Monday, September 28, 2020

The Winner-Take-All Problem with the Supreme Court Appointments Process

By Daniel Epps

Just like everyone else interested in constitutional law (and just like many other people in America, too) I've been thinking a lot about the Supreme Court in the wake of Justice Ginsburg's death. In particular, I've been continuing to reflect on what's wrong with our current system that has led us to this crisis. I discuss some of the system's defects in this Politico interview last week.  

In the interview, I talk about two failings with our system that should be pretty well understood by readers of this blog. First, there's the democracy deficit with our electoral system, in which the Senate and the Electoral College mean that the key decisionmakers may not represent the will of the majority of American citizens. Second, life tenure means that vacancies occur at unpredictable times, meaning that some Presidents get more opportunities to appoint Justices than others. These two problems interact with each other, making the problems worse. President Trump, who lost the popular vote and (so far) has only been elected President for one term, will get to appoint three Justices; Barack Obama, who won the popular vote twice only got to appoint two Justices. 

On reflection, though, I think there's another problematic feature of our structure that makes these first two problems much more salient: that's the winner-take-all nature of (1) our political process and (2) our judicial processes. 

Take the political aspect first. Senate elections are first-past-the-post systems. If you get the most votes, you win; it doesn't matter whether you win 51%-49% or 99%-1%. That can mean that small differences in elections can lead to huge differences in policy. In theory, we should see more moderate senators in states that are more closely split between liberals and conservatives. But that doesn't always happen. Ted Cruz won his 2018 Senate election by less than 3% over Beto O'Rourke; Cruz is one of the Senate's most conservative members. The fact that nearly half of the Texas electorate would have preferred the quite liberal O'Rourke isn't baked into Texas's representation in the Senate at all. So too with Presidential elections. Most states allocate their electoral votes on a winner-take-all basis. And then the winner becomes President no matter whether she wins by one electoral vote or gets nearly all of them. And again, even very close elections can lead to wildly divergent ideologies between the wining and losing candidates. 

And then consider the nomination and cconfirmation process. The President gets to pick whoever she wants; the fact that the prior election was close doesn't matter, except insofar as the Supreme Court choice might influence the President's reelection prospects. And then in the Senate, all that matters is whether the nominee can get the support of 51 Senators (or perhaps even 50 plus the Vice President, if, like Michael Ramsey, you disagree with Larry Tribe about whether the Vice President has the power to break a tie in votes on judicial appointments). 

Now take judicial procedure. What matters in a Supreme Court case is what five Justices think. The fact that there are four dissenters, or three, or zero, really doesn't matter. Sure, in some cases, those Justices might be able to persuade the majority to decide a case on a narrower ground. But there's nothing requiring a 5- or 6-justice majority bloc to do that. In practice, this means that some Supreme Court nominations are not that important, and others—such as the last two vacancies, which both shifted or will shift the Court's median Justice—are massively consequential. 

Combine all this with a legal and political system that are each becoming increasingly more polarized, and you have a system that allocates Supreme Court vacancies semi-randomly, in ways that are hard to justify democratically, and in a system that makes it possible for one side to capture the Supreme Court by installing a majority of committed partisans to the Court. As I keep saying, this system just doesn't make any sense, at least for the world we live in today. 

But I think if our system wasn't so winner-take-all, the other flaws with the system might be less troubling. Imagine if (1) Senatorial ideology more closely reflected the entire vote in their states; (2) presidential ideology reflected the entire vote of the electorate; and then (3) the confirmation process meaningfully incorporated the views of Senators outside the majority in some way. We'd probably end up with nominees that look pretty moderate, with smaller swings between the Justices appointed during Democratic versus Republican control. 

And then imagine that the Supreme Court itself made decisions that somehow better incorporated the views of Justices who aren't in the majority. Replacing or changing the median Justice on the Court wouldn't be quite such a big deal; it would still matter, but it wouldn't necessarily be such an earth-shattering shift. And while the democracy deficit and randomness problems would still persist, they wouldn't introduce such massive swings between possible outcomes that our current system does. 

