Gender analysis of land: beyond land rights for women?

Gender Analysis of Land: Beyond Land Rights for Women? 453

© Blackwell Publishing Ltd, Henry Bernstein and Terence J. Byres 2003.

Journal of Agrarian Change, Vol. 3 No. 4, October 2003, pp. 453–480.

Cecile Jackson, School of Development Studies, University of East Anglia, Norwich, Norfolk NR47TJ. e-mail: [email protected]

I would like to acknowledge the work of two recent research students that I have had the privilegeof working with, Disa Sjøblom and Nitya Rao. They are, however, in no way responsible for theuse I make of their work. I would also like to thank the journal’s referees and the editors for theirhelp in the preparation of the final version of the article.

Gender Analysis of Land:Beyond Land Rights for Women?

CECILE JACKSON

Gender analysts of development have worked on land and property relationsin poor rural areas for over two decades and the JAC 2003 special issue carrieda range of work reflecting some of these research trajectories. This article isboth a response to Bina Agarwal’s paper on ‘Gender and Land Rights Re-visited’, in which she reiterates her advocacy of land rights, and also an argu-ment for why we should temper her transformatory expectations, recognizethe complexity of what she sees as ‘social obstacles’ to women claiming land,and not rush to policy closure on land rights in all circumstances, or to blanketprescriptions. It argues for a renewed emphasis on reflexive ethnographicresearch with a focus on gender as social relations, on subject positions andsubjectivities, on the meshing of shared and separate interests within house-holds and on power residing in discourse as well as material assets.

Keywords: gender, land, property, development

The gender gap in the ownership and control of property is the single mostcritical contributor to the gender gap in economic well-being, social status,and empowerment. (Agarwal 1994a: 1455)

In the January 2003 issue of JAC, devoted to gender and land, Bina Agarwal’spaper ‘Gender and Land Rights Revisited: Exploring New Prospects via theState, Family and Market’ promises to examine why ‘gendering the land ques-tion remains critical, and what the new possibilities are for enhancing women’sland access’. In the course of this, she attributes a view to me that I would like toclarify; and I would also like to set out why I agree enthusiastically that genderingland questions is critical but, unlike Agarwal, do not see this as essentially reduc-ible to an argument for land rights for women.

I have no quarrel with efforts to increase opportunities for women to obtainbetter access to land, but I have doubts, based on a somewhat different analyticalstance, about whether land rights have quite the transforming potential for genderrelations that Agarwal invests them with, as indicated in the quote above.

454 Cecile Jackson

Agarwal’s paper makes it plain that she has been a passionate and effective advo-cate for land rights for women in South Asia and across the world, and she detailsat length the ‘catalytic’ effects of her 1994 book (Agarwal 1994b),1 which remainsa benchmark for any writing on gender and land. The JAC special issue has,however, prompted me to think through why my analytical disposition leadsme to different emphases and conclusions to Agarwal. I will first comment onthe question of women’s voices raised by Agarwal; then consider the limits tothe arguments she makes in relation to women’s land rights; and finally suggestanother approach to thinking about ‘gendering the land question’.

WOMEN’S VOICES

First, then, we may consider the vexed question of women’s voices and howthey are understood and represented.

Agarwal argues that ‘in many cases’ rural women are demanding land rights,but acknowledges that ‘there are also examples where women have not identifiedthis as a priority’ (2003: 189). I would reverse her implication that the latter arethe exception and the former the rule, but let us leave that aside for the moment.Agarwal goes on to quote a remark of mine, made in conference discussion,as an example of feminist academics arguing that women’s land rights do notdeserve policy attention, because relatively few women express a desire for land.She attributes to me the view that a ‘voiced concern (or its lack) is a sufficientindicator of needs and preferences, and an adequate basis for social policy’ (Agarwal2003: 189). This is a rather bizarre attribution, since I have been at pains to ques-tion such a view for some years now ( Jackson 1993, 1996, 1997) and recently( Jackson 2002: 503) explicitly criticized the World Bank Voices of the Poorbooks (Narayan 2000a, 2000b; Narayan and Petesch 2002) for epistemologicalinadequacies and failure to recognize the politics of speech. Many writers havedebated these issues in greater depth, i.e. issues of ‘false consciousness’, of howdoxa constrains expressed preferences, and the micro-politics of who says whatin particular circumstances and in relation to particular audiences (e.g. Raheja1996).

Reflexivity in research demands a continuous engagement with our ownpositionality as researchers and how this affects the research process and out-comes, since we all have intellectual (as well as emotional, social, political andeconomic) investments of one sort or another that potentially intrude on theanalytical interpretations we make. Agarwal’s paper illustrates this, and in theexamples that follow I am not arguing about the veracity of the statementsmade, but pointing to the dangers of ventriloquizing what ‘poor women’ say, ifwe as researchers are not vigilant in our reflexivity and our authorial practices.How quotation marks are used is one such instance. Agarwal puts into quotationmarks, and therefore directly into the mouths of women, a statement by poor

1 Cherryl Walker (2003), however, criticizes the wholesale adoption of Agarwal’s arguments forSouth Asia in South Africa.


peasant women of West Bengal in 1979, which is actually footnoted as a 1992personal communication from a fellow researcher (2003: 186). Another issue ishow the identity of a speaker is represented. On page 215, it is not clear if thespeaker is a member of DDS (the Deccan Development Society) or one of ‘thewomen’; words like ‘internalized’ in the quote suggest the former, but we arenot told, and the elision of the voices of organization staff or volunteers withthose of project participants is obviously problematic.

Agarwal says, in her argument against the emphasis on women’s voices as asole basis for social policy, that women ‘might shape what they reveal abouttheir priorities according to how an interview is conducted’ (2003: 189). Exactly.Yet she then reports herself (2003: 190) talking about land inheritance to ruralwomen in South Africa, and following a question about titles with ‘wouldn’t itbe better if the land was in your names or in joint titles since you are the onesfarming it?’, which surely illustrates exactly how the nature of a question signalsvery clearly the answer expected by the questioner. The evidence derived in sucha situation is compromised by leading questions that indicate what answer thequestioner would like to hear. But at the other end of the spectrum, the evidencefrom a very open question, such as ‘what constrains your farming?’ may becompromised by the taken-for-granted impossibility of land titles for womenobliterating an expression of interest in land titling. Somewhere between theselies a question such as ‘does it matter whose name is on the land title?’

But actually none of these is satisfactory if they are simply one-off encountersof the sort that sadly passes for adequate research ‘findings’. The real lessonabout research methods that I want to emphasize here is that getting underneaththe political inflections of talk requires longer-term work that builds up relationswith those we are seeking to understand. Only then can the contradictions oftalk in different circumstances, the content and meaning of talk versus thoseof observation and action, story and myth, and the triangulations possible withspeech from a range of actors, be brought to bear on the research questionin mind. Raheja gives an excellent example of the ways that context affectedresponses of women to questions about the desirability of emulating Sita, andconcludes that ‘we must regard the words of our interlocutors not as fixed andreified and essentialized mirrors of consciousness, but as shifting and purposefulnegotiations of identity and relationships which take different forms in differentsocial contexts and before different audiences’ (Raheja 1996: 170).

WILL LAND RIGHTS TRANSFORM THE POSITION OF WOMEN?

In this section, I begin to set out the doubts I have about the claims made for thetransforming potential of land rights, following the arguments in Agarwal (2003),where she justifies the call for land rights for women in terms of welfare, effi-ciency and empowerment gains for both women and men. These are then elabor-ated in the following section (‘Gendering Land Questions’), which also indicatesother starting points for ‘gendering the land issue’.

456 Cecile Jackson

The Scale of Transformation

The extent of public land available for distribution to landless rural women inIndia is very limited, and it can be argued that the substantial proportion ofwomen within landless rural households, generally considered the poorest, alsocannot benefit from any intra-family or intra-household land redistributiontowards women. Agarwal talks down the extent of effective landlessness, i.e.those households with no land or such small amounts as to be of marginal value.But even if one accepts her figure of about 20 per cent of rural Indian householdsas non-cultivators in 1992 (2003: 202), it is clear that there are many cultivatorsfor whom land holdings are miniscule. Most of the poorest rural women live ineffectively landless households, without natal family or marital land to contest,and the value of the land that is held may not be great, which is clearly a factorin weighing up the gains from such contestation. Furthermore, the pattern ofchange suggests increasing landlessness over time.

Agarwal, however, argues that ‘[a]s agriculture gets feminised, an increasingnumber of women will be faced with the prime responsibility for farming butwithout rights to the land they cultivate’ (Agarwal 2003: 195). She predicts agender divergence, with men shifting to non-farm livelihoods, that ‘an increas-ing number of households will become dependent on women bearing the largerburden of farm management’ (2003: 14–15), and that therefore women shouldown or control the land they cultivate. But Agarwal omits the word ‘landed’before ‘households’ – an important point when we consider what is meant by theterm ‘feminization of agriculture’. Land is worked by women under differentsocial relations – as labourers, as own-account household labour and as farmmanagers, to name just a few. The feminization of agricultural wage labour isnot the same as the feminization of farm management; the census data used referto the former and the case being made for women’s land ownership refers to thelatter. The former may be taking place in India, but different data would beneeded to establish the extent and the suggested increase in the latter. In landedhouseholds, the pattern of gendered agrarian transition may involve withdrawalfrom field-based agricultural labour, which may mean that such women become‘farm managers’ or it may not. In either case, the ‘land to the tiller’ justificationwould not hold for women who manage farms by managing hired labour. Aclass analysis is essential in understanding what these processes of change meanfor differently placed groups of women. Of course, one may still feel that womenin landed households who are not involved in field labour should have owner-ship of land, but the argument made by Agarwal here is on the basis of the rightsof those who actually work the land.

