[Chapter-delegates] ooXML - Corrupt countries more likely to support the OOXML document format

[Javier SOLA]

Numbers are always fun...

-------- Forwarded Message --------
From: Ville Oksanen <ville.oksanen at effi.org>
To: ipr list <ipr at mailhost.soros.org>
Subject: EFFI: Corrupt countries were more likely to support the OOXML
document format
Date: Wed, 05 Sep 2007 10:11:59 +0300

We did little research on the topic - of course association does not 
imply causation..:
http://www.effi.org/blog/kai-2007-09-05.en.html

Corrupt countries were more likely to support the OOXML document format

International Organization for Standardization (ISO) rejected in a vote 
on 2 September 2007 the controversial Microsoft-supported OOXML document 
format.

During the voting process, the reputation of ISO as a trustworthy 
technical standards body was called into question. For example, in 
Sweden a Microsoft representative was caught offering to recompense 
partners for voting yes to OOXML. There have also been questions about 
sudden and surprising interest of countries like Ivory Coast in document 
formats.

We studied the relation between the corruption level and voting 
behaviours of the countries. We found that more corrupted the country 
is, the more likely it was to vote for the unreserved acceptance of the 
OOXML standard proposal.

We used the 2006 CPI index (Corruption Perceptions Index) as a measure 
of corruption. CPI index is a number between 1 and 10. A small CPI index 
means that the country is perceived to be very corrupted, while a large 
CPI index means that the country is perceived to have little corruption. 
Haiti has the smallest CPI index of 1.8, while the countries perceived 
having the least amount of corruption (Finland, Iceland, New Zealand) 
have a CPI index of 9.7. In barplots below the CPI index has been 
rounded down to the closest integer value.

ISO received a total of 87 votes, of which 70 was given either by the 
Secretariat country (USA), participating members ("P Member") or 
observers ("O Member"). In various stages of the voting only these 70 
votes are taken into account, according to the complicated voting rules 
of ISO, and therefore in the graphs below only these 70 votes are shown. 
The remaining 17 countries - which are perceived to be relatively 
corrupted (median CPI index 3.0) - mostly supported the OOXML (approval 
13, approval with comments 2, abstention 0, disapproval 2).

Is this just a random coincidence? The median of the CPI index of the 
above mentioned 70 countries is 3.95. Of the most corrupted half (CPI 
index less than 3.95) 23 or 77% voted for approval (approval or approval 
with comments) and 7 or 23% for disapproval; 5 abstained. Of the least 
corrupted half (CPI index more than 3.95) 13 or 54% voted for approval 
and 11 or 46% voted for disapproval; 11 abstained - see the table below. 
This statistics supports with P value 0.07328 the hypothesis that the 
corrupted countries were more likely to vote for approval (one-tailed 
Fisher's Exact test). In other words, simplified a bit: the likelihood 
that there was no positive correlation between the corruption level and 
probability of an approval vote, that is, this is just a random effect, 
is about 7%. (The statistical test naturally does not tell anything 
about the reasons of the connection between the corruption level and the 
probability an approval vote.)