Sunday, January 19, 2014

Should We Believe the Egyptian Poll Results?

The Egyptian authorities claim that their new constitution was approved by 98.5% of those who voted. Even allowing for the effect of some opponents boycotting the election, I don't believe it, any more than I believed similar figures when they came out of elections in the Soviet Union and similar places.

The news stories I have seen seem mostly to take the results for gospel. It's possible that their authors know more than I do about how carefully the voting was monitored by outside observers, but given how easy it is to rig the process at one stage or another when one side is running it, I am skeptical.

Any readers have additional information to add, on either side of the question?


Tibor said...

I think the Soviet (and satellite) "elections" actually did turn out like this. Everyone knew that voting for anyone other than the communists (and you could only choose social democrats who were directly controlled by them anyway...that is in the Czechoslovakia, maybe they did not even have that in the Soviet Union) was pointless and still could possibly mean trouble for you, so if you had the guts to go against the communists you would simply not vote at all (even though, the penalties for that were even harsher).

However, Egypt does not seem to have such a well developed centralized control (but that is just my impression), so I expect the chances are the result is rigged.

Are the regional data from various election offices available? It would be interesting to try to test whether they are in accordance to Benford's law. I think the poll attendance numbers should be and if they are not, it is a strong evidence that the results are rigged. Of course, if you have smart riggers who know about Benford's law, you can pass that test anyway. But I think the chances are that you don't.

Tibor said...

And of course one could (probably should) say that a system where if you don't vote "correctly" you are punished (and it is widely known as a "public secret") is heavily rigged. But the statement "of those who voted 98% choose this and that" is technically true in such a system.

Anonymous said...

One thing I found suspicious was the the percentage were almost identical (98.1 vs 98.0) for votes inside Egypt vs votes from abroad.

(this source actually reports 98.1 for expats but if you do the math it's 98.0)

Anonymous said...

If I were going to rig an election, I'd rig it so that the vote was 78.93% in favor of the outcome I wanted. Nobody is EVER going to believe that 98% of everyone voted in your favor. Heck, you couldn't get 98% of Americans to agree that ice cream is tasty.

Power Child said...

It's helpful to remember that journalism doesn't really mean "information about what's going on" or anything like that. It just refers to a set of affectations, tropes, and mannerisms meant to signal disinterest.

Remember all those news reports about how "Mein Kampf" was a best-seller? Those turned out to be bogus.

So did all those reports about a "racist backlash" over the crowning of an Indian-American Miss America. (The racist backlash they referred to was a couple kids on Twitter. Twitter!)

Speaking of bogus reports about racism, how about the Oberlin white bedsheet incident?

Journalists are both gullible and lazy. As I've said before, the only peculiar skill of the journalist is his willingness to record the words of strangers. I should have added, willingness does not necessarily include gumption or critical thinking.

suckmydictum said...

Any statistic that has 98% of people agreeing on anything should be treated skeptically.

Tibor said...

Power Child:

Well, what distinguishes a good and a bad journalist is that the good one is able to do more than to be a voice recorder.

If investigative journalism is done properly, it can be really nice read.

Today, journalism is fast and everyone tries to report on everything ASAP (which reminds me of this wonderful Onion News parody which is however very accurate: ). That leads to a lot of nonsense which is sometimes later refuted and sometimes is not.

Also, you want to get a lot of clicks on your articles. So you post about anything that sounds unusual (such as Mein Kampf being a bestseller) without thinking about it or sometimes even with the knowledge of that being bogus.

However, I read some columns by Karel Čapek (a famous Czech journalist from the 1920s and 1930s and among other things the author of the word robot) who basically made comments in the sense that most journalists are looking for cheap sensationalism and do not check their information. I am not exactly sure, but I think Mencken made similar comments somewhere also. Which leads me to suspect that the problem is not as much the "rush of the internet age" but simply the fact that some journalists do their jobs properly and some are "in just for the clicks". However, internet made publishing cheaper, so probably also the quantity of bogus has increased...which does not mean there are no good journalists out there. I really liked one David's argument for unschooling (I consider it the strongest argument for that) which is that unlike the regular school (where you are presented and examined from the words of "wisdom" from an authority figure) it teaches you to filter information. And while before the invention of the printing press that was not a very important skill, finding out some information today place is often no longer a problem, the problem is to distinguish the true information from the false. I believe that mathematical or statistical way of thinking helps here even above everything else...but that may be just because that is my natural way to work (which does not mean that it never fails, especially when there are strong emotions associated as well).

