Spillover Effects of Minority Representation on Majority Bureaucrats’ Behavioral Change

• Published date: 01 November 2021
• Author: Danyao Li
• Date: 01 November 2021
• DOI: http://doi.org/10.1111/puar.13428

Research Article

Spillover Effects of Minority Representation on Majority Bureaucrats’ Behavioral Change 1071

Abstract: Representative bureaucracy scholarship has rarely examined whether passive representation of minorities

changes the behavior of majority bureaucrats. To address this omission, this article explicitly tests the relationship

between the two, in the context of traffic law enforcement. Using individual-level data over multiple years in

Washington and South Carolina, analyses show that minority representation has spillover effects on decisions made

by White officers. They are more likely to treat drivers of color similarly to White drivers, when working on a more

racially representative police force. These findings support an underexplored causal mechanism whereby representation

improves policy results for historically underprivileged groups, making a theoretical contribution to representative

bureaucracy. It also has managerial implications for practitioners who seek to reform future law enforcement for

greater racial equity in policing outcomes.

Evidence for Practice

• Behavioral changes on the part of majority administrators, engendered by minority representation in the

public organization, further contribute to equitable treatment of minority clientele. It is oftentimes an

overlooked benefit of representative workforce.

• It is important for public organizations to create a representative, diverse, and inclusive working environment

that encourages learning and adaptation of both individual employees within the organization and the

organization as a whole.

• Increasing representation and diversity on police force is not necessarily an unattainable goal. Under right

conditions, a small increase in representation can go a long way to produce and deliver desirable outcomes to

minority citizens.

The theory of representative bureaucracy

contends that a bureaucracy that is

descriptively representative of the constituent

population will make decisions that correspondingly

are responsive (Long 1952; Mosher 1968). Numerous

studies have empirically established the link between

demographic representation and beneficial outcomes

to the groups being represented (e.g., Atkins and

Wilkins 2013; Keiser et al. 2002; Meier 1993;

Meier and Stewart 1992). Work to date has almost

exclusively focused on two mechanisms that underlie

that link. One is “advocacy” by minority and

female bureaucrats—they likely embrace the “role

of representative,” under right conditions, and act

actively in the way that advances substantive interests

for bureaucratic clients who share their characteristics

(e.g., Meier and Nigro 1976; Selden 1997). The other

is “symbolization.” It hypothesizes that representation

may change bureaucratic clients’ attitude and

behavior, which eventually adds to their improved

outcomes (e.g., Gade and Wilkins 2013; Riccucci,

Van Ryzin, and Li 2016).

An important gap in literature remains, however,

because previous research has not been able to

conclusively decide whether passive representation

of minorities in public organizations also changes

the behavior of majority members, although scholars

have long speculated this theoretical possibility (see

Lim 2006 for a review). Lack of empirical evidence

may be, in part, due to data limitations. Findings

on active representation thus far draw heavily on

aggregate data at the organization level, which do not

allow scholars to disentangle behavioral patterns of

bureaucrats of different groups.

This article theorizes that increased minority

representation in public organizations will alter

majority bureaucratic decision-making about minority

clientele. This behavioral change is a result of several

plausible scenarios, including a reduction in their

implicit biases toward minorities after repeated

contacts with minority colleagues at work, peer

pressures, and/or change in organizational norms

and regulatory policies that curb discriminatory

Danyao Li

Indiana University Bloomington

Spillover Effects of Minority Representation on Majority

Bureaucrats’ Behavioral Change

Danyao Li is a doctoral candidate at

The Paul H. O’Neill School of Public and

Environmental Affairs, Indiana University

Bloomington. She is interested in questions

that concern issues of social equity

and justice. Her research focuses on

representation and diversity, bureaucratic

decision-making, and the performance of

public organizations, typically with a specific

application to law enforcement.

Email: danyli@indina.edu

Public Administration Review,

Vol. 81, Iss. 6, pp. 1071–1091. © 2021 by

The American Society for Public Administration.

DOI: 10.1111/puar.13428.

1072 Public Administration Review • November | Decembe r 2021

behavior. Using data on millions of traffic stops performed by the

State Patrol in two states over multiple years, Washington (2009–

16) and South Carolina (2005–16), within-officer analyses find

supporting evidence for this theoretical proposition. Specifically,

White officers, who work in police districts that are more racially

representative, demonstrate behavioral shifts toward drivers of color

that eventually reduce racial disproportion in policing outcomes.

Interestingly, these patterns are more evident and robust when

it comes to the effects of Hispanic representation. Although this

research is not able to pin down the source(s) of observed behavioral

alterations, findings provide empirical support for the theorizing

that minority representation likely has positive spillover effects on

majority members of the organization. Thus, this article addresses

a significant omission in representative bureaucracy literature and

advances our understanding of representation processes by unveiling

a causal mechanism that is rarely explored by previous work.

