Cutting a sorry figure
Collecting, analysing and projecting socio-economic data in Pakistan is beset by numerous ailments and there seems to be no remedy in sight
By Raza Khan Muhammadzai
By Raza Khan Muhammadzai
Statistics are as indispensable for socio-economic planning as diagnosis is for writing a medical prescription. Without appropriately gathered statistics, policy makers cannot be expected to make informed decisions. In Pakistan there has always been a problem about the authenticity of the figures collected by the government. Recently, there has been a great debate on the percentage of population living below the poverty line in the country. The government, the opposition and the non-government organisations are all coming out with different figures on the poverty's prevalence in Pakistan. Another proof of how bad Pakistan is in statistics is the recurring, rather chronic, objections by many international donor organisations and agencies like the World Bank, the International Monetary Fund and others. Some international bodies have conducted independent surveys in some cases to ascertain the validity of the official figures. It is needless to say that these surveys have more often than not proved the government figures wrong.
Not only that. The lack of authenticity of official figures has been highlighted by the incompatibility in the figures collected by different government agencies. This has happened on numerous occasions and on a number of subjects. This has created a sort of a severe credibility gap. Dr Nadeem Ul Haq, director Pakistan Institute of Development Economics (PIDE), tells The News on Sunday that this credibility gap should not be there. "The statistics ought to be credible not just because of their singular importance in socio-economic planning but also because they (statistics) form people's perception about various things." But he argues that the problem of erroneous statistics is not Pakistan-specific. It's the same all over the world, he says.
In Pakistan, it is the responsibility of the Statistics Division, working under Ministry of Economic Affairs and Statistics, to make plans and policies for statistical development and improvement of statistical infrastructure in the country. The division provides database to policy making bodies under the General Statistics Act, 1975, The Industrial Statistics Act, 1942, The Agricultural Census Act, 1958 and the Census Ordinance, 1959, which provide legal cover for the collection of data in different socio-economic fields all over the country.
Statistical division is assisted for collection, compilation and dissemination of statistical data in different socio-economic sectors of economy by its three attached departments -- Federal Bureau of Statistics (FBS), Population Census Organization (PCO) and Agricultural Census Organization (ACO). According to official sources, the Statistics Division employs a number of PhDs but they are not asked to do any research. It is very important to have research in Statistical Division so that authentic data is not only collected but also cross-checked, these sources say without wanting to be named. But they say the government seems least interested in research on the statistics it collects. "Even individual efforts by researchers are not encouraged or welcomed by their high-ups. They (high-ups) may have their own reasons for doing so but this practice is not good for appropriate planning," says a senior official of the Statistics Division, requesting anonymity.
The main arm of the Statistics Division for collecting various types of data is the Federal Bureau of Statistics (FBS) which collects mainly secondary data culled from sources like administrative records as well as records of the federal and provincial government departments. This is a method in which the already available statistics or data is gathered, collated, reviewed and projected. But keeping in view the incapacity of the concerned departments, it is very unlikely that they can collect and record authentic figures. In fact, when TNS asked a number of officials working in different departments if they had ever conducted data collection on their own, the response was more often than not negative.
FBS also gathers some primary data by conducting surveys/censuses. Besides collection of data in different areas like prices, industrial production, trade, agriculture, labour force and demographic changes, income, consumption and social changes, FBS is also responsible for preparation of the National Accounts of Pakistan.
Of course, it's not always possible for the bureau to carry out mass scale censuses due to human, technical and financial constraints as well as time limitations. Most of the primary data, therefore, is collected through sample surveys. In this process, there is certainly a lot of room for errors. No wonder, the government has allowed the bureau a 5-7 per cent margin of error in its data.
The process of data collection through surveys is hit by many snags. Though the FBS officials claim that they have the best available technical personnel in the filed available and use the most reliable statistical tools, at the same time they also admit that their statistics are not always correct. According to these officials, they conduct surveys by dividing urban and rural areas into blocks which are mapped and listed. Data in every reference period is gathered from the same blocks because formation of new blocks for every new survey is almost impossible. One senior official of the bureau, seeking anonymity, discloses that the system does not work well in the rural areas. Add to this the variable of staff indolence and complacency prevalent in every government department and you can have a fair idea of the data collection mechanism.
