Agarwal Participatory Paper

8/2/2019 Agarwal Participatory Paper

1/26

Participatory Exclusions, Community Forestry, and

Gender: An Analysis for South Asia and a Conceptual

Framework

BINA AGARWAL *Institute of Economic Growth, Delhi, India

Summary. The idea of people's participation has long been part of development thinking. Buttoday the management of local natural resources by village communities is widely accepted as aninstitutional imperative. It is therefore essential to examine how these institutions perform,

especially from the perspective of the more disadvantaged. Based on extensive eldwork amongcommunity forestry groups in India and Nepal, and existing case studies, this paper demonstrateshow seemingly participatory institutions can exclude signicant sections, such as women. Itprovides a typology of participation, spells out the gender equity and eciency implications of suchexclusions, and analyzes what underlies them. It also outlines a conceptual framework to helpanalyze the process of gender exclusion and how it might be alleviated. 2001 Published byElsevier Science Ltd.

Key words participation, community forestry institutions, gender, bargaining framework, South

Asia

1. INTRODUCTION

Central to the idea of people's participationin development, however diverse and contestedits denition and scope, is inclusivenesstheinclusion in decision making of those mostaected by the proposed intervention. There isalso an emerging consensus that eectiveparticipation requires people's involvement not

just as individuals but as a collectivity, such asa village community. We are thus seeing anincreasing emphasis on community participa-tion through group formation in all forms of

development interventions. Indeed, in thecontext of natural resource management (be itforests or water), devolving greater power tovillage communities is now widely accepted asan institutional imperative by governments,international agencies, and nongovernmentalorganizations (NGOs). Moreover, ruralcommunity forestry groups represent one of themost widespread and rapidly expandingattempts at participative development.

Ostensibly set up to operate on principles ofcooperation, such groups are meant to involve

and benet all sections of the community. Yeteectively they can exclude signicant sections,such as women. These ``participatory exclu-

sions'' (that is exclusions within seemingly

participatory institutions), constitute morethan a time-lag eect. Rather they stem fromsystemic factors and can, in turn, unfavorablyaect both equity and institutional eciency.This paper analyzes what underlies suchexclusionary outcomes and how outcomescould be improved. In also outlines a concep-tual framework that can illuminate the natureof gendered exclusion and the potential forchange. Where relevant, the interplay of class/caste with gender, in dening outcomes fordierent categories of women, is also indicated.

It is argued here that participation is deter-mined especially by rules, norms, and percep-tions, in addition to the endowments andattributes of those aected. These factors candisadvantage women, both separately andinteractively. Women's ability to alter them will

World Development Vol. 29, No. 10, pp. 16231648, 2001 2001 Published by Elsevier Science Ltd.

Printed in Great Britain0305-750X/01/$ - see front matter

PII: S0305-750X(01)00066-3www.elsevier.com/locate/worlddev

* I thank the anonymous referee and Janet Seiz for

comments on an earlier draft. I also thank the Ford

Foundation (Delhi) and ICIMOD (Kathmandu) for

their support towards my eldwork; Anupam Bhatia for

facilitating my Nepal trip in numerous ways; and the

many Indian NGOs that generously facilitated my trips

to their eld sites. Final revision accepted: 17 May 2001.

1623


2/26

depend on their bargaining power vis-a-vis theState, the community, and the family. Thelikely determinants of women's bargainingpower in these three arenas are spelled out.While the context here is community forestry,

the conceptual framework would also haverelevance in other contexts.

The paper is based largely on eldvisits andinterviews I undertook during 199899 to 87community forestry sites across ve states ofIndia (Gujarat, Karnataka, Madhya Pradesh,Orissa, and the Uttar Pradesh hills) and twodistricts (Kaski and Dang) of Nepal. 1 Infor-mation was obtained mostly through unstruc-tured interviews with village women and men,at times in separate groups, at other times

jointly, in addition to individual interviews with

key informants, especially oce bearers in theexecutive committees of the communityforestry groups (CFGs). This information issupplemented by existing case studies and someearlier visits in a few states.

In the sections which follow, the paperprovides a typology of participation; gives abrief background to South Asia's CFGs;traces the gender gap in CFG participationand the associated adverse implications forequity and eciency; outlines what constrainswomen's participation; and presents the

bargaining framework as a way of analyzinghow these constraints might be contested andalleviated.

2. PARTICIPATION: A TYPOLOGY

While the idea of people's participation haslong been part of development thinking, todayit has become almost mandatory in planningdevelopment projects. Views diverge, however,on how participation is dened, whom it is

expected to involve, what it is expected toachieve, and how it is to be brought about. 2 Atits narrowest, participation in a group is

dened in terms of nominal membership (e.g.,Chopra et al., 1990; Molinas, 1998), and at itsbroadest in terms of a dynamic interactiveprocess in which the disadvantaged have voiceand inuence in decision-making (e.g., Nara-

yan, 1995; White, 1996). In terms of objectives,at its narrowest participation is judged almostentirely by its potential eciency eects, and atits broadest by its ability to enhance equity,eciency, empowerment and environmentalsustainability (Upho, 1991).

While the earlier literature on participationlargely neglected how it was to be broughtabout, recent studies on collective action,especially in the context of natural resourcemanagement, are preoccupied with what makesfor successful collective functioning and coop-

eration between people.3

But the objective ofsuccessful cooperation remains largely narrow,viz. institutional eciency. Equality enters theequation less as a desired outcome thaninstrumentally: whether, for instance, socio-economic equality enhances prospects of coop-eration. 4 The implications of inequitableoutcomes for disadvantaged sections are littleexamined.

Common to both streams of work, butespecially characteristic of recent writings oncollective action by economists, is a striking

neglect of a gender perspective on who partic-ipates, what eects this has, and what factorsconstrain participation. 5 This paper oers sucha gendered analysis, wherein eective partici-pation is seen as important both in itself, as ameasure of citizenship and a means ofempowerment, and for its potential eects onequity, eciency and sustainability.

Participation can have a range of levels.Drawing on the typologies of White (1996) andPretty (1995), but also departing from them innotable ways, a typology of these levels

(broadly dened) is outlined in Table 1.Achieving eective participation would

involve a shift from the lower to the higher

Table 1. Typology of participation

Form/Level of participation Characteristic features

Nominal participation Membership in the groupPassive participation Being informed of decisions ex post facto; or attending meetings and listening in

on decision-making, without speaking upConsultative participation Being asked an opinion in specic matters without guarantee of inuencing

decisionsActivity-specic participation Being asked to (or volunteering to) undertake specic tasks

Active participation Expressing opinions, whether or not solicited, or taking initiatives of other sortsInteractive (empowering)

participationHaving voice and inuence in the group's decisions

WORLD DEVELOPMENT1624


3/26

levels, with levels being dened here not by howa group is initiated but by the extent of people'sactiveness. This is unlike some earlier typolo-gies, where, for instance, self-initiated activity isseen as the highest level. In fact not all self-

initiation need signify participatory success ofthe overall program. A set of disadvantagedpersons may, for instance, opt out of the maingroup where they have no voice and set up theirown group. But this exit option may notsubstitute adequately for their lack of voicewithin the main group, if the latter controlsmost of the resource. Moreover, participationis not the panacea many assume. There arelimits to what participation alone (even ifinteractive) can achieve in terms of equity andeciency, given pre-existing socioeconomic

inequalities and relations of power. Considerhow participatory exclusion plays out in thecontext of South Asia's CFGs.

3. BACKGROUND: COMMUNITYFORESTRY GROUPS

Forests and village commons have beenimportant sources of supplementary livelihoodsand basic necessities for rural households inmany parts of the world (including in Europe in

the late 18th and early 19th centuries:Humphries, 1990). In South Asia thesecommon pool resources have provided re-wood, fodder, small timber, and various non-timber products. Especially for the poor andwomen who own little private land, they havecontributed critically to survival. In India'ssemi-arid regions in the 1980s, the landless andlandpoor procured over 90% of their rewoodand satised 6989% of their grazing needsfrom the commons (Jodha, 1986). In thatperiod, rewood alone provided 6567% of

total domestic energy in the hills and desertareas of India and over 90% in Nepal as awhole (Agarwal, 1987). Today, rewood is stillthe single most important source (and for manythe only source) of rural domestic energy inSouth Asia, and is still largely gathered, notbought. 6

People's ability to fulll such needs, however,has been falling over the decades with thedecline in communal resources, due both todegradation and to shifts in property rightsaway from community hands to State and

individual hands. In India and Nepal thisdecline reached crises proportions by the early1970s, compelling the State to take remedial

measures, especially in the form of socialforestry programs. Implemented in a top-down,nonparticipatory fashion, with a predominantfocus on commercial species such as eucalyptus,these programs largely failed both in restoring

forests and in meeting people's needs (Agarwal,1986). In contrast were examples of successfulforest management with community involve-ment, initiated by some forest ocials orvillagers themselves (Poenberger & McGean,1996). These contrasting experiences contrib-uted in important ways to an emergingconsensus that successful forest managementneeded the participation of local communities.In addition, the formation of CFGs in recentyears represents a small but notable reversal inthe earlier processes of Statization and priv-

atization, toward establishing greater commu-nity control over forests and commons. Indeedthere is now a mushrooming of CFGs, 7 somestate-initiated, others self-initiated by commu-nities, and yet others catalyzed interactively byNGOs, villagers, and state ocials.

