Contemporary Approaches to Parking Pricing: A Primer
2.0 Pricing Overview
FHWA and local governments are looking at leveraging market forces by pricing transportation resources to reduce congestion. Pricing, if properly instituted, accomplishes three important objectives:
Pricing parking can be a powerful tool—especially when used in conjunction with other travel demand management strategies—to influence travelers' decisions about whether to drive alone, carpool, use transit, or use non-motorized travel modes. Reductions in drive-alone travel can subsequently reduce emissions and congestion and improve access and revenue generation.
This section of the primer discusses the two basic approaches to parking pricing: 1) free and fixed-rate pricing and 2) performance-based pricing. Within performance-based pricing there are two primary strategies: variable prices and escalating prices. These approaches can be used by cities to better manage parking supplies while simultaneously improving the travel experience of those who continue to choose driving. Depending on how parking revenues are invested, a parking strategy can more broadly improve access to an area where the desire to drive and park currently exceeds road capacity and/or parking supply.1
Cities own a tremendous amount of real estate that comprises the public right-of-way (ROW). While the value of the ROW as an asset is implicit in permit fees for uses ranging from block parties and construction to non-automobile storage, peculiarly, most cities allow residents and visitors to store their automobiles rent free on much of the ROW. In some instances, typically in business districts, municipalities will charge nominal parking meter fees. Because cars are parked about 96 percent of the time and because estimates of the number of parking spaces per automobile range from three to five, the 194 million registered vehicles in the United States take up between 5,200 and 8,700 square miles of parking space. The land devoted to parking in the United States could fill an area between the size of Connecticut and New Jersey—a valuable asset that is underutilized.
As noted in the introduction, the first parking meters were installed in Oklahoma City in 1935. Studies pre-dating the installation of those meters showed that vehicles parked on commercial streets belonged, by and large, to local merchants and their employees. Customers, who had begun to own automobiles at increasing rates, were left to circle around hoping that a parking space would become available. These drivers contributed to the incipient but fast growing downtown congestion problems. Civic leaders recognized that by renting the curb, rather than giving it away, they could shift the dynamic. Meters were first installed on only one side of each street. In the morning, as workers and merchants arrived, the free spaces quickly filled. By 10:00 a.m., as shoppers came downtown, metered spaces made up the majority of available parking. As customers completed their business and departed, the metered spaces were used by later arriving customers, who also paid for use of the parking spaces. By one account, merchants on the free side began to clamor for meters on their side as well (Popular Mechanics, 1935). The pricing strategy was very effective; instead of having all the spaces taken up by all-day parkers, ample turnover allowed many users to access the area. In short order, other cities adopted this approach to rationing the curb. All-day parkers resorted to finding spaces in slightly more remote areas, or left their vehicles at home, allowing the high-demand spaces to turn over repeatedly throughout the day. The outcome was good for business and good for street performance.
With inflation, however, the price of the meters effectively became lower and lower, and the rationing effect was eroded. Losing sight of the initial reason for meter installation, city leaders became dependent on parking revenues but lacked the political will to price the curb effectively to continue the initial success. Instead, downtown merchants, fearing competition from the burgeoning suburbs, fought for cheap or free parking.
Boston is an instructive case. In January 2011 Boston increased meter rates for the first time in 25 years, after "mulling it over for 10 years" (Andersen 2011). By the time of the rate increase, the pre-existing $1.00 an hour meter rate, set uniformly across the city, had lost value with inflation. The rate was effectively half what it had been when it was set in the mid-1980s. To restore the meter rate to what it had been the city would have had to double the rate to $2.00. Rather than use a performance-based pricing strategy, as explained below, the city simply increased the rate to $1.25 and justified the increase as a way to raise revenue for the city's general fund. Lacking a travel management rationale, the rate hike was seen by parkers as a tax on drivers.
