Income-Based Equity Impacts of Congestion Pricing—A Primer

Review of the Literature

Issues

The “fairness” question may be viewed within the context of the overall highway financing system, in which, in the absence of congestion fees, the costs of providing peak-period highway service are borne by all highway users, not just by those who travel during congested periods or on congested routes. In this context, placing more of the burden of paying for peak-period highway service on those who make use of peak highway capacity is being increasingly viewed as an equity improvement.

A well-designed value-pricing plan can be less burdensome to low-income citizens than current systems that are based on regressive taxes, such as car-registration fees, sales taxes, and the gas tax. For example, low-income drivers usually drive older vehicles that are not as fuel-efficient as newer models. They therefore must purchase more fuel per mile driven and consequently pay higher fuel taxes for each mile driven than do those who own newer fuel-efficient models.

A report by the U.S. Congressional Budget Office (1990) found that the tax on motor fuels was regressive relative to annual income. In addition, Schweitzer and Taylor (2008) noted in one study that most forms of transportation finance—fuel taxes, sales taxes, and tolls—are regressive forms of taxation in that they burden the poor more than they do the rich. Schweitzer and Taylor (2008) stated that, “Using sales taxes to fund roadways creates substantial savings to drivers by shifting some of the costs of driving from drivers to consumers at large, and in the process disproportionally favors the more affluent at the expense of the impoverished.”

Another equity concern is that congestion pricing may make it too difficult or too expensive for low-skilled workers to get to their jobs. Entry-level and unskilled jobs are often not well served by public transit. Even if service routes exist for jobs of this type, the work hours for such jobs often require travel during off-peak service times, making public transit use less appealing as an option. Many low-skilled workers need to drive to retain their jobs; however, any congestion-pricing system can be sensitive to the issue of affordability, as discussed later in this primer.

When congestion pricing relies on an electronic cashless technology, households that do not have credit cards, bank accounts, or cannot afford large deposits may be unable to set up toll accounts, which may limit their use of these facilities. The Auto Express System in Puerto Rico mitigates many of these barriers by allowing users to purchase transponders and replenish their accounts by using cash at numerous retail and convenience stores without the need to provide a checking account or a credit card number. A light on the tag indicates when funds in the prepaid account are running low. Customers then have the option of replenishing their accounts at any number of locations, including gas stations. In Texas, TxTag accounts may be opened with cash. Those replenishing depleted accounts with cash must currently do so at a customer service center, but TxDOT is working with retailers to make TxTag services available at many retail outlets.

Another equity concern is that most tolled facilities that use electronic toll collection offer discounted tolls to those who use transponders rather than using video tolling or booth tolling. In situations in which the purchase of a transponder presents a significant economic barrier, low-income travelers who cannot afford a transponder will face a regressive toll schedule. It is estimated that between 10 and 20 percent of the population is unable to overcome these barriers to transponder ownership (Parkany, 2005).

Impacts on Low-Income Groups

Congestion-priced facilities currently in operation in the United States include tollways and tolled water crossings with variable tolls and priced lanes along major transportation corridors that experience high levels of congestion (U.S. DOT, 2008). Such congestion-pricing projects are operating in California, Minnesota, Washington, Colorado, Utah, Florida, Texas, New Jersey, and New York. The data from priced lanes have shown that a wide range of income groups use the lanes at different levels of frequency of use.

The use of congestion-priced lanes by both high- and low-income users appears to be selective. If use of priced facilities was solely dependent on income, then low-income travelers would never use such facilities. Studies have indicated that roughly half of the users of congestion-priced lanes do so once a week or less. Weinstein and Sciara (2004) suggested that the impacts of congestion pricing are not necessarily related to income and can also be based on flexibility of time and routes available to users.

Graph. A bar graph showing the equity implications of hypothetical Los Angeles 5-cent vehicle-miles-traveled (VMT) fee, 1991. The light green colored bars indicate the "Share of Financial Burden" and the darker green colored bars indicate the reduction in "Daily VMT." The graph shows that in Quintile 1, the "Share of Financial Burden" was approximately 6%, and the "Daily VMT" was approximately 6.5%. In Quintile 2, the "Share of Financial Burden" was approximately 14%, and the "Daily VMT" was approximately 8%. In Quintile 3, the "Share of Financial Burden" was approximately 18%, and the "Daily VMT" was approximately 7%. In Quintile 4, the "Share of Financial Burden" was approximately 26%, and the "Daily VMT" was approximately 5.5%. In Quintile 5, the "Share of Financial Burden" was 35%, and the "Daily VMT" was approximately 2.5%.

Equity implications of hypothetical Los Angeles 5-cent vehicle-miles-traveled (VMT) fee, 1991.

A paper by the Rand Corporation and Volpe National Transportation Systems Center (2007) indicated that household surveys suggest that rush-hour travelers who travel in the busier direction—and thus are more likely to pay congestion charges—are the most affluent group within the larger category of street and highway users.

Congestion pricing clearly will create economic hardship for some households. Svadlenak and Jones (1998) found that of adult residents in the Portland, OR, area who travel during peak hours in single-occupant vehicles, approximately 3 percent are low-income commuters. Of all Portland-area commuters, 38 percent travel during peak hours in single-occupant vehicles and have relatively high incomes. Svadlenak and Jones (1998) suggested that of this 38 percent, most can afford tolls and would welcome tolls if they resulted in a commensurate improvement in travel time.

