Counting error: Why Pakistan’s first-ever digital census may be an exercise in futility

Strong data leads to efficient planning; a lack of it can lead to marginalisation, both economically and politically.
Published May 1, 2023

The primary goal of a census is to accurately count a population during a specific time frame — an exercise crucial for appropriate resource allocation, delimitation of electoral constituencies, electoral representation, assessment of living standards, public service provision, and policy planning.

Embedded within censuses are larger questions of “Who are we?” and “Where do we belong?”. Modern governance requires not just the quantification of human capital potential but also its quality. The answers to these questions are, therefore, essential for evidence-based decision-making.

Most countries produce population and housing data through a traditional census, which involves canvassing the entire country, reaching every household, and collecting information on all individuals within a given period. This exercise is usually carried by the government or a designated statistical agency to obtain accurate and reliable information about the size, distribution, and characteristics of the population.

South Asia’s first digital census

This year, the Pakistan Bureau of Statistics (PBS) is conducting the first-ever digital census in Pakistan — and South Asian — history. The self-enumeration process started on Feb 20, and field operations commenced on March 1. The deadline for the census has been extended multiple times and we are yet to see final results, even as various political parties cry foul. The timing and length of the enumeration period are crucial for a responsive and transparent census. However, conducting the enumeration process during Ramazan has posed challenges for both the enumerators and respondents.

The population census provides vital data on families and individuals residing in different provinces and cities. The National Finance Commission (NFC) too relies on this data to determine the distribution of financial resources based on population size. This helps ensure that provinces and territories with larger populations receive a fair share of resources, promoting equity in allocation.

Moreover, the census plays a critical role in determining the number of constituencies in national and provincial assemblies. Every 10 years, constituencies are reviewed and restructured based on changes in population size. This ensures that regions with larger populations have appropriate representation in the assemblies, promoting fairness in the democratic process.

Legal foundations of census-taking in Pakistan

The legal basis for conducting the census in Pakistan is provided by the Census Ordinance of 1959, the Census (Amendment) Act of 1991, and the General Statistics (Reorganisation) Act of 2011.

The Census Ordinance was the primary legal framework for carrying out censuses in Pakistan until it was modified in 1991. The ordinance outlines the extent and goals of the census, the duties of those in charge of carrying out the census, and the consequences for failing to comply or providing false information. The Census (Amendment) Act updated the legal framework for census activities in the country. It added questions to the census questionnaire covering topics such as health, education, and employment, while also introducing measures for ensuring the accuracy of the data and maintaining its confidentiality through quality control measures.

The 2011 Act established the PBS as the primary statistical agency responsible for collecting, analysing, and disseminating official statistics in the country. The act provides a legal framework for conducting various statistical activities, including censuses, surveys, and other data collection efforts. It also authorises the federal government to decide the census dates, formulate questionnaires, collect information on personal characteristics and housing, and disseminate census data. Additionally, the act ensures confidentiality and protects individuals’ personal information.

The country has traditionally conducted a census every decade, with the last one conducted in 2017. However, a 2011 amendment to the Statistics Act allowed the government to conduct the census whenever necessary.

The census enumeration method

There are two approaches to census enumeration — de jure and de facto. The de jure approach involves counting people at their usual place of residence, while the de facto approach involves counting people where they are found on the day of the census. The 1998 census applied both approaches simultaneously, but the data was tabulated and published on a de jure basis for comparison over time.

For the 2023 census, the de jure method is being employed. The PBS clarified that individuals who have lived in Karachi for six months or plan to stay for the next six months will be counted as residents of the city. Members of a household who are temporarily away for under six months will be enumerated at their usual place of residence whereas previous household members living elsewhere will be counted under their current place of residence.

It is, however, unclear how the questionnaires used for enumeration will address internal migration to determine an individual’s length of stay in a particular location. To accurately determine the population count, it is important to consider how the enumerator determines the respondent’s actual length of stay or future to stay in the city.

