By Sean Trende, RCP
This is the third part in a series exploring whether the GOP really needs to pass a comprehensive immigration reform package to continue winning elections (Parts one and two are here). It’s worth re-emphasizing that this series isn’t intended to offer a policy prescription. The purpose here is to challenge the almost universally unchallenged argument that the only real debate is whether the GOP will die a slow death or a fast one if it doesn’t back the bill.
One of the assumptions lurking behind the immigration debate is that Hispanics spurn Republicans in large part due to identity politics, with immigration as a primary motivator. But there’s usually precious little data offered up to buttress this assumption, aside from the occasional generic poll question of “How important is issue this to you?”
But dig down. When Pew asked Hispanic voters what the most important issue was in 2012, immigration ranked near the bottom of the list. At the top of the list was jobs, just as it was for whites. In the 2008 exits (the last presidential election for which we have full data), a majority of Hispanic respondents told the exit pollsters that they didn’t care much about immigration (regardless of how they ultimately voted), or that they cared a lot (but voted Republican).
Many point to George W. Bush’s performance with Hispanic voters, especially compared to Mitt Romney’s, as evidence that the immigration reform issue is crucial. After all, Bush was strongly in favor of a path to citizenship, and exit polls show him winning 45 percent of the Hispanic vote (though many analysts believe this likely overestimates his share of the Hispanic vote). But there are two counters here.
First, consider a different Bush, George H.W. He had many of the same “pluses” as Dubya did, with the added benefit of having been a part of an administration that signed comprehensive immigration reform in 1986. The result? He lost the Hispanic vote by 39 points, while winning nationally by eight. This is still the worst GOP showing among Hispanics relative to the national vote since the 1970s.
Second, and more importantly, there are a number of potentially important differences between George W. Bush and Romney that don’t directly involve a pathway to citizenship. For example:
– Bush was a governor of a state with a large Hispanic population, and had a history of interacting well with the state’s Hispanic community;
– Bush speaks Spanish reasonably well;Bush supported big government programs, and Hispanics tend to be more economically liberal than whites;
– Bush was a wartime president, and Hispanic voters tend to be relatively hawkish;
– Bush’s overall native appeal to downscale voters was greater than Romney’s;
– Bush didn’t say things suggesting that he would make life so miserable for illegal immigrants that they would “self-deport”;
– Bush was running against a boring, stiff, white candidate who lacked Barack Obama’s innate appeal to young voters.
I suspect that all of these factors contributed to Bush’s strong showing with Hispanics vis-à-vis Romney, in addition to his stance on comprehensive immigration reform. Bush was in many ways the perfect Republican candidate for minority outreach; Romney one of the worst imaginable.
I think the last two issues might be particularly salient here. Though supporting comprehensive immigration reform might be one way to help convince Hispanics that you aren’t a latter-day Tom Tancredo, it’s probably more important to not say things that are aggressive, and even offensive, to Hispanics.
Also, it’s important to remember that Obama had a more generalized appeal to younger voters, and that Hispanic voters are disproportionately young. And, of course, many Hispanic voters are nonwhite. To the extent that Obama had any generalized appeal as a nonwhite candidate, in terms of voting habits and turnout — and I think there was some such appeal, albeit only at the margins — it would affect the Hispanic vote and potentially be non-transferrable to another candidate.
My sense is that Romney died a death of a thousand cuts with Hispanics, and that his stance on immigration reform was just one of those cuts. Changing a stance on immigration reform heals one of those cuts, but there are plenty of others that can be addressed as well. It is often asserted that supporting immigration reform is a necessary precondition to addressing other issues, but that’s usually just that — an assertion.
In the big picture, I think what motivates Hispanic voters isn’t all that different from what motivates white voters:
This is the Hispanic vote according to exit polls broken down by income. Here is what the same chart looks like for whites:
At the end of the day, Hispanics tend to vote more Democratic than whites because they tend to be poorer than whites. There’s still plenty of room for GOP growth in the short-to-medium term — winning middle-class Hispanics by the same margin that he won middle-class whites would have almost delivered the Hispanic vote to John McCain in 2008 — but ultimately the GOP doesn’t need more Republican Hispanics so much as it needs more middle-class Hispanics (which should happen, as time progresses).