On-street parking and most municipally or publicly owned off-street parking, particularly at transit stations, has traditionally been free or set at fixed prices that vary little by location or time of day. In those cases where prices had at some time been established according to supply and demand, the failure of pricing to keep pace with inflation (and demand) has left the municipalities and agencies in charge of parking pricing without a sound justification for taking action. As in the Boston example, fixed-price parking, across time and geography, without respect to demand or inflation, is not very different from free parking in terms of congestion mitigation and access. Fixed-price parking has a benefit over free parking in that it does signal a fee for use of the space rather than simply an entitlement to the ROW, but it falls far short of its potential as an effective demand management tool.
In the 1960s and 1970s, cities concerned about competition from suburban merchants focused on trying to offer free and abundant parking instead of focusing on parking access. Boulder, Colorado, however, was an exception. Boulder's city leaders and merchants believed the essential ingredient to success was available parking for those who drove to the main business district, but, at the same time, they saw the value of rationalizing access, realizing that in order to offer abundant free parking, as suburban developers did, they would have to redevelop in a suburban style. By allowing the large-scale development of parking facilities, the city would essentially erode land values and become "suburban" itself. Instead, the city established the first parking benefit district, charging for parking, coordinating on-street and off-street fees, and using the revenue to enhance other transportation modes.
Rather than maintaining fixed pricing, increasingly cities are taking a holistic approach, as Boulder did. The current approach considers access broadly, taking into account all travel modes, and uses modern parking-management strategies to define and meet demand. These cities set parking rates to achieve specific occupancy goals or other objectives. Depending on the goals and local conditions, parking rates can vary by location, time of day, and presence of a special event. Policies and subsequent pricing are data-driven and designed to balance demand throughout neighborhoods and central business districts. Because they are data-driven, these policies allow city managers to adjust prices quickly based on economic and land-use changes.
Fixed-rate pricing has been the standard parking-pricing option for cities since the parking meter was introduced. While it may not adequately price parking, it does have advantages. The pricing scheme can be implemented with mechanical meters, requires no additional special equipment, and does not require the collection of data regarding parking utilization and availability. Unfortunately, fixed-rate pricing fails to manage parking supply.
Pricing parking based on performance goals for the street or transportation system, often called performance-based pricing, allows cities to better manage the parking supply. Parking experts generally agree that 10 to 20 percent (one or two spaces) of on-street parking per block should be vacant most of the time as a way to reduce or eliminate cruising for parking (BPR, 1956; Levy, et al. 2012). Higher vacancy rates may be a sign that pricing is too high.
While a vacancy rate of 10 to 20 percent might be the most common performance goal used by cities, other goals can be considered as well. Pricing can be set to drive turnover, maximize value extraction, and transition travelers away from private automobiles to more sustainable travel modes. In any case, the performance standard is met through various pricing schemes, including rates that escalate the longer a person is parked, prices that vary by location, prices that vary by time of day, or a combination of these options.
Implementing a performance-based pricing program begins with understanding the local parking context and establishing a balance between parking supply, both on street and off, and demand. Accurate and up-to-date supply and demand data are helpful to determine appropriate parking rates, but the rates can be set empirically as well. San Francisco and Seattle are good examples of cities that are empirically setting rates to reduce cruising. San Francisco has taken a complex approach with the aid of "smart" meters that can accommodate multiple forms of payment, charge variable parking rates, and record data regarding usage and duration of use; parking sensors; and a very advanced data collection system, whereas Seattle is experimenting with a low-technology approach that focuses on manual measurements of on-street parking conditions. Both cities seek to set rates that assure an appropriate level of available space (see the Seattle case study in section 7 and the San Francisco callout at the end of this section).
Parking rates should be allowed to vary across a variety of dimensions. One dimension should be geographical, as some areas of a city will have greater parking demand than others. Rates should also vary by time of day, which is already a common practice as meter rates are typically in effect only during daytime hours and overnight parking is free. A few cities, New York City and San Francisco being notable examples, have implemented differential parking rates that vary by time of day based on changes in parking demand. New York City implemented variable parking rates in two pilot neighborhoods. In one neighborhood the peak rate is charged between 12:00 p.m. and 7:00 p.m. and in the other neighborhood the peak rate is charged between 6:00 p.m. and 10:00 p.m. As in most meter applications, overnight parking is still free, leaving three distinct price regimes throughout the day. Rates should also vary across days of the week, as some areas will have higher demand on weekdays than weekends and vice versa. They should also vary across time more generally: as inflation erodes prices and as areas gain or decline in popularity, meter rates should fluctuate to reflect these realities.
