CFP'92 - Who's in your Genes?
Thursday, March 19, 1992
Chair: Madison Powers, Kennedy Center for Ethics, Georgetown
Panel: John Hicks, FBI Laboratory
Paul Mendelsohn, Neurofibromatosis, Inc.
Peter Neufeld, Esq., Chair, National Association of
Criminal Defense Lawyers Forensic DNA Project
POWERS: Good afternoon. My name is Madison Powers from the Kennedy Center for Ethics at Georgetown. I have several pieces of bad news for you at the front end. Two of our speakers who were listed on the program have been detained for reasons of weather and physical ailment. Tom Marr, from the Cold Spring Harbor Laboratory, will not be here today; and Phil Reilly, who was scheduled to be our Chair -- I learned 30 minutes ago that he has been snowed in in Boston. So I'm going to quickly take over in that role, and hope I know what I am doing for the rest of the afternoon. I trust that, given the remainder of our distinguished panel, I can't go too far wrong given who we have to speak before us. When I spoke with Phil last, he did say that he wanted me to open up, because Phil knew that I was a lawyer and a philosopher, and that that combination is bound to confuse everybody in the room. So we thought that we would get that out of the way at the front end.
More charitably, he hoped that I would do something along the lines of setting the stage to think about the kind of ethical issues that pervade our concerns about genetic privacy, and then think about how those ethical issues translate into concerns about the way we formulate public policy and our legal system's responses to concerns about maintaining privacy of genetic information.
Let me just start by saying that any kind of discussion about the ethical issues always raises deep concerns about how to approach a subject like this. If any of you have, to any degree, looked at moral philosophers, you know that that is one group of people who don't even agree on what the debate is about. The problem is that there are many, many different kinds of moral theories, and they each subscribe to different ways of looking at some of our most fundamental ethical problems. So what I will try to do is provide a kind of analysis that is meant to be largely, though perhaps not entirely, neutral among the range of competing ethical theories. I also know that many of you who have thought about privacy issues, because of your involvement with computers and other issues, know that many people spend a lot of time and spill a lot of ink thinking about the definition of privacy itself. I will not try to go through that conceptual morass other than to give you a very simple-minded, but hopefully useful, account of privacy which will aid us through the discussion that we are going to have this afternoon.
I should note that the account of privacy that I will mention will be one that suggests that the privacy notion is a highly metaphorical but very useful unified conception, a way of finding our way around a nest of philosophical and ethical problems. And I say it's a kind of metaphor, because it's meant to capture a wide range of different, underlying interests that human beings have that are all sort of caught under this umbrella of privacy. What I hope to do is think about some of those different kinds of interests that are at stake -- to sort of perform a bit of an analytical map of the different morally significant interests that we have in the protection of privacy of genetic information.
Now when I say that there are different interests, what is significant about drawing that map among a number of different kinds of morally significant interests, is that it means that with other things being equal, these different kinds of interests ought to be promoted or protected in the design of our public policies and of our laws. So, if my thesis is correct, and that is that there are a different set of interests at stake under this broad rubric of privacy, it may be an illusion to think that one particular way of formulating public policy, or one set of laws, is going to be adequate to protect and promote all the different kinds of interests that fall under this large umbrella concept.
Now when I say "other things being equal," you also know that other things rarely are equal. So when we identify some interest that lies behind our concerns with privacy, and say that other things being equal, they ought to be promoted and protected, what are we saying? Well, we are saying that they are of such importance that they, at least, have a very high threshold weight when they are to be balanced or put into comparison with other kinds of interests -- whether they be interests of other individuals, interests of society as a whole, or the like. So if we take the notion of privacy seriously and we take the notion that there are a number of different interests that stand behind this conception of privacy, then our task is to try to give them some special weight, and design our public policies and our laws that will take those interests very seriously.