How much of the Hispanic vote is enough
Some analysts like to claim that the GOP is unlikely to win the Hispanic vote regardless of what it does on immigration because Hispanics are not, as some have claimed, natural conservatives. They’re actually fairly liberal, relative to whites.
This is true, but it misses the point. If Republicans were actually to win the Hispanic vote outright, our question would quickly shift from whether the Republicans are going extinct to whether Democrats are going extinct.
Instead, the questions should be “What share of the Hispanic vote do the Republicans need to win in order to remain competitive with a given share of the remaining populations?” and “What do Republicans need to do in order to get there?” I suspect that to get in the neighborhood of 40 percent with Hispanics, Republicans would need a near-perfect candidate for Hispanic outreach, maybe Jeb Bush, Marco Rubio, or Susana Martinez. But if 35 percent is what they need, it might be enough to have a Democratic candidate with less intrinsic appeal to young voters, and a Republican who disclaims inflammatory rhetoric to reach that goal.
Let’s take a look at what we might see with modest Republican outreach toward Hispanics. In this scenario, the GOP makes some initial gains with Hispanics and Asians by running a candidate more-immigrant friendly than Romney, but loses the full benefits it might gain from a “Bush-like” candidate. It continues to make modest gains over time, but its gains are slowed among white voters from the present trend line.*
The end result is the same as the “polarization” scenario we discussed in Part 2: The GOP remains competitive in elections for quite some time.
What’s the real trend line with Hispanics?
But can the GOP really expect to do better with Hispanics? Hasn’t it poisoned the well?
The latter is possible, but I don’t think so. The truth of the matter is that the Hispanic population has gradually been trending toward Republicans over time, especially when you take account of where the country has been as a whole. This makes sense when you consider the socioeconomic tendencies in Hispanic voting, and the improved socioeconomic standing over time.
Here’s the presidential data. It’s a bit noisy, but even if you ignore the trend line, in 2012 Hispanics were, relative to the country, about five points more Republican than they were in the 1970s. In fact, relative to the country as a whole, Hispanics were more Republican in 2012 than they had been in 1972, 1976, 1984, 1988, and 1996.
In Congress, we see the same thing. In absolute terms, Republicans actually held Democrats to 61 percent of the (two-party) vote with Hispanics as recently as 2010. Through 1994, Republicans never once exceeded or matched 35 percent of the vote; since then, they have done so in 1994, 1998, 2000, 2002, 2004, and 2010. The latter group of numbers is especially intriguing, since many of those elections came right on the heels of the GOP’s mid-1990s battles on immigration.
Here’s the numbers for Congress in terms of PVI. They show the same basic results as the presidential numbers.
It’s worth noting, too, that there is probably a real floor for Republicans with Hispanics. It may surprise a lot of people, but there’s a substantial minority that sides with immigration hardliners in the GOP. Even the latest Pew poll, which has some reasonably favorable question wordings for the pro-reform side, finds a third of Hispanics support a “control the border first” approach. Substantial minorities — about a third — supported the three controversial California ballot initiatives from the 1990s. Jan Brewer managed to win about a quarter of Hispanic voters in Arizona in 2010; Romney did about as well
We pay lip service to the idea that there is no “Latino vote,” but there really isn’t one. It’s an incredibly broad array of voters lumped together for purposes of data analysis.
How much will the Hispanic population really grow?
For all the talk of the fast-growing Hispanic vote, it’s been pretty quiet of late. Net migration from Mexico stopped during the recession, and Michael Barone makes a compelling argument that it is unlikely to start up again. And the truth remains that you could depopulate Mexico — take every man, woman and child and move them to the United States — and non-Hispanic whites would still be a majority here.
Of course, Hispanics living in the country will have children, but the Hispanic birthrate has fallen dramatically. In other words, Hispanic immigrants are behaving an awful lot like older immigrant waves. This all has an impact on socioeconomic standing as well: As fewer new immigrants come in, the Hispanics who ascend to middle-class standing will make up an ever-larger share of the Hispanic population.
