Registraduría’s calendar sets Colombia’s first-round presidential election for May 31, 2026, and the final electoral roll is 41,421,973 voters. The current ballot is broad, not just the three names in the prompt; besides Iván Cepeda, Abelardo de la Espriella, and Paloma Valencia, it also includes figures such as Claudia López, Sergio Fajardo, and Luis Gilberto Murillo. That matters because fragmentation usually pushes the first-place raw vote total down even when the leader is clear. (registraduria.gov.co)

Historical baselines give the rough scale. In the first round, the leader got 3,759,971 votes in 2014, 7,569,693 in 2018, and 8,541,617 in 2022; total ballots cast were about 13.12 million, 19.64 million, and 21.42 million, respectively. The 2026 electorate is larger than in 2022, so if turnout merely matches the 2018-2022 presidential range, Colombia would see about 22.11-22.74 million ballots; my back-of-the-envelope translations of that turnout range imply about 8.40 million votes for a leader on 38% of ballots and about 9.21 million for a leader on 40.5% of ballots. (colombia.com)

As of late April and early May 2026, the polling picture points to a leader in the high-30s to low-40s, not an easy first-round majority. GAD3 had Cepeda at 36%, de la Espriella at 21%, and Valencia at 13% in late April; Atlas/Semana showed 37.4% of total responses for Cepeda (38.0% of valid votes), 29.4% for de la Espriella, and 20.9% for Valencia; Invamer showed 44.3%, 21.5%, and 19.8%; CNC showed 37.2%, 20.4%, and 15.6%; and an EL PAÍS summary of Guarumo put them at 38%, 23%, and 22.8%. Across methods, Cepeda is consistently first, but the right remains split and the exact margin varies a lot. (valoraanalitik.com)

Campaign reporting reinforces that interpretation: traditional parties are divided, many of their leaders lean toward Paloma Valencia, while the Char machine is moving toward de la Espriella. So the most likely outcome is still a Cepeda plurality, but not a clean 50%+ knockout; I therefore keep a meaningful minority scenario in which a conservative tops the field or Cepeda underperforms late. (elpais.com)

I model the vote total as a mixture of four scenarios: 15% chance a conservative candidate finishes first at roughly 7.9 million votes, 50% chance Cepeda wins a standard runoff-level plurality around 9.0 million, 30% chance Cepeda wins a stronger plurality around 9.9 million, and 5% chance of a near-outright-win surge around 11.4 million. The mixture mean is 9.225 million votes. This final paragraph is my inference from the electorate size, turnout history, current polls, and endorsement pattern, not a direct source claim. (registraduria.gov.co)

I treat this as a vote-total question, not a winner-identity question. Colombia’s first-round presidential election is scheduled for May 31, 2026; the Registraduría says 41,421,973 citizens are eligible to vote; and the ballot contains 14 presidential formulas. Current campaign coverage still describes a three-way race led by Iván Cepeda, with Abelardo de la Espriella and Paloma Valencia fighting for second. (registraduria.gov.co)

The best guide to the leader’s vote total is current polling plus a turnout model. The public polls I found show a fairly consistent pattern: CNC in March had Cepeda at 34.5%, Valencia at 22.2%, and de la Espriella at 15.4%; GAD3 in March had Cepeda 35%, de la Espriella 21%, Valencia 16%; Guarumo/EcoAnalítica had Cepeda 37.5%, de la Espriella 20.2%, Valencia 19.9% in March and 38.0%, 23.9%, 22.8% in late April; AtlasIntel in late April had Cepeda 37.4%, de la Espriella 29.4%, Valencia 20.9%; and Invamer, fielded April 15-24, had Cepeda 44.3%, de la Espriella 21.5%, Valencia 19.8%. Taken together, that points to a front-runner usually in the high 30s to low 40s, not yet a stable 50%+ candidate. (publimetro.co)

Recent first rounds give a useful base rate for absolute vote totals. The first-place candidate received 3,769,005 votes in 2014 (a low-turnout, fragmented race), 7,616,857 in 2018, and 8,542,020 in 2022. Turnout was about 40.07% in 2014, 53.11% in 2018, and 54.91% in 2022. That history suggests that with today’s much larger electorate, a normal contemporary plurality winner is more likely to land somewhere around the upper 7 millions to low 9 millions than anywhere near the 2014 outlier. (cmscedae.registraduria.gov.co)

My model is: leader votes = registered voters × turnout × share of ballots that become candidate votes after excluding blank/null/unmarked ballots × leader share of candidate votes. I anchor registered voters at the official 41,421,973. For turnout, I center near 54.5%: the March 2026 congressional election already reached just over 50.6%, 2022 congressional turnout was 47%, and 2022 presidential first-round turnout rose to 54.91%, so a low-50s-to-mid-50s presidential turnout is the likeliest band. For the conversion from ballots cast to candidate votes, recent presidential first rounds were fairly efficient after exclusions in 2018 and 2022, though 2014 is a reminder that blank votes can be materially higher in a fragmented contest. (registraduria.gov.co)

For the leader’s share of candidate votes, I center a bit below Petro’s realized 2022 first-round share of candidate votes and very near Duque’s 2018 realized share, because 2026 polling usually puts the leader in roughly the same zone once you mentally strip out undecided/blank responses, but the 14-candidate ballot and the fight between Valencia and de la Espriella create real fragmentation risk. Using a low case of 51.5% turnout, 95.5% candidate-vote conversion, and 35.5% leader share gives about 7.23 million votes; a base case of 54.5%, 96.2%, and 39.2% gives about 8.51 million; and a higher-turnout/higher-share case of 56.0%, 96.8%, and 41.5% gives about 9.32 million. For calibration, Petro’s 2022 first-round result works out to about 41.1% of candidate votes, and Duque’s 2018 first-round result to about 40.1%. (publimetro.co)

Putting that together, my median forecast is in the mid-8 millions, with Iván Cepeda the most likely first-place finisher but with enough uncertainty about turnout and late consolidation that a lower plurality around 7 million is still plausible. I therefore center the distribution at roughly 8.5-8.6 million votes, with most of the mass between about 7.2 million and 9.9 million, and only a thin tail above 10.5 million. That makes a final certified first-place tally around 8.54 million my best estimate.

