The reason for this is due to the simple fact that older voters die more regularly and also vote Republican more regularly. It's not any more complicated than that; the oldest 2012 voters backed Romney heavily and also see increasing mortality rates. Ergo, a shrinking pool.
Sort of. There are a lot of caveats that apply -- ones that suggest that this trend is not really something that should keep Republicans up at night. (Over the short term.)
The same pattern held after 2008
First and foremost, the exact same trend of older voters leaning Republican and dying at higher rates held in 2008, as well. We took exit poll data and compared mortality rates from 2009 (for the 2008 voters) and 2013 (for the 2012 ones) to figure out how many of the voters would have shuffled off their voting coil by the next election cycle. Our figures are lower than McGraw's, mostly because we used the bottom-tier mortality rate for each age group. As voters age, the mortality rates increase, so assuming a 60 year-old dies at the same rate as an 85 year-old is obviously underestimating things.
But we're not comparing age groups here; we're comparing elections. And 2008 and 2012 look very, very similar.
And yet Mitt Romney received about a million more votes than John McCain.
We have our own cascade of caveats here. A 65-year-old voter in 2008 will be 73 by the time 2016 rolls around. The Baby Boomers kept getting older and their mortality rates keep increasing. So it's safe to assume that the death rate will have ticked upward more between 2012 and 2016 than between 2008 and 2012. But that's the whole point: Other factors -- attitudes about Obama, the candidates who are running, the political moment -- played a much different role for Romney than for McCain.
Those differences made a much bigger difference than the number of people that died between elections. McCain won 53 percent of the over-65 vote; Romney won 56 percent of it. That was a net gain of almost 400,000 alone.
The population is still shifting older
We've talked about this before. Census projections suggest that the percentage of the country that is 65 and older will continue to increase through 2055 -- and at a faster rate than other population groups.
Those votes tend to vote more heavily than younger voters as well.
Again, the question comes back to how they vote. The Republican party has done a good job of solidifying the existing over-65 vote, but it remains to be seen if people who age into that demographic will vote increasingly conservatively as well. There's a lot at play: Party platforms, candidates and so on. It's not fair to assume that Republicans will keep the same lock on the 65-plus population. It's also not fair to assume that population will shrink.
We'll note here, by the way, that McGraw's point about how individual states' voting populations will have changed by 2016 is an interesting one. The numbers are squishy enough, though -- new retirees moving to the state, for example -- that they're worth taking with a grain of salt.
Other demographics shift at different rates, too
The last point is a big one: This only looks at the change by age. It doesn't include the shifts among racial groups, for example. A different study found that early death rates among blacks may have cost Democrats the 2004 election.
An electorate in any year is a complex mix of ages and demographics and voting interests and candidates. Would the Republican Party prefer to perpetually increase the number of older voters and the rate at which they vote for the party? Sure. Will both parties constantly evolve platforms and priorities to appeal to emerging voting blocs? Yes (but usually slowly). Between 2004 and 2008, a lot more Baby Boomers aged into the 60-plus age group, where Republicans have an advantage. But Barack Obama won easily.
Demographics is rarely destiny.