I worked on the Federal Reserve's staff macro forecast for a decade, starting in July 2008. I have learned how to be a vey good forecaster. I am. I affirm many of the points you made. Let me suggest some baby steps into the craft:
1) start with your 'most likely' forecast, that is, the 50% or the median forecast. That's easier than thinking through the entire distribution. state the conditions on your forecast. good forecasts are ALWAYS conditional. tell the 'story.' go beyond the outcomes, tell us what gets us there.
2) when you are ready to move off the most likely, think through some 'upside' and 'downside' risks. then tell us are the risks tilted to the downside or the upside.
3) a100% agree, re-elevate your forecast (most likely, risks, conditions) regularly. then UPDATE YOUR FORECAST. key is paying attention to ALL the data available. watch for blind spots.
4) Superforecasters are often'non experts.' why? one reason is that experts tend to have blind spots. they want so badly for their models to be right that they miss signs they are wrong. Fed staff grappled with that after missing the warning signs of the housing bubble. it's hard.
5) talk with non-experts, go out in the world. my walks have taught me and conversations with non-economists since Covid arrived have taught me so much. actively seek it out.
6) finally, be humble, especially, when your forecast proves right. after a really big hit, a senior officer came by my office and told me, "do not gloat in our next meeting. it could have been luck." and he said with a smile, "we all know you nailed it. you don't have to say a word."
I was fortunate and learned from the very best. no one has a crystal ball, but some of them give damn good advice.
This is content! I'm gonna brag to my partner that in a discussion of prediction accuracy, I am party of a community that includes thoughtful commentary from someone who literally worked at the Fed. Gadzukes this is sweet!!!!
I think the most important thing you said was about conditionality. One of the things that annoys me the most in pundits that predict or "warn of" growth/recession or of renewed inflation (this applies to Matt, too) is that they are not made conditional on Fed policy.
Shouldn't you or someone like you write a book about how and why the Fed has not kept average PCE inflation at 2%? Were there constraints on how much QE could be done? What role did TIPS expectations play in decisions?
great question and one I have thought a lot about. (I take my own advice):
1) CARES Act was bigger, faster, and broader than I expected. Congress delivered in March (not since). Shocked me. AND CARES WORKED MUCH BETTER. even I was surprised how much was spent out of the direct payments. same as in 2008. I did the research for 2020{ https://twitter.com/Claudia_Sahm/status/1295482998918459394?s=20. Others found similar bounce, even though many stores shut down. Plus that extra $600 a week in jobless benefits was a godsend and payroll protection loan program did (eventually) work.
2) I underestimated how hard it'd be to measure the unemployment rate in the household survey. The official rate shot up (but less than I had projected given jobless claims). we now know it was understated a) people were confused if they were on temp layoff or not so many laid off were not counted as unemployed b) even after measurement issues subsided, millions dropped out of the labor force--no child care, quarantines, Covid scare, virtual school, etc.--so they also don't count as unemployed. with adjustments it's still about 10% for the nation as a whole. We are still 10 million jobs less than February -- which is a bigger hole than in the Great Recession and jobless claims remain mindboggling large. Even so, early snap back, impressed me in a good way.
3) finally, I do not regret my forecast. I told a dire story (see above advice), we lived a dire year. and 2021 ain't gonna start well. I was pushing back on a very strong V-shaped recovery, which was all the rage in March and April. one macro leader even called Covid shutdowns 'Cape Cod in the Winter.' CARES Act was written for him to be right. he was not. key programs like the $600 expired at end of July. CARES worked but we were not back on track when that relief waned.
The prediction ought to be in terms of average inflation/price level. The farther back the start date, the lower my prediction of average inflation> 2% PCE going forward. Even starting in August (when the Fed made it latest iteration of targeting) I'll put average inflation 8/2020-12/2021 >2.3% CPI (~ 2% PCE) at 0.33.
What I mean by is interesting is: choosing the YoY CPI level in any given month is a pretty obscure thing to select, and so you must have had a reason for doing so.
But your specific prediction is very likely to be wrong, as anyone even vaguely familiar with inflation dynamics could tell you, so is the point of the prediction that you just don’t believe the “inflationistas” and wanted to get that on record?
Matt's post was good as are your suggestions. I'm often of two minds when I read predictions in the rationalist style. On one hand, I applaud the rigor: fierce debates over vague claims are the worst. On the other hand, I find myself quickly bored by the predictions and rarely survive the full list, whether it be Matt's or Scott Alexander's, at full reading attention.
Your excellent first suggestion helps me understand that boredom. Without the story, the explanation, a typical probabilistic prediction remains unfalsifiable in any practical sense (especially if the probability falls within moderate ranges, as they usually do). If someone predicts X with 70% probability, neither outcome tells us much: not-X is unsurprising and therefore mostly forgotten; and X could be a case of getting the right answer for the wrong reason. Yes, we can crunch all of the predictions and determine whether Matt is good or improving at predicting. But I don't care about that very much; I care about the substantive issues Matt writes about.
By contrast, a probabilistic prediction with an attached explanation is a lot more interesting and valuable. It helps rule out right-for-wrong-reasons successes. It might even rehabilitate predictive failures when the prognosticator tells a story that turns out correct, but the final outcome gets frustrated by something out of left field. Probably other benefits as well.
He's really gone, right? The web site has not been updated.
Has the NYT ever discussed why they would not play along with the pseudo anomality he wanted? Does it seem likely that that is really the reason he has disappeared?
Making fun of rationalists is a noble endeavor and probably worth a New Yorker article, although he's hardly the worst of them and the NYT is not the best place to do it either. The normal white guy rationalists seem to have got into it over minor disagreements with woke progressives but end up believing that they have no preexisting beliefs (only priors), but also that it's illogical to not be in an open relationship and we're all going to be enslaved by evil computers in the future.
Meanwhile his problem (which he seems especially sensitive about) is that he and, like, Tyler Cowen cultivated a bunch of scientific racists in their comment section, maybe because they enjoyed arguing with them more than with the woke people.
I found myself in disagreement with SA on many issues and have no issues with criticism of rationalists. However, with great power comes great responsibility. The NYT being THE leading paper in the country gives it vast power to impact people's lives and it handled that situation poorly.
I assumed it was more on the demand side than Tyler/Scott enjoying it but admit I didn't follow the boards closely. Racists have few mainstreamish places they are tolerated, so flock to libertarian rationalist boards that will tolerate anything as long as you provide a veneer of logic.
