Tyler Cowen explains today why he's skeptical of the lead-crime hypothesis. It turns out that it's because of a simple bit of confusion.
You'll have to click the link and read Tyler's post to understand my reply—so don't blame me if you're too lazy to do it—but the key thing is that the LCH is not a general explanation for crime. It is solely a theory for the unusually large rise and fall of violent crime in postwar America through about 2010:
As it happens, the LCH also turns out to explain the rise and fall of violent crime in some other places during roughly the same period. Since the lead in question derives from the ethyl additive in gasoline, it all depends on how much car ownership rose (a lot in rich countries, not so much in poor ones) and when leaded gasoline was phased out (the mid-70s in the US, generally later in other countries).
And two other quick things. First, neither left nor right is all that friendly toward the LCH. The right doesn't like it because it diminishes the role of favored theories like moral decay, the breakdown of families, broken windows, etc. Likewise, lefties dislike it because it diminishes the role of favored theories like poverty, structural racism, guns, and the supremacy of culture over "essentialism." I think lefties have come around a little faster, but that's probably just because it was a lefty who first popularized it.¹
Second, I too assume publication bias is everywhere. I just think that it's probably a little low in LCH studies relative to some other areas of study.
¹Me!
This is the regular old crime unaffected by lead or the pandemic. These are just regular old barbarians with guns and nothing better to do.
https://www.nytimes.com/2023/01/01/us/in-a-small-arkansas-city-crime-dread-and-an-emergency-curfew.html
"The police say they have traced the turbulence mostly to young people, many of them high-school age, who have been out on the streets at night, and skirmishes between cliques that escalate into violence."
The problems of rural America...
The first few minutes of 2023 in wildly violent West Michigan…
16 year old in Grand Rapids (no doubt gangs)
Two men shot dead by drunk man celebrating in rural van buren county. This is not a tragedy.
Guns, booze and stupid. No lead required.
I'm glad to hear that the new year started out right for you.
This moment, start generating income Spend more time with your family and loved ones by working from home on tasks that just require a computer and internet cd03 connectivity. Start earning between $75,220 and $95,430 every month. I began this job, which pays well, and I’m now trading it with you so you may do it too.
Here’s where I began————>>> https://brightfuture241.blogspot.com/
Guns, booze and stupid. No lead required.
The LCH doesn't claim the reduction in environmental lead means you should have seen a peaceful holiday weekend. Rather, it suggests, had environmental lead not been mitigated over the decades, your weekend would likely have been more violent still.
No, it does not. But Mr. Drum has a habit of minimizing current trends by highlighting (with charts!) the fact that stuff was worse before or the same. And look, he did it even in this post when he writes “First, neither left nor right is all that friendly toward the LCH…”
The implication is that neither the “left” nor the “right” have a leg to stand on regarding root causes and solutions to crime because, of course, crime is fine at these rates. Now I know Mr. Drum and the rest of us would be happy to see guns taken out of circulation but that’s not going to happen. Do the police or anyone in government have a role beyond picking up the bodies after the fact?
That town in Arkansas is struggling with some serious social dysfunction if the article is to be believed. Likewise these mid size cities in Michigan struggle with violent teens. Surely the left has something to say about all that. Yet it’s clear they have nothing but silly rhetoric about slavery and systemic racism. So much for racial solidarity.
“Please help us bring these senseless acts of crime to a stop,” Mayor Tomeka Butler pleaded in a brief video posted online on Dec. 27 to announce the emergency declaration.
We have a black mayor putting a small community in lockdown because the kids are out of control or something. Perhaps Mr. Drum should quit worrying about the crime rates in the 1990s now. There’s an emergency today in many places.
That’s why I posted about this violence today.
yes, the kids are out of control. the plot of a jd movie from the '50s. the ancient egyptians said the same thing 2,500 years ago.
No, it does not.
The LCH implies violent crime would be higher in Western Michigan than it currently is had we not stopped spewing vast quantities of lead into the air breathed by West Michiganders.
