In September last year, the Pakistan Tehreek-e-Insaf (PTI) government announced that the upcoming general election – scheduled for July 2023 – would be held on the basis of a new population census. On January 13, the Council of Common Interests (CCI) green-lighted the 7th population and housing census – to be held using smart digital devices. According to the Constitution, a census is supposed to be conducted after every 10 years. However, the last census that took place after a gap of 19 years in 2017 was rejected especially by the political parties of Sindh province that accused the Pakistan Bureau of Statistics (PBS) – the federal agency responsible for conducting headcounts – of underreporting the population of the province, particularly its urban areas including Karachi. As the government gears up to hold a new census, Bol News talks to Dr Mehtab S Karim who served as a member of the PBS governing council for four years and also chaired a committee to make recommendations for the 6th census – that was held in 2017.
A PhD from Cornell University of the United States, Dr Mehtab S Karim has been teaching demography – the statistical study of human populations – for over 40 years. As an eminent academic, Dr Karim has also been part of various research initiatives undertaken by Pakistani and foreign universities as well as the United Nations and the World Bank. He is the co-author of three books on demographic issues in Pakistan and Muslim countries and has published over 60 research articles. Earlier he served as an adviser to the Pakistan Census Organization on behalf of the UN Development Programme (UNDP) and later was a member of the steering committee constituted by the United Nations Population Fund (UNFPA) and the government of Pakistan to evaluate the 1998 census results.
| Q |
The government is due to conduct a digital census in days to come as it has been approved by the cabinet. How reliable it is and how will it take place?
Mehtab S Karim: In the earlier censuses, the data was collected on a form by enumerators while in the last census the form was printed on machine-readable paper. Later these forms were scanned and data was transferred to the mainframe computer. It was a tedious method, therefore, it took about six months to release the result of the census. The census advisory committee had recommended conducting a digital census in 2017 as well – as digital censuses have been conducted in developing countries like Bangladesh and Egypt –but the PBS had then chosen not to.
I am glad to know that the next census to be conducted this year will be a digital census, where the enumerators will collect data on tablets, which will be automatically transferred to a mainframe computer. This is a very good step and the PBS is all set to do so. However, before the census all enumerators will have to be trained to use tablets and computers. If they commit any mistake, it will be recorded accordingly.
| Q |
Headcount has remained a contentious issue with many people particularly of smaller provinces expressing reservations about its accuracy. How can the government dispel these reservations this time?
MSK: A census is a gigantic task where enumerators have to go to each and every house all over the country. In a census counting everyone is next to impossible. Actually “coverage error” often happens, which means a certain percentage of people are left out. This happens in the censuses conducted in developed as well as developing countries. To find out under-enumeration in the census a post-enumeration survey (PES) is conducted immediately following the census, where about 5% of households are revisited in sampled areas to find out whether any households or individuals living there were left out. This exercise not only helps determine the percentage of people who were not counted, but also helps improve the next census. For example, in the 2011 census of India, 2.5% of people were undercounted with wide regional variations, and in the 2011 census of Bangladesh, the undercount was 4.5%. The census advisory committee strongly recommended carrying out the same exercise after the 2017 census, but the PBS did not accept this recommendation. If a PES would have been conducted following the 2017 census, most of the critiques of the census would have been satisfied, if the under-counting in each of the provinces were not higher than 2%. The PBS has not decided to also hold a PES after the upcoming census which is a very wise decision. But for transparency, it should be conducted by a third party.
| Q |
The country has a comprehensive database of its citizens at the National Database Registration Authority (NADRA). Why do we have to conduct a census then?
MSK: There are two important issues in this regard. NADRA is not able to register all people below 18 years of age, especially the newborn babies are not immediately registered with NADRA and many parents do so a few years later. Similarly, most families don’t report to NADRA about the deaths of their family members. Besides, NADRA does not keep track of people who migrate from one place to another. Thus, the updated population of a province, a district, a village or a city – which is a piece of very important information – is not available at NADRA’s records.
