Even more specifically, mass shootings never or almost never happen when it's raining outside, only when the weather is clear. Mass shooters, in their twisted view of the world, think to themselves "today is a good day to carry out my plan" when the weather is nice and "maybe some other time" when it's raining that day. Maybe the correlation has less to do with heat directly and more to do with other weather issues.
I live in Singapore which lies near the equator and is hot throughout the year. And also a country with one of the world's lowest crime rate. Just saying.
So how doesn’t bode well in an era of climate change and little hope of gun reform.
For those interested in further reading on this topic, I suggest reading Jude Kieltyka's 2016 paper on "Ecologic factors relating to firearm injuries and gun violence in Chicago."
As many other commenters have stated, there are few studies that look at the mechanisms behind these statistical correlations. Many of us speculate the causes of these relationships, but without scientific evidence, we should not form policies around these opinions.
So, let's work on funding researchers from a variety of disciplines looking at the spread of and causes underlying gun crime and firearm violence in the United States. We are currently lacking evidence to support policy changes and, despite the desire of many researchers to publish on these topics, there is very little funding available for these projects.
Support your scientists and if anyone is looking for more literature on gun crime or gun violence, feel free to track me down. I have hundreds of papers saved on these topics.
I really like the photo captioned "The Chicago police investigate a shooting in August" by Tyler Lariviere/Chicago Sun-Times imbedded in the article, particularly the street lamp lighting.
1
If the weather is irritating and hot, it doesn’t mean that one should go and shoot some person or persons. This doesn’t make any sense.
I read somewhere that in Japan there is a provision for letting out anger by breaking some glasses in a room of course at some cost. In this way you are hurting none but your pocket.
Everyone gets angry at some point of time. When I am angry at the most I scold a person who has hurt me and not merely because of the weather.There are some great people, who simply avoid even that. Instead of creating a scene they walk away from that place.
1
So getting ‘hot under the collar’ truly means something tangible now. Hmm.
5
Without any statistical analysis, I maintain that school shootings are much more likely to occur in the spring time than any other time of the year. Likewise, street based revolutionary movements usually occur in the warmer months, especially in countries that are cold much of the year. More people would be outdoors and available to join street demonstrations that sometimes lead to major social and governmental changes, like revolutions.
As for shootings in the hotter months, studying that should not be difficult. The police reports would ordinarily contain some indication of a motive for the crime and this could be associated with the weather. A robbery can happen any time, but an increase in robberies tied to shootings in the hot months would indicate some causation.
I suspect there are a lot more random encounters in the hotter months than in winters. Drug deals, fights over women and other disputes are more likely to occur when people are outdoors and coming into contact with each other. If people stay away, fewer shootings.
Is there a decline in crime generally in the colder months? I've often thought about this when there is a major snow storm: what does a determined burglar do on days when the snow is high and people are home from work? If crime goes down, shootings should also decline.
Summer generally is the more productive time of the year for anything outdoors, like construction. If people are out doing more, then it follow that more violence would occur, too.
1
In a country where most parents value their guns more than they value the lives of their own children, that disparity is what should be investigated, not whether the consumption of ice cream, or murder, changes with the season. In fact, even more ominous that a link between human-caused global warming and murder is the FACT that mosquito-borne diseases will expand their ranges because most humans don't have the intelligence that natural selection gave wild sheep, and which artificial selection took away from them. There's a reason the Good Shepherd referred to his followers as his flock and held up the lowly domestic sheep as the model for good HUMAN behavior.
I love how San Francisco gun violence actually declines in hot weather. Everyone is just so relieved that the darn fog is gone for a change.
4
If you don’t already know how hard it is to get air conditioning in NYC public housing, look into it (prepare to be shocked). If you were a teenager in a big city on an airless summer night with no AC, wouldn’t you go outside to meet your friends? Add gangs and guns and we all know how the story ends. Here’s an idea: make sure every bedroom in every public housing project has a working air conditioner. Watch murder rates drop.
5
Where do you think the term “hot headed” comes from and what does it imply?
Where do you think the term “cool headed” comes from and what does it imply?
