Here is Part 3 or this four-part series:
The HAA prevents cities from denying or reducing the density of housing projects, but it doesn’t exempt projects from environmental review under CEQA. CEQA spells out time limits for the completion of environmental reviews, but as yesterday’s post explained, those limits have proven illusory in court. So if a city wants to deny a project that the HAA protects, what’s to keep the city from laundering the denial, as it were, through CEQA? Can the city keep asking the developer for additional environmental studies until, after squandering years and fortunes, the developer cries uncle and walks away?
That’s the million-dollar question raised by our running example, the San Francisco Board of Supervisors’ recent 8:3 vote sustaining a local gadfly’s appeal of the 469 Stevenson St. project. Rather than deny the project outright or reduce its density (likely HAA violations), the Board reversed the planning commission’s certification of the project’s Environmental Impact Report and directed the clerk to prepare findings that the EIR was inadequate.
Yet in view of what the Supervisors said at the hearing and afterwards, it’s pretty clear that the Board’s real objective was not to air out and mitigate specific environmental impacts but to nix the project. Most of the Supervisors who voted “No” argued that the project was not affordable enough and would cause gentrification – which is not an environmental impact and which is exceedingly unlikely to be caused by the project in any event. Supervisor Mandelman told a reporter that he’d “feel very good about this vote” if the site “become[s] a 100% affordable project,” but that if “15 years from now it’s still a parking lot, then I will not feel good.” That’s an explanation for a vote to deny, not a vote for further environmental study. Supervisor Melgar said the problem was that the developer hadn’t “negotiated a deal” with TODCO, a politically powerful nonprofit. That of course has no bearing on the adequacy of the EIR.
The supervisors who voted “No” also knotted themselves up with self-contradictory objections. For example, Ronen and Mandelman stressed that the developer didn’t have financing and that the project probably wasn’t economically viable (the implication being: “don’t blame us for blocking housing”), yet they also demanded that the developer reserve more units for low-income households – which would make the project even more difficult to finance.
The supervisor who came closest to voicing an environmental objection was Supervisor Ronen, who expressed concern that the project’s foundation might be inadequate. She pointed to another downtown project, the Millennium Tower, that had required an expensive retrofit, and she argued that the EIR for Stevenson St. should have fleshed out the seismic issues in detail. (The Initial Study treated these issues as “insignificant” because they’re addressed by the building code and an engineering peer-review required of all large buildings. Accordingly, the EIR did not further address them.) However, no one put any evidence in the record suggesting that a code-compliant, peer-reviewed project on the site would be an earthquake hazard to people or buildings nearby. Nor, as best we can tell, had Ronen or any other supervisor objected to previous EIRs that treated seismic impacts as adequately addressed through the building code and engineering peer review. In any case, contrary to Ronen’s claims to the press, the impact of an earthquake on the proposed building is not an “environmental impact” under CEQA.
All of this suggests that that the seismic safety issue – the only plausibly legitimate justification for the Board’s decision to reverse the CEQA certification – was pretextual. It was a fig leaf to cover up what the Board intended but was not allowed by law to do: to disapprove the project because it’s too big or not affordable enough.
Capitalizing on Administrative Law’s Achilles Heel
The strategy of laundering project denials through CEQA is nothing if not clever, for it takes advantage of two soft spots in administrative law: agency delay and agency bad faith.
Delay
The Board of Supervisors’ vote to reverse certification of the Stevenson St. EIR was tantamount to saying, “We haven’t made up our mind about this project, and we need more information before we can make up our mind.” When agencies say they need more time to gather information and make up their mind, courts normally let them have it. If an antsy plaintiff sues, the court will say that the suit is premature because there’s not yet a “final” agency decision, or because the plaintiff hasn’t “exhausted her administrative remedies,” or because the case isn’t yet “ripe.” After all, it would be a waste of judicial resources and a big practical problem for governance if anyone waiting in line for an agency decision could ask a judge to let him jump the queue.
