Census and election

I WAS surprised a few months ago when some otherwise very knowledgeable persons, well versed in the law and the electoral process, criticised the government of the day for delaying the completion of Census 2023 because they felt that it would deprive new voters of an opportunity to cast their vote in the coming election. Most of these critics saw some kind of a conspiracy in the delay because they felt that the new voters would likely vote for a particular political party and that the delay was deliberate to hurt the electoral prospects of a political party. Despite the clarifications offered at that time, the conspiracy narrative kept reverberating in social media.

The fact is that the electoral rolls (voters` list) and the population census are two entirely different processes, each carried out quite independently of the other by separate institutions of the state and governed by distinct legal provisions.

The census is included in the Federal Legislative List, Part II, which is regulated by the Council of Common Interests (CCI) under Article 154(1) of the Constitution. The census is also mentioned in Article 51(5) in the context of the distribution of National Assembly seats among the provinces but neverin relation to the voters`list.

The census is undertaken by the federal government through the Population Census Wing of the Pakistan Bureau of Statistics, which is attached to the Ministry of Planning, Development and Special Initiatives. The census is ordinarily conducted every 10 years. There have been seven censuses in the past 76 years, with the first one held in 1951 and the latest (seventh) in 2023. Census 2023 was organised as an extraordinary measure just six years after the preceding census, which was the shortest intercensus period in Pakistan`s history. There may have been political compulsions to do so, but technically there was no justification for prematurely carrying out such an expensive exercise.

The preparation and periodic revision of the electoral rolls is the responsibility of the Election Commission of Pakistan (ECP) under Article219(a) of the Constitution and Chapter 4 of the Elections Act, 2017. The revision of the rolls is an almost continuous process except for a brief period of about three months when the electoral rolls are frozen during the poll period. Voters are normally registered at the address given in their CNIC unlike the census which counts people at their normal place of residence.

Irrespective of the status of the census, all voters who are registered up to the time the electoral rolls are frozen, are eligible to vote in the election.

To be more specific, voters who have attained the age of 18 years and have their CNICs by Oct 25, 2023 the cut-off date announced by the ECP will be registered as eligible to vote in the general election scheduled for the last week of January 2024.

The story published in this paper on Oct 13 based on the findings of some CSOs about the perceived 13 million `disenfranchised` voters also seems to be the result of the mixing up of statistics of the census and the electoralrolls.In addition, the report which is the source of the story is based on the assumption that the adult population is 58pc of Pakistan`s total population. There is absolutely no basis for this highly exaggerated number. In Census 2017, the population of 18 years and above was merely 53pc of the total population. Although age-wise population figures have not been released by the 2023 census, we may use the same percentage (53pc) as in Census 2017 that drastically cuts the estimated number of unregistered voters to 2.9m as opposed to the13m that was erroneously reported. Most, if not all, of these 2.9m `missing voters` are in fact those persons who, despite the efforts of Nadra, the ECP, civil society and international donors, have opted not to get their CNICs made which is a legal prerequisite to be registered as a voter.

According to Census 2017 (sadly, the figures for Census 2023 are not available as yet) 18pc of the population had not obtained their CNIC. This percentage was around 10pc in men but, alarmingly, 26pc among women. It is, however, encouragingthat the percentage of unregistered women as a percentage of the total registered voters, had been steadily decreasing over the years from a peak of 13pc in 2013 to 8pc in July 2023.

In the 2018 general election, the difference between the adult population (110.897m) as reported in Census 2017 and the registered voters at that time (105.955m) was 4.941m that translated to 4.66pc of total registered voters. The estimated number of unregistered voters in 2023 is 2.29pc of the total registered voters way below the percentage in 2018.

An overall high number, such as 13m, of unregistered voters would have been truly a source of great concern but thankfully it is not the case.

The discrepancy between population and registered voters in districts is, however, possible as the persons are not necessarily registered as voters at the same place as they normally reside and get counted in a population census. Karachi and Sindh-based political parties had been complaining for long that there is a wide divergence between the number of voters and the population reportedin the census.

The statistics relating to the population and electoral rolls need to be taken very seriously and a scrutiny of these by the academia, civil society and political parties will improve their accuracy.

The timely availability of the data is, however, a prerequisite for analysis. Although the ECP regularly provides national and provincial gender-disaggregated statistics of the electoral rolls, its agewise breakdown should also be made public regularly.

Census 2023 was approved by the CCI on Aug 5 this year, but the detailed gender-disaggregated and age-wise data has not been made public despite the lapse of about two and a half months. • The wnter is president of the Pakistan Institute of Legislative Development And Transparency.

president@pildat.org X (formerly Twitter): @ABMPildat YouTube: @abmpildat