🔎 What the Study Does

Extends classical ideal point estimation to allow voters to express different preferences across distinct policy domains (for example, agriculture versus defense). The scaling procedure recovers estimated ideal points on a common scale so revealed preferences can be compared directly across domains. That comparison makes it possible to assess whether a member votes more conservatively in one domain than another and to evaluate the extent to which an individual’s voting behavior conforms to a uni-dimensional spatial model (i.e., whether preferences are constant across domains).

🔎 How Multiple Preferences Are Identified

• Employs a novel scaling approach that places domain-specific ideal points on a single, comparable scale.
• Crucially, the method estimates — rather than assumes — which legislators are “stayers” (those whose revealed preferences are constant across votes), removing a common restrictive assumption.
• Frames the estimation as a form of partial-pooling item-response theory (IRT) scaling, which borrows strength across items while allowing domain-level heterogeneity.
• Positions the approach as a tool for investigating the relationship between the basic space and issue space in legislative voting (Poole 2007).

📊 Where the Method Is Demonstrated

• Illustrates the model by estimating U.S. House of Representatives members’ revealed preferences across different policy domains (different sets of motions).

🧾 Key Methodological Advantages

• Enables testing of sharp hypotheses about domain-specific preference differences.
• The partial-pooling IRT framing reduces uncertainty in estimates relative to fully separate scaling.
• Offers a principled, unified solution to the ‘‘granularity’’ problem (the level of aggregation) in roll-call analysis (Crespin and Rohde 2010; Roberts et al. 2016).
• Allows direct comparison of individual legislators’ revealed preferences across issue domains.

🚀 Potential Applications

• Studying the relationship between committee and floor voting behavior.
• Investigating constituency influence and patterns of representation.
• Any roll-call setting where interest lies in domain-specific versus general ideological signals.

❗ Why It Matters

Provides a practical and testable framework for measuring heterogeneity in legislative preferences, improving inference about when legislators are consistent across issues versus when they shift, and clarifying how issue-specific behavior maps onto broader spatial models of roll-call voting.