As our op-ed in today’s Recorder charges, Alameda County
is playing with fire in the way it is attempting to change the county’s recall
law.
The wholesale adoption of the controversial 2022 law used by the state for general cities and counties may be warranted (matter for
another debate), but the fact that this is taking place at the exact moment
that the county is facing what may be the most significant recall effort in its
history against District Attorney Pamela Price means that voters will be suspicious
of the intent behind these changes. This is a significant problem, as we
discuss in this op-ed, as the voters are very happy to shoot down ballot
measures. In this case, by not choosing a limited method to handle the recall
changes, the county may be facing chaos when signatures are handed in. The
county has also not been clear on how what will happen with the existing recall
campaign (possibly the first real attempt since 1915), as the Warren Zevon rule for recalls means it is likely to be enmeshed in costs and increasingly complex litigation.
The op-ed deals with a few of the standout reasons for
concern. But I want to take a dive through the memo prepared by the County
Counsel and also touch on the FAQ page that they have apparently posted. The
memo seems to hide the ball on what is being proposed, which may make a ballot
measure failure that much more likely. Let’s jump right in:
Appointive
(or Appointed) Officers: Why is this a focus?
The first thing to notice is the focus on “appointive” or
“appointed” officers. The memo italicizes the word and it is the first actual
part of the law mentioned. They are correct that it is unusual to have the
recall law applied to non-elected appointive or appointed officers (think
police chief, City Manager, Health Director). But that is only part of the story,
and the reality makes the focus on appointive officers a real red-flag.
First, some history -- the first place to adopt a recall
law against appointive officers was Alameda – the City, not the County – in 1906
(according to Bird & Ryan’s seminal The Recall of Public Officers, who called this a
“revolutionary innovation.”). Other jurisdictions have the same type of
provision on the books, including San Francisco. But it is not a regular
feature of the recall. Among the 20 statewide recall laws, only Montana allows
appointed official recalls, though it has strong limits.
Looking at the 1911 legislative discussion, it makes
sense that California may not have wanted to wade into the fight. The big debate
at the time was over judges, so it could be that they did not want to expand
the fight to another controversial area. I don’t remember seeing it discussed
during the legislative debates.
From a theoretical point of view, recalls against
appointed officials may be unpopular because the targeted position is not
supposed to be subject to public approval (hence no direct vote) and, more
importantly, there is a simpler method of removal, namely that the appointing
body has the removal power. Unlike Impeachment, which has arguably failed as a
deterrent (at least on the national level), removal is simple and happens with
regularity. So a recall could be either overkill or a step too far for the process.
But the question here is why should Alameda County focus
on this provision now? Bird & Ryan note a City Manager recall in Long Beach
in 1922. I don’t know of any other ones anywhere. I don’t even know of a significant
threat of one. And not just in California – anywhere in the US. So highlighting
this point of the law that has never been used in Alameda and rarely used
anywhere else seems like a strange subject to focus on when vastly more
important questions are at hand.
Unconstitutional
Provisions:
The memo then mentions two unconstitutional provisions of
the Alameda Charter. Section 62 holds that petition circulators have to be
registered voters in the County. It also holds that voters have to vote for the
recall in order to cast a ballot in a replacement race. As noted, these are
both inoperative due to court rulings. It may be nice to clean up these
provisions, but it is certainly not necessary. Unlike many other lawsuits,
anyone suing to enforce these provisions will find themselves laughed out of
court.
Missing
Procedures:
The next section focuses on the pre-circulation
procedures. The memo notes that procedures are important for transparency
purposes. “They provide notice to the public, identify the target of the
recall, allow the target to answer the petition, give the elections official
the opportunity to vet the form of the petition and allow the public time to
challenge the form and content of the petition.”
Pre-circulation procedures were not used in the first
recall laws in cities throughout the state (Bird & Ryan call this “the
delightful informality with which anyone could start circulating a petition”)
and were added in first by Oakland. Berkeley had previously put forth the
innovation that petitioners should state, in 200 words, a reason for the recall
Since then, these procedures have become standard in practically all recalls in
the US, though frankly it is hard to say how much they are really needed. A
good part of the procedures are hoop-jumping that delays, adds costs and frustrates
the petition gatherers (Michigan is a prime example). As you’ll see below, some
of these provisions feel clearly designed for barrier-creating purposes. For
example, is spending money to publish notice in a newspaper needed these days?
The reality is that publication, alerting the officials
and other parts of the law are an added expense. It certainly seems that the
officials find out about the recall effort, and as the Price recall effort
shows, the de facto notification process seems to have been very successful. Having
these pre-signature gathering procedures may solve some problems for
petitioners, who are prevented from going out with deficient petitions. But notification
is not a huge problem when you are collecting the massive amount of signatures
needed for a county-level recall. Any change in the notification process,
especially after the petitioners have gathered over 70,000 (unverified)
signatures can wait until a less fraught time.
Unfeasibility
– the heart of the matter starts on page 4:
It is only on page four that we get to the heart of the
matter, and why the law needs to changed, unfeasibility. Our op-ed deals with
this at length, but the key parts are:
1) There
is no provision explaining how to replace the official if they are removed.
California voters probably expect the one-day/two-step process that they used
in the recall of Governor Newsom. This means, voters get to choose the
replacement candidate. But, as mentioned, this provision would be eliminated. Instead,
Voters will quickly discover that the state law for general officials uses an automatic replacement or by law model,
which most likely will result in an appointed replacement rather than an
elected one (admittedly, most states use this model). Not only is this not
explained in the memo, the FAQ (designed specifically to clear up issues),
avoids the issue and simply says that we should refer to Sections 8, 20 and 33
rather explain what those sections do (get rid of replacement elections).
