Hacker News new | comments | show | ask | jobs | submit login
A Basic Cohort Test of the Lead-Crime Hypothesis (motherjones.com)
49 points by curtis 6 months ago | hide | past | web | favorite | 45 comments



This appears to be a case of "I have my preconceived interpretation of the data, and I'm not going to see how well other hypotheses match the data."

The main thrust is that the cohort test supports the Lead-Crime Hypothesis because, on average, 70s kids were more violent than 80s kids were more violent than 90s. But later in the post, he mentions that there's an important confounding factor: the crack epidemic, which starts in 1987 and winds down in the mid-1990s. This would suggest that the increase in crimes over age for 1970s from 17 to 21 or so is fueled in large part (if we look at the later graph) by the crack epidemic. If you guesstimate that the trend should be flat-to-declining, then the trend line for the 1970s absent the crack epidemic is actually rather closer to the 1980s. Similarly the numbers for 1997 and 1998 (for 1980s) seem anomalously high, and should be lowered a bit.

Put it all together, and you'd estimate that removing the suggested effect of the crack epidemic would be that there is no substantial difference between the criminality of people born in 1970, 1980, and 1990.


This hypothesis is often put forth to in the US. The strongest studies of the lead hypothesis, however, compare data across multiple countries, which regulated lead-containing paint and gasoline at different times. These studies demonstrate that crime drops in each country at a similar period after the decrease in lead exposure, and at a rate that corresponds to the rate of decrease in lead exposure.


If you're looking for more reading on the Lead-Crime Hypothesis, Drum's "An Updated Lead-Crime Roundup for 2018" (https://www.motherjones.com/kevin-drum/2018/02/an-updated-le...) has a lot more information. Section 3 has some direct links to research papers, if that's your thing.


Undoubtedly lead exposure has had a profound effect on society, and on crime levels. Lead exposure leads to reduced IQ and increased aggression. But the crime landscape during the relevant period (the late 20th century) has too many confounding factors to extract data from it. For one, there's the "crack epidemic" crime wave of the late '80s and early '90s. During this period the rate of homicides committed by teens tripled and then went back to previous levels. Additionally, violent crime is correlated strongly with population density. And during the late 20th century in the US cities first began to empty out (due to "white flight") and then began to be repopulated as "yuppies" working office jobs and living in cities became more a more desirable "american dream" than the suburban family home and then later as urban centers became the core of the new tech economy. At the same time, police enforcement changed dramatically through the '80s, '90s, and later. The "war on (some) drugs (used by some people)", mass incarceration, etc. These are all tremendous confounding factors while the underlying "signal" from lead exposure is expected to be a gradual change over a fairly long period.


Separating out the confounding factors is definitely a big problem, but lots of effort has gone into just that. Drum has been following the Lead-Crime Hypothesis science since at least 2012. This blog post: https://www.motherjones.com/kevin-drum/2018/02/an-updated-le... provides an overview, but maybe not an ideal overview just because it's so long. Section 3 has links to some actual papers, and some of the other links may ultimately lead to papers, but there's so much of it that I didn't explore it deeply.


also the legalization of abortion, increase in food and medical subsidies for expectant mothers and young families, and the general rise in wealth.


“Lead exposure in childhood may have played a small role in rising and falling crime rates in the USA but it is unlikely to account for the very high percentage of the decline suggested by the ecological studies” [1]

[1] https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3829390/


I'm not sure why you're posting that link here. In academia, no single study is treated as conclusive. For example, when scientists observed a 50% decrease in men's sperm count in the last 30 years, they cited a meta-study which reviewed data from 185 other studies. [1]

If you look on google scholar. It's trivial to find well over 100 lead-crime studies. The one you linked to seems to be only published-online, and not a meta-study, so it's not clear on what basis you (cherry?) picked it.

https://doi.org/10.1093/humupd/dmx022


For me it's good to have a healthy dose of skepticism about these things, especially when it comes to social studies. I didn't like the conclusive tone of the mother jones article personally. The mother jones author does a fairly amateur analysis and concludes: "I’d say that basic cohort analysis supports the lead-crime hypothesis, while age analysis might not."

The article I posted is a counterpoint to this in which an academic cohort study is reviewed finding lead explains less than 1% of the variation in crime in a cohort, and another cohort study there reviewed found no association between lead exposure and murder -- so a relevant article to post here, I'd say.

Do you consider the results in the MJ article to be as robust as a meta-analysis and if not, do you have the same criticism of that article? (in which case, fair enough)


"Anyway, this is just a quick, amateur look at the data."

