'Private health, education to boost formal GDP' an interview with Dr Sajjad Akhtar, Acting Chief Statistician, Pakistan Bureau of Statistics

Dr Sajjad Akhtar is the Member, Economic Statistics, at the Pakistan Bureau of Statistics (PBS). He is currently also holding the charge of Acting Chief Statistician of the federal statistics body. Previously, he was associated with a host of organisations in a consulting capacity; these include the UNDP, ILO, DFID, World Bank and Population Council. He was also the Director, Policy & Research at Pakistan Institute of Trade & Development (PITAD), and Economic Advisor for Saudi Arabian Monetary Agency.
Earlier this month, BR Research caught up with him during his day trip to Karachi. This interview took place over a drive from Clifton to Jinnah International Airport and concluded over a quick cup of tea at an airport joint. The discussion surrounded the rebasing of CPI, GDP, and the population census. Below are edited transcripts.
BR Research: What's the progress to date as far as the re-basing of the CPI is concerned? And when will the series officially commence?
Sajjad Akhtar: We have constructed the rural CPI, the urban CPI as well as the national CPI. We are measuring it on a monthly basis. But we haven't released it because it is awaiting the approval of the governing council of PBS. Based on the recommendations of the governing council, PBS may have to seek the approval of the government for its official commencement and dissemination on a monthly basis. As per practice, the PBS will be releasing old CPI along with the new CPI for about a year.
A precise date for the official commencement of the series cannot be given at this point, because of the anticipated change in government.
BRR: The inclusion of items in the Family Budget Survey (FBS) based on which the CPI basket computed is an arbitrary affair, which is why the likes of Septran medicine and Sega games are found in the current basket. What is being done to make the FBS methodologically robust?
SA: The revision of the list for the current rebasing of the CPI started in 2014-15. A series of meetings were held with the technical committee. The technical committee is composed of academicians, demographers, and previous employees of PBS who have prior experience of the CPI-rebasing and the management of CPI time-series on monthly basis. The technical committee finalised the list for the survey (FBS). After that, PBS ran a pilot based on which certain items were deleted from the proposed basket and certain items were added.
BRR: But economists and academicians don't know the prevalence of medicines, for instance. So, for the health basket, the PBS should seek advice from a technical committee of doctors, say PMDC, chemists' association, etc; likewise, other experts for other specific baskets.
SA: You are right. There are valid points in what you are saying, and there are a lot of ways to improve the composition of the technical committee. But you have to consider the interview fatigue. Already the survey questionnaire is stretched to 20 pages; anything beyond that means that respondents run away or don't give proper answers. We can't include 50 medicines in the Family Budget Survey.
BRR: What's the update on GDP rebasing? Reportedly, the progress on the Census of Manufacturing Industries (CMI) is slow, particularly in Sindh and Baluchistan. Will that impair the GDP rebasing deadline of December 2018?
SA: We have not given up and we are still coordinating with Sindh; and they constantly agreed to make this CMI a success. They are working on proposals to increase the retrieval rate of the survey in Sindh. The PBS has various options to appoint any public body to get the CMI filled, and it also has the right to impose a penalty. So yes, by the last quarter of 2018, we hope to finish the field operations for CMI. Another 12-18 months will be needed for generating a rebased GDP series.
BRR: The CMI feeds into the re-basing of the GDP. Is there a way to audit the CMI survey ala the post-enumeration survey of the population census or a third-party evaluation?
SA: We have our internal checks and processes in place to verify and monitor the data quality and reliability but there is no third-party audit.
BRR: What new sectors you think would boost the GDP in this round of rebasing?
SA: The services sector would expand further. There is IT including hardware, software, and e-commerce; then there are private schools and universities; and also private hospitals and clinics. Their representation will increase as their numbers were few in the formally-estimated GDP in the last round.
BRR: What is the future of the population census, now that the post-enumeration survey (PES) has been delayed beyond the usual timeframe?
SA: The PES will be undertaken soon as per government's instructions. The world-wide practice is to conduct it within 60-90 days after the big count. Holding it 12-15 months after the main census puts into question its statistical validity due to shifts in population dynamics. Plus, although we don't expect average variation at the national level beyond 2 percent, any higher result will be due to the extraordinary lag in conducting the PES. Questioning and politicizing the census results based on PES results will be unfortunate and waste of valuable resources.
BRR: When would the PBS release final census results?
SA: The PBS would be ready by April-end 2018. The rest depends on the government, because these (results) will have to get a nod from the CCI before the data are made public.
BRR: Karachi's growth rate and total population released in the provisional census results invited a lot of criticism. How would you defend it?
SA: People are entitled to their opinions but not their facts. The critique on Karachi's results is influenced by politics and based on casual empiricism and squabbling with numbers rather than actual facts.
The critics need to look at the fact that within the Karachi Division, the average household size per district does not show any wide variation and is consistent and stable. This shows that there is no systematic under- or over-estimation of population across the districts of Karachi.
Likewise, those who raise suspicions over the drop in Karachi's household size should look at the data and see that average household size has declined in all the urban districts except in Peshawar. This is not some statistical anomaly, but a reflection of population dynamics in terms of migration, lower growth rate, smaller family size and tendency toward nuclear vs. extended family households in urban areas.
They also forget that the boundaries of Lahore city were updated and in 2015. Punjab government declared the entire Lahore district plus two Union Councils of Kasur district falling within the jurisdiction of Lahore city. The boundary of Karachi was not extended except urbanization of seven dehs.
With the expansion of Lahore's boundaries, nearly 20 percent of population in 1998 that previously constituted rural areas around the 1998 boundaries has been declared as part of Lahore City. In the latest census, the newly included areas in 2017 have 30 percent of Lahore's population. If the 2017 census were done on the basis of 1998 demarcation, then the annual growth rate of Lahore city would have been 2.4 percent, lower than Karachi's 2.5 percent.
The independent observers should focus more on the fact that while the inter-censual growth of Pakistan's population came down from 2.7 percent (1981-98) to 2.4 percent (1998-17), it is still the highest in the South Asian region. In terms of GDP per capita, it means that our per capita income is lower than Bangladesh.
BRR: Whose job is it to classify a region as urban or rural? And what definitions are being used?
SA: The job of the classification of an area as urban or rural rests with the revenue department of each respective province. The definitions were abandoned in 1988. It all depends on the land revenue department; if the land revenue department classifies a place as urban, then it becomes urban.
BRR: Is there a formal coordination mechanism between the PBS and the BoS, aside from as and when the PBS wants to call in a meeting for something? Or what is the quality of that mechanism?
SA: It is safe to say that the coordination mechanisms are weak between federal and provincial bureaus of statistics. They are ad-hoc and need-based.
BRR: How does it feel to become a producer of data from being a consumer of data under your previous hats?
SA: I am fully convinced that data are the lifeblood of decision-making and the raw material for accountability. Without high quality data, providing the right information on the right things at the right time, designing, monitoring and evaluating effective policies becomes almost impossible.
In light of this, as a producer of data I realise there is a considerable scope for improvement in how we produce the data. For example, with the effective implementation of IT alone, we can improve the quality of data manifolds. And the cost of getting that IT is not prohibitive.