By Ryan McNeely
There’s an annoying tendency in much of the reporting on poll results to ignore any sort of context or comparison, often in service of a narrative that is belied by the actual poll. From NPR we have “Hispanics: Cooling on Obama,” and the article details legitimate reasons why Hispanics might be upset with the President, including the lack of action on immigration reform and the disproportionate impact the recession is having on vulnerable communities. It then cites a poll about the President not “adequately addressing the needs” of Hispanics and says of the results: “Obama gets only lukewarm ratings on issues important to Hispanics — and that could bode poorly for the president and his party.”
Five paragraphs into the article you finally get the job approval number: “Still, 57 percent of Hispanics approve of the president’s overall job performance compared with 44 percent among the general population in the latest AP national polling.” This is held up against the 2008 election results, where Obama won 67 percent of the Hispanic vote. So, there has been some cooling! But, according to the poll results in the very same article, the cooling among the general public is almost exactly the same (53% of the vote down to 44% approval). So, in essence, Hispanics are like everyone else.
Then there’s the bizarre refusal of some journalists to internalize the fact that electoral politics (unlike policy) is zero-sum in the two-party system. If Obama’s “down,” Republicans have to be “up,” or the narrative falls apart. Further down in the article, we see that “[w]ith the first midterm congressional elections of Obama’s presidency in three months, the poll shows a whopping 50 percent of Hispanic citizens call themselves Democrats, while just 15 percent say they are Republicans.” Run for the hills! In fact, the only way for these issues to not manifest themselves in a zero-sum fashion is if someone like Tom Tancredo runs a third-party campaign in Colorado because the Republican party is insufficiently nativist.
Finally, there’s the odd tendency to ignore basic probability theory. The poll finds that “two years after witnessing Hillary Rodham Clinton’s White House bid, Hispanics are twice as likely to expect to see a woman than a fellow Hispanic become president.” The implication is that this shows a demoralized Hispanic community that is sadly bearish on the prospects of a Hispanic president. Actually, since Hispanics make up about 15% of the U.S. population while women are a majority, if you chose a person at random to become president it’s actually over three times as likely you’d pick a woman than a Hispanic person. So Hispanics are actually oddly optimistic about the prospect of a Hispanic president.