The author teaches law at Southern Methodist University Dedman School of Law, Dallas.
What, if anything, is wrong with these lyrics:
If every judge could have an iPad / Across the courts today
Then everybody’d be surfin’ / On Wikipedi-ay
They’d all be friending on Facebook / And using Google too
The record’s just the beginning / Surfin’ USA
OK, this version mangles the Beach Boys’ classic, but what about its vision of the role of the judge? Today it is second nature for most of us to turn to our laptops, smartphones, and tablet computers with their instant Internet access to look up just about anything we’re curious about. How did we ever live without the ability to check movie times, read restaurant reviews, and look up the name of the prime minister of Canada at the click of a mouse or tap of a finger? It would be profoundly countercultural for a judge who wants more information not to do the same. And while the Internet is not itself the issue—the issue is independent fact research outside the record, whatever the medium—the Internet has made research much easier and much more common.
Consider these examples from real cases:
These types of research are sometimes proper, sometimes improper but harmless error, and sometimes the basis for reprimand or reversal. Many times they are merely noted by the judge who has done the research without any sense of impropriety. But a judge who wants to do independent fact research swims in shark-infested waters: Rules of ethics and evidence limit a judge’s freedom to surf the Internet or roam the library in search of information relevant to a case. Just as is true in a number of areas, the role of the judge requires that the person holding that office act in countercultural ways. Why? Because to allow judges to base decisions on information gained through undisclosed or untested research threatens both the accuracy of outcomes and the sense of procedural fairness on which our legal system depends.
In a broad sense, judicial fact research is inconsistent with the roles implicit in the adversary system. While managerial judging has modified our image of the passive judge awaiting party input, greater judicial involvement in processing cases does not extend to independently acquiring information. One expert on judicial ethics has noted:
In many cases, discrepancies in the evidence probably tempt judges to conduct some “research” to resolve variations in the proof. While it is true that such contacts may assist judges in deciding issues and cases, American jurisprudence relies on the adversary process to resolve factual disputes.
Leslie W. Abramson, “The Judicial Ethics of Ex Parte and Other Communications,” 37 Hous. L. Rev. 1343, 1367 (2000). Beyond concerns about role expectations, however, lie ethics rules that directly address the lure of the Internet.
Concerned about the growing temptation to do factual research, the ABA Joint Commission to Evaluate the Code of Judicial Conduct explicitly addressed the research issue in the 2007 ABA Model Code. Rule 2.9(C) provides: “A judge shall not investigate facts in a matter independently, and shall consider only the evidence presented and any facts that may properly be judicially noticed.” The Internet rates an explicit mention, as Comment 6 notes that “[t]he prohibition against a judge investigating the facts in a matter extends to information available in all mediums, including electronic.” The rule sounds like a broad prohibition of independent research by judges, leaving the parties and the record as the primary sources of information.
By including the reference to judicial notice, however, the Model Code makes the prohibition both more limited and more complex. Judges may not independently investigate adjudicative facts—the facts that are at issue in the particular case—unless (in the words of Federal Rule of Evidence 201) they are “not subject to reasonable dispute” because they are generally known or “capable of accurate and ready determination by resort to sources whose accuracy cannot reasonably be questioned.” But they may independently ascertain and use information that meets the requirements for judicial notice, and they may investigate “legislative facts”—those that inform the court’s judgment when deciding questions of law or policy—to their hearts’ content, bound by no rules about sources, reliability, or notice to the parties. The cross-reference to judicial notice also tends to elide the ethics and evidence rules. (The latter are discussed in more detail below.)
It is important to note that the Model Code does not contain an exception for background information. When Rule 2.9 was being drafted, some judges argued that they should be allowed to surf the web for basic reference information, and the minutes reflect a consideration of “whether it would be inadvisable to restrict judges’ access to the Internet or other electronic databases in connection with particular cases when such activities are only for the purpose of obtaining background reference material.” The pro-Internet argument was rejected, however, and the ABA did not change the rule’s language to specifically allow such research. In fact, the provision allowing judicial notice was added to the current rule after judges raised the issue of background research, supporting an inference that background research is only permissible when that background is generally known or indisputable.
