Author's personal copy - Patrick Heller...73rd and 74th constitutional amendment, local governments...

This article was originally published in a journal published byElsevier, and the attached copy is provided by Elsevier for the

author’s benefit and for the benefit of the author’s institution, fornon-commercial research and educational use including without

limitation use in instruction at your institution, sending it to specificcolleagues that you know, and providing a copy to your institution’s

administrator.

All other uses, reproduction and distribution, including withoutlimitation commercial reprints, selling or licensing copies or access,

or posting on open internet sites, your personal or institution’swebsite or repository, are prohibited. For exceptions, permission

may be sought for such use through Elsevier’s permissions site at:

http://www.elsevier.com/locate/permissionusematerial

http://www.elsevier.com/locate/permissionusematerial

Autho

r's

pers

onal

co

py

Building Local Democracy: Evaluating the Impact

of Decentralization in Kerala, India

PATRICK HELLERBrown University, Providence, RI, USA

K.N. HARILALCentre for Development Studies, Trivandrum, India

and

SHUBHAM CHAUDHURI *

World Bank, USA

Summary. — Historically, local rural governments in India have enjoyed very limited powers andcitizens have been afforded very few opportunities to shape local development. In 1996, the stategovernment of Kerala initiated the ‘‘People’s Campaign for Decentralized Planning’’ devolvingnew authority and resources to panchayats and mandating structures and processes designed tomaximize the direct involvement of citizens in planning and budgeting. In both its scope and design,these reforms represent the most ambitious effort to build local institutions of participatory dem-ocratic governance ever undertaken in the subcontinent. This paper provides a detailed evaluationand analysis of the formative period of the reforms based on extensive survey data collected in 2002from a sample of 72 randomly selected panchayats.� 2007 Elsevier Ltd. All rights reserved.

Key words — democracy, decentralization, participation, South Asia, India, Kerala

1. INTRODUCTION

Historically, local rural governments in Indiahave enjoyed very limited powers and citizenshave been afforded very few opportunities toshape local development. The 73rd Constitu-tional amendment passed in 1993 aimed to rem-edy this democratic deficit by granting localrural governments new powers and makingthem more accountable to citizens. Implemen-tation was however left to the states, and as isalways the case in India, inter-state variationhas been pronounced. The most determined ef-fort at democratic decentralization has been inthe state of Kerala. In 1996, a coalition of leftparties led by the Communist Party of In-dia—Marxist (CPI(M)) returned to power andimmediately fulfilled one of its most importantcampaign pledges by launching the ‘‘People’sCampaign for Decentralized Planning.’’ All

1,214 local governments in Kerala—municipal-ities and the three rural tiers of district, block,and gram panchayats—were given new func-tions and powers of decision making, and weregranted discretionary budgeting authority over

* The research project was funded by a generous grant

from the Ford Foundation, India. We would like

to thank the project’s field survey coordinators—N.

Jagajeevan, T.N. Seema, T.P. Sreedharan, and

Kalyanakrishnan—as well Rajesh Puliyara and Aaron

Katz who managed the database. We are especially

grateful to Mark Robinson and Bishnu Mohapatra from

the Ford Foundation. An earlier draft of this paper was

presented to the Network on South Asian Politics and

Political Economy in Paris in July 2004 where we

received comments from many participants. We are in-

debted to four anonymous reviewers for their excellent

comments. Final revision accepted: July 14, 2006.

World Development Vol. 35, No. 4, pp. 626–648, 2007� 2007 Elsevier Ltd. All rights reserved

0305-750X/$ - see front matter

www.elsevier.com/locate/worlddevdoi:10.1016/j.worlddev.2006.07.001

626

Autho

r's

pers

onal

co

py

35–40% of the state’s developmental expendi-tures. In addition to devolving resources, stateofficials sought to directly promote participa-tory democracy by mandating structures andprocesses designed to maximize the directinvolvement of citizens in planning and budget-ing. In both its scope and design, the campaignrepresents the most ambitious and con-certed state-led effort to build local institutionsof participatory democratic governance everundertaken in the subcontinent. This paperprovides a detailed evaluation and analysis ofthe first four years of decentralization basedon extensive records and survey data collectedin 2002 from a sample of 72 randomly selectedpanchayats. 1

2. DEMOCRATIC DEEPENING

The debate on democracy has shifted fromthe study of transitions to formal democraticrule, to the study of democratic deepening (Linz& Stepan, 1996). Of the many conceptual refor-mulations this has called for, none has beenmore central than a reexamination of the ques-tion of political participation. Going beyond thetraditional focus on electoral participation, in-creased attention has been directed toward theimportance of direct participation—that is, di-rect forms of engagement with public decision-making entities and processes—in determiningthe quality and depth of democratic institutions.In the developing world, participation is seen ascritical to increasing the overall capabilities ofcitizens (Dreze & Sen, 1995), strengtheningfragile democracies, improving the quality ofgovernance, and countering the influence of or-ganized and powerful dominant groups(Avritzer, 2002).

The challenge of promoting direct participa-tion is particularly acute in contexts where theauthority of the modern, rational–legal stateis contested by various forms of social power(Migdal, Kohli, & Shue, 1994). Under suchconditions, even when representative demo-cratic institutions are well established, basicrights of association are distorted by pervasivevertical dependencies (clientelistic relation-ships), routinized forms of social exclusion(e.g., the caste system, purdah) and the uneven-ness and at times complete failure of publiclegality. In many democracies, the publicauthority of the modern state radiates out un-evenly, and ‘‘the components of democraticlegality and, hence, of publicness and citizen-

ship, fade away at the frontiers of various re-gions and class, gender and ethnic relations(O’Donnell, 1993, p. 1361).’’ The resulting dis-juncture between formal and actionable rightsis reflected in the predominance of oligarchicalparties and patrimonial bureaucracies. Theresulting institutional matrix favors vertical,back-door, and elite-based forms of participa-tion, over horizontal, transparent, and citizen-based forms of participation. This problembecomes even more acute when one considersthe weakness and even complete absence inmuch of the developing world of local demo-cratic government.

Any theory of democratic deepening must assuch explicitly take into account how unequalsocial relations and uneven institutional envi-ronments impinge upon the exercise of citizen-ship. Yet when one considers that this pointhas achieved axiomatic status in the literatureon democratic deepening in the developingworld (Huber, Rueschemeyer, & Stephens,1997; O’Donnell, 1993), it is surprising justhow thin the empirical literature remains andhow little we actually know about the variabil-ity of citizenship densities. In large part theproblem stems from the myopias of disciplinaryspecializations. Much of the political scienceand macro-sociology literature take nationalinstitutions as the unit of analysis, neglectingas such the vast underside of local political life.Conversely, the literature that explores socialrelations and politics at the local level usuallypays scant attention to the larger institutionalcontext and, based as it is on case studies, doesnot lend itself to generalization.

In this light, the recent wave of decentraliza-tion initiatives in the developing world presentsa new and critical opportunity for research andtheory. This is especially true in India, wherethe passage of the 73rd and 74th constitutionalamendments marks a natural experiment ofsorts: a single treatment of basic institutionalreform (empowering local governments), to becarried out by varied political units with differ-ent political configurations (states) across anextremely heterogeneous social landscape. 2

The opportunities for empirical evaluation areall the more tantalizing because of the relativelylow starting point: weak, if not altogether ab-sent, local democratic institutions.

One of the defining characteristics of post-colonial democracies is that they have compar-atively centralized states, with few, if any localarenas in which citizens can meaningfully par-ticipate. In the case of India, even though

BUILDING LOCAL DEMOCRACY 627

Autho

r's

pers

onal

co

py

citizens vote in much higher percentages than inmost democracies and subordinate group elec-toral participation is increasing (Yadev, 2000),and even though Indians are probably far morelikely to participate in contentious action thanis the norm for older democracies, the averagecitizen enjoys few opportunities to engagegovernment between elections. The sociologicalproblem has been extensively examined: theassociational field is marked by direct barriersto participation in the form of enforced socialexclusions and indirect barriers in the form ofgroup-bounded distributions of social, cultural,and educational capital. 3 The institutionalspace for the exercise of local citizenship is alsohighly constrained. Before the passage of the73rd and 74th constitutional amendment, localgovernments had little authority, few resourcesand were subject to limited electoral account-ability. 4 The development functions of localgovernments were limited to acting as imple-mentation agencies for line department schemesand ordinary citizens were afforded few oppor-tunities to directly engage in or influence deci-sion making about public allocations. 5 Thefirst chief minister of Kerala, E.M.S. Namboo-diripad, made this point succinctly when henoted that ‘‘if at the level of centre-state rela-tions the constitution gave us democracy, atthe level of state-panchayat relations the consti-tution gave us bureaucracy.’’ 6

Some commentators have gone so far ascharacterize the state in India as ‘‘democraticauthoritarian’’ (Jalal, 1995). If such an assess-ment is overly reductionist, ignoring thedynamics of electoral competition and signifi-cant zones of democratic engagement that for-mal rights allow for, the point remains thatthe day-to-day exercise of citizenship is madeparticularly difficult by the insularity and dis-tance of state authority and the overall uneven-ness of public legality. Thus even as Indiaboasts one of the most stable and vibrantdemocracies in the developing world, civic-based participation remains poorly developed,resulting in what Chhibber (1999) has describedas ‘‘democracy without associations.’’ Chatter-jee takes this reasoning one step further byarguing that most inhabitants of India ‘‘arenot proper members of civil society and arenot regarded as such by the institutions of thestate’’ (2001, p. 8). In light of these assessments,it is difficult to overstate the transformativepossibilities of democratic decentralization inIndia. Creating genuine institutions of localgovernment is tantamount to expanding the

associational field, and more specifically toincreasing the opportunities and the effects ofdirect citizen participation in government. It isequally difficult to overstate the significantobstacles that democratic decentralizationfaces.

3. THE CAMPAIGN

Decentralization in Kerala presents animportant test case of the possibilities anddeterminants of democratic deepening. ThePeople’s Campaign for Decentralized Planning(the ‘‘campaign’’ hereafter) was specificallyconceived as a vehicle for deepening democ-racy. 7 We focus on the first four years of re-form, a period that can aptly be called the‘‘campaign’’ because of the mobilizationalmode in which it was implemented. This periodmarked a decisive rupture with the past, settingin motion processes that have seen a very sig-nificant devolution of resources and authorityto the panchayat level. The Left DemocraticFront (LDF) was voted out of power in 2001(in keeping with an unbroken pattern in Keralaof incumbent governments losing elections),and the subsequent period under a UnitedDemocratic Front (UDF) Government led bythe Congress Party ushered in a very differentstage of the reforms, marked in particular bythe challenges of institutionalization andstrengthening of local governance. Thoughour data do not cover this period, it must benoted that in contrast to earlier efforts at decen-tralization in Kerala the reforms introducedwith the campaign have largely been sustained.In Kerala’s highly partisan political environ-ment, the campaign has been the subject offierce political debate, but fiscal devolutionhas been preserved and no key legislative provi-sions have been altered. A 2006 report by theState Planning Board found that ‘‘PanchayatiRaj had been mainstreamed in Kerala . . . andthat there is widespread consensus across thepolitical spectrum in Kerala that PanchayatiRaj must be strengthened and supported inKerala (The Hindu, May 11, 2006).’’ TheLDF government that was returned to powerin May 2006 has expressed its commitment todeepening the reforms.

