| 16th Count | |
| Cheung, Leland | 1775 |
| Maher, David | 1775 |
| Simmons, Denise | 1775 |
| Toomey, Tim | 1775 |
| Benzan, Dennis | 1775 |
| McGovern, Marc | 1679 |
| Kelley, Craig | 1565 |
| Mazen, Nadeem | 1549 |
| Carlone, Dennis | 1548 |
| vanBeuzekom, Minka | 1535 |
Nov 15, 6:55pm – The Official results in the City Council election gave the same winners as the Unofficial Results from last week. The only significant change due to the 7 additional ballots was a partial change in the Cheung surplus due to two additional #1 votes for Leland Cheung. The deciding margin at the end of the 16th Round grew to 13 votes separating Dennis Carlone and Minka vanBeuzekom. (Mazen had 1 more than Carlone.) In the final Round, Carlone surpassed Mazen to yield the following winners in the order they reached the election quota (1775).
City Council (in order of election): Leland Cheung, David Maher, Denise Simmons, Tim Toomey, Dennis Benzan, Marc McGovern, Craig Kelley, Dennis Carlone, Nadeem Mazen. [Detailed Report]
Incumbents defeated: Ken Reeves, Minka vanBeuzekom
Nov 16 Update: Minka vanBeuzekom has sent word to her entourage that she will be asking for a Recount.
School Committee (in order of election): Patty Nolan, Fred Fantini, Richard Harding, Kathleen Kelly, Fran Cronin, and Mervan Osborne. [Detailed Report]
Incumbents defeated: None
Official Final Election Results – City Council and School Committee (PDF)
Official Final City Council #1 Vote Distribution by Ward/Precinct (PDF) added Nov 20
Official Final School Committee #1 Vote Distribution by Ward/Precinct (PDF) – added Nov 20
Cambridge 2013 Election Count (in case you missed it)
Just in case you missed the live broadcast of The Count on Tues, Nov 5, the good folks at CCTV have now uploaded it in 3 parts for your viewing pleasure (not to suggest that will be your reaction, of course!). You’ll find the program (hosted by Susana Segat and Robert Winters) at: http://cctvcambridge.org/electioncount2013.
Bring on the recount!
Comment by patrick barrett — November 15, 2013 @ 7:40 pm
“Dennis Pearl Harbor Carlone” …. absolutely priceless Dr. Winters. I almost want to make a t-shirt.
Comment by patrick barrett — November 15, 2013 @ 10:35 pm
Robert,
Thank you for all your commentary and analysis, really appreciated. (On a minor visual note, I like the dice, don’t like them spinning all the time. They really distract from reading. At my age, I stop and start all the time, and rarely find exactly where I left off, and the spinning dice compound the problem. My wife, who is prone to motions sickness, can’t read the page at all.)
Re-count? Part of me says yes, always interesting, part of me says no, waste of time and money. Accept and move on. (Although, the 59 votes that separated Ward and and Reeves stills rankles me every time. No recount would have solved that problem, just a slightly better (1% better?) campaign.) Let’s move on to the intrigue of the next mayor.
Comment by Andy Farrar — November 16, 2013 @ 8:16 am
Andy – I swapped out the dice for something less distracting on http://rwinters.com.
I’m of two minds about the possibility of a recount. A candidate runs the risk of being seen less favorably for having forced a recount, and it’s no fun “losing twice” should that turn out to be the case. On the other hand, in a very close election there is a good chance that the supporters of a losing candidate will expect a recount and may be disappointed unless all such options are exhausted.
Recounts should always be about maximizing the accuracy of election results, and it’s certainly possible that some ballots could be reassigned upon visual inspection. For example, in the 2009 School Committee recount Nancy Tauber lost 7 #1 ballots, Fred Fantini gained 5, Richard Harding gained 5, Marc McGovern gained 1, Joe Grassi (who had sought the recount) lost 3, Alice Turkel gained 2, Patty Nolan gained 1, Alan Steinert gained 2, and Charles Stead lost 1. [http://rwinters.com/elections/school2009.pdf]
Shifting the initial distribution of #1 votes sets the stage for the rest of the count, but the real wild card remains the Leland Cheung surplus. It’s quite likely that some ballots would be either added to or removed from his stack of ballots and this would once again cause a change in which surplus ballots are chosen for redistribution.
On the question of choosing a Mayor, keep in kind that none of the new 4 councillors will be in contention, though I’m sure 2 of them may already be pledging “solidarity” in order to exert influence on the outcome. Unlike 2 years ago when there were almost 9 candidates for Mayor, the number this year will likely be at best 3 or 4.
