The numbers and the map beneath them

Weintraub opened with data he had scraped that morning, and the maps told a story of a country voting almost exactly along the lines it already knew. Cepeda's support mapped with striking precision onto Gustavo Petro's 2022 coalition — he flipped essentially no municipality where the combined right-wing vote had exceeded Petro's four years earlier — while De la Espriella's strength traced the same geography as the “No” vote in Colombia's 2016 referendum on the FARC peace agreement. Turnout, oddly, had risen more in Petro's 2022 strongholds than in the right's, which Weintraub read less as good news for Cepeda than as a sign his coalition may already be close to its ceiling, with more untapped voters available to De la Espriella heading into the runoff. Modeling out several scenarios for how Valencia's, Fajardo's, and López's voters might split, most of his projections had De la Espriella winning comfortably — though, as he noted, Colombian elections have a way of moving in the three weeks between rounds.

What the numbers don't show

Mejía pushed back on two points he felt the quantitative read was missing. First, the Pacto Histórico's cohesion: a Colombian left with a long history of fragmentation had, remarkably, stayed unified behind a famously self-centered leader in Petro — though Mejía flagged early signs that some in the movement were quietly urging Cepeda to distance himself, a tension he expected to shape the run-up to the runoff and, potentially, Colombian politics well past this election cycle.

Second, and more pointedly, Mejía argued against reading De la Espriella as simply Colombia's version of an internationally franchised populism — Bukele, Milei, Trump — with the usual explanations about fear and security anxiety layered on top. There's some of that, he allowed. But he made the case that De la Espriella had tapped something more specifically Colombian: a cultural premium on hard work as the individual, personally-earned marker of social standing — a value Mejía argued Petrismo's more abstract, structural language about labor and capital simply doesn't speak to — paired with a live-and-let-live, self-reliant streak that runs across class lines. He pointed to a viral interview De la Espriella gave the streamer Westcol, where he defended a truck driver's right to carry a gun against highway robbers in strikingly plain, non-ideological terms, as the clearest illustration of a message landing well beyond the demographic most analysts assumed was his base.

A clean vote, an anxious moment

Kaplan drew on firsthand observation, having been in Colombia both eighteen months earlier and again just weeks before the vote, where he watched supporters of every camp convince themselves their candidate would win. He was careful to note the good news first: Colombia's electoral authority and EU observers both called the vote clean, turnout was strong, and — in a detail that made the local news cycle that morning — Colombian voters get paid time off work for each round they vote in. But the deeper story, in his read, was a political system whipsawed by unmet expectations since the 2016 peace deal, Petro's stalled “total peace” negotiations, and a personal, legal feud between Cepeda and former president Álvaro Uribe layered on top of an election that had become, in effect, a referendum on Petro himself. Cepeda's central problem, Kaplan argued, echoed Kamala Harris's in the 2024 U.S. election: tied to an unpopular incumbent's record without ever having been able to fully break from it. Whether he could move toward the center — a softer task against De la Espriella than it might have been against the more centrist Valencia — was, in Kaplan's view, the pivotal question of the next three weeks. He closed by naming the deeper worry running under both campaigns: real, if differently shaped, concerns about what either winner might do to Colombia's political institutions.

A regional pattern, and a geopolitical wildcard

Curiel widened the lens. She was careful not to flatten every recent right-populist surge in Latin America into a single story, but she argued it's hard to ignore a regional pattern all the same — strikingly similar candidates, rhetorical strategies, and now fraud allegations recurring across countries (Honduras, Peru, Bolivia among them) whose underlying security and political problems differ substantially. She suggested it might be worth studying today's crop of outsider-right figures the way scholars once studied the diffusion of Latin America's leftist movements in the 1960s and '70s — as a phenomenon with its own cross-border mechanics, not just parallel domestic stories.

On the United States, Curiel expected rhetorical engagement from the Trump administration — De la Espriella is an open admirer who had already met with Trump officials in Miami — but was skeptical it would prove decisive; Cepeda, by contrast, might govern more in the pragmatic, arm's-length mode of Mexico's Claudia Sheinbaum, building on the more accommodating turn Petro's own relationship with Washington had taken since earlier in the year. She also flagged a looming energy question neither candidate had addressed in any detail: an anticipated El Niño-driven drop in hydroelectric capacity, and the resulting scramble for imported natural gas — from the United States or, potentially, Venezuela — that whoever won would inherit almost immediately. And she pointed to a genuine irony in De la Espriella's biography: a right-wing candidate whose past legal work had tied him personally to a figure associated with the Maduro government, complicating any simple ideological reading of what a De la Espriella presidency would mean for Colombia–Venezuela relations.

