Sharing Power, Controlling Issues: How Portfolio Allocation Shapes Cabinet Duration in Parliamentary Democracies

This paper argues that cabinet durability depends on how coalition partners allocate portfolios in a way that balances visible fairness with the policy risks created by delegating control over salient issues. Building on Bäck, Dumont, and Saalfeld (2011), who show that both proportional portfolio allocation and the assignment of salient portfolios to coalition partners influence cabinet survival, we develop a more explicit account of the mechanisms linking bargaining over portfolios to cabinet stability. Following Martin and Vanberg (2020), two mechanisms are central to this approach. First, proportional allocation of portfolios provides a visible signal of fairness that helps parties defend the coalition bargain to voters. Second, stability is shaped by policy risk: parties face greater risk when control over issues they find highly salient is delegated to ideologically distant partners. Matching portfolios to parties’ policy priorities reduces this risk, while leaving salient issues unresolved or in the hands of distant partners increases uncertainty, especially if exogenous shocks later force these issues onto the agenda, as modeled by De Marchi and Laver (2020) in their high-dimensional framework of government formation. From these mechanisms follow three expectations. (1) Coalitions should last longer when portfolios are allocated proportionally; (2) Governments should be more stable when parties receive the portfolios corresponding to their most salient issues; (3) These effects should interact: the stabilizing role of salience-portfolio matching should be stronger when proportionality is limited and when ideological disagreement within the coalition is high, since both conditions heighten policy risk. To test these hypotheses, we combine REPDEM data on government duration in 33 parliamentary democracies with WhoGov ministerial records and party-level policy salience and positions from CHES and MARPOR data. Event-history models are employed to evaluate the main and interactive effects of proportionality and matching of salient portfolios on cabinet survival across parliamentary democracies. The paper’s main contribution is to show that even coalitions which are not ideologically homogeneous can be relatively stable when portfolio allocation is both visibly fair and structured to minimize policy risk by matching control over issues with the parties that care most about them.