... Yes, yes it is. Statistics can tell you pretty much any answer you want to hear, if you are not careful.


Do people generally find mother jones a quality information source? it has always seemed to me to be a over the line into left-wing-nut territory - willfully misleading to appeal to their market.

I'd put it as basically a mirror of fox news on this chart :

https://www.codlrc.org/content/news-sources-journalistic-qua...


http://pressthink.org/2010/11/the-view-from-nowhere-question...

There are other, better, rebuttals but this is from someone who can't really be impeached on ideological grounds. Also, the idea that Mother Jones is "left-wing-nut territory" is ....bad. Unless your yardstick is "distance from more or less explicit white nationalism of breitbart" then Mother Jones is solidly center-left.

Finally the graph you posted is pretty bad. The Economist is an ideological chop shop with no bylines that started life as a forum for propoganda agitating against the Corn Laws. The idea that their minimally distant from an axis of "minimal bias" is not exactly born out by the facts.


My father is a liberal newspaper reporter and describes The Economist as the best hope for journalism. If you have to stretch to smear it by pointing out the lack of bylines and its start over 150 years ago, then I have to wonder what your biases are. The Economist has its slant for sure but your characterization of it is extreme.


Not at all, in fact my characterization of it was pretty restrained.

https://www.thenation.com/article/economist-has-slavery-prob... https://www.politico.com/magazine/story/2014/09/economist-re... https://www.colorlines.com/articles/five-problems-economists...

I could add in its tolerance if not open affection for Fidez the anti-semitic Hungarian political party, or its virulently anti-union posturing, or its hatred of the welfare state....

I could go on, specifically just about their ideology and how it permeates everything thats published there, but its a lost cause on this website.


Those articles are not really convincing - two are regarding one book review, one criticizes the British reader for mispronouncing Spanish names (among other nit-picks). Those sources are very obviously biased (which is fine) and don't nearly constitute objective, or moderate, or broadly-accepted source of commentary.

If you were a regular reader of the Economist you would know how they actually look at anti-semitism (scornfully), labor-relations (respectful of labor, critical of populism and unchecked power), the welfare state (maybe conservative in the UK, downright progressive for the US - see anything they cover on healthcare).

Their ideology does permeate all of their work. The reason they are so well respected is that their ideology is well-reasoned, and their research and writing is very high quality.


there's always room for more nuance, but the widely understood left-right model is useful. add more dimensions as needed. on a crude left/right i'd put the economist near the middle: they are far from neutral, but also far from either of those edges.


This particular submission is a post from Kevin Drum's blog. He was a fairly well known blogger before he started blogging for Mother Jones, and his general political outlook hasn't really changed before or since. So regardless of how you feel about Mother Jones, Drum is kind of his own thing. I don't generally read Mother Jones proper, so I can't tell you how he compares to their regular reporting, but having followed Drum for many years I'd place him squarely on the center-left and no where near left-wing-nut territory.


Ever since he was doing Calpundit, Kevin Drum has gained a well-deserved reputation as a thoughtful, highly moderate center-left intellectual. Not even remotely extreme. Lead is his personal hobbyhorse, and something he actually knows a great deal about. So I wouldn't be too dismissive of him.


ah cool, I hadn't encountered him before. thanks!


This article doesn't seem biased to me, just some statistical information relating to the well known debate about why crime dropped in the late 90's. The lead gasoline theory is an interesting one, in addition to the ones given credit in Freakonomics. Here are some of the theories: https://www.vox.com/2015/2/13/8032231/crime-drop


From the random sampling of Mother Jones articles I've read I would agree - even the headlines have a definite left-wing bent and you can tell within the first few sentences that the author has a certain world view that everything is filtered through.

It's nore of a collection of editorials than actual news pieces.


Is there any news source truly in the middle, high quality and not low volume? I find myself having to read the left version and right version, then approximate the truth.

Your referenced chart has NPR in the middle, for example. I do enjoy them, but the left bias seems obvious to me.


Journalistic neutrality, or "the middle" as you refer to it, as it is practiced in American media has some serious flaws. Sadly, it still has cachet with a sizable portion of what has been sneeringly referred to as the "mainstream media" for the last few decades. I highly recommend reading Jay Rosen, a Professor of Journalism at NYU, talk about the "View from Nowhere." There was one link posted above, and I'll post another.

http://archive.pressthink.org/2003/09/18/jennings.html

http://pressthink.org/2010/11/the-view-from-nowhere-question...


