Are Republicans Better at Gerrymandering than Democrats?

This paper investigates how partisan control over redistricting influences seat shares in the U.S. House of Representatives. Using state-decade level data from 1968 to 2016, combined with information on laws, legislative composition, and census-based district maps, the authors estimate the causal impact of unilateral party control of redistricting. They find that Republican control led to an average 8.3 percentage point increase in Republican House seat share in the three elections following redistricting in the past two decades, while Democratic effects were small and only significant in large states. Overall, partisan redistricting explains up to 54% of the partisan seat gap in the 2010s.

Full Citation and Link to Article

Coriale, K., Kolliner, D. A., & Kaplan, E. (2025). Political control over redistricting and the partisan balance in Congress (NBER Working Paper No. 33801). National Bureau of Economic Research. https://doi.org/10.3386/w33801

Extended Summary

Central Research Question

The central question of this paper is: how does partisan control over the redistricting process affect the partisan balance in the U.S. House of Representatives? Specifically, the authors seek to determine whether unilateral legal control of redistricting by either the Democratic or Republican Party translates into measurable gains in seat shares, how persistent these effects are over multiple elections, and whether they vary across time, state size, or party. The study is motivated by ongoing debates in political science and law over the extent to which gerrymandering distorts representation and contributes to partisan advantages in Congress .

Previous Literature

A broad literature addresses gerrymandering, with early theoretical work focusing on strategies like “packing” and “cracking” to maximize partisan advantage (Gilligan and Matsusaka 1999; Shotts 2001). Later contributions explored the role of uncertainty, turnout heterogeneity, and voter clustering in shaping redistricting strategies (Friedman and Holden 2008; Bouton et al. 2023; Kolotilin and Wolitzky 2023). Empirically, scholars have relied on simulated counterfactual maps to assess whether existing maps confer partisan bias beyond geographic clustering (Chen and Rodden 2015; Kenny et al. 2023). This line of work finds that some partisan bias reflects natural residential patterns rather than intentional design. Others have emphasized political sorting over time as a contributor to representational asymmetries (Bishop 2009; Hopkins 2017). Recent empirical strategies have included the “efficiency gap” metric, developed by Stephanopoulos and McGhee (2015), and regression discontinuity designs based on close elections for state offices (Jeong and Shenoy 2024). However, past work has been limited either by reliance on proxies for legal control, lack of statistical power, or inability to differentiate effects across parties. This study builds on that literature by directly coding legal control of redistricting at the state-decade level, carefully distinguishing partisan trifectas from independent commissions, and applying three complementary identification strategies.

Data

The study constructs a novel dataset covering the period 1968–2016. The main outcome variable is the Republican share of House seats at the state delegation level, collected from Congressional Quarterly. Control variables include state-level presidential vote shares. The key independent variable is legal control over redistricting, determined by the partisan composition of state legislatures, governorships, and the presence or absence of independent commissions. Data on partisan control comes from Klarner et al. (2013) for the earlier period, supplemented with National Conference of State Legislatures data from 2012 onward. Information on redistricting institutions is drawn from statutory and constitutional records, Spencer’s compilation of redistricting law post-2000, and independent research for earlier years. The dataset also includes ARC-GIS measures of district boundary changes, census data on racial composition, and lists of minority members of Congress. This allows the authors to evaluate not only seat shares but also shifts in minority representation and the geographic extent of redistricting. The final dataset covers 212 state-decades and more than 1,000 state-election years .

Methods

The authors use three primary empirical strategies. First, a two-way fixed effects panel model compares states with and without unilateral partisan control over redistricting within decades, exploiting the timing of decennial redistricting. They structure decades unconventionally, starting in years ending in 8 and ending in 6, to allow for two pre-redistricting elections and three post-redistricting elections. This specification enables estimation of both pre-trends and dynamic effects. Second, they apply a regression discontinuity design in gubernatorial elections for states where unified legislatures require gubernatorial approval for redistricting bills. Narrow gubernatorial victories of the majority party thus create quasi-random variation in whether a trifecta translates into legal control. Third, the authors develop a novel simulation-based binned matching (SBBM) estimator. This approach estimates the shock distribution of state legislative elections, simulates probabilities of partisan control, and matches treated and control state-decades with similar ex-ante probabilities of control. This design balances the tradeoff between statistical power and bias, offering a middle ground between the panel and RD approaches. Across all methods, the authors cluster standard errors at the state-decade level to account for heteroskedasticity and serial correlation. They also conduct robustness checks using alternative definitions of legal control, placebo tests, and randomization inference.

Findings/Size Effects

The results consistently show that Republican legal control over redistricting has had sizable effects in recent decades, while Democratic control has had much smaller or negligible effects. In the panel fixed effects model, Republican control increased Republican seat shares by 4.8 percentage points on average across the sample but by a much larger 8.3 percentage points in the past two decades. These effects persisted across the three elections following redistricting. Democratic control, by contrast, did not produce statistically significant effects overall, though in large states Democrats achieved some gains. The RD estimates reinforce these patterns: in states with narrow Republican gubernatorial victories, Republican seat shares rose by 19.7 percentage points in the recent period, though estimates were noisier due to smaller samples. The SBBM estimator also produced consistent results, with Republican control boosting seat shares by 10.9 percentage points in the recent decades. In contrast, Democratic control yielded coefficients close to zero across specifications. The authors also analyze mechanisms behind these partisan differences. They reject the idea that Republican gains reflect merely undoing Democratic advantages from earlier realignment or that Democratic underperformance results solely from creating minority-majority districts. While Democratic control has increased minority representation, particularly in large states, this does not fully explain partisan disparities. Instead, structural factors—such as denied trifectas in Democratic states due to Republican governors, and stronger Republican effects in smaller states—account for most of the asymmetry. Aggregate estimates indicate that partisan redistricting explained less than 10% of the partisan seat disparity before 2000 but rose dramatically to explain 54% of the gap in the 2010s. These effects are concentrated in small and medium states where Republicans exercised unilateral control, while large Democratic states often lacked trifectas or ceded authority to commissions. Importantly, the analysis finds no evidence that rising racial gerrymandering accounts for these trends .

Conclusion

The study concludes that partisan legal control of redistricting has had a significant and growing effect on the partisan balance in Congress, especially in the 2010s. Republican control has yielded large and durable increases in seat shares, while Democratic control has been less effective except in large states. This asymmetry stems from institutional and strategic differences, not solely from minority district creation or random variation. The findings suggest that partisan gerrymandering has become a more powerful driver of congressional representation in the modern era, explaining a majority of the partisan gap in the 2010s. At the same time, the results imply that institutional reforms such as independent commissions could meaningfully constrain partisan distortions. The authors highlight the broader implications for democratic representation, noting that unchecked partisan control over district boundaries threatens to entrench political advantages beyond what voter preferences alone would dictate. By providing robust evidence across multiple methods and over five decades, the paper contributes to scholarly debates, policy discussions, and legal considerations surrounding gerrymandering and electoral fairness.