Kenyan President Uhuru Kenyatta casts his vote last month in his home town of Gatundu, Kenya. (Ben Curtis/Associated Press)

Nic Cheeseman is a professor of democracy at the University of Birmingham and the author of “Democracy in Africa.” Todd Moss is senior fellow at the Center for Global Development and a former U.S. deputy assistant secretary of state for African affairs. Jeffrey Smith is executive director of Vanguard Africa, a nonprofit group that supports free and fair elections in Africa.

After one annulled presidential election and one passionately disputed do-over, Kenya is confronting a serious crisis of democracy. On Tuesday, the country’s Supreme Court made the first of several rulings regarding the outcome of the most recent vote, which was won by incumbent President Uhuru Kenyatta. If he is sworn in as president for the second time on Nov. 28, Kenya risks another round of destabilizing protests.

Kenya’s political leaders bear the lion’s share of responsibility for this mess. Yet international election observers also let themselves down, having failed to identify the many problems with the first of those two elections — shortly before it was canceled.

Sadly, this failure is part of a trend. Too often the election cops are letting the bad guys get away. Worse, they may be abetting the theft of many nascent democracies.

Election monitors regularly fly into a country to watch voting. Ideally, they are objective witnesses to a poll, helping to deter violence and to boost the credibility of the result. Monitors are supposed to help make democracy work.

Recent elections in Africa, however, suggest the opposite. Observers keep missing — willfully or otherwise — the actual ways in which elections are actually stolen. All too often they confuse peaceful voting with a “free and fair” election. This misplaced tolerance leads them to endorse illegitimate outcomes. Rather than expose rigging, monitors are too often helping undemocratic leaders stay in power.

Kenya is but the most recent example. In August, international observers missed the many warning signs and were embarrassed when judges declared the vote invalid, citing a rash of mistakes, plus concerns that the election commission’s computers may have been hacked. (The official in charge of the vote’s technology security was found tortured and strangled only days before the contest.) The surprise cancellation sparked heavy criticism of the international monitors, including the Carter Center and the European Union, whom the opposition accused of bias and incompetence — although it later pledged to give them a second chance.

Kenya isn’t the only African country to endure elections plagued by sophisticated thievery. Zimbabwe’s 2013 election was an early master class in how to subvert the vote. Recent elections in Uganda, Gabon and Zambia were also won by incumbents under highly suspicious conditions, despite the presence of international election monitors.

These and other examples all suggest that international election observation urgently needs a rethink. There is now a greater focus on long-term observation, but election-day monitoring still follows the same old model that was deployed in the 1980s: A small number of people arrive in the country a few days before voting and then fan out to watch the vote and tallying at selected polling stations. This has some tangible benefits. Observers can prevent brazen rigging or the most blatant intimidation of voters — at least in the small number of stations where they are present.

But autocrats across Africa and elsewhere have been highly creative in finding ways to subvert the will of the people. They use subtler methods of intimidation, “legally” change election rules at the last minute and manipulate voter rolls. Increasingly, too, they are hacking the vote. While observers are blissfully watching ballot boxes, the election is being peacefully stolen in cyberspace. In Kenya, for instance, observers were poorly placed to evaluate claims of election hacking that the Supreme Court found credible.

At least two significant changes must occur for election monitors to regain their rightful role in supporting democracy. First, teams need to become active far earlier in the process. They must make early assessments of the rules, identification requirements and changes to electoral registers well in advance of the vote —  and state their findings publicly. The safety of opposition candidates, their ability to campaign freely and their access to state media are also crucial to ensuring a truly free and fair vote. Too often monitors overlook these baseline requirements for the sake of ensuring “stability” in the short term. Such laxity must end. Observers must be ready to speak up loudly about these concerns in time to adequately fix them, or be prepared to openly question the vote’s overall legitimacy.

Second, like modern-day police now chasing cyber-thieves instead of masked bank robbers, international observers must adapt to new technologies. They have to shift from people-watching to data analysis, tracking digital processes, using statistical tests to identify potential abnormalities, and conducting forensic audits on the final results.

Adopting these more nuanced approaches not only will help to ensure credible and broadly acceptable elections but also will provide much-needed life support to sustain the faith that citizens have expressed in democratic processes across Africa.

In many important ways, the future of democracy is in the hands of international election monitors. And they must accept the need for dramatic reform. Otherwise they will only continue to be part of a growing problem.