The classic regression discontinuity (RD) design requires continuous variables, but legislative majorities are finite.
➡️ This paper explores an alternative: treating narrow one-seat margins as local randomized experiments.
➡️ The authors implement recent econometric tests and propose a new approach for US state legislatures.
➡️ Their method allows analysis of small seat changes without needing continuous running variables.
➡️ Key findings reveal insights into slim majority impacts on close district races.
➡️ This provides novel ways to estimate the effects of near-majority control in legislative settings.
