Update: Dean's text is fine, but there seems to be a problem with the image on his website, so I'm linking to Seth's copy of the graph instead. Click on the image for the [big] version, and on Dean's link for the details.
So all we have to do to reverse the obesity trend is to put people into smaller houses??
But before we implement such a drastic solution, we need to understand the system better. Who really benefits from houses being larger? Who really benefits from people being overweight? Who will support obese people in their valiant struggle against better health?
Update: Some other interpretations of the data are offered in Scribe's comments to this blog posting, as well as several comments to Dean's original post. I think this just reinforces my point - that we have to understand the system better before leaping in with ideas for system improvement.
Technorati Tag: POSIWID
It would certainly be foolish to launch into ideas based on the theory of a direct causal link between the 2. In the Roman times (I think), a larger person was symbolic of being more prosperous - in which case they'd be able to afford a bigger house. While the prosperity of America might be in dispute, maybe it'd be worth plotting some kind of purchasing power measure against these as well.
ReplyDeleteSecondly, I can think that there might also be things going on like bigger houses require more space, so tend to be built away from urban centres, so more time spent sitting in vehicles to get to these centres is required. Maybe a plot of "urban sprawl" would be handy...
None of this takes into accoun the overlap between who owns these houses and who's obese, of course. Are the mre obese people necessarily the ones with bigger homes?
Lastly, what is interesting is the way the 2 lines converge only in 1994-95. While obesity seems to be on a constant rise, there's only a sustained rise in house size since that point. Did the average size kick off larger houses for some reason at that point? Did somethihg else happen (e.g. the dotcom boom)? Did the way the statistics are determined change at all?
Nice stuff - shows what can be done with publically-accessible data and a few hours. Even if it's not 100% academically rigourous, it's at least the kind of thing that catches people's (and governor's/minister's) eyes. If only I had time to investigate further... :)