V -/- ----
GENDER ISSUES IN FARMING SYSTEMS
RESEARCH AND EXTENSION
INSTITUTIONAL AND POLICY PARAMETERS AFFECTING GENDER
ISSUES IN FARMING SYSTEMS RESEARCH IN TANZANIA*
Manasse Timmy Mtoi
Farming Systems Research Economist
Sokoine University of Agriculture
ap.*tmant of Bural r-onozm
P.O. Box 3007,
* Paper presented at a Conference on Gender Issues in
Farming &ysaseas Research and Extension held at
Gainesville, Florida, U.S.A., February 26 March 1,
The author is grateful to Dr. I.J. Mind,
Dr. G.I. Mlay, Mr. S.A. Shayo and Mr.
N.S. Mdoe of the Department of Rural
Economy, Sokoine University of
Agriculture, Morogoro for making useful
suggestions and comments on earlier
versions of this paper.
National policy objeotivwes iet-it~tional
arrangements and the organizational structures
of rural communities are recognized to influence
household decision making process. These are
posed to have an impact on the overall community
structure and hence the development of Farming
Systems Research (FSR) process in a recommendation
domain composed of men and women. This scenario
is particularly important both to researchers and
the target group in that women account for about
60 to 80 percent of the labour force in most
dpeeloping countries and they play a dominant
role in smallholder production, often performing
mna han half of the household and family duties
Policy objectives are considered to be
Jn ane of government action in a particular
area, whether based on explicitly outlined policy
objectives or reflections of the government's
implicit policy in a particular area. A dis-
tinction between two types of parameters is
made depending on the nature and characteristics
of rural farm communities in Tanzania:
i) Policy and institutional parameters
which aim at expanding the option
available in peasant decision
making. Principally they involve
endogeneous factors of the human
element namely land, labour,
capital and management; and
,/ In Tanzania with an estimated population of
21 million (1985) and GNP per capital of
about US $ 300, women account for about 50
to 65 percent of agricultural labour force
= 2 =
ii) Policy and institutional parameters
which change the external factors
i.e. the exogeneous factors of the
human element which include market-
ing infrastructure, pricing policy,
restrictions and controls, agri-
cultural extension, etc.
The economic implication of traditional gender
division of labour in the FSR project in Morogoro
District, Tanzania is reviewed together with the
risk associated with transferring excess female
labour during critical labour period to add to
male labour during the same period. The frame-
work of analysis looks at how FSR and Extension
address the diatinction between resultant and
expected roles of men and women, restrictions on
labimr Arnat~ c 4mohility) and access to credit
and marketing opportunities, Data pertaining to
this satdy were collected in 1982 through 1984
from the FSR project area in the Eastern Uluguru
Mountains of Morogoro District.g/ Quadratic
programming model was used as a tool of economic
Admittedly the subject matter of this topic
is relatively complex particularly due to the fact
that FSR has not adequately developed a proposal
for dealing with the culturalpolicy and support
system perspectives which affect the rural com-
munities and gender issues. This complexity is
further compounded by the fact that FSR
2/ The FSR Project in Morogoro District, Tanzania
is supported by grants from the International
Development Research Centre (IDRC) of Canada and the
Sokoine University of Agriculture. The project
started in 1981 and is expected to wind up in 1988.
activities are conducted under a great variety of
environments, institutional, political and socio-
economic settings 10o.
2. POLICY AND INSTITUTIONAL PARAMETERS REGARDING
THE ENDOGENEOUS FACTORS
These include those parameters aimed at
increasing efficient use of existing farm resources
namely land4 labour, capital and management. Thus
the parameters major role is to introduce new
technological packages which within the context of
FSR will improve the existing production system.
Figure 1 outlines a conceptual framework and
relationships of institutional, cultural and policy
parameters posited to have an effect on gender
issues in FSR process.