Of course, seeing the problem is easier than fixing it. There are certainly models for moving away from first-past-the-post in the election context. We could have a system of proportional representation for electing legislators and then a coalition process for choosing the head of government, as many other countries do. I don't see an obvious way to make the views of the entire Senate relevant to the confirmation process. It's hard to imagine a system of unanimous consent working. But requiring 60 votes for confirmation, as we did before the filibuster for Supreme Court nominations was abolished, certainly made the system a bit less winner-take-all. And then when it comes to judicial procedure, I again don't have an easy solution. You could move to a system requiring unanimity for a judgment. In practice, that might lead to extreme minimalism, or it would lead to horse--trading between cases according to some informal norms within the Court. (Or it would just lead to utter deadlock).

In any event, even if I'm not clear on the solution, I think recognizing this aspect of the problem with the Supreme Court nomination process helps us explain the magnitude of the crisis. Until we fix at least one of the problems—the democracy deficit, the random distribution of appointments, and the winner-take-all nature of the system that enables each side to try to capture the Court with partisans—things aren't going to get much better. 


Rob Strodtman said...

I am somewhat hopeful that ranked choice voting is an available solution. Theoretically, it should make moderate third party candidates more viable. Once you have some critical mass of states using RCV and sometimes electing a third party Senator, then getting a judicial nomination through would require building a voting coalition around the nominee.

The trick obviously is getting the existing political parties on board with RCV when it will almost certainly diminish their own political power in the future. Maine did it by referendum, which is not going to be an option in a lot of states.

Coyote said...

The winner-take-all system for electoral votes could theoretically be eliminated by a future liberal SCOTUS declaring the winner-take-all system to be unconstitutional for this purpose. I would support such a SCOTUS move as long as it was done very early in the presidential campaign and as long as there would be no de minimis threshold for gaining electoral votes if they were apportioned proportionately by state.

Another problem that you neglect to mention, of course, is the fact that some US states are disproportionately represented in the US Senate relative to other US states. Sure, the Democrats could be more competitive in more conservative US states by running more Joe Manchins, but there will still be a problem of the US Senate being more conservative than the US as a whole is going to be (at least so long as small US states are going to remain disproportionately conservative). Unfortunately, though, I just don't see anything short of a new US Constitution as actually fixing this problem since the US Senate's unequal apportionment is explicitly protected from amendment by the US Constitution and since the 17th Amendment is also clear on the question of each US state having two US Senators.

Joe said...

Ranked choice voting would have been an interesting thing in the 1992, 2000 and 2016 presidential contests.

kotodama said...

I generally agree with the overall thrust of the OP, which is naturally written quite nicely as well, but I do take issue with a few random discrete points:

(1) "the quite liberal O'Rourke": Not that it's undesirable to be quite liberal -- for me, at least, that's a big selling point in fact -- but I just wonder how you arrive at that label? From what I can tell, he seems to be more or less in the party mainstream, i.e., garden variety liberal. In fact, he seems to have some of Obama's annoying habit of catering to Rs, in this case, Will Hurd, despite ending up with almost nothing to show for it and basically no reciprocation. Not to mention, according to the latest rankings I've seen, Cruz is in the top 5 most conservative senators, which is really saying something in 2020. So even Jesse Helms would have looked quite liberal in contrast. In other words, Beto is hardly the second coming of that other Senator Ted (Kennedy).

(2) "The fact that there are four dissenters, or three, or zero, really doesn't matter.": On this, it's not really my place to quibble with a well-known SC expert, but I'll throw caution to the wind and do it anyway. Surely, it is true that the absence or presence of dissents has effectively no impact on the bottom-line outcome of the decision. That said, I had thought folks generally recognized at least a handful of lesser benefits of dissenting. For example, a dissent is read by some law students, who in turn may go on to become judges. Likewise, a dissent is of course persuasive authority. Otherwise you could find some fault with law review articles, which aren't the law either. Another example could be that a 5-4 decision has less staying power later on for stare decisis purposes. It also robs the decision of some legitimacy in the public eye. What if Bush v. Gore had been 9-0 instead? A more "selfish" benefit -- relevant to current events with RBG -- is that a justice is also thinking about how to fortify her or his individual legacy. Finally, and more generally, I think lawyers -- and most everyone for that matter, but especially lawyers -- just want to get a word in edgewise, even if ultimately they won’t prevail.

