I agree with the entirety of the court’s well-reasoned opinion except for the conclusion that the error in the environmental impact report (EIR) was not prejudicial. On this record, I cannot confidently infer that the EIR’s failure to measure impacts against a baseline of existing conditions did not deprive the public of relevant information about the project.
The court’s lucid analysis of the California Environmental Quality Act (CEQA; Pub. Resources Code, § 21000 et seq.) and applicable regulations firmly supports its holding that existing conditions comprise the normal baseline for measuring environmental impacts and that an agency may forgo analyzing impacts against a baseline of existing conditions only “if such an *479analysis would be uninformative or misleading to decision makers and the public.” (Lead opn., ante, at p. 453, fn. omitted.) Further, in light of Communities for a Better Environment v. South Coast Air Quality Management Dist. (2010) 48 Cal.4th 310, 328 [106 Cal.Rptr.3d 502, 226 P.3d 985], the court is correct that “an existing conditions analysis may take account of environmental conditions that will exist when the project begins operations; the agency is not strictly limited to those prevailing during the period of EIR preparation.” (Lead opn., ante, at p. 452; see id. at p. 453 [“[A] date-of-implementation baseline does not share the principal problem presented by a baseline of conditions expected to prevail in the more distant future following years of project operation—it does not omit impacts expected to occur during the project’s early period of operation.”].)
Here, the Exposition Metro Line Construction Authority (Expo Authority) used a baseline of existing conditions to measure most of the predicted effects of the light-rail project, but it used a baseline of conditions projected to exist in 2030 to measure the project’s expected impacts on traffic congestion and air quality. It is undisputed that the agency properly considered what the long-term impacts of the project would be in 2030. The issue is whether the agency properly considered those long-term impacts to the exclusion of any short-term impacts. In measuring traffic and air quality impacts solely against projected conditions in 2030, the EIR provided no analysis of such impacts against a baseline of existing conditions, including conditions in 2015 when the project is scheduled to begin operations.
As today’s opinion explains: “Even when a project is intended and expected to improve conditions in the long term—20 or 30 years after an EIR is prepared—decision makers and members of the public are entitled under CEQA to know the short- and medium-term environmental costs of achieving that desirable improvement. These costs include not only the impacts involved in constructing the project but also those the project will create during its initial years of operation. Though we might rationally choose to endure short- or medium-term hardship for a long-term, permanent benefit, deciding to make that tradeoff requires some knowledge about the severity and duration of the near-term hardship.” (Lead opn., ante, at p. 455.)
Here, there is “no substantial evidence supporting the Expo Authority’s decision to omit an analysis of the project’s traffic and air quality impacts on existing environmental conditions.” (Lead opn., ante, at p. 460.) “By focusing solely on the project’s operational impacts in the distant future, the EIR neglects to inform the public and decision makers explicitly of any operational impacts that could occur in the project’s first 15 years of operation.” (Lead opn., ante, at p. 461.) The fact “that project area population, traffic, and emissions of air pollutants are expected to continue increasing through and *480beyond 2030 does not justify the agency’s failure to analyze operational impacts under earlier conditions. The expectation of change may make it 'important for the agency to also examine impacts under future conditions . . . , but it does not constitute substantial evidence supporting a determination that an existing conditions analysis would be uninformative or misleading.” (Id. at p. 461.)
After reaching these conclusions, the court holds that the EIR’s failure to measure traffic and air quality impacts against existing conditions was harmless in this case. The court reasons that the EIR’s extensive analysis of traffic congestion against conditions projected to exist in 2030 “demonstrates the lack of grounds to suppose the same analysis performed against existing traffic conditions would have produced any substantially different information.” (Lead opn., ante, at p. 463.) But the fact that the project in 2030 is expected to have only a small effect on traffic congestion when compared to conditions in 2030 provides no reason to think that the project in 2015, at the start of operations, would have no greater impact when compared to conditions in 2015.
The EIR compared measures of congestion in 2030 if the project is built to measures of congestion in 2030 if the project is not built. But the measures of congestion in 2030 if the project is not built reflect significant predicted increases in congestion due to population growth. Thus it is not surprising that the project is expected to have little impact on congestion in 2030 when measured against the heightened congestion expected in 2030. But that finding sheds no light on the extent or magnitude of the project’s traffic impacts when it begins to operate in 2015, before the predicted increase in congestion due to population growth from 2015 to 2030. Without knowing how significant this transient impact on traffic congestion might be, how are the public and decision makers to decide whether the short-term pain is worth the long-term gain promised by the light-rail project?
It is not speculative to suggest that examining the project’s impact on traffic congestion in 2015 would yield different results. When the project begins to operate, ridership is expected to be at 77 percent of its eventual level. During that initial period, there may be an influx of cars to areas around the new transit stations, as people come to ride the train. While it is reasonable to assume that the worsening of congestion solely due to population growth is a more or less linear process, it is also reasonable to posit that the increase in congestion if the project is built would take the shape of a curve, with an initial steep increase due to an influx of cars and riders that later tapers off as the public adjusts to the new system. At the very least, it is not implausible to think that things may get worse before they get better. As Neighbors for Smart Rail contends, focusing solely on impacts in 2030 may mask earlier effects: intersections that are projected to worsen to critical levels of congestion if the project is not built may reach those levels sooner if *481the project is built. Or maybe not—but either way, CEQA does not permit the agency to simply leave the public guessing.
The EIR’s measure of air quality impacts suffers from the same problem. The EIR says the project, at full ridership, is expected to reduce vehicle miles traveled by 0.38 percent in 2030. The 0.38 percent figure reflects the differential between (a) vehicle miles driven in 2030 if the project is built and (b) vehicle miles driven in 2030 if the project is not built. From this, the court extrapolates that “the 77 percent initial ridership implies that initially the project will reduce vehicle miles traveled only by 0.29 percent.” (Lead opn., ante, at p. 464, fn. 11.) The court derives the 0.29 percent figure by comparing (a) vehicle miles driven in 2015 when the project begins operation with 77 percent ridership and (b) vehicle miles driven in 2030 if the project is not built. The proper comparison, however, is the differential between (a) vehicle miles driven in 2015 when the project begins operation with 77 percent ridership and (b) vehicle miles driven in 2015 if the project is not built. As with traffic congestion, there is reason to believe the project might actually increase vehicle miles driven in the short term, as new transit stations attract people from near and far to ride the light-rail. Further, without some analysis of the issue, we can only guess what portion of light-rail riders consists of people who would otherwise drive or ride cars to reach their destinations as opposed to new commuters who, but for the project, would not have traveled to their destinations at all, by car or otherwise.
For the reasons above, I respectfully disagree with the court’s conclusion that the EIR’s failure to measure traffic congestion and air quality impacts against a baseline of existing conditions “did not deprive agency decision makers or the public of substantial information relevant to approving the project.” (Lead opn., ante, at p. 465.) In all other respects, I join the court’s opinion.
Appellant’s petition for a rehearing was denied September 18, 2013.