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

TL;DR

I put YES at 83%. Paz survived a 53-day national blockade crisis and security forces still execute his orders, while the late-August protests remain far smaller and lack a common plan for succession (June crisis, current blockades). The main NO path is a September cascade in which the diesel conflict broadens, the Cerimedo investigation directly implicates Paz, and political or security elites defect.

Context

Bolivia’s official presidency site still lists Rodrigo Paz Pereira as president and says he assumed office on November 8, 2025; I found no official indication of a qualifying interruption since the question was submitted (Presidency of Bolivia). About 125 days remain before the resolution deadline.

Paz is weak rather than terminal. Most of his original cabinet has turned over, his former ally Samuel Doria Medina has moved into opposition, his vice president is openly hostile, and Congress forced out his economy minister (AP, Doria Medina’s rupture, ministerial dismissal). Yet there is no advanced constitutional removal proceeding, agreed successor coalition, or observable security-force break.

Evidence

The historical baseline favors survival. From democratic restoration on October 10, 1982 through August 28, 2026, I count five defensible early exits: Hernán Siles Zuazo’s negotiated shortening, Hugo Banzer’s health resignation, and the crisis departures of Gonzalo Sánchez de Lozada, Carlos Mesa, and Evo Morales (Georgetown presidential chronology, BTI Bolivia report). That is five events over 43.9 country-years. A constant-hazard conversion gives an unconditional 125-day exit rate of:

$$1-\exp\left[-\left(\frac{5}{43.9}\right)\left(\frac{125.4}{365.25}\right)\right]=3.8\%.$$

Counting only political or protest-related exits lowers it to about 3.1%. Paz’s present danger is clearly several times the unconditional baseline, but a normal short-horizon forecast would start around 96%–97% YES.

The broader regional record gives the same warning without justifying a low forecast. One study found 11 failed presidencies among 44 elected South American terms from 1978 through 2005, but that is a full-term rate, not a four-month rate (Hochstetler and Edwards). Comparative research identifies recession, protest and political radicalization as drivers of presidential falls, while institutional and military alignments shape whether a crisis actually removes the president (Pérez-Liñán and Polga-Hecimovich). I therefore use the history mainly to identify the mechanism: protests become terminal when political and security elites coordinate against the incumbent.

The strongest survival evidence is revealed behavior during May and June. The crisis lasted 53 days, reached roughly 90 blockade points across seven departments, and produced a final ombudsman-based tally of 22 deaths, 88 injuries and 583 arrests (conflict tally). Paz neither resigned nor lost control of the security apparatus. He signed an agreement with the COB on June 19, imposed a 90-day state of exception, and by June 23 the national road network was reported transitable (government–COB agreement, road reopening, Decree 5636). This shows high presidential pain tolerance and enough coercive capacity to survive a much larger challenge than the current one.

The late-August diesel conflict is serious but not yet comparable. Decree 5676 raised diesel for large-volume buyers from Bs 9.80 to Bs 18 per liter, about 84%, and Paz says it may be modified but not repealed (policy and government position). By August 28, producers maintained three blockade points on the Santa Cruz–Beni corridor; the previous day’s police-military intervention left at least five injured and two detained (Urgente.bo). The agricultural sector gave the government until August 30 to repeal the decree and threatened to ask the Comité pro Santa Cruz to convene an Asamblea de la Cruceñidad (El Deber). This raises the probability of another national protest cycle, but three localized blockades are still far from the May–June peak.

The organized labor signal is mixed and less alarming than the headlines. COB leader Mario Argollo rejected the diesel policy and an extension of emergency rule, but also acknowledged that the earlier blockades lacked public support and explicitly said the COB’s current position was not a call for sedition or overthrow (COB statement, admission on public support). The Comité pro Santa Cruz likewise described its goal as negotiation rather than war with the government (Red PAT). A COB national assembly expected in September remains a real escalation point, especially near the state of exception’s scheduled September 18 expiry.

Security-force behavior still favors Paz. Police and military personnel jointly attempted to clear the new blockades, direct evidence of operational compliance at the unit level (road-clearing operation). Government and PDC legislators are also considering seeking legislative approval to extend emergency rule, although opposition lawmakers are resisting (extension proposal). The OAS reiterated support for Paz’s constitutional government after its July mission, raising the diplomatic cost of an unconstitutional transfer (OAS, July 30).

Congress is a source of fragility, but its actions do not yet imply presidential removal. Paz’s PDC won 49 of 130 deputy seats and 16 of 36 senate seats, or 65 of 166 ordinary legislative seats, leaving it well short of a majority (IPU election record, official-result summary). Congress nevertheless assembled 91 votes out of 129 members present to censure Economy Minister José Gabriel Espinoza (censure vote). That proves anti-government coordination is possible. Removing a minister is still much easier than forcing a president through resignation or a final criminal judgment.

