Will Bolivian president Rodrigo Paz Pereira remain in power through the end of 2026?

TL;DR

I put the chance of YES at 82%. Paz stays favored because he survived the stronger May–June blockade crisis, the security hierarchy still follows the established command chain, and his opponents lack a shared successor. The main route to NO is a post-emergency fuel crisis that unites labor, Evo-aligned groups, Santa Cruz and security defectors before December 31.

Context

At the September 11 cutoff, Bolivia’s official presidency page still identified Rodrigo Paz Pereira as president, and current reporting showed him exercising executive authority. He had therefore survived continuously since the question opened on June 5, leaving 112 days and 45 minutes until resolution.

The question opened during the most dangerous part of Paz’s tenure. The confrontation lasted about 53 days, involved widespread blockades and explicit resignation demands, and led Paz to declare a 90-day national state of exception. The blockade campaign ended without presidential succession, but its economic causes remain active.

Evidence

The broad historical prior is low. A comparative dataset covering 100 Latin American presidential elections from 1980–2010 records 13 interrupted presidencies. Spreading that risk uniformly over an illustrative four-year term produces only about a 1% interruption probability over the remaining window. Bolivia is riskier: academic research finds that protests helped push four presidents out before their elected terms ended—in 1985, 2003, 2005 and 2019. Treating those as four events over roughly 44 democratic-era calendar years gives a crude 112-day political-exit prior near 3%.

Those priors cannot be used mechanically. Paz has the warning signs associated with presidential failure in the literature: mass protest, executive scandal, recession, low approval and weak legislative protection. But the successful Bolivian removal episodes also involved a deeper convergence of street pressure, elite abandonment and, most clearly in 2019, police or military rupture. That last threshold has not been crossed. Research on regional presidential failure likewise finds that protest becomes far more dangerous when combined with legislative vulnerability and executive wrongdoing rather than operating alone. (South American presidential-failure research; Bolivian historical analysis)

The full ACLED/HDX demonstration series used here covers January 2025 through the September 2026 cutoff, with 2,199 coded demonstrations. September is partial through September 11. These are event counts, not crowd totals or measures of how strategically important a blockade was. (ACLED datasets on HDX)

The series shows de-escalation after May, but it also shows why raw protest counts have limited predictive value. January produced more coded events than May without threatening immediate succession, while June’s count fell even though the blockade confrontation continued for most of that month. The data supports “no new national surge by the cutoff,” not “the crisis is over.” The Defensoría del Pueblo similarly found that exceptional measures rapidly reduced roadblocks but did not resolve fuel, living-cost and labor grievances.

Near-term risks are concentrated in September and October. The state of exception expires around September 18–20, although published accounts disagree on the precise legal date and no extension was confirmed at the cutoff. Evo-aligned groups scheduled a September 26 national meeting; producers had already installed September roadblocks; and Santa Cruz mototaxi organizations announced indefinite action beginning September 11. Santa Cruz’s broader civic assembly gave the government until October 10 to respond, but its president explicitly said it was not demanding Paz’s resignation.

Paz’s political base is weak. An Ipsos CIESMORI poll fielded August 19–31 among 400 internet-connected adults in La Paz, El Alto, Cochabamba and Santa Cruz reported 21% approval and 65% disapproval. The sample is urban, online and not nationally representative, but the decline is large enough to treat as real directional evidence. By August 21, only six of Paz’s original 15 ministers remained, and the PDC had split into three competing factions. One of those factions nevertheless reaffirmed support for democratic stability and Paz’s governability.

The constitutional routes are much weaker than the street route. Bolivia’s Constitution makes Vice President Edmand Lara first in succession and prevents recall until at least half the presidential term has elapsed. A private action has sought to force a responsibility trial over the Cerimedo scandal, but it was not a prosecutorial accusation, a two-thirds legislative authorization or a final conviction. Completing that chain before year-end remains remote. (Cerimedo-related action)

Security institutions remain the main buffer. Police commander Mirko Sokol resigned amid an investigation into leaked recordings, but Paz appointed a new command the same day. More revealingly, when Lara challenged an Army officer at Viacha, the defense minister and Army leadership backed the officer and the regular command structure. This does not prove enduring loyalty, but it shows that Lara cannot currently turn his position as constitutional successor into control over the security apparatus.

The economy supplies both risk and a modest stabilizer. Bolivia’s statistical institute reported August inflation of 1.10% month over month, 3.01% year to date and 5.02% year over year, covering prices through August 31 and published September 3. Fuel shortages remain politically sharper than that national CPI snapshot. The IMF’s US$1.9 billion, 36-month staff-level agreement offers financing but still requires Executive Board approval and can itself trigger resistance to reform.

My scenario model for the remaining period is:

The resulting political-exit probability is:

$$P(\mathrm{NO}_{political})=(0.40\times0.02)+(0.36\times0.11)+(0.18\times0.40)+(0.06\times0.90)=0.1736.$$

I reserve a separate 0.3% probability for death or definitive incapacity. Combining rather than simply adding the risks gives:

$$P(\mathrm{YES})=(1-0.1736)(1-0.003)=0.8239208.$$

What's non-obvious

The number of actors angry with Paz overstates the size of a removal coalition. Evo-aligned organizations want early elections and renewed political influence. Labor wants economic concessions. Santa Cruz wants fuel, resources and autonomy. Conventional opposition leaders gain little from automatically installing Lara. The same limited urban poll that put Paz at 21% approval put Lara at 10% approval and 75% disapproval. A coalition can agree that Paz is failing while still rejecting the legal consequence of his resignation.

The Viacha confrontation is more informative than generic declarations of military loyalty. Lara tried to assert direct authority over an officer, and the high command publicly defended the officer’s adherence to the established chain. Combined with Santa Cruz’s explicit rejection of a resignation demand, this shows that today’s unrest has not yet fused into the cross-institutional alignment that ended Bolivian presidencies in 2003, 2005 and 2019. (Viacha response; Santa Cruz position)

Uncertainties

• The precise end date and any extension of the state of exception were not settled in the available record. Reports give dates from September 17 through September 20, while the underlying decree states 90 calendar days from publication. (Decree 5636; extension analysis)
• ACLED’s recent data is revised and counts separate events, not the duration, crowd size or strategic leverage of roadblocks. Contemporary HDX snapshots also differed by two events for August, so 34 should not be read as exact beyond the current vintage.
• The approval poll has only 400 online respondents in four urban areas. A current nationally representative rural-and-urban poll would materially improve the estimate. (Poll methodology and results)
• Several Cerimedo-related recordings and corruption allegations remain unauthenticated or unadjudicated. A verified recording directly implicating Paz, a formal Fiscal General accusation, or a two-thirds authorization vote would sharply reduce the forecast. (Current legal action; prosecutorial treatment of related allegations)