Election Surprise

The reason why we count the votes is that the voters sometimes surprise us.

In the race for the 2008 World Science Fiction Convention, everybody who knew anything knew that Chicago was going to defeat Denver (worthy but starting late and short of personnel) and Columbus (perhaps not hapless but a couple of buckets short of hapful). Among those who knew were –

• Kent Bloom, chairman of the Denver bid, who didn’t bother to keep his bid’s suite for a Saturday night victory party,

• Seth Breidbart, perennial publisher of hoaxzines, who could think of no better joke headline than “Denver Wins!”, and

• me, who would have been willing to bet a moderate chunk of my net worth on a Chicago victory, had I found anybody foolish enough to take the other side.

Yet, when the votes were counted, Denver eked out a 12 vote margin, 672 to 660 on the third redistribution of ballots. (If you are familiar with Worldcon voting, you know what that means. If you don’t, you don’t care.) It was the closest outcome since the days when the selection was made by a hundred or so people at the business meeting.

What may have killed Chicago’s chances was turnout: the lowest for a contested election in a long time. In fact, there appear to have been fewer votes cast than when Chicago won the 2000 Worldcon unopposed. My impression, based on admittedly unscientific, anecdotal evidence, is that Denver’s strength lay among veteran convention-running fans, most of whom know and respect the bid’s chairman, while Chicago was more popular among the Joe Phans who don’t have much interest in the personalities behind the bid committees. The small size of the electorate magnified the influence of the former group. Why so few people voted, I can’t say. Perhaps L.A.Con’s unusual schedule, running Wednesday through Sunday instead of the traditional Thursday through Monday, threw off fans’ circadian rhythms.

I’ve heard other theories: that Chicago suffered a backlash for running “too soon” after Chicon 2000, that fans in search of variety preferred a city that hasn’t hosted a Worldcon in a quarter of a century over one that has held three during that period, that particular fannish subgroups hold grudges against Chicago for this reason or that, and so on. When the complete tally is published tomorrow later this morning, it may furnish insights. Or it may just add to the puzzlement.

Having served as treasurer of the Chicago bid, I didn’t leap up and cheer when I heard the results. OTOH, I thought after a while of the confident predictions that I read every day of Democratic gains in the November elections. Maybe the Worldcon election is an omen. At least, it’s a reminder that unexpected things can happen in the voting booth.

Update, 8/26/06: The vote count detail is full of the unexpected. Well, this is a science fiction convention.

First, it confirms that turnout really was low: 1,561 ballots cast, which is about what one would expect at an overseas site or in a particularly dreary uncontested election. Over 1,600 were cast when Chicago won the 2000 Worldcon unopposed at LoneStarCon (1997). L.A.Con has 2,000 more members yet somehow produced fewer voters in a hotly contested race. Very strange.

The first distribution of ballots was –

Denver: 494
Chicago: 489
Columbus: 433
Joke write-ins: 90
None of the Above: 10
No Preference: 45

The star of the write-ins was Hollister, California, the hoax bid pushed by various Bay Area fans. It got 79 votes, probably the best ever for a hoax. (I ignore those who insist that New Orleans in 1988 was the biggest Worldcon hoax ever.)

After redistributing the write-in and “None of the Aboves” (“No Preferences” are abstentions and are not redistributed), the count stood at –

Denver: 540
Chicago: 508
Columbus: 455

Thus the 100 redistributed votes broke 46 for Denver, 19 for Chicago and 22 for Columbus, with 13 making no valid second choice. This pattern suggests that Denver was a strong favorite among SMOF’s, as expected, since those are the people most prone to write in silly sites.

The Columbus votes were then redistributed. They divided in favor of Chicago, 152 to 132, with a remarkable 171 having no next preference. Chicago’s pickup wasn’t quite big enough. The final tally was Denver 672, Chicago 660.

My guess is that Columbus, which ran more strongly than anyone outside its own bidcomm expected, received support from a combination of Midwesterners, for whom Chicago was the most likely second choice, and fans looking for cheap hotel rates. Columbus’s promised $99 a night was well below the Chicago and Denver prices. Voters for whom that was the prime consideration probably didn’t care for either of the alternative sites or picked Denver as (slightly) less expensive.

