AMID controversy over the process, implication and outcome, the third-party validation exercise for auditing provisional results of the National Census 2017 seems to have reached its natural death.
The final results of the country’s sixth census, conducted last year, are now expected to reach the Council of Common Interests (CCI) for approval, without independent validation, before the current assemblies complete their term in a few days.
On Aug 25, 2017, the CCI approved provisional results of the census, which puts the country’s population at 207.77 million with an annual growth rate of 2.4 per cent. Various stakeholders immediately raised questions over the outcome of the exercise, as some results did not match their expectations.
To satisfy such objections, the CCI agreed to demands for revalidating the census results, and decided on Nov 24, 2017 to have an independent audit of 5pc census blocks.
Final results of Census 2017 are now expected to reach the CCI for approval, without independent validation, before the current provincial assemblies complete their term in a few days
Because of procedural challenges, the statistics division opposed an increase in third-party validation of randomly selected census blocks from 1pc to 5pc, and the hiring of an international or local third-party audit firm.
It said the bidding for hiring an audit firm of repute would take five to six months. Now six months down the road, the position of the statistics division stands justified for varying reasons.
As the selection process of an independent auditor was finalised and bids invited, the prime minister signed a nine-point agreement with senators for a monitoring and oversight committee and a three-member census commission to move the process forward. The agreement entailed random selection by computer of 5pc population blocks for a post-enumeration survey (PES).
A three-member census commission of reputed demographers was to be appointed through mutual consultation to supervise the PES under the monitoring of senators Mushahid Hussain Syed, Mir Hasil Bizenjo, Mushahid Ullah Khan and Taj Haider. The process of the census commission’s formation remained unimplemented.
By convention, a PES is conducted within one month of a census to determine if there are people who have been missed in the enumeration or were counted more than once. All censuses have these exclusion, multiple-inclusion and erroneous entries’ errors.
Even if we want to call it a validation/verification exercise in Pakistan, and not PES, the main objective of this exercise is to assess any under and/or over coverage of a geographical area and population in the census. Therefore, going back to 5pc of the census blocks was essentially aimed at cross-checking the number gathered.
International experts advised the government that given the length of the time elapsed since the census, the PES would simply be an exercise in futility because a year is too long a period for a population to stay put in the same number and location.
Also, it is almost impossible to recall accurately something that has taken place a year back. Retrospective questions after 12 months could be highly inaccurate.
Therefore, many believed the validation exercise would not be viable because of reasons including timing, costs and methodology. The results, in all probability, would add to the existing controversy instead of resolving it.
On top of that, billions of rupees would be needed for an exercise unlikely to guarantee a resolution to the issue.
There was also criticism over the methodology. Instead of paper forms, it was proposed tablets be used in the validation exercise, which was a departure from the census process. It might improve efficiency but could have lead to an even higher variability between the census and post-census figures, feeding further to the controversy.
There were also unresolved questions of whether the army needed to be engaged again. The rationale given the last time for the use of armed forces was to increase the accuracy and credibility of the census. But if it was not a part of the exercise this time, would it further reduce credibility?
Therefore, most demographic experts were against returning to the field. Instead, they proposed that data from the census be thoroughly analysed by experts to judge its quality. Doing so could provide a better way to adjust population numbers and remove any anomalies identified. There are many methods and techniques available to do so.
The Pakistan Institute of Development Economics did a major exercise in 2001, analysing the 1998 census data and found an under-enumeration by 2pc. A similar exercise can be carried out by a team of experts, with established expertise, credibility and neutrality to resolve the issue more accurately and efficiently.
This has also been recommended by the UN’s ‘Manual for Census Evaluation: Post Enumeration Surveys’, and in Henry Shryock’s The Methods and Materials of Demography, which is considered to be the bible of demographic methods, including censuses.
Published in Dawn, The Business and Finance Weekly, May 21st, 2018