For a little history here, while the
first recall laws just used a new election (like the type used in Wisconsin or
the UK), Bird & Ryan claim that in 1910 no-so-sweet Modesto seems to have
adopted the more familiar one-day/two-step process that California voters are familiar
with. There was also differing views on whether the removed official should be
allowed to run in the replacement race.
In the end, voters may not be happy to
discover that they will not be able to select a replacement – this massive
change alone is likely to jeopardize the ballot measure.
2) The
signature verification process must be completed in 10 days rather than the 30
days that the state allots. Additionally, there is no allowance to use the
statistical sampling model (described here – note that it is not allowed for
state level recalls, simply for local ones. This discrepancy was changed in
2017 to make state recall efforts take longer).
This short time frame may turn out to be a significant problem and once again
the Supervisors really need to deal with it.
3) Change
the time frame to actually hold the election. The county requires an election be
held within 35-40 days. The state gives a more leisurely 88 day to 180 day limit.
The extra days potentially allows the recall to be tied into a general election
(in this case, possibly the presidential election) in order to save money. Legitimate
issue here, though the elongated time frame may give voters a good complaint
about overlong delays.
Two practical points to
consider on the longer term political ramifications of these changes, both of
which may be somewhat surprising.
1) Using
an automatic replacement model may seem to be a way to cut down on recalls, as
the personal benefit of replacing the official may not be there. However, this
does not appear to be the case. Oregon uses the automatic replacement model and
it has the same number of recalls as California.
2) The
focus on tying a recall to a general election may seem like a good way to
improve the odds for an elected official. This was definitely the motivation in
2017, when they made some changes in the law to try and unsuccessfully avoid a
recall of State Senator Josh Newman. Theoretically, it makes sense. You would
think (as I did) that the official should be more likely to lose a recall on a
special election date – when the most motivated voters will come out – rather than
a general election date. However, this is incorrect. My research shows that
voters are slightly more likely to kick out an official on a general electionday than on a special election.
Key
provisions not discussed in the Memo:
The memo leaves out provisions of note, one of which will
likely further inflame voters against the ballot measure.
1) Signature Removal or Signature Strikes
Laws:
Adopting the state law would grant voters a signature strike law, allowing
petition signers to remove their names from the petition in a counter-effort by
the targeted official. Nothing wrong with this one, and, unlike the state-level
recalls it doesn’t have a supplement slowing down period. This was adopted in
2017 and we have already seen it work once, so something to keep in mind for
the Price and other recalls.
2) Funding Disclosure: The
state has a very detailed and quite onerous funder disclosure requirement. Petitioners
need to be show signers a document that explains the big financial backers of
the petition, and this has to be written in 14 point font with a 16 point
headline (note that 8 and 11 point fonts are listed for other provisions). Again,
this really feels like a way to discourage signatures (in some states, they
have this printed on the petition, in what feels like a push to make it so that
fewer signatures are able to be signed on each page of the petition).
3) Supplemental Signature Gathering Period:
The new law would remove the 10 day period to gather additional signatures if
the signatures are handed in, but don’t meet the bar. California doesn’t have
this law, though other places have it (or have a cure period, where they can
repair signatures that were invalidated).
4) Signature threshold changes:
Most important point is right here, and left off the memo, though discussed in
the FAQ. The county calculates how many signatures are needed to get on the
ballot. The current law requires signatures of 15% of voter turnout for the
position in the last election. The state law will change that to 10% of
registered voters. As a practical matter this will raise the signature totals.
For a putative Price recall, it would go up from 73k to 93K – though presumably
the signatures would be handed in before then. But the change would almost
certainly result in a lawsuit.
There’s no problem with
increasing the signature requirement. 15% of turnout is lower than most places
that have recall laws (25% of turnout is one that we often see used), so that should
not in and of itself be controversial. The possible problem may be that the 10%
number is tied to registered voters. A little awkwardness here – for years, I
have said that using registered voters, rather than voter turnout, is the right
way to go. This way, the recall is not tied to a variable number such as a low
turnout in the last election. Theoretically sound. But in practice, I was
wrong. In both the Los Angeles D.A. recall and in the New Orleans MayoralRecall we saw complaints and lawsuits over the rolls – specifically that there
are too many voters who died or moved still listed as registered voters. As a
result, this artificially increases the number of signatures needed (leading to
lawsuits). California (unlike Louisiana) generally has a good system to check
rolls, but I think in the interest of avoiding trouble, it is probably better
to use the turnout in the last election numbers.
But the big problem here is
that the memo does not explain how the changed signature threshold and other
provisions will impact the Price recall. The FAQ is frankly unclear on this as
well, suggesting that that it may impact and completely overturn the Price
recall effort, which has already spent a significant effort collecting
signatures under the existing law. In addition to guaranteeing further
litigation, this change during an active recall is a major problem for
petitioners who have already collected signatures under the existing law.
As we note in the op-ed, it is imperative to update the
recall law in the face of a prominent recall. But this needs to be done with
care, and should not be seen as a moment to try and avoid future recall efforts
(something that the change will probably not impact anyway). By wholesale adopting
the state’s local recall law, with the two potential poison pills of stripping
voters of the power to replace candidates in an election and at the same time
increasing the signature numbers, voters may well reject this ballot measure
and chaos will ensue. The Supervisors should just deal with the absolutely
necessary provisions in the least controversial manner (perhaps selective
incorporation) and then revisit wide-spread changes at a later date when an
active recall is not taking place.