Main Edition
US
FR
July 21, 2020 - 10:55am
Reading the latest copy of the New Yorker magazine, published exactly a week ago, I came across this sentence in a piece by Jill Lepore:
This in a 5,000 word feature on the history of policing in the United States, which draws a link between the early role of police in suppressing slave rebellions, and police killings of Black Americans in the twenty first century.
This sentence jumped out to me. How could it possibly be true that ‘two-thirds’ of all Americans aged 15-34 visiting emergency rooms had been injured by police or security guards, given the very many other reasons why people might present for emergency treatment? In the online version, there is no hyperlink to the research (although the article does contain hyperlinks), and the study’s authors are not named.
Jill Lepore could hardly be more eminent. She is a professor of American history at Harvard, the recipient of a long list of awards, and a longstanding staff writer at the New Yorker, as well as a contributor at many other well regarded publications. I love her writing, so much so that I bought several extra copies of her latest book These Truths to give as presents to friends and family. Given this, I thought at first that I might have misunderstood the sentence, and tweeted as much.
I sought out the study she was referring to, and found it: a 2016 paper, whose lead author, Justin Feldman, was a doctoral student at Harvard at the time. Soon after publication, the findings were described in a Harvard press release, and also reported on by The Guardian.
And it turns out I was right — the ‘two-thirds’ claim is not true. Not even close.
Lepore is right to draw a comparison between the rate of ‘legal intervention injuries’ (to use Feldman’s phrase) and the rate of pedestrians injured by motor vehicles, although this only applied to men aged 15-34.
But it’s not clear where Lepore got the ‘two-thirds’ figure from. Possibly she misunderstood a line from from the paper itself, which includes the finding that 61.1% of people injured by police fell into the 15-34 age bracket. Or from the Harvard press release, which reports that:
Join the discussion
Join like minded readers that support our journalism by becoming a paid subscriber
To join the discussion in the comments, become a paid subscriber.
Join like minded readers that support our journalism, read unlimited articles and enjoy other subscriber-only benefits.
SubscribeFrankly, this is why I have given up on mainstream media. I now use BBC only to follow the cricket and Bloomberg for stock prices. Any content is biased garbage.
Yep, me neither. BBC for the market data page and the weather forecast. Not for news and certainly not for opinion. And have stopped following the cricket after the team ‘took the knee’.
It’s a sad day when the weather forecast is the most reliable part of the service
I still subscribe to the New York Times crossword puzzle. It’s the only part of the paper I believe.
The crossword staff is all white. Expect it to die soon.
The crossword itself is black and white though, so perhaps the staffing is not quite so “problematic” that anyone will have to “take a knee” or be “cancelled” or “erased”.
I bought it on Saturdays for that and for the cartoons. Finally I quit because that alone wasn’t worth the money. Especially since they ditched the bridge column.
Unfortunately the electronic edition makes the crossword an extra cost. But I get the Sunday version in my local paper.
This is not an example of “journalistic bias.” It is an example of deliberate disinformation and it serves as additional evidence of the total corruption of “scholarship” that has taken place in the Ivy League schools o the US.
“How could it possibly be true that ‘two-thirds’ of all Americans aged 15-34 visiting emergency rooms had been injured by police or security guards,”
I can only assume the injuries were not why the men were at the emergency rooms, but just reporting the fact that at some point in their lives these young men had been ‘Injured’ by police or security, possibly by a harsh look or suspicious glance, or even an accusing posture. Youth are easily hurt by these microaggressions. Some scars only show on the inside.
You could have ruled out that claim right away, once she said that about the same number (2/3) are injured by motor vehicles. 2/3 plus 2/3 after all, adds up to more than 1. So the only way both claims could be true would be if police officers routinely injured people by hitting them with their squad cars!
You can’t rule it out like that (because “Americans between the ages of fifteen and thirty-four treated in emergency rooms” is not the same population subset as “pedestrians injured by motor vehicles”).
Edit
?! Let me try putting that more clearly: your set-counting argument doesn’t work because [the subset identified by] “pedestrians injured by motor vehicles” includes pedestrians younger than 15 and older than 34. So – obviously – it can amount to more than 1/3 [times the number of “Americans between the ages of fifteen and thirty-four treated in emergency rooms”].