Thus, given the extent of landlessness, and the size distribution of land hold-ings, a redistribution of land to women in landed households may affect fewerwomen than the proportions Agarwal suggests. Treatment of the feminizationof agriculture requires a proper analysis of what working on the land involvesand of the diverse social relations of female labour so engaged, and data on thefeminization of agricultural wage labour are inadequate for the argument madeby Agarwal.


The Welfare Gains from Land Rights for Women

In criticizing the failure of women’s organizations to recognize the importanceof land, Agarwal writes that ‘working on land without rights in it meant a highvulnerability to poverty’ (2003: 187). She argues that land ownership reduces therisk of rural poverty, and provides direct and indirect benefits. This may be thecase for households and for men as individuals, but, as I argue below, it is notnecessarily the case for women, since they experience poverty in very distinctiveways, and are differently placed as subjects in relation to property and liveli-hoods. Her position seems to be that land is good for poor men, so it will begood for poor women.

Agarwal considers rural women in India to be, in the title of one of her papers,‘disinherited peasants’ (1998). But this is a very problematic formulation, sincethe idea of the peasant carries an implicitly male subject, which has not beenthoroughly reformulated to either integrate the distinctive situation of womenin relation to key definitional elements (e.g. social relations of production andreproduction) or to recognize how male ‘peasants’ are not unmarked persons,but inhabit male gender identities and a variety of masculinities, some of whichare hegemonic. Rural women in small-scale farming households are not peasantsin the same way as men, their property relations are distinct and ‘disinheritance’,with its implication of a seizure by men of women’s socially legitimate entitle-ment, by no means describes their particular subject positions or the genderedsocial relations of land. Like Agarwal, my starting point does not accept thelegitimacy of a status quo that excludes women from land inheritance withoutquestion, but it also problematizes, rather than assumes, the extension of argu-ments for why land rights are desirable for poor rural men, to poor rural women.

Assuming that the centrality of land to male poverty is necessarily the samefor women’s poverty is mistaken. It has been argued that in Africa land accessis not a major cause of the poverty of women. ‘[I]n many (but not all) areas insub-Saharan Africa, women’s access to labour and to cash or other resources tomobilise labour are more important than access to land. The generally very smallsizes of women’s farms reflect their labour and cash constraints and not lack ofrights’ (Whitehead and Kabeer 2001: 11). This perception is what lies behind thecomment of mine and others that Agarwal refers to (2003: 189). In India, theconnections between land and household poverty are dramatically different tomost African contexts,2 as will be those between land and the poverty of indi-vidual women and men, but I believe that these are research questions not self-evident facts. It is certainly arguable that the absence of personal rights in land isnot nearly as great a poverty threat to Indian women as other factors such as thebreakdown of the marital household through death, divorce or separation. Lifefor women outside of marriage is potentially extremely vulnerable and risky;widows suffer material and social deprivation; children are frequently lost onseparation or divorce; and natal families do not always welcome the return of

2 Higher population densities, levels of irrigation and land values, and a range of tenurial forms aresome of the obvious differences, as well as the much greater degree of landlessness.

458 Cecile Jackson

daughters from failed marriages. The potential gains from claiming individualland rights from husbands will be set against the potential threat to maritalstability. Walker argues that in Natal women are committed to household mem-bership in a context where households are ‘increasingly vulnerable to dissolutionin ways that do not necessarily enhance women’s life chances’ (2003: 46).

Agarwal also maintains that, because of intra-household inequalities, womenshould have individual ownership and control of land, and presents individualownership as the ideal form of property control. She goes on to say that owner-ship of land by women would benefit child welfare more than ownership bymen. Again the evidence needs to be carefully used. In considering the impact ofasset ownership on child welfare, much depends on what kind of asset is underconsideration, and whether the beneficial effect of a mother’s home garden, forexample, is additional to a father’s fields, which would be unsurprising, orinstead of it, which would be a completely different situation. This matters. Theallocation of unused or underused land from the state to women for vegetablegardens would likely produce benefits for child welfare, but the transfer of landfrom within the family may not do so if the production by women does notexceed that of alternative users of the land, and child welfare could suffer.

The Efficiency Arguments for Women’s Land Rights

This brings us to the efficiency arguments. ‘[E]nhancing women’s land rightscould increase overall production’, since production inefficiency is associatedwith tenure insecurity and women with land rights and control of produce wouldbe motivated to ‘put in greater effort and investment into the land’ (Agarwal2003: 195). There is no Indian research on this, and I am not aware of any claimthat women with land titles in Bodhgaya (the 1979–83 land struggle in Biharwhich resulted in some landless women gaining title to land, and to which Agarwalrefers) have greater yields than comparable men. Agarwal, however, refers to aKenyan study (cited by Elson 1995) ‘in a context where men and women cultiv-ated both separate and joint plots’ to show that ‘disincentives can exist equallywithin the family, in relation to land rights’ (2003: 195), since a new weedingtechnology raised production on women’s plots by 56 per cent and men’s by15 per cent. I am not disputing the incentive effect of individual ownership, butit seems curious to imply that motivation is the essential missing factor, and thatwomen somehow could produce more by just trying harder, when the studiesused by Agarwal emphasize that the lower yields on women’s fields are clearlydue to lower levels of inputs, especially manure and fertilizers. In a situationwhere a de jure woman-headed household has secure land rights and outputcontrol, I expect that effort and output would be maximized, compared to asituation in which such a woman has insecure rights and uncertain control, andthese are good grounds for the land rights argument. But to propose that transferof land from male to female ownership within a landed household is justified onthis evidence is another matter entirely. If one assumes from her earlier state-ments that Agarwal is thinking of transfer of independent land title for women


not joint title, then one needs evidence that the incentive effect for women exceedsthe consequent disincentive effect for men, in order to produce efficiency gains.

Agarwal also says that titles would improve women’s access to credit, whichis indeed valid (although credit systems not reliant on collateral are now com-mon), and this may allow women to obtain the agricultural inputs that constrainyields. But when she suggests that ‘women might use resources more efficientlythan men’ (2003: 195), I am not convinced. The Udry et al. (1995) research inBurkino Faso used to make this point (separate women’s plots compared to maleplots within the same household) points out that yields are lower on women’sfields. ‘On average yields are about 18% lower on women’s plots than on similar[controlled for topography, soil type, distance from compound] men’s plotssimultaneously planted to the same crop within the same household’ (1995: 412).The explanation is lower use of inputs on women’s plots, especially labour andfertilizer. Logically, efficiency can be increased by either transfer of women’splots to men or men’s inputs to women’s plots, but I find it hard to see howAgarwal concludes that this study supports efficiency arguments for land rightsfor women. The significance of the Burkino Faso study is that land access forwomen within households where they have lower access to labour and fertilizersthan household men may well be associated with an overall reduction in house-hold efficiency. The meaning for India is far from clear, but it could easily beread to suggest that unless women are certain to have equivalent access to inputsas men, then the outcome of land transfers to women would damage householdlevel efficiency (also see Whitehead and Kabeer 2001).

However, whether such efficiency arguments are particularly meaningful isdoubtful. It is fiendishly difficult, methodologically, to attribute efficiency gainssimply to land rights, since other important factors intervene in a comparisonbetween yields on women’s and men’s fields, and separate and jointly cultivatedfields. These include not only the very significant control of outputs, but also theland quality and location, different crop choices and inputs levels, including,crucially, labour. Furthermore, a comparison of women’s and men’s plots withinhouseholds as if they are entirely separate enterprises is mistaken. For example,Agarwal notes the higher value outputs produced on women’s fields in BurkinoFaso, and indeed crops grown on women’s fields in West Africa, will often begroundnuts, bambara nuts or vegetables, by contrast to lower value staples onmen’s fields. But this does not mean that a transfer of land from household mento women would mean greater production of higher value ‘women’s crops’,given the household need, and markets, for staples. And the reason why womenare able to grow higher value crops on their fields is precisely because otherjointly cultivated fields produce staples: i.e. separate production by womenis enabled by joint production, and the boundary between the shared and theseparate should not be overdrawn. Finally, ‘the sub-optimal use of agriculturalresources . . . might make sense when evaluated as the management of complexfamily relations with positive spin-offs in the enterprise as a whole’ (Whiteheadand Kabeer 2001: 10), from which women benefit as members, particularly inrelation to food consumption.

460 Cecile Jackson

Efficiency arguments for women’s land rights are uncertain and make a numberof heroic assumptions. Where women lack proper access to the other inputsrequired for effective agriculture, land transfers to women within householdswill not enable a rise in household agricultural production. Whilst all may agreethat this precondition would be desirable, equivalent access to labour, ploughingtechnologies, manures and fertilizers, seeds and knowledge, crop storage andprocessing technologies – and most fundamental of all the cash to invest in theseinputs – is so distant in India today for the great majority of poor rural women asto be a vanishingly small possibility for the short to medium term.