He’s “cold as ice!”? He’s “hot under the collar!”?
Whether it’s weather-related or not!
Think about it!
3
This article is somewhat disingenuous. I'm in total agreement that weather is a major factor in violent crime, but race and poverty plays an even bigger role. The author of the article should have also looked the demographics of these cities. The cities with the largest increases in shootings during the summer months have large minority populations and that's a fact that cannot be discounted. Most urban cities always have high percentage of violent crime regardless of the time of the year, its always been that way and it will continue for the foreseeable future. SAD.
1
Temperatures are high in Summer, and also the days are longer. Perhaps if the data were normalized so one is looking at 'number of murders per daylight', then one can be sure that the cause is the temperature rather than the lengthened day.
It is also possible that economic activity (driven by sports, leisure etc) increases in summer/hot days, and hence the murder rate increases accordingly. Perhaps they normalized the murder rate data to, say, GDP growth during summer quarters etc.
1
The statistics for shootings per day are horrifying to read, whether for cold or hot weather. How could even one shooting per day be considered acceptable for a city?
Control guns. Control ammunition.
6
According to this study's data, it's not the heat - it's the humidity.
100% of the cities east of the Rockies in this survey had increased murder rates. But 100% of the cities in this survey west of the Rockies (where summers are dry) had decreased murder rates.
Or maybe only eastern cities have a rising murder rate, not western ones. In any case, it will take a study with better geographical sampling to move beyond speculation.
6
I am surprised the authors did not quote this excellent review article: Science, volume 353 issue 6304 (9 Sept 2016), “Social and economic impacts of climate.” The issue of crime as correlated to the temperature relative to the average is noted.
p.s. I am not associated with this paper, though I work at the same school as the authors.
3
When I lived in Brooklyn, other people's music/noise made me almost murderously angry. Not quite. I don't own a firearm, but I was tempted to throw bricks out my window at cars blasting music. A woman I dated went so far as to throw a pan of (room temperature) water on someone because of their music. When the weather is warm, people go outside and often make noise or blast music. How many of these homicides involve disputes between neighbors or people sharing city park space with rivaling music? Now, I live in Battery Park City. No noise. It gets pretty darn hot in SoHo (NYC), but I'm not sure the murder rate goes up.
5
Any street cop could have told you this without the data.
1
@Sarah True, cops have long had a saying, "The best policeman is bad weather" - by which they mean cold, snow and sleet.
3
I am not convinced. Phoenix has much higher temperatures all year round (reaching unbearable in summer) and much lower murder rate than Chicago,Baltimore, Detroit, St. Louis. Appalachian region is much poorer than the same cities as well, yet much lower murder rate (for those who says that poor equals criminal).
There is an elephant in the house that NYT and other media avoids talking about like a hot potato. Surpressing the truth it's not good for your health
6
@Jon - The leftist media suppresses these truths but mainly preach to the choir and fool fewer than they think; hence Trumpers being elected.
3
What about the non-weather factor that the closing of schools disrupts family lifestyles, which badly stresses the nerves. Teenagers and young adults lack the social and content structure that aids in controlling their hormonal surges, even as many are urged to keep an eye on primary-aged children while parents work. The wonder is not that so many murders take place in this chaos of identities, but that the invisible majority of these young people survive and develop solid characters.
Could this effect have anything to do with the Civil War? The southern states are warmer.
I worked a summer job for the Milwaukee County DA. There was definitely a correlation between heat and criminal offences. I recall in earlier times riots took place in “long hot summers.” I wonder what the NRA’s climate change policy is.
2
Really? You can't think of anything that happened about 4 years ago, that might have caused the Obama Administration to come down hard on police departments around the nation. I'm sure it you try, you might come up with something.
1
For clarity you should not use the word "gun violence" which includes suicide. The grouping of all firearm mortality together makes the numbers look larger and fools people into thinking homicide rates in places like Wyoming are also very high. The murder rates in places like Chicago are unconscionably high, and the way our society continues to ignore the loss of life of our youth needs to end. Sorry to sidetrack the main point of the article.
