TV causes autism? I doubt it.

An article in Slate yesterday argued that TV watching causes autism. The Slate article is based on research done by Cornell economists Michael Waldman, Sean Nicholson, and Nodir Adilov. You can download the academic working paper here.

The paper gives some theories why TV and autism might be linked, but the more interesting part of the paper is the data analysis. The researchers are trying to find a “natural experiment” that shifts around TV watching, but otherwise has no impact on whether a child is diagnosed as autistic. Rainfall is one of the things they use. In places where it rains a lot, kids watch more TV. Maybe rainfall doesn’t affect autism in any other way. This is a creative approach, although it suffers from the weakness (which they acknowledge in the paper), that rainfall changes other things, like how much time you spend indoors doing other things besides watching TV. They also use the arrival of cable TV in an area. This approach is potentially stronger, although it would be better if they used availability of cable TV, rather than the number of people who actually subscribe.

These are intriguing approaches, but personally I did not find the empirical evidence in the paper very compelling.

The rainfall evidence is based off of three states: Washington, Oregon, and California. It rains a lot in some parts of these states, but not others. There is more autism in the parts of the states where it rains more. The problem is that it rains on the coast of Oregon and Washington, and in Northern California. But there are a million other differences between the coast of Washington and the Eastern part of the state, and between Northern and Southern California. The researchers also look at how much rain there was when you were between the ages of 0-2, controlling for your county. This is more promising. The impact of rain gets smaller, but it is the most convincing evidence in the paper.

The data analysis of cable TV is limited to California and Pennsylvania and also finds positive results. The difficulty with the cable TV analysis is that there is an incredibly strong positive trend in autism. The cable TV data are basically on an upward trend. The regression analysis is going to have a very hard time sorting out between a steady rise in cable TV penetration and the time trend. In the current version they only include a linear time trend, which is an extremely powerful predictor. My guess is that if they generalize their specification to allow for non-linear time trends, the cable TV result will disappear.

The authors have done some interesting work, but the nature of the problem makes it a really hard one to answer convincingly. For instance, you might think that Oregon and California should have similar autism rates. Nope, Oregon’s rate is four times higher. That sort of gap is almost certainly due to differences in what is called autism in the official data in one state versus the other. The increasing time trend is also heavily influenced by what is labeled autism. When the outcome of interest is measured so poorly, it is hard to know what the analysis is really picking up — differences in the underlying symptoms or just in the reporting of them.

The more I thought about it, the more it seemed to me that there might be a causal link between rainfall, TV, and autism, but not the one suggested by the paper.

My theory: when it rains a lot, parents watch more TV, see more shows about autism, and this leads them to seek out a diagnosis of autism for their kids. They have the same kids, it is just that TV makes them believe that their kids are autistic.

I don’t mean to sound overly negative on this research. I applaud the authors for asking a daring hypothesis and gathering data to try to test it. My gut, though, tells me that this is a result which will not stand up to scrutiny.

scott cunningham

The Slate link is broken.

This is a fascinating study. That they even find these partial correlations for diagnosis and rain is itself interesting (even if they are, as you not, just picking up a causal relationship between TV and doctor visits, and not actual autism itself).

Autism Vox

Thank you for the thoughtful analysis. I am taking this study with some grains of salt as it relies so much on data analysis and not on actual observation and experience of autistic children and persons watching TV (which I tried to describe in this post).


The study is hogwash. Besides not establishing a causal relationship of any kind, it fails the most basic kind of test ... what about all the people that watch TV but don't have autism?

TV is so pervasive, it seems like it would be a correlating factor in any study. I also agree that someone with TV would be more likely to seek help when they see a display of autism on TV. Someone who doesn't have TV? Probably doesn't have the money to seek that kind of help.

I completely reject the notion, however, that this was a "daring hypothesis". TV is a firebrand. Why not go after something truly daring, like drinking water or some other environmental cause? Heck, even high voltage power lines are a tougher target than TV.

I'd like to see a study really go on a limb and pursue the results, rather than just present them. OK, so autistic kids tend to be in states where there is higher-on-average rainfall. That would tend to drive people indoors. Now lets look deeper. Sure, TV is a factor, but what are other factors? What are the relative ages of the homes? What building materials were in common use when those homes were built? What are the pipes made of? The insulation? TV was a cop-out result.



Did the study control for the occupations of the parents?

As for example, here:

High-functioning or borderline autistic people tend to be better at handling certain jobs than others. For example, engineering and computer programming are good fields for autistic people who may lack some people skills but may excel at math, visual thinking or memorizing facts.

Parents who may may have some autistic traits are more likely to have autistic children.

So areas of the country that just happen to be located near Silicon Valley, Microsoft, Boeing, etc. might just happen to have more of these types of people who are more likely to have autistic children.

Any slight trend in that direction is likely to be reinforced by the fact that an elevated population in one area of people who have an increased risk of having a child with autistism, there is going to be more couples where *both* parents have this increased risk. Furthermore, in an area where there is more autism, there will be a greater tendency for there to be resources allocated to testing for autism, services for autism, centers for studying autism, etc. These extra resources mean that there will be more screening for autism and more cases will be found. And even further, parents of autistic children may choose to deliberately move to such areas in order to make use of these enhanced services.



Nice work, sir.

How about something along these lines:

When it rains a lot, parents watch more TV and also spend more time with their kids, which makes them more inclined to seek out a diagnosis for possible signs for autism.

Good hand.