2.1 Land Ownership and Use
The Tanzanian "village act" requires each
village members male o female, to be allocated
separate plots of land and to cultivate crops
which are designated in the local village by-laws.
This is revealed, for example, in the way patriar-
chal relations have been reconstituted at the level
of nuclear families or much contracted extended
family network. Whereas centuries ago land was
controlled and allocated to household or extended
family heads by patriarch leaders of clans, today
land is owned by the State and allocated by village
government to village members. However, in many
areas of the country the practice of village
governments is to allocate land to male household
heads, who then allocate land to wives and unmarried
sons and daughters as well as to young, newly married
sons. In the FSR project area of Eastern Uluguru
mountains, user rights still predominate and the
original user, who is often a male, has some hold
over the land even if it is under fallow.
Major crops grown in the area include coffee,
maize, coloyams, bananas, rice, cassava, and pine-
apples. Table 1 shows the system of land use in
2.2 Traditional Gender Division of T,Ah-
Detailed comparative studies of the patterns of
the division of labour and the relative influence of
diff-erent family members on decisions relating the
farm are numerous in literature. Anthropological
studies have indicated that there is a high degree
of variation between cultures in the allocation of
different tasks to one sex or the other. However,
recognition of the importance of women's role in
African agriculture is patterned in several research
findings;3 5 9 ?1T. Labour time studies in Tanzania
have shown that the major labour input of women
peasants is into food production as compared to
production oriented solely to the market which is
dominated by men 4 11J. Table 2 shows tasks
performed by different sexes in the farming
system of Eastern Uluguru mountains,
=. 5 =
On average men tend to specialize predominantly
on cash crops which include cassava, pineapples,
bananas and cocoyams. Women on the other hand,
are mainly responsible for food crops namely rice
and maize. In 1983/84 cropping season men controlled
about 80 percent of the average gross income of US $
754 even though women contributed about 45 percent
of the total farm labour requirements (Table 1).
The exclusion of women from cash economy tends to
lower their position vis-a-vis men. To a certain
extent, cultural beliefs and norms happen to
exacerbate the situation. For instance, some
families in the F&R project area of Morogoro
District indicated some cultural restrictions on
female labour mobility from food crop to cash
crop production and marketing. On the other hand,
male labour mobility to other activities such as
fetching water, collecting firewood and cooking
is completely considered to be a taboo.
2.2.1 Economic Implication of Division of Labour
Risk-quadratic programming (QP) model was
used to test the significance of female labour on
expected risk and productivity in the farming
systems of the Eastern Uluguru mountains ./ The QP-
risk analysis is based on two systems of farm
production namely the traditional system and the
3/ QP as developed by Markowitz _73 has been suggested as
a useful model to consider expected returns under con-
ditions of risk and uncertainty in farm production. It
minimizes directly the variance of the objective function
given a set of linear resource constraints.
= 6 =
alternative system. The alternative system is an
emergent system of production following adaptive
experimentation results on cassava and pineapples
from 1981 to 1983 in the FSR project area.
The families surveyed in the area had an
average of 5.4 members with a spread of family
size of 2 to 11. Labour constraint formulation has
been considered in two categories; family labour
and hired labour. Family labour included all
working man-days available as estimated from the
farmer and his family. Furthermore, labour was
diaggregated into three parts: November to March;
April to June; and July to October; to reflect
different opportunity costs of both male and
female labour. Although labour restraints can be
formulated in the model for evey month of a year,
this implies a rigidity in the timing of farm
operations and consequently of labour use which
may be unrealistic. Thus labour restraints were
formed,to focus on.those periods of the year in
which labour allocation by gender is most critical.
The remaining non-critical periods were also in-
cluded to provide a complete accounting for labour
in the farming system.
The number of man-days available for farm
work per year was adjusted by taking into con-
sideration days that were solely committed to
non-farm activities, and public holidays as
shown in Table 3. The total average effective
adult equivalent man-days available for farm work
in the 1983/84 period was found to be 573. About
45 percent of this labour force was comprised of
The alternative farming system had efficient
packages of labour utilization with respect to two
cropping enterprises: cassava and pineapples.
