When is a cross not just a cross?

Hint: Almost always, and definitely when it’s 40-feet-tall!

The Establishment Clause expressly prohibits the State from making any “law respecting an establishment of religion.” Yet, for decades, the Supreme Court has applied this particular limit on government power only selectively and half-heartedly. On the facts, it seems obvious that a 40-foot-tall stone and concrete cross set astride a traffic median by the State of Maryland is, first and foremost, a religious display by the State. It is equally obvious that the decision to honor WWI veterans with this particular symbol is either pretextual or secondary to the religious symbolism of the Latin cross. Granted, a cross can be seen simply as a marker for the dead (although many find it exclusionary, and offensively so, in that context) but when the cross is 40-friggin-feet-tall, its predominant meaning is unmistakable. 

Facts versus fictions

As clear as American Legion v. American Humanist Association might be on the facts, we suspect that in this particular area of constitutional law, facts matter a lot less than beliefs about the inherent legitimacy of Christian values and identity as universal American values and identity.

The universalist claim is wrong in every possible way. The founders did not hold to it even when the population was overwhelmingly Christian, and there are now more non-Christians in America today than there are people in Thailand, the United Kingdom, France, or South Korea. Nonetheless, the fiction that symbols like the Latin cross and the Ten Commandments are simultaneously Christian and yet universal is an article of faith among many conservative Christians, and this almost certainly includes the five conservative Christian men who constitute the majority of the Supreme Court.

Unpredictable jurisprudence, predicable jurists

The Establishment Clause jurisprudence on questions such as the mandatory prayer readings and the display of religious monuments is inconsistent and unpredictable, but the justices themselves are not.

Our predictive model, based solely on the oral argument, is almost exactly what you would expect. We see Justices Alito, Gorsuch, and Kavanaugh strongly in favor of the Petitioners (The American Legion and The Maryland-National Capital Park and Planning Commission), joined by Justice Breyer (consistent with his live-and-let-live concurrence in Van Orden v. Perry) and Chief Justice Roberts (who admittedly comes out as a toss-up on the numbers, but there seems little doubt where his sympathies were).

The female justices of the Supreme Court, Justices Ginsburg, Kagan, and Sotomayor, seemed far more skeptical of the universalist assumption. It is probably no coincidence that Ginsburg and Kagan are Jewish and that Sotomayor is (in her own words) a wise Latina.

Bar chart showing predictions for each supreme court justice. Justices are ordered from pro-Petitioner to Pro-Respondent. Pro-petitioner represented by blue bars to the right of the 50 percent midline in the top half of the figure. Pro-Respondent represented by orange bars to the right of the midline in the lower portion of the figure. The results are discussed in the text.
ScotusOA.com prediction based on Supreme Court Oral Argument in American Legion v. American Humanist Association (February 27, 2019)8

Prediction: 6:3 for Petitioners

For Petitioner: Alito, Gorsuch, Kavanaugh, Roberts, and Thomas. Breyer concurring.

For Respondent: Ginsburg, Kagan, and Sotomayor

Ginsburg is back: is she feistier than ever?

Liberal feminist icon, movie star, survivor. That is how we think of the diminutive 85 year old Supreme Court Justice Ruth Bader Ginsburg. On November 8th last year, Ginsburg was hospitalized after fracturing three ribs in a fall in her office at the Supreme Court. She still turned up for work at the next oral argument in Apple v Pepper on November 26th. But after a CT scan of her ribs revealed signs of lung cancer, the Justice was forced to take time away from the Bench to recuperate from surgery in late December. Justice Ginsburg returned to the Court last Tuesday after missing all of the Court’s January cases (although she kept up with the cases remotely). This was her third bout of cancer but her first absence from oral argument since arriving in 1993.

Ginsburg’s Return

The stage for Ginsburg’s return was an obscure but engaging question of statutory interpretation as to whether the government—the U.S. Postal Service in this case—is a “person” entitled to use the post-grant patent review procedures under the 2011 America Invents Act.

Ginsburg was the first justice to ask a question in Return Mail, Inc. v. United States Postal Service on Tuesday and she followed up four more times over the course of the hour. Ginsburg’s first question forced Petitioner’s advocate Beth Brinkmann to concede that her point about estoppel, which she described as the “linchpin” of her argument, was not actually an essential element of that argument.

Ginsburg’s second question began one of the main themes of the discussion that followed:

Ruth Bader Ginsburg:

But why — why would — why would the government — why — why would Congress want to leave a government agency out of this second look if the idea is to weed out patents that never should have been given in the first place?

Ginsburg Watch

For those who care more about whether President Trump will get a chance to complete the Federalist Society takeover of the Supreme Court than they do about the intersection of patent law and administrative law, the burning question on Tuesday was simply: how was Justice Ginsburg?

NRP’s Nina Totenberg reported that at the end of the argument, Ginsburg got up carefully, descended the steps behind the bench and walked out on her own. Listening to the argument on Oyez.org, we were struck that Ginsburg’s voice sounded stronger and clearer than it has for a long time. But with all the hype surrounding “Notorious RBG” these days, we thought we should try to answer the question empirically.

The figure below shows the number of words that Justice Ginsburg spoke at each argument over the course of the 2018 Term thus far. We split the data around Justice Ginsburg’s injury when she cracked her ribs: the arguments prior to November 8, 2018, are represented by the orange bars; those afterward are represented in light blue. The superimposed gray dots represent the average number of words spoken by the nine justices of the Court in each argument.

Figure showing the number of words spoken by Justice Ginsburg each argument over the course of the 2018 Term
ScotusOA.com: Justice Ginsburg’s Participation in Supreme Court oral argument, 2018 Term

As the figure shows, Ginsburg usually speaks less than the average justice and her level of participation varies. Although she seemed slightly subdued after her injury in November, she actually contributed more than the average justice in two post-November 8 cases, Nutraceutical Corp v Lambert on November 18 (722 words) and Gamble v United States on December 6 (567 words). For the rest of the Court, these cases were not unusually quiet.