The campaign was the product of a long his-tory of broad-based demands for democraticreform emanating from many quarters, includ-ing Gandhian factions of the Congress partyand a range of Left social movements. Begin-

628 WORLD DEVELOPMENT

Autho

r's

pers

onal

co

py

ning with the first democratically elected CPIministry of 1957, there have been repeated at-tempts to decentralize. For reasons that havebeen examined elsewhere (Heller, 2005; Torn-quist, 1997), the pressures for decentralizationreached something of a critical political masswith the return of the CPI(M) to power in 1996.

As a set of far-reaching institutional reforms,the campaign evolved from a comprehensivecritique of the inefficacies of top-down, insu-lated, command-and-control bureaucraciesand of the myriad problems, both practicaland normative, of the local participation defi-cit. If the institutional goals of the campaignamounted to nothing less than creating localself governments with new resources andauthority, the political goal was to use planningas an instrument of mobilization (ThomasIsaac & Franke, 2002). As structured by theimplementing agency—the Kerala State Plan-ning Board (SPB)—the campaign was designedto create an active role for local citizens inshaping local development policy making andbudgeting. Thus, not only were local govern-ments charged with designing and implement-ing their own development plans (whichincluded designing and financing projectsacross the full range of development sectors),but they were also mandated to do so throughan elaborate series of nested participatory exer-cises in which citizens are given a direct role inshaping—rather than just choosing—policiesand projects. On both counts—local planningand citizen governance—the campaign goeswell beyond decentralization in West Bengaland Karnataka, the two most carefully docu-mented cases of successful decentralization inIndia (Bardhan & Mookherjee, 2004; Crook& Manor, 1998). These elaborate institutionaldesigns were publicized through intensive med-ia campaigns, conferences, and a massive train-ing program. The basic nested structure ofparticipation in the annual planning and bud-geting cycle consists of four discrete stages:

(1) Gram Sabhas are held at the ward level(there are 10–12 wards in a panchayat).The first Gram Sabha serves as an openforum in which residents identify local devel-opment problems, generate priorities, andform sub-sector Development Seminars inwhich specific proposals first take shape.Subsequent Gram Sabhas select beneficiariesfor targeted schemes. The Gram Sabhas areopen meetings, presided by local elected offi-cials, and facilitated by the Key ResourcePersons (KRPs). They are always held on

holidays, and in public buildings (usuallyschools). Preparations for the assembliesinclude extensive publicity, and the distribu-tion of various planning documents. Min-utes are kept, and each sub-sector grouppresents a report of its deliberations andproduces a list of ‘‘felt needs.’’(2) Development Seminars develop inte-grated solutions for various problems identi-fied at Gram Sabhas. The Seminars areconstituted of representatives selected bythe Gram Sabhas, members of panchayatsamithi, local political leaders, key officialsof the area, and experts from the localityand outside. The seminars were required toproduce a comprehensive planning docu-ment for the panchayat.(3) Task Forces are selected by the Develop-ment Seminars and are charged with con-verting the broad solutions of the seminarsinto project/scheme proposals to be inte-grated into the final panchayat plan. In gen-eral, a Task Force is constituted for each of10 development sectors, including women’sdevelopment, and includes a member of thepanchayat samithi, the relevant local officialand representatives selected by the GramSabhas.(4) The fourth phase of the annual planningexercise is the actual formulation of thepanchayat or municipal budget. Drawingon the shelf of sectoral projects designedby the Task Forces, the panchayat draftslocal plan based on available budgetaryresources, which include grant-in-aid (thelargest component), own resources (localtaxes and local resource mobilization) andstate or center-project funds.

Beyond its institutional design, the ‘‘mobili-zational’’ character of the campaign needs tobe highlighted. The state—and specifically theSPB—not only provided the procedural tem-plates, technical assistance and key oversightfunctions, and administrative capacity, but alsolaunched a massive training program of ‘‘KeyResource Persons’’ at the local and district levelthan involved over 100,000 people. In subse-quent years these training programs were ex-tended, and were more specifically targeted atwomen and Scheduled Castes and ScheduledTribes (SC/STs). 8 Civil society organizationsalso played a critical role in designing andimplementing the campaign, most notably theKerala Sastra Sahitya Parishad (KSSP)—thePeople’s Science Movement (Kannan, 2000;Parameswaran, 2000, 2001). 9 Parties and their


Autho

r's

pers

onal

co

py

affiliated organizations played a direct role inmobilizing participants. The KSSP suppliedmany field coordinators as well as the over-whelming majority of officials in the State Plan-ning Board. Also, in designing the campaignthe SPB relied heavily on a stock of practicalknowledge and ideas drawn from a 25-yearrepertoire of local level initiatives and plan-ning experiments led by the KSSP and otherNGOs. 10 During the course of the cam-paign, the KSSP and women’s groups were alsoinstrumental in organizing NeighborhoodGroups that were designed to increase women’sparticipation. In all of these respects, the cam-paign stands apart from most decentralizationreforms in the degree to which the boundariesbetween the state and civil society were blurred,both in terms of design and implementation. Inthis sense, the campaign is very much an exam-ple of the type of synergistic state–society rela-tions that Evans (1996) and others have arguedare key to sustained institutional reforms.

4. THE DATA

The data we report here were collected from72 (out of 990) rural panchayats selectedthrough stratified random sampling. The strat-ification was designed to wield a sample thatwas representative of regions and of the distri-bution of ruling political coalitions. The result-ing sample was found to be statisticallyrepresentative along a wide range of basic so-cio-economic factors (Chaudhuri et al., 2004).In each panchayat, investigators collected dataon socio-economic conditions and on theimplementation of the campaign (participationfigures, panchayats organizing activities, com-mittee composition, etc.). In each panchayat,we also administered a questionnaire to 13key respondents—politicians, government offi-cials and representatives of civil society organi-zations—in every panchayat. We chose keyrespondents over random respondents becausewe required informants with a detailed knowl-edge of local political and socio-economic con-ditions and most importantly with a detailedand practical understanding of the implementa-tion and impact of the campaign. Surveying keyrespondents or experts to develop profiles ofcommunities or complex institutional arrange-ments is a well established method for med-ium-N studies (Evans & Rauch, 1999). Wemoreover designed the sample of key respon-dents to maximize the range of opinions. First,

we surveyed members of both the ruling party(i.e., the party with a majority of the electedpositions in the panchayat council) and themain opposition party. Second, we spread oursample across the three sectors of government,elected representatives, and representatives ofkey civil society organizations. For reasons weelaborate throughout the paper, the differentstructural positions of each of these three sec-tors should shape (though not mechanicallydetermine) a priori interests vis-a-vis the cam-paign. Third, we recorded the political affilia-tion (or lack thereof) of each of ourrespondents. Given Kerala’s highly partisanpolitical scene, political affiliation should shapeviews of the campaign. In each panchayat,investigators identified and interviewed oneindividual in each of the following 13 catego-ries: panchayat member from ruling coalition;panchayat member from the main oppositionparty; head of the local branch of ruling party(but not panchayat member); head of the localbranch of main opposition party (but not panc-hayat member); three government officials(panchayat secretary and two line departmentofficials); six civil society activists (one civilianTask Force member and one representativeeach from a local union, a religious organiza-tion, a development NGO, a women’s organi-zation and an SC/ST organization). Thesurvey included a total of 220 questions andtook on average 4–5 hours to administer. Ques-tions covered three general areas: the character-istics of the panchayat (local social and politicalrelations, social movement history, level ofdevelopment), the process of the campaign(how participation and decision makingevolved), and the impact of the campaign ondevelopment, civil society, and social inclusion.In all, 858 key respondents were interviewed. 11

5. THE GENERAL IMPACT OF THECAMPAIGN 1996–2001

A number of studies have already establishedthat in institutional terms the campaign re-sulted in a significant reorganization of thestate and governance, and that the level andscope of decentralization surpasses what hasbeen achieved in most (and possibly all) Indianstates since the 1993 constitutional amend-ments (Thomas Isaac & Franke, 2002; Veron,2001; World Bank, 2000). First, there has beena very substantial fiscal decentralization: 35–40% of all developmental expenditures have


Autho

r's

pers

onal

co

py

been allocated directly to local governments(including village, block, and district panchay-ats, as well as municipalities). The largest com-ponent of these funds—grants-in-aid—areunder the direct control of panchayats. The in-crease in the discretionary portion of villagepanchayat budgets has been dramatic, jumpingfrom Rs. 1,000 million in 1996–97 (the year be-fore the campaign) to 4,204 million in 1997–98,and over 5,000 million in each of the three yearsthat follow (Government of Kerala, State Plan-ning Board, 2001). In comparative terms, thescale of the devolution is also significant. AWorld Bank report found that Kerala has thegreatest degree of local expenditure autonomyand is the most fiscally decentralized state in In-dia, and second only to Columbia in the devel-oping world (2000, vol. I, pp. 28–29). 12 In theirreview of the campaign, Chathukulam andJohn conclude that ‘‘the lack of finance willno longer be a major hurdle for local bodies’’(2002, p. 19).

Second, the campaign has brought very sig-nificant devolution of authoritative decision-making powers. The legislative architecture ofthe campaign empowers panchayats with fullauthority to plan, fund, and implement a fullrange of development policies and projects.As of 1999, Kerala was the only state alongwith West Bengal where there was no cap onthe powers of panchayats to sanction expendi-tures (Chaudhuri, 2006). Although during thecampaign projects were subject to sanction bya district level technical committee, these com-mittees acted mainly in an advisory capacity.The SPB has also stipulated broad patterns ofexpenditure (e.g., 40% of the budget must goto economic development) and imposed man-datory minimum expenditures on SC/STs andwomen. These guidelines notwithstanding,there is little doubt that panchayats enjoyedand exercised a wide degree of discretionaryauthority in budgeting, and that these preroga-tives have been preserved in the post-campaignperiod (2001–present). The degree and unifor-mity of institutional activity during the cam-paign is reflected in our finding (based onlocal records and the survey) that all the panc-hayats in our sample held Gram Sabhas on aregular basis, constituted Task Forces, devel-oped local plan documents, and created benefi-ciaries committees.