In India, these CFGs include: (a) groupsformed under the State-initiated Joint ForestManagement (JFM) program launched in 1990,in which villagers and the government share theresponsibility and benets of regeneratingdegraded local forests; (b) self-initiated groups,

started autonomously by a village council,youth club or village elder and concentratedmainly in the eastern states of Bihar and Orissa;and (c) groups with a mixed history, such as thevan panchayats (forest councils) in the UttarPradesh (UP) hills initiated by the British in the1930s. Some of them have survived or beenrevived by NGOs. JFM groups are the mostwidespread, both geographically and in termsof forest area. So far, 22 states have passedJFM resolutions which allow participatingvillagers access to most nontimber forest

products and to 2550% (varying by state) ofany mature timber harvested. Today, there arean estimated 36,000 JFM groups, covering 10.2million hectares (mha) or 13.3% of the 76.5mha administratively recorded as forest land(Bahuguna, 2000). 8 In addition, there wouldbe a few thousand groups of the other types.

Nepal's community forestry program, laun-ched in 1993, is largely State-initiated. Hereeven good forest is transferred to a set ofidentied users who form a forest user group(FUG) and who are entitled to all of the

benets. Today there are about 9100 FUGsinvolving one million households and covering0.66 mha or 11.4% (the target being 61%) of

PARTICIPATORY EXCLUSIONS AND COMMUNITY FORESTRY 1625


4/26

Nepal's 5.8 mha of forest land (Government ofNepal, 2000). In both India and Nepal, NGOscan act as catalysts or intermediaries in groupformation and functioning.

The criterion for establishing rights in the

commons is either membership (as in the State-initiated groups) or some other formal systemlaid down by a selected (often self-selected) setof villagers (as in the self-initiated groups).Unlike the older forms of communal resourcemanagement which typically recognized the userights of all village residents, these representmore formalized systems of inclusion andexclusion. The question then is: Are thesesystems inclusive and equitable in relation sayto women, especially the poor? Or are theycreating new property rights in communal land

which, like existing ones in private land, areelite and/or male centered?

4. PARTICIPATORY EXCLUSIONS

The State-initiated groups in both India andNepal broadly have a two-tier organizationalstructure: a general body (GB) which canpotentially draw members from the wholevillage and an executive committee (EC) ofsome 915 persons. Typically the GB meets

once or twice a year and the EC about once amonth. Both bodies, interactively, dene therules for forest use, the punishments for abuse,and the methods of protection (e.g., guards,patrol groups, etc.), benet distribution andconict resolution. Which set of persons has avoice in the GB and EC bears critically on howwell these organizations function, and whogains or loses from them.

Women's eective participation in CFGdecision-making would require that they notonly become members of the group, but also

attend and speak up at meetings, and can (atleast some of the time) ensure that decisions arein their favour. In other words, they wouldneed to move from being absent or just nominalmembers to interactive (empowered) partici-pants. Where are they placed currently?

(a) Nominal participation

With some exceptions, such as of all-womenCFGs discussed further below, most womenare not even nominal members: they constitute

less than 10% of most JFM general bodies; 9are usually absent in the self-initiatedgroups; 10 and are few or none in the van

panchayats. 11 Their presence in Nepal's FUGsis similarly sparse: in eastern Nepal, forinstance, Dahal (1994, p. 78) found that only3.5% of the recorded FUG members werewomen.

In India, the JFM membership criteria in theGB and EC vary by state. In 1993, of the 14states that had initiated JFM, six allowed GBmembership to only one person per household(Table 2). This was inevitably the male house-hold head. Today, this remains so for eight ofthe 22 JFM states on which there is informa-tion. In eight others (some due to rule amend-ments), both spouses, or one man and onewoman, can be members. But this still excludesother household adults. Moreover, where thewoman automatically becomes a member by

virtue of her husband being a member (as inWest Bengal), it is he who is seen as the primarymember. Only three JFM states allowmembership to all village adults. In the self-initiated CFGs the situation is worse, since theyreplicate the customary exclusion of womenfrom village decision-making bodies. In Nepal'sFUGs, again, the unit of membership is thehousehold, and in male-headed households theman's name alone is entered in the membershiplist (Seeley, 1996).

Women's representation in the ECs (barring

some exceptions) is also typically low. In 20JFM groups studied in West Bengal, 60% hadno women EC members, and only 8% of the180 EC members were women. Landless fami-lies, while present in most GBs, were barelyrepresented (Sarin, 1998). In many states, JFMresolutions today require the EC to includewomen, ranging from a minimum of two orthree to one-third (Table 3), but I found thatthe women so included were rarely chosen byother women as their representatives. Some-times male EC members even chose the women

in their absence and without consulting them.Such women are seldom active or eective. InNepal again, women's presence in most ECs isnominal. Often those who join are poorlyinformed of their FUG's activities; some areeven unaware that they are EC members. 12

(b) Passive participation

Without being members, women usually hearlittle of what transpires at GB or EC meetings,

that is, they are not even passive participants.Women across the dierent regions character-istically complain:



5/26

Table 3. JFM rules on women's representation in the EC

Rules: no. ofwomena 1993 N

14 1998 N

21 2000 N

22

Min 1 woman Punjab Punjab Punjab

Min 2 women Gujarat,b HimachalPradesh,c Jammu andKashmir, Karnataka,Maharashtra

Gujarat,b Himachal Pradesh,c

Jammu and Kashmir, Karnataka,Madhya Pradesh,d Maharashtra,Sikkim

Gujarat,b Himachal Pradesh,c

Jammu and Kashmir, Karnataka,Madhya Pradesh, Maharashtra,Sikkim

Min 3 women Andhra Pradesh,Orissa

Kerala, Orissa Kerala, Manipur Orissa,Rajasthand

Min 2 max 5women

Tamil Nadu Tamil Nadu

Min 2 max 5women

Bihar Bihar Bihar

Min 30% women Andhra Pradesh Andhra Pradesh

Min 1/3 women Haryana,d Uttar Pradesh Haryana, Madhya Pradeshd

Unspecied Haryana, MadhyaPradesh, Rajasthan,Tripura, West Bengal

Assam, Arunachal Pradesh,Nagaland, Rajasthan, TripuraWest Bengal

Assam, Arunachal Pradesh,Nagaland, Rajasthan, Tripura,Uttar Pradesh, West Bengal

Source: SPWD (1994, 1998) and recent state orders for Madhya Pradesh, Rajasthan and Sikkim.a Total specied EC members vary: typical number is 915 members.b Specication is for minimum two women in the ``working committee'' for managing the JFM micro-plan, buteectively it is interpreted as a minimum of two women in the EC.c Out of a total of 912 EC members, ve are village representatives, of whom 50% should be women. This works outto a minimum of 23 women in the whole EC.d These italicized states have moved toward more women-inclusive rules compared with the previous year(s).

Table 2. JFM rules for general body membership

Membershipconditionsa

1993 (N 14) 1998 (N 21) 2000 (N 22)

One personper household

Bihar, Jammu and Kashmir,Karnataka, Madhya Pradesh,

Maharashtra, Tripurac

Assam, Arunachal, Pradesh,b

Bihar, Jammu and Kashmir,

Maharashtra, Rajasthan,Sikkim, Tripurac

Assam, Arunachal Pradesh,b

Bihar, Jammu and Kashmir,

Maharashtra, Manipur,Sikkim, Tripurac

One male andone female perhousehold

Andhra Pradesh, HimachalPradesh, Orissa,West Bengala;d

Andhra Pradesh, HimachalPradesh, Madhya Pradesh,a

Orissa, Tamil Nadu, WestBengal,d Karnataka,a;dKerala

Andhra Pradesh, HimachalPradesh, Orissa, Rajasthan,a

Tamil Nadu, West Bengal,d

Karnataka,d Kerala

All villageadults/anyoneinterested

Gujarat Gujarat, Haryanaa Gujarat, Haryana, MadhyaPradesha

No clearrepresentationindicated

Punjab, Haryana, Rajasthan Punjab, Nagaland,c

Uttar PradeshPunjab, Nagaland,c

Uttar Pradesh

Source: SPWD (1994, 1998), and recent state orders for Madhya Pradesh, Rajasthan and Sikkim.a These italicized states have moved toward more women-inclusive rules compared with the previous year(s). In WestBengal's 1990 order only one person per household could be a member. This was modied in 1991 to include women.b Here the rule is one adult per family; and at least 30% of total registered members will need to be women.c In Tripura only families with at least one wage earner are eligible. In Nagaland only landowning households areeligible.d In West Bengal, if the husband is a member the wife automatically becomes a member; in Karnataka, if one spouseis a member the other automatically becomes a member.



6/26

Typically men don't tell their wives what happens inmeetings. Even if there is a dispute about something,they don't tell us; nor do they volunteer informationabout other matters (women to author, Khedipadavillage, Gujarat, 1999).

The men seldom inform us of discussions in meetings.When we ask them they say: ``why do you want toknow?'' If we were members we would be better in-formed (women to author, Jamai village, MadhyaPradesh, 1999).

