Cool but Disturbing Tool Maps U.S. Commutes

Want to know how many people commute from one zipcode to another?  Here’s a nice graphical tool by Harry Kao.  Input a zipcode and it shows the commute distribution by origin zipcode, from the 2000 Census.  I put in 94111, the San Francisco Financial District.

Commute map

The visual representation encodes a dreadful bias, though:  It assigns these commutes graphically to the road network, as though everyone travels by car!  In so doing, it appears to give evidence that San Francisco needs an Embarcadero Freeway!

Be careful with maps!  Good maps are designed to tell a story.  But automated mapping tools such as this one can make maps that are incredibly detailed, scientific-looking, and utterly misleading.   This happened often enough when authors of maps actually looked at their own maps. But with automated mapping we face a world in which “authors” of an automated mapping scheme never even see the map they produced — until it shows up in somebody else’s argument.

9 Responses to Cool but Disturbing Tool Maps U.S. Commutes

  1. TLP September 16, 2010 at 5:18 am #

    It would be nice to see modal splits too (or at least transit vs auto, but Google Maps can do walking and biking directions now) but that would only work in areas where Google has the transit network mapped. This is a nice start though to visualizing how people move at a fiarly granular level.
    Another difficulty is that the data is from 2000, and there isn’t really a way to update it. The author/mapper references this on the first page… The CTPP was superceded by the ACS, which doesn’t get down to the census tract or even zip code level; it’s a wider, less detailed survey. The 2010 census should give more detailed info, right?

  2. Ant6n September 16, 2010 at 6:44 am #

    I thought the american community survey goes all the way down to census block data on origin destination data.

  3. Rob Fellows September 16, 2010 at 6:48 am #

    You make a good point — but the other thing to take away from this is that the trips that the Embarcadero served didn’t just disappear when it went away, as many believe…

  4. Rob Fellows September 16, 2010 at 7:05 am #

    I guess to expand on my previous comment — people can make the wrong assumptions about travel by looking at the road system itself, not just at maps. I’ve always found it interesting to watch brainstorming sessions of local officials or the public trying to figure out the best place to put a rail transit line – and invariably the system maps they draw look exactly like the freeway system. That’s where they picture people traveling! Of course that leads to putting stations in the places pedestrians least want to be, and assuming that long distance regional trips are the best rail market.

  5. James D September 16, 2010 at 8:09 am #

    Is there a table in there somewhere that has the zip-codes ranked by number of inward commuters?

  6. Zoltán September 16, 2010 at 10:26 am #

    Are those red dots essentially acting as zone centroids, the zones being the zip code areas?

  7. Eric Fischer September 16, 2010 at 11:32 am #

    The 2010 census didn’t even ask about workplace, so the American Community Survey is all we have to go by any more.

  8. Digby September 16, 2010 at 3:24 pm #

    I plugged in a few zip codes of smaller cities where I’ve lived and the results were a bit screwy. The results could be badly misinterpreted if you don’t have a technical background. Still, it’s a good place to start looking at trip movements

  9. SpyOne September 19, 2010 at 2:12 am #

    I think there’s a more significant problem than the fact it assumes travel is along roads: it assumes commuters are all moving from the center of one zip code to the center of another.
    Looking at your map and taking it literally, it seems that Alameda needs only one transit stop to serve all the commuters going there.
    However, this is still a handy tool. One just needs to view the red dots as where the local surface transit needs to collect people from the surrounding area, then a system that connects the red dots rapidly serves the commuters.
    And while I agree there is some road bias, it seems to me that your map does not suggest the need for an Embarcadero Freeway so much as it suggest the need to move large numbers of people along that route.