The importance of census is highlighted in the 1973 Constitution. It is mandatory for the federal government to conduct a census after every ten years in order to determine the allocation of seats in the national and provincial legislatures. This is why the Supreme Court in 2016 took a suo motu notice of delay in the census and said the 2018 elections could only be held if a fresh census is conducted in the country. On the basis of the latest census results, 85% weightage was given for the distribution of resources to each province in the National Finance Commission (NFC) award.
| Q |
Why has the census become a political issue in the country? Why did some people start doubting the credibility of this massive exercise?
MSK: The 2017 census was not transparent enough, particularly due to absence of a validation process in the shape of the PES. Thus, nobody is sure where and how many people were left out. Besides, as was proposed through an all-party resolution adopted by the Senate, the PBS failed to do 5% auditing of the census results raising suspicion about the accuracy of results.
| Q |
Is the population boost the biggest problem of Sindh?
MSK: Certainly, Pakistan is the only country in South Asia, where the population growth rate is 2.2% per year, which is twice higher than all other countries in the region. This rising population eats up the major chunk of gains from an economic growth rate, which has varied between 3% to 4%. Poverty in the country is increasing due to about a 5 million increase in the population every year. Owing to the rapid surge in Pakistan’s population, the UN projections suggest that the country’s population will increase from 229 million in 2022 to 309 million in 2050. Sindh’s situation – particularly the situation of its urban areas like Karachi – is indeed alarming, where hundreds of thousands of people are migrating not only from the rural areas but also from other provinces every year.
| Q |
If a proper headcount is conducted in Sindh, will it benefit the province?
MSK: An accurate headcount should not only be done in Sindh but in all the four provinces. Since the 2017 census was much controversial, Sindh’s share in the federal resources did not increase. Strangely, Sindh’s share was 23% in 1998 and it remained 23% in 2017 as well. This is highly unlikely as the province has witnessed large-scale migration of people from other provinces and federating units. Whether Sindh’s share in the National Assembly will increase after the 2022 census can only be determined when the results of the census are officially released.
| Q |
Why are people so sensitive about their ethnicity and language?
MSK: In most countries of the world, a census also collects data on the social characteristics of the population. In America, censuses also include collection of data regarding peoples’ ethnicities and even their national origins. Questions on the language and religion of the population are being asked in every census conducted in the Indian Subcontinent since 1881.
| Q |
Political parties and people of Sindh had complained that at the last census the authorities did not take into account where a person was actually living. Why is the government reluctant to conduct the census on a de-facto basis?
MSK: These are two ways to ask questions about the people living in a household. The de-facto method is used when the purpose is to count whoever is living in the household as a member of the household irrespective of how long they are living there. The de-jure method only considers those as household members who are permanent residents. In the next census, the PBS will consider only those who are living in the household for more than six months as regular household members. The de-jure method becomes complicated where a lot of migration takes place. In such a situation, it is not an ideal method. The last census was conducted on a de-jure basis and not only political people but many demographers were not in its favour, since it could result in undercount in areas where people have migrated to and over-count in areas where they have migrated from. Perhaps for the same reason not only Sindh’s population was undercounted but also of urban areas such as Karachi and Faisalabad.
For example, the World Bank had estimated that about 45% of Pakistan’s population was living in urban areas but the 2017 census reported that only 36% was living in cities. Similarly, a well-known international organisation that keeps track of the population of the 30 largest cities in the world, had estimated that Karachi had a population of 21 million in 2016 while the 2017 census counted only 14.9 million people. Perhaps many people who had migrated to Karachi and were living there were considered its temporary residents and therefore were not counted due to the use of the de-jure method. There is another drawback of this method, as some people may not have been counted at either of the two places, resulting in overall undercounting of the country’s population. Although at the recommendation of the PBS, the federal cabinet has decided to use the de-jure method, in my opinion, an additional question could be added, asking if there are any other persons living in the household. The previous census committee had recommended using the de-facto method which the PBS did not accept.
| Q |
The urban population is rising day by day and people are migrating to cities, especially Karachi to get better services. Is the rural population decreasing in Sindh and other provinces of the country?
