Poor recall

It is now all but certain that there will be a recall election later in 2021 against California Governor Gavin Newsom. I oppose recall elections in principle, but this one is especially silly and likely counterproductive for its own promoters.

Recall elections only exacerbate the worst features of the presidential (including gubernatorial) form of government, in that they increase the already inevitably high personalization of the political process of such a system. As if all of what makes for (in)effective government and policy-making can be put on one individual.

In the particular case of Newsom, it is silly in that the number one issue the state (and all governments) have had to face over the past 15 months is the pandemic, and California actually has fared better than other large US states. Is that entirely due to Newsom? No, see my comment about the risks of personalizing government. But he probably deserves some credit.

Its promoters are, of course, Republicans. The Republican Party is so completely out of step with this state that it needs internal reform to make itself competitive again. One might think this would be the lesson it would learn from having won vote percentages in the low thirties in the last two presidential elections and having elected no statewide official since 2006. Newsom himself was elected with 61.9% of the vote in 2018. In recent cycles, the Republicans have struggled to win even a third of the seats in either house of the state legislature.

However, rather than learn the lessons of its irrelevance in this state, the California Republican Party has learned a different lesson. While it may not win state power the normal way, it can harness grievance, the possible low turnout of a special election, and a celebrity to pick off a Democratic governor now and then. But this isn’t the California of 2003, and neither Caitlyn Jenner nor Randy Quaid, nor any of the others in the “clown car” of candidates looks ready to be the next Governator.

As an institutional device, recalls sit poorly with the separate origin (and, normally, survival) of executive and legislative organizations that defines presidentialism. If you need a mechanism to enable early elections, logically you should have a fresh mandate for the legislature, too. Better yet, use parliamentary government (and if you also use proportional representation, you have the greater likelihood that the resulting coalition governments dissolve midterm when political conditions change, and early elections result). In a presidential/gubernatorial system, they just reinforce the worst aspect of the system–their personalization of the executive, and their potential for “populist” solutions. And I say that as someone who thought Arnold Schwarzenegger was a pretty good governor (although I voted against the recall of then-Governor Gray Davis, I voted for the new governor’s reelection in 2006). Or, rather, he was pretty good only after a rocky first year, which only buttresses my point about personalization and populism. He tried to govern by shear force of personality (and he has an unusual measure of that!) and through popular initiatives, including calling a special election for some of them. When it did not work, he eventually learned how to be a governor. The state can’t afford on-the-job training and exercises in populism as it emerges from the pandemic. This specific recall is an even worse idea than the institution of recall is generally.

It is nearly sure that it will fail, at great expense. And it likely will only push Newsom and the Democratic Party father to the left and into ever-greater embrace of unions and other constituent groups, while making the Republican brand even more toxic in the state. Not that I care too much about the latter. The California Republican Party can go hang itself. And if you’d rather replace that verb with another more pointed one, feel free.

33 thoughts on “Poor recall

  1. On the point about the potential for low turnout, I suspect that is something the promoters of a recall count on. But if 2003 is any guide, they will not achieve that. Consider the following total votes cast in 2002-06:

    2002 (reelection of Davis): 7,738,821
    2003 (recall of Davis, election of Schwarzenegger): 9,413,491
    2006 (reelection of Schwarzenegger): 8,679,423

    With no candidate likely to emerge with Schwarzenegger’s star power, probably the turnout bump will not be so significant. But I would guess this will greatly energize Democrats, which is obviously not good for the hopes of a successful recall.

  2. On opposition in principle to recalls–maybe they are okay in local government, given lower information (exacerbated by nonpartisan contests, which we also should dispense with). But not for state (or national) offices.

  3. The California recall system is unusually evil. Only 2 states hold a simultaneous election for a successor where the recallee is ineligible to stand.

    That gives the defending party an impossible choice. They must either abandon and run a new candidate against their own incumbent, or effectively concede the office to the attacking party if the recall succeeds.

    In 2016 the mayor of Fall River, Massachusetts was recalled and re-elected on the same day.

    There is a history in California of Democratic legislators being recalled and replaced by Republicans at low turnout special elections, and then being re-elected at the general election.

