How to Retaliate for Garland
by Michael Dorf
In my latest Verdict column, I explain why I declined to sign a letter from 72 former law clerks of Justice Kennedy in support of the confirmation of Brett Kavanaugh to the Supreme Court. To summarize, I argue: (1) Kennedy clerks do not have any special insight into Kavanaugh's qualifications in virtue of having clerked for Kennedy; (2) the letter purports to reflect a set of politically diverse views, but in fact nearly all of the signers are very conservative; and (3) the letter tacitly assumes without defending the controversial position that the role of the Senate should be limited to examining the professional credentials and judicial temperament of the nominee. Here I want to elaborate on why I think that assumption is wrong under current conditions.
As I explain in the column, there is a plausible argument for a norm under which Senators vote to confirm any professionally well-qualified nominee with an appropriate judicial temperament, regardless of party or ideology, so long as the nominee is not "extreme." The worry is that absent such a norm, we will end up with prolonged vacancies during periods of divided government and extreme nominees who lead to intra-SCOTUS polarization once confirmed on party-line votes. That's a real danger that appears to have been realized, I acknowledge, but Democrats who vote to confirm Kavanaugh in the hope of thereby restoring a reciprocal norm of bipartisan deference to the president are suckers.
Elaborating on a suggestion of David Leonhardt in the NY Times, I advocate what game theorists call tit-for-tat. In tit-for-tat, one cooperates with a rival so long as the rival cooperates, but if the rival unfairly takes advantage, one retaliates proportionately. As I cash out the strategy in the current context, Democrats need to retaliate for the Republicans' treatment of Merrick Garland before they can or should vote for any Republican president's right-of-center nominees. Voting against the confirmation of Republican nominees is necessary but not sufficient. So long as those nominees take their seats--as Justice Gorsuch did and as Judge/Justice-to-be Kavanaugh likely will--the Democrats will not have retaliated successfully. Successful retaliation requires actually blocking a Republican nominee.
Yet how can Democrats do that? It is conceivable that something could emerge in Judge Kavanaugh's record that would be sufficient to give pause to all of Senators Collins, Murkowski, Heitkamp, Manchin, and Donnelly (the two Republican Senators who call themselves pro-choice and the three Democrats from red states who face tough re-election campaigns, all five of whom voted to confirm Justice Gorsuch). That would have to be one hell of a smoking gun--something on the order of a memo from Kavanaugh to President Bush saying "waterboarding is torture but do it anyway, and also, by the way, I'll definitely vote to overrule Roe v. Wade if I'm ever a Supreme Court justice."
In the much more likely event that the Senate confirms Judge Kavanaugh, the Democrats will still be waiting for an opportunity to retaliate successfully for the Republicans' treatment of Judge Garland. That opportunity would not likely arise until Democrats control the Senate during the second half of President Trump's term or (gulp/ugh!) a second Trump term. Depending on the outcome of various elections and the timing of Supreme Court vacancies, the opportunity for successful retaliation might not arise at all for a decade or two.
That, in turn, raises a question about the utility of the tit-for-tat strategy in the current context. In the computer simulation settings in which tit-for-tat has proven itself effective, the penalty of non-cooperation immediately follows the offending conduct. In the real world, a delay on the order of ten or twenty years strains the theory. For one thing, many of the Republicans who denied a hearing to Judge Garland would no longer be serving in the Senate by then. For another, to some extent the strategy aims at the public, who have a notoriously short attention span. It is easy to imagine people--including journalists--in, say, 2032, being utterly befuddled when Democratic Senators justify their refusal to provide a hearing for a Republican president's Supreme Court nominee as retaliation for the 2016 treatment of Judge Garland. "Merrick who?," one imagines people saying.
Democrats may have an opportunity to take other kinds of retaliatory action before the stars align to produce a Republican presidency, a solid Democratic Senate majority, and a Supreme Court vacancy. As early as 2021, one could well imagine a Democratic president and Democratic majorities in both the House and Senate. At that point, various commentators have suggested, the Democrats should respond by adding two seats to the Supreme Court to counteract the Republicans' denial to Garland of his seat. (Two new seats, rather than one, would be required to restore the status quo ante: one seat would "cancel out" Gorsuch; the other would enable the Garland replacement to cast liberalish votes.)
The most straightforward version of such a program would simply add the two seats. A variation proposed by Ian Ayres and John Witt would have the two new seats expire after 18 years. Why 18? Apparently because under a proposal that Ayres and Witt also support, all justices would be limited to staggered 18 year terms, thus ensuring that a vacancy would arise every two years and eliminating the accidents of timing in appointments. (A justice who died or retired before the 18 years were up would be replaced by a justice who would complete that term, not a full one.) The original version of this proposal--by Paul Carrington and Roger Cramton--would have been accomplished by statute, but recognizing its vulnerability to constitutional challenge, Ayres and Witt propose adopting it by constitutional amendment if a statute is inadequate to the task. Although I too would support such a change, I'll set this more radical proposal aside here, because it is not a form of retaliation for the Republicans' treatment of Judge Garland. By contrast, the "court balancing" plan of adding two time-limited Supreme Court seats during a period of Democratic dominance would constitute such retaliation, as would the proposals to add two conventional seats.
