Constituency Size and Evaluations of Government

• Published date: 01 May 2022
• Author: Daniel Bowen
• Date: 01 May 2022
• DOI: http://doi.org/10.1111/lsq.12358

459

LEGISLATIVE STUDIES QUARTERLY, 47, 2, May 2022

DOI: 10.1111/lsq.12358

DANIEL BOWEN

The College of New Jersey

Constituency Size and Evaluations of

Government

This article examines the relationship between legislative constituency size

and opinions about the US state government. I show that over the course of US

history, the states have disconnected the size of their legislatures from population

change, resulting in a growing constituency size in nearly every state. I argue that

because district size structures the nature of constituent– legislator interaction

and levels of district heterogeneity, size influences the effectiveness of legislator

efforts to build trust and support, resulting in more negative evaluations of rep-

resentative government. Using a unique data set of nine Pew Research Center

public opinion polls on favorable opinions of US state governments from 1997 to

2012, I find that constituency size is indeed associated with more negative evalu-

ations. This relationship cannot be explained by other probable causes like leg-

islative professionalism, partisanship, opinions about the federal government, or

population size alone. Legislative institutions appear to moderate how residents

experience and evaluate state government.

On March 13, 2018, the Pennsylvania House of

Representatives voted to approve a constitutional amendment

shrinking the size the General Assembly from 203 representatives

to 151 and from 50 senators to 38. The reform would have reduced

the country’s second largest state legislative chamber in one of the

most populous states. Proponents argued that the chamber was

inefficient and expensive (Esack 2018). Opponents countered that

the amendment would limit effective representation by decreasing

direct communication between representatives and constituents,

reducing the voting power of geographically located population

groups, and increasing the reliance on staff to interact with voters.

As one member remarked during floor debate: “A smaller House,

bigger districts means that we lose that intimacy with our constitu-

ents. It also means that very important constituencies will be

© 2021 Washington University in St. Louis.

460 Bowen

swallowed up by larger districts and, therefore, have less of a voice

in this place.”1

While Pennsylvania was considering reducing the number of

seats in the legislature, reformers of other legislatures advocated

for seat increases. Political entrepreneurs have funded ballot ini-

tiative campaigns to increase the size of the California legislature

and to break up the state into several smaller states to address the

government’s perceived unresponsiveness, corruption, and dis-

tance from the people (Wildermuth 2018). Others have called for a

dramatic increase in the size of the US House of Representatives,

suggesting that an increased House would lead to greater citizen

influence on federal policy, improved representation, and less in-

fluence of money on politics (Drutman 2018; New York Times

Editorial Board 2018). Similar claims about the damaging effects

of large districts were made by the Anti- Federalists during the de-

bates over the ratification of the US Constitution (Storing 1981).

Despite its historical and contemporary relevance, scholar-

ship has not generally identified the number of seats in a legisla-

ture and corresponding constituency size as primary determinants

of representational outcomes or evaluations of government in the

United States, although important work has leveraged differences

in Senate constituency sizes (Hibbing and Alford 1990; Lee 1998;

Lee and Oppenheimer 1999) and intra- decadal population changes

in congressional districts (Frederick 2010). Recently, Wolak (2020)

extends this work to the US states, finding that survey respondents

are more likely to distrust their state governments when residing in

states with larger constituency sizes.

Building from this work, I argue that large districts, meas-

ured as the average number of residents per district, foster dis-

satisfaction with representative government. Because district

size structures the nature of constituent– legislator interaction

and levels of district heterogeneity, size influences the effective-

ness of legislator efforts to establish common interest with con-

stituents and build trust and support (Bianco 1994; Fenno 1978;

Lee and Oppenheimer 1999). This study makes two important

contributions to our understanding of legislative institutions and

1202 Commonwealth of Pennsylvania Legislative Journal 88 (2018). The

PA constitutional amendment eventually failed due to the inclusion of the reduc-

tion of the size of the PA Senate, which was added in the House version of the

legislation (see 202 Commonwealth of Pennsylvania Legislative Journal 1330-

1335 (2018)).

461Constituency Size

evaluations of government. First, I show that the American states,

following the example of the US House of Representatives, have

disconnected the size of their legislatures from the size of their

population. This disconnect is a recent development and has led to

a rapid growth of state legislative constituency sizes in the states.

Second, based on a pooled data set of nine Pew Research Center

surveys spanning 15 years, I create a dynamic measure of mean

favorability of state government by state using multilevel regres-

sion and poststratification (MRP). Using both state- level and

individual- level analyses, I find that larger districts are associated

with more negative evaluations of state government. This relation-

ship is not a product of attitudes toward the federal government

(Hetherington and Nugent 2001; Uslaner 2001), legislative pro-

fessionalism (Dyck 2009; Richardson, Konisky, and Milyo 2012;

Squire 1993), state constituency size outliers, or state population

size alone. Legislative institutions appear to moderate how resi-

dents experience and evaluate state government.

The Size of American Legislatures

The number of seats in a legislature is a fundamental feature

of legislative design. In the US context, conflict over the number

of seats in the House of Representatives and the size of congres-

sional districts played a central role during the debate over the

ratification of the US Constitution, with Anti- Federalist authors

arguing that a small House with large districts would produce an

unrepresentative national government dominated by wealthy elite

and distrusted by the public (Storing 1981). Despite such concerns,

the US Constitution was ratified with an original House size of 65

representatives. Congress regularly increased the size of the House

as the nation grew, reaching its current 435 seats after Arizona and

New Mexico were accepted as new states in 1912 (Kromkowski and

Kromkowski 1991). But after it failed to pass a reapportionment

bill following the 1920 Census, Congress instituted automatic re-

apportionment of current House seats (Balinski and Young 1982).

With the House frozen at 435 seats, the constituency size of con-

gressional districts has grown to well over 700,000 residents per

district after the 2010 apportionment.

While scholarship has examined changing congressional ap-

portionment, remarkably little attention has been paid to the size

of US state legislatures. Nearly every state has changed the size

of their state legislature at least once in accordance with growing