As if you need any reminders about close elections (See: HoCo Council, District One, Democratic Primary), I examined the last five General Election outcomes for the Howard County Board of Education. Nothing better than armchair psephology.
This post is a bit long, but bear with me, and there are some key takeaways at the end.
Factoring in the number of votes, expressed as a percentage of the overall turnout that cycle, between the “last place winner” (the person who finished 3rd of 3, or 4th of 4, depending on the year) and the person who placed in the next position, there was a less-than-5% margin of victory in all five cases.
In reverse chronological order:
In 2016, Mavis Ellis was 3rd and won a seat on the Board of Education with 58,431 votes. Janet Siddiqui, an incumbent, finished fourth and was thus voted off the BoE. She received 53,762 votes. A 4,579 vote difference may seem like a fairly comfortable margin of victory, even for a County the size of HoCo, but let’s take a look at the turnout. In that General Election, a total of 163,668 voters cast their ballots. 4,579/163,668 = only 2.8% of the overall turnout. Had a couple thousand votes switched between the two and/or Siddiqui managed to convince some additional voters (including former Siddiqui voters from previous cycles) to make her their first, second, or even third choice, Siddiqui might have pulled out a razor-thin W. Remember, there were a total of 345,781 BoE votes cast in that election (between the 6 candidates on the ballot + write-ins). That amounts to an average of 2.1 BoE votes cast per voter. Under-votes also make a difference. In that context, a 4,579 vote difference is hardly a chasm. Siddiqui actually obtained more Absentee/Provisional votes than Ellis, and secured only 1,390 fewer Early Voter ballots than Ellis…but lost on election day by 3,514 votes.
In 2014, the relative gap between the last winner and the top, well, non-winner was somewhat wider, a 4,447 vote difference out of only 106,158 votes (oh yes, the low-turnout 2014 general election…we all recall vividly the horrors of it). In this case, it was Christine O’Connor (37,561 votes) finishing ahead of Dan Furman (33,114 votes). As a percentage of the overall turnout, that works out to a 4.2% margin (the highest of the five most recent BoE General Elections). Again, the average voter cast their ballot for 2.6 BoE candidates that year. Had that figure been slightly higher, with more Furman voters…and/or had a couple thousand voters opted for Furman instead of O’Connor as one of their up-to-four choices, he would have been elected. Furman also ran only a few hundred votes behind O’Connor amongst Early Voters and 133 behind her with Absentee/Provisional voters, but trailed her by a larger margin among Election Day voters.
In 2012, the difference between 3rd (Win) and 4th (Lose) was 3,558 votes, with Ellen Flynn Giles (50,908) and Bob Ballinger (47,350) in those respective positions. That 3,558 vote total represents only 2.3% of the overall turnout (154,375 votes overall). Giles’ incumbency definitely helped enable her to secure re-election. Interestingly, Ballinger was behind Giles by only 209 votes among Election Day voters and by 425 votes among Absentee/Provisional votes, but lost by a wider margin among Early Voters.
The other two cycles are particularly interesting.
In 2010, the difference between 4th (thrill of victory) and 5th (agony of defeat) was a mere 393 votes out of 108,423 cast, which represents only 0.36% of the vote separating those two candidates. In this race, Cindy “Landslide” Vaillancourt garnered 30,990 votes, slightly ahead of David Proudfoot (30,597). Proudfoot actually received more Early Votes than Vaillancourt, but obtained fewer among Election Day and Absentee/Provisional voters. It is worth noting that the candidates on either side of them in the vote total column also finished with not-too-dissimilar numbers. Brian Meshkin had only a few hundred votes more than Vaillancourt (31,707 total) while Bob “Tough Luck” Ballinger was about a thousand votes behind Proudfoot (29,627) with Leslie Kornreich trailing him by a small margin (with 29,375 votes). Even the last-place finisher was not that far removed from the rest of the field, with Larry Walker placing eighth with 27,546 votes. Walker’s position is of particular note since he placed third, yes, third in the primary election that year. Noteworthy indeed.
Finally, in 2008, the difference was a scant 689 votes out of 146,304 cast (0.47% of the total). In this case, it was Allen Dyer with 54,148 votes outpolling the next-place finisher, Diane Butler (with 53,459 votes). Butler garnered more Absentee votes than Dyer, but lost on Election Day by around a thousand votes.
In 2008 and 2010 alike, we are talking about a shift of a handful of voters, per-precinct, changing the outcome.
So what does this all mean?
Candidates cannot rely on the support of external organizations, even powerful and popular ones, to ensure victory. While running on the Apple Ballot is extremely helpful in BoE primaries, it does not guarantee success in general elections (in 2014: see Dan Furman & Zaneb Beams; in 2012: see David Gertler & Jackie Scott). And with Hogan pulling his shenanigans this year with his Rotten Apple, will voters have doubts in their minds when they see the Official Apple Ballot, thus possibly tarnishing its value a bit this cycle? Stay tuned.
With no incumbents on the ballot this time around, I am anticipating a more even distribution of votes. The joys of parity. Without any mortal-lock heavies, that means it is quite likely that the gap between 4th and 5th place…or between 3rd and 6th place for that matter, might be relatively small. Perhaps one candidate pops head-and-shoulders above the others, but with voters facing up to four choices (and the average voter marking their ballot for, say, 2.5 of them, with no widely-known and well-liked incumbent on the ballot) I don’t think more than one will pull away from the rest of the field. Frankly, I would not be surprised if the election results for 2018 are similar to those of 2010 in terms of vote distribution. It’s a cliché, but so much comes down to turnout, turnout, turnout. Early voting, Absentee voting, and Election Day voting.
You can’t assume a strong primary election showing means a strong general election showing, especially if voter turnout in the General is higher than it was in 2014. If one has deep support, that is great…but it better be wide enough to appeal to voters who show up for general elections in mid-terms, but not necessarily primaries. GOTV strategies (and related communications) should reflect this reality.
While I have some thoughts on the likely order of finish, I believe that six of the eight campaigns have a serious shot at a top four finish…and it is not inconceivable that one (or both???) of the other two, albeit under some highly unlikely circumstances, could somehow find a way to win. That said, anyone who tells you that this race is a done deal has no idea what they are talking about. Historically, the margins have been close enough to be impacted by October communications, by interactions with and between voters at their homes, at campaign events, or at polling centers, by late-breaking events, by the last-minute voting booth choices of voters, and by the decisions of some to just not cast a ballot this time around.
We may be in for a long night of counting. Be prepared.