A few weeks ago, the Antiplanner reported on a questionable change in transportation data published by the Bureau of Transportation Statistics. An even more questionable change can be found in table VM-1 of Highway Statistics, an annual report published by the Federal Highway Administration.
Before 2009, Highway Statistics regularly appeared months before the Federal Transit Administration published its annual National Transit Database for the same year. There may have been good reasons for that: the highway data depended on reports from the 50 states and District of Columbia, while the transit data depended on data from nearly 700 transit agencies (as of 2008; more than 850 today). Collecting, reviewing, and collating all that data no doubt took a lot of time.
After Obama took office, a funny thing happened: the highway data started coming out after the transit data. In some cases, not just months, but years after. For example, the on-line 2009 Highway Statistics is still missing some minor tables, and one of the most important tables about highway finance, HF-10, is still missing from the 2011 edition.
When the first tables finally came out for 2009, I noticed a major change in table VM-1, which shows total vehicle miles traveled by type of vehicle. The table also has a line showing person-miles of travel based on the average occupancy for each type of vehicle. In the 2008 table, the average occupancy for “short-wheelbase, light-duty vehicles,” was 1.57 while for “long-wheelbase, light-duty vehicles” it was 1.73. These numbers came from the 2001 National Household Travel Survey, which collected data from about 70,000 households. The 1.57 was for cars and 1.73 was for light trucks, meaning pickups, SUVs, and full-sized vans, and while these aren’t exactly the same as short- and long-wheelbase vehicles, whoever put the 2008 VM-1 together decided to use the same numbers. (I have no idea why they switched from “cars” and “light trucks” to “short” and “long wheelbase,” a change that took place in 2007.)
In any case, the 2009 VM-1 reduced occupancies for short-wheelbase vehicles to 1.39 and for long-wheelbase vehicles to 1.34. The average for cars and light trucks together was 1.38, down from 1.64 in 2008. This reduced passenger miles in cars and light trucks by more than 15 percent. Occupancy assumptions for motorcycles, buses, and heavy trucks all remained the same.
Prevention is discount viagra important source always better than cure. However, it is advisable prescription canada de cialis to consult a doctor on the dosage consumed. Prevent the firm regarding damaging men and women. purchase generic viagra opacc.cv The British Council works with Hay at all its festivals around the world, providing enormous opportunity for global cultural discounts on levitra exchange between writers and readers.
When I discovered this, I contacted the Federal Highway Administration to find out where the new numbers came from. In an email exchange, the person who put together table VM-1 replied that they were based on the 2009 National Household Travel Survey, which collected data from more than 150,000 households. I pointed out that the survey found that average auto occupancies had increased, not declined, since the 2001 survey (see table 16 of the summary report).
They responded that the numbers reported in the summary used the “miles-based method,” whereas the occupancies they used for table VM-1 were based the “trip-based method,” which shows lower numbers. Since table VM-1 purports to show occupancies per vehicle mile, not per trip, I didn’t think the trip-based method was the correct one to use, but I asked to see their calculations. They refused to provide them, saying they were proprietary.
Having reached a dead end, I asked others to look into it, but they too reached a dead end. Since then, I have continued to use 1.67 as the average occupancy for cars and light trucks, which is the number reported in the 2009 survey. This also means changing the numbers in the latest iterations of National Transportation Statistics table 1-40.
Was reducing occupancies in table VM-1 a political decision? Did someone wanted to de-emphasize the importance of cars, or perhaps exaggerate the environmental impact per passenger mile? The Antiplanner can’t prove anything, but I hope the next administration will revisit this question.
Never ascribe to malice that which is adequately explained by incompetence.
The Federal Highway Administration “refused to provide [their calculations], saying they were proprietary”
You’ve got to be kidding. Dirty rat bastards.
How about an FOIA request?
Henry Porter,
Good point. Still have to wonder why they made the change when they used the NHTS results in all previous Highway Stats.
I agree with Frank. FOIA is your friend.
Calculations aren’t proprietary. That’s ridiculous.
FOIA should help. But….
Unfortunately, many governmental entities now treat FOIA as meaning they don’t have to give you anything unless you file an FOIA request. That was not the intent – the intent was that FOIA would provide a remedy in the (hopefully rare) event that the government would not turn over something it should.
Further, most government entities now routinely deny the requests knowing how difficult, expensive and time-consuming it is to appeal their denial.