What can we say about the proposal to add two seats and the Ayres/Witt variation? Consider four observations.
(1) These proposals are best viewed not as alternatives to tit-for-tat but as alternative forms of tit-for-tat. The term tit-for-tat implies that the game-theoretical responder does exactly to the bad actor what the bad actor did first. That works in the highly stylized computer models in which tit-for-tat has shown itself to be an effective strategy, but in the real world, actors often have different capacities. To give a current example, when the US puts tariffs on Chinese steel and aluminum, China does not respond by putting tariffs on US steel and aluminum, because China does not import substantial quantities of US steel and aluminum. Instead, China responds by putting tariffs on US agricultural products such as soybeans. The responder in a real-world tit-for-tat situation responds proportionately but usually not identically. Likewise here, if Democrats think that it could be decades before they are able to respond to the Republican treatment of Judge Garland by treating a Republican nominee identically, they may prefer some other form of retaliation, such as the addition of two seats to the Supreme Court.
(2) But would it end there? Responding to the Ayres/Witt proposal, Ilya Somin writes:
(3) There is also something peculiar or at least highly premature about Democrats talking about increasing the size of the Supreme Court now, when they control neither house of Congress nor the presidency. By legitimizing court-packing, Democrats run the risk of giving Republicans cover to do it immediately. I don't think that's likely, but it does strike me as odd to undercut one of the few longstanding norms regarding judicial appointments that is holding for now.
(4) Using the "nuclear option," Democrats abolished the 60-vote cloture rule for lower court and executive branch nominations when they last held the Senate. Then, during the Gorsuch confirmation process, Republicans abolished it for Supreme Court nominations. Although President Trump has occasionally tweeted his support for doing away with the cloture rule for ordinary legislation, Senate Majority Leader Mitch McConnell has thus far resisted. One might be tempted to think that if McConnell wants to keep the cloture rule, then it must be good for Republicans and therefore bad for Democrats. If so, then if the Democrats do retake the presidency and Congress in the 2020 elections but have fewer than 60 seats in the Senate, they should deploy the nuclear option again to abolish what's left of the cloture rule. Doing so would allow them to add two seats to the Supreme Court and would more generally make it easier to enact the next Democratic president's agenda.
Yet it is not obvious that abandoning the cloture rule would be good for Democrats. Perhaps McConnell's objection to doing so rests on respect for Senate traditions rather than partisan considerations. Or perhaps it rests on partisan considerations but he is mistaken in his overall judgment. On the basis of first principles, it is not clear where the advantage lies.
On one hand, other things being equal, Democrats are more likely to want government to act while Republicans are more likely to want to block government action. Against that baseline, it looks like the cloture rule serves Republicans' interests more than Democrats', because it raises the threshold required for action.
Yet the baseline is not government inaction. The baseline is the current legal status quo. A high threshold for government action means a high threshold for repealing legislation as well as enacting new regulatory programs. Preserving the 60-vote cloture rule for ordinary legislation makes it harder for Republicans to repeal or weaken the Endangered Species Act, the Clean Air Act, Social Security, and numerous other laws that Democrats cherish. And given that the Senate overweights sparsely populated red states relative to blue states, Democrats might conclude that over the long run they have more to fear from Republican-controlled Senates than to gain from Democratic-controlled ones.
Whatever judgment Democrats make about the cloture rule for ordinary legislation, they ought to make it based on a careful assessment of costs and benefits across the entire range of potential consequences. If court-packing is thought to be otherwise justified, it might not be justified if it can only be accomplished by first abolishing the 60-vote threshold for ordinary legislation.
In my latest Verdict column, I explain why I declined to sign a letter from 72 former law clerks of Justice Kennedy in support of the confirmation of Brett Kavanaugh to the Supreme Court. To summarize, I argue: (1) Kennedy clerks do not have any special insight into Kavanaugh's qualifications in virtue of having clerked for Kennedy; (2) the letter purports to reflect a set of politically diverse views, but in fact nearly all of the signers are very conservative; and (3) the letter tacitly assumes without defending the controversial position that the role of the Senate should be limited to examining the professional credentials and judicial temperament of the nominee. Here I want to elaborate on why I think that assumption is wrong under current conditions.
