What Happens When Voters, Courts, Armies and Constitutions Disagree About Democracy?
Something unusual is happening to elections around the world. In April, Hungarian voters ended Viktor Orbán’s 16-year rule. Yesterday, a student-backed movement defeated Serbia’s long-dominant political establishment, while Brazilian voters returned Luiz Inácio Lula da Silva to office in a closely fought presidential runoff. Tomorrow, Israelis vote in an election that could determine Benjamin Netanyahu’s political future, and one week later Americans decide control of Congress.
These elections do not point in one ideological direction. The center-right defeated Orbán in Hungary, the left retained Brazil’s presidency, Germany’s far-right Alternative for Germany nearly doubled its support in 2025, and the United States may soon produce divided government. What they reveal instead is something more important: the ballot box remains extraordinarily powerful, but it is not always the final authority in a democracy.
That is the contradiction worth examining. We often speak as though democracy means simply allowing the people to decide, yet every constitutional democracy places limits around what elected governments, parliamentary majorities and even voters themselves can do.
An Election Is an Event. Democracy Is a System.
An election answers a relatively narrow question: who received enough votes to govern? Democracy asks much more difficult questions about what happens before the vote, after the vote and between elections.
Could the opposition campaign freely? Could journalists investigate the government? Were the courts independent? Were votes counted honestly? Could the winner actually assume power, and could that same winner eventually be removed?
Those distinctions matter because countries can hold elections without possessing fully functioning democratic systems. Conversely, constitutional institutions can sometimes override elected officials without necessarily overthrowing democracy, provided their authority is lawful, reviewable and ultimately leaves citizens able to choose again.
The difficulty begins when these principles collide. What happens when voters elect a government that other institutions believe threatens the constitutional system itself?
Hungary and the Corrective Power of Elections
Hungary offers one answer. Orbán governed for 16 consecutive years after returning to power in 2010, during which critics in Hungary and the European Union accused his government of weakening judicial independence, restricting media and civil-society freedoms and concentrating executive authority.
Yet in April, Hungary’s electoral system still retained enough corrective capacity to remove him. Péter Magyar’s center-right Tisza party won an extraordinary 138 of the country’s 199 parliamentary seats, more than the two-thirds threshold required for major constitutional changes. Record turnout helped end an era that many observers had come to regard as politically entrenched.
The significance goes beyond Orbán. Democratic deterioration does not always become irreversible. Institutions can weaken while elections retain enough credibility and competitiveness to permit voters to change direction.
The harder question is how much erosion a democratic system can absorb before that corrective mechanism stops functioning.
Serbia: When Protest Becomes Electoral Power
Serbia has just provided another remarkable example. After nearly two years of demonstrations that began following the collapse of a railway-station canopy in Novi Sad that killed 16 people, a student-backed movement converted street pressure into electoral organization.
With nearly all votes counted, its candidates received 52.7 percent and won 139 of Serbia’s 250 parliamentary seats. Aleksandar Vučić’s Progressive Party-led alliance received 39.2 percent and 97 seats, while turnout approached 70 percent, the highest since the breakup of Yugoslavia.
This is democracy performing one of its most important functions: converting political dissatisfaction into peaceful transfer of power. Yet Serbia also exposes the next democratic problem, because winning an election is different from governing a country.
A protest movement can unite around what it opposes. Government requires budgets, compromises, appointments, institutional discipline and decisions that inevitably disappoint somebody.
Brazil and the Democracy of Losing
Brazil provides yet another variation. Lula won Sunday’s runoff with 51.79 percent of valid votes, compared with Flávio Bolsonaro’s 48.21 percent, a difference of roughly 4.43 million votes among more than 124 million valid ballots.
The election therefore produced a clear winner but also a country divided almost evenly between competing political visions. Bolsonaro has said he will not concede and intends to seek annulment of the runoff over a Supreme Court decision extending voting hours.
That brings us to a democratic principle that receives far less attention than voting: loser’s consent. Elections cannot stabilize political systems unless defeated candidates and their supporters generally accept that losing today does not eliminate their right to compete tomorrow.
Democracy therefore depends not only upon winners accepting limits. It also depends upon losers accepting temporary defeat.