I can make an argument that the neutral news media ideal from the golden age of mid-century reporting is what has led to the populist disaster we are experiencing -- and that it's really just a return to an old normal.

outside of the major news agencies neutrality got cargo-culted. the easiest way to be 'unbiased' is tell both sides, right? well, if one side is reasonable you have to go and find that opposing view to meet that (fair and) balanced standard. this gave a platform to and legitimized a host of formerly fringe ideas. all that noise and conflict sure helps meet the advertising numbers... and now we have fox news.

thing is, this polarized, biased, sensational media is what media has always been! with a rose-colored exception in conformist middle america starting in the long boom after the second world war. It's been gradually fading back to the old ways since the 70's.


I think that you're right about a return to the old normal. The press in past centuries was often unabashedly partisan. But I don't think that unbiased reporting is simply telling both sides. It's simply reporting the facts. No "sides." The problem with that is that reading the news is like reading meeting minutes. It's dry and boring. It doesn't sell.


Appreciate it. Am I off in thinking this used to exist decades ago? Reading, for example, the history of Watergate, it seems like there used to be more emphasis on neutrality.

Basically, that is, we lost something in the shift from printed press to online/social.


I believe the top publications today (NYT, Economist, WSJ,...) are far better than they ever used to be.

There was an article recently (no link because I'm on mobile, sorry) showing some of the factual errors that were published regarding Watergate at the time. None of the minor mistakes made by the major outlets with regard to the current Trump/Russia affair gets even close.

The practice of prominently correcting errors when they are found has also markedly improved. The internet is obviously a far better medium than print for corrections, as well.

It seems that people are expecting some sort of idealised "neutrality" from journalism today that is simply impossible. Journalism starts being subjective when decisions are made on what stories to cover. They have never been "phonebooks", simply listing facts, and journalists have always been more than typists: they are expected to call a spade a spade when they see it.

There's also a lot of criticism of "anonymous sources". If you're feeling nostalgic, I shall remind you that all of Watergate was build on an anonymous source. People seem to think of "anonymous sources" as a license for journalists to just invent facts. But that's completely wrong: the journalist's reputation is on the line every time they use such sources. If they are ever shown to have invented such a source, they would lose their job immediately. (As an exercise: read article from a year ago and see how many of those sources proved to be accurate).

The right measure for journalism is a sort of "good faith": there were untold numbers of editorials calling Trump out for claiming credit for the stock market's rise over the last year. And, sure enough, when the markets crashed last week, quality publications made a point of mentioning that Trump was just as irrelevant for the crash as he was for the rally before.


Regarding Watergate, there's a really fantastic podcast series entitled "Slow Burn" that just ended. It's 9 or 10 episodes long, and gives a great 'behind the scenes' look at the entire affair. I can't speak too much to the media's approach to Watergate.

> Basically, that is, we lost something in the shift from printed press to online/social.

This article from Harvard's Kennedy School of Government has some context on what happened: https://journalistsresource.org/studies/society/news-media/c...

One clipping from it:

One problem was the drive toward monopoly — or if not monopoly, then what economists call “concentration...” In the news business, this pattern threatens the very existence of diverse, and local, points of view. Clear Channel replaced more than one thousand individual owners, with different aspirations, different ideas about their civic roles, and different ideologies. The number of U.S. cities with healthy, competing daily newspapers, once in the hundreds, fell to about a dozen. By the year 2000, in American cities that had a daily newspaper, 99 percent had only one management in town. The fifteen biggest newspaper chains account for more than half the nation’s total circulation. In all, there were about 1,500 daily newspapers remaining in America. Of those, only about 350 were independently owned, and most of those were very small. The whole industry was heading in a direction that was at odds with the journalistic values of independence, localism, and competition.

The article also talks about the vast reduction in budgets for news outlets. This disadvantages organizations that do legitimate news gathering and reporting and advantages organizations that simply have an axe to grind (it's cheaper to spew your opinions than it is to put boots on the ground and see what's up).

There's more, but I'm probably getting excessively long for a comment.


That's helpful. I'm being a little flippant in print vs online, but there are obvious changes in the same time frame for other reasons, as you point out.


The history of Watergate, nearly 50 years later, might read neutral. The reporting at the time was hardly that. The press hated Nixon about as much as they hated Reagan and Bush and Trump. At the time there was no popular alternative to NBC, ABC, CBS so as an average TV news viewer you wern't aware of other points of view unless you actively sought them.


> Basically, that is, we lost something in the shift from printed press to online/social.

