Report No. 38901-IN India Rural Governments and Service Delivery (In Three Volumes) Volume III: Main Report March 7, 2007 Sustainable Development Unit South Asia Region Document of the World Bank Table of Contents VOLUME 111MAINREPORT I MAKINGDECENTRALIZATIONWORK:BUILDINGONTHECONTEXT . ........................ 1 THECONTEXT........................................................................................................................................ 1 Key services arefailing ..................................................................................................................... 3 Understanding thefailure of rural services ...... .10 BUILDINGONTHECONTEXT: ADDRESSINGFUNCTI .......................... 11 THECONCEPTUALFRAMEWORK.......................................................................................................... 14 Key public services ...... Criteriafor analysis .. Study methodology .... ............................................................................ 21 I1 . MAPPINGACTIVITIES: UNBUNDLINGRURALSERVICES ............................................... 24 BASICDEFINITIONS: SECTORS, SERVICESAND ACTIVITIES................................................................... 24 KEY SECTORS: SERVICESAND ACTIVITIES ........................................................................................... 27 Primary education .......................................................................................................................... 27 Health services......... ..................................................................................................... 27 111 . THE LEGAL FRAMEWORK:IMPLICATIONS FORTHE DIFFERENTTIERS OF GOVERNMENT ............................................................................................................................... 31 THESTARTINGPOINT: THECRITICALPROVISIONSOFTHE 73mAMENDMENT ...................................... Thehistorical background .............................................................................................................. 31 32 THEIMPACTOF THEAMENDMENT THE DEVOLUTIONOFPOWERS.................................................. Theheart ofthe legalframework .................................................................................................... 32 ON 34 THEIMPACTOF THE STATE ACTS:ANANALYSIS USINGTHE THREE F's .............................................. 38 Functions ..................................................................................... ................................... 38 Primary education: legislation and implications ............................................................................ 46 Health: legislation and implications ............................................................................................... 49 Rural water and sanitation: legislation and implications Employment programs: thepredominance of central gui INSUMMARY' DOES THE LEGAL FRAMEWORK ENABLE PRlS? ............................................................. 57 IV. THE REALITY ONTHE GROUND:KEY SERVICESAT THE LOCAL LEVEL ................59 FINDINGOUTABOUT THEGROUNDREALITY' METHODOLOGY............................................................. 59 PRIMARY EDUCATION: IMPLEMENTATION AND QUALITY ..................................................................... 60 An analysis of implementation: the limited role of PRIs and thepredominance of the state government...................................................................................................................................... 60 ARRANGEMENTSTHE HEALTH SECTOR............................................................................................ IN 68 RURALWATER SUPPLYAND SANITATION: FUNCTIONSINSERVICEPROVISION..................................... 70 EMPLOYMENT PROGRAMS: THE DOMINANCE OF THE RURALDEVELOPMENT AGENCY ......................... 75 INCONCLUSION: THEDEFACTOPICTURE ............................................................................................. 77 V. THE ASSIGNMENT OF FUNCTIONS,FUNCTIONARIESAND FINANCE:APPLYING THE CONCEPTUALFRAMEWORK .......................................................................................... 78 PRIMARY EDUCATION .......................................................................................................................... 78 Applying publicfinance criteria toprimary education 78 Accountability criteria applied toprimary education..................................................................... 84 I n summary: towards a coherent primary education delivery system............................................. 88 INFORMATION AND ENFORCEABILITY ................................................................................................. 90 11 .. Health sewices................................................................................................................................ 90 Applyingpublicfinance criteria to the characteristic marketfailures in health sewices...............91 Sewices and the appropriate level of government .......................................................................... 94 I n summary: optimal allocation of responsibility with accountability.. ......... 99 Expanding the role of PRls: some guidelines ............................................................................... 100 WATER SUPPLY AND SANITATION ...................................................................................................... 101 Applyingpublicfinance criteria to rural water and sanitation ................. 102 Applying accountability criteriaforjurisdictional allocation ....................... ..................... 105 Ensuring Accountability ................................................................................................................ 107 EMPLOYMENT PROGRAMS ................................................................................................................. 111 Applying accountability criteria to employmentprograms........................................................... 114 Assigning activities in employment programs: guidelines.,............................................. VI. TOWARDS BETTERRURAL SERVICES:RECOMMENDATIONS ................................... 123 I n conclusion: the context andfindings of the study ..................................................................... 123 Studyfindings ................................................................. I25 Recommendations ......................................................................................................................... 128 Listof Tables TABLE1.1. ESTIMATE INADEQUATE LEARNINGACHIEVEMENT* INMATHEMATICS. FOURSTUDY STATES. 6 OF TABLE 1.2: UNDERSTANDING THE FIRSTPRINCIPLES OFACCO~TABILI .................................................... 21 TABLE11.1. MATRIX SERVICESAND ACTIVITIES INTHE KEY SECTORS...................................................... OF 25 TABLE111.1.STATE-WISE STATUS OFTHE DEVOLUTIONOFFUNCTIONS ......................................................... 37 TABLE111.2: STATE-WISE COMPARISONOFFINANCIAL AND FUNCTIONALDEVOLUTION ................................ 42 TABLE111.3:MATRIX FORTHEDE JUREALLOCATION OF RESPONSIBILITIESINEDUCATIONINTHE FOUR STATES ...................................................................................................................................... 49 TABLE111.4: THEDEJUREALLOCATION OFFUNCTIONSFORWATER SUPPLYINTHEFOURSTATES.................54 TABLE111.5:THEDEJUREALLOCATION OFFUNCTIONSFORSANITATIONINTHE FOURSTATES...................... 55 TABLE111.6: THEDEJURE MATRIX FORALLOCATION OFFUNCTIONALRESPONSIBILITIESINTHE SGRY ................................................................................................................................. 56 TABLEIv 1 THEDE FACTO STATUSOFPRIMARY EDUCATION, KARNATAKA, KERALA WEST BENGAL62 . . PROGRAM NAD . TABLEIv.2:FUNCTION,ACTIVITY AND RESPONSIBILITYFORWATER SUPPLYINTHE FOURSTATES ..............71 TABLEIV.3:INSTITUTIONALMODELSOFWATER SUPPLYDELIVERY, KERALA .............................................. TABLEIv4: THEDEFACTOALLOCATION OFRESPONSIBILITIESINEMPLOYMENT PROGRAMSINTHEFOUR 72 STATES ...................................................................................................................................... 77 TABLEv.1.SIZES AND ENROLLMENTS OFSTATE, DISTRICT AND BLOCK-LEVEL SCHOOLS ............................ 80 TABLEv.2:PUBLIC FINANCEFIRSTPRINCIPLES FORRESPONSIBILITYALLOCATION ACROSS GOVERNMENT LEVELS INEDUCATION .............................................................................................................. 83 TABLEv.3:RECOMMENDEDALLOCATIONOFACTIVITIESFORRURALPRIMARY EDUCATION INDIA.......87 IN TABLEv 4: FIRSTPRINCIPLESFORALLOCATION OFFUNCTIONALRESPONSIBILITYBASEDON .............................................................................................. 88 TABLEv.5: PUBLIC FINANCEFIRSTPRINCIPLES FORRESPONSIBILITYALLOCATION ACROSS GOVERNMENT ACCOUNTABILITY INEDUCATION LEVELS INWATER SUPPLY ............................. .................................................................. 104 TABLEV.6: PUBLIC FINANCEFIRSTPRINCIPLES FORRESPONSIBILITYALLOCATION ACROSS GOVERNMENT LEVELS INSANITATION ........................................................................................................... 105 TABLEv.7' FIRSTPRINCIPLESFORALLOCATION OFFUNCTIONALRESPONSIBILITYBASEDON ACCOUNTABILITY INWATER SUPPLY ...................................................................................... 106 TABLEV.8: FIRSTPRINCIPLESFORALLOCATION OFFUNCTIONALRESPONSIBILITYBASEDON ACCOUNTABILITY INSANITATION ................................... .............................................. 107 TABLEv.9' RECOMMENDEDNORMATIVEASSIGNMENT OFFUNCTIONSFORRURAL WATER AND SANITATION ........................................................................................................................................................... 110 TABLEv.10:PUBLIC FINANCEFIRSTPRINCIPLES FORRESPONSIBILITYALLOCATION ACROSS GOVERNMENT LEVELS INEMPLOYMENT PROGRAMS ...................................................................................... 113 ... 111 TABLE .11 FIRSTPRINCIPLESFORALLOCATION OFFUNCTIONALRESPONSIBILITYBASEDON v . ...................................................................... 114 TABLEv.12: RECOMMENDEDINSTITUTIONALARRANGEMENTSFOREMPLOYMENTPROGRAMS.................119 ACCOUNTABILITY INEMPLOYMENT PROGRAMS TABLEVI.1. RECOMMENDEDMAINFUNCTIONSOFDIFFERENTGOVERNMENT TIERS INTHE KEY SECTORS 131 . ListofFigures FIGURE1.1. TEACHER ABSENCE: A COMPARISONOF MULTI-COUNTRY FIGURES AND THE FOUR STATES .........4 FIGURE1.2: ATTAINMENT ...................... 5 FIGURE1.3: STATE-WISE ABSENTEERATESINPRIMARY HEALTH CAREFACILITIES......................................... PROFILES INTHE FOUR STATES. RURAL/URBAN AND MALE/FEMALE FIGURE1.4: SHAREOFPRIVATESECTORINTHE HOSPITALAND PRIMARY HEALTHCAREMARKETSININDIA ..78 FIGURE1.5: INCIDENCEOFPUBLIC SUBSIDY INHEALTHININDIAAND THE FOURSTATES ............................... 9 FIGURE1.6: SHAREOFPUBLIC HEALTH CARE SUBSIDIES RECEIVEDBY INCOME GROUPS .............................. 9 FIGURE1.7 DEMYSTIFYING ELEMENTS ACCOUNTABILITYRELATIONS................................................... OF 16 FIGURE1.8: THEFOURPOSSIBLERELATIONSHIPSOFACCOUNTABILITY INVOLVED INTHEPROVISIONOF ............................................................................................ FIGURE1.9. ACCOUNTABILITYRELATIONSHIPSFORRURALPEOPLE: CANITWORK ININDIA?...................... SERVICESVIA THE PUBLIC SECTOR 18 19 FIGURE1.10:ANANALYTICALAPPROACH TO EFFECTIVE DECENTRALIZATION ............................................ 23 FIGURE111.1.PLANANDNON-PLANEXPENDITURE ELEMENTARY ON EDUCATION ....................................... 44 FIGURE111.2 EXPENDITURE RURALDEVELOPMENT BENGAL ...................................................... FIGUREIv.1 THEDEFACTOSTATUS OFINFRASTRUCTUREACTIVITIES INPRIMARY EDUCATION, KERALA63 ON -WEST ....45 FIGURE1v.2:THEDEFACTOSTATUSOFTHE TEACHING SECTORINPRIMARY EDUCATION, KERALA..............63 FIGUREIV.3:THEDEFACTOSTATUSOF INFRASTRUCTUREACTIVITIES INGOVERNMENTPRIMARY SCHOOLS, .......................................................................................................................... 64 FIGUREIv4: THEDEFACTOSTATUSOFTHETEACHINGSECTORINGOVERNMENTPRIMARY SCHOOLS, WEST WESTBENGAL BENGAL .................................................................................................................................... 65 FIGUREIV.5:THEDE-FACTO STATUS OF TEACHINGSECTOR--SSK .............................................................. 66 FIGUREIV.6:RESPONDENTPERCEPTIONOFSTATEGOVERNMENTROLEININFRASTRUCTUREACTIVITIES: GOVERNMENT PRIMARY SCHOOLSVS .SSK, WEST BENGAL ..................................................... 67 FIGUREIV.7.RESPONDENTPERCEPTIONOF STATEGOVERNMENTROLEINTEACHING SECTORACTIVITIES: .SSK, ..................................................... FIGUREv.1.FUNCTIONALALLOCATIONACCORDING FIRSTPRINCIPLESOFACCOUNTABILITY GOVERNMENT PRIMARY SCHOOLSVS WEST BENGAL TO ...............6785 FIGUREv.2:RECOMMENDEDFUNCTIONAL ALLOCATIONOFRESPONSIBILITIES FORRURALPRIMARY EDUCATION INDIA IN ................................................................................................................ 86 FIGUREV.3: MODELSDECENTRALIZATION OF .............................................................................................. 89 FIGUREVI.l. THEANALYTICALAPPROACH ................................................................................................ 128 ListofBoxes BOX 1.1. BRINGING GOVERNMENT CLOSERTO THE PEOPLETHROUGH THEPANCHAYAT .................................. 2 Box1.2: THESTANDARD APPROACH: FAILUREACROSS SECTORS................................................................ 11 BOX 1.3: THEFOUR STATESUNDER STUDY................................................................................................... 14 BOX 111.1.THE 29 SUBJECTS THAT STATES CAN DEVOLVE TO LOCAL GOVERNMENTS .................................... 33 Box111.2:THERANGA REDDYDISTRICT SAWANCHES JUDGMENT.............................................................. 36 BOX 111.3:TEACHER WAGES INGOVERNMENT SCHOOLS: A GENEROUSPACEOF INCREASE ............................ 47 BOX 111.4: EXAMININGTHE DESIGN OF EMPLOYMENT PROGRAMS .................................................................. 57 BOX Iv.1.SELECTION OF DISTRICTS FORSTUDY ............................................................................................ 60 BOX Iv.2:PNS INTHE HEALTHSECTOR: THE KERALA EXAMPLE 70 BOX v.1. .................................................................. INCREASINGACCOUNTABILITY FRONTLINE OF PROVIDERS: THEROLEOF FINANCE ..................... 90 BoxV.2: LOCAL MONITORING INDELIVERY HEALTH OF SERVICES ININDIA .............................................. 91 BOX V.3: THEGAINS ACTION ON PUBLIC HEALTH INCHINA OF ..................................................................... 92 BOX v4: LESSONS FROM INTERNATIONAL CASES OF DECENTRALIZED HEALTH ................................ CARE 98 BOX V.5: INTERNAT~oNALEXPERIENCE TARGETING BENEFICIARIES IN ...................................................... 111 1v o f India and state ga ESut rhls 1s not a m s f s w Box 1.3). I. MakingDecentralizationWork:BuildingontheContext India has made bold efforts to strengthen the voice of the rural poor through decentralization to local governments. The 1993 Constitutional Amendment created a decentralized government structure. But key services -basic education, health, drinking water and sanitation, and employment programs - are still not being effectively delivered to disadvantaged rural people. How do we use the decentralized structure - the variouspanchayat levels -- to ensure better services at the local level? Through little progress in implementing the decentralization laws, India is missing an opportunity. There are good reasons to believe that decentralization will greatly help in expanding services in education, health, water, sanitation and inthe implementationof ruralanti-poverty programs. This report discussesthe roles, responsibilitiesand relations between center, state and local governments that can help the PRl's, particularly the GP's, succeedinthis task. This chapter highlights the role of accountability of governments to the people - and of service providers to the people, either directly or through their government as the crux of the matter of improving services. It also shows that - the current, top-down systemof delivering i s an historical anomaly though one shared by most post-colonial developing countries. Services, however they are delivered now evolved from below in rich countries - and the top-down approach that has been inappropriately adopted, particularly for a country the size of India, represents an impossibleleap over this evolution. The context The importance o frural service delivery The needs of the ruralpoor inIndia include basic education, health, drinlungwater and sanitation, employment programs, and many others. If the quality o f their lives i s to improve, services in these key sectors need to reach the rural poor. Recognizing this responsibility, the central and state governments have been the major providers o f basic public semces since independence, spending a significant part of available resources on schemes for rural development. The delivery system The central objective of India's rural Like most countnes, India i s structured as a decentralization to give a voice to the centralized delivery system based on a large civil people of neglected areas and service. The minimumneeds program announced by communities. Bringing government the Government o f India (GOI) inthe FifthFive Year closer to the rural people means an Plan (1974-1979) identified eight areas for specific improvement in the match between intervention: pnmary education, pnmary health, water diverse local preferences and public supply, electnfication, roads, housing, environmental services. improvement and nutntion. This list was expanded in 1 r to the peoplethroughthe apt coordinates between urban and rural governments at the distnct level. The Constitutional Amendments specify a list o f 29 subjects that states can choose to devolve to rural governments at the distnct, block or village levels. The reality of failure India's effort to decentralize has been a bold attempt to address the need for better semce delivery -- by moving away from the centralized mode of provision to "a representative and democratic institution ... (thatwill) ...evoke local interest and excite local initiative in thefield of development."' But despite the sound rationale o f decentralization as strategy, despite the range o f central and state government schemes for rural development, the reality i s that semces continue to fail in rural areas. Where they are delivered, the quality is l o w Most states show regional dispanties in development, with rain-fed, forested and waste lands -- where most o f the rural poor are concentrated - at a disadvantage. Though central and state governments spend about seven percent o f GDP on rural development, the results are not commensurate with expenditures. Seven hundredmillionpeople still live inrural areas, most o f them invillages; and o f these, about 400 million are poor, the largest number ina single country inthe world. Key services arefailing Perhaps the most glanng indicator o f the failure o f service delivery i s the frequent absence of teachers. The 1999 Public evidence shows that the Report on Basic Education in India (PROBE) brought this current state basic problem to the spotlight with data and anecdotes o f pnmary education deliveryOfin rural schools infour states: India is far from effective -- in attainment, in provision of the basic services, and in learning achievement. When the investigators reached the primary school in Jotri Peepal (Bharatpur, Rajasthan) shortly after noon, no teacher was in sight. One teacher, who had apparently left for lunch, soon appeared, He said that the school actually had three teachers, but that the headmaster and another teacher had gone elsewhere on oficial duty. The villagers contradicted this story. They said that the two absconding teachers did not turn up at all. The only one who did was the one the investigators had met...He too was highly irregular and opened the school at will2 PROBE found considerable evidence o f similar absenteeism and "non-teaching." Subsequent nationwide studies3 using large samples and careful methodologies confirmed that teacher absenteeism inIndia is very highincompanson wlth other developing countnes (see Figure 1.1). Almost 25 percent o f teachers are absent from work at a given time in India.4 While Kerala did relatively well compared to other states, its absence rate was still higher than any other country in the sample (except Uganda). * PROBE 'The Balwantrai C o m t t e e , 1957 'Such Team 1999,quoted in Chaudhury et al, 2004. as Kremer et al, 2004. The percentageIS likely to be even higher if a rural-only sample1staken. 3 c: a of -c anti the states I - 3 Figure1.2: Attainmentprofilesinthe four states, rurayurbanandmale/female Rajasthan (India) 1998-99 Kerala (India) 1998-99 Attainment profile, ages 15-19 Attainment profile, ages 15-19 1 .8 .6 = E 4 .2 0 1 3 1 3 Grade 5 7 9 Grade 5 7 9 -Male Urban -Male Rural -Male Urban -Male Rural -$-Female Urban- Female Rural -.==&Female Urban- Female Rural 1 - 1- .8 - 5 .8 - e n. e 4- '2fi 2 - 0 5 7 9 O - l 1 1 3 5 7 g Grade Grade -Male Urban -Male Rural -Male Urban -Male Rural aMa~i Female Urban- Female Rural jd_ua* Female Urban-Female Rural Source: Filmer, 1999 (httD://www.worldbank.or~ell.esearch/~roiects/edattain/edattain.htm) 5 6 Let us consider absenteeism7 among medical care prowders. Figure 1.3 shows the results o f a large scale study o f surpnse visits to health facilities in all major states. The first thing to notice i s the very high average level of absenteeism for the country as a whole.. There are legitimate reasons for being away from posts such as leave or official duty However, the numbers claimed for being on leave are much higher than are legitimate given leave rules. The study also found that absenteeism was worst in the smaller sub-centers (for staff that were not supposed to be on home visits), followed by the pnmary care centers. The lowest absenteeism was in the few Community Health Centers (small hospitals) inthe sample. Figure 1.3: State-wise absentee rates inprimary health care facilities 70 BO 50 40 30 20 10 0 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Another evldence o f the failure o f public health service i s the dominance that pnvate health service has assumed inthe curative health. Figure 1.4 shows the share o f the pnvate sector inthe hospital and pnmary health care markets in India. The strilung feature is the significant presence o f the private sector in health provision, despite the existence o f a large public sector delivery system. Even in "traditional public health programs" with free semces and frequent campaigns, such as immunization, the private sector still i s very large inprovision. The relevant figures raise several strilung points with regardto the pnvate sector: 0 The predominance o f the pnvate sector for pnmary care in all states: For India as a whole, this hovers around 80 percent for all income groups. Among the study states, West Bengal stands out for a greater reliance on the pnvate sector for pnmary care. A substantial amount o f hospital care in the private sector: Inthe absence o f insurance, fewer people can afford the out-of-pocket expense o f in-patient care inprivate hospitals. Even so, the all-India average i sjust under 50 percent.* 0 Kerala's large share inpnvate hospital care: With a figure similar to the all India average, Kerala has, by far, the largest pnvate penetration o f the market for hospital care. Chowdhry, 2005. Higher income groups use medical services much more than the poor, so the average is close to the high end of the graph. 7 ure EA: 1Sbnreof p tar in tbe haspita Ith mar fadla Share of tire private services, rurof arms Share ofthe private sector in ~ ~ ~i ~~ ~i ~days,atrural e ~ ~ t a ~~ areas 8 of public health care subsidies r Understandingthefailure of rural services The needs, suptdv. civil service solution amroach Many developing countries, including India, approached the problem o f improving service delivery by relying on a top-down bureaucracy to supply a top-down and uniform public service." Inthis approachto semce delivery, a universal needwas met by a technical supply solution, and then implemented by an impersonal provider dnven by rules. Need was the problem, supply the solution and civil semce the instrument. This became the standard approach to solving public service concerns. The limitations o f the standard approach: the case o f rural water supply H o w does this standard solubon structure the key elements o f resources, information, decision- malung, delivery mechanisms and accountability7 Resources are centralized and canalized. The center collects nearly all The basic problem with the resources from general taxes, rents or aid, and then allocates supply, civil them into the budgets o f line ministnes. There are few user solution is that it treats all problems as amenable to the fees or local taxes. Information, if it exists at all, 1s tightly logic ofpolicies andprograms. controlled and only flows internally and upward. Decision- malung i s pnmarily by government agencies and their agents, with the discretion o f local agents only on paper. And it is tightly controlled by rules, regulations and mandates from the top. Delivery mechanisms are ma line agencies that reach directly from the center to the service prowder. The accountability o f service providers flows internally and upward. Accountability to citizens occurs only vla whatever political mechanisms exist to express discontent. Rural water supply i s a case inpoint. At first glance, this seems like a perfect case for the "needs, supply, civil semce delivery" paradigm. After all, what could be more o f a "need" than a biological necessity such as water, especially when the health The pattern of systemic consequences from insufficient or contaminated water sources are failure across sectors, from so obvlously harmful? What other problem could call more clearly water supply to basic for a supply-side solution such as low-cost engineering education, indicated the "appropnate technology," say a public standpipe that can be made need to Participatory available to all at virtually no cost? Indeed, the first round o f into the system, including: government intervention was to launch discrete (often donor- funded) projects that would create simple and inexpensive public 0 Incorporating local knowledge; standpipes. Sometimes it succeeded; but sometimes it failed, and badly For instance, one recent revlew o f 12,000 standpipes 0 Assessinglocal demand; and showed that breakdown rates were 50 percent when maintenance Creating transparent was the responsibility o f the national water corporation. But these conditionsof supply. fell to 11percent when it was under community control." But the failure was attributed to proximate causes and imperfect project design rather than lack o f community input. A new round o f "better o f the same" solutions were launched - better training, better technology, better central funding for maintenance. Only after the second round o f failures were the failures recognized as systemic. These decisions involved historical, social and political processes which tended to overlook the interactions among citizens, providers and the state. ''Narayan, 1995. 10 Analysts and policy-makers agree that three key factors affect the ability o fpanchayats to fulfill their role as institutions o f local government: For local governments to deliver functions, funds and functionanes, or the "three F'S.''~~services effectively, they must have Local governments must have clear functions, clear functions, sufficient funds, and sufficient funds and functionanes who are functionaries who are accountable. accountable. Local governments need to be accountable to people through the democratic election process and other means, and also to higher levels o f government when they are asked to deliver services on their behalf. They must be endowed with sufficient resources; the discretion to tailor services to local needs; and the capacity to deliver those services. Talung into account existing problems One reason for lack o f accountability i s because there i s confusion with the assignment o f roles to the public sector, and within it to different levels o f government. All levels o f government seem to intervene in the provlsion o f the same services. For example, a recent count inKarnataka listed 24 programs Overall, two problems continue to for drinkmg water supply and 72 programs for the . affect service delivery: Theconfused,overlapping welfare o f tribal people; five levels o f government were and incomplete involved in the provlsion o f these programs. Again, responsibilities of different there are close to 500 different public programs in tiers of government; and health, running through five levels o f government. 0 Weak and/ or ineffective Moreover, the failure o f these has led to around 80 systems for the "contracting" percent o f expenditure in the sector being with pnvate of service providers to provlders. H o w then i s India to renew its commitment to provide these services. the decentralization model and take it forward? The first step i s to define what local governments are supposed to do. This means defining roles vis-A-ms the state andthe center, and what i s to be done by the public sector and the pnvate sector. Most Indian states have made an attempt to devolve functions to local governments based on the subjects listed in the Constitution. In fact, most have devolved all 29 subjects. But this process has runinto problems for a vanety o freasons: Inadequate analvsis: There has not been enough analysis o f which services should be devolved, based on welfare economics considerations and the needto create accountability Deciding which level o f government should do what involves concerns about economies of scale and scope, capacity, externalities, national or local nature o f public good, and heterogeneity The creation o f accountability involves concerns such as monitonng, enforceability and proximity These are different for different services. Wholesale devolution o f functions: Functions such as pnmary education or sanitation were transferred to panchayats wholesale. But in India this i s irrelevant since public management i s carried out at the level o f activity or sub-activity (schemes or budget items). For example, most states have assigned responsibility for basic education to local governments. But the key activities and sub-activities that will deliver basic education, such as school construction or teacher hinng, have been left under the state line agency Ifthe assignment o f roles and expenditure powers i s to make sense, administration sectors need to be unbundled into activities and sub-activities. Assignments to different tiers o f government needto take place at this level o f disaggregation. l3Matthew,2000; Aiyar, 2001. GO1Task Force Report, 2001. 12 Overlauping: Most o f the devolved subjects have ended up in a concurrent list. Different tiers o f government share responsibility for the same things. These different levels may have a role to play in the provlsion o f key semces such as pnmary education. (The center or state may define the cmculum, the distnct may hire the teachers, and the village may repair the schools.) But different levels carrying out many o f these activities simultaneously (eg. repainng schools) has led to wastage o fresources and confused lines o f accountability Lack o f aupropnate involvement o f the unvate sector: India has taken important stndes towards pnvatization and liberalization in the past twenty years. But these developments have been ignored by the devolution process. Devolution has taken place entuely inline with the traditional predominant role o f the state in economic activity It i s now well recognized that wholesale privabzation has, in some instances, led to the collapse o f semces. But the ground reality suggests that the pnvate sector is, and can be, involved in the provision o f semces that are essentially private goods. Examples o f such services include operating tree nursenes or prowding curative health services. The studv framework W e have seen that there is a need for improvec. semce delivery in rural areas in the present Indian context. Such a need necessarily involves strengthening the mechanisms o f decentralization to local governments. It i s in this context that this study proposes a framework to assist This study addresses the issue of the GO1and state governments in identifying the roles institutional arrangements and the for the vanous levels o f panchayats. The approach roles of different levels of implied by such a framework i s then demonstrated by government in Providing services to this study for a small number of sectors that have been devolved topanchayats. The approach is, however, applicable to any of the other sectors eligible for devolution. Focus on Four States As discussed, the study focuses on Kerala, Karnataka, West Bengal and Rajasthan(see Box 1.3). Kerala has shown the deepest rural decentralization; Karnataka has taken a comprehensive approach to its decentralization exercise; West Bengal has been decentralizing sector by sector; andRajasthan has shown a slower move toward decentralization. 