5
Lack of air conditioning and low test scores are causing violent gun attacks and murders in certain cities. Of course, its the weather. Lets look at every reason to not believe that the people who kill young people and children, are criminals who have no reguard for human life.
4
Why was Chicago left out?
1
And don't forget the moon. Long ago the only job I could get - based on 10 days of instruction with an institution I still try to forget - I got a job as an emergency room orderly. I did trauma better than some of the docs.
It's always hot here, but when the 111º scorchers hit, we'd get more damaged people. But what we truly feared, was the moon.
Year round, despite the season or weather, when it went full the damaged and dead flooded in - haunting us under that malevolent light. We feared the moon.
And they say we aren't animals beneath the sun and moon.
4
@Robert huh, weird! I believe your n=1 anecdote, but it would be very interesting to see data -- maybe this paper of record can dig into full-moon data as a follow-up.
1
It is amazing that cities like Detroit have not embraced “stop and frisk”.
When 60 people are shot over a single weekend, it would seem to be a worthwhile tradeoff. Particualrly when so many victims are innocent bystanders. It would be interesting to see its impact upon crime statistics in these troubled locales.
The fact that it does not identify many gun toting individuals does not prove that it is ineffective. It functions as a deterrent, gun toters leave them at home.
5
Oh, this is totally true. In college, my friends and I used to joke at springtime that it was “about that time again”, the time when the creepers come out. Ask any young woman. Warmer days = more days harassed or worse.
3
Hi, I am a public school math teacher and like to bring real data into my class. We are finishing a unit on linear regression so I followed the links to the data you cite. I have created a graph and regression analysis to go a little deeper into the statement "If you look at national murder data from 2000 to 2016 and compare it with the average annual temperature across the United States per year, you could predict a national rise or fall in murder using just the average annual temperature in 11 of 16 years.” I created a linear model as seen at https://www.desmos.com/calculator/vn6lnxhdfg
May be I made a mistake.
However, if I didn't, I am perplexed as to why you only used 2000-2016 expect that if you add in 1998-2000, the correlation of determination goes down to less than 10% rather than about 30%. Is there a good reason to choose 2000 rather than some other date? (I see that the FBI changed reporting of some data in 1997, so I understand why you can't go back before then). I am trying very hard to teach my students 1) mathematics can help us understand the world and 2) we can be intelligent consumers of the news. Any insights will be appreciated.
8
@Linda Saeta
Hi Linda, having taken a look at your data it appears that in 1998 the number of gun violence incidents decreased by 1234 while the temperature rose by 2.03 Fahrenheit. If you include that observation in your data you'll get a smaller estimate for m of course (where using your notation y=mx+b with y change in # gun incidents, x the change in temperature and b the intercept). First of all, you're dealing with a very small sample. We don't even know whether the estimator is significant. To be a bit more robust you could try the leave-one-out estimator to ensure that no single observation drives the estimate m. This one is fun! Secondly, you cannot expect the parameter m to give you the true value. The model is incomplete to account for crime. The estimator should be biased because of omitted variable bias. For instance you may want to include state fixed effects, time trends but also policy dummies (as for instance stop-and-frisk policies or background checks). It may be that those policies are more effective during warm weather (i.e. omitted variables correlate with some included variables). Having a biased estimator one cannot deduce that a rise in x dgrs Fahrenheit increases the expected number of gun incidents by a number of y as produced by the model. To my mind the article is interesting nevertheless as it is suggestive of deeper empirical work that could be done in that area (if it hasn't been done already). As it stands it is merely toying around of course.
2
@Linda Saeta
I will try to clarify. First, Ms. Saeta's analysis is correct. The R2 (she calls it "correlation of determination", but she meant "coefficient of determination") for the 1998-2016 data is 0.0921, which translates to a correlation of 0.303. If you drop 1998 (I'll explain why you should below), the R2 increases to 0.2725; this is barely different from the R2 of 0.3088 for the regression using the 2000-2016 data. Both of the larger values are statistically significant at the 0.05 level; the smallest value, for the data including the 1998 changes, is not. (These R2 values correspond to correlations of 0.303, 0.522, and 0.556.)
