In this space, we would like to draw the attention of the stakeholders on a crucial issue of constituency delimitations which under the current practices, grossly violates the one-man-one-vote principle. We would also make a few comments on improvements in population census that have a bearing on the issue.
Population clusters are defined for the election constituencies of national and provincial assemblies. Last such delimitation was done in the year 2001, immediately before the 2002 general election. This delimitation was done based on 1998 population census. There is otherwise an elaborate procedure of constituency delimitation, which is subject to input and review by the public. And such delimitations are finalised after the due process. However, there is a fundamental flaw in the constitution delimitation process adopted by the Election Commission of Pakistan, whereby constituencies are forced-fit within the district boundaries without regard to the imbalances and inequalities generated in the process. Throughout the world, federations or unitary systems and constituency circles are free to spillover and beyond the district boundary. In federations, provincial boundaries are sacrosanct and constitutional units, districts are not.
Let me elaborate the procedure itself. Firstly, provinces are allocated the number of NA constituencies, in proportion to their share in the population, which is correct and fair. As a second step, districts are allocated the seats, which is neither correct nor fair. Why? Assuming that a district (Dir) has a population of 749,000, while a population of 300,000 gets an assembly seat, this district would get 2 seats if the population is 749,000 and 3 seats if the population is 751,000. With a population difference of just 2000, a district can get an additional seat, depending on the coincidence and circumstances, which are not rare. This violates the fundamental principles of human rights of one-man-one-vote. In this system one-man does have one-vote, but has unequal effect on electoral outcomes.
How to correct the flaw? The notion of taking district as a fundamental and sacrosanct unit has to be abandoned. Define independent clusters, with only some regard to Tehsil and districts. All NA and PA constituencies should have equal population. Only a village should not be divisible for logistics reasons which mean a margin of ± 1000 persons. Currently, the NA and PA constituencies have a large variation in sizes. District is too large a cluster to be considered indivisible. This is unprecedented anywhere in the established democracies. Electoral districts are independent of administrative districts.
In a country with a high population growth and high migration, the notion of established constituencies can at best be clung to only partly, if at all. There have been practical and wrongful implications of this policy of constituency delimitation. If I remember correctly and I think I do, the case of Swat-Dir area which got 2 NA seats instead of three they would have got, had the district sacrosancy policy not been applied. It would be farfetched to trace the current political problems in these areas to such issues, but certainly democracy is throttled and compromised, in numerous similar ways, which results in socio-political conflicts.
In a sample of seventeen districts and 150 National Assembly constituencies in Punjab, we analysed the constituency delimitation of the seats of the National Assembly. Average population per constituency in the sample was 490,808, while average number of voters per constituency came out to be 296578. Hafizabad with average voter strength of 201328 got two NA seats, while Vehari with more than twice the voter strength got one NA seat for that many votes. This variation in constituency size is mind boggling; from 201328 to 441,879 voters per constituency. Similar variations should be there in other provinces and also in the provincial constituencies, for the formula is the same. There ought to be some limit to the arithmetic simplification and the sacrosancy of the districts' indivisibility.
Liz Hadley's (USAID) study shows that 53% countries used population as basis, and 34% used voter strength as the corresponding basis. Voter lists are more dynamic than population census, and are subject to error and inadequacies and are never reported to be complete or error-free. In the US, strict adherence to equal population principle is applied but a tolerance of ±0.5% has been tolerated in limited cases. In Australia, the allowed deviation margin is 3.5% and in Germany 15%. In Germany, due to proportional representation system, the issue has much less relevance and intensity. Venice Commission on code of good electoral practices, recommends maximum deviation to be 10% with an occasional allowance of 15% in exceptional circumstances. European Commission report on Pakistan elections 2008, has also recommended equalising the constituency populations by modifying the constituency boundaries.
Constituency delimitation would have to be done anyway again in the wake of the new population census which has been postponed for unknown reasons. There is always an anxiety about the consequence of population census among the powers that be. The last census in 1998 was done after two decades. The interim census was mired with controversy and was rejected. There seems to be much ado about holding census. No body, even the political parties are talking about it. Holding of census is a constitutional requirement and the honourable Supreme Court has also not taken any notice. Word goes by, that the delimitations' basic work was done by the security/intelligence agencies, which usually have all the data, even on voting patterns of Mohallas. There appears to be some merit in the word that goes by; more the reason to revise the constituency delimitation to eliminate the shadows of Musharraf regime ala Baqiate Zia-ul-Haq.
Reportedly, voting lists have been improved by Nadra and are in the process of final compilation. Constituency delimitation can be done based on voters' list as well. This time it ought to be done more scientifically and fairly respecting the voting strengths of the constituencies and minimising the variations in size as has been discussed. Modern tools of GIS technology should be employed, which afford an interactive design of constituencies with the possible involvement of the stakeholders. This would minimise later objections and would save time, effort and resources.
And a few words about population census. Purpose of the census publishing is to let all and sundry, know the population and other characteristics of the smallest of population clusters in the country. The census data is used by myriads of local, national and international agencies, development personnel and private sector businesses. However, the data published is highly cryptic, at least, in so far as the urban clusters are concerned. Instead of referring the urban clusters by their names as Nazimabad, Saeedabad, Model Town etc it refers to the clusters as a number, which nobody knows. To find out what the number refers to, one has to visit the local or regional Federal Bureau of Statistics (FBS) or census office, which is very time consuming and wasteful. Population Census Organisation (PCO) could do one or more or all of the following to addressing the situation.
1. Replace the numbers by area names wherever feasible; organize and publish data for transparency. The data that is given out by local census office is usually in the form of disjointed list that should be streamlined.
2. Publish census maps, although PCO would face a lot of resistance from Pakistan's security establishment and its archaic security agencies, who shudder to publish any map at all. They can and should be persuaded, as almost everywhere in the world, census maps are published.
3. Census would become a tremendously efficient tool in development planning and business management for both private and public sector, if it is disseminated as maps, in the frame work of GIS standards.
4. Instead of printing 104 volumes, it should be released as a software data product and also as in web-based, interactive mode. And it should be free.
5. As per current practice, population data on Defence Housing Societies and Cantonment boards and other similar areas is not published. This practice ought to be done away with.
Comments
Comments are closed.