Reflections on the California Recall
The Lingering Questions Over Its Legitimacy, And Its Basis in the California Constitution

This column is part of a series by Professor Amar, sometimes co-writing with Professor Alan Brownstein ,on the California recall process. The other columns in the series may be found in the archive of Professor Amar's columns on this site .- Ed.

As the world now knows, the California gubernatorial recall campaign was successful, in the sense that Governor Gray Davis was recalled. California will now have a new governor, Arnold Schwarzenegger.

Probably the most recurring criticism that intellectuals on the left voiced against the recall effort was this: the recall device was intended to be used only in certain rare circumstances, and none of those circumstances was present in Gray Davis's case. Instead, with respect to Gray Davis, the State was faced with the not uncommon scenario of a Governor's prior opponents continuing to oppose him, and wishing for a new election so they could replace him. And, according to this criticism, the recall proved to be just that new election.

Is this criticism valid? In this column, I will take up this critique. Was the 2003 recall merely a chance to rerun the 2002 election to reach a different result? And if so, was such a rerun necessarily illegitimate?

Was This Election a Do-Over Of November 2002?

For critics, last week's recall election is best characterized as a "do-over" of November 2002 - when Democrat Davis beat Republican Bill Simon - instigated and ultimately carried by people who were unhappy with the 2002 results. Not only did California get rid of Gray Davis, these detractors urge, it also changed the rules under which the quadrennial election game is played.

There is clearly something to this account. In 2002, Gray Davis received about 48% of the vote, six percentage points more than Simon but still less than half of the electorate. Last week, about 45% of the people voted against the recall - voted, that is, effectively to keep Davis in office. These percentages, then, are quite comparable.

It is true, as has been publicized, that Davis lost over 20% of Democrats on Question 1 of the recall ballot (the one that asked for an up/down vote on the recall). But it is also true that almost 80% of Democrats voted against the recall, whereas almost 90% of Republicans voted in favor of it. (And Davis lost a lot of Democrats in 2002, too.)

In sum, people in 2003 pretty much voted with regard to Davis the same way they did in 2002 - along partisan lines. (Independents, however, broke in favor of the recall in 2003 in greater numbers than they did for Bill Simon in 2002.)

The Differences Between the 2002 Election and the 2003 Recall

Of course, there were also some differences between 2002 and 2003, the biggest one being that in 2003 Davis needed 50% to win (under the rules of the recall), whereas a year earlier he was able to prevail with just 48%.

Another and related difference was the structure of the ballot. In the recall election, voters first voted on the recall (and thus on Davis), and then voted separately on who should replace him in the event of a recall.

Unfortunately, the media focused overwhelmingly on Part Two of the ballot. As a result, the election essentially became a choice between four candidates (Davis, Schwarzenegger, Tom McClintock and Cruz Bustamante). Most media largely eschewed any focused inquiry on Part One. That is, they did not explore the nature and purposes of the recall device, and why, or why not, Davis's conduct implicated these purposes. In light of these realities, Davis's claim (made unsuccessfully in court) that his name should have been allowed on the second part of the ballot seems more sympathetic, if ultimately legally untenable.

It also turns out that Schwarzenegger was more popular than was Simon, attracting more of the anti-Davis votes. In part because of his celebrity, and in part because of his stand on social matters, Schwarzenegger may have been a more electable Republican in California than was Simon.

Even if the 2003 Election Was a Do-Over, Was It an Illegitimate Do-Over?

So suppose the election was, indeed, largely a do-over - as critics charge and a fair amount of evidence indicates. That leads to a second question: Is there anything wrong with that? Or, put another way, was this do-over, with Arnold replacing a less popular Republican, necessarily illegitimate?

Let us consider some responses. First is the response that since the California Constitution does not say what a good reason for a recall is, and indeed explicitly prohibits the Secretary of State from refusing to certify signatures because the reason for recall is insufficient, anything goes and all is equally legitimate.

In the same vein, the California Constitution's provision that says impeachment is reserved for "misconduct in office" establishes a standard, even if no court or bureaucrat can administer it. The fact that a standard is not enforced by some governmental body doesn't suggest that the standard is meaningless: A standard for when recall is proper, for instance, can provide a factor would-be recall-petition-signers and for recall voters should consider. It can also provide a benchmark around which debate can focus.

But - someone might quickly point out - there is a difference here. The impeachment provisions mention a standard, whereas the recall provisions do not. Doesn't the absence of a textual threshold for recall suggest that all is fair game?

Perhaps, but I think not. Text is but one of the important tools we use when giving meaning to constitutions, especially as relates to structural relationships between institutions. A comparative look at the U.S. Constitution will demonstrate why this is so.

Interpreting Constitutions: The Need to Look At Not Only Text, But Structure, Intent and History

Nobody thinks that a U.S. Vice President can preside over his own impeachment trial. But the text would suggest that he can. The U.S. Constitution provides that the Vice President presides over the Senate generally. And the sole textual exception to this speaks only to Presidential impeachment (over which the Chief Justice is to preside).

But of course, we all agree the Vice President can't preside over his own impeachment. Why? Because we look not only to the document's text, but also to structure, history, intent and common sense, all of which supplement the text of the document.

To take a more current example, suppose Bill Clinton were to run as Wesley Clark's running mate. And suppose that, during the campaign, Clark and Clinton were to make clear that if Clark were elected, he would resign and let his Vice President (Clinton) succeed him.