The legal doctrines that prevent plaintiffs from attacking agency delay have exceptions, but the exceptions are very narrow. For example, California courts excuse plaintiffs from exhaustion when further agency proceedings would be “futile” – but only if the plaintiff can “positively state” what the agency has decided (thus rendering further proceedings pointless). The courts have also waived exhaustion when the agency has no legal authority to conduct the proceeding at issue and when pursuit of further proceedings would result in irreparable harm. None of these exceptions fits the Stevenson St. scenario. The Board of Supervisors has carefully avoided “positively stating” its decision; there’s no question that the Board is authorized by law to be the city’s ultimate decider about the CEQA review; and the irreparable harm exception is applied “only in the clearest of cases.”
It’s also true that if the Legislature prescribes clear-cut timelines for an agency decision, a plaintiff can, in theory, use “traditional mandamus” to get a court order requiring the agency to act. But as we illustrated in yesterday’s discussion of Schellinger and the CEQA timelines, these cases make courts uncomfortable. At most, a court will order the agency to make a decision, as opposed to telling the agency what to decide. And if there’s an available equitable doctrine like laches that would let the agency off the hook, the courts will gladly invoke it.
Bad Faith
The other formidable barrier to a judicial fix for CEQA-laundered project denials is the principle that courts should review agency decisions solely on the basis of the reasons stated by the agency at the time of the decision, rather than probing to figure out the agency’s real reason and setting the decision aside if the real reason was not authorized by law.
To the extent that the Board’s decision to require further CEQA study of the 469 Stevenson project is reviewable at all, a court would normally uphold the decision so long as the “findings” prepared by the clerk include some legitimate reason for additional CEQA study. The stated rationale must also draw some support from the record of materials before the Board, but the evidentiary demand is lax. If a reasonable person could agree with the Board’s decision in light of the evidence in the record, courts generally will accept it.
In federal administrative law, there is a narrow exception to these general precepts. Upon a “strong showing of bad faith,” a court may peer behind the agency’s rationale and the record of contemporaneous materials the agency assembled to justify it. If the court concludes from this investigation that the agency’s stated reasons were pretextual, the court may set aside the agency’s decision – even if the stated reasons, if real, would have sufficed to justify it. This obscure doctrine enjoyed a moment of renaissance when Chief Justice Roberts invoked it to invalidate the Trump Administration’s addition of a citizenship question to the U.S. Census. But even as the Chief Justice insisted that courts “are ‘not required to exhibit a naivete from which ordinary citizens are free,’” he was at pains to limit the bad-faith exception.
It is for very good reasons that the bad-faith exception is narrow. Much like aggressive judicial review of agency delay, courtroom trials focused on the “real reasons” for agency action would gum up the work of government. Discovery requests and depositions would divert public officials from their primary charge. Courts would struggle to disentangle the mix of political and policy-minded considerations that shape agency decisionmaking – especially when the leaders of the agency in question (a city council) are elected officials who inevitably pay attention to politics even when acting in a quasi-judicial capacity (hearing a CEQA appeal).
Finally, it’s black-letter law that when an agency messes up, the judicial remedy is to vacate the agency’s decision and remand for a do-over. Even in the Census case, the Court did not strike the citizenship question from the Census: it just told the Commerce Department to try again. But what does this achieve if the agency is in bad faith? A court order telling San Francisco’s Board of Supervisors to rehear the 469 Stevenson St. CEQA appeal would be an invitation to re-launder the denial, minus the revealing tweets.
But the HAA’s a Game Changer, Right?
The foregoing ought to douse any hope one might have about using general legal principles to curtail CEQA-laundered project denials. But when the project getting laundered is a housing project, a court must consider the Housing Accountability Act as well. And the HAA gives the general principles of administrative law a real shakeup, reworking some and tossing others in the garbage:
The HAA expressly authorizes judicial inquiry into bad faith. “Bad faith” as defined by the Act includes “an action that is frivolous or otherwise entirely without merit.” This means that a court can find bad faith without subpoenas, depositions, or other searching inquiry into the mental processes of city council members. If the denial of a project was objectively frivolous, that’s enough.
In cases where a court finds bad faith, the HAA supplants the traditional do-over remedy. It authorizes courts to order the project approved--and to retain jurisdiction to ensure that this order is carried out.
The HAA provides at least a partial remedy for delay, by defining “[d]isapprove the housing development project” to include “[f]ail[ing] to comply with the time periods [for project review] specified in [the Permit Streamlining Act].”
