Intra-household dynamics and state policies as constraints on food production : results of a 1985 agroeconomic survey in Cameroon

Material Information

Intra-household dynamics and state policies as constraints on food production : results of a 1985 agroeconomic survey in Cameroon
Series Title:
Conference on Gender Issues in Farming Systems Research and Extenion, University of Florida, February 26 to March 1, 1986
Henn, Jeanne Koopman
Place of Publication:
Gainesville, Fla.
University of Florida
Publication Date:


Subjects / Keywords:
Africa ( LCSH )
Spatial Coverage:
Africa -- Cameroon

Record Information

Source Institution:
University of Florida
Holding Location:
African Studies Collections in the Department of Special Collections and Area Studies, George A. Smathers Libraries, University of Florida
Rights Management:
The University of Florida George A. Smathers Libraries respect the intellectual property rights of others and do not claim any copyright interest in this item. This item may be protected by copyright but is made available here under a claim of fair use (17 U.S.C. §107) for non-profit research and educational purposes. Users of this work have responsibility for determining copyright status prior to reusing, publishing or reproducing this item for purposes other than what is allowed by fair use or other copyright exemptions. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder. The Smathers Libraries would like to learn more about this item and invite individuals or organizations to contact Digital Services ( with any additional information they can provide.

Full Text

:.- __` l -'. -. --... 7 '

-- at the .jnivers ty londa
Conference on

Intra-Household Dynamics and State Policies

as Constraints on Food Production:

Results of a 1985 Agroeconomic Survey in Cameroon

Jeanne Koopman Henn

Bunting Institute, Radcliffe College
Economics Department, Northeastern University

Prepared for the conference on Gender Issues in Farming Systems Research and
Extension, University of Florida, Gainesville, February 26-March 1, 1986.

Compared to many African countries, southern Cameroon has a relatively

strong traditional food sector which is basically a women's farming system.

In a major review of the agricultural sector published in 1980, however, the

Ministry of Agriculture suggested that the trends of falling food output per

person reported for many African countries over the past twenty years seem

to be emerging in Cameroon (Cameroon, 1980; USDA, 1981). Having studied the

rural economy in southern Cameroon in the 1970s (Henn 1978), I wondered if

the apparent stagnation in national food output reflected, at least in part,

the effects of the tight.labor time constraints against which women farmers

in southern Cameroon must struggle in order to increase output and sales.

To investigate the importance of women's labor constraints, I conducted an

agroeconomic survey among men and women farmers in two villages with very

different conditions of market access. The hypothesis was that if there was

a binding labor constraint, one would not be able to observe a significantly

higher level of female food production labor in a village with good market

access than is observed in market isolated villages.

If women have too little time to expand food production, the state

would seem to have two basic directions for food policy: attempt to raise

the productivity of women's work (either in agriculture or in

non-agricultural activities) or develop policies to attract men into the

food sector. Historically, men's work on food production in southern

Cameroon has been extremely limited in comparison to that of women, but some

Cameroonian observers have suggested that the easing of food marketing
constraints which has resulted from Cameroon's extensive highway building

program of the past five years has attracted significantly more male labor

into food production. If this hypothesis can be supported, the alternative

of focusing efforts on mobilizing male labor in the food sector would seem

fairly promising.

This paper discusses the initial results of a random survey of 40 rural

families which was designed to explore these issues. It is organized as
follows: section one describes the farming system and the research method;

section two presents a comprehensive overview of men's and women's economic

activities, labor times, and associated incomes in an attempt to explain

gender differences in time allocated to food production; section three

explores the effects of good versus poor market access on both women's and

men's food production and sales; and the concluding section discusses the

implications of gender specific constraints for state policy.

Agroeconomic Survey of the Eton Farming System Lekie Department. Cameroon

The villages studied are located 50 and 70 miles northwest of

Cameroon's capital city on the northern border of the dense tropical forest

which covers the entire southern region of the country. The savannah begins

just north of the nearby Sanaga River. The region has two rainy seasons and

two growing seasons. Rainfall is normally adequate for production at just

over 1.5 meters per year. Temperature is relatively constant throughout the

year: the average is 25 degrees centigrade with an annual range of 3 degrees

and a nightly drop of 5 to 10 degrees. The laterite soils are said to have

been extremely fertile thirty or more years ago when low population

densities permitted fields to be left fallow 10 or more years. Today fallow

periods have been reduced to an average of 3 to 4 years, and one often hears

complaints that "the land is poor". Neither fertilizers nor pesticides are

used on food crops. (The great majority of farmers surveyed in this study

expressed the need for government research, extension, and input supply

policies in this area.)

The research sites were two villages with highly different conditions

of access to food markets. One, Bilik Bindik, is located on a major paved
highway which carries heavy daily traffic between the capital city and the

rapidly growing cities of western Cameroon. The second village, Mgbaba II,

has been extremely isolated from national and even local food markets.

(This area of Cameroon has few periodic rural markets, none of which are in

reasonable walking distance from this village.) Until the very month in

which this research was completed, Mgbaba's roads were only seasonally

passable and no public transport serviced the village. Food had to be

headloaded for 1.5 hours to gain access to vehicles capable of transporting

produce to urban markets.