As of May 11, 2026, Colombia’s first round is scheduled for May 31, 2026. The official electoral census for that vote is 41,421,973 voters, and the official ballot has 14 presidential formulas, not just the three names highlighted in the prompt. That matters because a 14-candidate ballot usually fragments the vote and makes it harder for the leader to push into the 50 percent-plus zone. (wapp.registraduria.gov.co)

For base rates, the top first-round candidate received 3,759,971 votes in 2014, 7,569,693 in 2018, and 8,541,617 in the 2022 official scrutiny. The broad pattern is straightforward: when Colombian first-round presidential turnout is high, the first-place raw vote rises a lot. The public tallies shown on the 2014 and 2018 result pages put turnout at about 40.1 percent and 53.4 percent, while the 2022 official scrutiny was just under 55 percent. (colombia.com)

Turnout is therefore the first major input. The March 8, 2026 congressional election drew about 20.9 million voters and 50.6 percent participation, which was above the 47 percent congressional participation recorded in 2022. Presidential first rounds historically run above congressional participation, and the voter roll has since risen to 41.42 million. So I center presidential turnout in the mid-to-high 50s rather than around 51 to 52 percent. (registraduria.gov.co)

The second major input is the leader’s share of candidate votes. Late-April polling is directionally consistent but numerically dispersed: GAD3 for RCN, fielded April 20 to 22, had Iván Cepeda at 36 percent, Abelardo de la Espriella at 21 percent, and Paloma Valencia at 13 percent; AtlasIntel for Semana, fielded April 25 to 29, had Cepeda at 37.4 percent, de la Espriella at 29.4 percent, and Valencia at 20.9 percent; Invamer for Caracol and Blu Radio reported Cepeda at 44.3 percent. AS/COA still describes the race as a competitive three-way contest. That combination argues for a leader share centered around roughly 40 percent of candidate votes, with a meaningful but not dominant upper tail into the mid-to-high 40s. (gad3.com)

The third input is ballot efficiency, because the resolution excludes blank, null, and unmarked ballots. In the 2022 official first round, blank, null, and unmarked ballots together were just under 3 percent of all ballots cast. The 2018 first-round public tally was also only a few percent outside candidate votes, while 2014 was unusually high because blank voting was much stronger. With 14 candidates on the 2026 ballot, I expect blank voting to stay well below the 2014 spike, so I place candidate votes in the mid-90s as a share of total ballots cast. (registraduria.gov.co)

Multiplying those pieces gives a central estimate near 9.0 million votes: 41.42 million registered voters times roughly 56.3 percent turnout times roughly 96.4 percent candidate ballots times a leader share a little above 40 percent lands close to that level. I then widen the distribution because there are two materially different ways this election can resolve: a more fragmented race where first place is won in the mid-30s, or a stronger Cepeda performance that gets close to an outright-majority attempt. My resulting distribution is centered just under 9.1 million, with most probability mass roughly from the mid-7 millions to the high-10 millions, and a thinner tail above 11.5 million. (registraduria.gov.co)

I model the outcome as: registered voters × first-round turnout × share of ballots that become votes for human candidates × the leader’s share of those candidate votes. Registraduría’s calendar sets the first round for May 31, 2026. Registraduría also says 14 presidential candidacies were registered, and the final electoral census for the presidential first round is 41,421,973 voters. (registraduria.gov.co)

Turnout is the first big lever. Registraduría’s own historical bulletin says presidential first-round turnout was 39.7% in 2014, 55.5% in 2018, and 55.0% in 2022. For the March 8, 2026 congressional election, Registraduría said participation rose from 47% in 2022 to more than 50% in 2026. That makes a reversion to the low-turnout 2014 pattern unlikely; I center presidential turnout in the mid-50s, around 56%, but keep some room above and below that. (registraduria.gov.co)

The second lever is the leader’s vote share. Recent public polls are noisy, but they are consistent that Iván Cepeda is ahead and that the opposition vote is split. The latest major polls I found show the leader at 37.2% in CNC, 38.0% in Guarumo/EcoAnalítica, 36% in GAD3, 38.0% in AtlasIntel, and 44.3% in Invamer. Because Abelardo de la Espriella and Paloma Valencia both remain substantial, the most likely first-place share still looks like high-30s to low-40s headline vote intention, not an outright-majority performance. (elcolombiano.com)

The resolution excludes blank, null, and unmarked ballots, so I need a further adjustment. Historically that adjustment is material but not enormous in recent Colombian presidential first rounds. In 2022, the official candidate table sums to roughly 20.8 million candidate votes, while MOE reports 21.44 million voters in the first round; in 2014, by contrast, vote in blank alone reached 770,610 votes, or 5.99% of the total. So I do not take poll blank-vote figures literally; instead I assume roughly 95% to 97% of ballots end up counting for human candidates, with a midpoint a bit below the very clean 2022 pattern. (biblioteca.registraduria.gov.co)

Putting those pieces together, a central case is a 41.42 million electorate, turnout in the mid-50s, a mid-90s candidate-valid share, and a leader share in the low 40s of counted candidate votes. That points to a result a little above 9.3 million votes. This is above Gustavo Petro’s 8,541,617 first-round votes in 2022 and Iván Duque’s 7,569,693 in 2018, which is exactly what I would expect from a larger 2026 electorate with participation still near recent presidential norms. (registraduria.gov.co)

My final distribution is therefore centered near 9.3 million, with meaningful tails on both sides: lower outcomes if turnout softens or the field fragments more than polls suggest, and higher outcomes if the front-runner consolidates undecided and soft protest voters late. In round numbers, I think the 80% credible interval is roughly 8.2 million to 10.6 million, and my median is about 9.3 million votes for the first-place candidate. (elcolombiano.com)

Colombia’s first presidential round is set for May 31, 2026. The Registraduría says 41,421,973 citizens are eligible to vote, and 14 presidential formulas were registered. Most current coverage describes the contest as effectively a three-way race, with Iván Cepeda leading while Abelardo de la Espriella and Paloma Valencia fight for second. (registraduria.gov.co)

For a base rate, the first-round leader got 3.76 million votes in the unusually fragmented 2014 race, 7.57 million in 2018, and 8.54 million in the official 2022 first-round count. The 2022 first-round turnout was 54.98%. So, absent an extreme fragmentation scenario like 2014, the modern baseline for a Colombian first-round leader is in the upper-7 to mid-8 millions before adjusting for the larger 2026 electorate. (electionguide.org)

The participation backdrop in 2026 looks stronger than in 2022. The Registraduría said turnout in the March 8, 2026 congressional election rose to 50.6%, up from 47% in the 2022 congressional election, and the presidential electorate is now 41.42 million versus 39.00 million registered voters in the 2022 presidential first round. That makes a presidential turnout in the mid-to-high 50s more plausible than another low-turnout outlier. (registraduria.gov.co)

For leader share, the late-April and early-May polls cluster in a fairly usable band, with one high-end outlier. CNC had the leader at 37.2%; AtlasIntel had 37.4% of total votes and 38.0% after excluding non-valid responses; Guarumo/Ecoanalítica had 38.0%; and Invamer had 44.3%, with blank vote at 4.8%. El País summarized the latest wave as putting Cepeda between 37% and 44%, with de la Espriella still ahead of Valencia for second, though narrowly. (elespectador.com)

I also think the right remains split enough to keep the first-place total from collapsing into a low-share outcome. Valencia has ruled out a first-round alliance with de la Espriella, and de la Espriella has likewise publicly rejected first-round alliances. That matters because the question is about the top vote total, not the winner’s identity; a durable split between the two strongest right-wing candidacies supports a relatively healthy first-place tally even if the race tightens. (infobae.com)

A small technical adjustment: the resolution excludes blank, null, and unmarked ballots. In the official 2022 first-round count, those categories totaled just under 3% of all ballots, and the 2026 polls still show blank vote mostly in the low single digits. Because null and unmarked ballots are not polled well, I do not make a sharp point adjustment for them; instead I widen uncertainty slightly. (registraduria.gov.co)

My midpoint arithmetic is: 41.42 million registered voters × roughly 56.5% turnout × roughly 39.5% for the first-place candidate, which gives about 9.24 million votes. A useful cross-check is Petro’s official 2022 first-round total of 8,541,617 votes; scaling that by electorate growth alone gets a bit above 9.07 million, and the somewhat stronger 2026 participation backdrop plus current polling nudges me above that. I therefore center the distribution in the low-9 millions, with meaningful upside into the 10s of millions if the high Invamer-style scenario is right, and a lower tail into the low-8 millions if turnout softens or the leader underperforms late polls. My median forecast is about 9.3 million votes for the first-place candidate. (registraduria.gov.co)

What number of valid votes will the first-place candidate receive in the first round of the 2026 Colombian presidential election?

