Lies, Damn Lies, and Statistics
Energy alternatives and infrastructure investment are likely to be important domestic policy issues in the coming year at the national level. When the Lehigh Valley addresses the issue of land-use and transportation planning, it will be important that the discussion be honest and wholistic. There will surely be plenty of statistics tossed about. In anticipation, here’s an excerpt from a post by (law prof.) Michael Lewyn at Planetizen:
After digging around through a big pile of statistics, I realized that there are so many different ways of measuring transit ridership that one can easily prove either that ridership is going up or that ridership is going down.
Similar statistical games can be played with highway transportation. If you want to “prove” that not enough of America has been paved over, compare the number of lane miles built to the number of vehicle-miles traveled. So if lane miles increase 300% in Sprawl City and vehicle-miles increase 400%, you can argue that we haven’t built enough roads- even if your region is honeycombed with expressways, every major street has eight lanes, and there is nary a bus or train in sight. Of course, this technique tends to create a kind of unending circle of construction: more highways mean longer commutes from new suburbs that the highways have opened up for development, which means more miles traveled, which in turn can be used to justify more highways.
On the other hand, if you want to argue that America has already been paved to death, focus on the raw number of miles built, or compare lane-miles to population. So if lane miles have increased by 300% in Sprawl City while population has only increased by 50%, obviously the highway lobby is out of control.
In sum, you can prove a lot with numbers- as long as you are careful which numbers to use.
This may not provide much in the way of clarification, but it highlights the importance of requesting sources of information/statistics and of understanding (or at least trying to understand, often in my case) how a particular set of numbers was calculated.