We're in the deepest recession in half a century, and people are cutting back, and from ONE DATAPOINT, effectively, we're supposed to believe that it's the "end of the car culture" ?
> To sort this out, I built a regression model that accounts for both gas prices and the unemployment rate in a given month and attempts to predict from this data how much the typical American will drive.
Note that this model is tuned using 15 years of data, and yet we're in a mini Black Swan event that happens every 50 or 100 years.
OF COURSE there's divergence of reality and model.
I moved to Boulder, CO on Saturday. In good part because I hate driving. It's hard to really grasp how much more enjoyable it makes life until you experience it. It's not just being car-free though, but the community that follows.
I lived in NYC for awhile and while it's car-free it's still pretty awful. You still have the same 30 minute commute -- there is no sense of geographic community. Unless you just refuse to make friends with people outside your neighborhood, every time you go out you're gonna end up on the train for who knows how long and-- oh wait, the cross-town is out of service and the train doesn't stop after 2am at blah blah blah. The stress level is easily on par with rush hour gridlock.
I've always found it amazing how people put up with rush hour traffic commutes. To waste a solid hour (or more), 5 days a week, doing something thoroughly unpleasant seems insane to me. I'd be willing to trade off a lot in life to avoid doing that.
Commuting on public transport is OK as long as I can read. Bikes are OK since they're fun.
Predictions via regressions should always have confidence bounds. I am skeptical of the conclusions given that he has prediction for 1980. In order to estimate predictions he probably ran his model on 1970 - 1979 data. The confidence in that prediction is a lot worse that the 1999 prediction and should impact all future predictions. I know it is an Esquire article, but I rarely see his posts go into these important details.
> To sort this out, I built a regression model that accounts for both gas prices and the unemployment rate in a given month and attempts to predict from this data how much the typical American will drive.
Note that this model is tuned using 15 years of data, and yet we're in a mini Black Swan event that happens every 50 or 100 years.
OF COURSE there's divergence of reality and model.
I'm amazed that anyone can take this seriously.