Work in progress

Selected working papers and pre-analysis plans.

Regime type and international cooperation

Varieties of Democratic Breakdown and IO Membership, 1816-2023.
With Karin Sundström, Carl Vikberg, and Jonas Tallberg

Abstract

While prior research has shown that democratization tends to increase states’ cooperation and expand their memberships in international organizations (IOs), the reverse process, democratic breakdown, remains underexplored. This is a critical gap given the current wave of autocratization and its potential to reshape global governance. Using new data on regime type and IO memberships from 1816 to 2023, we estimate the effects of different forms of democratic breakdown on state participation in IOs. Employing an entropy-balanced stacked difference-in-differences design, we find that democratic breakdown has a substantial negative effect, slowing IO membership growth by about 30%. However, this aggregate effect masks important heterogeneity: incumbent-led democratic breakdown has no discernible effect on IO memberships, whereas nonincumbent-led breakdown slows IO membership growth by about 37%. Notably however, even after nonincumbent-led breakdowns, states continue to expand their IO memberships in absolute terms. Our results suggest that only highly visible regime ruptures meaningfully reduce engagement with IOs, while incumbent-led breakdowns allow regimes to maintain institutional ties and potentially transform organizations from within, with important implications for understanding how the contemporary wave of autocratization is reshaping global governance.

The Autocrat in the Room: Effects of Autocratic Involvement in Human Rights IOs.
With Jonas Tallberg, Faradj Koliev, and Guðlaug Ólafsdóttir

Abstract

Autocratic regimes are surprisingly active in international organizations (IOs) that promote norms at odds with their own political practices. The UN Human Rights Council includes countries like China, Ethiopia, Qatar, and Sudan; Saudi Arabia has chaired the UN Commission on the Status of Women; and Iran presided over the UN Human Rights Council’s Social Forum. What are the consequences of such participation—for the legitimacy of the IO and the autocratic regime? We offer the first systematic study of these dual effects. Theoretically, we develop hypotheses on how autocratic involvement shapes the legitimacy of the IO and the autocratic regime, and how these effects are moderated by the regime type of the audience country, level of autocratic involvement, the degree of repression of the autocratic regime, and the response by democratic member states. Empirically, we examine these expectations through a population-based survey experiment to be fielded in six democratic and autocratic countries. The findings promise to advance literatures on IO legitimacy, authoritarian legitimation, and consequences of autocratization for global governance, while also informing policy debates about the trade-offs of inclusive multilateralism.

Elite views of state obligations to comply with international treaties.
With Jonas Tallberg and Faradj Koliev

Abstract

Are autocracies less likely to comply with international rules than democracies? This question has become increasingly important in the face of ongoing autocratization and the growing weight of authoritarian powers, yet it has not been investigated systematically. We address this gap with an elite-level survey experiment involving political and bureaucratic elites, who are central to the design and implementation of international agreements.

Our study covers sixteen countries across Europe, Africa, Asia, and the Americas, ranging from liberal democracies to full autocracies. The conjoint experiment presents respondents with hypothetical treaties that vary in six respects: adjustment costs, enforcement provisions, reciprocity dynamics, public opinion, leader opinion, and legal norms with democratic or autocratic orientations.

We advance three sets of expectations. First, we expect elites’ support for compliance to respond systematically to adjustment costs, stronger enforcement, reciprocity, leaders and public opinion, and the legal norms of the treaty. Second, we expect these effects to be conditioned by regime-related factors. Elites with democratic values, and those in democratic regimes, should be more responsive to public opinion and democratic norms, while elites with autocratic values, and those in autocratic regimes, should be more responsive to leader opinion and autocratic norms. Third, we expect democracies, and more pro-democratic elites, on average, to be more law-abiding than autocracies and less democratic elites, both because of their domestic institutions and their responsiveness to societal groups.

Our findings will offer a systematic assessment of how regime type shapes compliance and illuminate the implications of rising autocratic power for global governance.

