In the elections world, human and machine imperfection are variables that have to be factored in when discussing election results, turnout, and the like. Some of those imperfections can prevent every vote cast from being counted – a reality that became apparent during the Florida recount in the 2000 Presidential race. For this article, we are focused on the human side. More specifically, when an election consists of multiple races on the same ballot, there is a noticeable drop-off in voter turnout between the race at the top of the ballot and the race at the bottom of the ballot.
To illustrate, in the October 2011 statewide primary, 1,023,163 officially voted in the Governor’s race that saw Governor Jindal re-elected in a landslide. Also on the ballot were the races for Lt Governor, Secretary of State, Agriculture Commissioner, and Insuramce Commissioner. Did the same 1,023,163 who voted for Governor also vote in those other statewide races ? The answer to this question is “no”: while 1,023,163 voted in the Governor’s race at the “top of the ballot”, that number steadily decreased for those other offices also on the October ballot, as turnout varied from 890,786 who voted in the Secretary of State’s race (where Tom Schedler was narrowly re-elected) to 965,876 who voted in the Insurance Commissioner’s race that saw Jim Donelon comfortably re-elected.
While “ballot drop-off/fatigue” is not exactly news, what has not been covered until now is whether “ballot drop off” is more or less prevalent among particular voter demographics. To answer this question, we analyzed the October 2011 statewide races and the December 2014 runoffs in two Congressional districts (in that scenario, both the US Senate and Congressional runoffs were on the ballot). In looking at these races, we found that to varying extents, black turnout drop-off was noticeably higher than white turnout drop-off for downballot races.
To explain this analysis, turnout steadily dropped when going down the ballot from Governor to Lt. Governor, then to Secretary of State. And in those latter two races, turnout drop-off was even more substantial among black voters than among whites. When comparing turnout for the Governor’s race to the Lt. Governor’s race, white turnout dropped 4.4%, and black turnout dropped 17.2%. When comparing turnout for the Governor’s race to the Secretary of State’s race, white turnout dropped 9.5%, and black turnout dropped 27.0%. This racial disparity also happened when we looked at the December 2014 runoff for the 5th Congressional District: (northeast Louisiana/the Florida Parishes) between the US Senate and the Congressional race, where 1.7% less whites voted for Congress than for US Senator, while that figure was 4.9% for blacks. The differences were more muted, however, in the 6th Congressional district (areas around Baton Rouge and Houma/Thibodaux): white drop-off was 0.7%, while it was 1.3% in black precincts.
There is an anomaly in this analysis that deserves some mention: while the offices of Insurance and Agriculture Commissioner are actually after the Lt Governor and Secretary of State’s races on the statewide ballot, turnout drop-off was actually LESS than it was for Lt Governor/Secretary of State. For Agriculture, a 5.3% drop-off for whites and 9.9% for blacks occurred, while for Insurance, the drop-offs were 5.2% for whites and 9.2% for blacks. Why was this the case ? The Lt Governor and Secretary of State races were races between two Republicans, while there were Democratic candidates for Insurance and Agriculture. It’s entirely possible that partisan Democratic voters skipped those two “all Republican” races, while deciding instead to participate in races (Insurance/Agriculture Commissioner) where Democratic candidates were on the ballot.
So why would turnout “drop-off” even matter ? Given the racial polarization that exists in Louisiana elections, a higher black turnout benefits Democratic candidates, while a lower black turnout hurts their chances. And if we apply that to the 2011 statewide races and analyze the impact, we start off with the 2011 voter registration’s being 65-30% white/black, while the racial breakdown of those who voted (presumably in the governor’s race) was 72-25.5% white. However, the racial disparity that existed for the downballot races meant that the electorate for the Lt. Governor’s race was an estimated 75-23% white/black. For the Secretary of State’s race, it was 76-22% white/black, while those voting for Insurance/Agriculture Commissioner were 73-25% white/black. And in a closely contested “down ballot” partisan contest, this decrease in black influence of up to 3.5% could make the difference in some races, and that is why turnout drop-off matters.