That was the first thing I thought too. Having spent a lot of time in emergency wards in Toronto – a sick relative- I learned that a shocking 2/3 of patients were cycling accident victims. There were also a LOT of alcohol-drug related issues
Unfortunately the media also use bias by omissions as well as inaccuracies. Today the Guardian claimed an exclusive that the UK test and trace was in breach of data regulation and continued through a lengthy article. One…it wasn’t exclusive, the BBC ran it on their website yesterday and two…..they failed to report that the DHSC was engaged with the ICO on the issue and had been for a while having reported themselves for not having the high risk documents in place. MSM whether it be broadcast, online or print is so often driven by bias that it isn’t worth the candle.
I think this might be a fine illustration of Hanlon’s Razor: “Never attribute to
malice that which can be adequately explained by stupidity.” It is perhaps
possible that Jill Lepore and the various editors at The New Yorker may be
victims of another global pandemic: innumeracy. This sorry state of affairs has been completely unmasked by Covid-19.
So top-shelf journalism is now stuffed to the gills with Ivy League graduates but they don’t understand basic statistical analysis?
Wouldn’t you have to score something in the neighbourhood of at least 1500 on the SATs to get into Harvard or similar these days?
Depends on your protected characteristics. What Harvard giveth to some they also take away from others.
Gender studies grads. No surprise.
There’s something to that in that I doubt that Lepore deliberately included a bad statistic, but Perry’s point still stands because it seems that a wildly disproportionate number of these errors go in one direction.
Innumeracy is part of it, but what compounds it is that people don’t bother to fact check even outlandish numbers when they confirm the person’s pre-existing views.
Thanks for your article….I’m also a compulsive fact-checker. Disagreement is debate, a view opposing the facts is a nasty form of bias.
This kind of stuff is why people who it never occurred to assume falsehoods in MSM are now moving closer towards the opposite position. It blows my mind.
I hope that the heart of the issue here is simply they never thought it sounded wrong, so never checked the fact. But this is the difference between 2 people in a 1000 vs 666/1000 of a huge part of the population.
Have they highlighted a clear retraction?
That in itself is seriously alarming. The claim is so obviously untrue that one’s world-view would have to be radically skewed in order for it not merely seem plausible, but so obviously correct that it didn’t even require checking. And this skewed world-view must have been ubiquitous to every individual – writer, editors, sub-editors – who saw the piece before publication.
Spot on. It’s in fact the worst thing about it.
It is worrying some sub-editor may have been looking at the 2/3 and thinking ‘only that many?’
Exactly It is the most disturbing aspect…no person thought to question
such an alarming number ..one just needs to talk to someone who works in the ER
The exact same misinterpretations of data/death figures are being demonstrated and quoted by the intelligentsia in relation to covid-19. The fact that the elderly are largely the only people at risk of death from it seems completely lost on many of them, as they produce article after article that contain no mention of age bias at all, implying all age groups are equally at risk of severe illness or worse. It’s appallingly irresponsible journalism, but it mostly goes unquestioned as it plays into the level of fear required to allow governments to override civil liberties, presumably.
The reason this absurd statistic made it to print isn’t because Lepore and the New Yorker staff found it “plausible” enough to not subject it to any fact-checking; it’s because it fit the progressive vision, which is underpinned [irony intended] by the cynical musings of mid-20th century French intellectuals who dismantled the very idea of objective truth itself.
I don’t believe for a second that this was an honest mistake where several professionals who do this for a living missed something this glaringly half-baked. I find it odd that these kinds of oversights always benefit one political side, instead of having agnostic disbursement, which is what you’d expect if they were “honest mistakes”.
“Facts don’t care about your feelings.” Somehow this statement”more of a concept”which should be the manifesto of anyone purporting to be a journalist, has been pushed to the periphery as a right wing slogan, and its inverse has taken its place.
this is why, after years and years of subscribing to The New Yorker, I no longer do. I could tell by listening to the New Yorker Radio Hour that all I was getting was a more literate rehash of the same old bias that I had already read in the MSM. A sad commentary on where we are these days.
The MSMSM seems determined to create civil unrest, where they once lied by omission now it’s direct untruthfulness, as for Harvard, as long as the monies rolling in they simply DGAF.
