On Monday, millions of Ethiopians went to the polls.
For Prime Minister Abiy Ahmed, this election is supposed to represent something extremely important: democratic legitimacy.
Abiy was not originally elected prime minister in a national election. He came to power in 2018 after years of anti-government protests forced a political transition within Ethiopia’s ruling coalition. He promised reform, released political prisoners, welcomed opposition groups back into the country and ended a decades-long state of hostility with neighbouring Eritrea.
In 2019, he received the Nobel Peace Prize.
Three years after taking office, Ethiopians have finally had the opportunity to vote in the first national election under his leadership.
At first glance, this looks like democratic progress.
But there is a problem.
Not everybody was able to vote.
The entire region of Tigray, where Ethiopia’s government has been fighting a brutal war since November, did not participate. Voting was also postponed in a number of other constituencies because of insecurity and logistical problems.
Several important opposition figures are in prison.
Some opposition parties boycotted the election.
And the country holding this democratic exercise is simultaneously facing one of the world’s most serious humanitarian crises.
So what exactly are we measuring when we call an election democratic?
Because increasingly, I don’t think the existence of a ballot box is enough.
Abiy’s extraordinary political journey
It is worth remembering how different Ethiopia looked only three years ago.
Abiy came to power in 2018 after widespread protests against the Ethiopian People’s Revolutionary Democratic Front, the coalition that had governed the country for almost three decades.
His arrival initially seemed transformative.
Political prisoners were released.
Exiled opposition movements were allowed to return.
Restrictions on political activity were relaxed.
And, most dramatically, Abiy reached a peace agreement with Eritrea, formally ending a state of hostility that had existed since the two countries fought a devastating border war between 1998 and 2000.
That achievement helped earn him the Nobel Peace Prize.
If this week’s election had happened in that political atmosphere, it might have looked like the culmination of Ethiopia’s democratic transition.
Instead, the country voting in 2021 looks profoundly different.
The question is how it changed so quickly.
Ethiopia’s political problem is older than Abiy
Ethiopia is not simply one political community divided between government and opposition.
It is a federation organised substantially around ethnicity.
The country contains more than 110 million people belonging to numerous ethnic and linguistic groups. Regional states possess significant formal powers, and ethnicity has been deeply embedded within the country’s political institutions.
For decades, Ethiopia was governed by the EPRDF, a coalition of ethnically based parties.
Within that coalition, the Tigray People’s Liberation Front, or TPLF, exercised influence far greater than Tigray’s relatively small share of Ethiopia’s population.
That created resentment.
When Abiy came to power, he attempted to reorganise this political system. In 2019, he merged several components of the old governing coalition into a new national organisation called the Prosperity Party.
The TPLF refused to join.
That disagreement was not simply about the name of a political party.
It reflected two competing ideas about Ethiopia.
Abiy increasingly emphasised national political unity.
The TPLF defended a system in which Ethiopia’s ethnically organised regions retained substantial autonomy.
That raises a political problem with no particularly easy solution.
Too much centralisation can make minorities fear domination.
Too much decentralisation can weaken the common state.
A diverse country somehow has to create institutions strong enough to hold everyone together without becoming so strong that groups feel they are being controlled from the centre.
Ethiopia has not solved that problem.
Then political disagreement became war
The confrontation between Abiy’s government and the TPLF became much more serious last year.
Ethiopia’s national election was originally supposed to take place in 2020 but was postponed because of COVID-19.
Tigray’s regional government rejected that decision and organised its own election anyway.
The federal government declared the vote illegal.
Relations deteriorated further.
Then, in November, war began.
Abiy’s government said TPLF forces attacked a federal military command in Tigray. Federal troops entered the region, supported by forces from neighbouring Eritrea and fighters from Ethiopia’s Amhara region.
By the end of November, federal forces had captured Mekelle, Tigray’s regional capital.
At the time, the government presented that as effectively ending the major military operation.
It clearly did not.
Fighting has continued.
Thousands of people have been killed.
Millions have been displaced or require humanitarian assistance.
There have been horrifying allegations of massacres, sexual violence and other abuses involving multiple forces operating in the conflict.
And now, while that war continues, the rest of Ethiopia has held an election.
Can an election be national when an entire region is missing?
This is what I find most difficult about Monday’s vote.
There may have been perfectly practical reasons why voting could not occur in Tigray.
It is extremely difficult to conduct an election in the middle of a war.
But that does not make Tigray’s absence politically insignificant.
Imagine trying to determine the democratic will of a country while millions of people living in one of its regions cannot participate.
The election may still tell us something meaningful about political preferences in the places where voting occurred.
It cannot tell us exactly the same thing as an election in which the entire country participated.
That distinction matters because governments do not simply use elections to decide who holds office.
They use them to claim legitimacy.
After an election, a leader can say:
The people chose us.
But which people?
How many had realistic alternatives?
Were opposition parties capable of organising?
Could voters criticise the government without fear?
Could journalists report freely?
Could everyone participate?
Those questions seem much more important than whether voting technically occurred.
Democracy requires uncertainty
I think there may be a useful way of thinking about this.
A genuinely competitive election requires uncertainty.
The government has to possess some possibility of losing.
That does not mean every election must be close. A popular government can legitimately win overwhelmingly.