13 Box 1.3: The Four rtafes under study education and health. Even if the pnvate sector can, in pnnciple, provide these semces, it is unlikely that the government can escape assuming responsibility incases o f fai1~re.l~ Criteriafor analysis This study aims at malung recommendations on responsibilities within the decentralized government structure in rural India. For example, how do we decide how to allocate responsibilities to the vanous levels o f government for the array o f unbundled activities involved inproviding a system o f pnmary education? Who should be responsible for buyingdesks? The state, the district, the block or the GP7 W e use two related frameworks that advance cntena for deciding which level o f government should do what. One cntenon stems from public finance and welfare economics, the other from the principaVagent problem elaborated inWDR 2004. Public finance critena There are four tradihonal public finance cntena for choosing the size o f the jurisdiction that should be responsible for a public semce: Economies o f scale or scope: Since the unit cost o f production declines as the scale o f production increases, the general rule in public finance i s to allow for a sufficiently large market and avoid inefficient productioncosts. Scope o f the externalitv. An "externality" refers to a consequence o f one person's action on another. The geographic extent o f an externality can range from very local (the impact o f one person's radio on his neighbor's sleep), to the global (the impact o f one country's emissions o f greenhouse gases on another country's climate). The general rule in public finance i s to make the geographic extent o f the junsdiction coping with a problem large enough for the "external" effect to be "internalized" inthe junsdiction. For instance, ifadjacent cities share a mer, sewage dumped by one affects the cibzens o f the city downstream; while if it were one city, the effects are "internal" to the city Eauity Education is a good example o f this welfare economics cntenon. In India, for instance, the goal o f universal elementary education i s enshrined in the Constituhon, the law, and the policy pronouncements o f both central and state governments. Thus the system being designed cannot, intheory, be plannedto be less than universal. Heterogeneitv o f demand: Decentralization i s expected to improve service delivery when it allows a better correspondence between local conditions and preferences, and the activities undertaken by the government. Accountability cntena Accountability is a relationship among actors that has five features: delegation, finance, performance, information about performance, and enforceability Relationships o f accountability can be as simple as buying a sandwich or going to a doctor - and as complex as running a democracy For example, in buying a sandwich you ask for it (delegation) and pay for it (finance). The sandwich i s made for you (performance). You eat the sandwich (which generates relevant information about its quality). And you then choose to buy or not buy a sandwch another day (enforceability). Similarly, in going to a doctor, you go to the clinic to be treated l4California's recent electricity crisis IS one such example 15 e Elementsof t ability Relatians There are Five Features to Any Accountability Relationship ~- - --- -What . . ExampleI: Example2: Buying a Sandwich ' . Goingto a.poctot -I - - 1\ J l n f I~ the `agent' hasdone the which informsyou of its better you assess the - work ~ ~ ~ ~ ~ quality ~ e ~ oof the~ ~ n ~ r ~You~ gorto himnext Enforcing You reward goodpunish 0 You choose ~ ~ eto t bad ~ ~ oand m ~ nbuy~a the next time, $ a ~ ~ w ifrom the c h ttme (if he was good) r seller or chooseto go a ~ e c ~hisnorofits i ~ $omew~ereelse if not 16 maker must be able to transmit these demands to the actual prowder o f services and to make sure that the incentives for these providers are aligned with the ultimate preferences, or well-being more generally, of the citizens. The minister o f health does not personally give vaccinahons. She sets up rules, personnel policies, issues directives, payment and management systems, etc., to have these vaccinations done. The vaccinations themselves are given by real people with real constraints and preferences o f their own. The tnck is to ensure that the lncentives to the prowder continue to reflect citizens' interests. This two step process (i.e i s people to Government, and Government to service providers) and its companson with the direct transaction through a market was the essence o f the WDR approach to accountability Figure 1.8 outlines the relationships o f ac~ountability'~.Each o f the three accountability relationships (voice, compact, and client power) have the five features: delegation, finance, performance, information and enforceability In what ways do relationships o f accountability decide how Of service is Seen as public services should be delivered? Elaborating on this failures of accountability. set o f accountability cntena, WDR 2004 argues that failures in semce delivery can be traced to the weak accountabilities in the current system o f institutional structures involved inthe delivery o f a service. This systemic approach implies that attempts to improve education or health that deal only with "proximate determinants" o f learning, and expand the "business as usual" format, are bound to have only marginal impact. Ifthe system creates no incentives for effectwe learning, then n o budget increase or new knowledge i s likely to improve conditions. l5Inthis picture, lines of accountability are illustrated for the two ways to make sure that services are delivered. Services, o f course, go from provider to clients/citrzens. But the accountability mechamsms to ensure these services get delivered can go m one of two ways: directly to the provider as in a market or through the state (the policy-maker inthe diagram). W e call these the "short-route" and the "long-route'' o f accountability respectively. 17 Long routeaf accountab Short route of To ~ o ~ p lthings~~ ~ ~eh~e ~ .~eleot~onsiare ycontesxcd on a party basts, and there are atso i ~ u ~ r c l ~ ~ i o n s of~a~csc ~ ~ n t abetween~parry officers and between party ofiicers and politicians. h b i ~ ~ ~ 18 politicians is the greatest at the center and the state levels. The accountability of the pnvate providers and front line civil servant prowders is pnmarily to state civil servants. This breaks down the accountability chain. Figure 1.9: Accountability relationshipsfor rural people: Can itwork inIndia? ' ------- Providers Servant Providers The arguments so far have pointed out that when the accountability arrows do not connect, services are likely to break down. This discussion points at putting the services at the lowest level (where voice i s stronger) that can provide the service efficiently, make the providers accountable to that level. This is often called the pnnciple o f subsidianty But this still does not help us in deciding the level of government for service delivery To help us be more precise on 19 our recommendations we introduce three additional concepts: discretionary services and transaction intensive services, and observability Discretionarv services: The process o f semce delivery calls for malung discretionary decisions that are crucial to a successful outcome. Semces are discretionary to the extent that their delivery requires providers to make decisions on the basis of important information. It i s inherent to such information, however, that it i s imperfectly specified and incomplete, rendenng it impossible to mechanize. The nght decision depends on conditions that cannot be assessed easily, malung it difficult to evaluate how appropnate the decision is. Thus these decisions provlders have to make usually entail extensive professional or informal context-specific knowledge, gained vla training and/ or experience. Thus discretionary activities are those that require decisions to be made using individualjudgments inlocal context. Transaction-intensive semces: Transaction intensiveness refers to the extent to which the delivery o f a semce (or an element o f a service) requires a large number o f transactions, nearly always involvlng some face-to-face contact. School lunch programs, for example, require numerous cooks and cleaners to be present every day to prepare and distribute hundreds o f meals in a hygienic environment. Incontrast, a small committee can draw up the monthly menu at a single meeting. Observabilitv. At what level can performance be best observed? if the activity i s technically complicated, it requires `expert' (technical) evaluation, whereas if it i s simple and local, it requires local-level evaluation. Based on these definitions, we can come up with a two-by-two matnx, which classifies activlties as a mix o f transaction-intensive and discretionary So, whereas central banlung 1s In general, and other things being equal, the highly discretionary (since requires significant more and the moretransaction judgment calls from time to time), it 1s not intensive the activity, the appropriate transaction-intensive (it can be done by one good it is for a lowerlevelgovernment for Of central banker sitting inhis office). Immunization This is because the effects Of campaigns, on the other hand, are highly Services are much better observed at that transaction-intensive (they require thousands o f level, and can a lot health workers administenng millions o f vaccines situations at the local level), but are not discretionary (since they require a simple pre-packed vaccine to be delivered to each individual child o f appropnate age through a straightforward procedure). A service like ambulatory curative health care i s both discretionary and transaction-intensive. This discussion i s summanzed in(see Table 1.2). 20 1 ---- I I. I I II " 7 I I-ill" "I I Discretionary Transaction- .-.e? a??.! I Inferring .- Intensity . I an??_' I I ---- "l.ll_l --". Doesthe Service require decisionsmade in local context? (Discretionary) I many local I transactions? ! I (TransactionInte Activities that are Discretionary, Transaction- Implication Intensive and Locally Observable require local - .___ input for accountability I Assessing groundrealities The third step i s to address the question o f how services are actually being delivered, that is, what are the de-facto roles. Chapter IV investigates the behavior of agents engaged inservice delivery, including civil servants, politicians, and providers at different levels, to decide whether actions are consistent with legislation. Applying;first pnnciples The fourth step i s to apply the first pnnciples discussed above to provide some guidance about who should provide which semces. As discussed, Chapter V relies on two main bodies o f literature, one stemming from traditional welfare economics and public finance, and the other from the emerging literature on accountability, in particular the WDR 2004. The chapter also takes into account some internationalexpenences, eliciting lessons where possible andrelevant. Drawing conclusions The concluding chapter, Chapter VI, summarizes findings and analyses, establishing linkages that lead the way to recommendations -- onthe roles of the center, states, distncts, blocks and villages in the provision of pnmary education, health, rural water and sanitation, and employment programs. The analyticalprocess that leads to the recommendations i s summarized inFigure 1.10. 22 Figure 1.10: An Analytical Approach to Effective Decentralization Allocation and ecommendations What are the key functionsand these functions activitiesInthe and activities 8Disaggregateinto 8 Use First 8Analyze dejure Based on this functions and Principlesof Public functional allocation systematic activities (consistent Financefind optimal of eachtier by analysis, propose with the Indianpublic allocationfrom public studying existing changes to current administrationdelivery finance perspective legislation system system): Standards 8Use First Principles 8Analyze de facto Develop a *Planning of Accountabilityto functional allocation detailed proposal OAssest Creation address the (on-the-ground for reallocating *Monitoringand accountabilityissues picture) by responsibilities Evaluation conductingsurveys within different and interviews tiers that draws on the findings of our analyticalapproach 23 11. MappingActivities: UnbundlingRuralServices Unbundling services into activities helps to analyze reality and to develop a model of an optimal institutional system for service delivery. These activities must be mutually exclusive, and logically chosen on the basis of technical recommendations. Basic definitions: sectors, services and activities The exercise o f brealung down services into their components - or unbundling - calls for some basic definitions. Sectors: A sector refers to a collection o f semces that share aspects o f a technical discipline and are expected to lead to a certain outcome in terms o f human well-being. Rural educahon, health, drinlungwater and sanitation, and welfare are sectors; the outcomes these sectors are expected to 'have are healthy, educated, and less poor people. Services: Services are components o f the sector, separable in the sense that each service can contribute positively to the sectoral outcome even in the absence of other semces. Thus rural health would include semces such as immunization, vector control, pnmary curatwe care and secondary curative care (hospitals). Even without other services, immunization would improve health outcomes, though not as much as it would if these other health semces were also present. Education would include services such as primary, secondary and tertiary education and vocational training. Activities: Services are made o f activities. These cannot be separated from each other; unless all are camed out, an outcome o f the desired quality i s unlikely One example o f an activity i s the buildingor repair of schools -essential for the provision o f quality education. This classification is, o f course, arbitrary and for ease o f exposition. Buildinga Unbundling a service into school could also have been called a service. Thus it 1s activities should be down to a apparent that defining an activity i s not easy Activities, as level that enables those in used here, are very sector-specific. For example, operating a charge to produce a determined local health clinic i s an activity specific to health. There are many different ways o f unbundlingsemces into activities, depending on the charactenstics o f the service and the purpose o f the exercise. But it i s useful to identify identical activities across different sectors. Since our ultimate objective is to improve outcomes, the unbundlingo f a semce into activities should be down to a level that enables those in charge to produce a determined output or product. For example, if one necessary output for an education outcome i s a school, then one o f the unbundled activities i s building and repainng schools. Most activities in all sectors can be grouped inthis way into five activity groups: Policy and standards; Planning; Asset creation; Operation; and M&E. The next question is how these five categories can be appliedto each sector (See Table 11.1). 24 Standardsctting identity Priority Vtllages iectarrenx 25 I Is rtf 36 Key sectors: services and activities Primary education W e define pnmary education as a service. The first requirement o fpnmary education i s a teacher. The teacher must know the goal -- what it isthe child is to learn, i.e. the cumculum. The teacher must also knowthe subject matterto be taught;havemasteryo f at least one effective technique o f teaching the material; be able to assess whether a student has mastered the material; and be motivated to assist the child's learning. In addition, the teacher must be supported wth physical facilities adequate to the learning process, as well as instructional matenals. Building on the essential interaction o f child and teacher in the context o f elementary education leads to the classification o f the basic activities that any provider must undertake to deliver an educational or instructional semce. These activities can be mappedinto the five activlty groups identified earlier (Table 11.1). The process o f producing education i s a complex one, and there i s no reason to believe that each o f the elements should be camed out at the same level o f responsibility A discussion o f whether the effectweness o f elementary education would be enhanced if it were the responsibility o f the state, distnct or the GP i s neither realistic nor analytically coherent, since these activities do not belong at the same level. At the same time, merely allocating concurrent responsibility to all levels o f government without clanty on roles i s also not helpful. In thinlung through the allocation o f responsibilities for elementary education, it i s worth keeping in mind that the government production o f elementary education is a very special case o f a broad class o f educational or instructional semces. In other words, this genenc description o f the learning process applies to the range o f instructional settings: a pnvate tutor teaching a language, a firm providing training to its own workers, a firmprovlding training invocational slulls (e.g. computer training), a surgeon learning a new operation, an athlete learning a sport, or a not-for-profit elementary school. Health services Rural health involves a wide range o f services involving vanous types o f "provlders." For the sake o f discussion, we will classify these services into population-based activities; preventive and promotive activlties; pnmary health care; hospital-based care; and surveillance. The first and second are aspects o f traditional'6 public health; the third and fourth are aspects o f curative care. The fifth, surveillance, i s the systematic collection and dissemination o f information on the health status o fthe population. Population-based services Examples include pest control ("vector" control for pests that carry communicable diseases); safe water guarantee (not necessarily provision); and urban sanitation. These functions are often performed by people who may not even think o f themselves as part o f the health system, engineers for instance. l6Inthe Western sense. 27 Preventive andpromotive semces These include traditional public health activities such as health awareness campaigns, campaign style immunization, nutrition monitonng, and identifying and counseling pregnant women. These activities are, or could be, camed out by paramedics, nurses, specialized health educators and related professionals. Both lunds o f traditional public health services tend to be "supply-driven" inthe sensethat pnvatedemandis eithernonexistent, or, atbest, inadequate. Pnmarv health care This consists o f curative activlties initiated by patients and therefore "demand-dnven." The delivery o f primary health has two major components: the provlsion o f pnmary health care facilities including PHCs/ CHCs and hospitals; and disease control through health education. The keyactivities that determine the provision o fhealthcare facilities are: 0 Designing the critena for the location o f facilities; identifying the budget to run these facilities; and pncingpolicies for operation and maintenance (O&M); 0 Planning o f facilities; placing specialized equipment; developing operational protocols; and determining the eligbility o fpatients, and rates o f subsidy, ifrequired; 0 Creating assets, including constructing buildings and purchasing equipment, as well as investing inhuman and social capital; 0 Conducting the day-to-day operation o f facilities, including the purchase o f matenals, drugs and other consumables; systems for patients intake, registration and record keeping; referrals; and maintenance o f equipment, both specialized and otherwise; and 0 M&E to ensure quality medicaladvlce andpatient satisfaction. Hospital-based care Like pnmary health, this i s initiated by patients and i s demand-driven. The division between primary health and hospital-based care i s fuzzy, but it can be thought o f as "inexpensive" and "expensive" inways that we will clanfy later. Both are provided by medically trained personnel. Surveillance Generally speakmg, surveillance i s for higher levels o f government - state and central. In cases o f epidemic control, this requires the coordination o f surveillance activities. More important perhaps i s providing information to local decision-makers -- regarding the relative health status indicators in their areas, and research on the determinants o f health status and health system performance. This i s a role that has not been played systematically in India. But since this has emerged as a pnority in the roles o f central and state governments as they shift out o f responsibilities better left to PRIinstitutions, this will need detailed discussion. Rural water and sanitation Similar services are neededto deliver high quality rural water supply and sanitation (RWSS). The key activities in the delivery o f drinkingwater include the development o f water supply systems; their O&M; and the monitonng of their implementation. The essential sub-activities include the formulation o f specific water supply schemes; technical appraisal and approval o f these schemes; awarding contracts for the execution o f these schemes; monitoring and supervision o f the progress including quality control; establishment o f water testing laboratories and control o f 28 chemical and biogenic impunties; maintenance of water supply systems through the collection o f water charges from the users; and vigilance against illegal tapping and misuse. Similarly, in the sanitation sector, the key activities consist o f generating public awareness o f the use o f sanitary latnnes; providing resources to the needy to construct appropnate sanitary latnnes; designing and implementing policies on the provision o f subsidies; and maintaining common facilities such as community latnnes. Employment programs Two categories o f employment programpolicy Public policy aimed at employment programs can be divided into two distlnct categones. The first is commonly referred to as indirect policies or "growth mediated policies" that aim to accelerate growth on the one hand, and direct growth to benefit the poor on the other. Policies that direct public expenditures towards social sectors and subsidies fit in this category The second and more relevant category i s referred to as direct interventions or "support-led secunty" programs that aim to directly reduce poverty by enhancing the consumption entitlements o f the poor inthe short run.The collection o f these programs makes up the welfare sector. Policies that focus on self-employment and wage employment fit in this category These programs are narrowly focused and explicitly target the poor, inparticular the poorest o f the poor." Inline with these definitions, self-employment and wage employment can be called two services o f the welfare sector. Our special concern i s the wage employment sub-sector, which in India is dominated by the SGRY, the largest wage employment program currently implemented by the GOI. The components o f an effective employment program make up the key activlties: Targeting: Programs need to ensure that the funds allocated reach the intended beneficianes, usually the poorest o f the poor. Histoncally speahng, the inability to reachtarget populations has been the most significant problem o f employment programs of India.'* The debate on direct targeting versus self-targeting is also an example o f the cntical role that targeting methodology plays in employment program^.'^ Thus who and how to target, and whether to target at all, i s a key activity Choice o f activity through which incomes are to be enhanced: The activities chosen need to be labor-intensive to maximize the outreach o f the program. Like targeting, this i s an area where employment programs in India have failed.*' Strengthening rural infrastructurethrough asset creation i s a cntical secondary objective in most employment programs. The choice o f activlty i s thus crucial in ensunng that the activlties selected for infrastructure development reflect local needs and pnonties. The creation o f human and social capital through skill development and information dissemination: Effective information dissemination ensures that potential beneficianes are aware o f schemes so that they are able to partake o f them. This i s particularly important in self-targeting schemes. In programs where strengthening infrastructure i s a secondary objective, the process o fbuildingphysical capital is also cntical to a successful outcome. l7Rao, 2002; Snvastava, 2004. IntheJawahar rozgar yojana (JRY)program, for example, only 17percent of the totaljobs created reachedwomen as against a target of 30 percent. See Srivastava, 2004. ''See, for example, Imam,2000, and Datt and Ravillion, 1994. Central government evaluations of JRY and the Employment Assurance Scheme (EAS), for example, indicate that asset creation was not labor-intensive and involvedhigh material costs. See Srivastava, 2004. 29 0 Effective monitoring. and evaluation: Whether at the institutional or client level, effective M&E is cntical to any policyintervention. Employmentprograms are noexception. 30 111. The Legal Framework: Implications for the Different Tiers of Government The 73rd Amendment set in motion the process of institutionalizingPRIs in the Indian governancestructure. Through a series of mandatoryand recommendatoryprovisions, the 73`d Amendment set the parameters for PRIs to fulfill their role as "institutions of self- government." But actualpractice hasvariedacrossstates becauseof 0 Non-mandatorydevolutionoffunctions andoverlapsacrossgovernmenttiers; 0 Concurrencieswithinthe law; 0 Fundsnotfollowingfunctions; and 0 Lackof administrativecontrol. These implications of the Amendment, and other legislation such as the state Panchayat Acts, are apparent across the key sectors of primary education, health, water and sanitation, and employment programs. Also, though centralschemes aim to work within the decentralizationprocess, they tend to be implementedwith little or no involvement of the PRIs. The historical background At the time of its inception, the Indian Constitution envisaged a two-tiered federal structure of governance with executive and legislative powers shared between the center and the states. The Constitution did not assign any specific promions for panchayats. The role ofpanchayats was, however, arhculated: "the state shouldtake steps to organize village panchayats and endow them with such power and authonty as may be necessary to enable them to function as units of self government."" Later efforts attempted to revive Punchayah Raj, the traditional system o f local government inIndia. The first substantial effort, the Balwantrai Mehta Committee (1957) made recommendations on the devolution of power and resources to PRIs. However, its recommendations were not effectively implemented. The Ashok Mehta Committee (1978) recommended a two-tier system of Panchayati Raj with the block as the base, and the distnct as the key administrative unit for planning, coordination and resource allocation. The Committee also recommended the devolution o f financial powers to PRIs. However, it was not till the Singhvi Committee (1986) that the constitutional recognition o f panchayats became a serious issue for the government. The recommendations of this committee led to the introduction of the Panchayati Raj Bill in parliament in 1989 The billwas passedinthe Lok Sabhabutrejected inthe Rajya Sabha.A Joint Parliamentary Committee was then set up to prepare a revised bill. In 1992, the new bill --the 73rd Constitutional Amendment -- was introduced in the Indian parliament. The bill was finally passedin 1993, providing an exclusive legal basis forpanchayats. Article 20 of the directive (or non-justiciable) principles of the Constitution. 31 The startingpoint:the criticalprovisionsof the 73rdAmendment Theheart of the legalframework W e have seen that the PRIs are central to fulfilling the government's commitment to rural development. The 73rd Amendment - by virtue o f its according legal status to PRIs - i s the starting point o f any analysis o f the institutional arrangements for service delivery. What i s the nature of The 73'd Amendment of 1993 to the role (and limits) the Amendment envisaged for the the Constitution gave legal status PFUS? to PRIs. Thus the Indian government made a clear A legal definition o f panchavats: The 73rd Amendment commitmentto decentralizationas defines panchayats as institutions o f self-government that the basic institutional structure are elected for a period o f five years on the basis o f for development. universal adult franchise. Transfer o f subiects: The 731dAmendment assigned key development subjects to PRIs-including primary education, pnmary health, water and sanitation, and employment programs. The process o f transfer, and the specific role for PRIs in these subjects, was left to individual state governments. Box 111.1provides details on which subjects have been transferred to PRIs. 32 Box 111.1: The 29 subjects that states can devolve to local governments 1. Agnculture, including agnculturalextension. 2. Landimprovement, implementationof landreform, landconsolidationandsoil conservation. 3. Minor irngation,water managementandwatersheddevelopment. 4. Animal husbandry, dairymg andpoultry. 5. Fishenes. 6. Socialforestry. 7 Minor forest produce. 8. Small-scaleindustnes. 9. Khadi, village andcottage industnes. 10. Ruralhousing. 11. Dnnking water. 12. Fueland fodder. 13. Roads, culverts, bndges, ferries, waterways andother means of communication. 14. Rural electrification, includingdistribution of electncity. 15. Non-conventionalenergy sources. 16. Povertyalleviationprograms. 17 Educationincludingpnmary and secondaryschools. 18. Technicaltraining andvocationaleducation. 19. Adult andnon-formaleducation. 20. Libranes. 21. Cultural activities. 22. Marketandfairs. 23. Healthandsanitation, Including hospitals,pnmaryhealthcentersanddispensmes. 24. Familywelfare. 25. Women andchild development. 26. Socialwelfare includingwelfareof the handicappedandmentallyretarded. 27 Welfareof the weaker sections, and inparticularofthe scheduledcastesandtribes. 28. Publicdistribution system. 29. Maintenanceofcommunityassets. Provisions: Part IX o f the Constitution contains sixteen articles that constitute a complex of mandatory and recommendatory provisions for panchayats inrural India. It mandates the creation of a uniform three-tier system of local government -- at the village, block and distnct levels. Crucially, the Act accords legal status to the gram sabha. Inaddition, the Amendment deals with four mainissues: The mandatory and recommendatory 0 Composition o f panchayats: Articles 243C articles of the 73rdAmendment empowers state governments to make Definepanchayatsand create a provlsions for the constitution o f three-tier systemof local panchayats; Article 243D mandates the government; reservation of "not less that one-third" seats 0 Assign key subjects to PRIs; for SCs, STs and women on the basis o f 0 Mandateelections topanchayats and proportional representation. Reserved seats reservationsfor women, SCs and are rotated among the panchayat's different STs; and constituencies. 0 Address the continuanceof existing 0 laws relating topanchayats. Powers and authority. Articles 243G to 2435 deal with the functional and financial powers ofpanchayats. 33 Elections: Article 243K deals with elections to panchayats. It empowers state governments to appoint a State Election Commission, and stipulate, by law, the rules and conditions o f their service. Relationship with other laws: Article 243N addresses the issue o f the continuance o f existing laws with regard to panchayats. Thus any pre-Amendment legal provision relating to panchayats that i s inconsistent with the provisions o f Part IX o f the Constitution i s to be amended or repealedby the state legislature. The impactof the Amendment on the devolutionof powers Two readings o f the salient provisions: mandatoryvs. directory What are the constitutional provisions salient to the devolution o f powers to panchayats7 Article 2436, central to the Constitutional Amendment, articulates the powers, authonty and responsibilities ofpanchayats: ...subject to theprovisions of this Constitution theLegislature of state may, by law, endow a the panchayats with such powers and authority as may be necessary to enable them to function as institutions of self government and such law may contain provisions for the devolution of powers and responsibilities upon panchayats, at the appropriate level, subject to such conditions as may be speciJiedtherein, with respect to... Thepreparation ofplans for economic development and socialjustice; The implementation of schemes for economic development and social justice as may be entrusted to them including those in relation to the matters listed in the Eleventh Schedule. In legal practice, this Article has been interpreted in two distinct ways: that the provisions of Article 2436 are mandatory; and that they are merely directory, implying that the powers and authonty given to PRIs is entirely at the discretion o f state governments. The first interpretation i s anchored in the Article's use o f the term "self-government." The argument i s that if PRIs are defined as institutions o f self-government, they are autonomous institutions that can function without external interference. Thus all state governments are obliged to devolve functions to PRIs, allowing them to function as independent unitso f "self-government." This also means state governments are obliged to fulfill the recommendations o f Article 2436 in their entirety In a recent judicial intervention in the Andhra Pradesh HighCourt, Justice J. Raghuram argued along the lines that A state legislation that divests or withholds effective power of local governancefrom the elected representative of the panchayats would (be) tantamount to a legislation that subverts the Constitutionalpurposes of Part IX... The obligation of the state legislature to endow panchayats with powers, authority and responsibility is conditioned by the Constitutional instruction that the endowment of powers andfunctions must enable the panchayats to function as institutions of self-government. ...denuding the elected representatives of effective control over functions, finances, planning processes or personnel ... (leads to) disabling (and not enabling) the Panchayat Raj Institutions from functioning as institutions of Self government. "' The second interpretation - that the Article i s non-mandatory --was upheld by the Andhra Pradesh HighCourt in 2002. The argument here i s based on the use o f the word "may" inArticle 243G(a). The word "may" implies that the provisions o f the article are directory, and that state 22J RaghuramJudgment, AF' Judgmenton PRIs, 2002. 34 governments can set the limits on the role PFUsplay inrural development. This interpretation was also endorsed in 2002 by the National Commission for the Review o f the Constitution (NCRC), arguing that "Article 2436 along with the Eleventh Schedule indicates the kmdo f functions to be discharged by thepanchayats. It does not guarantee assignments o f an exclusive set o f functions to the PRIs.'"~ The impact o f the Act's non-mandatory nature Article 243G(b) provldes panchayats with powers to "implement schemes" that "may be entrusted" to them by state governments. The use of the words "implement" and "entrust" create the legal space in which panchayats can be interpreted to have powers to function merely as "implementing agencies," rather than as institutions o f self government. Moreover, panchayats only have powers over schemes devolved to them by the state legislature, not over schemes o f their own malung. Thus the language o f Arhcle 2436 clearly indicates that its promsions are recommendatory, and that the XIth Schedule i s merely an indicative list o f powers that state governments may transfer topanchayats. 