Even of the women who are GB or ECmembers, only a small percentage usuallyattend the meetings. If they do attend theyrarely speak up, and if they speak their opin-ions carry little weight. Some characteristicresponses are given below:

What is the point of going to meetings. We would onlysit silently (women to author, Panasa Diha village,Orissa, 1998).

Men don't listen, except perhaps one or two. Men feelthey should be the spokespersons (woman to author,Garbe Kuna forest, Kaski district, Nepal, 1998).

I attend van panchayat meetings, but I only sign, Idon't say much. Or I say I agree (woman van pancha-

yat member to author, Sallarautela village, UP hills,1998).

Even if women attend meetings, they cannot voicetheir opinions; they cannot speak against the opinionsof their seniors. When the men have nished speakingthat is the end of the meeting (Satibama, Nepal, citedin Hobley, 1990, p. 309).

Having a voice in the EC is important since thisis the site for discussions and decisions on manycentral aspects of CFG functioning. As mattersstand, women are not party to most decisions.Even those who are members remain passiveparticipants, far from the active or interactivelevel in my typology.

(c) Consultative participation

Equally, male CFG members and forestocials seldom consult women when framingforest use rules, or preparing micro-plans forforest development. Some women hear aboutthe plans through their husbands, others not atall (Guhathakurta & Bhatia, 1992). As themain fuel and fodder collectors, women areoften found to know more than men about theattributes of trees with such use value (Pandey,

1990), and to be better informed about the localenvironment where they gather and collect,while men are found more knowledgeable

about the species they use or those found indistant areas (Chen, 1993; Gaul, 1994). Failingto consult women means that their existingknowledge of diverse species does not enrichforest regeneration programs. At the same

time, women have less chance of acquiring newknowledge, such as about new silviculturepractices, since they are rarely part of the CFGteams that receive such training.

A related issue is gender dierences in pref-erences, say regarding when grass should be cutor which trees should be planted. I found thatin the rare cases that women were consulted,they often came up with more suitable alter-natives. A case in point is Simal village (UPhills) where the men had xed a date for grasscutting, but the women, when consulted, said:

``This period is not right. We have work nowand also have some dry fodder left. We shouldbe cutting when our store of fodder is deple-ted.'' So the committee rescheduled the forestopening. 13 But such consultations are rare.Nor is it common to incorporate women'spreferences in tree species, which typically tendtoward trees with domestic use value (as forfuel and fodder) while men more typically optfor cash yielding varieties (Brara, 1987; Hobley,1996).

Thus, despite their central stake as users and

their familiarity with forests, women have littlesay in CFG decisions. An analysis of ve JFMvillages in Gujarat revealed that all majordecisions on forest protection, use, distributionof wood and grass, and future planning, weretaken by men. The only joint decisions withwomen were about tree nurseries (Joshi, 1998).

(d) Activity-specic participation

Despite their absence from decision-making,women get drawn into specic activities, espe-

cially protection. In formal terms, protection ofthe bounded area is usually done by employinga guard, with CFG members contributing thewage, or by forming a patrol group fromamong the member households. A male guardor an all-male patrol is typical: these twomethods respectively characterized 45% and18% of the 87 sites I visited (Table 4). Femaleguards were rare, and only a small percentageof patrols had both sexes, or women alone.Mixed patrols were more common in centraland western India, and all-women patrols in the

UP hills and Nepal. Occasionally, there areshifts from all-men to all-women patrols, andvice versa.



7/26

More commonly, women protect informally,sometimes forming patrols parallel to men'sbecause they feel men's patrolling is ineective.Women's vigilance improves protection inimportant ways, especially because they are

better able to dissuade other women frombreaking the rules as well as catch femaleintruders. In re ghting, likewise, women joinin, and in several cases their alertness alonesaved the forest. None of this, however, addsup to interactive participation.

(e) Active and interactive participation

Within the stark scenario of low femaleparticipation in formal groups, there are somecontrasting examples of all-women CFGs and

mixed CFGs with a high female presence. 14The former are found especially in the UP hillsand parts of Nepal with high male outmigra-

tion (with a scattering in other regions). Buttheir numbers remain small. There are noconsolidated gures for India, but in Nepal all-women FUGs constitute only 3.8% of allFUGs, controlling 11.2% of all FUG-con-

trolled land (Government of Nepal, 2000).They typically receive small plots: some 50%control less than 10 ha each. Often this isbarren land needing tree planting, while themixed (male-dominant) FUGs usually receivelarger tracts and natural forest. Mixed groupswith a high female presence (say with 3050%women in the GB or EC) are similarlyuncommon, being found only in selectedpockets in both India and Nepal. 15

Usually CFGs with a large percentage ofwomen have been catalyzed by a local NGO,

forest ocial, or international donor. In rarecases women themselves have initiated formalCFGs either where there were none, 16 or

Table 4. Methods of forest protection: India and Nepala

Method No. of CFGs

Guardb 31 (3 women)Village employed (25); Forest dept employed (6)

Male patrolb 15

Female patrolb

7Joint male and female patrol 7Separate male and female patrols in same village 2Men or women plus guard 2Informal only 8

Changing methods 15 Male to female patrolling (2) Male to female patrolling to guard (1) Male patrolling to guard (6) Female to male patrolling to guard (1) Female patrolling to guard (3) Guard to female patrolling (1) Joint male and female patrolling to male patrolling (1)

Total CFGs visited 87

Most common protection method used

State CFGsvisited

Most common method

Gujarat 19 Village patrol (usually men, sometimes both sexes)Karnataka 13 Male guard employed by forest department for three years;

then informal protection or guard employed by villagersMadhya Pradesh 12 Village patrol (men only, or both sexes)Orissa 11 Village patrol (men only) or male guardUP hills 18 Village patrol (women only) or male guards (sometimes

female guards)

Nepal 14 Male guard employed by villagers (sometimes female guards)

Source: Bina Agarwal, Fieldvisits in 199899.a All patrols involve rotation among member households. Patrol size can vary between two and 10 persons.b Does not include cases that have shifted to these methods of protection from other methods. For such cases see``changing methods''.



8/26

where they were frustrated by their lack ofvoice in the existing male-dominant CFG andadditional forest land was available for settingup their own (eld observations in Orissa,199899). While the former cases are encour-

aging, CFGs formed as an exit option providean unsatisfactory solution to women's exclu-sion from the male forums which still controlmost of the resource.

More often, women self-initiate informalprotection groups to guard more eectively anarea being monitored by a male-dominatedCFG. Unlike formal groups which are clearlydelineated and often have the authority to makeand enforce rules, informal groups lack bothclear delineation and such authority. 17 In manyways, women's informal protection eorts are

extensions of their everyday forms of coopera-tion and social networking, and women'sabsence from men's formal groups reects thepower of gender exclusion that characterizesmany of men's other networks. Women's limitedinvolvement in formal CFGs, in turn, has wideradverse implications, as discussed below.

5. IMPLICATIONS OF PARTICIPATORYEXCLUSIONS

In so far as participation is important in itselfas a measure of citizenship rights and as a formof empowerment and voice, women's absence is

an indicator of a project's failure on thesecounts. In addition, excluding women whileincluding men could worsen power relation-ships and further disempower women. Indeedevery new participative institution that ignores

gender, be it forest user groups, water usergroups, or credit groups, can potentiallydisempower the women excluded. In addition,there are implications for distributional equityand institutional eciency. 18

(a) Equity

Table 5 broadly summarizes the kinds ofcosts and benets that tend to accrue mainly toone gender. Here I discuss a few to illustrate the

eects of participatory exclusion.Women's absence from CFG decision-mak-ing means that they have little say in theframing of forest closure rules. When protec-tion starts, most villages bar entry. Where theforest is highly degraded anyway, this maycause no extra hardship. But where earlierwomen could meet at least part of their fuel andfodder needs from the protected area, they arenow forced to travel to neighboring sites,involving additional time, energy, and the riskof being caught and ned. In the early years of

JFM, Sarin (1995) found that in some sites inGujarat and West Bengal, women's collectiontime and distances traveled for a headload ofrewood increased several-fold (Table 6).

Table 5. Main potential costs and benets of forest closure by gender a

Mainly aecting women Mainly aecting men

Costs Firewood shortages (more time and energy expended in

collection and/or cooking; and adverse health eects)Membership fee

Fodder shortages (more time and energy expended in collection) Patrolling time/guard's pay

Increased time in stallfeeding animals Fodder shortages (purchase) Informal patrolling time Loss of source for small timber Erosion of some livelihoods: e.g. rewood sellers, NTFP

collectors Erosion of some livelihoods: e.g. black-

smiths using woodfuel in furnaces Fines if caught stealing rewoodEnhanced (late entry) membership fee

Benets Firewood supply (a few weeks, if forest is opened) Small timber Fodder supply (a few weeks, if forest is opened) Housebuilding timberNTFP (seasonal)b Cash (if distributed) from sale of forest

productsUse of collective fund

a This is a broad outline of the main direct costs and benets. Not each of these need apply to every CFG. There mayalso be some indirect costs and benets. For instance, a greater supply of rewood indirectly benets the wholefamily.b NTFP: Non-timber forest products.



