by George Flynn
While I was working on the 1988 version of this article, Bruce Pelz asked me if I could compile a tabulation of site-selection voting figures. The two subjects seemed to go together well, so I decided to publish them together; this still seems appropriate. The total vote counts have already been included in Table 1, but these depend a lot on how many (and which) candidates are running. Of greater interest are the actual voting counts, to which this appendix is devoted. (It would of course be nice to have a compilation of the Hugo voting counts too, but that would take a fairly large volume.)
The present site-selection system, with both mail and at-convention balloting, has been in effect since 1972; before that the voting was done only at the Business Meetings (with a couple of exceptions for overseas cons). The last serious contest under the old system was in 1969 (at St. Louis), when Boston defeated Washington 169–119 for the 1971 Worldcon; as it happened, that was my first Business Meeting. (Note that the at-con voting continued to take place at the Business Meeting until the mid-’70s.)
As in the Hugo discussion, the years listed below indicate when the voting took place (with the voting sites in parentheses). Note that the 1971 through 1988 Worldcons were chosen two years in advance, and those since 1989 three years in advance. During the ’70s the numbers released were somewhat spotty, so I’ll start by discussing each year separately.
1972 (Los Angeles):
109 mail, 147 at-con ballots; total 256. At the last minute New York withdrew, and Washington was declared the unanimous winner; no voting counts were released.
Melbourne 330, Los Angeles 45; total 375. (If any separate mail and at-con figures were announced, I can’t find them. Here and for the next few years, I don’t have any numbers for “No Preference,” write-ins, etc., so the “totals” are probably low.)
|1st ballot||2nd ballot|
This was one of the only two times in the mail-ballot era when a second ballot has been required to determine the winner (1990 was the other one). I assume the extra 32 votes on the second ballot came from what I call “other” votes below.
Orlando 298, New York 111, Washington 91, Philadelphia 28; total 528. I don’t have the mail/at-con breakdown, except that the great majority of votes were by mail.
|1976 (Kansas City)|
1977 (Miami Beach):
884 votes cast. Brighton overwhelmingly defeated New Orleans; vote made unanimous without revealing vote counts.
Boston defeated Baltimore, but the breakdown wasn’t released; the “No Preference” votes were announced as for the hoax bid of Flushing. Here’s what was released:
|Boston or Baltimore||379||584||963|
(The invalid ballots mostly lacked the voting fee — probably because the ballot neglected to mention that one was required. As far as I know, there haven’t been anywhere near as many invalid ballots in any later year.)
From 1979 on, complete voting counts have been released, and since 1980 the WSFS Constitution has required their publication. In the tables that follow, the totals for “No Preference,” “None of the Above,” write-ins, and (sometimes) void ballots have been combined into a single “other” line to save space; they’re usually negligible except when there’s an unopposed bid (or one with no effective opposition). No further commentary should be necessary.
|1984 (Los Angeles)|
|1986 (Atlanta), for 1988|
|1986 (Atlanta), for 1989|
|1988 (New Orleans)|
|1990 (The Hague)|
(The “other” column comprised problem ballots that were counted separately; I presume they included both mail and at-con ballots.)
|1993 (San Francisco)|
|1996 (Los Angeles)|
|1997 (San Antonio)|
I note that every site-selection vote since 1981 (even in Australia) has had more at-con than mail ballots, and that the typical ratio seems to be getting more disproportionate in recent years. Is this due to an increasing distrust of the mails? (The highest at-con percentage was 83.10% in 1987, when there, ah, seems to have been a problem with the mail ballots; but 1998’s 83.07% was about as close as one can get.)
This suggests that my 1980 counts by date of receipt are no longer very relevant, but I might as well include the data anyway; the ballots went out on May 1, with a nominal mail-in deadline of Aug 15:
|Jun 29–Jul 5||39|
|Jul 27–Aug 2||31|
(1 mail ballot arrived after the con.) The major peak was of course at the Hugo/preregistration deadline.
For the at-con distribution of votes, we do have data from some recent years:
|1998||W: 342||Th: 488||F: 971||1801|
The first three averaged out pretty steadily to about 15% on Thursday, 36% on Friday, and 50% on Saturday. But in 1995 the breakdown was 22%, 36%, 43% (earlier voting because of the more central location?); while in 1996 (with a less central location) there was a shift in the opposite direction, giving 11%, 27%, 62%. Then 1997 had 15%, 33%, 51%; while 1998 (with the voting one day earlier) had 19%, 27%, 54%. The 1999 voting (with about 125 votes hand-carried) was completely atypical.