You are here: Home / Pakistani Politics / The Definitive Five Rupees Election Post
The Definitive Five Rupees Election Post

The Definitive Five Rupees Election Post

So it wouldn’t be a real election if there weren’t at least one pre-election post on FiveRupees. As a disclaimer though, this isn’t a prediction post – (one would have to be very stupid to put ungrounded fluff on the Internet a few days before an election) – but you can think of this as a stuff-to-watch-out-for guide. Yes, I know, we’ve all seen people giving party-wise seat-counts down to NA-272, but that’s mostly hot air sprinkled with snippets of anecdotal information. I say this based on the fact that nobody has any constituency-level survey data, or a systematic way to capture qualitative assessments in each electoral space. In a recent visit to Sargodha, every person I met told me they were voting for PML-N like they’ve always done, but I knew *some* of them were lying. Why? Because when you see polling station data from the 2008 elections, nearly every polling station in the district saw vote-fragmentation.

In any case, this is a tough election to call, and as the case happens to be, probably one of the most tightly contested – at least in Punjab and Khyber-Pakhtunkhwa. Elsewhere, I’d defer to the judgment of my far more well informed friends from Sindh, Balochistan, and FATA.

Point 1: Turnout

A lot of people have rightly identified electoral turnout as the key determinant of constituency-wise outcomes in this particular election.  This boringly obvious point stems from the view that the Pakistan Tehreek-i-Insaf is mobilizing first-time voters to come out and play their part in building a New Pakistan ™.

This is, by and large, true. The PTI has successfully managed to make inroads into the non-voter psyche and convinced them of the need to engage in this rather simple act of civic participation. The second, and less talked about facet in the turnout discussion though, is the impact of the Benazir Income Support Program. For those who don’t know, the BISP is Pakistan’s flagship unconditional cash-transfer program for females in the poorest households of the country. It was conceived by the previous PPP government, enacted through a unanimous act of parliament and given technical and financial support by DFID, Worldbank, and USAID. A fundamental requirement of the BISP is the possession of a National ID Card, which also happens to be a fundamental requirement of voting. Nearly 4 million women were enrolled under the BISP program and given ID Cards over the last few years. This will also have a yet-to-be-determined impact on voter turnout.

The turnout debate is premised on the fact that Pakistan sees a smaller percentage of the electorate participating in each election, compared to other countries in the region. This reality was captured really well in these two recent articles by Dr(s) Cheema, Bari, and Naqvi on voter turnouts, and female voting.

Overall Punjab Sindh KP Balochistan
























































Electoral Turnout over time, source: ECP data

Electoral Turnout over time, source: ECP data

Looking at historical turnout trends nationally, as well as in all four provinces, the fall looks quite stark. From the dizzying height of 60% in 1970, we’ve plunged all the way down to 37% in 1997, and for the last two elections, we’ve hovered around the barely respectable 40-45% mark. Going one-level below the national average though, the turnout story for each province is quite different. Punjab and Sindh have always been more involved in the democratic process – if one were to assume voting as an accurate proxy for it – while KP and Balochistan have seen a steady decline in their electoral participation.

The security situation in the two provinces explains the problem for the last decade, but not so much for the preceding one. Logistics could be a valid reason, since it’s probably harder to negotiate remote terrain in Balochistan and KP, and make it amenable to greater participation, as opposed to the flatlands of Punjab and Sindh. Anyway, given that terrorist-induced volatility has escalated over the course of election-season, specifically with the targeted attacks on ANP, PPP, and now on other parties as well, along with the TTP issuing repeated statements about its complete opposition to democracy, greater turn-out in KP looks doubtful. In Balochistan, the army’s been ‘sent-in’ to manage election related activities, which leaves me unclear as to what impact it will have on turn-out, even if I have an idea of what it’ll do to results.

The silver lining though, is that the province’s plethora of nationalist groups – National Party, BNP-M, PKMAP, and JWP – are all participating this time around, so that’ll probably push the number up from last time’s dismal showing.

For me, the interesting question is how high Punjab can go in terms of electoral participation. Everyone and their aunt’s been talking about how a higher turnout favors Imran – presumably because a statistical majority of new voters are PTI voters in Punjab, – but in a province where the turnout already hovers around 50%, I’m not sure we won’t run into a bounding problem on this front.

Not to make a prediction (god forbid) or anything, but people can bank on an increase in turnout in Lahore (37%), Faisalabad – urban (45%), and Gujranwala (37%).

(Just to put things into perspective though, 69 constituencies in Punjab (out of 148) had a turnout of higher than 50%, while 43 had a turnout of less than the national average of 45% in 2008.)