There’s a philosophical difference; the Bayesian people sort of like to act like the priors are just out there in the world and they’re constantly calculating everything from observation and first principles. I think that would be nice but they’re not claiming enough personal responsibility.
This is what I respect about 538. Say what you will about Nate Silver (I think he’s kind of an asshole) he makes a serious effort to make his models and predictions as accurate as possible. He’s not just trying to boost his preferred ideological position or something.
Funnily enough, the place you see this most clearly is the 538 sports models, which are often wrong (for instance they seriously underrated the Lakers in the last NBA playoffs), but it’s obviously not the case that they’re wrong because of some secret bias against the Lakers or something by the creators.
I do still strongly suspect Silver was consciously or unconsciously hedging his presidential model this year, knowing that for Trump to win after being forecast 95% to lose would be horrible for 538’s reputation. Of course, you could say, “he was right; the election ended up much closer than polls predicted,” but that doesn’t really work since by November his model was right where Morris’s was; the uncertainty he applied in the summer seemed to be based on caution that didn’t have an objective basis.
The changes silver made to his model, cheezy as they were, were clearly a response to the problems in 2016, in a way that I think is justifiable. 2016 saw a bunch of wild changes in the predictions where the confidence in the presidential result repeatedly swung between 50:50 and 85:15. Finding ways to dampen that out in an evidence based, back-testable way was the only responsible thing to do.
In one post Nate actually went back to his model's visualization and showed which result actually happened (at the state level) and it was near the expected outcome in his model.
I always saw the difference in Nate's model and Morris's model as their comfort in the assumption that a polling lead at a given time, conditional on all the covariates, would act similarly to the way they had in the past. Morris back-fitted all his data and there was no historical example for an upset. He was much more confident. Nate was less confident that assumption made sense and was open to the idea of unobservables still having time to change things. Maybe there weren't enough data points or that something about partisanship, the media, trump had changed to make that assumption wrong. I'm pretty sure it was more than just a fudge but that's how I saw it. but I'm just an amateur.
If you look at NFL and NBA games (which is all I look at, sorry for not stating that), their predictions are almost in line with the "unskilled forecast". The playoffs are better though. I'm not good enough at statistics to continue arguing with you since you might be right! It would be interesting to compare these predictions to other sports models and humans.
Matt, will you also be interested in writing down what were the biggest wrong-headed takes you did in past 20 years of your career and what you learned from them? Like this predictions thing that will also help with building some pundit accountability
He discussed this a bit in the Weeds episode w/Ezra. The main one I remember was believing that the US should be using its military might for humanitarian/liberal democracy purposes.
I feel like he may have overlearned this lesson, since sometimes this has worked really well (the Balkans, especially) and sometimes not doing it had such hideous consequences it’s hard to imagine action leading to worse outcomes (Rwanda).
"But a 15 percent chance of a global pandemic is really bad! We need people to be able to discuss moderately improbable bad events without sounding like the boy who cried wolf."
I've been thinking about this more or less constantly since reading The Fifth Risk. This conversation about how to talk about and address improbable but really awful outcomes is very important.
Exactly this. NASA funds the search for near-Earth asteroids not because a large asteroid collision is likely, but because if you can spend money to eliminate even the tiny chance of a truly catastrophic event, it’s worth it.
One of the problems with low risk high impact events is that our brains don't really work that way.
Post-COVID, Lewis looks like Nostradamus for having put into a book the thesis "it's dangerous to have Trump at the wheel because he'll screw up if a low-risk high-impact event happens."
What if the pandemic happened 5 years from now? It would be really easy to make the counter-case: "most of the protection from rare events comes from the system, not the guy in charge" or "the president doesn't have that much unique influence" or even "the Federal government isn't providing any value, we should shrink it." And that thesis would seem really plausible.
My mental model of people is that our risk is binary - we round everything down to 0% or up to 100%, and most arguments about improbable events are disagreements between the participants about which way the rounding went.
This is relevant to a topic I’ve raised here before: the validation of COVID models. I mean both the formal mathematical models and whatever lies behind media statements like things are going to get worse or better. More specifically, I’m thinking about the flurry of news about people travelling over Thanksgiving. It was a mess in that there was no precision about how much extra travel was being feared, whether that mount or more or less resulted and whether given the amount of travel there was, was there additional cases, serious cases and deaths.
Just looking at the aggregate data, I can’t see any “Thanksgiving effect.” The increase in cases before and after Thanksgiving look about the same to me and the relation of deaths to reported cases has not changed a lot. But I’m just a guy on the Internet. Why isn’t there a NYT story about how public health officials got Thanksgiving right or wrong and why. How did travel compare to predictions? Did more of those who travelled have negative tests before doing so than expected? And how id the incident affect the parameters of or the structure of the models that officials use?
The NYT might run such a story in 2021. But they won't run any story that might validate skepticism about the effectiveness of lockdowns while Trump is in office. I'm interested to see how they exercise editorial judgement in a Biden administration, or if they will act like some state-sponsored newspaper for the next 4 years.
I mean, cases peaked on Dec 11, about 2 weeks after Thanksgiving, so I'd probably want more than an eye-test to say whether or not you can claim a Thanksgiving effect.
I do generally agree that Covid prediction journalism has been horrendous, in a lot of the same ways science journalism consistently fails. The early models that predicted extremely high death tolls were upfront about predicting no government intervention - a worthwhile thing to model, but a case that specifically produces the upper bound of fatalities - and then media just reported the number straight up and the whole thing was less informative than if they had not written it up at all.
"The early models that predicted extremely high death tolls were upfront about predicting no government intervention"
Just a note of clarification, the early models "worst case" involved no action by both the government and individuals. They made sure to note that such a scenario was very very unlikely.
In 2009 I saw Lana del Rey, who wasn't calling herself Lana del Rey yet, play in a small club in Manhattan, and as we left I turned to my wife and said, "She's going to be playing arenas someday." That was pretty much my last prediction about anything, and I'd hate to break my streak.
In 1994, the minute that Nancy Kerrigan got nailed in the knee with a crowbar, I turned to my college roommate and said "Tonya Harding totally had someone do that". I think that stands as my best prediction like that.
Granted, I grew up in the same area that she did, and knew she was the kind of person who, after a small bit of fame, made a scene at a bar yelling "don't you know who I am?". So something something inside info.
Specifically, I recommend joining the challenges sponsored by The Economist and Fareed Zakaria GPS. I won both in 2018!
