The popular vote gap by Chuck Todd.

(Lifted in total from FirstRead.)

That brings us to the popular vote... According to our turnout estimates (based on past results and talks with the campaigns and also assuming a bump), it is unlikely that Clinton will completely erase Obama’s lead in the popular vote without factoring in Florida and Michigan. Here's a fairly rosy scenario for Clinton, which assumes big wins in her states and somewhat narrow wins for Obama in his states of the contests that remain.

	Total Votes	Clinton	Obama	Projection
PA	2,000,000	1,100,000	900,000	55%-45%
NC	1,500,000	675,000	825,000	45%-55%
IN	900,000	460,000	440,000	51%-49%
WV	400,000	240,000	160,000	60%-40%
KY	500,000	300,000	200,000	60%-40%
OR	600,000	270,000	330,000	45%-55%
SD	100,000	45,000	55,000	45%-55%
MT	125,000	56,250	68,750	45%-55%
TOTAL		3,146,250	2,978,750

So that projection gives Clinton a net gain of 167,500 popular votes for the rest of the remaining primaries -- if everything plays out by the projections above. Obviously, we encourage folks to play around with these numbers themselves. So Team Clinton couldn't get there with also adding Florida; they'd need Michigan, too... and maybe even try and count total Puerto Rico votes -- which, turnout-wise could look like Oregon, but in reverse. So MAYBE she could net another 60,000 votes out of Puerto Rico. Of course, if Clinton gets a big win out of Pennsylvania, she might be able to narrow the gap in places like North Carolina and Oregon and then suddenly her net popular vote take could increase. To have ANY chance of selling legitimacy to this popular vote game, she'll need to cut his lead without the use of Michigan. Maybe, she can include Florida; there are a chunk of superdelegates who would give her that, but most will not give her Michigan. So she needs to more than double the projected popular vote totals we've come up with to even start the popular vote conversation.