The Cerimedo scandal is the fastest-moving downside risk. Fernando Cerimedo faces preventive detention and investigations involving attempted femicide, alleged illicit enrichment and alleged protection of narcotics-linked flights (El País). On August 28 he identified Paz as his main contact in the presidential environment, while declining to answer many questions about their communications (Urgente.bo). This undercuts Paz’s attempt to portray the relationship as distant. It still does not constitute public evidence that Paz participated in Cerimedo’s alleged crimes.

The legal-removal route remains slow. Private citizens filed an action seeking to compel the legislature to consider a juicio de responsabilidades, but this is a request for institutional action rather than a prosecutorial accusation or authorization for trial (Infobae). The Constitution requires resignation, death, definitive incapacity, recall or a final criminal conviction for the mandate to end, and presidential trials require prosecutorial, judicial and supermajority legislative stages (Bolivian Constitution, Articles 169–170 and 184, TSJ procedure). A scandal-induced resignation is plausible by December; completion of an ordinary legal process is not.

Public opinion is bad but offers no popular successor. An online survey of 400 internet users in La Paz, El Alto, Cochabamba and Santa Cruz, fielded July 23–August 2, put Paz at 32% approval and 55% disapproval. It put Vice President Edmand Lara at 12% approval and 75% disapproval (Ipsos CIESMORI report summary and methodology). The urban online sample is not nationally representative. It nevertheless shows steep deterioration while also suggesting that Lara is a weak focal point for a broad succession coalition.

The economy provides modest breathing room but no clean escape. Bolivia’s July consumer-price index fell 2.79% month over month, putting twelve-month inflation at 4.93%; the statistics institute attributed much of the decline to restored supplies after the blockades (INE, August 4 release). IMF staff reached a 36-month, roughly US$1.9 billion agreement on July 29, but it still requires prior actions and Executive Board approval (IMF). The diesel reform is tied to that adjustment program, so the prospective economic stabilizer and the immediate protest trigger are two sides of the same policy.

I combined three models. A stress-adjusted historical model multiplies the 3.8% baseline hazard by 4.5 and gives 83.9% YES. A scenario model gives 52% weight to containment, 29% to renewed national unrest with loyal security forces, 14% to a broad protest-and-elite crisis, and 5% to a direct shock or security split; it produces 81.1% YES. An institutional model emphasizing the short horizon, legal barriers, weak successor and continued coercive compliance gives 86.0% YES. Weighting them 35%, 45% and 20% gives:

$$0.35(0.839)+0.45(0.811)+0.20(0.860)=0.831.$$

What's non-obvious

The current anti-Paz coalition does not yet exist as a single coalition. Groups demanding that Decree 5676 be repealed want different end states: some want subsidized diesel restored, while exporters and industrial groups favor wider market liberalization (El Post). The CAO’s fight is currently policy-specific. The COB says it is not calling for an overthrow. Lara is hostile but unpopular. Doria Medina calls himself constructive opposition, while Paz and Quiroga’s Libre have simultaneously opened the door to bargaining over economic legislation (Doria Medina, Paz–Libre talks).

The Cerimedo case cuts the other way. It is more dangerous than a routine adviser scandal because Cerimedo acknowledged direct presidential access, prosecutors are investigating money and alleged narcotics links, and the hostile vice president has made the case central to his attack on Paz. The case becomes presidency-threatening if authenticated records connect Paz or his family to criminal conduct, or if it gives otherwise divided blocs a common moral case for succession.

Uncertainties

• There is no reliable public information about private sentiment inside the police and military commands. Continued deployments show compliance today, not guaranteed loyalty through December.
• The latest detailed approval survey ended on August 2 and covered only 400 urban internet users, so it misses both the Cerimedo scandal and the current diesel dispute (survey methodology).
• The public does not know what evidence prosecutors or Nadia Beller possess, whether Cerimedo will cooperate, or whether financial records connect the alleged conduct to the presidential circle (current investigation).
• Four near-term decisions could move the forecast quickly: the Comité pro Santa Cruz’s response after the August 30 ultimatum, the COB’s expected September assembly, legislative action on extending emergency rule, and submission and approval of the IMF program (Santa Cruz deadline, COB plans, emergency-extension debate, IMF agreement).

My subjective 80% uncertainty range is 72%–89% YES. The center remains above four in five because the decisive collapse indicator—coordination among a national protest movement, political elites and the security forces—has not appeared.