In addition to these specific rules about research—which is functionally treated as an ex parte communication—the ethics rules also provide some limits based on bias and on the appearance of bias that can be created when a judge acquires information beyond the information in the case record. These limits stem from the fundamental need for the judicial system to provide—and to appear to provide—fair and unbiased decisions. The preamble to the Model Code notes:
The United States legal system is based upon the principle that an independent, impartial, and competent judiciary, composed of men and women of integrity, will interpret and apply the law that governs our society.
Therefore, judicial information gathering can run afoul of the ethics rules if the research would “appear to a reasonable person to undermine the judge’s independence, integrity, or impartiality.” Model Rule 3.1(C). Information gathering may also disqualify the judge from hearing a case if it gives the judge “personal knowledge of facts that are in dispute in the proceeding.” Model Rule 2.11(A)(1). Courts, however, have been hesitant to find that independent research on more general issues indicates bias or results in “personal knowledge of facts that are in dispute.”
The rules governing the conduct of federal judges are less specific about research but similar in impact. Canon 3(A)(4) of the Code of Conduct for United States Judges provides that “a judge shall not initiate, permit, or consider ex parte communications, or consider other communications concerning a pending or impending matter that are made outside the presence of the parties or their lawyers.” Nevertheless, ex parte communications are allowed “as authorized by law”—a provision that again pulls in the rules of judicial notice. Federal statutes also contain provisions regarding bias and personal knowledge of disputed facts that very much parallel the ABA Model Code provisions discussed above. “Any justice, judge, or [magistrate judge] of the United States shall disqualify himself . . . [w]here he has a personal bias or prejudice concerning a party, or personal knowledge of disputed evidentiary facts concerning the proceeding.” 28 U.S.C. § 455(a), (b).
Judicial discipline under the ethics rules for improper fact research is quite rare, and it usually involves conduct more closely akin to personal interactions than surfing the Internet. In the North Carolina example mentioned above, the judge was publicly reprimanded for his ex parte Facebook contacts with counsel and for allowing his opinion to be influenced by his own research into extra-record facts. In a more egregious case, the Washington Commission on Judicial Conduct censured a judge for initiating an ex parte investigation about gender reassignment surgery, including consultation with various medical societies about the procedure, and then using the information in the proceedings in which the petitioners sought a name change. The information the judge located was incorrect.
The result of the disciplinary proceeding was likely also influenced by remarks the judge made in the courtroom, in addition to recounting the results of his dubious research: “Before a crowded courtroom, Respondent made disparaging remarks about the petitioners and the reasons for each seeking an official change of name. Respondent’s remarks suggested that petitioners, if allowed to change their names, would pose a risk to those who ‘send their daughters into the ladies’ restroom.’” In re Hutchinson, CJC No. 93-1652-F-47, 1995 WL 902265, at *1–2 (Wash. Comm’n on Judicial Conduct Feb. 3, 1995).
Judicial research on rare occasion leads to a need for recusal. For example, a committee of the Fifth Circuit Judicial Council found that Judge James Nowlin, a former Republican state legislator in Texas, acted improperly when he conducted ex parte conversations with interested Republican political leaders and prospective Republican candidates for office concerning the court-ordered redistricting of the Texas Senate. Judge Nowlin subsequently recused himself from the case. Report of the Special Committee of the Fifth Circuit Judicial Council 7–15, In re Complaint of Earl Against U.S. Dist. Judge Nowlin (5th Cir. Judicial Council, filed Jan. 21, 1992) (unpublished report dated May 5, 1992, on file with the Fifth Circuit Court Clerk); Terrazas v. Slagle, No. A-91-CA-425 (W.D. Tex. July 21, 1992) (order of recusal).
