is a persistent characteristic of poverty. Indicator E5: Support for women’s entrepreneurship Almost by definition, poor people do not have enough to eat. When there is not enough to eat, women are typi- E5: Validity - Reliability - cally the most likely to suffer. In poor Lao households Ease of collection - Cost of collection - women are largely responsible for providing and prepar- ing food for their families. As UNIFEM (2008) explains, Support for women’s entrepreneurship is a qualitative when women have to increase their effort to produce indicator. To assess this support, the pilot test included more food or work to earn money to purchase food, the interviews and focus group discussions with women in household chores of their daughters often increase. The several PRF villages. One of the topics in these discus- result is that girls may be kept out of school, and the sions was the kind of support that women might need mother’s poverty is perpetuated though her daughters’ to create new businesses. lack of education. Among the women in the discussions, many looked This situation surely exists in Lao PDR. According to the favorably on PRF as a potential driver of increased most recent Lao Expenditure and Consumption Survey female entrepreneurship, and many recommended that (LECS) (World Bank 2010), over 25 percent of the coun- PRF’s training programs could be one means to support try’s rural population suffered from food poverty in 2007– entrepreneurship. Some women also suggested that 08. Nearly 40 percent of the population in rural areas existing entrepreneurs could be instructors in these without access to roads found themselves in this situation. programs. Moreover, families that are not able to grow or buy enough rice will also find themselves unable to afford the costs of As might be expected, the requirements for support to women’s entrepreneurship vary. Likewise, what other necessities, such as education for their children or constitutes entrepreneurship can differ from one vil- medical care and medicines for their ill members. lage to another, with the consequence that the con- Potential indicator: cept may be too abstract for some women to under- stand. Indeed, due to the concept’s complexity, the Indicator E6: Number of months in the past year a indicator is difficult to measure consistently and its household did not have sufficient rice26 validity is problematic. Unfortunately as well, data E5: Validity + Reliability + collection is work-intensive; it relies on identifying Ease of collection ++ Cost of collection ++ existing or potential female entrepreneurs and then lengthy interviews or focus group discussions with Rice is a staple crop in Lao PDR and among the most skilled facilitators. commonly grown and eaten foods. As a result, access to rice can serve as a suitable proxy for food security as well Recommendation as changes in economic security. No less important, Lao Despite the challenges associated with this indicator, PDR’s periodic expenditure and consumption surveys gathering data about the factors that motivate women (LECS) assess rice insufficiency by asking village chiefs The issue of the PRF’s relative benefits for males and females applies 25 not only to indicator E4, but to several other indicators as well. 26 This indicator was not included in the field test. 28 I m pa c t o f c o m m u n I t y- D r I v e n D e v e l o p m e n t p r o j e c t s o n G e n D e r about rice insufficiency. Should PRF decide to collect Scenery from information on rice insufficiency in the poor villages in PRF village in which it works, the data collected can be compared with Huaphanh province. information from the LECS. Photo © Anders Engvall/World Bank Recommendation Data on rice insufficiency can be collected routinely from women by asking a single question at the single- sex meetings (prior to the VNPA meetings in each vil- lage). To promote comparability with the LECS data, village chiefs can also be asked the same question. In addition to inquiring about food security, it would also be desirable to gather information about the nutri- tional and caloric value of what villagers consume. Unfortunately, there is no inexpensive and convenient way to do so. As an illustration, the nutrition portion of the 2007–08 LECS involved twelve questions about food consumption. 29 Women’s Access to Public Services GENDER-BASED ISSUES IN to travel to that hospital exceeded five hours (Ministry ACCESS TO PUBLIC SERVICES of Planning and Investment 2010). Access to a phar- macy is severely constrained for rural populations, A healthy and educated population is a fundamental even for those who can use all-year roads. For those requirement for development, and countries invari- without these roads, access is almost impossible. ably include attention to these areas in their national development strategies. Lao PDR is no exception. Access to education is equally important. Low levels Its National Socio-Economic Development Plan for of literacy are widespread, especially among ethnic 2006–10 emphasizes the importance of improving minorities and in rural areas without roads. Accord- the country’s education and health systems and, in ing to the LECS conducted in 2002–03, in 28 percent particular, women’s access to these basic services. of cases where children had never attended school in rural areas the primary reason was the schools’ Lao PDR has made remarkable progress in reducing distance from the children’s homes (King and van de infant mortality over the past decade as well as that for Walle 2007). Urban children are more likely to be in mothers and children under age five. Similar progress school than are rural children, and boys in rural areas is also evident in the fight against malaria and tuber- culosis (World Bank 2006a). Despite this progress, are more likely to be in school than girls, in particu- many Laotians continue to suffer from poor health lar non-Lao-Tai girls. Similarly, nonpoor children are and inadequate access to medical care. Women and more likely to be in school than are poor children. their newborn children are notably disadvantaged. There are notable differences among the country’s major ethnic groups. As an illustration, 90 percent According to the World Health Organization (2010), of Lao-Tai girls were enrolled in primary schools Lao PDR had a maternal mortality ratio of 580 in in 2007–08, compared with less than 70 percent of 2008, a rate far higher than in Thailand (48), Viet- Mon-Khmer and Sino-Tibetan girls. Similar dispari- nam (56), or Cambodia (290). Only 20 percent of Lao ties exist for enrollment in lower and upper second- women had their deliveries of children assisted by ary schools. skilled health staff in 2008 compared with 97 percent in Thailand and 88 percent in Vietnam. In villages Finally, access to clean water and proper sanitary without roads, only 7 percent of women gave birth facilities are inextricably linked to good health. With- in a hospital in 2007–08 (Ministry of Planning and out access to clean water, chronic and debilitating ill- Investment 2010). nesses are prevalent and often life threatening. Diar- rhea is rampant in areas without clean water. With- As with access to education in Lao PDR, there are con- out proper sanitary facilities, contamination of water siderable differences between rural and urban areas supplies is possible. In both instances, females are the in terms of access to health services. Whereas 81 per- most common victims. When women are responsible cent of the country’s urban population has access to for gathering water, their exposure to bacterial infec- medical staff only 28 percent of the rural population tions and unsafe water is increased. Girls may be with all-year roads has such access. In 2007–08, the reluctant to stay in school because toilet and wash- average village without a road was located more than ing facilities may not be available and when they are 25 km from the closest hospital, and the average time available they may not be private. Investments in 30 I m pa c t o f c o m m u n I t y- D r I v e n D e v e l o p m e n t p r o j e c t s o n G e n D e r sanitary facilities can provide enormous economic often considered to be an outcome.28 This is especially benefits. Research reported by UNICEF (2007) found true when these services were