Gender and Commercial Agriculture in Ecuadorand GuatemalaSarah Hamilton, Linda Asturias de Barrios,and Brenda Tevalan

Sarah Hamilton is Associate Professor and Director ofthe M.A. Program in International Development, Grad-uate School of International Studies at the University ofDenver, and Adjunct Senior Research Associate, Officeof International Research and Development at VirginiaTech. Linda Asturias is the Director of Estudio 1360 inGuatemala City. Brenda Tevalan is a research consul-tant with Estudio 1360.

It has become an article of faith among gender-and-development analysts that agricultural development agen-das focusing on commercialization among small-scaleproducers in Latin America are likely to result in theunintended economic marginalization of women. While thisexpectation is well-founded, given the male bias of devel-opment initiatives described below and micro-level resultsreported in earlier studies of Andean and Guatemalan popu-lations, new data suggest that institutional male bias ismediated by several factors, including ethnicity, mutualityof interests among women and men at the household level,and households' growing experience over time of particularmarkets. Rather than merely serving as unpaid labor inmarket-oriented household agriculture, women emerge asintegral partners in commercial production.

The productive contributions and constraints of LatinAmerican women farmers were largely ignored in the devel-opment policies, programs, and projects that effectively re-structured small-scale commercial agriculture during theopen-market or planned-economy modernization drives ofthe 1950s, 1960s, and 1970s. Women were largely excludedfrom public land reform programs that enabled small-scaleproducers to own or use lands formerly held by less-productive large estates (Arizpe and Botey 1987; Deere1986). State and donor-funded programs designed to in-crease commercial output in the smallholding sector chan-neled technology, credit, and marketing assistance to men,while women's programs focused on home-making and sup-plemental income-earning activities (Chaney and Schmink1976; Flora and Santos 1986). This pattern was intensified

during the 1980s and 1990s, as structural adjustment policiesresulted in decreasing public investment in small-scale agri-culture. Initiatives to privatize land ownership among hold-ers of redistributed lands and to restructure formerly publictechnical extension and credit services (often throughcontracting between farmers and buyers who offertechnology-credit-marketing packages) perpetuated the in-stitutionalizing of gender-differentiated access to the meansof commercial production (Hamilton 2002; Stephen 1994).

Case studies of both indigenous and mestizo popula-tions in the Andes and highland Guatemala reportedwomen's decreasing control of land, labor, products, or in-come in households that increasingly orient production tothe market (Bossen 1984; Bourque and Warren 1981; Deereand Leon 1982). Such cases informed the repeated gener-alization that men tend to control household production ofcash crops in Latin America (Alberti 1988; IDB 1995:64-69;Nash 1986), especially operations on relatively larger, weal-thier holdings in the small-scale sector where women's fieldlabor is not required (Deere and Leon 1987), those involvinghigher-level technologies (Warren and Bourque 1991), andthose producing high-value nontraditional export cropsthrough male-oriented production and marketing entities(Katz 1995; von Braun, Hotchkiss, and Immink 1989). Onsuch farms, where more of a household's economic re-sources are invested in commercial production, women areexpected to be marginalized from control of productiveland, labor, higher-level technologies, and incomes.

The research reported here calls into question the inevi-tability of women's decreasing control of the means, pro-cesses, and fruits of agricultural production in commodity-producing Latin American households. Indigenous womenin highland Ecuador and Guatemala have retained orachieved relatively egalitarian control of economic resourcesdespite the male bias of market structures. This paper ana-lyzes cross-sectional variation in women's control of land,technology, and incomes across levels of increasing house-hold commoditization (Ecuador) and nontraditional exportcrop production (Guatemala). It also explores the structuraland cultural bases of gendered resource control in these set-tings, where much of the earlier research supporting themarginalization thesis originated.

Culture & Agriculture Vol. 23, No. 3 Fall 2001

Gender and Commercial Agriculture in the CentralEcuadorean Sierra

The study region comprises three indigenous (Quichua)communities located on the eastern slopes of the inter-Andean basin in Canton (county) Salcedo, Cotopaxi Pro-vince. Quantitative measures are based on a 1992-93 three-round survey of a probabilistic sample of 116 households,derived from a population of approximately 470 households.Analysis will utilize data collected from 108 wife-husbandpairs. The population comprises nuclear-family households(80 percent) with access to small amounts of land that aver-age 2.6 hectares (standard deviation 2.6; median 2.0). Onlyone of the sampled households is headed by a single indi-vidual. Located some 10-20 km from the county seat, onlyone of the communities is directly served by a paved sec-ondary road. Education is limited to poorly-staffed elemen-tary schools. Public health clinics are chronically short ofmedicines and are only occasionally served by rotating med-ical personnel.

Although the population is primarily indigenous,Spanish is the dominant language (only elders also speakQuichua) and people no longer wear regionally-distinctiveethnic clothing. Syncretic folk Roman Catholicism is univer-sal and little pre-Conquest mythology has been preserved.However, traditional Andean social forms have survivedmany generations of enforced indentured service to a landedelite—service that ended only in the early 1970s. Bilateralkinship reckoning and gender-egalitarian inheritance of landare observed, as well as extra-household labor exchange andother forms of economic reciprocity negotiated by both maleand female household heads (Hamilton 1998:45-71,182-209).

The local economy is based on small-scale commercialproduction of potato, grains, and vegetables; crop incomesare supplemented by dairying and livestock production. Onaverage, households derive nearly 90 percent of their cashincome from agricultural production; local people have beencommercializing over 80 percent (by weight) of total agricul-tural yields for at least 25 years (SEDRI 1981: 150). Abouthalf of the sample also relies on men's cash incomes derivedfrom nonagricultural wage work, transportation, and com-merce. Local, nonagricultural, wage-labor opportunities forwomen are severely limited. During 1992-93, around one-fifth of male household heads migrated seasonally to urbanareas for work. This is a relatively low percentage by high-land standards (DeWalt et al. 1990).

