An Assessment of the Expected Impacts of City-Level Parking Cash-Out and Commuter Benefits Ordinances

Chapter 4. Results

For all cities, the project team estimated the VMT reductions as a percent of total citywide commute VMT. The results are in Table 3 and figure 2 for each city and each scenario. The overall level of VMT reduction is dependent on several factors, including, most notably, the existing mode shares (e.g., how many employees are currently driving alone) and how many employees currently receive free or subsidized parking and are subject to the ordinance.

Table 4 and figure 3 show the expected raw VMT reductions for each city and scenario. Table 4 shows the raw VMT reduction expected under a scenario where telework rates are 2x pre-pandemic rates, while figure 3 reflects these values as a range (1.4–3.2x) based on estimates presented by Mokhtarian et al. (2022) on post-pandemic telework expectations. Due to the sketch-level approach of this analysis, all estimated figures (VMT, congestion, emissions, and crash reductions) are subject to uncertainty.¹⁵

Additional results related to the extensions examining VMT reductions in affected commuters only for Scenarios 1 and 2, exempting small employers (i.e., for Scenarios 1A and 3A, which are otherwise identical to Scenarios 1 and 3 except that the extensions exempt employers with fewer than 20 employees), and investigating impacts of free versus partially subsidized parking impacts are presented in Appendix D. Additional Results. Detailed core results for each city are in Appendix E. Results By City.

Table 3. Percent reductions in daily citywide commute VMT by scenario and city.

City	S1: Monthly Cash-Out	S2: Monthly Commuter Benefit	S3: Monthly Cash-Out + Pre-Tax Transit Benefit	S4: Daily Cash-Out + Pre-Tax Transit Benefit	S5: Eliminate Parking Subsidies + $5 Non-Single Occupancy Vehicle (SOV) Subsidy
Boston/Cambridge, MA	10%	1%	10%	18%	29%
Chicago, IL	11%	7%	13%	18%	36%
Houston, TX	3%	2%	3%	7%	17%
Indianapolis, IN	5%	2%	5%	15%	24%
Los Angeles, CA	9%	5%	9%	17%	27%
New York, NY	3%	1%	11%	12%	36%
Philadelphia, PA	13%	9%	14%	21%	34%
San Diego, CA	6%	3%	6%	15%	25%
Washington, DC	4%	2%	6%	11%	24%

Table 4. Estimated raw reductions in daily citywide commute VMT by scenario and city (assuming “most likely” 2x telework scenario and approximated to the nearest one-hundred thousand)

City	S1: Monthly Cash-Out	S2: Monthly Commuter Benefit	S3: Monthly Cash-Out + Pre-Tax Transit Benefit	S4: Daily Cash-Out + Pre-Tax Transit Benefit	S5: Eliminate Parking Subsidies + $5 Non- Single Occupancy Vehicle (SOV) Subsidy
Boston/Cambridge, MA	1.0M	0.1M	1.0M	1.8M	2.8M
Chicago, IL	2.0M	1.3M	2.4M	3.3M	6.6M
Houston, TX	1.7M	1.2M	1.8M	4.1M	10.4M
Indianapolis, IN	0.6M	0.3M	0.6M	1.6M	2.6M
Los Angeles, CA	3.5M	2.2M	3.5M	6.8M	10.8M
New York, NY	1.2M	0.4M	3.7M	4.3M	12.6M
Philadelphia, PA	1.6M	1.2M	1.8M	2.7M	4.3M
San Diego, CA	1.1M	0.6M	1.1M	3.0M	4.8M
Washington, DC	0.5M	0.2M	0.6M	1.2M	2.6M

*M = million

Discussion

Across all of the scenarios, it is clear that local parking cash-out related ordinances could have a significant impact on reducing vehicle travel (along with subsequent congestion, emissions, and crashes) associated with commuting by employees. The impacts vary by scenario and by city based on a range of factors, including parking prices, transit fares, and the share of employees currently receiving free or subsidized parking. This section first discusses results framed around reductions in VMT. Then, results related to equity, congestion, emissions, and crashes are presented.