not previously available, that for every dollar invested in sanitation, up to $34 as is the case in many of Lao PDR’s villages that benefit more in health, education, and social and economic from PRF. development costs can be saved. Potential indicator: As Lao PDR recognizes, it is among the countries that Indicator S1: Number of beneficiaries and would benefit by increased attention to clean water and percentage that are female who have proper sanitary facilities. As much as 25 percent of Lao’s gained (improved) access to (a) health population does not have access to safe water in the dry services (b) education; (c) clean water; and, (d) appropriate sanitary facilities. season (Ministry of Planning and Investment 2010), and this percentage increases in rural areas without access S1: Validity ++ Reliability ++ to roads. Similarly, less than half of Lao PDR’s poor pop- Ease of collection ++ Cost of collection ++ ulation has access to a modern toilet, a squat toilet, or a dry toilet. PRF routinely collects data that identifies the infrastruc- ture—including social infrastructure—provided in its villages. The number of women and men in these vil- PRF GENERATED ACCESS TO lages is also available in PRF’s MIS. PUBLIC SERVICES One of PRF’s objectives is to increase access to pub- Recommendation lic services, including health and education. Between PRF is encouraged to count the number of beneficiaries 2003 and 2010, about one-quarter of PRF’s subproj- by gender in each village for each of the four service ects were related to education and almost 26 percent areas and to include the results in its overall summary to health. PRF has thus increased access to both edu- reports on subproject implementation rates.29 Access to cational and health resources. Nonetheless, measur- public services is an ideal indicator for annual reports ing access alone is insufficient if there are not also and can provide compelling evidence of PRF’s contri- benefits associated with PRF’s investments. In view of butions to poverty reduction and to Lao PDR’s overall the gender gaps noted above, a relevant gender out- gender strategy. Once data on access are collected it come for PRF would be a sex-disaggregated measure will also be possible to assess the relationship between for beneficiaries of the subprojects within the two access to the services and perceived economic benefits, social categories. such as on household incomes (indicator E1). Equally important, knowledge of this relationship can help PRF For health care there are additional considerations. Tra- to learn which subprojects are most and least effective ditional perceptions, language barriers, and cost con- in reducing poverty. siderations are potential barriers influencing health- seeking behaviors (World Bank 2006a). From a gender Attention to these issues can provide still another oppor- perspective, the degree to which women are able to tunity to enhance PRF’s M&E capabilities in terms of the influence the choice of seeking health services might selection of evaluation designs, data collection, and data also influence utilization. Given the focus on integrat- analysis. A primary purpose of evaluation is to promote ing women into PRF’s decision-making processes and learning and improve decision making. Identifying the women’s potential opportunity to influence the provi- relative effectiveness of different PRF investments can sion of health services to their communities, it is like- contribute to both. wise important to assess whether women also have an increased voice over the use of the health services that PRF provides. 28 As an illustration, the African Development Bank’s (2009) core out- come indicators for water and sanitation include access to improved Access to Public Services drinking water sources and improved sanitation facilities resulting Increasing the supply of public services is often defined from a project intervention (number), of which are female (percent- as an output. Nonetheless, access to water, sanitation, age). Health services are defined to include access to a nurse, a health clinic or center, or a dispensary where medicines can be obtained. education, and health services has such a strong rela- 29 PRF already collects information on the number of villagers in each tionship with improved well-being that such access is village in which it operates. I m pa c t o f c o m m u n I t y- D r I v e n D e v e l o p m e n t p r o j e c t s o n G e n D e r 31 Influencing Access to Health Services and Education Potential indicator: Indicator S2: Women’s influence on access to health services S2: Validity ++ Reliability ++ Ease of collection + Cost of collection - When to seek medical care and who to see about this care are typically important personal choices for women, at least in many developed countries. Elsewhere, how- ever, these decisions are subject to the availability and anticipated costs of the care and, often, the preferences of other family members, especially husbands. Such is the case in Lao PDR, as shown in Table 10. Respondents in the pilot test were first asked whether they had any health problems during the previous twelve months. Among Lao-Tai women, three-quarters reported having had a health problem compared with 92 percent of the ethnic minorities. These women were then asked whether they had sought help for that prob- lem and who had made the decision to seek or not seek help for that problem. The results indicate that hus- bands were overwhelmingly the primary decision mak- ers about their spouses’ health care.30 The need to con- sult other family members, especially husbands, before seeking health care may be due to the expenses asso- ciated with this care. Nationwide data from Lao PDR indicate that the cost of health services affects decisions 30 Among all 125 respondents, 123 were married and 2 were widows, so there were few decisions required about the health care of unmar- ried women. TAble 10 Persons making decisions to seek health care for women Lao-Tai Ethnic Women (%) Minorities (%) Respondent 40.0 21.7 The husband 78.2 91.3 Other family 23.6 13.0 member Nonfamily member 1.8 0.0 Portrait of villager in Khammoune province. N= 55 46 Photo © Meriem Gray / World Bank Note: Includes only those respondents who had a health problem in the previous twelve months and who had sought help for that problem. Respondents could select more than one option. 32 I m pa c t o f c o m m u n I t y- D r I v e n D e v e l o p m e n t p r o j e c t s o n G e n D e r Unlike access to health services, which some believe should be a personal decision, the same cannot be said about children’s access to education. Decisions about whether to enroll children in school are typically paren- tal, and this is the case in Lao PDR. Respondents were asked to indicate who made the decision to send their children to school. Over 80 percent of all respondents who had children in school reported that both parents had been involved in the decision making, and this was true regardless of whether the respondent was a mem- ber of a minority ethnic group or the majority Lao-Tai population. Recommendation If PRF’s objectives include the promotion of gender equity and increasing women’s opportunities, then Children in resettlement PRF may wish to consider how access to health services village, Nakai Plateau. can be changed. The construction of additional health Khammoune province. Photo © Meriem Gray / clinics is one way of doing so. More than thirty health World Bank dispensaries were constructed and staffed by trained nurses during PRF’s first four years.31 The construc- tion of additional health facilities is clearly desirable, but there is only a weak causal connection between about whether to use these services (GRID 2005; World construction and who makes decisions about accessing Bank 2006a). these facilities. There is little reason to believe the results related to PRF should continue to monitor and report data on indicator S2 are not valid or reliable. On the one hand, access to the health facilities it constructs. Information information about access to health services is already on who makes decisions about accessing these facilities part of PRF’s data-collection processes. On the other might best be left to thematic assessments or periodic hand, information about health-seeking behaviors and but infrequent surveys. Information on the indicator decision making is best collected through interviews, can be reported in anvnual reports as well as those on which can be time consuming. Gathering this informa- gender and participation. tion in the pilot test required about fifteen minutes per respondent. A choice thus needs to be made whether With regard to Indicator S3: Given what appears to be the value of the information about health-seeking appropriate roles for both parents in decisions about behaviors justifies the cost of its collection. enrolling their children in schools, there is no rea- son or rationale for PRF to become involved with this Potential indicator: process. Indicator S3: Women’s influence on access to education S3: Validity ++ Reliability ++ 31 None of the villages included in the pilot test had benefitted from Ease of collection + Cost of collection + health subprojects. 