For a city like Karachi, where a significant population of migrants and seasonal migrants move to the city for work and education, and many do not have official documentation of their residency, the de facto method seems to be more appropriate. It would provide a more accurate representation of the population that is physically present in the city at the time of the census as well as a significant part of the year.

Learning from the past: Insights from the 2017 census

The 2017 census was deemed controversial due to many errors. One of the biggest controversies is that the population of Karachi, the country’s largest city and industrial hub, is still hotly contested.

Data from the 2017 census revealed that the rural and urban population of Karachi was estimated at nearly 16 million. Experts and political parties contested the figure, claiming that it severely under-reported the population, with many claiming that the real figure was over 20m.

Prior to that, the 1998 census had reported Karachi’s population to be 9.3m. Comparing the 2017 census results with those of 1998, it appears that in the last 19 years, the population of Lahore city increased by 113 per cent, Islamabad’s by 90pc, and Peshawar’s by 100pc. However, Karachi city’s population increased by only 63pc, which is hard to believe, especially considering the scale of migration that the metropolis receives from other provinces and rural areas of Sindh.

During the 2017 census, experts and civil society representatives expressed concerns and pointed out shortcomings, but no progress was made due to the upcoming 2018 elections. Amazingly, some blocks in Karachi had fewer population counts in the census than the number of registered voters for the election. This discrepancy is impossible to explain beyond a fault in the enumeration process — even taking the time in between the census and election into account.

Another fault that many have pointed out recorded many residents of Karachi as living in their hometowns based on their permanent addresses, even though they reside in Karachi and use its resources. This is one of the reasons Karachi’s population was under-reported and led to an over-representation of other areas in the allocation of assembly seats and resources. However, the PBS hopes that counting where people live, and setting the six-month time frame will help overcome this issue.

The 2023 census so far

In the ongoing census, the accuracy of the count for Karachi is once again under question. Karachi’s population continues to grow, and it is estimated to exceed 30m if counted accurately.

People from all over Pakistan relocate to the city for employment, either permanently or recurrently for six to eight months (for seasonal employment). The city has experienced a significant influx of people from tribal areas and Khyber Pakhtunkhwa.

It is also home to a sizeable number of Afghan refugees and Bengali, Burmese Rohingya and Saraiki communities, many of whom are undocumented and thus ignored and unaccounted for. Over a decade since the 2010 floods, families affected by the disaster living in Karachi are still considered residents of their hometowns since their addresses remain unchanged.

The 63pc increase in the city’s population over 19 years, recorded in the 2017 census, thus failed to account for the actual growth that has taken place. Many residents, including the Bengali and Pakhtun communities, have either been ignored or inaccurately recorded in their hometowns, leading to an under-representation of Karachi’s population.

In the 2023 census, Karachi has been divided into 15,984 blocks, 1,290 circles and 365 charges. According to the PBS Karachi Coordinator, till March 2, the enumerated population of Karachi was reported to be 8.6m with 2.924m households counted during the first phase of the census. This indicates a 7pc increase in the city’s household count. On April 3, the PBS reported 2.94m households in their database, with 2.42 million households enumerated and a population of 12.33m. On April 10, the PBS suggested a re-check of enumeration in about 30,000 buildings of more than four storeys in Karachi and sent a list of those buildings to the Sindh government, whose officers have been conducting the census in the city.

This reflects poorly on the quality and standard of the census. As a GIS analyst, I believe that if the PBS had correctly geo-tagged all households in Karachi, they could check gaps in the data with one click. Similarly, on April 4, the day the census was officially supposed to end, the PBS announced that the population count in the leftover blocks would continue in Karachi. However, it did not mention the actual number of leftover blocks.