Let me start, then, with what I describe as a simple, if not simple- minded, conception of privacy relevant to this concern. Invasions of privacy occur when there is some reasonable expectation of non- intrusion into some sphere of a person's life, and in fact there is some intentional or inadvertent intrusion into that sphere. We think about privacy, we think about it protecting against, you know, people peeking through windows, people eavesdroping through microphones, people looking at what's in our electronic mail, people looking at a variety of ways of interfering in some sphere of our choice, of our information, of our life. When we think about privacy in the informational context, we're concerned about a special kind of unwanted, unwelcome intrusion, that someone getting their hands on information that we think we have some prerogative to be in control of. Or that we think we have some interests of substantial importance that this information should not be widely disseminated, and certainly not to people that we do not think should have it. To be concerned with privacy in the informational context is to be concerned with access that others have to us. And one of the most important kinds of access, and certainly one that we're concerned with here, is access to personal or intimate information -- information that may have a lot of bearing on how well our lives go, whether our interests are being served, whether we have opportunities, or whether those opportunities are being frustrated.
So let me then turn to four morally significant interests that I think ought to be reflected, at least in some measure, in any adequate public policy. The first two I include under the heading of intrinsic moral interest. What I mean by that is the following: we attach some inherent moral importance to our ability to limit and control access others may have to highly personal, sensitive, intimate information about ourselves. It is an intrinsic interest, because we're not just concerned with the consequences. Many times we're concerned about control of information, because we're concerned directly about predictable bad consequences. There are also additional kind of interests we have. We want to make sure that we're in control of information. It's part and parcel of our identity; it's part and parcel of our concept of self-control or autonomy.
The interests that I think of under this category of intrinsic interest would fall into several areas. First, I would argue that we have two kinds of rights that flow from these interests, which I collectively call the right of informational self-determination. The first part of that is this: we have an interest in controlling what information gets generated. That is to say, I want to decide for myself what information I or others may know about me. Notice I included myself in here. I want to decide for myself what information I or others may know about me. Secondly, I want to weigh for myself the risk of unwanted disclosure against the medical and other presumed social benefits of knowing such information. It's not enough always to be told that it is in my best interests that this information be shared. Or, it's in my interest that I know, and that I should know, certain bits of information, and have this information generated, regardless of whether I want the information or not. Without even thinking forward into the genetic context, you can already put yourself in an imaginative position to think about whether you want for yourself genetic information, say about Huntington's disease, a disease that has onset later in life that will not affect the way your life goes in a substantial way until that later date of onset. Do you want to know the future? That's one way of thinking about it. And if there is an interest we have in controlling our lives, an interest in autonomy, then one of the concerns we may have is to decide what we want to know as well as what others would want to know.
The second intrinsic interest that I have in mind is the interest in controlling or limiting further access to information once we have decided that it should be generated, and once we have already revealed that information to designated persons for very specific purposes. We may give that information to some people, but it's not necessarily a carte blanche or a license to give it to whomever else, or to use it for whatever purposes I might not have comprehended or contemplated at the threshold. Now that's an important element of privacy and confidentiality protection that's often noticeably absent from the laws in the medical context. We don't really, in most of our laws, give people a lot of control, in many contexts, for what kind of information can be generated. We tend to give people more control over how it can be used later, or we tend to give people some remedies if things go badly and information is spread in places we don't want it spread and it rebounds to our detriment.
A third morally significant interest that I want to mention is that we have an interest in information being inaccessible to others, even when our concern is not with controlling the information ourselves. Sometimes we cannot know for ourselves in advance what information ought to be protected from distribution, and so what we want is not necessarily that we control things, but that we want some protections and safeguards to be put in place, so that mistakes or the like will not happen in the future. DNA data banks provide one good example in this case. DNA data banks, as Phil Reilly likes to say (and I'm always supposed to speak about what Phil would always say), Phil always says, "There is no such thing as a data bank. We don't know what one is. We don't know if it's a few blood samples in a doctor's drawer somewhere, whether it is computer-generated digitized profile; whether it is a data bank containing all that computerized information, or the like."