Third, as a number of studies have docu-mented, participation in the campaign washigh. In an analysis of SPB data collected fromGram Sabha registries, Chaudhuri and Heller

(2003) found that in each of the first two yearsof the campaign, 1.8 million people attendedplanning Gram Sabhas. The social compositionof participation was also quite inclusionary,with women representing 41% of participantsby the second year, and SC/STs actually beingoverrepresented. 13

In sum, local governments in Kerala nowmake real decisions based on significant levelsof citizen participation. But this leaves opentwo sets of questions that need to be exploredmuch more closely. First, what has been thesubstantive developmental impact of local gov-ernment under the campaign? Have the de-volved funds been spent effectively? Did localallocation of development resources improveupon the prior model of department-drivendevelopment? Have there been, in other words,efficiency gains from devolution? The literatureon decentralization generally takes for grantedthe comparative advantage of local governmentin identifying and targeting local developmen-tal needs. In institutional terms, the informa-tional and allocative advantages of thesubsidarity principle are clear, but if gover-nance is as much about institutions as it isabout power, then allocative gains will dependon the institutional interface with local socialand political conditions, including well-knownproblems of local elite capture (Mohan &Stokke, 2000). As such it needs to be demon-strated that local allocative power does indeedproduce comparatively superior outcomes, ademonstration that has only rarely been madein the literature.

Second, to what extent did the campaign con-tribute to empowering citizens? Definitionalproblems with concepts of empowerment aside,the campaign was designed with the intent ofincreasing the level and the effect of directinvolvement of citizens in the decision-makingprocesses of local government. As measuredby attendance at Gram Sabhas, participationlevels were high. But while participation in pop-ular assemblies is in and of itself meaningfuland certainly demonstrates that the process car-ries a certain legitimacy (otherwise ordinary cit-izens would not make the effort to attend), onehas to be guarded when dealing with non-elec-toral forms of participation of conflatinginvolvement with actual influence. 14 In dealingwith this problem, we argue for three separatelines of analysis. The first, and most obviousin an inegalitarian society, is the depth of par-ticipation (i.e., the magnitude and social com-position of participation). The second is the


Autho

r's

pers

onal

co

py

quality of participation, that is, the extent towhich the participatory process approximatesa deliberative process that involves substantivepublic discussions and efforts to identify com-mon interests rather than just the aggregativelogic of voting or the mobilization of resourcesassociated with bargaining. The third is the pro-cess dimension of participation, that is, the ex-tent to which participatory inputs wind theirway through institutions and are translatedinto actual outputs.

6. DEVELOPMENT IMPACT OF THECAMPAIGN

Measuring the development impact of decen-tralization is notoriously difficult. Expendituredata for local governments are often unreliable(Bardhan, 2002), before and after comparisonsare hard to construct 15 and it is difficult toactually isolate the impact of local governmentgiven a range of other intervening variablesthat are either difficult to measure or go unob-served. The survey however did produce datafrom key respondents, who as informed localactors are well positioned to evaluate the devel-opmental impact of the campaign. The keyrespondents answered a series of questions

about the nature of services (such as healthcare, education, roads) and about developmentactivities (such as housing for the poor, supportof agriculture, efforts to improve income andemployment), both before and after the cam-paign.

As has been well documented in the litera-ture, Kerala leads all major Indian states in so-cial development indicators and in the past twodecades has seen the most rapid decline of ruralpoverty rates of any Indian state (Ravallion &Datt, 1999). Yet despite Kerala’s track record,our respondents felt that the initial quality ofservices and development in 1996—that is, be-fore the campaign was launched—was poor.For all but three areas (education, child care,and health), a majority of respondents judgedthe quality to be ‘‘low’’ or ‘‘very low,’’ withemployment generation and anti-poverty mea-sures receiving particularly low marks. It is alsonoteworthy that the lowest score was for ‘‘ef-forts to improve income and employment forwomen.’’ Our respondents were then askedwhether the quality of services and develop-ment, in each of 13 categories, had improved,deteriorated or stayed the same during the fiveyears of the campaign during 1996–2001. Asthe results in Table 1 indicate, for every cate-gory, a large majority of respondents felt there

Table 1. Respondents’ assessment of the magnitude and direction of the change in service delivery and development as aresult of the campaign

Indicator (quality, accessibility,and/or level, depending onthe indicator)

Fraction of respondents who said situation inpanchayat had

No. of respondents

Deteriorated No change Improved

Significantly Some Some Significantly

Primary health care 0.3 1.2 9.1 66.1 23.4 816Child care and child

development0.2 0.5 1.6 36.5 61.2 825

Primary education 0.1 0.7 16.8 60.9 21.5 806Drinking water 0.0 1.4 8.2 48.1 42.3 813Sanitation 0.3 0.4 9.3 48.5 41.7 811Housing for the poor 0.1 0.5 0.8 28.3 70.3 832Assistance to poor 0.0 0.3 10.5 56.9 32.4 809Roads 0.2 0.2 3.1 34.1 62.3 827Irrigation facilities 0.3 1.6 13.1 61.7 23.4 811Support for agricultural

cultivators0.2 1.1 8.5 61.8 28.4 821

Income and employment creation 0.3 1.1 25.9 62.7 10.0 807Income and employment

for women0.1 0.7 17.7 56.1 25.4 816

Income and employmentfor SCs or STs

0.3 0.6 21.1 56.9 21.2 807


Autho

r's

pers

onal

co

py

had been improvement (either ‘‘some’’ or ‘‘sig-nificant’’). In 5 of 13 categories, over 40% of therespondents felt that there had been ‘‘signifi-cant’’ improvement. The performance of panc-hayats was however uneven across areas. Thecampaign’s most marked successes were inbuilding roads, housing for the poor, andanganwadis (child services) where almost two-thirds felt the difference was significant. In con-trast, less than a fourth of respondents felt thatthe panchayats had made a significant differ-ence in economic development (employment,agricultural support, and irrigation).

Even more revealing than this overall findingis what we find when we disaggregate responsesby the identity and function of the respondent.Kerala has a long history of highly partisan lo-cal politics. Local elections to the panchayatare fiercely contested by all the major politicalparties, electoral participation is high, socialmovements and community-based organiza-tions are active (and often contentious) andsurveys show that Keralites are far more ideo-logical and likely to vote along party lines thanin other states (Beasely et al., 2004). We wouldas such anticipate sharply differentiated viewsof the campaign. Yet the overall positiveimpression of the impact of the campaign holdstrue across all respondent categories (see Table2). Thus for all 13 questions, a majority in allrespondent categories (which includes govern-ment officials, ruling politicians, oppositionpoliticians, and civil society actors) felt therehad been improvement. It is particularlyinstructive to note the response of oppositionpoliticians who logically would be more skepti-cal of the campaign’s impact. The Heads of theLocal Branch of the Opposition Party had anoverwhelmingly positive view of the campaign.Similarly, government officials from linedepartments, who stand to lose from the devo-lution of authority to local elected officials andcivil society, also had an overwhelmingly posi-tive view. In sum, irrespective of their functionor political affiliation, our respondents believethat the campaign led to improvement in devel-opment. 16

When we examine the percentage of respon-dents who expressed the strongest positiveopinions, that is, those who indicated ‘‘signifi-cant improvement,’’ the patterns reveal morepronounced differences of opinion (Tables 3aand 3b). As one might also expect, the Headof the local branch of the main opposition partywas consistently least likely of all respondentsto feel that there had been significant improve-

ment, to the point of being something of anoutlier. His or her political party counterpartwho serves on the panchayat samithi howeverheld views that were consistently closer to thoseof the ruling-party panchayat member. In otherwords, elected representatives from differentparties are more likely to share similar viewsof the campaign than elected and non-electedofficials of the same party. This indicates thatfunction plays a more significant role in shap-ing opinions than party. Given Kerala’s contextof highly partisan local politics, this is some-what surprising, but may speak to the fact thatsupport for an increased role for local govern-ment was bi-partisan.

The differences of opinion between govern-ment officials are also telling. Line departmentofficials had a lower evaluation across all cate-gories than the panchayat secretary. This gapmight be explained by the differences in theirties to the state. Panchayat secretaries have along history of working with and for local gov-ernment. Line department officials however an-swered to their departments before theirpositions were changed by the campaign. It iswidely reported that line department officialsresisted the campaign, both because of existingpatterns of line department controlled patron-age and the status loss associated with beingmade subordinate to local elected officials(Chathukulam & John, 2002).

The pattern across the civil society respon-dents (the bottom six rows of Table 3a) also re-veals some interesting trends. Overall, theserespondents had views that are closest to thoseof politicians in the panchayat samithi. Butwithin this group there is a clear outlier, therepresentative of caste associations or religiousorganizations. Across every category, theserespondents had lower evaluations than therepresentatives of functional organizations.One must be careful of reading too much intothis finding, but the trend is certainly consistentwith assumptions made in the democracy liter-ature that secular organizations are more likelyto favor inclusive forms of representation thanorganizations built on the strength of primaryidentities. In the Kerala context, where Muslimand Christian community organizations have along history of opposition to the CPI(M), thispattern may also simply reflect political posi-tions.

Finally, party affiliation clearly shapes viewsof the campaign. Those who identified withparties aligned with the LDF had distinctlymore favorable opinions of the campaign than


Autho

r's

pers

onal

co

py

those with non-LDF affiliations (Table 3b).Those without party affiliations (mostly gov-ernment officials) fell roughly in between. Itshould be noted that the difference may in partbe an artifact of context. LDF panchayats hadhigher levels of participation in the initial stagesof the campaign than UDF panchayats (Chau-dhuri & Heller, 2003) and it is widely believedthat LDF panchayats were more active in nur-turing local planning. Having said this, UDFrespondents still had an overwhelmingly posi-

tive view. In fact, the difference between LDFand UDF respondents is less pronounced thanthe difference between opposition party panc-hayat member and opposition branch headsexamined earlier (compare with Table 2). Thissuggests once again that function is moreimportant than party affiliation in shapingopinions of the campaign.