As an ER doc, I can tell you that my experience agrees with your data, not Lepore’s. It is hard to believe that someone as eminent as her could be so ridiculously, bogusly wrong.
Actually it is not hard to believe at all that someone of her supposed ’eminence’ could be so wrong. Most ’eminent’ people seem to be wrong pretty much all the time, and hopelessly biased to boot.
Whatever, one has come to expect nothing less than lies from the New Yorker and NYT etc. It’s a shame as I once loved the New Yorker. Today I glance at the film reviews and that’s about it, so absurd and biased has it become.
I have been a New Yorker subscriber for more than half a century, Since the nomination and election of Donald Trump, I question continuing my subscription with each new issue. Jill Lepore’s supposed “mistake” was deliberate, and is right in line with the “liberal elite” attitudes and biases relfected in each new issue. It’s just another jab at the President and his “basket of deplorable” supporters (like me). It is right in line with the New Yorker’s embrace of the “climate change ” and “carbon footprint” stupidities. I don’t believe for a moment that it was accidental.
In today’s world Lapore might be considered “brilliant,” but I would point out that no one ever felt it necessary to apply that or any similar adjective to the writings of James Thurber, Pauline Kael, Roger Angell, E.B. White, etc.
I read my parents’ NewYorker as a child and ended my subscription when Tina Brown came aboard and intelligence, urbanity, and wit disappeared from the magazine. It’s now the opposite of what it used to be: it’s now tedious, humorless, and my ultimate insult: boring.
Well done, it takes great courage to take down one of our heroes from their pedestal (if only temporarily). If only all of us could read so critically and with such lack of bias (and had the time to do so).
To think (and write) that 2/3rds of ER visits of Americans between the ages of 15-34 are due to injuries caused by law enforcement (including security guards) shows that Ms. Lepore has absolutely no grasp of reality. IMO, this is a telltale sign of the woke. Facts, truth, and common sense don’t hold a candle to their ideology, false narratives, and propaganda. I’m not saying the right is any better, but given that America’s new woke army includes professors, CEOs, professionals, the media, Hollywood, and other presumably educated people, it’s a huge disappointment that the left seems to have abandoned truthfulness as a core value. We need a new political party in this country.
“I’m not saying the right is any better”
It bloody is.
It’s a sad state of affairs that we have allowed this to regress to this point. The very publications that used to be the standard bearers are now leading the way in misinformation. And much of it is because the narrative is now considered of more value than the truth. And the truth is considered to be anything you want that moves the narrative forward.
Here are some facts never aired in the MSM. In 2018 Of the 2600-odd black murder victims 2200 were victims of other blacks. There were 9 killings of blacks by police. Mr George Floyd was a convicted felon who, with his confederates, invaded a black woman’s house and looted it whilst Mr Floyd pointed a revolver to her (pregnant) belly. He was on fentanyl at the time of his arrest. He had heart disease at the time of his arrest and died of a heart attack probably due to his drug use and the stress of his arrest. The arresting police failed in their duty of care to someone in their custody. (The last item is the only one reported).
I believe it is 9 unarmed Blacks killed by police…
All true, except for the famous New Yorker fact checking department. Yes, they have a very good fact checking department. But I can think off the top of my head of two instances, one non-consequential, one consequential, where fact checking utterly failed.
The latter case should be more famous. Jeffrey Goldberg, now the editor of the Atlantic , wrote an article in the run-up to the Iraq invasion that claimed that Saddam Hussein and Osama bin Laden had some tie. https://harpers.org/2008/05… The Great Terror was almost comic in feeding Cheney-style propaganda to the U.S. public – which may be why the white house cited it at the time – and Goldberg and the Nyorker have never apologized.
Less consequentially, back in 2000, Robert Friedman wrote a comically bogus scare book about the Russian Mafiya= Red Mafiya – an excerpt of which was published as an article in Tina Brown’s New Yorker. Talk about bad sources and tall tales! I reviewed the book myself in a now defunct magazine and marveled tjhat Friedman could tell a bunch of unsourced and implausible stories and get away with it simply because of the American fear of Russia. That fear has only gotten worse over time.
I should say, I’ve written brief reviews for the New Yorker and their fact checkers called me up and were very nice. But sometimes, they let through amazing howlers.