Where the center went, and what comes next

The discussion that followed ranged widely: why Paloma Valencia's campaign collapsed after polling near 20 percent just weeks earlier (the panel's rough consensus: a hard right that structurally tends to outflank the establishment right across the region, compounded by a Petrista strategy that concentrated its fire on De la Espriella as the tougher runoff opponent, leaving Valencia to absorb collateral damage); whether Colombia's center had simply disappeared with the 2016 peace agreement; and how much weight to put on U.S. involvement or even World Cup-driven national sentiment as runoff wildcards. Nobody thought either would be decisive on its own.

Asked directly what a De la Espriella or a Cepeda presidency would mean for the health of Colombian democracy, Weintraub was candid about his concern — less because either candidate is uniquely dangerous, he said, than because Colombia's political system has grown exhausted, its 1991 constitutional settlement showing diminishing returns and a public “willing to experiment” after the unfulfilled promise of Petro's own presidency. Curiel and Kaplan both located genuine grounds for hope in the same place: an unusually fragmented political landscape, without a single unchallenged caudillo on either side, and courts and regulatory institutions that had — so far — held their independence through repeated stress tests. On violence specifically, the panel's closing predictions diverged by scenario. Under De la Espriella, Weintraub and Kaplan worried that a rupture in Petro's ongoing peace talks would trigger retaliatory violence, that a return to aerial coca fumigation would collide head-on with the Constitutional Court, and that elite or paramilitary backlash would follow if Cepeda had instead pursued land reform and a constituent assembly. Under Cepeda, the concern ran more toward armed groups continuing to exploit ceasefires and extortion economies, and toward the durability of institutional checks on a government still tied to Petro's unfinished agenda.

Díaz-Cayeros closed the session on a note that has stayed with the panel since: a viewer's comment in the chat had framed the choice as one between “the devil you know and the devil you don't.” Fair enough, he said — but democracy remains one of the few systems humanity has built that lets a society get rid of either devil, peacefully, and try again.

Postscript: what actually happened

Set against the news since, the panel's read holds up better in its worries than in its point estimates.

The margin was the first surprise. Weintraub's scenario modeling, built the morning after round one, had De la Espriella winning comfortably in most transfer scenarios. The actual runoff, held as scheduled on June 21, was instead a nail-biter: 49.66 percent to 48.70 percent, a gap of roughly 246,000 votes out of some 26 million cast — under one percentage point, and close enough that Cepeda initially declined to concede, contesting results from close to a quarter of polling stations before formally accepting the outcome on June 24 with a pledge to “exercise a democratic, vigilant and constructive opposition.” Election night itself brought clashes between protesters and police in Cali, an early sign of the social tension the panel had flagged as a risk under either outcome.

Weintraub's own post-election analysis, published the day of the runoff, put hard numbers behind the panel's “entrenched spatial blocks” thesis — and, in the process, revised one of the panel's own central premises. Only 26 of Colombia's more than eleven hundred municipalities changed which candidate won between the two rounds (14 flipping to De la Espriella, 12 to Cepeda), and nearly all of those flips came in places that were already razor-thin in round one. The country that voted for De la Espriella — Norte de Santander at 77 percent, Casanare at 69, Santander at 65, Antioquia at 64 — and the country that voted for Cepeda — Chocó and Vaupés both above 80 percent, Putumayo at 79, Nariño at 77, Cauca at 76 — barely overlapped at all. Bogotá told a version of the story Weintraub had flagged live on the panel, now with exact numbers: Cepeda carried the capital 52.5 percent to 45.4, a margin of about 302,000 votes, solid but well short of what Petro had won there in 2022, even as Cepeda gained ground in every other major city between rounds — up 8.7 points in Medellín, 8.2 in Cali, 8.0 in Cartagena, 7.1 in Barranquilla.

The more important correction concerns the “pivotal center” the panel spent so much of its discussion on the day after round one — the roughly 2.8 million voters who had backed Valencia, Fajardo, and López. Weintraub's later numbers show both campaigns' gains between rounds exceeded that entire pool on their own: De la Espriella added nearly 2.6 million votes beyond his first-round total, more than Valencia's whole first-round universe by itself, while Cepeda added over 3 million, more than the combined Fajardo-López vote. The runoff, in other words, wasn't decided mainly by which side won over centrist defectors but by each campaign's ability to pull entirely new voters to the polls, a dynamic already visible in round one, when turnout had risen 2.25 million over 2022, disproportionately in Cepeda's Caribbean and Pacific strongholds. Between rounds, turnout climbed further, up 5.6 points to 63.5 percent (3.6 million more voters than the 2022 runoff), and Cepeda's operation mobilized more of them in absolute terms than De la Espriella's. It simply wasn't enough to close a gap that had already opened in round one. As Weintraub summarized: De la Espriella won a close, high-turnout election on the strength of an intact regional coalition plus modest gains in the center; Cepeda outgrew his first-round vote proportionally, but never closed the gap he faced.

