Assessing Credibility and Demeanor of Witnesses who Testify Through Interpreters

Judge Posner's decision in Apouviepseakoda v. Gonzales, 05-3752 (7th Cir., Feb. 2, 2007), which I mentioned yesterday has an interesting twist.

It's one of the great fetishes of US law that triers of fact, be they juries, judges, or administrative officials (including ALJs and Immigration Judges), deserve deference for their credibility determinations as they saw the live witness and the reviewer of the cold record did not. We don't question jury determinations — they're a black box, and juries are not called upon to give reasons for their decisions as to who to believe. The same is not true, however, of either judges or administrators. We expect them to issue reasoned opinions, and call foul when they fail to and when the reasons they give fail to hold water.

Indeed, one of the puzzles of administrative law is the so-called Universal Camera problem — suppose that an initial trier of fact reached one decision on the evidence but that the higher-ups in the agency appeals process reached an opposite conclusion. When the matter goes to the court of appeals, what weight should the court put on the trier's views? The issue implicates two competing values in administrative law: on the one hand the great deference to the front-line assessment of credibility, on the other hand the command in the Administrative Procedure Act that the agency, in deciding a matter, has full power to determine it (i.e. that its hands are not tied by lower-level officials). It's very hard for both of these to be true at once: if we give weight to demeanor then the front-line official has de facto power to limit the decisions of his/her bosses. If on the other hand we don't give any extra credence to the factual findings of the person who actually received the evidence, we've undermined an ancient principle of Anglo-American jurisprudence. We square that circle by saying that the trier's decision is part of the record that binds the agency. So it's free to make any decision, but its decision must be based on the record. The trier, we say only somewhat convincingly, hasn't bound the agency, he's just made part of the record.

Comes now Judge Posner in Apouviepseakoda to ask a really good question about credibility determinations by administrative agencies (and implicitly also by judges, although not juries):

The fact that she was testifying through an interpreter has a significance that my colleagues do not appreciate when they say that “The IJ spent 6 hours in a hearingroom, face to face, with Ms. Apouviepseakoda. We have never met her.” I take this to be an allusion to the common though not necessarily correct belief that being present when a witness testifies greatly assists a judge or juror in determining whether the witness is telling the truth. Even if so in general, it cannot be so when the witness is a foreigner testifying through an interpreter, especially if the judge cannot even hear the foreigner, but only the interpreter. Reading the facial expressions or body language of a foreigner for signs of lying is not a skill that either we or Judge Brahos possess.

We understand the dilemma facing immigration judges in asylum cases. The applicant for asylum normally bases his claim almost entirely on his own testimony, and it is extremely difficult for the judge to determine whether the testimony is accurate. Often it is given through a translator, and even if the applicant testifies in English, as a foreigner his demeanor will be difficult for the immigration judge to “read” as an aid to determining the applicant's credibility. Unfortunately, the Department of Homeland Security and the Justice Department, which share responsibility for processing asylum claims, have, so far as appears, failed to provide the immigration judges and the members of the Board of Immigration Appeals with any systematic guidance on the resolution of credibility issues in these cases. The departments have not conducted studies of patterns of true and false representations made by such applicants, of sources of corroboration and refutation, or of the actual consequences to asylum applicants who are denied asylum and removed to the country that they claim will persecute them. Without such systematic evidence (which the State Department's country reports on human rights violations, though useful, do not provide), immigration judges are likely to continue grasping at straws—minor contradictions that prove nothing, absence of documents that may in fact be unavailable in the applicant's country or to an asylum applicant, and patterns of behavior that would indeed be anomalous in the conditions prevailing in the United States but may not be in Third World countries—in an effort to avoid giving all asylum applicants a free pass. The departments seem committed to case by case adjudication in circumstances in which a lack of background knowledge denies the adjudicators the cultural competence required to make reliable determinations of credibility.

The concern that demeanor evidence is less probative — or is being judged by people who are not properly trained to assess it — when testimony is rendered through an interpreter, is one that Judge Posner has raised before, notably in Djouma v. Gonzales, 429 F.3d 685, 687-88 (7th Cir. 2005), and more thoroughly in Iao v. Gonzales, 400 F.3d 530, 534 (7th Cir. 2005), where speaking for the court, Judge Posner wrote:

We close by noting six disturbing features of the handling of this case that bulk large in the immigration cases that we are seeing:

…

4. Insensitivity to the possibility of misunderstandings caused by the use of translators of difficult languages such as Chinese, and relatedly, insensitivity to the difficulty of basing a determination of credibility on the demeanor of a person from a culture remote from the American, such as the Chinese. E.g., Lin v. Ashcroft, 385 F.3d 748, 756 n. 1 (7th Cir.2004); Ememe v. Ashcroft, 358 F.3d 446, 451-53 (7th Cir.2004); Mendoza Manimbao v. Ashcroft, 329 F.3d 655, 662 (9th Cir.2003); He v. Ashcroft, 328 F.3d 593, 598 (9th Cir.2003); Deborah E. Anker, “Determining Asylum Claims in the United States: A Case Study on the Implementation of Legal Norms in an Unstructured Adjudicatory Environment,” 19 N.Y.U. Rev. L. & Social Change 433, 505-27 (1992); Neal P. Pfeiffer, “Credibility Findings in INS Asylum Adjudications: A Realistic Assessment,” 23 Tex. Int'l L.J. 139 (1988). Behaviors that in our culture are considered evidence of unreliability, such as refusing to look a person in the eyes when he is talking to you, are in Asian cultures a sign of respect.

This issue of demeanor in translation (and the implications for deference on appeal) seems like an important question. There must be a good student note or two in here somewhere.