The two monthly cash-out scenarios—Scenario 1 (monthly parking cash-out) and Scenario 3 (a requirement that employers that don’t offer free parking offer a pre-tax transit benefit in addition to monthly parking cash-out for those offering free parking)—show significant potential for reducing daily VMT. Scenario 3 builds on Scenario 1 and applies an additional requirement for employers that do not currently offer free parking to offer pre-tax transit benefits. While the effects of offering pre-tax benefits, where employees set aside their own money for transit, is likely to be small compared to an employer-paid benefit, this policy is assumed to apply to a large population of employees in many cities who do not currently receive free parking. For instance, in San Diego (and most other cities analyzed), the pre-tax benefit requirement on top of the parking cash-out ordinance yields a relatively small incremental effect, since in San Diego it is estimated that 88 percent of employees work at sites with fully subsidized parking and would receive the cash-out offer instead of the pre-tax benefit. On the other hand, in New York City, where only 4 percent of employees work at sites with fully subsidized parking, the pre-tax transit benefit requirement applies to a much larger share of employees and yields a larger incremental benefit. Pre-tax commuter benefits may also exhibit greater reduction potential when paired with campaigns to increase commuter awareness.

Scenario 4 (a requirement that employers that do not offer free parking offer a pre-tax transit benefit in addition to daily parking cash-out for those offering free parking) shows greater reduction potential than Scenario 3 because it assumes that if employees are offered a daily cash-out, which is more flexible than a monthly option, more employees will take the offer and reduce their driving. The Scenario 4 analysis assumes an additional 16 percent shift away from driving alone with daily cash-out compared to monthly cash-out in Scenario 3. This does not translate directly to a linear 16 percent overall VMT reduction between the scenarios. For the population under each scenario eligible for cash-out, while there is a 16 percent decrease in drive-alone VMT between the two scenarios, some of this VMT is re-distributed to carpool and vanpool (among other modes), which still count toward total VMT (although less than driving alone, given higher vehicle occupancies). This slightly dilutes the VMT savings (i.e., to between a 10–15 percent decrease). From there, the overall citywide impact varies based on the prevalence of free-parking offerings in each city. That is, the difference between the results in Scenarios 3 and 4 is less pronounced in cities where a lower proportion of employees are offered the cash-out opportunity versus the pre-tax transit benefit (e.g., New York City).

Scenario 2 (the option of an employer-paid monthly transit/vanpool pass in lieu of free parking) shows more modest reductions than the monthly cash-out scenarios. Instead of offering cash, employers in this scenario are required to offer a tax-exempt transit or vanpool benefit in lieu of parking. Even in cities where the assumed average transit fare is high, driving reductions were smaller than in Scenario 1; fewer employees are likely to take a transit-only benefit as compared to a cash offer (with a tax-free transit option) and even most commuters accepting tax-free transit would, with Scenario 1, also be provided an additional taxable cash-out payment due to the market value of parking exceeding transit commute costs. Scenario 2 also applies to a slightly smaller baseline population than Scenario 1—beyond employees who receive free or subsidized parking (excluding those who already receive parking cash-out), it also does not apply to these commuters who already receive transit benefits.

Under Scenario 5 (a requirement that all employers eliminate subsidized parking and provide a universal $5 daily non-SOV commute benefit), employees who drive to work alone would suddenly face a new cost to their driving (parking). This scenario offers the greatest reduction potential in all cities, likely because it incentivizes non-SOV modes not considered in other scenarios (e.g., carpool, walking, biking) in addition to non-SOV modes already considered by other scenarios (transit, vanpool) for a greater number of employees, and it would yield the additional shift away from driving-alone also realized in Scenario 4 due to the provision of a daily benefit.

Under each scenario, the largest cities by number of commuters understandably demonstrate the greatest raw VMT reduction potential (e.g., Houston, Los Angeles, New York City). Cities that start with a high citywide drive-alone share show the smallest reductions relative to citywide commute VMT, especially when using the TRIMMS analysis methodology. If comparing two cities with the same drive-alone mode share, cities where a greater proportion of employees are offered free or subsidized parking (and thus would be eligible for cash-out) are expected to exhibit greater relative reductions under cash-out policies. For example, Houston and Indianapolis currently have drive-alone shares of more than 80 percent. When comparing those two cities, however, Scenarios 1, 3, and 4 have considerably larger impacts in Indianapolis than in Houston. This is because a much larger proportion of employees are offered subsidized parking in Indianapolis. Thus, offers directed toward current drivers receiving subsidized parking have more impact in Indianapolis than in Houston.

Examining the results across scenarios, Houston and Chicago reveal the impact of starting mode share and travel costs. These cities have relatively similar numbers of employee populations. Additionally, approximately 40 percent of employees in each city receive free parking at work, while around 10 percent already receive transit benefits. The estimated drive-alone mode share for employees receiving fully subsidized parking in Houston was 83 percent versus 53 percent for Chicago. While daily transit costs are estimated around six dollars in each city, the daily cost of parking in Chicago is approximately double that of Houston. Despite a number of similarities, the relative VMT reductions across the scenarios in Chicago are more than double those in Houston.