Though labour requirements were slightly higher
in land preparation and planting,. requirements
for weeding were less because of spacing and
mulching practices./ Table 4 summarizes labour
requirements in the alternative system of produc-
tion and the resultant yields. On average about
37 percent of the total labour requirements on
cassava production is expected to be contributed
by females (basing on traditional gender division
of labour). All labour required in producing
pineapples is contributed by males. Pineapples
and cassava yields increased by about 32 percent,
4/ The alternative plant spacing for cassava and
pineapples was so chosen on the assumption that
plant population is a basic determinant of crop
yield per unit area was not optimum in the
traditional cropping system. The choice of
plant arrangement was on the basis that both
crops have a shallow root system by virtue of
method of propagation (cassava) and the
monocotyledonous nature of crop species
(pineapples) so that a triangular configuration
coupled with mulching averted soil erosion:and
reduced weed growth.
The QP-risk model involved the transfer of 10
adult equivalent man-days of excess female labour
restraint for land preparation and planting during
the November-March peak period to add to the male
labour in the same period. The expected net farm
income (E) would increase from US S 230.56 to US
394.44 if the alternative system is undertaken.
This is an increase of about 71 percent when
compared to the traditional system. The expected
net farm income would decrease from US $ 394.44
to US $ 338.89 ( a decrease of about 14 percent)
if female labour mobility is restricted in the
alternative system. Table 5 and Figure 2 indicate
these changes would be effected at approximately
same levels of risk (measured in terms of
standard deviation of net farm income in US S
dollars The analysis further suggested that
female labour transfer would ensure relatively
higher efficiency ratios of E to farm resources
namely land, labour and capital 92 .
The substantial increase in E can be
economically and technically achieved in the
farming system of the Eastern Uluguru mountains.
This is because the alternative system does not
significantly contradict the existing system in
terms of additional requirements for farm
resources. The difference in the two systems
lies on the application of basic agronomic
practices such as plant spacing and
configuration, timing of critical farm operations
3. POLICY PARAMETERS REGARDING EXOGENOUS FACTORS
The distinguishing characteristic of the policy
parameters in this category is that they define
the socio-economic environment in which the
farming systems operates. Generally these policy
parameters are designed to steer production in
those directions deemed desirable by the national
policy-makers by providing guidelines for produc-
tion and creating a supporting infrastructure.
Thus they are likely, to influence the endogeneous
factors and genders'expected responsiveness to
technological change in view of the constraints
and expected roles of men and women discussed in
the preceding sections.
3.1 Marketing Infrastructure
Meaningful FSR requires an efficient market-
ing infrastructure that involves use of chemical
inputs, availability of credit, adequate trans-
port, etc. Technological packages associated with
significant use of purchased inputs usually become
a constraint to the limiting cash resources of
small farmers and hence would require credit
facilities. Two inter-related problems in this
area exist. Firstly, the failure of many small
farmers to adopt packages accompanied by
S 10 =
relatively large -a!.ounts of purchased inputs and
secondly, the widespread defaulting of many farmers
who have received concessionary loans for inputs.
Thus the successful adoption of an input in a
particular farming system will depend on whether
input use is rational for individual farm house-
hold in the light of its constraints and its
priorities as well as the complex set of gender
3.2 Pricing Policy
In Tanzania prices for export crops are set
essentially by estimating probable future export
prices and substracting marketing costs from
probable export earnings to arrive at unit
producer price. Prices for the preferred and
drought staples, and for oil seeds, and beans are
set with allowance for quality differentials.
There are economic as well as political impli-
cation regarding these prices both at micro and
macro levels. The large number of price controlled
crops and the great versatility of agro-economic
zoned over which they apply create a high pro-
bability of unintended production consequences in
some areas for some crops. Macro-level analysis
on crop production indicates that Tanzania
peasants are indeed price responsive J6 .
