how a household is classified as female-headed.6 Frequently,
households are defined as female-headed only if there is no
adult male present. Since male-headed households almost
always include at least one adult female, the distinction
between male- and female-headed households is both an
issue of the gender of the household head and of
household composition. The latter may be more
important. In addition, de facto female heads are included
in some studies as female-headed households and not
included in others, consequently, results may not be
comparable across studies. Finally, comparing female- and
male-headed households provides only limited information
about broader gender concerns because it ignores the
majority of women who live and farm in male-headed
households.
Although there is great heterogeneity among female-
headed households (Peters 1995), on average they tend to
be smaller than male-headed households, have lower
incomes, and are typically less likely to adopt new
technologies (Due and White 1986). It is difficult,
however, to disentangle the cause and effect relationships
among these factors. Almost by definition, female-headed
households are smaller than male-headed households,
simply because any household that contains both adult
men and women is considered male-headed. To the extent
that female-headed households are smaller than male-
headed households and household size is an important
determinant of productivity, we may expect to find that
female-headed households are less productive than male-
headed households (Larson and Kanyangwa 1990). Reverse
causality may also apply: a household may be female-
headed because the farm had low productivity and the
male head left to find better paying work. Thus, the low
agricultural income-potential of the household may
contribute to the household becoming female-headed. The
correlation can also arise when women have less access to
credit and other inputs, in which case a woman head of the
household lowers the household’s agricultural income
potential. There is also a correlation between household
income and technology adoption. Although adopting new
technology may increase household income, some
threshold of income and information may need to be
achieved before a farmer is willing to innovate and adopt
new technologies.
Thus, we would not necessarily expect empirical studies
to present a clear relationship between female headship and
the adoption of technology. Although in some cases there
may be a negative relationship between adoption of
technology and female headship, the finding that female
headship is not significant in explaining fertilizer adoption
in Zambia (Jha and Hojati 1993) should not be surprising.
Examining the gender of the head of household only
captures one component of the many gender-linked
barriers to technology adoption.
Health considerations
A final set of factors that affect the amount of available
household labor is the health of household members.
Illness of household members limits their ability to do
agricultural work and also may require other household
members, usually women or girls, to care for the ill. For
both men and women, it may be important to consider the
seasonal impacts of health as a production constraint. The
rainy season, which is the busy time for planting and
weeding, may also be the time when illness is most likely
(Eicher and Baker 1982). The high expenditure of energy
during this period may also result in energy deficiency and
weight loss (Bleiberg et al. 1979).
The number of healthy individuals in a household is
endogenous. In addition to decisions about fertility and
household composition, decisions about how to allocate
nutrition and health care are made within the household.
Finally, women face the additional burden of pregnancy,
which may limit their ability to perform heavy farm labor.
It is frequently noted that women prefer farm tasks that are
compatible with watching small children, but many tasks
may be quite difficult in the later stages of pregnancy.
Baksh et al. (1994) finds that pregnant and lactating
women contribute significantly less income-generating
labor over two years than other women. Women in
households with the lowest standard of living, while
pregnant and lactating, devote less time to labor-intensive
6 See Rogers (1995) and Peters (1995) for a detailed analysis of the problems with the concept of female headship.