Our data also provide some insight into theredistributive impact of the campaign. First,the high impact areas of the campaign were in

Table 2. Fraction (%) of respondents, by category and indicator, who felt situation in panchayat had improved as aresult of the campaign

Respondent category Primaryhealth

Childcare

Primaryeducation

Drinkingwater

Sanitation Assistance forpoor

Housing Other

Panchayat member fromruling party or coalition

92.0 98.7 88.2 96.1 97.4 97.4 89.5

Panchayat member from opposition 85.9 95.4 73.4 87.3 89.1 100.0 85.9Head of local branch of ruling party 87.7 95.4 87.7 92.3 92.4 100.0 86.2Head of local branch of main

opposition party80.3 90.3 65.6 79.4 77.8 95.2 79.4

Panchayat secretary 95.3 98.5 85.7 95.3 95.3 98.5 89.1Panchayat-level implementing officer 92.6 96.4 84.9 86.8 86.8 98.2 84.9Other line department official 90.8 100.0 84.4 87.5 84.4 98.5 92.1Civilian Task Force member 96.7 100.0 86.4 91.8 93.3 100.0 91.5Representative of caste or

religious organization74.6 100.0 71.7 88.5 86.5 96.3 84.6

Representative of labor union orfarmers’ association

96.7 98.4 84.5 88.3 86.4 98.4 95.1

Representative of development NGO 92.4 96.9 84.9 92.4 92.4 100.0 95.5Representative of women’s group 90.3 100.0 87.3 93.8 95.2 100.0 93.7Representative of SC or

ST organization85.7 100.0 83.6 93.6 91.9 98.4 91.5

Roads Irrigation Support foragriculture

Income and job creation

General Women SC/ST

Panchayat member from ruling partyor coalition

98.7 94.7 92.1 83.8 85.5 88.2

Panchayat member from opposition 95.4 78.5 85.9 59.7 81.5 68.8Head of local branch of ruling party 97.0 93.8 95.3 75.4 76.9 78.5Head of local branch of main

opposition party87.1 71.0 79.4 52.4 66.7 66.7

Panchayat secretary 100.0 92.1 92.1 76.6 84.1 79.4Panchayat-level implementing officer 100.0 89.7 90.0 70.9 71.4 74.1Other line department official 95.4 82.0 92.1 77.1 88.9 83.3Civilian Task Force member 96.8 77.8 90.3 80.0 88.5 86.4Representative of caste or religious organization 92.7 84.6 85.2 59.6 75.0 69.2Representative of labor union or

farmers’ association96.8 88.3 87.3 70.5 80.3 75.4

Representative of development NGO 100.0 86.4 96.9 80.6 85.1 80.6Representative of women’s group 96.9 83.9 93.8 78.1 87.5 87.5Representative of SC or ST organization 95.2 81.4 90.2 76.3 85.0 72.9


Autho

r's

pers

onal

co

py

housing to the poor, child services, and roads.All three of these areas have a pro-poor im-pact. 17 Second, 89% of the respondents indi-cated that the primary beneficiaries of thecampaign were the ‘‘socially and economicallydisadvantaged,’’ with only 4.0% indicating the‘‘socially and economically advantaged.’’ Simi-larly, when asked if ‘‘beneficiaries selected forvarious schemes under the campaign were moreor less likely to be poor than those selected un-

der earlier schemes,’’ 88% indicated more likelyunder the campaign.

7. BUILDING LOCAL DEMOCRACY

In this section, we discuss the depth, quality,and process of participation. The literature onparticipation has demonstrated that even in‘‘mature’’ democracies, participation rates are

Table 3a. Fraction (%) of respondents, by category and indicator, who felt situation in panchayat had improvedsignificantly as a result of the campaign

Respondent category Primaryhealth

Childcare

Primaryeducation

Drinkingwater


Housing Other

Panchayat member from rulingparty or coalition

26.9 67.9 26.0 49.4 47.4 75.9 40.5

Panchayat member from opposition 21.2 46.3 25.8 35.4 47.0 68.7 43.9Head of local branch of ruling party 25.8 68.7 22.4 47.8 52.9 80.9 40.3Head of local branch of main

opposition party10.9 40.0 15.4 25.8 26.2 42.4 19.7

Panchayat secretary 25.0 72.5 20.9 54.4 50.0 72.5 32.4Panchayat-level implementing officer 12.3 45.8 16.4 43.9 30.9 62.1 26.8Other line department official 17.9 55.1 19.7 33.3 36.4 66.7 21.5Civilian Task Force member 26.2 66.7 22.2 47.7 48.4 80.6 34.9Representative of caste or

religious organization14.3 56.1 11.1 30.2 30.2 60.6 24.5


25.0 71.8 18.0 44.4 45.2 75.8 35.9

Representative of development NGO 35.3 67.2 20.6 38.2 38.2 75.4 33.3Representative of women’s group 35.4 67.6 28.8 43.3 44.6 83.8 36.4Representative of SC or

ST organization26.6 68.3 29.0 50.8 46.0 71.4 31.7



General Women SC/ST

Panchayat member from ruling partyor coalition

70.9 23.4 34.2 13.3 28.9 26.3

Panchayat member from opposition 54.5 34.8 36.4 14.1 20.0 20.3Head of local branch of ruling party 70.6 32.8 30.3 9.0 23.1 20.0Head of local branch of main

opposition party46.2 12.7 12.3 6.1 17.5 15.9

Panchayat secretary 73.5 28.4 32.8 6.0 26.9 23.8Panchayat-level implementing officer 58.3 19.4 31.3 3.4 12.5 13.0Other line department official 62.7 22.2 32.3 12.7 23.8 21.7Civilian Task Force member 62.7 26.9 31.8 10.9 32.8 25.4Representative of caste or

religious organization51.8 13.2 20.0 9.4 23.1 19.2


61.5 22.2 28.8 10.9 19.7 19.7

Representative of development NGO 72.1 19.1 24.2 10.1 35.8 14.9Representative of women’s group 56.7 24.6 39.4 18.2 34.4 29.7Representative of SC or ST organization 63.5 18.0 27.4 10.0 28.3 23.7


Autho

r's

pers

onal

co

py

highly uneven across socio-economic groups(Verba, Schlozman, & Brady, 1995). Case stud-ies and surveys in India have found that casteand gender present particularly significant bar-riers to participation (Agarwal, 2001; Chhib-ber, 2003; Jayal, 2001; Seema & Mukherjee,2000). 18 Mechanisms of social exclusion in-clude coercion, intimidation, various forms of‘‘social purdah’’ or deficits in the individualskills, and self-confidence to participate directlyand openly in political life. Moreover, evenwhen ordinary citizens do participate, the selec-tivity of governance institutions intervenes be-tween citizen preferences and policy outputs.

Data collected by the State Planning Boardfrom all 990 panchayats for the first two yearsof the campaign show that 10.3% of the elector-ate participated in the first annual Gram Sab-has in 1996 and 10.6% in 1997. 19 The socialcomposition of the campaign improved drasti-cally in the second year. If in the first year ofthe campaign SC/ST 20 participation was wellbelow the average rate (relative participationwas 0.53 with 1.0 = participation rate of thegeneral population), by the second year it was1.44, meaning that SC/STs were participatingin greater proportions that non-SCs. Similarly,women’s relative participation increased from0.57 to 0.82, with women constituting 40% ofall participants in 1997–98.

The data from our 72 sample panchayatsshow that while overall participation has de-clined (falling to 4.7% of total population in1999 from 7.8% in 1997), its social compositionhas stabilized. In 1999–2000, women accountedfor 41% of participants, and SCs accounted for14% of participants, well above their propor-tion of the general population and their 11.5%

representation in the sample. STs on the otherhand constituted only 2.6% of participants, wellbelow their 3.7% of the sample population. Bycomparative standards these are very high lev-els of subordinate group participation. 21

We cannot take the democratic effects ofGram Sabhas for granted. Gram Sabhas arefairly large (average attendance in 1999 was1,271 per panchayat, or roughly 100 per ward)and while they serve important functions ofcommunicating basic information (criteria forbeneficiary selection, progress of the precedingyears’ budget, etc.) and for discussing and air-ing general demands, they do not directly makebudgetary decisions. Development Seminarsand Task Forces on the other hand are deliber-ative bodies charged by the Gram Sabhas withmaking concrete inputs into the planning andbudgeting process. Because they involve sub-stantive decision making and require moreskills, these bodies are more likely to be in-flected with power relations and one mightexpect a drop-off in subordinate group partici-pation in these bodies. And indeed, this is whatone study for the first year of the campaignfound in the case of women (Thomas Isaac &Franke, 2002).

Our data, which are for 1999–2000 and werecollected directly from panchayat registers, sug-gest that there has been an improvement in sub-ordinate group participation. In DevelopmentSeminars, we found that women representedon average 39% (78 out of 179 per panchayat)of total participants, only slightly lower thantheir participation in Gram Sabhas. Likewise,SC participation remains proportionately highat 14.6% (26 out of 179 participants). As wemove up the decision-making chain, we find

Table 3b. Fraction (%) of respondents, by party affiliation and indicator, who felt situation in panchayat had improvedsignificantly as a result of the campaign

Party affiliation Primaryhealth

Childcare

Primaryeducation

Drinkingwater


Housing Other

LDF affiliation 32.0 68.6 25.0 47.2 48.8 83.5 42.3No party affiliation 23.6 65.6 24.5 42.4 45.2 67.7 29.9Non-LDF party affiliation 12.1 50.7 17.6 35.3 33.3 54.2 25.1



General Women SC/ST

LDF affiliation 68.3 23.5 33.9 29.5 33.5 25.4No party affiliation 61.9 21.6 28.3 23.4 24.5 20.5Non-LDF party affiliation 50.5 27.0 23.6 18.7 13.2 17.6


Autho

r's

pers

onal

co

py

that in Task Forces women’s participation be-gins to decline. On average, women representedonly 30% of participants (24 out of 81). Thisdrop-off may in fact be an institutional artifactsince Task Forces are mandated to have anelected councilor and a government official asmembers, and both categories are overwhelm-ingly male. SC participation on the other handremains high at 12.3% (10 out of 81) and stillabove their representation in the sample(11.5%).

We can develop an even richer picture of thenature of participation in the decision-makingchain when we examine the sectoral composi-tion of Task Forces. The Task Forces were de-signed to be representative of all the keystakeholders in local democratic government,namely elected representatives, governmentofficials, and civil society actors. Since planningand development in India have long been dom-inated by politicians and officials, the challengehas been to incorporate civil society actors. Thecampaign appears to have done just that. Over56% of those who attended Task Force generalbody meetings in 1999–2000 were either civilianexperts or civil society activists, and 19% wereparty activists. Panchayat members and gov-ernment officials were roughly represented asofficially prescribed (i.e., one panchayat mem-ber and one official for each Task Force) withformer panchayat members (who could in prin-ciple be counted in civil society) making up thebalance. Moreover, our survey respondentsalso clearly found Task Forces to be account-able and representative. Fully 90% said thatTask Force members had been appointed inkeeping with established procedure (the factthat only 5% did not know is itself telling)and 70% judged that most of the Task Forcesin their panchayat were ‘‘representative of thestakeholders in that sector’’ with only 4% indi-cating that the Task Forces were not represen-tative.

