Post impeachment ergo propter impeachment
This is happening. House Democrats are going to pass articles of impeachment. But, in case you missed the variations of the most informative headline ever... check this variant. I just picked an anodyne-seeming one. "McConnell, coordinating with White House, lays plans for impeachment trial." It doesn't matter how guilty the ham sandwich is, there's no point indicting it when it will be coordinating the trial with the judge and jury.
What happens after the most thoroughly-rigged jury in history acquits Trump? The 2020 election. Will the impeachment acquittal help, hurt, or do nothing to 2020?
Beware the post hoc ergo propter hoc fallacy. After this, therefore because of this. Either Trump wins or loses. Whether he steps down upon losing is another matter... (the answer is "no"), but either he wins or loses.
If he wins, will that be because he was impeached and acquitted? If he loses, will that be because of the impeachment? Post hoc ergo propter hoc. You don't get to do that.
The causal question is based on the "counterfactual." What if the world had been otherwise? What if the whistleblower complaint had come out, we had the investigation, and the Democrats dragged out the investigation without impeaching Trump? That's the counterfactual. If Trump wins after an impeachment, would he have won by a bigger margin? A smaller margin? Not at all? If he loses after an impeachment, would he have lost by a bigger margin? A smaller margin? Not at all?
The question is about the baseline. We call this "the fundamental problem of causal inference"-- we can't directly observe the counterfactual of the hypothetical world in which the Democrats investigate without impeaching because they are impeaching him.
What can we observe? Basic underlying factors. The economy, and... Trump's approval rating.
Let's take a look at RealClearPolitics and Trump's net approval/disapproval numbers. Those numbers have stayed effectively unchanged since the whistleblower complaint came out. What does that mean?
It means the Democrats' efforts have accomplished exactly squat.
It also means that so far, there hasn't been a backlash. Ignore every poll you see about impeachment. The only polls that matter are Trump's approval numbers.
The question is what happens when Mitch McConnell gets his hands on the process? Here's what happens, as he coordinates with the White House. You know, as judges and juries are supposed to do with the defendant. Right?
Burisma, Crowdstrike, and the Bidens. Conspiracy-a-go-go, followed by a speedy acquittal. Pass go, collect $200, and go directly back to the White House.
This thing hasn't gotten started yet. McConnell doesn't run the show yet. He will.
Will that, plus the inevitable acquittal move the numbers? Will it reflect on the Democrats? That's what to watch.
And the question in 2020 will be our baseline. How well should Trump do, absent any impeachment-type-stuff? That's the counterfactual. The counterfactual is not the opposite of the end result. The counterfactual is the baseline expectation.
So, plug Trump's baseline approval ratings and economic data into our forecasting models, and see how well he should perform. Then, compare that result to what happens. Does he overperform that? Underperform that? That'll be the question.
To approach the question any other way will be to fall prey to the post impeachment ergo propter impeachment fallacy.
With that in mind, consider the concept of a scandal without impeachment. In congressional elections, scandals hurt. See, in particular, Scott Basinger's "Scandals and Congressional Elections in the Post-Watergate Era," from Political Research Quarterly (2012). That's just a visibility effect, though.
Add a formal acquittal? Messier. Add a Senate "trial," which will be all about Joe and Hunter Biden, and wacky conspiracy theories, followed by a formal acquittal, and will Basinger's findings apply? Or do we revert to the 1998 reverse effect of the failed impeachment?
Remember your baseline. Post impeachment ergo propter impeachment is a logical fallacy. That won't stop literally every journalist and pundit in the country from making it, and I detest people who misuse the word, "literally."
What happens after the most thoroughly-rigged jury in history acquits Trump? The 2020 election. Will the impeachment acquittal help, hurt, or do nothing to 2020?
Beware the post hoc ergo propter hoc fallacy. After this, therefore because of this. Either Trump wins or loses. Whether he steps down upon losing is another matter... (the answer is "no"), but either he wins or loses.
If he wins, will that be because he was impeached and acquitted? If he loses, will that be because of the impeachment? Post hoc ergo propter hoc. You don't get to do that.
The causal question is based on the "counterfactual." What if the world had been otherwise? What if the whistleblower complaint had come out, we had the investigation, and the Democrats dragged out the investigation without impeaching Trump? That's the counterfactual. If Trump wins after an impeachment, would he have won by a bigger margin? A smaller margin? Not at all? If he loses after an impeachment, would he have lost by a bigger margin? A smaller margin? Not at all?
The question is about the baseline. We call this "the fundamental problem of causal inference"-- we can't directly observe the counterfactual of the hypothetical world in which the Democrats investigate without impeaching because they are impeaching him.
What can we observe? Basic underlying factors. The economy, and... Trump's approval rating.
Let's take a look at RealClearPolitics and Trump's net approval/disapproval numbers. Those numbers have stayed effectively unchanged since the whistleblower complaint came out. What does that mean?
It means the Democrats' efforts have accomplished exactly squat.
It also means that so far, there hasn't been a backlash. Ignore every poll you see about impeachment. The only polls that matter are Trump's approval numbers.
The question is what happens when Mitch McConnell gets his hands on the process? Here's what happens, as he coordinates with the White House. You know, as judges and juries are supposed to do with the defendant. Right?
Burisma, Crowdstrike, and the Bidens. Conspiracy-a-go-go, followed by a speedy acquittal. Pass go, collect $200, and go directly back to the White House.
This thing hasn't gotten started yet. McConnell doesn't run the show yet. He will.
Will that, plus the inevitable acquittal move the numbers? Will it reflect on the Democrats? That's what to watch.
And the question in 2020 will be our baseline. How well should Trump do, absent any impeachment-type-stuff? That's the counterfactual. The counterfactual is not the opposite of the end result. The counterfactual is the baseline expectation.
So, plug Trump's baseline approval ratings and economic data into our forecasting models, and see how well he should perform. Then, compare that result to what happens. Does he overperform that? Underperform that? That'll be the question.
To approach the question any other way will be to fall prey to the post impeachment ergo propter impeachment fallacy.
With that in mind, consider the concept of a scandal without impeachment. In congressional elections, scandals hurt. See, in particular, Scott Basinger's "Scandals and Congressional Elections in the Post-Watergate Era," from Political Research Quarterly (2012). That's just a visibility effect, though.
Add a formal acquittal? Messier. Add a Senate "trial," which will be all about Joe and Hunter Biden, and wacky conspiracy theories, followed by a formal acquittal, and will Basinger's findings apply? Or do we revert to the 1998 reverse effect of the failed impeachment?
Remember your baseline. Post impeachment ergo propter impeachment is a logical fallacy. That won't stop literally every journalist and pundit in the country from making it, and I detest people who misuse the word, "literally."
Comments
Post a Comment