Nevertheless, He Persisted: White Men and the Links Between Incumbency and Group Descriptive Representation

• Published date: 01 December 2023
• DOI: http://doi.org/10.1177/10659129231173340
• Author: Christian D. Phillips
• Date: 01 December 2023
• Subject Matter: Articles

Article

Political Research Quarterly

2023, Vol. 76(4) 1691–1706

© The Author(s) 2023

Article reuse guidelines:

sagepub.com/journals-permissions

DOI: 10.1177/10659129231173340

journals.sagepub.com/home/prq

Nevertheless, He Persisted: White Men and

the Links Between Incumbency and Group

Descriptive Representation

Christian D. Phillips

1



Abstract

Descriptive representation is shaped by more than the number of seats group members win—it is also informed by how

group members hold on to seats long after election day. Drawing from the literatures on women of color, women, and

minorities in politics, this study argues that the relationship between incumbents and descriptive repre sentation is

different among groups that hold distinctive political positioning and power. To uncover those differences, the article

introduces a new measure called descriptive maintenance, which accounts for a group’s ability to retain descriptive

representation in a seat across unique group members and elections. This multidimensional approach expands current

conceptualizations by treating incumbency and descriptive representation as interrelated, group-level, phenomena. This

framework is tested using nearly 60,000 observations of state elections data in an analysis focused on a group of state

legislative incumbents that has rarely been explicitly examined, and yet outnumbers all others: white men.

Keywords

descriptive representation, incumbents, race, state legislatures, elections, candidates

Since the 1960s, high rates of incumbent success have

become a more consistent feature of American politics,

raising the specter of a grave political challenge: if an

incumbent receives electoral advantages that make it

extraordinarily diff‌icult for voters to dislodge that person

from off‌ice and replace them with a candidate of their

choosing, democratic mechanisms may be compromised.

For groups that are marginalized in politics, this demo-

cratic encumbrance can carry signif‌icant costs. For groups

that hold considerable power in politics, the same en-

cumbrance may be a boon that further solidif‌ies their

positions atop social and political hierarchies.

How might we think differently about incumbency’s

implications for representation if instead of focusing

only on individuals retaining seats, we also considered

how descriptive groups maintain their representation in

a seat? In this framework, there are not only individual

incumbents but also incumbent descriptive groups.

These descriptive groups may accrue some set of

electoral benef‌its because a group member has already

held a particular seat and is replaced by another group

member. Given the limited number of seats in any given

legislature, these group level benef‌its bear on the

representation opportunities of the group in question, as

well as other groups competing for seats in the elected

body.

The representation literatures focused on women and

women of color (Sanbonmatsu 2006;Hardy-Fanta et al.

2016;Reingold, Haynie, and Widner 2020;Schwindt-

Bayer, and Leslie, 2005;C.D.Phillips 2021) and racial

minorities (Preuhs and Juenke 2011;Fraga, Gonzalez

Juenke, and Shah 2019) suggest the necessity of ex-

plicitly accounting for descriptive groups of incumbents.

However, extant research has not yet examined how

descriptive groups’ability to maintain representation,

once a group member is the incumbent, may vary.

This study argues that the benef‌its of incumbency for

descriptive representation are multidimensional, and in-

clude both the retention of individuals who are members

1

Department of Political Science and International Relations, University

of Southern California, Los Angeles, CA, USA

Corresponding Author:

Christian D. Phillips, Department of Political Science and International

Relations, University of California, 3518 Trousdale Parkway, CPA

330, Los Angeles, CA 90089-0001, US.

Email: christian.d.phillips@usc.edu

of different descriptive groups, and the maintenance of

descriptive representation in a seat across unique indi-

viduals from the same descriptive groups. Because in-

cumbency benef‌its and descriptive representation are both

informed by group level processes, the relationships be-

tween incumbency and group representation are socially

and politically specif‌ic to each group. Once a member of a

descriptive group wins a seat, the groups’capacity for

maintaining group representation in that seat is shaped by

their power and political positioning, particularly within

elite networks that shape candidate pipelines.

Using a dataset encompassing 57,812 state legislative

general elections from 1996–2015, this paper introduces a

novel measure that captures the group-level links between

incumbents and representation: descriptive maintenance.