As I explain in the column, there is a plausible argument for a norm under which Senators vote to confirm any professionally well-qualified nominee with an appropriate judicial temperament, regardless of party or ideology, so long as the nominee is not "extreme." The worry is that absent such a norm, we will end up with prolonged vacancies during periods of divided government and extreme nominees who lead to intra-SCOTUS polarization once confirmed on party-line votes. That's a real danger that appears to have been realized, I acknowledge, but Democrats who vote to confirm Kavanaugh in the hope of thereby restoring a reciprocal norm of bipartisan deference to the president are suckers.
Elaborating on a suggestion of David Leonhardt in the NY Times, I advocate what game theorists call tit-for-tat. In tit-for-tat, one cooperates with a rival so long as the rival cooperates, but if the rival unfairly takes advantage, one retaliates proportionately. As I cash out the strategy in the current context, Democrats need to retaliate for the Republicans' treatment of Merrick Garland before they can or should vote for any Republican president's right-of-center nominees. Voting against the confirmation of Republican nominees is necessary but not sufficient. So long as those nominees take their seats--as Justice Gorsuch did and as Judge/Justice-to-be Kavanaugh likely will--the Democrats will not have retaliated successfully. Successful retaliation requires actually blocking a Republican nominee.
Yet how can Democrats do that? It is conceivable that something could emerge in Judge Kavanaugh's record that would be sufficient to give pause to all of Senators Collins, Murkowski, Heitkamp, Manchin, and Donnelly (the two Republican Senators who call themselves pro-choice and the three Democrats from red states who face tough re-election campaigns, all five of whom voted to confirm Justice Gorsuch). That would have to be one hell of a smoking gun--something on the order of a memo from Kavanaugh to President Bush saying "waterboarding is torture but do it anyway, and also, by the way, I'll definitely vote to overrule Roe v. Wade if I'm ever a Supreme Court justice."
In the much more likely event that the Senate confirms Judge Kavanaugh, the Democrats will still be waiting for an opportunity to retaliate successfully for the Republicans' treatment of Judge Garland. That opportunity would not likely arise until Democrats control the Senate during the second half of President Trump's term or (gulp/ugh!) a second Trump term. Depending on the outcome of various elections and the timing of Supreme Court vacancies, the opportunity for successful retaliation might not arise at all for a decade or two.
That, in turn, raises a question about the utility of the tit-for-tat strategy in the current context. In the computer simulation settings in which tit-for-tat has proven itself effective, the penalty of non-cooperation immediately follows the offending conduct. In the real world, a delay on the order of ten or twenty years strains the theory. For one thing, many of the Republicans who denied a hearing to Judge Garland would no longer be serving in the Senate by then. For another, to some extent the strategy aims at the public, who have a notoriously short attention span. It is easy to imagine people--including journalists--in, say, 2032, being utterly befuddled when Democratic Senators justify their refusal to provide a hearing for a Republican president's Supreme Court nominee as retaliation for the 2016 treatment of Judge Garland. "Merrick who?," one imagines people saying.
Democrats may have an opportunity to take other kinds of retaliatory action before the stars align to produce a Republican presidency, a solid Democratic Senate majority, and a Supreme Court vacancy. As early as 2021, one could well imagine a Democratic president and Democratic majorities in both the House and Senate. At that point, various commentators have suggested, the Democrats should respond by adding two seats to the Supreme Court to counteract the Republicans' denial to Garland of his seat. (Two new seats, rather than one, would be required to restore the status quo ante: one seat would "cancel out" Gorsuch; the other would enable the Garland replacement to cast liberalish votes.)
The most straightforward version of such a program would simply add the two seats. A variation proposed by Ian Ayres and John Witt would have the two new seats expire after 18 years. Why 18? Apparently because under a proposal that Ayres and Witt also support, all justices would be limited to staggered 18 year terms, thus ensuring that a vacancy would arise every two years and eliminating the accidents of timing in appointments. (A justice who died or retired before the 18 years were up would be replaced by a justice who would complete that term, not a full one.) The original version of this proposal--by Paul Carrington and Roger Cramton--would have been accomplished by statute, but recognizing its vulnerability to constitutional challenge, Ayres and Witt propose adopting it by constitutional amendment if a statute is inadequate to the task. Although I too would support such a change, I'll set this more radical proposal aside here, because it is not a form of retaliation for the Republicans' treatment of Judge Garland. By contrast, the "court balancing" plan of adding two time-limited Supreme Court seats during a period of Democratic dominance would constitute such retaliation, as would the proposals to add two conventional seats.
What can we say about the proposal to add two seats and the Ayres/Witt variation? Consider four observations.