Now, you're obviously free to believe the LCH isn't valid. But this indeed, is what the LCH postulates.
It is valid, I presume, for the time frame shown in the chart above. But that chart ends in 2010 and it’s 2023. The left and right propose root causes of crime today so LCH is not useful in that regard. And clearly I’m posting about crimes today too.
Are these criminals in 1995 are all zombies incapable of independent thought and free will? Are the criminals today similarly incapable of resisting criminal behavior?
I ask because I'm honestly curious: Do you have even a speck of free will at all?
It is valid, I presume, for the time frame shown in the chart above. But that chart ends in 2010 and it’s 2023.
No. It's valid now, too. Rather obviously, if we have less environmental lead pollution in 2023 than we had in (say) 1990, we'd expect violent crime to be lower now than it would be if we had never phased out leaded gasoline.
The fact that the ongoing trend has stopped doesn't mean that we don't continue to benefit from the one-time (which lasted about 25 years) reduction in environmental lead. We'll continue to enjoy the benefits of this indefinitely, unless we decide to re-introduce leaded gasoline.
It's not problems of "rural America", it's problems of "black America". Did you read the article. It's an overwhelming black town. Funny how you ignored that. Really funny!
Most of the people in that town are Christians. It's a problem of Christian America.
My buddy's mother makes $50 per hour working on the computer (Personal Computer). She hasn’t had a job for a long, yet this month she earned $11,500 by working just on her computer for 9 hours every day.
Read this article for more details.. http://incomebyus.blogspot.com/
Kevin, I have really enjoyed your postings on lead and crime. One possible reason people dismiss the correlation is they find it hard to wrap their head around the notion that environmental factors can affect crime rates.
But there are other things that may have been effected by changes to the environment. For instance, the reduction of lung cancer deaths due to efforts to reduce smoking. And reductions to car accident deaths due to increased engineering safety measures now built into cars.
"And reductions to car accident deaths due to increased engineering safety measures now built into cars."
Along with better design of roads and highways.
"One possible reason people dismiss the correlation is they find it hard to wrap their head around the notion that environmental factors can affect crime rates."
I think few argue lead had an effect, the issue with Kevin is that he thinks it is not just the main factor but the only factor. For him, lead is the one silver bullet that explains all violent crime and nothing else mattered.
I think you should re-read the original post. Kevin explicitly makes clear that lead is one factor, not the only one.
I think you should read the entire trend of Kevin's assessments, not just the specific words used in one article.
In particular read his original writings on the topic back when he was in MJ.
The graph he uses in this article is from one of those and he specifically mentions that the crime spike in the graph is due to lead, and over the years he has tried to downplay any type of other economic or policy factors suggested by either the left or the right as nothing but noise.
This is precisely the opposite of what I believe. I have said repeatedly over the years that (a) lead is just one factor in crime and (b) the lead era in the US is over. Crime today generally has nothing to do with lead.
Thank you for the response, since you rarely reply to anyone, so let me rephrase in the interest of clarifying my comment.
For you, lead is the one silver bullet that explains the violent crime spike in the US from the late 60s to the 90s, with every other cultural, economic, legislative, whatever factor suggested by the left or right being nothing but noise (i.e., they don't really matter for the increase and subsequent decrease).
Is that an accurate representation of what you think? Because that is the impression I've had from all your writing on the topic going back to your MJ days. If not, my apologies.
My knowledge of lead and crime is limited to the studies. While there are some interesting studies, that support a relationship, done in the US the international studies are far less robust.
For example, Japan had large industrial areas with lots of lead: however, the crime data and the study searching for a relationship with lead found no statistical relationship. A similar examination of Zurich, the most industrial area of Switzerland, also found no relationship.
If lead was a direct driver of crime, that relationship should likely be true in all settings/countries....
Japan had large industrial areas with lots of lead: however, the crime data and the study searching for a relationship with lead found no statistical relationship.
