FAS | Secrecy | Library || Index | Search |






If this recommendation is adopted, it will inevitably gut the secrecy agreement that is currently required as a condition of CIA employment. The report suggests that the broad-form prepublication review provision contained in this agreement has no value, because the malicious will disregard it anyway and the conscientious can safely be held to a less broad requirement. I do not believe that the historical record supports this suggestion, and I am mindful of the fact that DCIs have repeatedly affirmed, with reference to the current agreement or its predecessors, that the broad-form prepublication review provision is vital to the protection of intelligence sources and methods.


I do not believe that this recommendation should be adopted, if at all, without a much fuller accounting of the benefits that have been realized as a result of the obligations imposed by the CIA secrecy agreement, and the risks that would ensue if that agreement were to be modified in accordance with the recommendation.







I support the conclusion, reached in the main text, that the procedural safeguards available to military personnel and DoD civilians facing denial or revocation of security clearances should be the same. I would go further, however, in urging that different treatment for DoD government and contractor personnel also be eliminated. Elementary fairness requires that we provide uniform treatment for both classes of people.


Reaching this state of affairs requires that we bridge the gap between the two sets of procedures currently in place. For many of the reasons stated in the main text, the formal trial-like procedures, using the Federal Rules of Evidence as a guide, and available to anyone who requests it, whether or not there are any factual disputes that need to be resolved represents procedural overkill. And while the process is perhaps more expensive, and time and labor intensive than necessary at the front end, it is less generous than it ought to be at the appeals stage.


A common set of procedures for both government and contractor personnel should require provision of a full and complete statement of the reasons for the proposed denial or revocation and a clear statement about the right to counsel at all stages of an appeal.


Appeal of the denial of an initial clearance should be decided upon a written response without an oral hearing. Broader rights should be provided in cases involving the revocation of a clearance or the denial of a higher clearance. In these cases, so long as the person claims there is a factual dispute, there should be the right to an informal hearing before a hearing officer who neither has any involvement in the issue nor is within the chain of command of those responsible for the clearance adjudication. The hearing should resemble an informal arbitration, with a transcript and the right to call and examine witnesses. The Federal Rules of Evidence should not be used and the process should be expected to take one day or less.


A second, written appeal should be available in all cases. A board established to review these appeals should not be limited to strict scope-of-review limits but should be free to take a fresh look at the case in reaching its decision.



Appendix C.





The Commission struggled hard to reach a consensus on issues relating to polygraph testing for personnel screening purposes. In the end, however, I decided to go my own way on these issues, and to prepare this separate statement of my views. I did so not because I disagree with all of the Commission's recommendations and conclusions-indeed, there are a number with which I agree-but mainly because I do not believe that the report contains an adequate or well-reasoned analysis of the issues, and because I believe that shortcoming impeaches even those recommendations and conclusions with which I do agree.


Polygraph testing is an obviously invasive procedure, the more so in screening contexts than in other applications. In the more typical setting, there is a single factual issue that needs to be resolved, or some single event that is known to have happened and that is under investigation. Therefore the scope of the test is apt to be narrow, as is the class of persons who may have some relevant information to provide. Screening polygraphs have no such natural limits. Almost by definition they affect larger classes of persons and sweep more widely for information. The goal is not to find out the truth about some event that is known to have happened, but rather to find out about the background and personal history of the person being examined. Given that purpose, multiple topics are within the field of inquiry, and the questions may range across an entire lifetime or a substantial period of years and may begin for example with the words "have you ever" or "within the last five years have you ever." The breadth of the inquiry is one reason why privacy interests are so deeply implicated by screening polygraphs, and especially by the full-scope tests that include the so-called "lifestyle questions."


There is also the matter of the surroundings in which the tests are conducted. The atmosphere is clinical. The chair is no more appealing than a dentist's chair. The technology is apt to be mysterious, and only one of the three machine-to-body connectors, the blood pressure cuff, is apt to be familiar. There is an underlying premise that something about to be said, or already said in a personal history statement, may be a lie. The examiner is a stranger, and the entire session, including the pretest interview and any posttest questioning, is being tape-recorded or videotaped and is destined to become a government record. Those circumstances are almost bound to make the test an unnerving and intimidating experience, even apart from the extent to which the questioning encroaches on privacy zones.


Privacy interests, however, are not the same thing as legitimate expectations of privacy. At least as I see it, any analysis of the polygraph procedure, like any analysis of other invasive techniques that are used to screen government personnel, such as drug-testing programs in which urine samples are required to be given, must involve a balancing of such privacy expectations against the governmental interests that are at stake, and ultimately a determination as to whether the procedure is reasonable. My personal conclusion is that the procedure is reasonable. At least implicitly the Commission reached the same conclusion, but I get there by a different route.



