On Friday, Libyans were supposed to do something they had never done before.
Directly elect a president.
The date itself was symbolic. 24 December marked seventy years since Libya declared independence in 1951. After Muammar Gaddafi’s dictatorship, the uprising that overthrew him in 2011, a decade of political division and years of civil war, the election was supposed to provide Libya with something it has struggled to establish ever since the revolution:
one government with a broadly accepted claim to rule the entire country.
Instead, there was no election.
Two days before voting was supposed to begin, Libya’s electoral commission said the poll could not proceed and proposed postponing it by one month. The decision was not particularly surprising. For weeks, the electoral process had been becoming increasingly impossible to organise. Libya still did not even have a final agreed list of presidential candidates.
At first, postponing an election sounds obviously anti-democratic. Elections are how citizens choose governments. If politicians cannot agree, let the voters decide.
But Libya demonstrates why that principle contains an assumption we rarely notice.
For an election to resolve a political conflict, the people competing in it have to agree beforehand on what the election means.
They need to agree on who can stand, how votes are counted, which institutions resolve disputes, what powers the winner receives and, most importantly, that whoever loses will accept the result.
Libya had agreement on almost none of those things.
That makes Friday’s non-election more than an administrative failure. It exposes something fundamental about democracy.
Ballot papers cannot substitute for political rules.
Libya’s current crisis begins with the political order that followed Gaddafi’s overthrow in 2011. Gaddafi had governed for forty-two years without constructing the kind of institutions capable of easily surviving him. There was no established democratic party system waiting to take over, no normal constitutional mechanism for transferring presidential authority and no deeply institutionalised national military capable of monopolising force.
The revolution removed the person at the centre of the system faster than it could construct institutions to replace him.
The result was not immediate democracy.
It was competition over who possessed legitimate authority.
By 2014, Libya had effectively split between rival political and military centres. Different governments claimed legitimacy in the east and west. Armed groups controlled territory. General Khalifa Haftar developed a powerful military organisation in eastern Libya and eventually launched an offensive aimed at capturing Tripoli in 2019.
That offensive failed in 2020, and a ceasefire created the possibility of another political process. A UN-backed arrangement then produced the Government of National Unity under interim Prime Minister Abdul Hamid Dbeibah. Its central purpose was supposed to be temporary: reunify institutions sufficiently to take Libya towards elections.
December 24 was meant to complete that transition.
Instead, the transition has reached the election and discovered that the election itself contains almost every unresolved conflict of the previous decade.
Consider the candidates.
One is Khalifa Haftar, the eastern military commander whose forces spent more than a year attempting to capture Tripoli. To many supporters, Haftar represents order, secular authority and the possibility of building a state strong enough to control militias.
To many people in western Libya, he represents the military force that attacked their city.
Another candidate is Saif al-Islam Gaddafi, Muammar Gaddafi’s son.
For some Libyans exhausted by a decade of instability after 2011, the Gaddafi name may represent a period when the state at least appeared unified and functioning.
For people who participated in the revolution against his father, his candidacy represents something completely different.
Saif al-Islam is also wanted by the International Criminal Court over alleged crimes against humanity connected to the suppression of the 2011 uprising.
Then there is Abdul Hamid Dbeibah, the interim prime minister whose government was supposed to guide Libya towards elections.
When he entered that role, Dbeibah had pledged not to stand in the presidential election.
He stood anyway.
His opponents argue that remaining prime minister gave him access to state resources and visibility that other candidates could not match.
Each of these candidacies therefore raises a legal question.
It also raises a much larger political question.
If Haftar wins, will armed groups and communities in western Libya accept him as president?
If Saif Gaddafi wins, will people who fought his father’s regime accept the return of the Gaddafi family through an election?
If Dbeibah wins after breaking his commitment not to stand, will his opponents consider the process legitimate?
An election does not automatically solve these questions.
It may simply attach a number to them.
That is the distinction between electoral legitimacy and political legitimacy.
Suppose a candidate receives 52% of the vote in a perfectly counted election.
Mathematically, the result is clear.
But if the losing 48% controls armed groups, territory and institutions and believes the winner was never legally entitled to stand, the number itself cannot force them to accept the outcome.
Votes determine political authority only when enough people already agree that votes should determine political authority.
That sounds circular.
It partly is.
Democracy works because political actors expect everyone else to continue following democratic rules after losing.