Virtually all households have land for agriculturalproduction. However, distributions of both income and landholdings are positively skewed. The top economic quintile

is composed of families who own as much as 20 ha, owntrucks and farm machinery, practice high-technology agri-culture, own large concrete and tile houses equipped withgas ranges and perhaps refrigerators, provide post-secondary education for their children in nearby cities, anduse private medical care. The three-fifths of households com-prising the middle-income tier control 2 to 5 ha, own plowanimals or rent tractors for relatively high-technology agri-culture, live in two-room concrete houses, may own a smallgas stove and television, do not have indoor plumbing orlarge appliances, and must sacrifice to educate their childrenbeyond primary school or to use private health care. Thepoorest quintile comprises people who own 1 ha of land (orless), live in one-room houses (some of adobe and thatch),may have only rudimentary tools, and cannot afford sec-ondary education or purchased health care.

Relatively little of the commercial agricultural infra-structure created by governmental and quasi-governmentalagencies has reached the smallholding sector in CantonSalcedo. Local dairy and livestock production associationsprovide credit and supplemental income to around half oflocal households. Nearly all members of these associationsare male, as the groups were initially formed of men re-cruited by international development agencies (see Fernandez1988). Additionally, national law prohibited membership byhusbands and wives in any production cooperative (Co-operatives Law of 1973, see Phillips 1987). Publicly-subsidized cheap credit is not available to small-holders,who must pay around 60 percent interest on the short-termloans available from formal financial institutions. Thus,access to production-association credit funds can be a criticalbenefit of membership. Private extension of technology andcredit through contracts between producers and buyers hasnot been developed in the region.

Gender-egalitarian Ownership of Land and Divisionof Labor

Women and men own land equally. Around 80 percentof the women surveyed own land, a slightly higher propor-tion of the sample than men who report ownership. Half ofthese women have inherited land, while the other half ownland they purchased jointly with their husbands. Manywomen own both inherited and jointly-purchased plots.Among the women surveyed, the average amount of landowned is 1.8 ha (standard deviation 2.5). Surveyed men re-ported owning 2.3 ha (standard deviation 4.6). Mean differ-ences in the amounts of land owned by women and men arenot statistically significant (two-tailed probability of t = .430),even though men may have included some of their wives'

Culture & Agriculture Vol. 23, No. 3 Fall 2001

land in their reporting. Gender equality in landholding isobserved despite the fact that the research communities arelocated on lands of former haciendas and participated in theEcuadorean land reform program of the 1960s and 1970s. Bylaw this program permitted only "household heads," nearlyalways male, to receive hacienda subsistence plots formerlyworked by their households or to access communal landsexpropriated from a hacienda (Phillips 1987; Deere 1986). Inthese and other Ecuadorean land-reform communities, how-ever, traditional gender-egalitarian practice governs the in-heritance of land acquired during the reform (Alberti 1986;Stark 1979). Parents scrupulously apportion plots of equalquantity and quality among their children, regardless ofgender.

Household agricultural production is a family endeavor:all household lands are worked jointly by both householdheads and the income from all household production ispooled. There are no "women's crops" or "men's crops," nordo women specialize in livestock production to a greater orlesser degree than men. Nearly all women are full-timefarmers and participate in all agricultural activities through-out the production cycles. Contrary to practices observed inAndean mestizo populations (Deere and Leon 1982), wom-en's labor in agricultural production is not conditioned bythe wealth level of their households (Hamilton 1998: 144-166). Some forms of labor complementarity exist within indi-vidual households, but no productive labor domains areclosed to women. Both men and women report that the agri-cultural work of men and women is igual (equal, "thesame"). Both men and women insist that no activities aresuitable for—or performed well by—only men or women.Many families hire agricultural labor, where male andfemale workers earn an equal wage.1

Gender-egalitarian Intrahousehold Resource DynamicsEconomic decision-making is a consensual process in

which women have at least an equal voice with their hus-bands.2 Most of the women surveyed reported having anequal say in decisions regarding land use (when, what, andhow much to plant) and in the selection of agricultural tech-nology (Table 1). Nearly 90 percent of women reported atleast equal participation in the management of householdincome. Although surveyed women were not asked to iden-tify household members who manage income from allsources, men were asked to provide this information. Table2 shows that husbands credit their wives with the manage-ment of both agricultural and off-farm cash incomes, thelatter earned by husbands. Table 3 demonstrates that theseresource control patterns characterize households that

Table 1Women Report Gender-egalitarian Control" ofHousehold Economic Resources, Canton Salcedo(Ecuador)

Resource ControlDomain

Land UseAll IncomesAgricultural Technology

Percentage of WomenReporting Equal Control

84%88%71%

a Women's self-reported control of resource as equal to or greaterthan husband's control. N = 108.

Data Source: NSF/FUNDAGRO Household Surveys, Cant6nSalcedo, August 1992- July 1993.

derive more of their total income from market-oriented agri-culture, as well as households that derive less. Table 4 pre-sents statistical tests of association designed to test whetheror not the degree to which households derive their liveli-hoods from market-oriented production affects women'scontrol of household lands, agricultural technology, or in-comes. The proportion of total household income derivedfrom commercial agriculture was ranked into four quartiles.This indicator was used for the level of household commer-cialization because the reporting of income proved to behighly reliable in this sample, and this measure indicates therelative importance of commercial agriculture to householdlivelihood. Following the logic of the marginalization thesissummarized above, women's control of productive re-sources should decrease in households that derive more oftheir income from agriculture. Correlational analysis doesnot support this thesis, because the correlations are nearzero, and none is statistically significant.

