By Vikram David Amar
Published August 8th 2003 in Findlaw.com
On August 4, Governor Gray Davis filed a lawsuit involving the
California recall. In this lawsuit he argues, among other things, that
should a majority of voters opt to recall him, the federal Constitution
still entitles him to be a candidate on the second part of the ballot -
the part that asks voters to fill the recall-created opening.
The Governor's reasoning is simple: Suppose that, say, on the first part of the ballot 51% vote for his recall, and 49% oppose it. And suppose also that no candidate to be Davis' successor on the second part of the ballot gets 49% support. Davis contends that democracy requires that he should still remain Governor. After all, more voters prefer his remaining in office than any other choice.
As I will explain, this simple reasoning is also deeply flawed, and unlikely to prevail in court (for a number of reasons, including perhaps a lack of standing by the plaintiffs). But Davis, at least, has hit a nerve. It is reasonable to worry that a winner could get much less than 49% support - and not only for the obvious reason that the field is large, and wide open.
Is there any way to combat this worry and still have a one-day recall vote? I will show that indeed, there is - though unfortunately, it could be implemented only in the future, not in time for the Davis recall/successor election.
The Problem with Governor Davis' Argument
Davis's legal argument on the merits, while intuitively attractive, has some huge flaws. To begin with, although Davis may be right that he alone has to get 50% of the vote to be Governor whereas other aspirants (like Arnold) could win with, say, 30%, on the second part of the ballot, it is also true that Davis could win the recall with 51% of the vote on the first part even if 80% of the electorate would prefer Arnold in a head-to-head competition between the two. That is, just as Davis is excluded from the second part of the ballot, all others (including Arnold) are excluded from the first part. The two parts of the ballot effectively involve two distinct questions, so they are governed by different rules.
More generally, and more importantly, the Davis position disregards the fact that states exclude people who might be the most popular candidate from running, and winning, all the time.
Term limits, for instance, can have the effect of preventing someone (a popular incumbent) who might garner more votes than anyone else from running. So too do age and residency requirements.
So long as the candidate qualification criteria a state chooses are non-partisan and reasonable, they should be - and are - upheld in court. Here, California's interest in excluding Davis from the second part of the ballot is simple: It's telling someone who was so contentious that he just got recalled from the State's highest and most visible office that he now has to sit temporarily on the sideline, presumably to promote harmony and stability for the State, and to ensure the State survives whatever crisis brought on the recall intact. It may be true that should Davis lose on the first part of the ballot, he is being treated differently than are other candidates for the governorship, but that is because he is different - he is the only person who just got recalled.
In my view, the interest in keeping a recalled person on the sideline for a moment is actually a stronger interest than the interests in new blood that justifies absolute term limits (which have been upheld as constitutional by the state and federal courts in California). For this reason, Davis' suit should not succeed.
The Fear that the Recall Winner Won't Be a Popular Choice
But even if Davis will lose in court, his suit touches on a real fear: The fear that the recall election will yield a winner with the support of only a minority - perhaps a small minority - of California voters.
As of now (subject, of course, to yet another pending legal challenge), all one needs to do to have his/her name on the recall ballot as a possible successor is to come up with 65 signatures and $3,500. So far, over 300 persons have taken out the paperwork to try to do so. Tomorrow, August 9, is the final day for successor candidates to qualify. In the next 24 hours, scores of people may do so.
That has generated a number of dilemmas. For instance, over the last month, Democrats have faced a quandary. They say their preferred outcome is that Davis himself remain in office. But that puts them in a hard spot: Should they put a competing Democrat on the successor ballot, or not? Doing so might increase the chances of Davis' recall. (As of this posting, a few Democrats have indicated they will break ranks and offer themselves as successor candidates.)
Of course, this dilemma for the Democrats is in part a consequence of the fact that the recall/succession voting occurs on a single ballot on a single day. If there were separate elections, the problem would be obviated: voters (and Democratic party leaders) could make recall, and successor, decisions separately. But somewhat mystifyingly, the California Constitution does not allow that option: It prescribes a single-day recall-and-possible-replacement process.
