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

TL;DR

I assign 87% probability to YES: Rodrigo Paz Pereira remains president continuously through the end of the year, with no established qualifying interruption so far (official presidency). He survived the earlier removal-demand crisis and retains substantial legislative backing (September legislative vote). The main failure route is renewed national unrest combined with elite or security-force abandonment; the strict continuity rule also makes a legally effective interim succession a separate risk (constitutional succession provisions).

Context

At the supplied October 7, 2026, 03:04 UTC cutoff, Paz remains the officially identified president. The Presidency also documents his exercise of presidential functions at an October 5 ceremony, rather than merely listing him on an undated biography page (official October 5 report). The client’s dates leave approximately 86 days until resolution. I assess future survival separately from uncertainty about whether an interruption has already occurred since the question opened.

The government faces another test after ending the remaining diesel subsidy. External financing and legislative cooperation strengthen its position, while adjustment costs, executive infighting, and renewed mobilization threaten it. This is a vulnerable presidency, not an executive whose institutional support has already collapsed (AP, September 19).

Evidence

The historical backbone is Bolivia’s democratic-era presidential record, not its earlier military-government period. My reference window runs from October 10, 1982, through November 8, 2025: approximately 43.08 country-years. I count four clear unplanned departures—Banzer’s illness-related departure, Sánchez de Lozada’s resignation, Mesa’s resignation, and Morales’s resignation—and treat Siles Zuazo’s negotiated early departure as a fifth, definition-sensitive case (presidential chronology; CRS historical account; CRS on Morales’s departure). This is a retrospective classification with only four or five events, not a large statistical sample.

A constant-hazard conversion of that record gives roughly 2–3% exit risk over the remaining interval. For the four-event classification, my calculation is $1-\exp[-(4/43.08)(86.04/365.25)]$, using events per country-year and the remaining horizon in years. That is a starting benchmark, not the forecast: crises cluster, and the current presidency is riskier than an average president-day. The strongest historical warning is the combination of street pressure and loss of coercive support. During Morales’s departure, the military recommended resignation after sustained protests (CRS, January 7, 2020).

Paz has already survived a much sharper confrontation. The earlier resignation-demand protests lasted 53 days. On September 17, legislators approved another 90-day emergency extension with 100 votes supporting it and 58 opposing or abstaining. That demonstrates support for a central government measure, not unconditional loyalty (EL PAÍS, September 17). The decree extends emergency powers into mid-December, not through the resolution deadline (Decree 5707).

Security leadership remains publicly aligned with enforcement of the existing order. On September 28, Armed Forces commander Víctor Hugo Balderrama said military and police personnel were deployed along roads to prevent blockades. This is affirmative evidence of reported operational deployment; it does not prove how every unit would behave during a violent confrontation (September 28 reporting). International support also favors continuity: the OAS’s September 3 mission report called for respecting elected mandates and constitutional mechanisms. I read that as raising the diplomatic cost of irregular displacement, not protecting Paz against resignation (OAS report).

Financing has progressed beyond an announcement. The IMF approved an approximately US$1.9 billion, 36-month arrangement on October 2, with later disbursements conditional on reviews. The central bank confirmed receipt of a single initial disbursement of US$211.3 million on October 6. These are nominal financing amounts; the full facility is not immediately available cash (IMF approval; BCB confirmation reported by Bolivia TV). My inference is that this buys time against foreign-exchange stress. It does not remove the political cost of fuel reform.

The latest observed inflation release is less alarming than an undated annual forecast would suggest. September consumer prices rose 0.85% month over month; cumulative inflation from January through September was 3.88%. These are different measurement windows, reported from the official release in October, not forecasts of subsequent adjustment costs (EFE, October 6). I do not assume those readings capture the full future diesel-price pass-through or households’ accumulated loss of purchasing power.

Public support is weak. The September Ipsos CIESMORI snapshot gives Paz 21% approval and 61% disapproval, while Lara has 11% approval and 74% disapproval. The publisher released the assessment on October 3. Fieldwork ran September 18–28 among 400 internet-connected adults in La Paz, El Alto, Cochabamba, and Santa Cruz, using self-administered online questionnaires (pollster’s publication; sample and fieldwork details). This is not a national probability sample. It establishes serious urban dissatisfaction, but neither a national removal mandate nor an attractive successor.