    In 2016, Democratic state Sen. Josh Newman won a seat in a longtime GOP stronghold in north Orange County. In the summer of 2018, Republicans and anti-tax advocates mounted a recall campaign over Newman’s support for an increase in the state gas tax. A fired-up conservative bloc of voters turned out en masse, Democratic voters did not, and Newman was replaced by Republican Ling Ling Chang. 
    In that race, 66,197 voters, or 42%, opposed the recall and backed Newman. Chang, running against five other candidates, received fewer votes, 50,215, but still won the seat with 34%. In the next regularly scheduled election in 2020, Newman reclaimed the seat. 
    A sufficiently crowded and disorganized field could produce similar results in a 2021 recall race.

    https://calmatters.org/politics/2021/03/recall-rules-gavin-newsom-trouble/

    Only one California governor has ever been recalled. On the other hand no governor of California, and more importantly no US president, has ever been impeached and convicted. Flawed as the recall process is, the effective result of abolition would mean that, as with the presidency, there is effectively no way to hold a California governor to account for misconduct in office.

    A better solution would be to end the ban on the recalled officer running in the special election, and to hold the special election some time after the recall election. It’s also the twenty-first century, it’s not rocket science that election by plurality is a good idea.

    Recall of legislators is an equally bad idea. There may be a case for recall of legislative bodies as whole.

    • A recall (of a single office) is one car where Approval Voting would make sense.

      On Part A of the ballot, you tick yes or no on whether to retain the incumbent. (Maybe label the squares “KEEP” vs “REMOVE,” ie emulate UK 2016 on Brexit not Italy 1974 on divorce https://www.en.wikipedia.org/wiki/1974_Italian_divorce_referendum#Confusion_about_voting_methods).

      On Part B of the ballot, you can tick one or several (or all, or none) of the challenger candidates.

      Whichever of the options (challenger or incumbent) gets ticked by the most voters, but with two glosses:

      3.1 The incumbent is credited either with her actual number of votes, or with 25% of the number of enrolled/ eligible voters,* whichever is greater. (Ie, a minimum turnout and support quorum needed to oust an incumbent mid-term, ie effectively making them run twice due the same position)

      3.1 The incumbent is deemed re-elected, and every challenger losses, if the “REMOVE” votes are less than the incumbent’s deemed total. This would mean that if, say, the majority were to vote 52-48% to keep “Davis” it would be immaterial that, say, 57% of them had simultaneously ticked “Schwarzenegger.”

      In other words, it’s not simple Approval with an open field but a version tweaked to let the incumbent continue to enjoy the fruits of their regular-election victory unless both (a) a majority of those casting ballots on the question, being also (b) at least one-quarter of the total electorate, want her removed. Even then, she still gets a chance to stay in if (c) no rival candidate enjoys higher plurality support.

      No need for a second special election weeks later, you can still get it over and done in one polling day.

      So an incumbent can still be recalled if they are unusually unpopular (or some new rival is unusually popular and has only just emerged to throw hat in ring). But it should remove the incentive for a defeated minority party to give the electoral roulette wheel a second spin in the hope they might win this time via the vagaries of FPTP voting.

      Approval Voting is at root an attempt to extend the Yes/No logic of referenda on proposals to elections among candidates, in the hope that the binary ballot structure that ensures majority rule in referenda will ensure majority rule in elections as well.

      (NARRATOR: It doesn’t.)

      But Approval has more (at)traction in recall elections, which do (more closely) resemble referenda inasmuch both have a legitimate status quo option that should remain in place by default if no alternative can muster 50% of votes (plus some respectable percentage turnout).

      [* I’m not sure how to translate “percentage of enrolled/ eligible voters” in jurisdictions that have same-day registration or “motor voter” laws. Maybe, eg, “25% of all registered electors” could be replaced by, say, “20% of the total population of adult citizens who have not been disenfranchised by law as convicted felons”? You would still have adults who lose the franchise due to, eg, mental illness or dementia but even in Florida the number thereof must surely be too small to change the result.]

      • I am desperately trying not to mention The Thing That Should Not Appear In Every Single Thread, but it or a runoff would be a much cleaner solution for a same day election than Bonus Approval Voting. Bonuses are a bad idea in themselves and would be hell to explain to the electorate. So would d getting voters completely unfamiliar with Approval Voting to use it in recall elections but in no other election.

        Idaho has a Queen of the Hill rule. A recall needs a majority that is at least as many voters as voted for the incumbent at the last election.

      • Yeah, not a bad idea to have something like the “Queen of the Hill” rule. But even with IRO-AV for the special/ by-election following a recall, there should still be some kind of quorum. With IRO-AV it should apply once the preferential count is down to the last two [wo]men standing.
        The quorum could be framed more simply simply as “any challenger candidate supported by <50% of valid ballots and/or by <25% of all eligible voters is eliminated”, which may leave only the incumbent by default.
        Re: the other topic that Alan and I desperately avoid trying to drag into every thread unless it is absolutely absolutely essential to prove some point… UQ union council used to have a “15% of all eligible student voters must vote yes in the recall referendum” hurdle in the 1970s, but this had been repealed by the time I came on the scene. With turnout hovering around 20% this would have amounted to a de facto floating supermajority of votes.

