When should judges protect the people, and when should they defer to them? In countless contentious cases, courts have split: majorities invalidate laws to defend rights; dissents decry the decisions as undemocratic.1 Recently, in Martin v. City of Boise,2 the Ninth Circuit engaged in that familiar back-and-forth, this time sparring over the constitutionality of two city ordinances that banned sleeping and camping on public property. After a panel held that absolute bans violate the Eighth Amendment rights of homeless people,3 Judge Milan Smith, dissenting from a denial of rehearing en banc, accused his “unelected[] colleagues [of] improperly inject[ing] themselves into the role of public policymaking”4 and thereby creating chaos for “hundreds of local governments . . . and . . . millions of people.”5 But this fractious debate is belied by the panel’s narrow holding, which neither protected homeless people nor precluded democratic politics. Indeed, the incremental political achievements of Martin’s long litigation process may prove more significant than the panel’s divisive, undemocratic decision.
A decade ago, in Boise, Idaho, it was illegal to sleep in public. One ordinance banned “[o]ccupying, lodging or sleeping in any . . . place . . . without . . . permission”;6 another barred the “use [of] any . . . streets, sidewalks, parks or public places as a camping place at any time.”7 Janet Bell was cited twice, once for sitting on a riverbank with her backpack, another time for putting down a bedroll in the woods.8 She pled guilty and received a thirty-day suspended sentence.9 Robert Martin, who has difficulty walking, received a citation for resting near a shelter.10 He was found guilty at trial and charged $150.11
On October 22, 2009, Bell, Martin, and nine other homeless people sued the City.12 They claimed that the enforcement of the ordinances violated their Eighth Amendment rights, criminalizing them for carrying out basic bodily functions.13 Using § 1983,14 they sought expungement of their records, reimbursement for fines, enjoinment of enforcement, and a declaration that the ordinances were unconstitutional.15
On July 6, 2011, the district court granted summary judgment to the City.16 Retrospective relief, Magistrate Judge Bush found, was barred because the plaintiffs did not challenge their convictions in state court before bringing their federal case.17 Prospective prohibition of enforcement, he held, was moot because there was no longer any “reasonable expectation” of an Eighth Amendment violation.18 While he admitted that a “complete bar on sitting, lying, or sleeping in public at any time of day” would unconstitutionally criminalize homeless status,19 Magistrate Judge Bush highlighted three limits on the ordinances — all instituted in the immediate aftermath of the case’s filing. First, in November 2009,20 the City Council redefined camping, restricting it to overnight stays on public property;21 second, on January 1, 2010, the Boise Police Department issued a Special Order that prohibited enforcement of either ordinance when shelters were full;22 third, in the same Special Order, the Department formalized its policy not to enforce either ordinance during the day.23 The ordinances, Magistrate Judge Bush concluded, now criminalized voluntary conduct, not homeless status.24
In 2013, the Ninth Circuit reversed and remanded.25 Contra the district court decision, it held that retrospective relief was possible because the plaintiffs were challenging the City’s enforcement, not their state court judgments.26 Prospective enjoinment of enforcement was not mooted either, it held, because the Special Order was only an “internal policy,”27 not a “permanent change” to local laws.28
On remand, the district court, in two opinions, again granted the City summary judgment. In 2014, Magistrate Judge Bush barred retrospective relief because the plaintiffs had not contested their convictions before filing the case.29 In 2015, he prohibited prospective relief because the City Council incorporated the Police Special Order into the City Code and thus made unconstitutional enforcement unlikely.30
In September 2018, the Ninth Circuit affirmed with respect to the plaintiffs’ requests for retrospective relief, but it again reversed and remanded with respect to their requests for prospective relief.31 Judge Berzon, writing for the panel,32 began with the persistent procedural problems. First, she found that, despite the incorporation of the Special Order into the City Code, future prosecution was still possible: even on nights when beds were empty, the plaintiffs might not be able to sleep in shelters with religious or continual-stay restrictions.33 Second, she agreed with the District Court that retrospective relief was mostly barred by the plaintiffs’ failure to contest their convictions,34 but she found that prospective injunctive relief under § 1983 remained viable.35 Finally, she reached the merits of the Eighth Amendment challenge. Any ordinance that allowed for the “imposition of criminal penalties for sitting, sleeping, or lying outside on public property for homeless individuals who cannot obtain shelter,” she argued, unconstitutionally criminalized homeless status.36 This holding, she insisted, was limited to involuntary conduct: the Eighth Amendment bars punishing a person only “for lacking the means to live out the ‘universal and unavoidable consequences of being human.’”37
In April 2019, the Ninth Circuit denied rehearing en banc38 over the objections of two “dramatic” dissents.39 In a concurrence, Judge Berzon called the dissents “petitions for writ of certiorari on steroids, rather than reasoned judicial opinions.”40 Judge Bennett, in one, attacked the panel’s holding as contrary to the Eighth Amendment’s original meaning.41 The ban on cruel and unusual punishment, he wrote, was meant to invalidate “methods of punishment,” not to limit criminalization.42
Judge Smith, in the other, claimed the panel perverted precedent to enact its policy preferences.43 First, he claimed, no binding precedent held that laws that criminalize involuntary conduct unconstitutionally criminalize status.44 Second, he argued, the panel effectively invalidated prior convictions by granting prospective relief even though the Supreme Court bars that action in such cases.45 Finally, he explained, the panel allowed two plaintiffs without convictions to sue for retrospective relief even though the Court allows such relief only “after . . . a formal adjudication of guilt.”46 The panel’s precedent-stretching decision, he argued, could cause “dire . . . consequences.”47 Cities that did not “undertake [the] overwhelming financial responsibility”48 to provide adequate shelter for and accurate counts of homeless people would be unable to enforce bans on homeless people’s “defecation,” “urination,” and “use of hypodermic needles.”49 As a result, disease and despair would spread across the Ninth Circuit — home to some of the country’s highest concentrations of homeless people.50 “[T]he Eighth Amendment,” he concluded, “is not a vehicle . . . to critique public policy choices or to hamstring a local government’s enforcement of its criminal code.”51
Did Martin protect homeless people or preclude democratic politics? On closer inspection, the case neither nullified the range of laws that punish homeless people, nor prevented the political process from addressing “the serious societal concern of homelessness.”52 Instead, the panel’s holding is narrow enough that it may prove insignificant for homeless people in the Ninth Circuit, but the process of litigation led to positive political developments for Boise’s homeless residents. Indeed, by forcing the City to account for its actions in an adversarial forum, the case created opportunities for homeless people to participate in reforming policies designed to exclude them. In this way, Martin suggests that debates about judicial review may be too focused on judges’ decisions, missing the impact of the litigation process itself.