The research instrument was a single interview questionnaire which

structured two to three hour interviews with each adult in a sample of 40

randomly chosen households. The sample base was a village census prepared

with the aid of one young woman from each village who also served as
interpreter from Eton to French. I conducted all interviews while residing

in each village for periods of six and four weeks. The major food

fields--the peanut fields described below--were measured and their areas

calculated using the methodology of the national agricultural census.

The questionnaire specified field types and categories of farming

tasks, asking respondents to recall the number of days (or weeks) spent on

each task and to indicate how many hours per day that task was usually
carried out. From this data, annual labor hours were calculated. While it

is reasonable to view recall data of this type skeptically, it should be

noted that the results presented here on men's annual cocoa hours and

women's annual food hours are within 7% and 11% of the norms established by
a research method in which daily interviews were conducted throughout a

calendar year (Leplaideur, 1978 and 1981). On the other hand, my results

for average men's food hours are 26% lower than those found in the daily

survey. As will be seen, however, the number of hours men work on food is

very small, and since the 1978 study included a region where men grow rice

as a cash crop while only plantain and bananas (crops which require very

little labor) are produced by men in the Eton area, the 26% difference in

labor hours between our studies may be capturing a real difference rather

than reflecting a serious reliability problem. The questionnaire asked

about labor and returns for a comprehensive range of subsistence and

market-oriented production activities. Domestic labor tasks were included

for both men and women, as were major categories of expenditures.

Production data was estimated by asking for the number of sacks, basins

or pails of specified volumes harvested by season or by week. Containers of

various crops were weighed and production estimated in kilograms. Plantain

and banana harvests were estimated by number of regimes per "average month"

and average weights as determined by the national agricultural census were

used to estimate total production. Although an attempt was made to be

comprehensive and all major staples and many vegetables were covered, annual

production estimates could not be made for leafy vegetables, fruit, nuts and


The total sample included 34 men and 47 women. Six of the 40 household

surveyed were headed by women, seven were polygynous. One male respondent,


a wholesale trader and transporter whose business operations were based

outside the village, had to be eliminated from the analysis as an outlier
since his income was over 20 times greater than the average male income for

the remainder of the sample. His wives, however, were retained for parts of

the analysis since their personal incomes and labor times were not

atypical. Four of the 33 men in the analyzed sample held full time wage

earning jobs outside the village and two were essentially retired. All men

in the sample grew cocoa. Nearly all women, even the oldest, cultivated two

food fields per year.

The farming system of the villages studied is typical of patterns found

throughout the southern forest area of Cameroon (Westphal, 1981; Guyer,

1984). Cocoa is the only major export crop and is grown primarily by men.

Women rarely gain access to a cocoa plantation unless they inherit from a

husband to whom they are married in civil law. The 15% of women in this

study who owned cocoa had only 20% as many trees as the average man.

There are three basic types of food fields: an intercropped "peanut

field" which all women cultivate with minimal male help for at least one,

but usually two seasons a year; a "forest field" where men often plant new

cocoa trees as well as plantain and bananas and women plant the tuber,
cocoyam; and a small dry season field of vegetables cultivated by women in a

stream watered area. At least three quarters of the total value of food

consumed in rural households is produced in these fields, the greatest part

by far in the woman's peanut field.

The peanut field is oriented toward subsistence, although occasional

surpluses are sold and certain elements of its complex crop mix (peanuts,

corn, melons, leafy vegetables, onions, tomatoes, cassava, plantain, banana

and cocoyam) may be expanded to respond to market opportunities. The average

size of a woman's spring peanut field is 22 ares (1/2 an acre) whereas the
fall field is only 10 ares, reflecting both a shorter effective planting

season and demands on women's time for the cocoa harvest which coincides

with the fall peanut harvest. Peanut fields yield two harvests over the

course of two years: peanuts, corn and vegetables are harvested after their

three month growing cycle, some leafy vegetables continue to produce for a

few more months, and subsequently the plantain, bananas and cassava which

were originally planted with the peanuts take over the field and yield their

harvests during the second year. The peanut field both supplies the majority

of family food and accounts for the great bulk of women's food production

work. A husband may or may not help a woman clear her peanut field, and men

only very rarely help with other cultivation work. One man in the sample,

however, actually worked a peanut field himself, a most unusual phenomenon

in this area.

The forest field is highly integrated into the process of establishing

a cocoa plantation. Most men who clear a new area of heavy forest to plant

this field will eventually establish a cocoa plantation there. Beyond the

initial field clearing, relatively little labor is required to plant and

harvest the plantain, bananas, and cocoyams cultivated in the forest field;

the main constraint on its extension is access to land. Men normally claim

decision-making power over the disposition of plantain and bananas from the

forest field: some men turn over the harvest to their wives, some sell on

their own account. Because plantain and bananas are perishable and
difficult to transport, only 2/3 of the households in the village with poor

market access had established a forest field or intercropped plantain with

mature cocoa, but 90% of the households in the village on the paved road had
done so.

The dry season field is typically a very small field, perhaps 3 to 5

ares, cultivated in a marshy area where it will have adequate moisture

during the dry season. It is purely a woman's enterprise, has a three month

effective season and seems to be highly oriented toward market

opportunities. Thirty-nine percent of women interviewed in the good market

access village cultivated the dry season field whereas only one woman who

did not plant the spring peanut field planted this field in the market

isolated village.