Even though New York City has an employee population more than double the size of the city with the next largest employment population (over four million compared to Los Angeles’ two million), relative VMT reduction estimates are relatively small compared to some other cities (i.e., those which also have high parking costs and low drive-alone mode shares, like Chicago) in scenarios applied only to employees receiving fully subsidized parking (which is only around 4 percent in New York City). This is because these scenarios (Scenarios 1 and 2) are applied to a relatively small population, and consequently, the impact is small. By contrast, Scenarios 3, 4, and 5 apply to all employees, boosting the relative VMT reductions comparable to other cities with similar characteristics (high parking costs and low drive-alone mode shares).

Overall, travel impacts vary widely among scenarios and cities. Responses to the different scenarios generally depend on the attractiveness of alternatives to driving and parking and which segment of the employee population is targeted. Among the cities, trip and VMT reductions depend on a variety of parameters, including the size of the affected population, baseline mode shares of the employees, average trip distances, and existing parking and transit costs (see these core attributes in Appendix C. Data and Analysis Methodology, Table 6). Regardless of the exact strategy or city, the projected VMT reductions, along with reductions in driving related externalities, through any of these policy mechanisms are significant.

The presented analysis is based on existing data along with several assumptions and estimations, introduced in Chapter 3. City And Scenario Selection and outlined in more detail in Appendix C. Data and Analysis Methodology. Chapter 2. Literature Review highlights research by Shoup (1997), which provides results to compare to those modeled from Scenario 1. As previously presented, Shoup (1997) found parking cash-out programs were associated with a 12 percent reduction in VMT in a comprehensive analysis of eight parking programs in Southern California (mainly in and around Los Angeles), and subsequent analysis by researchers in Minneapolis and Seattle found similar results. While the results presented here reflect VMT reduction estimates citywide, additional analyses presented in Appendix D. Additional Results show Scenario 1 results for affected commuters only. The estimated reduction in commute VMT in Los Angeles for affected commuters under Scenario 1, the most similar policy to that studied by Shoup, is 11 percent. Across all the cities studied, the average percent reduction in commute VMT for affected commuters under Scenario 1 is 14 percent. Both figures are relatively close to Shoup’s (1997) earlier VMT reduction estimate.

Beyond Scenario 1, the other modeled scenarios lack implementation-based data to which a comparison to modeled results could be drawn. The results of Scenario 5 are striking, in particular, given estimates of VMT reduction between 17 percent and 36 percent; across the analyzed cities, Scenario 5 results reflect an average four-fold increase in estimated VMT reduction compared to Scenario 1, citywide. The results are plausible, though, considering the unique feature of Scenario 5 compared to the other modeled scenarios—the elimination of subsidized parking entirely—plus the addition of a new, daily benefit for not driving alone to work.

An INRIX (2017) study found that one-third of the total costs of vehicle ownership could be attributed to parking. When parking is subsidized, these costs are “hidden” to drivers. Subsequently, when those costs are revealed, such as through the elimination of parking subsidies, drivers may be likely to shift behavior given the cost increase relative to the overall total cost of driving. Additionally, elimination of subsidized parking entirely may make non-driving incentives more attractive. In a study of commuter benefits in the Washington, D.C., region, Hamre and Buehler (2014) found very significant impacts of parking and other subsidies, and their elimination, on commute mode choice. While benefits incentivizing public and active transportation commuting (e.g., transit benefits, showers/lockers, bike parking) were related to a decreased likelihood of driving, the provision of these benefits alongside free workplace parking reduced their effectiveness. For example, Hamre and Buehler’s (2014) found that when both free parking and transit benefits were offered together, the estimated drive alone mode share was quite high at approximately 83 percent.

Taken together, the high proportion of driving costs attributable to parking and the increased expected effectiveness of non-SOV incentives in the absence of free parking lend support to the estimated impact of Scenario 5 compared to the other modeled scenarios.

Each scenario has additional implications for equity, congestion, emissions, and safety, as discussed in the following subsections. Additionally, recall that Chapter 1. Background covered some financial impacts and benefits related to parking cash-out and commuter benefits policies. For interested readers, Appendix F. Non-Employee Financial Impacts: Employer Costs and Government Tax Revenues elaborates on some of these impacts, which may be useful for those considering implementation of such programs, or ordinances requiring such programs.