• My most likely prediction for the first-place candidate’s valid vote count is 8,727,500, with a 50% chance of falling between 8,002,500 and 9,525,000.
• The 2026 electoral census has expanded to 41.4 million eligible voters, a 6.2% increase from 2022, which naturally pushes potential vote totals higher for the frontrunner.
• Frontrunner Iván Cepeda leads current polling with a five-poll average of 38.6%, suggesting he is likely to match or slightly exceed Gustavo Petro’s 2022 first-round performance (8.53 million votes).
• Voter turnout is projected to be between 54% and 57%, which would translate to approximately 22.4 million to 23.6 million total ballots cast.
• Prediction markets and historical patterns indicate an outright win (>50%) is highly unlikely (approx. 5.5% probability), meaning the first-place count will likely remain below 11.4 million.
• The contest is defined as a competitive three-way race between Cepeda (Left), Abelardo de la Espriella (Right), and Paloma Valencia (Center-Right), which tends to fragment the vote and cap the leader’s ceiling.

Key figures

Historical context

The Colombian presidential election is a two-round system requiring over 50% for an outright win. Historically, first-round leaders have seen their absolute vote counts rise alongside population growth. In 2010, Juan Manuel Santos led with 6.76 million votes. In 2018, Iván Duque secured 7.57 million votes. The 2022 election set a record for a first-round leader, with Gustavo Petro receiving 8,527,768 votes (40.32% share). However, the 2014 election serves as a cautionary outlier for fragmentation, where Óscar Iván Zuluaga led with only 3.76 million votes (29.25% share). The 2026 cycle occurs in a context where the incumbent (Petro) cannot seek reelection, and the electorate has grown from 39 million to over 41.4 million eligible voters. Parliamentary elections in March 2026 recorded a turnout of 50.62%, roughly consistent with the 2022 legislative turnout (48.1%), suggesting no major shift in underlying participation trends.

Tailwinds

• The 6.2% increase in the electoral census (from 39M to 41.4M) provides a larger pool of potential voters than any previous election.
• Iván Cepeda’s momentum in late polling (reaching up to 44.3% in some surveys) suggests a consolidated base for the left-wing coalition.
• Historical trends show presidential turnout consistently exceeds legislative turnout, pointing toward a likely participation rate above 54%.
• Sharp improvements in consumer confidence (+19.3% in March 2026) and stabilizing GDP growth may encourage higher civic engagement.

Headwinds

• High levels of political violence (60+ leaders killed) could suppress turnout in rural regions.
• A highly fragmented field of 14 candidates may prevent any single candidate from capturing a dominant plurality.
• Low institutional trust (6.5/10) could lead to an increase in blank (voto en blanco) or null votes, reducing the tally for individual candidates.
• A technical tie or late-stage consolidation between right-wing candidates could split the anti-establishment vote more evenly.

Detailed reasoning

My prediction is derived from a systematic decomposition of three primary variables: the official electoral census, projected voter turnout, and the leading candidate’s polling share.

1. Electoral Census: The Registraduría Nacional del Estado Civil (RNEC) has confirmed a final census of 41,421,973 eligible voters for the 2026 presidential election. This is a 6.2% increase from the 39.0 million eligible voters in 2022. This expansion provides a higher floor for absolute vote counts compared to previous cycles.

2. Turnout Projections: Historical turnout for Colombian presidential first rounds has been relatively stable, with 54.98% in 2022 and 53.4% in 2018. The March 2026 legislative elections saw a turnout of 50.62%, and presidential elections typically see a 3-5% “bump” over congressional participation. Prediction markets (Polymarket) place the highest probability (43%) on a turnout bracket of 54-57%. Using a central turnout estimate of 55% results in approximately 22.8 million total votes cast.

3. Leading Candidate Share: Polling consistently identifies Iván Cepeda (Historic Pact) as the frontrunner, with averages ranging from 36.0% (GAD3) to 44.3% (Invamer). In 2022, Gustavo Petro placed first with 40.32% (8.53 million votes). In a competitive three-way race featuring Abelardo de la Espriella and Paloma Valencia, it is unlikely the leader will significantly exceed 45% without a major consolidation of the field. Polymarket estimates only a 5.5% probability of an outright win (over 50% of the vote).

4. Synthesis: Multiplying a central turnout of 22.8 million by a leading candidate share of 38-40% yields a range of 8.66 to 9.12 million votes. The most likely value of 8.73 million reflects a slight adjustment for “blank/null” votes, which are excluded from individual candidate tallies by the resolution criteria.

The 50% HDI (8.0M to 9.5M) reflects the core scenario of a multi-way plurality. The 80% HDI (7.1M to 10.4M) accounts for the possibility of lower-than-expected turnout or a late surge by the leader. The higher end of the distribution remains constrained by the very low probability of a first-round victory, which would require approximately 11.4 to 12.5 million votes based on expert turnout estimates. My weighting favors the current polling trend and demographic growth over more fragmented historical models like 2014.

Key uncertainties

• A late-stage withdrawal of minor candidates could lead to a sudden consolidation of votes for one of the top three contenders.
• Polling for second-place candidates has historically been volatile in Colombia (e.g., the 2022 Rodolfo Hernández surge).
• The impact of approximately 15% to 22% of voters remaining undecided just weeks before the election could shift the plurality winner’s final tally by 1-2 million votes.
• Official certification of results (E-26 form) can differ slightly from preliminary “preconteo” counts used in news reports.

Conclusion

• Expect a first-place result between 8.0 and 9.5 million votes, marking a high-water mark for a Colombian plurality leader due to census growth.
• Decision-makers should treat any figure above 11.5 million as a significant indicator of an outright win, though this remains an outlier scenario (approx. 5.5% probability).
• Monitor election day turnout reports: a participation rate exceeding 58% would likely push the winner’s count toward the 10 million mark.