The system therefore depends on an equilibrium of expectations.
I accept defeat because I expect you to accept defeat when you lose next time.
I allow you to govern because I expect another election will eventually give me the opportunity to replace you.
I do not use violence because I expect you will also refrain from violence.
Once those expectations become credible, peaceful transfers of power become normal.
Libya has not yet reached that equilibrium.
Too many political actors retain alternatives to accepting defeat.
Most importantly, they retain guns.
This is why holding elections in a country with competing armed organisations is so much more difficult than simply printing ballots and opening polling stations.
An election concentrates political stakes into one moment.
Before the election, several factions may share influence imperfectly.
After a presidential election, one individual is supposed to possess a national mandate.
For the winner, that is valuable.
For powerful losers, it can be threatening.
If losing means surrendering political influence, military control, access to state resources or protection from prosecution, accepting defeat becomes extremely expensive.
The more expensive losing becomes, the more difficult democratic competition becomes.
This is one reason strong democracies deliberately prevent election results from becoming existential.
The losing party loses office.
It does not normally lose its right to organise, its property, its physical safety or the possibility of competing again.
That lowers the cost of defeat.
Libya’s political system cannot yet guarantee all of those things.
Which means asking factions to accept an election result may involve asking them to accept risks much larger than politicians face in consolidated democracies.
This does not justify refusing elections forever.
It explains why institutional safeguards have to accompany them.
The electoral law itself has become part of the dispute. Libya’s political institutions failed to reach broad consensus over the rules governing the presidential election. There were disagreements over eligibility requirements, sequencing and the powers of the future president.
That last issue is particularly important.
Libyans were being asked to elect a president without complete agreement over what that president would actually be able to do.
Imagine voting in an election before deciding whether the winner will become a largely ceremonial figure or an extremely powerful executive.
The same vote has completely different political consequences depending on the answer.
Rules therefore need to precede competition.
Otherwise every actor has an incentive to design the rules according to whether they expect to win.
That is why constitutions are ideally written behind something resembling a veil of uncertainty: political actors should not know exactly who will benefit from each provision.
If I know my candidate will become president, I suddenly discover excellent arguments for strong presidential powers.
If I expect my opponent to win, parliamentary checks become extremely attractive.
Institutional design becomes credible when political actors are willing to accept the rules even when those rules later benefit somebody else.
Libya is trying to negotiate institutional rules while everyone already knows which political figures are competing for power.
That makes neutral design extraordinarily difficult.
Foreign governments have also placed enormous emphasis on elections.
The United Nations, United States and European governments have treated the December vote as a central mechanism for stabilising Libya.
Their logic is understandable. Libya needs legitimate national institutions, and elections are the most democratic mechanism available for creating them.
But there is a danger in treating elections as a technological solution to a political problem.
International organisations can provide ballot boxes.
They can train election officials.
They can monitor polling stations.
They can help construct voter registers.
What they cannot manufacture is agreement among Libyan factions that the result deserves obedience.
That has to be political.
This resembles the mistake international actors sometimes make in state-building more broadly. Because institutions have visible forms, outsiders reproduce the form and assume the underlying function will follow.
Build a parliament and parliamentary politics will emerge.
Hold an election and democratic legitimacy will emerge.
Train an army and state authority will emerge.
Afghanistan should have made us extremely cautious about that assumption.
Institutions are not only buildings, laws and organisations.
They are expectations about behaviour.
A court possesses authority because people expect its decisions to be obeyed.
An election possesses authority because people expect losers to leave office.
A constitution possesses authority because political actors expect everyone else to treat it as binding even when inconvenient.
Without those expectations, the formal institution can exist while its political function does not.
Libya also demonstrates why postponement is not necessarily the opposite of democracy.
There are circumstances in which holding a badly designed election can be more dangerous than delaying it.
That statement makes me uncomfortable because authoritarian governments constantly use similar arguments.
The country isn’t ready.
Security conditions are difficult.
Institutions need reform.
Elections must wait.
Those excuses can postpone democracy indefinitely.
So postponement needs its own constraints.
There should be a clear reason for delay, a credible process for resolving that reason and a new timetable that political actors cannot endlessly manipulate.
Otherwise temporary postponement becomes another mechanism for incumbents to remain in power.
Libya’s electoral commission has proposed 24 January 2022 as a new date. Whether one month can resolve disputes that accumulated over ten years is difficult to believe.