Equity Considerations

The parking cash-out and related commuter benefits policies examined in this analysis have various implications for equity. If free parking is traditionally only offered to specific subsets of commuters (e.g., commuters working in certain industries, at specific income levels, etc.), it would disproportionately benefit certain groups of commuters over others. Policies that include cash-out alone would only be offered to those commuters already receiving free parking at work. Even if parking benefits are not offered equitably, however, cash-out is equity-enhancing as it adds two groups to those receiving a benefit: 1) employees who were offered parking but could not take advantage of it due to not owning a car that is available for their commuting (either due to owning no vehicles or sharing a vehicle with other household members who may need it); and 2) employees living in locations where driving to work is not the most convenient alternative. Policies that include commuter benefits for other modes tend to be offered more broadly than parking (employers controlling a limited parking supply might pick and choose to whom they offer it) and thus are further enhancing of equity.

In this analysis, each scenario would impact the following baseline populations:

• Scenario 1: Only commuters who receive free workplace parking
• Scenario 2: Only commuters who receive free workplace parking
• Scenario 3: All commuters, with different benefits provided to those offered free workplace parking than to others
• Scenario 4: All commuters, with different benefits provided to those offered free workplace parking than to others
• Scenario 5: All commuters, with equal benefits offerings across non-SOV modes

In Scenarios 1 and 2, the population being impacted are only those employees with free workplace parking. Under both scenarios, those within the eligible baseline population can benefit if they can mode-shift from driving to work. Commuters expected to receive the greatest benefits under Scenario 2 are those who already commute via transit or vanpool, or who are located with access to quality transit for commuting purposes. Although the benefit would technically be offered to other employees in the baseline population, it may not be realized if commuters do not have access to transit or are unable to form a vanpool. In contrast, the full baseline population impacted could benefit from parking cash-out in Scenario 1, provided commuters who currently drive are able to shift their commute mode.

In Scenarios 3 and 4, all commuters would be eligible to receive some type of commute benefits. Those without free workplace parking would be eligible for pre-tax transit¹⁶ benefits, while those commuters who receive free workplace parking would be eligible for parking cash-out. Again, for commuters who would only be eligible for transit benefits and are unable (or unwilling) to commute via transit or vanpool, they may not see the benefit realized. However, these two scenarios may be more equitable than Scenarios 1 and 2 if a broader demographic of employees would become eligible for benefits.

In Scenario 5, all non-SOV commuters are eligible for equal benefits, with the same benefit offered to any commuter who travels via a non-SOV mode whether or not free/subsidized workplace parking was in place. This scenario is primarily different from the other core scenarios in that it eliminates parking subsidies entirely. While the commute benefit for Scenarios 3 and 4 for employees who did not receive fully subsidized parking might only be realized for those employees who are reasonably able to commute via transit or vanpool, the commute benefits offered under Scenario 5 could be realized by any employees not driving to work, regardless of if they received workplace parking subsidies or not. That is, although Scenarios 3, 4, and 5 would all technically offer benefits to all employees, the proportion of employees who are actually able to utilize the benefit offered is expected to be greatest under Scenario 5.

As such, in addition to maximizing VMT reduction potential, Scenario 5 is thought to maximize the benefits offerings across the analyzed scenarios, and reflects the policy expected to distribute benefits most equitably. For example, Census data shows that, in general, the lowest-income households exhibit lower rates of vehicle ownership and higher rates of walking or biking commuting compared to higher income households (see Appendix G. Additional Equity Discussion for more information). If these commuters cannot switch from walking or biking to another mode and do not receive workplace parking subsidies (and so are ineligible for cash-out under other scenarios), they would realize the greatest benefit out of Scenario 5 (e.g., while they may be offered transit or vanpool benefits under Scenarios 3 or 4, if they cannot switch modes, then they cannot actually use these benefits).

Scenario 5 also advances transportation equity by eliminating the false free cost of parking; free parking acts to subsidize automobile use, and can subsequently increase negative externalities related to congestion, pollution, emissions, and safety. Scenario 5 has the greatest VMT reduction potential, which will act to mitigate these externalities for the greatest number of people. Although this analysis is limited in that it is not designed to map VMT reductions spatially across the cities, reductions in emissions and pollution specifically should have citywide benefits; future work could examine how reductions in congestion and safety improvements could be mapped spatially and overlap with neighborhood demographics for a more robust look into equity impacts. Appendix G. Additional Equity Discussion explores considerations across the analyzed scenarios in more detail.

