Pakistan Bureau of Statistics

What it is

The Pakistan Bureau of Statistics (PBS) is the country's official statistics agency. Its website publishes reports, tables and, for many household surveys, the underlying anonymised microdata. For a data science learner that means real survey data with questionnaires, coding schemes and official reports to check your numbers against.

What you can find

These are the datasets and releases listed on the site as of September 2026:

• Population and Housing Census 2023: national, provincial and district reports, plus the separate Digital Census 2023 portal. The headline population is 241.49 million.
• Labour Force Survey (LFS): annual reports from 2001-02 to 2024-25, 52 statistical tables for 2024-25, questionnaires and coding schemes, and microdata in SPSS going back to 2005-06, with Stata versions for recent rounds.
• Household Integrated Economic Survey (HIES): rounds from 2004-05 to 2024-25, with the 2024-25 full report, questionnaires, 25 Excel tables, and microdata in SPSS and Stata with a coding scheme and manual of instructions.
• Pakistan Social and Living Standards Measurement (PSLM): the latest 2024-25 report, with district-level indicators on a separate dashboard.
• Prices: the monthly Consumer Price Index and the weekly Sensitive Price Index.
• Other releases: national accounts, large-scale manufacturing, external trade, the Agricultural Census 2024 and the Mouza Census.

Is it free?

The data dissemination policy says aggregate statistics, publications and anonymised survey microdata are normally released free of charge under an open licence. More detailed licensed files need an application and agreement, and fees apply only to exceptional customised tabulations and some specialised data, with waivers for students and publicly funded research. Requests go through the data request form on the site.

Why it is good practice material

Survey microdata is messy in exactly the ways real jobs are: coded categories you must map using a coding scheme, sampling weights you must apply, missing values, and SPSS or Stata files you read with pandas.read_spss or pandas.read_stata. Because PBS publishes its own tables, you can check whether your weighted estimate matches theirs, which is the best feedback a self-taught analyst can get.

What you can build

A labour market dashboard by province, gender and age from the LFS; an analysis of household spending shares from HIES; or a chart of monthly inflation from CPI releases with a short written explanation.

Common questions

Do I need to register? Published reports, tables and the public microdata files are linked directly from survey pages.

Where do I start? The LFS page: the files are recent, well documented and the questions are easy to explain.

Best for: learners who want real public data, and anyone analysing Pakistan's economy or labour market.