Jams, snarls, gridlock: These are traffic facts of life in metropolitan America, nowhere more so than in parts of California, home to some twenty-two million cars. Yet it doesn't have to be that way.
In grappling over the last twenty years to meet clean air goals, transportation analysts have come up with a number of workable schemes to reduce the number of cars on the road at any given time. These solutions include congestion pricing policies, whereby drivers pay to use freeways and major arteries during periods of peak demand.
As analysts and government officials have come to recognize, however, if and how traffic congestion is eased depends very much on those in the drivers' seats. Without public support, no plan to reduce congestion will work. For now, more drivers seem willing to sit in traffic than to pay to alter driving habits. But given time and more traffic, that attitude may change.
Meanwhile, interest in approaches like congestion pricing already is high among transportation planners, in large part because the old ways of dealing with traffic do not seem to be working any more. Building new roads to ease congestion, for instance, is no longer the obvious solution it once was. Public budgets are tighter and the environmental and quality-of-life repercussions of sprawl-induced travel are more in evidence.
These constraints pinch tightest in Southern California, where concerned groups across the political spectrum have established the REACH (Reducing Emissions and Congestion on Highways) Task Force to examine market-based alternatives for improving mobility rather than pouring more concrete. Organized by the Southern California Association of Governments, REACH includes reprecentatives from local government, state environmental and transportation agencies, local business groups, and environmental and consumer groups.
Last year, RFF Senior Fellow Alan Krupnick and I, along with Anna Alberini of the University of Colorado, worked with REACH to develop a telephone survey, which the California survey research firm Godbe Research and Analysis then administered to a sample of Southern Californians. Essentially, we asked respondents if they would be willing—with and without several different incentives—to pay a "user fee" to drive on freeways during rush hour. This article presents the results of that survey.
Attractive but Unpopular
From an economic point of view, collecting user fees to drive busy roads during peak demand is by far the most attractive way to curb traffic. If we add up all the costs, including the inconvenience associated with restricted travel and/or having to find alternative routes or modes of conveyance, we find that no other rationing method provides a given level of road service at lower total cost. Other approaches to rationing roadway use, most notably "high-occupancy-vehicle" (HOV) lanes restricted to cars carrying two, three, or four occupants, are less efficient. Yet it is HOV lanes that have been implemented in many places, while time-of-day user fees are rare. In North America such fees are collected in only two places: SR91 in Southern California, a private road built on public land (in the median of an existing freeway) and Route 407 in Toronto, Canada, designed from the ground up for electronic fee collection. [image - cars in traffic] Photodisc Inc.
If they are so attractive economically, why are congestion fees so unpopular? Our conjecture is that people have perceived the fees as tax increases with no discernible benefit. Most surveys that have tested public support for congestion tolls on freeways have been vague about how the revenues collected would be used, although it is generally understood that they would be used for public transportation. In fact, such revenues could be used to compensate for reductions in other existing taxes, such as those on sales and gasoline. Because we suspect that support for congestion fees would grow if people received assurances that a portion of the money collected would be returned to them in some specific fashion, we took a different tack in our survey. Among other things, we wanted to see how sensitive the level of support was to the amount of revenue returned.
The RFF Survey
After focus group pretesting and consultation with the REACH Task Force, the survey that the RFF team developed was administered to 1,743 freeway users (ages 18 or older) in the California counties of Los Angeles, Orange, Riverside, San Bernardino, and Ventura during August and September 1996. The objective was to estimate how the respondents would vote in a hypothetical referendum on alternative congestion fee policies with and without various revenue recycling options. Those surveyed were asked for their reactions to a "base fee policy" which would entail levying a fee on freeway travel during rush hour, the amount to range between 5 and 10 cents per mile depending on the level of congestion.
Compared with similar surveys, this one was unusual in offering respondents explicit information about options that might be implemented and projected benefits (for example, travel time might be reduced by "x" minutes per day) as well as about the fate of the fee revenues collected. Respondents also received estimates of what their annual fee obligations would be under the hypothetical base plan. These estimates were calculated by asking respondents to supply information about their commuter travel times and other driving behavior. To personalize the fees in this way required the use of a Computer-Assisted Telephone Interview protocol so that interviewers could enter data directly into computers and then calculate the fees.
Reactions to the Base Fee
Of the motorists surveyed, 38 percent reported that they would support the base plan, with 56 percent opposed and 6 percent undecided. Thus the results suggest that nearly two out of five commuting motorists in Southern California will support congestion fees on the region's freeways even without being told with any specificity how the revenues are to be used. If we consider the intensity of preferences, however, we see that a much higher fraction of the opposition was "definite," suggesting that a congestion fee presented without rebate or other inducement will enjoy soft support and face hard opposition.
The most common reason given for opposing the base plan was that it amounted to a new tax. Fully a quarter of all respondents gave this as their reason for opposition. Another 10 percent felt the time saved was not worth the estimated cost, and 8 percent were skeptical about the plan's ability to reduce congestion.
To isolate the influences on support, we built a statistical model to indicate the probability that an individual would favor the fee given his or her personal characteristics and socioeconomic situation, as well as commuting habits. It seemed reasonable to expect, for example, that respondents with higher incomes would support congestion fees because they would tend to value the time savings more. Likewise, we expected that because educated respondents would better understand the arguments in favor of user fees, they would support the policy more than those less well-educated.
We were in for some surprises. We found no correlation at all between support for the fee and income, while education was negatively associated with support. Perhaps educated respondents were more skeptical of an untested economic theory or had more doubts about the competence of governments to implement such a plan. The strongest demographic indicator of respondent support for the base fee policy was having an Hispanic heritage. At this point at least, we have no explanation as to why.