Both men and women emphasize the importance ofcollaborative intrahousehold decision-making by the "DosCabezas" (two household heads). Within most of the house-holds that were regularly observed, a dynamic balance ismaintained between two decision-making partners. Whenspouses disagree concerning the disposition of economic re-sources, neither wife nor husband always prevails concern-ing a particular resource or within an individual household,although women tend to have greater authority over fi-nances. The spouse with greater knowledge, past success,commitment or stubbornness tends to prevail in a givenresource-control decision. Although consensual decisionprocesses require a great deal of negotiation, often pro-longed when two equally-powerful householders go head tohead, most women and men express respect for the vigor

Culture & Agriculture Vol. 23, No. 3 Fall 2001

Table 2Men Report Gender-egalitarian Control of Household Cash Incomes, Canton Salcedo (Ecuador)

Percentage of Earning Householdswith Egalitarian Income Control"

92%93%89%93%

100%

73%100%100%

' N = 99 male household heads reporting income amounts (92% of sample of 108 male household heads).b Husband reports wife's equal or greater control of income per income source; in 91 of the 93 households that produced cash crops, the husbandreported who controls income derived from crops.c Income was earned by husbands.d Income was earned by husbands in most cases.Source: NSF/FUNDAGRO Household Surveys, Canton Salcedo, August 1992-July 1993

Income Sources

Crops (potato, grains)MilkAnimal salesWage labor, local0

Salary, localc

Wage or salary earned duringtemporary outmigrationc

Transportationc

Commerced

Households Earning(Sample N = 99')

93

706923

7

14136

and knowledge of their partners and satisfaction with theultimately egalitarian results.

Gender-egalitarian Control of Household Lands,Agricultural Technology, and Incomes: Case Studies3

Clemencia (age 32) and her husband Alejandro ("Alejo,"age 34) have both inherited land. During their 11 years ofmarriage, they have also bought land together, and farm atotal of 5 ha. Both are full-time farmers who invest in hiredlabor and modern machinery and tools. Nearly 100 percentof their cash income is derived from agriculture. Both arecommunity leaders, holding offices in a number oforganizations.

Clemencia participates equally with her husband indecisions concerning the choice of crops in which their landwill be invested. In making her recommendations Clemenciaconsiders soil, rainfall potential, pricing trends, resistance topests, labor and other input requirements, and benefits to bederived from crop-rotation patterns. Alejo does the same,and the couple usually reaches a consensus without pro-longed negotiation. This process is also followed withrespect to agricultural inputs, including expensive agro-chemicals. Because she earned a scholarship to attend atraining course in the US that included this information,Clemencia has specialized knowledge of the dangers ofhighly toxic chemicals. Alejo is more knowledgeable re-garding traditional forms of biological pest management.They pool their knowledge and have worked out an

integrated pest management practice that meets her require-ments for cost effectiveness and their mutual goals to reducehealth risks associated with pesticide application.

Clemencia is responsible for managing the householdbudget—determining how much can be spent for currentagricultural inputs and how much must be saved for futureinvestments—as well as doling out money for weekly ex-penses such as food. Although large savings and expen-ditures are negotiated between husband and wife, both menand women perceive that women hold the ultimate respon-sibility and veto power in managing financial resources.4

Clemencia's authority over financial resources includes in-come from all household crops, livestock, and dairy produc-tion, as well as loans and payments resulting from Alejo'smembership in the dairy production association. Two-thirdsof women surveyed market their family's products, butClemencia does not, as she dislikes dealing with potato mar-ket intermediaries. Although women's control of financialresources does not depend on direct control of marketing,other women in the sample derive considerable respect fromtheir husbands for their tough bargaining with intermedi-aries and knowledgeable negotiation with agricultural inputsuppliers and with lending institutions.

In the wealthier household of Alegria (age 35) andRuben (36), the female head moved to the region as a youngbride with 12 years of education but no land of her own.Alegria and Ruben purchased land to add to his inheritanceand jointly control the largest holding in the research

Culture & Agriculture Vol. 23, No. 3 Fall 2001

Table 3Gender Egalitarian Control of Household Economic Resources by Levels of Agricultural Commercialization,Canton Salcedo (Ecuador)

Percentage of Total Household IncomeDerived from Commercial Agriculture

Percentage of women reporting resource control 0 • 37% 40 • 60% 61 - 76% 77 • 100%equal or greater than husband's for: (N = 24) (N = 25) (N = 25) (N = 25)

Land UseAgricultural TechnologyAll Incomes

83%67%96%

88%76%80%

80%68%88%

80°/«68°/«88°/

Percentage of men reporting wife'sequal or greater control of:

Cash Crop Income

Percentage of Total Household IncomeDerived from Commercial Agriculture

7 - 37% 40 - 60% 61 - 76% 77 - 100%(N=16) (N = 25) (N = 25) (N = 25)

100% 91% 91% 94%

N = 99 households for which male head provided complete income information (92% of sample of 108 households), allowing computation of percentageof total income derived from commercial agriculture; in 91 of the 93 households that produced cash crops, the husband reported who controls incomederived from crops.

Data Source: NSF/FUNDAGRO Household Surveys, Cant6n Salcedo, 1992-93.

population (20 ha locally, plus 60 in the Amazon Basin). Bothare full-time farmers whose considerable income is derivedentirely from commercial production.

Ruben is eloquent on the need to "industrialize" agricul-tural production to enhance yields of their potato crop. Theyplant high-yielding varieties and apply heavy weekly dosesof chemical pesticides. Ruben is also eloquent on the subjectof his wife's managerial skill. He says that she is bienorganizada (well organized, a good manager) and that hermanagerial ability pays off in matters of recruitment andsupervision of hired workers and cost-benefit analysis ofinputs. Alegria has recruited several families from anotherprovince to work on their farms, and she also supervisesRuben and his father, as well as the hired laborers, duringharvests. While working in the field herself, she examinesthe output of each worker, exhorting the less industrious toincrease their loads. Ruben says that Alegria works harderthan he does because she has to organize things as well aswork in the fields. He tries to make up for this by increasinghis share of the cooking and child care.

Like Clemencia, Alegria is an equal partner in decisionsregarding savings, investments, and current expenditures.Like other relatively wealthy women, including womenmore than 20 years older than she, Alegria works in thefields. However, her family does not need her muscle powerand she could secure nonagricultural employment. But, inher view, she has a great deal of managerial authority inhousehold enterprises which she would not have if she took

an office job in town. Furthermore, she believes that her con-tributions boost agricultural earnings by a margin greaterthan she could earn off-farm. Alegria also prefers a partici-patory style of labor management, as do many other womenwhose attention to labor organization is valued highly bytheir husbands.