Obviously 1998 is an anomaly and should be excluded. Why? Because the 1997-1998 murder-rate change is unreliable, given the change in reporting from 1997 to 1998, as Ms. Saeta noted. (Good detective work!)
As Mr. Sandmann pointed out, this is clearly too simplistic a model for explaining changes in murder rate. It's entirely possible that there are other factors besides temperature that are related to the incidence of murder, which if included in the model could make the temperature effect disappear. That's an empirically testable question.
If anyone is interested in seeing my analyses, email me at [email protected].
1
Someone needs to Google "Ferguson Effect." From Wikipedia:
"The Ferguson effect is the idea that increased scrutiny of police following the 2014 shooting of Michael Brown in Ferguson, Missouri has led to an increased crime rate (or sometimes increased murder rate) in major U.S. cities."
While there has been some criticism of the idea, the timing is awfully suspicious, and the increase in murder rates tended to be steepest in cities with large black populations, which is exactly how you would expect something like the Ferguson effect to play out. The problem is that the idea is ideologically distasteful to people who run the liberal media, so it hasn't gotten the hearing it deserves, and probably never will.
8
The correlation between people being outdoors more, and being involved in acts of violence more, is easy to accept. However, the leap to "climate change" is tenuous. If that were true, then why has the violent crime rate been steadily falling for decades, while temperatures have tended warmer? Keep the conclusions narrow, for maximum credibility.
5
There seems to be some confusion here between number of murders and murder rate. The article shifts back and forth between the two. Murders per day may go up in years when the murder rate (murders per 100000 of population) goes down.
The*murder rate* went down for 16 years before rising in 2015 and 2016. Many explanations have been offered including the fact that young men commit most of the murders and their share of the population has been going down.
Even so, the young people who are still here can commit more murders per day if there are more guns per young person, or if there are more things to fight about, like drugs, or if there are more things to be angry about, like police-involved shootings, etc.
Statistics are wonderful but sometimes you have to average the different explanations in order to get the big picture.
Personally, I think it's more than a question of just the temperature, but also of the humidity -- I find high humidity/low barometric pressure make me depressed, edgy, short tempered, whether the temperature is above or below 85 (though it's worse above 85). There were plenty of days this summer here in the northeast when I felt like grabbing a knife and running out in the street and committing a felony as we had few breaks in the humidity. In a decent summer we may have 4 or 5 days of hot humid weather, but then a day or two when, even in the temp doesn't drop, the humidity does, and I feel lighter, more buoyant, happier. But we had few of these breaks this summer.
I lived in San Francisco for 12 years and the weather is very different there. More moderate year round, and on the few occasions when it is hot,like 90 or above, it is not humid. I never had air conditioning and rarely regretted it.
@ellienyc
You forgot to mention bariometric pressure in addition to humidity and temperature.
@5barris ...that's literally mentioned in the first sentence.
I have two friends who get migraines with changes in barometric pressure, so I can believe it has more subtle effects on other people.
People are more outside in warm weather than cold. Therefore, the more visible the easier to settle a beef. I've been told that personal beefs will be put off all winter and then come summer woof tickets are cashed in.
4
It's far more likely that the rise in gun violence in the major cities is dues to policies that hamper local police like stop-and-frisk.
8
It is interesting to see the correlation between heat and occurrences of murder. However, I thought this was already a well known phenomenon - David Simon outlined this in his book Homicide: life of the street back in 1991.
15
So what can be done? What’s within the control of the US to fix? Climate change to some extent over time. Sensible gun control laws are certainly the easier to address. With this administration there seems to be little appetite to address either.
2
... and Shakespeare in Romeo and Juliet and Spike Lee in Do the Right Thing. This all seems pretty obvious...people get edgy when it’s hot...
1
Definitely. This "news" is so last century.
1
Many of the articles in the NYT compare data between racial and ethnic groupings, it would be interesting to see if the heat thesis applies across all demographics.
Separately, I worked in a NYC law enforcement and justice related office in the mid 70s. I remember a seasoned cop telling me that the incidents of violent crimes was greatly influenced by the weather.
