Given that senior citizens are much more likely to vote Republican than younger people are, it's pretty clear that people who die between the 2012 and 2016 presidential elections will be disproportionately Romney voters.
Daniel McGraw for Politico becomes the first person I've seen to try to precisely run the numbers of this, projecting that the shift in population will give Democrats about 453,000 more votes nationally even if nobody changes their mind:
By combining presidential election exit polls with mortality rates per age group from the U.S. Census Bureau, I calculated that, of the 61 million who voted for Mitt Romney in 2012, about 2.75 million will be dead by the 2016 election. President Barack Obama’s voters, of course, will have died too—about 2.3 million of the 66 million who voted for the president won’t make it to 2016 either. That leaves a big gap in between, a difference of roughly 453,000 in favor of the Democrats.
Here is the methodology, using one age group as an example: According to exit polls, 5,488,091 voters aged 60 to 64 years old supported Romney in 2012. The mortality rate for that age group is 1,047.3 deaths per 100,000, which means that 57,475 of those voters died by the end of 2013. Multiply that number by four, and you get 229,900 Romney voters aged 60-to-64 who will be deceased by Election Day 2016. Doing the same calculation across the range of demographic slices pulled from exit polls and census numbers allows one to calculate the total voter deaths. It’s a rough calculation, to be sure, and there are perhaps ways to move the numbers a few thousand this way or that, but by and large, this methodology at least establishes the rough scale of the problem for the Republicans—a problem measured in the mid-hundreds of thousands of lost voters by November 2016. To the best of my knowledge, no one has calculated or published better voter death data before.
A couple of caveats. On the one hand, African Americans on average die younger than white people, which tends to tilt this calculus toward Republicans. But on the other hand, women tend to live longer than men, which should have an impact in the opposite direction. Consequently, don't bet your life on this 453,000 figure being exactly correct. But it's a decent ballpark estimate. Obama beat Romney in 2012 by about 5 million votes, so a death differential on the order of 10 percent of that margin isn't trivial.