The size of farms could not be measured directly due to dispersion and

the existence of several parcels of land in different stages of fallow and

(more rarely) in uncultivated forest. Farm size could, however, be roughly

estimated based on knowledge of the number of cocoa trees owned and seedlings

planted (1500 trees = one hectare) and on the method of multiplying the

measured areas of the peanut fields by their two year period of cultivation

plus the number of years the farmer reported to leave them in fallow. This

method produces a minimum estimate of cultivated land since it assumes that

all forest fields are simultaneously planted with cocoa seedlings. If a

household has uncultivated forest land, it would not be reflected in these

estimates. Using this method average landholding per household is roughly

estimated at 4.3 hectares (10.6 acres) with 1.6 hectares in mature cocoa

trees. The six women-headed households have an average of only 2.2 hectares

as contrasted to 4.7 hectares for male-headed households. The size

distribution for the 33 male headed households of the sample ranges from 2.3

to 9.6 hectares while the range for female-headed households is 1.2 to 3.4


From the data on average fallowing periods and from reports of the
unavailability of land for purchase, it seems clear that the highly
populated Lekie department faces a land constraint with current agricultural

technology. On the other hand, a land constraint on production is not yet a

general problem in southern Cameroon. This paper focuses on the more

pervasive problem of highly differentiated labor constraints between men and

women. The following section traces the sexual division of labor between

food, cocoa and other work and presents the survey data on annual labor

hours. It also assesses hourly income in various activities, an analysis

which sheds light on differences in men's and women's choices in allocating

their labor time between food production and other work.

The Sexual Division of Labor and Earnings

To understand the context in which men and women farmers make decisions

about allocating labor time to food production, we need to know about total

labor times. Table 1 presents men's and women's self reported estimates of

their annual working hours. In eliciting this data, detailed and

comprehensive questions were designed to capture seasonal labor patterns and

individual differences in time allocated by task. The results accord

reasonably well with labor time studies conducted in southern Cameroon on a

repeat observation basis (Sedes, 1966, and Leplaideur, 1976). Two sets of

data are presented for men: one for all 33 men in the random sample of 39

households, and one for the 27 full time cocoa planters, a group which

excludes six men--two full time wage earners and one retired man from each

vi llage.

The most striking contrast in Table 1 is between men's and women's

total labor hours. Men's weekly labor is less than 27 hours while women's

is over 63 hours. While much of this difference results from the huge but

not surprising difference in domestic labor time (where women spend 31 hours

a week cooking, gathering firewood and water, and washing clothes, and men

spend 4 hours a week sharpening agricultural machetes, repairing the house

and washing or ironing), a dramatic difference was also found in total

agricultural labor hours. Here women's 24 hours a week on food production

plus 2.2 hours weekly helping with cocoa adds up to an average agricultural

labor week of more than 26 hours, over twice the 12 hours a week all men

average on cocoa and food production tasks. Even full time sale farmers

work only about 14 hours a week in agriculture.

Most of women's agricultural labor time is spent producing food for

family consumption--over 1000 hours a year or 20 hours a week. In contrast,

men on the average spend only 2.6 hours a week on subsistence food

production. Half men's unpaid food work is devoted to helping with the

peanut field and half to cultivating plantain in the forest field. One

often hears that in Africa men do the heaviest work in food

production-cutting trees and clearing new fields in the forest. This

survey found that men did less clearing labor than had been expected. While

men did carry out about 2/3 of the labor necessary to clear a new forest

field (a project most men seemed to undertake every three to four years if

they had sufficient land), it was also found that women spend more than

twice as much time as men clearing the major food crop fields, i.e., the

peanut fields which are cultivated twice a year. Considering all the labor

required to establish the normal set of peanut, forest and dry season

fields, this survey found that men's total labor input into the farming

tasks necessary to produce family food is only about 13% of the time women
spend on this work. The sub-sample of full time male farmers work only 15%

as much time as women producing family food.

Women also spend considerably more time than men in producing food for

the market. Ninety four percent of the women in the sample sold food, but

only 27% of men did so. Food sales are women's major source of cash

incomes; average annual earnings are $271. Most men's primary source of

income is cocoa. The average man's net earnings in cocoa of $760 a year

exceeds by far his $134 of annual earnings from food sales. On the other

hand, men's earnings from food are very high when analyzed on the basis of

hourly return to labor. The low sample estimate of men's annual labor hours

in market-oriented food production (averaging only 33 hours a year) produces

an estimate of hourly earnings of over 1500 cfa or $3.75 per hour, an hourly

return which is over twice the average return to men's cocoa labor. Why,

then, don't men seem to be increasing their food work on a much more

extensive scale?

Several factors are important in the interpretation of the data on

men's market-oriented food production (Tables 1, 2 and 3). First the base

is small. Since only a quarter of the 33 men in the sample actually sold

food, the data has less than desirable statistical reliability. Second, an

extremely important element affecting the high return estimated for male

labor as contrasted with the return to women's market food production is the

fact that men sell almost exclusively plantain and bananas, crops with very

low labor requirements. Third, some men mobilize women's labor for the


plantain harvest with little or no remuneration, thus artificially boosting

estimates of their personal returns. (A similar situation obtains for
cocoa.) Fourth, there is a high risk to growing plantain for the market in

areas with poor roads. If headloading labor to transport the crop to

passable roads cannot be mobilized or if public transport is highly

irregular, output is likely to rot before it can be solds. In sum, the

marketing constraints, the extremely limited crop ix currently produced,

and men's ability to mobilize unpaid family labor have produced an

exaggerated evaluation of men's potential return to expanding marketable

food production. A more reasonable estimate of earnings potential in food

is provided by the data on women's marketed food output.

