1. The environmental review studied an overly narrow range of alternatives. Almost all alternatives considered in this document add a lane to the freeway, thereby increasing road capacity and likely future VMT associated with Yolo 80. Other realistic alternatives that could address the main rationale for the project (congestion) at far lower cost without widening the freeway and increasing capacity weren’t considered.

The DEIR studies 7 alternatives, with (a) and (b) options listed for most of these depending on whether median ramps and a flyover lane at the eastern end are included. Alternative 1 is the required No Build alternative. Alternatives 2-6 add a lane with various configurations of High Occupancy Vehicle (HOV), High Occupancy Toll (HOT), and transit use on the new lane. Alternative 7 takes an existing lane for HOV use (informed observers know that this is highly unlikely to be chosen since Caltrans has always constructed a new lane for HOVs in the past).

The most obvious alternative not included in the DEIR would be to price all lanes of the existing freeway. Pricing is widely acknowledged to be an effective means to discourage single-occupant vehicle travel (e.g. Small and Gomez-Ibanez, 2005; Clements, Kockelman, and Alexander, 2021). It can be easily implemented in California using FastTrak technology, with either a flat charge or variable congestion tolls, and is currently in use in northern California on the Bay Area bridges. Many other states nationally also use tolls to raise revenue or reduce congestion. Caltrans recently created an example of the sort of facility that could be built for Yolo 80 when it constructed a new automated toll facility in Martinez for the east-bound I-680 bridge. Equity impacts of pricing can be decreased by direct rebates to low-income vehicle owners and/or by using toll revenue for public transit or other services benefitting low-income communities.

Pricing is also the most cost-effective alternative. Instead of costing up to $465 million, pricing the existing I-80 Yolo causeway would generate in excess of $300 million annually (~150,000 vehicles/day x an $8 toll = $1.2M/day x 365 days/year =  $438M). In addition to equity rebates, this money could be used for transit and affordable housing near jobs in the corridor, further reducing traffic.

Pricing strategies are recommended by CalSTA’s CAPTI framework as mentioned above as well as by the Sacramento Area Council of Governments (SACOG)’s 2020 Metropolitan Transportation Plan/Sustainable Community Strategy. Policy 11 of the latter document calls for the region to “Initiate a leadership role in testing and piloting roadway pricing mechanisms, such as facility-based tolling and mileage-based fees.” That document also states (p. 73) “The roadway pricing mechanisms in the MTP/SCS are a critical component of the regional strategy to raise enough revenue to fund our transportation infrastructure, provide mobility benefits to residents, manage traffic, and help to achieve the region’s SB 375 greenhouse gas reduction target.”

Pricing alternatives would also have the benefit of improving transit performance. With tolls reducing overall traffic volume, buses would no longer be sitting in congestion. (A managed lane for both HOVs and transit, as proposed under several of the DEIR’s alternatives, could easily become congested as HOVs move into it from other lanes.) Pricing itself is not a panacea, and would likely result in a modest increase in VMT over the current situation since free-flowing lanes carry more vehicles than congested lanes. However, such an increase would be far less than that caused by widening the freeway.

A second main alternative would be constructing a dual express lane in each direction with single occupant vehicles tolled (one lane would be added, one existing lane converted). This alternative is used in SACOG’s travel demand model as part of its regional planning strategy, which Caltrans should have known about. It was also requested by the Yolo County Transportation District in its letter of May 4, 2022 to Caltrans, asking that “The Project Description…be written broadly enough to consider, and provide environmental clearance for, a multi-laned facility.” Tolling two lanes in this way would be more effective at meeting the project’s main goal of reducing congestion, and would have the additional advantage of generating additional revenue to mitigate VMT impacts.

A third potential alternative would be “using technology to optimize operations” as suggested by CAPTI. This approach is also a priority of SACOG’s Intelligent Transportation Systems (ITS) program, which seeks to use tools such as freeway ramp meters, dynamic message signs, closed-circuit cameras, and real-time information for the public to manage traffic. Congestion on I-80 could conceivably be kept manageable if all entrances featured ramp meters, with transit vehicles and HOVs allowed to bypass ramp queues. Incentives would then be strong for drivers to carpool or take transit, thus reducing VMT and congestion.

A fourth category of alternatives in the I-80 corridor would focus on dramatically better public transit. These alternatives might include better and more frequent bus service, better and more frequent rail service, and better feeder bus and van service in local communities connecting to trains and long-distance express buses.

