This is one of the Big Questions. Nate Silver was one of the biggest winners in last year's election and one is wary of suggesting that he's got this question wrong. Nevertheless, I rather suspect that he may have. He suggests that, in 2012 at least, the GOP could, perhaps should, consider giving up on the hispanic vote. The argument is that Republican weakness amongst latino voters didn't actually hurt the GOP all that much. That is, hispanic votes only make the difference in states that are trending Democratic anyway. Furthermore, there remains a path to 270 electoral college votes even if the Republicans concede Colorado, New Mexico and Nevada.
If you excise those three Southwestern states [Colorado, New Mexico and Nevada], you still have a menu of 159 EV from which to choose, of which you [the GOP] need 91. And the remaining states are noteworthy for their relative absence of Hispanic voters. If you could gain ground in the Midwest or the South by pursing an anti-immigrant, anti-NAFTA, "America First" sort of platform, you really wouldn't be putting all that much at risk by losing further ground among Latinos. Yes, you could make life (much) harder for yourself if you screwed up Florida or put Arizona into play in the process, but it's not a bad strategy, all things considered.
About half the Hispanics in the United States reside in California or Texas, and another 20 percent are in New York, New Jersey or Illinois, none of which look to be competitive in 2012. (Yes, the Republicans could lose Texas, but probably only in a landslide). There just aren't that many Hispanic voters near the electoral tipping point. This, I would suggest, rests upon a couple of near-heroic assumptions: firstly, taking the risk of putting Florida or Arizona into play by implementing such a strategy is, well, a bloody major risk! Secondly, and more importantly, while appealing to the white working-class is a necessary part of Republican renewal, a nativist and protectionist campaign seems likely to shift college-educated white voters towards the Democrats. What you may, putatively, gain on the swings you could lose on the roundabouts.
Immigration policy is one of those areas in which a hardline policy is popular in the abstract but becomes, I think, less popular when it becomes an actual campaign issue as many voters find themselves turned off by the stridency and perceived ugliness of those banging this paticular rhetorical drum. It offends the idea of decency that people who consider themselves decent hold dear. And understandably so.
In other words, suburbanites who might have an intellectual sympathy for stricter immigration policies might baulk at the actual consequences of those policies, not least as and when they imagine how such policies might impact the hispanics they know themselves. That in turn, helped by a media that would undoubtedly cast such a GOP campaign in unflattering tones, could easily lead to these voters either switching to the Democrats or simply staying at home on polling day. Talk radio hosts might despise the Colin Powell/Tom Ridge brand of Republican but there are still enough of those voters to make a difference.
Equally, if you cede the mountain west as well as the north-east and pacific coast states then you dramatically reduce your margin for mishap. American elections are a multi-front battle contested, on the one hand, in individual (and groups of) states and, on the other, demographic groups. Sometimes and in some places these overlap. But not always and not everywhere. In general and especially in a two party syestem (sorry libertarians!) narrowing the playing field is not a terribly good idea.
So while it may be the case that the GOP could ignore latino voters in 2012, doing so would seem to store up endless problems in the future. Minorities made up 24% of the electorate in 2008 and that percentage is only likely to increase in the future. A GOP victory in 2012 based upon nativist, restrictionist, protectionist sentiment might be possible but only, I'd hazard at the expense of destroying the Republican brand amongst hispanic voters in the future. Replicating the "Southern Strategy" in terms of latinos is obviously tempting but it's not at all obvious that this is really in the Republican party's long-term interest. Quite the contrary in fact. What about 2020? And beyond?