If the actual composition of participatorybodies points to a continuity of representationthrough the institutional system, it is alsoimportant to separately evaluate the actual pro-cessing of participatory inputs. What impactdid these bodies actually have in making bind-ing decisions? Specifically, to what extent werethe ‘‘felt needs’’ (to use the language of thecampaign) expressed in the Gram Sabhas actu-ally translated into concrete decisions? 22 Avery small minority (3.8%) judged the processto be unresponsive, while 30% saw the feltneeds of the community as only ‘‘occasionally

reflected’’ in the panchayat plan. Almost 2/3(64%) of our sample answered that ‘‘felt needsare always reflected’’ in the final plan, including44% of heads of local opposition party, therespondent group with the least favorable viewof the campaign. If we examine this question bypanchayat, we find that in over 70% of panc-hayats a majority of respondents gave this an-swer. Similarly, Task Forces were also veryeffective: 80% of our respondents said that TaskForce projects were ‘‘almost always’’ or ‘‘al-ways’’ included in the final panchayat plan. Itis important to remember that the elected mem-bers of the panchayat are not legally bound tobudget the projects proposed by Task Forces.Some observers have reported that elected rep-resentatives saw Task Forces as underminingtheir authority (Chathukulam & John, 2002).The fact that panchayats did incorporate TaskForce projects suggests however that electedrepresentatives feel accountable to the pro-cess. 23 More broadly—and this point is criticalto any argument for participatory institu-tions—this demonstrates that representativeand participatory democracy can co-exist.

The findings in this section show that theparticipatory institutions introduced by thecampaign were inclusionary, broadly represen-tative, and quite successful in processing popu-lar inputs through the decision-making chain.But newly established institutions, especiallyat the local level, are readily subverted. 24 Theliterature on decentralization has repeatedlyshown how susceptible local democratic institu-tions are to elite capture and how often theycan produce perverse results such as exclusion-ary practices or rampant rent seeking. 25 Thearchitects of the campaign were in fact candidin identifying the ‘‘politician-bureaucraticpatronage nexus’’ as the target of reforms andwere also acutely aware that in Kerala’s envi-ronment of highly competitive party politics,patronage constitutes a major developmentalproblem. 26 To what extent has the cam-paign been subject to such extra-institutionalinfluences?

The most obvious evidence of such influenceswould be distributional outcomes favoring lo-cal elites. Because of its history of land reformsand lower class mobilization, and because ofthe deep penetration of competitive party poli-tics into rural life, landed elites, or rural oli-garchs do not hold significant sway in Kerala(Heller, 1999). As such it is not surprising thatless than 4% of respondents felt that ‘‘powerfulprivate interests’’ exerted significant influence


Autho

r's

pers

onal

co

py

over project selection. They however also re-sponded that government officials had no influ-ence over project selection, a somewhat moresurprising finding given the historical powerof the development bureaucracy. 27

The extent to which patronage has been atwork is much harder to gauge. As we saw ear-lier, the campaign to date has largely benefitedthe poor, which in itself suggests a high degreeof insulation from elite interests. The survey didhowever yield some evidence of ward-basedpatronage (arguably the most visible form ofpatronage in a single member ward system).But as the results reported in Table 4 show,the pattern was hardly egregious, with only13% of respondents detecting clear ruling-partyfavoritism. 28

Historically, the manipulation of beneficiarylists for projects has been an important sourceof patronage. 29 The campaign introduced arange of procedures, including Gram Sabhasto discuss beneficiary criteria and prescribedperiods for public comment on published liststo minimize preferential selection. We foundthat almost all panchayats followed these pro-cedures. Our respondents moreover felt thatthis has reduced extra-institutional influence.

Thus 44% of the respondents identified GramSabhas as having the most influence over bene-ficiary selection, followed by panchayat mem-bers (17%) and Task Force members (14%).Even more pointedly, when asked ‘‘In practice,what do you think are the criteria by whichbeneficiaries are selected?’’ less than 1% re-sponded ‘‘caste or social status’’ and only6.6% pointed to political connections. Finally,when asked if corruption had increased or de-creased with the campaign, 74% said ‘‘de-creased’’ and 6% said ‘‘increased.’’ Even amajority of the heads of opposition partybranches point to a decrease.

We close this section on participation byaddressing its impact on governance. Participa-tion can improve the quality of governanceboth by providing better inputs (informationand ideas) and by holding politicians andbureaucrats more accountable. Accountabilitycan increase both as a result of a system ofchecks and balances (as in constitutional–legaltheories of democracy) and as a result of coop-eration between state and civil society actors (asemphasized in the literature on synergy andparticipatory publics). Both principles wereincorporated into the design of the campaign,

Table 4. Political patronage and the spatial allocation of plan funds

Respondent category Fraction of respondents (by category) who said spatial allocation offunds

Highly skewedtoward ruling-party

wards

Slightly favoredruling-party

wards

Equallydistributed

across wards

Based onprioritiesand needs

Panchayat member from rulingparty or coalition

4.0 26.3 36.8 32.9

Panchayat member from opposition 27.3 39.4 19.7 10.6Head of local branch of ruling party 4.4 27.9 35.3 27.9Head of local branch of main

opposition party40.3 32.3 19.4 6.5

Panchayat secretary 9.2 29.2 26.2 26.2Panchayat-level implementing officer 6.4 22.2 23.8 22.2Other line department official 12.9 27.1 21.4 17.1Civilian Task Force member 7.9 38.1 27.0 23.8Representative of caste association or

religious organization14.3 46.4 12.5 21.4


9.8 42.6 23.0 21.3

Representative of mass-baseddevelopment organization

10.3 41.2 26.5 16.2

Representative of women’s group 9.0 34.3 34.3 17.9Representative of SC or

ST organization14.3 39.7 19.1 23.8

Overall 12.9 34.1 25.4 20.8


Autho

r's

pers

onal

co

py

the first by creating new centers of authority tocounterbalance the traditional power of stateactors, the second in the co-production designof Task Forces and the planning process in gen-eral. It is not possible to disentangle the effectsof these two mechanisms, but the general im-pact appears to be clear.

When asked if the campaign had made‘‘elected representatives more responsive tothe needs and opinions of common people’’over 92% felt there had been an increase inresponsiveness. Not surprisingly, ruling-partyrespondents had a higher evaluation of theiraccountability and the head of the local opposi-tion branch a much lower evaluation (thougheven this group overwhelmingly found therewas an improvement). Most telling is that civilsociety respondents scored the increase inaccountability of representatives higher thanthe overall average. When asked the same ques-tion about government officials, only 75% feltthere had been an improvement. This confirmsthe wide spread assumption that governmentofficials have been more resistant than localpoliticians to the campaign.

Because of its successful land reforms and asocial structure that is far less unequal in casteand class terms than is true for most of ruralIndia, decentralization in Kerala was far lessvulnerable to elite capture. Nonetheless, localdevelopment projects in Kerala were histori-cally dominated by the department-basedbureaucracy and patronage politics. The cam-paign appears to have tackled these problemswith some success. Despite the fact that the re-sources and authority that were devolved weresignificant, our sample of local political and ci-vil society elites overwhelmingly judged thecampaign to have reduced existing levels of cor-ruption and to have increased transparency andaccountability of both representatives and offi-cials. 30 Of course, some of our respondentshave an ideological or even material stake inthe campaign. But the fact that these assess-

ments hold true across all our respondent cate-gories, including civil society actors, suggestthat the findings are robust. And overall, thereis little doubt that our key respondents believethe campaign has indeed improved the efficacyof local development. When asked to comparebefore and after the campaign, a vast majorityof respondents felt that development projectsunder the campaign were more appropriatethan before. This view moreover holds acrossthe respondents’ party affiliation (Table 5).

The increased responsiveness documentedhere must be balanced by a recognition thatthe challenges of institution-building remainsignificant. Though the campaign did by andlarge succeed in creating a nested structure ofparticipation in decision making, it appears tohave had less success in creating efficient mech-anisms of implementation. A number of criticshave pointed out that panchayats often had dif-ficulty in spending their money (Chathukulam& John, 2002; Kannan, 2000) and even thecampaign’s supporters acknowledge that thedegree of actual integration of projects was dis-appointing with plans tending to reflect ward orsectoral demands rather than a strategic vision(Thomas Isaac & Franke, 2002). Our surveydata also point to problems in the functioningof beneficiary committees. In most panchayats,beneficiary committees were elected to executeprojects and break the notorious nexus betweencontractors and corrupt officials and politi-cians. We have already seen that the constitu-tion of these committees was generally felt tohave been democratic. But our respondentswere much less sanguine about the efficacy ofthe committees. When asked whether benefi-ciary committees increased or slowed the paceof implementation, the responses were equallydivided. Similarly, the respondents were evenlysplit as to whether or not beneficiary commit-tees reduced the costs of implementation. Ourdata do not speak directly to what may havebeen the problem here, but some observers

Table 5. Appropriateness of projects implemented since the start of the campaign compared to earlier projects, by partyaffiliation

Party affiliation Fraction (%) of respondents who said

Earlier projects weremore appropriate

Not muchdifference

Projects since the start of thecampaign are more appropriate

LDF affiliation 1.2 2.5 96.3No party affiliation 3.7 9.8 86.5Non-LDF party affiliation 3.9 18.9 77.2


Autho

r's

pers

onal

co

py

have reported that many beneficiary commit-tees were in fact little more than fronts for con-tractors. Tellingly, for both questions anoverwhelming majority reported that it was‘‘hard to tell.’’ Bearing in mind that most ofour respondents have direct experience withthe campaign, we are reminded that assessingactual institutional efficiency presents signifi-cant informational and measurement problems.

8. CIVIL SOCIETY ANDDEMOCRATIZATION

The literature on associative democracy hasdeveloped rich theoretical arguments for therole that civil society can play in improvingthe depth and scope of democracy (Cohen &Joel, 1992; Mahajan, 1999). At the most funda-mental level, the argument boils down to creat-ing the spaces in which associational autonomycan flourish and can shape public choices. Suchspaces must be both differentiated from statecontrol and from primary social structures.And they must also be linked to public author-ity. This is indeed the double meaning of whatAvritzer (2002) has called ‘‘participatory pub-lics.’’ We have already examined the extent towhich inputs that emerge from civil societyare translated into policy outputs. In this finalsection, we turn to the question of associationalautonomy. To what extent has the campaignincreased the space within which previouslymarginalized or excluded groups can meaning-fully partake in public life?

One of the more intriguing claims made in fa-vor of participatory governance is that it haspositive spillover effects for civic engagement.Because more participatory structures make iteasier for civil society to have an impact onpublic affairs, civil society actors face fewer

costs and have greater incentives to organize.This crowding-in effect has for example beendocumented in the case of participatory bud-geting in Porto Alegre (Brazil), which has wit-nessed a significant increase in associationalactivity since the inception of participatorygovernance (Baiocchi, 2003). In order to assesswhat impact the campaign has had on civilsociety, we asked our respondents to evaluatethe level of activity of different civil society sec-tors before and after the campaign. Their re-sponses are summarized in Table 6.