Good Judgement Inc. annually asks the top performers to become “Professional Superforecasters”. I’ve been a “Super” for 6 years. Here are some of our public forecasts:
I just read the NYT article about substack, where it is claimed that you have bitterly broken with liberalism. That seems to me to be an overstatement. You've always examined the ideological shibboleths of liberalism skeptically as far as I can recall. I often disagree with you, however you are consistently interesting and with the money.
He seems to be breaking with the progressive camp, but doubling down on liberalism. I thought the progressive label was just marketing 5-10 years ago, but its a definitively different tactical mindset.
I think the possibility that the GA Senate races end with Perdue/Warnock as the winners is underrated! Clearly the races will be defined by partisan turnout but the margins could be razor-thin and Kelly Loeffler is a uniquely awful candidate (not to mention how appointed incumbents in general see less of an advantage than elected incumbents). Warnock also tends to perform slightly better in polls than Ossoff. I don't think there will be many Perdue/Warnock voters, but I think there could be some only-Warnock Democratic voters and only-Perdue Republican voters. Not many, but possibly enough. Since we're being accountable I would put my "the same party wins both GA Senate races" at 80%.
At the moment I'm probably thinking something like:
60% that Warnock beats Loeffler
40% that Ossoff beats Perdue
which I guess implies something like
35% Warnock/Ossoff
35% Loeffler/Purdue
25% Warnock/Purdue
5% Loeffler/Ossoff
And I'm only thinking this because the polls have shown the races as consistently close, with Warnock almost always leading and Ossoff often a little bit behind.
I don’t understand the optimism here or in Matt’s forecast. The 538 and RCP polling averages have both races essentially tied. Why isn’t all the rest just pointless speculation?
Yes but, as an outside observer, Georgia is a state normally won by Republicans where one Democrat won by narrow margins and the polls show a tied race. What information am I missing that someone who thinks Warnock and Ossoff are favored has?
While it’s true that republicans usually win Georgia, it is also clearly trending towards Democrats at a rapid rate (sort of the opposite of Wisconsin) and is now only 4 points more republican than the country at large.
Whether you think Dems will win the senate races probably comes down to whether you think the electorate will be more or less democratic than in November, and it’s easy to argue that Dems will actually have the advantage in a lower turnout election because their increasingly educated base is more likely to turn out.
Good point, I was thinking more about the prospect of a split decision vs. either party winning both seats. I would actually give Republicans the edge here, personally.
Isn't there the thing where voters need to register *again* to vote? Anyone know how this breaks down among Dems and Republicans? Who are more excited to vote?
Here’s a possibly dumb question but I don’t know statistics very well so I’ll ask... how do you check yourself on predictions where your confidence levels aren’t the same? If you have 10 predictions at 60%, that’s easy, you described it. But if you have 2 at 85%, 3 at 70%, 2 at 50%... is there a way to to combine all of those or do you just need to wait until you have enough of a sample size at each confidence level?
The method of "calibration" that is being discussed here only works if you have large numbers of predictions at each confidence level. Another method that is often used is the method of "scoring rules" (https://en.wikipedia.org/wiki/Scoring_rule), where you figure out whether each prediction was true or false, label those with 1 or 0, and then sum up some function of the difference between the prediction probability and the label (the two most common ones are either the square of the difference, or the negative logarithm of the difference). While someone can be perfectly calibrated if they give 50% as their prediction for 100 things, 50 of which are true and 50 false, the scoring rule only sums to a perfect score of 0 if you give 100% as your prediction for all the true things and 0% as your prediction for all the false things. The scoring rule gives you an incentive to aim towards extreme values if you are in fact very confident in that prediction, rather than hedging with a bunch of predictions at the same value. The difference between the quadratic score and the logarithmic score (and other rules) is based on whether they want you to be more careful about your predictions near the extremes, or near the middle of the range, or somewhere else.
"Twitter is full of bullshit" is an amazing sub-heading. But I view cable news punditry as the true home of the unaccountable crap prediction - tons of air time to fill, talking heads saying whatever they want, no one ever checks except rarely the Daily Show.
Twitter's it's own thing - it's performative group-on-group public fighting. The incentive isn't just to be wrong in a provocative way, it's to dunk on the other side.
Hi Matt, great to see you doing this. Here are some ideas based on Scott Alexander's experience:
* it's easier to check the calibration of your predictions if they're all multiples of 10. Most people can't distinguish between 21 levels of confidence anyway, so using numbers ending in 5 could simply make your work harder
* it will help check your calibration if all the probabilities given are 50 percent or higher. Your Lakers prediction could be rephrased as a prediction with 75 percent confidence that someone other than the Lakers will win
Superforecasting is a great book, and I agree more people should read it! Thanks again, and I look forward to seeing how your predictions turn out.
On the Lakers one, I think it matters that this is a case where there are several dozen possibilities, and he is listing the one that he thinks of as by far most likely, even though it's still less than 50%. If he had a bunch of predictions of this form, then it probably is more helpful to test calibration on the low probability specific predictions, rather than the high probability non-specific predictions.
I think the thing that surprises me most in this list is how bullish you are on the restoration of the JCPOA.
What's your thinking on Iran's incentives to rejoin at this stage? I have to imagine their internal pressures from the hardliners but also from just the general populace after the dual assassinations this year will make that difficult politically even if the leadership was so inclined, which also does not seem clear.
I had the exact same reaction. 80% for rejoining and compliance seems really high to me. I would say rejoining is probably around 80% and then actual compliance given rejoining is probably around 60-70% putting it more in the 50-60% range jointly.
I agree that most of the discussion around the attempted "coup" has been awful and hysterical (much like everything with Twitter) and I know this wasn't the main point of your post, but there were a number of prominent writers (Zeynep Tufecki in particular) to point out that Trump was:
a) Intentionally disrupting the mail to make it harder/more confusing to mail in ballots in states like PA
b) Having local Republicans insist that mail-in ballots be counted AFTER Election Day, thus making it seem like he was ahead (again, in PA)
c) Leaning on judges he appointed to stop the counting of mail-in ballots
And the election was pretty close in the Electoral College! 45k votes in GA, WI, and AZ and Trump wins! The point was that, had it been one state (or even two), Trump might have found more sympathetic judges to flip the count for him. But with so many states, it would've been politically untenable, and the smart judges knew they couldn't do it credibly. I think it's more than fair to say those fears were NOT overblown, and had Trump been successful, it would've constituted a coup (albeit not a military coup).
DeJoy's testimony before Congress about dismantling mail sorting machines? Postal workers not getting paid overtime despite increased volume, as was standard practice? Increased delays in delivery times? I could go on... there's tons of evidence, none of this is even remotely controversial outside of the right wing blogosphere.