23GOI, 2002c. 35 Box 111.2: The Ranga Rcddy Di iSrtite `Table 111.1: State-\\ise statu3 of the tle\dutiun of functions f 11 Ftinctioirs tft.voltvd i Andhra Pradesh 117 Artinclchnl Pradesh II 37 The concurrence o f powers could, inmany ways, prove to be enabling forpanchayats and ensure a more efficient delivery o f services. Schedule XI presents a range of subjects that require a complex interaction o f institutions at different levels o f government so that all schemes/ functions and matters related to the subject are implemented efficiently Consider the example o f pnmary education. For reasons ranging from economies o f scale to accountability, aspects such as cumculum design and monitoring need to be undertaken at the state and central levels. However, operational aspects -- such as school construction, procurement o f equipment and managing teachers --can and should be undertaken by local governments. In practice, however, this does not take place, and concurrence o f powers manifests itself as a problem inmost states. The impactof the StateActs: an analysisusingthe three F's W e have seen how cntical functions, funds and functionaries are to the role ofpanchayats inrural service delivery H o w has state-level legislation addressed the three F's? Functions Insome states, PanchayatActs havedevolvedto PRIsbroadfunctions relatedto subjects listedin Schedule XI. Butthis has been done without adequately considenng the specific roles o f each tier o f government inthe management and implementation o f the schemes and actiwties related to the subject. A task force o f the Union Ministry o f Rural Development (2004) states: "The functions devolved upon PRIs are in the nature o f `subjects' rather than in the form o f Activities or sub A c t i w t i e ~ . "In~other states, these subjects have been unbundled to their component parts, and ~ specific roles have been assigned to each tier based on the principle o f subsidianty2' This variation between broad and specific modes o f devolution can be illustrated by examining a few state acts. Broad devolution The Raiasthan Panchayat Act (1994): hjasthan's Panchayat Act mandates the devolution o f all 29 functions but in very general terms. Suppose we consider the example o f pnmary education once again. The Act assigns PRIs the following roles: 0 GP. "promotion o f public awareness and participation in village education committees for total literacy" and "ensunng enrollment o f boys and especially girls and attendance in pnmary schools and its management;" 0 BP. "running o f primary education including total literacy programs, especially girls' education" and "construction, repair and maintenance o f primary school buildings and teachers' quarters;" and 0 DP. "promotion o f educational activities including the maintenance o f upper primary schools." This broad devolution o f functions has meant that the state government has retained the power to negate any specific transfer o f functions and powers to PRIs. In 2004, for example, a government 24GOI, 2004a. 25The principle of subsidiantystates that what can be done best at aparticular level shouldbe done at that level andnot at a higher one. Ifthis principle is applied, the process of transferringfunctionsand powers should start from the level of the gram sabhas and Wards Comttees, going up to the Union Government. Only residual functions need get allottedto the higherlevel. 38 order o f the state government took away the DP's power to appoint teachers inpnmary schools. Insuch a situation, panchayats have no legal safeguards; the executive retains control over the functions assignedto PRIs. The West Bengal Panchayat Act (1973): This Act i s an example o f broad functional devolution that has created concurrencies in functional responsibilities across government tiers. Like Rajasthan, West Bengal has devolved all 29 subjects to the PRIs, again in general terms. Significantly, the Act mandates a similar role for BPs and DPs. A s in Rajasthan, the absence o f detailed devolubon o f subjects in the form o f actiwties or sub-activities across government tiers has led to the overlap o f functions, and ambiguities in the specific role to be played by different tiers. This in turn has meant that the state government has the power to assign and negate functions and powers to PRIs at its own discretion. Specific devolution Kerala and Karnataka: These states present a different story Both states have systematically attempted to unbundle the broad subjects listed in Schedule XI and devolve them across government tiers using the pnnciple o f subsidianty In Karnataka, this process i s still to be integrated into the Panchayat Act.26 But in Kerala, the Panchayat Act (1994) articulates the specific role that each PRI ought to play, and functional overlap is minimized. In the case o f primary education, the Act assigns the following: In the absence of exclusive Powers to the 0 GP. the management of government pre-primary panchayat, there can be no conflict schools; between the panchayat and a body exercising the same powers. 0 BP. no role assigned (in line with the pnnciple o f subsidiarity); and DP. the management o f government high schools (including upper pnmary schools). Contradictions and ambiguities: empowering parallel institutions There can be no legal conflict between the panchayat and a parallel body exercising the same powers. Instates where the Panchayat Acts have devolved certain subjects from Schedule XI to the PRIs, it could be argued that any existing state law that contradicts this would be in violation o f the Panchayat Act and must be repealed. But this analysis does not hold from the perspective o f junsprudence: the mere existence o f two separate laws empowenng parallel institutions with the same functions cannot be said to render the two Acts in conflict. The Supreme Court o f India has upheld this understanding o f the law -- any legislature i s entitled to enact two separate laws provlding for two different sources o f power. Ina case which dealt with a potential inconsistency between a state and central Act (La1 vs. the State o fUttar Pradesh, 1984), the Supreme Court held that "...there is no legal bar to creating two sources of power And there is no authority in principle or precedent for contending that one source of power is more valid than the other.'I2' The court argued that one o f the tests o f inconsistency i s whether the provisions o f one Act can be followed without disobeying the mandate o f any other law In other words, the provisions o f Schedule XI and the State Panchayat Acts would be in conflict with other laws only if the Schedule and Acts confer exclusive powers to thepanchayat. 26 It currently existsinthe form ofa governmentorder. 27 Supp SCC28, 1984. 39 The implications o f contradictions and ambiguities The presence o f these contradictions and ambiguities in Articles 2430 and Article 243N has meant that inmany cases, functions devolved to panchayats are also governed by state or central legislations. For instance, pnmary education in West Bengal i s governed by two main Acts, the Primary Education Act and the state's Panchayat Act. The Pnmary Education Act focuses on the developmental as well as regulatory aspects o f the delivery o f pnmary education. Its key objective i s to "make better prowsion for the development, management and expansion o f pnmary education wth a view to fulfilling the goal o f universal, free and compulsory education."** Among other things, the Act mandated the creation o f the West Bengal Board o f Pnmary Education, a body responsible for most o f the key functions relatedto pnmary education, ranging from settingpolicy standards and asset creationto operational functions. In1994,theWestBengal StatePanchayatRaj Act (1973) was modifiedtobnngitinconformity with the prowsions o fthe 73rdAmendment. Pnmary education was one o fthe functions devolved to the PRIs, and the Panchayat Act states that the "the Panehayat functions are also governed gram panChayat shall ah0 perform other fUnCtiOnS as by state/ central legislations. With the the State may assign in respect of primary exception of Kerala, which has chosen ed~cation."~~role has been enwsaged for the DP A an enabling interpretation of and the BP as well -- they have the authority to legislation, concurrency is prevalent. "undertake schemes or adopt measures including the givlng of financial assistance relatingto the development o fpnmary ed~cation."~'Thus the West Bengal Board o f Pnmary Education, the state department for education, as well as the PRIs, are legally empowered to perform similar functions inpnmary education. The problem o f concurrency i s prevalent inmost states. ButKerala has chosen to interpret Article 243N inan enabling way Based on the recommendations o f the Committee on Decentralization o f Powers (1997), Kerala amended its Acts to eliminate concurrency over powers devolved to panchayat~.~~ many as 44 Acts dealing with issues ranging from entertainment tax, land As development and education were amended, and the Kerala State Development Board Act (1971) and the Kerala Hackney Carnages Act (1963) repealed. Insummary. devolution offunctions Insum, the general case is that functional allocations overlap across government tiers, leadingto confusion inthe specific role to be played by each tier of government. The devolution of subjects to PRIs has been undertaken subjectively, with different states choosing to adopt different perspectives on the role o f PRIs in development. The specific role o f each panchayat tier in matters pertaining to the subjects devolved remains ambiguous. In practice, this has generally meant that the state government takes over the responsibilities o f matters listed in Schedule XI despite the State Panchayat Acts mandatingthe devolution o f subjects to the PRIs. 28West BengalPrimaryEducationAct, 1973. 29West BengalPanchayat Raj Act, 1973. 30Ibid. 3'The amendmentswere formally accepted inMarch2000. 40 Funds The financial Dowers o fpanchavats: conditions and limits For PRIs to perform the functions assigned to them effectively, they must be fiscally capable and autonomous. Funds must follow functions. The possible sources o f panchayat revenue include taxes, duties, fees and tolls, as well as grants-in-aid from the state. Tax assimmentsto panchavats Fiscal powers: Article 243H, which addresses the IfpRIsareeffectively, issue of tax assignments topanchayats, enables the functions to perform their assigned they must be fiscally capable and autonomous. Funds state legislature to authonze and set up procedures must follow functions. But as in the case through which a panchayat may levy and collect of functions, the directory nature of taxes. It also does the same for grants-in-aid from Article 243H allows state governments the consolidated state fund; and for the creation o f a to determine the "conditions and limits" panchayat-level fund that regulanzes the flow o f on the financial powers ofpunchuyuts. panchayat funds. Again, the role o f the state government in assigning financial powers to PRIs i s conditioned by the use o f the word "may" in the Article -- highlighting its directory rather than mandatory nature. And again, the implications for the panchayat level are similar. State governments have complete discretion over the financial powers to be devolved to the panchayats; and, in most cases, these are limited. On examining the condition of panchayat finances across India, the NCRC concluded: "By giving blanket powers to the state government, the said article has been made practically sterile. It i s not capable to serve its purpose since the state governments do not want to share their fiscal powers with the local government instit~ltions.~'~' Variations among states: Not surpnsingly, there are wide variations in the taxation powers o f PRIs across states. Rajasthan, for example, has a Panchayat Act with weak taxation powers; unlike Kerala's, which has a strong tax base. Though Rajasthan has devolved all 29 fimctions to the PRIs, the state's Act assigns only areas with low revenue potential to panchayats. The GP has the powers to impose taxes "subject to the rules and orders made by the state government in this behalf."33 In contrast, the Kerala Panchayat Act demonstrates an attempt to devolve an innovative list o f taxation powers to GPs. The GPs can collect property tax, professional tax, entertainment tax and advertisement tax. Inaddition, the GPs receive a share inthe taxes imposed bythe state government -- such as vehicles tax, stamp dutyand landtax. Grants-in-aid from the state Level of financial devolution: Arguably, grants-in-aid from relevant state-level ministries comprise an even more important source o f revenue for PRIs. We have seen that the taxing powers o f PRIs are limited in most states. And even in states such as Kerala where PRIs have a relatively strong tax base, the tax revenue o f PRIs barely covers establishment Yet again, the non-mandatory nature o f Article 243H means that the state government decides the extent to which relevant ministnes devolve finances to PRIs. The result, predictably, is that in some states, 32GOI, 2002c. 33RajasthanPanchayat Raj Act, 1994. 34World Bank, 2004a; Hadal, 2004. 41 Finance Commission, Article 280 was amended inconjunction with the 73rd Amendment. Article 280(bb) was introduced to mandate that the Finance Commission recommend "the measures needed to augment the Constitutional consolidated fund o f a state to supplement the resources o f the panchuyats inthe state on the basis o f the recommendations o f the SFC."36 Perhaps the most significant aspect o f Article 2431 is its use of the word "shall" -- indicating the mandatory nature o f the provlsion. Consequently, all states inIndia have formed SFCS.~' The impact o f the SFC mandate: Given the mandatory nature o f Article 2431, what has been the expenence in practice? Assessments indicate that implementation has been fraught with problems. 0 Most states have expenenced delays, both in constituting SFCs and submitting SFC reports. 0 The SFCs lack a common approach in the The innovative idea of setting methodology for malung recommendations. In most up SFCs has had problems in implementation - cases, they have confined themselves to the existing delays,the lack of a tax shanng structure.38 approach, and the non- From a legal perspectwe, part o f the problem IS that mandatory nature of their the preceding Article 243H, articulating the financial recommendations. capabilities o f the panchayats, i s non-mandatory. This allows SFCs the flexibility to not assign taxes from the state level to local bodies. The result i s significant dispanty inthe recommendations o f the SFCs. While some SFCs (West Bengal and Karnataka) have chosen to make a concerted effort to strengthen the fiscal capabilities o f panchayats, others (Rapsthan and Uttar Pradesh) have chosen to provide minimal finances topanchayat~.~' 0 While Article 2431indicates that setting up the SFC i s mandatory, it does not do the same for the implementation o f SFC recommendations. This has meant that these recommendations have been rejected by the state legislatures in some states.40 The impact o f central schemes on panchavat finances Any discussion o f the legal framework for panchayat finances would be incomplete without an examination o f the impact that CSS have onpanchayats. The nature o f Indian fiscal federalism grants the central government most o f the critical revenue raisingpowers while the state governments are responsible for the majonty o f expenditures inkey development areas such as education and health. To compensate for this financial imbalance, the Constitution allows for certain statutory provisions to transfer resources from the center to the states. A In fundsfor subjects css, proportion o f central revenues are thus transferred to to come fromthe states through the Consolidated Fund o f the States. center, and they are tied to These are untied grants to the states distributed guidelines that often bypassstate according to a formula incorporating measures o f need legislation or the role ofcss can and backwardness. Another provision allows the pancltayats.In thisway, center to give the states grant~-in-aid.~'Since the contributetothe problemof funds not followingfunctions. 36The IndianConstitution. 37www.Punchavat.nic.in,website of the Ministry ofPunchuyuts, GOI. 38Aiyar, 2001, GOI, 2002c; World Bank, 2004a. 39GOI, 2002c; Oomen, 2003. 40Aiyar, 2001, GOI, 2001b. 4'As articulatedinArticles 275-280 ofthe IndianConstitution. 43 1970s, these have taken the form o f CSS that transfer funds from the relevant line ministnes at the GO1level to state governments, to implement programs based on guidelines prepared at the center. Most of the CSS have to do with the subjects listed inthe state list and Schedule XI; and inmost cases they dominate the plan component of state budgets for these respective subjects. A recent report (2004) by a GOI-appointedtask force pointed out that there are as many as 154 CSS that deal with subjects devolvedto PRIs. For example, for 2002-0342,58 percent o f the Plan expenditure inelementary education came from the Center (mostly through Centrally-Sponsored Schemes), with only 42 percent coming from the States andUTs.Incontrast, the States paid for almost the entire Non-Plan expenditure. This is shown inFigure111.1below Considenng particular states: If, for example, we analyze the spending on Rural Development in West Bengal (excluding `Grants for Salary and Allowances'), we find that 61.9 percent o f this comes from the Center (again mostly inform o f CSSs), with the State puttinginonly 38.1 percent. This i s shown inFigure III.2 below Figure111.1:Plan and Non-Plan Expenditureon Elementary Education Plan Expenditure on ElementaryEducation II Non-Plan Expenditure on Elementary Education (2002-03BE, in lakh rupees) I (2002-03BE, in lakh rupees) I I Centre: I 0.01% I I I I I I ; + I 13%1 I I Source:World BankAnalysis based on estimates inAnalysis of5udgef Expenditure on Education,Minmfry of HRD, GO/ 42Budget Estimates, source: World Bank analysis based onAnalysis of Budget Expenditure on Education, Ministry of Human Resource Development, Government of India 44 Figure111.2 ExpenditureonRuralDevelopment-West Bengal State Expenditureon Rural Development:West Bengal (Released 2003-04, in Rs. lakh) "Schematic FJnds": Centre 63.0% Total :entre 61.9% Centre State Note: includesall `Schematic Funds' under `Panchay and Rural Development Departmenl and `Other Grants', excludes `Grants for Salaryand I wances' Thus! espite the increasing rhetonc o f federalism and decentralization, the Center continues to - exercise significant discretionary decision-malung powers on where and on what state governments can make incremental spending decisions. Inmost cases, CSS donot comply withthe devolution offunctions providedbythe state-level Panchayat Acts. As a result, even instates where all 29 functions have been devolved in accordance with Article 2436, fund flows ensure that the administrative arm of the state government continues to perform these functions. Inessence, funds do not follow functions. Functionaries The limited administrative control o f Danchavats Along with the transfer o f functions and funds comes the issue o f administenng them. If panchayats are to fulfill their role as agents of development, they needthe ability to participate in the process as administrative functionanes. The 731dAmendment does not explicitly address the issue o f panchayat functionanes so that the decision i s left entirely to the state governments. Currently, all administrative obligations o fpanchayats are fulfilled through staff deputed by and accountable to the state government. This raises two issues: 0 Just as in the case o f funds, there is a disjoint between the transfer o f functions and functionanes. Panchayat Acts in states such as West Bengal and Karnataka do allow PRIs to hire functionaries up to a certain level, to manage their day-to-day affairs. The West Bengal Panchayat Act gives power to the GP to "appoint such officers as may be 45 required by it" but subject to "rules as may be made by the state government." Thus the state government indirectly controls even the limited freedom o f the GP over its staff. In Kamataka, the State Act allows GPs to recruit their own staff provided they are paid for through the GPs' "own revenues." This naturally limits the GPs' powers to appoint staff since own revenues are limited. 0 PRIs have no authority over the deputed staff, including the authonty to hire, transfer or take disciplinary action. The issues o f recruitment and dismissal are determined by Articles 310 and 311, which grant these powers to the state governments. Thus the panchayat has no administrative control over its employees.43 The impact o f the legal framework on key sectors We have seen that the constitutional framework is weighted in favor o f the state government in that it allows state governments to decide on key aspects and the extent o f decentralization, and that central guidelines for CSS also affect actual decentralization to PRIs. This holds good for the key sectors -education, health, dnnlungwater and sanitation, and employment programs. Primary education: legislation and implications The Kerala example The Kerala Education Act (1958)44 allocated all responsibilities for pnmary education to the state government. These responsibilities included establishing and recognizing state government primary schools; teacher training; providing policy and administrative guidance to pnmary schools; and funding pnmary schools. The Kerala Panchayat Raj Act (1994)45 clearly delegated and shared out responsibility between the GP and the DP inthe management o f pnmary schools. The BP was not assigned any responsibilities in pnmary education. In2000, Kerala's legislative assembly amended the Education Act to incorporate the provisions o f the Panchayat Act in pnmary education. This was part o f the state's larger exercise to harmonize existing legislation with the mandate o f the Panchayat Act. Thus GPs were given the powers to manage pre-primary and pnmary schools, and DPs the mandate to manage upper-pnmary schools.46The amended Act also empoweredpanchayats to appoint temporary teachers inany school it managed. Despite this amendment, however, the legal framework still assigned many cntical activities, including the preparation o f schemes and assessing the need to establish schools, to the state government. Inessence, the legal framework for the provision o fpnmary education inKerala still favors the state government as far as the assignment o f functions is concerned. What distinguishes Kerala's legal framework from that o f other states i s the clear delineation o f responsibilities between the PRIs and the attempt at harmonizing the state's Education Act with the provlsions o f the PanchayatRaj Act. West Bengal: a nominal role for PRIs The West Bengal Primary Education Act (1973) empowers the state government to perform all key functions related to pnmary education through the state department o f education. It does not address the role o f PNs in primary education. On the basis o f the Act, the state department 43Kingdon and Muzammil, 2000. 44This Act was amended in2000 and i s now referred to as Act 16 (2000). 45This Act was amended in 1999. We refer to the amended act inthis discussion. 46Section 5A o f the Act (referred to as Act 16). 46 dcd Act 47 Raiasthan: ambiguity o froles Rajasthan has no Education Act, and the Rajasthan Panchayat RUJAct (1994), the only law that deals with pnmary education, arhculates only general roles for all three tiers o f government. The GP i s responsible for promoting public awareness and ensunng enrollment in primary schools. The BP i s responsible for "running" or managing primary schools, as well as constructing and maintaining schools; and the DP for all promotional and management related activities inprimary and upper pnmary schooling. The allocation o f responsibilities across PRIs is charactenzed by ambiguity in terns o f the specific roles to be played by each tier o f government, and by concurrency in the roles assigned. In addition, the state's Panchayat Act does not identify roles for any o f the critical functions associated with pnmary schooling. Inthe absence o f a state law to thiseffect, these decisions are left with the executive. The role o fpanchayats incentral schemes for educations Though the Constitution has entrusted the subject o f pnmary education to state governments, the central government has introduced specific CSS that create new institutional arrangements. There were seven central schemes4' relating to elementary education, but all o f them have now been integrated into a single scheme called the Sawa shikha abhzyan (SSA). This scheme, which aims at imparting education up to Class Eight to the age group 5-14 years by the year 2010, plans to bndge gender and social gaps using community participation. The central government has stated that the key activities of this scheme will be implementedthrough the PRIs. Insummaw the assignment ofrolesineducation Inadequate delegation o f functions: In education, the state delegates functions to lower administrative levels such as the distnct and block educational offices. But in the de jure allocation o f responsibilities (see Table 111.3), there i s very little delegation o f functions to the PRIs as autonomous units o f self-government. The few items delegated are usually the least critical elements o f schooling. Concurrent responsibilities: When explicitly delegated to the PRIs, functions often take the form o f concurrent responsibilities, with allocations among the three tiers left unclear. For instance, the maintenance o f school buildings i s allocated simultaneously to the state, distnct and block levels in both Karnataka and West Bengal. Who has the power or responsibility then depends on the specifics of implementation. Also, there are some key functions for which the law does not allocate responsibility This i s not necessarily a weakness, since all the details o f a sector neednot be legally specified. However, one o f the main problems inthe system i s that the monitoring and trackmg o f learning achievement, and meeting learning standards (not just enrollment targets), is, by and large, not clearly assigned to any given level. 49OperationBlackboard, Non-formal Education, Teacher Education, Nutntion Support to Primary Education, Lok Jumbish, ShikshaKarrmandthe Distnct PrimaryEducationProgram. 48 T allocftian of responsib inthe four sta ~p Ussr Groups Choice of students for targeting progs ~ ~ r ~ l l ~ e ~ ~ Code KamatPka Keraia Rajasthan West Bengal Kerda also deals with certain regulatory roles that PRIs can play in the health sector. For instance, Section 270A mandates that all pnvate hospitals and other medical institutions can only be established after it registers itself with the GP, and Section 271 allows the GP to collect an annual fee from pnvate hospitals. In addition, the Act mandates the organization o f managmg committees with supemsory powers over public health institutions within their areas o f junsdiction. West Bengal Though health i s a state subject, all health related activities in West Bengal are largely based on central Acts, which" are pnmarily regulatory and do not address the question o f functional assignments. The Panchayat Act i s ambiguous about functional allocations across the three tiers. The GP i s responsible for providing preventive and curative measures for malaria, small pox and cholera, and for performing functions related to rural dispensanes, health centers and child welfare centers. Both the BP and the DP have the power to "undertake schemes and adopt measures including financial assistance to promote public health and sanitation including the establishment and maintenance o f dispensanes and hospital^."'^ Karnataka Health issues inKarnataka are addressed pnmarily through the Mysore Public Health Act (1944), while the role o f PFUs i s articulated in the Panchayat Act. In addition, several central acts deal with issues o fregulation as well as drugprocurement. The state-level Public HealthAct stipulates the constitution o f a public health board to inspect, control and oversee health related activlties. The director o f this board has the authonty to recommend and implement measures to improve health administration; crucially, the director i s also responsible for hiring and transfernng personnel. The PRIs have an insignificant role to play in health related activities. In addition to drawing up annual and five year plans (mandatory duties o f all PRIs), the GP i s to implement family planning schemes, take preventive and remedial measures against epidemics, and regulate the sale o f penshable food items. The DPhas supemsory functions, including the management o f hospitals and dispensaries. The BP has no specific functions. Raiasthan The central acts provide the regulatory framework. At the PRI level, the Panchayat Act deals with the role o f the three tiers in fairly rudimentary terms. The GP is responsible for preventive and remedial measures against epidemics, implementing family welfare and immunization programs, regulating the sale o f penshable items, licensing, the destruction o f stray dogs and the regulation o f offensive and dangerous trades. The BP i s Except ~~~~l~ to extent, responsible for the mOnitOIlng O f immUniZatiOn programs the legal position has resulted in and organizing health and sanitation fairs. The DP i s state and central governments responsible for organizing activities related to health playing the predominant role in education, maternal and child health, family welfare and all stages of delivering health health camps, with the assistance o fthe BP and GP services, from planning to implementation. The role of PRIs remains insignificant. 50 Prohibition of Sex Selection Act (1994), the Food Adulteration Act (1954) and the Drugs and Cosmetics Act (1994) -- The relevant central acts -- including the Preconception and Prenatal Diagnostic Techniques Act (1994), the deal with certain issues related to drug procurement and the regulationof health related activities. 51West Bengal Panchayat Raj Act, 1073. 50 Insummary*theimplicationsofhealthrelatedlegislation Although the delivery o fpnmary healthhas been constitutionally entrusted to the states, the central government has been implementing a large number o fprograms5*for health and family welfare as well as disease control. On the whole, the legalposition on the delivery o f health services has led to the predominance o fthe state and central governments inall stages o f semce delivery, from planning to implementation. With the exception o f Kerala, the PRIs remain insignificant in several ways: Limited responsibility to GPs: With the exception o f Kerala, very few responsibilities are given to the GPs inhealth services. For example, inKarnataka, the GPs can "assist inthe formation o f village health committees." These committees are made up o f some village members, but also village health guides, trained birthassistants and mulh-purpose health workers - all o f whom are state employees. The ambiguity o f the word "assist," the fact that none o f these activities involves any money, and the responsibility o f supervising employees hired by someone else (the state) essentially means that the GPs have no real influence. Ambiguous division o f responsibilitv. The divlsion o f responsibilities among levels o f government and across tasks is often demarcated by ambiguous critena. For example, both DPs and BPs in Karnataka are to organize health camps (attended by people mobilizedby GPs). Lack o f administrative powers: Though Kerala i s an exception inmany ways, there is one problem it shares with other states: the GPs' lack o f administrative powers. This pattern persists throughout the system. BPs in Karnataka "assist" in the supervision o f sub- centers and the deployment o f field staff. DPs coordinate, but do not actually supemse PHCs. Ineach case, the staff in question are state employees, and supemsed, inreality, bythe state. Weak accountabilityto clients: The main line i s the state - at which level pay, promotion, transfers and other aspects o f employment are determined. For lower level staff, the multiplicity o f voices (clients complaining o f lack o f drugs, GPs aslung anganwadi workers to do additional jobs) may cause discomfort, but for medical personnel, accountability is to their employer, the state health department, not to the client. The bypassing o f PRIs by central schemes: These programs are invariably implemented through vertical formations without the involvement o f the PRIs, even though the subjects they cover now fall within the domain o f powers and functions transferred to the PRIs. There are some moves to increase the role for PRIs in central schemes through the GOI's recent Rural Health Mission (2005). Whether this vehicle can empower local leaders to take more responsibility for the health o ftheir constituents remains to be seen. Rural water and sanitation: legislation and implications Kerala's promess Kerala's legislation has been fairly consistent in devolving responsibilities regarding rural water to thepanchayats. Before the introduction o f the Panchayat Act, rural dnnlungwater supply was 52The major CSS under pnmary health include the family welfare programs, the vector-borne disease control programs, the TI3 ControlProgram, the BlindnessControlProgram and theHIV AIDS/STD programs. 51 the responsibility o f the Kerala Water Authonty (KWA).53Subsequently, the state's Panchayat Act (1994) specified functions to be performed by panchayats. GPs are to protect traditional drinhngwater sources, ponds and other water bodies, and set up and run water supply schemes covering one GP. DPs are to take up and manage water supply schemes covenng more than one GP inthe distnct. BPs do not have a role. Insanitation, the state's Panchayat Act assigns GPs the functions o f collecting and disposing o f solid and liquid waste, water drainage schemes, management o f markets, toilet facilities in public places and implementation o f sanitation programs. It assigns distnct-level coordination o f center and state sponsored sanitation programs to the DPs. As inthe case o f dnnhngwater, BPs have no role. Several other Acts and amendments have been modified to ensure consistency with decentralization andhave assigned specific water and sanitation functions topanchayats in Kerala: The Kerala Water Supply and Sewerage (Amendment) Act (1993) re-vested and re- transferred to the panchayats the powers and functions to cany out water supply and sewerage services. The amendment also transferred the right to recover arrears o f water charges, meter hire and any cost or fees due to the panchayats. A clause was added to stipulate that the salary, allowance, and other contributions due to state employees transferred to thepanchayats would be paid by the latter. The Kerala Panchayat Raj (Construction and Maintenance o f Public Latnnes, Urinals, Bathing Places and Sanitation o f Pnvate Premises) Rules (1998) assigns GPs the function o f providing and maintaining a sufficient number o f public latnnes in convenient places, and the GPs can issue licenses for this purpose. The GPs are also responsible for solid waste disposal. 0 The Travancore-Cochin Public Health Act (1955) vests panchayats with powers to ensure a clean and healthy enwronment. West BenPal: limited state-level legislation The state's Panchayat Act does provide some legislation in the area o f water and sanitation provision, and the role o f PRIs in development planning involves all three tiers o f local government. Insanitation, their role i s to ensure that buildingpermissions are withheld ifthe plan does not provide for a latnne or if it only prowdes for a dry latrine. The GP i s to prepare an annual plan for "the development o f human resources, infrastructure and civic amenities" in its area o fjunsdiction; set aside any public source o f water supply for dnnhng or culinary purposes; and take specific measures to prevent water pollution. The gram ~abha,'~ to be formed in every constituency o f the GP, i s to meet twice a year and guide the GP on all issues relating to socio- economic development, which would include water and sanitation. And the GP, in turn, i s to act on the decisions o f the gram sabha where possible. BPs and DPs are to undertake schemes and adopt measures, including financial assistance, relating to the development o f water supply, public health and sanitation. But since the Panchayat Act authonzes the state government to issue any direction to a GP, BP or DP regarding any function or power, the state government has a fair degree o f autonomy inassigning functions, finances and functionaries to local governments." 53 The role of KWA was defined by the Kerala Water Supply & Sewerage Act 14 of 1986, Amendments & other Regulations as the preparation, operation, maintenance, and financing of schemes for water supply provision as well as liquid waste disposal. It was also responsible for fixing and revising tariffs, taxes and charges on water supply and maintenance. 