Based on results of the recent elections in the US, I thought I would
try some comparisons with our own elections, to get a better idea of the
effective size of the Wikimedia electorate. What I did was to take the
elections results for cities in my general area, since I can find those
very quickly. I chose those cities where the number of ballots cast in a
contested local race (usually for city council) was closest to the
number of ballots cast in two of our elections (the Board of Trustees
election earlier this year, and the Arbitration Committee election last
December on the English Wikipedia). I figured that the population of
these cities would be a reasonable estimate of the effective size of our
community. Naturally, this is an impressionistic rather than a rigorous
statistical analysis.
At the time of the Arbitration Committee election, the number of ballots
cast on the English Wikipedia was about the same as several cities with
populations ranging roughly from 3,000-6,000 people. The upper end of
this range is quite close to the number of people who edited five times
or more last December, according to Erik Zachte's statistical reports.
So it seems this may be the number that is closest to a real-world
population for our community. Since those statistics have just been
updated (thanks, Erik) we can see that the English Wikipedia is now
nearly equivalent to a town of 15,000 people.
The turnout for the Board of Trustees election was comparable to a
real-world population of about 7,000-11,000 people.
It's a little harder to find an equivalent number in our ongoing
statistics, in part because project-wide participation is difficult to
capture (<cough>single login</cough>). Even the number of users with
five or more edits that month on just the Wikipedias collectively was
significantly larger than this.
--Michael Snow