The making of a census

Since the last census in 1998, policy makers in Pakistan are depending on estimates and projections based on figures available with them or random sample surveys carried out on a limited scale.

Carrying out a census is different in the sense that data is collected at a specified time from the entire population; in contrast to other surveys, in which information is collected from only a small part of the residents, and from that conclusions are reached regarding the general population.

In Pakistan, the methodology used by census officials in the past has been door-to-door visits to count the number of households and the people inhabiting them. To facilitate this process and distribute the burden of doing this exercise, the country is divided into blocks. Enumerators are engaged to compile record.

Then there are separate days and processes for counting the number of gypsies, the homeless, those travelling, and so on. Enumerators mainly depend on detailed questionnaires drafted for this purpose and handed over to the people to fill up.

This time, too, the country has been divided into around 166,000 blocks and an enumerator and a security personnel will be assigned to gather data from each block that comprises 200 to 250 households approximately. As the security personnel have to come from Pakistan army, there have been suggestions that the country can go for census in phases.

However, Asif Bajwa, the Chief Census Commissioner, has ruled out this possibility, saying phase-wise census will distort results due to the dynamic nature of population. He says the Pakistan Bureau of Statistics (PBS) has proposed completion of the house-listing and population count in 19 days. According to the proposal, in the first three days, house listing would be carried out while the remaining days would be used to count the population, leaving one day to count the homeless, including gypsies.

The housing census, if done properly, helps in identifying illegal immigrants who claim to be Pakistanis but cannot establish link with any household.

Sources in PBS share with TNS that there are two approaches to census enumeration. One is de-jure: in which people are counted at their usual place of residence and two de facto: in which they are counted where they are found on the census data. In 1998 census both the approaches were applied simultaneously during enumeration, but the data was tabulated and published on de-jure basis for comparability of data over time.

Irfan Ahmed, CEO Research Consultants (RCons), a local firm that provides research survey services, says accuracy of the data lies in creating a sense of responsibility among people and creating confidence among them. "Many people in the country are not aware of the utility of census and fear the data is collected for other properties, such as taxation."

Citing the example of South Africa, he says, "Section 16 of the Statistics Act (Act 6 of 1999) obliges a respondent to answer all questions put to them by an officer of statistics. The said clause guarantees confidentiality of the information and ensures that the data collected will be used for statistical purposes only and no one will be able to access it on an individual level."

PPP leader, Senator Taj Haider, differs with the idea of door-to-door census exercise and says that in modern times sources of information, like Nadra, should be relied upon for this purpose. "The declarations made at the personal level are no match to the electronically preserved data. Even the provincial census commissioner in Sindh had to reject the door-to-door exercise undertaken by his own department due to bogus figures," he says.

The opposing argument is that Nadra’s database is over-represented by those who need a national identity card or a passport and it misses out millions of those who do not feel the need to register themselves with it.

As the UN strongly pursues countries to hold population census in order to redefine their development and social uplift agendas, it also recommends best practices about how to carry out this exercise. For example, it recommends use of satellite imagery for household count, etc, as it covers a large area and can be cost-effective compared to other sources.

In some cases, it suggests, digital aerial photographs can be a cost-effective way of initiating some components of a geographic information system. Thirdly, a simple, hand-held global positioning system (GPS) receiver can give latitude and longitude coordinates with reasonable accuracy of key points. A GPS may also track linear features and, thus, be useful for mapping boundaries of areas covered. Though Pakistan is not well-equipped in this context, it can develop expertise in these technologies in future.

Around the world, population censuses try to answer two simple and straightforward questions: How many people live in the country? And, what are the conditions and needs of every household? To achieve accurate figures, different methodologies can be used provided there is no ulterior motive to manipulate the engineered figures.