What is striking about genetic information is that it reveals to people things that we never contemplated at the time when we give permission for some people to have that information. It tells things to people in the future that we do not know now and cannot reasonably predict. Every Tuesday in the New York Times, Gina Kolata writes another story about some putative new discovery of a genetic marker -- one week it's alcoholism, one week it's another. The next week the marker that was previously discovered was a mistake and there is no marker after all. Great concern -- I give away a genetic profile, and only in the future do people know what it means. I give people the key to a lot of information about myself and I do not know what I'm giving away. So my concern is that when information is generated, there be protections in place long past the point when I have control. Control is not everything that we are concerned about, although it often seems to be the watchword in privacy concerns.
The last interest I want to mention is that there are consequential interests. Even when we can't control the information, we want some kind of protection, some sort of remedy, in case the information gets into the wrong hands. We want some ability perhaps to sue or to have criminal prosecution against some people who misuse information or obtain it incorrectly.
In conclusion, I would simply say that we want a lot of things. Sometimes we want control of what information is generated; sometimes we want control of what's done with information and who gets it later, even after we have given it to some people. Sometimes it's not control that we want, but we want the information to be protected, perhaps in ways we never imagined. Finally, sometimes what we want is a remedy when control and protection fail.
There are many things that I could say, especially if I were going to speak for Phil Reilly about the nature of these data banks and the like, but that I'm not going to do. We're running on the time line since we started late. What I want to do is introduce the first of our next speakers who will tell us a great deal more about the kinds of banks themselves. You can think about how these kinds of moral issues fit into that.
Our first speaker is John Hicks. He's the director of the FBI Laboratory in Washington and will talk to us about the DNA law enforcement database. John...
HICKS: Those of you from the Washington, DC area may recall a case a few years ago -- a young girl by the name of Rosie Gordon. Rosie Gordon was a young ten-year-old child who was abducted during a very bold daylight abduction from her residence and is now dead. If law enforcement had the right tools in place at that time, there is the possibility that Rosie Gordon might be alive today. I'll get back to that case a little later in my talk.
There are three things that I would like to accomplish with you this afternoon. One thing is to tell you very briefly what law enforcement's interest is in DNA technology, particularly in data banks. I'd like to describe what's underway with respect to a national data bank that the FBI has been working on in cooperation with state and local police agencies around the country. Then I'd like to somehow draw a distinction for you between law enforcement's use of DNA technology and some of the other uses in the medical context that might be found for genetic information. I want to try to draw a clear distinction between a DNA data index used for law enforcement, and the so-called genetic databases that I think people are concerned may raise a number of very serious kinds of privacy issues.
First, let me just describe DNA technology. What it represents for law enforcement agencies is an identification tool much like fingerprints. In a forensic context, what we try to do is to review a crime as a transaction. In that transaction, something is typically exchanged between the subject of the case (the assailant) and the victim, particularly in violent types of crimes -- sexual assaults, homicides, and other kinds of assaults. The type of transaction that typically takes place could be hair and fiber-type things -- fibers from clothing, hairs, and very typically body fluids. As a consequence, in the FBI Laboratory, we began using DNA technology a few years ago. In December 1988, we examined about 5,000 cases. Overwhelmingly, the type of cases where DNA technology has the greatest utility is in the crimes of sexual assault and rape. Typically, we're analyzing the body fluids left behind by the subject in the case, typically the semen evidence.
Just to try and define the problem a little bit from a law enforcement context, every year there are over 100,000 sexual assaults in the United States. There are over 20,000 homicides -- again, where body fluids are the type of thing that is typically left at the crime scene which we try to characterize in some way so that we can identify the suspect or the defendant in the case. With respect to sexual assault and rape, because there is a lot of the publicity now concerning this type of crime, there is reasonable basis to believe that a fair number of sexual assaults are not reported. So these numbers, if anything, are on the conservative side. These are the rapes that were reported to the police. About 53%, or just about half of them, are cleared by arrests. That means there are more than 50,000 sexual assaults and rape cases in this country that do not have any suspects or defendants identified. No one has been charged with respect to the arrest. And figures for that type of crime have remained steady. The 1990 figures have increased from last year; and this year's figures have increased slightly over last year.
