It is an old adage that the party that wins the Punjab forms the government in the centre. Indeed since 1946, elections in the province have become the battleground for Pakistan. Hence, as the countdown to the 2008 election begins, it is pertinent to analyse lessons from the electoral history of the province between 1988 and 1997, particularly the electoral fortunes of the main political parties. It is also important to analyse the impact that the General (retd) Pervez Musharraf regime has had on electoral politics in the Punjab and the implications of this for the February 18, 2008 general elections. And last, at this juncture, it is imperative to consider the implications of Benazir Bhutto’s tragic death on electoral outcomes.

The past National Assembly elections in the Punjab reveal three salient facts. The first is the consolidation of the two-party competition in the province between the Pakistan Peoples Party (PPP) and the Pakistan Muslim League – Nawaz (PMLN) in the post-Zia and pre-Musharraf era. During this period the two main political parties consolidated their respective positions: together they held more than two-thirds of the National Assembly votes and seats in the province (see Table 1).

The second salient fact is the erosion of the PPP’s electoral position in the province: its share of the Punjab vote stagnated between 1988 and 1993 and declined considerably in the 1997 elections. In the same time period, however, the PMLN’s share of the vote in the province increased (see Figure 1).

Alternatively, a party’s performance can be judged from the proportion of “core” seats under its control, that is, seats that it has held consistently across different elections. During 1988 to 1993, the PMLN enjoyed a far larger share of such core seats (see Figure 2: the 1997 election was ignored as it represents an atypical sweep for one party and may not be a useful predictor for the coming elections). Such seats accounted for around 10 per cent of all National Assembly seats in the province.

This trend is also revealed from each party’s share of impenetrable seats, that is, seats that it could not even win once during these three elections. Such seats account for nearly 15 per cent of all National Assembly seats in the province between 1988 and 1993. PPP was far less able to win these seats than the PMLN. Hence, while two-party competition was being strengthened in the province, the PMLN was consolidating its position at the expense of the PPP. This partly reflects the woes of fragmented party organisation, and in part, a shift towards centre-right politics in the province.

The third fact is that a change in the seat share of the two political parties appears unrelated to changes in their vote share. Table 2 shows that the change in the PPP vote share between 1990 and 1997 did not yield a similar loss of seats. Likewise the increase in the votes won by the PMLN in 1990 and 1997 was far outstripped by the increase in the seats it obtained during these elections. Clearly, a small percentage change in the vote share of the non-incumbent party across elections resulted in a much bigger change in its seat share. This indicates that a fairly large number of constituencies in the province were “marginal” constituencies, or those where a mere 10 per cent increase in the votes polled by the runner-up is sufficient to change the election outcome.

By this yardstick, between a quarter to a third of all the seats in the province can be classified as marginal — excluding the 2002 general elections. Furthermore, such seats appear to be equally divided between the two main political parties (see Table 3). It is significant that it was the swing vote in these constituencies that resulted in the large scale switching of seats between the major political parties across different elections. This suggests that swing voters exist in the Punjab and are often central to determining electoral outcomes. It also suggests that the PPP continues to remain significant, electorally, because it has a chance of winning these marginal seats.

In addition it seems that each time the establishment has backed the PMLN in its bid for political power – 1990 and 1997 – this has led to an increase in the party’s vote share along with an increase in its seat share. However, when the establishment has opposed the PMLN’s bid for political power – as in 1993 – there has been a vote swing away from it. It appears that the swing voters in the Punjab either have a strong anti-incumbent bias or have been systematically second-guessing the establishment’s attitude towards the PMLN. Interestingly, Table 2 also shows that the swing voters were unwilling to realign with the PPP even when, as in 1993, they voted against their preferred party.

However, in what way has Musharraf’s takeover in 1999 changed all this? The 2002 National Assembly election results provide the answer. The Punjab’s two-party competition of the previous decades has given way to a three-party competition. The consolidated vote share of the PPP and the PMLN in 2002 was 43.2 per cent, while the vote share of the establishment backed Pakistan Muslim League – Quaid-e-Azam (PMLQ) stood at 34.3 per cent, indicating the emergence of a three-party competition.

Clearly, the PMLQ gained at the expense of the PMLN — the latter emerging as the biggest loser of the 2002 elections. As shown in Table 4, the creation of the King’s party, the luring away of the PMLN candidates, and the forced exile of Nawaz Sharif cost his party its historic vote share in the Punjab.

Moreover, the emergence of two Muslim Leagues seems to have benefited the PPP and the smaller political parties: on a number of seats a vote split between the two Muslim Leagues worked to the advantage of other parties, especially the PPP. This explains why the combined seat share of the two Muslim Leagues is significantly less than the 1988 to 1997 average seat share of the PMLN (see Table 4).

Another interesting by-product of the Musharraf era that will have implications for the forthcoming National Assembly elections is that a third of the seats won by the PPP, a quarter of the seats won by the PMLN and a fifth of them won by the PMLQ are marginal seats (see Figure 3). And these marginal seats will determine the outcome of the forthcoming elections in the Punjab. The choice of the historically pro-PMLN and anti-PPP swing voters will be pivotal in deciding the battle for these seats.

As argued earlier, the decision of these voters depends upon the weight they assign to the power configuration between the establishment and political parties relative to incumbent’s performance. If these voters place a higher weight on the configuration of power they will favour the PMLQ over the PMLN. But if the incumbent’s performance matters in their opinion, the result will depend on the measure of performance used. In this case, the verdict on PMLQ will depend on whether national growth matters more to these voters than food inflation, lack of electricity and gas, increasing inequality, constitutionalism and a dislike for US intervention.

If the latter issues matter more, then we will see a swing in favour of the PMLN, away from the PMLQ. However, to what extent will this benefit the PMLN or PPP depends on the magnitude of the swing: if it is small and the vote remains divided between the PMLQ and the PMLN the PPP will benefit, and if the swing is large the PMLN will experience a resurrection. And as the assassination of Bhutto is likely to result in sympathy vote in favour of the PPP, the PMLN will require a fairly large vote swing to increase its share of seats. In either event the coming elections will end up revealing more about the political proclivities of the voter that is likely to make or break politics in the Punjab.