While ethics opinions arising out of judicial fact research are unusual, case law more frequently applies procedure and evidence rules when analyzing the impact of fact investigation. A judge who brings in new facts may become a witness, violating Evidence Rule 605; a judge’s research may lead to a treatise (online or otherwise) that does not qualify as an exception to the rule against hearsay under Evidence Rule 803(18). Judicial notice in jury cases may violate the parties’ Sixth or Seventh Amendment rights. In appellate courts, independent research crosses another boundary: the case’s trial court record. Normally, any introduction of facts into the record occurs at the trial level. The appeal is a structured, stylized review of what happened below, complete with application of the burden of proof and carefully prescribed standards of review. Litigants are not generally allowed to introduce new evidence at the appellate level; an appellate judge doing his or her own factual research may be improperly committing the same error.
It is the rules of judicial notice that most often control the propriety of judges doing research. Unfortunately, these concepts turn out to be poorly calibrated to provide guidance about proper behavior. One reason for the lack of clarity is the distinction between adjudicative and legislative facts noted above. This distinction, with its accompanying terminology, comes from an article that Professor Kenneth Culp Davis wrote in 1942, in the context of administrative law. Adjudicative facts relate to the parties and their dispute-relevant activities. “When [a court] finds facts concerning immediate parties—what the parties did, what the circumstances were, what the background conditions were—the [court] is performing an adjudicative function, and the facts may conveniently be called adjudicative facts.” Kenneth C. Davis, “An Approach to Problems of Evidence in the Administrative Process,” 55 Harv. L. Rev. 364, 402 (1942).
The usual way to prove adjudicative facts is to introduce evidence at trial, and the more common use of judicial notice is for one party to submit information to the judge and argue that it meets the reliability requirements of Rule 201. Judges may do independent research regarding adjudicative facts under the rubric of judicial notice only if the facts meet those same reliability requirements and if the judge gives notice to the parties, with an opportunity to respond.
Another type of information gets the label “legislative facts.” Davis describes them this way: “When [a court] wrestles with a question of law or policy, it is acting legislatively . . . and the facts which inform its legislative judgment may conveniently be denominated legislative facts.” The most obvious examples of legislative facts are those used by the highest court in a jurisdiction when framing legal rules. Other state and federal courts, including trial courts, sometimes have occasion to use information legislatively. For example, the Washington Supreme Court used legislative facts about marital relationships in abolishing a cause of action for alienation of affection. Wyman v. Wallace, 615 P.2d 452, 453–55 (Wash. 1980).
Legislative facts are distinctive in two ways: The information to be noticed does not need to be indisputably correct, nor is the court required to give the parties any kind of notice or opportunity to be heard. A judge independently researching facts for purposes of making law and setting policy is therefore subject to no evidentiary restraints. Evidence scholars, however, have consistently argued that notice and response would be preferable even for legislative facts. See, e.g., Charles T. McCormick, “Judicial Notice,” 5 Vand. L. Rev. 296, 318 (1952).
One problem, of course, is that in practice it can be difficult to put judicial research into either the adjudicative or legislative category. It is not the nature of the information itself but the way in which it is used that distinguishes the two. Even general information may be an adjudicative fact in the context of a particular lawsuit. Judges setting out to conduct research, particularly on the Internet, cannot reliably predict whether the information they find will be used as legislative rather than adjudicative facts. In addition, the same information may be used for more than one purpose, so outside research may lead to nuggets of information that are used both adjudicatively and legislatively. For example, Davis’s influential article on judicial notice discussed cases raising the question of whether the Communist Party advocated the forcible overthrow of the government. He demonstrated that the courts’ use of that “fact” could be viewed as both legislative and adjudicative. The use of information to find adjudicative facts, even when that information is also used in a legislative way, requires any judicial research to meet the more stringent requirements for judicial notice of adjudicative facts.
