Many are called but few are chosen: the emergence of Latino congressional candidates

ABSTRACT

Prior work has shown that Latino candidates do not under-perform compared to other candidates, and that Latino underrepresentation is a function of supply rather than demand (Juenke, Eric Gonzalez. 2014. “Ignorance Is Bias: The Effect of Latino Losers on Models of Latino Representation.” American Journal of Political Science 58 (3): 593–603; Juenke, Eric Gonzalez, and Paru Shah. 2015. “Not the Usual Story: the Effect of Candidate Supply on Models of Latino Descriptive Representation.” Politics, Groups, and Identities 3 (3): 438–453). Building on this work we investigate what factors influence the prospects of Latino candidates from the supply side. Examining candidate emergence from the supply side requires that we examine not only the winners and losers, but also those who could have entered a pool of candidates but did not. Using an original dataset of latent and primary candidates running for U.S. House seats from 2004 to 2014, we show that Latino candidacy at the pre-emergence stage is sensitive to potential challengers and that party elites under-support Latino candidates relative to others. Latinos are predicted to receive fewer endorsements than non-Latinos, and those endorsements influence candidates’ chances of winning office. Our findings suggest that the poor integration of Latinos into partisan networks is largely responsible for their limited presence in candidate pools and elected office.

KEYWORDS:

• Latino politics
• representation
• political parties
• minority politics
• U.S. Congress

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes

1 Colorado Independent. Patch.com. “A record number of Latinos join State Legislature.” https://patch.com/colorado/across-co/record-number-latinos-join-state-legislature

2 For a counterargument to the minority empowerment and co-ethnic candidate mobilization theories see (Fraga 2016).

3 That list is derived from microsample data from the 2000 U.S. Census and is publicly available at http://names.mongabay.com/data/hispanic.html.

4 National Association of Latino Elected Officials, 2015. “National Directory of Latino Elected Officials.” https://d3n8a8pro7vhmx.cloudfront.net/naleo/pages/171/attachments/original/1440570181/2015_National_Directory_of_Latino_Elected_Officials.pdf?1440570181. For previous NALEO directories we directly contacted the organization, which provided access to its listings dating back to 2004.

5 https://ballotpedia.org/.

6 http://www.ourcampaigns.com/

7 While not the central focus of our project, it is worth noting that these numbers are significantly higher than the number of general-election candidates (or seated incumbents) in Congress who are Latino. That is, there is much more Latino political contestation than there are Latino political successes. This is consistent with prior work that finds that the greater the presence of minority candidates in a primary contest the higher the level of elecotral competitiveness (Branton 2009).

8 The modal piece of information the authors imputed was candidate age. Because resources like Our-Campaigns and Ballotpedia track the dates candidates received their college degrees, we imputed age under the assumption that candidates completed their bachelors’ degree at age twenty-two or their law degree at age twenty-five.

9 Rep. Jim Costa is of Portuguese descent. However, Costa self-identities as Latino, and as such, we have categorized him as a Latino legislator in this study. Two other additional members of Portuguese descent, Reps. Devin Nunes and David Valadao, are not categorized as Latino. Both of these members do not self-identify as Latino. http://www.foxnews.com/politics/2013/02/05/how-many-latinos-serve-in-congress-depend-on-whom-ask.html. We re-ran our analysis recoding Rep. Jim Costa as non-Latino and we did not find any substantive changes to our results.

10 We used a variety of search methods to identify almost-rans, prime among them were Google News, LexisNexis, and the crowd-sourced campaign website OpenCongress, where users could identify potential candidates for races. We did not observe any substantive disagreement between the news sources over who were each district’s potential candidates.

11 http://cookpolitical.com/house/maps

12 Including a dummy variable for whether the candidate is African-American produces a statistically insignificant coefficient on the interaction between challenger quality and race and the propensity to enter the candidate pool. This is almost certainly due to two factors: the geographic concentration of African American political candidates to areas where African American Democrats are competing with each other, and the comparatively higher network connectedness of African American political candidates. Our African American candidates are concentrated in the southeast, where the Democratic Party is generally considered to have a strong capacity and desire to support established African American candidates for higher office (Lieberman 2014; Schickler 2016).

13 If we consider perennial candidates to be someone who has run for office more than five times, our data include ten perennial candidates. While the average number of campaigns waged in our dataset prior to winning first office is 0.49, our most perennial candidates have run 8, 8, 9, 11, 13, and 19 times without apparently winning an office. Additional analyses suggest that the inclusion or exclusion of these candidates does not substantively change our results.