TO BE a democracy takes more than free elections. But no democracy can thrive without them. In some places votes are travesties, with incumbents sweeping the board; in others, free elections are entrenched. It is places in between—where multiparty elections are relatively new, the result is uncertain and the incumbents’ willingness to accept defeat cannot be presumed—where there is most to play for.
Hope is strongest in various of sub-Saharan Africa’s 48 countries. Not until 1991, in Benin and then Zambia, did the region see peaceful ejections of incumbent rulers at the ballot box (the long-democratic island of Mauritius excepted). Africa has now had decent transitions via elections at least 45 times. Plenty have slid back: in Zambia last year Edgar Lungu, the incumbent, passed the winning 50% mark by a suspiciously thin margin of 0.35%. His challenger, Hakainde Hichilema, was later jailed after his car convoy failed to give way to the president’s. But note recent big successes. Elections in Nigeria in 2015 and Ghana last year saw incumbents fall. In January a Gambian dictator had to accept the voters’ will. In June Lesotho’s prime minister more graciously bowed out.
Next up is Kenya, on August 8th. It is the commercial, diplomatic and strategic hub of east Africa, yet its post-colonial multiparty elections, held only since 1992, have been fraught. Post-election violence in 2007 left at least 1,300 dead and 700,000 displaced. That poll and the following one, in 2013, are widely thought to have been flawed. This time both Kenyans and foreigners are trying to ensure a fair contest. Procedures for ensuring cleaner elections, some using improved technology, will be on trial. If they work well, hope for fragile young democracies everywhere will be boosted. Failure, meanwhile, would be felt as a blow in Kenya and beyond.
This year’s nail-biter
According to Ezra Chiloba, CEO of Kenya’s Independent Electoral and Boundaries Commission, preparations are far better this time than last. A subsidiary of Safran, a French firm best known for aerospace technology, has delivered 45,000 tablets to check biometric voter identification at the 40,833 polling stations and to prevent multiple voting. Around 360,000 officials have been hired and trained to staff them and oversee the count. The voters’ register of 19.6m has been vetted by KPMG, an international auditor. No one claims it is perfect: births and deaths often go unrecorded in Kenya’s remote places. “But if the voter ID works it doesn’t matter how bad the voters’ roll is,” says Don Bisson of the Carter Centre, which is monitoring the elections. “Dead people don’t have voter biometrics,” says an official of the commission.
To prove their identities, voters must press thumb or finger on a tablet (shown here). Up come matching names and photographs. Officials in the polling stations will adjudicate in case of glitches. Votes are cast on printed ballot papers, once an identity is confirmed. The presidential result must be announced within seven days. If no one wins more than 50%, a run-off must be held within a month. (In 2013 suspicions rose when Uhuru Kenyatta squeaked past that mark by a mere few thousand votes, though he probably did genuinely win the first round.)
Last time half of the much clunkier devices in use failed to work on the day. Within a few hours most of their batteries had run out; this time polling stations will have spares. Kenya’s mobile-network providers are co-operating; 3G covers only 78% of the country’s territory but 98% of the stations are in range, says Mr Chiloba, and satellite phones may serve the few that are not.
By July 2nd about half of the 120m ballot papers for the six sets of elections (including for governors of 47 counties and for women’s special representation) had been printed, though a last-minute snag has arisen. On July 7th the high court accepted the main opposition party’s bid to nullify the tender for printing the ballots for the presidency amid accusations that the printing contract had been improperly awarded. If the commission’s appeal fails, a new printer will be needed in a hurry.
Each party and candidate will be entitled to put agents in polling stations to oversee the count, which will be transmitted electronically and also manually to one of 290 constituency stations. The supreme court has decided that, once the result has been declared there, it cannot be changed at the counting headquarters. In Kenya and elsewhere, much fiddling has happened centrally. So this ruling is hugely positive, says a leading observer. (By contrast in Zimbabwe in 2008, when Robert Mugabe lost the first round of a presidential election, his election commission in the capital sat on the ballots for weeks before declaring that the challenger had narrowly missed the 50% mark that would have given him outright victory. Such lethal violence followed that he withdrew.)
Another vital safeguard is “parallel vote tabulation” (PVT), whereby party agents and independent observers can witness the count in randomly selected polling stations and announce each result, which will be agreed upon, photographed by smartphone and transmitted. Elsewhere, and in Kenya in 2013, PVT has been very close to the final result (see chart), making it far harder for an incumbent to inflate his tally, at least by a large amount. “PVT is a highly effective check on the electoral commissions,” says an expert from the International Foundation for Electoral Systems (IFES). It is thought to have been vital for ensuring fairness in Ghana and Nigeria. Kenya’s main opposition leader, Raila Odinga, says he will put five agents in each of the polling stations. Even just one in each would be a boon.