Perhaps our subjective impression of Justice Ginsburg’s strong performance on Tuesday was a result of relief at having her back, or noting the strength of her voice, or wishful thinking, but it doesn’t really fit the objective data for the Term. Justice Ginsburg certainly made her presence felt in Return Mail, but she only spoke 123 words. To put that in context, she spoke more than this in 25 cases this Term and less in only 6 others excluding the January cases where she was absent. In fact, in 20 cases this Term, Ginsburg has had more to say to just one side of the argument (i.e., Petitioner alone or Respondent alone) than she said to both sides in Return Mail. Ginsburg is back, but only at 56% of her average for the Term.

We wish her well in her continuing recovery.

A truly unusual coalition: Predicting Garza v. Idaho

Back in late October last year, the Court heard argument in Garza v. Idaho on whether the “presumption of prejudice” for ineffective assistance of counsel should apply when a criminal defendant instructs his counsel to file a notice of appeal but counsel decides not to do so because the defendant’s plea agreement included an appeal waiver. In a time where pleas account for approximately 97% of resolutions of criminal defense matters, and when appeal waivers are increasingly demanded by prosecutors, the case has the potential to offer one small procedural protection in a context in which the usual procedural protections mostly do not apply. Garza v. Idaho also will potentially reveal some unusual coalitions among the justices, and offer an opportunity to learn more about the views of the Justice whose voting record we know least about, Justice Kavanaugh.

Our prediction is that any day now we may see the newest conservative justice cross the metaphorical divide and vote with his liberal brethren in Garza. Here is our prediction:

Supreme Court Oral Argument Prediction for Garza v. Idaho based on the oral argument of October 30, 2018.
Supreme Court Oral Argument Prediction for Garza v. Idaho based on the oral argument of October 30, 2018.

Elsewhere, one of us (Jacobi) has argued that what are often called “unusual coalitionsare not so unusual. Cross-ideological coalitions often occur where the rift is between pragmatists—such as Justice Breyer on the left and Justice Alito on the right—and formalists— most notably Justice Ginsburg on the left and Justice Scalia on the right.

Garza is interesting because, based on our prediction, the case seems set to bridge both the liberal-conservative and the formalist-pragmatist divide. It has pragmatists such as Breyer joining with formalists such as Justice Gorsuch, both seemingly favoring the pro-defendant side.

The central issue in Garza is the correct application of an earlier Supreme Court case, Roe v. Flores-Ortega. Tellingly, Ginsburg and Breyer split in that case—Breyer opting for a case-by-case approach, and Ginsburg joining a dissent in favor of an across-the-board rule.

It appears that in Garza, that split among the liberals is likely to recur. Breyer appears to be sitting strongly in Petitioner’s camp, whereas Ginsburg’s first question suggested that her preference for rules may be more determinative for her here than her usual sympathies for criminal defendants, asking whether the defendant isn’t trying to have his cake and eat it, too.  If we drew a map of the justices along both the liberal-conservative and the formalist-pragmatist dimensions, Ginsburg and Alito would occupy opposite corners, not just opposite sides. But in this case, Alito cross-referenced Ginsburg in the oral argument, something that we have shown to be a sign of likely coalition partners.

And where is Kavanaugh in this unusual mix? In Garza, the Courtroom heard Kavanaugh emphasizing practical concerns. He stressed both that an appeal waiver gives up very little in reality, and that the circuits that have applied a presumption of prejudice have not shown any serious problem. Eventually, he drew an acknowledgment from Respondent Idaho’s advocate that there would be no meaningful practical harm if the state lost on this question. In contrast, Gorsuch asked a series of questions that all had a formalist bent, hinging, for example, on whether a right was statutory in nature, and what the traditional division of decision-making responsibilities is between attorney and client. The two conservatives seem set to agree in ruling in favor for Petitioner, but for very different reasons.

It seems understandable to us that Petitioner could win on both formal and pragmatic grounds. Formally, if it is acknowledged, as almost everyone at the argument was willing to do, that there are some rights that are unwaivable—a plea waiver could surely not prevent an appeal for gross prosecutorial misconduct, for example—then which rights are waived in any circumstance is a question that must have some ambiguity, and thus require legal analysis. Consequently, it cannot be such an obvious question that an attorney can properly fail to respond to his or her client’s expressed desire to challenge. More pragmatically, Flores-Ortega emphasized that the client retains the choice over fundamental matters such as whether to appeal, and last year’s McCoy v. Louisiana embraced the client’s autonomy, even if the attorney in good faith believes it better to concede guilt in order to avoid the death penalty. Similarly, here, whatever an attorney may think of the potential foolishness of jeopardizing a favorable plea deal by appealing a potentially already waived right, it is the client’s choice to make that bad decision. The majority of the Court seemed to be saying that the Court already has practical a solution to the attorney’s dilemma: the attorney can always choose to file an Anders brief, no harm done.

Interestingly, the Court’s two newest conservative justices seemed willing to consider these arguments, even though they cut across the traditional ideological conservative favor for the prosecution. In contrast, the Court’s more senior conservatives seem unlikely to budge from that position. In the case of Thomas, like Ginsburg, this can be explained in terms of formalist methodology. In the case of Roberts and Alito, two of the more pragmatic justices, that decision is harder to square other than in terms of ideological preferences.

Prediction: 5:4 for Petitioner

For Petitioner (Garza): Breyer, Kavanaugh, Gorsuch, Sotomayor, Kagan

For Respondent (Idaho): Alito, Ginsburg, Roberts, Thomas

Graveyard Takings Case Rises From the Dead For Reargument

Knick v. Township of Scott, Take #2

When the Supreme Court called for re-argument in Knick v. Township of Scott—on the metaphysical constitutional law question of exactly when a takings claim comes into being—it seemed likely that the Court had been set to split 4:4. The re-argument was presumed to be largely for the benefit of giving newly minted Justice Kavanaugh the power to cast the decisive vote.