Yes people’s biases and preconceptions make a fertile bed for misinformation. It is easy to do….did you realize that Canadian sales of cosmetics are equal to the entire education budget of the PRC?
A bit of basic arithmetic tells anyone with half a brain cell that the figures are made up! 2/3 beaten by cops and 2/3 hit by cars is over 100%, unless 1/3 were beaten by cops then hit by a car or hit by a car then beaten by cops.
It does seem strange how the US Left has turned on its police – all those US cop shows and films portraying them as heroes (often flawed – go ahead make my day, do you feel lucky punk) keeping ordinary citizens safe from the all pervasive criminal fraternity seem to have been forgotten. What is more worrying is how those misconceptions and lies about US police have so readily transferred into UK even bigger misconceptions about our always much friendlier and more restrained bobbies on the beat.
It did not say the men were in the emergency room Because they had been beaten by police, just that they had been hurt by police at some point, and who knows why they were there. Possibly the indignity of a stop and search many years ago still lingers, and so the article may well be totally true. The writer above was the one leaping to conclusions. ‘Eats Shoots and Leaves’ stuff. I find the original article fact-checks as good, depending how it was meant.
Laughably unintelligent comment.
I think the misunderstanding or the article has more to do with a poorly written scientific paper than with some grand intrusion of political bias.
I’m very familiar with the literature on motivated reasoning and dual-process heuristics, so that interpretation is possible. But I ask that each of your readers really attempt to understand the original research paper, and see if they find it clear on this point.
Early in the paper, the authors state that “For the period 2001″“2014, 683,033 (95 % confidence interval “¢”¢”¢) legal intervention injuries treated in EDs occurred for US persons age 15″“34″¢”¢”¢”
But, in the Text describing results Table 2, they state “Time trends for legal intervention- and assault-related injuries treated in hospital
emergency departments”nationally representative estimates for US men and women ages 15″“34, 2001″“2014 (N = 683,033 injuries).”
So, in one place 683,033 equals the legal intervention injuries, and in the other, 683,033 equals the total of legal intervention and assault injuries. It’s pretty hard to come up with a solid understanding of what the statistics even mean.
And I have no idea how the Harvard Press release came up with their 64% ratio of cop-inflicted vs total assault injuries. I don’t think it’s correct.
I do agree with this critique that one would think that the the original Jill Lepore NewYorker statement was implausible on its face. They New Yorker troupe definitely needs to spend more time out of the office to get a read on reality. But if they tried to do due diligence and read the scientific paper, they may have just gotten confused, and I find that more plausible than biased intent.
So, if you are confused by a piece of research, as an author, you just wing it… rely on it anyway, aided by a sketchy press release? The research should be well-understood and verified to the extent reasonable or it should not be used. Where are you people from who don’t understand the basic rules of scholarship?
If we use the website at reference 17 of Feldman (2016) then we find that it does indeed indicate 683,033 legal intervention injuries for ER visitors aged 15-34, as the paper says. Some 64% of those injuries were coded as “Struck by/against”. So far so good.
But it reports 148,241,544 for all injuries in that age-group. So the legal intervention injuries constitute a little under 0.5% of all injuries in that age-group. The conclusion of the paper that ” legal intervention constituted an important contributor of visits for injuries treated in hospital EDs among persons age 15″“34″ is not consistent with the dataset used.
The press release says “Sixty-four percent of the estimated 683,033 injuries logged between 2001-2014 among persons age 15-34 resulted from an officer hitting a civilian.” but this is incorrect, since there were 17,420,046 inujuries coming from violence: as I said, this should read that 64% of the injuries logged as legal interventions resulted from an officer hitting a civilian.
Undergraduates in CS would flunk if they were studying data science and made any sort of mistake this egregious. Nobody with half an analytical mind would not consider this about policing and excessive force without first understanding how many arrests are made as a subset of detentions as a subset of encounters.
How on earth can any organization with a ‘city desk’ and editors who know their police departments get away with this? They’ve ventured too far away from the math.