The article below suggests the incidence of murder in Japan roughly halved between 1990 and 2020. Perhaps the reduction in leaded auto exhaust fumes has largely compensated for industrial processes?:
https://www.macrotrends.net/countries/JPN/japan/murder-homicide-rate
Also, Japan (like Korea and Taiwan) is highly exposed to atmospheric pollutants from China, so, it's likely that domestic progress on air pollution (including lead) would be at least partly undermined via this overseas (but nearby) source.
https://www.jstage.jst.go.jp/article/emcr/1/0/1_20200003/_html/-char/en
Jasper_in_Boston - when looking at Japan I believe its crucial to understand demographics. Males between the ages of 16 - 24 commit the majority of violent crimes. In 1990, the average Japanese was 37 years old: today the average is approximately 50 years old. Seems like age is a major factor in the decline in violent crime in Japan.
If lead was a direct driver of crime, that relationship should likely be true in all settings/countries....
Maybe, but I'd advise caution against sweeping statements in this manner.
The data from myriad countries and regions support (if not prove beyond all doubt) the validity of the Lead-Crime Hypothesis.
And yet our planet is a big place with lots of complicated processes going on. My understanding of how Occam's razor operates in the real world tells me that outliers don't obviate a statistically powerful trend. Rather, they should prompt a search for explanations as to what's driving them, or to what extent said outliers are the products of incomplete data or flawed interpretation.
The US was far above all other developed (OECD) countries in violent crime during the wave.
https://kieranhealy.org/blog/archives/2015/10/01/assault-death-rates-1960-2013/
Although the US has more guns, not all the other countries have strict bans on private guns. But why should the excess murders in the US during the crime wave have only been committed with hand guns?
https://en.wikipedia.org/wiki/Gun_violence_in_the_United_States#/media/File:Ushomicidesbyweapon.svg
Why wouldn't lead-crazed 19-year olds have just picked up the nearest object to commit murder? There are knives (for example) in every home.
There are many kinds of evidence which are inconsistent with the lead hypothesis, and Kevin just ignores them - he only mentions things which seem to be consistent.
The US was far above all other developed (OECD) countries in violent crime during the wave.
The LCH says nothing about differences in absolute violent crime numbers between different countries. It only speaks to the direction in which those levels are moving. IE, more lead means more violent crime, irrespective of what the rate of violent crime was before the lead pollution started.
Yes, a country where firearms (which are more lethally efficient than clubs and knives) are extremely easy to acquire is likely to see higher levels of murder, all things equal, than countries where this isn't the case. This is utterly independent of the environmental lead situation.
But if the LCH is valid, we'd expect similar increases in environmental lead pollution in different countries to prompt similar trends (ie, increases or decreases) with respect to violent crime. And this, of course, is exactly what we do see in a large number of countries. But in absolute numbers, violent crime statistics in different countries affected by environmental lead don't converge (why would they?), because the underlying conditions responsible for the base rate of violent crime (demographic, socioeconomic, policing, firearms availability, etc) are often quite different.
Lead damages the brain and neural system. Do people damaged by lead exposure behave the same in every society? I can imagine alternative outlets, for example, in areas of political instability, someone damaged by lead might be useful to the local militia. They might be restrained by their family in a society with strong family ties.
It's tough doing cross society studies.
Lead damages the brain and neural system. Do people damaged by lead exposure behave the same in every society?
Identically? No. Which is why I wrote "similar" trends, not "identical" trends. For one thing the level of lead pollution is going to vary from place to place. But if you dump gigantic quantities of a potent environmental neurotoxin into the air, you're going get an increase in antisocial behavior. In other words, the direction of the trend is highly likely to be the same, even if the slope of the curve varies, based on local factors.
As usual, no cited sources. Zürich is a medium-sized city whose principal ‘industries’ are banking and corporate headquarters. Less than 0.017% of the world’s population lives in the metropolitan area.