At micro-level the price formulation usually
influences the shift of farm resources to crops
which enjoy relatively higher prices. This
is likely to affect coffee, rice and maize in-
the farming system of the Eastern Uluguru
mountains and hence.gender's expected
responsiveness to'price-formulations. Obvious
problems':are likely to appear in a case where
FSR,:programme concentrates on a low priced crop.
3.3 Administrative Restrictions and Control&
', There is a range of.controls'which include
minimum,-aoreage by-laws, food movement' rest
rictions,. etc. An example of such a -ituatiQn
is the enforcement of a minimum acreage by-law
for coffee (export crop) in the PSR project
-area. Tricing signals, and:in some cases.
"-agronomic conditions, have worked against the
.growing of coffee and in favour of the more
reomnerative- and.:-ess risky cultivation of
pineapples: as a potential'cash crop. -
Adherence: to a -sich by-law. usually results
into planting-the-required coffee acreage,
subsequently leaving the plot unattended.
''hi ''ea3y -d onset e- tremendeouse waste
of both human resource, particularly male', .
labour involved' in. land preparation; -and --
-landr:resource. .Policy makers might ;e made..:.
:to recognized the'wasteful..effacts thPough.-
'intOrmation -cllected from PFER tworkr. in
th- nation -. :,
3.4 Agricultural Extension Services
The cultural environment in rural Tanzania
makes it difficult for male researchers or
extensionists to make effective contact with
women farmers. By virtue of their percentage
composition and long involvement in agriculture,
women have a store of indigeneous technical
knowledge which researchers could tape. /
There is lack of well organized and
articulated extension system to take advantage
of FSR programmes in the country. De Vries 123
recognized the necessity to link extension
programmes with FSR efforts. Though extension .
is coordinated at a national level, extension
programmes within regions could be organized
on a zonal or farming systems type basis thus
creating a good link between extension
planning units and PSR units.
This FSR study has revealed several
important facets which need to be incorporated
in FSR and National programmes in order to
improve -research efforts. Since there are
separate economies between males and females
2/ Only 4 out of the 23 participating farmers in
the FSR project in the Eastern Uluguru mountains
= 13 =
in rural households, any improvement in the
cash crop sector would mean a potential
increase in the income of males at the expense
of females who usually have to help in the
cash crop fields without corresponding income
benefits for their labour.
The study has also demonstrated that
improvements in labour efficiency during
critical labour deficit periods are likely
to raise labour productivity. Farmers need
to show flexibility in terms of labour
specialization by gender.
It is prudently necessary to appreciate
the organizational structures of rural house-
holds, if results from FSR are to bring
advantage to the community at large. The
starting points for such an understanding
would seem to be the pattern of resource
ownership and use, the character of the
productive system, the general intrahouse-
hold dynamics and the national policy
- 14 -
1. Boserup, Easter
1970 Women's Role in Economic Development New York: St.
2. De Vries, James
1981 Agricultural Extension and Farming Systems Research.
Paper presented at a Conference on Parming Systems and
FPaaring Systems Research in Tanzania held at AICC
Arusha April 14-16, 1981.
3. Due, Jean and 2. Anandajayasekeram
1982 Tomen and Productivity in Two Contrasting Farming Areas
of Tanzania. University of Dar es Salaam, Bean/Cowpea
ORST Project Report E-228, July 1982; Department of
Agricultural Economics, University of Illinois at Urbana-
4, Kamuzora, C.L.
1978 Constraints to Labour Time Availability in African Small-
holder Agriculture: The Case Study of Bukoba District in
Tanzania. University of Dar es Salaam, =9B Seminar paper
5. lemming, G.I,
1983 Vomen in Agricultural Cooperpatives, 7CARE D Pollow-up
Programme, Human Resources, Institutional and Agrarian
Reform Division, PAO, Rome, 1983.