Both men and women strongly state that, while partici-pation in field labor is not a requirement for entitlement todecision authority over land, labor, products, or incomes,both household heads are expected to contribute to decisionprocesses regarding any productive domain in which theywork. The association between labor and decision authorityis viewed locally as self-evident. This association has beenposited as an important indicator of egalitarian (versus patri-archal) household production systems (Deere 1995).

Both of these households derive their cash income pri-marily from commercial agriculture. If the paradigm posit-ing male control of household commodities production wereappropriate for this population, women would be margin-alized from control of productive resources and incomes.Clearly this is not the case for these market-oriented families.However, these relatively affluent cases represent less thanhalf of the research population. What happens in poorer, lessmarket-oriented households? Two cases demonstrate waysin which families with much less land also participate incommodities markets, and the ways in which women investcash resources, earned off-farm by their husbands, in com-mercial agricultural production.

Culture & Agriculture Vol. 23, No. 3 Fall 2001

Correlations: Women's Control of Household Economic Resources'"With Increasing Commercialization of Household Agriculture0, Canton Salcedo (Ecuador)

Dependent Variables:Resource Control Domains

Land Use8 (N = 99)

Input Selection a(N = 99)

Household Finance8 (all incomes) N = 99

Income from Crops" (N = 91 producers)

•Women's self-reported control of resource: control equal or greater than husband's coded as 1; less than equal control coded as 0.

"Women's control of crop incomes, reported by husbands: control equal or greater than husband's coded as 1; less than equal control coded as 0.c Percentage of total household income derived from commercial agriculture, ranked into quartiles.

Households grouped by percentage of total household income derived from commercial agricultural production:

Group 1 (N = 24) 0 - 37 % percent of total income derived from commercial agriculture.

Group 2 (N = 25) 40 - 60 %

Group 3 (N = 25) 61 - 76 %

Group 4 (N = 25) 77 - 100%

For women's control of crop incomes, as reported by husbands:

Group 1 (N = 16 male household heads reporting crop income) 1-37% of income derived from commercial agriculture;

Groups 2, 3, and 4 same as above.

Source: NSF/FUNDAGRO household surveys, Canton Salcedo, 1992-93

SymmetricMeasures

Gamma

Spearman

Gamma

Spearman

Gamma

Spearman

Gamma

Spearman

Coefficient

-.107

-.054

-.018

-.011

-.113

-.050

-.248

-.078

Std. Error

.201

.102

.166

.102

.200

.088

.245

.080

Approx.T

-.529

-.535

-.109

-.110

-.560

-.491

-.926

-.712

Approx. Sig

.597

.594

.913

.913

.575

.625

.355

.478

Susana (age 23) and her husband Nicolas (age 26) arerecently-resettled migrants who lived for several years nearQuito, where they worked in a cheese factory. They share-crop about 1 ha of land, some of which comes into thehousehold through her parents and some from his. Susanarecommended they plant one plot in garlic, an expensive,high-risk, and high-value crop. Nicolas readily agreed thather crop choice could maximize their earnings from a smallland area. He based his pro forma decision on his wife'sgreater knowledge of the crop, gained through her morecareful attention to their neighbors' experience. Although asoil-borne disease destroyed the crop, Nicolas supportedSusana's plan to try again once the land had "rested" suffi-ciently to be productive again.

Susana and Nicolas derive around two-thirds of theirtotal income from commercial agriculture. Nicolas workspart-time at the local dairy production association. His wagecovers household subsistence expenses and helped to payfor Susana's selections of expensive plant sets and agro-chemicals. Susana also arranged to feed a cousin's pigs in

exchange for a share of the piglets, and used the whey runofffrom cheese production as part of her feeding program. Thecouple heatedly negotiate the proportion of their agricul-tural income that must be reserved for future land purchase,but both reported that Nicolas had come to accept Susana'sjudgment concerning luxuries to which they were accus-tomed in the city but could no longer afford.

Mariana and Enrique, a poor couple in their late thirties,also hope to buy sufficient land to support their large family.Each inherited less than 1 ha and they derive a little less than40 percent of their annual income from commercial produc-tion. Although they have a small land base, they investEnrique's off-farm income in hired labor and agrochemicalsthat are applied, under Mariana's supervision, to the pro-duction of crops for sale. Enrique works construction in thecapital city for several months each year and relies onMariana to manage his salary, as well as production earn-ings. Mariana explains her role in budget management andaccumulation:

Culture & Agriculture Vol. 23, No. 3 Fall 2001

My husband is very good about bringing home themoney.... He spends very little on himself. So you mightthink I have a lot of money to spend, but I don't. Wewanted to join the [cattle production] cooperative. Thatcosts a lot and we had to pay it all at one time. And weare saving to buy land.... Well, I decide what we canspend. My mother taught me how to get along on verylittle.

This case is not unusual, as income earned by temporarymigrants is managed by their wives in nearly three-fourthsof households reporting such income. Off-farm income sub-sidizes high-input commercial production on many farms,including some of the smallest.

Gender Ideology and Women's Resource Controlin Highland Ecuador

Gender-egalitarian control of land, labor, technology,and financial resources is related to ideological constructionsof gender based in traditional Andean concepts and in localpolitical-economic history. Local people express strong pref-erences for collaborative, egalitarian social and economicinstitutions. Collaboration among households is widelyviewed as a survival mechanism that enabled local people tooutlast the hacendados' hold on their land and continues tooffer the best hope of social security in communities largelywithout public social infrastructure.

Collaboration between wives and husbands is valued inmuch the same way as cooperation among families. Men ex-press high valuation of their wives' (and other women's)agricultural expertise, economic judgment, physical strengthand hard work, political acumen, and bargaining skill in themarketplace. Women praise their husbands for similarstrengths. There does not appear to be a gendered divisionin the attribution of these strengths. There is a generalizedunderstanding that "two heads are better than one" in themanagement of household production, consumption, andaccumulation, because the pooling of skills is an importantpart of the enterprise. However, the preference for dualisticpower sharing in households goes beyond practical consid-erations, important though these are.

