A lively discussion in a recent post about highway taxes vs. highway spending boiled down to one question: “Who drives more: rural, suburban, or city residents?”
The data to answer this question is not easily accessible, but we can use available data to estimate it. We know from National Household Travel Survey (conducted by the Federal Highway Administration) results that about half of all vehicle trips are 4 miles or less in length, and three-quarters of those vehicle trips are 10 miles or less. Only 10% of vehicle trips are more than 20 miles in length.
With this in mind, we can use county-level VMT (vehicle miles traveled) data as a rough proxy:
This map shows 2013 VMT-per-capita by county, using 2013 VMT data from MnDOT and 2013 population estimates from the Minnesota State Demographer. The figures within each county are the value for that county. The lowest VMT-per-capita is in tiny Mahnomen County in northwestern Minnesota, while the highest is in Martin County (Fairmont area).
As you can see, the Twin Cities metropolitan area counties have low VMT-per-capita numbers compared to the rest of the state. Indeed, three of the bottom five counties are in the Twin Cities metro, and all seven core metro counties are in the bottom 11. Olmsted County (Rochester) and Blue Earth County (Mankato) also score relatively low. Meanwhile, all of the highest VMT-per-capita counties are Outstate. From this we can surmise that, on average, urban and suburban residents drive less, while exurban and rural residents drive more.
It would be interesting to parse out how much traffic is freight-traffic moving through the county vs trips generated within the county (or to a neighboring one) for work commutes, shopping, etc. I have to believe that a VMT/capita metric will swing much higher in low-population counties.
I only ask this because the debate has two sides to it: 1) we need good/numerous outstate roads for commerce (freight) – a straight VMT per lane mile argument should suffice, and 2) outstate families drive more than suburban/urban metro families and will thus be more adversely affected by increases to gas taxes/etc – VMT/capita may not capture this the best (though probably a close proxy).
Is there county-level household annual VMT out there?
To my knowledge, there is no county-level household annual VMT. There’s a national-level, and there might be a state-level, but nothing finer than that AFAIK.
Of the other datasets you mention, there are two that I know exist: county-to-county commuting, and freight VMT at the MnDOT district level. Also, VMT-per-lane-mile should be computable at the county-level from available MnDOT data.
MN’s cost of living adjustment has some formulas to approximate household driving rates. They use ACS commute time to determine total miles, citing the NHTS tables showing greater household driving distances for non-commute trips track fairly closely with commute trip length. Obviously, this is an approximation, but the counties calculated include:
Blue Earth: 18,277
St Louis: 20,492
Doesn’t seem to be a pattern between county HH VMT and total VMT/capita when I plot them out, but it’s also a small sample size.
VMT per capita is interesting. Of course, this is just VMT on the roads in these counties, not necessarily the VMT created by the residents of the counties. People cross county lines all the time so it’s tough to tie the VMT number to the county population figure. I think it is generally true that more urban areas have lower figures but it’s probably not appropriate to compare counties to each other with this data.
An interesting per capita analysis that does work is looking at the lane mileage per capita by city for the metro area. This gives you a sense of how much mileage is serving each cities’ population. I have a summary of this data if you want but you can also download the lane mileage data from MnDOT’s website, by city. The one thing about that data is that more rural areas will naturally have higher figures because of important rural roads that aren’t serving much population (which isn’t a bad thing). Comparing fully urbanized cities is more appropriate. Some cities, like Woodbury, are not fully developed but most developed. Not sure how the data should be interpreted for those cities. Teaser: Minneapolis and St. Paul are among the lowest in the metro area. The five cities above them have a combined population of 10,000.
Interesting presentation and good discussion. As Alex Cecchini points out, gasoline tax tends to be regressive. How many remember that Elmer Andersen modernized the highways with the taconite tax?
Only 10% of vehicle trips are more than 20 miles in length (according to NHTS) but they account for almost 40% of VMT. You can really see the influence of I-90 across the southern counties (high freeway VMT, low population).
NHTS would be the data source for household VMT, but the sample size is too low in MN (only 340 households statewide) to look county by county. You could do a similar analysis in states that had purchased add-ons (for example Wisconsin or Florida) and have enough data.
I don’t think the question boils down to who drives more. That’s important, but equally important is the question of whether that driving is productive or not. What is the value provided to the state of all the VMT?