In their dueling opinions, Judges Berzon and Smith advanced archetypal arguments about judicial review. Judge Berzon embraced judicial supremacy to protect a vulnerable minority, whose rights, she suggested, were too easily trammeled.53 Judge Smith insisted that the panel’s intervention was illegitimate — it used the Constitution to invalidate democratic decisions54 — and ill-advised55 — it constrained cities’ efforts to aid homeless people and protect other citizens.56 Despite their differences, then, both agree that judicial decisions are what matters.57
For advocates of judicial review, the Martin decision is likely to prove insufficiently protective. As Judge Berzon noted, “only . . . municipal ordinances that criminalize sleeping, sitting, or lying in all public spaces, when no alternative sleeping space is available, violate the Eighth Amendment.”58 Already, lower courts are following the panel’s lead: under Martin, cities can clear homeless camps,59 arrest those who refuse to leave,60 and force those arrested to show that shelters are full.61 Put simply, the panel left cities ample power to police and punish homeless people, as well as regulate and restrict their access to public space.62
Moreover, to effect the panel’s narrow holding, cities must enact only a minor policy. To satisfy Judge Berzon’s ruling that a city cannot prosecute homeless people for sleeping in public when there are more homeless people than available beds in shelters, cities need simply to create some way to know that shelters are full or, because of restrictions, effectively so. Judge Smith insisted that even this was burdensome,63 though it undoubtedly falls well short of the expense needed to build housing, increase shelters, or pursue other options, like expanding employment, providing healthcare, or increasing benefits.64 In short, the panel’s decision undermined popular power to produce little positive change.
Why, then, does Martin matter? As the plaintiffs recognized, Boise wanted “to drive . . . homeless individuals out of the City” by punishing them for staying still.65 This is a long-standing tactic: cities and towns have imposed penalties on those passing through — casual workers, day laborers, poor people — to keep them moving.66 The case’s most significant impact, then, was to limit cities’ ability to push homeless people out; by allowing them to stay somewhere within Boise’s boundaries, the panel turned homeless people into part of the City’s public. In this way, Judge Berzon’s decision, despite its substantive aura, is closer to a democratic-procedure-protecting judicial intervention.67
But, at least for Boise’s homeless people, this success did not require the court’s decision.68 In fact, it’s not clear that the court’s decision will have any effect on Boise’s actual policies.69 Instead, it seems, the litigation itself facilitated the inclusion of homeless people in the City’s political process. First, consider the immediate impact of the complaint. Less than a month after its filing, the City Council clarified the definition of camping, narrowed the scope of police discretion, and restricted enforcement to the time between “sunset and sunrise.”70 The Police Special Order, issued two months later, restricted officers from enforcing the ordinances on nights when no shelter beds were available.71 In 2014, the City Council voted to include these policies in the ordinances themselves.72 All this happened without a court deciding whether the plaintiffs had standing, let alone whether the ordinances were constitutional.
And policy changes were not the only consequences of the case. For one, the litigation created a record of sworn testimony about Boise’s policies: by the final filings, the City’s lawyers were ready to admit that issuing camping citations, let alone pursuing convictions, should be a “last resort.”73 The case also prompted the Department of Justice to file a rare Statement of Interest criticizing the City — and pitting one level of government against another.74 Finally, by challenging the City in a forum where they had some control over their stories, the plaintiffs spurred media coverage that questioned the City’s policies, highlighted the lack of shelter beds, and, generally, reframed their struggles as city-wide concerns.75 Again, none of this required a favorable ruling on the merits.
Maybe Martin wouldn’t have had any collateral consequences if judges didn’t sometimes make divisive, undemocratic decisions. Maybe the City needed the threat of a court’s coercion to come to the table; maybe the press needed the prospect of an aggressive ruling to cover the cause. But Martin shows, at least, that advocates and adversaries of judicial review overemphasize how pivotal judges are. For homeless people in Boise, the decision may make no difference; the case had already facilitated sought-after reforms. Even these, though, won’t come close to addressing the hardships of being homeless. What should Boise, and cities like it, do? That’s something no judge can decide.