From their website:
Observation of the Mongolian Party Elections
The Asia Foundation deployed 17 two-person teams to observe the parliamentary election. Other international observation efforts included a total of 26 teams deployed by locally based embassies and international non-governmental organizations, and 9 teams from overseas deployed under the aegis of the Asia-Pacific Democracy Partnership.
This report is now making its rounds in the press. The report consists of two parts, the first on the voting process and the second part on the voter sentiment. After I read the report and saw its pie-charts of voter survey, what I thought was "oh? really?". I am tempted to dismiss the second half of their report as complete nonsense, without blaming the Asia Foundation mind you, and not even close to being suggestive of reality.
Example: Only 12% of the surveyed thought the elections were not free / fair, while 67% said the elections were free / fair. Really?
And this? 70% thought this was not likely?
The main reason I am skeptical is its overly optimistic results.
I could try and explain the reasons for the optimistic results to a certain extent.
First of all, I don't know much about sampling and surveys, but how indicative is a survey taken from 200 voters? Second, who were the interviewers? Foreigners? While Mongolians may speak to the local press frankly and may express their dissatisfaction with the electoral process, they may not do so with foreign observers. It will not do to air our dirty laundry in the international press is the thinking. Most people, non-politicians, are reluctant to be quoted by name criticizing the system. Third, the report states the date as 29 June, the election day. 
I interpret this to mean that the interviews with voters were conducted at voting stations on the election day itself. Which explains why 67% thought "the elections will be fair and free". I certainly hoped it would be fair and free. Hmm... but still, surely we are not that deluded?
But perhaps I am missing the point here. These were simply the sentiments of 200 voters on the election day. What is more important in this report are the Asia Foundation observers' report on the voting process, which indicates to a large extent that the voting process looked fair and free. This does not, of course, rule out the possibility of vote-rigging during the counting process.
Funnily enough, in 14% of the visits, there were officials "loitering near the station". I found the phrase hilarious and perhaps quite fitting, for whatever reason.
In 29% of the visits, the voters were not given enough privacy to mark their ballots without being observed by others.
Now what are we to do with their findings? But more importantly what will the government do with this report? Better voting process including voter privacy and less loitering of officials would be a good start.
View the PDF report from their website here.
The report actually says that seven teams also witnessed the counting process. The question about vote-buying refers to the area the interviewee lives in. Probably there was vote-buying in some areas, but not in others? Finally, this may also have to do with what votzers define as vote-buying? Accepting a gift does not necessarily mean you sell your vote, IMHO.
And the report does not really say "In 29% of the visits, the voters were not given enough privacy to mark their ballots without being observed by others.". To me, it sounds more like "In 29% of the visits, the voters were not given enough privacy to mark their ballots without the potential of being observed by others."
In the report they mentioned that only one team stayed for an entire count at Murun, Khuvsgul. The other teams spent 4~8 hrs at each station.
They do not state where they surveyed most of their sample of 200, or whether they surveyed in all the 200 or so polling stations they visited. But that would make 1 person surveyed at each polling station which would make no sense.
That is true, they would not consider accepting gifts and small amounts of cash during the pre-election campaign hand-out as vote-buying. Most people think they'll take the gift/money and vote however they want or not vote at all. Unless they were paid specifically to turn up and vote for a particular candidate. But who would admit this.
As for the reliability of a survey of just 200 interviewee's, it is not very high. Within the European Union the preferred sample is about 1.000 interviewee's. 2.000 just makes marginal differencies. However, a sample of 200 is not uninteresting, despite the lower credibility. Speaking as a researcher, some data are better than no data. As you rightfully pointed out Bilguun, one could ask the question if it makes any sence to ask someone if they have been cheating in the election. If they have, why would they then admit it?
To my experience from making surveys in Mongolia I think that people leave rather straightforward answers, even to foreigners. The problem is rather a usual one: When asking people in surveys in the US and EU, the surveys tend to be conducted nearby research institutes and universities. People close to such institutions are accustomed to be asked questions and know how to answer them. You sort of educate people how they are supposed to respond to surveys.
In Mongolia this refers to that the election constituencies are very different; if you compare a wealthy neighbourhood, a ger-district or a constituency in the remote countryside. The quality of the answers differ considerably from constituency to constituency.
One problem with this survey is that the connection to exactly where the interviews where made is weak. Therefore it is hard to say anything about how they fit into a greater pattern. Still, it is better than nothing...
A more specific question ("Did candidates in your constituency hand out gifts or money?") might have been more meaningful, plus maybe a "Did this influence your vote". The cherry on the cake(?) would be a breakdown into regions/constituencies. But for statistical validity, this probably would require more than 200 interviewees. And research into these practices was probably not one of the Asia Foundation's priorities.