Question: what is the likely raw vote total for the first‑place candidate in Colombia’s 31 May 2026 presidential first round? Forecasters agree to decompose the outcome as 41.42M eligible voters × turnout × valid‑vote share × leader’s vote share; the fixed electorate (RNEC) and recent high congressional turnout (>50%) are important anchors. Polling consistently places Iván Cepeda first but with a method‑dependent band roughly 36–44% (Invamer high‑40s outlier), while the right remains split between Paloma Valencia and Abelardo de la Espriella—an arrangement that favors a concentrated first‑place lane for Cepeda. Central scenario arithmetic across teams yields a first‑place total around 8.3–9.4 million votes (most models’ central estimates cluster ~7.8–9.8M), with downside (2014‑style fragmentation + low turnout) pulling totals toward ~5.6–7.5M and upside (≥60% turnout or Cepeda ≥48%) pushing totals into the 10–12M range. Forecasters place modest mass on very low tails (≈15–22% for fragmentation/low turnout in one model, smaller in others) and small probabilities on extreme highs (>14M or >17.5M). Main remaining risks to update before election day are whether late polls keep Cepeda in the high‑30s vs. mid‑40s, any right‑bloc consolidation, and security‑related turnout effects.

Forecast and Reasoning

Forecast: Probability distribution:

• 5.00% chance of value below 6100000.0
• 10.00% chance of value below 6700000.0
• 20.00% chance of value below 7500000.0
• 30.00% chance of value below 8100000.0
• 40.00% chance of value below 8600000.0
• 50.00% chance of value below 9000000.0
• 60.00% chance of value below 9500000.0
• 70.00% chance of value below 10000000.0
• 80.00% chance of value below 10600000.0
• 90.00% chance of value below 11400000.0
• 95.00% chance of value below 12300000.0

Forecast Reasoning

We need the vote count (excluding blank/null/unmarked) for the first-place human candidate in Colombia’s 2026 first-round presidential election (May 31, 2026), per final RNEC certification.

Key drivers are (1) electorate size, (2) turnout, (3) fraction of ballots that are valid votes for human candidates, and (4) the first-place candidate’s vote share among those candidate votes.

Evidence from the analysts:

• 2026 eligible electorate cited at ~41.4 million (Forecaster 3 with an RNEC link), up from ~39.0 million in 2022.
• Recent first-round turnout anchors: ~53.4% (2018) and ~54.9% (2022) (Forecasters 2/3). A reasonable central turnout for 2026 is therefore mid-50s, with uncertainty of a few percentage points.
• Blank/null/unmarked appear to be a small share in recent elections (Forecaster 3 cites ~97% of “suffragantes” being votes for human candidates; Forecaster 2 assumes 5–8% non-candidate ballots, which seems a bit high versus the cited 2022 breakdown). I split the difference but closer to Forecaster 3 given the concrete 2022 component counts.
• Polling mentioned by Forecasters 2/3 puts the leader’s first-round share roughly high-30s to mid-40s; with a competitive multi-candidate field, a median share near ~40% is plausible, with downside if fragmentation increases and upside if late consolidation occurs.

A simple multiplicative frame: First-place votes ≈ 41.4M × turnout (≈0.52–0.60 typical) × candidate-vote fraction (≈0.95–0.98) × leader share (≈0.35–0.46 typical). This yields a central tendency around ~8.5–9.5 million votes, with most probability mass between ~6.5 million and ~12.5 million.

Reconciling the three forecasts:

• Forecaster 1’s median (7.2M) looks low given electorate growth and typical turnout; they also put substantial mass near/below 5M which would generally require an unusually low winner share.
• Forecaster 2 is tighter and somewhat low-to-mid, with a high p5 (5.8M) that may underweight a “high fragmentation + slightly lower turnout” tail.
• Forecaster 3 is the highest (median 9.5M) and is most directly grounded in electorate/turnout arithmetic and cited polling ranges; however, I pull the median slightly down from 9.5M to reflect that the leader-share could land closer to ~38–41% rather than consistently mid-40s.

I therefore center the distribution near ~9.0M with moderately wide uncertainty bands, keeping essentially all mass well within the provided [3,000,000, 17,500,000] open bounds (extremes beyond those bounds look very unlikely under any realistic turnout/share combination).

An amateur forecast often anchors on a single prior election result (e.g., “~8.5M like 2022”) or on a polling percentage without translating it into an absolute count using the electorate and turnout, which tends to understate uncertainty and can bias the median up or down depending on the chosen anchor. This forecast explicitly composes (electorate × turnout × valid-candidate fraction × leader share), using recent first-round turnout base rates (2018/2022), the larger 2026 electorate, and a wide leader-share range consistent with the cited polls. Confidence in improvement is moderate: the arithmetic is robust, but the largest remaining uncertainty is late-campaign movement affecting the first-place share.

Forecast rationale (numeric):

— Iteration 1 — Across the forecasts, the reasoning is built around a simple vote-share decomposition:

Valid votes for the first-place candidate ≈ registered voters × turnout × candidate share

Main factors used

• Electorate size: The starting point is Colombia’s registered voter base, about 41.4 million.
• Turnout assumption: Both forecasts expect mid-50% turnout, anchored in recent first-round presidential elections and tempered by more modest recent electoral participation.
• Leading candidate share: The top candidate is assumed to win roughly 39–41% of valid votes, based on late-April polling and the historically fragmented nature of Colombia’s first-round field.

Areas of consensus

• The central estimate clusters around 9 million votes.
• Both rationales rely on recent polling plus historical first-round turnout patterns rather than assuming a dramatic change in participation.
• Both view the race as fragmented enough that the winner likely falls well short of a majority, keeping the top candidate’s vote total in the high single-digit millions rather than much higher.

Main differences / uncertainty

• The forecasts differ more in how wide the uncertainty is than in the center.
• One rationale emphasizes a broader tail risk: very low turnout, unusual fragmentation, or an exceptional landslide could push outcomes far outside the central range.
• The other gives a more explicit 80% range of roughly 7.6–11.0 million, but still keeps extreme tails open to satisfy bounding assumptions.

Overall synthesis

The combined reasoning points to a first-place vote total near 9.2 million, with most of the uncertainty coming from turnout and the exact degree of vote fragmentation rather than from the size of the electorate.

— Iteration 2 — The core reasoning is that the first-place candidate’s vote total will be driven by three main inputs: the size of the electorate, likely turnout, and the candidate’s eventual share of valid votes. Using the official registered voter base of about 41.4 million, both estimates assume mid-50s first-round turnout, then subtract a small share of invalid or blank ballots to arrive at the pool of valid votes. From there, the candidate’s vote share is anchored to polling that shows a leading plurality in the mid-40% range, but trimmed slightly to reflect expected late-campaign tightening, the large field of candidates, and the tendency for polls to overstate concentration at the top.

There is strong agreement on the central estimate: roughly 9 to 10 million valid votes for the first-place finisher. The main uncertainty comes from turnout and vote fragmentation rather than from the electorate size itself. Lower-end scenarios reflect weaker turnout and a more dispersed vote, while upper-end scenarios assume stronger mobilization or partial consolidation around one candidate. The overall forecast therefore centers on a high-single-digit to low-double-digit million vote total, with wide tails to accommodate unusually low- or high-turnout outcomes.