Women's hourly return to the relatively wide mix of food products they

sell is estimated at about 500 cfa ($1.25) an hour. This may well be an

overestimate of net hourly earnings, however, because the data on food

earnings recorded actual receipts whether earned from sales in the village

or sales in various urban markets. Consequently some of women's transport

expenditures should be deducted from gross food earnings. If half women's
average transport expenditures (which in total account for almost 13% of her
annual earnings) are attributed to food marketing costs, the estimated

average return to women's work in marketed food production is 435 cfa

($1.09) an hour. This is the figure I will use as an estimate of the

opportunity cost of producing food for the market. It is important to

remember, however, that this return can vary considerably due to changes in

crop mixes and differing conditions of market access which affect village
level prices.

I turn now to a discussion of men's and women's alternative sources of

income in order to evaluate their effects as constraints on the allocation

of labor to food production. Men devote an average of 8.6 hours a week to
the tasks of cocoa cultivation--planting, clearing the 1.7 hectare

plantation of brush, spraying the maturing pods with fungicide, and

harvesting. Full time farmers spend just over 10 hours a week on cocoa

work. Male household heads also mobilize approximately the same number of

labor hours from family and friends, principally for harvest processing

where labor is remunerated exclusively with food and drink. Teenaged

children, wives and other relatives often work without direct payment

although the planter may be paying school fees, buying clothing and

providing shelter. The point here is that a cocoa planter typically

mobilizes more labor hours than he himself expends for a very low monetary

outlay. The cocoa depodding labor, which averages 277 hours per planter,
for example, costs an average of only 125 cfa ($0.31) an hour. The planter

himself therefore receives a relatively high net return to his own labor

hours (678 cfa or $1.70 an hour).

A wife's return for helping with her husband's cocoa may take the form

of a cash payment (received by 43% of the women sampled) and/or a gift of

clothing which two-thirds of all men reported giving their wives. Counting

both types of remuneration, the average hourly return to women's cocoa work

for husbands was only 217 cfa or $0.54 an hour which compares poorly to her

estimated opportunity cost to food production of 435 cfa ($1.09) an hour.

These data suggest that women's work on cocoa may be considered as a form of

marital obligation or that women consider their cocoa work as an obligatory
quid pro quo for men's food work.

Men engage in a wider range of income earning activities than are

available to women. The most frequently encountered male sideline activity

is the tapping of oil palm trees to produce a drink known as palm wine. Two

thirds of the men in the sample tap palm wine for one or more months a year,

but only half sell part of their production. Earnings vary considerably

from person to person but the average for those who sell is about a quarter

of men's average cocoa earnings. Hourly earnings to palm wine production

for the market is, at 386 cfa an hour, considerably lower than the return to

cocoa and also lower than the return to food.

Calculated returns to the 33% of the men in the sample who supplemented

their cocoa earnings with artisanal work such as house and furniture

building also varied considerably but again averaged out on an hourly basis

to less than two thirds of the average planter's return to cocoa. For some

men the hourly return was clearly less than the opportunity cost of food

production but on average is exactly the same, i.e. 435 cfa an hour. In

addition to the income generating activities already discussed, four men

from the village with excellent market access worked as manual laborers
digging sand from the Sanaga River for use in construction, one man

established a bar in a town twenty kilometers distant, and one developed a
modern poultry raising project on his farm. With the exception of the bar,
none of these activities brought a return to labor which was higher than

opportunity cost of marketed food production, although all participants

expected higher returns in the future.

Finally, four men in the sample (12%) hold full time wage employment

outside the village to which three commute daily and for which one rents a
room in town returns to the village once a month. These men all have some

cocoa trees and are continuing to plant, but the current hourly return to

their wage labor is only 315 cfa ($0.79), less than half the average return

to cocoa and only 72% of the opportunity cost of food production. It should
also be noted that three of the men have skilled jobs, and that their

average educational level, at 6.7 years, is nearly double the male average
of 3.4 years. The basic explanation for these men's choice of wage labor

can probably best be sought in the trade-off between leisure (or the

part-time which characterizes the typical male villager) and the opportunity

to earn a cash income for 2000 hours a year. The average cocoa planter

spends less than 1000 hours a year in income-generating activities. The men

with wage employment in this sample work an average of 2350 hours annually

(excluding unpaid labor) and their annual income (averaging $2150) is

considerably higher than the average planter's total earnings of $1600.

This review of men's complex mix of non-agricultural activities has
established that most men supplement their cocoa earnings with work that
brings them an hourly return to labor which is less than or close to the 435

cfa an hour women earn in producing food for the market. This being so, we

cannot identify a clear economic reason for men's failure to spend more than

a mere 30 to 40 hours a year producing marketed food. The problem is even

more vexing when one remembers that the particular crops men do sell,

bananas and plantain, bring a return to labor which is considerably higher

than 435 cfa, probably double that amount. In eschewing food production

work, most men seem to be placing a very high opportunity cost on their

leisure. I will return to the analysis of men's participation in food

production in the following section which examines the factor of market

access. First we need to look at women's cash earning activities.