Washington reacted exactly as Curiel expected, if more enthusiastically than her cautious framing implied: Trump declared on Truth Social that De la Espriella had “Won, BIG!,” and Secretary of State Marco Rubio called the president-elect to discuss regional security cooperation, immigration, and trade. Precisely, the rhetorical engagement Curiel predicted, delivered with more warmth than skepticism.

Despite initial post-election tensions, transfer of power went ahead on schedule, if uneasily. De la Espriella was sworn in on August 7, not in Bogotá, but in Cali, under an unprecedented security operation of some 11,000 troops and a drone ban, with Javier Milei and José Antonio Kast among the foreign leaders in attendance.

On policy, the new president has moved fast to confirm Weintraub's most specific fears. He has vowed to end the Petro government's peace negotiations with armed groups, resume aerial fumigation of coca crops (a practice suspended for more than a decade), and the exact flashpoint Weintraub predicted would trigger a “choque” with the Constitutional Court — and expand the military's role against criminal organizations, alongside new large-scale prison construction. Analysts are now warning, in terms that echo Weintraub and Kaplan's closing remarks almost exactly, that abandoning negotiations could backfire: an estimated 27,000 members of armed groups currently have no reintegration pathway if talks collapse entirely. On foreign policy, De la Espriella has aligned closely with the Trump administration, backing its military strikes on suspected drug vessels in the Caribbean, expediting extraditions, severing relations with Cuba, and restoring ties with Israel. 

Six weeks into the new administration, the question the panel closed on that afternoon has stopped being hypothetical and started being lived: can Colombia's institutions absorb what comes next? It's the story DAL's "Democracy at the Ballot Box" series intends to keep following.

Hanna Folsz then took a closer, domestic look at the Tisza Party and how they triumphed over Orbán. As Kelemen discussed, Orbán's new electoral rules strongly favored large parties with rural bases, the characteristics of the Fidesz party. The Fidesz Party also controlled the media and enjoyed advantages in party financing. However, the Tisza Party, led by Peter Magyar, dominated the 2026 election, despite the electoral system being stacked against opposition parties. 

Economic woes, corruption, and scandals surrounding Fidesz created broad voter discontent and set the stage for the Tisza Party’s victory. Tisza worked to create a broad coalition through extensive group-level campaigning, messaging that focused on competent economic governance and anti-corruption, and the idea of reclaiming patriotism. Magyar also extensively campaigned, holding rallies all over Hungary in localities of all sizes. The district candidates within the Tisza Party campaigned in a similar manner. 

The Tisza Party focused its policy proposals on extensive welfare, public services improvement, the elimination of corruption, strengthening relationships with the EU and neighbors, and largely avoided divisive topics. The Party also distanced itself from the discredited and divisive established opposition parties, and they did not coordinate with past opposition parties. 

Folsz outlined the lessons Hungary’s electoral outcome shows for democratic resistance against autocratization. The Hungarian case demonstrated the importance of connecting with voters and building credibility by campaigning a lot and across the country, including in rural constituencies. The Tisza Party also smartly presented a vision for a better future with concrete proposals, rooted in citizens’ core concerns– in this case, the economy and corruption, and distanced themselves from divisive opposition politicians and parties. The Tisza Party focused its messaging on unity and reclaiming patriotism from the far right.

The 2026 Hungarian election offered a rare example of democratic recovery in a system widely considered entrenched, raising important lessons for opposition movements confronting democratic erosion.

Introduction and Argument:

Many authoritarian countries hold elections where the incumbent might lose. The odds tend to be quite narrow, however, owing to the autocrat’s asymmetric control over economic resources, security forces, media, and so on. An important practical and theoretical question, then, is how the opposition can beat these narrow odds. 

Some scholars have argued that oppositions can defeat authoritarian incumbents by building broad, multiparty coalitions. Doing so should not only decrease the autocrat’s vote share but should also deter him from deploying state repression against the opposition’s supporters. Indeed, security forces will struggle or be hesitant to shoot at such large numbers of people, and doing so will likely attract international condemnation. All of this sounds intuitively plausible. 

In “When you come at the king,” Oren Samet shows how arguments for building big coalitions overlook a crucial possibility: If the opposition unites and performs well but still fails to defeat the autocrat, he may be “spooked” and react by doubling down on repression. This is because elections provide the government with information about its own and the opposition’s popularity. Too much opposition success further decreases the autocrat’s willingness to tolerate popular elections and freedoms. Therefore, the same strategy enabling oppositions to achieve a “stunning election” can — if the coalition fails to take power — lead to a “nearly stunning” election that further entrenches authoritarianism. Hence the paper’s title, a quote from The Wire’s Omar Little: when “you come at the king, you best not miss.” 