Major takeaways:

1)   Turnout will be higher in Punjab, but less than the 75% mark that some are predicting;

2)   Security concerns will mute new-voter sentiment (if there is any) in KP, Karachi, and Balochistan;

3)   Wheat threshing and cotton sowing, as well as ECP restrictions on private transport might prove to be problematic in rural parts of the country

4)   PTI’s admirable drive to bring out first-time voters and apathetic sorts is going to be a major factor, especially in the 45-odd wholly urban constituencies of Punjab;

5)   7.5 million families registered under the BISP, with 4 million of them getting new ID Cards. That has to count for something as well, right?

6)   My weather app tells me it’ll be 44 degrees Celsius on Election Day. Scoff all you want, but it’s a bigger factor than most will be willing to admit in public.

Point 2: Tight races in 2008

Since this isn’t America and my parents didn’t name me Nate Silver, using past electoral outcomes to predict what’ll happen in 2013 would be a wholly futile exercise. For starters, there’s the new PTI factor that changes constituency-level dynamics in at least two provinces. Secondly, there’s no way of assessing how anti-incumbency actually works for the PPP or the PML-N. What we’ve had in the past are anti-dismissal effects – i.e. whereby a segment of the electorate in certain regions rescinds support from a particular party because it feels it’s fallen out with the establishment/powers-that-be. Thirdly, I’ve yet to see a valid model that blends in an accurate estimate of party-vote and candidate-vote in Pakistan.

So let’s just go with simple math here:


Winning margins - Sindh and Pakistan (overall), source: CERP

Winning margins – Sindh and Pakistan (overall), source: CERP

There were two parties in Pakistan that won with huge margins in 2008, and both of them were in Sindh.


PPP: Distribution of Winning Margins

Winning Margin (%) Seats Total Seats Percentage






















MQM: Distribution of Winning Margins

Winning Margin (%) Seats Total Seats Percentage


















Some of these margins are actually quite phenomenal. What these tables show is that to displace the MQM on 12 of its strongest seats, the runner-up candidates would have to get at least 80% more votes. Same goes for 8 of PPP’s 31 seats.

In 1997, when the PPP suffered its biggest drop in vote-share in the province (a 15% negative-swing from 1993), it was reduced to 19 seats from 33 in the preceding election (out of 49). Let’s assume the PPP suffers a swing of twice that amount – i.e. 30% – in 2013: it would lose *only* 3 of its 31 seats. So basically, the PPP isn’t going anywhere anytime soon. Neither is the MQM, which has only consolidated its vote bank in the last decade, so a swing seems out of the question anyway. The new party knows this quite well, which is why its left Sindh to the old anti-PPP forces – nationalists, and Pir Pagara – , Karachi to the MQM, and concentrated on the two more electorally volatile provinces – Punjab and KP.

The PML-N won 61 seats in Punjab in 2008, out of which 15 were won with a margin of less than 10% votes. Around 29 were between 10% and 25%, while the rest were won much more comfortably. The PPP won 44 seats in Punjab of which 25 were won with a margin of less than 10%. Finally, the now-almost-defunct PML-Q won 28 seats in Punjab of which 24 were slim victories on the less-than-10% margin.


Tight contests in 2008 - Punjab, source: CERP

Tight contests in 2008 – Punjab, source: CERP

Not factoring in candidate-switching, and obvious changes in the party-vote component at the local level, one could say these are seats that would be most vulnerable to an assault by challengers. However, since that’s impossible to control for, all this tells us is how comfortable each party was in Punjab in 2008, and if say the PML-N, for example, holds on to these marginal seats, it’ll remain in reasonably good shape overall as well. All said and done, exactly 50% of all National Assembly seats in Punjab were won by a margin of under 10%, which makes the province far more prone to changes in voter preferences AND an increased turnout with a party-based skew.  PTI supporters should take hope from this fact.

The story in Khyber-Pakhtunkhwa is along the same lines. Victory margins in KP have been small, and well below the national average. In 2008, nearly 60% of the seats in KP were won with a margin of less than 10%, compared to 41.5% in Pakistan overall. No seat in KP was won with a margin of greater than 50%. Similarly, vote-shares have remained historically fragmented, and elections have been competitive for all major parties in the province.