Throughout the Andes, ethnographers have found evi-dence for a tradition in which cultural norms prescribe thathouseholds have two equally-powerful heads: one femaleand one male (Allen 1988; Harris 1978). Both balance anddynamism must characterize the relations between thesetwo heads if the household is to prosper. The desired bal-ance of power is maintained in households where neitherhead controls a greater share of material, social, or spiritualresources. Dynamism is achieved by a continuous and

egalitarian shifting of the balance, as neither head dominatesall decisions or decision-making domains. The perception isstrong that maintaining this dynamic provides the push thatenables the household to move forward and, in time, togrow and mature properly: literally the "power of balance."This fundamentally Andean preference for dualistic head-ship has been reinforced, rather than eroded, by the chal-lenges of making a living in small-scale commercial agriculture.

Gender and Agricultural Export Marketsin Highland Guatemala

The research population comprises two primarilyKaqchikel Maya communities in the Chimaltenango Depart-ment, a leading nontraditional agricultural export (NTAE)producing area in Guatemala. Chimaltenango leadsGuatemala in snow pea production for export and is also abroccoli export center. Quantitative analysis is based primar-ily on a 1998 probabilistic-sample survey of 141 householdsfrom a population of 406 households. The sample used foranalysis of women's control of land and other productiveresources includes 87 households in which an intervieweereported producing crops for domestic or export markets.Only interviewees in these households were asked to reportresource-control decision patterns. In 17 percent of house-holds, both male and female household heads were inter-viewed (some of these included widows and adult sons). In4 percent of households, only a female head was inter-viewed. In the remainder, only the male household headwas interviewed (4 percent of these men did not have aspouse or partner). All interviewees responded to a singlesurvey. If spouses supplied noncontradictory information,the common value was entered. In cases where values dif-fered between spouses, the woman's response was entered.

Among commercial producers, 98 percent are self-identified as Kaqchikel. Nearly all women and men are bilin-gual in Kaqchikel and Spanish. The majority of religiously-affiliated households are divided nearly equally amongRoman Catholic and Protestant denominations. Two-thirdsof households are comprised of nuclear families and 6percent are headed by single women. As in the Ecuadoreancase, social and economic infrastructure is severely limited.

The local economy is agriculturally-based, comprised ofmixed subsistence and commercial production in mosthouseholds, and characterized by a highly, and positively,skewed distribution of land among households. On average,households access (including rental) 1.2 ha. (SD 2.9). House-holds producing for agricultural markets access 1.5 ha. (SD3.6). Nearly one-sixth of households do not own land, and

Culture & Agriculture Vol. 23, No. 3 Fall 2001

an additional 60 percent own less than 1 ha. Only 3 percentof households own 5 or more ha and one household owns 33ha. This distribution is evened out somewhat through landrental. Land quality and limited access to needed irrigationare similar for the Ecuadorean and Guatemalan populations.Although their land base is much smaller, the Guatemalancommunities exhibit distributions of improved housing andownership of vehicles and other material goods that areremarkably similar to those reported by their Ecuadoreancounterparts.

Local producers have been growing high-value, non-traditional exports (primarily snow peas and broccoli) forsome 15-20 years. Smallholders move in and out of NTAEproduction for a variety of reasons, including increasingprice uncertainty. Despite U.S. rejection of produce with pestor pesticide-residue contamination, soil depletion associatedwith the overuse of agrochemicals and rising land pressure,and lack of insurance (Carletto, de Janvry, and Sadoulet1999; Barham, Carter, and Sigelko 1996), 66 householdsplanted these crops in 1998. Most NTAE producers alsoplanted commercial crops for the domestic market (potato,strawberry, and cabbage), as well as subsistence crops(maize, beans). The remaining 21 households commer-cialized production for the domestic market only.

Access to agricultural infrastructure is limited. Althoughmany Guatemalan smallholders produce and market non-traditional export crops through cooperatives, the localmarket is dominated by private intermediaries, with a mi-nority of producers contracting sale of crops (primarilybroccoli) with exporters. One-fourth of men and 5 percent ofwomen belong to groups self-organized to market nontra-ditional exports through contracts with intermediaries oragroexporters.

Intrahousehold Division of Land, Labor,Decision-making, and Incomes

The ownership of land, division of agricultural labor,and provisioning for household consumption needs aremuch more differentiated by gender in Guatemala than inEcuador. Women have traditionally earned incomes throughcraft production, storekeeping, small animal production, andselling both agricultural and nonagricultural products inregional markets, while men were primarily responsible forsubsistence and domestically marketed agricultural produc-tion. Separate budgets for subsistence (food, clothing, do-mestic technology) and agricultural production were man-aged by women and men respectively. Earlier studies fromthe region concluded that household adoption of NTAEsresults in increased field labor for women, who may

decrease the amount of time devoted to independentincome-producing activities (von Braun, Hotchkiss, andImmink 1989). Men dominate the marketing of NTAEsthrough co-ops and other forms of export contracting, whilewomen are responsible for stretching the household sub-sistence fund they administer to cover food and many do-mestic expenditures. Additionally, women depend on theirspouses to share receipts in a manner that compensates forany decrease in their own independent incomes. One studyfound that, although women did not give up independentincome-earning activities when they took to the NTAE fields,they received a smaller proportion of incremental incomederived from NTAE production than did women whosehouseholds' income increments were derived from othersources (Katz 1995). Kaqchikel populations have also beencharacterized as patriarchal with respect to land-holding andland-use decision-making (even on women's land) as well aswith respect to control of agricultural in-comes and othereconomic resources (Katz 1995; Nieves 1987).

In the study communities, only 22 percent of womeninherited or bought land individually—compared with 57percent of men—while another 29 percent bought land to-gether with their husbands. However, nearly half of thehouseholds report renting land, a strategy that enables theexpansion of NTAE production.