KenSchulz - I am not sure how familar you are with the Switzerland of the 1960's and 1970's: I lived in Zurich for two years in the 1970s. In the 1970's chemicals, pharmaceuticals (my father's industry), precision machinery were important industries. It was not always the land of chocolate, watches and banks....
Tell me, I'm curious: Why would you think that relationship should likely be true in all settings/countries? That's your hypothesis, BTW. So it's on you -- with the usual evidenciary standards -- to persuade us that your argument has merit. It is most definitely not on anybody else to prove you're wrong.
Which studies are these? I try to keep up, but I'm not familiar with either of them.
Well at least Tyler Cowen is honest about his own bias: "incentives, policy, and culture" . I like that he adds "culture" thereby making this as broad an umbrella as necessary to explain almost anything. Culture would include things like science which includes things like finding out that lead in gasoline is bad for us and then initiating a policy which bans it. Voila ! Lead crime theory now fits into Tyler's world view. But I don't think that's really what Tyler believes or wants to be true. I think he wants to find that all those liberal ideas of the 60's caused crime to rise until Reagan and the tough guys that followed initiated things like "broken windows" policing (and maybe stop and frisk?) based on the depravity of certain elements explains the changes in crime rates from 1960 to 2010. The fact that crime was rising before most of the supposedly liberal policies were in place and started the trend downwards before things like "broken windows" were initiated should make us wonder about the kinds of explanations Tyler Cowen likely favors.
BTW I think the meta study does raise questions about the lead crime theory but it's one study and seems to rely heavily on publication bias distorting results.
Yeah, I read the Cowan post. He introduces that flimflam about "incentives, policy and culture" as an a priori assumption, without backing it up.
I wasn't impressed. He also (as Kevin suggests) appears to have a somewhat strange understanding of the implication of the LCH.
As I recall, they took the lead out of the gasoline because it destroyed catalytic converters. They needed the catalytic converters to get rid of all the unburned hydrocarbons from the rich fuel/air mixtures needed to suppress nitrogen oxide formation. All this to combat smog.
Only later did they realize that most of the lead in children's bodies was a result of the air they were breathing and not from leaded paint. Not to say that leaded paint wasn't part of the problem, just not the biggest part.
Interesting. Sounds right.
That's my memory too, on both counts-- the lead (originally added to gas to lubricate and preserve cool seats) deposited on the catalyst pellets in the converters and stopped them from working.
In fact the whole gas-delivery system of the country had to be altered to prevent leaded gas from being used in cars with cat converters. Those cars had to use smaller-diameter filler-pipe openings that the nozzles on leaded-gas pumps were too big to fit into, while the required nozzles on unleaded gas pumps were small so they'd fit. Gas stations had to sell both for a while because you couldn't really run a car designed for leaded gas on unleaded; it would burn out the valve seats. Manufacturers had to switch to much harder valve seat inserts but that didn't help the cars that were already out there.
Meanwhile almost all the emphasis on lead toxicity in kids was on eating paint that chipped off walls in old houses. Newspapers were full of stories about this, ditto tv news. It so happened, though, that the neighborhoods where kids were developing high lead levels-- poorer, with old housing, etc-- were also cut up by major traffic arteries that carried heavy commuter traffic twice a day, but it took a long time for that penny to drop. I don't know when that was but it's easy to imagine how loud the collective dope slap must have been.
... added to lubricate and cool valve seats ...
The real reason is much simpler, I believe: Tyler Cowan is skeptical because he didn't come up with it. And this is precisely the sort of explanation that he wants to be known for (sorry). Tyler Cowan makes unseen connections that fly right by us lesser intellects. Y'know, likescientists and Democrats. Especiially the progressive types in those two categories.
Have I got this right? Tyler Cowen says the real causes of anything are "incentives, policy and culture" and since lead in the air isn't one of those causes, then lead can't possibly be an important factor?
I find his post very difficult to understand, not least because in his last sentence he seems to change 180 degrees and say lead could indeed be an important factor.
Huh?
