Here's what the 2008 and 2012 elections taught us: President Obama built a national political coalition -- the three main pillars of which were African Americans, Hispanics and young voters -- that Republicans couldn't come close to touching. Here's what the 2010 election taught us: That Obama coalition is not directly transferrable to all Democratic candidates. And here's the fundamental question for Democrats in 2016: Can anyone in their presidential field rebuild the coalition Obama pieced together in 2008 or 2012? Or is it unique to him?

The 2012 electoral map.


New data from an NBC poll pitting Hillary Clinton against Chris Christie in a hypothetical 2016 matchup suggest that if Democrats nominate the former secretary of state she would start out with much of that Obama coalition behind her.

In a matchup with Christie, Clinton wins the African American vote 83 percent to 4 percent, voters aged 18-29 45 percent to 31 percent and Latinos 44 percent to 33 percent.  Those numbers are roughly comparable to how Obama did among two of those three core constituencies. In 2012, Obama won black voters 93 percent to 6 percent and 18-29-year-olds 60 percent to 37 percent. Four years earlier, he won 95 percent of the black vote and 66 percent of the youth vote.

The one key Obama constituency where Clinton is not replicating (or coming close to replicating) the performance of the current president is Latinos.  Clinton's 11-point edge runs well behind Obama's 44-point margin among Hispanics in 2012 and even his 36-point edge in 2008. Still, the recent voting patterns of Hispanics would suggest that if the race were held today, Clinton would likely approximate Obama's margins; since George W. Bush won 44 percent of the Hispanic vote in 2004, Democrats have won the group by 39 points (2006), 36 points (2008), 22 points (2010) and 44 points (2012).

As potentially troubling for Republicans is that the only three subgroups in the poll where Christie leads Clinton are whites, seniors and the wealthiest Americans -- a demographic cluster that roughly approximates where 2012 GOP presidential nominee Mitt Romney did well while losing convincingly.

Before there's too much carping from Republicans, yes, there is no question that a poll conducted three years from the time when voters will actually vote shouldn't be regarded as entirely predictive of how the electorate might look by 2016. Christie, although he is probably the best-known Republican politician in the country, remains less well known and well defined than Clinton, who has been in the national spotlight for more than two decades.

Still, if you take this survey as a sort of generic Democrat vs. generic Republican matchup -- and it largely is that at this incredibly early stage, with voters not at all engaged in such a hypothetical race -- what the numbers suggest is that the demographic problems that Republicans faced in 2008 and that worsened in 2012 are not unique to the presence of Barack Obama on the ballot. And that's a major problem for a party desperate to reclaim the White House in 2016.