Women are heavily involved in household productionof nontraditional exports and other commercial crops.Among producers of snow peas and broccoli, 94 percent ofmen report that their wives work with them in either har-vesting (92 percent), planting (77 percent), fertilizer appli-cation (20 percent), and/or hoe cultivation (22 percent).Women also market crops in many households (Table 5),although only 17 percent of snow pea producers report thatwomen sell the crop and all broccoli was sold by men. Anethnographic study of three producer households in the re-gion indicated that an observable division of labor by genderand age guides work assignment in the fields. When disrup-tions of daily routines reduce availability of family labor,husband and wife together reallocate family members' agri-cultural work (Brenda Tevalan, personal communication; seeTevalan 1999 and IPM CRSP: 38-42). Only 4 percent ofwomen reported earning agricultural wages outside familyproduction. Male-biased wage differentials reflect the divi-sion of labor by task, with men performing tasks perceivedto require more strength (such as hoe use and spraying agro-chemicals from backpack sprayers). Given the access womenhave to proceeds from household agriculture (see below),their agricultural labor brings greater returns if applied tohousehold production.

Culture & Agriculture Vol. 23, No. 3 Fall 2001

Table 5Women's Marketing and Participation in Production Decisions by Level of Household Ntae* ProductionN = 87, Chimaltenango (Guatemala)

Hectares planted to NTAEsby Household

0b N = 21

.06-. 11 N = 3 7

.23-4.74 N = 29

Total N = 87

MarketBerries

%

73(N=15)

63(N=8)

67(N=3)

69 in(N=26)

MarketPotato

%

50(N=4)

38(N=8)

15(N=14)

27 in(N=26)

MarketSnow Peasc

%

0

15(N=34)

21(N=24)

17 in(N=58)

ControlNTAE Income'1

%

71

68

69

69(N=87)

ProduceAnimals d

%

24*

32*

55*

38(N=87)

ControlLand Use d

%

91

68

76

76(N=87)

ControlInputsd

%

48

27

24

31(N=87)

' Nontraditional agricultural exports (snow peas and broccoli).b Produce variety of commercial crops, for domestic market only.c Women did not sell broccoli in the 15 households that produced the crop.d Statistical tests of correlations between level of NTAE production and intrahousehold socioeconomic indicators were computed for households thatproduce commercial crops (N = 87).* Gamma is significant at p = <.O5.

Data Source: IPM CRSP Household Surveys, Chimaltenango, 1998

Despite the domination of landholding and NTAE mar-keting by men, three-fourths of commercial producers re-ported that land use decisions are made jointly betweenmale and female household heads (Table 5). Nearly one-third also reported joint decision-making regarding the se-lection of agrochemicals, an expensive input. This figurepartly reflects the contribution by 46 percent of women tothe purchase of agrochemicals, seeds, and other inputs. If aman does not have money for agricultural inputs, he mustask his wife to contribute from her own income or fromhousehold subsistence funds, even though the funds shecontrols may have originally derived from his crop sales.Women execute or share control of incomes derived fromnontraditional export production in 69 percent of producinghouseholds.5

Women in the Chimaltenango research communitieswere not marginalized from land-use decision processes, nordid they forego their own independent income producingactivities, in households that invested more of their re-sources in NTAE production (Table 5). Statistical tests on theassociation between indicators of women's resource controland levels of NTAE production demonstrate that women aremore likely to have independent income from animal pro-duction, and less likely to supply money for inputs, in

households that plant more land to NTAEs (Table 5). Thesehouseholds tend to have a larger economic resource base,thus women may be better able to invest in their own farm-based enterprises. The level of NTAE production is notassociated with the likelihood that women will make land-use and agrochemical selection decisions, manage incomesderived from NTAEs, or earn income from nonagriculturalenterprises.

Given the pronounced traditional gender division oflabor and management of household provisioning in theGuatemalan population, together with the male-dominatedmarketing infrastructure and reportedly patriarchal culturalsetting, it is surprising to find that women do not appear tobe marginalized from control of land, labor, and incomesrelated to NTAE production. A partial explanation may liein material payoffs from more symmetrical control of re-sources, such as that observed in highland Ecuador. In aQ'eqchi" Maya population in Belize, Richard Wilk (1990)found that households exemplifying a "mutual interest" pat-tern of shared economic decision-making, responsibility, andresources were able to accumulate more capital than house-holds exemplifying a patriarchal pattern of authoritariancontrol with intrahousehold bargaining over individually-controlled resources. When economic decision-making is

Culture & Agriculture Vol. 23, No. 3 Fall 2001

"conceived as a group decision over group resources ratherthan as a process of bargaining between individuals" (Wilk1990:340), each person has a stake in household productivityand becomes more willing to share tasks flexibly, to workhard, and to manage resources carefully. Perhaps householdpatriarchs in Chimaltenango have realized over time theeconomic benefits of shared decision-making. Further studyis needed to explore these decision patterns in Maya house-holds and communities with differing articulations to exportmarkets.

Conclusion

In the central Ecuadorean highlands, women and menhave compelling reasons, both politically and economically,for believing that egalitarian intrahousehold resource dy-namics have served them well. Both the perceived need tocooperate under conditions of past hacienda indenture anda fundamentally Andean preference for dualistic householdheadship underlie egalitarian intrahousehold relations inmarket-oriented households. A number of traditional con-cepts, values, and socio-economic structures enable the en-durance of egalitarian forms. Egalitarian inheritance patternshave been perpetuated together with the conviction thatwomen and men are equally worthy trustees of familywealth. The high valuation of women as economic actors isan important component of women's control of economic re-sources in market-oriented agriculture. This valuation isbased on both cultural tradition and perceived material ben-efits of gender-egalitarian relations of production.