I think a) and b) are backfilling conspiracies into unrelated events. B) definitely is, since it wasn't expected mail-in ballots would skew so heavily Democratic (otherwise the GOP wouldn't have been in support of the bill when they passed it); the rules for ballot counting were set well in advance of election day and there's no real evidence Trump was involved or interested in that process; and lots of states have slow counting processes for a variety of reasons - Pennsylvania wasn't even particularly late in certifying results. The second most likely tipping point state, Florida, being the literal first state to report substantive results doesn't help the theory either. Why weren't they let in on the anti-mail conspiracy?
A) I guess is still open to interpretation, but my takeaway from the DeJoy hearing was that he's a standard efficiency consultant who thought he could come in and twist a few dials (Send the trucks out on time! Cut overtime!) and make the USPS magically profitable. Maybe there was a deep conspiracy angle there but the Occam's Razor of DeJoy acting exactly like you would expect a Republican political appointee to act is much more plausible.
C) is actually concerning, but lazy, ineffectual corruption is so baked into Trump's persona that I don't see how that ever produces a different outcome.
It does seem to me like DeJoy was just innocently breaking everything, but it could be that the people who hired him knew he’d do that.
I still want to know what was up with Rex Tillerson coming in and apparently destroying the State Department with management consulting. Is there an accurate history about this yet? Did Putin actually get him appointed over Mitt Romney?
I thought that by August or September, it was clear that everywhere that mail-in ballots were an easy option, they would strongly skew Democratic, just because of the observed trends in all other walks of life during the pandemic, where in-person versions skewed Republican and home versions skewed Democratic.
Mail-in voting was a pre-Pandemic measure that passed with bipartisan support in 2019. The PA GOP couldn't have predicted covid would drastically increase it's usage, or that their president would directly tell people not to vote by mail, but the law as written could not have been part of a conspiracy because it was crafted under a completely different environment.
Not sure how dismantling mail sorting machines constitutes increased efficiency. There's no reason to give him or anyone else in the Trump administration the benefit of the doubt.
And Trump's strategy was pretty clearly to bank on PA being close and have everything else break his way (which it almost did). Why else would their entire army of lawyers descend on Philadelphia? If FL was lost, it would've been a landslide as Matt had said.
IIRC the argument was they were replacing the mail sorting machines with package sorting machines. The case behind it was that first class mail has been decreasing for years so they don't need mail sorting machines as much. In contrast, package delivery has been increasing and they needed more automation. I think there was also a lot of push back from the union fighting his efficiency program broadly and especially his reduced OT plans.
Obviously what Trump is doing is bad, and what other Republicans are doing/not doing is bad (and probably worse), but I've been against the "coup" description for several reasons. First, asking Courts to do something, which is most of what has actually occurred (rather than merely been discussed), doesn't seem to be what is usually meant by the word coup, but more important has been what is implied: no-one who has called this a coup really seems to acting like it is one.
If this was a coup, what would that imply people who oppose him and it should do? Should there be a counter-coup? Is violence not just an available option, and justified, but wouldn't a coup be so bad that perhaps failing to engage in some level of violence in response to it is actually bad? If there's a coup happening, why is the only response to it tweeting and op-ed-ing instead of rounding up coup supporters to stop the overthrow of democracy? etc.
I am happy to say I don't think this is a coup, and don't think violence is justified to stop it, but how do people who think this is a coup justify taking the same actions in response to it that I am (waiting for the courts to reject idiotic claims, waiting for the electors to vote, waiting for Congress to count the ballots, and waiting for inauguration)?
Violence isn’t justified because no one has expected the attempted coup to succeed. If we thought it had a chance, we’d be out in the streets. Courts have firmly stomped on lawsuits attempting to overturn the election. Military leadership has firmly stomped on any thought of becoming involved in the choice of president. Filing ridiculous lawsuits isn’t illegal, so mass arrests are obviously unjustified.
You put into words something I've been mulling over a for a while.
I think it's clear at this point that responding to Trump's behavior _as if_ it was a real coup would have had a worse outcome...there's still a disturbing amount of belief in Trump's claims, but we now have a court record of this stuff getting the door slammed on it, and that legal process ran without interference...a bunch of left wing "anti-coup" behavior would have fueled the fire.
I worked on the Federal Reserve's staff macro forecast for a decade, starting in July 2008. I have learned how to be a vey good forecaster. I am. I affirm many of the points you made. Let me suggest some baby steps into the craft:
1) start with your 'most likely' forecast, that is, the 50% or the median forecast. That's easier than thinking through the entire distribution. state the conditions on your forecast. good forecasts are ALWAYS conditional. tell the 'story.' go beyond the outcomes, tell us what gets us there.
2) when you are ready to move off the most likely, think through some 'upside' and 'downside' risks. then tell us are the risks tilted to the downside or the upside.
3) a100% agree, re-elevate your forecast (most likely, risks, conditions) regularly. then UPDATE YOUR FORECAST. key is paying attention to ALL the data available. watch for blind spots.
4) Superforecasters are often'non experts.' why? one reason is that experts tend to have blind spots. they want so badly for their models to be right that they miss signs they are wrong. Fed staff grappled with that after missing the warning signs of the housing bubble. it's hard.
5) talk with non-experts, go out in the world. my walks have taught me and conversations with non-economists since Covid arrived have taught me so much. actively seek it out.
6) finally, be humble, especially, when your forecast proves right. after a really big hit, a senior officer came by my office and told me, "do not gloat in our next meeting. it could have been luck." and he said with a smile, "we all know you nailed it. you don't have to say a word."
I was fortunate and learned from the very best. no one has a crystal ball, but some of them give damn good advice.
This is content! I'm gonna brag to my partner that in a discussion of prediction accuracy, I am party of a community that includes thoughtful commentary from someone who literally worked at the Fed. Gadzukes this is sweet!!!!
She also had a conversation with Matt on The Weeds back in January. For those who missed it: https://www.stitcher.com/show/voxs-the-weeds/episode/fix-recessions-by-giving-people-money-66991306
I think the most important thing you said was about conditionality. One of the things that annoys me the most in pundits that predict or "warn of" growth/recession or of renewed inflation (this applies to Matt, too) is that they are not made conditional on Fed policy.
Shouldn't you or someone like you write a book about how and why the Fed has not kept average PCE inflation at 2%? Were there constraints on how much QE could be done? What role did TIPS expectations play in decisions?