KP: Vote and Seat Share (%)



PPP (Vote-share)

PPP (Seat-Share)

Religious Parties (Vote-share)

Religious Parties (Seat-Share)

ANP (Vote-share)

ANP (Seat-Share)












































What should be worrying for the ANP – though less so than the fact that it’s literally being hunted down by the Taliban – is that small changes in vote-shares can have disproportionately large impact on seat-share percentages. So taking 1997 and 2002 as examples, a drop of 10 percent in vote-share saw a drop in 38 percent of the NA seat-share. This is where the PTI, or any other party, can make a dent in the province: capturing voting regions traditionally held by the PPP and the ANP in the province. The Muslim League appears to be in reasonable shape in Hazara division, as does the JUI-F in the south – though recent maneuvers by Imran in the trans-Indus districts has made Fazlu quite uncomfortable.

Major takeaways:

1)   Barring something catastrophic, the PPP and MQM will retain most of their seats in Sindh. Their victory margins were quite large, and while the MQM has the party machinery to grind out results in Karachi, the PPP has done enough in terms of patronage disbursement to ensure its reelection in the Sindhi heartland.

2)   The PPP has become largely uncompetitive in North and Central Punjab though, which is worrying for a party with federal credentials. It had 18 seats in the region, of which 10 were close victories and are being constantly challenged by N and PTI. Party stalwarts like Nadeem Afzal Chan, Qamar Zaman Kaira, and Imtiaz Safdar Warraich appear to be in trouble;

3)   The PML-N won comfortably in Punjab, but the downside for them is that it won in a province which has a history of vote-swings, and perception-based sweeps. This and the increased turnout is what the PTI is banking on to upstage Nawaz Sharif’s attempt at becoming Prime Minister for the third time;

4)   The PML-Q might survive because of its seat-adjustment with PPP in South Punjab, but it’s unmistakably converging towards a much-deserved demise. The electable candidates have mostly been absorbed into PML-N, as well as the PTI, and this process will continue unabated till nothing is left of the party except Ch. Zahoor Elahi’s progeny.

5)   KP remains fragmented and electorally volatile, and perhaps the toughest province to call. If analysts are to be believed, PTI has a chance of emerging as the single largest party in the province – though that does not necessitate the formation of a provincial government.

Point 3: Old versus ‘New’

I haven’t had the time to look at candidate lists in the other three provinces, but I’ve ran the NA nominees from all three (four – counting PML-Q) major parties through our secret election database to check whether they’re new-entrants, existing politicians (who’ve contested an MPA or MNA election before), or relatives-of-retired-politicians.


Party Total (/148) Existing New Relatives




















National Assembly Candidate Types - by party, source: CERP/IDEAS

National Assembly Candidate Types – by party, source: CERP/IDEAS

As the bar-chart shows, PTI has fielded a large amount of new candidates, as well as a not-insignificant amount of relatives of retired politicians. The rest of their pool is filled with fresh-entrants, and third-tier candidates from Jamaat Islami, PPP, and various factions of the Muslim League. Wasn’t specifically looking for old-timers, but I came across two candidates who contested on a PTI ticket in 1997 and 2002 – Hamid Khan and a candidate in Hafizabad.

PPP, more by default than by design, has had to field a host of new candidates as well. If anecdotal evidence is anything to go by, proposed candidates refused to contest on the party’s ticket – fearing a backlash given what’s happened in the last five years. Hence, what we see now is that it’s left the party with no option but to put up secondary, largely untested candidates. PML-N, on the other hand, has gone with tried and tested sorts in 136 NA constituencies. This is quite symptomatic of the party’s view on how to grind out election results in peri-urban and rural areas – i.e. forge local level alliances to produce a winning outcome, as opposed to running direct-voter-based campaigns through an institutionalized party machine.

Major takeaways:

 1)   Regardless of electoral outcomes, the number of new candidates contesting in this election will ensure the initiation of new (and competing) patronage networks aimed at displacing pre-existing ones;

2)   Whether the PTI does well or not is a secondary issue on this front. There will be candidates who establish personal credentials in tight-races across the country, which might help them to capitalize on local government elections when and if they take place.

3)   If – and that’s a pretty big qualifier – the PPP does well (>10 seats) with this candidate pool in North and Central Punjab, it’ll be a testament to the party’s bench-strength. That said, I just don’t see it happening.

Outside the numbers:

I’ve dealt with enough numbers to last me a life time, so I’m going to switch to the far more familiar genre of anecdote-driven grand-narrative. The PPP has ceased to be competitive on a majority of around 100 NA seats in Pakistan. It’s probably uncompetitive on more, but here we’re talking about an entire region, which houses a little over a third of the country’s population, that (mostly) despises the party and what it’s done in the last five years. This particular belt prefers service delivery to the politics of rights, specifically because it doesn’t have to face issues over the latter. It is also precisely why pro-economic growth and ‘sovereignty’ messaging carries greater resonance here than in other parts of the country.