9/26

Table6.

Impactofforestprotectiononwomen'seortstoprocurecookingfuel:IndiaandNepala

Situatio

ninearly1990s:FieldvisitsbyMad

huSarin(Sarin,1995)

Statea

ndvillage

Time/distanceforgatheringoneheadloadofrewood

Frequencyofcollection

Otherimpact

Beforeprotection

Afterprotection

WestBengal(BankuraSouthDivision)

1.Kam

ardanga

1.52hr

45hr

5days/wkexceptmonsoons

Partialswitchingtolantana,painfulto

collect

2.Bhadli

0.5km

45km

n.i.

Haveto`steal'fromother'sforest,hefty

nesifcaught

3.Bara

paccha

12hr

34hr

Dailyexceptmonsoons

n.i.

4.Karapara

0.5km

89km

n.i.

Partialswitchingtoleaves,dung,husk,

weeds

Gujarat

1.Vena(Panchmahalsdistrict)

0.5hr

34hr

1week/month

Harassmentandabuseby

FDsta

andresidentsofotherv

illageswhen

womengotounprotec

tedforest

furtheraway

2.Chari(Panchmahalsdistrict)

1hr

45hr

Dailyforonemonth/year

Abusebyresidentsofothervillages;fear

ofbeingbeatenbyown

men

3.Malekpurb(Sabarkantha

district)

12hr

Wholed

ay

n.i.

Abusebyresidentsofothervillagesfrom

whoseforestwomenco

llectrewood

SouthB

ihar

1.Sara

iya(Palamaudistrict)

n.i.

n.i.

n.i.

Switchingtoleaves,dung,lantana,arhar

sticks,somepurchasing

rewood

2.Ram

ua(Hazaribaghdistrict)

n.i.

n.i.

n.i.

Switchedtoleaves,lantan

a,dung,

thornybushes,somebu

yingcoal

3.Banaso(Hazaribaghdistrict)

n.i.

n.i.

n.i.

Switchedtodung,weeds



10/26

Table

6continued

Situatio

nin199899:FieldvisitsbyBinaAg

arwal

Stateandvillage

Totalpro

tectedarea(ha)

PerHHprotecte

darea(ha)

Firewood

availabilityandwomen'sresponse

s

Gujarat

(Sabarkantha,Bharuch,Surat,andPanchmahalsDistricts)

V1

190

1.88

Womenreportsevere

rewoodshortage;switchtobushes,tursticks,other

agriculturalwaste.M

anyinthevillagehavechangedtheircropping

patterninfavoroftu

randcastortogetmoreagricultura

lwasteforfuel

V2

141

0.59

Womenreportrewoo

dshortage;partialswitchtotursticks,biogasand

kerosene

V3

253

0.94

Womenreportrewoo

dshortage;partialswitchtotursticks,kerosene,

dungcakes

V4

304

0.82

Womenreportrewoo

dshortage;partialswitchtokerosene,dungcakes

V5

41

0.16

Womenreportrewoo

dshortage;stealfromanotherforest

V6

5

0.02

Womenreportrewoo

dshortage.Manystealfromanotherforestand

reportpayingnesw

hencaughtbytheguard

V7

100

0.37

Womenreportrewoo

dshortage.Manystealfromanotherforestandif

caughtbytheguard

losetheiraxesorpaynes

V8

115

0.52

Noinformation

V9

35

0.57

Womenreportrewoodshortage;partialswitchtoagriculturalwasteand

dungcakes

V10

60

0.68


dungcakes.Somenolongerheatbathwaterinwinter.

Afewhave

biogas

V11

140

0.50

Womenreportrewoo

dshortage;partialswitchtodungcakes

V12

344

1.60


biogas

V13

40

0.16

Womenreportrewoo

dshortage,especiallyamongthepo

or

V14

1500

n.i.

Womenreportrewoo

dshortage;partialswitchtodungcakesand

agriculturalwaste.A

lsodistancetravelledforaheadloa

dofrewood

reportedtohaveincreasedfrom1kmbeforeprotection

to5.7kmafter

protection.Landless

faceacuteshortage

V15

450.1

3.21

Womenreportrewoo

dshortage.Somehaveswitchedto

biogasand

kerosene,butmosta

lsouseagricuturalwaste.Distance

travelled

reportedtohaveincreasedfrom1kmto5kmdailyafterprotection

started

V16

70

0.31

Womenreportserious

rewoodshortageforatleast45months;partial

switchtoagricultura

lwasteandkerosene

V17

179.27

1.13

Womenreportrewoo

dshortage.Therehasbeenanincre

aseinthetime

takenforrewoodcollectionsinceprotectionstarted.Partialswitchto

tursticksanddungcakes



11/26

V18

10

0.50

Womenreportserious

rewoodshortage;partialswitchto

tursticksand

dungcakes

V19

127

0.554

Priortoprotectiontherewasasevereshortage.Womenreportthereisno

shortagenow

Nepal(KaskiandDangdistricts)

CF1

2.5

n.i.

Manyusebiogas.Noinformationonwhatthepooruse

CF2

24.8

0.38

Womenreportacuterewoodshortage.Theydonotlight

areforspace

heatinginwinter;sp

endmoretimeincookingtosavew

ood;steal

rewood

CF3

156.4

0.62

Womenreportrewoo

dshortage,especiallyamongthepo

or.Poor

womenstealatnight

CF4

19

n.i.

Womenreportacuterewoodshortage.Stealingiscommon

CF5

n.i.

n.i.

Thebetter-ohaveswitchedtobiogasplantsandstoves.Thepoorface

acuteshortage,uses

malltwigs

CF6

75

n.i.

Womenreportrewoo

dshortage.Thepoorusestrawand

stealfrom

neighbouringforestatnight

CF7

5.62

0.05

Womenreportrewoo

dshortage;partialswitchtogas,so

mebuy

rewood

CF8

5.5

0.09

Womenreportsevere

rewoodshortage.Switchtotwigs,d

ungcakes.No

longerheatbathwateroranimalfeedinwinter

CF9

257

1.04

Womenreportacuterewoodshortage.Switchtodrystra

wanddung

cakes.Nolongerheatbathwateroranimalfeedinwinter

CF10

1.75

0.02

Noinformation

CF11

10.8

0.09

Notmuchrewoodav

ailable

CF12

212.7

0.39

Theverypoorwhoare

dependentonrewoodsaleforaliving,nowsteal

atnight.Theirincom

esareinsucientevenforadequatefood

CF13

7.2

0.11

Womenreportrewoo

dshortage.Stealingrewoodiscom

mon

CF14

141.0

0.99

Womenreportrewoo

dshortage.Thepoorsteal,eventhoughtheyare

oftencaughtandhavetopayaheavyne

Sources:(a)AdaptedfromSarin(1995).(b)BinaAgarwalFieldvisitsin199899.

a

V:villa

ge;CF:communityforestry;HH:household;n.i.noinformation.

b

Situationinthersttwoyearsofprotection;somesubsequentimprovement

duetocutbackoperationsandrewooddistributionamongvillagers,afterwomen's

persisten

tcomplainingandthelocalNGO'ssupportledtheCFGtotakereme

dialmeasures.

*

All-wo

menCFGs.Asnoted,theirplotsaretypicallysmallanddegraded.



12/26

Oftentimes, school-going daughters had to jointheir mothers, with negative implications fortheir education.

As the forests regenerated, a less rigid closureregime might have been expected. But I found

this not to be the case even several years intoprotection. Of the 87 CFGs I visited in199899, 80 had rewood available. Of these 45(60%) had a ban on rewood collection,wherein 21 did not open the forest at all and 24opened it for a few days annually for drywoodcollection and/or cutback and cleaning opera-tions. The remaining 35 allowed some collec-tion on a continuing basis, but only of fallentwigs and branches (Table 7). Even after yearsof protection, women thus reported a persis-tence of rewood shortages in most villages I

visited, and in virtually all the study villages inGujarat and Nepal (Table 6). The exceptionswere regions that already had relatively denseforests when protection started, as in parts ofMadhya Pradesh and Orissa. Some character-istic responses in scarce regions are givenbelow:

We go in the morning and only return in the evening.Since the end of the rainy season, we have been goingevery day. I go myself and so does my daughter. Ear-lier too there was a shortage but not as acute (womanEC member to author, Kangod village, Karnataka,1998).

Men preach to women about not cutting trees, butwhat can women do? They cannot cook food withoutrewood and they cannot collect rewood from otherplaces (group discussion with women in KabhrePalanchok, Nepal, cited in Hobley, 1996, p. 147).

The women asked the men: should we burn our handsinstead of rewood to light the stove? (male NGOworker, Panchmahals district, Gujarat, 1999).

Where possible, women substitute other fuels:

a few can switch to biogas, but most have to useinferior fuels such as twigs, agricultural waste,even dry leaves (Table 6). These fuels need moretime to ignite and tending to keep alight, thusadding to cooking time and preventing womenfrom simultaneously attending to other work.Moreover, the additional fumes from inferiorfuels, in typically poorly ventilated conditions,have negative health eects. In some areaswomen also economize on fuel by forgoing awinter re for space heating (even in the subzerotemperatures of the Nepal hills), by not heating

bath water in winter or heating it only forhusbands, and so on.