Those are good points. Tweets aren't forecasts (sorta Matt's point I suppose), but did you ever get into what you missed when you predicted a depression in late May? https://twitter.com/Claudia_Sahm/status/1265318891749888003?s=20
great question and one I have thought a lot about. (I take my own advice):
1) CARES Act was bigger, faster, and broader than I expected. Congress delivered in March (not since). Shocked me. AND CARES WORKED MUCH BETTER. even I was surprised how much was spent out of the direct payments. same as in 2008. I did the research for 2020{ https://twitter.com/Claudia_Sahm/status/1295482998918459394?s=20. Others found similar bounce, even though many stores shut down. Plus that extra $600 a week in jobless benefits was a godsend and payroll protection loan program did (eventually) work.
2) I underestimated how hard it'd be to measure the unemployment rate in the household survey. The official rate shot up (but less than I had projected given jobless claims). we now know it was understated a) people were confused if they were on temp layoff or not so many laid off were not counted as unemployed b) even after measurement issues subsided, millions dropped out of the labor force--no child care, quarantines, Covid scare, virtual school, etc.--so they also don't count as unemployed. with adjustments it's still about 10% for the nation as a whole. We are still 10 million jobs less than February -- which is a bigger hole than in the Great Recession and jobless claims remain mindboggling large. Even so, early snap back, impressed me in a good way.
3) finally, I do not regret my forecast. I told a dire story (see above advice), we lived a dire year. and 2021 ain't gonna start well. I was pushing back on a very strong V-shaped recovery, which was all the rage in March and April. one macro leader even called Covid shutdowns 'Cape Cod in the Winter.' CARES Act was written for him to be right. he was not. key programs like the $600 expired at end of July. CARES worked but we were not back on track when that relief waned.
I am very very worried that the longer the weak labor market lasts (lost jobs, wage cuts, fewer hours, and more layoffs) the more persistent the damage will be. See my pieces at Bloomberg: https://www.bloomberg.com/opinion/authors/AUVWWumbCcs/claudia-sahm?sref=YAR8Qcu4 and NYT: https://www.nytimes.com/column/claudia-sahmon many of my worries - and some solutions!
oops here's the NYT link: https://www.nytimes.com/column/claudia-sahm
Only putting this here as Claudia may disagree.
I don’t think it’s very important, but I’d be very surprised if core CPI doesn’t have a month above 2% YoY in 2021
In fact I’d put it at > 80%
That itself is interesting to me
The prediction ought to be in terms of average inflation/price level. The farther back the start date, the lower my prediction of average inflation> 2% PCE going forward. Even starting in August (when the Fed made it latest iteration of targeting) I'll put average inflation 8/2020-12/2021 >2.3% CPI (~ 2% PCE) at 0.33.
I hope Claudia will make a prediction.
What I mean by is interesting is: choosing the YoY CPI level in any given month is a pretty obscure thing to select, and so you must have had a reason for doing so.
But your specific prediction is very likely to be wrong, as anyone even vaguely familiar with inflation dynamics could tell you, so is the point of the prediction that you just don’t believe the “inflationistas” and wanted to get that on record?
Matt's post was good as are your suggestions. I'm often of two minds when I read predictions in the rationalist style. On one hand, I applaud the rigor: fierce debates over vague claims are the worst. On the other hand, I find myself quickly bored by the predictions and rarely survive the full list, whether it be Matt's or Scott Alexander's, at full reading attention.
Your excellent first suggestion helps me understand that boredom. Without the story, the explanation, a typical probabilistic prediction remains unfalsifiable in any practical sense (especially if the probability falls within moderate ranges, as they usually do). If someone predicts X with 70% probability, neither outcome tells us much: not-X is unsurprising and therefore mostly forgotten; and X could be a case of getting the right answer for the wrong reason. Yes, we can crunch all of the predictions and determine whether Matt is good or improving at predicting. But I don't care about that very much; I care about the substantive issues Matt writes about.
By contrast, a probabilistic prediction with an attached explanation is a lot more interesting and valuable. It helps rule out right-for-wrong-reasons successes. It might even rehabilitate predictive failures when the prognosticator tells a story that turns out correct, but the final outcome gets frustrated by something out of left field. Probably other benefits as well.
Calculated Risk great example of this approach
Scott Alexander has been doing this since 2014 I think? See if you can outscore him!
I definitely cannot outscore him! That’s kind of the point. But hopefully next year I’ll be able to outscore how I did this year.
any other notable people that you think you could get to join you in this practice? Would be really cool to see it go at least a little mainstream
He's really gone, right? The web site has not been updated.
Has the NYT ever discussed why they would not play along with the pseudo anomality he wanted? Does it seem likely that that is really the reason he has disappeared?
He posted an update in September. He's working towards no longer needing the anonymity. Looks like he might move to Substack?
If he moves to Substack, he will earn more than his current job (prediction with 95% confidence)
Called it! https://astralcodexten.substack.com/
Best $100 I will spend all year. Slow Boring is a close 2nd. (Sorry, Matt.)
Making fun of rationalists is a noble endeavor and probably worth a New Yorker article, although he's hardly the worst of them and the NYT is not the best place to do it either. The normal white guy rationalists seem to have got into it over minor disagreements with woke progressives but end up believing that they have no preexisting beliefs (only priors), but also that it's illogical to not be in an open relationship and we're all going to be enslaved by evil computers in the future.
Meanwhile his problem (which he seems especially sensitive about) is that he and, like, Tyler Cowen cultivated a bunch of scientific racists in their comment section, maybe because they enjoyed arguing with them more than with the woke people.
I found myself in disagreement with SA on many issues and have no issues with criticism of rationalists. However, with great power comes great responsibility. The NYT being THE leading paper in the country gives it vast power to impact people's lives and it handled that situation poorly.
I assumed it was more on the demand side than Tyler/Scott enjoying it but admit I didn't follow the boards closely. Racists have few mainstreamish places they are tolerated, so flock to libertarian rationalist boards that will tolerate anything as long as you provide a veneer of logic.
Aren't priors preexisting beliefs?
There’s a philosophical difference; the Bayesian people sort of like to act like the priors are just out there in the world and they’re constantly calculating everything from observation and first principles. I think that would be nice but they’re not claiming enough personal responsibility.
Yes, that's exactly what they're doing.
Hey said he was hoping to start a substack in January. Maybe a bit later.
Link
https://www.reddit.com/r/slatestarcodex/comments/jwxtt4/another_blog_update_would_be_nice/gcuy44n?utm_source=share&utm_medium=web2x&context=3
The archive is back. Just use the “older” button at the bottom. But no new posts lately.
