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

TL;DR

Rodrigo Paz Pereira has an 86% chance of remaining continuously president through December 31, 2026. He already survived the larger May–June resignation campaign, while the military and the new police command still support constitutional order (El País; Los Tiempos; El Deber). The main failure path is a new fuel-driven protest wave that broadens, causes deaths, and triggers elite or security-force defection.

Context

Paz was still exercising presidential authority at the cutoff of September 1, 2026, 19:00 UTC. An official August 31 report recorded him meeting municipal representatives from all nine departments, leaving about 121 days until resolution (Bolivian Presidency).

His government is weak but functioning. Fresh threats include fuel protests scheduled for September 2 and 4, the Cerimedo scandal, an openly hostile vice president, a fragmented legislature, and heavy cabinet turnover. None had produced a legal vacancy, a formal prosecutor-backed presidential trial, or a security-force rebellion by the cutoff (Los Tiempos; Unitel; ChequeaBolivia).

Evidence

The historical baseline favors survival. I reconstructed every presidential spell since democratic government returned on October 10, 1982, using the Georgetown Political Database, the Archigos leadership dataset, and later public records. Siles Zuazo’s 1985 handover is classified as an early departure because political pressure shortened his mandate by roughly one year (Educa); Banzer’s 2001 cancer resignation also counts under the question’s broad criteria (Los Angeles Times).

Conditioning on having already survived Paz’s first 297 days leaves 12,239 historical president-days of comparable exposure and five qualifying exits. A constant-hazard model gives about a 5% neutral departure risk over the remaining four months; a Jeffreys-prior Gamma–Poisson model also gives about 5%. Excluding Siles’s unusual, pre-agreed early election lowers the estimate to about 4%. These are small-sample baselines, not a forecast of today’s crisis.

Bolivian research shows why the base rate must be adjusted upward. Protests can topple presidents when repression backfires and street mobilization is joined by elite, police, or military abandonment; protest alone is less decisive (Journal of Latin American Studies). Paz survived more than 50 days of blockades explicitly demanding his resignation. The government split the coalition, declared an emergency, deployed police and troops, and had reopened national roads by June 23 (El País). That is strong revealed evidence that he will not resign merely because protests are prolonged or unpopular.

The full weekly-updated national demonstration series from HDX/ACLED covers September 2025 through August 2026 and counts recorded protest or riot events, not participants (HDX methodology).

The data shows a sharp fall after May, but it deserves limited weight. January was higher than May, and the low June count conflicts with reporting that blockades continued for most of that month. I use the series as evidence that July and August were calmer, not as a precise measure of protest strength.

Security-force behavior favors YES. On August 7, armed-forces commander Víctor Hugo Balderrama said the military would defend the stability of the democratically elected government (Los Tiempos). Police commander Mirko Sokol’s August 31 resignation and the simultaneous command reshuffle are warning signals. Yet Paz replaced the leadership within hours, and incoming commander Adrián Álvarez pledged on September 1 to enforce the state of exception while prioritizing dialogue (El Deber). A viral claim of a police mutiny was fabricated from 2019 imagery (ChequeaBolivia).

Formal removal is a secondary risk. Recall cannot begin until half the five-year term has elapsed (Bolivian Electoral Authority). The criminal route requires a reasoned request from the prosecutor general, authorization from at least two-thirds of legislators present, a full Supreme Court trial, and a final criminal sentence (Bolivian Constitution, Articles 170, 184 and 240). The August 27 Cerimedo-related action came from three private applicants and explicitly represented their legal allegation rather than a judicial finding or prosecutor-backed accusation (Unitel). Completing that process by December 31 is unlikely. Resignation under pressure remains the main legal exit pathway.

The September protest risk is real, but the coalition was not yet unified. San Julián planned a blockade for September 2; Morales-aligned Cochabamba organizations planned a September 4 march; transport, agricultural, civic, and La Paz campesino groups were considering separate actions (Los Tiempos). Their immediate demands centered on repealing or changing the diesel decree and addressing shortages, not on a common, unconditional demand for Paz’s resignation. The decree set a Bs 18 reference price for bulk buyers while leaving direct vehicle-tank purchases at the previous subsidized price (Bolivian Presidency). A threatened tanker strike was also cancelled after an August 29 agreement (Economy Ministry). This shows that sector-by-sector bargaining still works.

The negative political signals keep the forecast below 90%. Only six of Paz’s original 15 ministers reportedly remained by August 21, his economy minister was forced out after a legislative censure, and Cerimedo’s arrest tied a rapidly widening criminal scandal to the presidential circle (Associated Press). An Ipsos CIESMORI online poll, fielded July 23–August 2 among 400 adults in La Paz, El Alto, Cochabamba, and Santa Cruz, reported 32% approval and 55% disapproval for Paz. Lara was weaker at 12% approval and 75% disapproval (Visión 360; Ipsos CIESMORI). The survey is small, urban, and internet-only, so its direction matters more than its exact level.

Economic evidence is mixed. The official July CPI, published August 4, fell 2.79% month-on-month and stood 4.93% higher year-on-year because food distribution normalized after the blockades (INE CPI release). The July producer-price index, published August 27, rose 5.09% month-on-month and 10.61% year-on-year, partly reflecting the new exchange-rate regime (INE producer-price release). The July 29 IMF agreement covers US$1.9 billion over 36 months, but it was still a staff-level agreement subject to prior actions and Executive Board approval (IMF). It is a potential relief valve, not money already available.

I combine two models. An adjusted historical model raises the normal Bolivian hazard by roughly three times for the active protest threats, low approval, cabinet churn, hostile vice president, scandal, and economic stress; that gives about a 14% departure risk. A scenario model assigns about a 42% chance of another broad national crisis, a 30% exit risk if that crisis occurs, and about a 4% exit risk otherwise; it gives about a 15% departure risk. Their blend produces 14% NO and 86% YES.

What's non-obvious

The police reshuffle cuts both ways. Losing Sokol and much of the command structure during a scandal is not normal turnover. But it also let Paz install an interim command that immediately pledged to enforce the emergency rules. The leaked recording promoted by Lara allegedly shows the first lady fearing that senior security officials could “turn” against Paz, but its speakers and context had not been independently authenticated (Brújula Digital). I treat it as a reason for wider uncertainty, not as proof of an imminent coup.

Paz’s likely successor makes removal harder to coordinate. Lara benefits directly from a vacancy, but the limited urban poll puts him far below Paz. Conservative legislators, business groups, and military leaders can oppose Paz while still preferring him to a Lara presidency. That helps explain why cabinet censures and harsh criticism have not yet become a presidential removal coalition.

Uncertainties

• The most important hidden variable is the real willingness of police and military commanders to confront another large national blockade campaign. Public ceremonies and statements show current obedience, but not how they would react after major casualties.
• The September 2 and 4 actions were still future events at the cutoff. Their turnout, geographic spread, duration, demands, and ability to attract urban residents were unknown (Los Tiempos).
• The Cerimedo evidence was moving quickly. Direct, authenticated financial or documentary evidence against Paz or his immediate family would materially lower the forecast; allegations against an informal adviser alone would not.
• There was no high-quality, nationally representative approval poll. The available 400-person online urban survey does not adequately measure rural organizations that can close Bolivia’s roads.
• IMF approval, legislative authorization, actual disbursement timing, and the decision on extending the state of exception were unresolved. Each could change fuel availability and the government’s capacity to manage unrest.