Published Mar 20, 2018 07:02am

Delimitation paradox

Tahir Mehdi

THE new delimitation of electoral constituencies was bound to generate heat. The National Assembly formed a committee recently to examine the preliminary proposals for delimitation and report back within a week. The speaker did so amid rising criticism on the returns of the exercise from the members of almost all the parties. The Election Commission of Pakistan (ECP), on the other hand, has issued an order to keep the committee from interfering in the matter and, instead, has asked the stakeholders to take the available legal recourse which asks voter/s to file objections.

The thrust of the criticism of the new proposals is that the new constituencies are unequal in population.

While the average population per National Assembly seat (quota) comes to around 780,000, there are seven constituencies in which the population exceeds one million and in 18 others it falls between 0.9m and 1m. Conversely, the population of five constituencies is less than half a million and in three others it is between 0.5m and 0.6m. (This excludes Fata where the size of constituencies is half that of the national average by law.)

This has happened despite Election Act 2017 calling for constituencies to be equal in population, as far as possible, and asking the ECP to provide justification if the size of the constituencies varies by more than 10 per cent. The ECP has provided justification where it considered it necessary. But in most instances, inequality is a direct outcome of the law itself that has set the principles of delimitation.

Inequality is a direct outcome of the law itself that has set the principles of delimitation.

The most significant of these is the rule that defines the patwar/tapedar circle in rural areas and the census circle in urban areas as the smallest and the district as the largest unit. This means that a constituency shall comprise a cluster of adjacent smallest units falling within the boundaries of a district. A constituency is not allowed to cut across district boundaries which also means that each district shall have a whole number of seats.

District populations, however, are not in exact multiples of the population quota set for a constituency. This leads to districts with a fractional share of seats which then has to be rounded off; this process leads to inequality in constituencies. For example, for the current proposal, Gujranwala district’s share in the National Assembly is calculated at 6.43 seats against a population of just over 5m. Rounded off to six, the average size of the district’s constituencies increases. The population of two of the proposed constituencies, Gujranwala 3 and Gujranwala 4, thus has surpassed the quota by 9pc and 11pc, respectively. Importantly, however, variation in the population amongst these six seats is plus and minus 5pc of the average for the district.

Hafizabad district was carved out of Gujranwala in 1993. Its share is now calculated at 1.48 seats which is rounded off downward resulting in the second largest seat of the country with a population of 1.16m for the district. The collective share of Gujranwala and Hafizabad stands at 7.91. Had they not been divided, or divided differently, they would have collectively secured one more seat in National Assembly. Both of them have lost half a seat in the new delimitation and ended up with constituencies larger than the provincial average.

The case of Jhang district and its newly separated part, Chiniot district, is even more ironic. According to the delimitation report, Jhang’s share of 3.52 had to be revised to three as the district’s fractional share, despite exceeding the half mark, had to be ignored because all available seats were already occupied by districts with higher fractional share. This gave birth to three large seats with an average population exceeding 0.9m.

In contrast, the share of Jhang’s erstwhile part, Chiniot, came to 1.76 giving it two seats that are obviously of smaller than the quota size. The two Chiniot seats have an average population of 685,000. The manner in which the two districts have been bifurcated has resulted in smaller constituencies in one district at the cost of making them larger in the other.

Bannu district, with a population of 1.17m and a seat share of 1.49 in KP is given one seat which is now the largest constituency of the country while Tank district with a share of 0.50 too got one and is the second smallest seat. The two most unequal seats thus are in close vicinity.

Compare this with Faisalabad whose share of 10.1 is almost a whole number that minimised the impact of the process of rounding off and facilitated the creation of 10 equal constituencies. The variation in the size of Faisalabad seats is well within the legal limit of 10pc. Same is the case with Lahore, Peshawar, Hyderabad and many other districts.

There are two main rules to be followed in delimitation. One requires the constituencies not to vary by more than 10pc in population, and the other calls for them to remain within district boundaries. They both have merit but certainly cannot be adhered to simultaneously. If one is followed, the other is violated.

Adding to the complexity is the fact that the creation of new administrative units or changes in the boundaries of existing ones is a prerogative of provincial governments and the delimitation authorities can only follow these.

More importantly, Elections Rules 2017 had clearly delineated the process giving priority to keeping the constituencies within the district boundaries. The draft rules were made public last October inviting objections from all stakeholders but no political party bothered to study and analyse their impact at that time and suggest changes.

The solution lies with parliament and the political parties. They need to come up with a comprehensive legislative plan to create administrative and political boundaries and align them with each other. Attending to just one could only result in the paradoxical situation that we find ourselves in.

But for the current delimitation the opportunity is already lost. Political parties can, however, use it to learn lessons and chalk out future plans and be mindful that taking the controversies surrounding the current proposals too far has the potential of harming the democratic discourse.

The writer is an independent researcher with an interest in elections and governance.

Published in Dawn, March 20th, 2018

Read Comments