— Iteration 3 — Across the forecasts, the reasoning was largely mechanical and data-driven:

• Start with the size of the electorate: Colombia’s registered electorate is about 41.4 million.
• Apply expected turnout: Recent first-round turnout was treated as a guide, with most estimates centering on roughly 53%–55%, implying about 22–23 million ballots cast.
• Adjust for non-valid ballots: A meaningful share of ballots is expected to be blank, invalid, or unmarked, reducing the pool of valid candidate votes to roughly 19.5–21.5 million.
• Estimate the frontrunner’s share of valid votes: The leader’s support was anchored to late polling, especially a snapshot near 44%, but tempered by regression toward historical Colombian first-round patterns and the possibility of vote fragmentation. This generally produced an expected winner share in the low-to-mid 40s, with some caution that it could drift lower.

Main point of agreement

The forecasts converge on a first-place vote total around 9–10 million valid votes, with 10 million as a central estimate.

Main sources of uncertainty

The biggest swing factors were:

1. Turnout — whether participation is closer to the historical low-50s or meaningfully higher/lower.
2. Poll durability — whether the apparent frontrunner keeps a share near the mid-40s or fades toward historical norms.
3. Fragmentation / blank voting — how much of the electorate ends up outside valid votes for candidates.

Range of views

• One forecast was slightly more conservative, centering near 8.9 million and using a wider range.
• The other centered near 10.0 million, with a somewhat tighter central estimate but still broad uncertainty bands.

Overall, the shared reasoning suggests a competitive first round, but one in which the leading candidate likely converts a substantial plurality into roughly ten million valid votes.

SUMMARY

Question: What number of valid votes will the first-place candidate receive in the first round of the 2026 Colombian presidential election? Final Prediction: Probability distribution:

• 10.00% chance of value below 6450000.0
• 20.00% chance of value below 7150000.0
• 40.00% chance of value below 8000000.0
• 60.00% chance of value below 8800000.0
• 80.00% chance of value below 9900000.0
• 90.00% chance of value below 10750000.0

Total Cost: extra_metadata_in_explanation is disabled Time Spent: extra_metadata_in_explanation is disabled LLMs: extra_metadata_in_explanation is disabled Bot Name: extra_metadata_in_explanation is disabled

Report 1 Summary

Forecasts

Forecaster 1: Probability distribution:

• 10.00% chance of value below 6450000.0
• 20.00% chance of value below 7150000.0
• 40.00% chance of value below 8000000.0
• 60.00% chance of value below 8800000.0
• 80.00% chance of value below 9900000.0
• 90.00% chance of value below 10750000.0

Research Summary

The research analyzes the expected vote count for the first-place candidate in the first round of the 2026 Colombian presidential election, scheduled for May 31, 2026. It highlights the current electoral landscape, noting that there are 13 candidates and approximately 41.57 million registered voters. Iván Cepeda from the Pacto Histórico is leading in polls with an expected vote share of 34-44%, followed by Abelardo de la Espriella and Paloma Valencia. Historical data from previous elections indicates that first-place candidates in contested elections typically received between approximately 3.76 million and 8.53 million votes, with a clear upward trend in vote counts due to increasing voter registration.

The research also projects turnout for the 2026 election, estimating it to be between 54% and 58%, which would result in around 22.4 to 24.1 million total votes cast. Valid votes are expected to account for 96-98% of total votes, leading to an estimate of approximately 21.9 to 23.6 million valid votes. Based on Cepeda’s anticipated vote share, the expected number of votes he might receive ranges from about 7.45 million to 10.4 million, depending on turnout and polling accuracy. The analysis emphasizes key uncertainties such as potential violence, right-wing fragmentation, and polling accuracy that could affect these projections.

Sources used in the research include:

• [1]
• [2]
• [3]
• [4]
• [6]
• [8]
• [11]
• [18]
• [20]
• [22]
• [24]
• [25]
• [27]
• [29]

RESEARCH

Report 1 Research

Research Rundown: First-Place Candidate Vote Count in the 2026 Colombian Presidential Election First Round

Current Electoral Landscape

Colombia’s first-round presidential election is scheduled for May 31, 2026. There are 13 candidates on the ballot [4], and 41,569,977 registered voters [6]. The race is a three-way contest:

• Iván Cepeda (Pacto Histórico, left): Leading polls at 34–44% depending on the pollster [1][3][6]
• Abelardo de la Espriella (Defensores de la Patria, right): Polling at 21–29% [1][3][6]
• Paloma Valencia (Centro Democrático, center-right): Polling at 13–23% [1][3][6]

All analysts agree it is extremely unlikely any candidate reaches the 50%+1 threshold needed to win outright in the first round, meaning a runoff on June 21 is near-certain [2][3][8]. Cepeda is widely expected to finish first in Round 1, though his “electoral ceiling” is considered well below 50% [3].

Historical Base Rates: First-Place Candidate Votes in First Round

This is the critical reference class for forecasting the vote count:

Sources: [8][11][18][25][27][29]

Key observations:

• In the most recent contested elections where no one won outright (2014, 2018, 2022), the first-place candidate received between ~3.76M and ~8.53M votes in the first round.
• The two most comparable elections (2018 and 2022) saw the first-place candidate receive ~7.57M to ~8.53M votes, with vote shares of ~39–40%.
• There is a clear upward trend in absolute vote counts, driven by growing voter registration (from ~36.8M in 2018 to ~39M in 2022 to ~41.6M in 2026).

Turnout Projections for 2026

• Polymarket turnout market gives the 54–57% bin a 50% probability and 60%+ turnout a 32% probability [20].
• An Invamer survey from late April 2026 found 56.5% of respondents “very likely to vote” [20].
• Historical turnout in competitive first rounds: 2022 was 54.98%, 2018 was 53.40% [29][18].
• The 2026 election is highly polarized [6], which could push turnout slightly higher, but Colombia historically has abstention rates exceeding 50% [6].

Best estimate for turnout: ~54–58%, implying ~22.4–24.1 million total votes cast.

Estimating First-Place (Cepeda) Vote Count

Using the data:

• Registered voters: ~41.57 million [6]
• Expected turnout: 54–58% → ~22.4–24.1 million total votes
• Valid votes (excluding blank/null): Typically 96–98% of total → ~21.9–23.6 million
• Cepeda’s expected vote share: 34–38% (central polling range) [1][3][5][6]

Calculation:

• Low end: 21.9M × 34% ≈ 7.45 million
• Central: 22.8M × 36% ≈ 8.2 million
• High end: 23.6M × 38% ≈ 8.97 million

If we use the most aggressive polls (Invamer shows up to 44.3% for Cepeda [6]):

• High outlier: 23.6M × 44% ≈ 10.4 million (though this seems unlikely given other polls)

Prediction Markets

Important caveats on prediction markets:

• The overall winner markets favor de la Espriella, but this reflects expected second-round dynamics where the right coalesces against Cepeda [3][9]. In the first round, Cepeda is the near-unanimous favorite.
• Analyst Sergio Guzmán of Colombia Risk Analysis has expressed skepticism about the predictive power of betting markets for Colombian elections, noting discrepancies with actual primary results [22].
• The margin-of-victory market ($11,288 volume) is very thin and less reliable [21].
• The overall election market ($28M) has meaningful liquidity [23][24].

