Aziz Rana

Voter Confusion and the Single-Subject Rule: Prop. 26 as a Test-Case-in-Waiting, Part Two in a Two-Part Series

Friday, December 17, 2010

My prior column, Part One in this two-part series, introduced readers to Proposition 26, a state constitutional amendment adopted by California voters in November that (1) reclassifies many regulatory fees as taxes, thereby subjecting them to supermajority- and referendum-vote requirements under the California Constitution, and (2) changes the trigger for supermajority and referendum votes on tax increases.

Previously, only measures that would increase public revenues faced this high hurdle; following Prop. 26, a bill that increases taxes on any person (or corporation) must pass by supermajority and/or referendum vote, even if the bill reduces taxes overall.

Prop. 26, I explained, may be challenged on the ground that it violates the state constitution's "single-subject rule"--a rule meant to protect voters from "be[ing] misled as to the overall effect of [a] proposed amendment" by "improper emphasis . . . placed upon one feature."

In their arguments in the ballot pamphlet, and in the "Statement of Findings and Purposes" included into the text of the proposition, proponents of Prop. 26 highlighted the first reform (reclassifying "fees" as "taxes") and entirely ignored the second (changing the supermajority/referendum trigger).

Yet my last column concluded that a single-subject challenge to Prop. 26 would most likely fail under the law as it stands today. California courts have applied the prevailing test for single-subject violations leniently, mostly insensitive to the risk of voter confusion.

In this column, which concludes the series, I will explain how the courts may refine the test for violations so as to better realize the purpose of California's single-subject rule: to prevent initiative proponents from duping the electorate into adopting reforms that voters do not understand and would not support.

"Purpose" as Represented to the Electorate

The black-letter test for single-subject violations requires that "all of the parts" of a challenged initiative be "reasonably germane to each other, and to the general purpose or object of the initiative." There remains considerable ambiguity, however, about how judges should go about identifying a measure's "general purpose or object" -- as I explained in my last column.

Given the original rationale for California's single-subject rule, the courts should hold that the "general purpose or object" of a ballot initiative is the measure's central, public purpose as represented by proponents to the electorateand understood by ordinary voters.

So long as the principal components of a ballot proposition reasonably advance the measure's public purpose as ordinary voters understand it, the measure's adoption is unlikely to result from voters being misled as to the initiative's "overall effect" by "improper emphasis . . . placed upon one feature."

To be sure, there will often be doubt about what voters understand to be the public purpose of a ballot initiative. And different voters may see the same measure rather differently. But it would be easy enough for courts to implement the proposed approach using an objective "ordinary reasonable voter" standard, and guided by appropriate presumptions.

For example, the courts could hold that the purpose stated by the initiative's proponents in the ballot pamphlet is rebuttably presumed to be the public purpose as conveyed to and understood by voters. (Research by political scientists shows that many voters peruse the ballot pamphlet to get the gist of proponents' and opponents' arguments, and to see who supports and opposes the measure, although very few read and assimilate the detailed legislative analyst reports.)

That presumption might be rebutted with evidence of purpose as conveyed by proponents' mass-media advertising, or, less probatively, by purpose as discussed in news coverage. Opinion polls could furnish evidence of public understandings.

"Reasonable Germaneness" as Contextual: Preferences and Politics Matter

One more step is necessary to realize the original purpose of the single-subject rule. The courts actually have two options; either would suffice (the difference is mostly rhetorical).

One option is to interpret the reasonable-germaneness test not as a requirement of minimal means/ends fit, akin to the rational basis test familiar from equal protection jurisprudence, but as a contextual inquiry under which the "reasonableness" of a given level of "germaneness" varies with the nature of proposition at issue, the campaign for and against it, the preferences of the voters who supported it, and the proposition's margin of victory.

The second option is take the minimal approach to "reasonable germaneness," while expressly recognizing an alternative basis for single-subject invalidation--to wit, a showing that an electorally decisive number of the measure's supporters were confused about one of its component parts, and that once informed, they object to that part being included in the proposition.

A Hypothetical Illustrates the Problem with the Minimal Approach

Imagine that firms utilizing certain hazardous substances in the workplace are required to file detailed emergency-response plans with the state occupational safety agency. Following a horrific and widely-publicized accident, petitions are circulated for a ballot initiative that would require these companies to file emergency response plans with local governments as well, and that spells out in enormous and largely redundant detail precisely what must be included in the emergency response plan. (Assume that, prior to the initiative, it was open to local governments to request and receive firms' emergency response plans from the state agency.) The initiative establishes a modest penalty for noncompliance.

