Here is part of the Radio Liberty article:

Магчыма, у апытаньні “Лявада-цэнтру” траціна апытаных адказала, што прагаласавала за Лукашэнку, 20% – што прагаласавалі за Мілінкевіча, 30%, як паведаміў спадар Гражданкін, адмовіліся адказваць, які выбар яны зрабілі. Калі гэтыя 30% адкінуць, то 33% за Лукашэнку як раз і ператвараюцца ў 47%, а 20% за Мілінкевіча – у 30%. Але гэтае адкіданьне сапраўды некарэктнае – гэтыя 30%, што адмовіліся адказваць – гэта “чорная скрыня”, калі усе яны былі за Лукашэнку, то ён атрымаў за 60%, калі скажам, усе за Мілінкевіча, то ён атрымаў 50%, Але як насамрэч разьмеркаваліся галасы, хто адмовіўся адказваць, невядома. Менавіта пра гэтую няпэўнасьць, відаць, і казаў намесьнік кіраўніка “Лявада-цэнтру”.

Source: http://www.svaboda.org/articlesfeatures/politics/2006/3/888C6F0B-49D1-4C55-A46F-48D3880E4BDF.html

It seems to me that Yuri Drakakhrust, who wrote the article, may be thinking along the same lines as I am.

As for the embassy voting in 2004, we tried to get the official figures from Minsk and not from the embassy. The latter distributed a result (which corresponded well enough to our polling result) to an emigré Russian website in Sweden, but there was no way of checking whether those same figures were reported to or used in Minsk.

/Tobias

> How did you get the official results from the embassies?

I don’t know. Some of my friends took part in this project, but I don’t know the details.

I have a file with exit polls results from 5-6 locations (Brussels, NYC, DC, Warsaw, Prague,…).

Actually, in most embassies they refused to reveal the data (I wonder if they were breaking the law by hiding the results?)

> Although my Belarusian is a bit weak, I gathered that 30 percent of respondents had refused to answer. 20 percent said they voted for Milinkevich and 33 percent for Lukashenko.

You got it a bit wrong.

30 percent refused to answer. But among those who AGREED to answer, the results were:

47% - Lukashenka
30% - Milinkevich

On Radio Liberty’s website I read an interesting article about the Levada Centre’s telephone poll conducted on election day in Belarus. Although my Belarusian is a bit weak, I gathered that 30 percent of respondents had refused to answer. 20 percent said they voted for Milinkevich and 33 percent for Lukashenko.

Do you trust these figures, and would you agree that it is plausible that most people who refused to answer voted for a candidate other than Lukashenko? If so, the real results could be rather even between Milinkevich and Lukashenko, or at least betwen Milinkevich plus Kozulin on one hand and Lukashenko on the other hand.

Regards,
Tobias Ljungvall

PS. Thanks for that document you sent me.

Official numbers:
Hajdukevicz – 2
Kazulin – 8
Lukaszenka – 158
Milinkevicz - 53
Against all – 7
Invalid bulletins - 7

(data taken from http://wolny.livejournal.com/228243.html )