19
The states with the lowest murder rates happen to be the colder, northern ones. New Hampshire, Vermont and Maine are the lowest.
11
And - elephant in the room - the whitest,
5
@PM Is climate the only variable you can come up with? Come on, THINK!
4
@PM
All those states are lo population density, and New England culture states. How do they compare with Mich, Wis, Montana, Idaho, Wyoming, the Dakotas and Minn? Wash and Oregon may be tempered by the Pacific ocean so I don't think they tend to be as cold.
Climate plays a role, but poverty is a much bigger factor.
And any ER physician can tell you that the murder rate
is up on weekends
26
@Gerhard
"And any ER physician can tell you that the murder rate
is up on weekends"
So are sales of "Mad Dog 20/20" and "44s." So much for the worn-out excuse about "poverty."
Inner-city liquor stores do OK -- why?
4
What's interesting to note is that the states with the lowest murder rates are colder, northern ones.
8
Shouldn’t this be obvious? More people get out on hot days and assuming constant probabilities of quarrels and ones which turn fatal more people would get killed.
Assuming 1% probability (super high - only for example), in cold days there were 100 people outside and one murder. On hot days there were 200 people outside and 2 murders and we get 100% increase!
13
@Fakrudeen
Of course, warmer days are also correlated with longer periods of daylight and more people out of their homes.
Those seems to be as significant and the temperature itself.
4
Hot weather tends to induce heated action. Sometimes, excessive, violent action.
This much seems to be obvious, but obviously needs to be qualified, examined.
The increased potential for violence when people are uncomfortably hot has been noted in our films and literature. Spike Lee's "Do the Right Thing" and William Faulkner's short story "Dry September" being just two examples.
I for one am not surprised by these results, but agree with the author that more research is needed.
11
@Ken L
" .. agree with the author that more research is needed .."
Sure. When the taxpayer debt, with Asia rising, is the worst 75 years -- why not keep spending on the obvious? Hey, Detroit got bailed out, right?
1
In SF, the fog can make some grumpy. On the rare day the sun comes out, good moods abound. So the drop in SF on warm days makes total sense to a local.
10
I read recently that there is a correlation between violent behavior and IQ. Low IQ individuals have less capacity for empathy towards others and less ability to reason with themselves, according to the psychologist who wrote the article. Maybe we could take this into consideration when making immigration quota decisions?
37
@me:
Where did you read that, and who was the psychologist? Please provide the source. Otherwise it is just fake news.
@me Do you agree that we should facilitate emigration for some current Americans based on their IQ profiles?
Remove the guns and you'll find death by harsh language to be very, very small.......
34
The air conditioning gap, folks. That’s, right. Call your senators. It’s time to pass the No Child Without Air Conditioning Bill. It’s really simple, actually. Once a week, a truck (Tesla, of course, so low environmental impact) will drive around to air conditioning-less communities and install window unit ACs. A separate day of the week will be the “non-window unit” day, as we need to take into account homes that do not have windows.
17
The data confirms what has always been obvious to anyone who works in an ER, trauma center, or police department.
34
This is a perfect example that correlation is not causation. There are plenty of other reasons that violence went up in 2015-16.
For this data to have any meaning, we would have to look at suburban and rural data as well.
15
@Cousy
No, this is NOT "a perfect example that correlation is not causation."
Because the data was examined in multiple ways, not just one, and in completely different datasets.
"Heat" and "murder rate" were positively correlated when examined across different cities, within individual cities, within ranges, and across time.
And their are logical causal mechanisms.
When data are this Consistent, they meet the Bradford hill criteria for causality.
See: https://en.wikipedia.org/wiki/Bradford_Hill_criteria
23
@SRP
Hill criteria never prove anything they simply provide evidence for something - and the fact that virtually no one outside of epidemiology used them and even that limited use is disappearing suggest they were never very good.
You can never get to causality without a causal model - all that you can ever do is suggest it is likely.
And when you cite multiple data sources - that is not Hill it is called a consilience of induction - whose best example remains On the Origin of Species - which I just happen to be reading now.
Ray Bradbury tells this story in Touched by Fire from the October Country anthology --