different products which were not of the same relative importance in Kingston and in the
rural areas, a larger proportion of the total being represented by potatoes among the ur-
ban households. Other evidence suggests that both Irish potatoes and sweet potatoes have
a considerably higher income elasticity than the other products in the group, and when this
is allowed for, a low, or even negative, elasticity among urban households for some of the
other root vegetables is not necessarily inconsistent with the survey evidence.

 It must be admitted that the statistical evidence on these points is not as conclusive as
might be wished. It seems reasonable nevertheless to take the differences in expenditure
on root vegetables between rural and Kingston households (expenditures are consistently
higher among rural households) as giving an indication (when considered in relation to the
difference in total expenditure between the same groups) of the change to lower expenditure
on root vegetables that would be associated with the increase in income experienced by
families moving from the rural to the urban section of the population. In relation to such
a change in income the elasticity of demand for root vegetables is obviously negative. Now
a continued rise in average income levels in Jamaica seems certain to be associated to a
certain extent with growing urbanisation (the term being taken in a broad sense to include
changes towards a more urban type of diet which may well affect households living in rural
areas as well as those which actually move from the country into the towns). Thus it seems
appropriate, in considering the future demand for roots and starchy vegetables, to assume
that part of the increase in average income which is expected in future years will be account -
ed for by an increase in the urban or urbanised section of the population, whose diet will
conform to the pattern indicated by the urban sample in the expenditure survey, while part
will be due to an overall increase both in urban and in rural income per capital. If so, the
elasticity for projection purposes should be an average of the different elasticities appro-
priate to these two changes.

 In the light of these considerations a small negative value was assigned to the income
elasticity of demand for root vegetables other than potatoes; for the latter, however, a
fairly high positive elasticity was assumed.

 The evident need, in this instance, to take account of the influence of urbanisation on
the pattern of consumption as a factor distinct from the increase of real income by itself
may seem to suggest the desirability of making separate estimates for all foods of the
possible change due to urbanisation and of combining these changes with those expected
from a uniform overall increase in average income in making the final demand projections.
Trial calculations showed, however, that for products other than root vegetables an almost
identical result was reached whichever method was used and that the attempt to take ex-
plicit account of the urbanisation factor added considerable complications to the process
of calculation without appreciably influencing its outcome. This is of course a reflection
of the fact that the urban-rural contrasts in consumption indicated by the 1958 survey, al-
though striking and important in themselves, do not, in respect of most commodities,
differ in any major way from those to be expected on the basis of the difference in total ex-
penditure between the urban and rural groups of households and the average demand elas-
ticities emerging from the analysis. The roots and starchy vegetables are the only major
exception to this statement, and this is the justification for departing from the usual pro-
cedure in estimating their demand elasticities.

 In general, then, the survey evidence has been accepted as the basis for the estimation
 of the income elasticities for each of the different product groups. But in projecting the