Thanks
This is what I respect about 538. Say what you will about Nate Silver (I think he’s kind of an asshole) he makes a serious effort to make his models and predictions as accurate as possible. He’s not just trying to boost his preferred ideological position or something.
Funnily enough, the place you see this most clearly is the 538 sports models, which are often wrong (for instance they seriously underrated the Lakers in the last NBA playoffs), but it’s obviously not the case that they’re wrong because of some secret bias against the Lakers or something by the creators.
I do still strongly suspect Silver was consciously or unconsciously hedging his presidential model this year, knowing that for Trump to win after being forecast 95% to lose would be horrible for 538’s reputation. Of course, you could say, “he was right; the election ended up much closer than polls predicted,” but that doesn’t really work since by November his model was right where Morris’s was; the uncertainty he applied in the summer seemed to be based on caution that didn’t have an objective basis.
The changes silver made to his model, cheezy as they were, were clearly a response to the problems in 2016, in a way that I think is justifiable. 2016 saw a bunch of wild changes in the predictions where the confidence in the presidential result repeatedly swung between 50:50 and 85:15. Finding ways to dampen that out in an evidence based, back-testable way was the only responsible thing to do.
In one post Nate actually went back to his model's visualization and showed which result actually happened (at the state level) and it was near the expected outcome in his model.
I always saw the difference in Nate's model and Morris's model as their comfort in the assumption that a polling lead at a given time, conditional on all the covariates, would act similarly to the way they had in the past. Morris back-fitted all his data and there was no historical example for an upset. He was much more confident. Nate was less confident that assumption made sense and was open to the idea of unobservables still having time to change things. Maybe there weren't enough data points or that something about partisanship, the media, trump had changed to make that assumption wrong. I'm pretty sure it was more than just a fudge but that's how I saw it. but I'm just an amateur.
They still haven't come up with a good explanation on last year's model in the NBA playoffs.
The explanation is: their sports models are terrible
They seem very well calibrated for sports forecasts! But it's hard to demonstrate much skill for individual matches which are highly variable: https://projects.fivethirtyeight.com/checking-our-work/
If you look at NFL and NBA games (which is all I look at, sorry for not stating that), their predictions are almost in line with the "unskilled forecast". The playoffs are better though. I'm not good enough at statistics to continue arguing with you since you might be right! It would be interesting to compare these predictions to other sports models and humans.
Matt, will you also be interested in writing down what were the biggest wrong-headed takes you did in past 20 years of your career and what you learned from them? Like this predictions thing that will also help with building some pundit accountability
He discussed this a bit in the Weeds episode w/Ezra. The main one I remember was believing that the US should be using its military might for humanitarian/liberal democracy purposes.
I feel like he may have overlearned this lesson, since sometimes this has worked really well (the Balkans, especially) and sometimes not doing it had such hideous consequences it’s hard to imagine action leading to worse outcomes (Rwanda).
+1 - I'd love to see "Matt's worst takes of 2020." It could be a private post to limit the embarrassment, we can all promise not to repost. :-)
I think there's _huge_ value in re-examining past thinking to see how it actually turned out.
"But a 15 percent chance of a global pandemic is really bad! We need people to be able to discuss moderately improbable bad events without sounding like the boy who cried wolf."
I've been thinking about this more or less constantly since reading The Fifth Risk. This conversation about how to talk about and address improbable but really awful outcomes is very important.
Exactly this. NASA funds the search for near-Earth asteroids not because a large asteroid collision is likely, but because if you can spend money to eliminate even the tiny chance of a truly catastrophic event, it’s worth it.
It's difficult enough to deal with probable bad events as long as they're a gradual process (e.g., global warming).
One of the problems with low risk high impact events is that our brains don't really work that way.
Post-COVID, Lewis looks like Nostradamus for having put into a book the thesis "it's dangerous to have Trump at the wheel because he'll screw up if a low-risk high-impact event happens."
What if the pandemic happened 5 years from now? It would be really easy to make the counter-case: "most of the protection from rare events comes from the system, not the guy in charge" or "the president doesn't have that much unique influence" or even "the Federal government isn't providing any value, we should shrink it." And that thesis would seem really plausible.
My mental model of people is that our risk is binary - we round everything down to 0% or up to 100%, and most arguments about improbable events are disagreements between the participants about which way the rounding went.
This is relevant to a topic I’ve raised here before: the validation of COVID models. I mean both the formal mathematical models and whatever lies behind media statements like things are going to get worse or better. More specifically, I’m thinking about the flurry of news about people travelling over Thanksgiving. It was a mess in that there was no precision about how much extra travel was being feared, whether that mount or more or less resulted and whether given the amount of travel there was, was there additional cases, serious cases and deaths.
Just looking at the aggregate data, I can’t see any “Thanksgiving effect.” The increase in cases before and after Thanksgiving look about the same to me and the relation of deaths to reported cases has not changed a lot. But I’m just a guy on the Internet. Why isn’t there a NYT story about how public health officials got Thanksgiving right or wrong and why. How did travel compare to predictions? Did more of those who travelled have negative tests before doing so than expected? And how id the incident affect the parameters of or the structure of the models that officials use?
The NYT might run such a story in 2021. But they won't run any story that might validate skepticism about the effectiveness of lockdowns while Trump is in office. I'm interested to see how they exercise editorial judgement in a Biden administration, or if they will act like some state-sponsored newspaper for the next 4 years.
Did they seem to you state-sponsored 2009-2016?
No, but that was a different NYTimes. They've changed dramatically since then.
Even so narrow an issue as Thanksgiving travel? Why 2021? What's the "political" angle right now?
I mean, cases peaked on Dec 11, about 2 weeks after Thanksgiving, so I'd probably want more than an eye-test to say whether or not you can claim a Thanksgiving effect.
I do generally agree that Covid prediction journalism has been horrendous, in a lot of the same ways science journalism consistently fails. The early models that predicted extremely high death tolls were upfront about predicting no government intervention - a worthwhile thing to model, but a case that specifically produces the upper bound of fatalities - and then media just reported the number straight up and the whole thing was less informative than if they had not written it up at all.
"The early models that predicted extremely high death tolls were upfront about predicting no government intervention"
Just a note of clarification, the early models "worst case" involved no action by both the government and individuals. They made sure to note that such a scenario was very very unlikely.
But what policy variables was doing more than nothing conditional on?
