What went wrong in the 1998 census?

It almost didn't happen again, but any doubts that still lingered were finally put to rest as the fifth nationwide census kicked off in full earnest on March 2. The fact that it managed to get off the ground does not, however, by any means signal the end of the controversy surrounding the exercise. Many already question the transparency of this latest head count, and as such the real problems may have only just begun.

A census has been due since 1991, and the last exercise took place ten years before that in 1981. In the interim, successive PPP and PML(N) governments have been unable, or perhaps unwilling, to get down to brass tacks and deal with a long pending job that is of such undeniable significance to the state and its federating units. Besides the obvious logistical hurdles, a major stumbling block has been the lack of consensus among the country's various political and ethnic groups. The issue is a highly sensitive one and there is clearly no dearth of those with axes to grind. As a result, mistrust and fears abound on both inter and intra-provincial levels.

The ongoing census, too, suffered a not-so-minor hiccup when the chief of army staff, General Jahangir Karamat, reportedly expressed di-satisfaction with the arrangements made by the civil administration. A letter he wrote to the government in this connection was "leaked" to the press, prompting some newspapers to conclude that the census had been put off indefinitely. In the letter, the COAS had purportedly objected that the director general of the database had not yet been appointed. This disclosure was followed by a clarification from the armed forces' Inter-Services Public Relations department, which pointed out that the COAS had merely communicated the ground realities to the government. Sources in the ISPR also claimed that while army personnel have been dispatched throughout Pakistan, the civilian enumerators had not reached the more remote areas of the country.

The government, for its part, maintained that the extended spell of cold weather was hampering work in the northern areas. The federal cabinet, however, subsequently decided that the census would go ahead, but no details were provided about the alleged "incomplete arrangements". After this official go-ahead, the country-wide head and house count began as per schedule on March 2 amidst reports of stiff resistance in parts of Balochistan, where data forms were torn up and thrown back in the faces of the enumerators. Tension also prevailed in some areas of the Punjab.

A census has been due since 1991, and the last exercise took place tenyears before that in 1981.

Such incidents are ample proof, if any were needed, that the grassroots-level hurdles which delayed the census for seven years are yet to be overcome. Various sections of the population still question the methodology of the exercise, which is hardly surprising given that the government has made little or no effort to build confidence among the people. As things stand, in fact, it is inevitable that the results of the census will not be acceptable to many of the country's political parties and linguistic groups.

The most skeptical are the leaders of rural Sindh, while the Baloch and Pashtuns in Balochistan also have their own reservations. The leaders of Punjab, meanwhile, are concerned about the prospect of losing some National Assembly seats if Punjab's population (relative to that of the other provinces) shows a decline. On this count at least, the Nawaz Sharif government has spared no effort to allay any and all fears. The federal cabinet, while confirming the census would go ahead, also decided that the results would not affect the allocation of National Assembly seats as well as the distribution of federal funds amongst the provinces.

This move, which came without prior intimation, is one that essentially negates the primary purpose of a census. The current allocation of National Assembly is based on the 1981 census, which was controversial to begin with. If the formula remains the same regardless of the new data, neither the will of the people nor the demography of the country will be truly represented in the next parliament. Over the years, there has been a marked population shift towards the urban centres in both Sindh and the Punjab, and as a result these areas now clearly require greater representation. But this, the government stubbornly says, is not going to happen, which effectively means that the next National Assembly elections, scheduled for 2002, will be held on the basis of a 21-year-old census.

The federal cabinet's decision not to redistribute NA seats in the light of the '98 census results is in clear violation of the Constitution. Article 51(3) states that seats in the national legislature shall be allocated to each province, the Federally Administered Tribal Areas and the federal capital on the basis of population according to the most recent officially published census. As such, if the Nawaz Sharif government wants to stand by its decision, it will have to delay the publication of the census results until after 2002. And that is bound to provoke a strong backlash from those who feel that they are being deprived of their due representation.

