Can A New Model Predict The Latino Vote?

A group of political scientists at Latino Decisions have come up with a model to take the guess work out of figuring our how the Latino vote will affect the 2012 presidential election. Up until now, that figuring has been mostly hocus-pocustry, educated guess work, experienced punditry, and hard-edged high-level political consulting – at a high price. So who are these guys and what have they got up their sleves?

Latino Decisions touts itself as a provider of  “independent data to facilitate critical decisions.” The difference seems to be that they don’t take the wide view; they dig among the trees at the state-level data. And what they’ve come up with is a model that makes room for shifting variables that can change the Latino vote out look on a regular basis.

Granted, the only outlook that matters is the one that registers on election day. But politicians and campaigns and all the hocus-pocus guys (and raise your hand if you’re a political data geek like me) are going to love this stuff. See, the back story here is that we’ve been talking about the importance of the Latino vote in the abstract, but now we’ll be able to make a clear case, in specific states.

The model will track Latino voter registration, the effects of voter ID laws, voter polls, changes in voter rolls. For instance, in swing states where the non-Latino vote is split among Republicans and Democrats, this kind of data will have a great value. New American Media explained it this way:

The political scientists explain it is not the size of the Latino voters in any particular state that determines their influence, but their turnout. In Virginia, North Carolina and Pennsylvania, Latinos are only between 3 to 5 percent of the electorate, but they can “tip” a state’s results if the non-Latino vote is almost evenly split, which could happen in places like Virginia and Pennsylvania.

There is a caveat though. The creators say this isn’t an end-all model. It may not be a wild guess about the Latino vote and what percentage of it a candidate needs to win each state, but it’s a best guess.