Cumulative Model

INRIX has released its 2008 Annual Report on congestion. (Via Tom Vanderbilt)

The key finding is "a decrease in the Travel Time Index of 3.5% from 2007; more than
reversing the increase of 1.9% between 2007 and 2006".

This conflicts with my experience and analysis of fuel economy. For most of 2008, I've felt that my drive times were longer, except in recent months. This agrees with the decline in fuel economy. For gasoline, fuel economy declined in most of 2008 and only rebounded slightly at the end of the year. It finished down for the year.

However, a little further thinking reconciles my analysis with the INRIX Report. When diesel and gasoline fuel economy is looked at, fuel economy does begin improving earlier in the year and does end as a net improvement for the year, in agreement with the INRIX report.


Here's what I think is happening.

I think there are two traffic flows that exist at the same time.
Car traffic may have gotten worse while commercial traffic became more efficient.

My thinking is that car drivers responded to gas prices by erroneously accelerating and driving slower, this worsened fuel economy for car drivers. Commercial driving habits probably didn't change much for non-freeway driving. Trucking probably benefited from decreased congestion and also improved fuel efficiency by driving at lower top speeds on the freeway. Car traffic suffered from decreased throughput at intersections and decreased speeds for non-highway driving (below 55MPH, higher cruising speeds are more efficient).

UPDATE (4/27/09): I e-mailed INRIX at the time of this post about some shortcomings in their report, I received no reply other than acknowledgement of receipt. Their reporting is based on data from commercial vehicles. Commercial vehicles behave differently than passenger vehicles. They accelerate more slowly, have cargo that shifts, hit more red lights, etc. Generally, most traffic flows around them, interfering slightly as they merge in front after passing (i.e. diminished passanger vehicle traffic flow could result in slight improvement in commercial vehicle traffic flow). Also, INRIX data compares traffic relative to "free flow" rates. A decrease in free-flow rates will exaggerate improvement in traffic flow. Their reporting also will not recognize diminished flow during non-congested hours, which can be even more costly than gains during congested hours.

I'll do Ed Glaeser one better.

Replace the Interest Rate Deduction with a Principal Payment Deduction. Sure, it'll break some banks with a bad mix of Mortgage Backed Securities holding prepayment risk, but fuck 'em.

This will bring a soft landing as people with large amounts of cash are encouraged to buy houses, without lots of leverage. Demand will be buoyed and banks will be flooded with cash and will have to lend at lower interest rates to compete. It will also move properties to less risky owners. Fast.