Supreme Court Justices routinely answer factual questions about the world – such as whether violent video games have a harmful effect on child brain development or whether a partial birth abortion is ever medically necessary. The traditional view is that these findings – often called “legislative facts” – are informed through the adversary system: by reviewing evidence on the record and briefs on appeal. Routinely, however, the Justices also engage in what I call “in house” fact finding. They independently look beyond the briefs and record to answer general questions of fact, and they rely on their discoveries as authorities. To be sure, judges have always done this. We have all heard the stories of Justice Blackmun holed up in the medical library at the Mayo Clinic during the summer of 1972 studying abortion procedures. And the Federal Rules of Evidence contain no rule restricting it; the rule about judicial notice specifically exempts legislative facts from its scope.
But times have changed. The world has recently undergone a massive revolution in the way it receives and evaluates information. No longer do Justices need to trek to the library to look up factual questions. Instead they can access virtually infinite amounts of factual information at the click of a mouse. If the Justices want more empirical support for a factual dimension of their argument, they can find it easily and without the help of anyone outside of the Supreme Court building.
My article, Confronting Supreme Court Fact Finding, (available here) discusses how that change in technology has and will affect the Court’s fact-finding practice. It collects over one hundred examples of factual authorities on which recent decisions of the U.S. Supreme Court relied that were found “in house” – i.e., that cannot be found in any of the party briefs, amici briefs, or the joint record. These are not insignificant rarities: almost sixty percent of the most important Court opinions in the last ten years (as identified by political scientists) rely on in-house research at least once. Virtually all of the Justices do this, and they do it on a variety of topics using a variety of different sources – from newspapers to law review articles to statistics culled from an advocacy group’s website. As I describe in my article, sometimes these authorities are used rhetorically – to show practical consequences of a different outcome, for example, or to emphasize the emerging significance of an issue. But other times these authorities are marshaled to support dispositive factual assertions – for example, that African-American children in integrated schools are more successful, or that walking the course is an essential component of the game of golf.
So what do we make of the observation that Justices need not rely on the adversary method to answer factual questions or mount factual dimensions of their arguments anymore? Assuming – as I think it is safe to do – that the digital revolution and dramatic change in the way we all access information means in-house fact finding will only increase over time, should it make any difference to our normative reactions about the process for this brand of judicial decision-making?
My answer – not surprisingly – is yes. The world looks very different from the way it looked when legislative facts were exempted from the scope of judicial notice rules in 1975. Of course there are benefits to letting judges research freely in a new digital age. Judges presumably make better decisions when they know more. But there are also troubling effects that accompany a robust practice of in-house judicial fact finding today. This article addresses three of them: (1) the systematic introduction of bias; (2) the possibility of mistake; and (3) concerns about notice and legitimacy. Certainly these risks have always existed when judges look outside the record on questions of fact, but the dangers are more potent in a world where information is easily accessed and freely traded.
First, in terms of bias, to “google” something is now common parlance and common practice for looking up an unknown fact. But internet searches like these may not present results in a neutral fashion. Since web companies – like Google – can gather vast amounts of information about their users and because they try to tailor services to our personal tastes, some claim the search engines filter the results depending on the searcher. A search for statistics on fatalities in police chases, or a search for the physiological effect of the chemicals used in lethal injection could produce different results for different chambers depending on, for example, the internet history (or Facebook profile!) of the users. The end result is worse than a Justice purposely finding something to cite that supports what she wants to argue; it is that she will only find factual authorities to support what it is she wants to argue. This opens a real possibility for the systemic introduction of bias – unrealized bias – into assertions of fact in judicial opinions.
As is true with the risk of incorporating bias into judicial opinions, the risk of factual mistakes is also one that is exacerbated by new modes of digital research. Not only is data easier to find thanks to the internet, but – importantly – it is also easier to post. Some factual information on the internet is trust-worthy, but some of it is not; and discerning the difference is not always easy. Moreover, Justices – like all of us – have a tendency to engage in “motivated reasoning” and to look for facts that support the argument they are building, wherever those facts may come from, and despite what other opposing authority is out there. This tendency may encourage the ad hoc and potentially mistaken evaluation of scientific findings – looking for what one wants to see – particularly if the studies to be used as authorities were never tested by the adversarial method or addressed by experts below. Couple this reality with the new instant ability to find facts to support almost anything, and confidence in judicial fact finding outside their areas of expertise diminishes significantly.
Finally, even if Justices are good at independent research in the digital age and even if worries about mistakes and bias are overblown, there remains a basic question of fairness. This concern has two components: (1) the short-term fairness question with respect to the parties; and (2) the long-term fairness question about the legitimacy of Supreme Court decisions generally. With respect to the parties, in-house fact finding threatens democratic values that accompany participation in the adversary system: the parties are denied the chance to weigh in and address potentially biased or mistaken factual authorities. In addition, as others have argued, there are good institutional reasons why legislatures and agencies – not courts – are typically given the responsibility of investigating facts on their own. Putting aside the practical reasons for this division of labor, there remains a question of whether judicially found factual authorities are legitimate in the eyes of the public.
My article concludes by offering preliminary thoughts on what should be done to fix this problem. It offers two independent and contrasting solutions: new procedural rules that restrict reliance on factual authorities found in house, or alterations to the adversary method to allow for more public participation. The first brand of solution could take many forms short of a remand. Possibilities include: additional briefing if a factual question remains hazy on the record, or a canon of avoidance that would mean that the Court should not decide a case in a way that would require extra-record fact finding. An entirely different solution is to make judges better at evaluating legislative facts without the parties’ help. Reform possibilities here include: additional training for judges on empirics, the creation of a research service, or even opening up the process completely, so that the Court would solicit opinions and evidence from all interested parties and encourage public participation in the amicus process – much like the notice-and-comment process in administrative agencies.
Regardless of which course is more persuasive, both are superior to the outdated procedural void that currently exists. As the pace of accessing information accelerates exponentially – and judges are understandably tempted to take advantage of it – we need to seriously contemplate the implications of in-house judicial fact finding and to update our approach to accommodate them.
Allison Orr Larsen is an Assistant Professor of Law at William & Mary Law School. Her research interests include administrative law, constitutional law, and the institutional and informational dynamics of legal decision making.
CLICK HERE FOR FULL VERSION OF THIS STORY