The HAA eliminates judicial deference to local governments on all questions about whether a housing development project complies with applicable standards.
The HAA’s stance is one of extreme distrust toward local governments. In 1982, the Legislature stated that “the excessive cost of the state’s housing supply is partially caused by activities and policies of many local governments that limit the approval of housing.” But as the Legislature noted in 2017, when it strengthened various provisions, “[t]he Legislature’s intent in enacting this section in 1982 . . . has not been fulfilled.” Hence the new policy going forward: “that [the HAA] be interpreted and implemented in a manner to afford the fullest possible weight to the interest of, and the approval and provision of, housing.”
But there’s a catch. While the HAA provides a powerful remedy for a bad-faith project denials, its only explicit remedy for delay is tied to the Permit Streamlining Act. Yet as noted in our previous post, the PSA clock doesn’t start to run until CEQA review has been completed, and another provision of the HAA states that the statute shall not “be construed to relieve [a city] from making ... findings required [by CEQA] or otherwise complying with [CEQA].”
How can a court make sense of these conflicting directives? In the rest of this post, we sketch three possible solutions.
Solution #1: Bad-Faith Delay Through CEQA Reversal as HAA “Disapproval”
A court following the Legislature’s command to “interpret[] and implement[]” the HAA “to afford the fullest possible weight to the interest of ... housing” could hold that a city’s delaying of a project in bad faith amounts to “disapproval” within the meaning of the HAA, at least if the delay occurs through a negative vote on a formal approval that a developer needs to reach the finish line.
The HAA’s definition of “disapproval” is broad. It includes “any instance in which a local agency . . . votes on a proposed housing development project application and the application is disapproved, including any required land use approvals or entitlements necessary for the issuance of a building permit.” The certification of an EIR or other CEQA clearance is one of many “approval[s]” or “entitlement[s]” which a developer must obtain before eventually landing a building permit. And it is an approval that a city council reversing a CEQA clearance “votes” to deny.
The HAA’s remedial provisions imply that the statute may be violated other than by final denial of an application for a project entitlement or building permit. A court that finds a violation “shall issue an order ... compelling compliance with this section within 60 days, including, but not limited to, an order that the local agency take action on the housing development project.” The “but not limited to” proviso suggests that a city may violate the HAA by taking unlawful action (or inaction) on ancillary matters necessary for the project to go forward, and it instructs courts to use their powers flexibly to remedy whatever violations a court finds.
On the other hand, the fact that the HAA doesn’t expressly list “legally inadequate CEQA analysis” as a permissible ground for disapproval of a housing development project suggests that the Legislature may not have thought that a city council’s reversal of a CEQA certification would qualify as a housing-project disapproval. But the HAA in its current incarnation is meant to be a super-statute, ”interpreted and implemented in a manner to afford the fullest possible weight to the interest of, and the approval and provision of, housing.” This interpretive instruction, together with the parallel legislative finding that local governments have for too long managed to evade the Legislature’s intent to ”meaningfully and effectively curb[] [their] capability ... to deny, reduce the density for, or render infeasible housing development projects,” suggests that the Legislature wants courts to read the statute flexibly as may be necessary to countermand evasive local tactics the Legislature did not anticipate.
A line-drawing problem remains: it can’t be true that every city council vote sustaining a CEQA appeal is a “disapproval” within the meaning of the HAA. Some appeals are meritorious. In other cases, a city council may reasonably believe that an appeal has merit, even if some judges would disagree. At what point does a city council’s reversal of a legally sufficient CEQA clearance become an HAA "disapproval’? The HAA’s remedial provisions point toward an answer: when the CEQA reversal is in bad faith. Like the party to a contract who commits anticipatory breach, the city that denies a CEQA clearance in bad faith signals that it has no intention of performing its legal obligation under the HAA.
If a court reads “disapproval” to include bad-faith denial of a CEQA clearance, and finds that San Francisco’s Board of Supervisors pretextually reversed the EIR certification for 469 Stevenson St., the court could order the project approved, because the HAA supplants the conventional do-over remedy in cases where a city has denied a project in bad faith.