Governmental interests and individual privacy expectations


At a threshold level, the analysis is pretty simple, and the balance is clearly in favor of the government. Not long ago, in l988, the Supreme Court said that the nation's security depends in large measure on the reliability and trustworthiness of CIA employees. That remark could just as well have been made with respect to others who occupy positions involving access to highly classified information. The self-evident point here is that the government has a compelling interest in assuring itself that such persons meet high standards. That interest necessitates a screening process. Individuals who seek intelligence agency positions, or other positions of equal trust, have every reason to understand and expect that such a process will be conducted, and that it will include a searching inquiry into their personal backgrounds. To be sure, there is room for disagreement about the appropriate scope of such inquiries, and as to the categories of information that are truly germane to the reliability and trustworthiness determinations that need to be made. In my opinion, however, so long as the inquiries stay within rational bounds and are carried out by lawful means, and with the consent of the persons affected, those persons can have no valid objections based on legitimate expectations of privacy.


Where the screening process entails a polygraph test, whether as a condition of initial or continued employment or as a condition of access, that fact is made known in advance, as are the topics to be covered. A decision to submit to the test is a matter of choice, requiring a voluntary consent by the person to be examined. In some cases that choice may be personally difficult, but then it is not the government's responsibility to make the screening process easy or painless. Nor can hard or difficult choices be equated with compulsion. A refusal to take a polygraph may have negative consequences, as for example the loss of a job opportunity at CIA or NSA, and there may be strong pressures to avoid those consequences, but this does not mean that a decision to take the test is forced or involuntary. While there are distinctions that can be made here between initial applicants for employment and persons who are already embarked on government or industry careers, and for whom therefore the pressures are undoubtedly greater, these distinctions are to some extent accommodated by the different test formats that are used and in any event it is still true that the tests are known-in-advance requirements, are conducted on a consensual basis, and not inconsistent with any fair expectations of privacy.



The relevance of the questions


However compelling the government's interest, the intentional collection of personal information unrelated to that interest, especially by invasive techniques, is not defensible. The issue here is therefore whether a rational link exists between the kinds of conduct that are probed by the "relevant" polygraph questions and the reliability and trustworthiness determinations that the government must make. In other words, the issue is whether these questions are "relevant" not just because they are so denominated in a polygraph test, but because they are tied to conduct about which the government has legitimate reason to be concerned and to inquire.


My own belief on this score is that, as the tests are currently structured, in both the full-scope format and the counterintelligence-scope format, all the relevant questions in the line-up deal with matters that are proper subjects of inquiry. Most of the controversy surrounds the so-called "lifestyle questions," which is the term commonly used to describe some of the questions that are asked when the test is given in the full-scope format, as it is to all applicants for CIA and NSA employment.


I view the term "lifestyle questions" as an unfortunate misnomer. The flavor of the term is that these questions have only to do with personal matters that are none of the government's business. In fact, however, the questions deal with such matters as prior criminal conduct, illicit drug use, alcohol abuse, and any history of serious financial or mental health problems. These same subjects are matters of inquiry on personal history statement forms and associated forms, and during background investigations. If they were judged to be irrelevant, they should be declared out of bounds on all these fronts, not just on the polygraph front. As I see it, however, all these subjects can readily be linked to reliability and trustworthiness concerns, and to established adjudicative criteria. Indeed it is hard for me to imagine a credible screening process in which these subjects were not pursued.


At the same time, it is my opinion that some of the relevant questions, including some of the "lifestyle questions," as currently approved for use in screening polygraphs, are overly general and too broadly worded. As a consequence, as these questions are discussed between the examiner and the person to be examined during the pre-test interview, there is a high likelihood that personal information will be elicited, perhaps embarrassing information, that could have no value in any adjudicative decision. I would therefore favor an effort to rework some of the questions, so that they would have a sharper and more narrow focus at the outset, and so that there would be a lesser chance of eliciting irrelevant personal information. I would also like to see it become an explicit objective of polygraph examiners to minimize the incidental "take" of such irrelevant information. I believe these steps would shorten the tests, make them less intrusive, and reduce the number of retests that need to be given, all without any offsetting disadvantage.





I agree with the Commission's finding that polygraph testing has high utility as a personnel screening tool. The utility evidence is varied. It consists partly of data showing that large numbers of significant admissions are made during the interview phase of the procedure that takes place before the polygraph machine is ever activated and during the questioning that may follow after the machine is deactivated. There are also less tangible but nevertheless important utility considerations having to do with the deterrent effects of the procedure in relation to both applicants and employees, with the mutual trust engendered among employees by their common polygraph experience, and with the fact that the procedure is seen as eliminating the need for other personally invasive security safeguards, as for example random drug testing programs.


