Peter Gordon blogs about a paper he presented at the Transportation Research Board conference in DC:
My friends and I just presented this paper at the Transportation Research Board meetings in Washington DC. We tested the effects of tolling Los Angeles’ freeways in the peak hours (we tested 10 cents and 30 cents per mile). It’s a simulation on a real network and many substitutions occur. As expected, peak-hour freeway speeds increase, some people switch to surface streets and that traffic slows, some switch to off-peak hours and some (very few) travel less. And politicians take in a lot of money! That’s for the 10-cent toll. The 30-cent toll overloads the surface streets. Many other options can be tested, including only tolling some of the freeways. Planners have voiced concern that tolling the freeways would overload surface streets. There is probably a “sweet spot” that can easily be found. We also plan to look for effects on freight travel as well as travel by income groups.
He’s established that “very few” people lessen their travel. And if the the number of people who switch to off-peak hours is small compared to the number of people who move to surface streets (and judging from my very cursory perusal of the paper, it seems like this is the case), then tolling is just shifting the burden from highways to local roads. This could be a problem since local roads, unlike highways, are paid for almost entirely out of general revenue, not user fees. It seems like the rational thing to do at this point is to argue for tolling local surface streets as well, perhaps through a congestion charge. Maybe I missed it (like I said, I didn’t read the presentation paper thoroughly), but his talk of a “sweet spot” in the summary makes me think he didn’t consider tolling local roads to be an option.