Actually, the sentence should have been flagged by the fact-checkers because it is mathematically impossible. According to Lepore: “two-thirds of Americans between the ages of fifteen and thirty-four who were treated in emergency rooms suffered from injuries inflicted by police and security guards,” and “about as many people as the number of pedestrians injured by motor vehicles. ” That’s 4/3 of all injuries treated in emergency rooms for those two causes. Shouldn’t that figure have raised alarm bells . . . ?
That was the very first thing that popped out to me as a red flag. Nice to know how to do basic math, isn’t it?
Sure inumeracy and overall low intellect are part of the issue – its more c**k up than conspiracy but does show something more than confirmation bias – the need to believe police etc are on balance negative + violent trumps the facts. Its pure ontology – like Daesh’s religious fervour. No testing required and falsification merely hardens the zealots resolve. They are always with us and are blooming right now but i agree with Mr Pinker that in the long run their power is on the wane. One of the few wise things Obama said was progress is not linear. I would add that at least its possible, except in the case of Jill Lapore, the KKK and DAESH to name a few of the usual suspects.
So isn’t it the Harvard press release that is wrong?
This just seems like a case of a very common lack of understanding of maths and statistics – not just among journalists.
“Sixty-four percent of the estimated 683,033 injuries logged between 2001-2014 among persons age 15-34 resulted from an officer hitting a civilian.”
does this mean 64% of ‘police-caused’ injuries or what? confused. also what are the alternatives to hitting?
Yes, I didn’t understand that bit either. It seems to substantiate the claim the article is challenging.
Yes didn’t quite get that either.
On further thought I think it is of all police-inflicted injuries (i.e. the 683,033) 64% were from officers hitting. The rest of the injuries are presumably from other methods of deliberate or inadvertent injuries caused by police.
Where the confusion is that the article refers to police injuries (15-34) then jumps back to vehicle injuries (15-34) then back to police without distinguishing well.
This is really confusing! I read it as saying: Emergency room data shows that roughly 683,000 injured people (not just men) between the ages of 15-34 went to emergency rooms between 2001 and 2014. Sixty-four percent of them, or 437,000 of those people, were in the emergency room because they were hit by police officers. That also would mean that only 246,000 injuries occurred in this population from all other causes. That’s an insane proposition. But not even that—if that many injuries resulted from (only) being HIT by police, then obviously there would be even MORE police-caused injuries than those 437,000. There would be lots of sprains and abrasions from being wrestled with. There would also be injuries from falls, tasers, tear gas, and shootings, to name scenarios off the top of my head.
Interesting. Is this not just a story about a bigoted liar.
Even if I was Left-leaning hook, line and sinker, the stat in question is so out of whack I would have taken it as a typo and moved on with my reading. I sure as heck would not use it in any supporting argument in favor of defunding or disbanding PDs. Stupid of the author to write it and invincibly ignorant for anybody to pass it along as fact.
Send a letter to the New Yorker via both Certified Mail & email: [email protected]. Demand a correction be printed, because the statement is both (1) demonstrably and massively wrong 0.2% compared to the claimed 2/3rds, and (2) defamatory to law enforcement personnel, and (3) incendiary and nearly as bad as falsely yelling “fire” in a crowded theater (which is illegal).
Send the letter to: Editor: David Remnick, New Yorker Magazine, 1 World Trade Center, NY, NY 10007.
The New Yorker strives for truth and is renowned and trusted for getting it right, but they are not perfect, unfortunately.
A few years ago they ran this story by Dave Eggers on Hollister Clothing.
https://www.newyorker.com/m…
Trivial, but also extremely untrue. Hollister Clothing is a fake surf and beachwear company dreamed up by Abercrombie and Fitch.
The name of the company is only indirectly connected to the city of Hollister, and has more to do with a surfing area south of Point Conception, called the Hollister Ranch.
I was pretty shocked by how wrong they are, and even brought it up with one of the New Yorker’s writers I met at Kelly Slater’s Surf Ranch contest.
So they might have screwed up on this story.
Although “One study suggests” is hardly carving that stat in stone.
Even the New Yorker makes mistakes.
I will say this about Trump:
1. He has made the New Yorker great again.
2. He has made Saturday Night Live Great Again.
3. Have Americans ever been more engaged in politics – good or bad – then now?
Simple case of another lying liberal-lefty with an agenda. They’re dangerous.
This is why UnHerd exists.
