[Ecommerce] Europeans-of-the-Year Poll Results Beggar Belief

[Seth Johnson]

(News from the information freedom arena.  Florian Mueller is a
fellow who got a lot of visibility in the fight to keep software
patents illegal in Europe.  Signs would seem to indicate that he
should have won the European of the Year title in the EV50 poll.
He instead was given the "Campaigner of the Year" award, beating
out U2's Bono.  -- Seth)


> http://slashdot.org/~FlorianMueller/journal


Result of Europeans-of-the-Year Poll Fails Plausibility Test

Tuesday December 13, @01:55PM


If Microsoft sponsors a series of political awards and two of the
ten prizes go to vocal opponents of software patents, then one
would usually think that it's an amazingly positive outcome.
That's just what happened with this year's edition of the EV50
Europeans of the Year
(http://www.ev50.com/noflash/lastyear.asp?id=3D7). Michel Rocard,
one of our key allies among the members of the European
Parliament (MEPs), became MEP of the Year. I received the
Campaigner of the Year award for my NoSoftwarePatents.com
campaign (http://www.nosoftwarepatents.com/). Basically, I had
been nominated as a representative of an entire movement that was
formed around the FFII (http://www.ffii.org/). The list of
previous winners shows that "my" award went to Pope John Paul II
in 2002, and to the then-president of the European Parliament in
2003 (http://www.ev50.com/noflash/charities.asp?id=3D5). Without a
doubt, software patent critics have a higher profile than ever.

However, what the organizers proclaimed as the result of the poll
is extremely difficult to comprehend. If a computer program were
fed with all of the information that is available from the
outside and checked an alleged result against that background,
then the plausibility test would either reject the entry
altogether or would only accept it after asking "are you sure?",
"are you still sure?", and "are you really sure that you're still
sure?".

That's why I returned the award trophy to the publisher of the
European Voice (http://www.european-voice.com/), a sister
publication of the vaunted Economist (http://www.economist.com/),
after the awards ceremony. It was a gesture of protest against an
election process that in my opinion was not fair. Since the
European Voice didn't announce anything concerning the poll other
than the "winners" (not even the total number of votes), I was
worried about what happened to tens of thousands of votes that
came from our community.

There are strong indications that our audience represented a
solid majority of all people who participated in that EV50 poll.
The European Voice newspaper, which organized and announced the
vote, only has a circulation of about 16,000 copies
(http://www.european-voice.com/downloads/EV_BPA_Sept2005.pdf), a
large part of which is distributed for free to the EU
institutions in Brussels. Most of the 50 nominees don't actively
campaign for votes, and none ever ran an EV50 campaign like ours.

For the first five weeks, voters received confirmation links that
contained an apparently sequential ID. We can't prove that it was
sequential, but during that period the number always went up,
never down, and after about two weeks after the start of the
poll, it was slightly above 19,000 and only moving very slowly,
on some days by less than 50. We asked different people to vote
at different times, and to then tell us the IDs they got.

Among those first 19,000 IDs, there must already have been some
number of software patent critics, as several IT websites
reported on my nomination on the day the poll started (22
September) and on the following days. However, we only got
started with official calls on our supporter base to vote when we
had a complete set of voting recommendations online in a dozen
languages, which was the case on 14 October (22 days into the
voting period). The confirmation ID was still below 20,000, and
it then surpassed 38,000 within less than two weeks. We were able
to see, on a daily basis, the impact of our mailings on our web
traffic and, quite proportionally, on the uptake of that
confirmation ID.

The organizers then switched to an encrypted ID, and at that
stage, we had some of our biggest promotions yet to come: a news
item here on Slashdot
(http://yro.slashdot.org/article.pl?sid=3D05/10/31/1212234&tid=3D155)
and mailings by the FFII to about 90,000 registered supporters.
Our web traffic stayed pretty high until the end of the voting
period on November 11, and even a few days beyond.

On the first day of the poll, people reported in Internet forums
that the voting server reported too many concurrent database
connections, and that happened pretty immediately after the
mailing of ThankPoland.info to more than 30,000 supporters
(http://thankpoland-info/), after some of my mailings, and
especially after a Slashdot report on our electoral campaign.

People in Brussels know about the tremendous numbers of people
who we can mobilize via the Internet. At a hearing of the
conservative European People's Party, a parliamentarian claimed
that he and his German colleagues received 75,000 emails within
only two weeks as part of an email campaign
(http://wiki.ffii.org/Epp050602En). On other occasions,
politicians mentioned similar numbers in the tens of thousands of
emails that they claimed to have personally received from
software patent critics.

