, , ,

Correlation, the close and coincidental association of paired sets of data,  does not equal causation, but it is a hell of a lot more likely to mean causation than no correlation.  Sometimes the causation is not one of the other, but of something else acting on both data, like obesity and heart disease.  Often, the causation is obscured by some much greater cause, like the amount of walking in a  perfectly designed neighborhood surrounded and served only by traffic.  Other times, the correlation indicates no causation, or at least a fanciful causation at best, like the correlation of murder rates and Internet Explorer browser share.

The reason I’m writing on this today is after coming across this fun graphic on Wikipedia this week.  



Of course, the zeroless X and Y axes are meant to deceive, but the point is clear.  There is some correlation between commute times and transit use.  This could say that transit does nothing for traffic, or even makes it worse, but I suspect that transit use is actually a reaction to long commute times.  When a place is successful enough that commutes get long, the time and location cost of transit become more appealing than sitting in traffic every day.  

There are two questions form this chart: what else correlates with transit use, traffic or commute times, and is the only thing that encourages transit use or construction traffic congestion?

More on that Monday…