33 Recommendations for Next Steps AMPLE REASONS AND tified though the indicators now in use? What are OPPORTUNTINIES FOR appropriate indicators of success in the Lao context? MEASURING PRF’S IMPACT ON This toolkit has addressed these questions and identi- GENDER fied several indicators of outcomes that can be added easily and efficiently to PRF’s ongoing approaches to Good intentions are never enough, especially with monitoring and evaluation. In most instances as well, regard to development. The same is true for efforts the indicators suggested in the toolkit are compatible to promote gender equality. Few people oppose such with PRF’s management information system. equality, and there is no shortage of documenta- tion espousing its virtues or desirability for poverty The indicators considered in this toolkit are not reduction. Nevertheless, much remains to be done meant to be exhaustive. Other gender-relevant indi- before gender equality is achieved, as the existence cators can be considered or continued depending on of the Poverty Reduction Fund acknowledges. PRF is PRF’s needs and changes in its implementation of an admirable effort to promote gender equality, and the Fund. Gender-balance among PRF’s employees Lao PDR’s commitment to this equality should be at all levels offers an example of an indicator that applauded. The Fund’s design and implementation PRF already tracks and should continue tracking. are well-suited to this commitment. PRF’s mandate— At the end of 2006, 74 percent of PRF’s 140 employ- that women play a central role in decision making ees were male. This percentage did not change in about the allocation of the Fund’s resources—cre- the years that followed. In early 2010, an identical ates essential opportunities that otherwise would not percentage of PRF’s employees at its headquarters be available—and for too many poor and disenfran- in Vientiane were male. This percentage was even chised women in Lao PDR have never been available. higher at the district level, where the employees are more likely to have contact with the villages in PRF. Creating opportunities for women is merely an ini- Among the seven provinces in which PRF operated tial step on their well-deserved path to equality. As in 2010, the percentage of female employees at the the country’s National Socio-Economic Development district level ranged from 0 to 38 percent. PRF has Plan recognizes, Lao PDR cannot realize its “goals of usefully extended its tracking of gender balance to reducing poverty and improving national education, include the percentage of koumban facilitators who health, and population indicators without the active are female (PRF 2010), but it may also wish to con- participation of all women, and particularly poor and sider providing similar information on the percent- ethnic minority women.� age of its district and provincial coordinators who are female. A step of at least equal importance is the identifica- tion and measurement of the results, especially out- PRF can choose from among the indicators to comes, of PRF’s gender-focused efforts. What are strengthen its gender focus and its ability to generate PRF’s benefits for women, and how can its advocates compelling evidence of PRF’s contributions to gender demonstrate its success? Are there gender-related equality. These gender indicators can also be useful political, economic, or social successes within PRF for capturing the impact on women from other CDD that have occurred, but are not well-captured or iden- projects in Lao PDR and elsewhere in Asia. 34 I m pa c t o f c o m m u n I t y- D r I v e n D e v e l o p m e n t p r o j e c t s o n G e n D e r NEXT STEPS donors, and key stakeholders. Table 11 summarizes all the indicators and recommends the ones that PRF The indicators discussed in this toolkit provide PRF might consider for priority attention. with multiple opportunities to demonstrate further its commitment to gender equality. The following steps are thus recommended actions for PRF to enhance its 2. D eveloping and implementing routine and cost- focus on gender quality and development: efficient means—or amending existing approaches— to collect and analyze the data associated with each of 1. Agreeing on gender-related outcomes to monitor the selected indicators. This toolkit provides some sug- and evaluate by reviewing the indicators and selecting gestions for how data can be collected (see discussion of those best-suited for the needs of PRF’s management, each individual indicator in prior chapters). TAble 11 Summary of gender-based indicators Indicator Evaluative Criteria Recommendations Ease Cost Should For use in Frequency of data of data the PRF monitoring or of data Validity Reliability collection collection use? Priority evaluation? collection P1 % of women at VNPA meetings ++ ++ + + Yes Medium Monitoring Annually at who speak or vote a sample of VNPA meetings P2 % of women at VNPA meetings ++ ++ + + Yes Medium Monitoring Annually at who are active participants a sample of VNPA meetings P3 % of women who are aware of + ++ ++ ++ Yes Low Monitoring Annually at PRF’s decision-making processes a sample of VNPA meetings P4 % of subprojects solely and ++ ++ ++ ++ Yes High Monitoring Annually initially selected at each single- and sex meeting at the village level evaluation and (a) forwarded to the koumban level; (b) selected at the koumban level; and, (c) approved for funding at the district level P5 % of subprojects that both single- ++ ++ ++ ++ Yes High Monitoring Annually sex groups selected and (a) and forwarded to the koumban level; evaluation (b) selected at the koumban level; and, (c) approved for funding at the district level P6 % of female PRF participants – ++ ++ ++ No --- --- --- who participate in non-PRF public decision-making institutions as a result of their experience with PRF P7 % of female leaders in local – ++ ++ ++ No --- --- --- non-PRF public decision-making institutions as a result of their experience with PRF I m pa c t o f c o m m u n I t y- D r I v e n D e v e l o p m e n t p r o j e c t s o n G e n D e r 35 In general, there are several ways in which PRF can address Second, PRF should consider the power and value of ran- cost efficiency and improve the practical value of the data dom sampling to collect individual or household-level it does collect. First, in its collection of gender-based data, data in its targeted villages. Depending on the margin PRF should consider using relevant questions already of sampling error that PRF can tolerate and the confi- pretested and used in the LECS. The questionnaire used dence level desired, data from a sample of less than 400 in the pilot test relied on several demographic questions people or households can be generalized to all people or included in the LECS, but there are other opportunities to all households in all of PRF’s villages. In other words, a do so with other items as well. Using items from the LECS random sample in which every person or household in the will allow PRF to place these findings into context and to population has an equal probability of being included in compare them with provincial and national-level data. the sample can produce results that are representative of TAble 11 Summary of gender-based indicators (continued) Indicator Evaluative Criteria Recommendations Ease Cost Should For use in Frequency of data of data the PRF monitoring or of data