Ironically, Benazir Bhutto's Pakistan People's Party- which itself came in for a lot of flak for its abortive attempt at a census in 1995- is also opposed to certain aspects of the ongoing exercise. "We have reservations about the census because the procedure adopted by the government is not transparent," said the PPP chairperson. She demanded that the government should first declare the results of the 1991 head and house count which, according to her, had showed a 12 per cent increase in the population of Sindh. She did not, however, care to fully explain why her own government chose not to announce the results of the abortive 1995 census. "It was because we had a coalition government," was all she said, skimping on the details. Does that mean that the PPP's coalition partners were not happy with the outcome of the census? Once again, Bhutto chose not to elaborate.

The changed demo-graphic realities of Pakistan are difficult to stomach for the feudal leaders of the rural areas and, on an inter-provincial level, for the nationally dominant Punjab.

The 1991 census was put on hold after the population of Sindh showed a massive increase during the house count, the first stage of the census. From 19.03 million in 1981, Sindh's population had ostensibly soared beyond the 50 million mark. The reason was obvious: the various ethnic groups in the province, particularly the Sindhis and the mohajirs, had both inflated the number of people living in their respective areas of influence. These figures were not acceptable to the other provinces, especially the Punjab, and even the planners and policy makers in Islamabad were refusing to accept that the population of Sindh could have increased by 171 per cent. Consequently, the census came to a halt and has since been postponed repeatedly, for various reasons.

The most compelling factor, obviously, is the fear in certain quarters that the demographics of the four provinces may have changed dramatically since 1981. People have migrated in large numbers from rural to urban areas, and from other parts of the country to Sindh. Even if the figures are not deliberately inflated, the population of the southern province is bound to show a marked increase. These realities are difficult to stomach for the feudal leaders of the rural areas and, on an inter-provincial level, the nationally dominant Punjab.

Controversy also surrounds the methodology of the census, and different lobbies have their own peculiar objectives. But even just completing the required paperwork is a confusing and exhausting prospect for the respondent. As many as five forms will have to be filled out, informed sources say. One is related to house listing, and another will go towards the Planning Commission's database. Interestingly, the Planning Commission form includes a column that contains inquiries about personal wealth, a thinly veiled attempt perhaps to build up an economic database of sorts.

The 1991 census was put on hold after the population of Sindh showed amassive increase during the house count, the first stage of thecensus.

The third and fourth forms, issued by the Census Organisation (CO) and to be filled in by the people themselves, will serve as the basis of the "big count" and the "sample count". The former will be distributed amongst 92 per cent of the population, while the sample count will be restricted to just eight per cent. In the big count, the choices available in the "mother tongue" category are Urdu, Punjabi, Sindhi, Balochi and Pashto. The sample count, however, includes two more languages, Hindko and Seraiki. But since only eight per cent of citizens - and that too spread all over Pakistan - will receive this form, chances are that these two languages may not receive their due representation in the final tally. "The sample count figures will be blown to 100 per cent to arrive at the actual figure," explains a source. The fifth form, intended as a counter-check, will be filled in by the army personnel accompanying the enumerators.

The Census Organisation forms are to be filled in with a pencil, since the OMR (Optical Mark Recognition) computer at the CO does not detect any other sort of impression. This has led to widespread fears that the information provided in pencil will be erased and substituted with false particulars. A more state-of-the-art scanner that recognises ball-point pen marks is also available, some experts claim, adding that the CO should have arranged for one considering the significance of the exercise.

In the case of people who are illiterate, the information they provide will be even more vulnerable to manipulation since someone else will have to fill in the forms on their behalf. Given the complexity of the documents, however, this task may be beyond even the most educated of Pakistanis. The presence of army personnel, whose job was initially restricted to maintaining law and order, is also resented in some quarters. Dr Qadir Magsi, chief of the Jiye Sindh Taraqqi Pasand Party, believes that the presence of the armed forces will result in a lower count in Sindh, especially in the rural areas.