A fifth category of alternatives would consist of regional Transportation Demand Management (TDM) programs. Air quality management districts in California have historically implemented these in order to address local air quality programs in locations such as the Bay Area and Los Angeles basin. Such initiatives typically include agreements with large employers to implement carpool and vanpool programs, to subsidize employee transit usage, and to charge employees for parking. They also include public education and informational strategies to decrease drive-alone commuting. To be most effective in the I-80 corridor, TDM programs should be combined with regional and local land use planning to create a better balance of jobs, housing, and services in local communities. Since 2008 California state policy pursuant to SB 375 has encouraged such planning, for example requiring regional planning agencies to prepare Sustainable Community Strategies.

Studying all of these alternatives is certainly not necessary for a robust DEIR, and the fourth and fifth would involve challenging inter-governmental coordination. But given the increasingly strong state policy framework against road capacity expansion and VMT increases, we request that the technically simple and cost-effective alternative of pricing all existing lanes be included in a recirculated environmental document.

2. The DEIR fails to adequately consider induced traffic. The Yolo 80 DEIR has a major flaw: its analysis fails to consider long-term increases in traffic volume stemming from widening the road, increasing capacity, and in turn influencing land use and behavior patterns. Again, the project would increase road capacity regardless of whether the new lane is an HOV lane, toll lane, or free-flow lane.

Road widening induces additional traffic in two main ways: 1) by changing short-term behavior, in particular as individuals see reduced congestion and choose to drive rather than using other alternatives such as carpooling, taking transit, or telecommuting; and 2) by changing long-term land-use patterns and behavior, in particular as individuals and businesses perceive that easy availability of commuting makes it possible to locate in certain places rather than others. Induced traffic is a well-established concept in the research literature dating back at least 50 years (e.g. Downs, 1962, Handy and Bournet, 2014; Hymel, 2019; Volker, Lee, and Handy, 2020). Many decades of experience in California also demonstrates the reality of this phenomenon, for example through the rapid growth of communities such as Vacaville, Dixon, and Fairfield which are almost entirely dependent on I-80 for long-distance motor vehicle travel.

Caltrans appears to have employed an identical future land use scenario for all DEIR alternatives. Agency materials state that “Land use inputs were not developed for each individual alternative. Instead, the SACOG 2020 MTP/SCS land use forecasts associated with specific model years 2016, 2027, and 2040 were used without modification….This approach limits the sensitivity of the traffic and revenue forecasts to any unique land use effects associated with each alternative.” (Caltrans, 2023b, 10).

Consequently Caltrans’ DEIR analysis appears to incorporate the first form of induced traffic but not the second. The DEIR shows a substantial increase in traffic at the 2029 opening, but only modest increases long-term. Table 2.2-9 (pp. 2-194-105) shows an immediate 9.2% increase in VMT for Alternative 2 compared with the No Build alternative, but a 2049 increase of only 4.2%. The corresponding figures for Alternative 3 are 9.2% and 4.3%. If induced traffic due to changing land use and lifestyle patterns were fully taken into account, these long-term figures would likely be significantly higher. Indeed, if NCST induced traffic figures of 495,400 VMT/day were simply added to the No Project alternative for these years, Alternatives 2 and 3 would have had increases of 11% in 2049. Actual increases would certainly be higher, since the DEIR shows the No Project alternative to be highly congested, reducing VMT, whereas alternatives would have greater throughput.

The DEIR flatly states (p. 3-38) that the project would have “less than significant impact” on population growth either directly or indirectly. This is particularly surprising given that the document previously includes the NCST data mentioned above showing more than 495,000 additional VMT/day for most alternatives. Here as in other locations the DEIR appears not to have incorporated the NCST data into analyses.

For this DEIR Caltrans District 3 relied on analysis procedures not compatible with more recent Caltrans Headquarters standards that require induced traffic be considered. Caltrans’ 2020 CEQA guidance states that

“[C]apacity-increasing projects generally need to be evaluated for their potential induced travel. The mechanisms by which induced travel occur include:

• Route changes (may increase or decrease overall VMT)

• Mode shift to automobile use (increases overall VMT)

• Longer trips (increases overall VMT)

• More trips (increases overall VMT)

• Location and land use changes (increases overall VMT)” (Caltrans, 2020a, 18)

Caltrans HQ adopted the NCST induced travel calculator in 2020 as an official tool, and the agency’s Transportation Analysis Framework document provides extensive guidance on how it is to be applied (Caltrans, 2020a). Table 2 on page 17 of this document specifically says that the NCST calculator is to be applied to Yolo County. But the DEIR fails to integrate NCST numbers for induced traffic into its analysis of long-term impacts of the project, thus understating VMT and GHG emissions while overstating congestion relief.

