THE more detailed census data being revealed by the authorities is throwing up further puzzling trends that need to be examined closely. The population of Karachi, for instance, has been reported at 14.9m, a figure that has baffled many who believed it ought to have been much higher. For almost a decade now, it had been argued that Karachi’s population was around 20m — although the estimate may not have been based on a comprehensive population count, a number of factors such in-migration led many to infer that this was a realistic figure. It is unfortunate though that many politicians are throwing numbers around as if they have counted all the people themselves; for example, Farooq Sattar of the MQM-P has claimed that Karachi’s population “cannot be less than 30m”. If we were to take that contention seriously, it would be interesting to see what sort of in-migration and fertility rates would be required to go from 9.3m in 1998 to 30m in 2017. The same is the case with much of the chatter from other political parties.

As the figures now being revealed show, Karachi’s population appears less than expected because two important districts that are integral parts of the city have not been counted: Malir and Korangi. The larger figure for Karachi division is 16.05m which includes the supposedly ‘rural’ population of the two districts mentioned above. The definition used for urban Karachi has excluded these districts. On the other hand, the definition used for urban Lahore has seen the inflating of the size of the urban population. This is the main reason why the urban populations of Sindh and Punjab appear out of sync with what people were expecting. Instead of rejecting the results out of hand and alleging large-scale undercounting (in Sindh’s case), it would be better to make a sound argument for why the new figure does not seem accurate.

What is even more puzzling are the provincial shares of the total population. The way Sindh has stayed constant, despite massive in-migration in the long intercensal period, invites questions. Even in the last census, the provincial shares had remained broadly the same over the intercensal period, leading many to claim that the numbers had been adjusted to ensure that no corresponding adjustment in the seat shares in parliament should be undertaken. This time, too, the total population share has not moved for Sindh, while the share of the rural population has increased markedly by more than 50pc. This is in sharp contrast to Punjab, where rural population grew by far less than the national average. These divergent outcomes need more focus. Changing definitions cannot explain most of this divergence, so a clearer picture is required from demographers and researchers.

Published in Dawn, August 30th, 2017