Political Survival Analysis
Also known as: Leader Survival Analysis, Government Duration Analysis, Selectorate Survival Model, Political Event-History Analysis
Political survival analysis applies survival and event-history models to the time leaders, governments, and regimes remain in power before failing. Methodologically it rests on the hazard-modeling apparatus codified for social scientists by Box-Steffensmeier and Jones in 2004 — the Cox proportional-hazards model and parametric alternatives such as the Weibull, estimated on duration data with censoring. Substantively it is anchored in the selectorate theory of Bueno de Mesquita, Smith, Siverson, and Morrow's 2003 The Logic of Political Survival, which explains the hazard of losing office in terms of the size of the winning coalition (W) and the selectorate (S). The model links institutional structure and performance to the risk that an incumbent's tenure ends.
Key highlights
- Models the timing and conditional risk of losing office directly, handling right-censoring that ordinary regression cannot accommodate.
- The Cox model estimates covariate effects without committing to a baseline-hazard shape, robust to unknown duration dependence.
- Accommodates time-varying covariates, so changing economic conditions and coalitions can be linked to changing survival risk.
- Connects cleanly to selectorate theory, giving institutional structure (W, S) an estimable role in explaining political durability.
Intuition
This section is available to Pro members. Upgrade to Pro
How it works
This section is available to Pro members. Upgrade to Pro
When to use it
Use political survival analysis whenever the outcome of interest is the timing of an exit from power — how long leaders, governments, cabinets, or regimes last and what raises or lowers their risk of falling. It is the appropriate framework when you have duration data with a well-defined start, a failure event, and right-censoring, and when you want to model duration dependence or the effect of time-varying conditions. It is especially apt for testing selectorate-theory predictions linking coalition and selectorate size to durability. The approach is less suitable when the outcome is a one-shot binary event with no meaningful time dimension, when spells are too few or too heavily censored to estimate hazards reliably, or when the exit process is better modeled as competing risks that must be distinguished rather than pooled into a single failure event.
Strengths & limitations
- Models the timing and conditional risk of losing office directly, handling right-censoring that ordinary regression cannot accommodate.
- The Cox model estimates covariate effects without committing to a baseline-hazard shape, robust to unknown duration dependence.
- Accommodates time-varying covariates, so changing economic conditions and coalitions can be linked to changing survival risk.
- Connects cleanly to selectorate theory, giving institutional structure (W, S) an estimable role in explaining political durability.
- The Cox model assumes proportional hazards; when covariate effects vary with tenure the assumption fails and must be corrected.
- Measuring the winning coalition and selectorate empirically is contested and relies on coarse institutional proxies.
- Unmeasured heterogeneity (frailty) across leaders or countries can bias duration-dependence estimates if ignored.
- Treating all exits as a single failure event conflates distinct processes — voluntary departure, electoral defeat, coups — that may have opposite covariate effects.
Common pitfalls
This section is available to Pro members. Upgrade to Pro
Applications
This section is available to Pro members. Upgrade to Pro
Frequently asked
Why use survival analysis instead of a logit on whether a leader was removed?
A logit on removal ignores how long the leader survived and cannot handle censoring — spells that are still ongoing or that end for unrelated reasons. Survival analysis models the hazard, the conditional risk of failure given tenure so far, using the timing of exits and correctly incorporating censored spells. This captures duration dependence (whether risk rises or falls with time in office) and lets time-varying conditions enter naturally, none of which a static binary model on the full period can represent without distortion.
How does selectorate theory explain who survives in office?
Selectorate theory turns on the winning coalition W (supporters the leader must keep) and the selectorate S (the pool they are drawn from). When W is small relative to S, supporters fear replacement and being shut out of private rewards, so loyalty is high and incumbents are durable — the classic autocrat. When W is large, the leader must provide public goods and good performance to retain support, and poor performance raises the hazard of removal. In survival analysis these enter as covariates, so the institutional ratio W/S and performance jointly shape the estimated hazard of losing office.
What is the proportional-hazards assumption and why test it?
The Cox model assumes each covariate multiplies the baseline hazard by a constant factor that does not change over time — covariate effects are proportional across the spell. If, say, economic performance matters more late in a tenure than early, this assumption fails and the single hazard ratio is misleading. The standard check correlates scaled Schoenfeld residuals with time; a significant relationship flags a time-varying effect, which is handled by interacting the covariate with time or stratifying. Testing it is essential because violations bias both estimates and inferences about durability.
Sources
- 1.Bueno de Mesquita, B., Smith, A., Siverson, R. M., & Morrow, J. D. (2003). The Logic of Political Survival. MIT Press.ISBN 9780262025461
- 2.Box-Steffensmeier, J. M., & Jones, B. S. (2004). Event History Modeling: A Guide for Social Scientists. Cambridge University Press.ISBN 9780521546737
You have read it. What now?
Cite this page
ScholarGate. (2026, June 22). Political Survival Analysis. ScholarGate. https://scholargate.app/political-economy/political-survival-analysis