Equality and Empowerment

We are told: ‘[P]ossessing land (especially titles) empowers women and placesthem in a stronger position to demand their due’ (Agarwal 2003: 196). I think itlikely, indeed, that ‘land rights can make a notable difference to women’s bar-gaining power within the home and community’ (2003: 197) although I might berather more tentative about this as a possibility rather than an inevitability, sincethe process of acquiring such rights may prove disempowering. Here I ask whetherland has been a major theme in women’s movements, what it means when itemerges as a theme of other political actors and whether the Bodhgaya experi-ence substantiates the empowerment claim for land rights.

Whilst acknowledging the complexity of interpreting ‘women’s voices’, itremains relevant to consider how far rural women are mobilizing around landrights. The reasons why land has not been prominent in the demands of women’smovements of many countries are diverse. In Brazil, rural social movementswere predominantly organized around the landless and unions, and demandsfocused on labour and social rights which were relevant to all women (Deere2003). In India, women’s movements have very rarely demanded land. Thosethat have been movements of women mobilizing as women rather than as landlesslabourers, dalits or adivasis have arisen around issues such as dowry murder,custodial and landlord rape, sati, alcohol and domestic violence and labour issues.The anti-alcohol movements are probably those which have been the closest tospontaneous protest and had the most concerted and widespread support fromwomen, yet they have had comparatively little attention. In Manimala’s wonder-ful account of the Bodhgaya movement (1983), she shows how the agenda of themovement was shaped by the debates amongst activists of the Chhatra YuvaSangharsh Vahini (a youth organization fighting for ‘total revolution’) and howland rights for women emerged out of mobilization around a number of otherissues, domestic violence, anti-alcohol, education and land distribution to landlessmen from a large landholding Math (monastery). Bodhgaya is quite rare, becauseland rights for women have not been common grounds for protest, and becausethey succeeded in getting some individual titles allocated to women, butManimala’s account makes it clear how difficult it was, even when dealing withland title to landless women rather than transfers from men to women withinhouseholds, to obtain some support from local men for women getting land title


in the redistribution of the 1000 acres, and how fiercely the women activists hadto struggle with male activists within the movement to win their argument.

Has the acquisition of land titles been empowering in Bodhgaya? Bodhgayainvolved a prolonged struggle, initiated and intensively nurtured by residentyouth activists, of landless men and women against a monastery controllingthousands of acres of land, exploiting labourers and sexually abusing womenlabourers. The demand for separate titles for women emerged when landlessmen given title to land became drunken and violent, provoking their wives todemand titles in their own names. However, those who received land said theywould leave it to their sons – and thus land ownership will revert to male controlbefore long.

The general absence of land demands from women’s movements in India isrelevant to the debate about how rural women see their own situation, but itshould not be overinterpreted because what gets on to the agenda of such move-ments also reflects what leaders, often from other social and physical locations,think that women want and need. The agenda is interactively formed betweenmovement leadership and membership, and there is still far too little institutionalethnography of such movements for it to be clear what derives from the precon-ceptions of leaders and what emerges from members’ priorities. Omvedt arguesthat the gap between leadership and membership of women’s movements wasgreater than in the dalit or peasant movements because, unlike in these cases,there simply was not an early generation of educated rural women to take onleadership roles – which were generally performed by older, upper caste highlyeducated women who ‘often sang of “going to the women of this country”,much in the same way as high-caste Nehru had written of his “discovery of India”– coming to the women from outside; they exhorted women to “join the strug-gles of the toilers”, as if they were talking to non-toiling women’ (1993: 79–80).

An example of land rights for women emerging into broader political debatewas in 1991, when Sharad Joshi of the farmers’ movement announced a policy ofencouraging farmers to gift land to their wives – known as Sita lands – to be usedfor low input agriculture producing subsistence crops (Omvedt 1993: 231). Butas Brass observes, the gifting of land to wives is useful to men as a strategy foravoiding land ceilings (1994: 35). Where land rights for women appears on apolitical platform, it is not necessarily women who put it there, or women whoare the principal beneficiaries.

However, faced with a fairly sustained indifference to land rights, one mustultimately consider why the logic of land for empowerment is not more attract-ive? Understanding land struggles is only partly about researching mobilizationand movements, however, for just as important, if not more so, is the ethno-graphic study of individual land disputes and conflicts, and how land fits into theeveryday business of struggles for resources. This raises the important issue ofthe shared and the separate interests of women and men in households, which Ireturn to below.

Overall, I support policy initiatives that strengthen the ability of ruralwomen to make land claims, such as legal education, reform of the judiciary and

462 Cecile Jackson

governance, public awareness campaigns, equality of treatment in settlementschemes and land allocation processes, equal access to agricultural credit andinputs, and so on. But I do not believe that land rights will prove to be the singlekey that unlocks the subordination of women. The arguments for such an expecta-tion are based on evidence that is too macro and too limited to capture meaning-fully the extent to which such rights will reliably increase welfare, efficiency andempowerment for women and their households, and there are conceptual reasons(set out below) that suggest that land rights for women may prove a bloodierbattle than Agarwal imagines, with less widespread support from women than shesuggests and more uncertain gains. This does not mean that it not worth pursuing,but it does suggest more modest expectations, a policy approach that is enablingand context-specific rather than prescriptive and generalized, and a research agendathat remains far more open than might be imagined from reading Agarwal. Thenext section addresses this last issue by indicating briefly what such an agendamight look like.

GENDERING LAND QUESTIONS3

Bina Agarwal discusses a number of kinds of land relations and scenarios, andmakes essentially the same case for all of them – that land rights for women aregood for efficiency, welfare, equity, empowerment. However, the renting ofland by landless women, or the allocation of wasteland to landless women, orthe titling of houseplots in women’s names, or the inheritance of family land bydaughters in landed households, all seem to me so deeply distinctive in theirsocial relations that they require, in addition to her broad-brush treatment, fine-grained, long-term study to reveal how they are gendered, what changes areunderway and what interventions may be promising. A gender analysis of ‘theland question’ needs both more disaggregated research questions and a moreopen analytical frame in which all roads do not necessarily lead to the singulardestination of the justification of land rights. I sketch in below some analyticalstarting points for such a project.

First, it is an old axiom, but nonetheless true, that land relations are socialrelations, and this section is mostly concerned with arguing for a stronger focuson social relations, and therefore inevitably also on social change. The largerpicture of how land is situated within livelihoods and what is happening to landrelations in the particular agrarian contexts in question needs proper considera-tion in order to understand how they are gendered (e.g. see Stivens 1985;Mackenzie 1990; Kandiyoti 2003). Secondly, it is important to disaggregate theterm ‘land’. The social relations that inhere in homestead land and gardens arevery different to those of intensively cultivated infields or lowland paddy, orextensively cultivated dry uplands, or land with permanent tree crops. Land ofdiffering value, location, soil type, topography, as well as land with differing

3 Agarwal’s 2003 paper considers prospects for women’s land rights via the state, family and markets,but I am mostly concerned with the issue of land relations within families.


tenure and production relations – owned jointly or individually, inherited, pur-chased, rented, borrowed or share-cropped – will have distinctive kinds of socialrelations, norms and discourses that pattern their use. For example, gender analystsin Africa have been especially interested in the gendered meanings and struggleswithin cultural designations of fields for household use and those for individualpersonal use (e.g. gandu and gayauna in Hausaland, maruo and kamanyango in theGambia), and for South Asia we need a comparable level of analysis of how thesocial relations that inhere in particular kinds of land (uplands, paddy lands, barilands) are gendered.

Men, Masculinities and Land

Male gender identities in relation to land remain little analysed, yet are verysignificant for understanding women’s land relations. Aspects of male identitiesof particular interest are lineage and caste ideologies, provider roles and strugglesbetween groups of men for land as both a material and a political and culturalasset. These seem to me essential for understanding the location and character ofmale resistance to women’s land rights.

The links between land ownership and lineage strength in Indian studies havebeen established for some time (see, for example, Shah 1993: 132), but have notbeen discussed in terms of what this means for how masculinities are experiencedby rural men, for whom land speaks of both descent and masculinity. Owningland is also central to particular caste identities: for example, Bhumihar men insouth Bihar are enormously proud of their caste identity as farmers, their statusas land owners and even their reputation for land grabbing. Their particularvariant of hegemonic masculinity puts land ownership in a place of great sym-bolic significance in the achievement of successful adult manhood. Being a suc-cessful man involves land ownership for some rural castes, in a way that women’sidentities very rarely seem to require.4 Disa Sjøblom shows how Bhil men inRajasthan, even when they have migrated to work in Gujerat, maintain theiridentity and community membership through continuing to return to cultivatetheir fields long after this is economically rational (Sjøblom 1999).

Male identification with provider roles and expectations are also connected toland and masculinities. Provision is not only about the production of a daily foodsupply, but also about intergenerational provision of heritable land assets forsons and dowries for daughters, since dowries are frequently financed indirectlythrough surplus agricultural production or directly via land sales.

