NHTS 2017
In a vehicle, per person
59 min
Average American, driver or passenger, including people who did not travel that day. People who actually went somewhere land closer to 80 minutes in ATUS/NHTS comparisons.
What the calculator leaves out
The calculator prices an extra vehicle-mile: congestion, crashes on strangers, air, climate, parking. These papers are about a landscape that requires driving — time in the day, a transit place versus a car place, inactivity, years of life, shops. They are not turned into ¢/mile here. How much driving there is in the US, Europe, and Canada is a third page. The sources are linked so you can disagree.
Time in a day
Your own hour in the car is mostly private — that is why Parry does not put it in the externality pack. Congestion delay you impose on other people is already on the calculator. This section is the hour itself.
Marchetti / Zahavi
A stubborn daily budget
~1 hour
Across cities and decades, people tend toward about an hour of travel a day. Faster roads mostly buy more distance, not a shorter day. Looked at closely, the “constant” is an average, not a law.
Induced travel
The speed dividend is spent on miles
FHWA’s Conditions and Performance report had Americans spending 88 minutes traveling per day in 2001 and 86 in 2009, about 80 of those in a vehicle. NHTS 2017 is 58.6 minutes per capita in a vehicle (Figure 7) — lower because it includes non-travelers. Marchetti 1994 restates Zahavi: mean exposure around one hour; people who own a car use it about an hour a day. Mokhtarian & Chen 2004 review the travel-time-budget literature and find the one-hour figure is much less stable once you leave national averages. The policy implication still holds at city scale: time “saved” by a wider road is often reinvested in a longer trip (Lewis–Mogridge / induced demand).
A transit place versus a car place
Clock commute times across large US metros sit in a narrow band. Miles, mode, and how many cars you own do not. Newman and Kenworthy call these walking, transit, and automobile urban fabrics — each roughly one hour wide at the speed it was built for.
ACS 2022
One-way commute, US
26.4 min
Census brief. Metro areas 26.8 minutes; longer outside principal cities. Large metros cluster around half an hour. New York is usually the long one, not because driving is slow everywhere there — because many people ride transit.
Atlanta SMARTRAQ
Inner counties vs outlying
59 vs 72 min
Same region: closer-in counties about 33 miles / 59 minutes in the car per day; outlying about 44 miles / 72 minutes. More car-place, more of the day sitting.
Brookings, 2005
Annual VMT per person
5,890 vs 11,200
New York metro versus Atlanta. Dated, and still the direction later maps show: Northeast and the Bay low, Southeast high. Clock time similar; miles not.
Do not read the mode averages as “transit is slower, so skip it”
Transit one-ways are longer because of where transit riders live
FHWA: about 49 minutes one-way on transit versus about 24–27 driving alone. That is a mix of cities, walk, wait, and congested corridors — not a fair door-to-door test in the same place. Census ACSBR-018 (2022 commuting). Atlanta inner/outer miles and minutes: SMARTRAQ / Frank. VMT: Brookings MetroPolicy table (2005). Fabrics: Newman, Kosonen & Kenworthy 2016. Household money, not minutes: location-efficient neighborhoods (compact, mixed use, transit) show lower transportation spending because they need fewer cars — CNT Housing + Transportation Index.
Health, sitting, and insurance
Sitting in a car is sedentary time. Medical bills get pooled in premiums, Medicaid, and Medicare — that is the insurance channel. Inactivity is all sitting (jobs, screens, cars), not a driving-only line. Litman has argued it may be the largest social cost of car travel; it is not in Parry’s table.
Frank et al., Atlanta 2004
Each extra hour in a car
+6% obesity odds
Each extra kilometer walked: −4.8%. Land-use mix had the strongest urban-form association. Cross-section, 10,878 people, SMARTRAQ.
Carlson et al. / CDC
US health spending tied to too little activity
11%
About 11% of adult health-care expenditures associated with not meeting activity guidelines (~$117–131B in that paper’s dollars). Later updates are higher. Cars are one reason people sit; not the only one.
van den Bijgaart et al. 2024
Add inactivity to the fuel tax
+44% US
People do not fully internalize the health gain from walking or cycling. Including it raises their second-best US fuel tax 44% (38% in the UK). Per active mile, they say the health benefit can dwarf unregulated carbon, air, congestion, and crashes.
The other insurance problem
A car landscape also prices out people without a car
Frank, Andresen & Schmid, AJPM 2004. Carlson, Fulton, Pratt, Yang & Adams, Prog Cardiovasc Dis 2015. van den Bijgaart, Klenert, Mattauch & Šulíková, Economica 2024. Litman’s compilation: If Health Matters. Urban Institute (2023): more than one in five adults without a vehicle and with poor transit forgo care because of transportation. That is a cost of driving everywhere even for people who do not drive.
Longevity
Hazard ratios and county life expectancy, not ¢/mile. Compact places look better after you control for income and smoking; that is still not a randomized trial.
Hamidi, Ewing et al. 2018
Compact vs sprawling US counties
~2.5 years
606 metro counties. Doubling their compactness index associated with about 3% higher life expectancy — roughly two and a half years on a 78-year baseline — via more VMT (crashes) and higher BMI, plus a leftover direct path they cannot name.
UK Biobank · Celis-Morales
Cycle commuting, all-cause mortality
HR 0.59
About 41% lower all-cause mortality for cycle commuters versus non-active commuting, after measured confounders. Walking commuting was associated with less CVD, not with all-cause mortality. England and Wales census follow-up over 25 years finds the same direction.
Sources
County sprawl, then commute mode
Hamidi, Ewing, Tatalovich, Grace & Berrigan, IJERPH 2018 — cross-sectional SEM, not a natural experiment. Celis-Morales et al., BMJ 2017. Patterson et al., Lancet Planetary Health 2020. Ding, Gebel, Phongsavan, Bauman & Merom, PLOS One 2014: 120+ minutes of driving a day had the strongest, most consistent associations with poor health behaviors and outcomes in mid-life and older Australians.
Local business
Street design and parking mandates, not the extra gallon. Merchants often overestimate how many customers arrive by car.
Shoup 2014
Required parking on a shopping center
+67% / +93%
Los Angeles: minimums raise construction cost up to 67% for structured parking, 93% underground. That cost is in the price of groceries whether you walked or drove.
NYC DOT 2013
Safer streets and sales tax receipts
Often up
Plazas, protected lanes, SBS: locally owned retail receipts often beat borough and control streets. One Brooklyn plaza: +172% versus +18% borough-wide (excluding auto-centric businesses).
Clifton, Portland
Spend per visit vs per month
Drivers ≠ more
Drivers often spend more per visit. Walk, bike, and transit customers come more often. Monthly spend is competitive or higher except at supermarkets.
Sources
Parking as a hidden tax on the shop
Shoup, The High Cost of Minimum Parking Requirements, 2014. NYC DOT, The Economic Benefits of Sustainable Streets. Clifton et al., OTREC / Portland State intercept surveys; Examining consumer behavior and travel choices. Review: Volker & Handy, Transport Reviews 2021 — bicycle and pedestrian facilities generally have positive or non-significant effects on abutting retail and food service.
Why this is not on the calculator
Parry priced a marginal mile in a given land-use pattern. Inactivity, shop sales, and years of life are mostly substitution — walking instead of driving, or living where walking is possible. There is no published ¢/mile in this file that I would put next to his 5¢ congestion line without mixing units. If a paper publishes one, it can go in estimates.json like the rest.
Cross-sections overstate if healthier people choose walkable places. Cohort commute studies are stronger on mode, weaker on “what if the suburb had a train.” National inactivity dollars include every couch, not only the driver’s seat.