Among the commuting behavior variables, only use of carpools and mass transit tended to translate into support, and only weakly. This weakness may be a consequence of the low frequency of transit and carpool use in the sample—on average only 0.12 and 0.56 days per week, respectively. Surveying a sample containing more carpoolers might yield a different outcome.
The variables of greatest interest to us were those that corresponded to the individual costs and benefits of congestion pricing policy: minutes saved and estimated cost per hour saved. We calculated the latter by taking each respondent's estimated congestion fee payments per week and dividing by the estimated time savings attributable to the fee policy both quantities being determined by the computer-assisted telephone interview program. We found that an increase in price caused support for the policy to decline.
We also found that support declined as the "minutes saved" variable increased, a result that at first glance appeared counterintuitive. After all, saving time on the road would seem to be a good thing and hence something respondents would vote for. But given the survey's construct, respondents were presented with a fixed quantity of time, at a fixed price, and asked if they wanted to buy. Thus it was not surprising to see the level of support drop as quantity increased. An individual may be willing to buy 5 pounds of potatoes for a dollar, or even ten pounds for two dollars, but not at all eager to buy 500 pounds for $100. Respondents favored a fee that would save them up to about an hour per week. Beyond that point, support dropped rapidly.
We also asked respondents whether they thought congestion pricing would cause the flow of traffic to move faster on roads and highways subject to a congestion fee. Positive opinions in this case were by far the most potent variable increasing support for the base fee policy. Thus an effective campaign to educate the public on the benefits of congestion pricing might garner more support for such a fee plan.
Reactions to Added Incentives
After getting their reactions to the base fee, survey participants were split randomly into three groups and asked to indicate their support for one of the following enhanced plans:
Congestion fees with fee/tax reductions. A percentage of the fee revenues (25, 50, or 82 percent of a respondent's fee payment) would be used to reduce other taxes, such as those on sales or gasoline, or used to reduce vehicle registration and license fees.
Congestion fees with coupons. A percentage of the fee revenues (25, 50, or 82 percent of a respondent's fee payment) would be returned as coupons that could be used to pay for transportation-related expenses such as public and private transit.
Fast lanes. Only the leftmost lane would be subject to fees on freeways; either an existing lane would be so designated or a new lane constructed. No rebates were associated with this option.
Reaction to the base fee plan was by far the best predictor of how survey participants would respond to any of the three plans enhanced with monetary and other benefits intended to attract support.
Congestion fees with fee/tax reduction. Among opponents of the base plan, 74 percent likewise opposed the plan when it was combined with a reduction in motor vehicle fee or sales tax. However, 20 percent of the base plan opponents were willing to support a congestion fee if combined with the tax/fee reduction. Among supporters of the base plan, only 10 percent rejected the enhanced plan. Thus this policy received more support than the base policy; it appeared to increase support for congestion fees by about 7 percentage points.
Interestingly, support for congestion fees that were combined with a fee or tax reduction varied significantly from county to county. The plan was extremely popular in San Bernardino County—by better than a two to one margin—but not at all popular in Riverside, with the other three counties somewhere in between. We have no hypothesis to explain these regional differences.
Not surprisingly, the level of support for the plan was higher among respondents who would receive a rebate worth more than the fee paid. Thus support for congestion fees might increase by designing plans so that most people will not pay in more than they get out.
Congestion fees with coupons. Unlike the tax/fee reduction policy, returning a portion of the revenues to the public in the form of coupons to cover transportation-related expenses did not in the aggregate improve support for congestion fees. Support for coupons was 36 percent—lower than support for the base policy across the entire survey sample. However, support for the coupons increased substantially as the aggregate dollar value of the coupons increased from 25 to 82 percent of the fees.
Congestion fees on fast lanes. More than 45 percent of the respondents said they would support congestion fees if an existing lane was designated as a fee lane (with 48 percent opposed). But support jumped to 54 percent when the fee would apply to a newly constructed lane, leaving all current lanes untouched. This latter congestion fee policy was the only one examined that won the support of a majority of respondents.
Off in the Future
The results of our survey suggest that congestion tolls and vehicle emissions fees can attract majority support from the public in Southern California, at least in a hypothetical referendum and under some circumstances. The survey also revealed that providing a rebate of some portion of the fees to individuals can increase support, although in our case not by enough to win a majority. Moreover, support for all congestion fee plans was weak in this sense: many more supporters said that they would "probably" rather than "definitely" support a fee plan, while opponents said the reverse was true. A referendum or similar proposal to adopt a congestion fee plan might be vulnerable to such qualitative distinctions—to the ferocity, say, of the opposition's sound bites during a related political campaign.
As far as we know, this survey is the first to elicit support for congestion fees by promising respondents a private benefit, in the form of cash, rather than a promise of public investment. With more experience, it is likely that policymakers will design more attractive rebate packages than are described here. In doing so, they may benefit from some of the information gained through this survey. Findings, such as where geographical and demographic support and opposition to these plans can be found, should help to target further investigation as well as campaigns to publicize such plans and inform the public about them.
Actual implementation of congestion pricing is probably still somewhat off in the future. Influential members of the California legislature, for instance, appear to continue to be implacable in their opposition. But meanwhile in the broader public forum, interest in and acceptance of such economic incentive policies appear to be growing. Time—and traffic—are on their side.
Winston Harrington is a senior fellow in the Quality of the Environment Division.
A version of this article appeared in print in the Fall 1997 issue of Resources magazine.