With women there seems to be no question of a high opportunity cost of

leisure. After devoting 52 hours a week to family food production and

domestic labor, the average woman extends her labor time by more than 4

hours a week in order to grow a surplus of marketable food. She also spends

an average of 3.5 more hours a week transforming cassava, a major staple

food of the area, into a less perishable form which can be stored and sold.

Women earn an average of $271 a year by selling food and an additional $167

from sales of processed cassava. Cassava processing is basically an

extension of women's market oriented food production. The amount of cassava

which can be produced in the second year of cultivation of the subsistence

oriented peanut field is usually well beyond family needs. Seventy five

percent of the women in the sample therefore use this surplus cassava to

produce and sell "couscous", cassava flour and "batons de manioc", a cooked

form of processed cassava.

Half the women interviewed also engage In part-time trading activities.

Twenty percent act as village food traders, taking their own and other

farmer's food to urban markets. The amount of time women traders spend in

this activity varies greatly from a maximum of two days a month to a minimum

of two days a year. Childless or husbandless women are highly represented

in this group because a food trader must be free to spend full days and

sometimes nights away from the village. Women farmers who sell in urban
markets, known in Cameroon as "buyem sellem", usually make $150-250 a year
from this work. On the other hand, village women who carry on petty trade

in beer, wine, soft drinks and/or cigarettes during the cocoa season make

much less, $25-50 a year. I cannot reliably estimate hourly returns to

trade since hours spent and profits earned were not well estimated by all of

the women who trade, but rough calculations would give the buyem sellem a

higher hourly return than the opportunity cost of food production. All of

these traders, however, also produced a good deal of food themselves. Trade
in food, like the processing of cassava, is an extension of women's

predominant economic activity in subsistence and market-oriented food

The total profile of women's labor activities is considerably different

from that of men. The major contrast, as emphasized earlier, is in total

work hours and in the share of labor devoted to unpaid tasks. Women spend

over 52 hours a week on family food production and domestic labor which

brings them absolutely no monetary return, whereas men spend less than 8

hours a week in unpaid work. Without engaging in any market activities at

all women work what is normally considered a "full time". The labor time

constraint on women's ability to expand marketed food output is clearly

tight. It may not, however, be binding. On the average women in this

sample work almost 12 hours a week beyond their 52 hours of family

obligations in order to earn a cash income they themselves can control. The
motivation for economic independence among African women is very strong.

Their willingness to sacrifice leisure in order to gain it is amply

demonstrated in a comparison of men's and women's labor times. The

following section pursues this topic by considering differences in food

production labor and earnings from sales resulting from differences in

access to urban food markets.

Market Access: Effects on Earnings, Output, and Labor in the Food Sector

The village of Bilik Bindik is located on a major paved highway with 40

minute access to an important regional market and with one hour access to

Cameroon's capital city. Public transport passes several times a day, and

private cars arrive regularly on weekends to purchase food. Two or more

wholesale buyers of bananas, which are exported to northern Cameroon,

purchase large quantities each week. In contrast, Mgbaba II was until

October, 1985, located on an only seasonally passable dirt road. No public

transport served the village, and to obtain transport for persons or produce

to regional towns or the capital city It was necessary to walk 1.5 hours to

the nearest well-maintained gravel road. Until the construction of the

highway currently serving Bilik Bindik was completed in late 1982, its

farmers had been as isolated from urban food markets as those in Mgbaba.

Since both villages are similar on agronomic and ethnic grounds, we can test

for a response to the change in market access experienced by the farmers in

Bilik Bindik by comparing the sample means of variables measuring labor,

food output and incomes between the two villages. Table 4 presents the

relevant data. Variables for which village means meet the statistical test

of significant differences at the 95% confidence level are indicated with

double asterisks. A single asterisk indicates a confidence level of 90%.

The response to-improved market access by men and by women was

apparently quite different. I begin with the data on women. Differences in

village means for three variables, total agricultural labor time (food and
cocoa), work on food alone (including both family and market food) and

annual hours on food produced expressly for the market, were all

significantly different at the 90% confidence level. Women with good market

access work an average of 1375 hours a year or 26.4 hours a week on food

production, while women who are isolated from major food markets spend an

average of 1148 hours a year or 22 hours a week growing food. These

differences are important evidence that women farmers respond to new

opportunities to increase the income they themselves could control by
significantly extending their labor times. The labor constraint on food

production cannot be said to be binding. One must remember, however, that

by raising their average hours in food production to 26.4 per week, women in

Bilik Bindik have extended their total work week to 67 hours. While the

labor constraint to this point has not been binding, surely it is now

extremely tight. Thus, even though many women told me that they were better

off since the opening of the road, many also complained bitterly of

exhaustion and overwork.

The data on food output also supports the conclusion that women respond

to the lifting of marketing constraints by Increasing food production. To

develop a proxy variable for comparing quantities of food produced and sold

across villages, I multiplied the estimated kilograms of production and

sales by crop by an average price for the two villages and summed across

crops. Subsequent comparison of the village means of the proxy for

quantities of food sold revealed significant differences at the 95%

confidence level. Women in the market-isolated village sell about 27% of the

food they produce, while the average for women who live near the paved road

is 43%.

