3 ms·
I thought it would be interesting to see what the commute time trends look like for cities on each end of the arrival times (early vs. late), so I grabbed some
by cwal37 13y ago
I thought it would be interesting to see what the commute time trends look like for cities on each end of the arrival times (early vs. late), so I grabbed some data from the Census Bureau and guessed at rough metro areas (since Nate doesn't clarify what each one consists of, could be MSA, could be city limits, etc.) using county-level data.
I only took the top 5 metros on each end because I had to do a fair chunk of this by hand and I wasn't feeling doing any more. I also threw in a bar of relative densities just for fun.
Here are those results (pardon my excel, for it was not worth opening IDLE): http://imgur.com/a/qCmwB http://imgur.com/a/qCmwB
A couple notes:
1. Big cities are certainly skewing this, both in terms of traffic issues and population density.
2. My metro areas could be totally wrong.
3. This is only calculating the commute of people within the areas I defined, while the article is talking about arrival time at work in (hopefully) the same areas.
4. Yes, I have the word "Average" too many times in the title of the first graph, sorry.
Still, I think the graph paints a neat picture of opposite trends.