In theory, it's still possible that our supporters didn't
represent a majority of the electorate. A confirmation ID doesn't
mean that the vote got counted, and maybe our computer-savvy
audience had a hugely higher rate of people who failed to click
on that link (despite the instructions we provided on
NoSoftwarePatents.com) than the EU folks who read the European
Voice. Then, the uptake of the ID could, by pure coincidence,
have been caused by other factors in addition to our own
activities. Those other promotions would have taken place in
parallel to whatever we did, and they would never have been
spotted by Google, according to which (dependent upon the keyword
and language one looked up) 75% to 99% of all EV50-related
Internet stories reported on our campaign. Of course, there are
other ways to reach voters than via the Internet, but it's quite
difficult to address a really large audience by other means
without someone writing about it on some website. But no matter
how unlikely, it is a possibility.

The publisher and the editor of the European Voice told me, by
email and at the EV50 event, that there were "many thousands" and
then even "dozens of thousands" of paper votes. It's true that
the European Voice had a coupon in several issues during the
voting period, and I don't doubt that among the readership of the
European Voice, we will have had very few votes. However, out of
a circulation of about 16,000, a number of people will not have
voted at all, and of those who wanted to vote, only a certain
percentage will have used the paper ballot instead of the
Internet, to which almost everyone in Brussels has access at his
or her workplace. But then, those assumptions may be wrong, and
the readership of the European Voice may be very special.

What surprised me most when the EV50 results were presented is
the combination of awards that our recommended candidates won and
those they didn't. I wonder how Dalia Grybauskaite would have
become Commissioner of the Year if not because of our votes. We
recommended her because she was least suspected of ties with the
pro-patent lobby, but of the five commissioners among which
people had to choose, she was definitely the underdog in the
election. Still she managed to beat G=FCnter Verheugen, a vice
president of the Commission who had won that award the previous
year and is considered the superstar of the present Commission.

It beats me how I could have won over Bono, the frontman of U2,
in a category of six candidates
(http://www.ev50.com/noflash/category.asp?id=3D6), while allegedly
having lost to Luxembourg's prime minister Jean-Claude Juncker in
a field of fifty for the main award
(http://www.ev50.com/noflash/european.asp?id=3D9). Bono received
votes from the fans of his music (U2.com and some independent U2
sites called on people to vote), plus from the supporters of his
anti-poverty campaign, and on top of all of that, everyone had to
make one choice per category
(http://www.ev50.com/noflash/votingrules.asp?id=3D12), so a number
of people will surely have picked him because he was the first
and most famous candidate in the category. Therefore, beating
Bono in the Campaigner category was one of the highest hurdles in
the entire election, if not the highest.

In contrast, it must have been much harder for a candidate like
Mr. Juncker to gather a huge number of votes in a field of 50
candidates. He is certainly a well-respected politician in the
EU, and he would be a logical choice for a jury to make because
he defended the proposed EU Constitution when others already
considered it dead (after the referenda in France and the
Netherlands). I talked to the president of LiLux, a free software
organization in Luxembourg
(http://www.lilux.lu/lilux/front_content.php?idlang=3D1), and he
had no explanation how Juncker could win a popular vote against
us because he and the other members of his organization didn't
notice any promotional activity for Juncker. That tiny country
only has 470,000 inhabitants, i.e. about 0.1% of the total
population size of the European Union
(http://www.cia.gov/cia/publications/factbook/geos/lu.html#People).

I know that some people only voted me for me in the Campaigner
category, but thought that someone else should receive the main
award. For instance, some voted for Michel Rocard, the other
anti-software-patent nominee. However, the emails I received
indicated that most of those who we asked to vote did follow our
voting recommendations, at least in the fields in which there
were candidates who were known for their anti-software-patent
stance.

All in all, one would be hard-pressed to find a plausible
explanation for the outcome of the EV50 vote. There are some
theoretical answers, but they aren't too likely. The organizers
decided that total intransparency is in their best interest, so
we're left in the dark as to what exactly happened. If this were
an official election and not just a poll by a relatively small
(but politically important) newspaper, then there would be ways
of finding out. In this case, we can have our opinions, but we'll
never be able to prove what the outcome should have been.

In the event that an independent evaluation of the poll would
have shown that I won the main award (possibly with more than ten
times as many votes as the runner-up), then according to the
rules someone else (probably Bono) could have won the Campaigner
category.

Since the European Voice continued to list me as the Campaigner
of the Year, I decided to accept that award anyway, and I've
meanwhile asked them to donate the prize money to the FFII on my
behalf. Most of our voters only spent a few minutes to fill out
the form, and the two category awards for our camp are a
significant step ahead that has been worth that kind of effort. I
also think that this jury, which picked the 50 nominees, seems to
have been reasonably independent
(http://www.ev50.com/noflash/advisors.asp?id=3D8). However, I can't
say that I'm going to be too enthusiastic about such polls in the
future, unless they meet a certain standard of security, and are
performed by trusted third parties that are only in the business
of organizing such polls.