54Calledgram sansadinWest Bengal. 55 This provision i s slightly (but not significantly) arrested by another provision in the statute that provides some autonomyofthe PRIs over state level staff. 52 Karnataka: emuhasis on service details, not functions The Karnataka Ground Water Regulation Act (2003) does not allocate any functional responsibilities for the provlsion o f water semces. Its main feature i s the prohibition o f bore- wells within 500 meters o f public sources o f h l u n gwater. The state's Panchayat Act allocates semces, and some functions to panchayats. For instance, the GPs are assigned the responsibility for semces including the construction, repairs and maintenance of water wells, tanks and ponds; the prevention and control o f water pollution; the maintenance o f general sanitation --the maintenance o f public latnnes and clean public roads, drains, tanks and other public places. The BPsare responsible for the establishment, repairs andmaintenance o fruralwater supply schemes, and the prevention and control o f water pollution, as well as the implementation o f rural sanitary schemes. The DPs are responsible for the promotion o f dnnhng water supply programs and rural sanitary programs. Raiasthan's emphasis on sanitation Inthe case ofruralwater supply, Rapsthan's Punchayat Act assigns GPs,ingeneralterms, the construction, repairs and maintenance o f dnnhng water wells, tanks and ponds, and the prevention and control o f water pollution. The Act i s a little more specific about sanitation. It assigns GPs the construction and maintenance o frural latnnes, facility parks and other public places; the maintenance o f general sanitation; cleaning o fpublic roads and drains; the disposal of unclaimed corpses and carcasses; the management and control o f washing ghats; and the implementation o frural sanitation schemes. The uredominance o f central schemes: imulications for rural water and sanitation Although drinhng water and sanitation is a state subject, there i s considerable intervention from the center through its schemes. Central initiatives to provlde adequate hnlung water inthe rural areas include the Sector Reform Project (SRP), introduced in 1999 in selected distncts o f the country to institutionalizecommunity participation; and Swajaldharu, which expands the scope o f sector reforms throughout the country Insanitation too, the center's presence i s significant. The Central Rural Sanitation Program is the major funding program o f state government initiatives, and the key GOI-administeredprogram i s the Total Sanitation Campaign (TSC), a program with the considerable planned outlay o f Rs. 37.5 billion (US$83.3 million). While many o f the institutional arrangements in CSS try to take decentralization further, central schemes can, in pnnciple, diverge from state legislation contributing to the confusion. Inthe case o f Kerala, the state government has addressed this through a government orders6 transfemng all single punchuyut schemes managed by KWA to the respective GPs. Nevertheless, inconsistencies continue. The assignment o f functions for water supply, as determined by the GO1guidelines and the associated state government orders, differs considerably, depending ontheprogram. Insummary. The transfer ofwater and sanitation functionsto uanchavats Some state governments have made credible efforts to transfer functions in the rural water and sanitation sector functions to panchayats (see Table III.4 and Table 111.5). But these efforts have been overshadowed by the CSS rules and guidelines that often bypass state legislation. Legally, the central government can only issue guidelines to states. But since the center contributes significant financial resources to the sector it can specify conditions underlying these transfers. 56GO MS.No. 44/2000/IRD dated 12.06.2000. 53 c e 55 ini,tiori crivi1.v Xn ry: Does the k eaabfe P 0 Some states have moved comprehensively (Kerala) while others have done it only for specific sectors such as rural water and sanitation. Funds do not follow functions. Since state governments are not obliged to transfer funds topanchayats, the fiscal capabilities ofpanchayats are weak. 0 Panchayats have little or no administrative control over functionanes, and inmost states, the accountability o f functionanes i s to their employer, the state government. Central guidelines tend to bypass relevant state-level legislation. Several major central schemes exist in the key sectors, and the dejure responsibilities for implementing these schemes are determined by central guidelines associated with the financial transfers. The result, in both theory and practice, i s confusion, even conflict, in the allocation o f responsibility inpanchayat or sectoral acts. 58 IV. The Reality on the Ground: K e y Services at the Local Level Across sectors, the defacto status favors the states even more than the legal: the state government plays the predominant role in implementing most key functions even in situations where these functions have been unequivocally devolved t o thepanchayats in state acts. The allocation o f responsibility to PRIs is limited, and even this i s not supported by financial o r administrative powers. As a result, the relationships of accountabilitv are weaker. and services do not meet local emectations. Finding out about the ground reality: methodology Local surveys H o w does the ambiguous legal framework for semce provision translate into reality? T o answer the question, this study carned out a survey o f local actors in the four states. Two sets of genenc quesbons were prepared for the field surveys. Questionnaire-I was aimed at understanding the de facto district and block-level position o f service delivery The respondentswere: 0 Presidents and chief executives o fthe DPs; 0 Chairpersons o f DP sub-committees and the dishct officers dealing with pnmary education, health, dnnlungwater and sanitation, and employmentprograms; 0 Presidents and secretanes6' o f the BPs; and 0 Chairpersons o f the BP sub-committees and the block-level officers dealing with pnmary education, health, drinlungwater and sanitation and employmentprograms. Questionnaire-I1 was aimed at understanding the defacto village-level position. The respondents included the president o f the GP; the chairpersons o f the sub-committees dealing with pnmary education, health, water supply and sanitation and employment programs in the GP; the GP Secretary; and about 10 villagers randomly selected from among the general, women and SCI ST categones. 6oBlockDevelopmentOfficers 59 Box JV.1: Selection uf districts fur study the results arc in Table W.X. 'S'ht: key Kerala: the exception ininfrastructure Vanation between urowsion of teaching semces and infrastructure: The survey revealed considerable vanation inthe institutional arrangements of infrastructure and the teaching sector. The GP's role in infrastructure actiwties: The GP i s perceived as playing either a predomnant role, or a somewhat cntical role, for most infrastructure related actiwties (see Figure What distinguishes Kerala from N.l).The survey respondents clearly identified the other states is that the PNS, GP as responsible for infrastructure related activities particularly the GPy have a such as maintenance, school construction and the defined to play in infrastructure preparation o f designs and estimates for school activities, with the state government buildings. The construction o f these school buildings playing a minimal role in this i s generally funded through the plan funds available process. at the GP level. However, there appears to be some confision among respondents on the specific flow of funds, with many identifying both the state government and the GP as the source o f funds. The state's role inteaching related actiwties: Incontrast, the state government i s perceived to be the key implementing institution in the teaching sector. It plays a predominant role in all teaching-related achvities, including hinng teachers and monitonng the quality o f teaching services (see Figure N.2). The explanation for this skewed result i s that teachers are essentially state government employees, deputed to the PRIs for a short duration. Ths creates a dual line o f authonty. inpractice, the implication i s that teachers are accountable to the state government (the department o f education),. rather than to PRIs. This issue o f the dual line o f authority between the state line departments and the PRIs holds true in most other states in India. The limited role of other local rrovernments: The DP i s seen as playing a minimal role inprimary education. The survey shows that the BP plays no role - inconsonance with the fact that the KeralaPanchayat Act does not devolve any o fthe pnmary education functions to the BP (see Chapter 3). 61 Figure IV.l The defacto status of infrastructure activitiesinprimary education,Kerala 70 Clear Responsibilityof GP Y 5 50 P ; P 0 8 40 s 0 30 20 10 0 Actlvltles Note: Responsesarenot mutuallyexclusive. Source:KILA, 2005 Figure IV.2: The defacto status of the teachingsector in primary education,Kerala 80 70 0BlockPanchayat Gram Panchayat 30 20 10 0 Salary TeacherTraining HiringJFiring LeaveSanction Service Quality & Teacher Disbursment Teaching Attendance Standards Activities Note: Responsesarenot mutually exclusive. Source:KILA, 2005 63 The dichotomous system inWest Bengal: mainstream vs. the Shishu shiksha karmasuchi(SSK) An examination o f the de-facto status o f semce delivery in West Bengal, highlights an interesting trend in the case o f pnmary education. At present, pnmary education in the state i s delivered through two key parallel systems. The first focuses on mainstream pnmary education which i s delivered pnmarily through funds obtained from the SSA program, the CSS discussed above. Most pnmary education activities are governed through the West Bengal Boardo fPnmary Education (WBBPE), which i s responsible for major policy decisions such as cumculum setting, preparation o f textbooks and teacher training. At the distnct level, the WBBPE has set up Distnct Pnmary School Councils (DPSC) responsible for all district-level operational activities including teacher recruitment, salary disbursement and procurement. The WBBPE and the DPSC are the organizational bodies that run the SSA program in the state. Parallel to this system, the state government introduced the alternative para-teacher scheme SSK in 1997-98, with the aim o f provlding primary educational facilities to students in areas without access to the formal education system. The state government as the maior player in the mainstream system: For both activities - infrastructure and the teaching sector - the survey respondents perceivedthe state government as playing a predominantrole inthe delivery o fmainstream pnrnary education (see Figures IV.3 and IV4). FigureIV.3: The defacto status ofinfrastructureactivitiesingovernment primaryschools, West Bengal 1 OState Gvt QZiiia Parishad OPanchayatSamiti Gram Panchayat EUser Group O j o Response 0Noone Activities Note: Responses are not mutually exclusive. Source: SIPRD, 2005 Stronger role for PRIs inSSK. Inthe case o f the SSK program, however, respondents identified a stronger role for PRIs and user groups in implementation. Respondents identified the BP as the key institution responsible for assessing the need for schools, In West Bengal, pRTshave a teachers and equipment, as well as for constructing school stronger inimplementing buildings.Again, inthe teaching sector, the BP was seen as SSK than in mainstream responsible for key functions such as hiring teachers, and education policies and supervising and monitoring their quality of service. But the programs. 64 respondents seemed confused on the specific role o f the BP and the state government in areas such as teacher training and salary disbursement. Overall, the survey identified a relatively stronger role for PRIs in implementing SSK than in the mainstream education policies and programs o f the state government (see Figure IV.5). This result shows again how Centrally Sponsored Schemes can distort the decentralization process. In West Bengal the SSA was channeled through the state line agency instead o fPRIs. FigureIV.4: The defacto status ofthe teachingsector ingovernmentprimaryschools,West - Bengal - OZiiia Panshad 0Panchayat Sarniti Gram Panchayat User Group No Response Activities 65 Figure IV.5: The de-Facto Status of Teaching Sector--SSK 80 w 70 I4 Clea esponsibility of State Govt I 60 3 50 mZilla Partshad E 8 0PanchayatSamiti 0Gram Panchayat K 40 8 r ElUser Group MTeacher 30 a: $ 0NoResponse 20 10 0 Activities Note: Responsesare not mutually exclusive. Source: SIPRD, 2005 The DP's limited role in both SSA and SSK. The DP has no substantive role to play in the administration o f primary education in its area o f junsdiction. This i s despite the fact that key policy and administrative implementation decisions are made at the distnct level through the WBBPE and DPSC. In SSK too, the general perception among respondents i s that the DP plays almost no role since all operational responsibilities lie with the GP and BP The DP, however, i s seen as playing a role infund disbursement and activities relatedto teacher training. Role o f the state in SSK. The expenence o f SSK on the other hand, reveals a different story According to our survey, SSK i s implemented primarily through PRI's. This i s in consonance with the policy guidelines (or de jure position). The state government however, is perceived to play a cntical role in supemsory activities particularly monitonng teacher attendance and disbursing teacher salanes. Figures 4.5 and 4.6 contrast the role of the state government in running o f government primary schools vs. SSK's. Both figures indicate clearly the diminishing role o fthe state government in SSK schools. 66 Figure IV.6: Respondent perceptionof state government role ininfrastructureactivities: government primaryschoolsvs. SSK, West Bengal 70 10 0 Actlvlties Figure IV.7: Respondent perception of state government role inteaching sector activities: government primary schools vs. SSK, West Bengal loo 901 80 20 10 0 Activities Note: Responsesare not mutually exclusive. Source: SIPRD, 2005 67 Karnataka: the dominant role o fthe state bureaucratic machinery InKarnataka, the state bureaucratic machinery is responsible for implementing most functions In Karnataka, the state related to the delivery o f pnmary education. This, like the bureaucracy plays a keyrole other states, i s partly explained by the fact that the centrally in all primary education sponsored SSA 1s the dominant source o f financing for primary activities. education in the state - approximately 50 percent o f the plan budget, and these funds are largely used through the state bureaucratic machinery DP level: The survey showed that neither the DPnor its standing committee on education has any role in policy-malung or in the day-to-day administration of primary schools. The Distnct Institution o f Education and Training,62 a state institution, makes decisions on cntical policy activities and DP members are not encouraged to participate inthe process. Other actiwties, such as the procurement and distributiono f textbooks, uniforms and scholarships, are all handledat the state government level. BP level: All implementation related activities are usually performed at the block level. The Block Education Officer (BEO) i s responsible for overseeing the implementation o f all key actiwties -- undertakmg need assessments for equipment and teachers; and overseeing the day-to- day management o f schools, including the monitonng o f teacher attendance. Most important, the BE0 is formally responsible for hiring teachers. This is usually undertaken on the basis o f approvals from the Deputy Director o f Public Instruction; in other words, a state-level body Since the BE0 i s an employee o f the state department for education, BP and BE0 operations are parallel to each another. But as in the case o f the distnct level, the survey indicates that at the block level too, it i s the state administrative machinery that controls all functional responsibilities. The BP has little or no role to play GP level: The survey highlighted that in most cases, GP members were either unaware of, or ambivalent towards, their responsibilities in primary education. The guidelines for SSA prowde for the formation, at the GP level, o f School Development and Management Committees (SDMCs) responsible for all operational actiwties. Although the SDMC i s expected to interact with the GP, inpractice, it tends to operate as a parallel organization. Insome cases, the SDMCs approach GPs for assistance in construction related activities -- toilets, kitchens for midday meal schemes and dnnlung water facilities. Interestingly, the SDMC also plays an active role in facilitating student enrollment, an activity that the Karnataka Panchayat Act assigns to the GP All assets at the school level are also maintainedby the SDMC. Thus the GP has a very limited role to play Arrangementsinthe healthsector Four key findings emerge from an analysis o fthe ground-level picture inthe health sector: 0 The domination o f state civil servants inthe public sector; 0 The dominance o f private health care provision; DIETSwere set up with the mssion of providing academc and resource support at the grassroots level to achieve quality and universalelementary educationand adult education. 68 0 A regressive pattern o fpublic subsidy across income groups; and 0 Varying extent and type o fPRIinvolvement across the study states. The domination o f state civil servants inthe public sector The state i s responsible for most health care provlsion that i s done by the public sector, and certainly most decisions that involve expenditure allocations. Staff salanes account for 78 percent o f the health budget, construction 20 percent o f the budget, and o f course 100 percent o f the capital account. Some o f this is by design and some by default. In the penod immediately following the passage of the 73rd Amendment, PRIs tended to feel insecure about takmg charge o f health as it was considered too t e ~ h n i c a l .Ten ~ ~ years later, opinions seem to be changing. This study's surveys o f GP members and village residents showedthat while villagers often complain about absent workers, inconvenient hours and other shortcomings, the pervasive opinion i s that nothing i s done about these complaints even when lodged with the department o f health. Similar to the finding in the education sector, this raises the question o f who the government health worker i s responsible to -the clients or his employer (the state government department o f health). Moreover, important functions are not specifically assigned to anyone. The study survey indicates that though they have little influence over it, GP members know about medical care; but they are not familiar with the importance o f disease prevention and health promotion. The irony i s that medical care i s not necessarily a government responsibility, local or otherwise, while disease prevention and health promotion are clearly public responsibilities. Though these responsibilities are probably best carried out locally, they are not recognized by local governments as their responsibility Vaned attempts at panchayat involvement inhealth We find evidence o f varying extent and type o f involvement o f PRIs in the health sector in the four states. At the two ends o f the spectrum, Rajasthan has made much less progress on decentralizing semces than most, while Kerala has gone much further. Karnataka seems to be quite typical o f much o f the rest o f the country on a number o f grounds -health status; the size o f the pnvate sector; the incidence o f public subsidies; and the degree o f decentralization achieved inhealthsemces. West Bengalhas, onthe books, a somewhat clearer division ofresponsibilities across levels o f government than most. Also, substantial change i s already takmg place. Still, one recent example from West Bengal illustrates the problem o f the division o fresponsibilitiesamong levels o f government.@ In this case, the ultimate decision was indeed made by the appropriate level o f government, but only after delays dunngwhich a substantial degree of confusion had to be resolved. The case o f Kerala i s still worth a closer look. Kerala has certainly gone much further than any other state in decentralizing health services. Several features may account for these results, including the size o f GPs, and arguably even more important, the direct accountability o f GP members to their constituency (see Box IV.2). Recent innovations in West Bengal are also relevant. GPs are now granted the authonty to buy semces from the pnvate sector. Some have opted to supplement state semces with much cheaper homeopathic services. Another approach i s that several GPs band together to hire a qualified MBBS doctor to visit each GP inthe group on a regular basis. The doctor (who must be pnvate given present rules) i s allowed to charge agreed- upon fees and can be monitored by local health committees. With individuals being charged, the payment from the GPs may be considered a "relocation grant" just to attract the provider to the "Chatterjee,1994. 64Though, inthis example, it is through municipalities, not PRIs. 69 It',2: PKIsin the h or: thc Keralx example Ruralwater supply and ~ a n ~ t a tfunctions in service provision i ~ ~ : 71 ARWSP SRP Swajaldhara KRWSSA Policy State GOY State GOY State Bank / State P l a m n g State DP State State Asset creation State GP/ User Group GP/ User Group GP/User Group O&M State User Group 7 6' User Group M&E State GP/ DP State SO/ State Source: Water and Sanitation Program, South Asia documents The ineffectiveness o f the Kerala Water Authonty (KWA): Although the State Act defines KWA as an autonomous institution, it requires prior approval from the state government for all cntical functions,68 rendenng its autonomy ineffective for all practical purposes. The result i s a situation inwhich KWA operations incur a significant loss that has to be borne by the state.69This seems to suggest a clear "business" rationale for KWA to withdraw from rural water supply. However, while the government has notified 1050 single village schemes to be transferred by KWA to the GPs, only 140 schemes had been transferred as o f July 2003. The situahon has not improved much since then. GPs are unwilling to accept the schemes without funds and functionanes, and KWA isreluctant to transfer the concerned staff or O&M funds to the GPs. Linkages between GPs and User Groups: For both the Jalnidhi and the SRP/ Swajaldhara-II programs, the BG i s the community level user and management group, while the GP performs a mix o f governance and management functions. The linkage between the BG and the GP is also clearly defined in the planning and asset creation phases. But the linkages are unclear in the post- implementation O&M phase. Fiscal difficulties: Having allocated 40 percent o f the plan budget down to PRIs, Kerala i s heralded as one o f the success stories in decentralization. However, the fund flows to panchayats is unreliable and restncted due to the poor financial status o f the state itself. This poor status may, in part, be the result of the state in effect running both centralized and decentralized service delivery vehicles. Responsibility without management and control: The priontization o f infrastructure needs i s undertaken at the local level. But the provision o f infrastructure and O&M services has remained largely with state-level functionanes. While PRTs have been given control over what to spend on, they cannot decide on the "how " For example, if they decide to spend part o f their funds on a RWSS program, this has to be implemented by the state line agency In addition, the state departments make appointments, postings and transfers o f functionanes, and the panchayats can only "recommend" disciplinary action against them. AddressinP water conservation and quality: The depletion o f water sources and the seasonal scarcity o f water resources highlights the need for water conservation and source protection. It also indicates the ineffectiveness o f solutions to date. Although groundwater regulation has been legislated by the Government o f Kerala,70 the functions to be performed by each tier o f 67 7Impliesthat as yet no schemes have enteredinto the O&M phase. 68For example, transfers & postings, tariffsetting, capital mobilizationand contract finalization. 69 An assessment undertaken by CRISIL in 2002 demonstrated that the deficit in KWA's urban operations was 19 percent, while the rural sector accounted for 81percent o f the total operational losses o f KWA. 70 Notified in March 2004, this legislation introduces a registration mechanism for the monitoring and surveillance o f groundwater extraction sources, and seeks to restrict interference with public dnnking water sources through a distance limtation o f 30 m. 72 government are not clear. On the matter o f the quality of water provided, one o f the senous concerns about the promotion o f well water sources i s the corroborated ewdence of bactenological contamination. However, studies have shown that people are boiling water for dnnlung, and that public sources o f piped water supply are also contaminated in spite o f investment in water treatment. At present, water quality is pnmarily a function o f the state government, but this is clearly ineffective. The issue o f bacteriologcal contamination i s considered more o f a regulatory issue than a semce delivery issue. Sanitahon Latnne coverage/ access: The 2001 census data places Kerala as the best performing state inIndia interms of access -- 81.3 percent of the m a l populationhave access to household latnnes. The next question i s whether these latnnes deliver a safe sanitary enwronment. Oualitv o f sanitation: Despite its high latnne coverage, Kerala has poor sanitation quality. The analysis o f the dnnlung water quality in wells suggests that around 90 percent o f the wells are bactenologically contaminated. The cause o f this water contamination is pollution from latnnes that have failed to adequately confine excreta.71 Given the dependence on "well water" as the pnmary source o f dnnking water, this is a major failure in sanitation service delivery The major problem i s that efficient solid and liquidwaste disposal is by and large non-existent. The state has initiated the Clean Kerala Mission (CKM)," a state level sanitation program that i s "conceptually" outcome onented. The challenge C K M faces i s the development o f localized "outcome" onentedprograms and extending these throughout rural Kerala. Funds and DroPram control: At the state level, sanitation is the responsibility o f the Rural Development Department (while water i s the responsibility o f the Water Resources Department). The state TSC program -- a high subsidy program that offers Rs. 2,000 per latnne to BPL households -- has been initiated in all distncts. The program mainly uses contribution from the pan~hayats.~~ this subsidy-driven program has its weaknesses. First, there i s very little But awareness raising. Second, the funds and control o f the program are in the hands o f state employees o f the Distnct Rural Development Agency (DRDA). This means that the DP and other PRIs do not have effective control over the significant funds allocated for TSC implementation. Despite their significant contribution of funds to the program, the panchayats have very little control over how that money is spent -- except to identify the beneficianes. The state does not have any consolidated cumulative data on the total number o f latnnes constructed on account o f departmental effort. But the indications are that about 70 percent o f the latnnes have been constructed without any subsidy element, through pnvate household initiatives. This illustrates the weaknesses o f the TSC program. West Bengal Assignment of functions: Although the GP has been assigned most o f the functions in water service delivery inthe state, defacto the BP plays the major role in the delivery o f water services. This is mainly because the latter is resourced with engineers; these engineers, however, are de- concentrated functionanes o f the state Public Health Engineenng Department (PHED). Only 71Surveys o f three lakh households in 43 GP and four distncts conducted by Jalnidhi indicated 83 percent coverage with latnnes, approx. 90 percent o f these latrines being deep pit latrines. Jalnidhi subsequently initiated a program to convert the deep pit latrines into shallow water-seal latnnes. 72The Clean Kerala Mission is pnmarily focused on solid waste management inurbancentres. 73DP = Rs.500, BP = Rs. 500, GP = Rs. 1000. 73 seven percent o f the access to water i s through piped water supply schemes; the rest i s through local sources. Given the prevalence o f hand pumps as the premier form o f access, the responsibility for governance o f oversight o f RWS should reside with the GP; and the service provlsion function should be the responsibility o f a body like the Village Water and Sanitation Committee (VWSC). Accountability. Since the project financing for RWSS is administered from the state PHED budget, and the funchons o f recruitment, transfer and discipline o f functionaries are also with the PHED, it appears that all accountability runs upward within the state government department. However, the stability o f the political environment in West Bengal and the political strength o f the local government do create considerable informal accountability o f the de-concentrated state functionanes to their respective tier o f PRI. This means that there are actually dual lines o f accountability - formal accountability of functionaries to their supenors in the line department, and some informal accountability to PRIs. Downward accountability i s also weak. A senes o f sample interviews conducted by the SIPRD concluded that most consumers were not billed and did not pay any user charges for the delivery o f water semces. Similarly, there was little no paymentby front-end providers for the bulk supply o fwater inlarger water supply schemes. Communitv involvement: Although water supply i s prachcally free, less than 50 percent of consumers were satisfied with the standard o f delivery Given this situation, it would be reasonable to expect that there might be great demand for community involvement in the management o f RWS. However, the sample surveys also revealed that neither consumers nor the GPs deem it necessary to raise funds or form user groups to improve the quality o f water supply service provlsion. Karnataka Water supply Gaps between allocation and practice: InKarnataka, the GP and VWSC have been assigned the major functions o f "oversight" and "service provision" The GP (and the VWSC) are allocated roles in the planning and implementation o f schemes in accordance with government orders. But the S P R D survey found that the DP i s performing almost all these functions. Similarly, where functions have been assigned to the BP, they are also being performed at the distnct level. Perhaps even more pervasive, the distnct-level execution o f RWSS i s governed by a Distnct Water and Sanitation Mission (DWSM) rather than a DP Although the chairperson o f the DP i s a member o f the DWSM, the execution o f the "mission" remains in the domain o f the supenntending engineer o f the KRWSSA. Insum, the DP has no formal control over the RWSS mission: it has no powers o f recruitment, transfer and discipline over the KRWSSA staff. Finally, while the state government has passed a water resource management act, the allocation o f responsibility for managing water resources i s not yet effectively implemented. This i s apparent in the general decrease in the water table across distncts due to the over- extractiono f groundwater, primarily for irngation purposes. Sanitation Karnataka has high subsidies and poor usage. Sample surveys conducted by the SIPRD reveal that only 50 percent o f the individual household latrines constructed under the high subsidy programs o f the state government are being used for defecation. However, indivlduals who have 74 constructed latnnes using their own funds were found to be regularly using them, and this once again illustrates the limitations o f the subsidyprograms.74 Employmentprograms:the dominanceof the ruraldevelopmentagency This study's investigahon o f employment programs reveals a picture similar to that o f the other sectors. The central and state government agencies dominate the scene, with PFUs playing an insignificantrole inplanning and implementation. The investigation i s based on an analysis o f the SGRY - the flagship CSS for employment programs, malung up about 50 percent o f the rural development budget o f the GOI. The findings are, however, applicable to other central and state employment programs. Work idenhfication, beneficiary selection and accountability Inmost cases, the DRDAplays the cntical role o f work identification and beneficiary selection. The indicate the Every aspect of the SGRY program 1s marked of employmentthat programs by the overwhelming presence o f line . is marked by department officials. The relationships of Control by line department officials; accountability between officials and clients, as Hiring of contractors against the . as pH members, are weak' This has guidelines; allowed for extensive leakages (including the 0 Works not reflecting local needs and hinngo f contractors); it has also meant that the priorities; Beneficiary listsnot including many of works identified do not reflect local needs and the poorest; prionties. Money has often been spent in areas Weak relationships of accountability; and on infrastructure that benefit only the local and elite. Moreover, the self-targeting mechanism o f Top-down M&E. the program i s diluted, wth officials drawing lists o f beneficianes and presenting them to the gram sabha. Very often, these lists do not represent many o f the poorest. Instead, they reflect the pnorities o f officials/ politicians and contractors. Top-down M&E systems have, o f course, created the possibility o f irregulanties in muster rolls and account books. Involvement o f multiple central/ state ministnes Multiple ministnes at the central and state level are often involved in implementing these programs; and the implementing authonty at the local level has to await a series o f sanctions and approvals. The result i s delays inthe release o f the grants, hence considerable delay inthe release o f grains and food at the local level, and delays in payment. Since activity plans can only be implemented on the receipt o f technical sanctions, any delay in sanctions from the center/ state can also mean a delay in the actual implementation o f the program. All these delays affect the synergy between the implementation of the program and the off-season requirement for labor and wages. Thus many programs tend to get implemented when most unslulled laborers have found alternative forms o f employment. This encourages the use o f contractors, the construction o f machine-intensive works, and in some cases, the inflation o f wage rates to attract laborers. Consequently, the benefits o f the program often do not reachits intended target. 74SIPRD, 2005. 