The formation of the “yellow‑green” Conte I government—an unexpected coalition between the Five Star Movement and the League after Italy’s 4 March 2018 election—appeared puzzling because the partners seemed far apart on the conventional left–right scale. This article shows that the puzzle arises from relying on a unidimensional left–right representation of party positions and resolves it by introducing a second policy dimension.

📊 What the Analysis Looks At

A long-run view of Italian coalition outcomes from 2001–2018 is used to assess whether spatial models explain which parties join government together.

🧭 How Party Positions and Coalitions Were Compared

• Statistical analysis of coalition formation across 2001–2018.
• Models include parties' policy distance on the left–right dimension and other control variables.
• The unidimensional (left–right) specification predicts coalition outcomes well up to 2013 but fails to account for the 2018 result.

📈 Key Finding: A Two-Dimensional Spatial Account

• Introducing a second policy dimension resolves the apparent anomaly of the Conte I coalition.
• The two dimensions are:
• Economic left–right
• A social/identity dimension capturing immigration, European Union positions, and social conservatism
• When parties’ positions are measured on both dimensions rather than a single left–right continuum, the 2018 coalition outcome becomes predictable and no longer anomalous.

🔍 Why This Matters

• Reliance on a single left–right scale can mischaracterize party proximity and coalition prospects in contexts where immigration, EU attitudes, and social conservatism form a distinct axis of competition.
• The findings call for spatial models of coalition formation that incorporate multiple policy dimensions to better explain contemporary party alliances, especially in Western European systems experiencing salient identity and EU debates.