Use the same method I did with "Brett." Why? Because I am consistent. That method doesn't always lead to the same result, but Bayesian methods are the right methods.
First, we begin with the observation that most accusations are true. This doesn't mean we convict on the basis of an accusation, but this isn't a criminal trial. This is a job interview. Same thing. An election is a job interview. With an incumbent, it is a contract renewal decision. I actually wrote a whole book about why this is the right way to think about elections. This was the point during the Kavanaugh hearings, and it's the point now. Most accusations are true. How does that factor into the contract renewal process?
The Bayesian prior about Brown's guilt, then, should be the baseline probability that any given assault allegation is true. That's gonna be high. Very high.
Can we update that probability? If you are a lefty, you want to update that probability downwards because Sherrod Brown is a loyal lefty, and blah, fucking blah.
No! Bad liberal! No kale!
I shouldn't have to remind you of all of the lefty hero shitbags who have turned out to be guilty, but the basic point is that you don't get to assume that your side wouldn't do it. Humanity is basically vile. Even the people on your side. (Welcome to The Unmutual Political Blog-- home of scholarly misanthropy on the web!) You also can't tell from people's personality or demeanor what they are like behind closed doors. As I have written in many contexts, the worst monsters are the ones who act politely. They trick you into thinking you are safe around them. Don't assume you can tell, the way all of Cosby's victims, and hell, most of America did.
So, Brown's politics, or frumpy demeanor or any of that? No, you don't get to update that prior based on any of that.
We know nothing about the source of the allegation, so we can't evaluate her credibility, which means we can't update the probability upwards or downwards based on her. So, we're still stuck at the baseline probability that any given allegation is true, right?
Not... quite...
There's still the little matter of Jim Renacci. Right now, the allegation is coming from Jim Renacci, effectively, not the woman herself. There are two elements here. Renacci's personal credibility, and the timing. Does Jim Renacci have personal credibility as an advocate for the rights of women who have been sexually assaulted, and does the timing tell us anything? The questions are related.
Where did Renacci stand on the Kavanaugh/Ford matter? Like all Republicans not named "Lisa Murkowski," he was a good, little Trumpkin who treated Christine Blasey Ford like she was either delusional or a liar because obviously a person whom the conservative movement/Republican Party had incentives to defend must be treated as innocent regardless of the facts. Kavanaugh blatantly lied about so many things that it looked like he was going for a Trump Award in the competitive sport of land-speed lying. Renacci didn't care. Neither did anyone in the GOP whose name wasn't "Lisa Murkwoski." They were more interested in smearing Ford for partisan and ideological reasons. Did Renacci do anything to earn any credibility here? Nope.
And without the woman making the accusations against Brown, we are left with Renacci's word, and Renacci's credibility rather than simply the baseline probability of an accusation being true. How often are sexual assault allegations false? Very, very rarely. That's not really the question, though. The real question is, how often do people like Jim Renacci lie?
All the fucking time. He has no record of taking sexual assault allegations seriously. The attorney, Laura Mills, does have a record handling sexual assault, but so did Rachel Mitchell, and that sleaze just oozed herself onto the chamber floor with no intention of doing a good faith job, then wrote up a hatchet-job document bashing Ford without even mentioning Kavanaugh or the many lies he told in his testimony. Mills is working for Renacci. So, does her record mean anything? Not when you factor in the observation that she is working for Renacci. See: Mitchell, Rachel. Once you knew that she was hired by Grassley, you knew that she was Kavanaugh's defense attorney, not a prosecutor, so her claims to being a prosecutor and any associated credibility went out the door. Ignore Mills. We have no accuser to evaluate. We just have Renacci.
The basic point here is the following question: would the party of birtherism and pizzagate make up a bullshit allegation against a Democrat, and run with it in an election? Birtherism. Pizzagate. This is the party of Donald Trump. They'll tell any lie, and we're not looking at the woman making the allegation. We just have Renacci. Yeah, he'd lie.
The complication here is that most women don't come forward. Why not? Look what just happened to Christine Blasey Ford. Suppose Brown did it. Would it be rational to come forward?
