Thursday, January 08, 2009

Do Houstonians really drive more vs. other cities?

Demographia has published a short white paper debunking the myth that Houstonians drive more than people in other cities based on flawed 2006 FHWA stats. Those stats show we drive 36 vehicle miles per capita per day, the most in the nation among major cities (table 1, page 3).
In fact, this data is incorrect. The FHWA 2006 data indicates that the Houston urban area has a population of 2,801,000. According to the United States Bureau of the Census, the population of the Houston urban area was 4,353,000 in 2006. It is true that the FHWA and Census geographical definitions vary, however, the land area of the FHWA urban area (1,476 square miles) is greater than the land area of the Census urban area (1,296 square miles). It is statistically impossible for a Houston urban area with a larger land area to have less population than a Houston urban area with a smaller land area.

Actually Houston’s driving is about average: If the urban area population is corrected to agree with the Bureau of the Census data, per capita driving in the Houston area is slightly below the national average for large urban areas. Houston would rank 19th out of 38 urban areas, with daily per capita driving of 23.2 miles, compared to the national average of 23.9 miles. Houston’s daily driving is only slightly more than urban areas with large rail systems, such as Boston (22.9), Washington (22.3) and San Francisco (22.0) (Table 2 and Slide 1).
They go on to point out that TXDoT data shows Houstonians drive meaningfully less (15+%) than the other big Texas cities Austin, San Antonio, and DFW - even with their larger rail network - then further theorize:
...Houston’s more market oriented land use system would provide a better transportation match between homes and destinations. Houston’s driving data is consistent with this interpretation. Without zoning in the city and the unincorporated suburban areas, Houston does not have the planning barriers that so often lengthen travel times from homes to work and other destinations.
Hear, hear!

Labels: , , , ,


At 9:36 PM, January 08, 2009, Blogger Michael said...

Wow, leave it to a group that advocates lack of zoning to posit that a study showing that Houston has high driving distances is wrong and that in fact we have low driving distances which we can owe to Houston's lack of zoning. Huzzah!

Also, as to the studies, none of the populations look correct to me - you would think if they got Houston's population wrong, then DFW's and on down would also be wrong (and indeed, the population figures still look too low from both versions of the study). Or you could just trust that they knew what they were doing as well as Wendell did...

At 8:03 AM, January 09, 2009, Blogger Brian Shelley said...


Your impression of the population being off is wrong. While all the cities have "urban area" populations below their respective MSA populations, Houston is clearly wrong.

Houston's MSA population is estimated at 5.6m currently, and the study shows 2.8m. Dallas' MSA is at 6.1m currently, yet the study shows 4.4m in the study. I assume the study is using a population area that is smaller than MSA.

Again Houston has a larger MSA population than Atlanta, Detroit, Miami, Seattle, San Diego, Phoenix, Boston, D.C., and San Fran/Oakland, yet the study shows all of these cities to have a larger "urban area" population than Houston.

None of the cities are as wildly inaccurate as Houston. Wendell is right and your bias is the one leading to the inaccurate viewpoint.

At 11:08 AM, January 09, 2009, Blogger Brian Shelley said...

One more point on the data. While Houston is incorrect and Wendell's data would push it to the middle, one can see from the table that cities that are more spread out do tend to have more driving. Wendell is probably right that our lack of zoning probably does mitigate the need for driving by better allocating land use. It appears to me that pro-suburban style zoning is more damaging to drive times than anything else.

At 11:13 AM, January 09, 2009, Blogger Michael said...


As I've said before I think the most accurate measure we have of metro populations is CSA, not MSA.

I would also trust the FHWA numbers because presumably they are using highway miles within the same area from which they were measuring the population. You cannot just go in and change one of those numbers without understanding how those numbers were used.

Also, the presumption that:
"It is statistically impossible for a Houston urban area with a larger
land area to have less population than a Houston urban area with a smaller land area."

is not necessarily true. It is only true if the larger area completely includes all of the areas that were in the smaller area, but somehow I doubt that's always the case with estimations of urban area, which seem to be as arbitrary as Tom Delay's political redistricting maps. For instance, we could have a small definition of Houston that still includes Clear Lake and the Woodlands, or a large definition that leaves these areas out and includes larger swaths of relatively uninhabited SE Texas and far east Houston.

The 2.8 million figure is awfully close to the city population of Houston or of Harris county population - perhaps that is what the original study was looking at.

Anyway, I generally don't trust studies when it is so easy to poke holes in the "givens". I'll just have to go with the FHWA and nearly everyone I've ever talked to in confirming that, yes, Houston is a place that is near the top of places in terms of daily driving distance.


At 11:52 AM, January 09, 2009, Blogger Brian Shelley said...


The population of the city of Houston is about 2.2m, and the population of Harris county is 4m (wikipedia)

The study claims the area used was 1476 sq. mi. So, in their study area the population density is thus 1897 person per sq. mile. In Harris County (1778 sq. mi.) the population density is 2250 persons per sq. mile. How can the study area call an area "urban" when it is less less dense than the county as a whole?

The original study has to have made a mistake. Either in tabulation or in orthodoxy.

At 1:01 PM, January 09, 2009, Anonymous Anonymous said...

Even if you use "urban area" instead of MSA data, and even if you go all the way back to 2000, Houston's urban area is 3.8 million. ( This is way above 2.8 million.

Is it that difficult to admit they could have made a mistake?

At 1:17 PM, January 09, 2009, Blogger Michael said...

>>Is it that difficult to admit they could have made a mistake?

I don't see any reason to believe that the US DOT would be any more mistaken or prone to error than a lone libertarian with a pocket calculator who also has an ax to grind. I bet that the 2.8 million number matches the Houston area for which they calculated highway miles (also notice that Wendell adjusts the area, adjusts the population, and leaves highway miles constant - nice) - and so long as they tried to do the same thing for every metro area, then their data is consistent - and their comparison is as valid as can be expected when comparing urban areas across the country.


At 6:11 PM, January 09, 2009, Anonymous Anonymous said...

lies, damned lies and statistics

At 10:06 PM, January 09, 2009, Blogger engineering said...

cannot imagine passenger rail is of any significance in places other than NY and Chicago.
regardless, people decide how far to drive thus the related costs are built in to their daily routine. neither their employer nor government are paying for it thus don't think there is a cost due to driving or delay except personal costs.
if length of driving affects peoples job performances then perhaps there should be a concern for related costs of driving.
also i have driven in several cities. houston's traffic is fairly good.

At 10:50 PM, January 11, 2009, Anonymous Anonymous said...

This is not just about personal choices and costs. There are several social costs associated with driving that are also important and effect quality of life. The environmental costs alone are very significant.

As far as the study, I can't say for sure but I do think they are using different definitions of the respective metro areas given that all of the cities seem to be "off." My personal experience is that I have found that I drive a lot more here than in other cities I have lived. The distances are just so vast and walking/public transit here is generally not an option (unfortunately).

At 9:23 PM, January 17, 2009, Anonymous Anonymous said...

Because if Wendell Cox wrote it, it just has to be true.

At 8:08 AM, January 24, 2009, Blogger Peter Wang said...

Dr. Peter Newman of Curtin University, Perth, Australia showed a slide this past week that showed Atlantans drive the most miles per person of any major city on the planet, followed a close second by, you guessed it, Houston. This is apparently driven by the densities, with Atlanta being somewhat less dense than Houston.

I don't know his data source. You could probably email him and inquire.


Post a Comment

<< Home