3. The DEIR fails to adequately take into account changing driver behavior. A second modelling inadequacy of the DEIR is that it uses a simplistic static trip assignment model rather than a more sophisticated dynamic traffic assignment model. The latter type of model takes into account a variety of feedback loops resulting from traffic congestion, including drivers changing the time of their trips so as to avoid congestion. Use of a static trip assignment model can have the effect of understating the increase in VMT. A related problem is that the model wrongly assumes that additional traffic would be routed off the congested highway up through Woodland, which is unlikely given the distance and likely congestion of that alternative route. Again, this points to a model with major inadequacies. Interviews conducted as part of Dr. Amy Lee’s dissertation The Policy and Politics of Highway Expansions show that Caltrans rejected better models (dynamic traffic assignment) because they would have shown increased VMT (Lee, 2023). The DEIR’s own technical appendices acknowledge limitations of the static assignment model:

“Another limitation of the SACSIM19 model is the use of static assignment rather than dynamic assignment of vehicle trips. With congested conditions, static assignment can result in volumes that exceed capacity for the analysis period. With dynamic assignment, trips are rerouted or shifted in time so that capacity is met.” (Transportation Analysis Report, I 80/US 50 Managed Lanes, November 2023, 78).

Dynamic travel assignment models are increasingly used elsewhere in the country for projects in urban areas with congested conditions. For example, Colorado’s Department of Transportation has concluded that

“DTA is useful when the analyst’s study area includes a congested transportation facility as well as its parallel facilities (or parallel capacity)….DTA’s assignment methods is more sophisticated than a traditional travel demand model as it accounts for bottlenecks. DTA also allows for temporal spreading (peak-hour spreading).” (Colorado Department of Transportation, 2023)

4. The DEIR relies on inadequate mitigations. The DEIR assumes that VMT/GHG increases can be mitigated if Caltrans funds projects in local cities. But its project list covers only 43% of its estimated VMT increase (which is likely low in that it appears not to include truck traffic), and it’s doubtful that such mitigations would be additional and verifiable. These mitigations would have to be implemented by other entities over which Caltrans has no control, and mitigation funding is questionable, with Caltrans committing to funding only 12.5% of the cost of trip reduction programs, for example. The mitigations therefore do not meet the “fully enforceable” standard required by CEQA.

At least one mitigation,“Build overcrossing at future Nishi Student Housing Development site” is a project that has long been planned by the developer of the Nishi student housing project and approved by the Davis City Council. This “mitigation” is not an additional GHG reduction and would simply give public funds to a private developer. Many other mitigations focus on increasing use of long-distance express bus services. Such services can have their own traffic-inducing impacts, for example if they encourage households to move to outlying suburban communities thinking that they can use long-distance express buses to commute to jobs in Sacramento. Such households then use their motor vehicles for many other types of trips, increasing VMT in the region. For such reasons Caltrans must identify fully enforceable and fully funded mitigations to offset the full amount of likely VMT and GHG increases in the DEIR.

5. The document falsely concludes no or less-than-significant impact for important topics. For NEPA and CEQA purposes the DEIR (Tables S-1 and S-2) wrongly states that the alternatives studied would have “no impact” on urban growth and population, air pollutants, and energy demand, and “less than significant” impacts on GHGs and state climate policy. These statements are not true and should be revised based on additional analysis and up-to-date modeling practices within a recirculated document.

1. Urban growth and population: The DEIR states (Summary-7) “Build Alternatives 2a and 2b would not remove an impediment to growth, provide an entirely new public facility, or provide new access to previously unserved areas. Build Alternatives 2a and 2b would not directly increase development of residential land uses, encourage growth outside of existing growth boundaries, or alter existing access to residential and employment areas; therefore, no adverse effect associated with population growth would be anticipated with implementation of Alternatives 2a and 2b.” (Emphasis added. Text for alternatives 3a and b, 4a and b, and 5a and b says “Same as effects described under Build Alternatives 2a and 2b.”)
   