Validity Reliability collection collection use? Priority evaluation? collection P8 Lessons learned by participating + ++ – – Yes Low Evaluation End of in PRF used in non-PRF decision program making P9 Perceptions of women’s roles in + ++ ++ + Yes Medium Evaluation Baseline decision making and end of program E1 Perceived economic ++ ++ ++ ++ Yes High Monitoring Annually improvements due to increased and access to economic infrastructure evaluation E2 Increased availability of time for + ++ ++ ++ Yes Medium Monitoring Annually women for nonhousehold tasks and evaluation E3 % of women in PRF villages ++ ++ ++ + Yes Medium Monitoring Annually operating small businesses E4 Benefits for women’s ++ ++ ++ + Yes Low Evaluation End of entrepreneurship program E5 Support for women’s – – – – No --- --- --- entrepreneurship E6 Number of months in the past + + ++ ++ Yes High Monitoring Baseline year a household did not have and sufficient rice annually S1 Number of beneficiaries and ++ ++ ++ ++ Yes High Monitoring Baseline percentage that are female who and have gained (improved) access annually to (a) education; (b) health services; (c) clean water; and, (d) appropriate sanitation facilities S2 Women’s influence on access to ++ ++ + – No --- --- --- health care S3 Women’s influence on access to ++ ++ + + No --- --- --- education 36 I m pa c t o f c o m m u n I t y- D r I v e n D e v e l o p m e n t p r o j e c t s o n G e n D e r the entire population with which PRF works, regardless ing whether a program or intervention has been suc- of the size of that population. Annex 3 provides addi- cessful. Target indicators, in contrast, are especially tional information on random sampling, including a brief useful for purposes of evaluation. They establish explanation of sampling errors and confidence levels. benchmarks or standards against which progress and accomplishments can be measured and judged. With- 3. Where appropriate, establishing baselines for the out target indicators it is not possible to determine new indicators in (a) all prospective sites where PRF whether a program, such as PRF, has achieved its goals, will work as well as in (b) the sites where PRF already other than anecdotally. operates.32 To address the recurring issue of attribu- tion, PRF should also consider the collection of appro- Table 13 displays PRF’s gender-related goal, and some of priate data, including baseline data, in villages in which the indicators discussed in this toolkit. More important, it does not operate. Doing so will not only strengthen the table also offers some illustrative target indicators PRF’s ability to claim responsibility for any improve- that PRF could use to judge its success. ments that occur, but it will also create opportunities to employ rigorous evaluation designs, including those 5. Monitoring for results in a way that identifies associated with impact evaluations. benefits and improvements in women’s well-being rather than merely counting outputs, such as the num- 4. Selecting targets for the outcome indicators that ber of people trained and the number of subprojects specify the magnitude of the changes desired and the completed. time period in which they should be achieved. The Mil- lennium Development Goals provide useful illustra- 6. Improving PRF’s capacity to collect and ana- tions of both of these desired characteristics. The goals lyze data to improve understanding of what kinds or are explicit about what should be changed or achieved types of infrastructure subprojects are most effective and by when. Table 12 provides some examples of how in promoting economic development and women’s these goals have been put into operation. economic opportunities. Enhancing this capacity might start with one or more workshops on evalua- Performance indicators, which do not specify what is tion designs, impact evaluations, sampling, and data to be achieved, are commonly monitored. By them- analysis, including bivariate and multivariate statis- selves performance indicators are not useful for judg- tics, which involve the simultaneous analysis of two or more variables. 32The PRF is also strongly encouraged to consider collecting comparable data in randomly selected villages that are not involved with the project. 7. Integrating key monitoring results into evalua- Doing so will create opportunities for rigorous impact evaluations, which tions that provide information on the benefits provided are especially useful in addressing attribution as well as the amount or and achieved. degree of change that can be ascribed to the PRF (as opposed to other possible explanations for changes that are observed). The World Bank’s Development IMpact Evaluation initiative provides useful guidance 8. Reporting and using findings for purposes of learn- about how to design and implement impact evaluations. ing, accountability, and improved decision making. TAble 12 Target and performance indicators for selected Millennium Development Goals Goal Target indicator Performance indicator Promote gender equality and Eliminate gender disparity in Ratios of girls to boys in primary, empower women primary and secondary education, secondary, and tertiary education preferably by 2005, and in all levels of education no later than 2015 Reduce child mortality Reduce by two-thirds, between 1990 Under-five mortality rate and 2015, the under-five mortality rate Improve maternal health Reduce by three quarters, between Maternal mortality ratio 1990 and 2015, the maternal mortality ratio I m pa c t o f c o m m u n I t y- D r I v e n D e v e l o p m e n t p r o j e c t s o n G e n D e r 37 TAble 13 Illustrative target and performance indicators for PRF Goal Illustrative target indicators Performance indicators Strengthen local institutions to 25 percent of infrastructure P4: Percent of subprojects solely support participatory decision- subprojects selected at women’s and initially selected at each single- making . . . at the village, koumban, single-sex meetings approved for sex meeting at the village level and and district levels, involving a broad funding at the district level by 2012 (a) forwarded to the koumban level; range of villagers, including women (b) selected at the koumban level; and the poor and, (c) approved for funding at the district level 75 percent of women in PRF villages E1: Perceived economic report economic improvements due improvements due to increased to increased access to economic access to economic infrastructure infrastructure by 2014 90 percent of women in need of S1: Number of beneficiaries and health services due to illness, disease, percentage that are female who have or child delivery can access these gained improved access to health services within two hours by 2015 services 39 References African Development Bank. 2009. “Guidelines for Kusakabe, Kyoko. 2005. “Gender Mainstreaming in the Use of Core Sector Indicators.� Tunis: African Government Offices in Thailand, Cambodia and Development Bank, Quality Assurance and Results Laos: Perspectives from Below.� Gender and Devel- Department. opment 13(2): 46–56. Alton, Charles and Houmphanh Rattanavong. 2004. Ser- Kusek, Jody Zall, and Ray C. Rist. 2001. “Building a vice Delivery and Resettlement: Options for Develop- Performance-Based M&E System: The Challenges ment Planning. Vientiane: UN Development Program. Facing Developing Countries.� Evaluation Journal of Buchhave, Helle. 2009. “Gender Data and M&E Prac- Australasia. 1(2):14–23. tice. An Inventory Report for PRF.� Unpublished Kusek, Jody Zall, and Ray C. Rist. 2004. Ten Steps to a report for the World Bank. Results-Based Monitoring and Evaluation System: Chattopadhyay, Raghabendra and Esther Duflo. 2004a. A Handbook for Development Practitioners. Wash- “The Impact of Reservation in the Panchayati Raj: Evi- ington DC: World Bank, Global HIV/AIDS Program dence from a Nationwide Randomized Experiment.� and Operations Evaluation Department. Economic and Political Weekly 39(9): 979–86. Lao National Commission for the Advancement of Chattopadhyay, Raghabendra and Esther Duflo. 2004b. Women. 2005. Lao PDR: Experiences in Implement- “Women as Policy Makers: Evidence from a Ran- ing the Beijing Platform for Action and the Outcomes domized Policy Experiment in India.