It is clear from these responses that there is awidespread perception that associational lifehas increased during the period of the cam-paign. Of course, we have to be careful aboutattributing these trends to the campaign itselfsince it is not possible here to control for otherfactors at work. Yet when we look at these re-sults more carefully, the distribution of associa-tional activity appears to match the campaign’sgoal of expanding non-traditional vehicles ofparticipation. Thus, while the civil society orga-nizations that have historically played the mostactive role in Kerala—labor unions, politicalparties, and political party-affiliated mass orga-nizations—have remained active, the highestlevels of activity come from religious organiza-tions and women’s organizations, which tradi-tionally have not been as central to politicallife in Kerala as unions and parties. The in-crease in activity of women’s organizationswould appear to be directly tied to sustainedefforts by the SPB to encourage women’sinvolvement in the campaign. The campaign’spublicity and policies have specifically targetedwomen with special fund allocations, trainingprograms, and mobilizational efforts. TheKSSP and women’s groups were particularlyactive in forming small Neighborhood Groupsas a way for women to deliberate and organize

Table 6. Impact of the campaign on associational activity

Type of association Fraction of respondents who said levels of activity are

Muchlower

Somewhatlower

Aboutthe same

Somewhathigher

Muchhigher

Religious organizations 2.5 19.8 18.6 10.6 48.5Labor unions and farmers’ associations 2.3 20.6 31.3 12.5 33.3Mass-based development organizations 1.7 16.9 35.1 17.8 28.5Party-affiliated women’s/youth organizations 2.5 17.9 34.0 14.1 31.6Women’s organizations 1.1 8.0 13.9 29.1 48.0Secular cultural associations 0.6 8.6 26.0 24.0 40.8Political parties 2.9 16.1 31.5 14.6 34.9


Autho

r's

pers

onal

co

py

in preparation for Gram Sabhas. A number oflocal studies have explicitly tied the rapid rise inself-help groups to the matching funds madeavailable by the campaign (John & Chathuku-lam, 2002; Manjula, 2000; Seema & Mukherjee,2000). Registration data we collected frompanchayats show a threefold increase in thenumber of self-help groups during the cam-paign.

The increased activity of religious organiza-tions is more difficult to interpret. On the onehand it may be entirely exogenous, tied to orga-nizing efforts of the Sangh Parivar, the allianceof Hindu nationalist organizations that havemarkedly stepped up their activities in Keralain the past decade. On the other hand, whenone considers that Kerala has a high densityof NGOs tied to the Christian, Muslim, andHindu communities, and that these organiza-tions have a long track record of civic-mindedactivity, it is just as plausible that the campaignprovided new opportunities for these groups tomobilize.

An increase in associational activity does notnecessarily translate into an increase in actualinfluence. To assess the extent to which civilsociety has actually acquired greater influencein public affairs at the local level, we askedrespondents the extent to which ordinary citi-zens ‘‘were more likely to raise questions anddemand answers from elected representativesand government officials’’ with respect to anumber of basic governance functions. The re-sponses reported in Table 7 paint a clear pictureof increased voice. In fact, only a very smallfraction of respondents felt there had been noincrease in citizen voice. If we examine the re-sponses by respondent category, it becomesclear that with the exception of the oppositionbranch heads, all other categories see a signifi-

cant increase in voice. Whereas governmentofficials and elected politicians had differentviews about the impact of the campaign onaccountability (reported earlier) their opinionof increase in voice is almost identical andespecially true of panchayat secretaries andpanchayat members. Also noteworthy is thatrepresentatives of SC/ST organizations andwomen’s organizations also perceived a high in-crease in voice.

We also asked a series of questions abouthow the campaign had affected women andlower caste and minority communities’ abilityto engage the public arena, including separatequestions about voice (willingness to make de-mands) and empowerment. As shown in Table8a, respondents detected a marked increase inwomen’s voice. When we control for the genderof the respondent, we find that women areslightly more likely than men to detect an in-crease. The lower score for lower castes (Table8b) may simply be a result of their uneven dis-tribution across panchayats.

The question on empowerment was as fol-lows: ‘‘In your opinion, compared to beforethe campaign, did the campaign and its activi-ties help bring women [or SC/STs and MinorityCommunities] more into public arena and em-power them in raising development issues?’’The responses with respect to women are sum-marized in Table 9a. Note that the response‘‘Drastic changes’’ is a translation of ‘‘valiatho-thil’’ in Malayalam that is more literally trans-lated as ‘‘in a big measure.’’ Fully two thirds ofthe respondents felt that there had been ‘‘dras-tic change,’’ with the representative of women’sorganizations choosing this response 82.1% ofthe time. Only a handful of respondents feltthat there had been no change. The respondentsdid not feel that the campaign empowered SC/

Table 7. Compared to before the campaign are ordinary people in the panchayat more likely today to raise questions anddemand responses from elected representatives and government officials. . .

Fraction of respondents who said

Lesslikely

No change Morelikely

Much morelikely

. . .if basic services such as primary health care, education, ordrinking water are not adequately provided?

1.2 2.1 63.7 32.9

. . .if basic infrastructure such as roads or irrigation facilitieswere not adequately provided or maintained?

1.7 2.0 59.1 37.2

. . . if they felt that the list of beneficiaries selected for variousschemes was inappropriate?

2.5 5.1 65.1 27.3

. . . if they suspected corruption or misuse of public funds inthe implementation of projects?

2.7 8.7 64.3 24.2


Autho

r's

pers

onal

co

py

STs and Minorities as much as it did women,but nonetheless 46.4% found the changes tobe ‘‘drastic’’ (Table 9b).

9. CONCLUSION

In both scale and design, the decentralizationreforms that were introduced in Kerala in 1996represent an ambitious effort to build localdemocracy. The impact of the reforms carriesimportant lessons for our understanding ofdecentralization, and even more importantlyof democratic deepening. The research reportedhere aimed to provide broad and robust mea-

sures of the campaign’s effect by collecting datadirectly from a sample of 72 panchayats andinterviews with 858 key respondents.

Our most important finding can be simplystated: the campaign has created structures ofparticipatory governance where none existedbefore. The simple fact that local governmentsin Kerala now have functions and resourcesthey did not have in the recent past representsa significant transformation. Until the passageof the 73rd and 74th constitutional amend-ments, local government in India (with a fewnotable exceptions) was little more than anempty institutional shell serving primarily asan extension of the planning and bureaucratic

Table 8b. Are members of SC/ST or minority communities in the panchayat more likely today to voice their needs anddemand responses from elected representatives and government officials?

Respondent type Fraction of respondents (by type) who said

Much lesslikely

Somewhatless likely

No realchange

Somewhatmore likely

Much morelikely

All respondents 0.7 0.6 7.5 71.6 19.6Representatives of SC/ST groups 0.0 0.0 8.1 69.4 22.6

Table 9a. Has the campaign helped to bring women into the public arena and empower them to raise developmentalissues concerning women?


No real change Slight change Drastic change

All respondents 0.5 32.2 66.5Representatives of women’s groups 0.0 17.9 82.1Female respondents (all categories) 0.0 32.5 65.7

Table 9b. Has the campaign helped to bring members of the SC/ST and minority communities into the public arena andempower them to raise developmental issues more effectively?


No real change Slight change Drastic change

All respondents 3.8 48.1 46.4Representatives of SC/ST groups 6.4 46.0 46.0

Table 8a. Are women in the panchayat more likely today to voice their needs and demand responses from electedrepresentatives and government officials?


Much lesslikely

Somewhatless likely

No realchange

Somewhatmore likely

Much morelikely

All respondents 0.7 0.1 2.9 56.2 40.1Representatives of women’s groups 0.0 0.0 0.0 41.5 58.5Female respondents (all categories) 0.0 0.0 1.2 57.1 41.8


Autho

r's

pers

onal

co

py

powers of states. Whatever authority local gov-ernments had was generally monopolized bylocal elites. Because of its history of land reformand social movements, Kerala departed some-what from this pattern of local elite dominance,but panchayats have nonetheless historicallybeen very weak and developmentally ineffective.That the campaign has irreversibly changed theimportance of local government in Kerala is be-yond doubt. Not only have resources been de-volved—and we know this both from officialsources, from direct data collected from oursample panchayats and from the almost unani-mous opinions of our respondents—but newinstitutions have been built, new processes of lo-cal decision making have been created and newchannels of participation have been opened up.In sum, new loci of governance and new spacesof citizenship now mark Kerala’s political anddevelopment landscape. One of the most re-spected and skeptical commentators on thecampaign, Kannan, has maybe offered one ofthe most succinct assessments: the campaignhas not only created a ‘‘public platform for avigilant civil society,’’ but has also ensured an‘‘enabling environment for development’’ (Kan-nan & Pillai, 2004, p. 39).

If the evidence presented here documents theformation of new institutions of governanceand democracy, it also suggests that these newinstitutions and practices have had a significantimpact on the quality, efficacy, and inclusive-ness of development.

(1) Contrary to Crook and Manor’s assess-ment that decentralized planning has littlechance of success in India (1998, p. 49), all ofour evidence points to the fact that the participa-tory planning took place in our sample panchay-ats. Participation figures show widespreadinvolvement in Gram Sabhas and broad-basedinvolvement at higher stages of decision making.All of our findings point to the fact that civilsociety inputs were processed through the sys-tem and significantly impacted final budgetaryallocations. The composition of Task Forces inparticular points to the incorporation of civilsociety, and as such marks a break with top-down and bureaucratic modes of decisionmaking. There has clearly been a decline in par-ticipation over the years suggesting that thecampaign may no longer be in a ‘‘mobiliza-tional’’ mode, and the inclusiveness of participa-tion decreases for women (though not for SCs)as one moves up the decision-making chain.Nonetheless, more people from a wider spec-trum of society are now involved in making deci-

sions about local development than ever before.The fact that elected panchayat members and lo-cal officials have a particularly positive view ofthe campaign certainly reflects their positionalbiases, but also the fact that they now enjoyfar greater responsibility and autonomy.

(2) There is also a clear evidence of the con-stitution of participatory publics. To the extentthat Gram Sabhas and Task Forces actually doinfluence budgetary outcomes and that panc-hayats have adopted participatory inputs, notjust as general demands, but as actual projectsdesigned and proposed by democratically con-stituted bodies of politicians, officials, and citi-zens, suggests that local spheres of publicopinion making (Gram Sabhas) are not justconsultative, but are in fact linked to authorita-tive institutions. This is critical, because as inthe well documented case of popular budgetingin Porto Alegre (Avritzer, 2002; Baiocchi, 2003)it marks a passage from the constitution ofpublic spheres (the focus of Habermas’ theory)to the institutionalization of participatory pub-lics, that is ‘‘institutional formats capable ofaddressing at the institutional level the issuesmade contentious at the public level’’ (Avritzer,2002, p. 7).