Let’s focus on those adjudicative facts, because that’s where the limitations on research lie. The absolute bedrock requirement for judicial notice is reliability of both information and source. A fact may be judicially noticed only if it is “not subject to reasonable dispute in that it is either (1) generally known within the territorial jurisdiction of the trial court or (2) capable of accurate and ready determination by resort to sources whose accuracy cannot reasonably be questioned.” Fed. R. Evid. 201(b). Until recently, judges and litigants typically used this provision to consult dictionaries, government documents, maps, encyclopedias, and well-recognized treatises. But further uncertainty is created when a wider variety of sources comes into play.
While the use of published reference works provided some institutional vouching for the accuracy of the contents, the same cannot be said of the blogs, wikis, and other information one encounters online. The reliability of the information varies dramatically. The reliability of the sources does, too. As Jon Stewart, host of the Daily Show, noted, “The Internet is just the world passing around notes in a classroom.” Information on the web changes, can be edited, and disappears. (Try it yourself: Go to the Wikipedia entry for your law firm or your court. Sign in. Insert this sentence: “The [lawyers/judges] of [your institution here] are the smartest and the best looking in the United States.” Submit the change. Presto! There it will be, as part of that Wikipedia entry, until you or someone else changes it. And it may or may not be true.) Information gleaned from social media also may be neither truthful nor accurate. See Jeffrey Bellin, “Facebook, Twitter, and the Uncertain Future of Present Sense Impressions,” 160 U. Pa. L. Rev. 331 (2012). It is possible that the very process of Internet research—making decisions on the fly about what links to click and what paths to follow—makes it more likely that the results are shaped by the researcher’s preexisting opinions and biases.
Conscientious judges seeking cases to guide their decisions about whether research would be proper would find little that is helpful or consistent. Research involving talking to humans is clearly improper. But when research is done on the Internet, judges sometimes lose sight of the fact that they are still investigating case-specific facts or that what they find in cyberspace may not count as a source whose accuracy cannot reasonably be questioned. In one case, a judge was reversed for independently checking the defendant’s website and a state insurance department website in ruling on a personal jurisdiction motion. NYC Med. & Neurodiagnostic, P.C. v. Republic W. Ins. Co., 798 N.Y.S.2d 309, 312 (N.Y. App. Term 2004). On the other hand, in Moore v. Landes, 2006 Unpub. LEXIS 1076, n.2 (Ky. Ct. App. 2006), an action regarding the sale of a horse, the court held that the Internet stallion register fell within the category of indisputable facts derived from sources the accuracy of which could not reasonably be questioned. And in some of the cases noted at the beginning of this article, the procedural posture of the case left the Internet research unreviewed: The judge who Googled names as an ethnicity check saw his opinion vacated, but on other grounds; the ruling of the judge who Googled “Strange Music” to decide whether to grant a preliminary injunction was not appealed. A federal appellate judge or state supreme court majority doing independent research would see that decision reviewed only in the unlikely event that the Supreme Court chose to grant certiorari.
Case-specific facts can be general as well as specific; they may apply not only to one particular case but also to other cases as well. This is particularly true when the fact at issue is a broad question of scientific principle. Cases considering judicial research into these types of facts are particularly inconsistent. For example, the Colorado Supreme Court reversed the intermediate appellate court for using medical treatises outside the record to assess whether an electric shock could cause serious injury without leaving a burn mark. The court commented that the appellate court “in effect assumed the role of an expert medical witness” because it used a treatise that “properly should be interpreted only by experts in the appropriate field.” Prestige Homes, Inc. v. Legouffe, 658 P.2d 850, 854 (Colo. 1983) (en banc). In contrast, the Eleventh Circuit found unremarkable a trial judge’s reading of medical journal articles on iron poisoning, which was central to the case, before hearing expert testimony on the issue, noting that “it is a matter of common knowledge that courts occasionally consult sources not in evidence, ranging anywhere from dictionaries to medical treatises.” Johnson v. United States, 780 F.2d 902, 910 (11th Cir. 1986). Sometimes the result turns more explicitly on whether the information used meets the requirements of Rule 201. Thus, in Gent v. CUNA Mutual Insurance Society, 611 F.3d 79, 84 n.5 (1st Cir. 2010), the First Circuit took judicial notice of general facts on the website of the Centers for Disease Control and Prevention concerning Lyme disease because those facts were “not subject to reasonable dispute.”