A Clarifying Re-Argument

The re-argument in Knick v. Township of Scott was clarifying. So much so that we wonder whether every case should be argued twice! In our initial analysis of Knick, we predicted a 4:4 split, but there was considerable ambiguity about the position of a number of the justices, as the figure below replicates.

Supreme Court Oral Argument Prediction for Knick v. Township of Scott (October 13, 2018).
Our earlier Supreme Court Oral Argument Prediction for Knick v. Township of Scott based on the oral argument of October 13, 2018.

As we discuss below, the advocacy in Knick II was much better, to the point where even though we were still unconvinced by Petitioner’s advocate, J. David Bremer, we at least understood what he was trying to say this time. In addition to the advocates being clearer, the positions of the justices seemed to clarify a great deal.

Our Supreme Court Oral Argument Prediction for Knick v. Township of Scott based on the reargument of January 16, 2019.
Our Supreme Court Oral Argument Prediction for Knick v. Township of Scott based on the reargument of January 16, 2019.

Not surprisingly for a case pitting individual property rights against government regulation, our model of Knick II suggests a mostly liberal-conservative split. However, there are some interesting shifts from the first argument.

In particular, Justice Breyer has moved from the 50/50 column to the clearly pro-Respondent column, and Chief Justice Roberts appears to have switched sides entirely. Also, Justice Alito’s strong support for the Petitioner was clearer on the numbers in the second argument, although, as we had mentioned in our previous post, it was always clear in substance.

Despite his apparent switch, we are not sure we believe the model’s prediction for Roberts. He seemed to be asking pointed questions of Petitioner initially, but after some tough questioning of Breemer by Breyer, Roberts — like Alito in the initial argument — jumped in to help the advocate out by asking friendly “questions” and leading the discussion back on track.

We think Roberts is likely genuinely conflicted over this case. As the head of the federal judiciary, he has little interest in flooding the federal courts with state takings claims where federal judges will have to opine on state property law, yet he is philosophically inclined to a pro-property and anti-regulation position. We don’t expect the Chief to be as concerned for the “dignity and sovereignty of the States” in this case as he was in Shelby County v. Holder (2013). But then again, why a federalist would caste a cynical eye over the congressional record that repeatedly reauthorized the coverage formula for the Voting Rights Act, but then want to remove every mundane question of state property law to the federal courts, is hard to fathom.

Based on the substance of his questions and comments, we also think that the model may be overstating Breyer’s pro-Respondent leaning. Like Roberts, but for different reasons, Breyer was also conflicted about the issues presented in this case. Breyer was concerned for individuals caught in a Catch-22 situation created by the Court’s own precedent in Williamson County. Under Williamson County, the property owner who believes that a state regulation intruded upon her rights to such an extent that it constitutes a “regulatory taking” has no claim for the violation of her rights until she has pursued a claim for compensation in state court and been denied. This makes sense because the Takings Clause is violated not by takings as such, but by takings without just compensation. However, Williamson County creates a Catch-22 for plaintiffs because once they have argued their case in state court, the doctrine of issue preclusion prevents them from re-litigating the same takings question in federal court.

This Catch-22 obviously does not sit well with Breyer, but like Roberts, he is a pragmatist little attracted to the potential flood of premature local and state takings claims inundating the federal courts.

When this case was scheduled for reargument, it seemed inevitable that Justice Kavanaugh would cast the deciding vote. In theory, an ultra-conservative, Federalist Society approved justice would be expected to reflexively side with the property owner in a case like this, but Kavanaugh’s position was a little hard to read. When Justice Gorsuch was in a similar position in the Janus case, he said nothing. In Knick II, Justice Kavanaugh had quite a lot to say, but he seemed to have issues with both sides of the argument. Our model, estimates that he is 57% likely to vote for Respondent, but that is basically a toss-up.

Our Prediction

Justice Breyer spent much of the argument searching for a reasonable middle ground that would address the Williamson County Catch-22 without making the federal courts the first stop for every takings challenge to state and local government property regulations. We think that Breyer will hold that a takings claim is complete for the purposes of Section 1983 when a regulation goes into effect without a reasonable mechanism to determine whether compensation is owed, and that somehow the usual rules of issue preclusion will not apply to plaintiffs who exhaust their state remedies. How many votes this opinion will attract is far from certain, but we do think that the middle position of leaving Williamson County intact with some softening on issue preclusion might attract the votes of Justices Kagan, Sotomayor, Ginsburg, and just possibly Roberts and Kavanaugh. Alternatively, we could see multiple opinions. Justices Alito, Gorsuch, and Thomas are a lock for Petitioner.

Advocate Performance

As a final note, the next figure shows the cumulative word count for the justices as a group and for each advocate.

Cumulative word count in Knick v. Township of Scott (January 16, 2019).
Cumulative word count in Knick v. Township of Scott (January 16, 2019)

The comparison between Petitioner’s advocate and the Solicitor General who was also on the side of the Petitioner is quite telling. As we discussed in a previous post, the Court tends to give the Solicitor General and even state solicitors general much more deference than regular advocates. In this case, despite his improvement from Knick I, Breemer still struggled to get a few words in between the justices. In contrast, Gen. Francisco, who was exceptional in both arguments, was allowed to talk for the majority of his allotted 10 minutes. Sachs for Respondent was also very good in the second argument and seemed to get a reasonable chance to make the points she needed to make.