      • From the site Alan linked to: “As opposed to the Queen of the Hill rule (used when several amendments are trying to impact a law, so the one with the most votes wins), a King of the Hill provision is when the last vote is the one that counts, no matter if an earlier one gets the most votes.”
        Okay, so “King” = basically the Swedish Riksdag method (where the Speaker decides the order in which motions are voted on, giving a slight advantage to the later movers if a Condorcet cycle exists: ie, if stone > scissors > paper > stone, paper wins if its pairwise contest against stone is held last) whereas “Queen” rules = basically the California referenda conflicts rule?
        The Swedish method could hold some attraction as a quick-and-dirty for clubs and associations etc if you first vote for one of the motions; if one gets 50% plus a quorum it’s adopted; otherwise, you then vote on them pairwise, starting with the pair that together have fewest votes and have not yet been voted on pairwise. This would give the last-mover advantage to the motions with more supporters if a Condorcet circular tie exists, but would not allow a minority to defeat a “beats-all” winner (if one exists) via a split vote.

      • The same day election is fundamentally unfair to the defending party, especially if the incumbent is barred from running. They must run an alternative candidate, effectively conceding the case for the recall, or run with the incumbent knowing they may be locked out of the election. And just quietly it is fundamentally unfair to the incumbent’s supporters to ask them to vote in an election not knowing if their incumbent has been recalled.

      • (1) How is a same-day election harsher on the incumbents than “run on two separate polling days if you want to retain a seat you won fair and square last November or the November before”?

        (2) It will also be noted that under versions like the model I proposed, the incumbent is not “barred from running”. It is basically Approval but with any challenger supported by less than 50% of valid votes or less than 25% of eligible voters being summarily eliminated. No room for any “Screws Cruz” tactics under this model that I can see.

      • Alan, if you mean the current California model as it exists, yes, it is poorly-conceived. But I don’t think those flaws apply to all same-day models. Can’t see why it’s unfair to ask supporters to “RETAIN GOV. DAVIS/ WALKER/ NEWSOM” before they know if he’s out or not. Is anyone going to say “I liked and supported the guy but, gee whiz, if 52.7% of votes went against him, maybe I should reconsider whether he’s a good leader or not”?

      • There have been cases in states where the incumbent is eligible for the special election, where officials have been recalled and re-elected on the same day. Evidently some votes say precisely what you insist they don’t.

      • Is there some logical contradiction between “Do I want Edwin Edwards to be Governor? Er, not really…” and “So it’s Edwards or David Duke? Okay, Edwards then”? And is it illogical to write rules that say “if the two questions give different answers, tie goes to the runner”?

  4. Playing devil’s advocate for a moment, here’s a counter-argument. In a parliamentary system, the legislature can remove the head of the executive branch (possibly with constraints about when and by what supermajority). In a presidential system without a recall provision, no one can remove the head of the executive branch until the next election, except by impeachment. As Alan just pointed out, impeachments are not exactly commonplace. I think this is an argument either for allowing recalls or for very short terms of office for the executive branch.

    But no quarrel with your analysis of the California Republican Party.

    • I thought MSS deserved high commendation for his restraint in describing the California Republicans.

    • The impeachment provisions in the California cosntiutiona re far more sensible than the federal provisions—no arcane language about high crimes and misdemeanours, and impeached officer is suspended until the trial ends—but impeachment is still a broken procedure.

      Article IV Section 18:
      (a) The Assembly has the sole power of impeachment. Impeachments shall be tried by the Senate. A person may not be convicted unless, by rollcall vote entered in the journal, two thirds of the membership of the Senate concurs.

      (b) State officers elected on a statewide basis, members of the State Board of Equalization, and judges of state courts are subject to impeachment for misconduct in office. Judgment may extend only to removal from office and disqualification to hold any office under the State, but the person convicted or acquitted remains subject to criminal punishment according to law.

      Article V Section 10:
      […]The Lieutenant Governor shall act as Governor during the impeachment, absence from the State, or other temporary disability of the Governor or of a Governor-elect who fails to take office.[…]

      • I doubt the wingnuts who deny that Trump’s many wrongdoings count as “high crimes and misdemeanors” would have much trouble equally denying that they were “misconduct” and dismissing the charges as politically motivated.

      • I know hindsight is silly, but including a second charge of Trump’s failure to act to defend the congress would have made a more difficult argument for the Senate Republicans to handwave.