75 Modificahon o f wages at local level Although the guidelines maintain that wages for SGRY should be set on the basis o f the legal minimumwage, wages are often modified at the local level. This happens for many reasons. In areas where productivity i s high,or where programs are implemented dunng the peak agricultural season, wages are increased in order to attract labor. Inareas where productivity i s low, or, most often, where contractors take over the implementation o f the program, wages are set far lower than the minimum wage. This has had two consequences. The inflation o f wage rates has crowded out the poorest o f the poor (because less poor people compete for jobs), thus weakening the self-targeting mechanism o f the program. Inareas where wages are below the state imposed minimum wage, beneficianes are clearly at the losing end. In the absence o f effectwe dissemination mechanisms and weak client power, beneficiaries are often either unaware o f thelr due, or unable to get it. All pervasive central rmidelines As a central scheme, SGRY operates, by definition, on the pnnciple o f "one size fits all." Although there i s space for negotiatmg the program guidelines at the state level, the central government i s responsible for all key decisions. Thus states often implement programs that do not necessarily apply to their local conditions. In SGRY, the grain component i s the main point o f contention. In states such as Kerala and parts o f Kamataka where food secunty i s not the most pressing concern, beneficianes prefer to be compensated in cash rather than kmd. This has encouraged funchonaries o f implementing agencies to sell food grains -- a compulsory component o f the program -- to wholesalers at high profits, thus bnngmg in the possibility o f misappropnation. Hiringofcontractors In all the states under examination, contractors are hired on a regular basis to implement the program, despite prohibition in the scheme guidelines. The contractors are hired either by line department officials or by PRI members and their presence i s often apparent at all stages from identifying works and beneficianes to the dissemination o f wages. This has many repercussions: 0 Contractors most often hire their "own" labor, who are often from outside the district and paid at rates far below the minimumwage rate. 0 Since profit i s the main motivation and because o f the timing mismatch between scheme sanction and implementation, contractors often identify machine-intensive works that do not require hinng much labor. This undermines the employment creating rationale o f these programs. 0 The grain component o f the program is particularly attractive to contractors. Rather than distribute the grain at the work site, contractors often sell the grain to the wholesale market at a profit. Occasionally, workers are compensated with cash in lieu o f grain but the cash component does not match the value o f 10kgnce inthe open market. The de facto functional allocation in employment programs i s shown in Table IV 4. The important role played by "service providers" (often contractors) in the actual operation o f the programs i s apparent. 76 77 V. The Assignmentof Functions,FunctionariesandFinance: Applying the ConceptualFramework H o w should functions, functionaries and finance for key services be assigned to the different tiers o f government? And how should the relationship between those tiers and frontline service providers be addressed? In response t o these questions, public finance criteria and accountability criteria were applied to the unbundled activities o f elementary education, health, water and sanitation, and employment programs. In general, services can be improved with decentralization if increased autonomy can be matched with greater accountability. Specifically, 0 A greater push towards decentralization in operations in education would create opportunities for substantial cost reduction and quality improvement; 0 Inhealth, a greater pushtowards decentralization of small-scale, transaction intensive preventive and promotive services coupled with operational responsibilities in curative care would lead to greater accountability; 0 Decentralizing all operational and planning related responsibilities in water and sanitation to the GP with powers to contract up for specific service can greatly improve service delivery; and 0 Accountability in employment programs can be achieved only if activity selection and asset creation are decentralized to the lowest level o f implementation, while monitoring and evaluation remainwith higher levels. The task: applying. first pnnciples If a state has chosen to move ahead with decentralization to the PRIs, how should it assign the three F's? This chapter addresses the appropriate assignment o f functions, functionaries, and finance for the key semces to different tiers o f government -- center, state, and the three PRI tiers; and thus the relationship between those tiers o f government and the frontline semce providers. Two sets o f criteria are applied to each sector. The first set o f public finance critena looks at the presence o f economies o f scale, externalities, and the heterogeneity and equity aspects. The second set o f critena tries to ensure that the appropnate relations o f accountability can be present. Primaryeducation Applying publicfinance criteria toprimary education Identifvnp catchment areas: It i s useful to introduce the concept o f a catchment area to apply economies o f scale to public service provision. It also applies to the other services discussed later. The nature o f the local services under discussion i s such that each person must come inphysical contact with the service provider (teacher, health worker) or the facility (water). In the case o f schooling, each child is expected to travel to the school each school day In rural areas, where transport costs are high, this means that nearly all the consumers o f a given service come fiom the same narrow geographic area, and nearly everyone in the same area uses the same narrow set o f providers. The key notion i s the ratio o f transport costs to the total value o f the service. And the choice o f the use o f a nearer versus more distant facility depends on balancing increased transport costs versus the benefits o f higher quality This is why the catchments area i s small for pnmary 78 schools, larger for secondary schools (including many who are willing to pay to board for higher quality schooling), and larger still for universities. Fixingsize iunsdiction: The first step indiscussing the allocation acrossjunsdictional levels is to fix sizes. The foundational unit is the semce delivery unit where providers and consumers o f semces physically meet. Ineducation this i s particularly easy as the delivery unit i s a pnmary or elementary school. Each school has a catchments area -- thearea from which it i s likely to draw ~tudents.'~ similar approach would apply to a local public or private clinic. A key concern in A decentralization i s balancing the costs and benefits of allocating responsibilities for vanous levels. Such a balance hinges on the sizes o f those junsdictions. Inthis context, Table V.l makes three simple points: The populations of Indian states are as large, or larger, than most countnes in the world. West Bengal would be the world's 12th largest country, just smaller than Mexico and larger than Germany Karnataka and Rajasthan are roughly the size o f France, Italy or Egypt, and considerably larger than Korea, Spain or Canada. Eventhe smallest state under consideration, Kerala, i s larger than 120 of the 150 countnes inthe world. With a population of around two million, even distncts in India are large units compared to some countnes, regions or cities elsewhere. For instance, New Zealand has a population of 3.6 million and Costa Rica 3.3 million; and many o f the new countnes from the former Soviet Union are even smaller. Inmany countnes o f this size, there are vlgorous debates about the extent to which education should be decentralized from the national level. Another companson i s to counties in the USA, which often assume many governmental functions including, inmany instances, schooling. The median size o f an American county in 1990was only 22,000 people, and only eight o f more than 3,000 counties had a population greater than two million. Less than 50 of the 3000 counties were larger than the smallest distnct in the four India states under study (78 1,000). Such compansons indicate that while decentralization to the distnct is often considered to be a very "low" level in the Indian context, we are still lookmgat a large unit ofjurisdiction. The typical catchments area of the services discussed in this Report, whether pnmary education, health or drinhng water, is very small relative to the junsdictional size. In education, for example, distncts have primary enrollments of more than 100,000 students and contain more than 1,000 schools. Evena block typically contains more than 100 schools, and a GP more than one school. ''Theseareas obviouslyhave "fuzzy" bordersand overlap. 79 Table V.l: Sizes and enrollments of state, district and block-level schools State I POP I52.8 I80.2 I 56.5 I 31.8 Economies o f scale H o w many efficiently sized service delivery units, along with their catchments areas, fit in the junsdiction o f a Panchayat Raj Institution (PRI)? In the education example, there are few economies o f scale beyond the single modestly sized school. This i s completely different, say, from urban piped water supply In the latter, the economies o f scale in the networked nature o f service distribution (that is, the pipes that carry the water to households), imply that the economically efficient solution will be large relative to even a large jmsdiction. The economically efficient thing would be to have one set o f pipes reaching any given household or neighborhood; the only question i s the arrangement for the ownership and management o f the assets o f this "natural" monopoly In contrast, declining average costs o f the provision of instructional semces does not dnve provision by a single large supplier (such as the government) covenng a large number o f schools. Are there economies o f scale involved inthe key service o fpnmary education? The typical ernpincal evldence o f cost effectiveness i s not that larger school systems are more effective than small systems. There appears to be a trade-off between many small school systems (which creates benchmark competition among public sector producers), and the potential economies of scale o f larger distncts. 0 There i s a large and growing private market for pnmary education supplied by for-profit producers. If there were economies o f scale to the production o f elementary education, one 80 would expect the emergence o f a few large firms responsible for a significant share of the market -- as the cost advantages would be enormous. Instead, it is apparent that though there are economies o f scale in the development o f learning matenals (e.g. textbooks), or in the setting o f standards, or in the provlsion o f testmg (e.g. testing semces), the typical non- government school system i s a small affair. A single stand-alone school i s not an unusual form o fproduction. Elementary education i s a sequenced cumculum across a number o f topic areas brought together into a single classroom due to economies of scope. Ifthere were economies of scale intnnsic to the produchon technology o f instructional semce~,'~once again, one would expect a few large firms to dominate service provlsion ina competitive market. Yet this does not hold true for any of these industnes. Instead, all these types o f instructional semces are provrdedby an overwhelmingnumber o f independent, competingproducers - as is evldent in any mediumto large sized city inIndia. By analogy from a similar market and production technology for instructional semces, the long history of elite educational institutions suggests an absence o f economies o f scale. In many other industnes, the acknowledged leading firms capture increasingly large shares of the market. In contrast, it appears that the top suppliers o f instructional semces do not expand. While this could be so becausethey are "not for profit" organizations, it is also true that growth is a powerful organizational imperahve that has little to do with profits per se. Another conjecture i s that the basic features o f the provision o f instructional services create disadvantagesfor large organizations that are not offset by lower costs. There are, however, four activities within education that have potentially substantial economies of scale. There are some economies of scale incurriculum design and service standard setting, though the existence o f one-school producers suggests that these are not determinative. When drawing on an already developed curriculum, the incrementalcosts are not substantial. Coordination issues in plans for expansion make for some economies o f scale at the system level, though they are not, perhaps, apparent at the unit level. Again, this i s only a concern in elementary education when the coordination ability of a market mechanism i s weak. For instance, if each individual producer chooses school locabon on the basis o f market potential, this is a powerful coordination mechanism if the market is "thick" or "dense" with a large number of "demanders" in a given area. In rural areas, however, the spatial separation of villages implies some gains to coordination. There are economies of scale in evaluation. Developing test instruments that are valid assessments of the desired learning standards i s a technically difficult endeavor requiring specialists. Hence it i s a costly exercise for any one unit. This would suggest that in a completely open market, there would be small schools and school systems, but a few large testingenterpnses that would contract to these school systems and schools. There are similar issues in the development of textbooks and learning material. There are economies of scale intextbook development, but this suggests there would be fewer textbook manufacturers than there are schools choosing which textbook and textbook series to use. This, in and of itself, i s no reason why the two activrties have to be integrated in the same enterprise/ organization or level o f government. 76But instructional services in elementary education are not different in kind from the instructional services in many other areas - from language instruction to training incomputers to pnvate tutonng for examnations. See Chapter 2. 81 Externalities There are two points relative to the scope o f externalities in allocating responsibilities for elementary education. 0 Most parents in India correctly perceive that the pnvate monetary and non-monetary value o f education i s high. The extent to which parental demand for education i s "too low" because they do not incorporate externalities that accrue at a higher level o f aggregation (e.g. district, state), is almost certainly 0 The two functional areas inwhich there are some externalities are in setting c m c u l a and standards, particularly with regard to the benefits o f socialization and preparation for citizenship; and in M&E, as evaluating multiple junsdictions on the same standards can provlde citizens with valuable information. 77There is a great deal o f debate about whether there are externalities to education at all, and there i s a rough consensus that the economzc externalities are small relative to the private benefits o f education (see Pritchett, 2005). This is not to deny the many non-monetary benefits of education, such as the effect o f the mother's education on child health; but generally, these are not "external" to the household, and the externality impact i s likely to be lirmted to a local level. 82 Table V.2: Public financefirst principlesfor responsibility allocationacross government levels ineducati ~ gb 4E. Policy Cumculum design NIS design/ Learmngachevement standardslgoals NIS standards Plamng Plansfor physicalexpansion Plansfor quality improvement Asset HumanCapital Creation Teacher Tramng D S/D SocialCapital PhysicalCapital School Construction D/B/GP Operahon BeneficiarySelechon GP Choice of students for targetedprograms NIS Promohonof universalenrollment NIS Recurrent expenditures(non-wage) Textbook choicelpurchase SPIG SPIGP Leamngmatenals P SPIGP Mamtenance SPIG Mamtenanceof schoolbuildingslfacilihes P SPIGP Personnel Compensahon SPIG SPIGP Hirmgofteachers P SPIGP Assignment ofteachers to specific classes SPIGP Performanceevaluation SPIG SPIGP In-servicetraimng P SPIGP Promotionacross grades SPIG P SPIG P SPIG P D D Momtormg Monitormg and Individualstudent progress SPIG SPIGP Evaluation Schoolprocesses P SPIGP Evaluabon SPIG Assessment of learnmg achievement P NIS DIS N=national (= 10 million), S=State (>5 million), D=Distnct (arou 1-2 mi (around few hundred thousand), GP=grumpanchuyut (around a few thousand), SP=Service provision unit (catchments area). 83 Both central and state governments have always been committed inpnnciple to universalpnmary education - a goal enshned in the Indian Constitution. Given this commitment, elementary education cannot be planned to be less than universal. The reality, o f course, does not quite measure up to this ideal. A commitment to equalizing educational opportunity may imply financial support to education at a geographic scope that allows for redistribution o f resources. But it does not imply anything about the size or ownership o f the units responsible for the production o f education. Heterogeneity Heterogeneity i s limited by the fact that all communities must at least plan for universal enrollment. But different communities or schools make different choices about the approach to instructional styles, pedagogical practices and the general vision o f the school. Research around the world shows better performance when educators perceive greater autonomy in putting their vision into practice in school envlronments. Obvlously this can happen only when individual schools or sets o f schools are allowed to choose their approach, when teachers are allowed to move, and leaders o f schools allowed some flexibility in choosing staff. Moreover, there i s heterogeneity in the demand for the socialization element o f schooling. As pnmary schooling i s an important component o f the socialization o f children, these elements are important to both parents and governments. This can create a tension between "centralizing" tendencies wanting to impose a socialization that creates a common national identity (language, culture, history, or "ascriptive identity") and more local forces that want schooling tailored to local preferences on those same dimensions. Accountability criteria applied toprimary education The vanous functions elementary education can be assessed on are the cntena of discretion, transaction-intensity and observability This will help inform what i s the appropriate junsdiction for these functions/activities based on the cnteria o f accountability This i s done in Figure V.l below, which points to the fact that highly discretionary, transaction-intensive, and locally observable activlties like Asset Creation and Operations (especially teacher operations) should be done under localjunsdiction. The operations function o f elementary education also falls into this category - it requires lakhs o f teachers teaching crores o f students in lakhs o f Indian villages (transaction-intensive), and what and how the teacher teaches, what pedagogical technique he/she applies depends to a large extent on the teacher (discretionary). What about observability? Inthe context o f pnmary education, observability i s a combination o f (a) actually obsemng the action and (b) knowing whether the action was appropriate. For instance, teacher absenteeism i s a particularly easy case. Ifthe teacher i s intended to be present and i s not, then it is clear to any direct observer o f the school (e.g. students, parents, other teachers) that the appropriate action for learning performance i s not being undertaken. But someone who i s not physically present cannot observe absenteeism. A more difficult case i s pedagogy For instance, "child centered 1eaming"- easy to observe, hard to know if it i s the "nght" technique that a teacher should be applying. Are parents capable 84 o f knowing whether a teacher i s doing a "good" job based on available observation? There are some who argue that locally elected officials should not have any say over schools because they might have "backward" attitudes relative to the "advanced" wews o f educational "experts." However, on the scale o f "observability" our judgment i s that a parent who wants to be informed can form a reasonably good idea o f the quality o f learning activity in a given school (especially given that the major problem i s that half o f the time no learning actmty at all i s going on, which is easy to observe). Figure V.l: Functional Allocation According to FirstPrinciplesof Accountability I Function I AccountabilityFirst Principle I Who Can BestInfer Discretionary? Transaction Intensive? (Technicalor Local)? StandardsSetting Low Low Technical Plannn g Medium Medium Local/Technical Asset Creation High Medium Local/Technical Operation - Nonteacher High High Local Operation Teacher - High High Local/ Technical Monitoringand Evaluation Medium Medium Local/ Technical The responsibility for setting standards, deciding on curriculum should remain at the state level (incollaborationwiththe centre). Ina decentralized environment the important additional responsibility of the state is to create a system o f monitonng the achievement o f the learning achievement standards. The state level government assumes responsibility for setting learning standards and goals and for monitonng progress towards those goals for each school, GP, and distnct. The Gram Panchayat (worlung with school specific or villagehabitation specific committees as member bodies o f the GP) should be responsible for proposing new school facilities, for supervising the construction o f such facilities, for all operational decisions at the school level (over maintenance, budget, utilization o f non-wage expenditures, purchase of learning matenals, etc), and for the assignment o f teachers to specific schools. 85 The overwhelming bulk of the resources to education should then flow pnmarily in untied form to GPs who, together with the school staff and the school andor village specific (VEC) or school specific (SMC) committees to make essentially all operational decisions'* The distnct should play a facilitating role o f (a) providing the pool o f qualified teachers fiom which GPs can choose (b) providing for the technical aspects of improvingpedagogicalpractice through in-service training, (c) coordinate the planning when necessary to prevent an over- expansion o f the system, and (d) supervise compliance with the processes of GP and school decision malung. FigureV.2 and Table V.3show our recommended allocation of functions and actimties for the vanous junsdictions. Figure V.2: RecommendedFunctionalAllocation of Responsibilitiesfor RuralPrimary EducationinIndia First Principlesof Public Finance First Princides of Accountabilitv FUISlb" I PVUk R-RnIRhcbb ' honornk OfEmm*lln H . - ~ d ~ o f sc* s y s t m e Eqdh D e m d States do Standards ngand itoring 0 PRlsassume responsibilityfor actual Operation - Asmuchas possibleas low as possible - Higher PRI tiers back-upon professionalism, technical 78 The scheme m West Bengal chose to use SMCs (which are school specific) msteadof VECs (which are village specific), which many believe work better as a tool for ensunng teacher accountability. It is also possible that SMCs are less politiclzed (and more focused in its role) than VECs (which are much more closely related to the Gram Panchayat and hence to local politics) and hence a GP with SMCs would providethe right mx. 86 Table V.3: RecommendedAllocationof Activities for RuralPrimaryEducationinIndia Standard Cumculum design State Setting Leamng achevement standarddgoals State Planning Plans for physical expansion GPlDistrict Plans for quality improvement GP with support fromDistrict on technical issues Asset Social Capital Creation Capacity building o f local committees District Physical Capital School Construction Operation Beneficiary Selection Choice o f students for targeted SchooYSVC/GP programs SchooYSVCIGP Promotion of universal enrollment Recurrent expenditures (non-wage) School/SVC/GP Textbook choice/purchase SchooYSVC/GP Learning materials Mamtenance School/SVC/GP Maintenance of school buildings/facilities Teacher District (on GP recommendation) Hiring SVC/GP/District Assignment District Training District (structure), SchooYSVClGP Salary (Payment) Supervision/ Performance evaluation SVC/GP/District Dismissal Monitorin Monitormg g and Individual student progress SchooYGP Evaluation School processes DistrictMate Evaluahon Assessment of learning achievement SVC-scho or village committee (operating in concert with the GP). ~~ 87 Table V.4: Firstprinciplesfor allocationof functional responsibilitybasedon accountability ineducation U E' j 4 3s 0 -!+ T g g O B v l v l =;'w 4s 0 c3 q g = % E 3 : -! 9 7 2. CD Standard Cumculum design Low Low Techmcal Setting Leamng achevement standarddgoals Low Low Techmcal Planning Plans for physical expansion Medium Low Mixed Plans for quality improvement Medium High Mixed Asset Creation Physical Capital High Medium Mixed School Construction Operation Beneficiary Selection Choice o f students for targeted programs High Low Local Promotion o funiversal enrollment High High Local Recurrent expenditures (non-wage) Textbook choice/purchase High Low Mixed Leamngmatenals High Low Mixed Mamtenance Mamtenance o f school buildings/facilibes High High Local Teacher H i m g High High Local Assignment High High Local T r a m g High Medium Mixed Salary Low Low Local Supervision/ Performance evaluation High High Mixed Dismssal High High Mixed I n summary: towardsa coherentprimary education delivery system The current system does not give autonomy to the front-line service organizations and providers; nor does it create accountability for performance. Thus all education reform proposals should be judged against the cntena o f cost-effective improvement o f the level and distribution o f learning achievement. The key problem in elementary education in India i s the high per student cost o f instruction. A greater push towards decentralization would create an opportunity for both substantial cost reduction and quality improvement -- but the political costs o f the reform will be highand the penod o ftransition long. Overall, the analysis indicates that the following allocation o fresponsibilities would help inthe move to a more coherent rural education delivery system: 0 The operation o f the schools should be pushed to the lowest possible level with the greatest possible autonomy -in financial control, the assignment o f teachers, and asset creation. 0 The district should be responsible for planning, coordination o f asset creation, hinng and providing technical and pedagogical support to teachers. 88 any Modelsof E d ~ ~~ a~ ~t e~n~t r~~ have nott worked well l i ~ a ~ ~ n o~era~iona~ controlovertea ina) Decentralize '\ India Experience ~ ~ 9 9 5 - ~ ~ ~ ~ ~ Functions,without -Devolved functions ~ i n cprimary educa~~on~ - i ~ ~ ~ ~ ~to PRls... devoivingfinance .But PRls do not have adequate finance or control over ~ ~ ~ c ~ to~~ rnn a~ lr~trtheir~~mandate i n~ e t Bust'*X2 Incre Py of E iit servictrs ts an health star P o ~ u l a t ~ o ~ i - ~ ~ s e d actwitm: 'IVe have c a t e ~ o ~ health services as p ~ ~ p u l a ~ ~ ~ nactivities;d ~ e d - ~ a s e pretentive and promot~~e act~vit~es~ primary health carc; h o s p ~ ~based carc; and s ~ r ~ e ~ ~ ~ a ~ ~ e a ~ (see Chaptcr 2). illrlnust all sertwcs in tkc first catcgory address c ~ ~ n ~ ~ i dtseosesct kao~i ~~~eh ~ n ~ ~~~a~~~ not fall under the a u t h o ~ iof' health policy makers. The degree of market fidure must do t ~ ~ be balariced against the ability of the public sector to implement the service in ytiesttori In this case. since the privare market is totally ~ n c a ~ aof~~~ er o ~public goods,~it ~r ~ m ~a~ ~ i s ~ ~ d i g u ~ ~ er ~e si ~ ~e n~ sregardless~ of~how d ~ f ~ ~ttc ISL ~to~~t n ~ ~ ~ eThecsituation 1s a little ~i ~~ ~ i ~ ~ n t . more ~ ~ p ~ thowever, siricec iiiany of the serwces in the first catcgory, such as swamp ~ ~ ~ t i ~ drainage and water s i ~ ~ pilnyf r a s ~ ~ ~ carc~"one-shot" tntertentions that are not as d i f ~to~ ~ ~ ~ t ~ r c , a c c o ~ p l i as~regular care. s rl Inilia. Once again, rhe ket failure and modest The secund cha~ac~erjmarket fatlure is the almost universal absence ofh e 2 1~~sur2rice. s ~ i ~ ~ ~ that people would be willing to pay, but cannot, because o f the lack o f health insurance. InIndia, rough calculations on this welfare loss have been calculated for different levels o f service.83 Using average costs inthe private sector as a benchmark, the welfare gain for covenng the typical hospital stay was fully 60 percent o f the expected cost o f the services4in companson with a relabvely poor person's income, as opposed to five percent inthe case o fprimary health care. The former i s certain to cover administrative costs for insurance; the latter i s unlikely to. Insurance programs are notonously hard to administer, however. While it i s certainly the direction inwhich India's health care system should be going,85 it i s probably very difficult to adopt at this time. In the absence o f insurance, the main public policy available i s capped fees in public hospitals.86 Publicly provided hospital semces compnse a reasonable, second best, option - if the incentive problem discussed above i s solved. But the ground realities o f public/ pnvate involvement inthe two categories o f curative care - pnmary health care and hospital-based care - need to be taken into account. Public/ mvate involvement inm m a r v curative care: Most curative care i s already in the pnvate sector (see Chapter 4). M o s t people can afford to pay the small amounts required for pnmary care. Very few can afford to pay the much larger sums associated with hospital care. There are state differentials in the relative size o f the pnvate sector in pnmary and hospital care, but in general, the share o f the pnvate sector i s much larger in outpatient than inpabent care - a direct result of the inadequacy o f the insurance market. On efficiency grounds, then, pnmary care services do not rank highon the list o fpublic responsibilibes. Imdementation abilitv. If pnmary care were easy to accomplish, the limited role o f government in prowding cheap curative care might be overlooked. But there is considerable ewdence of severe implementation problems in pnmary health care. A recent study, for example, showed absentee rates of approximately 40 percent across states when surpnse wsits were made (mostly) to pnmary care facilitie~.~'In short, the incentives to ensure attendance in public facilities are inadequate. On grounds o f implementation ability, hospital services also have an edge over pnmary care. Absentee rates are always lower in larger facilities because it i s easier for senior staff to monitor the attendance o f more junior staff. Also, there are non-monetary benefits to serving in hospital settings for medical professionals. Job satisfaction is higher when providers can use the s h l l s they are the The size of marketof problems failure and public trained for and have colleagues to confer with.@These are present inhospitals far more often than inPHCs. implementation capacity call for a strongemphasis on traditional health activities. Policies Distributional grounds: The one main drawback of public public have also to address the involvement in hospital care i s on distributional grounds. In handling of the catastrophic general, the pattern ofpublic subsidy across the states has been costs of hospitalization. regressive (see Chapter 4).89 The poor rarely use public 83World Bank, 1998. 84The actuarial fair cost of insurance. 85Peters et al, 2002. 86Optimal pncing for insurance is to have a small deductible and a cap on expenditureupto a maximum corresponding to "adequate" treatment and full payment for expenditures beyond that maximum. See Zeckhauser, 1970. None of the four study states is nearly as bad on these grounds as Bihar or Jharkhand, but all of them show fairly similar figures. This may be a little surprising given the good reputation that Kerala has inrunning its health system. See Chaudhuryet al, 2004. WDR 2004; FeedbackVentures, 2004. 89The case of Kerala in these figures shows that the bias against the poor i s not inevitable. The state seems able to charge higher fees to richer people as well as have a larger private sector with a much larger share of better-off people. This leavespublic hospital care to serve the poor. 93 hospitals, mainly because hospitals tend to be in urban areas where incomes are higher, but also because referral systems to higher-levelpublic facilities tend to be biased against the poor. In sum, based on both the size of market failure and the problems o f public implementation capacity, there should be strong emphasis on traditional public health activities and senous policy discussions on how to handle the catastrophic health care costs that hospitalizations impose. H o w these services should be diwded between different levels o f government is our next concern. Servicesand the appropriate level of government Population-based semces: case-bv-case basis allocation Population-based public health services comprise a mixed bag as far as the vanety o f services i s concerned, as well as the most efficient scale for such services. Therefore, responsibilities for specific programs will need to differ across levels o f government. For example, many mosquito control activities such as noticing and eliminating areas o f standing water where mosquitoes breed i s as local a "public good" as imaginable. Similarly, water quality can vary substantially between sources, sometimes within the same type o f source such as public standpipes invillages. Testing, maintenance and other remedial action (arsenic control inWest Bengal, for example) all require substantial local responsibility and control. Other services in this category are subject to substantial economies o f scale or to spillover effects -- where actions in one village, or even larger junsdictions, can affect its neighbors, and need to be monitored, coordinated or implementedby some higher level o f government. One idiosyncratic form o f "scale" inthe health sector is the need for relatively scarce, highlytechnical personnel handling specialized problems. Investigative epidemiologist^,^^ or some laboratory testing facilities (particularly for reportable infectious diseases), are not likely to be available or desirable in villages. It i s not possible, therefore, to be too sweeping and dogmatic injudging the appropriate level o f government for these actiwties. They must be treated on a case-by-case basis, balancing the monitonng ability o f different governments against considerations o f technical efficiency o f scale. The two justifications for the involvement o f higher level governments - spillover effects and scale economies -- differ on one important dimension relevant to hnding and the direction o f authonty. The former i s a Population-based health services purely "supply-driven" concern o f the higher level o f :'$& must be treated on a Case-bY-case government. The higher level sets standards and guidelines to ensure action by individual unitso f lower government tiers. So against b oconsiderations ~ ~ a of ~ ~ ~ ~ ~ e g sanitation disposal of an upriver community for the benefit of technical efficiency of scale. the downnver community must be handled without the upnver community (at least) aslung for help. However, testing and similar support services where the scale spans indiwdual GP or BP areas may benefit from "contracting up" either to a higher level o f government, or even to pnvate firms. Talung blood smears to check for malana, for example, may have to be done by personnel responsible to local government (though reporting the results may have to be ensured by the higher tier). However, whether the tests themselves are done in any particular government laboratory, or ina private one, could be a choice by the local government unit on the basis o f cost, reputation for reliability, speed o f results and other service charactenstics. This choice - a "demand-dnven" dimension, though the demander is, itself, a level o f government - introduces some competition that can keep costs l o w or reduce delays in reporting. A higher tier of People who track the emergence, spread and causes of disease outbreaks. 