Key Uncertainties & Risk Factors

1. Violence and security: An escalation of attacks by FARC dissidents (21 civilians killed in a recent attack [3]) could either depress turnout or boost right-wing candidates who promise “hard hand” policies [6].
2. Right-wing fragmentation: If de la Espriella or Valencia collapses, votes could shift, potentially lowering or raising Cepeda’s relative share [2][3].
3. Polling accuracy: Colombian polls have historically had notable misses — the 2016 plebiscite where “No” won despite “Yes” projections, and 2022 where Hernández vastly outperformed polls [2][25]. AtlasIntel has been among the more aggressive pollsters for Cepeda.
4. New polling regulations: Colombia’s National Electoral Council introduced new restrictions in 2025 causing delays in polling publication, meaning available data may be stale [22].
5. Turnout dynamics: If turnout exceeds expectations (above 58%), this could raise absolute vote counts across the board.

Summary Estimate Anchors

Based on historical base rates and current polling:

• Most likely range for first-place candidate votes: ~7.5 million to 9.5 million
• Central estimate: ~8.0–8.5 million votes (comparable to Petro’s 2022 first-round showing, adjusted for modest population growth and similar turnout/vote share dynamics)
• Outside chance of exceeding 10 million only if turnout surges well above 60% or Cepeda significantly outperforms polls

The 2018 and 2022 first rounds provide the strongest reference class, where the first-place candidate received 7.57M and 8.53M votes respectively, with vote shares of ~39–40% and turnout of 53–55%.

FORECASTS

R1: Forecaster 1 Reasoning

Conditional on the first round being held in 2026 and the question not being annulled:

(a) Time left: 20 days until the May 31, 2026 first round; preliminary results should be known that night, with final certified RNEC results likely within several weeks.

(b) Outside view: Recent comparable Colombian first rounds put the first-place candidate at 7.57 million votes in 2018 and 8.53 million in 2022. The broader 2002–2022 range is about 3.76 million to 8.53 million in non-landslide or modern elections, with higher registration pushing the 2026 baseline upward.

(c) Inside view: There are about 41.57 million registered voters. A 54% to 58% turnout, about 96% to 98% valid/non-null ballots, and a first-place share around 34% to 38% imply roughly 7.5 million to 9.0 million votes, with a central estimate near 8.4 million.

(d) If nothing changed: Current polling and turnout expectations would put the first-place candidate around 8.2 million to 8.5 million votes.

(e) If the current trend continued: A mechanical continuation of rising first-place vote totals and electorate growth could point closer to 8.8 million to 9.5 million, but this is partly offset by the current leader’s apparent share being below Petro’s 2022 share.

(f) Experts and markets: Experts broadly expect a runoff, implying no candidate near 50%. Markets and analysts suggest Cepeda is favored to place first in round one, while overall-winner markets reflect second-round dynamics. Turnout markets point mostly to mid-to-high 50s turnout, with some probability of 60%+.

(g) Low unexpected scenario: Violence, a major scandal, or protest/blank-vote dynamics depress turnout and fragment the leading bloc, leaving the top human candidate near 29% to 31% on low turnout.

(h) High unexpected scenario: One camp consolidates late, undecideds break heavily to the leader, and turnout exceeds 60%, pushing the first-place share into the mid-40s or higher.

My forecast anchors on the 2018 and 2022 first-round counts, then adjusts for a larger electorate, similar-to-slightly-higher turnout, and a likely first-place vote share somewhat below Petro’s 2022 level. That leaves my median at about 8.4 million, with meaningful upside if turnout or consolidation surprises upward.

Percentile 0.1: 3,800,000
Percentile 1: 4,700,000
Percentile 5: 5,800,000
Percentile 10: 6,450,000
Percentile 20: 7,150,000
Percentile 30: 7,600,000
Percentile 40: 8,000,000
Percentile 50: 8,400,000
Percentile 60: 8,800,000
Percentile 70: 9,300,000
Percentile 80: 9,900,000
Percentile 90: 10,750,000
Percentile 95: 11,600,000
Percentile 99: 13,350,000
Percentile 99.9: 15,500,000

Agent Forecast (numeric, confidence: low-medium)

• 9000000-10000000: 30.0%
• 8000000-9000000: 28.0%
• 10000000-12000000: 19.0%
• 7000000-8000000: 13.0%
• 12000000-17500000: 4.0%

Base Rate: In 2022, the first-round winner (Petro) received 8,527,768 votes (40.32% of 21.1M valid votes) with 54.91% turnout from 39.0M registered voters. In 2018, winner Duque received 7,569,693 votes (39.14% of 19.3M valid votes) with 53.4% turnout from 36.8M registered. Applied to 2026’s 41.3M registered base: at 40% share × 54% turnout ≈ 8.9M (8M-9M bin), at 44% share × 55% turnout ≈ 10.0M (9M-10M bin), at 37% share × 53% turnout ≈ 8.1M (8M-9M bin border). Central mass concentrates between 8M and 10M.

Initial Reasoning:

• The 2026 Colombian presidential first round is scheduled for May 31, 2026 (20 days away as of today).
• The first-place candidate will almost certainly be Iván Cepeda (Historic Pact, left), who leads in all polls — the most recent and rigorous Invamer poll (April 15-24, 3,800 respondents face-to-face) places him at 44.3%, while other polls show 35-40%.
• Historical base rates: the 2022 first-round winner (Petro) received 8,527,768 votes with 54.91% turnout from 39.0M registered voters; the 2018 winner (Duque) received 7,569,693 with 53.4% turnout from 36.8M registered voters.
• For 2026, there are 41,287,084 registered voters (5.9% more than 2022).
• If turnout is ~55% (historically consistent, matching prediction market’s most likely range of 54-57%), total votes cast ≈ 22.7M.
• Applying Cepeda’s polling range of 38-44% yields a first-place vote count of approximately 8.6M-10.1M.
• The central estimate is ~9.5M (9M-10M bin), accounting for poll uncertainty.
• The 10M-12M bin reflects the possibility the Invamer poll is accurate and turnout is in the 54-57% range.