The latest protest signals point in both directions. On October 5, Evo-aligned organizer Aquilardo Caricari announced mobilization beginning October 13 but excluded road blockades. That is an organizer’s stated plan, not a binding commitment from every participating group (October 5 interview). Santa Cruz urban transport leaders also suspended their planned stoppage after negotiations, while retaining an ultimatum (Red Uno, October 3). Conversely, organizer Nelson Virreira’s October 6 detention creates a fresh grievance as well as disrupting an opposition network. I do not count the arrest as an unambiguously stabilizing event (ERBOL, October 6).

Elite cohesion is another weakness. Cabinet disruption and the Cerimedo adviser scandal were documented in August; former Presidency Minister José Luis Lupo then declined the OAS appointment in an October 6 letter. The letter signals estrangement but does not establish a coordinated replacement coalition (AP, August 21; October 6 letter reporting). Ordinary recall is unavailable this early in the mandate, and a presidential criminal trial requires prosecutorial action and authorization by two-thirds of legislators present. Discontent is not equivalent to a parliamentary no-confidence mechanism (Constitution, Articles 184 and 240).

The September travel dispute deserves separate treatment. Lara claimed presidential substitution, and other officials endorsed a succession interpretation. But Ávila explicitly denied formally assuming the presidency and affirmed that Paz remained president in exercise (September controversy; Ávila’s denial). I distinguish those competing assertions from an operative constitutional transfer. Future exposure is concrete: the government announced Paz’s planned attendance at the November 4–5 Madrid summit. Attendance remains a plan, not an accomplished trip (official July 21 announcement).

My remaining-period scenario model uses the following judgmental assumptions. The percentages are rounded for display; the equation retains the calculation inputs. They are not fitted historical frequencies.

The weights reflect demonstrated survival, financing, and sectoral negotiations, balanced against adjustment costs and weak cohesion. I separately allow 2% for an already-occurring qualifying interruption or a later authoritative finding establishing one; 3% conditional risk for a future technical/legal transfer outside political collapse; and a small residual allowance for death or lasting incapacity. Defined sequentially, these avoid assuming that the pathways are independent or counting the same removal twice:

$$P(\mathrm{YES})=0.98\times(0.60\times0.985+0.28\times0.91+0.12\times0.60)\times0.97\times0.997.$$

This produces the reported 87% YES forecast. The extra digits in the probability field preserve arithmetic, not political certainty.

What's non-obvious

Disliking Paz does not mean supporting renewed paralysis or his successor. In the same September online sample, 53% supported extending emergency powers and 40% opposed it, despite weak presidential approval. El Alto respondents went the other way, which limits the reassuring interpretation (ANF, October 5, including methodology). My reading is that dissatisfaction and willingness to join a removal campaign are different things. Opposition coordination matters more than approval alone.

The legal story is also more subtle than the headlines. Decree 5675 expressly repealed the older remote-governance decree but reproduced the mechanism in Article 22. The replacement text directs remote exercise abroad and provides substitution when technological communication is impossible and the president communicates it expressly (operative decree text). The presumption of constitutionality supports application pending adjudication; it does not settle the constitutional merits (Constitutional Procedural Code, Article 4). This makes ordinary travel different from established succession, while leaving a real continuity risk under the client’s unusually strict rule.

Uncertainties

The largest gap is a complete official succession and court-record audit through the exact cutoff. The latest substantive court-status reporting I verified, published September 30, still described the remote-governance dispute as pending. That does not certify that no later ruling or uninspected transfer instrument exists (September 30 court-status report). Signed succession instruments, complete case dispositions, and authoritative clarification of the September episode would narrow this uncertainty more than another opinion article.

The other gaps are unit-level security loyalty, confirmed cross-sector participation in the next mobilization, and representative national opinion. Public deployment statements and organizers’ announcements are informative but incomplete. Reasonable downside and upside parameter sets produce approximately 76–94% YES; this is model sensitivity, not a statistical confidence interval. Verified security refusals, coordinated elite abandonment, or a binding succession decision would sharply lower the estimate. Continued sectoral settlements and institutional cooperation would raise it.