Congestion Reduction

The estimated reductions in daily congestion (i.e., percent change in peak period time delays experienced) for each scenario are shown in figure 4 (assuming the “most likely” 2x pre-pandemic telework rate scenario). The changes in vehicle hours of delay across scenarios generally follow the VMT patterns exhibited by the scenarios. That is, Scenario 2 exhibits the lowest impacts on delay reduction, while Scenario 5 has the greatest impact.¹⁷

While the reduction in vehicle travel should translate directly to reduced vehicle congestion, the level of delay reduction is uncertain and influenced by many factors, notably including the current level of congestion along individual corridors within each city. Here, delay reduction potential is more limited than VMT reduction potential. This is, in part, a function of the method applied (which relies on an elasticity showing a less than 1 percent reduction in delay for each 1 percent reduction in VMT). Additionally, recall that changes in delay are relative to all peak-time VMT (of which only 54.9 percent of peak-period VMT is commute related), while VMT reductions presented in figure 2 are presented relative to commute travel only.

This relationship of delay relative to VMT being less than one is not an overarching rule-of-thumb. In contrast, some research suggests that a small reduction in vehicle travel can have a disproportionately large impact on delay reduction. For instance, a study conducted for FHWA estimated that “in general a 10 percent to 14 percent decrease in traffic on congested freeways will reduce delay by approximately 75 percent to 80 percent” based on data from the Washington, D.C., region (The Louis Berger Group 2008).

Despite the more conservative percent changes in congestion compared to VMT in this analysis, however, the impacts are not insignificant. Considering the results from figure 4, the research team scaled the time reduced for Scenario 1 and Scenario 5 to an annual measure (assuming 19 working days each month) for each city. Then, using the 2019 value of time ($/hour) from the Texas A&M Transportation Institute’s (TTI’s) Urban Mobility Report,¹⁸ dollars saved from delay reductions (rounded to the nearest million) for these two scenarios are presented in Table 5.

Table 5. Estimated annual dollars saved for all commuters due to delay reductions.

City	Scenario 1	Scenario 5
Boston/Cambridge, MA	$10M	$32M
Chicago, IL	$25M	$85M
Houston, TX	$20M	$128M
Indianapolis, IN	$3M	$14M
Los Angeles, CA	$37M	$121M
New York, NY	$10M	$114M
Philadelphia, PA	$17M	$47M
San Diego, CA	$8M	$34M
Washington, DC	$6M	$35M

Emissions Reduction

Figure 5, figure 6, and figure 7 display estimated annual emissions reductions for each city and scenario (assuming the “most likely” 2x pre-pandemic telework rate scenario). The reduction potential trends within each city generally mirror those of VMT reductions, while the magnitude of reduction is greater in cities with greater numbers of commuters in general. These reductions are not insignificant. To put these estimates in perspective, 500,000 metric tons of carbon dioxide equivalents (CO₂e) (close to the average reduction across cities for Scenario 5) is equivalent to the energy use of more than 60,000 homes each year, the electricity use of almost 100,000 homes each year, and more than a million barrels of oil consumed.¹⁹ This is the amount of carbon sequestered by roughly 8 million tree seedlings growing over 10 years, or almost 600,000 acres of U.S. forests in one year.

Safety Impacts

Based on discussions with the peer review group consulted for this analysis, the research team decided it was appropriate to assume crash reductions trended linearly with VMT. That is, citywide crash reductions would mirror the relative VMT reductions presented in figure 2. TRIMMS v4.0 provides crash rates (crashes per million VMT) for each city. Based on those crash rates, expected annual reductions in combined fatal and incapacitating injury crashes (assuming 19 working days each month and the “most likely” 2x pre-pandemic telework rate scenario) are displayed in figure 8. While these estimates may seem relatively small in magnitude, any reductions in fatalities or incapacitating injuries on roadways improve safety; this is reflected in Vision Zero—a “strategy to eliminate all traffic fatalities and severe injuries”—initiatives adopted across over 45 communities across the U.S. (Vision Zero Network 2022).²⁰ The large impact of seemingly small crash reduction estimates is further demonstrated when considering the value of a statistical life (VSL), or “the additional cost that individuals would be willing to bear for improvements in safety (that is, reductions in risks) that, in the aggregate, reduce the expected number of fatalities by one” (USDOT, 2021, p.1). USDOT’s latest 2021 figure for VSL is $11.8 million (USDOT, 2022b). USDOT cost-benefit analysis guidance (USDOT, 2022a) asserts the fraction of VSL applicable toward incapacitating injuries as approximately 0.048 VSL or $564,365. These VSL estimates are applied here toward the crash reductions, assuming linear scaling based on the most recent USDOT VSL guidance (USDOT, 2021), for each scenario and plotted in figure 9.