Without exception, the senior agency officials consulted by the Commission, having direct responsibility for polygraph screening programs, gave it as their opinion that these programs were the single most useful screening tool at their disposal, and were the linchpin of their personnel security efforts. Granting that these opinions hardly come from neutral sources, they are still worthy of respect and are made all the more significant when considered in the light of the Commission's recognition that personnel security is the most vital ingredient in any security system.





The question that lurks behind the utility evidence, particularly insofar as it consists of data showing success in the elicitation of admissions, is whether the procedure is otherwise a sham, and succeeds only because it is orchestrated in such a way as to make it appear to persons being examined that they have only two choices, one being to make admissions assuming they have something to admit and the other being to practice deception and be detected. In other words, as I see it, the fundamental validity issue is whether the promise of detection is an empty threat, and therefore whether the whole procedure is a trick, or whether within some range of probability the procedure can actually distinguish a true answer from a false answer. By endorsing various expert pronouncements that "The scientific validity of the polygraph [when used for personnel security purposes] is yet to be established, "the Commission appears to come down on the first side of this issue. As a consequence, when it goes on to recommend that polygraph screening programs be continued with certain modifications, the report apparently adopts the position that, even though the procedure employed by these programs is or may be invalid, the programs should be maintained in any event because they are useful. If the lack-of-validity premise of that position is accepted, the programs are likely to be discontinued despite their utility.


I am not so ready as the Commission to write off screening polygraphs as lacking in scientific validity, in part because the Commission never explains what it means by that term, and even if I were ready to do so, I still would not quickly jump ahead to the separate conclusion that polygraph testing has no validity as a personnel screening tool. What follows is my own non-expert conception of the problem.


A polygraph machine monitors, usually on three channels, physiological reactions that are produced by persons as they respond to questions that can only be answered yes or no. The reactions show up as tracings on charts. The machine is not difficult to operate. There is no real dispute that it does what it is designed to do-which again is only to monitor physiological reactions and make them visible in the form of chart tracings-and that it does so accurately.


The validity problem arises not because the machine is fallible but rather because it requires an inference to derive some meaning from the charts, and because there are numerous important variables that bear on the correctness and strength of such an inference, the theoretical basis for which may itself be open to debate.


As the Commission notes in its report, there is no physiological reaction or combination of reactions that is known to be a unique earmark of lying or deception. In isolation, therefore, any reaction or set of reactions to any one question is meaningless. So, for example, if I were placed on a polygraph machine and asked only the single question whether I was an agent of the foreign intelligence service of country X, and the truth was yes but my answer was no, the best polygraph examiner in the business could not make heads or tails of my physiological reactions to that question. It is only in relation to my reactions to other questions that the examiner could begin to make sense out of my reactions to the key "are you an agent" question, and have some basis for an inference that my answer to that question was false. That inference would proceed on the theory that I would have a heightened concern about the key question and therefore react more strongly to that question than to others that were asked for the purpose of eliciting reactions that could serve as points of comparison.


All polygraph tests rely on this essential theory. The charts are diagnosed, or scored, and inferences thus drawn in favor of or against the persons being examined, by comparing the reactions to the relevant questions with the reactions to other questions. Different polygraph examiners, including CIA and NSA examiners, use different examination techniques, and different types of questions to elicit the reactions that are then compared with the reactions to the relevant questions in order to score the test. Each of the different methods has its champions, but nobody has ever discovered the magic formula. No matter which technique is used, no matter how skilled the examiner, and no matter what scoring system is applied, the resulting diagnosis may still be mistaken. If a truthful person is diagnosed as deceptive, the mistake is known as a "false positive." If a deceptive person is diagnosed as truthful, the mistake is known as a "false negative."


The accuracy and error rates of screening polygraphs are at best very difficult to estimate. The same is true in non-screening contexts, except in validity studies where mock crimes or some similar events are staged and the tests are then conducted in laboratory conditions, allowing the variables to be controlled. In such studies the guilt or innocence of the role-playing characters is known, although not to the polygraph examiner, and there is accordingly a stone tablet-a record of what is known in the business as "ground truth"-against which the examiner's conclusions can be cross-checked. Such tablets don't exist outside the laboratory, and even where they do exist, there is apt to be heated debate among experts about the design of the studies and about the extent to which their findings can be generalized.