No, we lost it in the long period of print media consolidation and newsroom slashing from the 1970s through the 1990s; the effects were already palpable and the subject of much discussion in the 1980s. Things have probably gotten better since the turn of the millenium (possibly even since the mid-1980s or earlier) in terms of available high-quality news (OTOH, they've gotten worse in terms of the volume of low-quality, widely-distributed "news" that the high-quality stuff is surrounded by, and the availability of signals which let the average news consumer distinguish the former from the latter.)


Reason.com


That is clearly a libertarian news site. They may be negative towards both US parties but it's still a right-leaning site, at least in an economic sense of right and left.


The question is whether they're biased towards one side. I think they're much less biased than mainstream media on both sides.


I'm not saying they are being untruthful. But they are very clearly biased to the right on economic issues, and significantly more than traditional news. All the articles I saw on regulation and taxes were negative.

I probably even agree with them on many things, but they are _not_ unbiased.


Biased towards an economic position is not the same as being biased towards a political party.


Thanks...spent 5 minutes on a very brief flyover, but it looks promising. Lightweight site too, didn't notice excessive JavaScript or ads. I really am interested in better sources.

Maybe I'm naive, but there seems like there's a reasonable, if not really lucrative, opportunity for someone to produce actual unbiased news in the US. It's worth a few dollars a month to me.


> Maybe I'm naive, but there seems like there's a reasonable, if not really lucrative, opportunity for someone to produce actual unbiased news in the US.

Unbiased news is an impossibility. Centrist bias relative to the current position of the Overton Window is possible, but probably not economically optimal since political views aren't normally, or even unimodally, distributed.


There's a difference between news and editorial that has been increasingly lost. To give an exaggerated example:

----

- Air strikes in the US killed 15 targets in Mideastland. Representatives for the US armed forces stated the targets were high priority leaders of Al Boom Boom. Domestic news in Mideastland claims that the strike was on a hospital. There has been no definitive identification of the victims yet.

- The never ending war machine of the US continues to ravage Mideastland as a brutal attack on a hospital where many women and children were being treated has been destroyed with mass casualties reported. Here's a sad picture of a baby by some rubble that's not related to this incident.

- An elite fleet of aerial forces unleashed a precision surgical strike in Mideastland decapitating important parts of the leadership of the extremist terrorist organization Al Boom Boom. Locals were reportedly seen celebrating and toppling a statue of the fascist organization while waving US flags and singing the US national anthem.

----

There will, of course, be an unavoidable bias in what is covered since not it's literally impossible to cover everything. But in how things are covered, removing bias is not difficult and it's also testable. If an average reader can, with reasonable accuracy, state the biases and inclinings of the author of a news piece then it is biased.

The extreme bias in the news today is very much a contemporary thing. Here [1] are a list of online newspaper archives. Most are free and some go all the way back to the 1800s. Compare then and now, and the difference is quite extreme. I expect we're going through a time today that people will look back on with some degree of bewilderment, not dissimilar to how we might think of things like the two red scares we've gone through. We, collectively, seem to lose our ability to think coherently and impartially quite easily, but at least historically it tends to right itself quickly enough.

[1] - https://en.wikipedia.org/wiki/Wikipedia:List_of_online_newsp...


There is no "extreme bias" in today's top publications. To use something terribly close to your example:

> Dozens of Russians Are Believed Killed in U.S.-Backed Syria Attack

https://www.nytimes.com/2018/02/13/world/europe/russia-syria...

First paragraph:

Four Russian nationals, and perhaps dozens more, were killed in fighting between pro-government forces in eastern Syria and members of the United States-led coalition fighting the Islamic State, according to Russian and Syrian officials.

The article makes absolutely no judgement on the strike itself. It stresses the limits of information currently available, and quotes dozens of sources from all sides.

It then goes on to show, with lots of evidence, that such irregular Russian soldiers actually are in Syria. While people will probably latch onto this part as somehow being biased against Russia, it would be journalistic malpractice not to mention, for example, Crimea, where the exact same dynamic played out. I. e. denials of irregular troops being involved quickly being proven to be lies with Russia's official annexation of Crimea.


There is still bias in the form of choice of news to report. You may agree or disagree with that bias, and certainly some do a better job than others in reducing it, but it simply is not possible to avoid it, not least because many things are not objective.

E.g. "pro-government forces" is factual, but many supporters of the "members of the United States-led coalition fighting the Islamic State", which is also factual, will consider it pro-Assad bias to use such neutral language about both sides, and vice versa, exactly because it leaves out the background.

And there is not single, objective unbiased solution to that, even if you ignore that their very selection of what to report on also inevitably will be biased.

Personally I prefer openly biased sources, because then I don't have to deal with reporters pretending to be neutral while including or omitting information on the basis of biases anyway.