6. Marketing Development Bureau
1976 Price Policy Reoommendations for 1977/78, Agricultural
Price Review, Dar es Salaam, July, 1976.
7, Markowitz, H.
1959 Portfolio Seleotion: Efficient Diversification of
Investments. New York: John Wiley and Sons, 1959.
8. Masoarenhaa, Ophelia and Marjorie Mbilinyi
1983 7omen in Tanzania: An Analytical Bibliography.
Scandinavian Institute of African Studies, Upsala, SIDA,
9. Mtoi, Manasse
1984 Comparative Analysis of Resource Allocation Under Risk
and Uncertainty in Morogoro, Tanzania: Unpublished I.So.
Thesis submitted to the University of Guelph, Canada,
10. Mtoi, Manasse and Akwilin Tarimo
1985 Bringing National Policy Objectives to Bear in the Course
of the Locally Oriented Farming Systems Research Process.
Paper presented at the Regional 23S Workshop held at
Egerton College, Kenya, August 19-23, 1985,
11. Ngalula, T.K.P.
1977 Women as a Productive ?orce in the Tanzanian Rural Society:
A Case Study of Duhungwa Village in Mwanza District.
Unpublished M,A. dissertation, University of Dar es Salaam,
12. Saohs, Carylony
1983 "The Invisible Farmers": Jomen in Agricultural Production
Roman and Allan held, Totawa, New Jersey, 1983.
Figure 1.Conceptual framework of institutional,cultural and policy factors affecting intra
household dynamics in FSR.
Self sufficiency food N
Improving the standard of
Increasing net farm income
SFair and stable returns to
* Better use of available reso
Restrictions and controls
Cultural parameters e.g.
norms, beliefs, division of
' Objectives & priorities
SEndogeneous factors e.g.land,
labour & capital
: AVailbl tehdnoogy
- Products coaumption
Inventory of available
Figure 2.E,V Relationships generated by traditional and alternative farm
= Alternative with labour transfer
= Alternative without labour transfer
100 150 200
250 300 350
Expected net farm income (E US.#
Table 1. Monthly labour requirement for form operations for major Crops in Eastern Uluguru Mountains a/
Crop Mean Total per Total Gender Contribution
area hectare MALE FEMALE
ha man-days JAN FEB MAR APR HAY JUN JULY AUG SEP OCT NOV DEC man-days
Coffee .44 110.0 5.5 4.9 3.3 12.7 7.7 3.9 4.9 5.5 48.4 20.8 27,6
Maize .64 72.5 9.6 7.2 6.4 6.4 8.4 8.4 46.4 8.4 38.0
Cocoyams .56 141.0 29.4 6.2 6.2 6.2 31.1 79.0 60.4 18.6
Bananas .36 106.5 2.3 4.9 3.5 3.7 3.5 3.1 3.2 4.1 4.0 5.1 38.3 28.1 10.2
Rice .52 145.0 11.4 11.4 8.8 8.8 2.6 2.6 12.7 2.7 4.6 75.4 10.0 65.4
Cassava .64 117.0 14.8 17.8 8.0 8.0 8.0 11.7 6.6 74.9 47.9 27.0
Pineapples .40 118.3 10.6 5.5 4.5 4.5 9.6 12.6 47.3 47.3 -
Labour requirements 72.2 51.7 36.8 13.3 18.9 25.1 20.0 3.1 11.9 35.8 16.4 73.8 409.7 222.9 186.8
Labour available 51.7 51.7 51.7 51.8 51.8 51.8 51.8 51.8 51.8 51.7 51.8 51.8 621.0
Surplus/Deficit b/ (20.5) 0.0 14.9 38.5 32.9 26.6 31.8 48.7 39.9 15.9 55.4(22.0) 211.3
Source: Farming Systems Research Project, 1984.
a/ Labour requirements estimated from a sample of 46 farmers.
b/ Figures in parenthesis represent deficit labour.
