Income Inequality and Ideological Positions in the U.S. Congress

• Published date: 01 September 2021
• Date: 01 September 2021
• DOI: 10.1177/1065912920922318
• Subject Matter: Articles

https://doi.org/10.1177/1065912920922318

Political Research Quarterly

2021, Vol. 74(3) 599 –614

© 2020 University of Utah

Article reuse guidelines:

sagepub.com/journals-permissions

DOI: 10.1177/1065912920922318

journals.sagepub.com/home/prq

Article

Two of the most critical, dominant, and studied national

trends in modern American politics are increasing income

inequality (APSA Task Force Report 2004; Atkinson,

Piketty, and Saez 2011; Piketty and Saez 2003) and

increasing ideological and partisan polarization, particu-

larly in the U.S. House of Representatives (Aldrich 1995;

Fiorina 2002; Jones and McDermott 2009; Rohde 1991;

Theriault 2008). As each of these trends has continued to

rise, there is also an emerging literature addressing vari-

ous aspects of these two phenomena. These include the

effects of the rise of partisan class differences (Stonecash

2000); the effects of income inequality on ideological

polarization (IP) in the U.S. Senate (Garand 2010; Hayes

2013) and state legislatures (Chaker 2017); the increasing

influence of district income on the ideological positions

of the members of the U.S. House (McCarty, Poole, and

Rosenthal 2006, 34–44); the effects of income and/or

income inequality on public opinion, citizen partisanship,

and voting (Bartels 2008; Gelman 2008; Gilens 2005;

Kelly and Enns 2010; McCarty, Poole and Rosenthal

2006; Newman and Hayes 2019; Soss and Jacobs 2009);

and the direct national-level relationship between the two

aggregate trends (Duca and Saving 2015; Kwon 2015).

All of these studies, and others, offer important theo-

retical and/or empirical insights. And, many studies—

like the popular impression—often suggest that the

increase in both of these aggregated national-level trends

are causally connected. Yet, no study has directly theo-

rized and tested the individual-level causal relationship

that must underlie these well-known aggregated national-

level trends: specifically, the effect of district-level

income inequality on the ideological positions for indi-

vidual members of the U.S. House of Representatives.

This is especially troubling for at least three main rea-

sons. First, although the aggregated national-level trends

seem graphically convincing, the actual underlying indi-

vidual-level relationships should not be inferred from

them. To infer individual-level behavior from aggregate-

level trends runs the risk of ecological fallacies. Second,

much of the current literature that discusses or analyzes

income inequality often imply, assume, or posit different

economic contexts and different theoretical relationships.

When these are logically compared, they actually lead to

a number of different, but typically unstated and unex-

plored, individual-level political outcomes. Interestingly,

many of these outcomes do not logically suggest an

increase in ideological and partisan polarization of

922318PRQXXX10.1177/1065912920922318Political Research QuarterlyLadewig

research-article2020

1University of Connecticut, Storrs, USA

Corresponding Author:

Jeffrey W. Ladewig, Associate Professor, Department of Political

Science, University of Connecticut, 365 Fairfield Way, U-1024, Storrs,

CT 06269, USA.

Email: jeffrey.ladewig@uconn.edu

Income Inequality and Ideological

Positions in the U.S. Congress

Jeffrey W. Ladewig1

Abstract

Over the past twenty years, there has been much discussion about two of the most important recent trends in

American politics: the increase in income inequality in the United States and the increase in ideological and partisan

polarization, particularly in the U.S. House. These two national-level trends are commonly thought to be positively

related. But, there are few tested theoretical connections between them, and it is potentially problematic to infer

individual-level behavior from these aggregate-level trends. In fact, an examination of the literature reveals, at least,

three different theoretical outcomes for district-level income inequality on voter and congressional ideological

positions. I explore these district-level theoretical and empirical possibilities as well as test them over decades with

three different measures of income inequality. I argue and demonstrate that higher district levels of income inequality

are related to higher levels of ideological liberalism in the U.S. House. This stands in contrast to the national-level

trends, but it tracks closely to traditional understandings of congressional behavior.