The independence of the electoral commission and the integrity of the supreme court as arbiter of disputes are crucial for a decent election. Monitors have become a powerful force, too. In Kenya the most serious are from the European Union, the Carter Centre and the National Democratic Institute, an American organisation. The Commonwealth and African Union are also sending teams. Among the heavyweights lending extra credibility are Thabo Mbeki, South Africa’s ex-president, for the AU; John Mahama, Ghana’s recently defeated president, for the Commonwealth; and John Kerry, America’s former secretary of state, for the Carter Centre.
Just as vital is that local citizens play a part in oversight. In Kenya the Elections Observation Group, an umbrella of 19 independent outfits, including church and human-rights groups, is set to send 6,000 watchers into the polling stations, compared with a few hundred foreign ones.
And yet…
If this all sounds too good to be true, it may be. Nic Cheeseman of Birmingham University, an expert on elections in Africa, has warned against what he calls the “fetishisation” of technology. “In some cases the complexity of digital processes may actually render elections more opaque and vulnerable to manipulation—or at least the suspicion of manipulation,” he has written. Election machines in Ghana in 2012 failed more often where no observers were present, suggesting tampering with the intention of forcing a fallback onto the more easily fiddled manual system. Technology can even facilitate fraud. In Azerbaijan in 2013, the election commission accidentally jumped the gun by releasing an electronically verified result—a day before the vote.
“You can’t digitise integrity,” says John Githongo, a veteran Kenyan anti-corruption campaigner, implying that the corrupt politicians who still dominate the country’s politics will not let technology get in the way of fiddling the result if it goes against them. “The manual count is definitive,” says one foreign observer. “The electronic one is a backup”—and should not be considered a fail-safe. Yet paper ballots are always liable to be lost, stuffed or falsified.
The register, too, can be manipulated, for example by signing up under-age people in areas where the government is popular or by making it harder for officials to register voters where opposition is strongest. KPMG has recently expressed anxiety about loopholes. In countries where trust in authority is low, and fear is high, voters may even think that technology will let the government know how they voted.
Above all, technology cannot prevent some pervasive forms of election-rigging. Incumbents in Africa won 88% of direct presidential elections since multiparty elections became common a generation ago until 2010, says Mr Cheeseman. That was partly by using state resources to outspend the opposition, often commandeering the civil service and sometimes the army. In Uganda the perennial challenger, Kizza Besigye, has repeatedly been arrested during campaigns. Incumbents often ensure biased media coverage. Technology cannot stop vote-buying or bribery.
“I don’t think technology will everguaranteecredible elections,” says one of the world’s most experienced monitors, who does not wish to be named. “The best it can do is increase the transparency and accountability of the data. By exposing the data to broader scrutiny there is some hope of creating broader acceptance of close outcomes.” PVT, for instance, depends on reliable technology.
Dispatching party agents to every Kenyan polling station will be hard. “You won’t find many Luos [who overwhelmingly back Mr Odinga’s coalition] wanting to be sent as agents to polling stations in the heartland of the Kikuyus [where their leader, Mr Kenyatta, will prevail]—or vice versa,” says a white farmer. “They’d be chased out or murdered.” It is unlikely that Mr Odinga will be able to put an agent in all 41,000 polling stations, let alone five in each. Nor are foreigners likely to observe polling stations in parts of the north-east, where Somali terrorism may be a threat.
Mr Odinga’s campaign has made much of accusations of unfairness, sighs a Western ambassador. It is widely believed that Mr Odinga was robbed of victory in 2007, and that in 2013 he genuinely trailed in the first round but probably not by so much that Mr Kenyatta truly won outright. This time at virtually every step he has accused the authorities and commission of bias against him, implying, for instance, that the printers are likely to print extra papers to aid ballot-stuffing, or that the returning officers are likely to be government stooges.
Good losers required
In the end, an unrigged election requires the protagonists’ goodwill and willingness to accept defeat. Mr Kenyatta may be sincere in saying he will step down if he loses. But it is widely surmised that the Kikuyu old guard would stop at nothing to keep Mr Odinga out of power. “The one thing we all hope for,” says a foreign monitor, “is that the margin of victory, one way or another, will be wide.” Alas, it may be close.
It is in transitional democracies—countries struggling to embed a tradition of fair polls—that trust and transparency are most needed. In countries where the incumbent blatantly fixes the vote, nobody bothers with the effort that is going into Kenya’s poll. If it is fair and peaceful, like Ghana’s last year, it will mark a massive advance for east African democracy. Though a large dose of scepticism is warranted, it is still a hopeful moment. Hold your breath.
This article appeared in theInternationalsection of the print edition under the headline"How to unrig an election"
