How much democratic backsliding will citizens accept to secure a policy they want? We argue the answer depends on whether the trade-off itself is partisan or policy-based. In April 1969, Charles de Gaulle bundled decentralization with a Senate reform that stripped it of legislative power, tying his presidency to the outcome and putting this trade-off to French voters directly. We isolate the policy-based mechanism, focusing on France’s center-periphery cleavage, with digitized referendum returns (1958–1969), a 1967–1969 panel survey, and de Gaulle’s campaign speeches. Regionalist departments moved further toward the reform than the rest of the country; sympathy for sub-national self-determination predicted the vote net of partisanship; and voters and de Gaulle himself split along the same line: the reform’s substance at the periphery, his rule elsewhere. Policy-based trade-offs may be harder to resist than partisan ones: citizens who accept a weaker check for a good they want are not excusing their side, but see the exchange as legitimate.

Territorial prejudice — negative affect toward residents of other regions — represents an understudied source of intergroup tension in (polarized) federal and decentralized democracies. We examine whether indirect reciprocity can reduce such prejudice under favorable conditions, exploiting the large-scale deployment of cross-regional firefighters following the catastrophic October 2024 Valencia floods. A pre-registered survey experiment among 1,500 Valencian residents randomly assigned participants to conditions varying whether their neutral or most-disliked outgroup region had provided assistance, measuring willingness to reciprocate in a hypothetical future disaster. Territorial prejudice significantly reduced cooperative intentions toward disliked regions by 3.5%, rising to 5% among respondents with the most negative outgroup affect, even amid overwhelming prosocial behavior and salient interregional solidarity. Indirect reciprocity produced no countervailing effect. These findings suggest that territorial prejudice constitutes a robust attitudinal disposition that persists even under conditions most favorable to its attenuation, with important implications for intergovernmental cooperation in polarized, decentralized democracies.

Influential literatures presume institutions forge territorial identities: decentralization is expected to build regional attachment, globalization to breed cosmopolitans and a nationalist backlash. We test these premises with the joint EVS/WVS (92 countries), measuring identity from a five-item closeness battery (town, region, country, continent, world) and classifying each respondent into a typology: exclusively attached to one level (subcountry, country, or supracountry), plural, or unattached. Most people are plural identifiers, and this plurality is largely world-inclusive; exclusive cosmopolitans are rare and concentrated among immigrants, while exclusive country identification reflects grievance, not tradition. The two institutional dimensions behave asymmetrically: globalization sharpens exclusive attachments and, in formative years, broadens identity into world-inclusive plural profiles rather than building exclusively cosmopolitan ones, whereas decentralization builds subcountry identity nowhere. Migration reshapes identity, toward the country in less-globalized destinations, beyond it in more-globalized ones, inherited by the second generation. We propose an interactive context–composition account.

Federations and decentralized countries divide their geography and people into autonomous territorial units. These units are different in a number of dimensions and the historical development of these asymmetries shapes the territorial co-existence in these regimes, where inter-unit sympathies and rivalries are a defining feature. However, a rival territorial unit might turn into an out-group when its members prompt negative sentiments among the members of an in-group unit. We define as Territorial Affective Polarization (TAP) the extent to which citizens like people living in their unit more than people living in the rest of the territories. Moreover, we argue that this ‘affect gap’ might become a threat for the functioning of a decentralized democracy once individual-level party identification and inter-territorial feelings overlap. We utilized data from a survey-experiment developed in Spain in 2021 to show the prevalence of TAP and to analyze the extent to which TAP can hinder democratic accountability in highly polarized territorial regimes.

As natural disasters become increasingly frequent under climate change, understanding how they reshape politics has become more urgent. This paper theorizes and examines the political consequences of natural disasters in an increasingly prevalent context: advanced democracies with low levels of political trust. We argue that when political trust is already weak, disasters act as performance shocks that further erode political confidence and disrupt conventional mechanisms of democratic accountability. Rather than rallying around incumbents or shifting support toward mainstream opposition parties, citizens may disengage from representative institutions and develop stronger anti-political attitudes. Using the case of the October 2024 floods in Valencia, Spain, and drawing on observational, experimental, and panel data, we show that highly exposed residents exhibit significantly lower levels of political trust and stronger populist attitudes. Electorally, this dissatisfaction translates primarily into persistent electoral demobilization, alongside suggestive evidence of increased support for the radical-right party VOX in the most severely affected areas.

Who is held accountable for human rights violations after a fragmented conflict? Dominant theories hold that post-conflict accountability depends on the outgoing regime’s leverage over the new government. Yet multi-faction transitions complicate this logic. We argue accountability reflects not only the power balance but also each faction’s relationship to the implementing party. We illustrate this framework using an original dataset of amnesty decisions from South Africa’s Truth and Reconciliation Commission, comparing applicants assessed by the same judging panel for the same crime. Decisions favored the implementing party (ANC)—not other anti-apartheid factions without leverage (PAC/APLA), nor opponents holding leverage (apartheid officials). Opponents without leverage (Inkatha) were the most likely to be rejected. Consistent with a career-incentives mechanism, bias concentrated among judging panels with less established legal careers. Results show factions’ relationship to the implementing party and their leverage jointly shape accountability in fragmented transitions, even without direct government intervention.