The State Secrets Doctrine: In Violation of International Law? (Part 3)
Having set the scene with some basic remarks on the international rules involved in the first post, and applied these rules to the political question doctrine in the second, I now turn to the state secrets doctrine. This operates where a case involves state secrets, the disclosure of which would adversely affect national security. The doctrine commands that such disclosure must be avoided by all procedural means available to the court; if there is no way of preventing such disclosure, the plaintiff’s case depending totally on classified information, the case must be dismissed (see United States v. Reynolds, 345 U.S. 1 (1953)).
This state secrets privilege would appear to be primarily a rule of evidence, but the term of the ‘state secrets doctrine’ may be given a broader meaning: where a case depends for its cause of action or for its essential facts on classified matters, the case will be non-justiciable (see Totten v. United States, 92 U.S. 105 (1876); Tenet v. Doe, 125 S.Ct. 1230 (2005), and the discussion of the two cases in El-Masri v. Tenet, available here).
I already stated in the introductory post that there is a right of access to a court under Article 14 ICCPR and that a dismissal based on any of the American doctrines is in principle capable of constituting an interference with the right of access to a court. This is borne out by the jurisprudence of the European Court of Human Rights, which decided as much in two cases concerned with state secrets and court proceedings: Tinnelly and Sons Ltd and Others and McElduff and Others v. United Kingdom, at paras. 58-63, and in Devenney v. United Kingdom, at paras. 20-22.
Again, this leaves the issue of justification.
It is obvious that there are some state secrets in every state. The considerations of national security behind such secrecy are also a legitimate aim for restrictions on the right of access to a court to pursue (Tinnelly and Sons Ltd and Others and McElduff and Others v. United Kingdom, supra, at para. 76), if the classification of relevant material was ordered in good faith. Nevertheless, where state secrets are not disclosed to a court of law because of these interests, the resulting state of affairs is ‘tantamount to a removal of the court’s jurisdiction by executive ipse dixit’ (i.e. by the executive branch’s own say-so: Devenney v. United Kingdom, supra, at para. 28, citing Tinnelly and Sons Ltd and Others and McElduff and Others v. United Kingdom, supra, at para. 77). This is not easily acceptable (see also United States v. Reynolds, supra, at 9-10: ‘Judicial control over the evidence in a case cannot be abdicated to the caprice of executive officers.’).
The right of access to a court therefore demands that the classified information be made available to the court (see El-Masri v. Tenet, supra, where such a possibility under the Classified Information Procedures Act was briefly considered), perhaps with the assistance of special advocates instructed by the court to act on behalf of the parties, while observing the secrecy of the information (see R v. Shayler [2002] UKHL 11, [2003] 1 AC 247, at para. 113, citing with approval Secretary of State for the Home Department v. Rehman [2000] EWCA Civ 168, [2000] 3 WLR 1240, para. 31, where the instruction of such special advocates was regarded as possible even without statutory authorisation, under the inherent jurisdiction of the court; see as to this concept Taylor v. Lawrence [2002] EWCA Civ 90, [2003] QB 528, at paras. 50-53). Alternatively, it may be enough for ‘other mechanisms of complaint’ to be available (Devenney v. United Kingdom, supra, at para. 28), but these would always have to be judicial proceedings before independent and impartial judges (Tinnelly and Sons Ltd and Others and McElduff and Others v. United Kingdom, supra, at para. 77; see also Waite and Kennedy v. Germany (ECtHR), at paras. 68-74).
It would also be permissible in such cases for the court to hold private sessions. While Article 14 (1) (2) ICCPR provides for the right to a ‘public hearing’, public order and national security are explicitly recognized as reasons for excluding the public.
Whatever the solution adopted, the state must always closely examine whether any relaxation of secrecy or adaptation of court procedures can be made in order to secure the right of access to a court at least in part. A court faced with an application in such a case would, for its part, have to examine whether the classification of material had been appropriate (as the court in El-Masri v. Tenet, supra, did); if not, then the legitimate aim of protecting national security would not be engaged, or would not be sufficient to displace the human rights of the private party (see also on this whole issue of judicial oversight in the operation of the state secrets doctrine Jack Balkin’s post at Balkinization; as will appear shortly, I completely agree with his conclusions).
