The difference in margin between 2000 & 2016 is 33 points (between 2000 & 2012 is 25 points), so let's break as much of it down as we can:
1) Racial demographic shifts - Between 2000 and 2010, Mecklenburg went from being around 60% non-Hispanic white to being 45% non-Hispanic white. Factoring in the shifts among white and non-white residents while maintaining identical levels of support over that time period produces a swing of 7 points in favor of Democrats.
2) White demographic shifts - I'm not entirely sure how to calculate this, but over a 10-year period, roughly 50,000 whites older whites would have died in Mecklenburg. Considering that there is almost an identical number of raw whites in the county in 2010 as there were in 2000 (~415,000), they've been replaced by a combination of births and (mostly) influx of white professionals from outside the state. If we assume that the whites who passed on were 75% Republican in 2000 and the whites who replaced them would've been 50/50, between 2000-2010, this would produce an average swing of 6 points in favor of Democrats.
3) Actual campaign infrastructure - While NC was still largely Democratic at the state level during the Bush years - and even had a Democratic Senator - the kinds of investments that come with presidential campaigns are above/beyond anything that would likely come from the state or local level. Considering Mecklenburg specifically was heavily-targeted by multiple presidential campaigns, this produces a swing of 3-6 points in favor of Democrats*.
2008-12 Only:
- 4) Improvements in non-white support - While we accounted for the increase in the share of non-white voters above, we also have to account for the increase in support for Democrats (at least for 08/12, as 16 looked a lot like 00/04). Assuming that non-white voters' support of Democrats swung by 15 points between 2000-2012 (say, from around 77% to 85%), this produces a swing of 5 points in favor of Democrats.
2016 Only:
- 5) The standard suburban swings of '16 - In southern metropolitan areas like Mecklenburg, we saw massive swings to Clinton compared to 2012. While Charlotte is an urban area, Mecklenburg has plenty of suburbanesque areas in it as well. In similar metro areas (Dallas, Atlanta, etc), we saw some pretty consistent swings of around 15 points in favor of Democrats.
This accounts for a 21-24 point swing in favor of Democrats from 2000-2012 (scenarios 1-4) and a 31-34 point swing in favor of Democrats from 2000-2016 (scenarios 1-3 & 5).
That's right on par with Clinton's 30-point win in the county in 2016 and Obama's wins of 24 and 22 points in 08/12.
* According to conventional political wisdom, a good groundgame/field op improves a candidate's performance by "3 points" - I always forget if that's a 3-point improvement in candidate's margin or 3-point improvement in candidate's share of the vote, hence the range
Thanks President Griffin for an extremely detailed explanation and analysis....
Do you have any particular data on the wealthiest parts of County, so maybe we can pull some precinct results and look at the '12 > '16 shift, since there's another thread where many of us were pulling up detailed precinct level data on the Obama '12/ Clinton '16 swings, and the fascinating thing is that it looks like there was something like a 15-20% swing regardless of regions of the country, Metro Area, nor traditional Party affiliation....
Now, I believe that I managed to locate a comprehensive precinct level dump that includes NC in 2000, so there might be some data to pull on that part of your post, as opposed to Macro-Level County Demographic shifts.....
