With transportation coffers barely able to maintain highways, let alone adding new capacity to relieve congestion, many transportation economists and urban planners have concluded that the best solution to U.S. freeway congestion is to implement variable pricing on all congested freeways. At the same time, many political scientists and other political observers consider doing so to be politically impossible, primarily due to strong opposition from taxpayers, voters and highway user groups. Federal efforts dating back to the 1970s to induce one or more urban areas to price its freeways have all been unsuccessful, which reinforces the political scientists’ skepticism. But even if the political difficulties could be overcome, the conventional model of freeway congestion pricing may not be optimal, since it could well create more losers than winners.
Analyzing this conundrum requires a closer look at two assumptions implicit in the standard congestion pricing model: uniformly applied (variable) pricing and all lanes treated the same way. Charging the same price to all users during rush hour ignores the huge variability in people’s value of time (and their value of the reliability of trip times). This can vary for the same person from one rush-hour trip to the next, depending on that trip’s purpose and the person’s circumstances at that time. A uniform price applied to all rush-hour travelers will overcharge many and undercharge others, leading to suboptimal outcomes in which the losers will likely outnumber the winners. In other words, only a minority of those using the freeway might find the value of time savings to be greater than the cost of the toll; the others would divert to already-congested surface streets or to much slower transit alternatives. Similarly, the assumption that all lanes must be the same, serving all kinds of vehicles, prevents consideration (on multi-lane expressways) of reserving some lanes for particular kinds of vehicles or trips.
Once those two constraints are relaxed, it becomes possible to consider a multi-pricing approach for freeways. The current trends toward converting HOV lanes to HOT lanes and adding new priced express lanes are the potential first steps away from the conventional model of pricing. The second step should develop complete networks of priced lanes, aimed deliberately at those trips with the highest value on any particular day.
Once the network of premium lanes is in place (paid for by the new toll revenues), the network’s performance may help generate support for expanding pricing to other lanes of the expressway. Drivers still in congested general purpose (GP) lanes may realize that if more of the lanes were priced, the average price in those lanes would be lower, making them more accessible to those with somewhat lower values of time. This could generate political support for converting an adjacent GP lane to pricing, creating a larger constituency for priced lanes.
Finally, if congestion is still a problem on the remaining GP lanes, the benefits of pricing on the (by now) larger system of premium lanes will be obvious to a larger fraction of the population. This could lead to support for implementing modest peak pricing on the remaining GP lanes, flattening the peaks and shifting some trips to alternate modes.
In this evolutionary manner, the entire freeway eventually gets priced, but with at least two different pricing regimes, premium and regular, increasing the likelihood that winners outnumber losers. Political support would also be greater if all the revenues from pricing were dedicated to building, operating, and maintaining the expressway system on which prices were being charged. In selected freeway corridors with a high percentage of truck traffic, specialized truck-only toll lanes would also be implemented. As separate lanes, they could be tailored more specifically to the needs of heavy trucks—e.g., with shallower grades, more durable pavements, etc. In this limited number of freeway corridors, the end result would be three tiers of lanes and pricing: regular, premium and truck-only.
Given the long history of political failure in implementing the standard top-down model of alllanes, one-price-fits-all freeway pricing, this evolutionary, bottom-up model of freeway pricing should be seriously considered.
Conventional approaches toward implementing congestion pricing on U.S. freeways have made little headway. Besides not being able to overcome the political resistance from highway users (motorists and truckers), conventional pricing proposals are likely to create more losers than winners. Research suggests that social welfare would be maximized not with a single price for all freeway users but with several choices of price and service level for various categories of user. This paper has proposed one way to implement such a system, with separately priced lanes for premium-service motorists, regular motorists and heavy trucks. It has also sketched out a possible evolutionary approach to implementing such a system, in which each step can be justified on its own merits, and each creates preconditions for later moving on to the next step. This approach seems more likely to succeed in the U.S. freeway context than conventional “big bang” approaches.