It is furthermore difficult to see how it could be necessary to refuse disclosure even to the judge in chambers (as the US Supreme Court contemplated in United States v. Reynolds, supra, at 10). Disclosing evidence to a judge alone will produce only a very low risk to national security. Furthermore, a court ruling confirming the secrecy of information on certain alleged acts does not imply that such acts have indeed taken place (contra Arar v. Ashcroft, available here, at p. 72 of the PDF file); it means simply that the relevant complex of information is secret. On the other hand, if even the judge in a case is left in the dark, this ‘cannot be said to be conducive to public confidence in the administration of justice’ (Tinnelly and Sons Ltd and Others and McElduff and Others v. United Kingdom, supra, at para. 78). The disadvantages involved in the disclosure of classified information to a judge therefore seem entirely negligible.
Care must also be taken in defining what interests can give rise to legitimate considerations of national security. This is all the more so since the judge will make his or her decision on this point away from the democratic control inherent in the publicity of court proceedings and judgments, and since the outcome will hide even more information from the watchful eyes of the general public. It is certainly not enough that the information concerned, or the outcome of the litigation following the use of such information, may acutely embarrass the government of the day (see Arar v. Ashcroft, supra, at pp. 71-76 of the PDF file, and the comments on the case by David Luban at Balkinization, some of which he repeated here); embarrassment is simply not a danger to national security (this argument does, however, share some ground with the political question doctrine – which I criticized in an earlier post).
For considerations of national security to be properly brought into play, there must be some concrete information on the effects of any disclosure or decision on the dangers facing the state or the operations of state authorities against such dangers (see Lustig-Prean and Beckett v. United Kingdom, where the ECtHR held (at para. 82) that a state had a certain margin of appreciation on the grounds of national security if there was ‘a real threat to the armed forces’ operational effectiveness’, ‘substantiated by specific examples’).
Of course, there is still a danger that ‘national security’ may be used by the government as a catch-all defence (as Julian Ku suggests it might). This is all the more incentive for the courts to exercise their powers of review extensively and their power of dismissing a case under the state secrets doctrine sparingly. Above all, the courts must take care not to become ‘more executive minded than the executive’ (to borrow the famous phrase from Lord Atkin’s ultimately vindicated [see paras. 139-141 of the linked case] dissent in Liversidge v. Anderson [1942] AC 206).
However, it may be that all this scrutiny and care in allowing for some form of judicial process will lead nowhere if properly classified information is so central to a given case that the court could not possibly decide the case, except by extensive reference to the classified material. This would be the case where the cause of action (e.g. a contract) is itself a state secret (as in Totten v. United States, supra), or where the facts complained of in a tort action are completely secret (as in El-Masri v. Tenet, supra). This is where push really comes to shove, as it were, that is to say, where the opposition of state secrets and the right of access to a court is at its most dramatic. In such cases, legitimate reasons of national security may possibly outweigh the right of access to a court enjoyed by a private party to a court case.
The state secrets doctrine may therefore be permissible, so long as the dismissal of the case is only resorted to where no other means of protecting properly classified material is available, and the propriety of treating material as secret is duly considered by the court.
This state secrets privilege would appear to be primarily a rule of evidence, but the term of the ‘state secrets doctrine’ may be given a broader meaning: where a case depends for its cause of action or for its essential facts on classified matters, the case will be non-justiciable (see Totten v. United States, 92 U.S. 105 (1876); Tenet v. Doe, 125 S.Ct. 1230 (2005), and the discussion of the two cases in El-Masri v. Tenet, available here).
I already stated in the introductory post that there is a right of access to a court under Article 14 ICCPR and that a dismissal based on any of the American doctrines is in principle capable of constituting an interference with the right of access to a court. This is borne out by the jurisprudence of the European Court of Human Rights, which decided as much in two cases concerned with state secrets and court proceedings: Tinnelly and Sons Ltd and Others and McElduff and Others v. United Kingdom, at paras. 58-63, and in Devenney v. United Kingdom, at paras. 20-22.
Again, this leaves the issue of justification.
It is obvious that there are some state secrets in every state. The considerations of national security behind such secrecy are also a legitimate aim for restrictions on the right of access to a court to pursue (Tinnelly and Sons Ltd and Others and McElduff and Others v. United Kingdom, supra, at para. 76), if the classification of relevant material was ordered in good faith. Nevertheless, where state secrets are not disclosed to a court of law because of these interests, the resulting state of affairs is ‘tantamount to a removal of the court’s jurisdiction by executive ipse dixit’ (i.e. by the executive branch’s own say-so: Devenney v. United Kingdom, supra, at para. 28, citing Tinnelly and Sons Ltd and Others and McElduff and Others v. United Kingdom, supra, at para. 77). This is not easily acceptable (see also United States v. Reynolds, supra, at 9-10: ‘Judicial control over the evidence in a case cannot be abdicated to the caprice of executive officers.’).
