« Field experiments for cultural protectionism | Main | Very good sentences »
Which are the most obese American cities?
Memphis wins the competition, but:
Had we included every area on the list, the smaller cities of Huntington, W.V., and Ashland, Ohio, on the West Virginia, Kentucky and Ohio state borders would have far outpaced every city on the list with obesity rates of 45%. Of the 50 cities we did rank, Boston entered last, with only 19%.
Here is the full story. Residents of San Antonio are the most likely to patronize fast food restaurants, with an average (or is it median?) of 20 fast food days a month. I'll note that Latino fast food is better than average and it involves a smaller health penalty, relative to the non-fast food. Here are photos of the most obese American cities, though oddly they show the buildings far more than the people. Would an article about tall or wide buildings show only the people? Could they not find heavy people? Or do they think we simply don't want to look at them?
Posted by Tyler Cowen on November 25, 2007 at 08:55 AM in Data Source | Permalink
Comments
...so these heavily reported "statistics" on American geographical obesity rates 'might' be a little erroneous ?
Duh.
A brief look at the reported statistical methodology is enough to ignore all the media hype on this "story".... and any serious discussion of its supposed societal implications
These obesity rankings were compiled from a very limited, non-random sample of self-reported data collected by uncontrolled telephone calls. Get that... casual phone calls to a slapdash selection of people and totally relying on these people to honestly state their weight & height -- then mindlessly expanding this bogus sample to apply to vast populations of Americans.
Objective basis for these obesity 'facts' does not exist. Of course, the American media never let facts get in the way of a good story.
Posted by: DouglasH at Nov 25, 2007 9:44:46 AM
I almost dropped my fork when I saw my lovely hometown of Milwaukee ranked #17 on this list. Not only do we have a spectacular selection of food choices here, we have 960(!) bars and taverns. Milwaukee, and Wisconsin in general, has a long tradition of hearty appetites and thirsty palates. It is something we are actually take pride in.
If you remember an episode of "The Simpsons", Springfield was named "America's Most Obese City". Homer exclaimed "Eat your heart out, Milwaukee!" Our reputation is well-deserved, but I am afraid we may be slipping.
I invite everyone to visit Milwaukee. You will love it here.
Posted by: Stephen F. Thiel at Nov 25, 2007 9:51:36 AM
It's like spitting into the wind to ask but: what business is it of government's how fat we are?
Posted by: ricpic at Nov 25, 2007 10:11:51 AM
This guy:
http://alltalksports.files.wordpress.com/2007/08/ray-lewis.jpg
http://sports.espn.go.com/nfl/players/profile?statsId=3542
is obese. BMI of 33. I know most people aren't in the Ray Lewis category, but BMI can be a poor measure of obesity for those in shape. People who tend to have a lot of muscle tend to have high BMIs. Perhaps later in life all that muscle will turn to fat and those people will truly be what we consider to be obese, but I don't think anyone would call Ray Lewis obese today.
And this guy is "overweight":
http://sports.espn.go.com/nfl/players/profile?statsId=3664
http://www.washrun.org/clubnews/news/terrell_owens.jpg
He might be a complete lunatic, but I'd say he is in pretty decent shape.
Posted by: AZ at Nov 25, 2007 10:14:45 AM
"what business is it of government's how fat we are?"
negative externalities.
Posted by: odograph at Nov 25, 2007 10:37:17 AM
I'm not sure what kind of "brief look" DouglasH took, but it is not obvious to me from looking at the methodology information on the CDC site that the data can be immediately characterized as bogus. Of course, any telephone based survey will have (well-known) biases, but this particular study does not appear particularly slap-dash in its design or execution. My initial question with the sampling issue is whether the data was weighted appropriately for ethnicity, gender, and so forth on the back end. Another question is whether the statewide samples of 4,000 allowed sufficiently large city-specific samples.
Citing a number like 31.3% without a confidence interval is potentially misleading and there is no way to know whether that differs significantly from the 27.5% in another city. But that's an issue of interpretation and presentation and more a critique of data reporting in the popular press than an indictment of the CDC's methodology.
I will admit that the conspicuous lack of Houston (and big cities in general) on this list is suspicious, however.
Maybe DouglasH or someone else can point to a more thorough analysis of the methodology to determine how much credence to place on the data. I don't think the amount of time any of us is likely to take will decide that issue.
Given the fact that 32% of people reported height and weight numbers that put them in the obese category, and that you would expect the self-report bias to go in the direction of underreporting of obesity, I don't see any reason to assume that the numbers are over-estimating the prevalence of obesity to any significant extent.
Also, can we sort of give the whole "BMI is a poor measure of fat" thing a rest? We all know that BMI is used as a proxy, and that athletes and all people reading this blog who have a BMI over 25 are just mega-buff, muscular individuals for whom the normal rules of thumb do not apply. But surely we can agree that for the majority of those other people in the country, having a high BMI reflects being actually fat, particularly when we're limiting ourselves to those with BMI>30.
Posted by: Sally at Nov 25, 2007 10:42:13 AM
"Would an article about tall or wide buildings show only the people? Could they not find heavy people? Or do they think we simply don't want to look at them?"
Nah, not at all, I think the pictures were just to pad out the article (no pun intended). It looks like those are all stock photos, most of them Shutterstock (no royalties). If you go there and search for each of the cities, the pictures in the article are usually one of the first few hits. (The Memphis picture is the exception, thanks to Egypt and Graceland)
Posted by: Dolohov at Nov 25, 2007 11:59:17 AM
I'm from West Virginia and now live in Boston. The difference in body types is just continually striking to me. (For one thing, here I'm chubby and dumpy, whereas when I visit my parents I'm svelte and stylish. It's discombobulating.)