Prediction:

For Petitioner Knick: Alito, Gorsuch, Thomas, Kavanaugh

For Respondent Township of Scott: Breyer, Ginsburg, Kagan, Roberts, Sotomayor

Most likely to switch: Kavanaugh, Roberts

Prohibition, pragmatism, and protectionism

Predicting Tennessee Wine and Spirits Retailers Association v. Blair

Wine retailers seeking to operate in Tennessee argued last week that the state’s highly restrictive condition requiring 10 years of prior residence effectively discriminates against out-of-state retailers. The case, Tennessee Wine and Spirits Retailers Association v. Blair, seems quite simple to us: the Court in Granholm v. Heald addressed the almost identical question in 2005, finding in favor of out-of-state wine producers on the basis of the Dormant Commerce Clause. Since then, two thirds of the personnel of the Court has changed and yet we see similar patterns emerging in the oral argument in Tennessee Wine and Spirits.

In Granholm the Court split 5:4, but not along the usual ideological lines. In the majority were Kennedy, Scalia, Souter, Ginsburg, and Breyer; the dissenters were Stevens, O’Connor, Thomas, and Rehnquist — the dissent encompassed both extremes of the Court. The numbers on the current Court look similar, as the following figure shows.

Supreme Court Oral Argument Prediction for Tennessee Wine and Spirits Retailers Association v. Blair (January 16, 2019)

That Alito and Sotomayor were closely aligned in this case was apparent not just from the numbers but also from the content of the argument. Early in the argument, Alito stepped in to buttress Sotomayor’s arguments and questions of Petitioner, a retailers’ association suing to prevent the state agency from granting licenses to the out-of-state retailers.  Alito asked “what is the — the basis for thinking that the purpose of or a purpose of Section 2 of the Twenty-First Amendment was to authorize the states in this one area, dealing with alcohol, to engage in protectionist activities that wouldn’t be permitted with respect to any other commodity?”

Despite the predictions of our model, we think that Kavanaugh and Kagan are in agreement in this case. Kavanaugh argued persuasively that the Twenty-First Amendment gave states and localities the power to remain dry, but not to discriminate. Borrowing Kagan’s phrasing, he pointed out that otherwise, “the sky is the limit” in permitted discrimination.

We don’t predict another 5:4 ruling in the Tennessee case. Gorsuch, Breyer, and Kagan (and very briefly, Roberts) all raised concerns regarding the slippery slope of finding for Respondents here, asking what is to stop the Amazon of wine retailers from operating and making the Twenty-First Amendment redundant? But Kagan and Breyer also made similar arguments in challenge to Petitioner, querying what is to stop more and more protectionism if this level of discrimination is permitted. Breyer asked “Suppose you  [require that a]ny liquor store has to use paint made in Tennessee, asphalt made in Tennessee for the parking lot, neon — you know, I can go on.”

Breyer also wrung his hands about not wanting to upset the history of extensive permitted state regulation of these matters, but Granholm also gave him a chance to let these and the pragmatic concerns sway him, and he took the other side.

We think that Kagan and Breyer will rule in favor of Respondent, and only Gorsuch and Thomas (who said nothing, just for a change) are likely to be swayed by Petitioner’s argument, even though Gorsuch is said to be a formalist.

As we have discussed elsewhere, our predictive model of oral argument works better in divided cases than in unanimous or highly lopsided decisions. If the justices mostly agree on the merits of one side, at least a few of them are going fill the vacuum and ask questions of the side they favor.

The clearest sign of how unbalanced we expect this case to be comes not from the justices, but from the advocates. By our calculations, Carter Phillips is the fifth most experienced advocate in the modern era of the Court. He sat down with 10 minutes to spare, after giving a quick conclusion and asking if the justices had any further questions. We do not believe Phillips ran out of arguments; his early exit was a sign of supreme confidence. We trust Phillips’ expert judgment and think those missing 10 minutes are the strongest signal of the likely outcome of the case.

To further examine the back and forth between the justices and the three advocates — the other two being Shay Dvoretzky for Petitioner and Illinois Solicitor General David L. Franklin as amici curiae representing multiple states in support of Petitioner — the next figure shows the cumulative speech episodes for each advocate (in various shades of blue on the bottom) and the justices as a bloc (in orange on top).

The figure shows the Cumulative wordcount in Tennessee Wine and Spirits Retailers Association v. Blair in three segments, petitioner, amici and respondent.
Cumulative wordcount in Tennessee Wine and Spirits Retailers Association v. Blair (January 16, 2019)

Phillips and Dvoretzky spoke for roughly the same amount of time, around 20 minutes each, and had a similar ratio of words spoken by them versus the justices. The far broader region of the graph for Phillips indicates the much more rapid back-and-forth between the justices and that advocate. Phillips was effectively controlling the argument, in a way that Franklin, and particularly Dvoretzky did not. For instance, when Gorsuch began his less-than-concise invitation to give Phillips “the opportunity” and “one final shot” to convince him on the history, Phillips cut him off effectively but not rudely.

Here is the interaction:

Neil Gorsuch: . . . And I — I just want to give you one more shot —
Carter G. Phillips: Sure.
Neil Gorsuch: — at the — at the history —
Carter G. Phillips: Yeah.
Neil Gorsuch: — and dealing with the Wilson Act and Webb Act and —
Carter G. Phillips: Thank you, Justice Gorsuch.
Neil Gorsuch: — those sorts of things.
Carter G. Phillips: Appreciate the opportunity.

Phillips then went on to give a substantive answer.

In terms of deference given to the advocates, the first few minutes of Dvoretzky’s time were dominated by Sotomayor and Alito, not by the advocate. In contrast, both Franklin and Phillips were able to give extensive introductions before the justices jumped in. Solicitors General get more deference from the justices, and in this case at least, that was also extended to the state Solicitor General, for whom the ratio between advocate time and justice speaking time was the most favorable of the three. We do not expect that this deference will amount to success, Franklin played a poor hand as well as he could, we just note it for interest.

Prediction: 7:2 for Respondent (Blair, Interim Director of the Tennessee Alcoholic Beverage Commission, et al.)