The limits on the use of independent judicial research are not mere technical parsing of facts, law, and the nature of what Google provides. They are part of the sea of due process values that keeps our court system afloat. Beyond the question of the reliability of the information discovered is a question of procedural safeguards, embodied in rules giving parties notice and an opportunity to respond to new information. These fundamental values are both constitutionally based and embodied in a number of ethics and litigation rules. See, e.g., Model Code Rule 2.9(B) (requiring judges to give notice of inadvertent ex parte communication); Model Code Rule 2.11 cmt. 5 (requiring judges to notify parties of information they might find relevant to disqualification); Fed. R. Evid. 201(e) (requiring judges to provide parties an “opportunity to be heard as to the propriety of taking judicial notice and the tenor of the matter noticed”); Fed. R. Civ. P. 26(a)(2), (a)(3) (requiring advance disclosure of witnesses, exhibits, and expert witnesses).
Disclosure of independent research by judges also helps deal with the danger of misunderstood information, which is particularly strong when information is independently acquired and used without cross-examination or supplementation. Many experts question whether judges, however intelligent and well schooled in law, can properly evaluate the kinds of scientific and technical information often involved in the judges’ research:
There is no reason to think that [a judge] who is not a scientist would have, or could have, or should have found the time to gain the enlightenment in oncology, epidemiology and pediatrics needed to render his decision . . . with the assurance of correctness. . . . Today, even a trained scientist is barely more knowledgeable than a layman about the almost innumerably proliferated specialized scientific and technological areas outside the scientist’s own specialty.
Edward Gerjuoy, “Judicial Understanding of Science,” in Science and Technology Advice to the President, Congress, and Judiciary 470, 470–71 (William T. Golden ed., 1988).
The same is true in the social sciences. “Few judges are trained in statistics, demography, psychoanalysis, cognitive psychology, or whatever the relevant social science material might be.” Michael J. Saks, “Judicial Attention to the Way the World Works,” 75 Iowa L. Rev. 1011, 1026 (1990).
For these reasons, the evidence rules require that a judge who intends to take judicial notice must provide the parties with notice and an opportunity to contest both the propriety of taking judicial notice and the accuracy of the underlying “facts.” These procedural rights allow a party that would be adversely affected by the information gathered by the judge to question its indisputability and the reliability of the sources, which may help the court to get it right. Appellate courts have therefore quite correctly remanded even cases in which the information may turn out to be accurate when no notice was given. Just this past February, a Seventh Circuit panel reversed a judge’s reliance on web-based information about the Consumer Price Index and its use as a multiplier for an hourly fee: “Given that the Internet contains an unlimited supply of information with varying degrees of reliability, permanence, and accessibility, it is especially important for parties to have the opportunity to be heard prior to the taking of judicial notice of websites.” Pickett v. Sheridan Health Care Ctr., 664 F.3d 632, 648 n.22 (7th Cir. 2011).