Also, this article deals with this (bad and good journalism and the internet) and I think is quite interesting:

Power Child said...

@Tibor Mach:

I am not so charitable. I don't think there used to be some golden age of journalism back when the pace was slower and the tone was calmer. News back then was, if anything, even more stilted and biased. Look at any newspaper from 100 years ago for evidence of this. My favorite is reporting on "marihuana" from the 1920s, which today would make tabloid editors blush.

1. Objectivity is impossible.

1a. Even if a journalist tries to cover both sides, he has inserted bias by ignoring a third side.

1b. Even if a journalist can cover all the sides anyone knows about, there will be sides that people have not thought of, so the final picture will still be incomplete.

1c. Because the audience is human, and because journalists are human, human interest must be the focus of any story even if it can only subjectively be considered the most important aspect. (Google "onion autistic reporter train" and play the first video for an excellent demonstration of this problem.)

1d. The reporter typically relies entirely upon other sources which are themselves biased for all the reasons above. Even if the reporter was immediately involved in the story, he will not have perceived everything that happened and will have to rely upon other sources to fill in the picture.

2. Reporters are mostly just English majors with no real expertise in anything.

3. Reporters generally, and investigative reporters in particular, have preexisting biases they bring to the table.

4. Journalism does a worse job on covering many topics than individual writers, such as bloggers. For example, at a convention of intelligence researchers, Steve Sailer was ranked way above just about every other source for intelligence reporting, with a score of about 7.5 out of 10. The next best (with a score of about 5) was some other blog, and regular newspapers didn't even factor into the list until a few more positions downward, with scores starting around 4 and getting worse from there.

5. The tone of journalism tricks people into believing qualitative things about its content. A more muted, hushed tone = higher quality, less biased content. Thus why people believe (falsely) that BBC or NPR are less biased than MSNBC or Fox News.

6. Read/watch/listen to the news. Why is it delivered in that strange trope? Why that strange way of writing? Why this standard setup of someone at a desk with a bunch of video screens behind him? These strange but ever-present patterns signal the existence of a highly stylized posture. It is the posture of objectivity. Objectivity is impossible (see #1) so the posture of objectivity is therefore a deliberate obfuscation.

7. Journalism is, at its essence, nothing but this posture.

8. Journalism is nothing but the studied art of lying.

Tibor said...

Power Child:

Of course the journalists are biased. As is anyone else. I like to say that the only unbiased people are those who don't care. But that does not mean everyone is just a cheerleader for their team.

I don't think you need much formal university education to be a good journalist (or, well...good anything). The fact that someone is an English major, or journalism major or has not studied anything at all does not mean he cannot be good.

I agree that most journalists are not good. At the same time, that is probably true of most programmers for example, or most of anyone or anything.

It is hard for me to define exactly what good journalism is supposed to be. I would say one is the watergate story. Another is an article I read few months back which was about the practice of policemen who pretty much mugged people on the grounds that they could confiscate property of criminals. It was about 20 pages long, it consisted of the reporter talking to a lot of people of both sides of the story and covering quite a wide area (geographically). It was stylistically well written and so it was a good read. It was probably is anything. But a biased estimate is not necessarily a bad one, especially if you know the side of the can couple it with one which is biased in the opposite direction and get a fairly balanced opinion ... the only problem is you are biased as well, so you can't, but you can try to keep your bias in limits at least...which is also what a good journalist should do.

Oh, and Michael Falk of the ONN? He is awesome, more reporters like him please :D
However, I'm not sure why any story that is not full of "heartwarming" clichés needs to be autistic, so I don't think it is a good example.

As far as TV news go, I don't watch that very much (with the exception of ONN :) ). I agree that the way TV reporters talk is ridiculous.

Anonymous said...

Surely somewhere someone has shown that these types of election results are statistically impossible