In the late nineteenth and early twentieth centuries, several
Punjab Settlement Officers attempted to estimate food consumption rates.
These estimates, based on direct observation and ad hoc guesses, were
made partly out of academic curiosity, but more urgently, as an aid in
establishing the land revenue (i.e., tax) rates. The pre-1926 estimates
are summarized in Table I, expressed in pounds of wheat and other
foodgrain consumption per person per year1. Broadly speaking, the later,
more systemtic observers (e.g., Sir Ganga Ram and C. B. Barry), found
lower consumption levels than the earlier observers. It was generally
accepted that the rural populace ate better than urban dwellers. Despite
the ingenuity of the early Settlement Officers, their compiled estimates
suffer from all the difficulties of haphazard small sample observation.
Given the revenue purpose of the estimates, they may be biased towards
the able-bodied, economically active, population. Further, the very
early estimates may have confused dry weight with cooked weight,
including water.
Using Micro Data to Assess Average Tax Rates