Key Uncertainties:

• Polling overshoot bias: Colombian front-runner polls historically overstate final share by 2-4 points; Invamer’s 44.3% may translate to 40-42% on election day
• Poll-of-polls vs. single-poll anchoring: averaging Invamer (44.3%) with other polls (35-40%) yields ~39-40%, materially lower than the Invamer central case
• Right-wing consolidation in final 20 days could draw 2-4 points from Cepeda toward a unified opposition candidate
• Turnout volatility: while 53-55% is the historical base, mobilization dynamics could push turnout to 50% (low end) or 58%+ (high-stakes engagement)

Evidence Sources (8):

• Colombia: Informe postelectoral primera vuelta 2022 - CELAG
• IFES Election Guide - Colombian Presidency 2018 Round 1
• Colombian Presidential Elections 2026 – Heading Toward the First Round - Medellin Advisors
• Colombian Presidential Elections: voting intention 2026 - Statista
• Cepeda Surges to 44.3% in Colombia’s Biggest Pre-Election Poll - Rio Times
• Will Colombia’s 2026 First Round Turnout Hit 54-57%? - Lines.com
• In Colombia’s Election, Two Conservatives Fight to Face Cepeda - Americas Quarterly
• Resultados elección Presidente 2022-26 primera vuelta - Colombia en España

Grounding Facts (passed to reviewers):

1. The first round of the 2026 Colombian presidential election is scheduled for May 31, 2026.
2. 41,287,084 Colombians are eligible to vote in the 2026 presidential election.
3. The latest Invamer poll (April 15-24, 2026) shows Iván Cepeda leading with 44.3% of voting intention.
4. In the April 2026 Invamer poll, Abelardo de la Espriella received 21.5% and Paloma Valencia received 19.8% of voting intention.
5. Iván Cepeda’s poll support increased from 37.1% in February 2026 to 44.3% in April 2026.
6. The turnout for the first round of the 2022 Colombian presidential election was 54.91% (21,418,631 votes cast) and 53.40% in 2018 (19,636,714 votes cast).
7. In the 2022 first round, 21,146,287 valid votes were cast, with Gustavo Petro receiving 8,527,768 votes (40.32%).
8. In the 2018 first round, 19,336,134 valid votes were cast, with Iván Duque receiving 7,569,693 votes (39.14%).
9. Prediction markets indicate a 51% probability for a 54-57% turnout range in the 2026 first round.
10. The right’s combined support, as per the April 2026 Invamer poll, is 41.3%.

Draft: 9000000-10000000: 35.0%, 10000000-12000000: 24.0%, 8000000-9000000: 22.0%, 7000000-8000000: 10.0%, 12000000-17500000: 4.5%

Reviewer Panel:

Reviewer (reasoning)

The forecast’s weakest link is its reliance on a single Invamer poll for both vote share and a narrow turnout assumption (~55%), leading to an overconfident central estimate around 9.5 million votes. It ignores polling aggregation, the standard ~40 percent share historical base‐rate for front-runners, and the polling house’s likely overstatement of the lead. As a result, it overweights the 9–10 million bin and underestimates downside risk.

Objections:

• Relies on one poll (Invamer) rather than a poll‐of‐polls, ignoring variation and methodological differences across survey firms.
• Assumes turnout tightly around 54–57%, ignoring tail risks (e.g. 50–60%) driven by mobilization dynamics or blank‐vote mobilization.
• Takes the 44.3% poll share at face value without adjusting for historical bias (front-runner polling overestimates tend to overshoot final vote share by ~2–4 points).

Reviewer (calibration)

The forecast relies heavily on the latest Invamer poll showing Iván Cepeda at 44.3%, which, while recent, might represent a peak in support rather than a guaranteed outcome. Polls can be volatile, and a slight decline from this high or effective consolidation among right-wing candidates could reduce his final vote percentage. The reasoning could benefit from explicitly considering historical poll-to-election accuracy in Colombia.

Objections:

• The forecast may be overly reliant on the peak 44.3% support from the most recent poll, without fully accounting for potential shifts or a plateau in Cepeda’s momentum in the 20 days leading to the election.
• The analysis does not fully explore the impact of potential consolidation among the right-wing candidates. If a significant portion of the 41.3% combined right-wing support coalesces around a single candidate, it could draw more votes away from Cepeda than currently modeled.
• The forecast does not explicitly consider historical accuracy or typical biases of Invamer polls (or Colombian polls generally) when translating voting intention to final election results, which could affect the precision of the percentage applied to total turnout.

Reviewer (fact-checker)

3 unsupported claim(s). Verdict: minor_issues

Unsupported claims:

• Iván Cepeda (Historic Pact, left)
• leads in all polls
• Cepeda’s polling range of 38-44%

Adjustment Rationale:

Adjusted in response to reviewer critiques. (1) Reasoning reviewer correctly noted overreliance on Invamer (44.3%) and that front-runner polls in Colombia historically overshoot by 2-4 points; I now weight a poll-of-polls central case (~39-40%) more heavily, shifting probability from 9M-10M (-0.05) toward 8M-9M (+0.06) and 7M-8M (+0.03). (2) Calibration reviewer correctly flagged peak-support risk and right-wing consolidation downside; I trimmed 10M-12M (-0.05) and added mass to 6M-7M (+0.015) and 5M-6M (+0.005) for downside tail. (3) Turnout uncertainty widened beyond 54-57% by giving more mass to flanking bins. (4) Confidence relabeled ‘low-medium’ to reflect that the distribution should be broader given a 20-day horizon, only one rigorous recent poll, and known polling biases. Central estimate remains ~9M (mode 9M-10M at 0.30) but the 8M-9M bin (0.28) is now nearly co-modal, reflecting poll-of-polls anchoring rather than Invamer-anchoring. Fact-checker’s flags about Cepeda’s affiliation/lead are supported by evidence already in the draft (Invamer poll, multi-poll consistency).

forecast_sdk | 468.0s | research $0.6389 + reviewers $0.0164 + reconcile $0.1582 = $0.8134

Extracted Percentile Values: {0.1: 2900000, 1.0: 5700000, 5.0: 6500000, 10.0: 7000000, 15.0: 7350000, 20.0: 7650000, 25.0: 7900000, 30.0: 8100000, 35.0: 8280000, 40.0: 8440000, 45.0: 8580000, 46.0: 8610000, 47.0: 8650000, 48.0: 8690000, 49.0: 8730000, 50.0: 8780000, 51.0: 8830000, 52.0: 8880000, 53.0: 8940000, 54.0: 9010000, 55.0: 9080000, 60.0: 9350000, 65.0: 9650000, 70.0: 9950000, 75.0: 10300000, 80.0: 10700000, 85.0: 11200000, 90.0: 11800000, 95.0: 12700000, 99.0: 14600000, 99.9: 17600000}

Model’s Answer: Step one is turnout. The strongest hard data here is the RNEC voter roll: 41,421,973 registered voters. March 2026 congressional turnout was materially higher than in 2022, which nudges me toward a somewhat higher presidential turnout than the last cycle rather than lower. A reasonable base case is roughly 51 percent to 56 percent turnout, with the center around 53 percent to 54 percent. That implies about 21 million to 22.5 million total ballots cast.

Step two is converting ballots into candidate votes. This market resolves on votes for the first-place candidate, excluding blank, null, and unmarked ballots. Historically, those excluded categories are meaningful but not dominant. A reasonable base case is that about 92 percent to 95 percent of ballots become candidate votes. That gives a central range of roughly 19.5 million to 21.3 million valid candidate votes.