None of this, however, leads me to believe that the use of polygraph testing for screening purposes is an unreasonable procedure. To say that polygraphy may not be an exact science is not at all to say that polygraphers cannot reach credible and reasoned opinions, let alone that such opinions can be dismissed as wild guesses. We are not dealing here with a procedure in which an examiner simply hooks up a machine, looks at the charts, and delivers a verdict. We are dealing instead with a much more careful procedure, one in which both the relevant and other questions are previewed and discussed with the person to be examined, and in which the examiner then seeks to adjust the relevant questions so as to eliminate possible causes of high-stress reactions not attributable to deception. We are also dealing with a procedure in which equally careful efforts are made, following a run on the machine that does not produce a "clear chart," to again eliminate, by further adjustments in the relevant questions, any high-stress reactions to those questions that could have causes or explanations other than deception. At the end of the procedure, if the high-stress reactions remain, there at a minimum is a rational basis for an inference that deception is the most probable cause of those reactions.


Where the Commission's report goes wrong, it seems to me, is in its apparent suggestion that the validity of polygraph testing is an all-or-nothing proposition. The sense of the report is that one or another of two propositions must be accepted-either the procedure is able to distinguish truth from deception with scientific accuracy, or it isn't able to distinguish anything at all.


If matters were this simple, the policy choices would be far easier than in fact they are. If polygraph testing produced results that were no better than random chance, say no better than the results that could be obtained by flipping coins, the arguments against it would be much stronger and might even be overwhelming, despite the utility evidence and the government's compelling interest in conducting an effective screening process. On the other hand, if polygraph testing results had the same degree of certainty as, say, the results of the testing of urine or blood samples, the arguments in favor of it would be much stronger, although for different reasons the technique would still be controversial. As it is, however, at least in my opinion, the reality is somewhere in between, probably much closer to the high end of the scale than to the coin-toss end but nevertheless at a point on the scale where there is some significant chance that opinions may be mistaken. The hard policy problem for any manager or adjudicator then becomes: how much credence can or should be given to such opinions, and who should bear the burden of the doubt, the government or the individual.


The Commission's report does not lay any of this out, but instead sidesteps and masks this policy problem by its treatment of polygraph validity as an all-or-nothing proposition, and leaves what I regard as a false impression both as to the state of the art today (the inference being that validity is zero) and as to the promise of research tomorrow (the inference being that something approaching absolute validity might be established.)


I am a strong supporter of further basic research, but I have also come to appreciate the challenge of designing high-yield research projects in this field, and I believe that any advances in knowledge will come slowly and in small increments. Again, in my view the opinion products of polygraph testing, assuming the competence of the examiner, are rational inferences either that a person is probably telling the truth or probably being deceptive, or perhaps that the results are too inconclusive to support an inference one way or the other. It may well be that a procedure that is so dependent on the competence of an examiner, and that deals in inferences about probabilities, could never meet exacting standards of scientific accuracy, no matter how extensive or well designed any future research projects might be.


If my conceptions are right, any DCI, Director of NSA, or Secretary of Defense who wishes to maintain polygraph screening programs, now or in the foreseeable future, will have to accept the uncertainty of accuracy rates, and the inevitability of some false positive outcomes, as facts of life. Likewise inevitable are some false negative outcomes. On that side the possibility that the polygraph can be "beaten," by physical countermeasures or otherwise, adds something, although nobody can say how much, to the accuracy rate uncertainty. Insofar as polygraph testing results may play a decisive role in connection with security approval decisions, these uncertainties mean that some deserving individuals will be screened out, and some undeserving individuals, conceivably even a trained foreign agent from whom we have the most to fear, will make their way through.


These uncertainties, however, need to be kept in perspective. While polygraph tests may not be scientifically exact, the other available means of investigating a person's background are anything but foolproof themselves. Personal history statements, personal interviews, and background investigations can be, and often are, carriers of information that is false, distorted, or misleading, purposely or otherwise, and record checks are not guaranteed to be reliable either. Even in the best of circumstances, the information derived from these other sources does not meet, nor is it expected to meet, any scientific accuracy standards, and may be low-grade in terms of its value and credibility. If anything, polygraph testing is less open to being faulted on these grounds, particularly considering the fact that it so often leads to admissions that have undoubted reliability. Given a choice between two screening regimes, one of which would involve a personal history statement and the other traditional non-polygraph means of investigation, and the other of which would involve a personal history statement plus only polygraph testing, my guess is that CIA and NSA would vote for the second every time. However, there is no reason to make that choice, because better decisions are likely to be made when all sources of information are used in tandem.


Whether I am right or wrong in any of this, I do not think that any major policy shifts should be based on non-expert judgments concerning a set of issues that are as technically complex as the issues related to the validity of polygraph testing procedures used to screen personnel.