For Respondent: Alito, Sotomayor, Kavanaugh, Roberts, Ginsburg (absent from the argument), Kagan, and Breyer

For Petitioner: Gorsuch and Thomas

Most likely the switch: Roberts

How SCOTUS argument transcripts and recordings became widely available

The data that makes our analysis possible is a resource that had to be created and nurtured

All of the empirical analysis at ScotusOA.com is based on the SCOTUS transcripts. We constructed a dataset drawn from the text of every Supreme Court oral argument from 1955 to 2017, based on 1.7 million speech events by justices and advocates. All of this came from Oyez.com. The Supreme Court makes some of these transcripts available but not all—only 15 Terms of the six decades that Oyez provides—and many are not fully identified as to who the speaker is. Oyez developed a program to identify the speaker, went back through the transcripts to 1955, and continues to improve the quality of the transcripts, checking for accuracy. How all of that data became available is an interesting story.

The history of the development of this invaluable resource is an interesting tale of the somewhat obsessive focus one person, Jerry Goldman, on what many thought was a very obtuse idea for a long time, but is now recognized as a great legal resource. It is also the story of the sometimes-odd interaction between the very staid institution of the Supreme Court and quickly changing technology.

Humble beginnings as a set of HyperCards

Before all else, Alderson Reporting Service created transcripts of SCOTUS arguments. These documents found their way to Lexis and Westlaw beginning in about 1980, but they did not identify the individual justices and they were text-only resources. Sometime in the 1980s, Jerry Goldman, then a Professor of Political Science at Northwestern University specializing in law and courts, had the idea—during a Chicago Cubs baseball game—to create a digital Court resource much like baseball cards. Jerry’s idea was that instead of photos of favorite players, their biographical information, and fun sporting trivia, Court nerds could enjoy the same but with justices and Supreme Court cases. Initially, the project was called “The Hitchhiker’s Guide to the Supreme Court.”

The Internet was yet to be, so Jerry set up the cards using HyperCard, a software tool that stacked and interconnected digital cards, with text, images, and other simple concepts. The Hitchhiker’s SC Guide used bitmapped, black-and-white images of the justices and gave information on the cases and the Supreme Court building.

In the early 1990s, Jerry made came to an agreement with the National Archives that allowed him to use their audio recordings of Supreme Court oral arguments, so as long as they would not be used commercially. In 1994, with the help of Joe Germuska, self-designated “Chief Nerd” of Northwestern’s Knight Lab, Jerry created a small prototype of a digital audio workstation that would eventually allow him to make the Supreme Court audio accessible. He received a small grant from NUL, and a subsequent NSF grant in 1995-1996, to build the equipment, and “OYEZ, OYEZ, OYEZ” was created. Another grant in 1996 from the NEH’s Teaching with Technology program allowed Jerry to slowly add more cases, using simple technology that involved cutting and pasting when updating changes to materials. Although the site only had about 100 select cases—all significant constitutional law cases—the server was struggling with the unwieldy technology, so Jerry and Joe decided to create a database of case records and justice records, with the audio on a designated audio server.

The Supreme Court creates its own monster

In 1996, at the same time that Jerry was developing what became Oyez.com, another resource, “May it Please the Court” launched, in which 23 edited versions of the oral arguments, with commentary, were made available for purchase on cassettes. That project gained enormous media attention because the Supreme Court threatened to sue the developer for using the case audio contrary to the arrangement with the National Archives. Ultimately, the Supreme Court gave in and actually changed its policy, permitting broader use of the recordings. The controversy sparked the media’s interest in using audio recordings of oral argument as part of their general news coverage, and that is when the use of Supreme Court oral argument audio really took off.

That appetite for Supreme Court audio hit another high in 2000, in response to Bush v. Gore. The prospect of the Supreme Court deciding a presidential election led to a sudden and enormous demand for a live feed of the two cases (Bush v. Palm Beach County Canvassing Board and Bush v. Gore). Knowing that the Court would be very resistant to a live video or audio feed, Jerry wrote to Chief Justice Rehnquist on November 25 to try to convince him to release the audio the same day as the argument, as a substitute for a live feed. As seen in the fax reproduced below, Rehnquist wrote back, agreeing to release the recordings in these exceptional cases on the same day as the argument. The remaining cases, orders etc. still had to wait until the end of Term for release to the National Archives. Eventually, the Court adopted the end of the week release of all audio, with occasional same-day release for other highly salient arguments, such as the Obamacare arguments and some of the gay marriage cases.

Justice Rehnquist's letter to Jerry Goldman regarding Bush v. Gore oral argument (described in the text)
Justice Rehnquist’s letter to Jerry Goldman regarding Bush v. Gore oral argument

Now, a multi-million user interface

Gradually, with some assistance, Jerry went backwards through the cases, collecting earlier arguments. Leveraging various grants, and collaborating with several scholars, by 2003, the collection extended back to 1980, and by 2014, the Oyez collection was made comprehensive back to October 1955. Today, Oyez also includes not only oral arguments but opinion announcements and speaker-identified transcripts. Jerry says: “The metadata for every case—including our abstracts—remains a work in progress but every case that is in Spaeth is represented in Oyez plus a substantial number of pre-Spaeth (i.e., pre-1953) cases.” Along the way, SCOTUS phased out its reel to reel recordings and went digital in 2004, and Jerry had created a standardized process for making the data available quickly.

Since 2010, Jerry reports that website traffic has increased at approximately 10% to 15% per year. Traffic has been growing particularly strongly in the last two years, going from 3.3 million users (in 6.9 million sessions) in the November 2011 to November 2012 period, to 7.2 million users (in 14.2 million sessions) over the same period in 2017-2018. That is a lot of people who owe a debt to Jerry for pursuing his vision with such determination.

The Supreme Court and technology: A sometimes fraught relationship

Jerry defied traditional academic incentives and has made a lasting contribution as a result. Jerry was advised by colleagues not to build a resource that he would get little credit for and would take away time from his research. Fortunately for those of us who enjoy not only empirical analysis such as that done at ScotusOA, but those who value timely analysis of the arguments on ScotusBlog and other forums, Jerry was stubborn enough to continue.