94 government may play a facilitating role to help GPs cooperate and obtain these services -- aslong as it does not steer the GPs to use government facilities when pnvate options are cheaper. Local ability to contract up makes use o f a short route o f accountability The local government takes the role o f the client with the highest stake inthe outcome. It can monitor providers more closely and get better results if it is their money being paid. Whether this can be done in a purely market context, or whether separate GPs (or higher) levels need to band together to generate sufficient demand for a provlder, depends on the nature o f the service. Again, "contracting up" may be another way o f malung doctor services available in areas where indivldual villages are too small to support a qualifiedphysician. Surveillance: involvement o f higher tiers Surveillance i s another important public responsibility that requires the involvement o f hightiers o f government, either exclusively, or with substantial guidance o f and cooperation with other tiers. Again, there are two aspects o f surveillance - one that is widely recognized in current legislahon; and one that i s essentially left undone. The standard surveillance o f disease, particularly o f communicable disease, can be done in different ways that will interest different levels o f government. Since diseases do not respect political boundanes, state and central government institutes need to be involved and informed o f the incidence o f diseases so that epidemics can be averted and information flows back to people. This i s particularly useful if diseases are preventable. Protection from insects i s a long-standing example, and information concerning the spread (and means to prevent) HIV/AIDSi s a more recent addition. What i s the surveillance expenence in other countnes? Inncher countnes, the current practice i s for epidemiologists in public semce to undertake routine surveillance and investigate unusual events. These activities occur at different levels of government (usually at all levels o f government as part o f a systematic reporting procedure), in accord with local vanations in economies o f scale. Closer to India, there i s also the success story o f monitoring doctor attendance inPalustan. InIndia, standard surveillance i s hampered intwo ways: 0 While some states have laws that require pnvate providers to report cases o f particular diseases, this i s not well enforced. Given the predominance o f the pnvate sector as the first contact with health care, and the daunting task o f regulating the sector, this i s a senous obstacle. 0 There are legtimate concerns about the diagnostic ability o f medical providers in bothpublic and pnvate sector^.^' The first problem calls for the involvement of a government representative - either a functionary from the higher tier o f government assigned to local government offices, or employees o f local government with clear reporhng responsibilities to higher levels. Which level o f government officials are best able to disseminate information to individual pnvate and public providers is unclear. The second major aspect o f surveillance has barely begun inIndia. It involves the regular, routine and systematic collection o f information on disease incidence, mortality, and the wide range of possible determinants o f disease and death. Discovenng the factors that cause illness, and which can be influenced by policy, i s a clear "public good" across government levels. Since health determinants are varied and not limited to health services, this monitonng will have to be based on multi-purpose surveys that can identify education, income, sanitation and water among many See Das and Hammer, 2005, though this IS based on a study in an urban area 95 other householdcharactenstics. No single local body would (and inmost cases, could) undertake researchthat could be useful inareas other than their own. Thus it is the responsibility o f the state and central governments to collect, analyze and disseminate the results to local officials and prowders. Preventive and promotive health services: the uerfect case for decentralized authonty Preventive and promotive health services are the most amenable to decentralized government authority -- both in principle and in the light o f political constraints. On the one hand, these sermces are inthe greatest need of public intervention. On the other hand, local variation i s often important for setting pnonties. One substantial constraint on improved health among the rural poor is the persistence o f harmful traditional health practices. Standard Information, Education and Communication (IEC) interventions are important to counter such beliefs, but they must be tailored to the specific area. They are also most effective if carned out by people who pare part o f the community and have their trust. While support in the form o f basic information and suggestions on dissemination techniques might come from higher levels o f government, this is clearly a local responsibility One area o f preventive care that could be an exception to local government junsdiction i s immunization. The campaign style o f immunization appears to work, possibly because the value o f this semce does not vary much across jurisdictions, malung it easy to monitor even by relatively highlevels o f government. And, once again, there are cross-border effects o f mishandledvaccine -- preventable disease. Drugpurchases could be done at higher tiers o f government ifbuyinginbulk allows the public sector to get lower pnces. Otherwise GPs could do it. Technically difficult drugdistribution, such as vaccines that require a cold chain, are more easily provided or regulated at higher levels o f government with or without extra subsidy (above overall devolved funds). But even inthis case, the cooperation o f a locally employed and responsible health worker supported by a higher level o f government is a more likely route to success. Curative care: Dnmary health semces "Money follows the patient". We have seen that the role o f provision o f pnmary health care by government - local or otherwise - i s debatable. Even the financing o f cheap interventionsg2may not be a high pnonty, since people already pay substantial sums for curative care, and there i s considerable pnvate demand. Governments may want to prowde financial support to the poorest, butthe pnnciple o f "money follows the patient" alwaysrequires government to cede the choice o f prowders to individuals. This i s consistent with the need for direct accountability to the entity best able to represent the public's interest. Inthe case o f primary care -- an excludable and nval good - this entity i s likely to be the patient herself. One common problem associated with government health facilities and their monitonng i s solved as a by-product o f lettmg people choose facilities. This i s the problem o f "catchment areas" o f pnmary health facilities not being coincident with political boundanes. Ifpeople can go where they want and the money goes with them, the optimal location o f facilities will probably anse from providers moving so their location can be convenient enough to attract the most clients. This obviates the need to decide which local government should monitor which facility 92 Inrich countries, while the vast majority ofpruvzszun of services is done bythe private sector, the ultimate source of finance is almost always public (with the United States the big exception). 96 The exceptions - arrruments for local control: Two exceptions to the general rule o f money following the patient argue for local authonty over PHCs. One exception is ambiguous in effect butmay argue for amonitonngrole for higher levels ofgovernment. 0 Many rural areas are too small to support competitive markets for qualified doctors. In that case, local government (or a User Group with this delegated authonty) may want to choose a provider on a regular, perhaps annual, basis on the understanding that continuation o f the contract i s contingent on acceptable service. Inthis case, there is competition for the market (that at each moment may be a monopoly), rather than competition in the market that i s not sustainable in sparsely populated areas. Whether the provlder (now a monopolist) i s paid by salary or by patients, or a combination, i s a matter o fnegotiation. 0 The only exception to the rule that pnmary care provlders in rich countnes are never paid with salanes is Sweden.93But Swedish medical provlders are employees o f local government -- based on the idea that it i s the only level close enough to the point of semce to monitor performance adequately. Even then, dissatisfaction with medical services has lead to innovativereforms that increasecompetition and the rule o f money following the patient. The reformis limitedto Stockholm, which has a large enough market to support ~ompetition.'~ International compansons on the decentralization o f health semces, particularly of health care, are hard to make for two reasons. First, nch countries had a completely decentralized system before they amalgamated separate pnvate practices and pnvate insurance systems into national systems. They retain a totally decentralized system ofprovision of services - decentralized down to the level o f individual choice of provlder, with only thePnanczng of health care undertaken by government at a level highenough to spreadnsks. Second, inpoor countnes with systems similar to India's (because of a colonial past or the recent adoption o f a centralized government system), expenences of actual decentralization are either uncommon, or too recent to be fully evaluated. In either case, they have a mixedrecord, and these successes and the failures do have some lessons to tell. As far as decentralization goes, the main evldence infavor from other countnes i s indirect. There have been a great many experiments with alternative methods of changing peoples' health related habits. But the conclusion i s that the degree of expenmentation, flexibility and range of options that can lead to better options requires the autonomy to try new ideas. It is in this sense that GPs can be the "laboratones" for the states and the nation. 93Here a distinction that is important inthe framework o f the WDR 2004 must be drawn. The lack o f a salary payment applies to the provider organization and not necessarily to each individual within the organization. For example, a private practice can require several people on staff and may pay any or all o f them with a salary. The nurse that a doctor hires to assist her in the practice mght be paid a salary. This is an internal management decision and a choice of the organization. The payment o f salary i s likely to be chosen and be a successful means o f remuneration if close monitonng by a manager (in this case the doctor) i s possible. 94At three million people, Stockholm is slightly larger than the average district in India and half the size o f Midnapur inWest Bengal. It was considered too large an area to be adequately monitored by government. This was in a country with among the most competent and honest governments in the world. Transparency International ranked it as having the sixth most honest government. India places 90th inthat ranking. 97 98 distncts for secondary hospitals, and states for teaching hospitals with subspecialties. In the absence o f health insurance, hospitals are best handled outside the GP. The GP may, however, want to be responsible for emergency transportation to such facilities. The issue o f transportation raises a further difficulty. the rules, procedures and incentives for proper referral to hospitals. Referrals: Most people do not use a hospital as their first point o f contact with the health system, though there are many except~ons.~'The bypassing o f lower level facilities, particularly by city dwellers, i s one reason the distribution o f public subsidies i s skewed so badly towards the relatively well o f f (see Chapter 4). Achieving the most efficient sorting o f patients means handling the problems o f referral, as well as the first-line prowder level at which treatment i s cheaper. Two lundso fmistakes can be made: 0 Referrals may be made too readily when prowders are either not available or do not want to deal with the patient; and 0 Referralsmay not be made, even when appropnate, because the prowder does not want to lose a customer. The first i s charactenstic o f public pnmary care centers, the second o f the pnvate sector.96The critena for referral are made at the state level, which i s perfectly reasonable. H o w this gets translated into prachce is more complicated. Even if the local government were able and willing to use some o f its funds for transportation, they need to be in a position to judge whether the case warrants the referral. Insurance: If health insurance does become an important part o f the Indian health system, it i s also likely to be managed at a fairly highlevel o f government. The benefits o f insurance are the result of the much smaller vanability o f overall costs for a large population, relative to the very highvanability an indiwdual or a small group would face. Also, since infectious disease is still a major health problem, there i s substantial covanate nsk for small groups o f people subject to the same problem. Epidemics compnse the most dramatic example. The more covanate the risks, the larger the insurance pool needed to diversify that n~k.9~ However, each o f the study states i s large enough to reap practically the full benefits o f diversification. I n summary: optimal allocation of responsibility with accountability The causes o f centralization To date, most aspects o f health, particularly health care, have been state and central government responsibilities with little role for local governments. In spite o f the fact that several o f the 29 functions delegated to the PRI institutions relate to health, the states and the center have been reluctant to yield much authonty to them. Health has remained centralized for two major reasons. Linkwith other allied services: Healthcare is only one o f a large vanety ofpossiblesemces that improve people's health. The indications are that other services - safe water, sanitation, health 95 Accidents, heart attacks, urbanites near a hospital, and people who mume that they will either be referred to a hospital or that a PHC will not have drugs or staff available. The literature on how people decide on how to seek care i s quite large and often points to quite sophisticated strategies concerning expectations about where they will get the most a propriate treatment. See Leonard, 2004. ''"DasIn and Hammer, 2005. case of a perfect correlation o f nsks, no population size will lead to the diversification o f nsk needed to justify insurance. In the recent tsunarm, for example, the GO1 and the government o f Tamil Nadu were necessary to provide protection to everyone inall the districts affected. 99 and nutntion education as well as other preventive and promotive services - influence health status as much as, ifnot more than, health care. The relative importance o f each of these services should, and does, vary substantially across geographic locations. Delegating responsibility for varying the mix o f services to local areas is, clearly, a core aspect o f the health sector. Oddly, it i s not a charactenstic o fthe Indianhealth system. Money not following the patient: The system o f paying doctors and other health care providers by salary, and from a level o f government far removed from the point o f service, is an anomaly that stands in the way o f any meaningful r e f o m in the sector.98 The role that PRIs can play in improving the health system i s tightly circumscribed by this overnding problem, particularly since the lion's share o f publicly spent health resources goes to curative care. But keeping in mindthe Kerala expenence and some recent innovations o f local hiringo f medical personnel in West Bengal, some tentative options for expanding this role have been included in the following l i s t o f guidelines. Expanding the role of PRIs: someguidelines Block m-ants to GPs: Block grants, much bigger than those currently given, can be managed by GPs for the health and well being o f their constituents. Concern that GPs may ask for unwarranted work from local health workers99 can be handled by recourse to higher levels o f government for technical as well as moral support. The monitonng o f outcomes and standard auditingpractxes will also have to come from higher tiers o f government. Public money for "public goods". Overall, central and state governments should spend more money on real public goods that cannot be provlded by the pnvate sector, such as the control o f communicable disease and increased surveillance. Combating communicable disease with more public money i s a two-part process: large- scale pest control that i s either monitored and regulated, or implemented, by the appropriate level -- state, district or block; and small-scale, transaction-intensive preventive and promotive activities, such as health education and immunization. These small-scale activities are definitely GP responsibilities, given the variation o f needs across areas, and the greater effectiveness o f information dissemination among known networks o f people. Technical support for local health educators, such as teaching materials, could come from higher tiers. But the workers themselves must be paidby, and be accountable to, local governments. Increased surveillanceio0 requires data collection from households and not simply from management information systems, since so little curative care i s performed in the public sector, and since many determinants o f health status are found within the family This data should be analyzed to determine the causal links between policies and health status, and widely disseminated inan easily understood form. Where i s the extra money for these public goods to come from? One possibility is curative care, particularly since the pnvate sector i s thving inthis area. '*By default, much of the healthcare systemdoes abide by the principle of money following patients.The evidence is there, for example, inthe ovenvhelrmngshare ofthe pnvatesector inambulatorycare. 99 Anganwadi Worker and Auxiliary NurseMidwife (ANM) in the current system, though these responsibilitiesrmght change as local governmentsdecideontheir own m x o fneededservices. loo Standard surveillance, the process of identifyng and reporting diseases to control epidemcs, is already being expanded. 100 Pnmarv health semces - long-term and short-term possibilities: As far as government role in curative care i s concerned, there i s a long term and a short-term set o f actions that should be considered. 0 As India becomes ncher, it is likely to pay its provlders for the semces delivered. This will reduce absenteeism and improve at least the non-clinical aspects o f care if a competitive environment i s created. This change involves political problems inthe short run,but such a reformis necessary for improvement inthe public health care system in the long run. 0 There are several short-term possibilitiesavailable, depending on the level of semce. For pnmary health care, two models (with modifications) are possible. The first is the Kerala model inwhich GPs take on substantial responsibilities for pnmary care operations. The modification necessary i s putting the GP in charge o f pay, hinng and dismissing o f personnel. The second i s the model being introducedin West Bengal - inwhich GPsjoin together to hire the previously unaffordable a n d or unavailable semces of qualified doctors. This can be a lump-sumgrant for the doctor's vlsits, monitored by GPs or their delegated health committee; but with patients' fees charged on a sliding scale (or any other arrangement mutually agreeable to the provider and the GPs). Alternatively, the GPs may choose to pay for all semces from these provlders -the discipline coming from renewable contracts. If either o f these approaches i s adopted, funding must be shifted to GPs from states through block grants. Hospital services: Until insurance i s a vlable option,"' public hospitals will need to continue to be run by levels o f government commensurate with the type o f services offered, with higher levels talung on the more specialized services. The role o f GPs i s limited. And so i s that o f the state government -- as only the most specialized teaching hospitals require direct state-level involvement. The state can help with setting fees as well as better referral guidelines for hospital admission. Treatment can be free (or much cheaper) for those with appropnate referral. The GPs may want to use part o f their block grant to provlde emergency transport incases o f accidents and complicateddelivenes. Water supply and sanitation The domain o f water supply involves not only water service delivery, but also water quality and the development and regulationo f water resources. The sanitation service involves environmental hygiene, excreta disposal, wastewater management and solid waste management. High-quality RWSS aims at providing access to these two services talung into account the full range of this domain. The form o f water supply provision vanes with settings - in some settings, a quality water service may be makershave to beinvolved most reliably provided through individual access to in defining the quality of household wells. hother settings it may be hand pumps, parameters, and in providing an environment in which those or mini-piped schemes or regional schemes. These forms parametersmight be met. o f access should be pnmarily defined by user preferences, willingness to pay, collective involvement in ownership and management o f schemes. Inthe case o f sanitation, there i s an even greater need to define the lund o f access necessary for a high quality service - given that the present emphasis on latrine provision has not had the desired results. A high quality sanitation service requires a focus on sanitary outcomes as being the provlsion o f a clean and healthy sanitary environment. It i s inthis context that policy makers have lo'Insurancewould have to be admnistered at a high level of government. 101 to be involved in defining the quality o f access parameters, and in providing an environment in which those parameters mightbe met. Applying publicflnance criteria to rural water and sanitation Economies o f scale Approunate unit o f production: What do economies o f scale tell us about the appropnate unit o f production for RWSS? Larger regional water supply schemes theoretically have the potential to capture economies o f scale and facilitate higher levels o f semce at lower costs. They could also offer long-term solutions to the ground water scarcity and contamination issues faced by many regions. However, the expenence in RWS in India suggests that the potential for improved productive efficiency o f larger regional schemes has been more than offset by failures to achieve allocative efficiencies in semce prowsion. There have been significant increases in the expertise levels and costs associated with more complex schemes, higher coordination costs, greater legal requirements, greater likelihood o f leakages and the higher transaction costs. Moreover, larger World Bank reviewio2 o f the RWS sector in India suggests that the typical per capita costs for and more complex schemes do not necessarily incur financial economies o f scale inrural India. A regional water supply schemes are significantly higher than those for "small" or "mini" piped water supply schemes. The choice o f increasing the unit o f production in rural water supply in India is generally not dnven by financial considerations. Water quality problems, water resource scarcity a n d or a desire for higher service quality constitute the pnmary driver for larger schemes. Centralizing design and production aspects: There are some economies o f scale associated with centralizing the design o f programs, the setting o f standards and the development o f standard designs in RWSS. This i s also true o f more complex forms o f water quality testing, and the development and production o f IEC materials in sanitation. But in a more comprehensive approach to sanitation, the extent o f these economies o f scale may not be so significant. There are also some economies o f scale associated with centralizing the production o f sanitary facilities through rural sanitary marts (RSMs). West Bengal has had considerable success in centralizing the performance o f this function at the block level and managing this predominantly from the distnct level. The service i s outsourced to NGOs, hence a higher tier o f government i s better suited to manage the function. Externalities Auuropnate level o f functional assignment: What can we infer from the externalities in RWSS about the appropriate level o f assignment o f the functions in service delivery? The most relevant externalities are related to health impacts o f poor water quality or lack o f sanitation. In water supply, the scope of the externality is dictated pnmarily by the size o f the scheme and proximity o f habitations. These health externalities are likely to be reasonably internalized at the village or habitation levels, since within these people are concentrated, but they are far from other villages. IO2World Bank, 1999. 102 Equity The central government and vanous state governments are committed to delivenng a basic level of access o f dnnlung water to all citizens. GO1norms define such a level;lo3 in fact, many states have adopted higher standards o f guaranteed access. The demands for a basic "safety net" may call for some cross subsidies for the poor, particularly in areas where costs are prohibitively high. But this does not necessarily infer anything about the size o f the unit o f production. Ensunng equity o f access to sanitary outcomes i s a pnonty in the design, delivery and monitonng o f sanitation programs. For instance, the promotion o f latrine facilities through targeted subsidies for BPL households has led to these families investing some o f their own resources to construct latnnes. But in the many situations where the above the poverty line (APL) families have not invested in latnne facilities, the situation for the BPL families is not necessarily any better - especially where the BPL households are located adjacent to the "open defecation" sites. Equity in access to solid waste management "outcomes" implies that the public element of garbage disposal i s equitable, rather than the prowsion o f a particular service. For instance, a fixed charge on all households for waste disposal services is dispropofionately loaded against the poor, since they generate less solid waste than the nch. Equity in access implies that the garbage disposal sites are not adjacent to SC and BPL landholdings; and that street sweeping or garbage disposal services (where prowded by local government) are available in all areas. Equity in access to bathing facilities and bunal sites similarly implies that SCs/ BPL families have commensurate access to facilities, though not necessarily to the same facilities. Heterogeneity The need for RWSS i s homogeneous in that everyone requires access to dnnlung water and a place for defecation; but the demand for facilibes i s anything but homogeneous. People's choices about the form o f their access are heterogeneous. The combined need to manage externalities while ensunng equitable access suggests multiple responses to heterogeneity o f demand - including the provision o f a range o f choices in the means o f accessing services; and the decentralization o f decision-malung on the form and the functioning o f the system so that the "collective" i s able to respond to specific local demands. Functional assignment for RWSS on the basis o fuublic policy pnnciules What do public policy principles infer about the allocation o f responsibility for rural water and sanitation? (See Table V.5 and Table V.6.) 0 The planning, program design and standard setting functions for RWS have to be sufficiently centralized to establish an appropnate means o f promoting equitable access. 0 The M&E functions for RWS need to be centralized away from the semce provision functions to ensure impartiality and avoid conflict o f interest; yet sufficiently decentralized for the adequate identification o f pnority villages for investment, and detailed observation o f system performance. The M&E o f "who has access" to water services i s best managed at one tier above that o f the semce provision unit. 0 At the local level, it is those with the most powerful "voice" and those with more "choice" who are more able to articulate their demand for water services. Thus promoting IO3Access to a m~nimumof40 lpcdper 250 persons, withina single source locatedat a distanceof 1.6 kmor 100m of elevationfrom everyhousehold. 103 Tsble v.5: Pu ifn~eflrst principlesfar respunsibi~ lacrass ~ ~ ~ ~ 1ft4 Function Activity Public Policy Prtnclpter for ::6 E o o a - 0 % g?i g 3. rc g 0 % u Policy, lcentify P~~o'it)Areas S:ate G3 Gp Planning 8, 3efne ProSraT7n7es B Budgets S:ate G3 GP Standards Standard Setting State G3 C;= Planning dentifit Pro'itj L Ilazes GP iG3 ,Asset Creation Human CaptCaaf I !I Creation of cernano S:ate GP G3 Social Capttai I Format 01of user Group GP G= i Physrcai Capfrat Owrlers9 3 of ?f.asl,-ciurc GP SPU I Technical Sa-ct OP of lnvejtireni SP G" I Des gn of Sanr'ar, Fac l i t ~ ~ s ZP SPU Location Qi'ai tt o ' C 3 - s ~ ~ c f i m 53 G3 I Financial ACID-om1 ZF G? O&M of Operatrons Sanitary S,stern operation GD SP, Faolitie s Findic al ranagerent GP SPL Staff +ring F rn(i 0' star GP S3" Sala., GP sp v M amtcnance Va,or repa 's re$aceirenr GP SP, Fiiiancing GP SP S?J Monitoring and P~r~rsrcrna Schwne Status SP PS SP Evaluation Status c'serrce de my SP "S SP A,&: of Acx,-ts SP zs SP State Table V.8: First prin d on aCCO insrrnitlttinu Function Activity Accountability Principles for Aflocati 0 - * 0 1 0 ) f cTJ & - = mTJ E. e. =c E0 0 0 & 3 Policy, Identipi Prsoritv Areas tt gh & -on L o r Technicaf & Local Planning & Define Progranimes & Gudgets H gh & -ow Low Technicai Standards Standaro Setting LO& LOW Tecltnlcaf & Local Planning Identify Prioritv V i lages H gh PA eaium Technical & total Asset Creation Human Capita/ Creation of aernand H gh High Social Capital Formation of user group H 9h High Physical Capiral Ownership of infrastiuctJre H gh LO"\ Technical Sanction of Inveslnient H gn Low Design of Sanitary Facilities i+gh F.1 ea ium Location i Quality o! Construct on high F.1 eaium Fin3ncial ApproLa high Low OBM of Operations Sa nita ry System operation high High Facilities F nancial management high Lo* _oca1 Staff Hir.ng Flr,ng of staff high High Salary t i igh hleaium Maintenance Major repairs treplacement high High r Financing high Log; Monitoring and Programme Scneme Status High LO,& rwhnrcal 8 Local Evaluation StatLs of sei'., ce oelivery High High -oca1 A..dit of Accotlttts LoN Lor\ rechnicaI LtClon of the polrcy out, It1 all thrcc, rh 107 the local government for arbitration, facilitation and oversight. But the semce provlsion function needs to be clearly delineated from the governance function. Delegating down: Enabling some separahon downwards between the local policy-maker and provider can be achieved through institutional separation (e.g. Kerala); role separation (e.g. Maharashtra); or separation by assignment,, which enables a balance between performance objectives and standards for staff assigned as local government functi~nanes.'~~ 0 Delegating uu: A degree o f financial capacity and autonomy in determining whether to purchase expertise from higher tiers o f government i s essential in enabling lower hers of government to deliver improved semces to consumers. The success o f any delegation upwards is dependent on the extent to which a local provider can hold a higher-level provider accountable for the delivery o f requisite services. The procurement o f a professional service needs to be differentiated from that o f a professional public institution in one sense: the higher tier o f government must be responsive to the client power exerted by the lower tier if the latter i s to hold the former ac~ountable.'~~ Delegating out: In delegating out, government generally performs an oversight function, enabling the appropriate management o f services by a higher tier. InWest Bengal, most o f the provision functions in sanitation are delegated out to the NGO that runs the RSM. The RSMs are hired and monitored at the distnct level on behalf o f the block level.lo6 One o f the objectives in separating the policy function from semce provision i s generating transparency on the quantum o f public funds flowing to the provlder. Inboth delegating down and delegating up, ensmng that the client (or client government) can holdprovlders accountable for high-quality services requires that the provider be placed under some hard budget constraints. Indelegating down, ahardbudget constraint on aprovider strengthens downward accountability to users. Ideally, the GP should support funding only for the public good, and the private good element should be completely funded by users. In delegating up, hard budget constraint will strengthen the higher tier's accountability to the GP This also requires some separation between the higher-level service provider and the policy maker. Any financing from the policy function in the higher tier should be transparent and predictable, and less than the revenue receipts o f the lower tier service providers. Information. Once service provlsion has been separated from policy-malung, policy-makers need information to ascertain the achievement o f the requisite outcomes. Performance. InIndia, the transfer o f the O&M responsibility has not been clearly linked to the responsibility for re-investment, and higher tiers o f government have incentives to create more assets. Thus there i s a tendency to build an asset, run it down, and then rebuild it. But the ewdence indicates that the focus should be on sustaining existing assets through improved O&M. IO4 There i s a need to separate the accountability o f local government employees for performance (to their respective tier o f local government) from their admnistrative accountability to higher tiers o f government for meeting quality standards. This will ensure appropnate incentives for performance. I O 5InMaharashtra, this isbeingattemptedthrough a process ofcorporatization ofthe state water board, thus seeking to make it dependent on the VWSC for financial survival. IO6 There i s strong downward accountability to the clients in this model. This i s achieved through an arrangement that aligns the financial viability o f the RSM with their mobilization capacity, and the quality of the service that they provide to both BPL and APL consumers. The market forces element in this model could be strengthened if local government invited competitive bids for the nght to operate a RSM. This could involve NGOsi contractors submitting tenders on the lowest price for the establishment o f an RSM, or the lowest price per latnne installed. By enabling NGOs/contractors to innovate inreducingpnces, the efficiency of service delivery may be improved significantly. 