Recommendations of the Commission


I will turn now to the various recommendations contained in the Commission's report. Before doing so, however, I want to comment about one of the other statements in the Commission's report with which I strongly disagree. In its catalogue of pro-polygraph arguments, the report includes an alleged argument relating to "cost-effectiveness," and goes on to say that both CIA and NSA present a good case that "[w]hen admissions made by a subject during a polygraph test result in a disqualification, these agencies are saved the considerable cost and time of conducting a background investigation. "As far as I know, neither CIA nor NSA has ever said that polygraph testing is conducted in order to save money. What they have said is that it makes more sense to conduct the testing, as they do, at the front end of the screening process, rather than as a last step in that process, because when things were done in the reverse sequence, as was formerly the case, too often the background investigation would be successfully completed only to find that the applicant made disqualifying admissions during the polygraph test. The real argument here is that polygraph testing often turns up information that background investigations do not. Cost effectiveness has nothing to do with whether such testing is conducted, only when it is conducted. Counting cost effectiveness as a pro-polygraph argument is incorrect and only serves to belittle the serious pro-polygraph position.


Scope. The Commission's first three recommendations relate to the scope of the relevant questions to be asked on screening polygraphs conducted by DOD and intelligence community agencies.


The first recommendation is that all such testing be limited to the so-called "CI-scope" questions, except in the case of applicants seeking staff positions at CIA or NSA. As I understand it, this recommendation is principally aimed at the testing of contractor personnel, and specifically NSA contractor personnel and some CIA contractor personnel, who today are required to take the so-called "full-scope" tests. I agree with the recommendation. My reason for that agreement is that, as I see it, contractor personnel are in a somewhat different position, so far as concerns their legitimate expectations of privacy, than applicants for full-time staff positions at CIA or NSA. The latter are seeking careers that would give them continued and wide-ranging access to highly classified information over a long period. The former are apt to be persons who are already embarked on careers in industry, which they may well have undertaken without any reason to believe that their personal backgrounds would ultimately be the subject of searching inquiry by the government, and who in any event may have only less wide-ranging and only temporary access to highly classified information. In my view these considerations support the recommendation.


The second recommendation is that the testing of applicants for staff positions at CIA and NSA be limited to the so-called "CI-scope" questions plus questions about serious criminal conduct and recent drug use. The rationale is that the other questions currently asked on the so-called "full-scope" tests do not produce much useful information and therefore should be eliminated, producing a cost-free benefit in the form of a reduction in intrusiveness. In my judgment, as I have said, the other questions are not objectionable on relevance grounds, and I would be slow to discard them without a fuller cost-benefit breakout than I think the Commission has ever seen.


The third recommendation is that all reinvestigation polygraphs be limited to CI-scope questions. This recommendation would simply continue current practice.


Reciprocity. The Commission's fourth recommendation is that "the polygraph should not serve as a bar to clearance reciprocity or to the exchange of classified or sensitive information." This recommendation is not explained in the report, and I am not sure what problem it is meant to correct, or what the correction would be.


Control questions. The fifth recommendation is a large mosaic of several ideas: that "the intrusiveness of control questions be minimized;" that there be strict oversight to prevent abusive control questions; that information elicited by control questions not be kept in a permanent record unless it relates to criminal activity; and that appropriate compliance procedures be adopted and enforced.


The predicate of this recommendation is a finding in the report that "control questions are frequently identified as the most intrusive aspect of the polygraph." I do not agree with the finding, which I believe is based on several misconceptions, but I do agree that there is probably room to narrow the scope of control questions, just as I believe that there should be some narrowing of the relevant questions. So far as concerns the idea of keeping no permanent record of information elicited by control questions, I am very doubtful that this idea makes any sense, although it may deserve further study. If the idea were to be implemented, it presumably would require that the audiotape or videotape be edited. This would involve the partial destruction of these records, even though one of the purposes for which they are kept is to assure their availability in the event of any complaint about misconduct or overreaching by the examiner. Further, these records are held very closely, and I am unaware of any evidence that came before the Commission of any instance in which there was an improper release or any misuse of the kind of information to which the recommendation relates. While the recommendation calls for implementing procedures, it is impossible to know what sort of procedures the report might have in mind.


Over-reliance. The Commission's sixth recommendation is that "physiological reactions without admissions, to questions during a polygraph examination should not be used to disqualify individuals without efforts to independently resolve the issue of concern." This recommendation is low in clarity. What kinds of efforts would be required to "independently resolve the issue of concern," and what could happen if those efforts failed? Suppose there were two equally well qualified applicants for the same position, and the polygraph tests resulted in an examiner's opinion of probable deception in one case but not the other. Would that then mean that, absent some confirmation of the probable deception opinion, these results had to be ignored in making the decision as to which applicant to hire? The recommendation raises more questions than it answers, and provides no useful guidance.


