Recent work on legislative politics has documented complex patterns of interaction and collaboration through network analysis. Field experiments, a growing approach in state legislatures for causal identification, face unique challenges when social networks exist.
Data & Methods
* Use recently developed methods that account for interference caused by network effects
* Specify spillover models defining how legislators influence each other
* Identify the precise network structure facilitating treatment spread
* Analyze data from two previously published field experiments on state legislatures
Key Findings
* Existing research shows mixed evidence for persistent spillover effects in legislative settings.
* Networked social interaction demonstrably affects outcomes, requiring special consideration beyond standard assumptions.
Why It Matters
These findings highlight the necessity of incorporating network considerations into experimental designs and analyses studying state legislatures. The replication analyses demonstrate practical ways to test for interference using advanced techniques.
| Considering Network Effects in the Design and Analysis of Field Experiments on State Legislatures was authored by Bruce Desmarais and Sayali Phadke. It was published by Sage in SPPQ in 2019. |
