Did Emmerson Mnangagwa win Zimbabwe’s 30 July presidential election? Not according to the MDC Alliance and its presidential candidate Nelson Chamisa. Even before the official results were announced, which gave Mnangagwa 50.8% of the vote to Chamisa’s 44.7%, the Alliance claimed that they had polling station returns (V11s) showing a ‘resounding’ victory for Chamisa.

When the Zimbabwe Electoral Commission (ZEC) released a spreadsheet of the presidential results, granulated to polling station level, the Alliance persisted in its claim that the figures published by ZEC had been altered and didn’t match the V11s in their possession.

The true numbers on the V11s would soon be revealed in a petition to the Constitutional Court, the Alliance said. This would show that ZEC had fraudulently deflated the number of votes allocated to Chamisa and inflated Mnangagwa’s tallies, and that Chamisa had won.

Yet when the matter came before the Constitutional Court, it appeared that Chamisa had all but abandoned this claim. After the court case, several thousand V11s were posted online. A check of a random sample of the several hundred listed in the court petition proved ZEC’s numbers accurate in each instance and Chamisa’s claim false. This could be why none of these V11s were presented to the court and the claim not advanced with any enthusiasm.

Instead of pursuing this allegation, Chamisa pointed to two other numerical anomalies to try to prove his case. The first of these was that the state broadcaster, ZBC, had announced on polling day that by 5pm, about 105 000 people had voted in Mashonaland Central Province.

The MDC Alliance recruited statisticians to show that it was impossible for another 370 000 people to have voted by the time voting closed at 7pm. Therefore, the MDC claimed, the final tally of 475 000-odd was fraudulent.

This allegation required one to believe that Mashonaland Central had an incongruously low voter turnout, that ZEC had fraudulently allocated several hundred thousand votes to Mnangagwa not reflected on the V11s, or that the V11s had been altered by ZEC officers at over 1 000 polling stations – all without anyone noticing. The other possible explanation was simply that the ZBC announcement was wrong.

The MDC Alliance’s second smoking gun supposedly proving manipulation was a claim that the presidential and parliamentary tallies didn’t match. Three polls are held simultaneously in Zimbabwe’s general elections: presidential, parliamentary and local government. In the normal course of events, voters are given three ballot papers. The total votes cast in each of the three elections should match.

There were 40 000 more votes in the presidential poll than in the parliamentary, Chamisa alleged. However, apples were being compared with pears – the total votes cast in the presidential poll with total valid votes cast in the parliamentary.

Furthermore, ZEC admitted that they had inadvertently omitted (without affecting who had won) 19 000 votes in a single constituency from the parliamentary tally. The remaining difference is easily accounted for by spoiled ballots omitted from the Alliance calculations.

However, Chamisa pointed out other indisputable discrepancies in ZEC’s figures, compelling the electoral management body to revise the margin of Mnangagwa’s win downwards to a wafer-thin 31 000 votes above the absolute majority of 50% plus one vote required to avoid a run-off (50.67%). Some of ZEC’s admitted ‘errors’ were of the same order of magnitude as Mnangagwa’s margin of victory – five figures.

With these errors exposed only through Chamisa’s court action, what confidence could the public have that there weren’t other discrepancies in the tabulation which would have taken Mnangagwa’s tally below the threshold required to avoid a run-off? If a run-off had taken place, Chamisa would have looked like a possible winner, gained greater financial backing than he had in the first round, and the momentum would probably have carried him to victory.

Having admitted significant errors in the tally, ZEC should have shored up confidence in its final figures by subjecting the count to an independent audit. That it didn’t, suggests that ZEC feared what such an audit might reveal.

Although Chamisa’s claims on the numbers were easily discounted, the Constitutional Court was also reluctant to address them. It dismissed all his arguments on account of the failure to put evidence in the form of the V11s before the court – though several of the claims of manipulation didn’t rely on the returns.

The odd failure to adjudicate on the substance of Chamisa’s clearly dubious claims left the door open for the Alliance to continue to dispute Mnangagwa’s win. The Alliance strategy has been ‘to declare victory early and not to stop declaring it’. MDC legislators walked out of Parliament last Tuesday as Mnangagwa gave the official opening address, stating that they would not stay and listen to an ‘illegitimate’ leader.

In this way, Chamisa and the Alliance hope to remain relevant. Their argument is that Mnangagwa must be accepted as legitimately elected before desperately needed foreign investment will flow into the country. As the Alliance would have it, since Mnangagwa’s victory is questionable due to ZEC’s dodgy numbers, only the Alliance can confer that legitimacy. Mnangagwa, then, needs to accommodate the Alliance.

ZEC’s results spreadsheet has been subjected to intensive scrutiny by numerous MS Excel ninjas intent on securing the glory of ‘exposing the fraud’ and by academics and political analysts who predicted a Chamisa win. The latter have the uncomfortable choice of either admitting that their analysis was wrong or claiming that the numbers were fiddled.

With so many having failed, despite considerable effort, to prove fraud, Mnangagwa probably did secure just enough votes to avoid the run-off. Unfortunately, when one is considering the legitimacy of a president, as the Alliance tactics have shown, ‘probably’ is just not good enough.

Written by Derek Matyszak, Senior Researcher, Peace and Security Research, ISS