We say "long-term" because population can only grow so fast -- i.e. it's not necessarily an imminent problem for the GOP in 2016.
Or maybe it is, according to a new study from the left-leaning Center for American Progress.
The study, conducted by policy analyst Patrick Oakford, ran three simulations of the 2016 election:
1) Racial and ethnic groups turn out to vote at 2012 levels and vote for Republicans and Democrats at 2012 levels
2) Racial and ethnic groups turn out at 2012 levels, but vote at 2004 levels (when George W. Bush was reelected)
3) Racial and ethnic groups turn out at 2012 levels, but whites vote at 2012 levels, while racial minorities vote at 2004 levels
You'll have to bear with us a bit here, because this is a little dense. But the main takeaway is this: Republicans lose in all three of these simulations -- handily.
Here's what No. 1 (a repeat of 2012 turnout and exit polls) looks like, courtesy of the great 270ToWin.com. The Democrats won the Electoral College 332-206 in 2012. Under this simulation -- with nothing changed from 2012 except the passage of time and demographic shifts -- they would add North Carolina to the "win" column.
Under simulation No. 2, things get a little closer, because using the Bush-Kerry exit polls means Republicans take about four in 10 Hispanic voters -- versus the 27 percent Mitt Romney got in 2012.
The Electoral College, though, still favors Democrats, 291-247. So the 2012 electorate, even if it voted at 2004 levels (under which a GOP president was reelected) in 2016, would elect a Democrat.
And finally, simulation No. 3. This would seem to help Republicans, because they get the stronger white vote they had in 2012 but the stronger Hispanic vote they had in 2004. Yet Republicans actually do slightly worse here than under simulation No. 2.
They lose here 303-235, losing Colorado and Virginia while picking up Missouri.
The reason for the Democrats' success here? Growing minority populations. In North Carolina, which flips to Democrats in 2016 even at 2012 turnout and voting levels, the Latino eligible voting population is projected to increase from 3.1 percent to 4.5 percent between 2012 and 2016, while the black population moves from 22 percent to 22.7 percent and the Asian-American population increases from 1.4 percent to 1.8 percent. The white population, meanwhile, drops by two points, from 71.3 percent to 69.2 percent.
That means a state Romney won by two points in 2012 suddenly goes for Democrats by 0.3 points -- again, assuming the vote goes the same as 2012.
Over that same four-year period, the Hispanic population is projected to increase by three points in Florida (from 17.1 percent to 20.2 percent), by nearly three points in Nevada (15.9 to 18.8), and by about two points each in Colorado and Virginia. Almost every state loses at least two points off its white population.
Those are only a few-point shifts, but these states could be decided by only a few points.
A few caveats here.
First, this is from a left-leaning group, and it is obviously a highly hypothetical exercise, based on imperfect exit polling data. Thus, odd things happen, as with Democrats winning Missouri under simulation No. 2. (Democrats didn't compete for Missouri in 2012, and they probably won't in 2016 either.)
Secondly, and perhaps most importantly, all three simulations match turnout to 2012 levels, which featured higher-than-usual turnout among racial minorities -- particularly among African Americans, but also among Hispanics. That helps Democrats in a big way.
Here's how turnout changed over the last four presidential elections. You'll note that not only are minority populations growing, but they are also voting at higher rates among all eligible voters. White turnout, meanwhile, was down from 2004 levels in 2008 and 2012.
If Democrats can get these groups to show up and vote in the numbers that they did for President Obama, that will clearly be huge advantage. But it's also not clear that black voters will turn out at the same levels once the first black president is no longer on the ballot.
At the same time, we would note that all three of the biggest racial minority groups were already increasing their turnout in 2004, so it's not clear that it's solely or even mostly attributable to Obama. Perhaps Democrats are just doing a better and better job at getting low-turnout minority groups to the polls.
But it's also important to emphasize that 2012 was a very good electorate for Democrats, and it might not be repeatable in 2016. Using 2004 turnout levels, for example, would probably produce a much different result.
If nothing else, the data back up the idea if Democrats can produce something close to the Obama Coalition in 2016, they will be pretty strong favorites to retain the White House.
And that Republicans might need to address their problems with Hispanic voters sooner than their huge 2014 wins might indicate.