Food production data provide important measures of welfare (nutritiona
l status). They also provide important policy signals to governments.
The estimation of crop production entails estimating both the area pla
nted to the crop and production per unit area (yield). The usage of ob
jective (i.e., crop-cutting) methods to estimate yields has replaced s
ubjective methods (estimates by farmers and government officials) and
has probably produced improved estimates of yields. Area estimates hav
e not attained obvious improvements in recent years. Consistency check
s between annual sample survey data and census data have been used to
assess the reliability of production data. In addition, consistency ch
ecks between food consumption, production, trade stocks and costs data
provide another index of reliability. These checks, when applied to f
ood production estimates from South Asian countries, show that some im
provement in reliability has probably occurred over time but that cont
emporary estimates of food production remain subject to considerable e
rror. The estimation of food production in the form of livestock produ
cts (milk, meat, eggs) is rudimentary at best in South Asia and this i
s likely to be the case for many other developing countries as well. M
ethods for improved estimates exist and are in use in many developing
countries. The policy value of better estimates for policy making in S
outh Asia and other developing countries is sufficient to call for mor
e resources and better methods to measure food production.