14
wife of the household head as opposed to a
junior wife or a household head wife, may be an
important proxy for intrahousehold bargaining
power."
A number of studies have suggested that
household food intake is a poor proxy for indi-
vidual intakes, as correlations between them
may be quite low. Wise (1992), for instance,
cites research by Garcia and Senauer (1992)
from the International Food Policy Research
Institute (IFPRI) which indicated that linkages
between household and individual food con-
sumption measures are quite low in the Philip-
pines, with the correlation between preschooler
calorie adequacy and household calories per
capita estimated to be only 0.42. None of the
household-level indicators measured in their
study were determined to be good proxies for
the nutritional status of individual high-risk
members of the households in the Philippines
food subsidy pilot program. Wise also cites
Staatz et al. (1991) who found that individual
nutritional well-being could not be accurately
measured by household level indicators.
Unfortunately, little is known about the re-
lationship between what is produced and pur-
chased by whom, and what is actually con-
sumed by individual family members (Wise
1992). Researchers have seldom tried to mea-
sure food consumption by individual family
members with some exceptions (e.g., Haddad
and Kanbur 1990; Behrman and Deolalikar
1987; Pitt et al. 1990). Instead, surveys of Af-
rican households have tended merely to ob-
serve total household consumption, and use in-
herently untestable assumptions about
distributions within the household (Thomas
1992; Hoddinott and Haddad 1991).
An important question addressed in the lit-
erature has been to what extent do increases in
household calorie intakes correlate with calorie
intakes for children. For instance, Kennedy
(1989) found in a Kenyan study that, although
increases in household income and calorie con-
sumption is associated with increases in
children's calorie intakes, the link between them
is weak. In fact, Kennedy and Bouis (1993)
observe that "a doubling of household income
in Kenya and the Philippines resulted in an
increase in preschooler energy intake of only 4
percent and 7 percent, respectively. This was in
areas where the child's diet was 20 to 30 per-
cent below recommended levels. Thus, quite
large percentage increases in household income
would be needed to fill the energy gap via the
income/household calorie/child calorie link."
Bouis and Haddad (1990) have also pointed
out in their Philippines study that a large share
of "the extra calories that were available at
higher incomes went to adults, who were al-
ready meeting their recommended intakes of
calories.
*
Preschool children (once breastfeeding
had been stopped) at all income levels con-
sumed well below their recommended calorie
intakes.... Regressions show calorie intakes of
preschoolers to be positively and significantly
related to their nutritional status. Yet higher-
income households choose to purchase non-
food items and higher priced calories at the
margin, while preschoolers continue to con-
sume well below recommended intakes."
The findings from Kenya and the Philippines
in the previous paragraphs may certainly warrant
a reassessment of the methodology used. These
results may be consistent with a number of causes,
in addition to the possibilities implied by these
analyses of widespread callousness or ignorance
by parents regarding their children's needs (a
suggestion which seems intuitively unlikely to
the author of this report). While there may be
behavioral factors involved, there may also be
methodological reasons for these findings. In par-
ticular, the issue of the appropriateness and va-
lidity of the Recommended Daily Allowance stan-
dards being used should be closely considered.
This issue is discussed further in section 4.2.
Another complexity in examining links be-
*
However, Strauss (1993) points out that this could
be due to higher incomes being associated with
higher physical exertion which could make energy
requirements higher than so-called recommended
calorie intakes.