Comparisons of output by crop showed significant differences in the

quantities of tomatoes and bananas produced in the two villages. Both of

these crops are perishable and require good market access for successful


sales. Bananas are rarely consumed locally since plantain is much

preferred, so the increased output of bananas in Bilik Bindik was a clear
response to changed market factors.

Average plantain output for women was not significantly different in

the two villages, but when men's and women's production was combined,

village means were different at the 90% confidence level. Some men are

allowing their wives to sell plantain from the forest field where men tend

to control the tree crops, but others are reserving the marketable output

for themselves. Women also grow plantain in their peanut fields but

quantities are considerably less than those usually grown in a forest

field. Overall, plantain production by households increases when market

constraints are lifted.

Finally, it should be noted that the average total size of the two

peanut fields cultivated by women each year was not different in the two

villages (31 ares in Bilik Bindik and 30.4 ares in Mgbaba). Furthermore,

women's average output of peanuts (a major ingredient in the daily diet) and

corn did not differ between the two villages. On the other hand, evidence

from sales data suggests that women in Bilik Bindik are growing more leafy

and specialized vegetables in their peanut fields with an eye to market


Women's average earnings from food sales is considerably different in

the two villages. Women in Bilik Bindik earn an average of $398 a year from

food sales compared to average earnings among women in Mgbaba of only $149.
When transport costs assumed to be associated with food sales (50% of
women's annual average expenditures on transport) are netted out of these

figures, and they are divided by estimates of annual hours spent to produce

marketed food, we find that women's economic incentive to grow food for the

market is clearly higher in Bilik Bindik than in market-isolated Mgbaba.
Women's net return to labor on marketed food is over 500 cfa ($1.25) in

Bilik Bindik but only 300 cfa ($0.75) in Mgbaba. Part of this difference is
attributable to the better village level prices and lower unit transport

costs which characterize food markets along the paved roads and part results

from the fact that with easy access to public transport, farmers can shift

their marketed crop mix toward the more lucrative plantain and banana


We have already seen that women significantly increased their labor in

food production in response to the economic .incentives resulting from an
improved marketing situation. Did men respond similarly?

The data for men is contradictory. Some men living along the new

highway now plant and sell plantain and bananas expressly for the market.

Table 4 shows dramatic differences in the mean quantities of food sold and

on the mean incomes from food sales reported by men in the two villages.

These differences, however, cannot be proved significant with a difference

of means test because with only 39% of men in Bilik Bindik and only 12% in.

Mgbaba selling any food at all, standard deviations of the means are

exceedingly large.

Since 39% of the men along the paved road have begun to sell food, we

would expect to find higher labor hours in food production among men in

Bilik Bindik than in Mgbaba. It is here, however, that we discover a major

anomaly in the data. Men's average annual hours producing both family and

market-oriented food are higher in Mgbaba, the market-isolated village (198

hours a year) than in Bilik Bindik (141 hours). These differences do not

test out as statistically significant, again due to the tremendous

variability in men's participation in food production, but they do indicate

that it is highly unlikely that men in the village with excellent market

access have significantly increased their work in food production.

With such variability in the data, it is difficult to assess

differences between the two villages. Nonetheless, it is notable that

women's reports of help from their spouses on food results in an identical

average in each village (59 hours a year compared to 1147 and 1374 hours on

food for women). It is probable that men's input into women's peanut fields

doesn't change as market access changes. The higher average male labor on

food in the market-isolated village (which is calculated from men's own

reports of their labor times) can probably be attributed to the one man in

that village who cultivates his own peanut field for family subsistence.

(It is very rare for men to work a peanut field. The one man in the sample

who did so was the only man in either village to cultivate this classic
"woman's field".)

A better clue to men's changing activity with improved market access

comes from data on quantities of food produced and sold by men in each

village. Men in the market accessible village sell 79% of the food they

produce while men in the market isolated village sell only 18% of their

output. The proxies for quantities of output show about 2.5 times as much
production per man where markets are good, but the amounts which are

available for family consumption are less. It may be that men are taking

advantage of good market access to sell what they would have otherwise

contributed to family subsistence, thereby transferring even more of the

responsibility for feeding the family to women. Such conclusions must be

extremely tentative, however, since we are evaluating changes from a very
small base.

What we can unambiguously conclude from the comparison of labor time,

output and sales data in villages with very different marketing constraints

is that women respond to improved opportunities to sell food both by

increasing output and by increasing labor time allocated to food

production. There is no evidence that women withdraw labor from family food

production or other work in order to produce marketable food. Instead they

simply extend their working day. Thus, while family welfare is not reduced

by women's increased activities in food production, women's personal gain

from the new income obtained is mitigated by their loss of leisure and their

increasing burden of overwork.

Summary and Conclusion

In reviewing the economic circumstances which affect men's and women's

work in the food sector in two villages of southern Cameroon, this paper has

emphasized several points regarding the intrahousehold relations of

production affecting food:

1. Women's total labor time is exceedingly high, while men's is

relatively low.

2. Women's agricultural work is double that of men, principally because

women spend 20 hours a week simply to produce adequate food to feed

their families. Men's labor time contribution to this objective is

only 15% as great as women's.

3. Men produce a very limited range of food products, whereas women

produce nearly the entire range of agricultural products consumed

by rural families.