The history of Oyez.com shows how technological change can bring about institutional change. The Supreme Court has resisted the call for television access for decades, and it is clearly going to continue to do so. But the development of the Internet created the demand for a different type of access, one that was harder to resist. Once it became clear through Jerry’s efforts that online access to Supreme Court transcripts and audio was possible, a new constituency developed around the demand for such access. The Court’s reasons for rejecting televisions in the courtroom did not seem to apply to this new medium. The Court had to either develop new reasons for why it would not make transcripts and audio available, or else had to concede the point. In the end, the Chief chose transparency, but the Court may not have anticipated all of the implications of that move.

In particular, it probably never occurred to the members of the Court that some day social scientists would be using text data mining to place their conduct at oral argument under the microscope. The justices may not relish empirical findings such as that they favor one side over another at oral argument or that some justices’ votes are highly predictable based on whom they interrupt. The justices may have also not appreciated that their behavior at oral argument was already becoming more performative (partly due to increasing political polarization) and that the increased transparency brought by Oyez might accelerate this trend.

Lawyers, law students, journalists, and academics who study the courts all benefit from Jerry’s work in founding Oyez.com. It is hard to think of one person who has done more to make an aspect of the Supreme Court’s decision-making process more transparent, available, and accessible to the broader public.

This post is based in part on a telephone interviewwith Jerry Goldman on November 13, 2018, and follow-up emails.

The Supreme Court eats its copyright vegetables

Listening to Tuesday’s argument in Fourth Estate Public Benefit Corp. v. Wall-Street.com, it was striking that most of the justices would rather have been somewhere else, or at least deciding some other case. Maybe a trademark a case (see below)? Ironically, the justice with the most interest in the rights of copyright owners, Justice Ginsburg, was absent from oral argument recovering from surgery. But sometimes, even Supreme Court justices need to eat their vegetables.

Fourth Estate is exactly the kind of case that the Court should take. True enough, the policy issues are not earth shattering, the statutory interpretation issues are a little dull, and the controversy is so copyright-specific that it has no real implications for other areas of law. However, it is ridiculous in a national copyright system that the Fifth and Ninth Circuits allow copyright claimants to file a lawsuit based merely on filing an application for registration, whereas the other circuits require an actual registration or a rejection thereof (i.e., a decision on whether the work is copyrightable). This is a significant difference because the process takes several months on average.

Text likely to win over policy

Oral argument saw Petitioner’s flimsy statutory interpretation and more sympathetic policy position pitted against Respondent’s strong textual argument and less compelling policy stance (discussed in more detail below). If the justices were voting on their policy views alone, this case would probably be 9:0 or 8:1 in favor of Petitioner, but Respondent will be hoping that the law still matters. The tension between law and policy resulted in mostly even handed questioning at oral argument; as a consequence any predictions based on transcript metrics are quite speculative.

Supreme Court Oral Argument Prediction for Fourth Estate Public Benefit Corp. v. Wall-Street.com,  January 8, 2019) (summarized in the text)
Supreme Court Oral Argument Prediction for Fourth Estate Public Benefit Corp. v. Wall-Street.com, January 8, 2019)

As the figure above shows, we predict Justices Kagan and Sotomayor, as well as the Chief Justice, to vote with Respondent. Justices Thomas and Alito (both silent during oral argument) will probably join them, as will Justice Breyer. A three-justice minority would not be surprising (based on the argument and Ginsburg’s history of favoring copyright owner interests), but the outcome is more likely to be 9:0 for Respondent (perhaps 8:1 with Ginsburg dissenting).

Ultimately, we predict that in this case at least, a straightforward reading of statute will carry the day.

How clear is the textual argument?

The issue in Fourth Estate is the correct reading of Section 411(a) of the Copyright Act, entitled “Registration and civil infringement actions.”

In simple terms, the most natural reading of the section is that it bars a copyright owner from instituting an infringement action until the Register of Copyrights (i.e., the Copyright Office) has either approved or refused registration. The relevant text of the section consists of three sentences. The first sentence reads:

… no civil action for infringement of the copyright in any United States work shall be instituted until preregistration or registration of the copyright claim has been made in accordance with this title.

The second sentence provides that if “the deposit, application, and fee required for registration have been delivered to the Copyright Office in proper form and registration has been refused” the applicant may then institute its civil action. The third sentence allows the Register of Copyrights to join in that litigation and defend its refusal.

Petitioner argued heroically that “registration … has been made” means simply that the copyright claimant has submitted an application in proper form to the Copyright Office. This is hard to square with the second sentence that talks about a registration having been refused. Basically, Petitioner wants “registration … has been made” to mean the exact same thing as the following 18 words from the second sentence: “the deposit, application, and fee required for registration have been delivered to the Copyright Office in proper form.” This requires the Court to treat the same word, registration, as meaning vastly different things from one sentence to the next; it also asks it to accept that vastly different expressions within the same section mean the same thing. This is both counter-intuitive and anti-canon.

The policy question is not clear-cut

The policy question in Fourth Estate is more finely balanced. Petitioner’s best argument was that being forced to wait several months for a registration to either be granted or denied puts the value of their copyrights at risk and denies them the chance to take swift injunctive action. This clearly scored some points with a few of the justices, particularly Justice Kavanaugh, who demanded a drawn out explanation.

However, Respondent and the Deputy Solicitor General did a good job explaining why such dramatic unfairness was unlikely in view of the availability of preregistration and expedited review (special handling). They also explained how registration as a precondition to litigation plays an important role in encouraging timely registration with the Copyright Office and deposit with the Library of Congress.