   This is not true. Common sense tells us that since no alternative through-route exists congestion on I-80 is a prime determiner of urban growth in the corridor. If the project reduced congestion upon opening, access to residential and employment areas along the corridor would be greatly altered. It could take, for example, 25 minutes rather than 1 hour to get from Vacaville to Sacramento in the morning rush hour. Such a time difference would have a large impact on people’s travel and residency decisions, fueling urban growth and population increase in some locations rather than others.
   
   These “no impact” conclusions, like many other DEIR conclusions discussed below, arise in part because Caltrans did not consider induced traffic within its model. However, this is a well-established concept in the research literature, much of which is summarized by Volker, Lee, and Handy, 2020. The DEIR itself acknowledges the induced traffic concept by including data from the National Center for Sustainable Transportation (NCST) showing an additional 495,000 vehicle miles travelled (VMT) for most alternatives (Table 2-1-26). If induced traffic had been considered, the DEIR’s conclusions about no effect on urban growth and population would not have been supported.
   
   
2. Air pollutants: Sacramento and Yolo counties are state non-attainment areas for 8-hour ozone and PM 2.5 air pollutants. In terms of pollutant increase, the DEIR (Summary-16) finds “no impact” for the No Build alternative. This is unlikely to be true and shows deficiencies in the underlying analysis. Indeed, Table 2-2-34 shows a 19.9% increase in congestion for the No Project alternative in 2049 (idling cars usually increase pollution), and Table 5 In Appendix J shows increases of 3.5% and 22.2% respectively for PM10 in the years 2029 and 2049 versus the baseline year of 2019, and an increase of 4.5% for PM2.5 in 2049. (That appendix does not show results for other pollutants such as carbon monoxide and ozone. The increases for particulate matter are dismissed with the following statement: “the difference between Build and No Build would be not significant in terms of PM10 and PM2.5 in regards to the increase of total AADT [Average Annual Daily Traffic] between Build and No Build with a HOV-HOV connector.”)
   

   The DEIR also finds “less than significant” air quality impacts for all other project alternatives. This is also unlikely to be true. Table 5 in Appendix J shows increases of between 2.9% and 13.0% for PM10 in the opening year of 2029 for alternatives 2-7b compared with the 2019 baseline. For 2049 it shows PM10 increases of 0.5% to 9.5% for alternatives 2-7a and 6.1% to 26.9% for alternatives 207b.  For PM2.5, the increases range from -1.9% to 6.7% for the “b” alternatives. It is hard to see how these increases can be called “less than significant.” Such figures also do not take into account induced traffic and likely increases in congestion resulting from that well-documented phenomenon, which would likely increase air pollution further.

3. Energy demand: The DEIR states (Summary-12) “When balancing energy used during construction and operation against energy saved by relieving congestion and other transportation efficiencies, the project would not have substantial energy effects. Therefore, no adverse permanent effects are anticipated.”

This statement is likely false, since it assumes congestion would actually be relieved in the long-term which is inconsistent with the literature on induced traffic (much of which is summarized in Volker, Lee, and Handy, 2020), and the historical experience of freeway-building in California.

Even without considering induced traffic stemming from secondary land use change and population/employment location decisions, the DEIR’s modeling shows sizeable traffic increases from roadway dynamics alone (less congested lanes attracting more drivers). These VMT increases are in turn likely to require additional energy, making a finding of “no adverse impacts” inappropriate. (The form of energy will of course shift as vehicles electrify, though electric vehicles will still have impacts.)

The exact traffic increases from existing conditions under DEIR modeling are difficult to calculate since the DEIR fails to provide existing traffic volume (measured as Average Annual Daily Traffic) anywhere. SACOG gives this figure as 136,700 vehicles in 2017 at the Yolo/Solano county border (https://vitalsigns.mtc.ca.gov/indicators/traffic-volumes-at-gateways). Taking this as the baseline and using Table 4 in Appendix J figures for future AADT, 2029 traffic increases from roadway dynamics compared to the recent baseline would range from 15.0% for the No Project alternative to 14.5%-27.1% for the various alternatives. 2049 traffic increases would range from 31.9% for the No Project alternative to 27.3% to 39.6% for the various alternatives. Again, it is hard to see how such large increases in traffic would produce “no adverse impacts” in terms of energy.