Usually women from both middle and poorpeasant households report such domesticenergy problems, since even the former seldompurchase rewood or have enough private treesfor self-suciency. Women of landless orlandpoor households are, however, the worsto, since without private land they have nocrop waste or trees of their own, and few cattlefor dung. 19 As a poor woman in Khut village(UP hills) told me: ``We don't know in the

morning if we will be able to cook at night.''Can the regenerated forests alleviate these

shortages? Some argue, yes. Shah (1997), forinstance, suggests a norm of 0.2 ha of forest perhousehold in Gujarat for meeting rewood andother household needs. In my Gujarat sample,

Table 7. CFG rules of rewood collection by region: India and Nepal

Rules Gujarat Karnataka MadhyaPradesh

Orissa UttarPradesh

Nepal Total

Drywood collection banned 6 4 3 7 1 21

Drywood collection bannedbut forest opened for alimited period each yearto allow collection and/orcleaning/cutbackoperations which providesome rewood to CFGmembers

9 2 1 2 10 24

Drywood collection allowed(typically only fallentwigs and branches)

3 5 12 7 6 2 35

Little rewood available 1 1 2 1 5Unspecied or no

information 1 1 2

Total 19 13 12 11 18 14 87

Source: Bina Agarwal, Fieldvisits in 199899.



13/26

15 out of 19 villages have more than thisminimum (Table 6). Some others, however,give much higher estimates: e.g., Lal (1992)suggests 0.47 ha per capita, but as an all-Indiaaverage and covering forest products for all

uses. What is less contested is that much morerewood can be extracted sustainably than theconservative closure regimes currently allow.Other solutions (discussed further below) couldalso be found. The neglect of this persistentproblem appears to largely reect women's lackof bargaining power in the community.Women's time and eort is invisible unpaidlabor that does not elicit automatic communityresponse.

Similarly, there are gender inequities inbenet sharing. In most cases, cash benets are

not distributed. In the 29 CFGs across sixIndian states examined in Table 8, for instance,only three distributed any funds to members.Most commonly, the funds were put to usesfrom which women were unlikely to benet,such as youth club repair, purchasing commu-nity utensils, rugs, drums etc. (which the menused or leased out), and travel by EC members.Only seven CFGs allocated funds for a forestguard or for community needs such as schoolrepair or income-generating enterprises. Insome regions, such as Orissa, I found that

spending on religious functions and youth clubswas especially common.

Women have little say in fund allocations. 20

Many resent this:

The money obtained from grass and rewood is keptby them in their fund. We have not seen one penny ofit. We buy grass, which is auctioned by bundles. Thecommunity forest belongs to the men. We own noth-ing (women to author, Ghusra village, Dang district,Nepal, 1998).

Even where the CFGs distribute some cashbenets, women of nonmember householdsreceive none, since entitlements are linked tomembership. These are often the poor whocannot easily contribute to the guard's wages or

to patrolling (where they migrate for work).But even in member households, money givento men does not guarantee equal sharing, oreven any sharing, within the family. As foundin non-CFG contexts (e.g., Dwyer & Bruce,1988), here too the men typically spend asubstantial part on gambling, liquor, or perso-

nal items. 21 Women gain if the benetsdistributed are in kind (such as rewood orgrass), or where drywood collection or grazingis routinely allowed. But entitlements areusually on a household basis, so even when

both spouses are members they get only oneshare. Women, when asked their preferences,tend to opt for equal and individual shares foreach spouse (Sarin, 1995; my eldwork inGujarat, 1999).

Inequities also arise from the distributiveprinciples followed. Forest products can bedistributed either according to willingness topay (i.e., market determined), or by contribu-tion, or need. While seemingly neutral, theseprinciples have signicant gender and classimplications since people dier in their abilities

to pay or contribute, or in their needs. Practicessuch as auctioning grass to the highest bidder(the market principle) tend to be both unequaland inequitable, since those unable to pay getnone, even if they have contributed to protec-tion. This adversely aects both the poor andwomen in general, since rural women even inrich households have less access to nancialresources than men. But even distribution bycontribution (say toward protection) can prove

Table 8. CFG funds: principal uses, India

Use of fundsa W. Bengal N 20 Other states N 9b All N 29

Purchase of community utensils,rugs, drums, etc. for CFG useand leasing out; travel by ECmembers; local club repair

14 4 18

Religious and social functions 2 2 4Forest guard's pay, school repair;

income generation 7 7

Loans to families 3 2 5Distributed to members 1 2 3Not used so far 6 1 7

Source: Compiled from Ecotech Services (1999).a Some CFGs use the funds for more than one purpose.b Andhra Pradesh (2); Karnataka (1); Madhya Pradesh (2); Rajasthan (2); Uttar Pradesh (2).



14/26

inequitable where the poor and women gener-ally are less able to contribute toward theguard's pay or to patrol duty, or because theyare more dependent on the commons. Only ifsome account is taken of dierential needs,

such as by allocating poor women rights in anadditional grass patch, would the distributionbe relatively equitable.

In practice, I found contribution (in terms ofmembership fees, protection, etc.) to be themost common basis for distribution, with allcontributing households getting equal access tograss or fuelwood when the forest was opened.But there were also cases of auctioning either ofgrass, as in the UP hills and Nepal, or of otherforest products as in Orissa. Need rarely guideddistribution. Hence for poor women the

outcomes of forest closure have been especiallyinequitable.

Of course, not all the noted inequities can becured solely by women's presence in decision-making. For instance, that rural women (ratherthan men) bear the main burden of rewoodshortages stems from the pre-existing genderdivision of labor wherein rewood collectionand cooking is usually women's work. Thisdivision of labor is unlikely to change solely bywomen's participation in CFG decision-mak-ing. Participation can, however, reduce several

other dimensions of the noted gender inequitiesin costs and benets, such as:

It could lead to more equitable distribu-tion rules. I found, for instance, that womenacross class lines were vehemently opposedto forest products being auctioned andwanted rules which allowed everyone equalaccess to grass or rewood. Women's re-sponse in Chanauta village (Nepal) is illus-trative: ``We should not have auctions. Letone person per household go and get asmuch grass as she can carry and let everyone

pay the same amount as an entry fee.''Again, with greater say in decisions, womencould push for shares in their own right,rather than on a household basis. In addi-tion, poor women could put pressure onCFG committees to have their particularneeds taken into account.

It could give women greater say incommunity fund allocations.

It could induce communities to formulateless conservative closure policies so that,where possible without sacricing sustain-

ability, more could be extracted from theregenerated forests. In some sites this hashappened only after women forced the issue

by constant complaining or ghting with theforest guard.

It could pressure the CFG to nd solu-tions to domestic fuel shortages, such as byallocating part of the forest to fuelwood

plantations; using the community funds tosubsidize alternative fuels such as biogas,or to subsidize rewood purchase from else-where, etc. In other words, if women had amore eective voice, rewood shortageswould be seen as a community concern andnot just the concern of individual house-holds, or of women alone.

(b) Eciency

Women's lack of participation in CFG deci-

sion-making similarly has eciency implica-tions. At one level, many CFGs have hadnotable success in forest regeneration. In somecases, replanting is undertaken, but if therootstock is intact even simply restrictinghuman and animal entry can lead to rapidnatural revival. For instance, within ve toseven years of such restrictions many severelydegraded tracts in semi-arid India are foundcovered with young trees; and areas with littleand declining vegetation show signs of goodregeneration. In fact in most ecological zones,

CFGs show such benecial results, and inseveral cases incomes and employment arereported to have increased, 22 and seasonaloutmigration fallen. 23

At the same time, there are gaps between thegains realized and those realizable; or cases ofgroups failing to take o, or being unable tosustain their gains. In broad terms, such inef-ciencies can stem from ineective protectionand inecient planning for forest development(see also, Agarwal, 2000a,b).

(i) Ineective protectionAt present, there are at least three indications

of ineciencies in protection. One, there arerule violations. Almost all the villages I studiedreported some cases of violation, at times as afrequent occurrence. Violations by men areusually for timber for self-use or sale (the latterin areas with commercially valuable trees).Violations by women are typically for rewood.Where a CFG bans collection withoutconsulting the women or addressing theirdiculties, many are under pressure to break

the rules, given their daily need for fuelwood.Sometimes, acute need forces women intopersistent altercations with the guard. 24 In one



15/26

Gujarat village I found that only when theguard threatened to resign as a result did theEC agree to open the forest for a few daysannually. In Agrawal's (1999) study of a vanpanchayat village, women constituted 7080%

of the reported oenders during 195191, mostbeing poor and low caste.

Women's greater participation in rulemaking could reduce tendencies to rule break-ing, by allowing them to formulate rules thatare not only fairer but also perceived by them tobe so. As women in the UP hills reasoned: ``Themale members of the committee have dicultiesimplementing the rules. Women could discussthese problems with the men. Perhaps more`mid-way' rules would be, in the long run, moreeective. . . more viable'' (cited in Britt, 1993,

p. 148). Bardhan's (1999) study, althoughungendered and relating to water users groups,is also a pointer: he found that rule compliancewas linked to participation in rule formulation.