In a sample survey a household is asked questions for the purposes of collecting data. The most important element in the sample selection is that it should not only be representative of the population or the area of the study but its size should be reasonably big. It should be as big as the financial resources and time allows it to be. The FBS officials say they try their level best to make their samples both represetative and large. But they also reveal that sometimes their sample size can be as small as 16 households in rural areas and 12 households in urban areas. It's difficult to understand how such a small sample can be called representative, accruing correct data on sensitive issues like poverty and income trends.
Some FBS officials point out another problem. In any kind of surveys, they say, eliciting true information from the respondents is an uphill task. For instance, people are sometimes very apprehensive and do not give right information regarding their income and expenditure lest they are taxed, these officials say. Though, they say, they have a mechanism of asking supporting questions to verify the respondents information, still the possibility of the data they collect being incorrect cannot be ruled out.
The wrong information by respondents or failure on the part of the data gatherers is a fundamental problem in knowing true socio-economic trends prevalent in the country. It also effects the whole edifice of planning and development structures and processes. Perhaps, it's because of this reason that there is always a huge gap between the ground realities and the picture presented by official statistics.
Different people see different reasons behind people's reluctance to reveal true information about their lives. But low literacy, ignorance and a general lack of awareness about the importance of socio-economic data figure in almost every assessment of the reasons for incorrect statistics in Pakistan along with cultural factors like ego, reluctance to let outsiders know about the inner details of a household. Cultural factors in rural areas of the country as well as in the largely underdeveloped provinces of Frontier and Balochistan are big problems in data collection. These factors mar the data collection process, leave alone the authenticity of data collected under these restricted circumstances.
According to officials of the FBS, the lack of awareness is particularly a problem in rural areas. One of them is of the view that awareness about the importance of data collection is not that good in the so-called urban areas either. An assistant director at the provincial Bureau of Statistics in NWFP narrates how he has faced many ugly situations in far flung areas. On some occasions he was even beaten up and his female colleague misbehaved with. "In many places, local people see as having come from NGOs and do not cooperate with us and on many occasions, the concerned district authorities ask us to pack up because they think that our presence may take the situation beyond their control," he tells TNS.
Whatever the flaws in data collection, the government aggravates them by how it treats them. Sometimes figures gathered from the grassroots by putting in a lot of effort are altogether ignored when it comes to using them for policy formulation. But they are not made known to the people of the country, especially when they show the government doing poorly in a sector. "Extreme care is taken that economic statistics only show the government in a better light. Sometimes, statistics are even juggled, twisted and distorted," says a senior official at one of the organisations, wanting to keep his identity secret. On many occasions, population growth rate has been shown coming down, sometime significantly, (against the recorded trend) because the government had taken funds from international donors for bringing down the population growth rate in a specified period which it failed to do, he reveals.
Likewise, the figures that directly create a perception about the government's performance, like the level of poverty, literacy, unemployment, are also sometimes doctored. On poverty and literacy, the government also changes 'operational' definitions of what is meant by these terms to show positive trends about them.
A retired chief economist of NWFP, on the condition of anonymity, claims Altaf Gauhar, one of Pakistan's top bureaucrats in 1960s changed the definition of literacy to show enhanced literacy level to a visiting a World Band team.
Even today, a lot of distrust is being shown by people and international donors on figures of economic growth and poverty and literacy levels and somebody out there may feel tempted again to change the definition to overcome this distrust, wrongly.
The problem of collecting and projecting authentic data are numerous and complex and there seems to be no remedy for them in sight. "The solution is the creation of an independent body on statistics, which should evolve its credibility over time and then it should educate people how statistics are gathered, what role they have in economic planning and management and above all in their own lives," suggests Dr Nadeem Ul Haq.
