Census results raise questions: edit in Dawn, August 27, 2017
AFTER a prolonged delay, the census exercise is finally over and the first provisional results have popped a number of surprises. The population growth rate has actually increased by 30pc since the 1998 census, and the implications are truly alarming. The country’s total population now stands at more than 200m; with a growth rate of 2.4pc it will double again in less than 30 years if the growth rate continues unabated. This flies in the face of repeated assertions that the growth rate has been reduced to 1.89pc as stated in successive Economic Survey documents. “Increasing population growth raises the dependency ratio and puts pressure on education, health system and food supply”, the last Survey noted after announcing that the total population was 199m.
But now, the figure of 207.77m for the total population means the population growth rate has been far higher than what the authorities estimated and used as the basis for their population welfare programmes and to plan service delivery. The figure is staggering, and is an even larger challenge — almost at the level of an emergency — for our policymakers as the country’s ability to feed, house, clothe and provide opportunities for this enormous mass of humanity simply cannot keep pace with the increase. Since the last 1998 census, it was assumed that the population growth rate had come down to 1.8pc — a figure that was heralded at the time as a huge success. That assumption now stands reversed, presenting the authorities with a hard task in making family planning policies, and aiming for a sharp decline in infant mortality rates. Above all, it means the Rs8bn that the government has been allocating for population welfare programmes in recent years will need to be increased dramatically, and the design and delivery of the programmes themselves strengthened, if the emergency is not to swamp us in the foreseeable future.
Beyond the aggregate numbers, the details given in the provisional data bring further surprises, leading some to allege ‘number fudging’ at this early stage. Whereas the reflex to attack the credibility of the data needs to be resisted at this point, it is critical that the Pakistan Bureau of Statistics move quickly to make the full, disaggregated data available to independent researchers. The surprises are there in many indicators. For example, Punjab’s population share has fallen sharply, that of Sindh has risen marginally, while KP and Balochistan have surged, rising by 1.3pc and 0.9pc respectively. These are enormous increases, and they defy research that has shown fertility and birth rates coming down across the country in the 19 years since the last census. They also beg the question of where this additional population has come from since both provinces, KP in particular, have seen large out-migration in these years. Have the fertility rates been much higher than originally thought, or has in-migration been even higher? Even the urban-rural proportions beg an explanation. The last census changed the definition of an urban area to mean those areas presided over by municipal or metropolitan corporations, which resulted in large swathes of peri-urban settlements around cities falling outside the definition of urban. That census famously underestimated the urban population of the country, leading to massive distortions in all those policies that rely on this distinction, from property taxation regimes to planning for urban service delivery, as well as allocations from the centre for rural-based industries.
It is not known what definition of an urban area the latest census has used, but if the last definition hasn’t been changed, then once more large swathes of urban populations have not been counted as such. The provisional estimate for Karachi’s population, which reportedly has come in at 16.5m, is far too low to be credible, considering it would take a growth rate lower than the national one to keep the total population figure at this level in 2017. Either Karachi has seen drastic reductions in fertility rates or in-migration has fallen since 1998 — both beggar belief. The PBS ought to make it a priority to release the disaggregated data so the more granular details of the result can be tested against the prevailing research. Until that happens, no policy decisions should be based on the census data, while attacks on its integrity should also be resisted. https://www.dawn.com/news/1354229/census-results-raise-questions

Figuring it out so far: edit in The Express Tribune, August 27th, 2017.

It is unwise to extrapolate too much from the data coming out of the provisional summary results of the 6th Population and Housing Census 2017; but even the raw data is startling. It is 19 years since the last census when the population was a little over 130 million. There has been an increase of 57 per cent at an annual rate of 2.4 per cent, and unless that changes as in drops further and faster, then Pakistan is in trouble at an accelerating rate. There are fewer women than men and the reason for that is yet to be explored. Punjab remains the most populous province and Karachi the largest city — though not as large as some were predicting. This is still predominantly a rural population, and by a significant majority, and the urban population has only grown by a little over 3 per cent in almost two decades.
Demographers and statisticians are now going to have to work on the final analysis of the figures, hopefully by October, and if by then the numbers are settled it may be that changes can be incorporated in boundary divisions prior to the next election. Despite this there is much to ponder. Bluntly put — there are too many babies. Terrorism may occupy the attention of the media but the threat presented by overpopulation (coupled with an existential water crisis let us not forget) is far greater than anything that the bombers can inflict.
There were 27 million of us at the point of Independence. Today — over 200 million, a doubling every 25 years. If we continue to proliferate at the same rate in 25 years there will be 400 million. As the demographers point out exponential growth is not going to happen, but the decline in the birth rate to 23.19 in 2014 is not going to be enough to avert that figure of 400 million in a quarter century. On available resources and livable land that is unsustainable. Within the lifetimes of children born this very day there will be famine on an unimaginable scale. There are many national emergencies — but contraception is the one nobody is talking about. https://tribune.com.pk/story/1491895/figuring-it-out-so-far/