Two, there are problems in catching trans-gressors. In virtually all the regions I visited,all-male patrols or male guards were unable todeal eectively with women intruders becausethey risked being charged with sexual harass-ment or molestation, especially where non-member women, or women from neighboringvillages, were caught. In some incidents, women

and their families registered false police casesagainst patrol members, or beat them up.Involving women in protection would thus bemore eective, as some villagers recognize.

When only men protected, women would not listen tothem. They would say it is our forest so you can't stopus. But women can persuade other women. They canalso physically stop them, whereas men can't. Oncewhen a man was protecting and saw a woman bring-ing wood, he asked her to put the wood down. Herhusband later beat him with a stick and broke hishead. If women were protecting this would not have

happened (woman leader to author, Vejpur village,Gujarat, 1999).

Equally, however, women on their own nd itdicult to patrol at night or confront aggres-sive male intruders. The most eective solutionappears to be a patrol team constituted of bothsexes. Recognizing this, in some regions malepatrol groups have inducted women, but this isatypical.

When women voluntarily set up informalpatrols, even where there is a male guard or

patrol, protection eciency can improve nota-bly. In their study of 12 van panchayats, Sharmaand Sinha (1993) found that all the four that were

``robust'' had active women's associations. Butin so far as women's groups are typically infor-mal, they lack the authority to punish oenderswho still have to be reported to the formal(typically all-male) committees. This separation

of authority and responsibility can introduceineciencies in functioning. For instance, I cameacross several cases where the women hadabandoned their eorts and the forest conditionhad worsened because the male EC members didnot punish the culprits women caught. In addi-tion, when women catch intruders, they areseldom party to discussions or decisions onappropriate sanctions. This also means thatconict resolution is often ineective.

Three, often the male guard or patrol fails tonotice resource depletion. In several cases,

women's informal patrols took me to theirpatrol site and pointed out illegal cuttingswhich the men had missed. Part of this genderdierence arises from the fact that women, asthe main and most frequent collectors of forestproducts, are more familiar with the forest thanmen.

(ii) Inecient forest planningThe second set of potential ineciencies lie in

inadequate consultation with and informationseeking from the women. For instance, a failure

to involve women in forest development andnew planting decisions means missed opportu-nities for gaining from the particular knowl-edge of plants that we noted rural women inmany areas possess. This could reduce pros-pects for regeneration and biodiversity. More-over, if no account is taken of genderdierences in tree preferences, it can undercutthe project's ability to fulll household needsand the commitment of excluded members tothe initiative.

Basically, in ignoring gender concerns, theCFGs are violating many of the conditionsdeemed by several scholars as necessary forbuilding enduring institutions for managingcommon pool resources. This includes ensuringthat those aected by the rules participate inframing and modifying the rules; that the rulesare kept simple and fair; that there are gradu-ated and appropriate sanctions against oend-ers; that there are eective mechanisms formonitoring the resource and resolving conicts;and so on. 25 All these conditions would be

fullled better with women's greater participa-tion in all aspects of CFG functioning, withpositive eciency eects.



16/26

Of course, despite women's low involvement,forests can show regeneration, but many mightnot reach their full potential for growth, or forbiological diversity or benet provisioning; andsome CFGs might not be sustained at all. While

these eciency eects require more empiricalverication, even the noted evidence providesstrong pointers, and makes it imperative tounderstand what determines participation andhow it can be improved.

6. WHAT DETERMINESPARTICIPATION?

In broad terms, my study suggests thatwomen's participation in CFG functioning

depends especially on the following factors: Rules of entry: the criteria deningmembership in the GB and EC.

Social norms that dene who should at-tend and speak up at meetings; who shouldform the patrol; how men and womenshould behave in public, and so on.

Social perceptions regarding women'sability to contribute to CFG activities.

Entrenched territorial claims.Personal endowments and attributes (e.g.,educational levels, property status, marital

status, age, etc.). Household endowments and attributes(which dene where women fall in the struc-tural hierarchies of class and caste).

(a) Rules

In formal CFGs, such as India's JFM groupsor Nepal's FUGs, rules determine membershipin the GB or EC. As noted earlier, rulesallowing membership to only one person perhousehold eectively exclude women. Those

allowing one man and one woman per house-hold to join are more inclusive; but the mostinclusive, which allow all adults to join, arerare. A lack of awareness of rules, or of changestherein, can also constrain women's participa-tion. For instance, a study of 19 West BengalCFGs showed that even four years after themembership criteria was amended to allowwomen's inclusion, barely two-fths of themembers knew of the change (Raju, 1997).Among the self-initiated groups (that lackformal membership rules), long-standing

conventions, which traditionally excludedwomen from public decision-making forums,deny women entry into the CFGs.

(b) Social norms

Even when entry rules are favorable and wo-men join the CFGs, they seldom participate inmeetings due to a range of pre-existing restric-

tive social norms, such as those described below.

(i) Gender segregation of public spaceIn general, village spaces where men congre-

gate (such as tea stalls and the market place) arespaces that women of ``good character'' areexpected to avoid (Agarwal, 1994). Restrictionis somewhat less for older women, but neverentirely absent. These notions carry over toformal village meetings. A fear of reputationloss or family reprimand makes most womenuncomfortable with attending CFG meetings,

unless the men explicitly invite them:

They don't call us, so we don't go (women to author,Roopakheda village, Madhya Pradesh, 1999).

The meetings are considered for men only. Womenare never called. The men attend and their opinionsor consent are taken as representative of the wholefamilyit's understood (woman in a van panchayatvillage, UP hills, cited in Britt, 1993, p. 148).

Rural women and men can't sit together. But we con-vey our decisions to them (man to author, Chattipurvillage, Orissa, 1998).

(ii) The gender division of laborThe fact that women bear the main respon-

sibility for childcare and housework, in addi-tion to their share of agricultural work,cattlecare, etc., makes for high work burdensand logistical constraints. Men are reluctant toshare not just domestic tasks and childcare, buteven cattlecare. This seriously restricts women'sability to attend lengthy meetings held atinconvenient times:

There are problems in attending meetings since weneed to cook and serve the evening meal. The meetingis long. We also have to feed the cattle (woman toauthor, Barde village, Karnataka, 1998).

Women have a lot of problems. They can cut grassquickly, but who will give grass to the bualoes ifwe come to a meeting? Women's work is a constraint(Arti Shrestha, grassroots organizer, to author, Kaskidistrict, Nepal, 1998).

We want women to come to our meetings, but whowill take care of the children? (man to author, Basapurvillage, Karnataka, 1998).

In the van panchayat villages that Mansingh(1991) studied, women's attendance thinned



17/26

over time since, as most women said, they didnot have time to ``sit around for [the] fourhours that it took to have a meeting in themiddle of the day.''

(iii) Gendered behavioral normsNorms of both acceptable male behavior and

acceptable female behavior impinge onwomen's participation. For a start, womenhesitate in attending meetings due to aggressivemale behavior:

When we open our mouths, men shout us down (wo-men in Harimari village, West Bengal, cited in Raju,1997).

If men drink and say something to us, we don't like it.They ght with us, so we don't go to the meetings (wo-

men to author, Khabji village, Gujarat, 1999).

Men drink a great deal here. . . and start abusing us. . .If one woman is abused, ten men stand up and agreewith the abuser saying, yes, she deserves this (womento author, Deolikhal village, UP hills, 1998).

The men I interviewed usually agreed withwomen's observations that men often tended todrink and ght in the meetings. Some said sooutright: ``There are often ghts in the meeting,so women will not come'' (Hanasur village,Karnataka, 1998).

Equally, women's participation in male-dominated CFGs is constrained by the socialstrictures on their visibility, mobility, andbehavior, whether internalized by them orimposed by threat of gossip, reprimand orviolence. Female seclusion norms are the mostrestricting. Although traditionally uncommonamong hill and tribal communities, such norms(e.g., veiling) have been adopted by some,under inuence of upper-caste Hindu practices.More pervasive, however, is the social emphasison self-eacement, shyness and soft speech that

many women imbibe.Gendered behavioral norms also restrict

women by creating subtle hierarchies, such asrequiring them to sit on the oor while men(especially older ones) sit on cots or chairs.Even where everyone sits on a level, typicallywomen (including EC members) sit on one sideor at the back of the meeting space where theyare less visible. This makes them less eective inraising a point, while issues raised by malemembers who sit in front receive priority.Moreover, when senior family males are

present, women hesitate in attending meetings,or speaking up at them, or opposing the menpublicly. The hierarchy that marks ``respectful''

family behavior gets carried into communityspaces. 26 In addition, the forest service sta islargely male (Narain, 1994), so that lines ofauthority and gender overlap.

The collective action literature has typically

emphasized the enabling positive side of socialnorms, but from women's viewpoint theseexamples reect the disabling ``dark side'' ofmany social norms.

(c) Social perceptions

Incorrect perceptions regarding women'sabilities also impinge on men's reluctance toinclude women. Men often view women'sinvolvement in CFGs as serving no usefulpurpose and tend to downplay their potential

contributions. During my eld interviews withmixed groups, for instance, while menanswered questions women would listen atten-tively and keep the children quiet. But when Iwould begin interviewing the women, the menwould smile, implying: Why ask her, what doesshe know? On my persisting, they would beginchatting among themselves, or interrupt thewomen and answer questions on their behalf,or just get up and leave.