� Econometrica of the Twenty-Third Special Session of the United 72(5): 1409–43. Nations General Assembly. Vientiane: Lao National Department of Social Welfare and Development. 2009. Commission for the Advancement of Women and Trials and Triumphs: Communities Fighting Poverty the United Nations Population Fund. Through KALAHI-CIDSS. Quezon City: Depart- Lao National Commission for the Advancement of ment of Social Welfare and Development. Women. 2009. Concluding Observations of CEDAW Francisco, Josefa S. 2007. “Gender Inequality, Poverty Committee: Towards the Combined 6th and 7th Peri- and Human Development in South East Asia.� Devel- odic Report of the Lao PDR. Vientiane: Lao National opment 2007, 50(2): 103–114. Commission for the Advancement of Women. GoL. 2006. National Socio-Economic Development Plan Mansuri, Ghazala and Vijayendra Rao. 2003. Evaluating (2006-2010). Vientiane: Committee for Planning and Community-based and Community-Driven Develop- Investment. ment: A Critical Review of the Evidence. Washington, GRID. 2005. Gender Profile Lao PDR. Vientiane: World DC: World Bank. Bank. Ministry of Planning and Investment. 2010. Poverty in Iarossi, Giuseppe. 2006. The Power of Survey Design: Lao PDR 2008. Vientiane: Ministry of Planning and A User’s Guide for Managing Surveys, Interpreting Investment. Results, and Influencing Respondents. Washington, Narayan, Deepa, ed. 2005. Measuring Empowerment DC: World Bank. Cross-Disciplinary Perspectives. Washington, DC: King, Elisabeth, and Dominique van de Walle. 2007. World Bank. “Schooling, Poverty, and Disadvantage in the Lao National Statistics Center. 2004. Lao Expenditure and People’s Democratic Republic.� Paper presented at Consumption Survey 2002-2003 (LECS 3). Vientiane: the annual meetings of the Population Association National Statistics Center and Ministry of Planning of America, New York City, March 2007. and Investments. 40 I m pa c t o f c o m m u n I t y- D r I v e n D e v e l o p m e n t p r o j e c t s o n G e n D e r National Statistics Center. 2005. Population Census. World Bank. 2000. World Development Report 2000/2001: Vientiane: National Statistics Center and Ministry of Attacking Poverty. New York and Washington, DC: Planning and Investments. Oxford University Press and the World Bank. National Statistics Center. 2008. Socio-Economic Atlas World Bank. 2001. Engendering Development: Through of the Lao PDR. Vientiane: Swiss Agency for Devel- Gender Equality in Rights, Resources, and Voice. New opment and Cooperation and National Statistics York and Washington, DC: Oxford University Press Center. and the World Bank. Petesch, Patti, Catalina Smulovitz, and Michael Walton. World Bank. 2005a. The Effectiveness of World Bank 2005. “Evaluating Empowerment: A Framework with Support for Community-Based and -Driven Devel- Cases from Latin America.� In Deepa Narayan, ed. opment. Washington, DC: World Bank Operations 2005. Measuring Empowerment. Cross-Disciplinary Evaluation Department. Perspectives. Washington DC: World Bank. World Bank. 2005b. Community Driven Development Poverty Reduction Fund. 2003. Manual of Operations. and Social Capital. Designing a Baseline Survey in Vientiane: PRF. the Philippines. Washington, DC: World Bank. Poverty Reduction Fund. 2007. Annual Report 2006. World Bank. 2006a. Lao PDR: Poverty Assessment Vientiane: Prime Minister’s Office. Report: From Valley’s to Hilltops - 15 Years of Poverty Poverty Reduction Fund. 2008a. Annual Report 2007. Reduction. Washington, DC: World Bank. World Bank. 2006b. “Implementing the Bank’s Gender Vientiane: Prime Minister’s Office. Mainstreaming Strategy: Annual Monitoring Report Poverty Reduction Fund. 2008b. Manual of Operations. for FY04 and FY05.� Washington, DC: World Bank, Vientiane: PRF. Gender and Development Group, Poverty Reduction Poverty Reduction Fund. 2008c. Social and Environ- and Economic Management. mental Guidelines. Vientiane: PRF. World Bank. 2006c. Gender Equality as Smart Econom- Poverty Reduction Fund. 2010. Annual Report 2009. ics: A World Bank Group Gender Action Plan (Fiscal Vientiane: Prime Minister’s Office. years: 2006-10). Washington, DC: World Bank. Reinharz, Shulamit. 1992. Feminist Methods in Social World Bank. 2008a. Community-Driven Approaches Research. New York: Oxford University Press. in Lao PDR: Moving Beyond Service Delivery. Vol- Swiss Agency for Development and Cooperation and ume II: Main Report. Washington, DC: World Bank, the World Bank. 2009. “Poverty Reduction Fund: Human Development Sector Unit, East Asia and Aide Memoire of Implementation Support Mission, Pacific Region. November 9-18.� Vientiane: World Bank. World Bank. 2008b. Impact of CDDs on Women’s United Nations Development Fund for Women (UNI- Empowerment—An M&E Initiative. Concept Note. FEM). 2008. Making the MDGs Work for All: Gen- Unpublished. Washington, DC: World Bank. der-Responsive Rights-Based Approaches to the World Bank. 2008c. Social Development How to Series. MDGs. Bangkok: UNIFEM East and Southeast Asia Vol. 1. Washington, DC: World Bank, Social Devel- Regional Office. opment Department. UNICEF. 2007. “International Year of Sanitation 2008: World Bank. 2009. World Bank Group Agriculture Action A Plea for Dignity and Health for All.� http://www. Plan 2010-2012. Washington, DC: World Bank. unicef.org/wes/index_41911.html World Bank. 2010. Lao PDR Gender Profile. Vientiane: Wassenich, Paul and Katherine Whiteside. 2004. CDD World Bank. Impact Assessments Study: Optimizing Evaluation World Bank, Food and Agriculture Organization, and Design under Constraints. Washington DC: World Bank. International Fund for Agricultural Development. White, Howard. 2009. Some Reflections on Current 2009. Gender in Agriculture Sourcebook. Washing- Debates in Impact Evaluation. Working Paper ton, DC: World Bank. 1. New Delhi. International Initiative on Impact World Health Organization. 2010. Trends in Maternal Evaluation. Mortality: 1990 to 2008. Geneva: WHO. 41 Annexes Annex 1. Methods Used in the Pilot Test Sampling trade. Other sources of income, such as forestry or fish- The field work for the piloting of the proposed indi- ing, were of minor importance. This pattern is broadly cators was conducted in 2009 in villages in which representative of the situation in all rural areas in Lao PRF has been active since its first annual cycle, which PDR (National Statistics Center 2005). began in February 2003. Six villages in two districts were included in the pilot test. To capture regional dif- All the women were interviewed using an identical ferences, three villages were randomly selected from questionnaire (see Annex 2. The questionnaire was PRF-targeted communities in the Sukhuma District of based on the forms PRF currently uses. The question- Champassak Province in southern Lao PDR and three naire thus conforms, as much as possible, to forms PRF villages in the Huamuang District in Huaphanh Prov- and the Department of Statistics use.33 This makes it ince in northern Lao PDR. Thus the sampling frame possible to add the questionnaire, or parts of it, to PRF’s consisted of (a) one district in southern Lao and one in existing monitoring and data-collection system. northern Lao; and (b) three randomly selected villages in each district where PRF has been active since 2003 The questionnaire used in the pilot test contained six and remains active. modules: 1. Respondent profile, which solicited information on Several methods were used to gather information about the respondents’ gender, age, marital status, number PRF in the six villages. of children, number of people in the household, eth- Individual Interviews nicity, and main sources of income. Data were collected from 125 women, all of whom had 2. Participation in PRF, which solicited information all participated in PRF’s processes within their villages. on participation in PRF’s activities, including its Some of the women had also represented their villages meetings. 