Finally, land is part of relations of solidarity and competition in the inter- andintra-household relations of men. Marilyn Strathern (1988) has questioned theauthorship of some aspects of patriarchy, and shown how the position of womenis not necessarily motivated by a desire to subordinate women, but can be aneffect of competition between groups of men. Nitya Rao (2002) finds some

4 Matrilineal societies are much rarer than patrilineal, and even here land is generally controlled bymen, even if inheritance is traced in the female line.

464 Cecile Jackson

support for this in her gender research on Santal land relations, where strugglesto assert the rights of adivasi men, against land alienation to male Hindu incomers,have involved the assertion of a version of traditional community land relationsthat, by necessity, erases divisions within the community – including those ofgender – and individual women’s rights. Thus male solidarity based on ethnicitycan have implications for women’s access to land.

Inter-household relations between men can also leave women caught in thecross fire, because of competition between men differently positioned within alineage, for example where primogeniture differentiates son’s rights to father’sland. Even when there is land sharing between sons on inheritance, in a contextof land scarcity a kind of de facto head of the land-owning corporation maydevelop, signalling differential power of men to position themselves in relationto land rights. This issue of fragmentation illustrates the complementary differ-ence between Agarwal’s perspective and mine. She is rightly concerned to dis-miss the fragmentation argument against land rights for women (2003: 196), butall-India indices of fragmentation conceal the interesting social processes under-lying land divisions and consolidations, and their gender implications, that areonly uncovered by the careful forensic work of researchers studying land recordsover time. Both Sjøblom (1999) for Bhils in Rajasthan and Rao (2002) for Santalsin Bihar find that under conditions of increasing land scarcity in the twentiethcentury, registered title holders are no longer those who are de facto using andcontrolling land, but individuals representing land-holding kin groups. Whatthis means for gender is that the masculine descent group may consolidate itspower over land, and develop new narratives of kinship-based entitlement, thatexclude other kinds of entitlements based on marriage that have in the past beenthe basis of women’s land access. In other words, what is interesting aboutfragmentation is not the issue of efficiency, but that of how changing land avail-ability, commoditization and other trends, are mediated through social relationsof kinship and marriage and produce gendered effects in land access and control.Making a related point in her study of gender and land in Tanzania, Yngstromobserves that over the course of the twentieth century ‘where male identity hasbecome increasingly more significant as a channel for men to gain access to landthrough the lineage, gender identity [for women] has become more significant asa way of excluding women from inheriting lineage land’ (1999: 262).

Lastly, within households, the hierarchies of men mean that many men inlanded households lack the secure rights to land that they desire. The ideal Hausadomestic groups in northern Nigeria used to be the gandu, where the head of agroup of households (a compound head) controlled collective food production onjointly farmed gandu fields, and junior males (and in ‘pre-Islamic’ times, women)were allocated private gayauna fields for personal cultivation and control. The gandudeclined over time with the decline in the ability of compound heads to controlenough land to sustain the allegiance and labour obligations of junior males. Manymen may have much more precarious, contingent and anxious relations withland than may appear if we lump all men together, as the inevitable beneficiariesof patrilineal inheritance (see also Cooper 1997 for the Hausa in Niger).


In both her account of her intervention with a South African NGO, whereland rights for women was taken up enthusiastically and ‘men’s resistance wasneither that strong nor insurmountable’ according to a staff member (Agarwal2003: 191), and of the Deccan Development Society (DDS) in Andhra Pradesh,which established collective farming by women on leased and sometimes pur-chased land, and where a staff member says ‘When women acquire land, there isa win-win situation for everybody’ (Agarwal 2003: 215), Agarwal minimizesmale resistance. Certainly the DDS situation of land leasing for landless womenmay not be heavily contested, but the redistribution of land within householdsand families will mean that men in various subject positions – sons, brothers –will be losers, and contestation of change will be likely.

If we focus on gender as a relational form of analysis, then we are interested inthe interactions between women and men, both as they are structured by socialnorms and institutions, e.g. kinship and marriage, and reconfigured by indi-vidual agency over time. How then does land control enter into these genderrelations? How are they changing? How are the identities and subjectivities ofwomen as actors implicated in land relations? I want now to suggest two argu-ments. The first is that the social relations of women in distinctive subject posi-tions, and with subjectivities in which land has quite particular significations,critically differentiates their land relations, and that these shift over life coursesand historically. The second is that analysing gendered land relations requires afocus on social change in order to understand how all these elements cometogether to produce trends and outcomes.

Distinctive subject positions and subjectivities

Gendering the land question involves a fundamental rethink of how propertyrelations are reconfigured by a female subject (Whitehead 1984). I see too little ofthis in Agarwal’s work, which considers kinship and marriage in the course ofher comprehensive survey of gender and land relations in south Asia (1994), butdoes not follow through the implications into the rest of her analysis.

She wonders ‘what obstructs women from realising their claims in familyland? The obstacles are partly legal and in part social and administrative’ (2003:203). I would give much more emphasis to what I think she means by the socialhere, but also would see the social as sets of relations which are not only obstruc-tions, but subject positions which may be imbued with desirable attributes aswell as serving as impediments to land claims.

For clarity, I divide this sub-section into arguments for taking account of thesubject positions of women,5 which I use to indicate the particular locationswithin social structures and discourses occupied by women; and argumentsfor an equal attention to subjectivities, i.e. to that mental interior that is bothpsychological and social. An individual’s subjectivity ‘is an amalgam of both

5 The term ‘subject positions’ implies a much more complete determination of social action bysocial structures than I subscribe to.

466 Cecile Jackson

psychological dispositions developed out of their unique biographical circum-stances and the habits, customs and orientations that reflect their involvement inparticular social groupings’ (Layder 1997: 27).

The subject positions of rural women are many and various, but patriliny, asthe commonest form of descent, creates some uniformities. For example, a con-ventional distinction between men’s and women’s land rights in Africa is that theformer have primary and the latter secondary rights, because men have kin-based rights legitimated by patriliny, and women have marriage-based rightsgrounded in the conjugal contract. Whitehead and Tsikata (2003: 97) questionthe value of this distinction. The formulation, indeed, is too crude, but it doesnevertheless try to capture something of the gender difference in kinds of tenure.Both patriliny and the conjugal contract are social relations, of course, and sub-ject to interpretation and contestation, but inherited male rights are possiblyfirmer and less open to discretionary interpretation than marital ‘rights’. There isclearly variation here. Yngstrom (1999: 267) argues from her Tanzanian researchthat women’s land access within marriage is considerably more secure and rigidthan in, for example, the Cameroonian case noted below, and male lineage rightsare also contested. But the latter often appear more ascribed, whilst a woman’smarital rights appear more conditional, since they have more obviously to beearned, evaluated and achieved.

The analysis of gendered land relations in particular social contexts requiresconsideration of the diverse subject positions of women, the relations they involve(e.g. with fathers, brothers, husbands) and the implications of these in relation toland. The situation of daughters in relation to parental land is very different tothat of wives to marital land. A girl in a patrilineal society is positioned in herkinship group very differently to a boy, both in terms of the inheritance ofproperty but also of the descent ideology. A gender analysis of land is interestedin both of these factors, for both women and men. In the Indian context, boysmay be held to be the rightful inheritors of land, those who remain in place andcontinue the line, whilst girls marry out and reproduce other lines, but areentitled to dowry in order to set them up in favourable households. Dowry isnot of course proper pre-mortem inheritance, since it is controlled not by thedaughter herself but by her husband/husband’s family, but the cultural logic isclear – land for boys and dowry for girls are the appropriate ways to establishone’s offspring.

The variations in extent of inherited land amongst daughters range from 8 percent in northern India to 18 per cent in southern India, where constrainingfactors are weaker (a better legal position, village endogamy and the absenceof seclusion), with an overall average of 13 per cent. However, 51 per cent ofwidows of landed men inherited from their husbands (Agarwal 2003: 202). Thisdoes suggest that marriage is a more successful need claim, for women, than theparental contract (see also Rao 2002). The implication is that entitlement tosupport from a husband, or his family, is a more reliable and less risky basis forgaining access to land, and that investment in a reputation as a deserving wifeis potentially valuable. Hypergamy – i.e. the ideal of women marrying up into


better placed households than those they are born into – also will tend to meanthat for women there are more resources potentially available in marital house-holds than natal ones.

The simple presence of patriliny and patrilocal marriage is only a startingpoint for considering the web of social relations experienced by women in rela-tion to potential to land claims. Patrilocal marriage takes many forms – theextent of exogamy, the distance of movement from the natal home and thedegree of absorption into the marital home are all factors that influence the roomfor manoeuvre. Marriage in north India, which I generalize about here to simplymake the case for the wide variations in the experience and meanings ofpatrilocality, and its impact on property relations, is culturally constituted as amuch clearer rupture from the natal family, and more akin to adoption, with agreater loss of support and contact with kin for the marrying woman, thanpatrilocal marriage in many African cultures. In the north Indian context, it isshameful for a man to accept help from his wife’s family, wife-takers beingsuperior to wife-givers. Although being considered a stranger in the maritalcommunity is common to the experience of inmarrying women in African andsouth Asian research, in north India we see a combination of dramatic rupturewith the natal family, combined with simultaneously being a stranger amongstthe family one marries into, despite the formal ‘adoption’.