Oversight. The seventh recommendation is that a new independent and external mechanism be established to investigate and track polygraph complaints. It is a given that polygraph programs should be subject to rigorous and effective oversight. This recommendation is made, however, without any real review of existing oversight structures, or any real effort to show how or why those structures might be inadequate, or any indication of how the new "mechanism" would be expected to operate. If the existing oversight is ineffective, obviously it should be improved. But within CIA, for example, there is already oversight within the Polygraph Section of The Office of Security, and there is also a special oversight panel (The Polygraph Complaint Oversight Board) which includes a representative of the Office of General Counsel and that was formed in mid-1992 for the explicit purpose of resolving polygraph-related complaints, not to mention the Inspector General's office. Surely any recommendation calling for additional oversight should be based on some showing, which the report does not contain, that these checks and safeguards are insufficient.


Standardization. The Commission's eighth recommendation is that "standards be developed to ensure consistency in the administration, application and quality control of screening polygraphs." There is already a trend in this direction, and I agree that further steps should be taken. I do not understand, for example, why the relevant questions, in whichever of the two basic formats the tests are given, should be different depending on which agency is conducting the test.


The different practices to which this recommendation relates, however, are overshadowed by circumstances that the Commission's report barely even mentions.


Polygraph screening programs are not in effect, and have virtually no chance of being placed into effect, in parts of the government where highly sensitive national security information is handled on a steady basis. So, for example, no screening polygraphs are given to State Department employees at any level, or to officials in the national security apparatus at the White House, or to members of the defense and intelligence committee staffs in the Congress, although many of these persons have access to much of the same information as intelligence agency employees, or to equally sensitive information. Even in DOD, the program has a very spotty application, if only because of the numerical limit on screening polygraphs imposed by the Congress. Among other things, high-ranking civilian employees are essentially exempt, and many high-ranking military personnel are also unlikely to be affected.


If the programs are truly important to the protection of national security information, the question that obviously waits to be asked is why the programs don't have more general coverage and acceptance. If they are needed in one place, why not in another? The Commission's report never asks this question. Instead it cites, and singles out for criticism, various differences in the ways in which polygraph screening programs are administered at CIA and NSA. These differences are small matters, however, compared to the double standard that exists by virtue of the fact that such programs are used in one form or another by both these agencies, and seen by both as indispensable security measures, but are not used in any form by other agencies whose personnel have access to the same or equally sensitive information. From a broad policy perspective, it is this double standard, not the much more minor differences cited by the Commission, that has real significance, because it points to a security system that taken as a whole is lacking in coherence and logic.


I am frankly at a loss to know where any of this leads, but there is at least a need to raise these considerations and make them part of the debate.


Certification. The Commission's next recommendation is that "certification of polygraph examiners under the auspices of a single entity should be mandatory" and that "mandatory requirements for recertification also should be established." I do not know what this recommendation means. As I understand it, polygraph examiners who complete the training curriculums at the DOD Polygraph Institute or at the CIA polygraph school already receive certificates reflecting their successful completion of training programs approved by the American Polygraph Association. Further as I understand it, that Association views these programs as the finest of their kind in the country. I agree of course that superior training is a must, because competence and professionalism on the part of examiners are key elements in any polygraph program, but here again I have no basis to be critical of the way in which DOD or CIA polygraphers are trained, and the report provides no such basis.


National polygraph institute. The Commission's next recommendation is that "the CIA polygraph school be consolidated into the DOD Polygraph Institute to form a national polygraph institute that would conduct all training and certification of government polygraph examiners." This recommendation does not appear to have any cost cutting rationale, since none is mentioned in the report. Instead the stated objective is to "enhance the quality of polygraph training provided by the government." If such was the likely outcome, I would favor the recommendation, but here again the report provides no supporting reasons that point to such a likely outcome, and the recommendation has the feel of one that was made just for the sake of moving some furniture around.


Research. The Commission's last recommendation is that "a robust interagency-coordinated and centrally funded research program should be established with DOD/PI as executive agent," and that this program "concentrate on the development of valid and reliable security and screening tests and standardize their use." I have already said that I am a strong supporter of further basic research. DOD/PI already conducts a broad research program, however, and I am not sure how the Commission would want to see this program redirected. Nor do I understand how it could be the function of any research program to "standardize" the use of polygraph tests. Only management decisions could have that result. Further, the wording of the recommendation suggests by implication that polygraph screening tests, as currently administered, have no validity or reliability, and I do not agree with that implication, which may not have been intended.