The 2008 Worldcon will be the first since 1988 to be chosen only two, rather than three, years in advance. When the shortened lead time was adopted, I was alone among emeritus Worldcon chairmen in arguing that two years isn’t long enough for an event that now tries to do quite a bit more than 20 years ago. I hope that my worries are proven wrong. Happily, this first experiment in foreshortening has about as battle-tested a chairman as one could hope for and is sure to draw plenty of help from the Permanent Floating Worldcon Committee. So, in the first light of day, I feel, as Ronald Reagan used to say, cautiously optimistic.

Comments

[Kevin Standlee]

Presumably by "overseas" you actually mean "Non-US," as there were about that many ballots cast at ConAdian (San Antonio vs. St. Louis).

But anyway, I was initially also concerned about the low turnout, but I think that there are multiple factors here. As others said above, there were probably people who, despited everything they were told, assumed that voting would continue through Saturday no matter that the Worldcon was a day earlier than usual this year. It was hard enough for me to stick to Worldcon Daylight Time, and I knew what was going on.

But the factor that I think really affected this election is that it is the first one to be fully affected by both "no-zone" and 2-year bidding. In the old zonal/3-year system, Worldcon sites in North America were restricted to the same zone as was administering the election. Therefore, there was probably a disproportionate interest in the location from people within that zone. The biggest example of this I can think of was the 1991 race between Louisville and Winnipeg held in Chicago, where there were 2,107 ballots cast, split nearly evenly between the two candidate sites.

It took several years for "grandfathered" bids to work their way through the transition to "no-zone," and then we also shortened the lead time. In any event, we now require that bids be no closer than 500 miles/800 km from the host site. This means that there's far less "local vote" factor than there probably was when one or more of the candidates was a relatively short distance from the host city.

In other words, the lower turnout may not be a bug, but a feature, of the no-zone system. We at least in theory want people voting for reasons other than "It's near where I live," after all.

[RJ Johnson]

I think one of the contributing factors to the low turnout (in addition to the possibility of people thinking they could vote one day later than they could) was the unobtrusive location the polling place had in the exhibit hall. It was placed so as to fairly guarantee no walk by traffic, either to remind people to vote or to catch the curiosity of first time attendees who might not know much about the voting process.

I'm surprised about the amount of animosity that still remained towards Chicago as to the Art Show from 2000. I shouldn't be, I suppose --people have long memories when it comes to gripes. I'll be curious to see how the collective fannish memory is after five days of walking a minimum of 3-4 blocks from their hotel to the convention center. At altitude.

[Bruce M. Miller]

Having all the party suites on one floor is a better reason than you think. Even if you don't attend parties yourself, it still cuts the lines at the elevators by 60-90%.

[Bruce M. Miller]

My nutshell analysis: None of the three bids was good (hence the low turnout), but Denver's flaws were less public than the other two bids.
- Columbus probably has the best facilities, but were conspicuous by their absence (e.g., at Glasgow and Marcon).
- Chicago had two problems, both involving ChiCon 2000 (which was also conspicuous). Many people were tired of going to Chicago. Even more voted against Chicago on account of the Chicon 2000 art show. The Chicago bid committee thinks they addressed that sufficiently, but as far as I can tell, that opinion was only held by the Chicago bid committee. All I can say is that within a half hour of hearing that Denver won, two people told me they voted for Denver because they expected me to run the art show; two who voted for Columbus said that had they known who would have run Chicago's art show, they'd have voted for Chicago. That's a four vote swing just among people I talked to in half an hour. This doesn't sound "sufficiently addressed" to me.
- Denver has the worst facilities and has other problems I considered serious, but they largely remained hidden - none were publicly on display at a Worldcon.

[KD]

Someone who voted for Columbus told me that he voted for them because they had all the party suites on one floor, rather than having to spread them around on the few available suites on different floors. Although I find this an odd reason to choose a convention venue, I suppose there are even odder ones out there.