The Impact of Voter Turnout: 2016 Presidential Election Results by County, City, and Precinct — Michigan
With the upcoming midterm elections, I have seen a lot of discussion about turnout and the impact it will have on the election results. With that in mind, I wanted to take a look at a state that had a close result in 2016, and see just how much a difference turnout may have made. While finalizing this, just in the last couple of days, I have read articles citing the approve/disapprove polling for President Trump amongst black voters, and also regarding gerrymandering and claimed voter suppression tactics in Michigan. Both of these are either reflected in and/or could have an impact on the state’s results.
First, let’s look at the 2012-to-2016 change in turnout by county in the map below (also Tableau link if you are interested). The color of each county represents how it voted in 2012. Turnout mostly increased in red counties and decreased in blue counties, with some pretty big differences in the bluest and reddest counties. Notice especially the dark blue in the lower right-hand corner of the state (at the base of the thumb, as Michiganders say). One solid blue district had +5% turnout change, and the one next to it had -5% turnout change. While it is impressive that the lighter blue county (Washtenaw) saw a turnout increase, its blue margin in 2012 was one-sixth that of the darker blue county (Wayne). So obviously +5% in Washtenaw County does not have the same impact as-5% in Wayne County.
To illustrate the point, the election margin for the entire state was less than 11,000 votes (red). The 2012 margin in Wayne County was 382,000 votes (blue). Wayne County still voted overwhelmingly blue in 2016, but because turnout dropped in the county, the 2016 blue margin was less than 290,000 votes, or about 92,000 worse than in 2012. That’s remarkable given the 11,000 state margin. In other words, if Wayne County turnout had only dropped 4% instead of 5%, all other things being equal, Hillary Clinton would have won the state. Would Democrats want to target the entire county with turnout strategies, or where would those strategies potentially have the most impact? Let’s look at that county in more detail.
Here is a view by city in Wayne County (Tableau link). Again we see some blue cities with increasing turnout, and some with decreasing turnout. But the largest dot (where size of dot represents 2012 blue-ness of the city) is Detroit, and its turnout dropped a whopping 15%.
We can look at Wayne County by precinct, as well. Here is a map I plotted (with the help of Tableau, of course — link here) showing the blueness or redness of the 2016 vote in each precinct in Wayne County. That dark blue area toward the right and near the top is around Detroit. Much of the rest of the county is mixed between red and blue. It’s easy to see, then, that if the city of Detroit had just about the largest drop in turnout that it would be bad news for the Democrats.
For some added insight, I was also curious to see if there was any correlation between the voting pattern by precinct in the county and the income level for each precinct. Below is a map I plotted showing income level (Tableau link). The lightest green areas are the areas with the lowest per capita income. That includes Detroit, the darkest blue area above. The darkest green are the precincts with the highest incomes, and corresponds to many of the red precincts above.
Here is a side-by-side view to make it easier to spot the parallels. You can even see the small red area in the northeast of the first map aligns with the dark green area in the second map (Grosse Pointe). The spot of dark red in the middle is the City of Dearborn. There really is nothing as blue as Detroit west of there except for the cities of Inkster and Romulus. It’s pretty easy in all of these cases to see the alignment between the two maps.
If you want to interact with these maps, they are all available at Tableau Public with the links provided above.
So, to bring it to a strategic level, if I was advising Democrats in 2018 (Senate) and 2020 (Presidential), I would focus on turnout in Wayne County, specifically the city of Detroit. It seems to me the state vote could again hinge on that. This is why the news about gerrymandering and voter suppression makes such headlines here.