108 The success of this alternative approach'" to O&M will require movlng away from tied funds that favor the creation o f assetsrather than their operations and maintenance. Table V.9 summanzes the recommended normative assignment o f functions for RWSS on the basis o f analysis o f the sector. The relevant guidelines are as follows. Policv and design standards: The governance o f the policy and design function for RWSS should be clearly assigned to the state. The role o f the center shouldbe to support the states bycreatmgincentives that promote sustainable access to high-quality services. Planning: The responsibility for planning for water supply should be assigned to the DP This should include the provision o f a receptive envlronment for "bottom-up" planning by GPs that "self-select" for the uptake o f community-managed water supply schemes. The responsibility for the governance o fplanning processes for sanitation should reside with the GP, under a system o f incentives that pnoritize improved sanitation outcomes. The role o f mobilizing GPs to take up this responsibility should reside wth either the DP or the state. Asset creabon: The responsibility for governance inthe creation o f RWSS assets should be clearly assigned to the GP The GP should then delegate the responsibility for managing asset creation to the relevant sized semce provlsionunit.The responsibility for the governance o f HRD,accounting andtechnical design services -as well as service provlsion-shouldbe unambiguously assigned to the DP However, semce provision can be either on a supply- driven basis (funds directed to the DP level and semces provldedto GPs); or on a demand- driven basis (funds flow to the GPs who purchase services fiom the range o f HRD, accounting and design services offeredby the DP). The responsibility for the governance o f social capital (community mobilization,) should be assigned to the GP Operationand maintenance: The governance o f O&M should be clearly assigned to the GP, which inturn should give the actual responsibility to service provisionunits o f appropnate scale. 108Irrespective o f the location o f this service unit, the semce i s provided to the GP on a "fee for service basis." The responsibility for fundingmaintenance lies with the User Groups, though the GP may offer loans or grants inthe initial stages to help finance major maintenance costs. For sanitation, the GP i s appropnately positionedto provide pure public services (street sweeping), while the habitation-level users o f open defecation sites are best positionedto manage the public good element o f the defecation practices and solid waste management services o f individual households. Monitonng and evaluation: M&E shouldbe assigned to the BP This may require a small cross-sectoral team that reports to the BP on the performance of a wide range o f schemes. The BP then pushes this information, either down to the GP to improve the compliance o f User Groups, or up to the DP to bringincentives/ disincentives to bear on GPs for compliance. The role of the state inM&E i s predominantly one o f impact evaluation, providing support to the BP on the form o f M&E, and bnngingstandards to bear across the state. The governance role o f the BP inmonitonng does not preclude functionanes at the block level from acting as the de concentrated arm o f the DP or the state. lo'In Karnataka, the transfer o f the O&M o f existing assets to GPs has been funded through an O&M window. Since the state government has begun strengthening the upward accountability o f GPs for the use o f these funds, GPs find they need to strengthen their downward accountability to consumers. Some GPs are now establishing User Groups at the habitation level to operate schemes and raise revenues from users. At present, these GPs are fundingthe deficit in the O&M account o f the User Groups, but intime the GPs should be able to introduce a hardbudget constraint, provide incentives for performance, and re-invest inthe upgradation o f assets. logA habitation-level user group, for instance, seems appropriate for most single village network schemes. However, private providers (such as farmers for pumped schemes) are also appropnate scale providers o f O&M services. For multi-village schemes, complex water testing services and complex maintenance undertakings, the unit o f service provision may be appropnately located at the distnct or state level. 109 th`bt 1 IIff Under the existing a ~ ~ o c a ~ ~ ~ 3 nto states, the greatest ~ r ~ ~ g h ~ agiven to states with the formula IS g e largest rural an& or tribal popu~ations,and the highest nu~iibcrof Ne/PC1Iiand q u a ~ i ~ y - a ~ f ~ c ~ e ~ ~ ~ ~ a ~ ~ t a t iKowvcr, with India having achieved BS percent coverage in ~ n ~ a s ~ ~access, o n s . c ~ ~ar a 1 higher NCI PO percentage might be ~ ~ d ~ c a toi fvthe m ~ ~ ~ a n a g eof ienn~tr a s ~ i c t urather than e ~ r e an absence of ~ n v e s ~ ~ eThis. suggests r ~ ~ ~ athci NGI PV indicator. What ~ n t ~ n g ~ ~ e ~ a ~ ~ ind~catorssould best indicate a state's c o ~ n ~ ~ t to ~ ~RWSSt reforms? Options include local e n g o v e ~ revenue receipts as a per~en~age state government e~pendi~ure RWSS; GP ~ ~ n ~ of an (tfWK) e ~ p e n d ~on~RWSS ~ ~ e v ~funds~plusdown ~ ~ r e 3 l ~ ~ as a p~e r c ~ n et a ~oef the state ~ ~ e ~ ~ ~ s ' ~ ~en ~~ e~ net ~on~KWSS; and GPe~ V ~ re~renue) ~ ~ t ~ r S ~ for RWSS as a percentage of total revcnuc for RWSS. The first of thcse options s e e m to offer the s I ~ p l e ssystem o ~ d ~ ~ ea ~ i ~ ~ ~ ~ i g ~ state's c ~ ~ i ~to RWSS t ~ ~rcfi3mis.~States~that have de~eIopeda ~ r e d ~reform p r o ~ r and i c ~ ~ l e a ~ signed the.Motl will not be bound by the dictates of central schemes. But for states that have not devclaped credrbic reform prc>gams, i ~ c r ~ the~S~~ ~nr a~j ~ l ~ haa~r~ao c a t ~and redwing the a * n ARWSP ~~Iacation help shift them closer to the refom agenda. will The goal o f the SGRY program IS to p r o ~ i add~d i t ~ o ~wage as asset ~ sclcction, asset creation and hf&E _-arc analyzed by applying the first p ~ n c i p ~ eofs publi~ r The four unb~ndleda ~ ~ ~of~an~e t~ p~~l oe~ sm e~program --~t a ~ ~ e ~activity i ~ l ~ m ~ l o ~food~security~ ast wcll ~ c , ~ a ~ ~ n t ~ i n g ~ finance and a ~ ~ * t ~ n t a btol ~allocate r g s p o n s i ~acrossi ~~~ ~ ~a tiers~o fog ou ~ ~e ~ m e (see ~ t y ~ ~ ~ n t Table V.19 and Table V.11). Economies o f scale Inthe coordinationofany CSS, there are significanteconomies ofscale inthe setting o f standards, and intargeting and budgeting, suggestingthat these activities should be undertaken by the center and the state. Since SGRY i s implementedat the DP, BP and GP levels, cost efficiency suggests that both activity selection and asset creation should be localizedat the level of implementation. Butthe economies of scale inM&Elie with higher levels of government for activities such as records of assets as well as financial audits; and with the lowest levels o f government for social audits and the physical verification of assets. Externalities Infood- andwage-for-work programs, externalities aremostlikelyinthe actual creation of assets. Ifa GP decides to builda road connecting a few villages, the social networks o f the laborers from these villages are enhanced. And once the road is built, employment opportunities and trade markets inthese villages could improve. Equitv Central schemes bear testimony to the government's commitment to equity through employment programs that reduce poverty. Ina caste-dominated country like India, equity issues, including the imbalance o fpoverty indicators across the states, shouldbe dealt with at the highest level o f government. The supervision and the hiring/firing ofpersonnel shouldalso be dealt with at the highest level to ensure the inclusion of ST/ SC inthese programs. 112 le\ els in employment prograr 4Cfivity "rsset scacion c 113 Table V.11:Firstprinciplesfor allocationof functional responsibilitybasedon ?olicy Targetmg IesignI Standards Budgetmg Standards Activity pnontzatiod Actionplan ? l a m g Activity Selection Techcal k Local Human Capital 4sset SkillDevelopment Medium High rechnical Clreation Social Capital InformationDissermnation High High ,oca1 Physical Capital Public Works High High rechnical Beneficiary Selection Iperafion Identificabon ofbeneficianes High High ,oca1 >oca1 edium ,oca1 edium edium r e c h c a l edium ,oca1 edium r e c h c a l Repairs High High -oca1 VIomtormg Assets md Record o f Assets Medium Low r e c h c a l ?valuation Physical Verification o f Assets Created Medium High ,oca1 Audits Financial Audit Medium Low r e c h c a l Social Audit High High ,oca1 Applying accountability criteria to employmentprograms Ideally, the relationship between citizens and policy-makers in employment programs should be such that the citizens delegate by choosing the executive in government, participate in planning processes to have a say inpolicy design, and pay taxes to finance the program. The policy-makers and service provlders perform by implementing schemes. Based on information from BGs on the 114 impact o f semces, the state/ central and local governments create enforceability by rewarding penalizing the providers. Infact, international expenence intargeting beneficianes highlightsthe possibility o f using communibes to target benefits to households. Communities have information that can be used by local decision-malung structures for better targeting without the high cost o f either improvlng administrative targeting or self-selection targeting. The second relationship i s the compact between policy makers and the organizational provider. Inthe decentralized system o f governance, the accountability relationshiphas to apply to each level. But these ideal relationships o f accountability remain far from ground realities (see Chapter 4). Arguably, SGRY policy design does reflect an attempt to address the issue of accountability Unlike most central schemes, SGRY emphasizes the role o f PRIs in planning and implementation. But the failure to address key issues -- client power and the compact between policy-makers and providers -has led to concurrency, weak delegation and poor performance. Analyzing SGRY with the features o f accountabilitv Delegating: Client power i s weak, particularly over the DP and BP, which prepare the action plans to commission public work in their respective areas o f junsdiction. The clients (indivldual beneficianes in this case) play no role in preparing these plans presented at the gram sabha. Participationinthe program is, o f course, based on need and availability Hence there is no delegation, and implementing agencies are neither directly aware o f client needs, nor responsive to them. Financing: As a central scheme, SGRY finances are provided by the center and the state. To this extent, they are unrelated to any local-level taxes paid by the client. The provlders themselves have no control over the finances, and this takes the client one step further from the financing process. Performing: The local governments, line departments and, occasionally, the semce providers enter into a relationship o f accountability with the client by performing their function of implementing the scheme. But the specific role o f the PRV line department and semce provlder remains unclear. Informing: Relatively weak information dissemination means that beneficianes h o w little about the scheme guidelines or its benefits. Enforcing: Given the weakness in the relationship between clients and providers, electoral choices are not necessarily connected to the impact o f the scheme. And while there are some provlsions for enforceability through social audits, these remain localized at the village level, and citizens have little or no ability to enforce accountability at either distnct or block level. The relationship between clients and the GP i s somewhat stronger. The guidelines require that GPs regularly consult beneficiaries and share information with them through the gram sabha. The gram sabhas conduct yearly social audits and can lodge complaints at the DP level. The question i s to what extent these provlsions translate into accountability inpractice. In sum, the center and the state enter into a "contract" with PRIs by devolving functions and resources as mandated by the guidelines. The center and state hold the PRIs accountable - financial flows are contingent on the presentation o f audited accounts. But several problems are evident: The involvement o f multiple ministnes undermines the clarity o f roles and encourages functional overlap or concurrency - hence increasing the possibility o f misappropriation (see Chapter 4). 115 Though the guidelines attempt strong upward accountability relationships,' ''downward and external accountability are missing at the DP and BP levels. Clients are already one step removed from policy makers at the BP and DP levels, and client voice i s hrther weakened by patronage politics. The potential for collusion between PRIs and organizational providers i s somewhat checked by the weak accountability relationshipbetweenPRIs and providers. Butthis also means that despite the guidelines, the state government controls SGRY implementation inmost cases. 0 The PRIs have no control over day-to-day activities since they have no administrative power over frontline provlders and cannot holdthem accountable. Assigning activities in employmentprograms: guidelines We have seen that ideally, an employment program has to ensure effective targetmg; efficient and labor-intensive activity selechon; provide for widespread information dissemination and local capacity building; and ensure functioning monitonng systems. What then should be the ideal institutional arrangements for the effective implementation o f SGRY? We propose a system where the GP, in consultation with the gram sabha, has complete authority at every major step - beginning with activity selection and prioritization, asset creation and operationalization, to M&E (see Table V.12). This proposal holds even when technical skills are required. In such cases, the GP shouldbe able to outsource these skills. The GP's role: The bulk o f resources should be transferred directly to the GP,li3 which should makes all the major planning and operational decisions in consultation with the gram sabha. This implies that the GP will be responsible for the creation o f all public works, the planning aspect in particular. All plans should be prepared by the gram sabha on the basis o f the estimated budgetary allocations before the funds are sanctioned. This will mean a reduction o f implementation delays, as only financial sanctions will be required once the funds have been devolved to PRIs. Moreover, the active role o f the gram sabha in decision-malung will localize asset creation. This will promote beneficiary awareness o f plans prior to their implementation, leading to effective targeting. The crucial factor of an effective gram sabha: The efficiency o f this system i s dependent on the effective functioning o f the gram sabha. There i s the possibility -- and the reality -- o f the gram sabha either not being held; or when held, being ineffective because o f poor parti~ipation."~But giving the GP greater financial powers will help enhance participation, since clients will have greater stakes in the proceedings o f the gram sabha. There i s also the question o f the quality o f participation. Often, power relations within the gram sabha keep the poorest -- the key beneficiary group -- outside the system. This possibility needs to be met with building the capacity o fthe GP and the gram sabha, as well as monitonngcompliance. Capacity buildingo f GPs and gram sabhas: The process o f strengthening the capacity o f GPs and gram sabhas involves a phased approach o f scaling up the role o f state-level training institutions such as SIPRDs by. 0 Conducting training inall GPs and ensunng regular follow-up; 'I2This is cntical for addressing the problems associatedwith patronage politics, particularly because infrastructure relatedprogramslike SGRY canbe susceptibleto patronagepolitics. See WDR, 2004, pp. 159-163. 'I3 is This as opposedto the current distributionoffinances, which is 20 percent to DP, 30 percentto BP and 50 percent to GP *I4Pandeet al, 2005. 116 0 Improvlng current training modules to focus on the issue o f gram sabha participation; and Creating the skill set gram sabhas require to conduct social audits through direct interactions with NGOs, and raising local awareness o fthe gram sabha. Monitonng: compliance: Giving the GP greater financial responsibilities makes monitonng and supervision cntical. The DP, and to some extent the BP, have to play an overall supemsory role and ensure that regular and effective gram sabhas, and social audits, take place. Creating the nght incentives for each o fthe three tiers to function effecbvely involves following some guidelines: 0 The DP has to undertake regular random visits to ensure that GPs hold gram sabhas on the dates agreed upon. 0 Based on benchmark performance indicators decided by the state government, GPs holding gram sabhas are rewarded with a combination of monetary and non-monetary incentives such as grants and awards. 0 One option could be a financial penalty as an incentive for the GP to hold the gram sabha. The role o f the BP and DP. Often there are circumstances that call for intervention beyond the junsdiction o f a single GP (or BP). Inpublic finance terms, this i s an argument o f economies o f scale. The roles o f BPs and DPs are similar to their current roles, but the voice element between client groups and the BP/ DP must be strengthened. To this end, we propose a system o f contracting upward - in which the GP contracts the BP and DP to implement programs in their respective areas of junsdiction. This would ensure that the BP and DP have proximity to the client through the GP and the gram sabha, thus creating conditions for strong accountability The role o f functionanes: One o f the cntical gaps in the current system i s the weak relationship betweenproviders and PRIs. This leads to the broader issue o f the relationshipbetween PRIs and their functionanes -- who are state government employees and thus not accountable to the PRIs. The solution to this problem lies in a broader structural change: functionanes should be transferred to PRIs, and PRIs given complete authonty over these functionanes. In the intenm, and specifically for SGRY, transferring resources directly to the GP could create a system in which the GP (along with the BP and DP when required) could outsource the necessary technical expertise to a distnct pool o f experts. The role o f the center vis-a-vis the state: Ideally, the state should be responsible for setting standards, and for the design elements o f the program (particularly the minimum wage). However, the inherent nature o f CSS has meant that the center too plays a role in this process. There are, then, three possible options: 0 Offer states a menu o f programs: This radical option would allow states to choose programs according to their needs, creating a more transparent system. This is also consistent with givlng block grantsii5 to states, and moving from a schematic to a thematic approach. 0 Identify broad principles: The central government must play a role in addressing equity concerns. But in view o f vaned conditions in the states, state governments should design their own programs. In addition to reflecting local conditions, these designs must be based on key pnnciples such as targeting SC/ ST populations and women. This recommendationhas emergedfrom the GOI. See GOI, 2004c and GOI, 2001b. 117 0 Set flexible guidelines: This option, probably the most feasible one, would have the center set guidelines in consultation with states as it does now. But the state would have flexibility on cntical aspects such as wage rates and the form o f compensation. In sum, the most feasible option has the central and state governments set standards, articulate broad policy guidelines, develop skills, and monitor the program financially All other operational decisions and processes are with the GP, with decisions being made inconsultation with the gram sabha. The BP and DP play a supervisory role and are responsible for holding the GP accountable. Such a setup would enhance voice, compact, client power, and inter-governmental relahonships among policy-makers. 118 Table \'.I Kcconimcntletl institiitional arrangt 2: ent program! Function kctit7ir)t The assignmentof financing116 This study has defined assignment o f roles in terms o f who is in a better position to deliver quality semces in a perfect world, that is, assuming that many other things come into place at the same time. Once expenditure powers are assigned to the level that i s best suited to deliver those functions or activities, Government will needto address the financing issue. In India, like most developing countnes, there is a large imbalance between revenues and expenditures. The Center collects most revenues, but if the recommendations of this study are implemented, the states, the distncts, the blocks and above all the Gram Panchayats, should be responsible for most expenditures. Hence a large proportion o f central revenues will need to be passed down to states and lower levels to enable them to meet their responsibilities. The way these funds are devolved d o h will have substantial implications for how services are delivered. As discussed above, central funds are currently transferred down through two mechanisms: the Consolidated Fund o f the States (CFS) and Centrally Sponsored Schemes. The CFS distributes funds among states as block, untied grants, according to a formula that reflects population, poverty, backwardness and other parameters. Centrally Sponsored Schemes are matching grants to the states, tied to a certain type o f expenditure (e.g. drinhng water supply) that can be implementedaccording to a variety o frules that are scheme specific. Sometimes the Center also makes small allocations directly to lower levels o f government (e.g. 1I*Finance Commission Grants). However most revenues o f local governments are tied grants from the state. Tied means it i s assigned to a quite precisely defined type o f expenditure leaving little discretion for local governments to decide on how to spend the funds. Revenues raised by Panchayats range from small to negligible and we do not consider them inthis discussion. One way o f categorizing the different ways o f transfemng central or state revenues down i s the following: 0 Untied =ants--where we leave it up to local governments to decide where to and how to spend the money For example, given a hard budget constraint, a local government may decide to put up a dnnlungwater system and not another class room for the school. 0 Matchinn m-ants--where higher level government will share local government expenditures hence creating an incentive for local governments to divert funds into a desired activity 0 Conditional =ants--where funds are tied to specific expenditures. For example, teachers salanes or ina more modem set-up tied to outcomes such as attendance inschools. 0 Own source revenues.--GPs can levy some taxes, notably property taxes, and feedcharges This is an underdeveloped set o f financing sources and more revenue 116 This study dealt exclusively with service delivery and expenditure assignment, that is, who should be responsible for which services, which is where the reform of the intergovernmentalsystem should begin. Once assignment is decided, the next step is to address financing issues ("finance follows function"), which will be the subject of a forthcomng reporting using West Bengal as a case study. In this section we offer only a general description of the principlesguidingassignmentof financingresponsibilities. 120 powers need to be transferred down. At present, it accounts for very little o f the financing. The central question for the vanous services discussedthrough out this report i s whether transfers should be made conditional on the provlsion o f those services at certain levels. When local governments serve as agents o f the center or state conditionality is obviously needed. This would apply to services that have very large externalities relative to the value o f the benefits accruing locally Vaccination would be an example. On the other hand, where the value o f externalities i s low compared with the value accruing locally discretion should be left to the local government o f whether or not to finance the semce, and an untied grant would be appropnate. This might be the case with a local health clinic or a dnnkmg water pump. These would be a high pnonty for the local population, therefore an accountable local government would have strong incentives to deliver them from its pool o f untied resources; and at the margin it might need to pnontize one versus the other, again in response to local preferences. Where the value o f externalities or positive spillover effects i s reasonable, matching grants would be the appropnate form o f financing, and inthis case the match from the state or center shouldbe commensurate with the value o f the externalities associated with the service. For many o f the semces discussedinthis report externalities are relevant and some element o f matching would be necessary, for example for school buildings or instructional services. The above discussion assumes that all local governments have been equalized in terms o f the amount o f funds they have at their disposal. This means that they have been brought to a minimumlevel o f revenues (combining own revenues and equalizing transfers) that enable them to meet a basic level o f services. For the 5 tier Indian system, the vast majonty o f services discussed should fall at the state level or below This implies that the Consolidated Fund o f the States, rather than Centrally Sponsored Schemes, would be the ideal vehicle for the Center to channel funds for those semce~"~.. However, the intergovernmental system i s far from perfect and i s in transition. There are many information and governance failures. There are insufficient mechanisms o f checks and balances and lines o f accountability over decisions. This makes the case for maintaining key CSS to ensure semces can happen where they are not being delivered and tying some o f the grants going downwards to specific activities and outcomes. It i s important that the system move in the nght direction, albeit gradually As an intermediary step the government may consider: 0 Substantially reducing the number o f CSS at the central level and channel released resources through the CFS. 0 Make remaining CSSs thematic rather than tied to specific expenditures. General themes would be pnmary education, pnmary clinics, etc. Discretion should be left to lower levels how to manage and operate these. 0 Consider ear-marhng certain percentage o f the spending on capital and recurrent within this thematic blockgrant. '" Some of the work on financing is being done m the context of an ongoing study-Fiscal Decentralization to Rural Governments in West Bengal-and will present clear guidelines on how to designthe transfer systemdunngtransition. 121 0 For distncts and blocks, which are not real governments since they lack both discretion and autonomy all transfers shouldbe tied (For example for the training o f teachers. Some o f the state funds now passing to distncts could be devolved to GPs, therefore increasing the revenue budget o fthe lowest level unit 122 VI. TowardsBetter RuralServices: Recommendations Taking decentralization as a given, the study conclusions make a strong case for a general shift to 0 Giving the lowest level of government -- the GP -- the responsibilities of asset creation and O&M, while involving it in the planning process through the gram sabha; Giving the middle tiers, such as the DP, responsibility for human capital developmentandM&Eof outcomes; 0 Giving higher levels of governmentthe responsibilityof policy and standards; and 0 Gradually moving away from rule-based CSS towards fiscal transfers through the state's consolidatedfund, conditionalon reforms. I n conclusion: the context andfindings of the study The context This study began wlth the question o f who should do what inthe five-tier government system in India. The intention was to feed the current debate on the allocation o f funckons, funds and functionanes among these tiers of government. Several states have, or are, formulating "activity maps" to assign spending powers to local bodies, and the GO1itself has provided a bluepnnt to states for this exercise. The failure o f services That key services are failing rural India i s a foregone conclusion. The general expectation i s that economic growth and urbanization will help deal with the service delivery cnsis; that higher incomes will allow better access to pnvate prowders; and that rural-urban migration will bnng people closer to points o f service delivery But India has 700 million people living invillages in rural areas; 400 million o f them are poor. Such poverty means that a complete reliance on private providers i s fraught with uncertainty; indeed, there are many public services that the pnvate sector should not be involved in.As for migrationbnngingthe rural poor closer to urban services, even if most o f this population migrated to urban areas, providing them with access to urban semces would take a long time. Moreover, this would impose substantial external costs on the existing urban population. Clearly, the task at hand i s to plan and implement better rural services. Around 70 percent o f the world's poor live inrural areas, and around 60 percent o f the poor will still live in rural areas by the time half the population o f the developing world lives in urban areas.Il8 This statement applies to India more than any other country The failure ofthe top-down approach It is also well known that large bureaucracies, top-down, supply-driven, and based on a civil service, have failed in many countnes. However, it is the top-down bureaucratic approach to service delivery that has dominated in India - a result o f histoncal, social and political processes inwhich the interactions among citizens, the state and providers have tended to be overlooked. Many features of Indian public decision-malung were "transplanted" dunng colonial rule, or adopted from the central planning approach o f socialist countnes. This approach has failed inthe `I8Ravallion,2004 123 delivery o f basic education, health, drinlung water, sanitation, employment programs and many other public semces. While the civil service based approach has worked in the developed countnes, inmost cases this built on a decentralized structure o f service delivery. The Indian alternative India's alternative to the civil semce approach i s the Panchayah Raj system. This is a homegrown system that grew out o f many public and political debates to combine traditional and modern features. Having learned from the successive failures o f the centralized civil service approach, most developing countnes have embarked on a number o f expenments with alternative models o f semce delivery Participation, community-driven development, empowerment, and decentralization are among these new approaches. They are not mutually exclusive; in fact, India has advanced along all these lines, while opting for decentralization as the way to reconfigure the state. The panchayat system builds on the central pnnciple o f democratic representation at all levels. The decentralization model that emerged out o fthis system was enshrined inthe Constitution and inthe Panchayat Acts ofthe vanous states. Itwas also recognizedthat for thepanchayat system to get o f f the ground, all states and parties would need to move together. If a state or party pursued decentralization alone, it would be less easy for reformers to claim credibility for any gains in semce delivery and development since people would attribute those to local politicians. Thus emerged a consensus on getting Parliament to impose order on the system through the ConstitutionalAmendments. The next question confronting national legislators was whether the devolution o f powers and resources to the local level would undermine the states, violating the underlying pnnciple o f decentralization. The response gave states the powers to decentralize further, while stipulating some minimummandatory conditions in the Constitution. These conditions include democratic elections, gram sabhas, and reserved seats for disadvantaged groups. The states were given the freedom to choose the powers and resources to be actually devolved. The dilemma o ftwo models Thus India firmly established local democratic governments. Today there are about three million locally elected politicians, many o f them women or SC/ ST -- a political force to reckon with. But meanwhile, the civil semce bureaucracy was not dismantled or restructured. The result was the precanous coexistence o f two different - almost incompatible - systems to deliver services and development. This dilemma i s linked with several problems in the implementation o f the decentralized approach: Most reforms were hesitant and incomplete; they could easily be undermined or reversed. Subjects were devolved wholesale. For example, some Acts assign responsibility for pnmary education to panchayats. But pnmary education includes many different activities. Should all o f them be devolvedto the same level o f government? Most o f the Acts devolved subjects wholesale, but left specific activities as the responsibility o f line agencies -- rendenng the devolution o fresponsibilities meaningless. Expenditureassignments were made wholesale. 