Alert Sign Dear reader, online ads enable us to deliver the journalism you value. Please support us by taking a moment to turn off Adblock on

Alert Sign Dear reader, please upgrade to the latest version of IE to have a better reading experience


A hazy crystal ball

May 09, 2013

MOST voters will not decide on Election Day about the candidate for whom they’re going to vote.

If someone has been voting for party ‘A’ there’s a good chance they will vote for party ‘A’ again.

Old habits die hard.

Galileo insists that the universe cannot be read until we have learned the mathematical language in which it is written. Statisticians study possible outcomes or a ‘voter swing’ in an election on the basis of comparative data from past elections results and public perceptions about the recent performance of political parties. Similar techniques are used to decipher post-election fiascos.

When a voter who previously voted for party ‘A’ decides to vote for party ‘B’, he or she becomes part of a small percentage of voters, usually as low as 5pc, giving party ‘B’ a 10 percentage point lead.

However, any voter swing away from the earlier winning party is not automatically a swing in favour of the runner-up. It could be divided by new entrants such as Imran Khan.

Let us begin with the past performance of leading political parties contesting in each of the provinces separately while maintaining a category of ‘others’ for independents and smaller parties. For optimal size of data, standardise the number of electoral seats for at least three past elections, assuming that the number of seats in earlier elections were at the 2013 level. This is done by an appropriate scale change of the actual number of National Assembly seats for 1997 to a comparable number of National Assembly seats in Punjab, Sindh, Khyber Pakhtunkhwa and Balochistan at present, i.e. 148, 61, 35 and 14 respectively.

The size of the available data does not allow analysis for the two Islamabad seats or the 12 Fata seats where the past elections were conducted on a non-party basis.


The actual vote share and seats obtained by the various parties are listed in the table below.

In 2008, the 50 most urbanised seats were captured by the PML-N. Although incumbency has helped its outreach in rural Punjab, the PPP could still spring a surprise here, not least because of the latter’s seat adjustments with the PML-Q.

While the PML-N may be a formidable adversary, Imran Khan’s promised tsunami through the PML-N heartland may turn the upcoming elections into the most rigorously contested since 1970.

The PPP could suffer losses due to its poor performance during the last five years. However, in the 2008 elections, the party was able to bag 46 National Assembly seats in Punjab against 34 runners-up from the PML-Q. Similarly, there were 12 PPP runners-up against 28 seats won by the PML-Q. Seat adjustments in such constituencies could help the two parties.

Imran Khan’s Pakistan Tehreek-i-Insaf is clearly betting on the urban youth vote in Punjab, which pitches the party directly against the PML-N.

While the PTI claims that the youth vote in the country is 45 million, the fact is there are only 13 million registered voters between the ages of 18 and 23. This translates into 48,000 fresh voters on each of the 272 constituencies.

Historically, young people have not been the most dependable of voters. Even if the charged political environment in Punjab does mobilise the youth in favour of PTI candidates, would this translate into a deciding vote for the PTI or only help Imran Khan act as a spoiler for the PML-N?

Relatively unknown PTI candidates would need far larger number of votes to cause real injury to the PML-N electables, two-thirds of whom bagged victory with more than 70,000 votes in 2008. To cut into the PML-N’s loyal vote bank in Punjab, the PTI will need a very high voter turnout.

In the absence of data for PTI’s past performance, calculating ‘swing’ is only possible in favour of the PML-N as given in the table below.


The PPP and the MQM have been rivals for political power in Sindh for over 25 years. However, the burden of poor governance along with allegations of sycophancy, corruption and a dubious alliance between the two could cost them in the form of lower voter turnout in Sindh.

The rural poor tend to vote for the PPP. However, restrictions barring parties and candidates from providing transport to the voters may hurt PPP candidates the most.

National Assembly seats, actual vote share and seats obtained by the various parties are listed in the table below.

During the last general elections the winning margins of PPP candidates exceeded 20pc and more. Even with a 10-party alliance opposing the PPP and up to 10pc vote swing against the party, it may not shed a significant number of seats from Sindh, provided it is able to bring its remaining voters to the polling stations.

In 2008, the MQM was able to bag 19 seats in the National Assembly from Sindh out of which 17 were won from Karachi alone. However, the 2002 elections, when the Jamaat-i-Islami (which boycotted the 2008 elections) was competing against the MQM in Karachi and Hyderabad, may better represent the party’s actual strength. With limited swing effect, if the MQM is able to mobilise its vote bank, the party could bag close to 14 or 15 NA seats in Sindh with 12-14 seats from Karachi alone.

Khyber Pakhtunkhwa & Fata

The actual vote share and seats obtained by the various parties in the National Assembly elections from Khyber Pakhtunkhwa are listed in the table below.

In 2008, of the 35 NA seats from Khyber Pakhtunkhwa and 12 from Fata, except for four seats from the Hazara division claimed by the PML-N, one each in D.I. Khan, Bannu and Lakki Marwat won by the PPP, the MMA and the PML-N, the remaining 40 seats saw a very low voter turnout. These seats could be easy prey for the PTI or the PML-N.

On the other hand if religious parties capitalise on 3-7pc swing in Khyber Pakhtunkhwa over their 2008 vote, they could go from four seats to six or even seven seats.


The actual vote share and seats obtained by the various parties in the National Assembly elections from Balochistan are listed in the table below.

Political parties are drawing fresh battle-lines here. However, in the current law and order situation where returning officers are hesitant to perform their election day duties, ultimately the deputy commissioners could be deciding the winners in consultation with local commanders.

In the final analysis, predicting the outcome of Saturday’s elections is very complex. An unprecedented environment of fear in the three smaller provinces and the unpredictable element of militant attacks in Punjab are not helping the situation.

No one may have found the silver bullet to this conundrum, but one thing is for sure: voter turnout will play as important a role as ‘vote swing’ in the outcome of the upcoming elections.