But opposition parties must at least have a realistic ability to organise, campaign and persuade voters.
Otherwise, an election becomes less a mechanism for choosing a government and more a mechanism through which an existing government demonstrates support for itself.
Several major opposition organisations have raised serious complaints about the conditions surrounding this election.
Some have boycotted.
Prominent Oromo political figures remain imprisoned.
The government argues that individuals facing criminal charges are being prosecuted for legitimate reasons rather than because they oppose Abiy.
Perhaps some cases genuinely involve criminal conduct.
But this is exactly why institutions matter.
When the government competing in an election is also responsible for imprisoning important competitors, voters need extraordinary confidence that courts, prosecutors and security forces are acting independently.
Without that trust, even legally justified actions can become politically destabilising.
Democracy depends not only on rules.
It depends on people believing those rules are not being selectively applied.
Elections can legitimise governments. They cannot legitimise everything governments do.
There is another distinction that seems increasingly important.
Suppose Abiy’s Prosperity Party wins an overwhelming majority when the final results are announced.
That would give his government electoral legitimacy.
It would not automatically make every government policy legitimate.
Winning an election does not make allegations of atrocities in Tigray disappear.
It does not mean restrictions on opposition become acceptable.
It does not prove that the government’s approach to Ethiopia’s ethnic divisions is correct.
This is something democracies themselves sometimes forget.
Winning an election gives a government authority to govern.
It does not give it an unlimited mandate.
A majority cannot vote away every minority right.
Otherwise democracy becomes simply the rule of whoever can assemble the largest number.
Institutions, constitutions and rights exist partly because some things should remain protected even when they are unpopular.
The Nobel Peace Prize makes this story feel especially strange
It is difficult to ignore the contrast.
In 2019, Abiy stood in Oslo receiving the Nobel Peace Prize.
Less than two years later, his government is fighting a war inside Ethiopia.
It would be tempting to turn this into a simple story about hypocrisy.
Peace-prize winner becomes wartime leader.
But I don’t think that explanation tells us very much.
People change.
Circumstances change.
Political incentives change.
And perhaps international recognition sometimes encourages us to turn complicated political leaders into symbols before we actually understand them.
Abiy’s agreement with Eritrea was significant.
That achievement did not guarantee that every future decision he made would promote peace.
A Nobel Prize recognises an action.
It cannot certify someone’s character forever.
The same should probably apply to political leaders more generally.
We like categories.
Reformer.
Autocrat.
Peacemaker.
Nationalist.
Democrat.
But politicians can be several of those things at different moments, sometimes simultaneously.
Maybe the more useful question is not “What kind of leader is Abiy Ahmed?”
It is “What incentives and institutions constrain what Abiy Ahmed can do?”
That produces a much less satisfying story.
It may produce a more accurate one.
There is also a dangerous argument hiding inside the crisis
Governments facing rebellions often make a powerful claim:
National unity must come before political freedom.
The argument is understandable.
If a country is fragmenting, allowing every regional movement unlimited political freedom can appear dangerous. Governments therefore justify stronger central authority as necessary to prevent chaos.
But this creates a trap.
Political repression can itself strengthen the grievances that produce separatism or rebellion.
A government centralises because it fears fragmentation.
Groups fear centralisation, so they resist.
The government interprets resistance as proof that stronger centralisation was necessary.
Resistance grows.
Each side’s behaviour becomes evidence for the other’s fears.
That kind of cycle can become extraordinarily difficult to break.
And Ethiopia’s ethnic federal system makes the danger particularly serious.
The country needs enough national unity to function.
It also needs enough regional autonomy that its different communities believe remaining inside Ethiopia protects rather than threatens them.
Military victory cannot permanently answer that constitutional question.
What should we actually look for after this election?
Not simply who wins.
Abiy’s Prosperity Party is widely expected to perform extremely strongly.
The more important questions come afterwards.
Will opposition parties accept the process?
Will postponed constituencies eventually vote?
Will political space expand or contract after Abiy receives his electoral mandate?
Can the government find a political solution to the Tigray conflict?
Can humanitarian organisations reach people who need assistance?
And, perhaps most importantly, does electoral victory make the government more willing to compromise because it feels secure — or less willing because it believes the election has vindicated its approach?
That will tell us much more about Ethiopia’s democratic direction than the size of the eventual parliamentary majority.
Democracy is a process, not an event
I think I previously understood democratisation too much through visible moments.
A dictator falls.
Political prisoners are released.
An election is announced.
People queue outside polling stations.
Power changes hands.
Those moments matter.
But democracy appears to depend much more on what happens between them.
Can an opposition newspaper criticise the prime minister on an ordinary Tuesday?
Can somebody organise a political party without being arrested?
Can a court rule against the government?
Can a region disagree with the centre without the disagreement becoming military?
Can a government lose?
And if it does lose, will it leave?
Those questions are much less photogenic than an election.
They may be much better measures of democracy.
Millions of Ethiopians voting this week is significant, particularly in a country whose previous elections have often lacked genuine competition.
It should not be dismissed.
But nor should the existence of ballot boxes prevent us from noticing the country surrounding them.
An election can choose a government.
It cannot, by itself, create a democracy.
For that, the people who lose must possess rights too.
