President Obama’s decision to put off executive action on immigration until after November was perceived as a response to political pressures. Recent polling indicates that support for immigration reform has declined and voters are evenly split with respect to which party is more likely to do a better job handling immigration.
In this context, the political calculations underlying Obama’s decision rests upon three assumptions: Executive action is a net negative for the Democrats’ Senate candidates; any potential decrease in Latino turnout that the delay engenders will have little effect because Latino voters, outside of Colorado, are not sufficiently concentrated in states with competitive Senate races; by acting after November the Democrats will win back the support of immigration-centric voters angered by the delay.
I examine the evidence relevant to the first two assumptions.
Support for the first assumption is evident in polling, since public opinion on this issue is somewhat muddled. In a recent New York Times/CBS Poll, 10 percent of respondents indicated that immigration was the most important issue shaping their vote for Congress (14 percent indicated it was the second most important issue) and over two-thirds supported allowing unauthorized immigrants to stay in the country, 54 percent supported a pathway to citizenship, and 51 percent supported executive action. At the same time when asked if they would be more or less likely to vote for congressional candidates who support a pathway to citizenship 30 percent said this would make them more likely, 39 percent indicated it would make them less likely, and 26 percent responded it would make no difference.
In contrast, for most Latinos immigration reform is the animating issue. A June poll conducted by Latino Decisions for the American Progress Fund suggests that enthusiasm for voting among Latinos would decrease by 54 percent and their support for Democratic candidates would decrease by 57 percent if executive action were not taken before the election. Although these data are from a national sample and may not project to particular state contexts, they suggest the centrality of immigration to the political participation of Latino voters.
As for the second assumption, clearly Colorado is the state where any decrease in Latino turnout hurts the Democrats’ prospects. To examine this potential, Table 2 presents the partisanship of registered Colorado Latinos, sorted by their probability of voting as estimated by Latino Decisions and L2’s micro-targeting models. While 72 percent of Colorado Latinos are likely Democratic voters, just 45 percent have a high probability of voting in November. What about the Latino voters who favor the Democrats or are persuadable, but have lower probabilities of voting (the blue-shaded cells)? If these voters were mobilized in response to executive action, then the pool of potential Latinos voting Democratically would more than double. Absent executive action, there may be little incentive for these voters to participate in November.
At the other end of the spectrum is Arkansas; a state with a fraction of the number of registered Latinos as Colorado (see Table 3). Note that roughly half of Arkansas Latinos are independents. As a consequence, less than 20 percent of the Latino electorate has a high probability of voting Democratic — a number that could be potentially increased twofold if Latino voters in the blue cells were mobilized.
Although North Carolina (Table 4) is home to a growing Latino voting population, just 10 percent of these voters have a high probability of voting Democratic in November — a total that is close to the share of Latinos who are predicted to vote Republican. To be sure, there are significantly more Democratic leaning or persuadable Latino voters (the blue cells) than Republicans in North Carolina. The lack of executive action coupled with Democratic Sen. Kay Hagan’s statements urging Obama not to act may give these voters little reason to participate in November.
The contours of Latino participation for the three other states—Georgia, Kansas, and Michigan—are similar to North Carolina. With the exception of Georgia, likely Democrats and those who might be persuaded to vote Democratic far outnumber likely Republicans. Among Latinos with a high probability of voting, the gap narrows considerably and in Georgia our model predicts more Republican than Democratic high probability voters.
More generally, these data highlight the lack of strong partisan identities for many Latinos. In three states (Arkansas, Georgia, and Michigan) independents are the modal category and in Kansas and North Carolina likely Democrats are a plurality. Thus, only in Colorado, where the Democrats have made significant investments in engaging Latino voters in the last three election cycles, do the Democrats have lopsided support.
This analysis underscores three points.
First, growth in the Latino electorate and these voters’ potential influence on electoral outcomes is occurring everywhere.
Second, while in recent elections Latinos have overwhelmingly favored Democrats, many Latinos, particularly naturalized Latinos and those who have not been engaged by either party, are politically independent. To this end, a consistent finding of Latino Decisions polling is that half of Latino voters have voted Republicans at some point.
Third and most important, as compared to any other voting bloc, immigration is the most important issue for Latino voters.
The third assumption — that action on immigration after the election will be sufficient to win back the support of Latinos angered by the delay — is an open question entirely. The anemic response to these voters’ most pressing issue, combined with limited partisan outreach, suggests that Latinos have not yet been fully integrated into the political system and that their participation and support should not be taken for granted in 2014 and beyond.
David Damore is a political scientist at University of Nevada, Las Vegas, and a senior analyst at Latino Decisions.