124 Resvondingto the context Inresponse to this context, the study first unbundledthe four sectors into services and activities. Most activities can be aggregated into five main activity groups: policy and standards; planning; asset creation; operation; and M&E. Followng this, the study examined the dejure situation - what the law says. Next, the study investigated the reality on the ground. Two approaches were used to study the defacto situation: one was based on field surveys o f politicians, local civil servants and citizens on how services were being delivered in their areas of junsdiction. The second approach relied on results from other studies and reports on the status o f service delivery in India. The conceptual framework - consisting of public finance cntena as well as accountability cntena - was then applied to the findings to reach conclusions about both ideal and vlable allocahon o fresponsibility to the different government levels. Studyfindings The de lure situation State variations: By examining the prescnptions implicit inthe Constitutional Amendments, State Acts and other relevant legislation, the study confirmed that the states have moved at varying speeds. Kerala has devolved substantial powers and resources to the local level, and attempted restructuring line agencies so that they can play a new role. West Bengal and Karnataka have made less dramatic progress - though they have made some noticeable moves toward decentralization despite most o f the civil semce apparatus continuing to do "business as usual." Rajasthan i s an example o f those states in which local governments appear to be accessories without real powers and financial resources. Concurrencv and overlaps: The intergovernmental system is charactenzed by concurrency and overlaps to a significant extent. Concurrency stems from the concurrent list --the list of subjects under the preview o f both central and state governments. Concurrency i s not necessarily a problem, since all levels o f government may have a role to play inmany key semces. But these roles do need to be clearly delineated - and this i s what does not happen. At times all five levels o f government are responsible for the same, or very similar activlties (such as buildingschools or roads). This has created confusion among citizens, undermined accountability, and led to the duplication o f efforts and the wastage o f resources. Overlapping i s acceptable in Indian jmsprudence. The Supreme Court has ruled that two laws empowenng parallel institutions with the same functions do not render the two Acts inconflict. But any state trying to improve service delivery through decentralization will need to address inconsistencies the way Kerala has done. After approving its Panchayat Act, Kerala revised 44 Acts to remove inconsistencies, and ensure a clear legal framework to drive the institutional setup after decentralization. Domination o f CSS: While many states have attempted to assign roles through the panchayat or semce Acts, the emerging institutional picture i s very much driven by CSS - grants-in-aid from the center to the states onginating in the 1970s. These schemes were conceived as a means o f complementing the fiscal resources o f the states; but they have evolved into the channel through which most public resources for the delivery o f key services flow The bulk o f CSS are concerned with items in the state lists and Schedule XI (items under state responsibility or the concurrent list). More important, most CSS flow through the budgets o f line agencies and carry their own guidelines, which often bypass the state panchayat or service sector laws. Some CSS have, in fact, givenpanchayats a more predominant role than i s accommodated in some state laws. More importantly, the CSS have undermined the responsibility of the states to design their own 125 decentralization: they shape institutional responsibilities, for example, among the three tiers o f panchayats inways that may not necessarily be inconsonance with state wishes and laws. Poor fund transfers: States are not required to transfer funds to the panchayat system. All the states have set up the constitutionally mandated SFCs to recommend revenue powers and grants- in-aid for panchayats. But the SFCs only make recommendations. Since the states are not obliged to follow these recommendations, it i s rare for them to do so. Hencepanchayats have few resourcestransferred to them, and they have weak revenue powers. Administrative control and accountability. The Constitutional Amendment and most State Acts do not address accountability in the relationship between panchayats and functionanes. Local governments have no authonty over functionanes and service providers (teachers, anganwadis) even where these havebeen deputedto them. Implications for the key sectors: In pnmary education, wth the exception o f Kerala, the assignment o f responsibilities in the law i s onented towards the state. Inhealth, the assignments in all states are heavily oriented towards the state. Indrinkmgwater and sanitation, Kerala and Karnataka have devolved responsibilities to panchayats. But this sector is dominated by the guidelines o f CSS that do not necessarily approach the role o f panchayats in consonance with state laws. Employment programs are completely dominated by central schemes that channel resources through parallel instruments that fall outside the panchayat system, and to some extent, the state bureaucracy Finally, only Kerala and Karnataka have attempted to unbundled sectors and devolveresponsibilities at the level o f activity, rather than at the level o f the semce or sector. The de facto situation Education: In all the states, the central scheme SSA i s the main vehicle for delivering pnmary education, and the state bureaucracy dominates the implementation o f this program. In terms o f implementation, the key findings in Kerala are the variation between the provision o f teaching services and infrastructure: the respondents recognizedpanchayats as playing a role inproviding infrastructure activlties such as building, repainng and maintaining schools, but teaching activities were perceived as being a responsibility o f the state. InWest Bengal, where education i s delivered through both the mainstream system as well as the para-teacher scheme SSK, a stronger role for PRIs was identified in implementing the SSK than inmainstream policies and programs. InKarnataka, the state bureaucratic machinery is responsible for implementing most of the functions for education delivery What has been the impact o f the top-down civil service approach in the delivery o f basic education? Interns o f quality, an accountability-based diagnostic indicates that the current state o f education delivery i s far from effective - in the provision o f basic semces, in attainmenu completion, and in learning achievement. Teacher absenteeism i s rampant despite Indian government schoolteachers being among the best paid in the developing world. With the exception o f Kerala, educational attainment i s l o w -- with Rajasthan showing a greater gender gap, and Karnataka and West Bengal a wider rural-urban divide. The quality o f learning achievement i s low and highly variable. InKarnataka and Rajasthan, for instance, more than 90 percent o f children have inadequate learning achievement. Health: The key findings include the following: Interms ofadministrative andfinancialcontrol, statecivilservantsplay the dominantrole in the public sector inall the study states. 126 Despite the existence o f a large public sector delivery system, the pnvate sector has a significant presence inhealthprovision. The figures for the incidence o f public subsidy inhealth in each o f the states shows that more public money goes to the relatively well o f fthan the relatively poor. On the one hand, most spending inthe Indian health system goes to curative health -- pnmary and hospital care. But the amount o f public spending i s dwarfed by the amount o f pnvate spending, which i s about eight times larger. On the other hand, very little is spent on health public goods -- populabon services, disease control, healthpromotion and surveillance. There were many compIaints from GP members and citizens that health services are not worlung; but the perceptionwas also that no one acts on these complaints. Local officials and politicians revealed little knowledge and awareness o f the importance o f disease control and health promotion, though these are semces are clearly inthe public domain. The evidence indicates the varying extent and type o f involvement o f PRIs in health in the different states. For example, InWest Bengal, there are examples o f GPs coming together to hire the semces o f a pnvate doctor. InKerala, GPs have a role to play in the running o f pnmary health clinics, buying drugs, maintaining facilities and filling temporary vacancies. However, they cannot hire functionaries (other than on a temporary basis) or dismiss them, and this seems to have ledto some fnction between personnel and local governments. Rural water and sanitation: Some major moves away from the top-down civil semce approach have taken place, but problems persist. Service prowsion still takes place at the state level, despite state laws, government orders and CSS guidelines specifying the contrary Also, rural water i s regulated by state laws, CSS guidelines as well as legal agreements in donor-funded projects. Hence there are many different models coexisting at the same time and inthe same place. Kerala, for example, has four insbtutional models for the delivery o frural water. Though the allocation o f responsibilities to panchayats i s clear, the Kerala Water Authonty has transferred only a small number o f village schemes to them. This is in spite o f the fact that the Water Authonty i s losing money supplying rural water. Moreover, the effect o f decentralization on the state line agency seems negligible since it has retained control over technical staff. The situation i s even more senous in sanitation -- despite several schemes in place, solid and liquid waste disposal i s the exception rather than the rule. The control over sanitation services remains at the state level though both the center and some states have tned to pass them down to the panchayat system. Ironically, panchayats appear to contribute substantial financial resources to sanitation, but have limited say on expenditure. But the fact that they are willing to spend reveals the importance they attachto sanitation semces. E m d o w e n t Dropams: The institutional arrangements for employment programs are defined by the guidelines o f CSS. The study's examination o f the SGRY program revealed several major problems in implementation. The findings indicate that implementation i s largely controlled by line department officials. The DRDA controls most o f the process. The choice o f beneficianes is, for instance, supposed to be a responsibility o f the gram sabha. Butthis responsibility i s reduced to mere ratification in cases where the DRDA staff prepares the lists, sometimes excluding the most needy Wages are often modified at the local level. And because sanctions and approvals from the center are often required, there are substantial delays in reaching the beneficianes. Contractors are frequently hired, despite this being against the guidelines. The grain component o f the program i s particularly attractive to contractors, who often sell at a profit what i s meant to be distributed. The possibility o f irregulanties i s also exacerbated by the top-down M&E in the program. 127 I28 cntena support this result. Aspects such as teacher absenteeism and students ulth special needs are more observable at the local level. Higher levels o f government: Cumculum design, textbooks, the development o f learning matenal, and M&E should be the responsibility o f higher levels o f government. There are economies o f scale in cumculum design, evaluation, and textbook and learning matenal development. Again, there are externalities in cumculum design, the setting o f standards, and M&E. The question is how equity considerations are to be addressed. The answer seems to be that the center or the state should make sure o f allocating budgets in consistency with the inequalibes they want to address. Health The role o f public funds: Inhealth, the first pnonty i s for most public money to go to preventive and promotive health semces ingeneral. This i s particularly true for the control o f communicable diseases, where there are large externalities. Many o f these semces (e.g. sanitation) do not fall under the authonty o f the health department, and reforming towards decentralization would create opportunities to ensure that these semces are also provlded. The shift of public money to preventive and promotive health services involves a two-part process: large-scale pest control at the appropnate level; and small-scale, transaction-intensive preventive and promotive activlties, such as immunization and health educabon at the GP level. Technical support for local health educators, such as teaching materials, could come from higher tiers. But the workers themselves could be paid by, and be accountable to, local governments. In the case o f testing and similar support services, one option is to let local governments contract up for the services o f technical specialists. In general, when semces are subject to substantial economies o f scale and highly technical personnel are required, the hgher levels o f government need to be involved. For example, the higher levels should set standards for health education activlties. Curative care: As far as government role in curative care i s concerned, the long-term option i s to pay its providers for services delivered. But in the short term, two models are possible -- the Kerala model o f givlng GPs substantial responsibilities for pnmary care operations, but with the modification o f administrative control to GPs; and the emerging West Bengal model o f GPs joining together to hire qualified doctors. If either o f these approaches i s adopted, funding needs to be shifted from states to GPs through block grants. The monitonng o f outcomes and standard auditing practices will have to performed by the higher tiers o f government. Water supplv and sanitation Shift o f emphasis: Inwater supply, the pnonty has to move from access to the O&M o f facilities, and the quality o f water. The focus in sanitation has to shift from latnne construction, which has not solved the sanitation problem, to creating a healthy environment. The top-down civil service approach to sanitation has to change if this outcome o f "total sanitation" i s to be achieved. Local-level involvement: Public finance cntena suggest that the operation o f rural water facilities should be at the local level. There are no economies o f scale in operating rural water facilities, and larger and more complex systems do not carry a financial advantage inrural India. To some extent, externalities in dnnlung water are captured at the local level. There is considerable heterogeneity o f demand indnnlungwater -- reinforcing the need to decentralize decisions on the type o f design, location and operation to the lowest level. Most activities in sanitation should be carned out at the lowest level - the village or the household. The construction and operation o f sewage systems is, for example, clearly a matter for the village. There are no economies o f scale. 129 There are externalities associated with the spread o f disease but these are also sufficiently internalized at the local level. There i s considerable heterogeneity since sanitation i s a pnvate activity charactenzed by many elements o f local customs. Regarding equity, the same considerabons as those for water apply here. Distributional issues can be addressed through the fiscal transfer system. Accountability critena yield similar recommendations to those suggested by public finance cntena. The GP and/ or the users should be responsible for asset creation and O&M; andthe DPfor human capital development. Higher levels: As in other sectors, economies o f scale occur at the policy and standards level, M&E and human resource development. There are considerable externalities associated with broader water resource management and this should be a responsibility o f higher levels o f government. The same applies to M&E o f hnlung water outcomes. Access to drinlungwater i s a universal nght in India, and it can be accomplished through appropnate fiscal incentives from higher levels o f government. In the case o f sanitation, the state should be responsible for policy and planning. B:AccountabilitycntenasuggestmovingawayfromCSStowardstheuseoffiscaltransfers through the state's consolidated fund. Transfers from the center should be untied, and conditional on a set o f reforms that have been agreed upon and that can be monitored. This includes the assigning o f responsibilities to the appropnate level. At the state and lower levels, GPs should be responsible for asset creation and O&M, the state for policy and standards, and the DP for planning and human resource development. M&E shouldbe placed at the block or distnct levels. Employment programs There are economies o f scale in policy setting, planning and M&E, and externalities in asset creation. There i s a breakdown of accountability relahonships in current institutional arrangements - particularly in the cntical aspects o f downward and external accountability. On this basis, this study recommends a drastic revamping o f employment programs away from rule- based CSS. All the major decisions shouldbe delegated to the GP and gram sabha levels, and the DRDAs dismantled. Funds should be transferred to the states through the Consolidated Fund. What i s needed, above all, i s an independent and legitimate system to monitor access to public jobs at the local level. Equally important i s monitonng the quality o f gram sabhas in terms o f participation and transparency 130 'Table 1'1.1 :Rccommrncied niain functions of diffi ent goternmcnt tiers in the ke! wcturs I;uncrion L elo/&vcpmmcJt~ I f 131 References Ackerman, S.R. 1999 Corruption and Government: Cause, Consequences, and Reform. New York: Cambridge Umversity Press. Aiyar, M.S. 2002. `%nchayah Raj: The Way Forward." Economic and Political Weekly. Banegee, A., S. Cole, E. Duflo and L. Lmden, L. 2003. Remedymg Education: Evidence from Two Randomued Expenments inIndia. Unpublished. Banerjee, A. andDuflo, E. 2005. Improvmg Healthcare Delivery inIndia. Mimeo, MIT. Center for Education Research and Innovation. 1998. Education at a Glance: OECD Indicators 1998, Orgarnabon for Econormc Co-operation and Development. Chaturvedi, I.S. 2004. Elementaly Education in Bihar. Towards Better Management. MPA/ID SYF'A, Kennedy School of Government. Chaudhury, N.,J. Hammer, K.Muralidharan, M.Kramer, and F.H.Rogers. 2004. Teacher and Health Care Provider Absence: A Multi-Country Study. Unpublished. Chatterjee, M. 1989 Pnmary Educabon and Pnmary Health: Dlrections and Intersections. Paper prepared for the World Bank Country Econormc Memorandum on Poverty, Employment and Social Services, The World Bank, Washmgton D.C., processed. Chatterjee, M. 1990. Decentralization o fHealth Care: How Does HealthPolicy Fit? Paper presented at the Conference on Decentralizabon, NationalInstitute o f Advanced Studies, Bangalore. Clarke, P 2003. Education Reform zn the Education Guarantee Scheme in MP, India and Fundescola, Brazil. World Bank, March2003. Clarke, P., and J. Jha. 2004. Improvmg Educabon Service Delivery m Rajasthan: Collaborating between Civil Society and Government. Mimeo, The World Bank. Constitution ofIndia, Seventy-Thud Amendment Act, Article 243 (A-0) of Part IX. Coulson, A. 2003. "Implementing Education for All: Moving from Goals to Action." Paper presented at the Fondazione Liberal's Second International Education Conference, Milan, Italy, May 17th, 2003. Das, J. and J. Hammer. 2004a. "Whch Doctor? Combmmg Vignettes and Item Response to Measure Doctor Quality." Policy Research Worlung Paper Series, #3301. The World Bank, forthcommg inJournal of Development Economics. 2004b. "Stramed Mercy: Quality of Medical Care 111Delh." Economic and Political Weekly, February28,2004. Dasgupta, A. 2004. "Alternative Policies on Education and Health with Special Reference to the State o f West Bengal and India." Presentation made at Harvard Umversity,November 2004. Datt, G. and M. Ravallion. 1994. "Transfer Benefits From Public Works Employment: Evidence for Rural India." TheEconomic Journal, Vol.104, No. 427, 1346-1369 Deshmgar, P and Johnson. 2003. "State Transfers to the Poor and Back: the Case o f the Food for Work PrograminAndhra Pradesh." Workmg Paper 222, Overseas Development Institute, www.odi.orE.uk. 132 Dobbm, F., forthcomng. "The sociological view o f the economy." InF Dobbm (ed.), TheNew Economic Sociology: An Anthology. Pnnceton, NJ: PrincetonUmversityPress. Duflo, E. and Hanna, R. 2005. Momtonng Works: Getting Teachers to Come to School. Mimeo, MIT. Feedback Ventures. 2004. Draft RuralWater and Samtation Sector Assessment Study o fAndhra Pradesh. Filmer, D. 2005. Educational Attainment and Enrollment Profiles: A Resource "Book" based on an Analysis o f Demographc and Health Survey Data. (Penodically updated). Development Research Group, The WorldBank. see: http://www.worldbank.org/research/projects/edattam/edattam. htm Filmer, D. and L.Pntchett. 1998. Educahonal Attament Profiles of the Poor (and hch): DHS Evidence fiomAround the Globe. Mimeo, The World Bank. (http://www.worldbank.org/research/projects/edattai/edattam.htm). 1999 "Educational Enrollment and Attainment inIndia: Household Wealth, Gender, Village, and State Effects." Journal of Educational Planning and Administration, 13(2): 135-64. Foster, V. 2002. "Ten Years o f Water Service Reform in Latin America: Towards an Anglo-French Model." In Paul Seidenstat, David Haarmeyer, and Simon H a h m (eds.), Reinventing Water and WastewaterSystems: Global Lessonsfor ImprovingManagement. New York: JohnWiley and Sons, Inc. Galasso, E. and M. Ravallion. 2003. "Social Protection in a Crisis: Argentina's Plan Jefes Y Jefas." Policy Research Worhng Paper Senes 3165, The World Bank. Government o f India (GOI). Website o f the Mimstryo fPanchayats: www.panchavat.nic.in, Website o fthe Department o f DnnlungWater Supply. www.ddws.mc.in. .1993. EighthFive Year Plan (1992-1997). Plamng Comssion, GOI. 1998. NinthFive Year Plan (1997-2002). Plamng Comssion, GOI. 1999a. Guidelines on Central Rural Samtation Campaign, Total Sanitation Campaign. .1999b. Rajiv Gandhi National Dnnlung Water Mission, Department o f Drlnlung Water Supply, MhstryofRuralDevelopment. 2000. Status o f Elementary Teachers of India, Year 2000. Assessment o f EFA, Mhstry o f HumanResource Development. 2001a. Review o f the Worhng o f the Constitutional Provisions for Decentralization (Panchayats): A Consultation Paper. National C o m s s i o n to Review the Worlung o f the Constitution (NCRC), GOI. 2001b. Report o f the Task Force on Devolution o f Powers and Functions Upon Panchayan Raj Institutions. Mimstry ofRural Development: Department o f Rural Development (Panchayah Raj Division), GOI. 2001c. TenthFive Year Plan (2002-2007), Chapter 5. Plannmg Comssion, GOI. 2002a. Comptroller and Auditor General of India Report on Distnct Primary Educahon Program, MinistryofHumanResource Development. 133 2002b. Pradhan Mantn Gram Sadak Yojana: Scheme and Guidelines. Mhstry o f Rural Development, GOI. 2002c. Report o f the National C o m s s i o n to Review the Workmg of the Constitution (NCRC). MhstryofLaw, Justice andCompany Affam, DepartmentofLegalAffam. 2003. Draft M O U Guidelines 06/11/03. Rajiv Gancllu National Dnnkmg Water Mission, Department o f Drlnlung Water, Mimstry o f Rural Development, GOI, http://ddws.mc.in/Data/SwaialiDO- JS.htm. .2004a. Annual Report o fthe Mimstry o fRural Development. Government o f India. 2004b. Guidelines: Sampoorna Grameen Rozgar yojana (SGRY), Mbstry of Rural Development, GOI. 2004c. Report o f the Task Force o f Officials in Charge o f Punchayah Raj m States to Examne the Centrally Sponsored Schemes.MimstryofPunchayahRaj, GOI. Government o f India. 2001. National HumanDevelopment Report 2001. Plamng Comrmssion, Government o f India. 2001 Government o f India. 2004. Analysis o f Budget Expenditure on Education, Department o f Education, Plamng and Momtormg Umt, MinIstryo fHuman Resource Development, Government o f India Government o f Karnataka. 2001. Government Order N o RDP 145 PPM 98 dated 19th June 2001, Department o f Rural Development andPanchayat Raj. 2002a. Government Order N o RDP/384/GPA/2002 dated 231d November 2002, Department o f Rural Development andPanchayat Raj. 2002b. Report of the Workmg Group on Decentralization. Department o f Rural Development and PunchayahRaj. Government o f Kerala. Website o fthe Department o fPanchayats: httr,://www,localgovkerala.netllsgd/LinkslResearchiStatusofpanc~u~.htm. 1999 Final Report o f C o m t t e e on Decentralization o f Powers, Part A, October, 1999 2001. The Kerala Water Supply and Sewerage (Amendment) Act, 1993. Government o f Rajasthan. 1999 Pragatl Pratlvedan 1998-99 Department o f Elementary and Secondary Education, Jaipur. Government o f West Bengal. 1973a. West Bengal Panchayat Act, Law Department, Government o f West Bengal. 1973. West Bengal Primary Education Act. Law Department, Government o f West Bengal. Hanushek, E. 1995. "Interpreting Recent Research on Schooling in Developing Countries." The World Bank Research Observer, August 1995, pp.227-246. Harilal, K.N. 2004. "Refomng Local Governance Structures in India: Lessons from the Kerala Expenment." Paper Presented m the International Symposium, Citlzen's Participation in Local Public Policy Makmg: European Experiences in Global Perspectives, http://lorc.ryukoku.ac.jp/docs/Harilal- paper.pdf. 134 Heaney, J. P., D. Sample and L.Wnght. 2000. Cost Analysis and Financing of Urban `WaterInfiastructure, Chapter 10. InR. Fieldet a1(eds.), Innovative Urban Wet-Weather Flow Management Systems, Lancaster, PA. Technomcs Publishmg Co. Howes, S. and R. Murgai. 2004. Subsidies and Salaries: Issues in the Restructuring of Government Expenditure in India. World Bank. Imam, K. 2000. "Targetmg Versus Universalism: An Evaluation o f Indirect Effects -- the Employment Guarantee Scheme inIndia." Department o fEconomcs, RoyalHolloway, Umversity o f London. International Institute for Population Sciences. 1999 IndiaNationalFamily and Health Survey-2. 1999 Institute o f Development Studies. 1998. Bnngmg Citizen Voice and Client Focus into Service Delivery: EducahonGuarantee Scheme, Madhya Pradesh, India. Umversityo f Sussex, UK. Isham, J., D. Narayan and L. Pritchett. 1995. "Does participation improve performance? Establishmg causality wth subjechve data." WorldBank Economic Review, 9(2): 175-200. Jalan, J. E. and Glinskaya. 2003. Improvmg Primary School Education 111 India: An Impact Assessment o f DPEP-Phase I. Unpublished. Jenluns, M.W. 1999. PhDthesis, Umversity o f Califoma at Davis, Department o f Civil Engmeering. Kerala Institute o f Local Admmstration. 2005. Survey on State o f Panchayah Raj Instituhons and Public Services inKerala. Survey done for World Bank. Kerala Insbtute o f Local Adrmnistrahon. 2005. Kingdon, G.G. 1996a. "Pnvate Schooling in India: Sue, Nature and Equity Effects." Economic and Political Weekly, 31, No. 51.3306-3314, December 1996. 1996b. "Student Achevement and Teacher Pay: A Case Study of India." Discussion Paper No. 74, Development Economcs ResearchProgram, STICERD, London School o f Economcs, August 1996. Kingdon, G. G. and M. Muzammil. 2000. "A Political Economy o f Education inIndia: The Case o f U.P " Mimeo; first halfpublishedmEconomic and Political Weekly, August 11,2001. Kremer, M., N. Chaudhury, J. Hammer, K. Muralidharan, and F.H. Rogers. 2004. Teacher Absence in India: A Snamhot. Mimeo submttedto the Journal o f the EuropeanEconomc Association. Kremer,M., S. Moulin and R.Namunyu. 2003. "Decentralization: A Cautionary Tale." PovertyAction Lab Worlung Paper. Leonard, Dutch. 2002. Personal commumcation cited m L. Pntchett and M. Woolcock, 2004. "Solutions When the Solution I s the Problem: Arrayng the Disarray in Development." World Development, 32.2 (February 2004): 191-212. Leonard, K., G.R. Mliga and D.H. Manam. 2002. "Bypassing Health Centers inTanzama: Revealed PreferencesFor Quality." Journal ofAfiican Economies, vol. 11, issue 4, pp.441-471, Mahal, A., 3. Singh, F Afiidi, V Lamba, A. Gumber and V Selvaraju. 2002. Who Benefits from Public Sector Health Spending in India. New Delhi:National Council for AppliedEconomc Research, 2002. Mathew, G. et.al. 2000. Status of Panchayati Raj in the States and Union Territories of India 2000. Indian Institute o f Social Sciences.New Delh:Concept. Mathur, 0.P 2000. "State Local Fiscal Relations." In D.K. Snvastava (ed.), Fiscal Federalism in India: Contemporary Challenges. National Institute o f Public Policy. As quoted m GO1 2001, "Review o f the 135 Workmg o f the Constitubonal Provisions for Decentralization (Panchayats): A Consultation Paper," NationalCoxmussion to Review the Workmg o fthe Constitution. Mehrotra, S. and P Buckland. 2001. "Managing School Teacher Costs for Access and Quality m Developmg Countnes -A Comparative Analysis." Economic and Political Weekly, December 8, 2001. Mehrotra, S. 2002. "Basic Social Services for All? Ensurmg Accountability Through Deep Democrahc Decentralization." BackgroundPaper for HumanDevelopment Report 2002. Misra, A. 2005. Report for Expenditure Trackmg Exercise for SGRY and Mid-Day Meals. Report prepared for SouthAsia Rural Development Department, World Bank Misra, R.K. 1997 National Civil Service System m India: A Cntical View. Paper presented at `Civil Service Systems m Comparative Perspective', Indiana Umversity, Bloomngton, Indiana, April 5-8, 1997 Murray, C.J.L., K. Styblo and A. Rouillon. 1993. "Tuberculosis." In D.T. Jamson et a1 (eds.) Disease ControlPriorities in Developing Countrzes. Oxford UmversityPress for the World Bank, New York. Narayan, D. 1995. "Contribution o f People's Participation: Evidence from 121 Rural Water Supplies Projects, Envuonmentally Sustainable Development." Occasional Paper Senes, No.1, World Bank, Washmgton D.C.. Nayak, R., N.C. Saxena and J. Farmgton. 2002. "Reachmg the Poor: The Influence o f Policy and Admmstrative Processes on the Implementation o f Government Poverty Schemes in India." Workmg Paper. Oommen, M.A., forthcomng. Fiscal Decentralizabon to Rural Local Bodies: A Case Study o f Kerala. A World Banksponsored study, 2003. Pande, R., R. Vijayendra and T. Besely. 2005. Participatory Democracy 111Action: Survey Evidence from South India. World Bank. Peters, D.H., AS. Yazbeck, R.R. Sharma, G.N.V Ramana, L.H. Pntchett and A. Wagstaff, 2002. Better Health SystemFor Indias Poor. Findings Analysis and Options. World Bank, Washington D.C.. Pntchett, L. 2004. "Towards A New Consensus for Addresslng the Global Challenge o f the Lack o f Education." Copenhagen ConsensusChallenge Paper. Pntchett, L. and M.Woolcock. 2004. "Solutions When the Solution I s the Problem: Arraymg the Disarray inDevelopment." WorldDevelopment, 32.2 (February 2004): 191-212. PROBE Team. 1999 Public Report on Basic Education zn India. NewDelh: Oxford UmversityPress. Ramachandran, V and H. S e h 2000. Rajasthan Shiksha K a m Project: An Overall Appraisal. New EducabonDivision Documents No.7, SIDA, Embassy o f Sweden, New Deh. Rao, M.G. 2002. "Poverty Alleviation under Fiscal Decentralization." In M.G. Rao et a1 (eds.), Development, Poverty and Fiscal Policy. Oxford UmversityPress, 2002. Ravallion, M. and J. Jalan. 1999 "Income Gains to the Poor from Workfare: Estimates for Argentina's Trabajar Program." Policy ResearchWorklng Paper Series 2149, The World Bank. Ravallion, M. and Q. Wodon. 1999 "Does Child Labor Displace Schooling? Evidence on Behavioral Responsesto an Enrollment Subsidy." Policy Research Working Paper Series 2116, The World Bank. Registrar General o f India. 2003. SRS Based Abndged Lefe Tables, SRS Analytical Studies, Report No. 3 o f 2003, New Delh: Registrar General o f India 136 Scott, J. 1998. Seeing Like a State: How Certain Schemes to Improve the Human Condition Have Failed. New Haven, C T Yale UmversityPress. Seh, G. (ed.) 2004. Fiscal Decentralization to Rural Governmentsin India.Published for the World Bank by OxfordUmversityPress, NewDeh. Shariff, A., P.K. Ghosh and K.S. Mondal. 2002. "Indian Public Expenditures on the Social Sector and Poverty Alleviation Programs m the 1990's." Worlung Paper 169, Overseas Development Institute, www.odi.org.uk. Singh, AS., Jain, V.K., Gautam, S.K.S., and Kumar, S. 2004. L e a m g Acluevement o f Students at the Endof Class V Departmentof Educahonal Measurement and Evaluahon, National Council of Educational Research and T r a m g Snvastava, P 2004. "Targetmg Measures Used m Anti-Poverty Programs." Discussion Paper No. 5, ADB Institute, www.adbi.org. State Institute o f Panchayan Raj and Rural Development. 2005. Survey on State o f Panchayah Raj Inshtutions and Public Services mWest Bengal. Survey done for World Bank. State Institute o fPanchayan Raj and RuralDevelopment. 2005 The Pratich (India) Trust. 2001. The Praticlu EducationReport. TheStatesman, Kolkata, November 23, 2004. Tilak, Jandhyala B.G. and R. Sudarshan. 2001. "Pnvate Schooling mRural India." NCAER Worlung Paper No.76, Nahonal Council for Applied Economc Research, New Dellu. Tooley, J. 2001. "The Entemnseo f Education." Occasional Paper No.6, LibertyInstitute, India. Umted Nations Development Program (UNDP). 2002. Rajasthan Human Development Report 2002, Chapter IV. Primary Education mRajasthan. 2004. West Bengal Human Development Report 2004, Chapter 7. Pnmary Education m Rajasthan. UNICEF and WHO. 2004. Meeting the MDG Drlnlung Water and Samtation Target: A Mid Term Assessment o f Progress, August 2004. UmtedNations Development Program. 2004. HumanDevelopment Report 2004. UNDP Water and Samtation Project, South Asia (WSP-SA). 2005. Scaling-Up Rural Samtation m South Asia: Lessons Learnedfrom Bangladesh, India and Palustan. WSP Regional Study, May 2005. Wmkler, D. and A.I. Gershberg. 2000. "Educahon Decentralization in Latin Amenca: The Effects on the Quality of Schooling." LCSHD Paper Senes, World Bank. World Bank. 1996. India: Primary Education Achievement and Challenges. South Asia Country Department 11, Waslungton D.C. World Bank. 1998. Reducing Poverty I n India: Optionsfor More Efective Public Services. A World Bank Country Study. World Bank. 1999a. Project Appraisal Document for Rajasthan Primary Education Project. Report No: 18934-IN7Education Sector, South Asia Region, Washington D.C. 137 World Bank. 1999b. Beyond the Center. Decentralizing the State. World Bank Latm Amenca and Caribbean Studies Viewpomts, L A C Region, Washmgton D.C. World Bank. 2000. Overview of Rural Decentralization in India - Volume I,Rural Development Umt, South Asia Region. World Bank. 2003. WorldDevelopment Report 2004. Making Services Workfor Poor People. Washmgton D.C. World Bank. 2004a. India: Fiscal Decentralization to Rural Governments. New Deb: Oxford Umversity Press. World Bank. 2004b. "Project Appraisal Document on a Proposed Credit to Republic o f India for an Elementary EducahonProject." World Education Forum. 2000. "Education for All: Meeting Our Collective Comtments." The Dakar Framework for Action, UNESCO. Zeckhauser, R. 1970. "Medical Insurance: A Case Study of the Tradeoff Between hsk Spreading and Appropnate Incentives." Journal of Economic Theoly, Vol. 2, pp. 10-26, 1970. StateActs Karnataka EducationAct, 1983, amended 1995 Mysore Public Health Act, 1944 KamatakaPanchayat Raj Act, 1994 Kamataka Ground Water Regulation Act, 2003 Kerala Panchayat Raj Act (amended 1994) Kerala EducationAct, 1958 (now Act 16,2000) Kerala Water Supply and Sewerage Act 14, 1986, Amendments and Other Regulahons Kerala Water Supply and Sewerage (Amendment)Act, 1993 Travancore-Cochm Public HealthAct, 1955 Kerala Panchayat Raj (Construction and Maintenance o f Public Latrines, Urinals, Bathng Places and Samtation of Private Premses) Rules, 1998 Rajasthan Panchayat Raj Act, 1994 West BengalPanchayat Raj Act, 1973 West Bengal Pnmary EducationAct, 1973 Judgments C. Rangaswarmah vs. State o f Karnataka (1998) 6 SCC 66; State o f Punjab vs. Inder Singh (1997) 8 SCC. Central Inland Water Transport Corporahon vs. Brojo Nath Ganguly (1986) 3 SCC 156; DTC v DTC Mazdoor Congress (1991) Supp (1) SCC 600. J. Raghuram Judgment Para 113-126 111Ranga Reddy District Sarpanches Association vs. Government o f AP and Ors WP Nos. 12348 and 12356 o f2002; decided on: 29.01.2004, reportedas (2004)2 ALD 1. Secy, State o f Karnataka vs. Umadevi (2004) 7 SCC 132. La1vs. the State o f Uttar Pradesh, Supreme Court Judgment, SUPP SCC 28, 1984. 138