Living in a village some distance from a natal home where fields are locatedcreates a multitude of practical problems, since distance may make direct cultiva-tion impossible: cultivation through hired labour may be simply too expensive,and cultivation through kin relies on the cooperation of brothers or male kin,precisely those who are likely to have rival claims to such land. A mechanicalinterpretation of how elements of marriage practice affect ability to claim land isunwise though. Distance to natal home is unlikely to be significant in ability tomake land claims where the possibilities for independent cultivation of land areprimarily ruled out by powerfully patriarchal cultural contexts.

African women do not only acquire land via marriage (Whitehead and Tsikata2003: 78), but it is a very significant route compared to India where they seldomacquire land within it (given that widowhood ends a marriage). The same dis-course of need that recognizes widows’ claims in India, based on a powerfulobligation for men to provide for wives who are much more fully absorbed intomarriage and contact with natal families attenuated, is also the basis of a refusal torecognize the ‘need’ for wives to have land access whilst married.

Finally, different identities are inhabited simultaneously, which creates crosscutting interests. As a daughter, a woman appears to have the obvious interestsin claiming a share of parental property that Agarwal outlines, but as a wife shemay also be against the land claims of her husband’s sister, and as a mother shewill not necessarily support a daughter against the claims of a son. These inter-secting interests, which reflect the multiple subject positions occupied by anyone woman, mean that we should expect a diversity of opinions on the desirab-ility of land rights for women, and by no means anticipate universal supportfrom women, once the scales of false consciousness have fallen from their eyes.

468 Cecile Jackson

Writing about dowry in north India, Ursula Sharma observed that women asmothers-in-law and as mothers have conflicting interests in this institution andthat ‘Feminists are hampered in their efforts by the fact that in this sphere, as inso many others, property divides women among themselves’ (1984: 73).

I have briefly discussed above the question of women’s demands for land, butreturn to it here in an equally brief discussion of women’s subjectivities. BinaAgarwal has said that ‘Land rights are unlikely to be granted to women by mostcommunities, on any significant scale, without women collectively demandingand agitating for them’ (1994b: 477). Whether they do demand and agitate forland will turn to a large degree on their subjectivities, on how women feel aboutland and land claims. That is to say, it will depend on the ways in which thepersonal worlds of identity, memories and emotions, motivations and desires, thatare not simply given by discourses and are always in flux and emergent, impingeon social action. Indian women’s sense of themselves, their personhood, seemsto me rarely to implicate land ownership in the ways mentioned for men above,and I doubt that it holds a place in their subjectivities to rival the importance ofchild bearing, marital success, a respected occupation or social position, an achievedlevel of security, personal autonomy and so on, in this regard. It would beinteresting to know what kinds of associations different kinds of property havefor women, for example jewellery with its mobile and transferable character, andsignifications of love, rather than assuming they have the same passion for landas men.

Whilst clearly relevant to all areas of land relations, it is particularly import-ant to register the entanglements of subjectivities in intra-household and intra-familial land relations. Such relations are much more complex than relations ofemployer and employee, or buyer and seller, and involve levels of intimacy, andlong-term emotional investments that simultaneously confer costs as well asbenefits. For a wife who is part of a production and consumption group based onland owned by her husband, the collectivity of everyday work and life confirmscommitment to a marriage, whilst insistence on joint title signals distrust andmay seem to question the marriage itself. Intimacy, particularly sexual intimacy,with husbands establishes solidarity. In the north Indian case, this is in the con-text of efforts to assert the conjugal relation over patrikin relations (Raheja 1996:160) and thus a move to claim land from a husband may be problematic inrelation to the larger struggle. Whilst conjugal solidarity is in the process ofbeing established, there may seem to be significant risk in such a move and,conversely, where a marriage is working well for a woman and delivering mater-ial, social, emotional and sexual satisfactions there may be little interest in de-manding joint title or individual land allocations from husbands.

There are examples of African cases where women acquire property and leaveit to daughters and sisters (Lastarria-Cornhiel 1997: 1328), but in India it remainsthe case that where women obtain land rights they invariably wish to leave suchland to their sons. The motivation to make emotional as well as social andeconomic investments in sons is a powerful force in many women’s subjectivitiesthat needs to be acknowledged, since it is an important factor in the exercise of


agency. Conversely, the emotional motif of relations with daughters is one ofloss, captured poignantly, for example, in the departure songs sung by northIndian women leaving home on marriage for a socially distant place where theirparents will not be welcome. The repeated refrain of one, ‘Call me back quickly,mother, Beg with folded hands’ (Raheja 1996: 157–8), expresses this vividly.

The shape and emotional content of relations between siblings will be asvaried as those between spouses or parents and children, but a significant factorin subjectivity as a sister may well be the value of a brother as a protector, as wellas the closeness and concern for his wellbeing and his ability to make a goodmarriage. For a daughter to make a claim for parental land is for her to riskalienating brothers, whose share of land would be diminished, by reducing theirlevel of material wellbeing. Agarwal (1994a) suggests that brothers do not in factoffer their sisters much support, but here I am concerned with how the emo-tional investment in the desire for an ideal brother enters into women’s thinkingabout making land claims. Social relations with close kin are invested with mater-ial, symbolic and emotional significance that are not lightly put under strain, andthe consequences of breakdown in these relations are potentially severe. Raheja(1996) analyses the contradictions within the formal stress on the priority of awoman’s ties to her husband’s kin and yet the permanence of ties to parents andsiblings. The longing for natal kin expressed in songs and speech, and the experi-ence of ‘foreignness’ in the husband’s home, suggests an emotional charge torelations with natal kin that must impinge on decisions to claim land rights.What would be interesting to establish is how women who successfully claimparental land manage relations with brothers, and how much this is a factor forthose who do not. Is it the case that women’s land claims generally emerge insituations where relations between siblings are already breaking down, and theinvestments in such relations look to be forfeit anyway?

What does this admittedly speculative excursion into women’s subjectivitiesin relation to land suggest? For India, my guess is that many rural women’ssubject positions, lived experiences, emotional investments and culturally spe-cific subjectivities do not encourage desires and claims for land.

Gendered land relations and social change

Understanding this terrain requires more detailed studies of social change.One example is van den Berg’s Cameroonian study (1997) which shows howwomen’s strategies for securing access to land mainly operate via cultural con-tracts, particularly marriage, and involve managing marital compliance, success-ful reproduction, discursive manipulations to widen notions of access, andsometimes recourse to local authorities, i.e. the local state representatives, mayorand customary chiefs. Kinship and marriage are simultaneously the primary con-straints to secure land access, and the main means by which it may be achieved.Local authorities offer another avenue, as does direct land clearance, but thesesignal failures of marriage as the primary route to land – local authority access isless secure (van den Berg 1997: 245), and clearing your own land is an act of

470 Cecile Jackson

autonomy that indicates the imminence of divorce (1997: 263). In 1973 womenwere said not to inherit rent, loan or clear land, but 20 years later they weregaining land through all of these means (1997: 181), as well as applying to localauthorities for allocations. Yet even in dealing with local authorities and chiefs incharge of land allocation, corruption and the expenses of recourse to courtsrequired women to have strong bases of supportive male kin (1997: 205).

Another African case is covered in Han Seur’s research in a matrilineal com-munity of Zambia (1992). When land was farmed through shifting cultivation,matrilineal ideology did not apply to land, which was plentiful. But as it becamescarce, matrikin groups developed ideologies extending matriliny to include land,and thus a political system organizing descent was transformed into one forcontrol of material assets (1992: 370). Over time, however, men’s interestsin their conjugal units have come into conflict with those of matrikin groups,because in order to command the allegiance of sons, they need to assure them oftheir inheritance rights in their parental property. Fathers often wish to leavetheir land to their sons rather than their brothers, as matriliny dictates. A dis-course of entitlement based not on descent, as in matriliny, but on the rights ofthose who put the work into developing land (i.e. the conjugal group) hasemerged (1992: 346–55). These rival discourses remain in tension, for men maybe tempted to exercise their rights as members of matrikin groups, whilst also, asmembers of conjugal groups, wishing to see a different disposition of property.A respondent explains that if a man dies and his brothers claim his property, asthey are entitled by matriliny,

the children of these brothers will become distant with him [their fathers]or even hate him. Do you think these children will work with their fatherto develop their farm? No chance! Only when a father makes it very clearthat he does not want to follow these traditions, only when he leaves theproperty of his older brother to that older brother’s children, will his ownchildren assist him. Now looking at wives: in this modern world, it isnormal to demarcate fields, women having their own fields. Men whoreason have come to understand that . . . they depend on their own wifeand their own blood; their own children. . . . Many couples have found away to overcome this problem of matriliny, by pumping in some funds toa woman to enable her to maintain her own fields for the good of thechildren and herself . . . [I]t seems in this way that the matrilineal system isfully intact but in fact . . . it will be patrilineal full time, in the sense that ifI invest a lot of money in my wife I know I will enjoy this money while Iam alive, because she knows she started from my pocket, my relativesreceive less when I die, and the lot will remain with my wife and children.(Seur 1992: 272–3)

The position of women is interesting. One might expect them to be completelycommitted to matriliny (reading Agarwal), which appears to put women into abetter position in relation to property. However, the triumph of conjugalityover matriliny benefits them as wives, since they stand to inherit from husbands,


but threatens their control of children. If they seek to divorce and establish theirown farms, their children, of a strongly ‘conjugal’ unit, may prefer to remain withtheir fathers (where their land rights may appear more secure) instead of follow-ing their mother as matriliny indicates (Seur 1992: 372). Respondents felt thatwith the changing inheritance practices, children will soon only be prepared tohelp their mothers (or leave with them) if they are very wealthy or better farmersthan their fathers, and that eventually there may be both a drop in the percentageof women deciding to set up their own farming enterprises and in the divorcerate. Seur suggests that, therefore, a greater dependence and subordination ofwomen may result not from the ability of a man to control his wife’s labouror the fruits of her labour, but by a new alliance between a man and his children(1992: 291). Independent women farmers rejected some aspects of matriliny,such as the demands made by matrikin who were not willing or able to con-tribute to the development of their farming enterprises, and emphasized otherssuch as a mother’s control of the labour of children (1992: 288).