That's mystifying mostly because this isn't a primary - it's the general election. Primaries are often won by a plurality winner, and sometimes even a low-percentage plurality winner. But in the general election that follows, the winner, in our basically two-party system, generally receives a majority, or at least something close to it. That's not likely to be the case in this recall.
As a result, a nagging fear plagues not just Democrats, but many Californians of every political stripe: What if the winning successor candidate receives only a small percentage of the vote?
This is very possible, for not only is this a general election, but there is no runoff. The California Constitution clearly says that "the candidate who receives a plurality [that is, the most votes, no matter how few] is the successor." (Emphasis added.)
The winner, thus, may only garner a relatively small percentage of the overall vote. In a field of 100 candidates, even a person who gets only about 15% of the vote may stand a decent chance.
And worse still, that winner's percentage of the recall-successor vote might not reflect his or her actual support among the State's population -- which might be much lower -- because most people don't vote. A candidate with narrow but deep support could win, especially if there is low voter turnout.
Moreover, a plurality system doesn't take account of voters' dislike for a particular candidate. Thus, the California recall winner not only may enjoy only narrow support, but also may well be reviled by a supermajority of voters.
Put another way, a "fringe" candidate could win the recall-successor election - leaving a huge majority of California voters both effectively disenfranchised, and deeply dismayed.
In the Future, Single Transferable Voting Should Be Used To Allow Instant Runoff
Is there any solution to this mess, short of amending the California Constitution to break up the recall process into two separate elections on two separate days? Fortunately, the answer is a loud yes - though it isn't an answer that will be able to help with the current, troubled recall.
The solution is Single Transferable Voting (STV), or an "Instant Runoff." My brother, Akhil Reed Amar, and I have explained the system in an earlier column.
Here's how it works: When a voter casts her ballot, she lists all her candidate preferences: First choice, second choice, third choice, and so on. She could still just list her first choice; she's not forced to rank everyone. But if she wanted to, she could express all the relative preferences she has.
How does this help? Consider a simple illustration. Suppose that 4 candidates are running in an election in which 100 persons vote. A gets 31 votes, followed by B with 30 votes, C with 29 votes, and D with 10 votes. The plurality winner obviously is A. But A wouldn't necessarily be the winner in an STV system.
STV repeatedly drops the lowest vote getter, but doesn't exclude those who voted for that person. Instead, it transfers their votes to their second choice.
So suppose, in our example, that everyone who voted for D ranked C as their second choice. D (the lowest vote-getter) drops out. But his ten votes go to C. Now our line-up looks like this: C has 39 votes; A has 31 votes; B has 30 votes.
Again, the lowest vote getter drops out - this time, it's B. Now B's 30 votes are assigned to those voters' second choices. Suppose 20 named C as a second choice, and 10 listed A. That gives us our new line-up: C has 59 votes (39 plus 20 second choices), A has 41 (31 plus 10 second choices). And C - not A - is the winner.
That makes sense, for A is considerably less popular than C, when voters' second choice preferences are taken into account.
STV can be used for virtually any election, but let's focus on governorships for a moment. In 1998 Jesse Ventura, won the Minnesota governorship with a bare plurality in a three-way race--37% compared to 34% for Republican Norm Coleman and 28% for Democrat Skip Humphrey. But under STV, in comparison, Ventura would have won only if a very good chunk--almost half--of the Humphrey voters who put down a second choice picked him and not Coleman. As a result, 63% of the electorate would not have had their preferences totally ignored.
STV itself is not without problems. It might promote fringe candidates to run in the first place, because they might get more initial first-choice votes if supporters knew that their votes would not be wholly wasted, but quickly reassigned to more major candidates. But at the end of the day, STV would generally make it harder for a fringe candidate to win with a simple plurality in a three-way race, like the one that elected Ventura, or a 30-way race, as is likely in California.
