Doubling Down: Inequality in Responsiveness and the Policy Preferences of Elected Officials

• Date: 01 August 2018
• Author: Matthew S. Mendez,Christian R. Grose
• DOI: http://doi.org/10.1111/lsq.12204
• Published date: 01 August 2018

MATTHEW S. MENDEZ

CSU, Channel Islands

CHRISTIAN R. GROSE

University of Southern California

Doubling Down: Inequality in

Responsiveness and the Policy

Preferences of Elected Officials

Is bias in responsiveness to constituents conditional on the policy preferences of

elected officials? The scholarly conventional wisdom is that constituency groups who

do not receive policy representation still obtain some level of responsiveness by legisla-

tors outside of the policy realm. In contrast, we present a theory of preference-induced

responsiveness bias where constituency responsiveness by legislators is associated with

legislator policy preferences. Elected officials who favor laws that could disproportion-

ately impact minority groups are also less likely to engage in nonpolicy responsiveness

to minority groups. We conducted a field experiment in 28 US legislative chambers.

Legislators were randomly assigned to receive messages from Latino and white constitu-

ents. If legislators supported voter identification laws, Latino constituents were less

likely to receive constituency communications from their legislators. There are signifi-

cant implications regarding fairness in the democratic process when elected officials fail

to represent disadvantaged constituency groups in both policy and nonpolicy realms.

Is bias in responsiveness to constituents conditional on the policy

preferences of elected off‌icials? Are legislators who support restrictive

voter identif‌ication laws more likely to exhibit bias in responsiveness

against Latino constituents? Theories of inequality and minority political

representation have focused on the policy representation that minority

and identity groups achieve in legislatures (e.g., Casellas 2011; Grofman,

Handley, and Niemi 1994; Mansbridge 1999; Minta and Sinclair-

Chapman 2013; O’Brien 2015; O’Brien and Rickne 2016; Rouse 2013;

Williams 1998), but they have concentrated much less frequently on

nonpolicy responsiveness by legislators to constituents. Some argue

that constituency groups who do not receive policy representation

from legislators may still receive some level of representation via actions

taken by legislators outside of the policy realm (e.g., Cain, Ferejohn, and

Fiorina 1987; Canon 1999, Grose 2011). Others argue or show, through

LEGISLATIVE STUDIES QUARTERLY, 43, 3, Au gust 2018

DOI: 10.1111 /lsq.12204

V

C2018 Washington University in St. Louis

457

experimental audit studies, that political elites have biases toward

minority constituents when engaging in nonpolicy representation (e.g.,

Broockman 2013; Butler 2014; Butler and Broockman 2011; Butler and

Dynes 2016; Costa 2016; Perez 2015), but they do not link this respon-

siveness bias to legislators’ policy preferences.

We argue that inequality in political representation may occur con-

currently across different dimensions of legislative behavior. The same

legislators who are unlikely to respond to requests for assistance from

minority constituents are also the ones least likely to vote in favor of

public policies that serve the substantive needs of minority constituents.

There are signif‌icant theoretical implications regarding fairness and

accessibility in the democratic process if elected off‌icials fail to represent

disadvantaged constituency groups in both policy and nonpolicy realms.

We present a theory of preference-induced responsiveness bias by

public off‌icials. Due to electoral or personal biases, legislators are more

likely to favor some groups in their constituencies over other groups

when it comes to responsiveness, which we def‌ine as responding to con-

stituent requests for assistance (or, more broadly, any attempt to respond

to the needs of a constituent other than through policymaking or position

taking). Our theory argues that legislator biases in responsiveness in

favor of a majority group of constituents over a minority group of con-

stituents are most likely among legislators with exogenous preferences

for public policies that could have a disparate impact on the minority

group. Thus, responsiveness biases are conditional upon and exacerbated

by legislator policy preferences. This is in direct contrast to the scholarly

conventional wisdom that legislators whose policy preferences

diverge from a constituent will nevertheless engage responsively

with the constituent in other ways. While we use the term bias to

describe the lack of responsiveness by legislators to certain groups

of voters relative to other groups, it is important to note that this bias

in responsiveness could be due to electoral considerations where a

legislator does not want to mobilize a nonsupporter or due to per-

sonal discrimination by the legislator.

Empirically, we examine the case of restrictive voter identif‌ication

laws proposed or passed in US state legislatures. These laws are more

likely to be implemented in states with large minority populations

(Bentele and O’Brien 2013; Biggers and Hanmer 2017), are more likely

to be passed by legislators conditional on the minority population of their

districts (McKee 2015), and these and other policies are supported more

frequently when the issue is racialized or framed in terms of group self-

interest (Mangum 2008; Tesler 2012; Wilson, Brewer, and Rosenbluth

2014). It has also been claimed that these laws have the potential to have

2 Matthew S. Mendez and Christian R. Grose

458

a differential impact on Latino citizens relative to white citizens (Barreto,

Nu~

no, and Sanchez 2009), as Latino citizens are less likely to have voter

identif‌ication and thus may be less likely to vote in states with these

laws.

1

To test our theory’s implications, we conducted a f‌ield experiment

with state legislators as subjects in 28 legislative chambers in 14 states

with large Latino populations. Field and natural experiments of elected

off‌icials are fairly uncommon in the study of political institutions (Costa

2016; Grose 2014; O’Brien and Rickne 2016; Yoshinaka 2 016),

although they are becoming more frequent in work examining questions

of representation (e.g., Bergan 2009; Butler 2014; Butler and Broockman

2011; Butler and Kousser 2015; Butler and Nickerson 2011; B roockman

2013; Grose, Malhotra, and Van Houweling 2015; K alla and Broockman

2016; Mendez 2015; Preece and Stoddard 2015). Field ex perimental

studies have shown that legislators are generally less likely to respond to

black constituents when compared to white constituents (Butler and

Broockman 2011; Broockman 2013) and that local election registrars are

less likely to give complete information to Latino voters regarding voter

identif‌ication requirements (Faller, Nathan, and White 2015).

2

No study has attempted to assess whether revealed behavior by

legislators is associated with support for restrictive voting rights laws,

which is a question central to the policy debate over voter identif‌ication

and to our theoretical understanding of political representation. This is

the f‌irst study to examine the responsiveness of legislators to constitu-

ency entreaties conditional on legislators’ policy preferences and one of

the only experiments to examine legislator responsiveness to Latino ver-

sus white constituents. In our study, legislators were randomly assigned

to be treated by both Latino and white constituents. We compared the

differential in legislator response rates to the constituent treatments by

whether the legislator supported voter identif‌ic ation laws during the same

legislative session. If the legislator supported voter identif‌ication, Latino

constituents received legislator responses at much lower rates than white

constituents. This result holds even when conditioning on the legislator’s

political party and other variables. Supporting our theory, we f‌ind that

responsiveness bias is conditioned by legislator policy preferences.

Theory: Responsiveness Bias to Constituents

and Legislator Policy Preferences

Responsiveness to constituent needs and requests is an important

component of a legislator’s representational repertoire (Cain, Ferejohn,

and Fiorina 1987; Grose 2011). Responding to constituency queries is

3Doubling Down 459