A somewhat controversial approach is to vary prices in real-time, which the District of Columbia is proposing to pilot for some on-street commercial vehicle parking. This approach is analogous to a travel lane that is priced to ensure a particular travel time. As parking utilization on a given block increases, the price escalates from a base price. The practice is more controversial with respect to parking as there is a value-driven belief among most city leaders that people should have a reasonable a priori expectation of prices. Also, it may be counterproductive to keep the price low for people who arrived during a period of high availability. That outcome would encourage people to arrive early and stay for longer periods.
Often used in off-street parking, escalating rates increase the longer a vehicle is parked at a location. The rate structure is designed to discourage long-term parking, thereby increasing parking turnover and availability. Differentiated rates are common practice at airports. Airport operators typically divide parking into short-term and long-term lots. The spaces nearest the terminal are well suited to people who are dropping off or picking up passengers and will only use the space for a short time. People who will be parking overnight or for multiple days are often accommodated farther away so the airport operator has adequate parking supply for those who need the more convenient spaces. The way that airports enforce the distinction is by setting different prices. Frequently they will set an escalating price in the short-term lot to discourage long stays. Atlanta's Hartsfield-Jackson airport charges short-term parkers $2.00 per hour for the first 2 hours and then $3.00 per hour for the next 4 hours. At Chicago's O'Hare Airport they charge $2.00 for the first hour, $3.00 for the second hour, nothing for the third hour, and then a steep rise to $5.00 for the fourth hour and $19.00 for the fifth hour. An escalation like this encourages people to park only for short periods to accomplish a task. In the case of an airport, the task is to pick up passengers.
Cities may also use this model where they wish to encourage additional parking turnover. For example, they may wish to use this approach in commercial areas that have many deliveries. If deliveries can be accomplished in 1 or 2 hours, having a third and fourth hour charge that is very high will discourage all-day parkers, allowing an adequate turnover rate so that deliveries can be accommodated. In New York City, certain spaces throughout the city are designated for commercial vehicles. The rate for these spaces is $4.00 for the first hour, $5.00 for the second hour, and $6.00 for the third hour.
Cities that have adopted explicit performance goals usually seek to achieve a certain level of parking turnover or a certain level of parking availability. At the heart of each goal is the objective that people wishing to park should be able to do so with minimum search costs. As a practical matter turnover may be hard to measure (especially if space sensors have not been deployed), which means it is difficult for a city to know if it has met a turnover-related performance standard; however, availability may be simpler to measure (e.g., by occasionally conducting manual counts and supplementing such counts with meter-payment data). Turnover is also harder to enforce. Many cities adopted time limits on metered spaces to meet their turnover goals. Anecdotally it is unclear that citizens understand that a time limited meter is to be vacated at the end of the time limit. Many people think they need only return to their car and "feed the meter" in order to be in compliance with the regulation. The evidence shows that meter time limits are frequently violated (Weinberger et al., 2010). Time limit enforcement used to rely on agents placing chalk marks on the tires of parked cars. Newer approaches use license plate recognition technologies but still require an enforcement agent to make frequent passes along the streets.
Technological advances have made obtaining data, setting prices, and adjusting prices much easier. New smart meter systems can accommodate multiple forms of payment, charge variable parking rates, and record data regarding usage and duration of use. These meters can be supplemented with parking sensors and license plate reader technology, both of which have been used by cities to determine occupancy with varying degrees of success.
Instituting different pricing strategies does not require advanced technology but it is made much easier and defensible with emerging meter and data collection technology. These technological advances are the subject of the next section.
1In many instances of apparent parking under-supply it is infeasible and/or unsound to add parking capacity. This occurs where the street system is also congested and where adding parking would require reducing active uses of the land. [Return to note 1.]
United States Department of Transportation - Federal Highway Administration