Keywords

U.S. Congress, ideology, income inequality

600 Political Research Quarterly 74(3)

members of Congress (MCs). Third, there is often an

incongruent matching of theory and data: income is

sometimes used when an actual measure of income

inequality is posited, national-level data are sometimes

used to describe district-level phenomena, and the theo-

retical discussions rarely differentiate among these.

Given the attested importance of both income inequality

and ideological positioning—much less the combination

of the two—such a mix of inferences, theories, and results

calls for a treatment that tries to bridge the differences in

the current literature as well as theoretically and empiri-

cally examines directly any potential relationship between

district-level income inequality and the ideological posi-

tioning of MCs.

In order to fill these critical gaps between these two

vital issues, this manuscript makes two primary contribu-

tions. First, I argue that the literature on income inequal-

ity and MCs’ ideological responses suggests two critical

questions: How does the national-level income inequality

trend disaggregate to congressional districts? And, in

light of income inequality, do MCs still represent their

constituencies as the traditional congressional literature

would suggest? These questions are foundational to any

study of congressional representation, in general, and of

these two phenomena, in particular. Second, a possible

reason why there has been no studies that directly analyze

district-level income inequality and the ideological posi-

tions of members of the U.S. House is that there is little

consistent district-level data on income inequality over

time. The U.S. Census has now offered a measure of con-

gressional district-level income inequality since 2006.

This is a good start. But, to get a longer time series that

can better capture some of the critical moments in both

time series, I also develop and employ an original data set

that measures income inequality at the district-level from

1993 to 2016.

In sum, I argue and demonstrate that the increasing lev-

els of income inequality at the national level are generally

replicated in most congressional districts, but these dis-

trict-level increases are related to higher levels of ideo-

logical liberalism among all MCs—Democrats and

Republicans alike. As such, despite the popular percep-

tions as well as the graphical appeal of the national-level

trends, income inequality does not seem to be a direct

source of partisan IP. This conclusion, however, does track

closely to the traditional theoretical and empirical under-

standings of congressional and voter behavior. Specifically,

the higher the level of district income inequality, the

median voter is by definition relatively poorer, and thus,

the greater the constituent pressures on the MC for some

form of economic liberalism. In these ways, this study

takes significant steps forward in better understanding

both of these vital phenomena in American politics today.

Theoretical Relationships between

Income Inequality and Ideological

Positions

The topics of income inequality and of congressional

polarization have been prominently intertwined by aca-

demics and journalists. Many of these discussions tap

into, but also deviate from, traditional understandings of

constituent representation and congressional behavior.

The possible connection between income inequality and

congressional polarization is largely inferred from the

two national-level trends being graphically quite similar.

The Gini coefficient (GC) is probably the most recog-

nized measure of income inequality. It typically scores

the income variation among individuals, families, or

households within a given geographic area. The scores

range from zero, which indicates a perfectly equal distri-

bution of income, to one, which indicates that a single

measured unit earns all of the income and the rest of the

units do not earn any income. The annual GC for families

in the United States indicates that the level of income

inequality was at a relatively low level (average GC =

~.36) and often decreasing from the beginning of the

post–World War II era through the mid-1970s. But, in the

mid-1970s, there is an inflection point, and the GC

increases rather steadily since then to a series high (GC =

.452) in 2016 (see Figure 1).

The figure also shows that IP in the U.S. House fol-

lows a very similar pattern (r = .95). The IP measure can

range from zero, which indicates that the two parties have

identical ideological medians, to 2, which indicates that

the median member of one party is perfectly conservative

Figure 1. National income inequality and ideological

polarization: 1947–2017.

Source. The annual Gini data are from the U.S. Census (see https://

www2.census.gov/programs-surveys/cps/tables/time-series/historical-

income-families/f04.xls). The ideological polarization scores for each

Congress are derived from the DW-NOMINATE data (see https://

voteview.com).