My sense is that this election, like arguably all recent ones, is being dictated by the politics of Punjab’s middle class – be it on the issue of corruption, drones, self-sustenance, or increased civic participation. More than that, there’s very little that differentiates the two Punjab-centric parties – i.e. the PTI and the PML-N – as far as their ambitions for the country are concerned. The only difference is that the PTI’s ruling coalition – i.e. people calling the shots apart from Imran – has greater representation from the landed elite, and the urban white-collar class. The PML-N’s decision-making body is a largely urban club, with a largely urban though wholly nativized leadership (think white shalwar-kameez, gold watches, and three expensive cell-phones). As my friend Ammar Rashid points out rightly, it is this failure to sufficiently accommodate a burgeoning suburbanized segment of society in Punjab that has led to the PTI experiencing a form of ‘ideologically vague populism’. The nature and inclination of that populism is a separate debate all together.

In this evolutionary model, the PPP is two steps behind. Not only do they have no traction with the suburban class of Punjab, they’ve completely failed to make inroads into the urban class as well – i.e. the traders, shopkeepers, and others associated with a transitioning economy. Failing to adapt to the demographic reality of rapid, liquid urbanization – characterized by greater demands, less patience, and increased self-centeredness (in the name of the country, of course) – the party is being elbowed out of a space it carved out in the first place.

Those better informed will be able to enlighten readers about similar processes taking place in other parts of the country. My personal hunch is that most extant views on urbanization in Sindh are also quite static and outdated. Recent visits to southern Sindh were around a year ago, and I remember making a mental note of the investments made in physical infrastructure – specifically in road-networks and higher education infrastructure. Such investments don’t just happen because ‘waderos’ need to travel to their estates, but usually because there’s a tangible demand coming from actors involved in commerce and commodity-exchange. If say 15 years down the line, the PPP continues to be the most popular party in Sindh, it will show that it’s remained in-sync with the political economy of the province – something it has largely failed to do in Punjab.  Bottom line is that there’s probably a middle class story in Sindh as well, and it’s not only confined to Karachi and Hyderabad.

Where does the Establishment stand?

Unlike previous elections, especially in the 90s, the military establishment appears to have fewer electoral cards this time around. There have been moments that can be characterized as a ‘confluence of interest’ between the PTI and the army, or the army and the judiciary, but there’s been very little overt gaming of the sort we saw in the 90s. This is also probably why the election is a lot more competitive in Punjab, because people are unsure as to which party has ‘unseen’ support. Assuming that the military establishment has one clear-cut goal, i.e. control over certain policy affairs and the unfettered ability to expand its economic power, it would, ideally, want an unwieldy, incoherent coalition at the center – think PPP + PMLN, or PMLN with MQM and Sindhi nationalists. This would prevent the party with the plurality from exercising greater authority on a unilateral basis.

Finally, a generic prediction of sorts:

1)   PML-N might be on the defensive these days but it still appears to be strong in Punjab and Hazara division, and will most likely be the biggest party in parliament. Their aim was to remain within 30 seats of the magical 137-seat mark. I don’t see that happening, so expect a bits-and-pieces coalition if the PML-N does come out on top;

2)   The PTI’s election campaign has been impressive, and Imran Khan’s rapid-fire rally mechanism, and of course now THE FALL, has given them that second wind they were looking for. That said, it’s hard to predict what their overall haul would be, even if one factors in all sorts of reports coming in from the field. Personally, I’d be very surprised if they got less than 40 seats out of 272 across the country.

3)   PPP (31 seats) and MQM (21 seats) to retain their government in Sindh, barring a huge upset by the 10 party alliance led by Pir Pagaro.

4)   JUI-F to do well in the trans-Indus districts of KP, parts of northern Balochistan, and maybe even on a couple of seats in Sindh – (I’m quoting Saba Imtiaz on this).

5)   The Jamaat Islami will be decimated, and we can all be very happy about it.

6)   Nationalists will pick up a plurality of the 14 seats in Balochistan. They will surely be a part of any coalition made at the federal level, as will the MQM.

Note: All data used here has been constructed using original data from Dr. Ali Cheema, Dr. Faisal Bari, Dr. Farooq Naseer, and Dr. Ali Asjad Naqvi at the Center for Economic Research in Pakistan (CERP) and the Institute of Development and Economic Alternatives (IDEAS). Many of the data-based ideas were given by the same gentlemen, as well as by Omar Qasim, Anum Malkani, Bisma Haseeb Khan, and Dr. Hassan Javid.

About Umair Javed

Umair is a political economy researcher based at the Center for Economic Research in Pakistan (CERP).

Comments are closed.