Litigators, both trial and appellate, are strong believers in this right to be heard when a judge independently injects new information into a case. Resources permitting, however, there is a better approach. Rather than waiting to hear that the judge has gone surfin’ and merely hoping for an opportunity to dislodge some belief the judge has formed, litigators can catch the wave themselves and provide the judge in advance with the information the judge is likely to require. Sometimes, having lived with a case for months or years, lawyers fail to provide undisputed information about technical issues and vocabulary because they have forgotten what it was like not to know it. Instead of leaving the judge with the need to self-educate, trial lawyers could present the judge with an agreed presentation on basic concepts. Even in disputed matters, counsel can do more to give the judge a sufficient comfort level so that judicial research is not so tempting. There are certain recurring areas where a more robust presentation of evidence would calm the waters:
There is one other type of information that lawyers often overlook, and some judges have begun to seek it out: the visual. The law is such a verbal profession that we tend to forget that some information is better conveyed in pictures. Judge Richard Posner, an enthusiastic Internet researcher, refers to this kind of information as “comprehension-enhancing facts (which help the reader understand a case, often by providing a visual version of a description, such as a map or a photograph).” Thus, in a later discussion of the Brother Jim case mentioned at the outset of this article, Judge Posner explained that he did the research because of the lawyers’ failure to answer questions that are likely to occur to appellate judges bothered by gaps in lawyers’ narrative of a case. One . . . mystery was the layout of the university, and why exactly Brother Jim thought the walkway in front of the student union was an inadequate venue for his preaching to the college community. . . . The satellite photo did support Brother Jim’s contention that the library lawn was a far superior venue for preaching or other speechifying than the cramped walkway to which he was confined by the university authorities (in fact the library lawn appears to be the only open area on the campus), but since the university won the case anyway, our photographic assist to Brother Jim could not be thought prejudicial.
The same article notes his use of online maps to illustrate the route of some Chicago demonstrators (largely uncontroversial), and use of a satellite photo and fact research in a case where the key issue was whether shooting a gun into the air in a particular location in Indianapolis posed a substantial risk of bodily injury to others. Judge Posner made the following argument:
Indianapolis Shooter was another case in which the lawyers failed to answer questions that they should have realized would occur to many judges. The briefs did not describe the location or surroundings of the nightclub outside of which (in the night-club’s parking lot) the defendant shot into the air (for example, were there tall buildings nearby?), or the model of the FN Herstal pistol that the defendant was using, or the type of ammunition the gun fired. . . . The satellite photo appended to the opinion reveals that there are high buildings near the nightclub, which would make shooting into the air more dangerous unless the shooter shot at a steep angle. The website of the pistol’s manufacturer confirmed that it shoots armor-piercing ammunition, which would increase the danger from shooting the gun randomly. These facts, gleaned from Internet research, could well be regarded as adjudicative facts. But the purpose of obtaining and publishing them was not to sway or bolster the outcome; it was to provide a fuller picture of the crime and the crime scene. For the defendant had not denied that the bullets he fired were armor-piercing or that there were high buildings in the vicinity of the nightclub or that the nightclub was downtown rather than in a rural area.
Richard A. Posner, Appellate Judges and Internet Research (Mar. 26, 2011) (unpublished manuscript on file with author); see also Grayson v. Schuler, 666 F.3d 450 (7th Cir. 2012) (research on dreadlocks and photo of Bob Marley).
One can debate whether all of this research was proper and, even if improper, whether it was harmless error. For example, in the physical world, it is error for a judge to conduct a view outside the presence of the parties; is a satellite photo more like a view or more like a map? Cf. Lillie v. United States, 953 F.2d 1188 (10th Cir. 1992) (collecting cases considering whether improper views were harmless error). Even if the information researched meets the requirements for judicial notice, the parties were given no warning prior to the publication of the opinions that notice would be taken, and no opportunity to make arguments regarding the inferences to be drawn from the information. According to Judge Posner, “[l]awyers think they should be warned, whenever the appellate judge is minded to inject something into an opinion that the lawyers hadn’t thought to argue; but if judges did this, the appellate process would be protracted beyond endurance.”
But here’s the point: Judges are articulating the desire for fuller information, and it is that desire that sends them on their search for new facts. Why be put at a disadvantage as a result of judicial research when it is possible to choose the source and get it in the record?