I was intrigued by DKE’s story from last Tuesday detailing how a Jim Crow era blurb in the Mississippi Constitution might cost Democratic Attorney General Jim Hood the gubernatorial election next year even if he wins a majority of the vote. This is because Mississippi requires the winner of the popular vote (with a majority of the vote) to also win a majority of the “electoral vote” in order to win outright. The electoral vote is determined by the winners of each of the state’s 122 House of Representatives districts. If the winner of the popular vote does not win a majority of both the popular vote and the electoral vote, the House of Representatives gets to decide who will be the next governor. Given that Republicans currently hold a 72-48 majority, even if Hood wins a majority of the popular vote, I’d be surprised if he was seated if he fails to win a majority of the electoral vote. I agree with Stephen Wolf’s speculation that there’s a good chance this could be struck down under the Voting Rights Act, but even so, I figured it was worth taking a look at the voting patterns of the state house districts. While DKE has wonderfully calculated presidential results for legislative districts in 48 states, Mississippi isn’t one of them, so for this project I had to do some good old-fashioned grunt work.
One rather glaring omission from the original story is this little snippet from the Mississippi Constitution: “All other state officers shall be elected at the same time, and in the same manner as provided for election of Governor.” In other words, the Lieutenant Governor, Secretary of State, Attorney General, Treasurer, Auditor, Insurance Commissioner and Agriculture & Commerce Commissioner are all elected under this same system, where they need to win majorities of the popular and electoral votes to win outright. To my knowledge, this procedure has never been used to seat the popular vote loser, but with today’s Republican Party, that isn’t likely to stop them if the chance arises. One important feature of this system is that we can examine how Hood has performed in his four elections as Attorney General using the electoral vote:
2003: Won the popular vote with 62.7%, winning the electoral vote 100-22
2007: Won 59.8%, won 95-27
2011: Won 61.1%, won 98-24
2015: Won 55.3%, won 66-56 (the total column says 69-53, but if you check the data, that’s inaccurate)
A majority of Mississippi’s electoral vote is 62, so if Hood had won five fewer state House districts, the election would have been thrown to the state House to decide the winner, despite Hood having won by double digits statewide. And this was precariously close to happening. Eleven districts were decided by 4% or less, and Hood won nine of them. He won 48.9%, 49.2%, 50.3%, 50.4%, 51.2%, 51.2%, 51.6%, 51.6%, 51.8%, 51.8%, and 51.9% in these districts. This was an incredible stroke of luck, and is insanely impressive when comparing Hood’s results to that of the Democratic gubernatorial candidate, Robert Gray. In these same 11 districts, Gray’s performance ranged from 14.7% to 30.8%. That’s not a typo. Hood won districts where Gray took 13.1%, 13.8%, 14.7%, 16.9%, 16.9%, 17.9%, and 19.3%. In other words, Hood’s victory in the electoral vote depended on winning districts where his ticket mate was winning less than one in five voters!
Speaking of Gray, it seems likely that his weakness as a candidate contributed significantly to Hood’s relatively lackluster (but still impressive) performance in 2015. Gray received 32.2% of the vote, the only time a Democratic gubernatorial nominee in Mississippi has ever received less than 39% (which belongs to 2011 nominee Johnny DuPree). Since Hood will be the one driving turnout in 2019, I doubt that we will see a lackluster result in the gubernatorial race, and almost certainly the best result since at least Ronnie Musgrove’s 45.8%-52.6% loss in 2003. Polls of Hood vs. Tate Reeves, the Lieutenant Governor, have shown 5 and 7 point Hood leads and a 7% Reeves lead. In a mark of how weak a candidate Gray was (and how popular incumbent Phil Bryant was), Hillary Clinton overperformed Gray in 117 of the 122 House districts, by an average of 8.2%. However that was still only enough for Clinton to win 45 districts (compared to Gray’s 37). So while I believe Hood’s 2015 electoral vote performance might have been a motivating factor in declaring his gubernatorial bid (in addition to the national environment and the lack of an incumbent), I don’t think he’s necessarily in as dire shape as the trendlines would indicate. One tidbit I found interesting was related to ticket-splitting. While Hood’s largest raw overperformances came in districts where Gray did poorly (Hood beat Gray’s vote share by 25%+ in 35 districts where Gray lost), he relatively overperformed Gray more in Democratic areas. Ignoring the minimal undervote (less than a 2500 vote difference between the races), Hood won an average of 46.6% of non-Gray voters in the 42 districts where Gray took at least 45% of the vote. When you look at the other 80 districts, Hood only carried 30.0% of non-Gray voters. It is hard to attribute this to gerrymandering (apparently Republicans did draw the maps, so this is a case of effective gerrymandering), although racial gerrymandering certainly plays a part, since Democrats controlled both chambers when the current legislative maps were passed, but I see this as confirmation that the map does indeed place Hood at a relative disadvantage. The concentration of Democratic votes is even more harmful to Hoods chances; for example, in 2015, Hood won 70%+ of the vote in 39 districts, while he took less than 30% in only four districts.
It’s no secret that voting patterns in Mississippi are highly correlated by race. Part of why Hood has continued to win is that he can still win an appreciable number of votes from the non-Hispanic white population of Mississippi. As you can see in the chart below, Hood’s (green) overperformance compared to Gray (orange) was similar no matter how white the district was, although his overperformance in whiter districts was more sporadic. When compared to Clinton’s (blue) returns, Hood’s relative strength among white voters sticks out, although it is worth noting that he clearly overperformed with POC as well. Hood won greater shares of the vote in 109 districts than those district’s percentages of POC residents. In nine of the 13 other districts, he still won at least 69% of the the vote. Three of the remaining four are heavily conservative, while the fourth, the 40th district, is perhaps the most obvious pickup opportunity for Hood. Comprised of most of the Memphis suburbs of Southaven and Horn Lake, the district is rapidly diversifying and is the only Clinton district in the state where Hood lost. At 27.3% Gray, 45.1% Hood, 49.5% Clinton, and 51.3% POC, this is ripe for the taking, even though it’s located in historically uber-conservative DeSoto County.
If you assume voter turnout is proportional by race (which it is not, so the following analysis very likely undersells Hood’s performance with white voters), Hood won at least 40% of the white vote in eight districts. Two of those are heavily POC and Democratic districts. His best performance with whites comes from his home district, the 22nd, a 34.0% Gray, 73.8% Hood district that is 58.0% white, so Hood probably received the votes of at least 55% of whites. The other four are also in the northeast corner of the state, so it seems he retains substantial regional appeal (Gray took 13.8%-19.3%, Hood took 47.9%-54.6% and won three). All four are among the twelve whitest in the state, and the district where he took 47.9% is the single whitest in the state (the next whitest is the most conservative in the state, at 22.6% Hood, 7.8% Clinton and 6.3% Gray).
If you want to look at a district in more detail, you can find the maps here. Without further ado, here are the details on the districts: