For the first time in more than two decades, Pakistan has produced a national economic census. It had more than 7.1 million establishments and adjusted to almost 40 million structures, which makes it the highest digitalization exercise in southern Asia. Government officials have presented it as a historical step to build a reliable statistical framework of the economy. But behind the main figures, experts say that the report also exposes serious blind spots and contradictions that ask questions about how useful the data set for policy formulation.
The last economic census in Pakistan was carried out in 2000. Since then, the country’s economy has been transformed, but political leaders have had to trust obsolete estimates and irregular surveys.
That is why Bilal Gilani, executive director of Gallup Pakistan, sees the new census as historical. “It is wonderful that this has finally happened after 23 years. The world has changed dramatically at that time, however, we still trust obsolete figures. This provides a digital baseline that Pakistan has never had before,“ said.
The census was integrated with the seventh population and housing census of 2023. When combining the two, the government says it saved RS7 billion. Enumerators equipped with digital tablets on label 38.3 million structures, of which 79 percent were residential and 13.4 percent were economical. A total of 7,142,941 establishments were documented, using 25.3 million workers.
Dr. Lubna Naz, economy professor and director of the Center for Business and Economic Research of IBA, described the scale of effort as impressive and unfinished. “It is both a milestone and a work in progress. The integration of seven million establishments in a single digital framework and a geographical label of 40 million structures is not preceded, but the data need to strengthen even more before it can serve as a solid base for politics, ”he said.
However, how the census, its methods, tools and checks were made are as important as the numbers themselves.
Methodology, discrepancies and the role of AI
The census marked Pakistan attempt to completely digitize the economic enumeration. When integrating the exercise with the population census, the enumerators were able to collect data on companies and at the same time register household information, using digital tablets for geogging.
However, reliability became a challenge. More than half of all tickets, 52 percent, were initially recorded as “others” due to descriptions of vague activities, spelling errors or Roman Urdu. To solve this, the Pakistan Statistics Office (PBS) resorted to artificial intelligence, automatic learning and natural language processing to clean and classify data.
The tests showed unequal results. In Lahore, the first essays revealed that the algorithm reached only a 24.68 percent accuracy in establishments coinciding with their correct industrial codes. After the reviews, the rate improved at 85.23 percent in Karachi and 86.50 percent in Faisalabad.
At the national level, the final model was implemented with an estimated precision of around 80 percent. In a data set of more than seven million establishments, that margin means that up to 1.4 million could be classified badly.
Dr. Naz questioned the approach: “The process depended on limited evidence in three urban centers, which restricts its representativeness. Much of Pakistan’s linguistic and regional diversity may not have been properly captured,” he explained.
Gilani took a more indulgent opinion. “What matters is that a baseline has been created. Even with some erroneous classifications, this is a leap forward compared to not having any census,” he said.
What the Census leaves
Beyond the classification challenges, experts say that the biggest problem lies in what the census excluded. The methodology covered establishments in fixed structures and home -based activities, such as embroidery, tailoring, poultry, registration and production of home food. But he did not count mobile suppliers, street stalls, small beauty salons or the rapid growth sector of independent workers and digital companies.
Dr. Naz warned that this bias the image: “Many unstructured activities are found in services, retail or small production. When leaving them out, the census changes the apparent weight of the economy towards more formal sectors. The contributions of women, in particular, are statistically visible but economically invisible, because there were no financial values,” he said.
The report recognizes this weakness. Domestic activities were recorded, but 40.49 percent fell into a “miscellaneous” category to catch. This means that almost half of the domestic economy is still unknown and effectively unknown.
Contradictions and anomalies
A deeper review of the report reveals inconsistencies between different data tables and classifications.
-
EDUCATION OF EDUCATION: Add the types of unit offers a total of 290,729 educational establishments (242,616 schools, 11,568 schools, 214 universities and 36,331 madrasas). However, under industrial classification, educational establishments number 326,868. This leaves more than 36,000 inexplicable establishments.
-
Health mismatch: Hospitals are counted as 119,789 establishments, but under industrial codes for human health and social work, the number increases to 123,973. The report does not provide a breakdown to reconcile this gap.
-
Mosque workforce anomaly: The census has 600,403 mosques that use 2.06 million people, more than the entire factory sector. But it is not clear if this includes volunteers, caregivers or teachers, which makes the figure difficult to interpret.
-
Structures versus establishments: The census documents around 6.3 million economic structures but more than 7.1 million establishments. This implies an average of 1.13 companies per structure, confirming that many places, markets and several -story buildings house multiple companies. However, the report does not analyze this density, losing the opportunity to shed light on the commercial group in the cities of Pakistan.
These contradictions point to the weaknesses in how they cleaned, tabulated and presented the data.
PAkGazette-ETIQUATED: A first digital
One of the most widely announced achievements of the census was the PAkGazette-stage. The enumerators recorded the GPS coordinates of almost 40 million structures in Pakistan, which allows for the first time a mapping of referenced PAkGazette companies.
Gilani sees this as a digital advance: “It is an important achievement. For the first time, each store and establishment is literally on the map. This can change the way Pakistan plans its urban and commercial growth.” said.
Dr. Naz urged caution: “The Geo-Estiplined is a real promise for urban planning and investment orientation, but it is as good as the attached classifications. If the activity codes are inconsistent, then link them with coordinates does not solve the problem of reliability,” he said.
Missing depth: finance and gender
Another important gap is the absence of gender financial and disaggregated data. The census records the number of establishments and their workforce, but does not measure the sources of income, investments or financing. This makes it impossible to evaluate the economic value of home -based companies or their sustainability.
Dr. Naz emphasized the implications: “By not monetizing home -based work, Pakistan loses crucial ideas about companies at home, which limits the capacity of policy formulators to design effective credit programs, training and commercial support,” he said.
The lack of gender data is also a concern. Without the breakdown of male and female workers in establishments, the census cannot accurately capture the role of women in the economy. This makes your contribution visible in numbers but invisible in value.
The risk of being still
Worldwide, economic censuses are carried out regularly. India and Bangladesh already use them to capture financing, investment and digital businesses. Pakistan’s exercise, although important, runs the risk of being unique if it is not institutionalized.
Dr. Naz warned: “If Pakistan does not repeat and improves his economic census, he will face blind policy points and lose competitiveness in emerging sectors such as electronic commerce and independent work, where women and young people are more active.”
Gilani agreed that continuity is essential, although he emphasized the importance of progress already done. For him, the census marks the beginning of a digital statistical system in which Pakistan must build now.
A basis for building
The 2023 economic census is undeniably historical: the first in more than two decades, the first digital and the largest of its kind in southern Asia. It provides a reference image of Pakistan’s economy that did not exist before.
However, their defects are equally significant. Contradictions between tables, anomalies in the numbers of the workforce, the exclusions of the informal and digital sectors, and a great dependence on the Imperfecta raised questions about how much weight the policy formulators should put in their findings.
As Dr. Naz summarized, he looks better as an advance and a work in progress. The real test will be if Pakistan can refine, repeat and expand this effort to a reliable system. For now, the country has a digital framework of its economy. If that framework is resistant enough to support solid economic decisions remains uncertain.