A much more convincing study is here

They find a significant correlation between age of the father and autism. They use a uniform definition of "autism".

This paper reminds of some studies done in the 1930's and 40's, showing the "link" between automobiles sold and lung cancer. A significant correlation existed between the two.


I think the paper attempts new ground and we should have a space for it. Note that we have an outdated model of truth for thes items. If the paper were entitled, "We're f#$@ed: why the data on autism is bogus", no one would read it.


This is what epidemiology refers to as an "ecologic" study, for reasons lost in the depths of history. The primary problem is that it is a group-level study, ie there is no direct link between individuals with and without autism, and tv watching or lack thereof. In other words there is no way to determine if the autistic kids are getting the higher TV hours or that their non-autistic counterparts are, I don't know, playing video games instead.

The preferred approach, since the question has arisen, would be to assemble two groups of subjects, one group with autism & another without, then look at whatever exposure you're interested in. Folks in Seattle read, especially used books, and drink coffee, especially dark roast, a lot more than people elsewhere; maybe those are possible causes.

With this type of study, one can uncover interesting ideas for further work, but equally can overreact to results and get led far astray.



Slate posted a follow-up where a bunch of commentors leveled some withering criticisms at the report:

Here's one I particularly liked:

Easterbrook consistently refers to the increase in autism that begins "around 1980, about the same time cable television and VCRs became common". If Easterbrook had done his homework he would have found that 1980 was also the year the diagnosis of "Autism" actually became a diagnostic entry in its own right in the DSM-III. It was reclassified from being part of "psychotic" disorders (like schizophrenia) to having a specific heading in "developmental" disorders. Furthermore, in 1980's the 'autism spectrum' (autism, PDD-NOS, and the newly-minted Asperger's syndrome) definitions were changed or created to include persons with normal range IQ, and less severe symptoms. All of these changes to diagnosis would significantly impact the makeup of the population called "autism spectrum disorders" (ASD)-- sometimes just called "autism" by the media. (to the confusion of many readers). All of these changes coincided with the increase in television-watching, but (am i going out on a limb here?) weren't caused by television. […]



TV as a contributor to autism sounds like a long shot, but autism is such a severe problem that it deserves to be investigated. So, here's a more promising historical natural experiment, with a much sharper increase in television watching:

The South African experience. The regime allowed no television until experimental broadcasts in 1975, followed by full-time broadcasting in 1976. South Africa is unique among countries with a large modern health system in its late, sudden adoption of television during an era when autism's defintion was more clear than during the 1940s and 1950s when TV-watching was growing in America. I have no idea what the autism trends were in South Africa, but this might provide a good test.


DNA and the Brain -- If you're looking for offbeat but 'genius' insights into the causes of autism, watch this Google video of Dr. James Watson. He talks about the high incidence of autism in Silicon Valley and relates it to the many super-smart people living and reproducing there.


As a breast cancer survivor, I have become increasingly cynical about these types of studies. I suppose it's acceptable that some researchers at a university did a study that concluded women who ate french fries before the age of 5 were more likely to develop breast cancer (they used the empirically sound method of depending on the respondents' memory) -- What bothers me is that they are using scarce grant monies and seem to have no problem getting their questionable correlations published in the main stream media.

There is seemingly no end to the worthless flawed “studies” that get air-time these days.


To clarify my previous comment -- this study reminds me of the one that was published last year about breast cancer and french fries.


I didnt have the time to properly look for any studies that may support this but could it be possible that an increase in rainfall could result in a higher birth rate?


There may or may not be something to this study, but I suspect that in any case it is probably a good idea to limit TV viewing for children 3 and under. BTW, all autism by definition manifests by the age of 3.

Remember the rainbow man and how much TV he watched? That's anecdotal, but suggestive. I really do believe in moderation in all things. Even things that are good for you in moderation become bad if you get too much. That's not to say that I always follow my own advice, but I should.

Justin Ross

Could it be that people with autistic children find some advantage to being subscribed to TV? I don't know. However I believe this is the case with ADD and ADHD. I like the authors creativity, and it adds a small piece to a big puzzle. Seems to me though there are a lot of people who are letting their emotions on the issue get in the way of evaluating the paper and ranting about other studies. Before they should be allowed to rant about questionable empirics, they should have to identify themselves as qualified to do so. Something simple, like answering the question: "What is b=(x'x)^(-1)x'y?"


What about power failure during rains?How has that been accounted for?


I have two daughters, one has an Autism Spectrum Disorder, while the other doesn't. (Asperger's, very mild, high functioning.) Anyway, she rarely if ever watches TV, while the other, non-ASD child is a couch potato.

Go figure.

(This anecdote about as valid as the study for drawing meaningful results.)

tim worstall

I have to admit that I don't think it's got much to do with TV either. I tend to the same thoughts as meomaxy up above. This paper does a good job:

And yes, I'll admit I'm biased. I write a blog based on Simon Baron-Cohen's EQ/SQ theories. But still, the inheritability and assortive mating sound a much more likely explanation to me.


What about mold and fungus and other side effects of rainy places? As you say, the cable TV one makes no sense since it has been steadily trending upwards with time and so has Autism. Unless you can go into a country with no TV that added it, you cannot say there is more than time based correlation (two things trending together with time).

But, rain and modern house construction makes more sense. In rainy places, people are trapped in humid indoor environments with modern construction materials and semi-air-tight houses. So, you have more exposure to the same possibly contaminated air, etc. One ought to look at cases of Autism and the age of the houses in the area to see if there is enough correlation to look for causality.