4. Women allocate seven times as much labor as men to produce food for

the market. When marketing constraints are eased, women respond by

increasing both labor and food output, whereas men do not seem to

allocate more hours-to food but do sell a greater share of their


The hypothesis that women's labor constraint is binding has been

rejected because women have been able to extend their labor days in order to

respond to market incentives. The hypothesis that men will respond to

market incentives to increase food production cannot be unambiguously

accepted because men's increased sales of bananas and plantain under

conditions of good market access seem to have been accompanied by reduced

labor on food for family subsistence. Because men's labor on food varies

considerably by individual and because only a small percentage of men

produce food for the market, larger sample sizes will be necessary to draw

statistically supportable conclusions on this issue.

The data on women's labor, production, and food sales presented in this

study is much stronger. It's confirmation of women farmer's extremely long

labor days highlights the need to pursue policies designed to raise the

productivity of both food and domestic labor if women are to continue

increasing their marketed output. Cooking, firewood and water collection,

and other domestic tasks take up more of a woman's labor day than her

agricultural work. In addition, four fifths of women's food work must be

allocated to providing family subsistence. Even though women spend


increasing amounts of time in market-oriented activities, the patriarchal

traditions which attribute nearly all the unpaid labor necessary to

sustaining a family to women do not seem to have changed. Unless the

productivity of women's work can be increased, the labor constraint on their

ability to expand food output may soon begin to bind.

One might also cite the continuing impact of an implicit patriarchal

devaluation of women's work as a factor inhibiting male participation in

food production. The economic incentive of returns to labor in food

production for the market which are as high or higher than many other male

sideline activities does not seem to have attracted significant investment

of male labor into food production. Thus policies designed to attract male

labor into food may have to envisage a much higher return to labor than

policies designed to stimulated increased food output from women farmers.

Finally, there are equity and family welfare grounds for advocating

policy efforts to increase food production by easing the constraints

currently experienced by women farmers. With cultural traditions of

separate male and female budgets, women's and children's consumption needs

are not expected to be fully covered by men's contributions to household

maintence and food expenditures. Both men and women contribute monetarily

to family welfare. The data which showed women responding to improved

marketing conditions by increasing their work week from 62 to 67 hours

indicates that women are highly motivated to increase their own incomes.

Policies designed to help them do so may contribute to intrahousehold

equity, especially if by raising the productivity of women's food

production, they facilitate both a reduction in women's overwork and an

increase in their incomes.


Bilan Diagnostique du Secteur Agricole
Yaounde: Ministry of Agriculture.

de 1960 a 1980.

Guyer, Jane
1984 Family and Farm in Southern Cameroon. Boston University, African
Research Studies No. 15.

Henn, Jeanne
1978 "Peasants, Workers, and Capital: the Political Economy of Labor
and Incomes in Cameroon." PhD Dissertation, Harvard University.

Leplaideur, Alain
1978 "Les travaux agricoles chez les paysans du Centre Sud de
Cameroun: les techniques utilises et les temps necessaries."

Lepalideur, Alain, G. Longuepierre and A. Waguela
1981 Modele 3C: Cameroun-Centre-Sud-Cacao-Culture ou Simulation du
comportement agro-economique des petits paysans de la zone
forestiere Camerounaise quand ils choisissent leur system de
cultures. Paris: IRA7.

1966 Le niveau de vie des populations de la zone cacaoyere du
Centre-Sud Cameroun. Paris: Secretariat d'Etat aux Affaires

U.S. Department of Agriculture
1981 Food Problems and Prospects

Westphal, E.
1981 L"Agriculture Autochtone au
Zonen B.V.

in Sub-Saharan Africa.


Cameroun. Wageningen: H. Veenman &


M. N4

04 N

61 0 CM C-) -0
0%l %a M M -t
J ON 0 O
v4 -

O Wv

cm V

Z 2 $



< aa

a ca
0 2

I0 <

0 e






eC O 00 N

rC' n )O O

a s- o o4
0,.. 0O
0'..Ou 4

N 0
ir n

NO 0o

0 r- en n
a-i <'

N r

C', 'C
Nu 0

0% m

.. f
a 10








0 0
o 0
0 0

U 0
o o
0 40

o a Li

*O 0c

0 0
O o


= &
O <
0 .
B) -


W 0

L wa -



o .

u- ,

a c
c 3 9
'- 0

o0 1

1.,- 0




CO fn
<- ("


,4 I-



1r4 4

* r-

4 a
0- CO
0 CMr.t


Ir N
* *
r- r-

O -4--
0' NO
-4l N'C

4 a

N s

co C
0 r-
0" em

0 '

0'. '.

uM r-
co '0 0
- No





Li (

0 0
C 0

0 0.
5- C

C ,
1.4 0
o =
m ca

u <

0 0
S 6-

). ^. r-- O0
g 4 r% 0
0 0 441). C40C O'
re Q


a- 1. -4
a b rh ra- i o 4 O
0 a rl en c m o

o a co 'a \o oo
we 04 @o v







t v

w 0

u 0
x a

Z 0





0 0


4 C-4

14 "

