Why An Important Part of the California Recall Process Is Unconstitutional, According to U.S. Supreme Court Precedent
|By VIKRAM DAVID AMAR AND ALAN BROWNSTEIN
Friday, Jul. 25, 2003
This column is part of a series by the authors on the California recall process. - Ed.
California is, if nothing else, always an exciting place, what with all the earthquakes, wildfires, mudslides and the like. And when Mother Nature fails to provide any fireworks, we Californians make our own, often through the device known as direct democracy.
The most common form of direct democracy entertainment in California in recent years has been the statewide initiative. But this summer we are being treated for the first time its close cousin -a statewide gubernatorial recall campaign.
And where there is unusual direct democracy, there is bound to be litigation. As Richard Hasen pointed out earlier this week, the movement to recall Governor Gray Davis has already generated a number of lawsuits, and is likely to generate many more.
In a series of columns beginning with this one, we will attempt to analyze the merits of various legal claims that have been brought, or may soon be brought, relating to the recall process. In particular, in our column today, we focus on one very forceful argument that suggests at least part of California's recall process, California Elections Code §11382, violates the United States Constitution under U.S Supreme Court precedent.
This argument is just beginning to surface in the various litigations and media discussions (in part based on conversations we have had with various persons). In any event, we think it holds more promise for recall opponents to obtain some success than any of the other challenges currently being pursued or talked about.
How Recall Works in California
Before we get into the specifics, some general background is required. Added to the California constitution at the beginning of the 20th Century, the recall power was one piece (along with direct election of U.S. Senators and other populist measures) of the Progressive era reforms.
A recall in California for any state official is initiated by the delivery to the California Secretary of State of a petition alleging reasons for the recall. (Insufficiency of the reasons, however, is not a basis for rejecting the petition.)
The California constitution then provides that, to place a recall on the ballot for a statewide office such as a governor, recall proponents must obtain the signatures of a number of California voters equal to at least 12% of the persons who voted at the last election for the office in question. In current Governor Gray Davis's case, 12% of the number of voters in the last gubernatorial election is somewhere in the neighborhood of 900,000 persons. A California statute provides explicitly that only registered voters in the State can solicit or collect the required signatures.
When the requisite number of signatures are gathered, turned in and certified as valid (as was done the night of Wednesday, July 23), an election shall be called, to be held no less than 60 days, and no more than 80 days, from the date of signature certification. The election shall determine whether to recall the officer and, "if appropriate," to elect a successor.
The ambiguous phrase "if appropriate," over which people are now wrangling, appears in the California constitution itself. Most people assume that in a gubernatorial recall alternative candidates may appear on the ballot; some people, however, are arguing that if a governor is recalled, the Lieutenant Governor - here Democrat Cruz Bustamante - would automatically succeed him, and no other candidacies may be voted upon.
If a majority votes to recall, the officer is removed, and if there are successor candidates (assuming it is "appropriate" to have successor candidates on the ballot), the one with the most votes - no matter how few - becomes the successor.
Under the terms of the constitution, a recalled officer may not himself be a candidate for successor. (In other words, Gray Davis could not be considered to fill the vacancy his own recall creates.)
What if the Governor Has More Votes Than the Recall Election Winner?
This framework could give rise to a number of thorny legal questions.
For example, suppose that 49% of the voters favor keeping Davis, but he is recalled because more than 50% vote against him. Suppose further that there are successor candidates on the ballot, and that the leading vote getter among them - call him candidate A (for Arnold?) - gets only 10% of the successor vote. Can it be that a candidate with 10% support will oust a governor with 49% support?
If so, could this result be squared with the venerable constitutional principle of one-person-one-vote, where everyone's vote is supposed to be counted equally? And what about Bush v. Gore, in which the U.S. Supreme Court reaffirmed that the Equal Protection Clause of the Fourteenth Amendment requires that each voter's ballot be treated similarly to all others'?
These are non-frivolous questions. In the end, though, we think that there is no federal constitutional equality problem with a 10% candidate beating a 49% governor.
The key lies in the fact that the governor is not eligible to be a candidate once he is recalled. In effect, the disqualification of recalled officials in California's constitution acts like a Term limit - it prevents someone from running, even if he has popular (or the most) support.