One might object that this gloss on HAA “disapproval” would “relieve[]” the city of compliance with CEQA. Not so. San Francisco’s planning department prepared a full EIR for 469 Stevenson St., which the planning commission certified as complete. So long as the court concludes that the EIR was in fact legally sufficient, an order directing the city to approve the project would do no violence to the HAA’s CEQA-preservation clause. The court could also allow the Board of Supervisors a brief window of time to decide whether to impose any additional mitigation requirements on the project, in light of the findings of the EIR. This would honor CEQA’s policy that elected officials bear final responsibility for deciding what to do about identified environmental impacts.
Another counterargument is that the Board in voting to reverse the EIR certification didn’t actually determine whether the project could go forward or what its density would be. It just said it wanted more information. This argument would be a strong counter under general administrative law principles. But in taking a practical, real-world approach to “disapproval,” the HAA undercuts it. For example, delay beyond the time limits of the Permit Streamlining Act is explicitly an HAA disapproval, even though such delay doesn’t entail any concrete act or statement of reasons by the city. A formal vote reversing a CEQA clearance is much closer to the conventional paradigm of a discrete, reviewable agency action.
In its first letter to San Francisco after starting to investigate the 469 Stevenson St. debacle, the Department of Housing and Community Development signaled support for reading “HAA disapproval” to include pretextual CEQA-clearance reversals. If an agency that the Legislature has authorized to enforce the HAA concludes that bad-faith denials of CEQA clearances are "disapprovals within the meaning of the HAA, a court need not go out of its way to conclude the same.
Solution #2: Enforcing CEQA Timelines in Light of the HAA
Without reaching the question of whether bad-faith denial of a CEQA clearance is “disapproval” within the meaning of the HAA, a court could hold that the Legislature’s refashioning of the HAA as a super-statute warrants revisiting – and limiting or rejecting – the Court of Appeal’s decision in Schellinger Brothers v. City of Sebastopol. Burying Schellinger is necessary to give practical effect to the HAA’s incorporation of the Permit Streamlining Act’s timelines into the definition of disapproval.
As we explained in yesterday’s post, Schellinger held that judges may not order a city to certify an environmental impact report (as opposed to ordering the city to make up its mind about whether to certify it). The court also said that the project applicant had, by cooperating with the city and making revisions well past CEQA’s deadline, forfeited its right to enforce the deadline.
The most basic problem with Schellinger is that it makes a hash of the statute’s definition of “disapproval.” As noted, the HAA defines disapproval to include noncompliance with the PSA deadlines, but the PSA clock only starts to run after CEQA review is done. So if there’s no practical way of forcing cities to comply with CEQA’s deadlines, then the delay-oriented piece of the HAA’s definition of disapproval is a dead letter. That doesn’t befit any statute, let alone one which the Legislature has declared to be super.
As for Schellinger’s “equitable” holding (that the developer who cooperates past a deadline forfeits her right to enforce it), equitable doctrines are not supposed to be used in ways that “nullify an important policy adopted for the benefit of the public.” Whatever might have been said about the HAA when Schellinger was decided in 2009, there is no gainsaying that, today, the Act’s policy of expeditious permitting is “important” and inures to the “benefit of the public.”
CEQA allows one year for completion of an EIR. A recent study of housing project entitlements in twenty California cities found that the median project in San Francisco took 27 months to entitle; only 5% were entitled in under a year. 469 Stevenson St. is more of the same. The final EIR for the project wasn’t certified by the planning commission until nearly three years after the developer’s submission of the project application. And then came the appeal to the Board of Supervisors, resulting in further delay.
Bearing these facts in mind, and reading CEQA in light of the newly “super” policy of the HAA, a court might reasonably hold (1) that the CEQA deadlines are enforceable regardless of whether the developer has cooperated with the city past the deadline (contra Schellinger), and (2) that if the CEQA deadline has passed and a legally sufficient environmental review document has been prepared, the city must certify it.
The second holding might seem to depart from the background norm that a court can only order an agency to act, rather than telling it how to act. But sometimes only one course of action is available to the agency, in which case a court may direct the agency to do what the law requires it to do. And what we're proposing is that courts read the CEQA deadlines, in light of the HAA, as creating a ministerial duty to certify any legally sufficient environmental review document once the deadline for completing CEQA review has passed.
