worse off. In Africa, several such cases have been
documented. Udry (1996) finds that in Burkina Faso, by
shifting inputs from plots controlled by men to plots
planted to the same crops controlled by women, the total
level of production increased. In Cameroon, Jones (1983)
found that labor was not allocated efficiently across men’s
rice fields and women’s sorghum fields. If the household
could be treated as a single economic actor, then we would
not expect to see these inefficient allocations. The
explanation that is frequently given for this is that women
are not willing to provide additional labor if they do not
receive compensation. Thus, projects that require
additional female labor but provide the remuneration to
the male household head may fail.
Such inefficiencies should not occur if people within
households can trade freely with one another. In the
Burkina Faso case, we might ask why men don’t “sell”
inputs to their wives for part of the increased profit that
would result. The economic models do not seem to resolve
this question: Why isn’t there full trade within the
household? The case studies from Africa suggest that
individuals are not just concerned about maximizing
profits or output in the short run, but that they are also
concerned about maintaining their control over resources
in the long run. Trades within the household might
increase productivity in the short run but change the
balance of resources in the long run. The intrahousehold
models have important implications for empirical analyses
of technology adoption and impacts. They insist that it is
important not to simply assume that the household is the
appropriate unit of analysis. Questions should be
addressed to the farmer; however, the status of the farmer
within the household will affect the constraints faced by
that particular individual. Constraints may differ among
household members. For example, women may have less
ability to mobilize household labor, but this ability may
vary between women depending on their status as wife
(and, within polygamous households, depending on her
status as junior or senior wife), mother, daughter, or in-law
of the household head. Since household resources may not
be pooled, the farmer may not have access to all of the
household income and wealth. Thus, measures of income
and wealth of both the farmer and the household should
be included in analyses.
One way that some studies have tried to incorporate
intrahousehold dynamics is by asking which household
member makes the decisions about different aspects of
agricultural production. This information is particularly
important in studies of technology adoption. Yet,
household decision-making may also respond to economic
changes. For example, Abbott (1976) finds that in Kenya,
the relative decision-making power of the wife increases
significantly when the husband is a migrant. However, the
relationships are more complex than this simplification
might suggest. Some women who are coresident with their
husbands have high levels of decision-making power, e.g.,
older women whose in-laws are no longer living and
women who earn substantial income through
entrepreneurial activities. Meanwhile, some women whose
husbands are absent have low levels of decision-making
power. They live on extended homesteads, have been
married a relatively short time, or have husbands who earn
relatively high incomes. It is difficult to disentangle the
cause and effect relationships. For example, entrepreneurial
ability and household decision-making skills may be
related. Women with high levels of decision-making power
within the household may use their independence to start
their own businesses. In addition, the ability to earn a cash
income may increase her power over decisions within the
household. Yet, the correlation tells us something about
how different groups of women may respond to changes in
their economic environment.
Differences in the decision-making power of women
may also be based on age and status within the household
(Safilios-Rothschild 1985; Zuidberg 1994). Among the
Luo, a recently married woman will farm under the close
supervision of her mother-in-law for many years. Thus,
older women are able to influence agricultural decisions
much more than younger women (Potash 1981). In
northern Ghana, Warner (1997) identifies five measures of
social status within the household for women. Women in
these different classifications have different levels of
decision-making power, including the allocation of their
own labor. Marital status, however, was not found to be
significantly useful in predicting men’s activities (Warner
and Warner1996).
20