Political leaders who weaken democratic institutions do not automatically succeed. Autocratization can provoke resistance from courts, legislatures, civil society, political rivals and voters, leaving the leaders driving it exposed to removal from office. A new peer-reviewed study suggests that one factor may substantially improve their chances of staying in power: ideology.
Writing in the European Political Science Review, Claudio Balderacchi and Luca Tomini examined autocratization episodes across the world from 1900 to 2024. Their survival analyses found that stronger ideological legitimation was consistently associated with a lower hazard of an incumbent losing office during an autocratization process.
Across the researchers’ Cox proportional hazards models, a one-unit increase in ideological legitimation was associated with a reduction in the risk of leader exit ranging from 33.8% to 48.5%, after accounting for the controls included in the models. The relationship became particularly striking where horizontal accountability was high. In those settings, ideological legitimation was associated with a 59% lower hazard of removal.
Autocratization is risky for the leaders driving it
Much research on democratic erosion focuses on what autocratizing leaders do to institutions. This study asks a different question: what helps those leaders survive politically while they are doing it?
That distinction matters. Autocratization is not simply a destination from democracy to autocracy. It is a process during which political rules, constraints and balances of power can change. Leaders attempting to concentrate authority can encounter institutional opposition, public mobilization, elite defections or electoral defeat before they consolidate their position.
Balderacchi and Tomini propose that ideological legitimation can help leaders navigate this vulnerable period. Ideological narratives can provide a broader justification for political actions, connect policy changes to a coherent worldview and potentially strengthen cohesion among supporters and elites.
Importantly, the argument is not that a particular ideology produces this effect. The researchers examined ideological legitimation as a political strategy, rather than treating one ideological family as inherently more effective than another.
The study followed autocratization across more than a century
The researchers drew on the V-Dem Episodes of Regime Transformation dataset and examined autocratization processes lasting longer than one year that began in liberal democracies, electoral democracies or electoral autocracies between 1900 and 2024.
The dataset covered 135 autocratization episodes across 98 countries. Because leadership can change within an episode, the survival analysis included leader-level observations and distinguished cases in which an incumbent lost power from cases in which the leader remained in office until the episode ended or was still ongoing.
The study first used Kaplan-Meier survival estimates and log-rank tests to compare patterns associated with different forms of legitimation. It then used Cox proportional hazards models with clustered standard errors to estimate the risk of leader exit while incorporating political and institutional controls.
The authors considered ideological legitimation alongside personalistic, performance-based and legal-rational forms of legitimation. They also examined factors including regime type, presidentialism, political polarization, divided party control, accountability and democratic mobilization.
Ideology stood out from other legitimacy strategies
The first notable result emerged before the multivariate models were introduced. Among the legitimacy measures compared in the Kaplan-Meier analyses, ideological legitimation was the only one showing a statistically significant difference in incumbent survival.
Personalistic, performance-based and legal-rational legitimacy claims did not show consistent statistically significant effects on survival. This does not establish that those strategies are politically irrelevant. It does, however, indicate that they did not predict incumbent survival in the same robust way within this dataset and research design.
The multivariate analysis strengthened that result. Across model specifications, stronger ideological legitimation remained associated with a significantly lower hazard of leaving office. Depending on the model, the estimated reduction in exit risk associated with a one-unit increase in ideological legitimation ranged from 33.8% to 48.5%.
That consistency is important because a relationship visible in a simple comparison can disappear once other political conditions are considered. Here, the association survived the researchers’ progressive addition of controls and several robustness checks.
Strong institutions did not erase the apparent advantage
One of the most interesting parts of the study concerns accountability. A reasonable expectation might be that ideological appeals matter most where institutions are already weak. The results point in a more complicated direction.
Where horizontal accountability was low, ideological legitimation was associated with a 33% reduction in the hazard of removal, corresponding to a hazard ratio of 0.67, but this effect was not statistically significant. Where horizontal accountability was high, the estimated hazard ratio fell to 0.41. That represents a statistically significant 59% reduction in the hazard of incumbent removal.
The survival curves make the contrast easier to understand. Under conditions of high horizontal accountability, incumbents with strong ideological profiles retained a survival probability above 70% after ten years. Those with weak ideological profiles were almost entirely removed by the same point in the analysis.
The authors interpret this pattern as evidence that ideology may be particularly useful when leaders still need to justify their actions in environments where courts, legislatures or other accountability institutions retain meaningful capacity to challenge them. In settings where those constraints are already weak, leaders may have less need for ideological legitimation to remain in office.
Democracies showed a larger estimate, but the difference was uncertain
The researchers also compared democratic regimes with electoral autocracies. In electoral autocracies, ideological legitimation was associated with an estimated 43% reduction in the hazard of losing power, with a hazard ratio of 0.57. In democracies, the estimated reduction was 56%, corresponding to a hazard ratio of 0.44.
However, the interaction by regime type was not statistically significant. The authors therefore do not treat the difference between 43% and 56% as conclusive evidence that ideology works differently across those regime categories. They note that statistical power may be limited, and the safer conclusion is that the accountability interaction provides stronger evidence than the regime-type comparison.
What the findings do and do not show
The study provides evidence of a robust association, but it should not be read as a simple formula in which adopting an ideology automatically keeps a leader in power. Political survival is shaped by institutions, opposition strength, economic conditions, elite relationships, elections, crises and many other forces.
There is also an important measurement issue. The ideological legitimation variables rely on V-Dem expert-coded data and its Bayesian measurement framework. This makes a century-spanning global analysis possible, but historical and cross-national political concepts cannot be measured with the precision of a controlled laboratory variable.
The outcome also combines different ways in which leaders can leave office. Losing an election, resigning, being removed by institutional processes or being displaced through other political events are not equivalent outcomes. The authors acknowledge that collapsing exit modes prevents them from estimating whether ideological legitimation is more protective against some forms of removal than others.
Most importantly, the analysis is observational. The models, controls and robustness tests strengthen confidence in the relationship, but they cannot establish causality as cleanly as randomized evidence could. Leaders capable of constructing strong ideological narratives may differ from weaker leaders in other ways that are difficult to capture fully in cross-national historical data.
Why ideology may matter during democratic erosion
The study adds a useful dimension to research on democratic backsliding. Institutions matter, but political competition also involves narratives about why institutional changes are necessary, legitimate or desirable. An incumbent who can place controversial actions within a coherent ideological story may be better positioned to maintain supporter loyalty and elite cohesion while facing resistance.
This does not mean strong accountability institutions are ineffective. In fact, the study begins from the premise that such institutions create meaningful risks for autocratizing incumbents. The finding is instead that ideological legitimation may help some leaders withstand those risks.
For researchers and democratic institutions, that suggests monitoring formal institutional changes alone may provide an incomplete picture of autocratization. The justificatory narratives surrounding those changes may also be part of the mechanism through which political power is consolidated.
Future work could separate different modes of leader exit, examine how specific ideological messages reach voters and elites, and test more directly the mechanisms linking ideology to political survival. It could also investigate whether the same narratives that help incumbents survive during autocratization have lasting effects after the process ends.
Source Information
Study: A risky business: navigating autocratization through ideological legitimation claims
Authors: Claudio Balderacchi and Luca Tomini
Journal: European Political Science Review
Publisher: Cambridge University Press
Published online: 22 September 2026
Study design: Global observational survival analysis of autocratization episodes from 1900 to 2024 using V-Dem Episodes of Regime Transformation and legitimation data.
Coverage: 135 autocratization episodes across 98 countries.
Source: Cambridge Core, European Political Science Review.
