Conflict, Information, and Regime-Change

Bosco D., Colombo L., Femminis G., 2025 – European Journal of Political Economy

In recent years, political science has highlighted how information transparency—particularly media freedom—shapes collective action and, consequently, the political stability of authoritarian regimes. Information influences both citizens’ awareness of their grievances and their capacity to coordinate collective responses, while simultaneously shaping how regimes perceive and address discontent.

In a recent article, “Conflict, Information, and Regime-Change”, published in the European Journal of Political Economy, Davide Bosco, Luca Colombo, and Gianluca Femminis add to this literature by formalizing, within a global-games framework, how a regime’s choice between media freedom and censorship interacts with citizens’ incentives to revolt, the distribution of private discontent, and the prospects of regime survival.

The authors consider an autocratic regime that faces potential protests whenever discontent with the status quo becomes sufficiently high. Discontent represents the model’s key state variable: it reflects citizens’ private assessments of regime performance in providing public goods or addressing grievances. Importantly, individual discontent is private information, unknown to both other citizens and the regime itself. Citizens differ in their level of discontent and revolt if it exceeds a threshold determined by their personal cost of revolting and their beliefs about how many others will join.

The regime must choose whether to allow free media or to impose full censorship. Public information plays two contrasting roles. On the one hand, it allows the regime to better gauge aggregate discontent and adjust public goods provision accordingly, thereby mitigating grievances when repressing a revolt is possible. On the other hand, it lowers coordination costs among citizens by reducing uncertainty about others’ discontent and likely actions. In settings characterized by strategic complementarity — where an individual’s incentive to revolt increases when others are also expected to revolt — better information can thus facilitate collective coordination and ultimately weaken regime stability.

Building on this framework, the analysis yields several key insights. First, a regime’s preference for media freedom versus censorship depends non-monotonically on its ability to withstand a revolt. Very strong and very weak regimes tend to prefer media freedom: in both cases, the informational benefit of accurately gauging discontent and fine-tuning public goods provision outweigh the coordination risk arising from more precise information. For strong regimes, even well-informed citizens struggle to mount a successful revolt; for weak regimes, by contrast, well-targeted policy responses can be effective at reducing discontent. Regimes of intermediate strength, however, face the greatest danger from improved public information. Neither fully secure nor close to collapse, they are the most exposed to the coordination effects of public signals and therefore tend to prefer censorship, as the risk of enabling coordinated dissent exceeds the gains associated with better policy targeting.

A second major result concerns the role of policymaking itself. If a regime could not adjust public goods provision in response to information — so that media freedom improved knowledge without enabling corrective action — all regimes would unambiguously choose censorship. This highlights that autocrats find media freedom useful not for the sake of information itself, but because it enables them to address hidden grievances and maintain sufficient popular support.

The predictions of the proposed framework help explain recent developments across authoritarian and semi-authoritarian contexts. As the model suggests, regimes of intermediate strength – neither fully secure nor close to collapse – have the strongest incentive to impose tight media controls, since public information effectively facilitates citizens’ ability to coordinate collective action.

Bangladesh under Sheikh Hasina offers a clear example, but a similar logic would apply to the ayatollahs’ regime in Iran. For over fifteen years, the regime upheld strict information controls and used public spending to reward its core supporters. This strategy remained effective while the regime remained in the intermediate range predicted by the model. When mass student protests over public sector recruitment rules undermined the regime’s stability, censorship weakened: more extensive coverage in both traditional and online media facilitated coordination, contributing to Hasina’s downfall in August 2024.

Stronger regimes face fewer coordination risks and therefore rely less on strict censorship. China offers a clear example: although far from a liberal media system, some criticism and public debate are tolerated. Similar patterns appear in other partially or non-democratic systems, such as the `illiberal democracy’ or `soft authoritarianism’ of Mubarak’s Egypt and Erdogan’s Turkey. Even in Putin’s Russia, somewhat independent sources of information, such as the Levada Center, still operate despite being designated as a “foreign agent” since 2016.

Some regimes, finally, leverage media openness to enhance governance. Singapore stands out in this respect. Although formally a democracy, the People’s Action Party has ruled since 1965 while allowing substantial media freedom. Crucially, informational openness is paired with highly responsive and well‑targeted policies that reduce discontent — precisely the conditions under which our model predicts a preference for media freedom.

In conclusion, the model shows that information in authoritarian regimes has a dual effect: it helps rulers detect and address discontent, but also enables citizens to coordinate collective action. Whether media freedom stabilizes or undermines a regime depends on its underlying strength and its capacity to respond with well‑targeted policies. Recent events — such as the protests in Iran and Bangladesh — illustrate how different informational environments shape beliefs about others’ willingness to dissent, influencing the scale of mobilization. Understanding these mechanisms is crucial for anticipating when coordinated action and regime change may occur.