Ethan Steinberg writes:
I thought you might find this funny given your past blog posts about related subjects (Iran & Benford’s law, preregistration, etc).
Announcing that an election is fraudulent due to Benford’s law on data that doesn’t even exist yet seems like the perfect encapsulation of crazy research.
The link points to this news article, “As Newsom leads California recall polls, Larry Elder pushes baseless fraud claims,” which reports:
Republican Larry Elder appealed on Monday to his supporters to use an online form to report fraud, which claimed it had “detected fraud” in the “results” of the California recall election “resulting in Governor Gavin Newsom being reinstated as governor.”
The only problem: On Monday when the link was live on Elder’s campaign site, the election hadn’t even happened yet. . . .
“Statistical analyses used to detect fraud in elections held in 3rd-world nations (such as Russia, Venezuela, and Iran) have detected fraud in California resulting in Governor Gavin Newsom being reinstated as governor,” the site reads. “The primary analytical tool used was Benford’s Law and can be readily reproduced.”
The site added on Monday afternoon a disclaimer saying it was “Paid For By Larry Elder Ballot Measure Committee Recall Newsom Committee,” with major funding from Elder’s gubernatorial campaign. . . .
Elder’s campaign homepage under the title “stop fraud” links to the website, StopCAFraud.com. The site solicits donations for him and asks supporters to sign a petition “demanding a special session of the California legislature to investigate and ameliorate the twisted results of this 2021 Recall Election.”
The site seems to presuppose the outcome of the race, claiming fraud has already been detected in the election and that Newsom won — even though Election Day is not until Tuesday.
The page suggests voters may turn to the “ammo box” if they can’t trust the ballot box. . . .
Kinda funny, also kinda scary.
Here’s the webpage linked from Elder’s campaign site:
Ya gotta love the legalese at the bottom of the page. They’re willing to brazenly lie about hypothetical fraud and make up documentation on votes that haven’t yet been tallied, and they’re threatening to shoot you if you don’t accept their fabricated claims—but they still have time for the legal mumbo-jumbo.
But I didn’t see anything there about Benford’s Law. So I went a few days back in time on the Internet Archive and found it:
I guess they didn’t run the text by Putin. Calling Russia a “3rd-world nation” . . . that’s so insulting!
I eagerly await their Benford’s Law analysis. Funny that it “can be readily reproduced” even though it hasn’t been produced the first time. Kind of a paradox of reproducibility.
P.S. Elder has some endorsements:
David Mamet!
Hey . . . wasn’t there a Benford’s Law scheme in The Spanish Prisoner? The sequel will be awesome. Eastwood directing, Mamet with the script, and Norris can still do his own stunts, no? I don’t know if they’ll be able to get Alec Baldwin on board, though; I’ve heard he’s a Democrat.
Hmm, and most sequels do have relatively low final digits. If they next make The Spanish Prisoner VII, we’ll know something’s up.
John:
Now that you say this, I’m thinking there might be something fraudulent involved with the Leonard series.
Not to mention the Surf series
https://en.wikipedia.org/wiki/Surf_II
Thiotimoline statistics
and with all due respect to Norris’ athletic accomplishments, when you’re over 80, you really should go with the stunt double.
Or do age-appropriate stunts.
He could reprise his role as “Walker, Texas Ranger” and do stunts with, you know, like a real walker …
Now I want to see Mamet do some old-guy shtick! I’m picturing the entire movie. They could bring in Mary Rosh as a statistical consultant to demonstrate Benford’s law.
As horrendous as this story is, I don’t think it is being reported accurately. It is not true that there were no results at the time of Elder’s announcement upon which statistical analysis could be done. On Sept 13, early data on ballots (mail-in) returned were released – they shows much higher responses from the Democratic areas than the Republican ones – I know, a real shocker. But the point is that there were some numbers that could be analyzed.
This in no way means that Benford’s law would tell us anything about the election at that point in time. Nor does it make it reasonable to declare that Newsom had been reinstated prior to any votes being counted. Nor does it excuse the word “reinstated” for someone who never left office. So, the abuses are severe and very troubling. But I do think the media overplayed the story a bit by suggesting that the detection of fraud was based on nonexistent data.
I’ve seen this as an ongoing troubling pattern, where the media overplays every possible incident. And it has been happening by all sides. Much as I detest the Republican ongoing attempts to subvert democratic processes and cast unsubstantiated doubt on our elections, I worry about using that to justify overstating the facts. Explaining why Benford’s Law would not tell us about fraud in this case is much more complicated than simply saying that there were no results to analyze. But let’s not continue to dumb down the story because people are not smart enough to understand complicated reality.
Dale:
I disagree.
When writing the post, I thought about this objection you raised, but (a) it’s not clear how Benford’s law would be applied to partial votes, (b) no Benford analysis was ever posted or linked to from that site, and (c) after the news story came out, the entire thing about evidence of fraud and Benford’s law was scrubbed from the site.
This all suggests strongly to me that there was never any data analysis using Benford’s law or anything else; that is, the claim (since scrubbed) on that website was a lie. I guess you’re right that it’s theoretically possible that there was a data analysis and it was just never posted anywhere and then the mention of it was removed, but I doubt it. Given what we’ve seen so far, I don’t think it’s “dumbing down the story” to conjecture that these people made up a claim of fraud based on zero evidence.
I don’t disagree with anything you say. My point is that reporting it as they claimed fraud before there was any data is not quite accurate. There are plenty of reasons to bash what Elder did, and I have no issues with that. My issue is with the media portrayal which was geared to oversimplify and misrepresent because it was a better story (in their mind). They did not misrepresent that Elder did horrible things – they overstated what those horrible things were. We don’t know what analysis might have been done and we strongly suspect that there was none. That would be fine to say – but declaring that there was no analysis because no votes had been counted yet was an appeal to a gut level “understanding” rather than your more complex explanation.
Quote from Larry’s website: “have detected fraud in California resulting in Governor Gavin Newsom being reinstated as governor”
I’m not sure how you could talk about Gavin Newsom being reinstated before the election results were out.
Re: Elder and reality about the election – https://twitter.com/brianstelter/status/1438136944492429315
But the best thing that happened to Newsom in the run-up to the recall vote was that Elder became the GOP front-runner, and was given a lot of air time. The more he talked, the higher Newsom’s numbers. You personally may agree or disagree with Elder, but most of his positions are not ones that resonate with Ca. voters, look at the election results (in fact they probably scared people into being certain to vote).
And speaking of “accurate” political takes -https://twitter.com/NateSilver538/status/1429865259804790784
Roy:
Regarding your second link: That’s funny. Nate’s a poker player, you’d think he’d understand some game theory.
Yup. And speaking of accurate political takes:
https://twitter.com/moonbeamvotes/status/1437983412380332032
And quotes that should have received more attention:
https://twitter.com/BruceBartlett/status/1438191515717750792
Also a lot of discussion of whether Elder outran the polls. He did if you only consider the votes on part 2, he didn’t if you consider all the votes, since almost 50% of the voters didn’t vote on the second question. It’s that old base-line question, you have to use the proper denominator.
Roy,
OK, I wrote my post on Nate’s take yesterday and scheduled it for tomorrow, but if everyone’s talking about it already maybe I should just post it now…