Hacker News Reader

Comment by dredmorbius | Hacker News Reader

transportation systems in the US have co-evolved with the growth and spread of its population

For the post-1950 period, I disagree. The car was established. Development followed auto-driven growth dynamics.

Land use is a regional issue at the scale of metro areas (or CMSAs). Not nationwide.

Average annual mileage for the typical American is 15,000 miles/year. That's 40 miles/day, or 20 miles one-way radius. Even assuming no weekend travel (all commutes are work related) you're still at 75 miles, or 37.5 miles one-way.

That is: transportation use is a local phenomenon. Distribution of population centers across the country doesn't matter, land use within those centers does. And those patterns have developed and overwhelmingly been driven by auto use for the past 65 - 75 years, with very few exceptions (e.g., NYC).

View on HN

Your figure makes more sense as a per driver figure. There are about 100 million Americans who do not have a license to drive (and reason to believe that most of them do not drive illegally):

http://www.fhwa.dot.gov/ohim/onh00/onh2p4.htm

The passenger vehicle fleet is also about 250 million (estimates arriving at that figure are including pickups that are used as cars).

I guess the typical American might be a modal driver, but it seems odd to use a definition of typical that discards ~1/3 of the population in question.

That utterly misses the point that 1) most households have at least one driver, and that the typical distances travelled on a daily basis are consistent with suburban, auto-centric land-use patterns.

You can corroborate this with other sources, e.g., US Census commute patterns. Insurance and US DOT should also have driving stats.

It doesn't miss any point, it's just pedantic. The mean American vehicle doesn't even go 15,000 miles a year (there's only ~12,000 available according to gov statistics), it's a little impossible for the mean American to do so. Of course average and typical have other meanings than the arithmetic mean (but I pointed out a problem with using the numbers as the mode in the previous reply).

I'm not going so far as to argue about whatever point it is you are making, I'm pointing out that somewhere in either your characterization of the numbers you are using or the numbers themselves, there is an inaccuracy.

I've largely abandoned all hope of achieving some reasonable understanding here, but just what case are you making?

That the monthly miles driven should be 12k rather than 15k? A 20% delta doesn't change my fundamental argument at all. 15k vs. 1.5k vs. 150k would.

That I'm incorrect in my description of regional growth dynamics 1950-2015? You've not argued that at all, and it's the information I'm principally interested in.

Here's the fucking FHWA US DoT stats: http://www.fhwa.dot.gov/ohim/onh00/bar8.htm

Range: low of 4,785 (females 65+), high of 18.9 (males 35-54), total combined average: 13,476. That appears to be from a 2000 study, and alternate data shows growth since then:

http://www.advisorperspectives.com/dshort/updates/DOT-Miles-...

Oh, here's the most recent report year, for 2011.

3 trillion vehicle miles, 246 million registered vehicles, 210 million licenced drivers, 68.5% of the total population.

Maths gives us 12,195 miles/vehicle, 14.285 miles/driver.

I think we've litigated that point to fucking death, and shown my initial round-number estimate was reasonably accurate.

I was loose in saying "per American" rather than "per vehicle" or "per driver", but again, it doesn't materially change my main point concerning urban-area growth (the vehicle / passenger miles data are used to illustrate it).

If you are in fact making a minor / pedantic point, it helps to call that out. I'll preface my own such comments with an "NB:", "ObPedant", or similar comment.

Otherwise, I actually do appreciate corrections and clarifications. But prefer they're clearly given as such (though I'll accept them even when they're not).

My objection is that you pasted a veneer of numerical support around your conclusion.

If you had said "I think people tend to commute a medium distance on a daily basis and community development has followed along that dynamic" I wouldn't have anything to object to, the problem is that you took a coarse estimate (the 15,000 miles) and then eventually stated it as a possible "37.5 miles one-way". That should be an "about 30 or 40", not a 37.5. Given the wide variance, it's also not a great place to use the mean as a representation of the typical.

So my point, if you are going to present your conclusions as being based on/supported by numerical analyses (which you do all the time...), please do be careful to not to exaggerate the precision with which they are done.

I posted an estimate and a check of that by way of sensitivity analysis, and have supported the point with further data.

For someone grousing about numeric precision your own lack of verbal precision (or accuracy) in noting issues is richly ironic.

I'd appreciate an explanation of where I've been inaccurate.

I agree that I raised a concrete objection in my initial post and did not mention the abstract motivation for raising the objection, so no need to pursue that aspect of it.

> Development followed auto-driven growth dynamics.

Development was also driven by things people wanted: Their own land. Their own house. Escape from the noise, pollution, and general harshess of industrial zones. Land was more affordable in the US than in Europe and Japan. Building probably was as well. Cars enabled the kind of development that people wanted.

Reply on news.ycombinator.com