(3) The campaigns’ participatory structureshave had a significant positive impact on devel-opment performance. This finding must bequalified because of the nature of our sample,local political elites. There are however a num-ber of factors that allow us to be fairly confi-dent that this is a robust finding. First, thepositive response of our sample was true acrossa wide range of questions, even as respondentsexpressed some doubt about certain aspects ofthe campaign. For example, while an over-whelming majority felt that the campaign pro-duced projects that were more appropriatethan in the past, many expressed skepticismabout the efficacy of beneficiary committees.Our respondents also clearly differentiated be-tween areas of development, giving panchayatshigh marks for roads, housing for the poor andchild services, but low marks in promoting eco-nomic development. Second, opinions variedacross respondent categories, but were gener-ally consistent. Thus, even the group that wasmostly likely to be critical—local oppositionpoliticians—and indeed consistently providedthe least positive judgments, nonetheless gavethe campaign high marks overall. Both mem-bers of and representatives of subordinategroups generally shared the opinions of the lar-ger sample.


Autho

r's

pers

onal

co

py

(4) The campaign had demonstrably positiveeffects on social inclusion. Our data on partici-pation leave no doubt that SCs and womenhave established a new presence in the localpublic sphere. These groups represented a highproportion of Gram Sabha participants, andwere active in Development Seminars and TaskForces. Also, a vast majority of our respon-dents—including women and SCs—felt thatdisadvantaged groups had benefited from thecampaign. This was true both in material termsand with respect to growing voice and empow-erment.

This brings us to a final question which is notdirectly addressed in this study, but calls forcomment. To what extent are the developmentsdocumented in this study sustainable? To whatextent has local democratic government beeninstitutionalized, and more specifically to whatextent has participatory planning translatedinto greater capacity for local governance? Thisstudy focused on the launching of decentraliza-tion in Kerala, a period of four years duringwhich the CPI(M)-led LDF was in power. A re-form of governance structures of this magni-tude requires significant political initiative andthe campaign was by all accounts made possi-ble by a fortuitous alignment of political andsocial forces, including a sustained mobiliza-tional effort. In their infancy, all institutions de-pend on a delicate equilibrium of politicalforces. The history of decentralization in India(as well as in Kerala) is a history of reversals. Achange of government has often seen the dis-mantling of decentralization initiatives.

As the declining levels of participation inGram Sabhas indicate, the campaign has movedbeyond its mobilizational phase and now de-pends more on the commitment of active citi-zens, local politicians, and local officials. Thereform stage of breaking with the past througha strategy of large-scale mobilization mustnow give way to the more incremental processof systems consolidation, and most notably ofbuilding local governance capacity, a stage thatpresents unique political and institutional chal-lenges of its own. One of our findings—thatbeneficiary committees have not been effec-tive—underscores the difficulty of institutional-izing local governance. But the likelihood ofroutinizing participatory democratic prac-tices—which we would argue is a necessary firststep in any process of democratic deepening—has been increased by two key experientiallegacies of the campaign. First, the proceduraland substantive achievements documented here

have significantly increased the legitimacy ofdemocratic local government and widened thepolitical space for local politicians and civil soci-ety. Citizens now expect more of local govern-ment, and such expectations are what sustainsdemocratic practices. Second, the campaignhas had a significant experiential impact on awide range of actors. It has created a new con-stituency of over 14,000 elected local representa-tives (39% of whom are women) and thousandsof local government officials who have a vestedinterest in the new institutions, and it has giventhousands of ordinary citizens a direct taste oflocal development planning.

The Congress-led UDF government thatcame to power in 2001 showed far less enthusi-asm for decentralization than its predecessor.Although the Congress government officiallyembraced the new institutions and did not alterany of the legislative architecture of the cam-paign (leading Chathukulam & John (2002) toargue that its position is in fact a continuationof the LDF’s policies) it did fully downshift thecampaign’s mobilizational mode. The masstraining programs, voluntary technical commit-tees and the movement-like activism of theState Planning Board all come to an end. Ker-ala’s worsening fiscal crisis has also slowed theflow (if not the proportion) of revenues down-wards. The overwhelming electoral victory ofthe LDF in local government elections in2005 and its return to power at the state levelin May 2006 may yet infuse the process of localgovernment democratization with a new politi-cal vitality. In any event, the future of localdemocratic government in Kerala remains crit-ically dependent on shifting political configura-tions. These political vicissitudes (which havebeen the historical norm in Kerala) notwith-standing, the campaign has produced valuablelessons and experiences, and has quite possiblytriggered and amplified societal developmentsthat are conducive to democratic deepening.The production of local development plans,the discussion and the design of thousands ofprojects, and the resulting innovation have in-jected new ideas, new energy, and new dis-courses into local development. Gram Sabhas,and maybe most significantly Task Forces,have drawn thousands of citizens directly intothe process of problem-solving and even con-crete decision making, opening up new spacesfor the deliberative logic of civil society. If pol-itics are as much about practices as they areabout institutions, then these experiences willsurely have a lasting impact.


Autho

r's

pers

onal

co

py

NOTES

1. The full results of the research were first presented ina report submitted to the Ford Foundation (Chaudhuri,Harilal, & Heller, 2004).

2. A similar point can also be made for Brazil andSouth Africa, where recent reforms have created a widerange of local democratic opportunities.

3. See in particular Mahajan (1996, 1999) and Jayal(2001).

4. For an overview, see Chaudhuri (2006). John andChathukulam argue that before the constitutionalamendments Kerala ‘‘was behind most states in Indiain the implementation of Panchayat Raj’’ (2002).

5. Exceptions of course abound. Decentralization inWest Bengal has provided for significant local decisionmaking, though not integrated planning as such. Theimpact of decentralization in 1980s in Karnataka hasbeen well documented by Crook and Manor (1998),though many of the reforms were subsequently rolledback. More recently, Madhya Pradesh has made impor-tant strides, both by devolving new powers to GramSabhas and through the rise of a new educated middleclass of intermediaries that have effectively broken thetraditional brokering monopoly of local elites (Krishna,2002). In the aggregate, the insignificance of localgovernment in India is readily summarized: the Reportof the Eleventh Finance Commission found that during1995–98 the average annual per capita expenditure at theGram Panchayat level in India was a paltry Rs. 69 percapita (Chaudhuri, 2006, p. 186).

6. As cited by V.K. Ramachandran at the InternationalConference on Democratic Decentralisation, Trivan-drum, May 27, 2000.

7. For the most comprehensive overview of the Cam-paign’s origins and design, see Thomas Isaac and Franke(2002). Most of the campaign’s official documents areunusual in the candor and directness with which theyaddress existing problems of development and democ-racy. The two volumes of the Committee on Decentral-

ization of Powers report is not only a testament to thecare and detail with which legal and regulatory infra-structure of the campaign was developed, but also standsas a first rate analysis and indictment of top-downbureaucratic governance.

8. These figures are provided by the SPB. Mediareports and the authors’ own experience leave littledoubt as to the vast scale of the initiative.

9. With its 50,000 strong membership recruited pre-dominantly from the white-collar professions of civilservants and school teachers, the KSSP has an organizedpresence in a majority of villages in Kerala and is by farthe most active and influential non-party affiliated,secular organization in the state.

10. These included comprehensive participatory plan-ning exercises in selected panchayats, experiments inwater shed management, the development of newparticipatory techniques of research and planning,including local resources mapping, and a range of localenvironmental projects. Many of these have beendocumented by the Centre for Development Studies’Kerala Research Programme for Local Level Develop-ment. See http://krpcds.org/.

11. The total is lower than the full sample because insome panchayats some of the categories were missing(e.g., a caste association representative), some respon-dents could not be located and some respondents haddual identities (e.g., panchayat president and unionleader).

12. In a study of six states, Jha found that in 1999–2000, the average size of the Gram Panchayat budget inKerala was Rs. 8,900,000. The next highest of any statewas Rs. 200,000 (2002, p. 2614).

13. Beasely, Pande, and Rao (2004) using householdsurvey data from the four south Indian states ofKarnataka, Andhra Pradesh, Tamil Nadu, and Keralafind that individuals were three times more likely toattend Gram Sabhas in Kerala.

14. Fung and Wright (2003) address this problemdirectly in developing their model of ‘‘empoweredparticipatory governance.’’

15. In the case of Kerala, reliable data on pre-campaign local expenditures are available only at thedistrict level.

16. One particularly interesting result is how govern-ment officials evaluated improvement in areas thatrequire a high degree of technical capacity. Much ofthe criticism of the campaign concerned the lack oftechnical know-how at the local level and the campaign’sreliance on voluntary technical experts. In two suchareas—roads and irrigation—over 80% of line depart-ment officials felt that things had actually improved.Since both these sectors were previously to a largeextent controlled by the Public Works Department, a


Autho

r's

pers

onal

co

py

department notorious throughout India for high (andorganized) leakage, this represents an improvement froma possibly very low baseline.

17. Anganwadis in Kerala provide not only childcare,but also after school services. Since participation ofwomen in the workforce increases as one moves downthe occupational and caste ladder, childcare is a subsidyfor the working poor. Many economists have arguedthat local roads have a poverty-alleviation impact. Thispoint is open to debate, but in Kerala, where the ruralroad infrastructure is already highly developed, it hasbeen argued that new roads primarily benefit marginal-ized communities.

18. Alsop, Krishna, and Sjoblom (2000) present datafrom Madhya Pradesh and Rajasthan that suggest thatcaste and wealth are not as decisive in shaping partic-ipation patterns as generally assumed. Chaudhuri andHeller (2003) argue that categorical inequalities have acertain plastic quality, lending them to surprisingly rapidtransformation under the right social and institutionalcircumstances.

19. Note that this is attendance at only one of three orfour annual Gram Sabhas meaning that the percentageof individuals who attended a single Gram Sabha in agiven year is even higher. Comparisons are by definitionproblematic, but this represents the highest level ofparticipation in any Indian state (World Bank, 2000)and compares favorably with the city of Porto Alegreand the province of Rio Grande du Sul in Brazil, themost celebrated and carefully documented case of directparticipation in budgeting (Baiocchi, 2003).

20. Scheduled caste is the official term for ‘‘untouch-ables’’ and Scheduled tribe for designated tribal groups.The preferred terminology for these groups is nowrespectively Dalits and Adivasis, but we use SC/STbecause this is how official data are reported.

21. Bardhan and Mookherjee report that while SCsand Muslims are evenly represented in West Bengal,women’s attendance rate was only 9% (2004, p. 10).