The courts could also give cities a brief window to decide what changes to the project or other mitigation should be required in view of the environmental study. This splitting of the baby – letting the politicians choose mitigation but not legally unnecessary environmental study past the CEQA deadline – would go a good distance toward reconciling CEQA with the HAA. It would breathe some life into the PSA deadlines (which the HAA incorporates into its definition of disapproval), without impinging on municipal authority to impose mitigation conditions on development approvals (which the HAA countenances so long as they don’t reduce the project’s density).
Solution #3: Levering “Pretext” for Judicial Review of the City Council’s CEQA Reversal
Our third solution is inspired by Chief Justice Roberts’s opinion in Department of Commerce v. New York. Instead of putting an expansive gloss on HAA ”disapproval,” or battling Schellinger to make the CEQA deadlines judicially enforceable, a court would hold that CEQA reversals are reviewable for pretext in limited circumstances. Specifically, a plaintiff’s “strong showing of bad faith” would render a city council’s CEQA reversal reviewable, and, if the court determines that the city council acted in bad faith, the court would hold the council’s decision unlawful.
This solution invites a number of questions. First, is it even available in California? Second, once the door has been opened to pretext inquiries in this context, what’s to keep them from spreading across all of state administrative law, at a high cost to courts and agencies alike?
Third, would this solution make a difference, given that the standard remedy in CEQA cases is a remand for a do-over – which is basically an invitation for the bad-faith agency to better cover its tracks?
As to the first question: The solution is available in the sense that it hasn’t been ruled out by California Supreme Court. Although there’s a pretty strong norm against looking behind the official record assembled by an agency, the Court has reserved the question of whether there might be a “limited” exception for “agency misconduct.”
The second question – whether pretext claims can be cabined – is serious but not hard to answer. The HAA and the institutions now being erected to enforce it offer guardrails. For example, a court could hold that the HAA’s concerns about municipal good faith warrant recognizing “CEQA pretext” claims vis-a-vis HAA-protected projects, if not otherwise. Going a step further, it could hold that the pretext claim is available only if HCD or the Attorney General makes the preliminary “strong showing of bad faith,” or otherwise raises serious concerns about the city’s development-review processes.
The remedy question concerns us more. If a court finds that a city’s CEQA reversal was pretextual, must it send the whole thing back and give the city another chance to dress up its decision, exactly as the U.S. Supreme Court did with the Census case? Not necessarily. The California Supreme Court has endorsed the “inherent power” of a trial court to send only part of a decision back to the agency, while retaining jurisdiction to issue judgment later. Perhaps a court in a pretext case could treat a CEQA certification as mostly complete (and valid), retain jurisdiction, and give the city a short period of time to address any legitimate concerns identified by the court on a limited remand. This would light a fire under the city and ensure that the case comes back to the same judge.
As motivation for this or another nonstandard remedy, consider what courts do when a decision-maker is found to have prejudged the facts or otherwise manifested bias in violation of due process. Normally the court disqualifies the biased arbiter and remands for a fair hearing before another hearing officer. The Court of Appeal has said that a city’s “malicious[] or arbitrar[y]” refusal to certify a CEQA document violates the developer’s right to due process. If that’s right, a city council’s bad-faith reversal of a CEQA certification violates due process too, and the biased decision-maker should be disqualified on remand. But a court generally cannot disqualify the whole decision-making body that must decide the case (as is true under CEQA), so there is no analogous remedy if a quorum of the council has shown bad faith. Hence the need for innovation beyond the usual do-over remedy.
All that said, the judicial norm against telling agencies what they must do is very strong, and without specific textual authorization – e.g., the HAA directing courts to order projects approved, or CEQA specifying deadlines for completion of environmental review – we fear that judges would be reluctant to deviate from the standard remedy, even in a pretext case.
One more point about remedies is worth mentioning. A bad-faith CEQA reversal that violates due process would make the city liable for damages. The prospect of having to compensate a developer for holding costs, and for the expense of the additional environmental studies, might be enough to discourage some cities from trying to launder housing denials through CEQA.
* * *
After a forty-year saga, the HAA is at a moment of truth. Will courts nodding to background principles of administrative law stand by while city councils deny 500-home projects on frivolous environmental grounds? Or will courts wake up to the HAA’s ditching of the old ways and appreciate – finally – that housing is the rare domain in which city councils are not to be trusted at all?
Comentários