Elections are the key moments in a democracy when the people themselves speak, and there is much to be said for allowing them to speak clearly and communicate their nuanced views - all their preferences, not just their first choice. STV may be a bit complicated, but voters can handle it. After all, they're handling the two current two-part California recall ballot already.
Moreover, the idea of second choice is a simple, familiar one: If Ruffles are sold out, get Pringles. If Americans can handle this level of complexity as shoppers, why not as voters? We're not talking about a butterfly ballot here. This system is intuitive enough that, with some voter education, it could be used even in a high school student council election.
Would STV have helped Davis himself - assuming that he, miraculously, could prevail on his legal claim to be included as a successor candidate? The answer is almost certainly no.
If 51% of the voters voted to recall Davis, they'd probably rank him low as a successor choice as well. But they also might prefer him strongly to a "fringe" candidate - and an STV ballot would let them register that preference.
Actually, if STV were in the scheme, Davis might be more likely to be recalled in the first place, because voters who preferred to recall him - with all their preferences taken into account - would no longer have to fear ending up with a "fringe" candidate whom they like much less.
Ironically then, what is good about Davis' lawsuit - the idea that democracy should mean that voters' preferences are taken into account insofar as possible - wouldn't be good for Davis. A truly preference-favoring system like STV might only make his plight worse.
The Hope for STV in the Future: San Francisco's Attempt to Use the System
If this recall election shapes up to be the debacle some predict, California may be spurred to use STV the next time. There will doubtless be resistance: Last week, the State put the kibosh on San Francisco's efforts to move to STV in its own municipal elections. But that resistance shouldn't be the last word - STV, in the end, is the most deeply and truly democratic solution there is.
Voters in San Francisco approved the STV device as a local experiment 17 months ago. But state officials say their plans are still too complicated and logistically expensive to be implemented in time for local elections next spring.
State election officials may be right that STV can be expensive and somewhat foreign right now. But as the recall melee illustrates, so are the alternatives. And STV's democracy-promoting virtues far outweigh the modest burdens of its slightly greater complexity.
The Governor's reasoning is simple: Suppose that, say, on the first part of the ballot 51% vote for his recall, and 49% oppose it. And suppose also that no candidate to be Davis' successor on the second part of the ballot gets 49% support. Davis contends that democracy requires that he should still remain Governor. After all, more voters prefer his remaining in office than any other choice.
As I will explain, this simple reasoning is also deeply flawed, and unlikely to prevail in court (for a number of reasons, including perhaps a lack of standing by the plaintiffs). But Davis, at least, has hit a nerve. It is reasonable to worry that a winner could get much less than 49% support - and not only for the obvious reason that the field is large, and wide open.
Is there any way to combat this worry and still have a one-day recall vote? I will show that indeed, there is - though unfortunately, it could be implemented only in the future, not in time for the Davis recall/successor election.
The Problem with Governor Davis' Argument
Davis's legal argument on the merits, while intuitively attractive, has some huge flaws. To begin with, although Davis may be right that he alone has to get 50% of the vote to be Governor whereas other aspirants (like Arnold) could win with, say, 30%, on the second part of the ballot, it is also true that Davis could win the recall with 51% of the vote on the first part even if 80% of the electorate would prefer Arnold in a head-to-head competition between the two. That is, just as Davis is excluded from the second part of the ballot, all others (including Arnold) are excluded from the first part. The two parts of the ballot effectively involve two distinct questions, so they are governed by different rules.
More generally, and more importantly, the Davis position disregards the fact that states exclude people who might be the most popular candidate from running, and winning, all the time.
Term limits, for instance, can have the effect of preventing someone (a popular incumbent) who might garner more votes than anyone else from running. So too do age and residency requirements.