22. The specific question was: ‘‘Have the felt needs ofthe community, as expressed in the Gram Sabhas, beenreflected in the plans submitted by the panchayat sincethe beginning of the campaign?’’

23. Similarly, in Porto Alegre the ‘‘popular budget’’that emerges from neighborhood assemblies has no legalstanding. Yet for 13 consecutive years it has beenadopted by the city council.

24. Inverting James Scott’s weapons of the weak,Scheper-Hughes has argued that footdragging, lies andfalse compliance are the weapons with which theprivileged can subvert demands from below (1992, p.514).

25. In their detailed study of decentralization in Kar-nataka, Crook and Manor report that local politicianswere quite open in explaining how they prevented SCsfrom exerting influence and even obtaining fundsearmarked for SCs (1998, p. 39).

26. The Committee on the Decentralisation of Powers—which played a key role in developing the campaign’sinstitutional structure—reports that an informal opinionsurvey among panchayat presidents before the campaignfound that 40% of funds spent on road construction (anexpenditure preferred for its skimming returns) ‘‘do notgo into the work for various reasons’’ (GOK, 1997, p.46).

27. If government officials have limited sway overproject selection, they clearly can still impact implemen-tation. John and Chathukulam report that governmentengineers still exert significant power in beneficiarycommittees charged with implementation (2002, p.1942).

28. In a comparative study of decentralization in foursouth Indian states, Beasely et al. (2004) also found thatthere was a lower incidence of ward-based targeting inKerala than in other states.

29. The Committee on Decentralisation of Powers waspointed in its condemnation of the widespread manip-ulation of beneficiary committees noting that it was ‘‘aninsult to the literate and politically aware public of theState’’ (GOK, 1997, p. 46).

30. Harilal and George (2000) have also shown that thecampaign’s required sectoral allocations maximize inclu-siveness (and have a pro-poor bias) and minimize therisk of patronage.

REFERENCES

Agarwal, B. (2001). Participatory exclusions, communityforestry, and gender: an analysis for South Asia and

a conceptual framework. World Development, 29(10),1623–1648.


Autho

r's

pers

onal

co

py

Alsop, R., Krishna, A., & Sjoblom, D. (2000). Are grampanchayats inclusive? Report of a study conducted inRajasthan and Madhya Pradesh. Background Paperno. 3. World Bank.

Avritzer, L. (2002). Democracy and the public space inLatin America. Princeton: Princeton University Press.

Baiocchi, G. (2003). Participation, activism and politics:the porto alegre experiment. In A. Fung, & E. O.Wright (Eds.), Deepening democracy. London: Verso.

Bardhan, P. (2002). Decentralization of governance anddevelopment. Journal of Economic Perspectives,16(4), 185–206.

Bardhan, P., & Mookherjee, D. (2004). Decentralizationin West Bengal: origins, functioning and impact.Paper presented at Conference on Democracy andhuman development: A global inquiry, Boston Uni-versity, November 12.

Beasely, T., Pande, R., & Rao, V. (2004). Panchayatsand resource allocation in south India. World Bank.

Chathukulam, J., & John, M. S. (2002). Five years ofparticipatory planning in Kerala: rhetoric andreality. Economic and Political Weekly, 37(49),4917–4926.

Chatterjee, P. (2001). Democracy and the violence of thestate: a political negotiation of death. Inter-AsianCultural Studies, 2(1), 7–21.

Chaudhuri, S. (2006). The 1994 Panchayati Raj Act andthe attempt to revitalize rural local government inIndia: what difference does a constitutional amend-ment make?. In P. Bardhan, & D. Mookherjee(Eds.), Decentralization and local governance indeveloping countries: A comparative perspective.Cambridge: MIT Press.

Chaudhuri, S., Harilal, K. N., & Heller, P. (2004). Doesdecentralization make a difference? A study of thepeople’s campaign for decentralized planning in theIndian state of Kerala. Thiruvananthapuram: Centrefor Development Studies.

Chaudhuri, S., & Heller, P. (2003). The plasticity ofparticipation: evidence from a participatory gover-nance experiment. Columbia University ISERPWorking Paper, New York.

Chhibber, P. K. (1999). Democracy without associations:Transformation of the party system and social cleav-ages in India. Ann Arbor: University of MichiganPress.

Chhibber, P. K. (2003). Why some women are politicallyactive: the household, public space, and politicalparticipation in India. International Journal of Com-parative Sociology, 43, 409–429.

Cohen, J. R., & Joel (1992). Secondary associations anddemocratic governance. Politics and Society, 20(4),393–472.

Crook, R. C., & Manor, J. (1998). Democracy anddecentralisation in South Asia and West Africa:Participation, accountability and performance. Cam-bridge: Cambridge University Press.

Dreze, J., & Sen, A. K. (1995). India: Economicdevelopment and social opportunity. Delhi, NewYork: Oxford University Press.

Evans, P. (1996). Government action, social capital anddevelopment: Reviewing the evidence on synergy.World Development, 24(6), 1119–1132.

Evans, P., & Rauch, J. E. (1999). Bureaucracy andgrowth: a cross-national analysis of the effects of‘‘Weberian’’ state structures on economic growth.American Journal of Sociology, 64, 765–784.

Fung, A., & Wright, E. O. (2003). Deepening democracy:Institutional innovations in empowered participatorygovernance. London: Verso.

GOK. (1997). Committee on Decentralisation of Pow-ers: Final Report: V. 1, part A and B, December 23.

GOK, Kerala State Planning Board (2001). EconomicReview, 2001.

Harilal, K. N., & George, M. S. 2000. Prioritisation inlocal level planning: the Kerala experience. Paperpresented at the International conference on demo-cratic decentralisation, May 23–27, Thiruvanant-hapuram.

Heller, P. (1999). The labor of development: Workers andthe transformation of capitalism in Kerala, India.Ithaca, NY: Cornell University Press.

Heller, P. (2005). Reinventing public power in the age ofglobalization: the transformation of movement pol-itics in Kerala. In R. Ray, & M. Katzenstein (Eds.),Social movements in India: Poverty, power andpolitics. New York: Rowman and Littlefield Pub-lishers Inc.

Huber, E., Rueschemeyer, D., & Stephens, J. D. (1997).The paradoxes of contemporary democracy: formal,participatory, and social democracy. ComparativePolitics, 29(3), 323–342.

Jalal, A. (1995). Democracy and authoritarianism inSouth Asia: A comparative and historical perspective.Cambridge, New York: Cambridge University Press.

Jayal, N. G. (2001). Democracy and the state: Welfare,secularism and development in contemporary India.New Delhi: Oxford.

Jha, S. (2002). Strengthening local governments: ruralfiscal decentralisation in India. Economic and Polit-ical Weekly, (June 29), 2611–2623.

John, M. S., & Chathukulam, J. (2002). Building socialcapital through state initiative: participatory plan-ning in Kerala. Economic and Political Weekly, (May18), 1939–1948.

Kannan, K. P. (2000). People’s planning, Kerala’sdilemma. Seminar, 485, 92–97.

Kannan, K. P., & Pillai, V. N. (2004). Development asfreedom: an interpretation of the ‘‘Kerala model’’No. Working Paper 361. Centre for DevelopmentStudies, Thiruvananthapuram.

Krishna, A. (2002). Active social capital: Tracing theroots of development and democracy. New York:Columbia University Press.

Linz, J. J., & Stepan, A. C. (1996). Problems ofdemocratic transition and consolidation: SouthernEurope, South America, and post-communist Europe.Baltimore: Johns Hopkins University Press.

Mahajan, G. (1996). Gender equality and communityrights: paradoxes of liberal democracy in India.Journal of Gender Studies, 5(2), 169–176.

Mahajan, G. (1999). Civil society and its avtars: whathappened to freedom and democracy. Economic andPolitical Weekly, 34, 1188–1196.

Manjula, B. (2000). Voices from the spiral of silence: acase study of Samatha Self Help Groups of Ulloor.


Autho

r's

pers

onal

co

py

Paper presented at the International conference ondemocratic decentralisation, Thiruvananthapuram,May 23–27.

Migdal, J. S., Kohli, A., & Shue, V. (1994). State powerand social forces: Domination and transformation inthe third world. Cambridge [England], New York:Cambridge University Press.

Mohan, G., & Stokke, K. (2000). Participatory devel-opment and empowerment: the dangers of localism.Third World Quarterly, 21(2), 247–268.

O’Donnell, G. (1993). On the state, democratization andsome conceptual problems: a Latin American viewwith glances at some postcommunist countries.World Development, 21(8), 1355–1359.

Parameswaran, M. P. (2000). Role of the Kerala SastraSahitya Parishad in the Movement for DemocraticDecentralisation. Paper presented at the Interna-tional conference on democratic decentralisation,Thiruvananthapuram, May 23–27.

Parameswaran, M. P. (2001). From voters to actors:people’s planning campaign and participatorydemocracy in Kerala. Kerala Sastra Sahitya Pari-shad, Trichur.

Ravallion, M., & Datt, G. (1999). When is growth pro-poor? Evidence from the diverse experiences ofIndia’s states. World Bank Development ResearchGroup Poverty and Human Resources and SouthAsia Region Poverty Reduction and EconomicManagement Sector Unit, Washington, DC.

Scheper-Hughes, N. (1992). Death without weeping: Theviolence of everyday life in brazi. Berkeley: Universityof California Press.

Seema, T. N., & Mukherjee, V. (2000). Gender gover-nance and citizenship in decentralised planning.Paper presented at the International conference ondemocratic decentralisation, Thiruvananthapuram,May 23–27.

Thomas Isaac, T. M., & Franke, R. W. (2002). Localdemocracy and development: The Kerala people’scampaign for decentralized planning. Lanham, Md.:Rowman & Littlefield.

Tornquist, O. (1997). Making democratisation work:from civil society and social capital to politicalinclusion and politicisation. Theoretical Reflectionson Concrete Cases in Indonesia, Kerala, and thePhilippines. Research Programme on Popular Move-ments, Development and Democratization, Univer-sity of Oslo.

Verba, S., Schlozman, K. L., & Brady, H. E. (1995).Voice and equality: Civic voluntarism in Americanpolitics. Cambridge, MA: Harvard University Press.

Veron, R. (2001). The ‘‘new’’ Kerala model: lessons forsustainable development. World Development, 29(4),601–617.

World Bank (2000). Overview of rural decentralizationin India: approaches to rural decentralization inseven states. Volume I.

Yadev, Y. (2000). Understanding the second democraticupsurge. In F. Frankel, Z. Hasan, R. Bhargava, & B.Arora (Eds.), Transforming India. New Delhi:Oxford.


Date post:	01-Dec-2020
Category:	Documents
Upload:	others
View:	2 times
Download:	0 times

Author's personal copy - Patrick Heller...73rd and 74th constitutional amendment, local governments...

Documents