Thursday, March 7, 2013

Size Doesn't Matter

A week ago or so when news swirled again about cities going broke, I wrote about Harrisburg and Stockton and their giant public investment gambles being merely the straw the broke the camels back, much like the Guggenheim was merely the cherry on top of Bilbao's sundae.  In other words, there were underlying issues behind the simple explanation.  Harrisburg and Stockton merely pushed the remainder of their chips into the center of the table and lost.  However, their going bust holding 2-7 unsuited doesn't explain their gambling problem or why they were dwindling chips beforehand.

Surprisingly, this data scrolled across my computer from a 2006 Brookings Inst. study ranking metropolitan areas by Vehicle Miles Traveled per capita.  In terms of Metropolitan areas, the greater the per capita VMT, the greater the amount of local money leaves the area, and even worse, the greater the imbalance between infrastructure and tax base.  Stockton and Harrisburg were 4th and 5th in the country (If Richmond, Little Rock, and Jackson, MS aren't worried about their future, they ought to be).  Incidentally, New York City metro is last (1st?) in the country in least amount of VMT per capita.

So I began digging for more data to substantiate the theory that more highway infrastructure per capita tends towards lower density and greater vehicle miles traveled, i.e., things further and further apart meaning too much infrastructural cost burden combined with too dispersed tax base.

Without further ado, here are a series of charts illustrating this information produced from a number of tables I assembled this morning:







The above three charts show the top 50 metropolitan areas, 51-100, and 101-150 in terms of Vehicle Miles Traveled per capita in relation to Highway Miles per capita.  In each case, there exists a direct relationship.  The more highway miles per capita, the more the citizens of the metro are driving.  The largest metros have the greatest amount of noise, but there still exists a direct relationship.  In metros 51-100, it is remarkable how little noise there is compared to the other two graphs.


This chart shows Lane Miles per Capita in relation to percentage of commuters using private automobile (single occupant + carpool).  Interestingly and predictably, having charted all forms of commuting, percentages by metro that carpool or telecommute are the most stable from area to area.  I suspect this is because these forms have the least to do with infrastructural network design and more to do with other issues such as economy, i.e. San Fran/San Jose have a high number of telecommuters.



The above shows non-vehicular commuters (transit, pedestrian, bicyclist) by population density.  There is some correlation, but you'll also notice that NYC and LA are all over the place.  There is quite a bit of noise as the r-squared is lower than just about anything.  Detroit and Dallas-Fort Worth are last of the top 20 MSAs.  Despite having the most extensive light rail system in the country, DFW has the lowest transit ridership.  Square pegs and round holes and all that.



Lastly is highway lane miles to density.  This is the only inverse relationship of all of the above as you see.  More highway lane miles per capita, the lower the density.

Some other data of note:

  • Of top 20 MSAs, DFW has more driving commuters than any other metro except for StLouis and Detroit.  Good company.  I suspect given the exsanguination effect, DFW has merely only recently had their throat cut, while StL and Detroit have long since bled out.  Never fear.  It hurts only but a little.
  • DFW despite the extent of the mass transit, have lower percentage than every other top 20 MSA commuting via transit.  Some of these other metros barely have transit service.
  • DFW highway lane miles per capita seem pretty low (7.39 in 2009 figures), but of the top 20 metros only Houston (10.18) and StL (10.53) have greater.
  • Having calculated the most highway lane miles per capita of all of the top 150 metros, the top 5 are: Greensboro, NC (1.449), Little Rock (1.409), Tulsa (1.246), KC (1.254), Wichita (1.226).  Look at that Kansas hustle.  This is not a list you want to be on.
  • The top 5 in terms of all types of roadway miles per capita of the top 150 MSAs are: Pensacola, Augusta GA, Jackson MS, Little Rock, Birmingham
  • Having charted road miles per capita to density, again there is a very high indirect correlation.  For once, LA and NYC actually have something in common.  They both have very low total road miles per capita.
While transportation planners will tell you it's the lack of density that demands more highway miles (and due to inertia, this is somewhat true), but rather it's the additional infrastructure that then disperses the housing and (delayed) real estate markets, thus demanding even more highway infrastructure.  It feeds upon itself.  However, this belief that transportation planners are merely serving the market as it currently exists without having an additional effect upon that market is like scientists unaware that their observation of tests can change the actions of the test subject they're observing.  Only, this is far more direct.