So long as the candidate qualification criteria a state chooses are non-partisan and reasonable, they should be - and are - upheld in court. Here, California's interest in excluding Davis from the second part of the ballot is simple: It's telling someone who was so contentious that he just got recalled from the State's highest and most visible office that he now has to sit temporarily on the sideline, presumably to promote harmony and stability for the State, and to ensure the State survives whatever crisis brought on the recall intact. It may be true that should Davis lose on the first part of the ballot, he is being treated differently than are other candidates for the governorship, but that is because he is different - he is the only person who just got recalled.
In my view, the interest in keeping a recalled person on the sideline for a moment is actually a stronger interest than the interests in new blood that justifies absolute term limits (which have been upheld as constitutional by the state and federal courts in California). For this reason, Davis' suit should not succeed.
The Fear that the Recall Winner Won't Be a Popular Choice
But even if Davis will lose in court, his suit touches on a real fear: The fear that the recall election will yield a winner with the support of only a minority - perhaps a small minority - of California voters.
As of now (subject, of course, to yet another pending legal challenge), all one needs to do to have his/her name on the recall ballot as a possible successor is to come up with 65 signatures and $3,500. So far, over 300 persons have taken out the paperwork to try to do so. Tomorrow, August 9, is the final day for successor candidates to qualify. In the next 24 hours, scores of people may do so.
That has generated a number of dilemmas. For instance, over the last month, Democrats have faced a quandary. They say their preferred outcome is that Davis himself remain in office. But that puts them in a hard spot: Should they put a competing Democrat on the successor ballot, or not? Doing so might increase the chances of Davis' recall. (As of this posting, a few Democrats have indicated they will break ranks and offer themselves as successor candidates.)
Of course, this dilemma for the Democrats is in part a consequence of the fact that the recall/succession voting occurs on a single ballot on a single day. If there were separate elections, the problem would be obviated: voters (and Democratic party leaders) could make recall, and successor, decisions separately. But somewhat mystifyingly, the California Constitution does not allow that option: It prescribes a single-day recall-and-possible-replacement process.
That's mystifying mostly because this isn't a primary - it's the general election. Primaries are often won by a plurality winner, and sometimes even a low-percentage plurality winner. But in the general election that follows, the winner, in our basically two-party system, generally receives a majority, or at least something close to it. That's not likely to be the case in this recall.
As a result, a nagging fear plagues not just Democrats, but many Californians of every political stripe: What if the winning successor candidate receives only a small percentage of the vote?
This is very possible, for not only is this a general election, but there is no runoff. The California Constitution clearly says that "the candidate who receives a plurality [that is, the most votes, no matter how few] is the successor." (Emphasis added.)
The winner, thus, may only garner a relatively small percentage of the overall vote. In a field of 100 candidates, even a person who gets only about 15% of the vote may stand a decent chance.
And worse still, that winner's percentage of the recall-successor vote might not reflect his or her actual support among the State's population -- which might be much lower -- because most people don't vote. A candidate with narrow but deep support could win, especially if there is low voter turnout.
Moreover, a plurality system doesn't take account of voters' dislike for a particular candidate. Thus, the California recall winner not only may enjoy only narrow support, but also may well be reviled by a supermajority of voters.
Put another way, a "fringe" candidate could win the recall-successor election - leaving a huge majority of California voters both effectively disenfranchised, and deeply dismayed.
In the Future, Single Transferable Voting Should Be Used To Allow Instant Runoff
Is there any solution to this mess, short of amending the California Constitution to break up the recall process into two separate elections on two separate days? Fortunately, the answer is a loud yes - though it isn't an answer that will be able to help with the current, troubled recall.
The solution is Single Transferable Voting (STV), or an "Instant Runoff." My brother, Akhil Reed Amar, and I have explained the system in an earlier column.
Here's how it works: When a voter casts her ballot, she lists all her candidate preferences: First choice, second choice, third choice, and so on. She could still just list her first choice; she's not forced to rank everyone. But if she wanted to, she could express all the relative preferences she has.