Democratization, Autocratization, and the Effectiveness of International Shaming.
With Faradj Koliev

Abstract

How does regime type affect states’ propensity to comply with international agreements? We distinguish between proactive compliance, where states conform to international standards without external prompting, and reactive compliance, where states improve only after being publicly criticized. Using ILO supervisory data from 1970–2020, we examine how democratization and autocratization affect reactive compliance—states’ responsiveness to international shaming for labor rights violations. We operationalize regime change as significant movements on V-Dem’s Polyarchy Index and employ linear probability models with country, year, and convention fixed effects. Democratization increases the likelihood of receiving positive ILO compliance notes by 4–8 percentage points, regardless of starting regime type. Autocratization shows no systematic effect. We argue that democratization enhances reactive compliance by empowering domestic compliance constituencies, creating credible commitment incentives, and disrupting institutional equilibria that sustained prior non-compliance. These findings reveal that the effectiveness of international shaming depends on the direction of domestic political change, not merely static regime characteristics.

Autocratic politics and political elites

Tools for adaptation. How elections inform and legitimize elite turnover in autocracies.
Single-authored
Previous version (PDF)

Abstract

Dictators rely on capable and loyal elite coalitions to stay in power, but shifting political landscapes often require adjustments to this coalition. However, detecting gradual political changes and implementing elite turnover without provoking backlash is challenging. I argue that dictators can use elections to mitigate both problems. First, elections provide information on the strengths and loyalties of elites and regime outsiders, helping dictators assess whether coalition adjustments are needed. Second, elections institutionalize and legitimize elite turnover, reducing the risk that turnover is perceived as a power-grab, thus causing backlash. Empirical analyses demonstrate that election-years, on average, increase cabinet turnover by around 12 percentage points and have a similar effect on military purges. Additional analyses show that the effect of elections is stronger when their informational value is higher, and that while elite turnover increases the risk of successful coups, this risk is mitigated when elite turnover happens during election years.

Monuments of might: Symbolic political structures across the world.
With Haakon Gjerløw, Felix Haas, and Carl Henrik Knutsen

Abstract

Physical symbols that represent institutions of power, nations, or individual leaders have been a feature of politics since ancient times. When and why do governments erect such politically symbolic buildings and monuments, and do they have tangible consequences on, e.g., citizens’ beliefs or behavior? Systematic data-based answers to these questions have been obstructed by the lack of comparable measures that extend across countries and over time. Hence, we collect and make publicly available detailed data on three ubiquitous features of symbolic politics pertaining to, respectively, institutionalized power, nations, and leaders: (1) government and legislative buildings, (2) mega-flagpoles, and (3) leader statues. We elaborate on the concept of symbolic political structures and discuss how they may be conceived as credible commitment devices, before we detail and evaluate our new data. To illustrate how the data may be used, we show that investment in symbolic structures meaningfully varies with leader- and regime features. For instance, personalist dictators are particularly likely to build more and higher statues and change residencies than democratic leaders. Our new data helps to advance comparative research on symbolic politics and provide new measures for hard-to-observe regime features, such as regime legitimation.

Public opinion and economic policy

Political Support for Economic Growth Policies in the Global South.
With Fulya Apaydin and Gizem Arikan

Abstract

How do citizens evaluate competing economic growth strategies, and when do those evaluations translate into political support and trust? We will conduct parallel online survey experiments with 4,000 adults in each of Argentina, Malaysia, and Turkey (N ≈ 12,000) to answer these questions. The survey combines two randomized components. First, respondents evaluate economic plans that vary six attributes: growth strategy, household credit conditions, support for firms, financing source, macroeconomic risk, and trade policy. Second, respondents are randomly assigned to a three-arm vignette that holds the broad content of an economic plan constant while varying whether the government communicates and implements it credibly and predictably. The conjoint identifies which substantive components generate support, while the vignette isolates credible commitment as a complementary political mechanism. Our central expectation is that policy packages with more components matching the incumbent government’s documented policy baseline across all six randomized attributes will receive greater support among incumbent voters.