It is truly disgusting that neither the elitist professor nor the magazine have issued a correction. It would undo the spreading of the error but it would be honorable and also, hopefully, stop it from being cited further.
So I see where you marked up Professor Lepore’s error to political bias. And I see where you got in touch with the author of the study she misinterpreted. What I don’t see in the article is that you actually contacted the professor to find out what the hell happened. Did you? Because I think that’s something I’d have done before leaping to a conclusion about the professor’s private motivations, then making heavy weather about them.
Dear Louise, weren’t you biased here as well? You wrote a whole article discrediting the magazine based on one incorrect sentence (thou it is critical, it doesn’t invalidate the magazine, the article or the author).
I love your work but this one reminded me of internet trolls that in a lack of arguments pick on the first error they see (grammar errors or a wrong data) to discredit someone – of course a much more elaborated and pleasant to read troll – I know you are no troll, but it reminded me of troll methods.
Again, I agree it was a critical misleading sentence but still it does not invalidate the magazine or the article. I read other articles and of course she has an humanitarian agenda but I doubt she did it with misleading intention. She is also a scientist and we usually hate getting facts wrong or propagate fake news.
I understand that its a “fact check” piece but it seems a stretch to generalize it to whole magazine or article as untrue – generalization bias.
Thanks for reading and I will love to hear different thoughts on this.
Best,
Joao
I disagree. A mistake of that magnitude certainly does discredit that article (and as she intimates, to an extent the magazine), especially in the present climate which is so hyped up in regard to this issue. And she goes to great lengths to put the author in the best possible light – to the point of steel-manning. I’ve never witnessed a “troll” perform such a reasonable, self-defeating feat.
Thanks for answering, Poyton.
I still think it’s a disproportionate call to discredit a 60 paragraph (5000 words) and a magazine because of a wrong sentence. Specially when the specific wrong number being corrected or omitted is not key to understand the whole article – specialy in the present climate where we shoild be debating and not canceling ideas.
I read the article and the paper. She used this data to support her evidence but she uses lots of other data that Louise or anybody here would never be able to fact-check as wrong (because they are not). What I mean is the thesis of her article is not essencialy wrong because of those wrong numbers. One can disagree on her conclusions and opinions (just like you and I are doing here) but it’s a bit of a stretch to get to the exact opposite conclusions because of a sentence with wrong data (that’s trolling methodology in my opinion).
And yes, I like Louise writings and she is very polite (I wouldn’t say self-defeating) but still tried to discredit the author, the article and the magazine because of a sentence in a 60 paragraph piece.
Well, the New Yorker used to have a reputation for fact-checking. So one egregious mistake like this is a blow to that reputation. But let us suppose that the single mistaken sentence had been “One study suggests that two-thirds of Americans between the ages of fifteen and thirty-four who were treated in emergency rooms suffered from injuries inflicted by …” and that the mistake had been to end it with “Joao Eduardo Daud Amadera” rather than “police and security guards”. Would you be equally relaxed about that single mistake?
But we tend to take all the claims in the article on trust, because we haven’t the time and resources to fact-check each of them individually, and, as I say, we tend to trust the New Yorker’s reputation for aggressive fact-checking. So if the one thing that catchers our eye as seeming as if it can’t be right, on checking, turns out not to be right, what credibility do we place on the remanining assertions that we haven’t checked. You refer to “data that Louise or anybody here would never be able to fact-check” but then assume that they are right. How can we be sure about that, and why can you?
Thank you Richard for taking the time to explain something which I didn’t have the energy to do and couldn’t have done so clearly anyway.
Thanks for your extensive reply, Joao. It’s a fair point and you seem to have gone into the background of the article to a much greater extent than me, so I will give credit on trust to your viewpoint. However, I would refer you to Richard’s reply to your comment – just what I wanted to say (although I thought it would be obvious to most).
For the record though: the poverty and other hardships that black people face in the US and the rest of the West have very little to do with police brutality. They are way more deep-rooted and intractable than the cynical BLM (ie. regressive left) movement is selling to us.
Btw – you can call me Dan. Apart from your lack of an ‘n’ butchering my name, and it being rather anti-social to call someone by their surname in the Anglo world, I’ve always thought my last name was pretty weird anyway (handed down by my rather crazy Irish pioneer ancestors or something).