Germany and the Voter We Do Not Understand
Germany presents a different challenge. In the 2025 federal election, the Alternative for Germany received 20.8 percent of the second vote and 152 Bundestag seats, nearly doubling its 10.4 percent share from 2021. Turnout reached 82.5 percent.
The easy response is to dismiss voters supporting parties outside the political mainstream as irrational, angry or manipulated. That is also an intellectually dangerous response because democratic systems cannot praise the wisdom of voters only when voters produce outcomes established institutions prefer.
Citizens vote for complicated combinations of economic interest, immigration policy, cultural identity, insecurity, distrust, taxation, ideology and dissatisfaction with governing elites. Democracy has never depended upon every citizen reaching the same conclusion about those questions.
The democratic test is not whether voters always choose wisely. It is whether institutions remain democratic after voters choose.
Israel: When Winning Still May Not Be Enough
Israel votes on October 27 in an election widely regarded as a test of Netanyahu’s continued leadership. Polls indicate that his Likud-led bloc and the opposition are closely matched, but neither currently has a clear route to the 61 Knesset seats necessary for a governing majority.
That makes Israel an important reminder that ballots do not translate directly into governments. Parliamentary democracy requires votes to pass through parties, coalition negotiations and legislative arithmetic before executive authority emerges.
Netanyahu could lose seats and remain politically relevant. An opposition party could perform strongly and still fail to construct a governing coalition.
This is not necessarily a defect in democracy. It demonstrates that democratic power is mediated through institutions rather than transmitted directly from a single vote to a single leader.
America and the Return of Divided Power
The United States will vote on November 3 with Republicans controlling the presidency and both houses of Congress. Democrats need a net gain of four seats in both the House and Senate, although the House currently appears more favorable terrain than the Senate.
A Reuters/Ipsos survey conducted in October put Democrats ahead 47 percent to 35 percent among registered voters on the generic congressional ballot. Republicans hold a narrow 218-seat House majority, and Reuters identifies roughly 40 competitive House races, while the Senate remains structurally more difficult for Democrats despite the national environment.
If Democrats capture one chamber while Donald Trump remains president, Washington will become more confrontational and probably less legislatively efficient. Yet constitutional friction is not necessarily constitutional failure.
America’s system was deliberately constructed to divide power. Sometimes voters use elections not to replace a president but to place another institution between the president and unrestricted political control.
Algeria: When the State Overrules the Voter
History shows that intervention can be much more dangerous. Algeria held its first multiparty parliamentary election in December 1991, and the Islamic Salvation Front, or FIS, won 47.5 percent of the first-round vote and 189 of the 231 seats decided immediately. It appeared virtually certain to secure a parliamentary majority after the second round.
The military intervened before that second round occurred. President Chadli Bendjedid was forced from office, the electoral process was halted and the FIS was eventually banned.
The justification was essentially preventive: allowing an Islamist movement to acquire democratic power might endanger the political order. The consequence was catastrophic, as Algeria descended into a brutal conflict between the state and armed Islamist groups that lasted through much of the 1990s; estimates of the death toll have reached roughly 150,000.
Algeria leaves behind a question democracies still struggle to answer: can you destroy an election because you fear what the winners might do with democracy?
Egypt: When the Army Becomes the Referee
Egypt confronted a related dilemma two decades later. Mohamed Morsi, supported by the Muslim Brotherhood, won the 2012 presidential runoff with 51.73 percent against Ahmed Shafiq’s 48.27 percent, becoming Egypt’s first democratically elected president.
His presidency quickly became deeply polarizing, and enormous demonstrations demanded his removal. In July 2013, the military under Abdel Fattah al-Sisi removed Morsi, suspended the constitution and eventually established a new political order dominated by Sisi, whose government has since been accused by human-rights groups of extensive repression.
Egypt illustrates why mass protest and electoral legitimacy cannot easily substitute for one another. Millions of people in the streets may demonstrate profound political opposition, but allowing the military to determine when popular dissatisfaction has invalidated an election creates an institution more powerful than the voter.
Once the army becomes democracy’s referee, persuading it to return permanently to the sidelines can prove extremely difficult.
Romania: Can an Election Be Cancelled to Save Democracy?
Romania presents perhaps the most intellectually difficult modern case. Its Constitutional Court annulled the 2024 presidential election after declassified intelligence alleged campaign manipulation, opaque financing, coordinated online promotion and Russian-linked interference surrounding the surprise first-round victory of far-right candidate Călin Georgescu. Romanian intelligence also reported more than 85,000 cyberattacks targeting election infrastructure.