In the Guatemalan communities that were studied, tra-ditional gender complementarity in work and resourcecontrol and a more patriarchal baseline do not afford womenthe structural and ideological bases for egalitariancommercial-agriculture resource control observed in theAndean population. Yet women and men cooperate in a lessasymmetrical manner than expected, despite the male orien-tation of many institutions. Women are more likely to havea farm-oriented independent productive base in householdsthat rely more on nontraditional agricultural export produc-tion, rather than losing ground as household agriculturebecomes more oriented to export markets. Although incomesharing takes the form of two-way transfers, rather than thefemale-administered pooling observed in the Andes, womenappear to have greater access to the income derived fromagricultural export production than expected.

The paradigm positing increasing male control of eco-nomic resources as a general, and perhaps inevitable,

corollary of increasing market orientation among small-holders in Latin America proves inadequate to encompassthe dynamics of intrahousehold resource control in thesesettings. These cases demonstrate that particularities in localpolitical-economic history and socio-cultural institutions, aswell as macro-economic policy environments, profoundlyinfluence gendered outcomes of household agriculturalcommercialization. The "myth of the masculine market"(Hamilton 2000) obscures the productive potential and con-straints of Latin American women working in householdcommodity production and marketing and should not berelied upon in policy formulation.

Acknowledgment

Research in Ecuador was carried out from August 1992-August 1993 in conjunction with National Science FoundationProject, "Farming Systems and Socio-cultural Determinants ofChild Growth in Two Ecological Zones of Ecuador," KathleenDeWalt, Principal Investigator. This project was instituted in col-laboration with Fundacion para el Desarrollo Agropecuario, Quito(FUNDAGRO), a foundation engaged in agricultural research andextension. Quantitative data reported from Ecuador were col-lected by an interdisciplinary team headed by James P. Stansbury.A Fulbright-Hays Dissertation Research Grant supportedHamilton's qualitative field work. Research in Guatemala was car-ried out under the auspices of the Integrated Pest ManagementCollaborative Research Support Program, funded by the U.S.Agency for International Development Global Bureau (AgreementNo. LAG-4196-G-00-5001-00). This paper does not necessarily re-flect the views of the agency.

Notes

1. Men share reproductive labor to an extraordinary degree(Hamilton 1998, Chapter 5). Although women do more of the child-care for the youngest children, men share the burden of foodpreparation, laundry, house cleaning, and fuel gathering.

2. This is according to survey responses from both female and malehousehold heads, informal interviews and behavioral observationsfrom a sub-sample of 10 women and their husbands.

3. These cases describe women in their twenties and thirties.Resource-control patterns are similar in households whose femaleheads are 40-65 years old. Individuals' names have been changed.

4. Women's control of household consumption and accumulationin the study communities is not unique within indigenous com-munities in Ecuador. Ethnographic studies from throughout thesierra report that income is pooled and that women have equalparticipation in establishing household accumulation goals andgreater than equal participation in managing daily expenditures,even in households heavily dependent on migrant males' wages(Alberti 1986; Barsky et al. 1984; L. Belote and J. Belote 1988;Poeschel 1988). However, Mary Weismantel finds that in one

Culture & Agriculture 10 Vol. 23, No. 3 Fall 2001

region of Cotopaxi Province, men no longer feel obligated to turnover wages to their wives, with the result that women may haveaccess to their husband's cash earnings only when the men are athome (1988).

5. The survey did not quantify the proportion of this incomedirectly controlled by women. The proportion of household in-come controlled by women was quantified in a 1994 study of aKaqchikel community near Guatemala City, which showed thatwomen in NTAE-producing households directly controlled 58percent of all incomes. In households that derived all of their in-come from agriculture, women and men each controlled half of theincome. It should be noted that, although many households wereaffiliated with a male-oriented production and marketing cooper-ative, women delivered to the co-op or marketed snow peas in 40percent of producing households and French beans in 60 percentof producing households. (Asturias de Barrios and Tevalan 1996).

References Cited

Alberti, Amalia M.1986 Gender, Ethnicity, and Resource Control in the Andean

Highlands of Ecuador. PhD dissertation. StanfordUniversity.

1988 From Recommendation Domains to Intra-HouseholdDynamics and Back: Attempts at Bridging the GenderGap. In Gender Issues in Farming Systems Research andExtension. Susan V. Poats, Marianne Schmink, and AnitaSpring, eds. Pp. 61-72. Boulder, CO: Westview Press.

Allen, Catherine J.1988 The Hold Life Has: Coca and Cultural Identity in an

Andean Community. Washington, DC: Smithsonian Insti-tution Press.

Arizpe, Lourdes, and Carlota Botey1987 Mexican Agricultural Development Policy and Its Impact

on Rural Women. In Rural Women and State Policy.Carmen Diana Deere and Magdalena Leon, eds. Pp. 67-83. Boulder, CO: Westview Press.

Asturias de Barrios, Linda, and Brenda Tevalan1996 Production Agricola en Santa Maria Cauque. Guatemala:

Asociacion de Investigacion y Estudios Sociales (ASIES).

Barham, Bradford, Michael R. Carter, and Wayne Sigelko1996 Agro-export Production and Peasant Land Access:

Examining the Dynamic between Adoption and Accum-ulation. Journal of Development Economics 46:85-107.

Barsky, Osvaldo, Lucia Carrion, Patricia de la Torre, and LuciaSalamea

1984 Modernizacion hacendal y nuevos roles de la mujercampesina. In Mujer y transformaciones agrarias en lasierra ecuatoriana. Susana Balarezo et al., eds. Pp. 47-146.Quito: Centro de Planificacion y Estudios Sociales(CEPLAES).

Belote, Linda, and Jim Belote.1988 Gender, Ethnicity and Modernization: Saraguro Women

in a Changing World. Michigan Discussions in Anthro-pology. Ann Arbor: Department of Anthropology, Uni-versity of Michigan.

Bossen, Laurel1984 The Redivision of Labor: Women and Economic Choice in

Four Guatemalan Communities. Albany: State Univer-sity of New York Press.

Bourque, Susan, and Kay Barbara Warren1981 Women of the Andes: Patriarchy and Social Change in

Two Peruvian Towns. Ann Arbor: University of MichiganPress.