3. Political Sphere in koumban- or district-level meetings. About 60 per- 4. Economic Sphere cent of the women were Lao-Tai, the majority popu- 5. Social Sphere lation in Lao PDR. The other 40 percent represented 6. Survey Assessment, which interviewers completed Hmong (26 percent), Xuay (14 percent), and “other� (1 immediately after each interview. percent). A majority of respondents were in the “mid- dle-age bracket,� older than age 25 but not yet 46 years Focus Group Discussions of age. About one-fifth were under age 25; a similar In addition to the interviews with 125 women, focus- portion of respondents were older than 45. All respon- group discussions were organized in each village with dents were either married (98 percent) or widowed (2 women who had participated in PRF’s processes. The percent). Almost all respondents have children; only discussions gathered information about proposed five respondents were childless. More than a third of qualitative indicators of gender outcomes. To provide the women had five or more children. Finally, almost all respondents derived their primary income from agri- The Department of Statistics was formerly the National Statistics 33 culture, and almost a third derived some income from Center. 42 I m pa c t o f c o m m u n I t y- D r I v e n D e v e l o p m e n t p r o j e c t s o n G e n D e r a means for data triangulation, the focus group discus- Case Studies sions offered the participants the opportunity to talk Case studies were conducted in the six villages to assess about a range of issues, challenges, observations, and situations in which PRF has affected women’s opportu- comments on their lives as well as PRF’s processes and nities and their engagement in local political, economic, perceived impacts. and social spheres. The study team used local knowl- edge to identify suitable participants for the case stud- Interviews with Village Representative ies; women who participated actively in the villages’ eco- Each of the six villages has two PRF representatives, nomic, social, and political life were the main targets. one man and one woman, and each was interviewed. They represent their villages at the koumban-level The need to substantiate the relevance of the proposed meetings and have a central role in voicing village- gender-based indicators for measuring PRF’s impact on level concerns and priorities for subprojects. The women motivated the case studies. The studies offer interviews focused on the nature of active participa- narratives about women’s engagement in local deci- tion of community members and village represen- sion making outside PRF. The case studies provided an tatives and tested the relevance of the items in the opportunity to gather opinions from women active in questionnaires used in the individual interviews, the PRF process and their perspectives about how PRF which are discussed above. is influencing their lives. I m pa c t o f c o m m u n I t y- D r I v e n D e v e l o p m e n t p r o j e c t s o n G e n D e r 43 Annex 2. Questionnaire, English Version Gender Empowerment Pilot Field Survey Individual Questionnaire Identi�cation Code Province ___________________________________________ District ____________________________________________ Village ____________________________________________ Date of interview _____________________________________ Respondent's name __________________________________ Interviewer's name ___________________________________ Time at start ________________________________________ Time at nish ________________________________________ Good morning/afternoon! We are researchers conducting a study on how projects such as the PRF made an impact on women’s lives. This study will allow future projects like PRF measure the effects of their projects on women and their families. We will ask you questions which may be personal, but your answers and your name will not be made public. Thank you very much! Do I have your permission to continue? [ ] Yes [ ] No 44 I m pa c t o f c o m m u n I t y- D r I v e n D e v e l o p m e n t p r o j e c t s o n G e n D e r I. Respondent Pro�le No. Question Alternative Answer Answer code 1 Are you male or female? Male = 1 Female=2 If "Male", go to next respondent 2 What is your relationship 1=Head of household to head of household? 2=Spouse 3=Parent/parent in law 4=Daughter 5=Daughter in law 6=Sister 7=Sister in law 8=Other relatives 9=Non relative 3 How many persons live in your Total number of persons household? 4 How old were you at your last Years birthday? Ask for year of birth and calculate competed years 5 What is your 1=Never married marital status? 2=Married 3=Divorced/separated 4=Widowed 6 How many children do you have? Number of children Number of children living in or outside the household 7 What is your ethnic origin? ……………..……………..… Enter code based on code list 8 What is your main source of A = Agriculture, crops A income? B = Agriculture, livestock B C = Forestry C D = Fishing D E = Industry/handicraft E F = Trade F G = Construction G H = Government employee H I = Other speci�c I ……………..……………..… Mark all relevant in alternatives order of importance. List from 1, 2, 3, etc. Go to II. Participation in PRF I m pa c t o f c o m m u n I t y- D r I v e n D e v e l o p m e n t p r o j e c t s o n G e n D e r 45 II. Participation in Poverty Reduction Fund No. Question Alternative Answer Answer code 1 Have you participated in any 1 = Yes PRF village prioritization 2 = No meetings? If "No", go to next respondent 2 How many times have you 1 = One time 1 participated in PRF village 2 = Two to three times 2-3 prioritization meetings? 3 = Many times 4< 3 Have you ever represented the 1 = Yes village in a koumban 2 = No prioritization meeting? 4 Have you ever been a member of 1 = Yes the Village Implementation and 2 = No Maintenance team? If "No" go to end of module 5 What was your role in the Village 1 = Implementation coordinator Implementation and Maintenance 2 = Material and equipment team? of�cer 3 = Operations and maintenance coordinator Go to III. Political Empowerment 46 I m pa c t o f c o m m u n I t y- D r I v e n D e v e l o p m e n t p r o j e c t s o n G e n D e r III. Political Empowerment No. Question Alternative Answer Answer code Time at beginning of module 1 Lets talk about the last PRF A=Talk in front of group A village prioritization meeting B=Vote on sub-projects B you attended. How did you C=Only joined meeting without C participate in the prioritization talking or voting meeting? Mark all relevant alternatives 2 What type of subproject did 1 = Remember project suggested the female village prioritization 2 = No, don't know/don’t group agree to suggest? remember If 2, go to Question 4 Project code 3 Please enter type of project ………………………………………. A stated by respondent ………………………………………. B ………………………………………. C Ask for 1 or more projects 4 Would you be interested in 1 = Yes representing the village in the 2 = No next koumban meeting? 3 = Don’t know/unsure 5 Lets leave the PRF. Do you 1 = Yes participate in any other group 2 = No making decisions for the people in the village? (Use examples to illustrate) If "No" go to end 6 What other groups making A = Village credit group/revolving A decisions for the people in fund the village do you B = Village meeting B participate in? C= Lao Womens Union C D= Village water group D E= Village health facilitator E F = School board F G = Lao Front for Development G and Reconstruction H = Youth Union H I = Labor Union I J = Other, specify J ………………………………………. Mark all relevant alternatives Time at end of module Go to IV. Economic Empowerment I m pa c t o f c o m m u n I t y- D r I v e n D e v e l o p m e n t p r o j e c t s o n G e n D e r 47 IV. Economic Empowerment No. Question Alternative Answer Answer code Time at beginning of module Module to be repeated for each PRF subproject implemented in village 1 Subproject number (order the subprojects from 1, 2, 3…) Type of subproject ___________________________________ Enter subproject code 2 Your village has been provided 1 = Yes with (type of subproject). Does 2 = No your household use/access this? 3 Has (type of subproject) 1 = Increased income had an effect on your income? 2 = Decreased income (also non-monetary income) 3 = No effect 4 = Not applicable 4 Has (type of subproject) affected 1= Increased cost of living your household cost of living? 2= Decreased cost of living 3 = No effect 4 = Not applicable 5 Has (type of subproject) changed 1= Increased time spent the time you spend on daily 2= Decreased time spent household tasks? 3 = No effect Example: fetching water, 4 = Not applicable collecting �rewood, cooking 6 Do you operate a business on a 1 = Yes regular basis? 2 = No Not including sales of Enter type of business own grown crops ………………………………………………… ………………………………………………… If "No" go to end 7 Has (type of subproject) affected 1 = Yes your business operations? 