If people didn't exist they wouldn't be poisoned with lead, ergo it's culture?
One way to explain the different crime rates in other countries may be that the U.S. has greater crime opportunities. IOW, lead could be the proverbial straw that broke the camel’s back. Japan, for example, requires individuals to jump over more steps to commit crime due to several factors including its homogeneity. In the US, because of its ample size and socio-cultural factors (e.g., guns) it is just easier to commit crimes. The lead crime theory should not be hard to understand since it is a fact that lead affects the impulse control of individuals and such would impact a
“freer” society like the US.
One way to explain the different crime rates in other countries may be that the U.S. has greater crime opportunities.
This has little or no bearing on the debate regarding the validity of the LCH. Said hypothesis obviously doesn't claim that, absent environmental lead, all countries would suffer from an identical level of crime. The LCH claims that lead is an important factor in a marked increase (or decrease, once environmental lead is mitigated) in violent crime levels, above what the base level would be absent the lead pollution. In other words, environmental lead pollution should trigger similar trendlines in different countries in terms of increasing crime, but doesn't have any connection to what the base level of crime was beforehand. To cite one example: the murder rate in Japan decreased by about 50% in the three decades after 1990. This is similar to the reduction achieved in the United States. But Japan, needless to say, both started with and ended up with a much lower incidence of murder than the US (even if the trends in both countries were similar).
I read the Cowen post, and concluded that he just simply doesn't want to believe it and that's all there is to it. He seemed determined to answer a question that wasn't asked, in order to avoid the real question. Sounds like a politician.
Give the guy a break! in all fairness, ya gotta admit Tyler Cowan didn't come up with it; that's surely a point or two in his favor.
Oh, so you mean Tyler Cowen was being stupid?
Who said that in the last post on--oh yeah.
IT WAS ME.
What Cowen and some of the commentators don't seem to understand, maybe because they don't want to, is that the LCH was always trying to explain one specific phenomena, that is the alarming rise of crime in the US, especially violent crime between roughly 1960 and 1990 and the equally surprising decline of crime between 1991 and the 2010's. It seems to me that the LCH does a pretty good job of explaining this , bolstered by evidence from different countries, though its certainly possible, maybe likely that other factors were involved. The meta study Tyler Cowen cites pushes back on I think Kevin raises reasonable concerns about that. We need more science, not Tyler Cowen's vague suspicions and nearly vacuous generalities.
OTOH, most of the most popular alternate explanations have even bigger holes.
I remember the early 90's when crime was a huge issue, hence the 1994 crime bill Many supposed experts got this pretty wrong.
From a 1995 Office of Juvenile Justice and Delinquency Prevention OJJDP publication, noting that crime seemed to be going down.
"While policymakers might consider the latest figures as cause for cautious optimism, criminal justice prognosticators are warning that the downturn could merely be a lull before the next storm of juvenile violence. Two of the most vocal espousers of this theory are John J. DiIulio, Jr., a professor of politics and public affairs at Princeton University (New Jersey), and James Alan Fox, a professor in the College of Criminal Justice at Northeastern University (Massachusetts). They argue that while a slight downturn in juvenile violence is inevitable, the Nation, in Fox's words, is "on the verge of another crime wave that will last well into the next century."
i just did read cowen's post and it is almost incomprehensibly stupid and ignorant. the idea that he doesn't understand that the "LCH is not a general explanation for crime" yet still wrote a blog post about it seemed literally unbelievable, but from his post it clearly is true. it's amazing that any pays attention to anything he ever says.
If Cowen has ever written a non stupid post/article/etc., I am not aware of it. It remains a mystery why Kevin cites him so often.
he was also on ezra klein's podcast recently. i truly don't get it either.
Not a single comment -and not Kevin's blog- explains what "LCH" stands for. Neither Cowen nor Neeva gave me a clue. Very glad it doesn't matter!
the first line of kevin's post is "Tyler Cowen explains today why he's skeptical of the lead-crime hypothesis", ie the LCH.