One of the many valuable services performed by the folks at Pew Research is their biannual polling on
party identification in America—in other words, how many people identify as Democratic, how many as Republican, and how many as independent. Looking at those statistics isn't unusual;
Gallup, for instance, polls regularly on that too, and more frequently. However, Pew talks to far more people about this question (more than 25,000, for the 2014 survey), and asks a variety of demographic questions that let you drill down to the party ID for any ridiculously specific group of people. Want to know how many
Hispanics earning more than $75,000 per year, or
Asian millennials, or
black midwesterners, or
white agnostics, are Democrats and how many are Republicans? Now you can find out!
Over the fold we'll talk more about how party ID breaks down along different demographic lines, but the finding that seemed to get the most ink, at least in the beltway press, was the steadily increasing number of self-described independents. Independents are at their highest point in decades, at 39 percent of the population (Pew surveyed all adults, not just registered voters). Democrats are at 32 percent, and Republicans at 23. The number of independents hit its low-water mark in 2004 (only 30 percent), but has risen steadily since then.
However, you need to look at the trend lines on the graph to get the full picture. If you do, you'll notice that the Democratic line has been pretty flat for the last 20 years—they were at 33 percent of the population in 1992, so they've fallen only 1 percent over the course of two decades. Instead, most of the movement seems to be from the Republican column to the independents. GOPers went from 28 percent in 1992 to 23 now, while indies went up from 36 then to 39 percent now.
That seems to confirm what recent polls showed about the pool of independents becoming more conservative as the GOP shrinks. Remember in 2012 when Mitt Romney was suddenly getting a much larger share of independents in polls than Republican presidential candidates in previous races had, and Republican pundits became convinced that was evidence that Romney was on track to win (despite swing state polling that showed he was clearly going to lose)? In retrospect, it looks like the alternative hypothesis was totally right—an increasingly large number of hard-right, tea party types could no longer abide calling themselves Republicans and were dropping their party label, even as they still planned to vote for Republican candidates in the next general election.
There's more over the fold.
Another way you can tell is by looking at Pew's numbers that push leaners, when they ask independents whether they "lean" toward the Democrats or the Republicans. In the end, you can see there really are very few true independents. Most independents have distinct ideological preferences but still can't bring themselves to fully associate with the party they gravitate toward (often, apparently, because the party isn't pure enough for them, rather than because it's too extreme for them). With leaners, 48 percent of respondents identify with the Democrats, and 39 with the Republicans, which means that the independents are split almost exactly down the middle (16 percent are indies who lean toward the Democrats, taking them up from 32 to 48, while 16 percent lean toward the GOP, taking them up from 23 to 39. Only 7 percent of the population are independents who truly claim not to lean one way or the other).
And if you follow the trend lines, despite the gain in indies, it's clear we aren't moving toward some sort of David Broder/Ron Fournier-style centrist paradise where we turn away from partisanship and toward Those Who Lead. The 48 D/39 R split with leaners is almost identical to the 49 D/42 R from 22 years ago. The big difference is that some of those voters who always vote Republican are now disguised as independents. (To their credit, some of the more perceptive beltway pundits picked up on this "same as it ever was" situation too.)
You might still be wondering why the Democrats aren't in a more dominant position if the nation breaks down 48/39. Bear in mind, again, that these are adults, not registered voters. Non-voters, as polls have repeatedly shown us, are more liberal than voters, and, as we were rudely reminded in November of last year, people who only vote in presidential years are more liberal than people who vote in every election.
Let's turn now toward more of a deep dive into the different demographic categories, and how they break down in terms of party ID. Pew's write-up helpfully turns the most strongly Democratic and strongly Republican segments of the population into an infographic (seen above). What you see probably won't surprise you: white people, especially Mormons (who break 70/22) and evangelicals, are most likely to be Republican, along with the Silent Generation (those 69 and older). (Amusingly, the nation's most prominent Silent Generation white Mormon male, Harry Reid, is a bit of an outlier.) People of color, millennials, and the highly educated are most likely to be Democratic.
The gender gap is quite pronounced, but if you look closely, men (45 percent) are much likelier to be independents than women (35), while women (37) are much likelier than men (26) to be Democrats. The percentage of both men (24) and women (23) who are Republicans, however, is consistent. Where the gender gap really shows up is when you push leaners, as many of those male independents seem to lean toward the Republicans; with leaners pushed, men break for the Democrats only 44/43, while women break for the Dems 52/36.
When you look at the graph for race, what might leap out at you is that Hispanics (at 44 percent), surprisingly, are more likely than whites (40 percent) to consider themselves independents. However, when you push leaners, you'll see that Hispanic independents—unlike white independents—seem to break pretty heavily toward the Democrats. The Republican share among Hispanics goes up from 13 percent to 26 when leaners are pushed, but the Democratic share goes up from 34 percent up to 56. Contrast that with whites, where the Republican share among whites shoots up from 30 percent to 49 when leaners are pushed, while the Democratic share goes up only from 25 percent to 40.
As I've pointed out before, level of education presents itself in a weird, U-shaped curve. People with less education (high school or less) are more likely to be Democrats, but also people with more education (college graduates or more) are more likely to be Democrats. It's the "some college" crowd who are more likely to be Republicans. Pew finds, though, that the strongest Democratic trend is among persons with not just college but also some sort of post-graduate experience. This group is very unlikely to be Republican any more (down to only 20 percent of them, currently). If you push leaners on this particular group, you'll see that post-grads broke 47 D/46 R in 1992 ... but now break 56 D/36 R today.
Finally, let's turn briefly to one of the most interesting but often-overlooked demographic phenomena, the
generation gap. You're probably familiar with the canard that sensible people get more conservative as they age, which is not only
misattributed to Winston Churchill but also consistently proven wrong by polling. Instead, generational cohorts tend to stay pretty ideologically stable over time, which is why it's important to lock in people's beliefs while they're still young.
Millennials (current 18-33-year-olds), interestingly, are by far the generation most likely to be independents (48 percent). In fact, each previous generation is less likely to be independents (least of all the Silent Generation, who are only 29 percent indie). Nevertheless, those millennial indies are much likelier to lean Democratic than other generations' indies: when you push independent leaners, Democrats have a 51/35 advantage among millennials. That makes them the most Democratic-friendly generation, by far. The least Dem-friendly is the elderly Silent Generation. Even though there are few independent Silents, the ones who are, lean heavily toward the GOP. They're the only generation where the GOP currently has an advantage when you push leaners: 47 R/43 D.
Even though the millennials are the most Democratic-friendly generation (when you push leaners), don't get too excited thinking that we've finally produced a universally tolerant, post-racial generation of young adults. Instead, it's largely a product of interaction with other variables, namely, the millennials are a much less white generation than previous generations. This becomes pretty stark when you look at the millennials broken down into white and non-white segments. White millennials, when you push leaners, still opt for the Republicans, by a 45/43 margin (granted, that's a smaller gap than among white adults in earlier generations). And non-white millennials, surprisingly, are actually a smidge less Dem-friendly than non-whites in previous generations (backing the Democrats by a 61/23 margin).
But the fact that there are simply so many more non-white millennials than, say, non-white Gen-Xers or non-white Baby Boomers makes them a much more Democratic-leaning generation than previous ones. Assuming that trend continues—and there isn't some unpredictable event that shocks the as-yet-unnamed generation that follows the millennials into all becoming Republicans—that should give us even more hope for the long-term future, considering that, for the first time, more than half of all public K-12 students are now non-white.