Read more: At 90pc, census data shows ‘dip’ in Karachi’s population

By April 21, the population count in Karachi had reached up to 16.5m, while experts are estimating that the maximum figure of Karachi’s population will reach around 17.5m based on PBS household data. The current figure shows a population growth rate of only 2.5pc since 2017. While comparing this with other divisions of Sindh, this value is exceptionally low; no division in Sindh has below 12pc increase in population, with Larkana recording the largest increase (25pc). Beyond the migration aspect, this figure is hard to believe even just taking the natural growth rate of the population into account.

As a result of under-counting, planning for the city’s basic infrastructure, such as schools, hospitals, housing, roads, and utilities, have been hindered. It has created feelings of exploitation and discontent among its residents. In order to ensure effective planning and to estimate the city’s actual population, census authorities need to improve transparency in the process.

Incomplete questions lead to incomplete answers

The 2023 census lacks options in questions on important factors such as water usage and alternative energy sources. Respondents from Karachi are facing confusion while selecting the light source for their homes, since many utilise multiple options such as Karachi Electric, solar panel, UPS, and batteries. However, the census form only allows them to select one option from electricity, solar panel, kerosene oil, gas lamp, generator, and others. This is problematic as nearly 50pc of Karachi’s population is facing 12-hour load-shedding every day, which means they rely on multiple means of electricity.

Similarly, there seems to be no question about domestic water usage, with the survey only asking about drinking water inside and outside the house and its source. The options for inside the house include tap water, motor pump/hand pump, protected well, unprotected well, and others. The outside-the-house section includes bottled water, spring, canal/river/pond, filtration plant, tanker, and others. It is essential to note that Karachi is experiencing severe water scarcity, and people are increasingly relying on groundwater for domestic use. The city also frequently experiences heatwaves, and this increasing trend in temperature is linked to water availability and use.

The questionnaire also has a problematic approach in the housing material section. For instance, in the question about the construction materials of the roof, they have merged cement and iron sheets into one option, which does not make much sense as cement sheets and iron sheets have different impacts. This approach will not provide useful data to frame policies about climate change vulnerability and resilience. Moreover, the question about the construction materials of walls has also clubbed together materials such as baked bricks, blocks, and stone, which have different impacts in case of heat waves. The options provided are limited to Katcha/Pakka (undercooked/cooked), which is concerning. These questions lack specificity and consideration, limiting their potential use in research and policy-making.

Without strong data, there can be no successful planning

As outlined above, the 2023 census questionnaire has several gaps, making it difficult to evaluate the social, economic, and demographic situation of cities with the acquired data. Meanwhile, errors in census data are common in developing countries due to incomplete enumeration, inaccurate information provided by respondents, or computer errors in compiling data.

What is important to remember is that census data plays a crucial role in determining the future of cities. Karachi’s development pace is lacking, and without accurate data, identifying necessary development projects becomes even more difficult.

While the city’s population is underestimated, problems faced by its citizens are increasing as millions are forced to contend with major civic issues such as a lack of clean water, housing, waste management, storm drain cleaning, public transport, poor roads, inadequate health and education facilities, electricity shortages, and a lack of open spaces. Moreover, the city is prone to urban flooding, and identifying low-lying areas and drainage blockages is challenging.

Without accurate knowledge of a city’s population size, it is impossible to estimate the resources required for the missing population, or to gauge the extent of the housing shortage for the city’s poor. Additionally, strong census data answers questions such as how many people are homeless or have been evicted from their homes, the number of people involved in the informal economy, the needs of unskilled and semi-skilled workers, the status of children’s health and education, the problems women face in using public places and transportation, the number of senior citizens in the city and the facilities required for them. Real data is needed for sound planning and allocation of resources; without it, there can be no successful planning.

An accurate and comprehensive population count is not only important for the demarcation of boundaries, but is also a complex and politically sensitive issue. Strong data leads to efficient planning; a lack of it can lead to marginalisation, both economically and politically. The authorities must recognise the importance a census holds and give it the diligence it requires.

Header image: This is an AI-generated image using Shutterstock.