This new measure is used to analyze patterns of over-

representation among white men, who make up the largest

incumbent group of descriptive representatives. As

Rainbow Murray and others have argued (Murray 2014;

A. Phillips 1998), explicitly reframing studies of repre-

sentation around the problem of overrepresentation

generates intellectual traction to evaluate particular

shortcomings in democratic processes, which may be less

accessible in studies of underrepresented groups.

The analysis demonstrates that the relationship between

white men’s status as an incumbent group and their main-

tenance of descriptive representation across unique group

members is quite different from that of other groups . This

f‌inding contributes directly to our understanding of the

advantages that incumbency may encompass, by showing

that they may extend beyond the individual, and should not

be treated as fungible across groups. These f‌indings also

underlie a concept discussed in the conclusion, labeled

“descriptive conservatism,”whereby the maintenance of

descriptive representation by a group that is historically and

numerically dominant may contribute to patterns of repre-

sentation that preserve their political presence and power.

By analyzing the links between incumbency and de-

scriptive representation as a group-level set of processes,

this study cautions that analytically treating political

institutions—like incumbency—as seemingly “universal”

in their advantages may obscure how they contribute to

the continuing domination of white men in politics.

Bringing the Group In: Incumbents and

Descriptive Representation

In the American gender and politics literature, individual

incumbency advantages are often theorized as directly

informing candidates’perceptions of the risks and op-

portunities of a given seat or election. Candidates for

off‌ice are strategic in deciding where and when to run for

off‌ice, and have other choices for professional gain and

political leadership available to them if electoral oppor-

tunities seem too high risk (Fulton 2012;Crowder-Meyer

and AdrienneSmith 2015;Fulton et al. 2006;C.D.

Phillips 2021). By dramatically raising the risk of running

for an occupied seat relative to an open seat, incumbents’

well-documented electoral advantages at the individual

level constrain the opportunity context for other indi-

viduals seeking off‌ice (Stone et al. 2004). Studies of how

incumbency functions as a structural impediment to

women’s representation (since most incumbents are men)

often begin from this individual opportunity logic (Darcy

1994;Palmer and Simon 2001;Carroll 1994;Lawless and

Theriault 2005).

The emphasis on individual advantages also often

extends to empirical analyses of the representative op-

portunities taken up by incumbents. For example, studies

of incumbency and women’s representation tend to rely

on individual measures of incumbents’ability to maintain

their seats—individual candidate vote shares, retention or

exit, fundraising, etc.—and their relationship to women’s

levels of candidacy or representation. This treatment—

aggregating individual level data to the group level—

recognizes that the scale of one descriptive group of in-

cumbents has consequence for the scale of another.

However, that measurement approach does not neces-

sarily theoretically tie differences in the effects of in-

cumbency to the identities or relative group power of

either men or women (c.f. Schwindt-Bayer and Leslie,

2005);

1

incumbency is treated as the same type of in-

stitution with the same types of effects on representation

opportunities for women and men. In this way, studies

relying on individual measures of incumbency’s rela-

tionship to representation may underestimate, or obscure,

group level differences in the advantages often ascribed to

incumbency.

To illustrate, retention rates have been frequently used

as an outcome variable in studies of women’s represen-

tation. This measure is focused on an individual outcome

(whether or not candidate X was re-elected), but is often

used in the aggregate to illustrate group-level differences

in the experiences of women and men incumbents (Palmer

and Simon 2005;Schwindt-Bayer and Leslie, 2005;

Fulton 2012). Table 1 reports the mean retention rate

among state legislative general election winners across

eight race-gendered groups (Hawkesworth 2003;Brown

2014;C.D.Phillips 2021).

Incumbents, from any of the groups above, are clearly a

good bet to win in state legislative elections. The rates

presented in Table 1 rely on pooled data across nearly two

decades of elections, but these rates do not vary signif‌i-

cantly over time. The statistically signif‌icant difference in

the mean retention rates (p> 0.001) among these groups

provides an initial suggestion, often raised in the women

in politics literature, that the electoral benef‌its that

1692 Political Research Quarterly 76(4)