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

TL;DR

I assign an 80% probability of YES: Rodrigo Paz Pereira remains president continuously through the end of the specified period. His survival of prolonged unrest and recent legislative victories favor continuity (El País). The main danger is economic protest that fractures institutional support; the disputed rules for temporary presidential succession add a separate risk (AP, Constitution).

Context

Paz remains the functioning president. The Presidency’s latest verified dated record shows him promulgating legislation alongside governors and the Senate president on September 30, 2026 (official announcement). I found no verified, institutionally accepted replacement during the qualifying period. That finding is not a complete audit of every succession document.

This is a forecast of continuous officeholding, not successful government or restored popularity. About three months remain. I apply the client’s distinction between ordinary travel or delegation, which does not disqualify Paz, and an actual constitutional transfer of the presidency, which does—even temporarily.

Evidence

The historical baseline favors survival over a short horizon. My retrospective coding of Bolivia’s presidential record from October 10, 1982, through November 8, 2025, finds four abrupt departures: Banzer, Sánchez de Lozada, Mesa and Morales. Adding Siles Zuazo’s negotiated early handover gives five. The exposure is approximately 43.08 country-years, not 43 independent presidencies (presidential chronology, historical analysis, CRS account of the Morales departure). A constant-hazard calculation gives roughly 2%–3% departure risk over the remaining horizon. This is only a starting point: the sample is small, exits cluster in crises, and Paz is already under severe pressure.

His strongest survival-positive evidence is revealed behavior. He survived the earlier campaign demanding his resignation, reported retrospectively as lasting 53 days and causing 22 deaths. The Assembly then approved another emergency extension with 100 votes from 158 legislators present (El País, September 17). Those votes show a working coalition on a survival-critical issue. They do not establish permanent loyalty.

The economic adjustment supplies a fresh trigger. Congress approved the US$1.9 billion IMF agreement on September 18, after which Paz announced the immediate end of diesel subsidies. The program still required IMF Executive Board approval before disbursement in that report (AP, September 19). For customers previously paying the regulated price of Bs9.80 per litre, the new Bs17.95 price represents an 83% nominal increase. Some large consumers had already faced different prices, so the shock was not identical for every user (El Día, September 19). Financing improves the government’s options, but the immediate price increase raises its political risk.

Protest coordination remains incomplete. Reporting on the September 28 transport stoppage found uneven participation and normal operations in La Paz and El Alto; part of that assessment came from the government, so it is not a neutral participation census (Xinhua, September 29). Against that, Santa Cruz transport groups announced further stoppages for October 5–6, Morales threatened mobilization from October 13, and cooperative miners declared readiness to mobilize after their September 30 assembly (transport announcement, EFE, Opinión). These are evidence of mobilization intent, not proof that a unified national removal campaign has formed.

Public support is weak. The August survey published by Ipsos CIESMORI showed Paz at 21% approval and 65% disapproval; Lara was weaker, at 10% approval and 75% disapproval (pollster’s September 5 publication). Reported fieldwork ran August 19–31 among 400 online respondents in La Paz, El Alto, Cochabamba and Santa Cruz (sampling details). This is an urban, internet-connected sample—not an October national reading. I read it as evidence of vulnerability, while Lara’s poor standing weakens the case that succession would provide an attractive alternative.

Security support is more consequential than approval alone. The military commander publicly stated on September 28 that the armed forces would enforce constitutional order and free transit (reported statement). That is affirmative evidence of operational alignment, not proof of private loyalty throughout the forces. Meanwhile, the arrest of Attorney General Roger Mariaca and Lara’s subsequent allegation of a plan to arrest him create another institutional flashpoint. The allegation is not verified (AP, October 1, Europa Press, October 2).

The continuity question requires a separate legal check. Constitution Article 169 provides for temporary vice-presidential assumption during presidential absence, while Article 170 distinguishes events ending the presidential mandate. Recall cannot begin before half the term has elapsed, so it is not an available route during this forecast window (constitutional text). Decree 5675, dated August 13, instead permits remote presidential government and conditions temporary substitution on technological inability and written presidential instructions. It expressly replaced the predecessor decree frequently cited in the dispute (decree text). The conflict between constitutional text and subordinate regulation is real.

But Lara’s assertion did not establish an accepted handover. Paz continued leading virtual cabinet business while abroad, including approval of Decree 5727; Senate President Diego Ávila rejected the claim that he had become president (official broadcaster, September 25, contemporaneous account of Ávila’s refusal). The constitutional challenge remained unresolved in the latest verified court reporting, which also described a substantial backlog (September 24 report).

I translate this evidence into three mutually exclusive forward political scenarios. These are judgmental assumptions, not fitted historical frequencies. Political departure includes resignation, permanent removal, coup and permanent judicial displacement; technical interim succession is handled separately.

The model gives approximately 86% political survival. I then assign 3% risk that a past qualifying interruption is established and, conditional on accepted past continuity and substantive survival, 4% risk of a further recognized interim transfer. A small residual allowance covers death or permanent nonpolitical incapacity. The accounting is:

$$P(\mathrm{YES})=0.97\times(0.50\times0.99+0.35\times0.86+0.15\times0.45)\times0.997\times0.96\approx80\%.$$

The factors represent conditional checks, not four independent risks. The displayed JSON precision preserves the arithmetic; uncertainty in the assumptions is much larger.

What's non-obvious

An adverse constitutional ruling would not automatically establish a past interruption. Article 14 of Bolivia’s Constitutional Procedure Code protects acts performed under a norm while it was presumed constitutional. Court jurisprudence recognizes exceptions, including certain pending or challenged matters (statutory text, official constitutional-court compilation). This does not settle presidential succession. It does mean that “the remote-government decree is unconstitutional” and “Paz already lost the presidency” are distinct propositions. That distinction, alongside the observed institutional non-recognition of Lara’s claim, supports a limited rather than dominant retrospective-disqualification allowance.

The financing and fuel-distribution agreements also deserve less immediate credit than their headlines suggest. The government announced a tanker agreement intended to restore supply, but subsequent reporting still documented a local blockade announcement over shortages (official September 30 account, Red Uno, October 1). I credit the government’s bargaining capacity without assuming supply was already repaired. Paying more while still queuing for fuel is a dangerous combination.

Uncertainties

The largest documentary gap is a complete official travel-and-delegation ledger, together with the constitutional docket and any binding orders governing temporary succession. Those records would sharpen the distinction between a disputed assertion, authorized delegation and legally effective replacement. The statutory protection for past acts is relevant, but its application to this exact officeholding dispute remains untested in the evidence retrieved.

The largest political gap is private security-force and coalition cohesion. Public statements and recent votes show current cooperation, not how those actors would respond to sustained blockades or another lethal confrontation. I also could not verify completed financing disbursement or a nationwide fuel-availability snapshot at the cutoff. The available approval survey predates the latest diesel shock.

Sensitivity is substantial. A stressed model with weaker political survival and larger continuity risks gives 64.6% YES; a favorable model gives 91%. These are assumption tests, not a statistical confidence interval. The central estimate remains 80% because institutional continuity currently outweighs the removal mechanisms, while the economic shock and strict interim-succession rule prevent a near-certain forecast.