The right of access to a court therefore demands that the classified information be made available to the court (see El-Masri v. Tenet, supra, where such a possibility under the Classified Information Procedures Act was briefly considered), perhaps with the assistance of special advocates instructed by the court to act on behalf of the parties, while observing the secrecy of the information (see R v. Shayler [2002] UKHL 11, [2003] 1 AC 247, at para. 113, citing with approval Secretary of State for the Home Department v. Rehman [2000] EWCA Civ 168, [2000] 3 WLR 1240, para. 31, where the instruction of such special advocates was regarded as possible even without statutory authorisation, under the inherent jurisdiction of the court; see as to this concept Taylor v. Lawrence [2002] EWCA Civ 90, [2003] QB 528, at paras. 50-53). Alternatively, it may be enough for ‘other mechanisms of complaint’ to be available (Devenney v. United Kingdom, supra, at para. 28), but these would always have to be judicial proceedings before independent and impartial judges (Tinnelly and Sons Ltd and Others and McElduff and Others v. United Kingdom, supra, at para. 77; see also Waite and Kennedy v. Germany (ECtHR), at paras. 68-74).
It would also be permissible in such cases for the court to hold private sessions. While Article 14 (1) (2) ICCPR provides for the right to a ‘public hearing’, public order and national security are explicitly recognized as reasons for excluding the public.
Whatever the solution adopted, the state must always closely examine whether any relaxation of secrecy or adaptation of court procedures can be made in order to secure the right of access to a court at least in part. A court faced with an application in such a case would, for its part, have to examine whether the classification of material had been appropriate (as the court in El-Masri v. Tenet, supra, did); if not, then the legitimate aim of protecting national security would not be engaged, or would not be sufficient to displace the human rights of the private party (see also on this whole issue of judicial oversight in the operation of the state secrets doctrine Jack Balkin’s post at Balkinization; as will appear shortly, I completely agree with his conclusions).
It is furthermore difficult to see how it could be necessary to refuse disclosure even to the judge in chambers (as the US Supreme Court contemplated in United States v. Reynolds, supra, at 10). Disclosing evidence to a judge alone will produce only a very low risk to national security. Furthermore, a court ruling confirming the secrecy of information on certain alleged acts does not imply that such acts have indeed taken place (contra Arar v. Ashcroft, available here, at p. 72 of the PDF file); it means simply that the relevant complex of information is secret. On the other hand, if even the judge in a case is left in the dark, this ‘cannot be said to be conducive to public confidence in the administration of justice’ (Tinnelly and Sons Ltd and Others and McElduff and Others v. United Kingdom, supra, at para. 78). The disadvantages involved in the disclosure of classified information to a judge therefore seem entirely negligible.
Care must also be taken in defining what interests can give rise to legitimate considerations of national security. This is all the more so since the judge will make his or her decision on this point away from the democratic control inherent in the publicity of court proceedings and judgments, and since the outcome will hide even more information from the watchful eyes of the general public. It is certainly not enough that the information concerned, or the outcome of the litigation following the use of such information, may acutely embarrass the government of the day (see Arar v. Ashcroft, supra, at pp. 71-76 of the PDF file, and the comments on the case by David Luban at Balkinization, some of which he repeated here); embarrassment is simply not a danger to national security (this argument does, however, share some ground with the political question doctrine – which I criticized in an earlier post).
For considerations of national security to be properly brought into play, there must be some concrete information on the effects of any disclosure or decision on the dangers facing the state or the operations of state authorities against such dangers (see Lustig-Prean and Beckett v. United Kingdom, where the ECtHR held (at para. 82) that a state had a certain margin of appreciation on the grounds of national security if there was ‘a real threat to the armed forces’ operational effectiveness’, ‘substantiated by specific examples’).
Of course, there is still a danger that ‘national security’ may be used by the government as a catch-all defence (as Julian Ku suggests it might). This is all the more incentive for the courts to exercise their powers of review extensively and their power of dismissing a case under the state secrets doctrine sparingly. Above all, the courts must take care not to become ‘more executive minded than the executive’ (to borrow the famous phrase from Lord Atkin’s ultimately vindicated [see paras. 139-141 of the linked case] dissent in Liversidge v. Anderson [1942] AC 206).