In 2009 I saw Lana del Rey, who wasn't calling herself Lana del Rey yet, play in a small club in Manhattan, and as we left I turned to my wife and said, "She's going to be playing arenas someday." That was pretty much my last prediction about anything, and I'd hate to break my streak.
In 1994, the minute that Nancy Kerrigan got nailed in the knee with a crowbar, I turned to my college roommate and said "Tonya Harding totally had someone do that". I think that stands as my best prediction like that.
Granted, I grew up in the same area that she did, and knew she was the kind of person who, after a small bit of fame, made a scene at a bar yelling "don't you know who I am?". So something something inside info.
For those interested in forecasting, I recommend signing up for the commercial offshoot of Tetlock’s work:
https://www.gjopen.com/
Specifically, I recommend joining the challenges sponsored by The Economist and Fareed Zakaria GPS. I won both in 2018!
Good Judgement Inc. annually asks the top performers to become “Professional Superforecasters”. I’ve been a “Super” for 6 years. Here are some of our public forecasts:
https://goodjudgment.io/superforecasts/#1418
Metaculus is another popular forecasting platform:
https://www.metaculus.com/questions/?show-welcome=true
I just read the NYT article about substack, where it is claimed that you have bitterly broken with liberalism. That seems to me to be an overstatement. You've always examined the ideological shibboleths of liberalism skeptically as far as I can recall. I often disagree with you, however you are consistently interesting and with the money.
He seems to be breaking with the progressive camp, but doubling down on liberalism. I thought the progressive label was just marketing 5-10 years ago, but its a definitively different tactical mindset.
I think the possibility that the GA Senate races end with Perdue/Warnock as the winners is underrated! Clearly the races will be defined by partisan turnout but the margins could be razor-thin and Kelly Loeffler is a uniquely awful candidate (not to mention how appointed incumbents in general see less of an advantage than elected incumbents). Warnock also tends to perform slightly better in polls than Ossoff. I don't think there will be many Perdue/Warnock voters, but I think there could be some only-Warnock Democratic voters and only-Perdue Republican voters. Not many, but possibly enough. Since we're being accountable I would put my "the same party wins both GA Senate races" at 80%.
At the moment I'm probably thinking something like:
60% that Warnock beats Loeffler
40% that Ossoff beats Perdue
which I guess implies something like
35% Warnock/Ossoff
35% Loeffler/Purdue
25% Warnock/Purdue
5% Loeffler/Ossoff
And I'm only thinking this because the polls have shown the races as consistently close, with Warnock almost always leading and Ossoff often a little bit behind.
I don’t understand the optimism here or in Matt’s forecast. The 538 and RCP polling averages have both races essentially tied. Why isn’t all the rest just pointless speculation?
Polls aren't always right and whether party labels trump candidate differences is relevant to campaign strategy.
Yes but, as an outside observer, Georgia is a state normally won by Republicans where one Democrat won by narrow margins and the polls show a tied race. What information am I missing that someone who thinks Warnock and Ossoff are favored has?
While it’s true that republicans usually win Georgia, it is also clearly trending towards Democrats at a rapid rate (sort of the opposite of Wisconsin) and is now only 4 points more republican than the country at large.
Whether you think Dems will win the senate races probably comes down to whether you think the electorate will be more or less democratic than in November, and it’s easy to argue that Dems will actually have the advantage in a lower turnout election because their increasingly educated base is more likely to turn out.
Good point, I was thinking more about the prospect of a split decision vs. either party winning both seats. I would actually give Republicans the edge here, personally.
Isn't there the thing where voters need to register *again* to vote? Anyone know how this breaks down among Dems and Republicans? Who are more excited to vote?
No need to re-register, but if voting absentee, you need to submit a new ballot request.
Here’s a possibly dumb question but I don’t know statistics very well so I’ll ask... how do you check yourself on predictions where your confidence levels aren’t the same? If you have 10 predictions at 60%, that’s easy, you described it. But if you have 2 at 85%, 3 at 70%, 2 at 50%... is there a way to to combine all of those or do you just need to wait until you have enough of a sample size at each confidence level?
The method of "calibration" that is being discussed here only works if you have large numbers of predictions at each confidence level. Another method that is often used is the method of "scoring rules" (https://en.wikipedia.org/wiki/Scoring_rule), where you figure out whether each prediction was true or false, label those with 1 or 0, and then sum up some function of the difference between the prediction probability and the label (the two most common ones are either the square of the difference, or the negative logarithm of the difference). While someone can be perfectly calibrated if they give 50% as their prediction for 100 things, 50 of which are true and 50 false, the scoring rule only sums to a perfect score of 0 if you give 100% as your prediction for all the true things and 0% as your prediction for all the false things. The scoring rule gives you an incentive to aim towards extreme values if you are in fact very confident in that prediction, rather than hedging with a bunch of predictions at the same value. The difference between the quadratic score and the logarithmic score (and other rules) is based on whether they want you to be more careful about your predictions near the extremes, or near the middle of the range, or somewhere else.
Thank you, this is very helpful.
"Twitter is full of bullshit" is an amazing sub-heading. But I view cable news punditry as the true home of the unaccountable crap prediction - tons of air time to fill, talking heads saying whatever they want, no one ever checks except rarely the Daily Show.
Twitter's it's own thing - it's performative group-on-group public fighting. The incentive isn't just to be wrong in a provocative way, it's to dunk on the other side.
Hi Matt, great to see you doing this. Here are some ideas based on Scott Alexander's experience:
* it's easier to check the calibration of your predictions if they're all multiples of 10. Most people can't distinguish between 21 levels of confidence anyway, so using numbers ending in 5 could simply make your work harder
* it will help check your calibration if all the probabilities given are 50 percent or higher. Your Lakers prediction could be rephrased as a prediction with 75 percent confidence that someone other than the Lakers will win
Superforecasting is a great book, and I agree more people should read it! Thanks again, and I look forward to seeing how your predictions turn out.
On the Lakers one, I think it matters that this is a case where there are several dozen possibilities, and he is listing the one that he thinks of as by far most likely, even though it's still less than 50%. If he had a bunch of predictions of this form, then it probably is more helpful to test calibration on the low probability specific predictions, rather than the high probability non-specific predictions.
I think the thing that surprises me most in this list is how bullish you are on the restoration of the JCPOA.
What's your thinking on Iran's incentives to rejoin at this stage? I have to imagine their internal pressures from the hardliners but also from just the general populace after the dual assassinations this year will make that difficult politically even if the leadership was so inclined, which also does not seem clear.