The danger is that postponement now creates another legitimacy crisis.
Dbeibah’s government was supposed to be temporary.
Its mandate was connected to the electoral process.
If elections do not occur, how long does the interim government remain legitimate?
Political opponents may argue that its authority has expired.
The government may argue that somebody has to govern until elections happen.
Both positions can create competing claims to legality.
Again, temporary political arrangements become dangerous when their endpoint disappears.
This is particularly serious in Libya because the country has already experienced rival governments claiming legitimacy simultaneously.
The election was supposed to resolve that problem.
Its failure could reproduce it.
There is also oil.
Libya possesses Africa’s largest proven oil reserves, and petroleum dominates government revenue and exports.
That means control over national institutions is connected directly to control over enormous economic resources.
Political competition is therefore not merely ideological.
The state is the mechanism through which oil wealth is distributed.
That increases the value of controlling it.
Resource wealth can make political compromise harder because the prize of national power becomes larger.
If government principally collected taxes from citizens, rulers would need a functioning relationship with those citizens to finance the state.
Oil changes that relationship.
A government can receive enormous revenue from exporting a resource rather than from taxing a broad domestic economy.
This can weaken the normal fiscal bargain between citizens and rulers: we pay taxes, therefore we demand representation and services.
It can also make control of oil facilities and institutions a source of political leverage for armed groups.
Again, resources are not destiny.
Norway has oil and strong democratic institutions.
The question is whether institutions are strong enough to prevent resource wealth from becoming a prize over which political actors fight.
Libya’s are not yet.
Foreign powers complicate the picture further. Turkey, Russia, Egypt, the United Arab Emirates and European states have supported different actors or pursued different interests during Libya’s conflict.
That means domestic factions do not operate only with domestic resources.
External support can make compromise less necessary.
If a political actor believes a foreign patron will provide money, weapons or diplomatic protection, the cost of refusing a domestic settlement falls.
This is another reason civil wars become so difficult to end once they internationalise.
The conflict may occur inside one country’s borders.
The bargaining table effectively extends far beyond them.
Libyans therefore face a paradox similar to several countries I have written about this year.
Everyone says Libya needs one sovereign national government.
Multiple foreign governments simultaneously support relationships and interests that make unified sovereignty harder to construct.
The international community wants stability.
Individual states want influence.
Those objectives do not always align.
So was postponing Friday’s election the correct decision?
Given the absence of an agreed final candidate list, disputes over electoral law and genuine danger that powerful factions might reject the outcome, I think proceeding simply because 24 December had been printed on calendars would have been irresponsible.
A deadline is valuable because it forces political actors to make decisions.
It is not valuable enough to justify an election whose rules are unresolved.
But postponement is defensible only if the additional time is actually used to solve the problems that made the election impossible.
That means agreement on candidate eligibility, clearer constitutional rules, credible dispute-resolution mechanisms and enough security guarantees that losing an election does not become an existential threat.
One month may not be enough.
The alternative cannot be indefinite transition either.
This is the difficult middle ground democracy repeatedly occupies.
Hold elections too early and the competition can destabilise institutions that are not strong enough to absorb defeat.
Wait for perfect institutions and elections may never occur.
There is no objective point at which a country becomes “ready for democracy”.
The better question is whether enough political actors believe they have more to gain from following electoral rules than abandoning them.
That is ultimately what Libya needs to construct.
Not simply voting.
Losers.
More specifically, political actors willing to lose.
Democracy depends on winners accepting limits, but perhaps it depends even more fundamentally on losers accepting temporary defeat.
A candidate has to believe that losing today does not mean disappearing permanently.
An armed faction has to believe that surrendering political control does not mean becoming defenceless.
A region has to believe that a president from another region will still govern it fairly.
A political movement has to believe it can contest the next election.
When those expectations exist, ballots can replace bullets.
Without them, ballots may simply determine who reaches for the bullets afterwards.
Libya was supposed to vote on Friday.
Millions of Libyans had registered, and many undoubtedly wanted the opportunity.
The failure to hold that election is disappointing.
But the deeper democratic failure occurred earlier.
Libya reached election day without agreeing on what everyone would be required to do the morning after.
That may be the most important rule of democracy.
The election determines who wins.
The political system determines whether losing is survivable.
Until Libya can guarantee the second, the first may remain dangerously difficult.