Doing nothing carried obvious risks if foreign interference had materially distorted the contest. Cancelling the election carried another risk: unelected judges nullifying millions of votes.
There was no cost-free solution.
That is precisely why Romania matters. Democracies increasingly face threats that arrive not by tanks surrounding parliament but through information systems, social media, covert finance and manipulation of the democratic process itself.
South Korea and the Difference Between Reversing an Election and Restraining a Winner
South Korea demonstrates an important distinction. President Yoon Suk Yeol had been legitimately elected, but his December 2024 declaration of martial law triggered impeachment proceedings, and the Constitutional Court removed him from office in April 2025. The constitution then required a new presidential election within 60 days.
The institutions did not declare that voters had chosen the wrong president. They concluded that the president voters had chosen had exceeded constitutional limits.
That difference is fundamental.
Constitutional democracy does not give elected leaders unlimited power simply because they possess electoral legitimacy. Elections confer authority, while constitutions define its boundaries.
Senegal: When Institutions Protect the Ballot
Not every intervention weakens popular sovereignty. Senegal demonstrated the opposite in 2024 when President Macky Sall attempted to postpone the presidential election by ten months.
The Constitutional Council and Supreme Court repeatedly rejected efforts to extend the delay. The election proceeded, opposition candidate Bassirou Diomaye Faye prevailed, and both the outgoing president and the governing coalition’s candidate accepted the outcome.
Here, courts constrained political power precisely so voters could exercise theirs.
That is constitutional democracy operating as intended: institutions did not replace the electorate but protected the electorate’s opportunity to decide.
Five Things a Democratic Election Must Actually Do
The cases suggest that judging an election merely by whether ballots were printed and polling stations opened is inadequate. A meaningful election must allow citizens to choose, remove, replace, reconsider and try again, and all five functions matter.
Citizens must have genuine alternatives from which to choose. Incumbents must be removable, winners must actually be able to assume power, yesterday’s majority must be allowed to change its mind, and today’s losers must remain free to contest tomorrow’s election.
That last principle may be the most important.
An election that allows someone to win but makes it impossible for that person ever to lose is not the completion of democracy. It is the beginning of something else.
Who Protects Democracy From Democracy?
This is the uncomfortable question running from Algeria and Egypt through Romania and South Korea. Armies, courts, legislatures and electoral commissions sometimes claim that intervention is necessary to protect constitutional government from dangerous political choices.
Sometimes that claim has merit. Sometimes it is the vocabulary through which democracy is dismantled.
The appropriate questions therefore concern process rather than ideology. Was the institution acting under constitutional authority? Was the evidence public and credible? Was the intervention proportionate? Could the decision be reviewed, and could political opponents continue organizing afterward?
Above all, citizens must eventually be allowed to vote again.
The power to protect democracy from an elected government can itself become one of the most powerful instruments for destroying democracy.
The Morning After the Election
Election night receives most of the attention because it provides drama, numbers and winners. Yet the morning after may tell us considerably more about the health of a democratic system.
Does the loser accept defeat? Does the winner take office? Do judges remain independent, journalists remain free and opposition parties continue organizing? Will another competitive election take place, and could the new government genuinely lose it?
From Budapest to Belgrade, Brasília, Jerusalem and Washington, elections are once again demonstrating both the extraordinary resilience and the inherent fragility of democratic government. The ballot can remove entrenched leaders, strengthen parties outside the mainstream, restrain presidents and peacefully transfer power, but it can also produce choices that frighten institutions charged with protecting constitutional order.
Democracy’s greatest strength is therefore not that voters always get elections right. They do not, and neither do governments, judges or political elites. Its deeper strength is that political decisions are supposed to remain reversible, power can change hands without violence, constitutional restraints survive whoever wins, and citizens retain the right to return later and change their minds.
That is why the ballot is indispensable but never sufficient. Democracy is not proved merely when citizens are permitted to vote; it is proved by what the political system permits those votes to change, what it prevents winners from destroying, and whether everyone is still allowed to come back for the next election.
Patrick Machayo is a retired U.S. Army officer, public policy analyst and author of Stolen Youth: How Kenya’s Young Are Reclaiming Their Future.