(3) "In practice, this means that some Supreme Court nominations are not that important": Maybe I'm in the minority (no pun) here, but I'm just puzzled as to how you draw this conclusion. That just doesn't seem right to me and I'm inclined to wager that folks across the D-R spectrum would feel similarly. Unless court packing happens -- and that doesn't seem like a sure thing at all even under current circumstances, and even assuming ACB is confirmed before inauguration day, which is looking pretty likely right now -- then you're stuck with a 9-justice court. In that case, you would always prefer to be closer to 5 than farther away. So in that sense, All Justices Matter.

(4) "We'd probably end up with nominees that look pretty moderate": This goes back to (1). At least on the D side, I'm curious as to which nominees you think *weren't already* pretty moderate? Not RBG I think. Definitely not Breyer or Kagan, or Garland for that matter. Not even Sotomayor really. It's not like she's Wild Bill resurrected. So if there is any problem of lack of moderation, to me it lies solely with the Rs, because they just seem to keep upping the ante on putting forward increasingly extreme individuals.

@Coyote: I fully agree, and it's not just the Senate that has long been woefully deficient. To me, Marshall put it best in his bicentennial speech, and he even left out the genocide and suppression of the indigenous population. As I see it, squeezing any more mileage out of a document created in those circumstances -- even with all the later amendments -- is like not only trying to keep a 1945 ENIAC up and running in 2020, but also expecting it to do weather simulations on a par with an IBM Summit. At some point it just can’t work.

kotodama said...

[apologies for posting seriatim but I had a verbosity attack and exceeded the single-post character limit]

@Unknown: I'm just struggling to grasp the current infatuation -- in some circles -- with RCV. To me, the famous words of the Sage of Omaha, albeit spoken in a different context, are quite relevant here: "Beware of geeks bearing formulas." In this case, RCV would be the formula, and the reason to beware is that the assumptions supporting RCV seem to be entirely untethered from current reality. As it appears to me, RCV assumes, among other things, (1) perfect voter participation, (2) perfect voter information, (3) that a voter could meaningfully identify 2nd, 3rd, etc. choices, (4) no voter suppression, (5) that RCV inherently produces more "moderate" (by whose definition?) outcomes, and (6) no bad intentions by third-party candidates, e.g., to merely be spoilers or as we've seen recently and even more egregiously, to actually collude with a major party (and you could add (7) of no gerrymandering in cases where it applies). I don't think any of those line up with reality or with the empirical evidence. Especially for (1) and (2), I think most (but maybe not the Rs!) would agree that two of the biggest problems in all U.S. elections are lack of participation by eligible voters and lack of voter education. (4) is likewise a huge problem in my estimation and only getting worse these days. For (3), at least in a head-to-head election, I just don't see how that works. No Trump voter would have listed Hillary as a second choice and likewise for a Hillary voter. And all the third-party voters seem to be the ones who despise both major parties, so why would they want to give them an assist with a 2nd or 3rd choice selection? For (5), whether or not everyone could agree on a definition of "moderate" (seems most unlikely), at a minimum I think no one would confuse Ross Perot, Ralph Nader, Gary Johnson, or Evan McMullin with a moderate. Likewise -- and apologies for being very inside baseball at this point -- in the recent D primary for my congressional district with a very crowded field, regardless of whether one was happy with the outcome, I think all would agree that the *most moderate* candidate won *even without* RCV. (6) doesn't require further elaboration I think.

On the flipside, although I don't agree with the silly argument from some circles that "gov't can only do one thing at a time", it is true that gov'ts, especially state and local gov'ts, do have limited resources in terms of time and money, and implementing RCV would seem to consume a nontrivial amount of those resources. That's to say nothing of the further costs in voter education and inevitable voter confusion -- or even discouragement -- that would result. So to me, RCV is not just harmless in the sense of being a solution in search of a fictional problem (or, more generously, simply awaiting a future problem). Rather, given all these urgent, existing issues with voting outlined above, it would seem to be malpractice to pursue RCV and thereby deplete finite resources that could be put toward solving those issues.