If Term limits are permissible at the state executive level (and they are), then there is no problem with invoking them here to prevent Davis from filling the vacancy his recall creates. So we are left with candidate A, who is the leading vote getter among the legitimate candidates. And there is nothing wrong with giving an election to the leading vote getter, even if the field is so large that the winner has a relatively small percentage of the vote. (We do it all the time in primaries.)
All of this may be why the first major lawsuit filed by recall opponents did not focus on the inequality of Davis losing to a less popular candidate.
Was the Signature Gathering Process Proper?
Instead, in papers filed last week in Los Angeles in the case of Robbins v. Shelley, recall opponents challenged the validity of the signature gathering that took place.
The plaintiffs' Complaint alleges a number of problems with the way signatures had been gathered. But no allegation is more central than the contention that the gatherers were not all registered California voters, as California statutes require. In particular, Elections Code §11045 says that "[o]nly registered voters of the electoral jurisdiction of the officer sought to be recalled are qualified to circulate or sign a recall petition for that officer."
One big response to plaintiffs' contention in the Robbins case is that the California law, §11045, is displaced by the U.S. Supreme Court ruling in Buckley v American Constitutional Law Foundation, which we will refer to as ACLF. In ACLF, the Court struck down a requirement in Colorado that signature gatherers for ballot initiatives (as opposed to recalls) there be registered voters. The Court said the requirement violated the First Amendment rights of gatherers - preventing them from collecting signatures and petitioning the government for change simply because they may have chosen not to register to vote.
In ACLF, the State of Colorado suggested that a voter registration requirement does not impose much of a burden on individuals. Crucially, the Court pointed out, in response, that a person may choose not to register not out of laziness, but rather out of political conviction - a "form of private and public protest." And "for these voter-eligible circulators, the ease of registration misses the point."
Reasoning that people should not have their right to petition government taken away simply because they make an ideological choice not to register, the Supreme Court invalidated Colorado's law.
The ACLF case may be wrongly decided. After all, Colorado wasn't really punishing people for not registering to vote - it wasn't throwing them in jail or anything. Instead, it was just telling people that if they wanted the signatures people collected to count for placing an initiative on the ballot, the collectors had to be registered voters. Colorado doesn't need to provide an initiative process. Surely, if it to chooses to have one, it can have content-neutral rules for placing initiatives on the ballot.
But right or wrong, ACLF is a United States Supreme Court case that is fully binding on California and its courts. Unless initiative processes can be distinguished from recall processes (and we're not sure they can), California courts will likely have to invalidate §11405.
The state trial court before whom the Robbins case was filed does not have the power under California law to enjoin implementation of a California statute until an appellate court has specifically invalidated that statute. Yet surely the trial court's decision last Friday to deny the plaintiffs a preliminary injunction stopping the signature counting was based, in some measure, on ACLF .
The Robbins case is on appeal this week, to the California Court of Appeal. We would not be surprised to see the Court of Appeal, following ACLF, strike down §11405 and reject plaintiffs' challenge.
Suppose the Court of Appeal does strike down §11405 as unconstitutional. What happens to the rest of the recall scheme, which now is missing a provision? The courts must then decide whether the section that was struck down is "severable." Put another way, the courts must ask, is that section separable from the rest of the recall scheme, so that the offending provision may simply be removed, or must the entire scheme all fall because of one unconstitutional part? We expect courts would sever §11405.
The Part of the Recall Process That Is Likely to Be Held Unconstitutional
In any event, constitutional law is often a double-edged sword. Recall proponents in California may be correct that the ACLF case undercuts plaintiffs' challenge to the signature gathering process. But the reasoning underlying the ACLF decision calls into question another aspect of California's recall scheme - Election Code §11382.
Section 11382 provides that "No vote cast in the recall election shall be counted for any candidate unless the voter also voted for or against the recall of the officer sought to be recalled." Put simply, this requirement conditions the right to vote for a successor to a recalled official on the voter's willingness to weigh in on the recall itself, by voting for or against the recall measure. But what if the voter's ideological preference was to abstain from voting on that issue?