2 = No 3 = Not applicable If 2 or 3, then go to end 8 How has the (type of subproject) A = Costs A affected your business B = Access to goods B C = Number of customers C D = Sales D 1 if higher, 2 if lower, 3 if no change Repeat module for each subproject implemented in village When all subprojects has been covered, go to V. Social Empowerment Time at end of module 48 I m pa c t o f c o m m u n I t y- D r I v e n D e v e l o p m e n t p r o j e c t s o n G e n D e r V. Social Empowerment No. Question Alternative Answer Answer code Time at beginning of module 1 If you want to sell assets that you A = Decide by myself A own, do you ask other persons B = Ask husband B before you sell them? C = Ask other family member C D = Other person, specify D Use �rewood for KIP 30 000 …………………………………………………. as example Mark all relevant alternatives 2 If you have collected savings, do A = Decide by myself A you ask other persons B = Ask husband B before you use the savings? C = Ask other family member C D = Other person, specify D Use KIP 50 000 for the …………………………………………………. example Mark all relevant alternatives 3 Lets talk about health. Is there 1 = Yes any nurse, clinic, health center 2 = No or dispensary in the village? 4 Have you had any health problem 1 = Yes during the past 12 months? 2 = No If "No", go to 8 5 Did you seek help 1 = Yes for your health problem? 2 = No If "Yes", go to 7 6 Why did you not seek help? A = Not serious enough/ A wanted to wait B = Dif�cult to get there B C = Too expensive C D = Not good quality D E = No cure possible E F = Other, specify F …………………………………………………. Mark all relevant alternatives 7 Who made the decision to seek A = Myself A or not to seek help? B = Spouse B C = Other family member C D = Non family member D Mark all relevant alternatives I m pa c t o f c o m m u n I t y- D r I v e n D e v e l o p m e n t p r o j e c t s o n G e n D e r 49 8 Type of PRF health subproject in village __________________________________ Enter code of PRF health subproject in village If no health subproject in village, Go to 11 9 There has been a (type of health 1 = Yes subproject) in your village 2 = No Have you ever used this? 10 How has (health subproject) 1 = Easier to get help changed your ability to get help 2 = More dif�cult to get help when you have health problems? 3 = No difference 11 Lets talk about schooling. Is 1 = Yes there any elementary school in 2 = No the village? 12 Did you attend school as a child? 1 = Yes 2 = No If No, go to 14 Level Class 13 What was the highest level 1 = Lower primary 1-3 you completed? 2 = Upper primary 4-5 3 = Lower secondary 6-8 4 = Upper secondary 9-12 5 = Vocational training 6 = University/institute Go to 15 14 Why did you not attend school? A = Too expensive A B = No interest B C = Need to work C D = School too far away D E = No teachers/supplies E F = Illness F G = Language G H = Other specify H …………………………………………………. Mark all relevant alternatives 50 I m pa c t o f c o m m u n I t y- D r I v e n D e v e l o p m e n t p r o j e c t s o n G e n D e r 15 Who made the decision for you A = Father A to attend or not attend school? B = Mother B C = By myself C D = Other family member D E = Non family member E Mark all relevant alternatives 16 Number of children between Number of children 6-14 age 6-14 in household? If "0" go to end 17 Number of children in the Number of children 6-14 household age 6-14 that attend that attend school school? If all children go to school, go to Question 19 18 Why did the child/children not A = Too expensive A attend school? B = No interest B C = Need to work C D = School too far away D E = No teachers/supplies E F = Illness F G = Language dif�culties G H = Other specify H …………………………………………………. Mark all relevant alternatives 19 Who made the decision to send A = Myself A or not to send the children to B = Spouse B school? C = Other family member C D = The child D E = Non family member E F = Other specify F …………………………………………………. Mark all relevant alternatives 20 Type of PRF education subproject in village ______________________________ Enter code of PRF education subproject in village If no education subproject in village, Go to end. 21 How has (education subproject) 1 = Easier to go to school made it easier for children 2 = More dif�cult to go to school in the village to go to school? 3 = No difference Thank the respondent for participating in survey. Time at end of module Also note time of completion on cover page Fill in the survey assessment I m pa c t o f c o m m u n I t y- D r I v e n D e v e l o p m e n t p r o j e c t s o n G e n D e r 51 VI. Survey Assessment No. Question Alternative Answer Answer Code Answer Code To be �lled in by the �eld researcher immidiately after the interview Module Question no. Dif culty 1 List all questions that the I. Respondent pro�le …………….. …………….. respondent answered with …………….. …………….. dif culty or were unable to …………….. …………….. answer. …………….. …………….. II. Participation in PRF …………….. …………….. Indicate the type of dif culty …………….. …………….. A = Unclear wording of question …………….. …………….. B = Language dif�culty due to …………….. …………….. respondent not fluent in Lao III. Political Empowerment …………….. …………….. C = Respondent lack information/ …………….. …………….. respondent don't know …………….. …………….. D = Respondent unwilling to …………….. …………….. reveal information/sensitive IV. Economic Empowerment …………….. …………….. question …………….. …………….. E = Other, specify in "Dif�culty" …………….. …………….. column …………….. …………….. V. Social Empowerment …………….. …………….. …………….. …………….. …………….. …………….. …………….. …………….. 2 What is your overall judgement 1 = Very reliable of the reliability of the responses? 2 = Somewhat realiable 3 = Somewhat unrealiable 4 = Unrealiable If 3 or 4, please provide reason ……………..……………..………………..……………..……… ……………..……………..………………..……………..……… ……………..……………..………………..……………..……… 3 Please provide any other ……………..……………..………………..……………..……… suggestions for improving ……………..……………..………………..……………..……… the questionnaire ……………..……………..………………..……………..……… ……………..……………..………………..……………..……… ……………..……………..………………..……………..……… ……………..……………..………………..……………..……… 52 I m pa c t o f c o m m u n I t y- D r I v e n D e v e l o p m e n t p r o j e c t s o n G e n D e r Subproject Codes No. Subproject Type Items Eligible Please inform us about all projects implemented in village Number of since the village started participating in the Poverty sub-projects Reduction Fund implemented A. Community Infrastructure 1 Access A = Small bridge B= Footpath C = Track D = Culvert E = Ramp F = Pier G = Road repairs and upgrading H = Other, specify ………………………………………. 2 Community electrical supply A = Mini-hydro generator B = Wiring C = Line extension D = Other, specify ………………………………………. 3 Primary health care facilities A = Health centers building and furniture B = Supplies and medicines C = Allowances for contracted nurse or midwife D = Village medicine kit E = Traning/scholarships F = Medical equipment G = Other, specify ………………………………………. 4 Domestic water systems A = Well B = Gravity water supply C = Latrines D = Other, specify ………………………………………. I m pa c t o f c o m m u n I t y- D r I v e n D e v e l o p m e n t p r o j e c t s o n G e n D e r 53 5 Education A = School and nursery building B = Allowance for contracted teacher C = Supplies, equipment or furniture D = Training and scholarships E = Textbooks F = Musical instruments G = Other, specify ………………………………………. 6 Agricultural infrastructure A = Weirs B = Ponds C = Canals D = Bunds E = Gates F = Spillways G = Other, specify ………………………………………. 7 Markets, community halls A = Buildings B = Drainage C = Wells D = Furnishings E = Other, specify ………………………………………. B. Training Programs 8 Training programs Specify type of training program ………………………………………. ………………………………………. 