(1) These proposals are best viewed not as alternatives to tit-for-tat but as alternative forms of tit-for-tat. The term tit-for-tat implies that the game-theoretical responder does exactly to the bad actor what the bad actor did first. That works in the highly stylized computer models in which tit-for-tat has shown itself to be an effective strategy, but in the real world, actors often have different capacities. To give a current example, when the US puts tariffs on Chinese steel and aluminum, China does not respond by putting tariffs on US steel and aluminum, because China does not import substantial quantities of US steel and aluminum. Instead, China responds by putting tariffs on US agricultural products such as soybeans. The responder in a real-world tit-for-tat situation responds proportionately but usually not identically. Likewise here, if Democrats think that it could be decades before they are able to respond to the Republican treatment of Judge Garland by treating a Republican nominee identically, they may prefer some other form of retaliation, such as the addition of two seats to the Supreme Court.
(2) But would it end there? Responding to the Ayres/Witt proposal, Ilya Somin writes:
What matters is how the "court-balancing" plan will be perceived by Republicans. If they agree that it is merely a one-time "temporary intervention" justified by the GOP's actions against Garland, then the Ayres-Witt ploy will work. But if they see it as a major escalation in the judicial nomination wars, then they will almost certainly retaliate in kind as soon as they get the chance. I think it's pretty obvious that the overwhelming majority of Republicans (as well as many independents) are likely to take the the latter view. They won't be mollified by the supposed distinction between "court balancing" and "court-packing," in part because they really aren't all that different, and in part because they don't believe that blocking Garland was wrong in the first place. At best, the GOP might limit its retaliation to appointing two new 18-year justices of their own, as opposed to life-tenured ones. At worst, they could choose to escalate further, such as by appointing justices to longer terms, or by appointing a larger number of new justices than the Democrats added.I agree with Somin. Even if Democrats think that court-packing or court-balancing is simply tit-for-tat, Republicans will perceive it as escalation, to which they would almost certainly retaliate proportionally or by escalating further. And because the whole point of game theory is to act in a way that anticipates competitors' responses, that makes court-packing/balancing a dangerous strategy.
(3) There is also something peculiar or at least highly premature about Democrats talking about increasing the size of the Supreme Court now, when they control neither house of Congress nor the presidency. By legitimizing court-packing, Democrats run the risk of giving Republicans cover to do it immediately. I don't think that's likely, but it does strike me as odd to undercut one of the few longstanding norms regarding judicial appointments that is holding for now.
(4) Using the "nuclear option," Democrats abolished the 60-vote cloture rule for lower court and executive branch nominations when they last held the Senate. Then, during the Gorsuch confirmation process, Republicans abolished it for Supreme Court nominations. Although President Trump has occasionally tweeted his support for doing away with the cloture rule for ordinary legislation, Senate Majority Leader Mitch McConnell has thus far resisted. One might be tempted to think that if McConnell wants to keep the cloture rule, then it must be good for Republicans and therefore bad for Democrats. If so, then if the Democrats do retake the presidency and Congress in the 2020 elections but have fewer than 60 seats in the Senate, they should deploy the nuclear option again to abolish what's left of the cloture rule. Doing so would allow them to add two seats to the Supreme Court and would more generally make it easier to enact the next Democratic president's agenda.
Yet it is not obvious that abandoning the cloture rule would be good for Democrats. Perhaps McConnell's objection to doing so rests on respect for Senate traditions rather than partisan considerations. Or perhaps it rests on partisan considerations but he is mistaken in his overall judgment. On the basis of first principles, it is not clear where the advantage lies.
On one hand, other things being equal, Democrats are more likely to want government to act while Republicans are more likely to want to block government action. Against that baseline, it looks like the cloture rule serves Republicans' interests more than Democrats', because it raises the threshold required for action.
Yet the baseline is not government inaction. The baseline is the current legal status quo. A high threshold for government action means a high threshold for repealing legislation as well as enacting new regulatory programs. Preserving the 60-vote cloture rule for ordinary legislation makes it harder for Republicans to repeal or weaken the Endangered Species Act, the Clean Air Act, Social Security, and numerous other laws that Democrats cherish. And given that the Senate overweights sparsely populated red states relative to blue states, Democrats might conclude that over the long run they have more to fear from Republican-controlled Senates than to gain from Democratic-controlled ones.
Whatever judgment Democrats make about the cloture rule for ordinary legislation, they ought to make it based on a careful assessment of costs and benefits across the entire range of potential consequences. If court-packing is thought to be otherwise justified, it might not be justified if it can only be accomplished by first abolishing the 60-vote threshold for ordinary legislation.
* * *
I am left to conclude that there may be no effective way for Democrats to engage in effective tit-for-tat retaliation for Republicans' treatment of Judge Garland. If so, however, that does not mean that Democrats should simply capitulate by pretending that we live in ordinary times. So long as the Garland wound remains reasonably fresh, non-cooperation should be the order of the day.