Yes, of course, if "it is factual, but people will call such neutral language biased" then there's mo argument here, and we can all go home and it doesn't matter if we have journalists or just read a random number generator's output.

But I was making an argument within the context of the parent, which was giving some absurdly biased "examples" and passing them of as something typical for today's top publication. And that's just not true.


Thanks for the article. Let's look over this since I see it rather differently.

The article states, "there are hundreds if not thousands of contract soldiers in Syria whom the Russian government has never acknowledged." According to whom and why? And if their estimate is so unreliable as to have an error margin on the order of magnitudes would it not be more accurate to state an unknown number? Anyhow, they not only take their controversial statement as a given, but then go on to offer a completely bizarre explanation for why they were deployed. According to the article, "They were deployed both to help keep the official cost down and to avoid reports of casualties, especially with a March presidential election in Russia fast approaching.". Again according to who? And idea that these secret soldiers are because of an election seems dubious, at best. Putin's approval rating is around 75% with 0 substantial opposition. His reelection is little more than a formality regardless of what happens in Syria.

Let's now look at some of their named sources for which you reference "lots of evidence." One bit of evidence was literally what a source, described only as "a woman from central Russia", said in "a brief online chat." And then they reference what they, again literally, describe as "investigative bloggers." And on top of all of this the article originally stated Russians were killed in the airstrike, when the source actually said Syrians. Regardless of whether or not that was a genuine mistake, it really should make you question their editorial standards for such a key fact to be night and day wrong.

There are many other such issues throughout the article, but let's stop there since I think the point is more or less clear. Look at it this way. Imagine this article was discussing an issue for which you had less personal biases, and from an outlet you also had no biases of brand recognition and trust towards. You would consider it to be dubious, at best. When we read things from sources we trust or that confirm our own biases, we turn off our ability to think critically.

-----

Granted the above is a tangential issue to bias - reporting quality. But the two tend to be strongly connected. When you want to push a story but the data to support such a tale isn't there, you have a choice of either moving on or turning to lower levels of support for your view. And in times past I think the choice there would have been dead obvious for the New York Times. Citing what somebody, who is again literally described as "a woman from Russia", wrote online is insulting the intelligence of your readers and instills a sense of incredulity for the article in the mind of anybody who's not taking what you write as beyond reproach. Nonetheless, they chose (and have regularly chosen now a days) to go down this path. And this is something new.


> There's a difference between news and editorial that has been increasingly lost.

Perhaps, but it doesn't change the fact that unbiased news is impossible; even rigorously bare-facts reporting expressed bias in the decisions of which facts are worth reporting, both on the macro scale (what events get stories) and the micro scale (what aspects of the event and reported.)

> The extreme bias in the news today is very much a contemporary thing.

No, it's not; the diversity of biases found in sources with wide distribution is (and it's particularly a change from the period of extreme media consolidation in the decades just before the explosion of online media), but the degree of bias is not.


I do not think there's very much diversity of bias in today's publications. It mostly breaks down into a very binary partisan split. You have 'liberal outlets' that mostly sound identical and 'conservative outlets' that mostly sound identical.

In the past there was less child-like black and white conflict, but there was vastly more disagreement on views and papers willing to publish things that fell outside the expected grain. The New York Times was a real leader in this regard. My favorite example is probably from an article from 1920. The NYT ran a featured editorial arguing that rockets could not actually work in a vacuum, like space, since 'there would be nothing for its thrust to push back against.'

They were obviously very wrong, but nonetheless were willing to consider views that ran against the grain even for the time. And that is a good thing. Many of the things they were right on were also equally 'out there' at some point, but I'm not mentioning those as we have the bias of 'well that really happened' so the truth doesn't seem as bizarre as it really is.

The problem we have now a days is that the media has become so collusive and incestuous that they rarely publicly disagree, again beyond the partisan split which I can only describe as child-like. And, in my opinion, this is likely intentionally done under a belief that having a unified front would increase their apparent integrity or confidence. Collusive groups like JournoList turned CabaList would work as some evidence towards this. However, at the same time it also completely destroys any notion of competence when everybody gets something so completely wrong. A very recent example of this would be the so-called 'sonic weapons' used 'against' US diplomats in Cuba. In spite of the media going into a unified frenzy of speculation and finger pointing, literally no major outlet took the logical, even if outsider, position that this probably was not even an attack in the first place. And now that it seems like that it indeed was not, it leaves the entire media system looking like a joke. Even when there are only two options (is a weapon, or not a weapon), nobody managed to get it right. That's sadly impressive and again something that was far less frequent an occasion in the past.




Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact

Search: