Party Unity, Ideology, and Polarization in Primary Elections for the House of Representatives: 1956–2012

• DOI: http://doi.org/10.1111/lsq.12092
• Published date: 01 November 2015
• Date: 01 November 2015
• Author: Nicholas Pyeatt

NICHOLAS PYEATT

Penn State-Altoona

Party Unity, Ideology, and

Polarization in Primary Elections

for the House of Representatives:

1956–2012

Increasing party polarization in Congress is a vexing phenomenon for political

scientists, as it offers a theoretical conundrum. Members of Congress have become

increasingly ideologically divided by party in recent years, which seems counterintuitive

as the public electorally punishes representatives for excessive partisanship and ideologi-

cal behavior. One explanation for this result is that members receive benefits for such

behavior during primaries. This article examines the effect of ideological and partisan

behavior on primary challenges and primary vote totals for incumbent House members.

The results show that incumbents receive benefits in the primary from greater levels of

partisanship but not greater levels of ideological extremity. This finding is substantively

important as it provides further insight into the motivation of congressional incumbents

and offers a partial explanation for the rise in congressional polarization.

What I have had to consider is how productive an additional term would be. Unfortu-

nately, I do not realistically expect the partisanship of recent years in the Senate to

change over the shortterm.

—Senator Olympia Snowe (R-ME)

On February 28, 2012, Senator Olympia Snowe (R-ME), one of the

most moderate members of the Senate, announced that she would not

seek reelection at the end of her term (“Snowe, Kerrey Complicate GOP

Goals,” 2012). This led to an active discussion about the implications for

the Republican goal of retaking the Senate, but perhaps the more interest-

ing issue is what this retirement ref‌lects about the status of moderate

politicians in Congress. Snowe, while elected from a state that has trended

Democratic at the national level in recent years, was in no signif‌icant elec-

toral danger; most pundits considered her reelection in 2012 to be almost

LEGISLATIVE STUDIES QUARTERLY, 40, 4, November 2015 651

DOI: 10.1111 /lsq.12092

V

C2015 The Comparative Legislative Research Center of The University of Iowa

perfunctory (e.g., Miller 2012). Snowe’s retirement ref‌lects that the insti-

tution of Congress has become less hospitable to moderates and that most

of the remaining moderates are the vestiges of an earlier era.

The increasing ideological gap between Republicans and Demo-

crats in Congress, known as polarization, is precisely the reason that

Snowe used to explain her retirement from the Senate. Polarization has

been heavily studied by academics, who have sought to understand the

sources of this phenomenon (for an extensive literature review on the

subject, see Fiorina, Adams, and Pope 2010; Theriault 2008). Regardless

of the exact causes, using a variety of measures, it is clear that parties in

Congress have become further apart ideologically over the past several

decades (McCarty, Poole, and Rosenthal 2006; Poole and Rosenthal

1991). This phenomenon seems curious as the public has been found to

electorally reward members of Congress for more moderate behavior

(Canes-Wrone, Brady, and Cogan 2002; Erikson 1971).

Given that the public electorally punishes members of Congress as

they become more ideologically extreme, and because they are highly

focused on reelection (e.g., Arnold 1990; Mayhew 1974), what explains

the movement of Congress to the extremes? One explanation is that pri-

mary elections create incentives to be more extreme in order to appeal to

party activists (Aldrich 1995; Ansolabehere, Snyder, and Stewart 2001;

Burden 2001, 2004), especially with the increase in safe seats since the

1970s (Abramowitz, Alexander, and Gunning 2006). Additionally, pri-

mary elections may reward extremity in terms of presenting incumbents

with fewer in-party challengers and easier nomination contests (Brady,

Han, and Pope 2007).

The view that ideological extremity is a penalty for incumbents in

general elections has recently been challenged. Findings from Carson

et al. (2010) show that the strength of the incumbent’s partisanship has a

greater impact on vote share than the incumbent’s ideological position.

Their conclusions point to the electorate’s ability to differentiate between

the ideological and partisan behavior of incumbents. Though this work

focused on general elections, it brings into question whether ideological

extremity in primaries is actually as advantageous as previously

observed. As voters in primary elections are principally partisan voters

(i.e., their motivation is principally partisan in nature), it seems reason-

able that they should reward partisan behavior but not necessarily

ideological behavior. For these voters who have self-selected to partici-

pate in a party-nominating contest, incumbent partisanship should be the

dominant factor in their vote.

In the analysis to follow, looking at data over the majority of the

postwar period (1956–2012), greater levels of partisan voting by

652 Nicholas Pyeatt