To rig or not to rig? This is perhaps no longer the question agitating people about the forthcoming elections. By all accounts, significant pre-poll rigging has already taken place. From a caretaker government stacked with PML-Q members, the role of partisan nazims and civil administrations to a less than forthcoming election commission, the odds are clearly against the opposition.

The general perception is that in spite of pre-poll rigging, the opposition has a chance of pulling off an electoral victory if rigging on the polling day is contained. This perception is grounded in the observation that the return of Nawaz Sharif and the assassination of Benazir Bhutto has created a momentum for the opposition, coupled with the hole that the Musharraf dispensation has dug for itself in the last year. An array of recent opinion polls also corroborate the popular mood.

Of the many mechanisms of rigging on polling day, one that appears to have a pattern is that of a high voter turnout in key constituencies. The mechanism is that once the voting is over, with the help of, or under pressure from the ‘angels’, the presiding officer (or his staff) stuffs the ballot boxes with votes for the favoured candidate. While the presiding officer is supposed to provide a certificate to the polling agents regarding the exact vote count, such certified copies would not be provided for all the ballot boxes at the station. Obviously, this happens at a handful of polling stations amongst the 200 plus stations for a typical national assembly constituency and in select constituencies.

With the help of the database compiled by the Dawn Election Cell, we have tried to statistically discern the pattern of such rigging by looking at the beneficiaries of high turnouts. The analysis has been done for the National Assembly seats in the Punjab for all elections from 1988 to 2002. The choice of the Punjab is simply because of the sheer number of seats in the province – 54.4 per cent of all NA seats – the electoral outcome is instrumental in the formation of the government in Islamabad. Moreover, the party (or candidates) who are favoured by the ‘angels’ have tended to be headquartered in the Punjab.

The turnout in the Punjab for national elections over the last five elections has ranged between a high of 46.7per cent in 1988 and a low of 40.5 per cent in 1997. If we take an unusually high turnout, say 50 per cent and more, is there a pattern amongst winners of these seats in terms of them belonging to establishment or anti-establishment parties or coalitions? The IJI and PML-N (until 1997) and the PML-Q as well as the National Alliance in 2002 were the pro-establishment parties, while the anti-establishment grouping is centred around the PPP traditionally but were joined by the PML-N in 2002.

Table 1 shows there were 188 seats in the Punjab and where the turnout was 50 per cent or greater, the pro-establishment parties have over time won roughly two-thirds of the seats as opposed to only one-fourth won by the anti-establishment parties. Since independents also tend to go along with the wishes of the establishment, the weight of the establishment vote in such constituencies increases further.

It can be argued, however, that the winning constituencies of the PML factions generally have a higher voter turnout than others. To further corroborate the above finding, we aggregated the results of elections where the pro-establishment parties won and those where the PPP or its coalitions were successful in forming the government in Islamabad.

The 1990, 1997 and 2002 elections are perceived as those where the role of the establishment in creating positive results was much greater than the other two.

Table 2 shows that in these elections, more than three-fourths of the seats with more than 50 per cent turnout went to the ruling PML-led groups. On the other hand, in the elections of 1988 and 1993 – won by PPP and its coalition partners – as shown in Table 3, the distribution of seats won by both parties are more evenly distributed than those when PML factions formed the government. Furthermore, of all the constituencies in the Punjab, there are only three (two in Bhakkar and one in Attock) where turnout has been consistently over 50 per cent. As such, the argument of a constant stronghold of PML factions in high turnout constituencies does not hold.

This analysis does not imply that high turnout automatically means that some form of rigging goes on, it only suggests that we can statistically corroborate some of the allegations in terms of a pattern. While transparent ballot boxes and the promise to certify the results at the polling station this time round may mitigate the recurrence of this phenomenon, intimidation and violence still loom large for this particular form of poll day rigging to take place again. Watch out for high turnout constituencies on Monday.