The Impact of Partisan Gerrymandering on Political Parties

• Published date: 01 November 2020
• Author: Nicholas O. Stephanopoulos,Christopher Warshaw
• Date: 01 November 2020
• DOI: http://doi.org/10.1111/lsq.12276

609

LEGISLATIVE STUDIES QUARTERLY, 45, 4, November 2020

DOI: 10.1111/lsq.12276

NICHOLAS O. STEPHANOPOULOS

Harvard Law School

CHRISTOPHER WARSHAW

George Washington University

The Impact of Partisan

Gerrymandering on Political Parties

The relationship between votes and seats in the legislature lies at the heart

of democratic governance. However, there has been little previous work on the

downstream effects of partisan gerrymandering on the health of political parties.

In this study, we conduct a comprehensive examination of the impact of partisan

advantage in the districting process on an array of downstream outcomes. We

find that districting bias impedes numerous party functions at both the congres-

sional and state house levels. Candidates are less likely to contest districts when

their party is disadvantaged by a districting plan. Candidates that do choose to

run are more likely to have weak resumes. Donors are less willing to contribute

money. And ordinary voters are less apt to support the targeted party. These re-

sults suggest that gerrymandering has long-term effects on the health of the dem-

ocratic process beyond simply costing or gaining parties seats in the legislature.

Introduction

In recent years, there has been a resurgence of scholarship

about redistricting and partisan gerrymandering. Almost all of

this literature, however, has focused on measuring the extent of

gerrymandering (e.g., McDonald and Best 2015; Stephanopoulos

and McGhee 2015; Warrington 2018b) or distinguishing the in-

tentional manipulation of district lines from the impact of a ju-

risdiction’s political geography (e.g., Bangia et al. 2017; Chen

and Rodden 2013; Cho, Wendy, and Liu 2016). Somewhat sur-

prisingly, there has been little political science scholarship on the

© 2020 Washington University in St. Louis

Replication files for this article are available from LSQ’s Harvard Dataverse

at https ://doi.org/10.7910/DVN/WVVJFZ.

610 Nicholas O. Stephanopoulos and Christopher Warshaw

downstream effects of gerrymandering on candidates, voters, or

political parties more generally.

In this article, we focus on the impact of partisan gerryman-

dering on political parties. From a theoretical standpoint, it is

plausible that parties disadvantaged by gerrymandering become

less able to conduct several of their key activities. Why? Because

parties exist, in large part, to win seats (and by winning seats, to

enact their preferred policies). Gerrymandering prevents parties

from winning as many seats. Faced with a higher probability of

electoral defeat, both party elites and rank-and-file voters are thus

likely to demobilize. Elites are apt to run less often for office, to

have worse credentials when they do toss their hats in the ring, and

to contribute less money to long-shot candidates. Ordinary voters,

similarly, may be expected to cast fewer ballots for a party with lit-

tle hope of controlling the legislature.

Moreover, while all party activity may be dampened by par-

tisan gerrymandering, we anticipate that functions carried out by

elites are hampered even more than the behavior of the party in

the electorate. This is because elites (party officials, candidates for

office, major donors, and so on) are more aware of, and affected

by, the electoral environment than regular voters. Compare the

situation of a prospective candidate in a gerrymandered state to

that of a typical voter. The putative candidate is likely to be highly

conscious of her (low) odds of prevailing if she runs and highly

sensitive to the cost (in time, money, and reputation) of a failed

campaign. In contrast, the voter probably has top-of-the-ticket

races in which to cast ballots, and even if not, faces only the usual

opportunity cost of going to the polls.

Turning to our empirical analysis, we conduct a comprehen-

sive evaluation of the downstream effects of partisan gerryman-

dering on both elites and voters. In doing so, we have the benefit

of an unprecedented dataset of partisan bias in districting: the

crucial independent variable in our analysis.2

Our dataset spans

congressional and state house elections from 1984 to the pre-

sent. For each election, the dataset includes four distinct meas-

ures of partisan districting bias: the efficiency gap (McGhee 2014;

Stephanopoulos and McGhee 2015), the declination (Warrington

2018b), the mean-median difference (Best etal. 2018; Krasno etal.

2019; Wang 2016), and partisan asymmetry (Gelman and King

1994a). We also complement the dataset with a variety of longi-

tudinal indicators of party health, capturing activities performed

611The Impact of Partisan Gerrymandering on Political Parties

by both elites and voters. These include candidate entry, candidate

quality, fundraising, and voter decision-making.

Using a series of dynamic panel models, we find that par-

tisan disadvantage in districting indeed impedes numerous party

functions. These results hold no matter which measure of partisan

disadvantage is used and at both the congressional and state house

levels. The results also hold whether partisan disadvantage is calcu-

lated using endogenous legislative elections or exogenous presiden-

tial races. Candidates are thus less likely to contest districts when

their party is handicapped by a districting plan. Candidates who

do choose to run are more likely to have weak resumes. Donors are

less willing to contribute money. And ordinary voters are less apt

to support the disfavored party. These effects, moreover, are gen-

erally quite large. A one standard deviation increase in a district-

ing bias metric, for instance, is associated with roughly a 5-point

decline in a party’s share of campaign contributions. Consistent

with our distinction between elites and voters, however, the impact

of districting bias on the electorate is somewhat less impressive

(though still statistically significant).

Our findings complement a small set of related studies that

have examined the consequences of partisan gerrymandering

(Seabrook 2017). These studies have shown that partisan bias in

a districting plan affects the ideological composition of state leg-

islatures (Caughey, Tausanovitch, and Warshaw 2017) and con-

gressional delegations (Cameron, Epstein, and O’Halloran 1996;

Stephanopoulos 2018). It also affects the policies implemented by

state governments (Caughey, Tausanovitch, and Warshaw 2017).

Our results show that gerrymandering not only skews legislative

representation and enacted policy, it also hinders parties in the

performance of several of their core functions.

Our findings also have implications for litigation on partisan

gerrymandering. In a concurrence in a high-profile 2018 Supreme

Court decision on gerrymandering (Gill v. Whitford, 138 S. Ct.

1916, 1938 [2018]), Justice Kagan emphasized the possibility of

downstream effects of gerrymandering. Gerrymandering, she

wrote, may cause “difficulties” for a party in terms of “fundrais-

ing, registering voters, attracting volunteers, generating support

from independents, and recruiting candidates to run for office.”

These difficulties may “ravage” a party, “placing [it] at an endur-

ing electoral disadvantage” and “weaken[ing] its capacity to per-

form all its functions.” Our findings provide empirical support for

Justice Kagan’s speculation.