54 I m pa c t o f c o m m u n I t y- D r I v e n D e v e l o p m e n t p r o j e c t s o n G e n D e r Annex 3. The Virtues of Random Sampling Random sampling provides an efficient means to col- past year with a 95 percent confidence level that the lect data from a relatively small number of people or actual values were at least 60 percent but less than 66 households and then to generalize the results from a percent. sample to an entire population. To produce a sample that is truly random, every member of the population To determine the appropriate sample size, it is neces- must have an equal probability of being included in the sary to select a desired or acceptable confidence level sample, regardless of the potential challenges associ- and sampling error. There are several online tools that ated with that sample. Simple random sampling means can be used to calculate sample sizes, and one of the that respondents in a survey cannot be selected merely easiest to use can be found at http://www.raosoft. because they are conveniently located, because they com/samplesize.html. Using this sample-size calcu- have volunteered to participate, or because they are lator produces the sample sizes shown in Table A-1. friendly with the data collectors. A population need not As the sample sizes suggest, the cost of achieving low be all Laotians or all people in a province. A population margins of error and high confidence levels is often is simply all the “units� of interest. Units can be people, prohibitively high and rarely worth the expense, espe- households, villages, rice paddies, school children, etc. cially when high levels of precision are unnecessary. For PRF’s purposes, a population can be all females in For many of its purposes, PRF could reasonably and PRF villages over the age of 15 or all villagers who have justifiably choose a margin of error of ±5 percent and accessed health services in the past year. a confidence level as low as 90 percent when sampling villagers, households, and even villages. Conversely, In deciding the size of a random sample, there are two there is no need or justification for PRF to use a mar- key considerations. The first involves confidence levels, gin of error of ±1 percent or a confidence level above which indicate how confident one should be with the 95 percent. results of a random sample. A 90 percent level is often appropriate, but a typical confidence level is 95 per- It is important to remember, however, that these sug- cent. This confidence level indicates that if data were gestions are based on an assumption of high responses collected from twenty independent samples from the rates among anticipated respondents. Likewise, sample population, similar results would be obtained nineteen sizes should be increased if the PRF anticipates that it of twenty times. Likewise, a confidence level of 99 per- will disaggregate the data in several ways, such as by gen- cent means that similar results would be obtained 99 der, Lao-Tai ethnicity versus minority ethnicity, or geo- times from 100 independent samples. The higher the graphic location. To illustrate, consider a sample of 1,066 confidence level desired, the larger the required sam- villagers divided equally between males and females. ple size. For the entire sample, the margin of error would be ±3 percent with a 95 percent confidence level, as shown in The second consideration involves sampling error. Ran- Table A-1. For the 533 females, the margin of error would dom samples produce estimates of the characteristics of increase to approximately 5.6 percent and the confidence a population, so random samples reflect some amount of “error.� As an illustration, a survey using a random sample might find that 63 percent of women have had an illness requiring medical attention within the past TAble A-1 Sample sizes for large year. The sampling error might be plus or minus (±) 3 populations percent, which would mean that the actual values might Confidence Level be as low as 60 percent but as high 66 percent. For PRF’s Margin of Error 90% 95% 99% purposes, a sampling error of ±5 percent is sufficient. The smaller the margin of error desired, the larger the ±5 271 384 663 required sample size. ±3 751 1,066 1,838 Confidence levels and margins of error are commonly ±1 6,696 9,466 16,181 reported together. In the example just discussed, a Note: Assumes a 100 percent response rate and a population size of report might indicate that 63 percent of Lao women in 660,000, the approximate total population in PRF’s villages in 2009– PRF’s villages have required medical attention in the 2010. I m pa c t o f c o m m u n I t y- D r I v e n D e v e l o p m e n t p r o j e c t s o n G e n D e r 55 level would decrease—because the subsample of females socialresearchmethods.net/kb/sampling.php and at is much smaller than the overall sample. http://www.stat.berkeley.edu/~census/sample.pdf. Iar- ossi (2006) also provides a useful and comprehensive There may be instances in which other considerations, introduction to sampling and the management of sur- including cost, are important, and alternative methods veys. Considerable expertise with sampling and data of sampling are possible or desirable, such as cluster and collection also can be found within the Ministry of Plan- stratified sampling. There are many online sources of ning and Investment’s Department of Statistics, which information about sampling procedures, including the is responsible for conducting the Lao Expenditure and “Research Methods Knowledge Base,� at http://www. Consumption Surveys. 56 I m pa c t o f c o m m u n I t y- D r I v e n D e v e l o p m e n t p r o j e c t s o n G e n D e r Annex 4. Community Participation at Various Stages of KDP (data collection form) Name of FK: Date: Kecamatan: District: Province: No. of Persons Participating in Activity Quality of Date(s) of Participation Activity Activity Total Male Female Poor (circle one) General Comments UDKP I (First Village Very Active Development Unit) Active Fair Poor MUSBANGDES I Very Active (First Village Active Discussion Forum for Fair Development) Poor SOCIALIZATION Very Active DUSUN/DESA (sub- Active village/village) Fair Poor MUSBANGDES II Very Active KHUSUS (Second Active Women Specific Fair Discussion Forum on Poor Village Development) MUSBANGDES II Very Active (Second Village Active Discussion Forum for Fair Development) Poor UDKP II (Second Very Active Village Development Active Unit) Fair Poor MUSBANGDES III Very Active (Third Village Active Discussion Forum for Fair Development) Poor PROJECT Very Active IMPLEMENTATION Active Fair Poor PROJECT Very Active MAINTENANCE Active Fair Poor NO. PERSONS ON Very Active O&M COMMITTEE Active Fair Poor I m pa c t o f c o m m u n I t y- D r I v e n D e v e l o p m e n t p r o j e c t s o n G e n D e r 57 Instructions: This form is to be completed by the Kecamatan Facilitators (FKs) each month. The form reports upon community participation throughout KDP’s project cycle. Please fill out this form as completely as possible. 1. Write your name, date, and names of the kecamatan, district and province. 2. Date(s) of Activity Fill in the date of each activity. If the activities occurred over several dates, write down the range of dates, i.e., 4 – 20 June, 2000. 3. No. of persons participating in Activity From the attendance lists, write down the number of persons attending each activity, how many male, female, and poor. 4. Quality of Participation Circle one of the following on the table to describe the quality of participation for each activity. 77 Very active 47 All or the vast majority of villagers (over 70%) were involved in the activity. 47 Everyone felt free to speak up and play an active role. 47 Women and poor groups participated in the activity. 77 Active 47 Over half (51-70%) of the villagers were involved in the activity. 47 Most members participated actively and felt free to speak up and play an active role. 47 Villagers asked questions during the activity and showed interest. 77 Fair 47 Participation was still limited to some or the minority of villagers. 47 The elite and some community members were involved. 47 Only a few members felt free to speak up and play an active role. 47 There was very limited involvement of women and poor groups. 77 Poor 47 Participation was limited to one or two influential persons or the village elite. 47 No members felt free to speak up and play an active role. 47 There was hardly any involvement of women and poor. 5. General Comments Please include any general or overall comments about each activity. THE WORLD BANK Social Development Unit Sustainable Development Department East Asia and Pacific Region 1818 H Street, N.W. Washington, D.C. 20433, USA Telephone: 202 473 1000 Facsimile: 202 522 1666 Website: www.worldbank.org/eapsocial