The most interesting question, though, is how much “racial attrition” will occur? Reihan Salam observes:
[W]hile virtually all third-generation Mexican-Americans with three or four Mexican-born grandparents identify as being of Mexican descent, Duncan and Trejo observe that only 79 percent of those with two Mexican-born grandparents do the same. For those with only one Mexican-born grandparent, the share falls to 58 percent. Only 17 percent of third-generation Mexican-Americans have three or four Mexican-born grandparents, so the ethnic attrition rate is quite high: 30 percent of Americans with at least one Mexican-born grandparent do not identify as being of Mexican descent.
It seems at least somewhat inevitable that Hispanic population growth will slow somewhat (and this has significant implications for voting, given our earlier observations about socioeconomic status and voting). If it does, it makes a significant difference in the baseline for the projections we’ve been using. Here’s the above “modest nonwhite gains” scenario played out with significantly lower Hispanic and Asian population growth (albeit still higher than white population growth):
Conclusion: Arizona, Rubio, or somewhere in between?
Basically, the choice the GOP has here can be summed up as whether it wants the Arizona model, what we might call the “Rubio model,” or something in between. Arizona has recently operated as something of a poster child for everything you could possibly do wrong as a Republican-governed state bordering Mexico. George W. Bush won the state by 10 points in 2004; hometown hero John McCain won by eight in 2008.
But Republican Gov. Jan Brewer, who spearheaded the state’s controversial immigration law, won by 12 points. Romney won by almost the same margin as McCain (it was a closer race nationally, but Romney didn’t have McCain’s “home team advantage”). The Hispanic vote grew, and dropped off substantially for Brewer and Romney, but it was compensated for by a surge in white support, particularly among non-college educated whites.
A similar effect is probably occurring in Texas. Even if Republicans aren’t losing ground among Hispanics (we don’t know, because we don’t have exits for 2012), the state should still be trending blue due to demographic change. But it isn’t. Romney won it by 16 points, McCain won by 12, Bush won by 23 and 21 points. When you account for changes in the national environment and Bush being a hometown governor, that actually works out to slight movement toward Republicans.
On the other hand, the GOP could go “full Rubio”: Back immigration reform, nominate a Hispanic-friendly candidate, and see what happens if Hispanic population growth slows. This approach works pretty well, even if Republicans sacrifice some progress with white voters:
I don’t mean to leave the impression that the GOP will win no matter what it does. Tweak some of these assumptions, and you get plenty of Democratic wins too. And it may not matter what the GOP chooses. The most dispiriting possibility is that racially diverse electorates may inherently add racial cleavages to otherwise “neutral” issues, and that polarization becomes inevitable. That’s certainly the experience of Northern cities during the great immigrant wave of the early 1900s, as well as of the American South.
My point is simply that there are a slew of realistic scenarios where Republicans do very well in the future. In most of the scenarios I consider reasonable, the elections stay close enough that either party could win most any individual election for the foreseeable future.
My overall view of presidential elections is that they are like giant algebra problems that suddenly simplify down to three or four variables at the end. Both sides have reasonably good arguments and appeals, run decent campaigns, nominate competent (if not outstanding) candidates and raise enough money to be heard. This end result is that this tends to cancel out.
Things like the economy, wars and incumbent fatigue create a fairly narrow playing field. The other factors, under normal circumstances, can probably move things three or four points in either direction. The fundamentals did a pretty good job predicting 2004, 2008, and 2012, without any regard to demographic shifts. I suspect this will continue in the future.
The GOP and Democrats should pursue the policies they believe are best for the country. If they govern competently, the coalitions will take care of themselves.
* I performed this using Nate Silver’s electoral projection calculator. I’ve made some baseline changes here: 40 percent of immigrants become citizens and vote (as I explained here), the black vote falls to 90 percent Democrat post-Obama, the growth of the black population is slowed down a tick to try to compensate for the likely turnout surge in the Obama years, and the white population growth is increased a tick to try to compensate for the return of some white voters who passed on 2012.
This articles appeared on Real Clear Politics on Friday 28th, 2013.