I had the exact same reaction. 80% for rejoining and compliance seems really high to me. I would say rejoining is probably around 80% and then actual compliance given rejoining is probably around 60-70% putting it more in the 50-60% range jointly.
I agree that most of the discussion around the attempted "coup" has been awful and hysterical (much like everything with Twitter) and I know this wasn't the main point of your post, but there were a number of prominent writers (Zeynep Tufecki in particular) to point out that Trump was:
a) Intentionally disrupting the mail to make it harder/more confusing to mail in ballots in states like PA
b) Having local Republicans insist that mail-in ballots be counted AFTER Election Day, thus making it seem like he was ahead (again, in PA)
c) Leaning on judges he appointed to stop the counting of mail-in ballots
And the election was pretty close in the Electoral College! 45k votes in GA, WI, and AZ and Trump wins! The point was that, had it been one state (or even two), Trump might have found more sympathetic judges to flip the count for him. But with so many states, it would've been politically untenable, and the smart judges knew they couldn't do it credibly. I think it's more than fair to say those fears were NOT overblown, and had Trump been successful, it would've constituted a coup (albeit not a military coup).
"...Trump was:
a) Intentionally disrupting the mail to make it harder/more confusing to mail in ballots in states like PA"
Is there any evidence for that at all? (And, BTW, Pelosi's and Biden's lies do not constitute evidence.)
DeJoy's testimony before Congress about dismantling mail sorting machines? Postal workers not getting paid overtime despite increased volume, as was standard practice? Increased delays in delivery times? I could go on... there's tons of evidence, none of this is even remotely controversial outside of the right wing blogosphere.
"DeJoy's testimony before Congress about dismantling mail sorting machines?"
You mean the testimony where he said, "your accusations are actually outrageous"? That testimony?
The entire USPS/Election kerfuffle was nothing more than a wild conspiracy theory.
https://www.businessinsider.com/usps-not-possible-to-restore-dismantled-mail-sorting-machines-2020-9
What, specifically, is it you believe that story demonstrates?
I think a) and b) are backfilling conspiracies into unrelated events. B) definitely is, since it wasn't expected mail-in ballots would skew so heavily Democratic (otherwise the GOP wouldn't have been in support of the bill when they passed it); the rules for ballot counting were set well in advance of election day and there's no real evidence Trump was involved or interested in that process; and lots of states have slow counting processes for a variety of reasons - Pennsylvania wasn't even particularly late in certifying results. The second most likely tipping point state, Florida, being the literal first state to report substantive results doesn't help the theory either. Why weren't they let in on the anti-mail conspiracy?
A) I guess is still open to interpretation, but my takeaway from the DeJoy hearing was that he's a standard efficiency consultant who thought he could come in and twist a few dials (Send the trucks out on time! Cut overtime!) and make the USPS magically profitable. Maybe there was a deep conspiracy angle there but the Occam's Razor of DeJoy acting exactly like you would expect a Republican political appointee to act is much more plausible.
C) is actually concerning, but lazy, ineffectual corruption is so baked into Trump's persona that I don't see how that ever produces a different outcome.
It does seem to me like DeJoy was just innocently breaking everything, but it could be that the people who hired him knew he’d do that.
I still want to know what was up with Rex Tillerson coming in and apparently destroying the State Department with management consulting. Is there an accurate history about this yet? Did Putin actually get him appointed over Mitt Romney?
"It does seem to me like DeJoy was just innocently breaking everything, but it could be that the people who hired him knew he’d do that"
The people who hired him knew they needed someone to dramatically cut operating costs. That's what he was doing and that's all he was doing.
There is absolutely no evidence of a conspiracy to have the USPS disrupt the election.
I thought that by August or September, it was clear that everywhere that mail-in ballots were an easy option, they would strongly skew Democratic, just because of the observed trends in all other walks of life during the pandemic, where in-person versions skewed Republican and home versions skewed Democratic.
Mail-in voting was a pre-Pandemic measure that passed with bipartisan support in 2019. The PA GOP couldn't have predicted covid would drastically increase it's usage, or that their president would directly tell people not to vote by mail, but the law as written could not have been part of a conspiracy because it was crafted under a completely different environment.
Not sure how dismantling mail sorting machines constitutes increased efficiency. There's no reason to give him or anyone else in the Trump administration the benefit of the doubt.
And Trump's strategy was pretty clearly to bank on PA being close and have everything else break his way (which it almost did). Why else would their entire army of lawyers descend on Philadelphia? If FL was lost, it would've been a landslide as Matt had said.
IIRC the argument was they were replacing the mail sorting machines with package sorting machines. The case behind it was that first class mail has been decreasing for years so they don't need mail sorting machines as much. In contrast, package delivery has been increasing and they needed more automation. I think there was also a lot of push back from the union fighting his efficiency program broadly and especially his reduced OT plans.
Obviously what Trump is doing is bad, and what other Republicans are doing/not doing is bad (and probably worse), but I've been against the "coup" description for several reasons. First, asking Courts to do something, which is most of what has actually occurred (rather than merely been discussed), doesn't seem to be what is usually meant by the word coup, but more important has been what is implied: no-one who has called this a coup really seems to acting like it is one.
If this was a coup, what would that imply people who oppose him and it should do? Should there be a counter-coup? Is violence not just an available option, and justified, but wouldn't a coup be so bad that perhaps failing to engage in some level of violence in response to it is actually bad? If there's a coup happening, why is the only response to it tweeting and op-ed-ing instead of rounding up coup supporters to stop the overthrow of democracy? etc.
I am happy to say I don't think this is a coup, and don't think violence is justified to stop it, but how do people who think this is a coup justify taking the same actions in response to it that I am (waiting for the courts to reject idiotic claims, waiting for the electors to vote, waiting for Congress to count the ballots, and waiting for inauguration)?
Violence isn’t justified because no one has expected the attempted coup to succeed. If we thought it had a chance, we’d be out in the streets. Courts have firmly stomped on lawsuits attempting to overturn the election. Military leadership has firmly stomped on any thought of becoming involved in the choice of president. Filing ridiculous lawsuits isn’t illegal, so mass arrests are obviously unjustified.
You put into words something I've been mulling over a for a while.
I think it's clear at this point that responding to Trump's behavior _as if_ it was a real coup would have had a worse outcome...there's still a disturbing amount of belief in Trump's claims, but we now have a court record of this stuff getting the door slammed on it, and that legal process ran without interference...a bunch of left wing "anti-coup" behavior would have fueled the fire.