That is true. It would be interesting if they publish information on where the surveys were conducted, and which social / income-level the surveyed belonged to.
What you said about surveys being conducted on people educated in answering them is intriguing. Does it not mean then that the surveys become a little too self-referential? Subsequent surveys become surveys on how people answer to surveys more than anything else? Just a thought.
I think, if just one independent (I do mean-independent) organization did the survey, it will help tremendously. In real life, just talking to people, you get the feeling that the majority didn't think the elction was free and fair. I personally do not think so, because I personally observed how a governor wouldn't let non-MPRP people into the station. He just chased them away. Mongolians (there were about 4 people) being humble (in the country) and submissive left the election site murmuring. I mean, what "free and fair" eletions are we talking about?
The report is nothing short of laughable, to be honest. As was Asia Foundation head Bill Infante's statement to various international media during the riot that the election was largely free and fair. A handful of so-called observers (and I use the term lightly) - some of whom are woefully ignorant of Mongolian culture - standing around at a few polling stations, interviewing a tiny amount of people makes a mockery of "election observation". The Asia Foundation has completely failed to address the alleged corruption issues pre-election, such as the many false voter registrations.
This report is not worth the paper it is printed on, and shame on the Asia Foundation for making such definitive statements as the election was free and fair. Given the many allegations of corruption, at best the election was questionable, at worst it was completely unfair.
The Asia Foundation has done itself a grace disservice by publishing such rubbish, and people should not put an ounce of stock in it.
But I guess that is what happens when a bunch of people who are so removed from the realities of Mongolian life are put in charge of such a task.
The sad fact is that they are touted as the experts. And they are far, far from it.
As the comment above indicate it is not easy to conduct scientific investigations. It is claimed that Mongolians are culturally different from others. This generally known as the argument of the relativist. In practice it means that a Mongolian do what he or she do simply because they are Mongolians. Consequently, a Japanese do what the Japanese do and the Chinese do what Chinese do, and so on… The problem with this is that in the end does not explain anything.
I have undertaken scientific studies in Mongolia for seven years and one of the most intriguing questions have been: How do Mongolians respond to questions? This is not an easy question because you have to find this out, otherwise you will get answers that does not reflect what you were after. Therefore, it is not enough to say that Mongolians are different, you have to find out in what respect they are different.
Turning back to the election; if you think about it logically, then you realise that if you are about to cheat in an election, then you do it where the risk of exposure is the least, combined where the effect of the fraud is the greatest. This is not easy as these goals tend to contradict each other. But if we think about it Mongolia is a rather peculiar country. It is the 18th greatest in size, while only having 2.6 million people, which means about one person per square kilometre. This means that in most of the country it is extremely difficult to efficiently conduct election observations. Because the population is concentrated to Ulaanbaatar and that the funds of the election observers are limited, most of them will observe the election there. The second choice will be in aimags that are easy to access. In the rest of the country there will be no observations by international observers. Remember that the ideal country to conduct election observations in is Belgium, a small country, and relatively densely populated.
Then we have the time limit. Observers cannot observe for unlimited amounts of time. The rational cheater will concentrate on the pre-election, i.e. the registration procedure, and the post-election, i.e. the counting procedure, the eventual re-counts, re-elections and court proceedings. In both these time-frames the election observers have few if any possibilities to perform their observations.
This means that there are too few international observers and they are not coordinated to observe the election in the most efficient manner.
For instance, in the parliamentary election 2004 there were complaints about people taking two or more voting bills and thereby they cheated. In the following local and regional election the screens were lowered and combined with the design of the voting bill the voter secrecy was severely violated. This was repeated in the Presidential election in 2005. Despite complaints to the General Election Committee (GEC), this was not addressed as a problem. But it was. At that time the GEC was dominated buy the MPRP and the only one that was not from the MPRP was Ms. Radnaa Burmaa. I interviewed her after the local and regional election and she was very disappointed about the lack of treatment of the complaints. The GEC Chairman commented on the violated secrecy of the voters: “- I see no problem here! Of course we have to be able to see to that people do not cheat in the election!”
This problem was addressed by the GEC in 2008 and the screening was adjusted and the design of the voting bills changed, though this was as a consequence of the new election system. But in one respect the problem continued, as the screens were too close, so that the one voting in the closest voting booth could easily see how you voted. Of course, this is not as bad as if the staff at the election facility can see how you vote. But it is still not good!
One last remark: There is no such thing as a perfect election system and perfectly run elections. Even in the old democracies there are frauds, cheating and corruption, but generally this ends up in court and there are undertaken re-elections at those constituencies.