Step three is the leader’s share. The best recent polling picture is that Cepeda leads clearly but remains well below a likely outright first-round win. The weighted La Silla Vacía aggregation puts him around 38 percent to 40 percent in headline intent, Invamer is higher at 44.3 percent, AtlasIntel lower at 37.4 percent, and markets strongly favor him to finish first while still expecting a runoff. With only about three weeks left, I put most weight on a first-place share of candidate votes around 40 percent to 44 percent, with some chance of a late consolidation pushing a bit higher and some chance of polling error pushing lower.

Multiplying those pieces together gives a center of gravity around 8.3 million to 9.0 million votes for the first-place finisher. My median is a little above the midpoint of that range because the electorate is larger than in prior cycles and participation appears somewhat energized.

The most important assumption is that turnout stays near recent presidential norms, slightly elevated by the stronger March participation. I think there is about a 25 percent chance this is meaningfully wrong. If turnout underperforms, the first-place total falls quickly into the high 6 millions or low 7 millions.

The next key assumption is that the polling picture is directionally right and the leader ends up with about low-40s percent of candidate votes rather than, say, mid-30s or upper-40s. I put about a 20 percent chance on meaningful polling miss or late swing. If undecideds break heavily toward the leader and blank vote is modest, the first-place total can push into the 10 millions; if the race tightens, the top total can be much lower even if the same candidate still finishes first.

The third assumption is that the blank, null, and unmarked share stays within ordinary Colombian presidential ranges. I assign about a 10 percent chance this is materially wrong. More invalid or blank voting lowers the resolving total even if turnout is healthy.

A low outcome would come from lower-than-expected turnout, a tighter race than polls indicate, and somewhat elevated blank or invalid voting. In realistic terms, that means something like 6.5 million to 7.3 million. I think that sort of low-but-not-crazy scenario is around a 10 percent to 15 percent possibility. An extreme low below 3 million would likely require severe disruption or a radically different election environment, so I only place the minimum tail mass there.

A high outcome would come from strong turnout, left consolidation after Murillo’s move, and a late drift of undecideds toward the frontrunner, producing something like 10.5 million to 11.8 million. I think that broad high scenario is also plausible but not the base case. An extreme above 17.5 million would require a near-plebiscitary blowout with very high turnout and unusually low vote fragmentation, which is so implausible that I again only place the minimum required tail mass there.

Overall, my distribution is centered on the high 8 millions, with a fairly wide but still realistic middle, and very thin forced tails outside the practical range.

Percentile .1: 2900000 Percentile 1: 5700000 Percentile 5: 6500000 Percentile 10: 7000000 Percentile 15: 7350000 Percentile 20: 7650000 Percentile 25: 7900000 Percentile 30: 8100000 Percentile 35: 8280000 Percentile 40: 8440000 Percentile 45: 8580000 Percentile 46: 8610000 Percentile 47: 8650000 Percentile 48: 8690000 Percentile 49: 8730000 Percentile 50: 8780000 Percentile 51: 8830000 Percentile 52: 8880000 Percentile 53: 8940000 Percentile 54: 9010000 Percentile 55: 9080000 Percentile 60: 9350000 Percentile 65: 9650000 Percentile 70: 9950000 Percentile 75: 10300000 Percentile 80: 10700000 Percentile 85: 11200000 Percentile 90: 11800000 Percentile 95: 12700000 Percentile 99: 14600000 Percentile 99.9: 17600000

Forecast: 8,578,983 valid votes for the first-place candidate in the 2026 Colombian presidential election.

• Voter Turnout: Expected turnout is projected between 54% and 58%, translating to approximately 22.4 to 24 million total votes cast. This significantly influences the number of valid votes available.

• Candidate’s Share: Iván Cepeda is polling at around 37% to 40% of valid votes, which is crucial for estimating his total. If he maintains this share, it will lead to a higher vote count.

• Right-Wing Vote Dynamics: The potential for a split in the right-wing vote between candidates de la Espriella and Valencia could benefit Cepeda, allowing him to secure a larger share of the votes.

• Late Campaign Effects: Factors such as security issues and endorsements (like the recent Liberal endorsement of Valencia) could impact voter sentiment and turnout in the final weeks leading up to the election.

• Historical Context: Previous elections provide a reference point, with recent first-place vote totals ranging from about 7.6 million to over 8.5 million, suggesting that the current forecast is consistent with historical trends while accounting for a larger electorate and expected turnout.

Overall, the forecast reflects a balanced consideration of current polling, historical data, and potential late campaign shifts.

Forecast (percentiles, aggregated):

• p10: 7200000.0
• p25: 8200000.0
• p50: 9050000.0
• p75: 9900000.0
• p90: 11000000.0

Per-model forecasts:

• 1: p10=7000000.0, p25=8000000.0, p50=9200000.0, p75=10500000.0, p90=11800000.0
• 2: p10=6000000.0, p25=7200000.0, p50=8500000.0, p75=9800000.0, p90=11500000.0
• 3: p10=7100000.0, p25=8200000.0, p50=9300000.0, p75=10400000.0, p90=11500000.0

Rationale: Summary The central estimate of 9.05 million votes rests on a combination of historical turnout trends, an expanded voter roll, and current polling dynamics. Colombia’s electoral roll has grown from approximately 39 million voters in 2022 to 41.4 million for the 2026 election. Assuming a turnout rate of around 55%—consistent with recent presidential elections—the total number of valid votes cast is projected to land between 22 and 23 million. The race is currently a competitive three-way contest, with leading candidate Iván Cepeda generally capturing between 36% and 45% in polls. Applying a roughly 40% to 42% vote share to a base of 22.3 million valid votes yields an expected first-place finish of around 9 million votes. This represents a logical progression from historical benchmarks, where the first-place candidate received 7.6 million votes in 2018 and 8.5 million votes in 2022.

Strongest Arguments for Higher Values

• Surging voter turnout: Some projections and recent congressional trends suggest turnout could be historically high, potentially reaching 23 to 25 million valid votes.
• Candidate consolidation: In a polarized environment, strategic voting could cause the electorate to heavily consolidate behind a single frontrunner. If the leading candidate secures closer to 45% or higher of a large turnout, their total could easily approach or exceed 11 million votes.
• Structural growth: The addition of over 2.4 million registered voters since the last election inherently raises the ceiling for the maximum number of votes a candidate can receive.

Strongest Arguments for Lower Values

• Voter fragmentation: With prominent candidates on the left, right, and center-right, the vote could fracture significantly. If the first-place candidate only secures a narrow plurality (e.g., 33% to 35%), their absolute vote count will be strictly constrained.
• Lower turnout: If voter fatigue or dissatisfaction sets in, turnout could revert to historical lows closer to 45% or 50%, drastically shrinking the pool of valid votes.
• High invalid or blank votes: A polarized but unpopular field could lead to an unusually high number of blank or null votes, which are excluded from the candidate’s valid vote total.

Key Uncertainties

• Final voter turnout rates, which dictate the absolute size of the valid vote pool.
• The degree of late-stage voter consolidation and strategic voting around the leading candidates.
• The proportion of total ballots cast that end up being classified as blank, null, or unmarked.