📌 New Method for Hidden Political Communities

A new methodology—the bipartite link community model (biLCM)—infers political actors' latent memberships in communities of collective activity that shape observable interactions. Unlike existing approaches, biLCM is designed for two distinct actor sets and captures overlapping and multi-faceted interaction structure.

🔬 How the Model Captures Complexity

• Applies to two groups of actors (a bipartite setting).
• Allows actors to belong to more than one community (overlapping membership).
• Permits a pair of actors to interact in more than one way (multi-relational links).

🗂️ Constructing a Legislative Lobbying Network

An original dataset links politicians who sponsor congressional bills with the interest groups that lobby on those bills. This linkage is based on more than two million textual descriptions of lobbying activities, addressing a common empirical obstacle: the difficulty of observing direct ties between interest groups and politicians.

📈 Applied Case: 113th U.S. Congress

The biLCM is used to characterize legislative communities of special interest groups and politicians in the 113th U.S. Congress, producing quantitative measures of community membership for both sets of actors.

🔎 Key Findings

• Community memberships range from narrow, targeted interaction patterns aligned with industry interests and committee jurisdictions to broad, multifaceted connections spanning multiple policy domains.
• The model yields fine-grained, quantitative descriptions of how interest groups and legislators cluster around shared policy activity.

💡 Why This Matters

By combining a bipartite, overlapping-community framework with a large, text-derived lobbying dataset, the approach overcomes limits of prior work and provides a scalable way to map the hidden structure of legislative lobbying networks and their policy-relevant communities.