0 co0 Ln
r- 1-

4, W-



0 0' 0
0 0 0
N -O

-4 'C10


0 W-4
L 0
N '.

co co


O iO

r- cn

II O 4 c
CM l
0 -4 r- '0

W a **- *-*

d 0b -o 4o so


eO n
-a. -4

. | cU o 4-- n %

n < a n%
C', NCt
l'-4 -
'a a

00 c 0 0
r- oo o 0
fl %Q %0 O'

-4 co ro .4
Sto0 rs i
'- N

0 c0
0 re

0 1 0 %
0 40

tS 0

co ON

r 00
i- c 4

E E o*
0 0 E
o 0 0

0 a *
0 0 d -. 4
S0 0

S' 0 E E m

0a rw m cc0 c 0_ 0w
0 02 0 02 0 2 C w $

u w e) 0
-e o E u 0 0 -
0Ijr O0r- 0n 0 -N 0 0 "
0) O ON V r- O n w1 0 'a 0
Z 4. v CL CLV Lu 0 E4

























* u

* Ll


0 r-

1. 4 0

E II t0Q



0 z

i- KI

0 0

c -. -4 W.-
e 0 0
o V cV .0
S *

Ul )
o p 0 c 1 og
1. 0. 4 f1 0 0

0A (U O U

0 0 v 1
0 0 C (

= 0 *a 0

wo W 4 0

a) 0 0
M O CM 0 0 -.

0 0 0 0 0)
B U 9 .0 0 a *O
&W 0 0 0)

ca = = 0 C L 3 O
0 00 0 0 e

4* 3 Ua m C 0
o 0A0 0v I 0) 0 cc

SZ O z 0 u 0. *
E 4JU 4*1 0 5) 441

(A 0 0 ) (A 5

0 0. 0- 0 8 0 *
z-' 2' ( 0. s 4c<

C 0

Q ) t V N oI 0 04 N 0 )
S o o4 co n m rc- L >
u N M 4 o 0 co CV '0 c Ma N V
U %0 -4 i o
C) h- 1 CO sO -l M 0
U rs 5 N '-
C tT 0 .C
S 11 0

c c C z
0 0. 4

U CY 0r 40' A0 0.( 4
M 4 C) co r r 4 0 o c c -0 o u
0 4i' 4 4 ('t N M 0' 4 cO LA ci 0.
SU e .4 o *W 0o o
0 4 4 CO 0 0' 0 l
*N 4 -4 N VA IA U

l 0 0

) 0
dco co
Z (0 @n
-4 C
( to (0 0
to- C)
>E 54
0C O
cc 0
0 .0 0 *

0 0 O
-4 C U U)I CN4O -4 0' N m 0C r-e CN 0 %0 0.
I ct C C) enen N a' V4a N a 4 o0 o coo
Z a ) o) co Cn en in 1a 13

w C 0' e n- wO
Z 0. -4 11 0|

_ 0 0 E
U)C >
-I 0 4 Z
0 *
C-4 0 4
. 0 C) u a 0

U) 5.4 C) U N '0 '4 0 4 o 0I OD e.n %) )a>
m-' 0 u U s' .. =
C L. C C; '0 0 CA "A -
E E C 0w rC7N C0M A4-'
si. 0 0 0 C V
w C. 0 c
eC s w
0 cc
$* 4 m C

ni un E O O O o- OW 0 O U>
0- 0'0 0 4 0 = 0
-0 *V4 0 I 0 O .0 MW 0 V 0
CO-i 41 0 .0 w 0 C. 3
0 in (A 0 to V w a
-4. w w 9u 4 4-1 4AJ 0 < 0 A *
*0 = = = V t < -4 0 0 C 0 w o0
C O 0 0 0 'O 4*1 C 4u 1Z1 0 w 0 0 a0
0- 4- C) CO C C r r- 0 0 w 0u
it L u a) 3 z 3 i O > C O ws u O
>o o ** *a OB o' O' E i E o 0
..0 wl 0 w- C) P0 0C 0 0 W )O -
(coo m o 0 to w 4 4) ) o u E 5 SC)
L < a 0 o0 0 0 t 0 C 0 w- 0 U 0)
wU 4 U-I a -U0 3 4 -w O U ) u* U
B-4 w 0 >0 0 >1C C) -0 u S 45.4
80 0 XA 3 XU X O U 0 0 (0 .0.
C CCC 0 0 0o uO U w4 *
W 4 0 4 4 4 w w -1 w = *
Ad CL ad 93 I-& CL -i Z z .,.4


o0 c

0 *

0 0
o n

Ia 0 0

.4 n n

Nt u

0 0 00
c 0 0


II M -

Z 00

Cd Q

|0 10
S in -o



0 C

C.) 0
2- P

o 8.
3i4 &

LM 0 0 o c 4 rM "' -T

-4 -
4-* *0 oo an i 4 i 0

60 0 0 0 0 0 0 0 0
o 0 0 co 0 %D D Go
o2 is u n 00) -;T N

; : -4 -4 -4 -4 -4 'V0

?! B- N #4 S 4 #4
is 0c r- co o r- 'o

* 4* *.

0 0 0 4 0 1 -

4 L e n LM 0

mln o- o vrr^ o n<
r- o -i M r



0 x

w -

40 0
lE0 0
j-4| 0

a Q


C 5
4J 00

0 0
a 4.

0 00
1 -1

0 O .L 0

= M U.



b- C

0 u A
S 0

0 C 0C

C) 0
o C. 0)