This is from the Washington Post.
http://www.washingtonpost.com/politics/the-9-most-fascinating-numbers-from-national-exit-polls/2014/11/23/7fceb424-7320-11e4-a5b2-e1217af6b33d_story.html?wpisrc=nl_politics&wpmm=1The 9 most fascinating numbers from national exit polls
By Chris Cillizza November 23 at 10:33 PM
The best thing about e******ns if you are a numbers nerd like me is the massive amount of raw data about the American public and what they think, as well as the how and why of it. Yes, only about 36 percent of eligible v**ers cast b****ts three weeks ago, but that amounts to almost 77 million people a pretty great sample to sort through.
I sifted through the national exit poll in search of clues about the state of the American e*****rate. The nine numbers that jumped out at me are below.
■ 4. Thats the margin by which Democrats beat Republicans among women nationwide in the v**e for the House. Thats a significant decline from President Obamas winning margins among women (11 in 2012, 13 in 2008), though its an improvement from the 2010 midterms, when Democrats lost the womens v**e by a point. Still, the massive focus of Democratic candidates across the country on the Republican Partys supposed war on women clearly didnt persuade large numbers of female v**ers to abandon the GOP. Assuming Hillary Clinton is the Democratic p**********l nominee in 2016, the historic nature of her candidacy as the first female may erase any doubts about a shrinking g****r-gap edge for Democrats. But, in midterm e******ns at least, women are simply not an overwhelmingly Democratic constituency.
■ 62. Thats the percentage of the v**e for Democrats among those who said they never attend any sort of religious services; Republicans won just 36 percent among that group. Compare that with the 18-point edge Republicans enjoyed over Democrats among those who go to some sort of religious service weekly and you see that ones religiosity continues to be among the most reliable predictors of how they will v**e. Consider yourself a religious person, or, at least, someone who attends religious services regularly? Theres a strong likelihood you are v****g Republican. Not a churchgoer? You are v****g Democratic.
■ 54. A majority of Americans who went to the polls Nov. 4 believe that the government is doing too many things better left to businesses and individuals, while just 41 percent think the government should do more to solve problems. Those numbers suggest that the long-running battle over what government can and should do (and how much it should do) is tilting back toward the smaller-is-better crowd that dominated in the mid- and late 1990s. The early years of Obamas presidency were defined by a belief that government might need to do more than it had during the latter days of the George W. Bush administration. (Think Hurricane Katrina.) But the pendulum is in the process of swinging back to the government-shrinkage folks.
■ 78. How people feel about their government is this number, which represents the percentage of people who say you can only sometimes (60 percent) or never (18 percent) trust Washington to do what is right. Thats stunning. The number is partly attributable to a Republican-flavored e*****rate and the natural suspicion among many within the GOP of the federal government particularly when its run by a Democratic president. But the number is so high that its hard not to see the problems surrounding the Internal Revenue Service, Veterans Affairs and, of course, the National Security Agencys spying revelations as part of that deepening distrust among broad swaths of the American public.
■ 48. Thats the percentage of people who said same-sex marriage should be legal in their state, the same number that said it should be illegal. That even split is a significant break from most polling on the issue; Gallups most recent poll, in May, on same-sex marriage showed 55 percent of Americans supporting it while 42 percent opposed it. What explains the discrepancy? Maybe the difference between asking whether same-sex marriage should be legal in your state versus asking, more broadly, about its legality. Or maybe that people are less willing to tell someone over the phone what they really feel about a d******e social issue than in person. (That seems odd.) Maybe the difference between legal (the exit poll language) and valid (Gallups language). Or that this midterm e*****rate was a Republican-friendly one.
■ 75. Three-quarters of the 2014 e*****rate was white (and they v**ed for Republicans by 22 points) on Nov. 4. That might seem like great news for Republicans. Its not. W****s made up 77 percent of the 2010 e*****rate and the decline in w****s as a percentage of the overall e*****rate is happening in p**********l cycles, too.
■ 38. The percentage of the white v**e that Democratic candidates won nationwide. Thats the same percentage Democrats got among white v**ers in the 2010 midterms and virtually equal to the 39 percent Obama won in the 2012 e******n. Thats a trend and a downward one for Democrats. From 1996 to 2008, the Democratic p**********l nominee always won 41 to 43 percent of the white v**e.
■ 36. Thats the percentage of the Hispanic v**e that Republicans won Nov. 4, an improvement on the 34 percent they won in 2010 and a major step up from the 27 percent that Mitt Romney took in 2012. It was the strongest showing for Republicans among Hispanic v**ers since Bush won 44 percent of the Latino v**e in the 2004 e******n. (Bushs showing was, by far, the best performance for a Republican p**********l nominee since 1972.) It remains to be seen how Obamas executive action on immigration and the Republican response to it will affect those numbers in the long run, but 2014 was a step in the right direction for Republicans among Latinos.
■ 53. A majority of v**ers who identified as moderates (four in every 10 v**ers) cast b****ts for Democrats. Republicans got 45 percent of the moderate v**e. Thats a good reminder that a) moderate does not equal independent (Republicans won independents by 12 points) and b) this e******n was not decided by the middle. It was decided by the Republican base or, put another way, the no-show of the Democratic base.