Jumping Ship: The Story of Party Switching in Pakistan

Over tea and sandwiches, a prominent politician from Punjab emphatically declared to my colleague and me, the irrelevance of political parties to Pakistan’s electoral setup. Parties, he opined, are simply vehicles for ‘electables’ to gain power.  A brief look over his political career seemed to confirm how little political parties mattered to him, having contested his 5 national elections under the banner of 3 different political parties.  Such successive party switching is not uncommon in Pakistan and, perhaps, tells a sad story of the cynical leveraging of the party-system for personal interests. However, it is important to study the impact of such defection on a politician’s career before dismissing parties outright.

In each Pakistani election there exist some candidates who have changed their party from the previous election year. These defectors number anywhere from 500-800 per election, amounting to an average of 19% of total candidates per election.  Almost 60% of the defectors across the year met with reasonable success pre-defection, ranking in the top 3 in their previous election.

Around 50% of these defections occurred in Punjab, with Sindh and KPK standing at roughly 20% and 17% respectively. While these numbers may suggest that Punjab rewards party switching (with the province already having a bad name in some quarters for opportunistic politics), this claim cannot be made without comparing how successful candidates have been in securing a victory post-defection, in each province. The table below details the province-wise percentage of defections and non-defections that resulted in a win from 1990-2008. Column 1 of the table shows % of total occurrences of defections from 1990-2008 which resulted in a win and Column 2 shows % of total occurrences of non-defections from 1990-2008 which resulted in a win.

25% of total defections in both Punjab and Sindh resulted in a victory, while the corresponding percentages for Balochistan and KPK were 27% and 23%, respectively. This indicates that a higher number of defections in Punjab does not necessarily correspond with a better political career for Punjabi politicians. In fact, juxtaposing these numbers with the success rates of non-defectors, weakens the argument that defection increases the likelihood of a politician winning an election.

So if defection does not necessarily increase chances of victory and comes with the risk of a politician being branded as a turn-coat, then why is party switching so common?

One possible answer can be found in our politician’s assertion of parties not mattering in Pakistan’s electoral landscape. However, confirming whether politicians defect because they don’t need parties, or because they DO and see the winning party as integral to their success would become a complicated task. Simply looking at a defectors success post-defection would not isolate the impact of a politician’s own electoral success, from the impact of a change of party.  For instance, a post-defection victory for a politician who consistently won his/her constituency before defection cannot be lumped with a post-defection victory for someone who did not rank very well in the previous elections. In the case of the former, defection might be seen as having little effect, while in the latter case, success could, in part, be attributed to their party. The table below breaks down the winning defectors and non-defectors in each province, by their rank-mobility.

In Punjab, 11% of total defections were converted into a post-defection win by someone who ranked 2 or 3 in the previous election whilst in 15% of the cases where an individual did not defect to another party, he/she rose 1 or 2 positions to win the next elections. So in all provinces except Balochistan, a non-defector has a higher chance of electoral–rank mobility than a defector. But specifically looking at payoffs for defections,  Punjab and KPK give a lower-ranked defector a decent chance of improving his/her rank, so clearly in these provinces there is a fair share of  the ‘party vote’ captured by defectors once they switch parties.

We can also look towards the direction of defection, to gain better insight into the incentives of party-switching. It is unlikely that ideological bent is an important factor in the consideration since, across the years, politicians are mostly defecting to incumbent parties. The exception is 2002 where PML-N, the incumbent party, had for all effects and purposes been disbanded with the Sharif brothers in exile. The wave of defection into the winning party, PML-Q, was unprecedented. However, PML-Q did seem like the default choice with the other two political parties languishing without their leaders.

An explanation of this desire to join the ‘winning party’, was provided by a Senator we interviewed who considered party-switching as important to providing patronage. So politicians join the party they feel would win the elections, because being part of the treasury benches improves their constituency’s budgetary allocation. Another politician stated that electables don’t want to be part of the losing party, as that would be considered a weakness by their constituents.

However, this tactic also seems to backfire, as incumbency is not a good indicator in Pakistan of electoral success. So, the wave of politicians who joined PML-Q hoping they will replicate their 2002 victory in 2008, would have been sorely disappointed. The same trend can be seen throughout the years as a number of politicians tend to join losing incumbent parties.

So in part our Punjabi politician was correct in seeing parties as vehicles driving electables to victory. But the numbers do not speak much for the soundness of this strategy. Perhaps, investigating the evidence for the value of defection would help the politicians gain some foresight before blindly following the bandwagon of party-switching.

Disclaimer: The views expressed here are those of the authors and do not necessarily represent the views of the Institute of Development and Economic Alternatives