The necessary work on deconstructing the household, which gender analystsof development have focused on, may blind us to a bigger picture in whichthe conjugal household is of increasing importance in relation to other socialinstitutions of kinship. In African studies, the change from the idea of marriageas a contract between two lineages and towards one of a contract between twoindividuals (Seur 1992: 286) is frequently commented upon. What does this meanfor changing gendered property relations? And for India, are relations betweendomestic groups and kin groups changing? Are household forms and functionsshifting and what implications does this have for how women relate to land? TheHindu family operates at four levels: as household, as a group of householdsconstituting a property group, as a wider group defining the limits for ritualpurposes and as a dispersed genealogical group (Madan 1993: 420). It has beenargued for some time that nuclearization of the family is taking place (Beteille1993: 440) and, in a study of one location over 60 years in western Uttar Pradesh,that richer landed households are increasing in size whilst landless householdshave become more nuclear (Wadley and Derr 1993: 414). And in her classic workon kinship in a Tamil village, Gough (1993: 171–2) shows that Brahman lineagesare deep and unifying because they are based on land ownership, whilst the non-Brahman landless low castes have very shallow patrilineages, which are notstrictly patrilocal, and in which wives are not assimilated to their husbands in thesame ways but retain strong links to natal families, and to brothers in particular.Crudely, given that intra-household gender relations are quite different inextended and in nuclear households, it seems to me important to look at howwomen within larger land-owning households may gain material well-being atthe expense of personal autonomy, whilst those in smaller landless householdsare materially disadvantaged but more autonomous. The project of a genderedanalysis of land questions should be looking at how this changing social institu-tional architecture is both affecting women’s relations to land and how, conversely,gender relations impact on land – for example, how far nucleation is driven byintra-household conflict: ‘A major focus of discussion in Karimpur is the breaking

472 Cecile Jackson

up of families. When nyare (separation) occurs, it is a sign that the men could notreach consensus on economic issues and that the daughters-in-law were fighting’(Wadley and Derr 1993: 401). If women are less powerful in landed households,how likely is it that they will demand land or that husbands will relinquish it?We need more class-disaggregated gender analyses to unravel how class andgender interests may clash.

This kind of research, which situates analysis of gender and property relationsin the shifting relations between social institutions and actors in the everyday lifeof rural women and men, must be the basis of ‘gendering the land question’.Kinship and marriage should be the analytical heart of this project, rather than anaside on ‘social obstacles’ faced by women in their assumed need and desire forland, and the ways social relations of family and marriage, and personal subject-ivities both strengthen and weaken claims should be brought into the analysis.

In summary, I have argued in this section that thinking of gender in terms ofsocial identities and relations, rather than focusing on women as a category, caststhe gendering of land issues in a different light, and poses research questionsaround the diverse and overlapping subject positions of women and men inrelation to land, the importance of subjectivities in relation to desires for land,the ways household structures, lineage ideologies and marriage practices are experi-enced by different kinds of women, and the need to understand how women asactors make and are made by patterns of social change.

Shared and Separate Interests within Households

It is a central weakness of the Agarwal argument that she does not adequatelyrecognize that the household is itself a collective enterprise, and I return here toissues already touched upon above. Agarwal calls for independent land rights forwomen and collective cultivation by women from different households. If col-lective cultivation is desirable between women, why is it not so between womenand men of the same household? Presumably this is because of an assumptionthat relations between the genders are marked by exploitation and within gen-ders by solidarity. Nancy Folbre famously observed, in an essay on assumptionsin economics, that ‘There is a delicious political paradox in the juxtaposition ofnaked self-interest, which presumably motivates efficient allocation of resourcesthrough the market, with a fully clothed altruism that presumably motivatesefficient allocation of resources within the family’ (1988: 252). But Agarwal’sinverted focus on self-interest within the household and altruism in the extra-household collective farming by women is also problematic.

Who would disagree with Agarwal when she reiterates the point that ‘landaccess by men alone cannot be assumed to benefit women and children equit-ably’ (2003: 194)? For women in landed households may be subject to some ofthe worst forms of neglect as a consequence of the land/dowry/sex ratio nexusand may be worse off in this regard than women in landless households. How-ever, it is mistaken to assume a complete separation of male and female interestswithin a household, as she seems to. Access by individuals within a household to


household resources cannot be assumed to be equitable, and intra-household ill-being exists even where the household is not poor, but this is not to say thatthere is no relationship between a woman’s level of well-being and that of herhusband or father. The life of a woman married to a severely poor man willlikely be materially more circumscribed than one married to a less poor man,parents seek better-off husbands for daughters, and so on. In other words, withina household, women may not get their full due of access to resources, but thisdoes not mean that they are not materially advantaged by marriage to a wealthierman, and that they have no interest in their spouse’s level of wealth or landownership. Many women have a stake in their husbands performing the role ofprovider effectively, and may, for good reason, not wish to see this challengedor disrupted. And this is part of why many women in landed households maynot judge it in their best interests to seek joint or individual title to marital land,whilst a marriage exists. The interests of women and men within marriage areboth joint and separate, which is why gender struggles are so complex and whywomen may not act as might be expected from an analysis based on the idea ofwomen as a separate category with interests entirely separate from men’s.

In addition to this account of how women’s self-interest, in landed house-holds, may lead them to buy in to the status quo of male ownership of land, thequestion of ‘altruism’ in the household remains. Well-being tends to be conceptu-alized in a very individualistic manner, and Amartya Sen’s much quoted remarks(Sen 1990) on how Indian women lack a strong sense of personal well-being –obliterating personal well-being, seeing it in terms of the well-being of familymembers – has signalled a new emphasis on self-interest. But the experience ofpersonal well-being is also situated in social relations that are meaningful in ourlives, for our subjective experience of well-being is profoundly affected by thewell-being of those we care about and depend on (Sugden 2000). This worksin two ways: for pragmatic, instrumental reasons (the well-being of one youdepend on is a means to personal well-being) and for altruistic and emotionalreasons. We should not be so willing to represent social relations between spousesas simply contractual arrangements, because many will be long-term relations ofintimacy, marked by affection and regard. Both uncritically extending westernnotions of romantic love into other cultural contexts, and assuming relations ofemotional commitment are entirely absent, are equally mistaken.

Gendering the land question involves a critical consideration of shared as wellas separate interests, labour and well-being, in relation to a range of kinds ofcultivation and ownership of land. Shared interests and shared consumption ofcollectively own-produced food are clearly central to rural women’s lives infarming households (see for example Whitehead and Kabeer 2001: 6). Back-grounding shared interests can therefore underestimate the extent to which womenhave rational commitments to household arrangements that appear gender in-equitable. Walker, in her South African work, argues that ‘the focus on individualrights for poor women needs to be tempered by a deeper appreciation of theimportance of household membership in poor women’s lives . . . While a minority[of women] were interested in the idea of independent rights in land, delinked

474 Cecile Jackson

from that of their husbands or families, few saw this as the solution to their prob-lems. They were more interested in mechanisms for securing, even extending,their rights within their households, including through such mechanisms as jointtitle and individual copies of title deeds’ (2003: 143). Naila Kabeer (personalcommunication 2003) also remarks, in relation to Nigera Kori, Bangladesh, anorganization that Agarwal cites in support of her own position, that land remainsthe most important single issue for its members, but it is most important in areaswhere there is government land available. She remarks that ‘it was a collectiveprocess of reflection that has led landless women to view land rights for womenas important, [and] my own view is that their struggles – and I found women tobe as fierce as men about the struggle for khas land in Char Jabbar – are to ensuretheir families’ access to government land. Joint or sole entitlements may be apriority with widows etc. but there too so that they have land to pass on to theirsons’.