The paper provides both cross-national data and an in-depth case study of Cambodia to show how the logic of nearly stunning elections poses a serious dilemma for democracy promoters: When oppositions cannot defeat autocrats, then they must achieve a “sweet spot” of neither too many votes (which scares the incumbent into autocratizing) nor too few (which fails to threaten the incumbent and compel him to make democratic concessions). Yet deliberately planning to hit this sweet spot is simply not possible. Samet thus offers an important challenge to the claim that bigger is better in authoritarian elections.

Cross-National Findings:

Samet’s argument about the pitfalls of nearly stunning elections implies three hypotheses. First, and as previous scholarship suggests, coalitions should outperform individual opposition parties in authoritarian elections. Second, absent the incumbent’s defeat, autocratization is more likely as the opposition’s vote share increases. And third, absent the incumbent’s defeat, countries with high-performing oppositions should witness (a) an increase in state repression and (b) decreases in the quality of elections in the years following a nearly stunning election. Samet then analyzes all elections from 1990 to 2022 in cases where the same authoritarian leader or party had ruled for at least 10 years. This yields 286 elections: 58 (20%) featured coalitions, and 28 (10%) featured electoral turnovers. These numbers alone paint a bleak picture of the prospects for beating dictators. 

The statistical results broadly support Samet’s hypotheses. Coalitions do in fact perform better at the ballot box, winning a median vote share of 36% (compared to just 13% for individual parties). In addition, and consistent with the idea that hitting the “sweet spot” will encourage autocrats to make concessions, Samet finds a positive association between moderately strong opposition performance and democratic change. Importantly, however, levels of democracy decline sharply as the opposition vote share approaches 50%. Nearly stunning elections thus appear to provoke autocratization, both in the short- and medium-term. Finally, the relationship between nearly stunning elections and repression or electoral fraud is somewhat weaker. This may be because the autocrat has more than just these two tools at his disposal — he might limit the number of seats that can be contested in future elections, prevent the opposition from accessing state media, and so on.
 

Image

FIGURE 2: Opposition performance in authoritarian elections. Note: Density plots display the frequency of specific opposition (opp.) vote shares (left) and vote margins (right) broken down by coalition status.

Image

FIGURE 3: Opposition performance and electoral democracy change. Note: Displays elections that did not feature turnovers, plotted along two dimensions: opposition vote share and change in electoral democracy. ∙ (dot) denotes election with coalitions; × denotes election without. The gray lines plot LOESS regressions fit to the data, with gray shading indicating 95% confidence intervals.

Image

FIGURE 6: Predicted change in repression and electoral manipulation, with controls. Note: Displays predicted post-election changes in (1) V-Dem measure of electoral irregularities, comparing election at t with election at t + 1; (2) V-Dem measure of government intimidation of opposition, comparing election at t with election at t + 1; (3) Fariss (2014) physical integrity rights measure 3 years after election, using the lagged score as a baseline. Includes all elections that did not result in turnovers. Dotted lines indicate 95% confidence intervals.
 

The Cambodia Case:

Samet concludes by showing how his theoretical process — oppositions uniting, nearly winning an election, frightening the incumbent, then increasing authoritarianism — is borne out in Cambodia’s recent political history. Throughout the 2000s, Cambodia’s opposition was fragmented, in part due to deliberate actions by its authoritarian Prime Minister, Hun Sen. Ahead of the 2013 elections, in the face of mounting popular dissatisfaction with the government, the two largest opposition parties coalesced. Hun Sen was confident in his position, in part because Cambodia’s strongest opposition party had won just 22% of the vote in 2008. As such, he pardoned Sam Rainsy, one of the country’s most prominent opposition leaders, whom he invited to return from exile. 

The coalition did remarkably well in 2013, winning around 45% of the vote, but then alleged fraud and refused to take their seats in parliament. Opposition supporters then took to the streets in protest, where they were met with state violence. Yet Hun Sen made a number of concessions to successfully quell the protest crisis, including reforming the election commission.

By 2015, however, signs of autocratization became glaring. Opposition lawmakers were publicly beaten by the personal bodyguards of Hun Sen, who withdrew his prior pardon of Rainsy. Other opposition leaders faced politically motivated legal cases. Ahead of the 2018 elections, Hun Sen’s government hired an external survey firm, which found the opposition had become even more popular among Cambodians since 2013. Hun Sen’s fears were aggravated by a strong opposition performance in the 2017 local elections. 

Samet argues that this was the last straw: the government responded by promptly arresting and exiling opposition leaders and dissolving their coalition. All of this constituted the most dramatic instance of autocratization in Cambodia since the 1990s. Hun Sen’s allies then ran unopposed in the 2018 elections. By this time, the opposition was once again divided — particularly as regards how to face a government whose elections could barely be characterized as anything other than window-dressing. “When you come at the king” offers an important if distressing lesson for practitioners and scholars of democracy.

*Brief prepared by Adam Fefer