However, it may be that all this scrutiny and care in allowing for some form of judicial process will lead nowhere if properly classified information is so central to a given case that the court could not possibly decide the case, except by extensive reference to the classified material. This would be the case where the cause of action (e.g. a contract) is itself a state secret (as in Totten v. United States, supra), or where the facts complained of in a tort action are completely secret (as in El-Masri v. Tenet, supra). This is where push really comes to shove, as it were, that is to say, where the opposition of state secrets and the right of access to a court is at its most dramatic. In such cases, legitimate reasons of national security may possibly outweigh the right of access to a court enjoyed by a private party to a court case.
The state secrets doctrine may therefore be permissible, so long as the dismissal of the case is only resorted to where no other means of protecting properly classified material is available, and the propriety of treating material as secret is duly considered by the court.
posted by Tobias Thienel at 17:06
2 Comments:
Once more, thank you Tobias. I like the proposed role for 'special advocates.'
You may be interested in an article in today's Los Angeles Times by Louis Fisher, author of a forthcoming book: In the Name of National Security, in which he examines US v. Reynolds (1953). Among important items are the mention of Wigmore's treatise on evidence. I think an examination of the Pentagon Papers case provides some helpful perspective here, especially in light of the fact that government attorneys 'discovered photocopies of newspaper articles and presidential speeches' among the many documents and 'could not understand why such materials--already in the public domain--were classified.' [!] When prospective witnesses were asked 'to identify the specific documents in the Pentagon Papers study that would jeopardize national security,' they were informed by Defense and State Department representatives that they knew such documents existed, but they could not explain which of the documents in the forty-seven volume study presented specific risks to national security. See David Rudenstine's The Day the Presses Stopped: A History of the Pentagon Papers Case (Berkeley, CA: University of California Press, 1996).
I've pasted here material from the second half of Fisher's piece:
'The fact is that the branch of government that runs the courtroom and decides questions concerning privilege and the introduction of evidence is the judiciary, not the executive branch. Judges have a constitutional duty to function as neutral referees to allow each side to present its case fairly. A court that automatically accepts the government's argument about state secrets, without ever looking at the documents in question, aligns the judiciary with the executive branch and eliminates any chance of a fair trial.
The "state secrets privilege" reached the Supreme Court in 1953 in United States vs. Reynolds, a case in which the widows of three civilians who died in the crash of a B-29 bomber sought damages. They asked the government to give them the accident report and the statements of three surviving crew members. Lower courts held that the government must surrender the secret documents to the trial court or lose the case. The judge would examine the documents in his chambers to determine if the government's demand for secrecy was valid.
Refusing to release the documents, the government lost at the district and appeals court levels. But when the case got to the Supreme Court, the justices reversed course and accepted the government's argument about state secrets — without ever looking at the documents. Even having the judge examine the evidence alone in his chambers could jeopardize national security, the majority ruled. When the documents were declassified in the 1990s, it became evident that the Supreme Court had been misled: The documents contained no state secrets.
In his authoritative 1940 treatise on evidence, John Henry Wigmore concluded that the executive branch is entitled to protect state secrets but that in cases in which classified information is at issue, it is up to the judge to decide whether such evidence qualifies as legitimately secret, and thus legally privileged. A court that "abdicates its inherent function of determining the facts upon which the admissibility of evidence depends will furnish to bureaucratic officials too ample opportunities for abusing the privilege," Wigmore warned.
I believe that Wigmore was right and that the Supreme Court decided the Reynolds case wrongly. Some federal judges agreed, and they have continued to recognize their obligation to examine disputed documents in their chambers before passing judgment on matters of evidence and privilege.
In 1971, the U.S. Court of Appeals for the Federal Circuit argued that "an essential ingredient of our rule of law is the authority of the courts to determine whether an executive official or agency has complied with the Constitution and with the mandates of Congress which define and limit the authority of the executive."
Claims of executive power "cannot override the duty of the court to assure that an official has not exceeded his charter or flouted the legislative will," the court said.
The executive branch may assert a right to keep its secrets, but this must not be viewed as an absolute right, only a privilege granted by the court when appropriate. Otherwise, even in the age of terrorism, there can be no judicial independence and no fair trial for Americans who would challenge their government.'
See also my addendum on the subject in a later post (which links here and therefore appears below, under Links to this post).
Post a Comment
<< Home