      • What happens in Nebraska with a unicameral legislature? How does impeachment work in presidential countries with one chamber? Are impeachment procedures less complex with one chamber vs two?

      • In Nebraska the supreme court sits as the court of impeachment except when a justice is impeached. Justices are tried by 7 lower court judges elected randomly.

        In Costa Rica the trial is before the supreme court.

        In Ghana, 1/3 of the national assembly ask the chief justice and the 4 most senior justices to investigate if there is a prima facie case. If the justices find there is a case the national assembly can remove by 2/3 vote.

    • Yeah, in a similar vein to what Bob’s saying, Ganghof presents a convincing argument in favour of recall for presidents. He certainly doesn’t argue for presidentialism, but he suggests that if you have it, it’s logical that you should be able to remove the chief executive through recall, because the president should be responsible to the voters in the same way that PMs are to the assembly.

  5. The recall system is clearly a joke. However I cannot say that I am opposed in principle to the ability of the voters to remove an otherwise effectively unremovable and unchecked chief executive. Parliamentarianism is the obvious answer, but in its absence, a sensibly designed recall would have its advantages.

    • During the Zuma years in South Africa repeated attempts to force him from office failed. so we should remember that there’s a romantic view of parliamentarism that assumes that parliamentary heads of government are easy to dislodge. There was considerable support for making the presidency elective on the theory that Zuma was unelectable. There was even one proposal that the president be recallable, forcing the ruling party to name someone else after a successful recall.

      • Could a President and/or Governor submit to voters vote of no confidence vote? That way they could preempt recall provisions. The question could be asked, should the president/governor be dismissed from this office, with a yes or no vote?

        If there is a successor elected at the same time, the president/governor could be dismissed by the first provision, but continue in office by having the most votes and/or majority by the second provision.

    • Yes, “sensibly designed” is the key. Some of the initial comments make some suggestions along those lines. I could consider bending my principle if one of these were used instead of the terrible California procedure.

  6. The form of [redacted] voting I admire is not the YES/NO variant, for the simple reason that all candidates could end up with negative scores, as part of the downward spiral of cynicism in politics. I’d prefer voters to be able to put a X next to as many or as few candidates as they would find acceptable as their MP for the subsequent term.

    I’m not a fan of recall. Of course, it’s been introduced in a way even in a parliamentary system in the UK in recent years. The first attempt at a recall petition was unsuccessful, the second triggered a by-election in which the incumbent followed advice not to re-offer as an independent candidate, and the third caused another in which the incumbent was allowed to run again under his party label and did so.

    The two elections that took place saw a Labour hold and a Lib Dem gain respectively, but both seats returned to the Conservatives at the subsequent general election later that same year. So politically it looks like the reverse effect from the one you’d expect from the Californian experience. Although if it was all part of a Tory ‘long game’ then it was a cleverer and less transparent one than that of the Cali GOP.

    • Olivier, you could incorporate that as well. “A non-incumbent must be approved by more than half the valid ballots (these being defined as those that contain a tick for any one or more candidates, or you could even include blank ballots) to escape elimination”.
      My main point is, all the candidates should go on the same menu but it is legitimate to spot the incumbent a few yards in terms of turnout if they are still serving their original term

      • Yes, that would have merit. Donald Horne (author of THE LUCKY COUNTRY) proposed in a 1975 tract that instead of a Governor-General sacking an unpopular government, the parliament be dissolved if 40% petition for it. Horne knew perfectly well that Kerr didn’t sack Whitlam because the government was unpopular – a popular government would still need to go to an early election if an upper house was obdurate in blocking some essential bill, as in 1974 or the UK in 1911, while an unpopular government can soldier on if there is no opposition upper house majority willing and constitutionally empowered to block supply, like Qld in 1988 or NSW in 2010-11 – but he wanted to offer an alternative to those who thought the electoral ends justified the legal means in November 1975.
        I am conflicted over the precise threshold of signatures that should be required to dissolve the entire legislature (or at least its lower house, or each elective chamber thereof). 40% seems the absolute minimum, since with the Internet and a very polarised electorate, the losing side could probably attain that within a year or two, rather than putting its energies into organising Tea Party rallies or a “Convoy Of No Confidence”. So 50% seems to be the right number, symbolically and practically.
        But verification of the signatures would be a nightmare, since the result would not just be “one more proposition gets put to the voters on the November ballot-paper” or even “a single official has to re-apply for their job” but “all (elected/ lower house) legislators will be sacked en masse and required to re-apply for their current positions”. I hear disturbing reports about how widespread fraud is in the US initiative petition process (usually, interestingly, from the same people who assure me that fraud cannot possibly be widespread in the US voting process because “There are heavy fines if you get caught!”) and would not envy an Electoral Commission official who draws the straw of deciding whether the 5,000,001st signature on a “Recall Parliament” petition is valid, or of having to phone that purported voter to verify, or indeed of being any of the signatories getting doxxed by angry supporters of the incumbent majority.
        It’s true that all of the above can happen with an initiative or even a recall petition, but then the stakes are slightly lower because the result is to send the decision (back) to the voters and the side that last won, hasn’t had that victory taken away from the yet. Whereas if the current legislature is dissolved (assuming we just go straight to another election and don’t bother with a separate “Do you wish the legislature to face a fresh election, yes or no?” referendum question either two weeks before or on the same day), then the opposition party(ies) have just scored a big, palpable hit on those who defeated them a year or two earlier.

