Read on Twitter
1/ U Mich Prof Walter Mebane did a statistical analysis of S Korean's Nat'l Ass'y election & finds that fraud may have changed some results.
http://www-personal.umich.edu/~wmebane/Korea2020.pdf
Friends have asked me what I think. I don't know enough to believe it, but I also know too much to dismiss it.
http://www-personal.umich.edu/~wmebane/Korea2020.pdf
Friends have asked me what I think. I don't know enough to believe it, but I also know too much to dismiss it.
2/ The paper is only 14 pages long. Mebane has done statistical analyses of other elections, finding evidence of fraud in some cases & not in others.
https://www.isr.umich.edu/cps/people_faculty_wmebane.html
https://www.isr.umich.edu/cps/people_faculty_wmebane.html
3/ Based solely on his statistical modeling, Mebane concludes that 10.43% of the ruling Democratic Party votes were fraudulent. That's ... a lot.
He also concedes the possibility of benign explanations, & the limits of relying on statistical analysis alone.
He also concedes the possibility of benign explanations, & the limits of relying on statistical analysis alone.
4/ I'm the first to admit that I don't understand statistics well enough to understand Mebane's methodology or conclusions. Here's video of Mebane explaining it.
He seems qualified, credible & sane, but I'd appreciate other views from statisticians.
He seems qualified, credible & sane, but I'd appreciate other views from statisticians.
5/ One factor Mebane doesn't consider in his analysis is the coronavirus pandemic, which might have skewed turnout.
Only, I'd think a pandemic would have *depressed* turnout. Yet it was the highest in 28 years, & that is very weird. https://www.cnn.com/2020/04/15/asia/south-korea-election-intl-hnk/index.html
Only, I'd think a pandemic would have *depressed* turnout. Yet it was the highest in 28 years, & that is very weird. https://www.cnn.com/2020/04/15/asia/south-korea-election-intl-hnk/index.html
6/ I'd also expect opposition voters to be *more* energized to turn out in a non-presidential election, especially with the economy in a freefall before COVID-19 hit.
Polls last year showed that the ruling party had lost much support from young voters who supported it in 2017.
Polls last year showed that the ruling party had lost much support from young voters who supported it in 2017.
7/ What could legitimately explain this sudden shift?
By the time of the election, an effective response to the pandemic had overcome a lot of bad press from the govt's failure to prevent it in the 1st place.
It's still hard to believe this was such a strong net positive.
By the time of the election, an effective response to the pandemic had overcome a lot of bad press from the govt's failure to prevent it in the 1st place.
It's still hard to believe this was such a strong net positive.
8/ What observed behavior supports allegations of fraud? I've seen claims of ballot tampering, but not enough independent analysis to draw conclusions. https://twitter.com/AuthorJCha/status/1259345488521748480?s=20
9/ Ideally, any persuasive analysis should be informed by an understanding of statistics, polling, voting demographics, & S Korean voting procedures.
Extraordinary claims require extraordinary evidence.
Extraordinary claims require extraordinary evidence.
10/ One known fact that causes me not to dismiss the allegations out of hand is that Blue House officials & ruling party pols illegally manipulated the Ulsan mayoral election, covered it up, & totally got away with it. https://freekorea.us/2020/01/09/moon-jae-ins-wednesday-night-massacre-threatens-the-rule-of-law-in-korea/comment-page-1/
11/ That not only disentitles the Moon admin to a presumption of good faith, it undermines my confidence that journalists would investigate credible allegations.
Not knowing if elections are fair, gov't is honest, or the press is objective is a bad place for any democracy to be.
Not knowing if elections are fair, gov't is honest, or the press is objective is a bad place for any democracy to be.
