I thought it'd be interesting to explore how the youngest and oldest cohorts are currently voting and measure the discrepancy between those two groups' margins as a way of peeking into the future.
I decided to look at 2014 for multiple reasons, including the fact that it is the most recent election and that these individuals (especially going forward for the young cohorts) are/will be the core, reliable voting blocs for each subset of the population. I could have used 2008 data, but sheesh: it's 8 years old now.
Obviously we don't have exit polling data for every state, but you can use it to observe general trends for several regions of the country, and definitely so for the South. All but two states' (ME & WV) youngest voters are more Democratic than their oldest voters; in WV, the margin difference was 2 points.
Shades indicate the difference between 18-29 & 65+ voters' margins in 2014. In states where exit polling was available for both gubernatorial and senatorial races, both outcomes were averaged together to produce the result.
The five states with the biggest discrepancies between 18-29 & 65+ groups:
| State | '14 Mar-Diff (Pts) | 18-29 | 65+ |
| CO | 81 | D+61 | R+20 |
| SC | 58 | D+19 | R+39 |
| MS | 43 | D+3 | R+40 |
| GA | 43 | D+13 | R+30 |
| TX | 42 | R+3 | R+45 |
As far as trend (in the Atlas sense of the word) is concerned, I looked at the national House vote exit polling, which showed a margin discrepancy between the two groups of 27 points (18-29: D+11 & over 65: R+16), and then made this map to show which states had discrepancies that were larger or smaller than the national House vote discrepancy. I know, it's a bit of apples and oranges, but that's all the store had. There is no Gov/Sen national exit poll.
Shades indicate trend in 2014 margin discrepancy shown above relative to national House margin discrepancy for the two groups. White equals no trend. Both maps show same thing (one with gradient, one w/o):
There are a couple of outliers on the map based on this measurement - which is to be expected given the nature of this - but I think it's interesting nonetheless. I'm wondering if all of that Latino infusion into western KS is starting to produce small but noticeable dividends in the youngest voting blocs.
Bumping an interesting post. I wonder how plugging 2016 data into this would look.
Poll