Closing thoughts


I am not blind to the fact that screening polygraphs, for many people, are hateful experiences. The one such test that I took in my own life, which was one of the full-scope models, was certainly no picnic. It is only natural for people to think of themselves as patriotic, and fit to serve in government positions of trust should the opportunity to do so come along. All probably resent the idea that their honesty or integrity might be impugned by a polygraph examiner armed with a set of form questions and a strange technology. But there are higher stakes here, because mistakes can have fateful consequences for the country. Somewhere among us (no reference here of course to any members of the Commission) there are some bad apples. Others among us, whatever we may think of ourselves, do not meet the standards of reliability and trustworthiness that the government is entitled to set, and indeed must set if there are to be any personnel security controls at all rather than a system in which all comers are accepted, no questions asked. The standard-setting alone is a difficult job, and judgmental to the core. So is the sorting process. I end up believing that polygraph testing is a reasonable step in that process.


I am also well aware of the fact that polygraph testing has a high potential for abuse. There are few clear roadsigns here, however, and except in obvious cases, as for example if an examiner pursues unauthorized lines of inquiry, abuses are hard to define. I favor an effort to develop an agreed set of ethical guidelines, beyond any that exist today, that would apply to the conduct of screening polygraphs. I also favor the other steps to which I have referred in this statement, but in substantial part I do not favor the Commission's recommendations, and for that reason and the others I have already stated, I concluded that I could not join in the Commission's report.



Appendix D.





AECA Arms Export Control Act


ASPP Acquisition Systems Protection Program


ASPWG Acquisition Systems Protection Working Group


ASSIST Automated Systems Security Incident Support Team


C3I Command, Control, Communications, and Intelligence


CCISCMO Community Counterintelligence and Security Countermeasures Office


CCVS Central Clearance Verification System


CERT Committee of Emergency Response Team


CI Counterintelligence


CIA Central Intelligence Agency


CIO Central Imagery Office


CISARA Counterintelligence, Security Countermeasures and Related Activities


CMS Community Management Staff


COPS Committee on Physical Security


COTS Committee on Technical Security


CSE Center for Security Evaluation


CTC Counterterrorist Center


CTTA Central TEMPEST Technical Authority


DCI Director of Central Intelligence


DCID Director of Central Intelligence Directive


DCII Defense Clearance Investigations Index


DDEP Defense Development Exchange Program


DIA Defense Intelligence Agency


DICOB Defense Industrial Security Clearance Oversight Board


DII Defense Information Infrastructure


DIS Defense Investigative Service


DISA Defense Information Systems Agency


DISCR Defense Investigative Service Clearance Review Office


DoD Department of Defense


DoDD Department of Defense Directive


DoDPI Department of Defense Polygraph Institute


DoDSI Department of Defense Security Institute


DoE Department of Energy


ENTNAC Entrance National Agency Check


EO Executive Order


FBI Federal Bureau of Investigation


FFRDC Federally Funded Research and Development Center


FOIA Freedom of Information Act


FOCI Foreign Ownership Control and Influence


FORDTIS Foreign Disclosure and Technical Information System


GAO General Accounting Office


G&A General and Administrative


GOVIND Government-Industry Restricted Information


GSA General Services Administration


IACSE Interagency Advisory Committee on Security Equipment


INFOSEC Information Systems Security


IOSS Interagency Operations Security Support Staff


ISOO Information Security Oversight Office


ISM Industrial Security Manual


ISPG Intelligence Programs Support Group


LIMDIS Limited Dissemination


MASINT Measurement and Signature Intelligence


NAC National Agency Check


NACI National Agency Check with Inquiries


NAG/SCM National Advisory Group/Security Countermeasures


NCS National Communications System


NDP National Disclosure Policy


NDPC National Disclosure Policy Committee


NFIP National Foreign Intelligence Program


NII National Information Infrastructure


NISP National Industrial Security Program


NISPPAC National Industrial Security Program Policy Advisory Committee


NIST National Institute of Standards and Technology


NOAC National Operational Security Advisory Committee


NOFORN Not Releasable to Foreign Nationals


NPC Nonproliferation Center


NRO National Reconnaissance Office


NSA National Security Agency


NSD National Security Directives


NSDD National Security Decision Directives


NSTISSC National Security Telecommunications and Information Systems Security Committee