@Joe: I guess it might have been, although I'm not entirely convinced, as that goes to my (3) above in response to Unknown. But if you're going to indulge a counterfactual, why not just go the whole enchilada and assume nat'l popular vote? In that case Bubba still wins, but instead we also get Gore and Hillary.

Joe said...

I'm welcome to the idea of a national popular vote, but the specific issue of some sort of ranked choice voting was on the table. Also, if there was a national popular vote, if the difference between the two candidates is less than the third party vote, ranked choice voting is important too.

On that front, 2016 is tricky. Clinton as some highlight was the plurality winner. But, what if you toss in the third party? One would think if you put a gun to their head a typical Jill Stein voter would rather her than Trump. But, we are talking idiosyncratic voters for which that is not guaranteed. But, even if every Jill Stein voter chose Clinton as a second choice, that won't get you a majority. Basically, doing the math, it would turn on libertarians. How THAT would split is unclear too. Of course, different system, the candidates would run differently too.

As to the benefits of the system, granted this is not my area of even amateur expertise, but it is my understanding there is some evidence it is beneficial. That voters generally can offer second choices and the like. I think it might work best in primaries. You might have four or five candidates with the winner getting at best thirty or so percent. Runoffs are not always done & if done they are costly affairs. Plus, the top two vote getters very well might not actually win using some sort of ranked choice voting.

My city is starting it for local races next year, so I will get a personal taste.

Rob Strodtman said...

@hardreaders: FWIW I don't think you're correctly characterizing RCV in your critique. I'm talking about this section:

"As it appears to me, RCV assumes, among other things, (1) perfect voter participation, (2) perfect voter information, (3) that a voter could meaningfully identify 2nd, 3rd, etc. choices, (4) no voter suppression, (5) that RCV inherently produces more "moderate" (by whose definition?) outcomes, and (6) no bad intentions by third-party candidates, e.g., to merely be spoilers or as we've seen recently and even more egregiously, to actually collude with a major party (and you could add (7) of no gerrymandering in cases where it applies)"

(1): RCV does not require this. You do not need to rank every candidate. If you rank one candidate as #1 and rank nobody else that is functionally exactly the same as the voting system we are all used to.

(2): RCV does not require this, you simply need to have an orderable preference given a slate of candidates. I don't think you can articulate why RCV needs "perfect voter information," whatever that means, and our current voting system deals better with under-informed voters.

(3): I do not think this is hard. E.g.: "5 fruits are presented, please state your preference for which fruit you'd like to eat in order from strongest preference to least. If you absolutely would not like to eat one or more of the fruits, leave them off your list." That is not a hard instruction set, and yields a ranked vote, just replace fruit with politician and eat with "serve in office X".

(4): RCV can't fix suppression, I agree. It's not a panacea.

(5): It doesn't *inherently* produce more moderate candidates, but does make it so third party candidates are not spoilers and therefore can attract many more "1st" preference votes. It removes strategic voting where votes for a third party are "wasted". To the extent any third party candidates are ever elected such that single party doesnt get a majority in a house, that would result in moderation because they or the other side would HAVE to be bargained with to get a voting majority.

(6): Spoiler candidates are not possible with RCV, that's the entire point, so long as someone who prefers the would-be spoiler selects at least one second preference. Spoilers would have to actively convince their supporters to not prefer anyone else. With minimal voter education, such a ploy would be nakedly obvious.

(7): RCV can't fix gerrymandering.

Elsewhere you also state that you don't see how it would work in a head to head matchup. Well it doesn't need to because if there are truly only two candidates on the slate, RCV is wholly unnecessary because one WILL have a majority of votes. Easy.

With your other concerns about implementation: Maine will be a case study. I have not seen the hiccups or resource issues you're describing thus far. We'll see on election day I suppose.