Tuesday, May 13, 2014

Statistical Surreality

Being armed with facts only makes you dangerous if you don't know how to use them.

For example, one might make the assumption that highways are a necessary part of urban life and too, building more highways would relieve congestion, somehow mysteriously clean the area, and reduce commute times.  One might also make the assumption that since Texas is growing in population (or at least has - as a state and DFW as a region) therefore more highways are necessary.

These assumptions are rife within a certain sect of seemingly libertarian minded people that really, really like freedom.  Not yours.  Just theirs.  And the only thing they like more than free-dom is free-things.  Like free highways and lower taxes and low density housing.  How any of these things are paid for when interconnected is beyond me.  I choose the freedom from thought.

So let's unpack some of these misconceptions:

We need more highways because... growth.



The above is highway capacity per capita within city boundaries.  Most of this data tends to cover the metro.  However, the above was measured by me as I keep a growing table of the burden urban centers must bear for the sprawl their governments subsidize.

There is belief that new people will mean more people on the road and thus we'll need more highway capacity.  That is true, of course, if you're only capable of understanding one form of city living: car-dependence.  Portland and Seattle, those fiscal conservatives, have built a third of the highway capacity per capita as Dallas.

Building more highway capacity only serves to increase the amount people drive, as shown below:



The more highways we build, the more people drive.  And the more people drive, the more they ARE the congestion.  Therefore, we're not solving the problem of congestion only exacerbating it.

We need more highways because...congestion.


















The above is an excellent bit of research by Shane Phillips.  The red line shows the increasing number of billions spent each year on highways.  The green line represents the national travel to work time, which has been steady at 25 minutes.  The funny thing is, this is about the typical travel time throughout the history of cities if you're to believe Professor Peter Newman's hour-wide city theory.

The reason for the increased spending on highway capacity, supposedly, is to reduce congestion.  However, the cost of congestion isn't decreasing according to TTI's annual Urban Mobility Study, having increased from $115B to $120B in three years.  In effect, we're spending hundreds of billions each year to marginally reduce the $100 billion or so congestion is costing the country each year, and that policy is not even not successful, but actively making the problem worse.

We're spending hundreds of billions with no discernible positive outcome, but plenty of negative outcomes and externalities.



While we should be focusing on the real waste which is car-dependence, costing the country $2 trillion per year.



Let's see, does population density affect commute times?



Nope, not really.  How about highway capacity?  Does that speed up commutes?



Nope.  Commutes hover around 25-30 minutes.  The longer ones are accompanied by significant increases in median income as well, the outliers being San Jose and NYC.  In other words, people are willing to go above and beyond average commute time for significantly higher pay.

Well, maybe building more highways and all that public money is what we need to jumpstart the economy and increase GDP...

We need more highways because...jobs.


Well, no.  There is no correlation between between vehicle miles traveled and GDP.  And since we know there is direct correlation, er in fact causation, between highway capacity and VMT, we can assume that highway capacity does not affect GDP.  Meanwhile, those urban centers heavy on highway capacity are hemorrhaging jobs: KC, Detroit, Dallas, Atlanta...

Assuming that more highway capacity to address the issue of congestion is also to assume that more highway capacity has no affect on the world around it, which in fact, it has extreme impact.  It's like the scientist who doesn't take into account his own affect on the experiment.  The only thing inner-city highway capacity is proven to do is make people drive more.  That doesn't sound like freedom to me.

The last assumption is the one that is the most annoying to me from the various pseudo-statisticians and economists, that growth lasts forever and carrying capacities don't exist.  Therefore we can extend trend lines in perpetuity.  The reality is there are carrying capacities and Texas is not exempt from this certainty.  There are issues of water, ag production, and the , but also issues of debt.  Sure we're growing now, but who will want to move here when Texas can't afford to maintain all that new shiny highway infrastructure (let alone service the debt) in a low tax political climate?