That's where ACLF become relevant once again. Recall that ACLF held that Colorado could not constitutionally burden participation in political conversations by conditioning signature gathering for initiative petitions on an individual's registration as a voter. By the same logic, courts may well hold that California may not constitutionally condition the right to vote on an individual's prior voting history.
Of course, conditions on the exercise of fundamental rights do not always abridge constitutional guarantees; not every burden on the exercise of a right is sufficiently substantial or undue to require invalidation. Some administrative requirements serve such obviously important state interests, and impose so slight an obligation on the individual, that no infringement of the right is recognized.
But the burden on the individual is not measured exclusively in time and effort. Rather, ACLF makes clear that when a court assesses the weight of a condition for constitutional purposes, the ideological cost of compliance must also be taken into account.
Thus, the Court explained in ACLF that the state's emphasis on the ease by which a person may register to vote "misses the point" because "the choice not to register implicates political thought and expression." For those persons as to whom the refusal to register to vote constituted a form of political protest or communication, the Colorado registration requirement struck down in ACLF imposed an unconstitutional burden on their participation in the political process.
There is a direct parallel with §11382. Some individuals conscientiously oppose recall elections as a matter of principle. They may not wish to participate in the recall - even by voting "No" - because participation implicitly legitimates the recall process, and because a "No" vote suggests support for the challenged official.
In addition, some people may be truly agnostic about Gray Davis - completely on the fence - and do not want to vote on his recall, since either a "yes" or a "no" vote would misstate their position.
Section 11382 imposes a serious burden on such individuals. They can sacrifice their integrity by voting on the recall itself, against their convictions. Or, they can lose the right to participate in deciding who will be chosen to replace a recalled official.
Indeed, the burden §11382 places on voters to betray their ideological convictions is surely greater than that presented by the Colorado statute implicated in ACLF. How many people, after all, opt out of registration for political reasons and then work to qualify a measure for a ballot they have opted out of, as in the Colorado case?
The ACLF Ruling May Play an Important Role in the Recall Battle
The ACLF ruling thus casts doubt on a state's ability to burden residents' ideological choices in the way that §11382 does. At a minimum, ACLF suggests that conditions of this kind must further state interests of sufficient importance to justify the burdening of a right - in this case, the right to participate in the selection of a successor. It is not clear to us that S 11382 furthers any legitimate state interest - much less one that justifies this condition on the right to vote, and the right to demonstrate one's opposition to an electoral recall mechanism by refusing to participate in it.
We recognize that some may suggest differences between ACLF and the California recall issue we have identified. And we will address those possible distinctions at some length in Part Two of this column series next time.
For the moment, we will mention only one such difference. In ACLF, the Court evaluated the challenged registration conditions as limits on political speech - petitioning government by gathering signatures. Section 11382, by contrast, burdens not the right to gather signatures, but rather the right to vote itself.
This difference could arguably be significant to the fate of S 11382. The Court has often suggested that the need for flexibility on the part of states means that ballot access requirements, and other electoral regulations, may sometimes receive a slightly more lenient level of scrutiny than that which is ordinarily applied to regulations of the right to speak and petition.
Nevertheless, the standard of review applied to electoral regulations is quite a serious one - for the right to cast one's vote is a fundamental right. To be a constitutionally permissible regulation of voting, a state regulation must be "reasonable," in that it must substantially further a significant governmental purpose.
Moreover, as the Court has made clear, the reasonableness test has some firm guideposts: "[A] [r]easonable regulation of elections does not require voters to espouse positions that they do not support." And yet that is precisely what §11382 may in fact do, and for no good reason at all.
If a federal court were to hear a facial challenge to §11382 brought by an appropriate plaintiff, we think it may ultimately feel compelled to strike the statute down based on the precedent of ACLF. (A facial challenge attacks the statute in general, not in the context of any particular application to any particular person or circumstance.)
If a court were to strike down §11382, what would happen next? Again, the questions would concern severability. Should the rest of the recall statutory scheme be severed from the tainted §11382, or struck down along with it? And even if the rest of the recall scheme could be severed, must California rewrite the ballots before the impending election in a way that would comply with the federal Constitution? We take up those matters in our next column, as well as some other potential challenges to the recall scheme, including whether California can require a would-be successor candidate to obtain signatures from a political party to qualify for the ballot, when he may not be a member of any party.