Exponential Growth: edit in The Nation, August 27, 2017

An exercise that took the back seat in the past two decades, the completion of the 2017 census is a major step taken by the government that it must be commended for. To Pakistan Muslim League-Nawaz’s (PML-N) credit, they are also the party that last organised the national headcount in 1997, with two successive governments (one of them democratic) neglecting to do so afterwards. On its part, the Pakistan Bureau of Statistics must also be congratulated on a tremendous effort to conduct the tally and compile the results as effectively and diligently as possible, amid threats from terrorists.
A provisional result has been released, with the final report to come later. The most startling aspect of the summary of results is the fact that the country’s total population increased by a whopping 57 percent in just nineteen years – from 1998 to 2017.Out of the total population of 207.77 million, 110 million are in Punjab, 47.9 million in Sindh, 30.5 million in KP, 12.3 million in Balochistan, 5 million in FATA and 2 million in Islamabad. Data on Gilgit-Baltistan and Azad Jammu and Kashmir has not been released in the provisional results, which is why the total might increase even further with the final report.
Another interesting fact to note is that there are now over 10,000 transgender persons in the country, and 106.45 million males and 101.31 million females. While the gender distribution between the three sexes has remained roughly the same in terms of percentage, having 10,000 recognised transgender persons in the country means that the state no longer has any excuses to not look for the welfare of the marginalised third gender. Provision of education and employment opportunities, alongside protection from threats and general improvement in the lives of the community is a necessity, with a number as large as this.
With a 57 percent increase in almost two decades, Pakistan’s overpopulation problem has now reached newer and more terrifying heights. There is a dire need for the state to change common perceptions about contraceptives and family planning, and to discourage households from having any more children than necessary. If this keeps up, Pakistan might soon have to employ drastic measures such as a restriction on the number of children allowed to be born to a family. Thankfully, we are not quite there yet, but the results make one thing obvious – we do not have twenty years, or even a decade more to lose.http://nation.com.pk/editorials/27-Aug-2017/exponential-growth

Cross-Census Survey: edit in The Nation, August 31, 2017

On Tuesday, the Senate Committee on Privatisation and Statistics urged the Pakistan Bureau of Statistics (PBS) to verify one percent of the 168,000 blocks of the census to increase credibility of the national headcount. The parliamentary committee also claimed that the army had different census results as compared to the PBS. Chief Statistician (Asif Bajwa) clarified that PBS and the army had the same results, and that multiple institutions were involved specifically to increase the validity of the census. For certain sections of the parliament to claim the results are invalid is just prolonging the inevitable delineation of constituencies. Increasing the credibility of the census is no real excuse, it is clear that the parties have an issue with the results themselves – the overwhelming increase in the population of Punjab and Lahore has led to fear among opposition parties of even greater support for parties such as PML-N, however pointing holes in the census will not do much good.
All those sitting in parliament currently arguing against the results were aware the exercise was a mere head count and would provide basic demographics; but supposedly due to lack of time, it was going to skip internal migration, taking into account expatriates, disability ratio, and mortality rates, etc. The reason was that mainstream parties would benefit from no drastic changes in the statistics, because that would have meant a dire challenge to the voting strength of these parties.
Now that the result to a long-delayed exercise is finally out and it it is not in their favour, opposition parties cannot stop pointing out irregularities without any substantial proof; if they had issues with the process itself, where were they when the framework was being laid out? They have finally realised that a change in the demographics of rural Punjab could be disadvantageous, but a cross-census survey is only a waste of time and resources.
Political parties need to understand that the time for raising issues over the headcount has passed, everything should not be turned into a political gimmick. Chief Minister Shehbaz Sharif has also urged leaders of the federating units to not make the census controversial. The margin of error in this situation is merely two percent, which is why focusing on only 1 percent of the blocks is likely to give cause for opposition parties to raise even more of a ruckus. With the next election coming up, the Election Commission of Pakistan (ECP) also needs time to work on delineation based on the new results. If that process is not completed, the new elections will be termed illegal according to the constitution of Pakistan. The opposition will have to live with the results and look towards actually doing something about the population problem.http://nation.com.pk/editorials/31-Aug-2017/cross-census-survey

Integrity of census data: Editorial in Dawn, August 31st, 2017

EVEN before the larger data set has been released, politics has taken over and the provisional results of the latest census exercise are being called into question. Considering this is the first census since 1998, a long-awaited one that comes after at least one abortive attempt by the previous government to conduct the exercise, it is too early for public figures to start contesting the numbers. Census data is voluminous in scope and its takes time to examine it. In this case, matters are especially complicated as definitions have not always been applied uniformly across the country; it was up to the respective provincial governments to decide what definition they would apply to some census categories. As such, much confusion is circulating around the figures for the urban-rural divide, which is crucial in the formulation of many policies because provincial governments have demarcated their urban areas differently. Such definitional issues can be resolved quite easily without impacting the integrity of the data. But it is the job of experts, such as demographers, to do this, and not the politicians.
The PPP was hasty in announcing that they ‘reject’ the results of the census, even before the results were fully in. Although it is understandable that many people feel that the data is very different from what they expected, especially given the varying estimates that came out during the intercensal period, it is also true that census numbers should not be a matter of political consensus or arbitrary views. If people believe that Karachi’s population ought to have been more than what the figures reveal, the data must be challenged on the basis of hard facts and not simply because expectations were different. Data integrity issues can be addressed through closer examination, followed up with a rerun of the exercise in a small number of census blocks to tally the new findings with the existing ones. If a uniform variation is observed across a large number of blocks, then it can be surmised that there might be issues with the integrity of the data.
Many political players have a vested interest in pulling the data in one direction or the other, but it should be statisticians and demographers who should have the first say when it comes to examining the findings. One of the biggest implications the census data has is for parliamentary seat shares and the delimitation of constituencies, in addition to resource-allocation decisions. For the time being, many of the political apprehensions driving the politics around the data can be settled if an agreement is reached to freeze seat shares and constituencies for the next general election. This would be a pragmatic decision. www.dawn.com/news/1355013/integrity-of-census-data