Technically, textually, such a result would not violate the Constitution's Twenty-second Amendment (which prevents a person from being "elected to the office of the President more than twice"). But even Bill Clinton, I think, would agree that this result would violate the structure and spirit of the Amendment. Even Bill Clinton would believe that we should change the Constitution before he should be allowed near the Oval Office again.

So too with California's recall power. Although there is no explicit text governing the standard for its invocation, text is not the only evidence relevant to its meaning. We must consider, also, its place in the larger framework of the Progressive agenda (which included the initiative, the referendum and the direction election of United States Senators). And we must consider the intent of the people who ratified the 1911 amendments to California's constitution (which inserted the recall powers). All of these sources indicate that recalls were designed first and foremost to deal with politicians who were corrupted by big corporate or other special interests.

In sum, there is strong evidence that, under the California Constitution, the recall power was meant to be limited to two scenarios. The first is post-election corruption. The second is striking post-election promise-breaking that makes the campaign by which the official had been elected in the first place seem misleading.

Under the Constitution's Existing Standard, It's Hard to See Compelling Reasons to Recall Davis

People can argue back and forth about whether Gray Davis was captured by special interests. But certainly he was no more captured in 2003 than he was when he was elected in 2002. And although he may have reneged on certain key promises, again these disappointments preceded his reelection last year. So neither of the two main reasons for recall appears to be applicable here.

The only issue on which Davis arguably changed his views after being reelected was the size of the budget deficit, and that is something early recall proponents tried to publicize when they were attempting to shoehorn this recall into the mold of earlier precedents.

But if Gray Davis' 2002 election was flawed because he lied about the deficit, so too was the election of each and every legislator; all were more or less equally cagey about the budget crisis in an election year. And it is hard to imagine that voters would want to recall all the lying legislators, even if given a chance.

In the end, it's plain (to me at least) that Gray Davis wasn't recalled for corruption, or for shifting his agenda. He was recalled because he was singularly unpopular, and he was singularly unpopular because he was singularly identified with failures - both policy and personal - that went far back before November 2002. The big reason he wasn't booted in the 2002 election is that voters didn't feel that they had an acceptable or attractive alternative..

A Majority of Californians Could Agree to A Do-Over - But Did They?

Even if this recall was indeed an unprecedented "do-over" election because of dissatisfaction with the last one, can't a majority of Californians agree to have a do-over?

Indeed, doesn't the fact that a majority of California voters, in an election with a high turnout, voted in favor of the recall blunt any concern about whether the recall device was intended for such things? A majority of people can change the California Constitution - they can add new grounds for recalls - so why isn't that the best way to construe what was done here: an addition of a new basis for recalls, a do-over election when the last one was really dissatisfying?

That has been the case from the beginning, and the right of the people to alter and abolish their forms of government is the basis for the legitimacy of both state and federal Constitutions. Consider an anecdote from history that illustrates how fundamental this idea was: When the original federal Constitution was proposed for ratification, it was criticized on the ground that it disregarded the rules for change that were laid down in each of the state constitutions and was thus "illegal." However, federalist proponents from James Madison to Alexander Hamilton to James Wilson responded by pointing out that so long as the new document was approved by the people, it could not be "illegal;" ultimate legality - by definition - meant only conformance with the will of the people. As Madison put the point in Philadelphia, "[t]he people were in fact, the fountain of all power, and by resorting to them, all difficulties were got over. They could alter constitutions as they pleased."

So the people of California can alter - and expand - the state constitutional grounds for recall. But they should do so knowingly and deliberately. The popular sovereignty right of people to alter and abolish their forms of government presupposes a careful, thoughtful, considered judgment about what the ground rules for everyday government ought to be.

(Indeed, that is why the original federal Constitution was ratified by "conventions" - groups of people specially selected for the sole and important task of focusing on a single set of issues, namely the wisdom of the Constitution's ground rules. Such conventions were moments of high, rather than ordinary, politics. Even the most ardent supporters of popular sovereignty acknowledge that majority rule is not the same as majority will, and certainly not the same as majority whim.)

And so far, the wisdom of allowing for recalls that are simply do-overs hasn't been considered nearly carefully enough. No majority of California voters has expressly approved the "do over recall" theory. Instead, as argued above, so far there are only two agreed-upon grounds for recall: Postelection corruption, and postelection agenda change. Postelection dissatisfaction, up until now, has not been a ground.

Should the Grounds for Recall Be Expanded?

But should it be? Perhaps. There may be good arguments for adding a "do-over" category for recalls. And perhaps such arguments are especially strong when, as here, the victor at the last election unduly influenced the choices put before the voters by interfering in an unprecedented way with the other party's primary election process. (Gray Davis did this by spending millions of dollars to ensure that Simon would prevail over fellow Republican Richard Riordan, whom Davis feared more in a general election.)

Viewed in those terms, the recall of Davis does not commit us to a broad "do-over" category, but rather a narrow one, focused on campaign misdeeds. This narrow understanding would not necessarily open the door to so-called "sore loser" recall campaigns that might be harmful to the long-term stability of government.

But that discussion - focused on why a do-over in Davis' case was particularly warranted and focused on the long-term consequences of creating new categories for recalls - was one that we as Californians should have been having but did not.

Indeed, as I noted earlier, there was very little sustained discussion about the fundamental aspects of ballot Question 1 at all - what a recall is for, and whether these were appropriate conditions for one. With respect to Question 1, debate centered overwhelmingly not on these basics, but instead on a comparison of Gray Davis to the others vying for the governorship. In other words, most of the California press covered the campaign as if it were supposed to resolve only the single question: which of the four major candidates should be governor?