Comment by Robert Winters — November 16, 2013 @ 9:05 am
Cheung Mayor, Simmons Vice Mayor. For various reasons, I don’t expect the newbies to vote as a bloc but perhaps the Clean Slate (Carlon, Mazen) vote together. I also don’t expect a recount since it seeks unlikely it changes the votes. Minka will have to run that extra 1% better as another commentator said up thread. Cheung has already been Vice Mayor, he ran an aggressive campaign to get those #1’s, and you know he wants it. I suspect he gets some if the newbies. It’d be great if it went to just one ballot this year too…
Comment by James Conway — November 16, 2013 @ 9:57 am
The fact that as little as one extra ballot awarded to someone with a large surplus changes the set of ballots that are redistributed such that it can alter the outcome is a very poor practice. An election system where the order that people voted or what ward-precinct they came from can affect the outcome is just plain wrong. It is time to address this problem. Perhaps if Minka does ask for a recount and she wins as a result will be an extra impetus for change, but the data that you have shown should be good enough to start the process.
Comment by John Gintell — November 16, 2013 @ 12:53 pm
Does the city make the actual ballot data available? It would be interesting to see how much the results change with different ballot order, or what the effect of different counting schemes would be.
Comment by Kent Johnson — November 16, 2013 @ 6:17 pm
The ballot data will be available at some point. I have the data for all of the Cambridge municipal elections we’ve done so far via computer tally. In preparation for a TV broadcast, I ran some experiments last Monday (Nov 11) to investigate what can happen with different ballot order. I also gathered some known information from past elections comparing what can happen when just a few additional ballots are included. What I put together is posted at http://rwinters.com/experiment/.
Comment by Robert Winters — November 16, 2013 @ 7:09 pm
Toomey for mayor. Also, based on the email I just got it looks like a recount is a definite.
Comment by patrick barrett — November 16, 2013 @ 11:31 pm
When was the last time a recount for the municipal election in Cambridge actually resulted in any different winners?
The most recent two for School Committee (especially 2009) were statistically (and otherwise) unnecessary.
Comment by Fred Baker — November 17, 2013 @ 4:12 pm
In every Cambridge municipal election recount since the adoption of the Plan E Charter the results did not change from the original. It’s important to note, however, that in the 2001 School Committee election the results were so close that the recount could easily have produced different winners. In fact, based on the original ballot data, use of fractional transfer would indeed have produced a different result. The election would still have likely resulted in a recount which could have produced different winners.
In many places, a near tie triggers an automatic recount.
Comment by Robert Winters — November 17, 2013 @ 4:31 pm
Robert,
Thanks for your hard, shared work on this.
However, it would be very helpful (at least for me) if you could explain the various systems for transferring ballots. I cheerfully admit to being an ignoramus on such things as the Cincinnati method, etc., etc. How do these things work, and exactly which system is the City using? Perhaps you have posted this info in the past.
Comment by Martha Older — November 20, 2013 @ 9:04 pm
In the original method used by Cambridge (the Boulder Method) if a candidate had 2400 #1 votes and the Quota was 2000 votes, the last 400 ballots would be the ones chosen for redistribution to other candidates based on the next available choice. This method puts a lot of bias on the order of the precincts. That’s why we do the “random draw of precincts”.
In 1973, the Election Commission voted to adopt the Cincinnati Method instead. In this method, since 400 ballots are to be pulled out of the original 2400, they would pull every 6th ballot. If the actual numbers were not so even, they would round off to the nearest integer (n) and pull ever nth ballot. This removes the bias of the precinct ordering since you’re pulling ballots more uniformly spread through all precincts. The down side is that if an error is discovered that causes even one ballot to be inserted, this shifts the ballots and causes a very different set of ballots to be selected for redistribution to other candidates.
A more preferable alternative is to use a fractional transfer method in which each ballot starts with a “weight” of 1 and a fraction of ALL of an over-quota candidate’s ballots are transferred to other candidates. In the original example, 5/6 of every one of the 2400 ballots would remain, and 1/6 of every one would be transferred to the next available choice as expressed on the ballot. It’s a little strange at first to see vote totals that are not whole numbers, but the advantage of no longer having any dependence on ballot order far outweighs the aesthetics of using weights rather than whole ballots.
Comment by Robert Winters — November 20, 2013 @ 9:37 pm
OK, I think I follow, and it does seem better — but in the method you recommend, what happens after the transfer of #2 choices? How do you transfer the #3 choices (likely to be needed) — when you would probably then have additional candidates who’ve made quota?
Comment by Martha Older — November 26, 2013 @ 4:25 pm