![Sample our Social Sciences journals, sign in here to start your FREE access for 14 days]({"type":"image" , "src" : "/sda/110822/TOC-Subject-Sample-2021-Social-Sciences.jpg"})
ABSTRACT
In contrast to strong evidence that increases in district-level African American population size leads to greater support of African American issues in Congress, 40 years of research has been decidedly mixed on whether increases in district-level Latino population size translate to greater support of Latino issues. This paper provides an update to this literature by analyzing members’ of Congress (MCs) support of Latino issues through roll call voting, as collected by the National Hispanic Leadership Agenda (NHLA) in the 113th (2013–2014) through 115th (2017–2018) House of Representatives. First, we find no relationship between NHLA scores and Latino district-level population size among Democratic MCs. The reason is that Democrats’ NHLA baseline support is already very high, regardless of Latino population size. But Republican MCs’ support for NHLA positions in the 113th through 115th Congresses increases as the size of their Latino constituency increases. Interestingly, we find that this relationship is tied to a district’s Latino noncitizen population as well as their Latino voter-eligible population. So, while Democrats on the whole may provide greater substantive representation to Latinos than Republican MCs, Republican MCs are not entirely unresponsive to Latino interests, particularly if they are given sufficient reason to be responsive.
Disclosure statement
No potential conflict of interest was reported by the author(s).
Notes
1 We use the identification terms Latino and Hispanic interchangeably throughout the manuscript.
2 This includes North Dakota, which accounted for eight of the top 20 counties with the fastest growing Hispanic populations between 2000 and 2014 (Stepler and Lopez Citation2016).
3 Despite its wide use in the literature, the application of NHLA scores in scholarly discourse is not without controversy. Among the most prominent issues are: (1) the difficulty with defining Latino interests/issues; (2) whether there is consensus between elite and mass agendas; and (3) the correlation between issues of Latino concern and the progressive agenda. The key question is whether NHLA scores adequately reflect the diversity of the Latino community. While we agree that NHLA may not accurately reflect the complexity of the Latino community writ large, MCs are primed by the NHLA to believe that it does. Indeed, their advocacy letters are designed to accomplish exactly that; to formally notify MCs of the “pro-Latino” position on a given bill and/or vote. So, even if NHLA agenda items and positions are not a perfect reflection of Latino interests, MCs may still behave as if they are. Still, it would be worthwhile for future research to determine the extent to which NHLA’s agenda actually correlates with Latino opinion, particularly given their scorecards’ wide use in the representation literature.
4 See the online appendix for a complete list of the votes flagged by NHLA as relevant to the Latino community.
5 See the online appendix for more details regarding our control variables.
6 Our results are not sensitive to changes in the methodological approach. For example, in light of the cross-sectional time series nature of the data, we also utilized a random-effects model to test our hypotheses. The results are nearly identical.
7 This interpretation does depend on the idea that members of Congress are, in fact, aware of the size of their district’s noncitizen population. We do not find this to be an unrealistic assumption. First, noncitizen population estimates are readily available from the Census Bureau’s American Community Survey and often referenced by media, particularly when reporting on salient issues such as health care, voting laws, and immigration reform. Indeed, the fact that these estimates are likely undercounts (Brown et al. Citation2018) bolsters our findings, as an undercount would produce more conservative results than an overcount. Second, MCs maintain close contact with a wide range of public (e.g., non-profit groups) and private (e.g., business) interests, many of whom rely on, support, and/or oppose noncitizens. Given interest groups’ role in providing expertise to MCs (e.g., Austen-Smith and Wright Citation1996), it is reasonable to assume organizations keep MCs informed on key demographic changes in the district, including those related to noncitizen populations. Third, and most important, few are as knowledgeable about congressional districts as MCs and their staff (e.g., Henderson et al. Citation2021). MCs, after all, have strong electoral incentives to be knowledgeable, particularly when it comes to factors related to economic interests in the district. We therefore find it reasonable to believe MCs have at least an approximate idea of the size of their district’s noncitizen population and how it changes over time.
8 Another benefit of modeling Republican and Democrat separately was that it allowed us to control for ideology, a proven predictor of congressional position taking including NHLA scores (Preuhs and Hero Citation2011). The variance inflation factor for MC Ideology is only about 1.3 in both models. In our judgement, this level of multicollinearity is not so serious as to necessitate eliminating it from the model.
9 Following Kalaf-Hughes (Citation2020), we also controlled for MCs’ previous vote share to account for the possibility that roll call positions on Latino interests are conditioned on incumbents’ electoral security. The variable was statistically insignificant in all models; all other results remained unchanged. In the end, we omitted the variable from our models due to its relatively high collinearity with Democratic Presidential Vote Share in each party model.
10 This brings to attention yet another potentially important variant of our model. The heterogeneity of the Latino community may warrant breaking % Latino into distinctive national-origin populations; namely Mexican, Cuban, and Puerto Rican populations (e.g., de la Garza et al. Citation2018). Our preliminary results indicate positive effects across all three populations for Republican MCs, with district-level Cuban and Mexican population sizes registering the strongest and most significant effects. Sadly, we are unable to give the dynamics of these relationships the space it deserves in this paper and must wait for future research to address this important subject of study.
11 There are discrepancies between two frequently cited sources over whether MCs of Portuguese decent are Hispanic. For example, in their annual publication of elected officials, the National Association of Latino Elected Officials (NALEO) does not list Representative David Valadao as Hispanic. The House Clerk, however, does. The same discrepancies exist for Republican Representative David Nunes (CA-22) and Democrat Jim Costa (CA-16). All three MCs are Portuguese, which NALEO does not include in their tally. We rely on the NALEO list given their affiliation with NHLA. While adopting the Clerk’s tally changes some effect magnitudes, it does not alter our main conclusions.
12 As we would expect due to the smaller sample size, the confidence interval for Latino MCs is larger than the interval for non-Latino MCs. However, the interval is not so large as to render the substantive effects of the interaction meaningless. Furthermore, they stay relatively tight across the entire range of both variables. In all, there is a clear difference in the slopes of each plot and the results are robust to a variety of model specifications.