Others claim that villagers are afraid of the army and will therefore not take part in the census. But official sources insist that such fears are unfounded. "The army personnel are not there to harass the people," says a source. However, it is not clear what will happen if the figures compiled by the civilian enumerator and the army man do not match. "If the figures are alarmingly different," one source says, "then a fresh count might be ordered in that particular area." Officials at the Census Organisation see no reason why there should be any discrepancy, but that does little to ease the minds of the various nationalist groups.

Past experience shows that census data can be doctored at more than one stage. First in this chain of command of potential wrongdoing are the enumerators, then those who sort the forms and, finally, the central tabulators. There is no guarantee that similar malpractices will not take place this time round too.

The "Punjab factor" is another major cause for concern. At present, Punjab officially comprises 62 per cent of the country's total population, but these statistics are based on 17-year-old data. According to demographic experts, Punjab's population as a percentage of the whole is bound to fall by 12 to 15 per cent if the census is fair. "There has been continuous migration from the Punjab to Sindh, particularly Karachi," says an expert, adding that Sindh's population now stands at roughly 32 million.

The question is will the powerbrokers in the Punjab allow this to happen? Even if resource and NA seat allocations are not going to be affected by the census results, which is what the government is insisting, demands for a new, more equitable formula are bound to grow following any major change in the Punjab's relative population. And if the census does not reflect any real change, the results will immediately be rejected anyway by the other provinces, especially Sindh. When he was in the opposition, interior minister Chaudhry Shujaat Hussain went on record to say that his party was opposed to a census conducted by the Benazir Bhutto government, the reason being that the PPP would inflate the population of Sindh. With the Punjab-led PML(N) now in power, many Sindhi leaders are convinced that the census will be rigged to the detriment of the smaller provinces.

Similar ethnic tensions prevail in Balochistan. The Baloch leaders fear that the large number of Afghans settled in Balochistan will be counted in the census to alter the complexion of the province. The presence of Afghan refugees is of concern also to the Pashtun leaders, but for markedly different reasons. Their worry is that on the pretext of discounting Afghan refugees, many legitimate Pashtun residents of Balochistan may be excluded from the census. There is a demand, from both sides, that the question of who is an illegal immigrant and who is not should have been decided ahead of the head-count.

Past experience shows that census data can be doctored at more thanone stage.

The issue of illegal aliens is sensitive also in Sindh, particularly Karachi, which a large number of Afghans, Bengalis, Burmese, Sri Lankans and other foreign nationals have made their home. According to a rough estimate, there are some 1.4 million aliens in Karachi alone. "Nearly 100 localities have been identified in Karachi where aliens live side by side with bona fide Pakistanis," says a Census Organisation (CO) source. "It is for the enumerator to decide whether a particular person is an alien or not," he adds. That is a difficult task, CO officials admit, but then they don't see any alternative either.

The 1998 census results, sources say, will be compared to those from 1981. "We have a rough estimate of how much the population has increased over the years," says the same source. The results, he says, should more or less match these estimates, which have been drawn up keeping in mind the migration and growth rates throughout the entire country. But since the credibility of the 1972 and 1981 censuses is disputed to begin with, whether or not the results of this latest exercise match the official estimates will be irrelevant for many people. Regardless of the outcome, one interest group or the other is bound to feel victimised. "Some people will feel aggrieved even if the head-count is absolutely fair," says a CO official. "The census, however, has to be conducted. It is crucial to future planning."

Almost every single political party and pressure group in the country questions the legitimacy of Census '98, for varying reasons. Yet in all these discordant voices runs a common grievance: the government, they say, should have taken them into confidence before embarking on an exercise of such monumental significance. This, by all accounts, is a serious mistake on Islamabad's part. With this backdrop, while the outcome of the trial by numbers is still unclear, the reaction that awaits the results is all too predictable.

This article was originally published in the Herald's March 1998 issue. To read more subscribe to the Herald in print.

The writer was a staffer at the Herald.