In the absence of a national registration of births and deaths. and the lack of
comparability between population census figures at the lower levels of enumeration.
it is impossble to establish for Tanzania the spatial variation in the net gain
or loss in population resulting from migration.! This is likely to remain so for
several decades to come and it seems worthwhile, therefore, to try to estimate the
relative net migration patterns from the available census data. Indeed such an
attempt to map areas of net in- and out-migration has already been made for
Ghana.2 Using the 1960 census data. Hunter found a correlation. at Local
Council level. of r = +0.789 between immigration (as represented by persons
'not born in this locality') and sex ratio of the 15-44 years age group (Le. 'economically
active persons'). He then proceeded to use adult sex ratio as the basis
for delimiting net in- and out-migration regions in Ghana. 3 Furthermore on a
local scale. Hunter has plotted adult sex ratio by Chiefdoms to identify the varying
emigration rates within Nangodi. northern Ghana.4 In a similar way, District
variations in total sex ratio have been related to patterns of in- and out-migration
in Tanzania: high sex ratios being associated with urban centres and rural areas
offering wage employment. and low sex ratios associated with areas of outmigration.