We took a deep dive with DDHQ’s 2016 election data, looking at Michigan specifically. Michigan was one of the three crucial states that flipped the election to Donald Trump (along with WI and PA). A state that Republicans had often targeted but never able to break through in, Donald Trump was able to eek out a 11K vote victory. The following maps explore some of the trends that help explain Michigan’s rightward shift in 2016.
Each map has data presented at the precinct level, so we’re able to see much more finer details than what is seen with basic county level data. For instance, the Detroit area is generally regarded as a Democratic stronghold (and it is), but zooming in on the greater Detroit area shows major differences between different parts of town.
The first map has 35 different variable layers that are relevant to voting patterns in 2016. Below, we go more in-depth on specific patterns that can be discerned by comparing different variables.
Map 1. Title: Election Results and Shifts
We start with the basics: the two-party share of the 2016 vote, 2016 turnout, and party shifts from 2008 to 2012, and 2012 to 2016. We see a very strong rural-urban split in the two-party vote coupled with weak urban turnout. The highest turnout levels are in suburban areas- historically Republican but less so in 2016. Ultimately, rural turnout was higher than urban core turnout, and that was one of the several factors that allowed Trump to narrowly flip Michigan.
The swing maps are especially striking when compared to each other. In 2012, most areas moved slightly towards Republicans, and Democratic gains were concentrated in cities but otherwise disbursed. By contrast, in 2016, rural areas moved decidedly towards the Republican ticket, the historically light red Republican suburbs moved substantially towards the Democratic ticket, and cities were mostly flat, but shifted slightly towards Republicans.
Map 2. Title: Democratic weakness since 2008
In 2008, Obama won Michigan by over 16%- a very strong result and the modern day high point for national Democrats in Michigan. The wave began to recede in 2012, when Obama won by slightly less than 10%. And the bottom fell out for Clinton in 2016, becoming the first Democrat to lose Michigan since Dukakis in 1988.
The rural shift is striking in Clinton’s map. She is below 30% nearly uniformly across the state, with the exception of suburbs and cities. Areas that Obama won voted for Trump over Clinton by double digits, and areas where Obama held his own in narrow losses were completely lopsided this time around.
Map 3. Title: Republican resurgence in rural areas
The inverse of the previous maps, this series demonstrates the growth of Republican strength in Michigan- and the few areas where support actually dropped. While the base of Republican support in Michigan used to be near Grand Rapids, several rural areas throughout instead display the same level of support. And some areas, like Ann Arbor, actually move significantly away from Republicans. On the whole, the shifts clearly benefit Republicans.
Map 4. Title: Primary carryover into the general election
Many of Clinton’s weakest areas of support in the general are Bernie Sanders’ highest areas of support in the primary. In the primary as in the general, Clinton was favored by polling- before then losing narrowly.
Trump’s primary map tracked his level of support in the general in some areas (upper peninsula, areas north of Lansing), while they foreshadowed areas that would not be as receptive to Trump in the general as Republican candidates before him.
Map 5. Title: The divide between racial groups and education
The African-American population in Michigan tends to be clustered in cities and Detroit/Ann Arbor suburbs, which were areas of strong Clinton support. The Hispanic population is centered to the east of Dearborn, and in Southwestern Michigan.
The White population can be viewed most sharply through an education lens. A long stretch of educated white voters can be found between Ann Arbor and Macomb outside of Detroit- and this long stretch matches the largest support increase from Obama to Clinton in the entire state. The same dynamic can also be found outside of Grand Rapids, Lansing, and Kalamazoo. On the other hand, strong areas of whites with no college degree throughout Michigan strikingly track Trump’s largest gains perfectly.
Map 6. Title: The myth of economic anxiety?
However, the talk of economic anxiety predicting support for Trump may be overplayed. Yes, his surge of support is predicted by education- but no college degree does not necessarily mean economic hardship, even in 2016. Home values are highest in the suburbs around Detroit and Grand Rapids, in very educated areas, but tend to be very low in inner city Detroit, Flint, Lansing, and Grand Rapids. On the other hand, home ownership rates are highest in the rural areas- so even though the homes are not worth as much as in upscale suburbs, rural residents tend to have assets that those in cities do not. Income inequality, too, tends to be higher in cities, and poverty rates are among the lowest in rural areas, with some exceptions near Big Rapids.
Map 7. Title: The difference in occupation industries
Manufacturing jobs in Michigan are clustered near Holland, the areas of historical Republican strength. Ironically, this area actually moved away from Trump in 2016, despite his emphasis on manufacturing. Technology jobs tend to be clustered in white college-educated areas, which makes sense. These areas too moved towards Democrats in 2016.