How does this help? Consider a simple illustration. Suppose that 4 candidates are running in an election in which 100 persons vote. A gets 31 votes, followed by B with 30 votes, C with 29 votes, and D with 10 votes. The plurality winner obviously is A. But A wouldn't necessarily be the winner in an STV system.
STV repeatedly drops the lowest vote getter, but doesn't exclude those who voted for that person. Instead, it transfers their votes to their second choice.
So suppose, in our example, that everyone who voted for D ranked C as their second choice. D (the lowest vote-getter) drops out. But his ten votes go to C. Now our line-up looks like this: C has 39 votes; A has 31 votes; B has 30 votes.
Again, the lowest vote getter drops out - this time, it's B. Now B's 30 votes are assigned to those voters' second choices. Suppose 20 named C as a second choice, and 10 listed A. That gives us our new line-up: C has 59 votes (39 plus 20 second choices), A has 41 (31 plus 10 second choices). And C - not A - is the winner.
That makes sense, for A is considerably less popular than C, when voters' second choice preferences are taken into account.
STV can be used for virtually any election, but let's focus on governorships for a moment. In 1998 Jesse Ventura, won the Minnesota governorship with a bare plurality in a three-way race--37% compared to 34% for Republican Norm Coleman and 28% for Democrat Skip Humphrey. But under STV, in comparison, Ventura would have won only if a very good chunk--almost half--of the Humphrey voters who put down a second choice picked him and not Coleman. As a result, 63% of the electorate would not have had their preferences totally ignored.
STV itself is not without problems. It might promote fringe candidates to run in the first place, because they might get more initial first-choice votes if supporters knew that their votes would not be wholly wasted, but quickly reassigned to more major candidates. But at the end of the day, STV would generally make it harder for a fringe candidate to win with a simple plurality in a three-way race, like the one that elected Ventura, or a 30-way race, as is likely in California.
Elections are the key moments in a democracy when the people themselves speak, and there is much to be said for allowing them to speak clearly and communicate their nuanced views - all their preferences, not just their first choice. STV may be a bit complicated, but voters can handle it. After all, they're handling the two current two-part California recall ballot already.
Moreover, the idea of second choice is a simple, familiar one: If Ruffles are sold out, get Pringles. If Americans can handle this level of complexity as shoppers, why not as voters? We're not talking about a butterfly ballot here. This system is intuitive enough that, with some voter education, it could be used even in a high school student council election.
Would STV have helped Davis himself - assuming that he, miraculously, could prevail on his legal claim to be included as a successor candidate? The answer is almost certainly no.
If 51% of the voters voted to recall Davis, they'd probably rank him low as a successor choice as well. But they also might prefer him strongly to a "fringe" candidate - and an STV ballot would let them register that preference.
Actually, if STV were in the scheme, Davis might be more likely to be recalled in the first place, because voters who preferred to recall him - with all their preferences taken into account - would no longer have to fear ending up with a "fringe" candidate whom they like much less.
Ironically then, what is good about Davis' lawsuit - the idea that democracy should mean that voters' preferences are taken into account insofar as possible - wouldn't be good for Davis. A truly preference-favoring system like STV might only make his plight worse.
The Hope for STV in the Future: San Francisco's Attempt to Use the System
If this recall election shapes up to be the debacle some predict, California may be spurred to use STV the next time. There will doubtless be resistance: Last week, the State put the kibosh on San Francisco's efforts to move to STV in its own municipal elections. But that resistance shouldn't be the last word - STV, in the end, is the most deeply and truly democratic solution there is.
Voters in San Francisco approved the STV device as a local experiment 17 months ago. But state officials say their plans are still too complicated and logistically expensive to be implemented in time for local elections next spring.
State election officials may be right that STV can be expensive and somewhat foreign right now. But as the recall melee illustrates, so are the alternatives. And STV's democracy-promoting virtues far outweigh the modest burdens of its slightly greater complexity.