Copyright and trademark are not the same

For intellectual property lawyers who are bemused, enraged, or amused by the inability of non-specialists to understand the difference between copyright, trademark, and patent (it is like confusing Star Wars with Star Trek, or Lost In Space), the oral argument provided a couple of triggering exchanges:

John G. Roberts, Jr.
Well, that’s enough assuming that the registrar has registered the mark.
Aaron M. Panner
Again, the registrar does not have to register the mark.
The — the — not the mark, the copyright. …

John G. Roberts, Jr.
  So you could go back, the registrar hasn’t even registered the mark, and you can go into court and say, hey, I get the benefits of having registered my mark?
Aaron M. Panner
  The copyright claim, yes, Your Honor.

Prediction

Petitioner: Sotomayor, Kagan, Roberts, Breyer, Thomas, Alito, Gorsuch, Kavanaugh, Ginsburg
Respondent: none
Most likely to switch: Ginsburg, Kavanaugh, Gorsuch

The emerging formalist-pragmatist divide on the Supreme Court (Biestek v. Berryhill forecast)

Unheralded case may be emblematic of the Supreme Court’s 2018 Term

Based on the ebb and flow of oral argument in Biestek v. Berryhill, we predict a formalist-pragmatist division between the justices, rather than one based upon the usual liberal-conservative fault lines.

Biestek was a relatively unheralded and unremarkable case about administrative procedure and the denial of disability benefits. The hearing generated very little press coverage and may have even seemed a little dull to spectators without a keen interest in the outcome or in administrative law. However, for this very reason, Biestek may end up being emblematic of the 2018 Term, a Term in which the Court appears to be trying to stay out of the limelight and avoid the impression that every case is a pre-determined political contest.

The results of our predictive model based on the oral argument transcript are shown in the figure below. The model captures the formalist-pragmatist division as reflected on the opposing views of conservative Justice Gorsuch and equally conservative Justice Alito.

Supreme Court Oral Argument Prediction for Biestek v. Berryhill (December 4, 2018)

However, the model does not capture every nuance of the argument.

“I have evidence, but it’s a secret”

Some additional background helps explain the issues. Biestek applied to the Social Security Administration for Supplemental Social Security Income and benefits in 2010. In a hearing to review the denial of his claim, a vocational expert testified that there were jobs available to Biestek in spite of his disability. The vocational expert based her testimony on private data that she refused to produce. Nonetheless, based on this secret evidence, the Administrative Law Judge (“ALJ”) found there was work available to Biestek and denied his claim.

The Sixth Circuit affirmed on the basis that the strict requirements of the Federal Rules of Evidence do not apply to Social Security disability proceedings. Petitioner acknowledges that the Federal Rules do not apply, but contends that the ALJ’s decision failed the substantial evidence threshold. Thus, the essential question in the case is whether a man should be denied disability benefits when the expert evidence against him is based solely on data that the expert declines to reveal.

Or, as Justice Gorsuch put it: 

If we were in federal district court [and] if on the key issue in the case the evidence depended upon the testimony of an expert, and the expert said, ah, I’m not going to give you my underlying data, it’s secret, I don’t think we would hesitate to find that no rational jury could sustain a verdict in favor of the party propounding that expert. Why isn’t the same true here?

The arguments against the Petitioner are largely pragmatic. As Anthony Yang explained on behalf of the government:

Each year, there are about 2.6 million initial disability claims that are filed with SSA, and at the third level of review, the SSA conducts 670,000 hearings. That’s about 200 — 2500 a day. Over 1 million people are waiting for just a response for their hearing, and they wait, on average, about 605 days. There is no adjudicatory process on a scale comparable to this.

The Chief Justice and Justice Alito did not seem convinced that the secret basis of the vocational expert’s opinion was any great problem. Justice Breyer, on the other hand, had clear reservations about the reliability of the testimony of vocational experts but thought that this issue would be better addressed by a test case of some kind.  

Prediction

The case is hard to predict. Justices Gorsuch, Ginsburg, and Sotomayor seemed clearly inclined to the view that secret evidence was no basis for an administrative decision. However, Justice Kagan raised telling objections on both sides of the argument and Justice Kavanaugh’s limited comments to the Respondent did not contribute much.  

Prediction:

Petitioner: Gorsuch, Ginsburg, Sotomayor, Kavanaugh, Kagan

Respondent: Alito, Roberts, Breyer, Thomas

Most likely to switch: Kavanaugh, Kagan, Breyer

Timbs forecast: Many roads to incorporation

In Timbs v. Indiana, the Court saw a particularly lopsided oral argument in favor of Petitioner, who argued that the  Eighth Amendment’s excessive fines clause is incorporated against the states under the Due Process Clause of the Fourteenth Amendment (or alternatively, the Privileges and Immunities Clause).

The Supreme Court continues to ignore pressing social issues

There are many reasons why Petitioner Tyson Timbs is a sympathetic litigant, and why his constitutional argument is persuasive. He is the prototypical victim of the opioid epidemic: having become addicted to hydrocodone for foot pain, he took to dealing small amounts of heroin to fund his habit. He used his one considerable asset, a $42,000 SUV bought with money inherited from his father’s $73,000 estate, to drive to the sites of two deals of value $225 and $160. Timbs was convicted and put under house arrest, and subject to fines totaling $1,200. But the state of Indiana seized his $42,000 SUV as an instrumentality of the crime, subject to in rem forfeiture.

According to Indiana, there is no meaningful limit to stop police directly profiting from seizing cars used in drug crimes, or even cars going 3 miles over the speed limit. According to Timbs, the constitutional guarantee against excessive fines should be incorporated against the states because freedom from excessive fines is “fundamental to our scheme of ordered liberty.” This constitutional protection is vital with respect to fines because, unlike other forms of punishment, fines are particularly prone to abuse as a potential profit center for the state. Furthermore, as Timbs’ brief argues persuasively, such forfeitures have been used for centuries to oppress the poor, perpetuate Jim Crow long after the Civil War, and give problematic police departments, such as that in Ferguson, Missouri, a mechanism of harassing racial minorities.