Buried in the measure's fine print is an exemption from all state taxation for the next one-hundred years for firms that meet two conditions: (1) the firm employs hazardous materials in the workplace; (2) during the last ten years or during the next ten years, the firm won (wins) an award from a named industry trade group for "outstanding safety practices."

There is little organized opposition to the measure, but the "Yes" campaign is pretty well funded--by a handful of corporations that recently won the award in question. The measure passes easily, on the strength of support from voters who still have terrifying images of the recent accident seared into their brains.

Does this initiative run afoul of the single-subject rule, thanks to the tax exemption? It certainly should.

The exemption does relate to the initiative's public purpose as understood by voters. (The public purpose is to prepare local governments to deal with industrial accidents, and the tax break arguably incentivizes firms to improve the plans they file with local governments.) But there can be little doubt that average voters who discerned and understood the tax exemption detested it, and that the exemption was not remotely necessary to achieve the initiative's public purpose.

Indeed, in light of the funders of the "yes" campaign, the fact that the bestower of the award (with its related tax break) is an industry trade group, and the doubtful relationship between the initiative as a whole and its public purpose, the initiative seems a paradigm case of "the potentially deceptive combination of unrelated provisions at which the [single-subject rule] is aimed."

Lessons for the Courts

This hypothetical suggests that courts concerned with voter bamboozlement should apply the "reasonable germaneness" test in a relatively stringent fashion where (1) the allegedly "separate" subpart of the proposition is substantively objectionable for well-informed voters who support the rest of the measure; and (2) the "No" campaign was poorly funded, or there is other evidence indicating that many "Yes" voters misunderstood or were unaware of the objectionable subpart.

Other circumstances will call for relatively lax applications of the reasonable-germaneness test. For example, if the allegedly separate subpart is found in a comprehensive, complex initiative that responds to a large, difficult problem, the courts should treat it gently (other things equal). Inevitable imperfections of drafting and judgment should not lead to the invalidation of reasonable efforts to use the initiative as a battering ram against social and political problems that the legislature won't touch. (See the concurrence of Justice Werdegar in Manduley.)

Nor should ballot initiatives be invalidated simply because they contain a section which, though functional vis-à-vis the measure's purpose and other parts, lacks independent majority support. (A point recognized in Brosnahan v. Brown.)

And courts should be more lenient whenever the proposition's description on the ballot itself fairly discloses the putatively "separate" part at which the single-subject challenge aims. (Recent work by political scientists indicates that ballot-level descriptions of initiative measures have a substantial effect on vote choice.)

Implications for Proposition 26

How does Prop. 26 fare under the framework I have sketched? Not well.

The first and potentially decisive point is that the public purpose of Prop. 26 was to close the loophole that enabled state and local lawmakers using clever nomenclature (labeling taxes "fees") to end-run the California Constitution's supermajority- and referendum-vote requirements.

Subjecting so-called "hidden taxes" to "the same vote requirements" that apply to regular taxes was the only objective stated in the proponents' argument in the voter guide, and in the findings-and-purposes section of the proposition itself. No doubt it was also the purpose conveyed to the electorate through mass-media advertising.

If "purpose" within the meaning of single-subject doctrine is purpose as represented to the electorate and understood by ordinary voters, then Prop. 26 looks quite vulnerable. Changing the existing trigger for a supermajority or referendum vote is not even rationally related to the stated objective of subjecting "fees" to the trigger that had applied to taxes.

This, arguably, is the end of the matter. A court ruling on Prop. 26 need not reach the question of whether "reasonable germaneness" is properly understood as a contextual inquiry that accounts for the likelihood of electorally decisive voter confusion, or whether the single-subject clause supports a separate, stand-alone test--a supplement to the "reasonable germaneness" inquiry--for claims predicated on voter confusion about subparts to which well-informed voters object.

But if the courts were to look further, they would discover additional reasons to invalidate Prop. 26--if, as the California Supreme Court has said, the "principal purpose" of the single-subject rule is "to avoid confusion of . . . voters . . . and to prevent the subversion of the electorate's will."

Most importantly, as my prior column explained, conservatives, liberals, and moderates alike who understand the new trigger are likely to find it objectionable on the policy merits. It will benefit the tobacco, oil, and alcohol interests that funded the Yes-on-26 campaign, but not ordinary taxpayers or principled ideologues. In the scholarly lingo, the new trigger is a rider, a disfavored public policy piggybacking on an otherwise popular reform.

