Politico's Dylan Byers made a mistake last October. His critique of Nate Silver, suggesting that The Times's data guru might be a "one-term celebrity" if his predictive model wasn't borne out on election day, woke the beast. Byers's column, a benchmark in the history of ill-advised, unclear, erroneous attacks, has led Silver to take every single available opportunity to humiliate and mock Politico, particularly when it engages in data analysis. It's like Silver's echoing that line from No Country For Old Men, in which Josh Brolin's character tells Javier Bardem's Anton Chigurh that he plans to make him "a special project of mine." But Silver fits better as the Chigurh character, a man set on ruthlessly obliterating his opponents, far past the point at which any further destruction is necessary. "You don't have to do this," Politico pleads. Silver smiles — "People always say the same thing" — and pulls the trigger.
Last night, Politico posted new analysis of the way in which providing a path to citizenship for undocumented immigrants might affect electoral politics. The story's point, in short: There will be more Latino voters, which likely benefits Democrats. This isn't a complex argument, of course; one of the reasons that the Republican Party is putting a new emphasis on working out a deal on immigration is that it wants to bolster its support among Latino voters. Those who might follow a new path to citizenship are only a small percentage of that entire population, but every vote counts.
POLITICO "analysis": no way to know how many illegal immigrants would get citizenship and vote, so let's assume 100%. politico.com/story/2013/04/…— Nate Silver (@fivethirtyeight) April 23, 2013
The quotes around "analysis" are a subtle dis. Calling out the assumption is a direct one. Silver is referring to this section of the article:
If one adds 11 million new Hispanic voters after immigration reform but applies 2004 percentages, the damage to Republicans is real but much less severe: Romney would have still won border states Texas and Arizona, albeit by smaller margins, while Obama would have held other Latino-heavy swing states like Nevada and Florida by slightly larger margins than the ones he did win by.
The POLITICO analysis is intended to reflect the GOP’s broader dilemma on immigration issues; it is not meant to be specifically predictive. There is no way of knowing how many of the estimated 11 million undocumented workers would ultimately succeed in gaining citizenship, nor any certainty of what their turnout percentages would be once they gain voting rights.
There are a variety of ways in which this is odd analysis. Demographics don't translate elegantly between presidential elections; four years is a long time. Nor do voting percentages. A George W. Bush-versus-John Kerry contest is very different than a Mitt Romney-versus-Barack Obama one. And, to Silver's point, not every single undocumented immigrant would seek (or be eligible) for citizenship. Nor, as he noted in a later tweet, are they all Latino.
Also love how Politico's "analysis" assumes that all illegal immigrants are Hispanic. politico.com/story/2013/04/…— Nate Silver (@fivethirtyeight) April 23, 2013
Silver is not the only pundit to find fault with that Politico article. The Daily Caller's Matt Lewis, for example, suggested that the outlet's motives were suspect.
Strikes me as very curious that Politico is sooo worried about how immigration reform could hurt Republicans. Talk about concern trolling.— Matt Lewis (@mattklewis) April 23, 2013
But Silver is the only one to attack Politico with such glee. We have saved the best for last.
Politico attempting to use statistics is like Taco Bell attempting to cook French food.— Nate Silver (@fivethirtyeight) April 23, 2013
You don't have to do this, Nate. But we're enjoying it nonetheless.