With a more nuanced approach to separate and shared land management andcontrol within households, where household members both cultivate fields jointlyand have fields that are separately cultivated, how then do we understand therelationship between jointly cultivated land and separate fields? Does the desirefor the separate come into play when the shared is failing? Does the voicing ofa desire for the separate constitute a challenge to male providers? The separatefields farmed by women in many African contexts should not be seen as an auto-matic entitlement, but as a conditional part of a conjugal contract. In a detailedrecent ethnography of gender and land in northern Cameroon, van den Berg(1997) shows how, even where separate portions for women’s personal cultiva-tion is commonplace, it takes some years of successful marriage, the productionof children and skilful management of marital harmony, before women cansecure such fields. Women here also acquire land through independent clearing,but this form of access signals marital discord. Thus comparisons between men’sand women’s productivity on joint and separate fields is methodologically ex-tremely complex because personal and joint fields are part of a package deal, andthe women with personal fields are simultaneously successfully managing jointconjugal production.

Han Seur (1992: 269) finds, in Zambia, that the development over time of a‘separate fields for women’ practice is due to many reasons – women’s desire forindependent income, men’s desire to insulate themselves from the demandsof their matrikin, and polygamy. He observes that intra-household tension isimplicated in the farming of separate fields, but not always from the women’sside (1992: 277), involving friction and trajectories towards divorce, but also adesire for monitoring of work levels (1992: 275), which is possible on separatefields. He also comments on cases where separate fields are initiated in order tomaintain an impression of separation (to limit demands from matrikin), butmerger of yields takes place later, and in such a case wives and husbands arecolluding to establish joint conjugal interests under cover of apparently separatedkinship interests. The gendered meanings of land controlled independently bymarried women within households are clearly very complex and not simply an


indicator of women’s autonomy, or of their power and influence as wives, or oftheir success as farmers.

Shared and separate interests, assets and well-being mesh in complex waysthat are violated by an assumption of autonomous individual self-interest. Bal-ancing the recognition of this with a critical awareness of how outcomes thatdisadvantage women are produced by the discourses of joint interests, sharedutility and collective well-being, which prop up an inequitable gender order, isthe challenge for a gender analysis of land.

Concepts of Power

‘Land defines social status and political power in the village, and it structuresrelationships both within and outside the household’ writes Bina Agarwal (1994a:2). She goes on in this passage to say that ‘Instead employment is taken as theprincipal measure of women’s economic status, obscuring what has been com-monplace in measuring the economic status of men or of households: propertyownership’ (original emphasis). But why should what defines men’s status andpower also define women’s? And why should a foundational aspect of classanalysis automatically play the same role in relations of gender? Gender relationsbetween women without land, and men with it, are not a mirror of class relationsbetween men without land and those with. And the meaning of being ‘landless’for women is very different to that for men.

Agarwal’s expectations of the transforming potential of land rights stems froma view that power relations are ultimately determined by assets (for a critique seeKapadia 1996: 171). But far from all power and influence in gender relationsresides in material advantage (Moore 1992; Fraser 1989), and also what is appar-ently material, such as land as a physical asset, has other significations that arepart of gender ideologies. As Strathern puts it ‘wealth symbolises aspects ofpersonhood’ (1984: 171).

To take the first point: rights to land and needs for land have to be establishedthrough discursive struggles to validate, define and interpret these rights andneeds. A range of idioms, vocabularies and arguments will be used for thesepurposes, and their effectiveness will ebb and flow over time as different actors –husbands, wives, state officials, NGO workers, political movements – invokethem to take forward their own interests (as well as those they have investmentsin) (Fraser 1989). Norms, ideas, values and words are resources that can deliverpower just as surely as ownership of means of production or material assets.Bargaining positions and material outcomes are not simply determined by owner-ship of assets. Similarly, the ownership of assets does not by itself confer power. Awoman holding independent title to land certainly has the potential for strengthen-ing her position in a range of social relations, but this also depends on the extentto which she can secure cultural interpretation of this ownership as legitimateand appropriate. If she cannot, then she may be prevented from exercising thisright, or at least social disapproval may mitigate the advantages of land title, anda high level of contestation may bring overwhelming disadvantages. A number

476 Cecile Jackson

of factors are likely to mediate the extent to which support for women’s landclaims are secured, and Rao’s work (2002) shows the variability with whichwomen are able to deploy rights and needs talk, in relation to land, to theiradvantage. Furthermore, the fact that most women who manage to obtain landin south Asia leave it to their sons, where they can, indicates a powerful dis-course of need that privileges males. This discourse is the fundamental basis ofmale power in relation to land, which is as important as the de facto control ofland. ‘Women . . . may have rights to pooled resources within the household,but their ability to exercise those rights is likely to be determined by variouscultural and contextual evaluations of need’ (Moore 1992: 134).

The second point is that land is a symbol of wealth that is tied to notions ofpersonhood and not only a material asset, and this is discussed in more detailabove. The dominance of productionist approaches to land can obscure this ideaand lead to an emphasis on land as a source of income (directly or indirectly) thatunderestimates the level of resistance that might be encountered by a projectdesigned to peel male fingers off their land holdings in favour of wives anddaughters.

This might be a rather pessimistic view, since changing gender cultures is along-term process and not immediately tractable to government policies andprojects, unlike the claims made for land rights by Agarwal. It suggests that theproject of shifting land from the state to landless women is worthwhile, but maywell produce a second generation of landed sons with a greater investment inpatriliny. And whilst it is important that government provides an enabling envir-onment for women to exercise land rights, the expectation that many marriedwomen will therefore claim land within landed households in south Asia isunrealistic. A number of studies of how claims are adjudicated at communitylevel suggest that marriage alliance and successful performance of gender rolesare factors in successful claims to land (Bosworth for Uganda, cited in Whiteheadand Tsikata (2003: 96); van den Berg (1997) for Cameroon). In studying dis-courses of need (Fraser 1989) in relation to land, Rao (2002) shows, with adetailed Indian case, how village authorities are open to land claims made on thebasis of appropriate and effective gender performance, and to an acceptance ofneed where abandonment and widowhood leave women without means of sup-port. In these discourses the presence of children, especially sons, as evidence ofsuccessful role performance can be decisive. Patriarchy rewards those that playby its own rules. I think that it is from the women who are not offered much bypatriarchy, e.g. in landless households where dependence is not an option, thatmost gender experimentation and change will come.

CONCLUSIONS

Land rights for women is not the inevitable conclusion of a gender analysisof land, which needs a more open terrain of possibilities, richer analyticalframeworks, a more critical approach to existing data and more contextuallygrounded research to make progress. Research on gender and land relations has


moved forward in fits and starts since the 1980s, with different approaches andemphases in particular regions, but, I would argue, remains in its infancy, withan exciting agenda that has yet to be fully explored. Research recommendationsmade by Agarwal include work on efficiency implications of gender-equal accessto land (2003: 196), but I would prioritize detailed ethnographies that focus onsocial change, that consider the diversity of subject positions and subjectivities ofwomen in relation to land, and situate gendered property relations within abroader context of marriage, kinship, livelihoods and life courses.

The policy implications of a broader focus on gendering land questions,rather than treating this as essentially about land rights for women, and a moreethnographic approach to social relations and subjectivities around land, perhapsinclude the following. A singular focus on land as the magic bullet that willtransform the position and condition of women is as mistaken as the over-emphasis on credit that Agarwal criticizes (2003: 217). Advocacy of a specificpolicy such as land rights for women runs the risks of both obliterating the needfor a consideration of land rights as a constraint in relation to other constraintsfaced by women, and giving blanket policy advice that is not properly con-textualized in specific places. In a South African gender analysis of land reform,Cherryl Walker argues that the Department of Land Affairs Gender Unit ‘has notanalysed the specific contribution of land reform to the fundamental transforma-tion of gender relationships that it presents as the goal. How important is landcompared to other resources, such as jobs or education or health services or thereform of discriminatory laws and customs?’ and criticizes it for adopting aninternational orthodoxy, derived from south Asia, without relating it to SouthAfrican conditions (Walker 2003: 128).

Of course women should receive land in land reform initiatives for landless orin resettlement schemes, but claiming land from the state is very different toclaiming it from fathers or husbands. The policy proposals offered by Agarwal’sperspective are grounded in a view of the family and households as only institu-tions of repression and exploitation, obstacles to be overcome. I think a view ofthese institutions as gender inequitable, but also the source of much that is sus-taining and satisfying in women’s lives, confers a more reliable basis for policythinking. Agarwal objects that ‘the assumption of family-based farming under-lies all forms of land distribution to the landless’ and suggests collective invest-ment and cultivation by women as an alternative, substituting women’s solidarityfor that of men and women within a household. Collective cultivation by womenis unlikely to become widely popular and successful (see also van den Berg 1997:240), although the DDS projects supporting land leasing and purchase by particu-lar groups of poor landless, mostly single, women makes sense (Agarwal 2003:209) in the ways that many group-based income-generating projects do, both formaterial gain and for the benefits to confidence, skills and social support. But thefamily, household and kinship are extraordinarily enduring institutions. In policydiscussion, rather than wishing the family and household away, more detailedunderstanding of them is necessary. Certainly, the recommendations made byAgarwal (2003: 216) for government to make available (usufruct) land for

478 Cecile Jackson

collective farming by in-marrying women, and for group investment by womenwith land rights, needs first a much clearer and deeper sense of the complexity ofgendered land relations. There is a risk of land rights for women becominganother ‘ready, fire, aim!’ policy.