MSK: Yes! For want of employment opportunities, people are constantly migrating from rural areas to cities. Consequently, not only in Sindh but in all other provinces population is increasing much faster in urban areas as compared to rural areas.
| Q |
Has a free and fair census ever been held in the country so far?
MSK: After Independence, the first census was conducted in 1951. Since then censuses have been conducted in 1961, 1972 (it was delayed for a year due to civil war in East Pakistan), 1981, 1998, and 2017. Prior to each population census, a housing census was conducted. Pakistani demographers believe that the 1981 census was the most accurately conducted census, perhaps for the reason that the census staff by then had gained substantial experience of conducting proper census. However, due to unnecessary delay in conducting the 5th and 6th censuses, senior staff members of the PBS retired and proper experienced persons were not available to plan, supervise and monitor the census operations. In fact, for the last many years the PBS has not had any well-trained senior demographer with expertise in census operations, leaving all decision-making in the hands of bureaucrats.
| Q |
What’s the role of the Federal Bureau of Statistics (FBS) as the FBS used to conduct a demographic sample survey annually?
MSK: Till 2011, the organization was known as the FBS which was part of the Statistics Division of the Ministry of Finance. However, in 2011 a bill was passed by the Parliament to make the FBS an autonomous organization and consequently, six independent members (3 senior demographers and 3 senior economists) were inducted as members of the governing council of newly formed PBS. I had the privilege of being one of the members of the governing council which allowed me to look at the functioning of the PBS very closely. Unfortunately, nothing changed in the organization with the exception of the appointment of a chief statistician who had recently retired as the secretary of the statistics division.
Not only the census which was to be conducted in 2011 but the Pakistan Demographic Survey which was conducted every year since 1974 has not been conducted after 2007. This survey used to provide important demographic statistics for the country and its provinces every year. However, due to its absence, Pakistan’s population growth rate could not be accurately known and resultantly all the Planning Commission documents are now based on estimates.
| Q |
How do you foresee the future role of the PBS?
MSK: The original idea was to make the PBS an autonomous and professional organization and to appoint an experienced professional as its head, with a PhD in a relevant subject. Unfortunately, it could not happen earlier. However, fortunately, after a gap of about 10 years, a professional has been appointed to lead the PBS. It is expected that things will now move in the right direction and an accurate census will be conducted which will be acceptable to all the stakeholders.
| Q |
What are the flaws in various columns of census form?
MSK: I think there are no flaws in columns of the census about minorities. Enumerators ask the households about all family members’ religion and the information is recorded as reported.
| Q |
Any suggestions you have for the coming census?
MSK: During the past 40 years, we have conducted only two censuses. There has been a dearth of proper demographic data and the last census has become controversial for the reasons explained above. I am glad that the government has decided to conduct the next census before the next general election. As a demographer and a researcher, I am eager to see quality census data. Fortunately, some of the recommendations given by demographers for the previous census have been accepted by the PBS. However, since the census is an expensive exercise and has long-term implications, it is very important that no compromises are made in the quality of data and information collected remains uncontroversial. Therefore, it is important that all stakeholders; such as the provincial governments, political parties, and the end-users, particularly researchers, remain on board throughout the exercise.
Regarding conducting the census more professionally, my advice to the PBS will be that the housing census should be held at least a month before the population census so that when the census enumerators go for data collection, from house to house, each household is properly identified on the map. Since this process was not followed in the 2017 census, the enumerators when they went into the field found that many households, particularly in urban areas – where there are high-rises with multiple households – were not registered prior to the census.
However, a long questionnaire has been designed by the PBS similar to one developed for the 2017 census. The population census does not need to collect too much information from each household as it affects the quality of data. Therefore, nowadays in most countries, population censuses are based on six to seven questions. More detailed information is usually collected from sampled households (such as every 10th household) or even later on a yearly basis, through well-designed sample surveys.
