NOW that the census has begun in earnest, it is important for observers to bear in mind that it is too late to start asking for additional items to be included in the list of data that is being gathered. There has been an entire process for preparing the form, and for weeks leading up to the exercise there was an open debate focusing on what is and what is not to be included, with representatives from the provincial governments participating in the conversation. Many in the media, including this newspaper, argued that removing Form 2A from the exercise was a mistake, but none in government or opposition bothered to look into the matter. Now that 55m forms have been printed, and a system has been agreed upon through which to gather the data, a number of politicians have suddenly started issuing proclamations about what is and is not being counted in the census. Functionaries of the state are not errand boys to be tasked at whim, and the census exercise is certainly not a shopping list.

It is true that the credibility of the data is essential for the exercise to be meaningful. But now that the process has been launched, the only way to ensure it is credible is for the provincial governments to set up complaint centres where people can report malpractices. The Pakistan Bureau of Statistics says people should report any malpractice, but there is little information on where these ought to be reported. Second, the credibility of the data will become obvious when it is placed next to data from the previous census and glaring discrepancies are seen, such as how the population shares change between the provinces, and how the ethnic composition of certain provinces alters. While the process is under way, there is little point to issuing statements that evince dissatisfaction at the absence of certain preferred data points, or that ask for the duplication of forms (‘carbon copies’) for verification after the exercise. There was ample time for making those preferences known in the weeks leading up to the exercise, and simply politicising the process at this point will prove counterproductive. An imperfect census is still better than no census at all, and it has taken a lot of prodding to get to even this point. Greater care should now be exercised when issuing statements that cast a pall of uncertainty over the results.

Published in Dawn, March 20th, 2017