Equally indicative of their perceptions aresome of men's direct responses:

There is no advantage in having women in the EC. Wehave been told by the forest ocials that we must havetwo women in the committee, that is why we have in-cluded them (male to author, Pathari village, Karna-taka).

Calling all the women to the meeting just hampers thework of the agenda. . . It is true that women are thereal users of the forest but our women have not yetparticipated in the meetings. They don't know much,they don't give solid opinions. I am a man, I attendthe meeting. If I am prepared to make the femalemembers of my family act according to what I say,why should they attend the meeting? (male respondentin Nepal, cited in Hobley, 1990, p. 310).

Women can't make any helpful suggestions (man toauthor, Arjunpur village, Orissa, 1998).

Women are illiterate. If they come to meetings, wemen might as well stay at home (EC chairman toauthor, Ghusra village, Dang district, Nepal, 1998).

In some cases I asked the men who justiedwomen's exclusion on the ground that womenwere illiterate, whether they themselves were

literate. It turned out that many weren't! Arecent three state study of 843 women repre-sentatives in India's village councils (Buch,



18/26

1999), indicates that illiteracy in itself is not theconstraint to women being eective leaders,that it is argued to be.

Women encounter negative perceptions notonly from village men but also from male forest

ocials. In West Bengal, for instance, womencomplained to Narain (1994) that male ocialsdiscouraged them from coming to the forestoce, and rebuked them if they came in theevening. Roy, Mukerjee, Roy, Bhattacharya,and Bhadra (1993, pp. 1516) similarly notethat ``the forest ocers put very little value onwhat [women] say and always crosscheck withthe men to verify the truth of [women's] words.''

(d) Men's entrenched claims and control overcommunity structures

Opposition to including women is strongerwhere men's claims are already entrenched.For instance, where CFGs begin with onlymale members, or where men feel they have aprior claim to the forest, they resist newclaimants. In Basapur village (Karnataka) menreacted to the idea of including women asfollows: ``Women have DWARCA, 27 theyhave savings groups, why don't you leave theCFGs to us men?'' In Kudamunda village,Orissa, women wanted their own separate

patch for protection because (they said): ``Ifwe have our own forest, we would not need toask the men each time for a bit of wood.''``They are not willing to give us even a patchto protect. Why would they be willing to giveus a whole tree if we asked?'' Elsewhere, in theUP hills, women from Bitholi village told methat when they closed o a patch of opengrazing land, the men objected and insisted ongetting the grazing reopened.

Men are also able to assert their interestsbetter through their prior control over

community structures and village publicforums. Where there are caste/class dierences(as discussed further below) low-caste poormen too are disadvantaged in this respectrelative to the upper-caste rich.

(e) Personal endowments and attributes

The fact that rural women typically lackpersonal property and are seldom wellconnected politically, also reduces the weight oftheir opinions; and limited experience of inter-

action in public forums undermines theireectiveness. Some of these disadvantages canbe overcome if the women are older, married,

have leadership qualities, and the self-con-dence to speak up. In many CFGs, the activewomen members are mostly widows, or oldermarried women living in their parental homes(Britt, 1993; Narain, 1994), women whose

social positioning, as Britt observes, ``enablesthem to travel more freely, speak more loudly,and assume the posture of local leaders.'' Suchwomen also carry a lower burden of domesticwork than do young married women.

(f) Household endowments and attributes

Finally, the class and caste position of thewoman's household is likely to matter wherethe village is multi-caste, with a dominance ofthe upper caste, or where the CFG is consti-

tuted of several villages that are caste/classhomogeneous in themselves, but dier hierar-chically in this respect from other villages in the

CFG. 28 But, the impact of caste is not linear.On the one hand, being low caste and poor canreduce a person's bargaining power within apredominantly upper-caste community; on theother hand, low-caste women are also lesssubject than upper-caste ones to norms thatrestrict their mobility and stie their speech.

Strikingly many of the above constraints are

common across the dierent regions studied.This is also indicated by the quotations given, inwhich women from quite diverse regions ofIndia and Nepal describe their constraints invery similar ways. This does not imply thatcultural diversity is irrelevant in deninggendered norms and perceptions. But in relationto public decision-making forums traditionallywomen's exclusion was near universal acrossSouth Asia. Even in matrilineal communitiesdecision-making authority in communitycouncils was vested in men (Agarwal, 1994). On

this count therefore cultural and geographicvariation does not appear to make a notabledierence. Moreover, the range of regionaldierence in women's social situation is rela-tively small within the context of CFG func-tioning, given (as noted earlier) that most CFGsare still located among tribal and hill com-munities. Regional/cultural dierences could,however, prove an important determinant ofwomen's participation as more CFGs areformed in mixed caste contexts. Such dierencescould also aect CFG responsiveness to NGO

attempts at promoting women's participation,with regions and communities with histori-cally less gender bias being more responsive.



19/26

7. ENHANCING PARTICIPATION: THEBARGAINING FRAMEWORK

How can the noted factors be changed toimprove outcomes? Some factors (e.g.,

membership rules) are embedded in the designof community forestry programs; but severalothers have deep economic and social rootsthat predate the programs. The programscould, however, either further entrench eventhe historical factors, or serve as an opportu-nity for weakening their hold. A promisinganalytical framework for examining the possi-bilities of positive change is that of bargaining.In terms of this framework, women's ability tochange rules, norms, perceptions and endow-ments in a gender-progressive direction would

depend especially on their bargaining powerwith the State, the community and the family,as the case may be.

(a) Bargaining: some conceptual issues 29

Given the substantial literature on bargain-ing, here it suces to mention only the broadparameters of bargaining situations. Broadly, abargaining situation is one in which interactionbetween two parties contains elements of bothcooperation and conict. Both parties cooper-

ate in so far as cooperative arrangements makeeach of them better o than noncooperation.But, many dierent cooperative outcomes arepossible that are benecial to the negotiatingparties relative to noncooperation. Among theset of such outcomes, some are more favorablethan others to each partythat is one party'sgain is another party's losshence the under-lying con ict between those cooperating.Which outcome emerges depends on the rela-tive bargaining power of the parties. Insightsinto the determinants of bargaining power can

illuminate both why inequality persists andhow it may be reduced.

Traditionally economists have applied theseideas within the game-theoretic mode and withlittle attention to gender. Recent interest inincorporating intrahousehold gender dynamicshas yielded some interesting formulations ofbargaining models, but with little application toarenas such as the State and the community,and inadequate attention to the impact ofqualitative factors on bargaining outcomes. 30

In our discussion at least two types of shifts

from this standard approach appear necessary:one, taking account of qualitative factors suchas social norms and perceptions; two extending

the framework to include extrahouseholdnonmarket arenas, in particular the State andthe community. Both extensions are likely toneed a less restrictive formulation than of aformal game-theoretic model. Here it is useful

to distinguish between a bargaining approachand formal models. 31 A bargaining approachthat is not constrained by the structure thatformal modeling requires, would allow us tomore freely develop and apply the concepts ofcooperation-con ict and bargaining power tothe noted extrahousehold arenas. It wouldalso allow a freer engagement with thecomplexities of gender analysis and with qual-itative factors.

In the present discussion, applying thebargaining approach to interactions within

three major arenasthe State, the communityand the familyis especially relevant. Forinstance, some CFG rules, such as eligibility toJFM membership are made at the State level,others that relate to say forest closure orproduct distribution are made largely at thecommunity level. Social norms, social percep-tions, and endowments are constituted in allthree arenas. What would aect women's abil-ity to bargain eectively in these arenas? 32

(i) The State

Consider rst bargaining with the State. Tobegin with, the State itself can be seen as anarena of bargaining at multiple levels. It may,for instance, formulate gender-progressivepolicies for community forestry at the highestlevel, but the local bureaucracy with lesscommitment to gender equality may resist theirimplementation (as revealed by the attitudes ofsome of the forest ocials described earlier). Orsome State departments (such as women'sministries), or key individuals within depart-ments, may pursue gender-progressive policies

within an overall retrogressive State structure.The elements of the State more committed togender equality would be the ones that a groupof women (or a group representing women'sinterests) could eectively cooperate with, evenwhile being in conict with other elements. TheState, in turn, would have an interest inresponding positively to demands from gender-progressive groups because of political pressure,or pressure from international aid agencies, orbecause it recognizes the inecacy of its ownmachinery in implementing some essential

development programs. In India, the State'srecognition of NGO contribution in commu-nity forestry is a case in point.



20/26

We would expect village women's bargainingstrength with the State to depend on a complexset of factors, such as whether they function asa group or as individuals, and the cohesivenessand strength of the group. The bargaining

power of such a group is likely to be higher thelarger and more unied it is, the more thepolitical weight carried by the castes of which itis composed, the greater its command overeconomic resources, the more the support it canmuster from gender-progressive elementswithin the State and from NGOs, the media,academics, and international donors, and themore State ocials are inuenced by gender-progressive ideas.