Carletto, Calogero, Alain de Janvry, and Elisabeth Sadoulet1999 Sustainability in the Diffusion of Innovations: Small-

holder Nontraditional Agro-Exports in Guatemala. Eco-nomic Development and Cultural Change 47(2):345-369.

Chaney, Elsa, and Marianne Schmink1976 Women and Modernization: Access to Tools. In Sex and

Class in Latin America. June Nash and Helen Safa, eds.Pp. 160-180. South Hadley, MA: J. F. Bergin Publishers.

Deere, Carmen Diana1986 Rural Women and Agrarian Reform in Peru, Chile, and

Cuba.. In Women and Change in Latin America. JuneNash and Helen Safa, eds. Pp. 189-207. South Hadley,MA: Bergin and Garvey.

1995 What Difference Does Gender Make? Rethinking PeasantStudies. Feminist Economics l(l):53-72.

Deere, Carmen Diana, and Magdalena Leon1982 Women in Agriculture: Peasant Production and Proletar-

ianization in Three Andean Regions. Geneva: Interna-tional Labor Organization.

1987 Introduction. In Rural Women and State Policy. CarmenDiana Deere and Magdalena Leon, eds. Pp. 1-20. Boulder,CO: Westview Press.

DeWalt, Billie R., et al.1990 Dairy Based Production and Food Systems in Mejia and

Salcedo. The Research Extension and Education ProjectBaseline Surveys Report No. 1. Quito: Fundacion para elDesarrollo Agropecuario.

Fernandez, Maria E.1988 Technological Domains of Women in Mixed Farming

Systems of Andean Peasant Communities. In Gender Is-sues in Farming Systems Research and Extension. SusanV. Poats, Marianne Schmink, and Anita Spring, eds. Pp.213-222. Boulder, CO: Westview Press.

Flora, Cornelia Butler, and Bias Santos1986 Women in Farming Systems in Latin America. In Women

and Change in Latin America. June Nash and Helen Safa,eds. Pp. 208-225. South Hadley, MA: Bergin and Garvey.

Culture & Agriculture 11 Vol. 23, No. 3 Fall 2001

Hamilton, Sarah1998 The Two-Headed Household: Gender and Rural Devel-

opment in the Ecuadorean Andes. Pittsburgh: Universityof Pittsburgh Press.

2000 The Myth of the Masculine Market: Gender and Agricul-tural Commercialization in the Ecuadorean Andes. InCommercial Ventures and Women Farmers: IncreasingFood Security in Developing Countries. Anita Spring, ed.Pp. 65-87. Boulder, CO: Lynne Rienner Publishers.

2002 Neoliberalism, Gender, and Property Rights in RuralMexico. Latin American Research Review 37(1):119-143.

Harris, Olivia1978 Complementarity and Conflict: An Andean view of

Women and Men. In Sex and Age as Principles of SocialDifferentiation. J. La Fontaine, ed. Pp. 21-44. London:Academic Press.

Integrated Pest Management Collaborative Research SupportProgram (IPM CRSP).

1997 Fourth Annual Report. Blacksburg, VA: Virginia Poly-technic Institute and State University.

Inter-American Development Bank (IDB)1995 Women in the Americas: Bridging the Gender Gap.

Washington, DC: IDB/Johns Hopkins University Press.

Katz, Elizabeth G.1995 Gender and Trade within the Household: Observations

from Rural Guatemala. World Development 23: 327-342.Nash, June

1986 A Decade of Research on Women in Latin America. InWomen and Change in Latin America. June Nash andHelen Safa, eds. Pp. 3-21. South Hadley, MA: Bergin andGarvey.

Nieves, Isabel1987 Exploratory Study of Intrahousehold Resource Allocation

in a Cash-cropping Scheme in Highland Guatemala.Washington, DC: International Center for Research onWomen.

Phillips, Lynne1987 Women, Development, and the State in Rural Ecuador. In

Rural Women and State Policy. Carmen Diana Deere andMagdalena Leon, eds. Pp. 105-123. Boulder, CO:Westview Press.

Poeschel R., Ursula1988 La mujer Salasaca: Su situation en una epoca de

reestructuraci6n economico-cultural, segunda edition.Quito: Ediciones Abya-Yala.

SEDRI (Proyecto de Desarrollo Rural Integral Salcedo)1981 Volumen I. Diagnostics Gonzalo Munoz and Roberto

Susman, et al., eds. Quito: Instituto Interamericano deCooperation para la Agricultura (IICA).

Stark, Louisa1979 Division of Labor and the Control of Economic Resources

among Indian Women in the Ecuadorian Highlands.Andean Perspectives 3:1-5.

Stephen, Lynn1994 Accomodation and Resistance: Ejidatario, Ejidataria, and

Official Views of Ejido Reform. Urban Anthropology 23(2-3):233-265.

Tevalan, Brenda1999 La division del trabajo agricola por genero en Santa

Maria Cauque, durante la segunda mitad del siglo XX.Licenciatura thesis.Universidad del Vallede Guatemala.

von Braun, Joachim, David Hotchkiss, and Maarten Immink.1989 Nontraditional Export Crops in Guatemala: Effects on

Production, Income, and Nutrition. Research Report 73.Washington, DC: IFPRI.

Warren, Kay B., and Susan C. Bourque1991 Women, Technology, and International Development

Ideologies: Analyzing Feminist Voices. In Gender at theCrossroads: Feminist Anthropology in the PostmodernEra. Micaela di Leonardo, ed. Pp. 278-311. Berkeley: Uni-versity of California Press.

Weismantel, Mary1988 Food, Gender, and Poverty in the Ecuadorian Andes.

Philadelphia: University of Pennsylvania Press.

Wilk, Richard1990 Household Ecology: Decision Making and Resource

Flows. In The Ecosystem Approach in Anthropology.From Concept to Practice. Emilio F. Moran, ed. Pp. 323-356. Ann Arbor: University of Michigan Press.

Culture & Agriculture 12 Vol. 23, No. 3 Fall 2001