Yet, as is so often the case in criminal procedure cases, few of these compelling arguments appeared to shape the justices’ inquiries in the case. Only Justice Sotomayor made passing reference to these pressing social concerns.

The real point of interest in the oral argument was how justices from various points on the ideological spectrum seemed to be converging on a decision in favor of Petitioner while maintaining starkly different motivations and reasoning. It was particularly fascinating to see that the differences among the conservative justices and among the liberal justices were more sharply drawn than the distance between the conservative and liberal camps.

A variety of ways of getting to a new state constitutional right

Among the conservatives, Justice Gorsuch and Justice Kavanaugh seemed to be competing as to who would come out swinging most against the state. Kavanaugh repeatedly asked a version of the question “aren’t all –all the Bill of Rights at this point . . . incorporated?” Gorsuch clearly won, however, immediately jumping into a 138 word monologue, demanding that Respondent advocate “at least agree” that the Excessive Fines Clause is incorporated against the states, and subsequently goading the advocate, asking if he “really wants” the Court to address a secondary question.

In contrast, Chief Justice Roberts and Justice Alito pushed hard against Petitioner’s argument, questioning whether any fine can be excessive in comparison to prison time. The figure below reflects this division, with Roberts and Alito favoring Respondent. However, there is reason to think that they may not vote that way: for institutional reasons, the Chief is unlikely to want to be on the losing side of a case recognizing a constitutional right. Also, by the end of the argument even Alito seemed troubled by the implication of Respondent’s position that individuals pursuing Second Amendment rights would have to establish each aspect of the right both at the federal and state level.

ScotusOA Predictive Model for Timbs v. Indiana

Model predictions for Timbs versus Indiana showing votes likely in favor of Petitioner and respondent.  Ranging from Gorsuch 95% Petitioner to Roberts 93% Respondent.
Supreme Court Oral Argument Prediction for Timbs v. Indiana (November 28, 2018)

There was also division among the liberal justices, but over the reasons for and extent of support for Petitioner. As mentioned, Sotomayor did consider the practical reality of civil and criminal forfeiture; Justice Kagan, in contrast, repeatedly raised a question of great interest only to constitutional scholars: at what level of generality should a constitutional right be recognized? Justice Breyer raised strong institutional concerns with Petitioner’s position, particularly regarding the role of stare decisis and the logical conundrum of protecting against excessive fines given the lack of a meaningful proportionality requirement in sentencing. In contrast, Justice Ginsburg so clearly came to the aid of Petitioner’s advocate, Wesley Hottot, that he subsequently committed the faux pas of referring to her as having “assisted” him earlier. Based solely on the oral argument, one might think Breyer was a vote for Respondent, but although he was clearly weighing some broader institutional issues, our reading of Breyer’s constitutional theory leads us to the opposite conclusion. Breyer’s congenial verbosity makes him harder to predict than many of the other justices.

Unanimous and lopsided votes are difficult to predict

Thus, while the numbers show a 5:3 split in favor of Petitioner, we expect that the case will be much more lopsided, perhaps even unanimous, since even ultra-conservative, ever-silent Justice Thomas called for the Court to hear the issue. Very lopsided and especially unanimous cases are harder to predict on the numbers from oral argument: even if the justices all expect to vote in favor of one side, we would still expect them to talk during that advocate’s time, since modern norms of Supreme Court oral argument involve a lot of justice speech. This makes the numbers less reliable in lopsided cases, showing the importance of both qualitative and quantitative analysis. We do, however think that the figure reflects a real division between the justices, particularly between the Chief and Alito versus the rest of the Court. This division just may not be reflected in the final votes. We expect that either Roberts and/or Alito will vote for Petitioner on the question presented but write a narrowing concurrence.

Prediction: 8:1 or 9:0 for Petitioner, Timbs

Most likely to switch: Alito

Visualizing the content of justices’ and advocates’ cross-references

In our previous two posts (this one and this one), we examined the implications of justices cross-referencing other justices at oral argument. In this post, we take the analysis a step further and focus on the content of those cross-references.

The figure below is a word cloud derived from every sentence in the Supreme Court oral argument transcripts from the 1985 to the 2017 Terms in which one justice refers to another sitting justice by name. (We also treated references to “the chief” and “the Chief Justice” as a reference to Rehnquist or Roberts, depending on the era.) The word cloud provides a visualization of the relative importance of the 100 most frequent words in that set, excluding common words such as and, the, it, to, …

Supreme Court Oral Argument, 1985 – 2017 Terms: Justice cross-reference word cloud 

Word clouds are the text mining equivalent of a Rorschach test. The frequency of each individual word in a “bag of words” is of course an objective fact; but the significance each viewer gives to that fact depends on his or her expectations. Our interpretation of the word cloud above is that when justices are referring back to other justices in the modern era, they are overwhelmingly attempting to go back to a specific question raised by that justice. “Question” is the dominant token in this word cloud, occurring more than twice as often as the next two tokens, “answer” and “think.” Likewise, “answer” suggests a reference is being made to an earlier question; “think” is more generic.


Supreme Court Oral Argument, 1985 – 2017 Terms: Advocate-justice cross-reference word cloud 

Looking at the same visualization for advocates in the above word cloud, we see questions are still important, but not so clearly dominant. In the advocate word cloud, “question” is only the third ranked token, after “think” and “court.”

Previously, we have shown that the modern era of oral argument is defined more by comments than by questions; we have also shown that questions are more often asked of one’s friends, and comments are put to one’s foes. Finally, we have shown that judicial cross-references are generally made between two justices who ultimately agree with one another in the case at hand. As such, it follows logically that justices’ cross-references tend to focus substantially on fellow justices’ questions. This content analysis further buttresses this prior evidence of the relationship between questions and agreement, comments and disagreement, and judicial cross-references and agreement.