(ii) The community

Within the community, again, there are ele-ments of both potential conict and potentialgain from cooperation between individuals orbetween subsets of individuals. Implicit orexplicit bargaining can occur between an indi-vidual (or a subset of individuals) and othercommunity members over the formulation and/or enforcement of rules and norms governing,say, economic resource use and social behavior.Noncompliance with CFG rules could be seenas a form of implicit bargaining.

As with the State, women's bargaining power

with the community would be enhanced if theyhave support from external agents such asNGOs and the State. Group cohesiveness andstrength would again be important. Forinstance, it is easier to penalize an individualwoman who breaks seclusion norms by castingaspersions on her character, than if a group ofwomen transgress the norms (Agarwal, 1994).It is also important to examine what histori-cally-constructed gendered norms are prevalentin the region. In addition, in a multi-caste/class-heterogeneous village, we would expect

women's bargaining power to depend on theirgroup's socioeconomic composition. In thesharing of common pool resources, forinstance, the negotiating strength of a groupconstituted of low-caste or poor peasantwomen would tend to be weaker than one ofhigh-caste or rich household women whosecaste or class commands greater village power.

(iii) The familyIntrafamily bargaining for a more equitable

sharing of benets or tasks, or for greater

freedom to participate publicly, is the mostcomplex. As spelled out in Agarwal (1994,1997b), broadly we would expect especially four

types of factors to aect woman's intrahouse-hold bargaining power: her personal endow-ments and attributes (educational level, whethershe earns an income or owns property, her age,marital status, etc.), her ability to draw upon

extrahousehold support from friends, relatives,the State, women's groups and gender-pro-gressive NGOs within and outside the village,social norms (which dene who gets what, orwho does what within the household), andsocial perceptions (say about deservedness).

In the above formulation, external agentsupport, group strength and social norms andperceptions are seen as the common determi-nants of bargaining power in all three arenas.Here village women's group strength derives

not merely from the number of women seekinga change in rules and norms, but also fromtheir willingness to act collectively in theircommon gender interest, over and above thepossible divisiveness of caste or class. Thecreation of such group identity would need tobe part of the process of improving outcomesfor women.

Social norms can aect bargaining powerboth directly and indirectly. For instance,female seclusion norms reduce women's abilityto bargain for CFG rule changes directly by

restricting women's presence in public spaces,and indirectly by reducing women's ability tobuild contacts with NGOs and State ocials.Norms can also inuence how bargaining isconducted: e.g., covertly or overtly. Wheresocial norms stie explicit voice, women oftendepend on covert (and less eective) forms ofcontestation within the family, such as persistentcomplaining. Social perceptions aect women'sbargaining power in so far as perceptionsdiscount women's actual contributions. Muchof what village women do is typically under-

valued by families and communities. Moreover,if women internalize these perceptions they canself-restrict their options and eorts for change.To enhance women's bargaining power, anecessary step would thus be to alter both howthey perceive themselves, and how their familiesand communities perceive them. Ground expe-rience reveals some of these key elements of thebargaining framework.

(b) Bargaining: ground experience thus far

(i) Bargaining with the stateJFM experience indicates that pressure from

external agents such as NGOs, donors, and key



21/26

individuals can prove signicant in changingthe initial rules of entry. For instance, since theearly 1990s several states have made their JFMmembership rules more gender inclusive, andmany new JFM states have adopted such rules

(Table 1). More recently the Central govern-ment has passed new guidelines directing allstates to ensure that 50% of JFM general bodymembers are women (Government of India,2000). Here village women did not have tobargain explicitly for change. External agentsupport and the greater gender sensitivitybrought about by the women's movement inSouth Asia gave village women implicitbargaining power. This process can be carriedfurther if the Federations of CFGs catalyzed byNGOs to ``scale up'' CFGs cross-regionally, 33

are more gender sensitive. These Federationsenhance the bargaining power of local groupsvis-a-vis the State. More female forest serviceocials at levels that interact closely withcommunities would also help village womenbetter represent their concerns to the State.

(ii) Bargaining with the communityEnsuring that women-inclusive membership

rules are implemented by the community, orthat women have more voice in CFG forums, isfound to be more dicult than rule changes at

the State level. External agent support hashelped to some extent. Some Indian NGOs, forinstance, have used their bargaining power onwomen's behalf, by making CFG formationconditional on high female membership. OneGujarat NGO insists on 50% women whenstarting new CFGs. Similarly, some Indianforest ocials have substantially increasedwomen's CFG presence by stipulating thatthere should be at least 30% women in the GB(Viegas & Menon, 1993), or that meetingswould only be held if the men also invited the

women. Women, on being so invited, often turnup in strength (Sarin, 1998).

External agent bargaining power appears towork best when women's participation ispushed from the beginning. Once men's ``terri-torial interests'' get entrenched, women's entrycan prove dicult, even if the formal rules arefavorable. Involving women from the start canalso reduce subsequent gender conict overrules. How rules are perceived and interpretedagain makes a dierence. For example, oneNGO sta member in Gujarat, when forming a

CFG found that when he said: ``there should beat least two women'' (the stipulated minimum),the villagers perceived this as implying a

maximum of two women, so only two wouldjoin. Now he says: ``anyone who wants to, canbecome a member,'' leading to larger numbersof women joining (personal communication bythe NGO sta member). Similarly some gender-

sensitive NGO personnel have increasedwomen's voice in mixed meetings by solicitingtheir opinions and giving them weight, or havebrought up women's complaints about re-wood shortages at CFG meetings, leading tothe forest being opened for a few days annually(personal observation, Gujarat, 1995).

Not all regions, however, have externalagents committed to gender equity. A largerand sustained impact is likely to need women'sactive input. Left to themselves women typi-cally rely on covert and individual forms of

protest (their ``everyday forms of resistance''),ranging from simply ignoring the closure rulesor challenging the authority of the male guard,to persistent complaining. Sometimes, thisbears fruit:

After our complaints, women and men had a jointmeeting and decided to open the forest for a few daysfor rewood collection, since everyone has to cook(women to author, Asundriya village, Gujarat, 1999).

The seven-day opening was inadequate, so womenusers complained. Now they open the forest for eleven

days (women to author, Laxmi Deurali village, Kaski,Nepal, 1998).

But complaining or breaking rules are seldomfound to be the most eective ways of changingrules. For eective change, women are likely toneed more formal involvement in rule makingand the bargaining power to ensure changes intheir favor.

Ground experience suggests that for morevoice in mixed forums and for challengingrestrictive social norms and perceptions, would

need, for a start, a critical mass of vocalwomen, and a sense of group identity. As somewomen in the UP hills, interviewed by Britt(1993, p. 146), stressed: ``without a goodmajority of women present it is impossible toexpress opinions.'' The importance of a criticalmass has also been strikingly noted in thewestern context. Dahlerup (1988), for instance,in her study of Scandinavian women politiciansfound that gender stereotyping and even openlyexclusionary practices were common whenwomen were few in number. But once women

became a signicant minority (passing athreshold of some 30% seats in Parliament orlocal councils), there was less stereotyping and



22/26

open exclusion by men, a less aggressive tone indiscussions, a greater accommodation of familyobligations in scheduling meetings, and agreater weight given to women's concerns inpolicy formulation.

In South Asia, however, an additional initialstep appears necessary, namely buildingwomen's self-condence in public interactions.There is a growing consensus among NGOsand elements of the State that for increasingwomen's self-condence, and their numbersand voice in public forums, the formation ofseparate women's groups is essential. MayaDevi (a Nepalese activist with long experiencein group organizing) puts it emphatically:

In mixed groups when women speak men make fun of

them, so women need to learn to deal with this. . .When women join a [separate] group they graduallylose their fear of making fools of themselves whenspeaking up. . . Women need their own small groups.This is what I know from my 22 years of experienceworking with the government and NGOs.

It is less clear, however, what type of groupsthese should be. In some cases, all-womenCFGs have been formed, many of which havecome close to fullling the empowermentfunction of participation in my typology. AsPratibha Mundergee notes from 16 years ofexperience with a rural development NGO inKarnataka:

Yes, all women's CFGs have changed women's lives.Women now go to government oces themselves,including the Block Development Oce. They talkto the forest department ocials, they meet the rangeocers and the forester. Earlier they did not have thecondence to do so (personal communication, 1998).

All-women CFGs (as noted) are special cases,however, still small in number and area protec-

ted. To enhance women's presence and voice inthe more typical mixed CFGs, other kinds ofeorts appear necessary. Toward this end, someNGOs have formed all-women savings-and-credit groups; others have formed more multi-functional groups, such as mahila mangal dals inthe UP hills or amma samuhs in Nepal.

On the positive side, such separate women'sgroups have enhanced women's self-condenceand sense of collective identity. They have alsoimproved male perceptions about women'scapabilities and weakened social norms that

earlier dened only the domestic as legitimatefemale space. The responses below are fairlytypical:

Initially men objected to our going to meetings. Butour amma samuh helped men understand better. Whenwe women became united in the amma samuh, mensaw we were doing good work. That also helped (wo-men to author in Tallo Goundonda village, Kaski dis-trict, Nepal, 1998).

Men used to shut us up and say we shouldn'tspeak. Women learned to speak up in a sangathan(group). Earlier we couldn't speak up even at home.Now we can be more assertive a

Agarwal Participatory Paper

Documents

Transcript of Agarwal Participatory Paper