Deakin and Harvey (1996) found that, if a 5-cent vehicle-miles-traveled fee were to be imposed in Los Angeles, CA, the lowest income quintile (i.e., 20 percent of users) would bear only 7 percent of the financial burden, whereas the highest income quintile would bear 35 percent of the financial burden.

Safirova et al. (2003) estimated the impacts of a high-occupancy toll (HOT) lane network in the Washington, DC, area. They found that the lowest income quartile would pay 5.2 percent of tolls, whereas the highest income quartile would pay 50.3 percent of tolls.

Transek (2006) found that, in the case of the Stockholm city center congestion-pricing scheme, affluent men in the inner city pay the most in congestion-pricing charges. Because high-income individuals use their cars more frequently, it was found that high-income households were more likely to incur the congestion charge compared with the average household. This analysis indicates that, if the revenues are used for public transportation, those who gain the most from the pricing scheme are young people, low-income individuals, single people, women, and residents of the inner suburbs. These groups pay relatively little in congestion charges on average and use public transportation more often than do other groups.

Public Opinion

Taniguchi (2008) provided results from a survey of public opinion on paying for transportation infrastructure with tolls versus taxes. The survey found that support for tolls was higher among low-income individuals (58 percent support for tolls) than among high-income individuals (42 percent support for tolls). Support for taxes was 32 percent for low-income individuals compared with 45 percent for high-income individuals.

Morallos (2006) found that, although limited, evidence from the successfully operating VPP projects clearly demonstrates that the most valued feature in tolling and pricing projects is that of providing people with a choice of whether to use priced lanes. Studies have shown that lower income individuals face the greatest financial harm when they are denied adequate travel choices. Lack of choice to pay a toll in exchange for reliable travel times can result in lost wages or late fees for daycare that could have been avoided.

Even when priced lanes are seen to be used more heavily by high-income users than by low-income users, a broad spectrum of income groups still express approval of the projects (as documented later in this primer) because they are given the choice of choosing the tolled route, an alternative free route, or a different transportation mode. Although high-income motorists do use the priced lanes more often, all income groups value the choice of a reliable trip travel time that is now available to them, serving their needs when they absolutely have to get to their destinations on time (e.g., getting to a daycare center before late fees kick in).

Graph. A bar graph showing the support for tolls versus taxes in King County, WA. The graph illustrates that low-income households prefer tolls over taxes. The light green bars represent the percentage of people who prefer tolls. The darker green bars represent the percentage of people who prefer taxes. The red bars represent the percentage that didn't answer. Among households with annual incomes less than $35K, 58% preferred tolls, 32% preferred taxes, and 9% didn't answer. Among households with annual incomes $35K-$55K, 51% preferred tolls, 36% preferred taxes, and 13% didn't answer. Among households with annual incomes $55K-$100K, 45% preferred tolls, 41% preferred taxes, and 14% didn't answer. Among households with annual incomes less than $100K, 42% preferred tolls, 45% preferred taxes, and 12% didn't answer.

Support for tolls versus taxes in King County, Washington. Low-income households prefer tolls over taxes.

Addressing Equity Concerns

Research has identified strategies for addressing equity concerns through redistribution of toll revenues. These include distributing rebates or credits, or revenue transfer to transit and carpooling services in the priced corridor. To ensure that at least some surplus toll revenue is used to improve transit, some areas have passed legislation to dedicate a portion of the surplus revenue to transit, whereas others have created special transit accounts.

A particularly important consideration in evaluating congestion-pricing options and their equity implications is the use of revenues generated by tolls. Toll revenues can be used to compensate those who might otherwise consider themselves “losers” as a result of congestion pricing. Compensation can come in a variety of forms. Toll revenues may be used to finance highway improvements (particularly in the corridor where the tolls are levied) or to pay for improvements in transit service. In cases in which effects on low-income drivers are felt to be particularly severe, toll exemptions or toll rebates may be offered to eligible drivers, or other forms of monetary compensation may be offered, such as tax rebates that provide reimbursment for tolls paid or income supplements.

Each of these approaches has been used or considered for use in congestion-pricing programs. For example, revenues from area pricing in Central London were used in part to improve bus service into the priced area, thereby enhancing transportation services to low-income groups and other users of those systems. The statutes in California mandate that 18 percent of toll revenues from the Bay Area Toll Authority be transferred into three accounts controlled by the Metropolitan Transportation Commission, a multimodal planning agency for the region. The Port Authority of New York and New Jersey likewise uses surplus toll revenue to subsidize transit services. When New York City considered a cordon-pricing scheme, it proposed a tax rebate for drivers who qualified for the federal-earned income tax credit. In the case of a proposed congestion-pricing scheme on the San Francisco Bay Bridge, tolls were to be raised from $1 to $3 per trip, but the proposal called for a reduced “lifeline” toll rate of $1 for low-income users.

Schweitzer and Taylor (2008) suggested that if policymakers are worried about low-income, peak-period commuters paying tolls, one way to address this would be to provide discounted “lifeline” pricing based on income levels, as is done by utility companies for qualifying customers. As an alternative, they could provide travel credits to low-income commuters.