...I've put fashion in the last paragraph when it oughtn't to be the same thing as obesity, but they all get wrapped up in my mind; people here just care more about their bodies. They smoke less, have better teeth, and just look like they cared before they left the house in the morning. Partly I hate it; obsession over image drives me nuts. Partly the WV approach just depresses me.
Posted by: Andromeda at Nov 25, 2007 12:17:05 PM
"I'll note that Latino fast food is better than average and it involves a smaller health penalty, relative to the non-fast food."
What exactly is that supposed to mean? Is there a reference?
Posted by: Izzy at Nov 25, 2007 2:15:37 PM
Very few people have large amounts of muscle. I hardly seem them even at the gym. Thats not whats inflating bmi.
and Ashland, Oh is in north-central ohio.
And the photos seems appropriate, in that a lot of what is going to cause one city's residents to be fatter than another's is urban planning and layout. Those cities with NOONE walking around in them? they are going to be where fat people live.
Posted by: yoyo at Nov 25, 2007 2:24:09 PM
You have to go all the way to San Diego at #10 or thereabouts to see any mention of race or ethnicity. Clearly, however, race is a contributing factor; blacks and Hispanics have higher rates of obesity than whites, and many of the high-ranking cities have substantial minority populations.
Posted by: Peter at Nov 25, 2007 2:47:29 PM
Obese people rarely crowd in front recognizable landmarks.
Posted by: Yan Li at Nov 25, 2007 5:14:15 PM
> > what business is it of government's how fat we are?
> negative externalities.
What are those? Manufactured by government?
Posted by: rluser at Nov 25, 2007 7:42:33 PM
When did "we all know" anything, much less that "BMI is used as a proxy"? I'm pretty sure that I can find someone who doesn't know that BMI is used as a proxy - my guess is that many people might believe that weight is used as a proxy for health (and even subconsciously adjust for height), but hearing something as technical sounding as BMI suggests that it is "the truth" (whatever that means) and not a proxy. And I think if society is getting healthier (building more lean muscle mass) that BMI would become less correlated with "health" - so I could make the argument that high BMIs suggest better health and not be incorrect based on the data provided, even though I don't think that is what is really happening.
This link provides some sources on the correlation of BMI with body fat, but note that the latest source is 1990.
http://www.fao.org/docrep/t1970e/t1970e03.htm#P277_35684
What are we supposed to say here? If just one person learns that BMI is a proxy for obesity or "fat" or "health" at the individual level then we have accomplished our goal ... of course the simplest thing is to look in the mirror and be honest with yourself - what did Potter Stewart say about the "p" word (I wonder what he'd think about Nip/Tuck)? Same thing applies to obesity.
Posted by: AZ at Nov 25, 2007 8:19:11 PM
AZ, sorry if I wasn't clear that by saying "we all know that BMI is used as a proxy," I only meant those of us reading this post, not the general public, are familiar with the issues because it seems that every time obesity comes up on the blog, we read the same (not inaccurate, but not highly significant when taken across the entire population) criticisms of BMI and people rush to demonstrate how some particular person or persons with high muscularity are deemed "overweight" using the BMI methodology and so forth. It gets kind of tedious after a while.
Posted by: sally at Nov 25, 2007 10:25:06 PM
I'm looking for a source but I'm pretty certain that I've read the opposite about Latino fast food. Perhaps it was specifically about Mexican fast food.
Posted by: frank at Nov 25, 2007 11:11:39 PM
I'm looking for a source but I'm pretty certain that I've read the opposite about Latino fast food. Perhaps it was specifically about Mexican fast food.
Posted by: frank at Nov 25, 2007 11:13:02 PM
Yes! Finally Alabama is ranking high on something!
Posted by: Matthew at Nov 26, 2007 2:40:39 AM
Peter: Re: the ethnic component. It may be that obesity afflicts blacks and hispanics at higher rates. However if you think this is a not a problem for white folks too, then you haven't been to the South. Take a look in any Wal-Mart down here.
Posted by: John S. at Nov 26, 2007 7:30:33 AM
Perhaps I am missing something, but should the emphasis of the story be somewhat different?
If 32% of Americans have a BMI of > 29.9, but only one city in the survey has > 32% of its population with a BMI > 29.9, isn't the real story that cities are comparatively healthy environs (at least wrt BMI) viz non-cities? Putting aside methodological and reporting problems (touched on by Sally, above), the real story in this data is not what Memphis schools serve their students (perhaps that is a real story, but it is not so b/c of this data), but rather how do Memphis schools differ from their non-City neighbors?
Posted by: j.. at Nov 26, 2007 8:22:34 AM
> > > what business is it of government's how fat we are?
> > negative externalities.
> What are those? Manufactured by government?
Um, because it's welfare reducing because fat people are less pleasant to look at? Seriously, though, I've never seen such a patronizing article in Forbes. The word "crisis" is used about twenty times too many.
Posted by: mobile at Nov 26, 2007 10:43:21 AM
Having been raised in Ashland, Ohio, I can firmly attest to one thing: it's people are about 200...
...miles north of the West Virginia/Ohio/Kentucky border. Ashland, Kentucky, fits the bill for that.
Posted by: Mike D at Nov 26, 2007 10:47:50 AM
any attempts to analyse these rates against education levels?
Posted by: anon at Nov 26, 2007 1:35:08 PM
There are a lot of obese people in the world but as far as geographic location, I don't think it really matters. There are some people that you can look at and they don't seem overweight at all but according to BMI, they are. But I do have to say, there are a whole lot of lazy people in the country and lots of McDonald's.
Posted by: at Nov 28, 2007 4:33:58 PM
Posted by: 翻译公司 at Feb 25, 2008 8:07:49 AM