It is one thing for voters to be confused about a matter to which they are indifferent on the merits; it is quite another for them to be misled about a reform to which they would object.

Also significant is that the election was very close. A vote swing of 2.5 percentage points would have changed the outcome. Given that most "Yes" voters likely would disfavor the new supermajority/referendum trigger in a world of full information, voter confusion probably determined the election outcome. This assumes, of course, that a significant share of "Yes" voters were obtuse to the trigger issue.

Prop. 26's Supporters Have Some Arguments to Raise -- But None Is Persuasive

Defenders of Prop. 26 have a couple of cards they could play in response (accepting for now the doctrine I have urged). One, they might point to the legislative analyst's statement in the voter guide, which clearly explains the change to the supermajority/referendum trigger.

But as I argued earlier, legal analysis under the single-subject rule should reflect the practices and capacities of "ordinary reasonable voters," and few propositions would command wider assent among political scientists than that ordinary voters do not read and assimilate the legislative analyst reports in California's mammoth voter guides.

Second, Prop. 26 defenders could point to the financial resources of the "No on 26" campaign. While the "Yes on 26" campaigns raised nearly five times as much money as the "No" campaign, the "No" campaign did amass several million dollars. Whether opponents had enough money to reasonably inform the electorate about the new trigger is doubtful, however, given that the issue is technical and not amenable to catchy sound-bite advertising, and given the expense of mass-media advertising in California.

Needless to say, anyone thinking of bringing a single-subject challenge to Prop. 26 would do well to conduct opinion and knowledge surveys of California voters posthaste, so as to better understand what "Yes on 26" voters knew about the reforms they wrought.

Larger Objections, Briefly Answered

There are at least two substantial objections to my confusion-centric approach to the single-subject inquiry. One is that, as an interpretive matter, it is too far removed from the text and history of the California Constitution.

Professor Daniel Lowenstein maintains that the drafters of the 1948 single-subject amendment meant to apply to constitutional ballot initiatives the same standard that had long been applied to legislation. This plausible understanding was accepted in a contemporaneous decision of the California Supreme Court.

Moreover, the "plain meaning" of the constitutional text arguably calls for an abstract inquiry into the "subject(s)" of a challenged measure, not an assessment of voter beliefs and political context. (But my approach does not flaunt the text. It requires, among other things, that the initiative contain a discrete "part" about which voters were confused.)

Whether one sides with Lowenstein's approach or my own probably boils down to essentially unarguable differences of jurisprudential disposition. If you believe, as I do, that constitutional provisions should be read flexibly, in light of changed conditions, to realize the ratifiers' purposes, then you'll favor my approach. (Consider the role of the initiated constitutional amendment in California governance circa 1948, and circa 2010.) If you don't, you won't.

The other objection is that my approach would politicize single-subject adjudication. Professors Rick Hasen and John Matsusaka have provided evidence that judicial application of the single-subject rule is heavily influenced by judges' partisan affiliations in states with demanding single-subject tests, but not in states like California with deferential doctrines.

I have three responses to the politicization worry. First, as many scholars have noted, the conventional "demanding" approach to the single-subject rule consists of a subjective inquiry into whether the parts of a ballot measure have something in common.

My approach provides vastly more guidance to judges. It centers the "purpose" inquiry on what the initiative proponents represented to the electorate, and it ultimately resolves into a question that can be answered more or less objectively: Did the initiative contain a discrete subpart that a decisive number of "Yes" voters did not understand and (once they understand it) do not like? There would be hard cases from time to time, but this is true of most areas of law.

Second, it would be a mistake to characterize single-subject adjudication as predominantly partisan or ideological. Professor Michael Gilbert's ingenious empirical research shows that legal variables (both express and implicit) explain much more of the variation in single-subject decisions than do ideological variables, even under the current mushy standard. To the extent that the legal test can be focused and refined, extra-legal considerations are likely to have even less sway.

Finally, if the California Supreme Court becomes anxious about judicial partisanship in single-subject adjudication, it is open to the Court, as a prudential matter, to bar voter-confusion/single-subject invalidations except in cases where the "secondary part" of the proposition is substantively undesirable from both liberal and conservative perspectives. This would make the doctrine harder to apply, with a straight face, in a manner that narrowly advances judges' personal political preferences. But it would not save Proposition 26.

Chris Elmendorf is a professor of law at the University of California, Davis. A graduate of the Yale Law School and a former clerk to Judge Guido Calabresi, he writes principally on the subject of election law.

Ads by FindLaw