OADR Originating Agency's Determination Required


OMB Office of Management and Budget


OPM Office of Personnel Management


OPSEC Operations Security


ORCON Dissemination and Extraction of Information Controlled by Originator


OSD Office of the Secretary of Defense


OSPG Overseas Security Policy Group


PERSEREC Personnel Security Research and Evaluation Center


PEP Personnel Exchange Program


PROPIN Proprietary Information


PSEAG Physical Security Equipment Action Group


PSWG Personnel Security Working Group


R&D Research and Development


REL TO Releasable To


SAP Special Access Program


SARF Special Access Required Facility


SCI Sensitive Compartmented Information


SCIF Sensitive Compartmented Information Facility


SCM Security Countermeasures


SIGINT Signals Intelligence


SIOP Single Integrated Operations Plan


SOR Statement of Reasons


SPECAT Special Category


SSA Special Security Agreement


SSBI Single Scope Background Investigation


SSII Suitability and Security Investigations Index


TEMPEST Transient Electromagnetic Pulse Emanation Standard


TIARA Tactical Intelligence and Related Activities


TS Top Secret


TSCM Technical Surveillance Countermeasures


USSS United States Secret Service


WNINTEL Warning Notice-Intelligence Sources and Methods Involved


Appendix E.




The Joint Security Commission is pleased to thank the following individuals and organizations for advice, counsel, and support in the preparation of its report:


AEGIS Research Corp.

Aerospace Corporation

Aerospace Industries Association

American Bar Association

American Civil Liberties Union

American Defense Preparedness Assoc.

American Federation of Government Employees

American Polygraph Association

American Society for Industrial Security

Analytical Systems, Inc.

ARCA Systems

Armed Forces Communications Assoc.

Arthur D. Little Corp.

AVCO/Textron Defense Systems

BDM International, Inc.



Bolt Barenek & Newman, Inc.

Booz-Allen & Hamilton, Advanced Decision Systems

Bristol-Myers Squibb Co.

BTG, Inc.

Central Imagery Office

Central Intelligence Agency

Charles Stark Draper Laboratory

CODEM Systems, Inc.

Communications Security Establishment of Canada

Computer Sciences Corporation

Contractor SAP/SAR Security Working Group

C. S. Draper Labs

Cray Research, Inc.

DCI Center for Security Evaluation

DCI Community Management Staff

DCI Counterintelligence Center

DCI Counterterrorist Center

DCI Non-Proliferation Center

Defense Information Systems Agency

Defense Intelligence Agency

Defense Investigative Service

Department of Energy

Department of Defense

Department of Justice

Department of State

DoD Polygraph Institute

Electronic Warfare Associates, Inc.


E-Systems, Inc.

Federal Bureau of Investigation

Federation of American Scientists

Galaxy Computer Services, Inc.

Dr. Robert Gates

GDE Systems, Inc.

General Dynamics

General Electric Co.

General Research Corp.

Grumman Corp.

GTE Government Systems

Hoffman-LaRoche, Inc.

Hughes Aircraft Co.

Hughes Information Technology Co.


Information Security Oversight Office

Intelligence Programs Support Group

Department of Defense

International Information Integrity Inst.

ITT Aerospace

Joint Chiefs of Staff

Knoll Pharmaceuticals

Knollsman Instruments, Inc.

Litton Systems, Itek Optical

Lockheed Missiles and Space Company

Lockheed Sanders, Inc.

Logicon Ultrasystems, Inc.


Massachusetts Institute of Technology, Lincoln Labs

Martin Marietta

Mattel Toy Company



MRJ, Inc.

MVM Group, Inc.

Mystech Associates

National Classification Management Society, Inc.

National Communications System, Office of the Manager

National Federation of Federal Employees

National Institute of Standards and Technology

National Intellectual Property Law Inst.

National Reconnaissance Office

National Security Agency

National Security Archives

National Security Council

National Security Industrial Association

National Treasury Employees Union

Naval Criminal Investigative Service

Naval Post-Graduate School


Office of Government Ethics

Office of Management and Budget

Office of the Secretary of Defense

Office Technology Assessment


President's Foreign Intelligence Advisory Board

Rand Corporation

Raytheon Co.


Dr. Roger Schell

Schering Plough

Secure Computing Corporation


Security Affairs Support Association

Software Products Association

SRI International


Treasury Board of Canada

Trusted Information Systems


United Technology Corporation

US Air Force

US Army

US Atlantic Command

US Central Command

US Coast Guard

US House of Representatives

US Marine Corps

US Navy

US Secret Service

US Senate

US Space Command

US Special Operations Command


United Technologies

Vitro Corp.

XEROX Special Information Systems

Back to JSC Top Page

FAS | Secrecy | Library || Index | Search |