Table 2. Tasks Performed by Gender, by Crop and by Activity in the Parming Systems of EaStern Uluguru Moun1
Category Cassava Pineapples Bananas Coffee Cocoyams Rice Maize
Planting Weeding Weeding Weeding Land Preparation Land Pre
Women Weeding Harvesting Weeding Weeding'
Transporting Bird Scaring Soaring
Processing Harvesting Harvestir
Drying Transporting Transpor
S Cooking Cooking
Land Preparation Land preparation' L/prephration I/preparation'
2 Weeding Planting Planting .'Plonting
Men Harvesting. Weeding scaring wild Marketing
Transporting Harvesting animals
Marketing Transporting Harvesting
preparation Land prep
'I i --- -
- ~~ ~--iL--i----L-.~CI~~-13---iI
Table 3. Estimation of Effective Adult-Equivalent Man-days Available per Month for Farm Work.
Days Unavailable for Farm Work Man-days
Month Days Sundays b Public Rainfl Days Males Females
and Fridays Holidays Days
November 30 8 2 1 1 18 26.17 21.58
December 31 8 4 2 3 14 20.35 16.79
January 31 8 4 1 4 14 20.35 16.79
February 28 8 2 1 1 16 23.26 19.19
March 31 8 1 4 18 26.17 21.58
April 30 8 1 1 5 15 21.80 17.99
May 31 8 1 1 3 18 26.17 21.58
June 30 8 2 20 29.07 23.98
July 31 8 2 1 20 29.07. 23.98
August 31 8 1 2 20 29.07 23.98
September 30 8 1 21 30.54 26.18
October 31 8 1 22 31.98 26.38
Total 365 96 22 10 21 216 314.00 259.00
Source: Farming Systems Research Project, 1984.
a Wednesday are committed to communal activities such as building dispensaries, schools, etc.
Friday are committed to marketing.
Rainfall days are based on 1971 1980 average from Morogoro District.
Table 4. Labour Requirements and Yield on the Alternative System of Production.
Yield per Labour Requirements. Gender's Expected Contribution
Hectare Land Planting Wdin Harvesting
Category Preparation eeand Total ale Female
Pineapples; Technology: .75m x .75m
Researcher Managed Trial 5119.00 28.15 25.03 32.36 33.08 118.62 118.62
Researcher-Farmer Managed Trial 4912.00 28.15 25.03 32.36 32.59 118.13 118.13
Farmer-Managed Trial 3890.00 24.44 25.78 44.44 25.19 119.85 119.85 -
11Cassava; Technology: 1m x 1m
Researcher-Managed Trial 5870.00 39.88 26.68 42.97 54.21 163.24 102.38 60.86
o Researcher-Farmer Managed Trial 5438.00 35.57 24.13 42.71 38.30 140.71 95.23 45.48
Farmer-Managed Trial c/ 4458.00 26.68 15.49 35.30 36.20 113.67 80.53 33.14
Source : Farming Systems Research Project, Morogoro District 1984.
SBased on expected roles of men and women in the farming systems of Morogoro District.
SBased on farmer's traditional spacing of 0.92m x 0.92m.
Based on farmers traditional spacing o 1.14m x 1.14m.
Based on farmers traditional spacing of 1.14. x 1.14m.
= 21 =
Expected Returns and Standard Deviation of
the Two Systems of Production.
System of Expected Net Farm Income ..Standard Deviation
Production Minimum Maximum Minimum Maximum
US. $ US.
System 41.67 230.56 7.65 10.88
System AL 150.00 338.89 7.90 10.30
System B/ 172.22 399.44 8.01 11.16
.a/ Alternative system A refers to an an alternative
production system without female labour transfer.
b/ Alternative System B refers to an alternative
production system with female labour transfer.