  7. EDD is such a fiasco, impossible to call anyone to get help, one has to pay Claimyr to get a hold of a representative, then the representative drops the calls, nothing gets solved.

  8. Speaking of Approval Voting…
    “… Part of what made her victory so historic was the new method of voting used by the city, known as approval voting. Approval voting is a non-partisan way to give voters the chance to vote for as many candidates as they’d like, making the single winner the most approved of candidate.
    Approval voting is seen by many as a better alternative to the more traditional first-past-the-post system, where voters choose only one candidate and the candidate with the most votes wins, regardless of whether they receive a majority of votes. Approval voting, therefore, opens room for third party challenges and also prevents vote splitting, where one candidate will win a majority of votes in their base areas but will receive little to no support in other parts of the city.
    This kind of vote split was seen in the 2017 mayoral race, where Lyda Krewson was narrowly able to clinch the votes, beating Tishaura Jones. Before approval voting went into effect, Krewson won her race by just 888 votes, though that mainly consisted of just the Southwest corridor of the city. Jones won citywide this year, including north of Delmar, despite their voter turnout being substantially lower…”
    Zoë Butler, “Tishaura Jones Makes History: After a close-fought election, Tishaura Jones was sworn-in as St. Louis’ 47th mayor and the City’s first Black female mayor on April 20,” The University News (1 May 2021), https://unewsonline.com/2021/05/tishaura-jones-makes-history/

  9. As an old loyalist of progressive populists like Oregon’s W.S. U’ren, who gave that state the recall and the short ballot as well as the Initiative and Referendum (and who struggled mightily, though in vain, for proportional representation there too), I disagree that recalls need be essentially personal. They simply reflect the democratic ideal that citizens should not be stuck with representatives whom they don’t want.

    I hope Newsom is not recalled myself. But it isn not–an should not be–up to me.

  10. Also, regarding impeachment v. recall, Anthony Trollope (of all people!) was right on the money, I think. Just as the Civil War was breaking out in the U.S., Trollope wrote this:

    “We know that he can be impeached by the Representatives and expelled from his office by the verdict of the Senate; but this in fact does not amount to much. Responsibility of this nature is doubtless very necessary, and prevents ebullitions of tyranny such as those in which a sultan or an emperor may indulge; but it is not that responsibility which especially recommends itself to the minds of free men. So much of responsibility they take as a matter of course, as they do the air which they breathe. It would be nothing to us to know that Lord Palmerston could be impeached for robbing the treasury, or Lord Russell punished for selling us to Austria…. We are anxious to know, not in what way they may be impeached and beheaded for great crimes, but by what method they may be kept constantly straight in small matters… [T]hey must be…of one mind with the public. Let them be that; or if not they, then with as little delay as may be, some others in their place. That with us is the meaning of ministerial responsibility. To that responsibility all the cabinet is subject. But in the government of the United States there is no such responsibility. The President is placed at the head of the executive for four years, and while he there remains no man can question him… There are no reins, constitutional or unconstitutional, by which he can be restrained. He can absolutely repudiate a majority of both Houses, and refuse the passage of any act of Congress even though supported by those majorities. He can retain the services of ministers distasteful to the whole country. He can place his own myrmidons at the head of the army and navy, or can himself take the command immediately on his own shoulders. All this he can do, and there is no one that can question him.” (North America, 1862, Vol. 2, Ch. 10)

    And, Trollope adds with disdain, “Seeing that Mr. Buchanan has escaped any punishment for maladministration, no President need fear the anger of the people.”

    Recent events here have certainly borne this out. Much more on this matter in my book, “Democracy Naturalized.”

    –WH

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.