The rest of the article starts with a brief review of representative

bureaucracy theory and connects it to the large literature on race

and policing. It then examines causal mechanisms of representation

to develop the hypothesis. Next, it introduces the empirical context

of traffic policing, data, measures, and methods. After presenting

results, it concludes with a further discussion of the findings,

contributions, implications as well as limitations that may inform

directions for future inquiries.

Representative Bureaucracy: A Brief Overview

The theory of representative bureaucracy argues that a

demographically diverse bureaucracy will lead to democratic

outcomes through responsive public policies and administrative

actions (Kingsley 1944; Meier 1993; Meier and Nigro 1976;

Mosher 1968; Selden 1997). Mosher (1968) distinguishes between

the concepts of “passive representation” and “active representation.”

The former notion concerns the degree to which bureaucrats mirror

demographic origins of the local population, in terms of race/

ethnicity, gender, class, or other characteristics (Kingsley 1944;

Mosher 1968). Research on passive representation constitutes

the focus of early representation studies and is greatly motivated

by a normative view that descriptive representation is necessary

for organizational legitimacy. Active representation, however,

refers to the bureaucratic outputs that are responsive to those

being represented (usually historically underprivileged groups)

(Meier 1993). The distinction, but more importantly, a presumed

connection between the two, has informed subsequent analyses and

theoretical advancements in representative bureaucracy literature for

decades.

A large and still growing body of research across sectors of public

service has led empirical support to the notion that passive

representation can translate into substantive policy benefits for

the represented clients when certain preconditions are met. These

policy domains include public education (Meier and Stewart 1992;

Nicholson-Crotty, Grissom, and Nicholson-Crotty 2011), federal

contracting (Fernandez, Malatesta, and Smith 2013), farmers loan

application (Selden 1997; Selden, Brudney, and Kellough 1998),

child support (Wilkins 2006; Wilkins and Keiser 2004), local

police (Meier and Nicholson-Crotty 2006), Equal Employment

Opportunity Commissions (EEOC) charges (Hindera 1993),

environmental justice (Liang, Park, and Zhao 2020), and many

others. Scholars have identified several conditions under which

representation likely manifests, such as the identity saliency of the

demographic group under discussion, discretion that bureaucrats

have in implementing policies, and direct and substantial impact of

bureaucratic actions on the group in question (Keiser et al. 2002;

Meier 1993; Meier and Bohte 2001; Riccucci and Meyers 2004;

Sowa and Selden 2003).

Race, Representation, and Policing in the United States

Policing is one of the frequently examined contexts by representative

bureaucracy scholarship (e.g., Nicholson-Crotty, Nicholson-Crotty,

and Li 2019; Theobald and Haider-Markel 2008; Wilkins and

Williams 2008). This body of research is intertwined with work

from other disciplines that seeks to understand racial disproportion

in policing outcomes, but goes on to discuss, from the viewpoint of

public policy and administration, whether increasing representation

is a remedy available to address the stark racial disparities that

widely exist and sustain in U.S. policing. Before discussing

where empirical evidence stands on the effectiveness of police

representation in reducing racial inequalities, it is beneficial to start

with a discussion of the broader literature on race and policing.

It goes beyond the scope of this article to thoroughly review that

long-running research tradition, but a brief overview is necessary

because policing is related to both the force responsible for policing

and of the characteristics of the population being policed. It gains

us a better understanding of how representation relates to the larger

conversation on racial justice in law enforcement.

Race and Policing in the United States

Previous work in sociology and criminology, among other

fields, has identified three major categories of contributors

to racial disproportion in policing outcomes. These include

factors at psychological/individual, sociological/situational, and

organizational/institutional levels. Research in the first category

examines how social identity, prejudice, and many other elements

in individual officers predict their behavior at the frontline. The

second stream of research turns attention to social dynamics and

situational cues in police-citizen encounters that are used by officers

to inform their decision-making. The third approach argues that

characteristics of institutions, such as police culture, administrative

practices, and larger legal contexts, impact policing results in a more

systematic and broader way (see Nicholson-Crotty and Nicholson-

Crotty 2020 for a review).

Note that components at three levels usually co-exist, interact, and

contribute collectively to racially biased policing outcomes. First,

police targeting at minorities may be driven by officers’ personal

stereotypes or overt racism in some cases. For example, social

categorization is commonly used by individuals as a psychological

mechanism to make sense of a vast social world. Differences

between individuals evoke a subjective classification process that

divides others into in- and out-groups. This categorization acts

as a system of orientation for self-reference and a guide to action

(Tajfel 1982; Tajfel et al. 1971). One detrimental consequence of it,

however, is that “categorization, per se, propels the individual down

the road to bias” (Wilder 1986). As Tajfel and Wilkes (1963) show

in their experiments, it leads to an accentuation of within-group

similarity and between-group difference. Combined with “us vs

them” mentality that is deeply ingrained in police since very early