D. GHGs: The DEIR itself shows that CO2 emissions would increase by between 2.2% and 10.9% for the various project alternatives in the 2029 opening year (Table 7, Appendix J). That is a substantial amount; it is hard to see how this can be called “less than significant.” In the year 2049 the document shows all alternatives reducing GHG emissions versus existing and No Build conditions, but in many cases these reductions are small. For the “b” alternatives, for example, they range between -1.4% and -4.9%. Most importantly, these modelled GHG emissions reductions do not take into account induced traffic.

If induced traffic is taken into account, the GHG increase would be large. The DEIR itself includes data from NCST modeling showing likely increases of 495,300 vehicle miles traveled daily from most project alternatives (Table 2-1-26). Based on these NCST figures and average GHG emissions figures for passenger motor vehicles, widening I-80 in Yolo County would add at least 79,545 tons of CO2e emissions annually (218 tons/day)—equivalent to increasing Davis citywide emissions by at least 14%, or unincorporated Yolo County emissions by at least 7.3%. (This figure would be higher still if the higher-emitting truck percentage of up to 29% of vehicles were included.) Building on Table 7, the increase in GHG emissions produced by the I-80 project in 2049 would then range between 15.4% and 29.5% for alternatives 2-5 (those that add a lane) compared with the No Project alternative. This is a very large amount for a future date when the state plans to be carbon-neutral, and hardly “less than significant.”

The DEIR analysis of greenhouse gas emissions is thus highly inadequate. Caltrans should bring its model up-to-date by including induced travel from freeway projects, and should revise and recirculate this document based on a more accurate model taking this phenomenon into account.

E. State climate policy: Under its 2022 Scoping Plan, California has a state climate planning goal of achieving carbon neutrality by 2045, which means cutting GHG emissions 85% compared with 1990 as well as sequestering large amounts of emissions in natural landscapes and below-ground. In contrast, the I-80 project is likely to produce sizeable increases in GHGs both in the short-term (2029) and the long-term (2049), as shown above. Thus this project, like other freeway widening projects that Caltrans continues to pursue, cannot be said to have “less than significant” impacts on state climate policy.

For the above reasons, the Yolo 80 DEIR should be revised and recirculated.

Thank you very much for your attention.

Sincerely,

Stephen M. Wheeler, Ph.D.

On behalf of the Yolano Group of the Motherlode Chapter of the Sierra Club

REFERENCES

Caltrans. 2020a. Transportation Analysis Under CEQA: Evaluating Transportation Impacts of State Highway System Projects. First Edition. Sacramento.

Caltrans. 2020b. Transportation Analysis Framework. First Edition. Sacramento.

Caltrans. 2023a. Interstate 80/US Highway 50 Managed Lanes Transportation Analysis Report. Sacramento.

Caltrans. 2023b. Interstate 80/U.S. Highway 50 Managed Lanes Traffic and Revenue Report. Sacramento.

Clements, Lewis M., Kara M. Kockelman, and William Alexander. 2021. Technologies for congestion pricing. Research in Transportation Economics. 90. 100863. https://doi.org/10.1016/j.retrec.2020.100863 .

Colorado Department of Transportation. 2023. Traffic Analysis and Forecasting Guidelines. Available at https://www.codot.gov/safety/traffic-safety/assets/traffic_analysis_forecasting_guidelines/traffic_analysis_forecasting_guidelines.

Downs, Anthony. 1962. The law of peak-hour expressway congestion. Traffic Quarterly. 16 (3) 393-409.

Handy, Susan and Marlon G. Bournet. 2014. Impact of Highway Capacity and Induced Travel on Passenger Vehicle Use and Greenhouse Gas Emissions: Technical Background Document. Sacramento: California Environmental Protection Agency and California Air Resources Board.

Hymel, K. M. 2019. If you build it, they will drive: Measuring induced demand for vehicle travel in urban areas. Transport Policy, 76 57-66.

Lee, Amy Elisabeth. 2023. The Policy and Politics of Highway Expansions. University of California, Davis. Doctoral Dissertation.

Small, Kenneth A. and Jose A. Gomez-Ibanez. 2005. Road pricing for congestion management: The transition from theory to policy. In Oum, Tae Hoon, 2005. Transport Economics. New York: Taylor & Francis.

Volker, Jamey M. B.; Lee, Amy E.; and Handy, Susan. 2020. Induced vehicle travel in the environmental review process. Journal of the Transportation Research Board. 267 (7) 468-479. DOI: 10.1177/0361198120923365.