Um... there is a difference. The Democrats are more likely to clean house when there is a credible accusation within their ranks. Al Franken and John Conyers come to mind. Pulling the shit that the GOP did on Ford? Not as likely when the allegation is made against a Democrat. That doesn't mean the woman would be acting rationally to come forward, but this does complicate the issue. We cannot rationally expect a woman assaulted by Brown to come forward, if it did happen. Yes, lefties, it is a real possibility, and the fact that Renacci is a Trumpkin doesn't mean you treat every word out of his mouth as definitively false.
Still, from a Bayesian perspective, Renacci's lack of credible record on the issue, the constant lying from the GOP in the Trump era, and the fact that the allegation rests, not on the credibility of the woman making the allegation, but on Jim Renacci, means we must update the probability of Brown's guilt downwards.
And we're not done. Does the timing look odd to you? Like, right after that phrase, "what goes around, comes around," from the Kavanaugh hearings?
Trump admitted that he didn't care whether or not Kavanaugh did it, and plenty of Republicans in surveys admitted that they didn't care because... Republicans don't really care about rape. I have written about this before, and it is worth stating again. Rape is a partisan issue, not a gender issue. However, in order to tell themselves that it was morally OK to vote for Kavanaugh, Republican Senators told themselves that Kavanaugh must be innocent, and that it was all a Clinton plot, like that batshit crazy, lying sack of shit, Justice Kavanaugh screamed during his testimony. "What goes around, comes around."
If they really believed that Kavanaugh were innocent, and just smeared by some secret Clinton plot, what would they do?
Make up rape accusations against Democrats. And of course, there would be no public revelation of anyone making the accusation, because a false accusation would risk falling apart, and revealing the scheme.
Yes, this would be vile and conspiratorial. Also, way illegal.
I remind you that Donald Trump is President. His campaign manager, who was in debt to Russian oligarchs, and son met with a Russian spy to try to get dirt on his opponent, while the Russians were working to elect Donny, whose finances are still completely unknown to us, and... I'm not even going to bother continuing this because the point is that there is so much criminality that these people accept and cover up that really? Making up a false accusation for retaliatory purposes would be nothing compared to normal politics in the Party of Donald Trump.
So, here's the Bayesian problem. We begin with a baseline probability that most allegations are true. We cannot typically update that probability upwards or downwards without the ability to assess the woman making the allegation, but she won't come forward. Normally, we wouldn't infer anything from that because women making allegations of sexual assault risk more than the men, usually. However, given the Democrats' record of cleaning house on their own recently, the Republican embrace of all-lying-all-the-time, Jim Renacci's willingness to embrace Kavanaugh despite far more credible accusations against him all things considered, and the suspicious timing, we have to treat this as a lower probability of guilt than the baseline.
Did Brown do it? I don't know. I am very skeptical, not because I distrust women, and not because I trust Brown. My default is to treat an accusation as having a baseline Bayesian probability of being true, and to disregard any assessment of the person being accused because such assessments are bullshit. All we really know that is relevant here is that the accusation is being made by someone we really shouldn't trust at a very suspicious time.
However...
As proper Bayesians, should more come out, we need to be prepared to update our assessments of Brown's guilt upwards. He may have done it. It's just that in the current era, my default assumption is that the Republican is lying through his fuckin' teeth.
Until they start challenging Donny Trump on the fact that he is clearly guilty of multiple sexual assaults, until they deal with Kavanaugh in a more serious way, until they stop doing this circle-the-wagon shit more generally, none of them save "Lisa Murkowski" have any credibility with me on sexual assault, and in the era of Trumpian lies, no Republican ever has a right to say, "trust me."
No. Not under any circumstances. If Brown did it, then I understand the dilemma the woman faces, and encourage her to come forward. If not, I understand, but Renacci's word means less than nothing to me, and I write this as someone who has zero loyalty to Sherrod Brown and defaults to trusting the women who make these allegations. If you don't have me on your side, the only way you get anyone on your side is with blind partisan hatred of Democrats.
Bayesian methods. They're the right way to go.
