Poll inquiry talks are deadlocked on minor issues. Logically, PML-N should accept PTI’s demands about a comprehensive inquiry and opening of constituencies given how contested its mandate has become.
But this may require constitutional amendments allowing a commission to conduct such inquiries since the apex court has recently rejected three election petitions. This is because Article 225 designates tribunals as the primary fora for poll disputes.
The party should also agree to dissolve assemblies if massive rigging is proved, even if not carried out by PML-N, since its mandate would still be non-authentic. These issues are diverting attention from bigger problems.
The first relates to whether the Supreme Court should get involved given that three senior judges have headed the ECP since August 2013.
Also read: Justice Nasirul Mulk resigns as acting CEC
The chief justice recently eschewed an election case given this. But could even other (junior) judges unbiasedly evaluate the post-poll performance of an ECP headed by fellow senior judges? It seems a case of institutional conflict of interest.
A probe into alleged poll rigging can be problematic.
This politically charged case is so complex that the commission may fail to provide watertight justifications for its judgments, thus exposing the court to bias allegations.
A one-time commission of respected Pakistani or UN experts makes more sense, though finding such Pakistanis may be difficult while UN involvement may lead to sovereignty concerns.
Secondly, sound legal grounds must be established for the inquiry to avoid setting bad precedents. Statistics show no party submitted petitions against more than 8pc of the contested seats, thus suggesting none had rigging proof against 92pc seats.
Tribunals have to date adjudicated on 80pc of the petitions. Few appeals have been made against tribunal decisions. So, it must be shown that new evidence has emerged after the relevant deadlines passed.
The most vexing issue is about how to undertake such a vast inquiry, in a short space of time. Few countries have conducted investigations of entire elections. Afghanistan merely recounted around eight million votes, that too inconclusively.
Also read: ECP tests biometric system for voter verification
The demands here, eg, biometrically checking fingerprints, are more complex. Investigations must cover three types of rigging (vote fraud, suppression and miscounting) and determine whether its actual levels justify re-election. There are five main inquiry options, all rife with complications.
The most time-consuming option is verifying thumbprints. While opening constituencies, statistical standards must be followed which dictate random and sufficiently large samples, not PTI-style hand-picked, small samples.
Going by statistical confidence and probability levels which even most routine surveys use, some 150 constituencies must be opened just for the National Assembly.
This could take months. While further vote-level sampling could be done within constituencies to reduce time, such multi-level sampling must be done with expert input.
Complications could arise if most of the votes everywhere remain unverifiable. Automatically considering unverifiable votes as fraudulent is unreasonable.
At current percentages, that would mean millions of fraudulent votes were cast, which seems impossible.
Thus, evidence of widespread election-day voter identity fraud is needed to consider unverifiable votes fraudulent.
Physically retaking thumbprints from around 20,000 random voters and comparing them with Nadra records could remove confusion about unverifiable votes.
The second option is reviewing whether electoral irregularities’ levels documented by observers (eg Fafen) justify nullifying results. However, this will be highly judgmental since for most irregularities it is difficult to quantify the extent of rigging they caused.
Observers have not tabulated or analysed their irregularity data strategically, making such review difficult.
The third option is reviewing ECP vote tabulation forms, as Fafen suggests, though this will only cover vote miscounting, not suppression and fraud.
Fafen suggests nullifying elections in constituencies where major forms are missing. But in such cases, physical recounting of votes should be done before ordering re-elections.
The fourth option is reviewing tribunal decisions. Again, random sampling and adequate sample sizes will be needed, which could prolong this exercise.
The fifth option is interrogating returning officers, candidates, voters and other stakeholders. The issue here is the extent of admissible evidence they could provide.
Would commission findings be appealable and how many rigged constituencies would be required in each assembly to justify its dissolution: numbers that make the majority a minority?
It may be wise to leave most such decisions to the commission while recognising that it would have no magic wand to resolve all complications fully and issue widely acceptable judgments. Both parties must recognise these complexities to ensure a smooth inquiry and ready acceptance of inquiry results.
The writer is a governance specialist.
Published in Dawn, November 2nd, 2014