This paper may seem rather pessimistic since I do not see women mobilizingactively around land demands, I am unsure of the extent and character of gendertransformation that would follow land rights for women, and I can see soundreasons why making land claims from husbands and kin is a very complex andcontradictory course of action for many women. However, I also think thatthe valuable detailed historical research emerging from doctoral researchers andothers suggests patterns of change that do not disadvantage women in relation toland nearly as uniformly as is sometimes alleged, and as Agarwal assumes, andshows how women can and do prise open patriarchal control of property whenopportunities and subjectivities coincide. One abiding source of optimism I havelies in the contradictory character of patriarchy and the ability of poor women tomake use of these contradictions. And whilst women’s movements and ruralactivism may not have directly assaulted male control of land, they have beenvital in changing the discursive environment, making gender an issue of localdebate and refiguring ideas of needs and rights in ways that have aided andabetted women in taking forward their own such projects.

REFERENCES

Agarwal, B., 1994a. ‘Gender and Command over Property: A Critical Gap in EconomicAnalysis and Policy in South Asia’. World Development, 22: 1455–78.

Agarwal, B., 1994b. A Field of One’s Own: Gender and Land Right in South Asia. Cambridge:Cambridge University Press.

Agarwal, B., 1998. ‘Disinherited Peasants, Disadvantaged Workers: A Gender Perspectiveon Land and Livelihood’. Economic and Political Weekly, 28 March, 33 (13): A2–A14.

Agarwal, B., 2003. ‘Gender and Land Rights Revisited: Exploring New Prospects via theState, Family and Market’. Journal of Agrarian Change 3 (1 and 2): 184–224.

Beteille, A., 1993. ‘The Family and the Reproduction of Inequality’. In Family, Kinshipand Marriage in India, ed. P. Uberoi, 435–51. Delhi: Oxford University Press.

Brass, T., 1994. ‘The Politics of Gender, Nature and Nation in the Discourse of the NewFarmers’ Movements’. Journal of Peasant Studies, 21 (3 and 4): 27–71.

Cooper, B., 1997. Marriage in Maradi: Gender and Culture in a Hausa Society 1900–1989.Oxford: James Currey.

Deere, C.D., 2003. ‘Women’s Land Rights and Rural Social Movements in the BrazilianAgrarian Reform’. Journal of Agrarian Change, 3 (1 and 2): 257–88.

Elson, D., 1995. ‘Gender Awareness in Modelling Structural Adjustment’. World Devel-opment, 23 (1): 1851–68.

Folbre, N., 1988. ‘The Black Four of Hearts: Towards a New Paradigm of HouseholdEconomics’. In A Home Divided: Women and Income in the Third World, edsD. Dwyer and J. Bruce, 248–62. Stanford: Stanford University Press.

Fraser, N., 1989. Unruly Practices: Power, Discourse and Gender in Contemporary Social Theory.Minneapolis: University of Minnesota Press.


Gough, K., 1993. ‘Brahman Kinship in a Tamil Village’. In Family, Kinship and Marriagein India, ed. P. Uberoi, 237–56. Delhi: Oxford University Press.

Jackson, C., 1993. ‘Women/Nature or Gender/History?’. Journal of Peasant Studies, 20 (3):389–418.

Jackson, C., 1996. ‘Rescuing Gender from the Poverty Trap’. World Development, 24 (3):489–504.

Jackson, C., 1997. ‘Post Poverty, Gender and Development?’. IDS Bulletin, 28 (3): 145–55.

Jackson, C., 2002. ‘Disciplining Gender?’. World Development, 30 (3): 497–509.Kandiyoti, D., 2003. ‘The Cry for Land: Agrarian Reform, Gender and Land Rights in

Uzbekistan’. Journal of Agrarian Change, 3 (1 and 2): 225–56.Kapadia, K., 1996. ‘Property and Proper Chastity: Women’s Land Rights in South Asia

Today’. Journal of Peasant Studies, 23 (4): 166–73.Lastarria-Cornhiel, S., 1997. ‘Impact of Privatisation on Gender and Property Rights in

Africa’. World Development, 25 (8): 1317–33.Layder, D., 1997. Modern Social Theory: Key Debates and New Directions. London: UCL

Press.Mackenzie, F., 1990. ‘Gender and Land Rights in Murang’a District, Kenya’. Journal of

Peasant Studies, 17 (4): 610–43.Madan, T.N., 1993. ‘The Structural Implications of Marriage Alliance in North India’. In

Family, Kinship and Marriage in India, ed. P. Uberoi, 273–86. Delhi: Oxford UniversityPress.

Manimala, 1983. ‘Zameen Kenkar? Jote Onkar!: Women’s Participation in the BodhgayaLand Struggle’. Manushi, January–February, 14: 2–16.

Moore, H., 1992. ‘Households and Gender Relations: The Modelling of the Economy’.In Understanding Economic Process, eds S. Ortiz and S. Lees, 131–50. Lanham: Univer-sity Press of America.

Narayan, D., 2000a. Voices of the Poor: Can Anyone Hear Us? New York: Oxford Univer-sity Press and World Bank.

Narayan, D., 2000b. Crying out for Change. New York: Oxford University Press andWorld Bank.

Narayan, D. and P. Petesch, eds, 2002. From Many Lands. New York: Oxford UniversityPress and World Bank.

Omvedt, G., 1993. Reinventing Revolution: New Social Movements and the Socialist Traditionin India. New York and London: Armonk.

Ongaro, W., 1988. ‘Adoption of New Farming Technology: A Case Study of MaizeProduction in Western Kenya’. PhD Thesis, University of Gothenburg, Gothenburg.

Raheja, G.G., 1996. ‘The Limits of Patriliny: Kinship, Gender and Women’s SpeechPractices in Rural North India’. In Gender, Kinship, Power: A Comparative and Interdis-ciplinary History, eds M. Maynes, A. Waltner, B. Soland and U. Strasser, 149–76. NewYork and London: Routledge.

Rao, N., 2002. ‘Standing One’s Ground: Gender, Land and Livelihoods in the SantalParganas, Jharkhand, India’. PhD thesis, School of Development Studies, Universityof East Anglia, Norwich, UK.

Razavi, S., 2003. ‘Introduction: Agrarian Change, Gender and Land Rights’. Journal ofAgrarian Change, 3 (1 and 2): 2–31.

Sen, A.K., 1990. ‘Gender and Cooperative Conflicts’. In Persistent Inequalities, Women andWorld Development, ed. I. Tinker, 123–49. New York: Oxford University Press.

480 Cecile Jackson

Seur, H., 1992. ‘Sowing the Good Seed: The Interweaving of Agricultural Change, GenderRelations and Religion in Serenje District, Zambia’. PhD thesis, Wageningen University,the Netherlands.

Shah, A.M., 1993. ‘Lineage Structure and Change in a Gujarat Village’. In Family, Kinshipand Marriage in India, ed. P. Uberoi, 119–45. Delhi: Oxford University Press.

Sharma, U., 1984. ‘Dowry in India: Its Consequences for Women’. In Women and Property:Women as Property, ed. R. Hirschon, 62–74. London: Croom Helm.

Sjoblom, D., 1999. ‘Land Matters: Social Relations and Livelihoods in a Bhil Communityin Rajasthan, India’. PhD thesis, School of Development Studies, University of EastAnglia.

Stivens, M., 1985. ‘The Fate of Women’s Land Rights: Gender Matriliny and Capitalismin Rembau Sembilan, Malaysia’. In Women, Work and Ideology in the Third World, ed.H. Afshar, 3–36. London: Macmillan.

Strathern, M., 1988. The Gender of the Gift. Berkeley: University of California Press.Sugden, R., 2000. ‘Team Preferences’. Economics and Philosophy, 16: 175–204.Udry, C., J. Hoddinott, H. Aldeman and L. Haddad, 1995. ‘Gender Differentials in Farm

Productivity: Implications for Household Efficiency and Agricultural Policy’. FoodPolicy, 20 (5): 407–23.

Van den Berg, A., 1997. ‘Land Right, Marriage Left: Women’s Management of Insecurityin North Cameroon’. Research School CNWS, Leiden University, the Netherlands.

Wadley, S. and B. Derr, 1993. ‘Karimpur Families over Sixty Years’. In Family, Kinshipand Marriage in India, ed. P. Uberoi, 393–415. Delhi: Oxford University Press.

Walker, C., 2003. ‘Piety in the Sky? Gender and Land Reform in South Africa’. Journal ofAgrarian Change, 3 (1 and 2): 113–48.

Whitehead, A., 1984. ‘Women and Men: Kinship and Property: Some General Issues’. InWomen and Property: Women as Property, ed. R. Hirschon, 176–92. London: CroomHelm.

Whitehead, A. and N. Kabeer, 2001. ‘Living with Uncertainty: Gender, Livelihoods andPro-Poor Growth in Rural Sub-Saharan Africa’. IDS, Sussex, Working Paper 134.

Whitehead, A. and D. Tsikata, 2003. ‘Policy Discourses on Women’s Land Rights in Sub-Saharan Africa: The Implications of the Return to the Customary’. Journal of AgrarianChange, 3 (1 and 2): 67–112.

Yngstrom, I., 1999. ‘Gender, Land and Development in Tanzania: Rural Dodoma 1920–1996’. DPhil thesis, University of Oxford.

Gender analysis of land: beyond land rights for women?

Documents