Another aspect: we know from criminal behavior, just based on a variety of studies, that certain behavioral types commit certain crimes, and sexual assault is one of those. The individuals who engage in those violent types of crimes have a fairly high recidivism rate. Just within the FBI Laboratory, I mentioned that we've completed about 5,000 cases. We gathered these figures, based on having completed testing in 2,900 cases, then we went back and re-examined those cases. These were cases that came to us from all over the country, typically from state and local police agencies -- these are not FBI investigations per se. The FBI Laboratory simply provided technical support to those agencies. We found that in those cases that were presented to us, 20% of the cases typically involved serial offenses -- in other words, the same suspect was believed to be involved in more than one of those cases. There's a lot of other data that demonstrates that kind of recidivism rate.
As a result of the recidivism rate, there are now 17 states that have laws which call for any individual convicted, typically of a sex offense (although the laws vary somewhat from one jurisdiction to another), that upon conviction, that individual must provide a blood sample. The blood sample will then be used to develop a DNA type. The DNA type is typically expressed as a number much like a very long Social Security number. That type is retained by law enforcement agencies, so that if and when that individual is released on probation or parole, the law enforcement agency already has the index information in the event that individual gets involved in that kind of activity again.
So, for example, if a rape occurred in some jurisdiction, one of the first investigative steps that might be taken -- particularly in the jurisdiction that has one of these types of laws -- is that the crime scene material, the semen evidence (typically a vaginal swab taken from the rape victim) would be taken by the medical people and turned over to a forensic laboratory for testing. The DNA type would be developed, and then one of the first investigative steps might be to compare it with the DNA types found in this index to determine whether or not this crime may have been committed by someone who has previously been convicted of that type of offense. These are not arrestees -- these are people who have been through the judicial process. These are the people who have been convicted, not suspects.
When you are dealing with technology like this, if you're going to try to share this information within law enforcement, it's very clear that we have a very mobile society. Even within the Washington, DC area, you have more than a dozen investigative jurisdictions. And they don't routinely share information back and forth about potentially random events that might have occurred in their areas. So, in order to try to establish the ability to do that, you have to have compatible standards. You have to have a degree of coordination, and you have to have a communications mechanism to accomplish that. That's a role we believe that the FBI as a national agency can fulfill. We can facilitate the establishment of the standards necessary to permit the exchange of investigative information, crime scene information for these types of data files. The FBI, with its own staff and also using contractors, has developed standardized software that reads the image so that it comes out in a standard format. The interpretation of the DNA pattern is done using a computer. The FBI also provides software to manage and store that information, and is also promoting the establishment of a communications network to exchange that information between crime laboratories typically associated with law enforcement agencies.
Again, as I said, the critical element here is compatibility and reliability, so there has been a considerable effort put into a cooperative effort with other crime laboratories around the country to build reliability into the system. We have the group which is referred to as the Technical Working Group on DNA Analysis Methods, which has defined quality control standards, proficiency testing standards, and a number of other procedures to assure the reliability of the data that goes into the system. We also provided, with respect to the last item, legislative guidelines. Because we were seeing different states where legislatures with an interest in that type of a system defined their own laws, some inconsistencies existed in the way some of the legislation was being drafted. We have tried to provide some general guidelines. We're not drafting legislation for any state, and we feel that we need to be somewhat careful in the role we play here. We try to identify some issues that other states have dealt with in their development of legislation on this issue. So we do have legislative guidelines available; we do have prosecutor assistance available because it is a complex technology. You may have read some of the controversy in the press -- it has been subject to a scorched earth attack by the defense and particularly obnoxious lawyers, I might add. No offense intended, Peter. (laughter) But we have provided some prosecutor assistance, again, to help prosecutors who are dealing with a lot of cases at any given time, to try to be prepared quickly with the kinds of issues that they will be confronted with by the defense with respect to presenting this novel kind of evidence.
The system that we refer to as CODIS -- the acronym for Combined DNA Index System -- is truly an index. There is no personal information in it. It simply is an index of identification or identifiers for individuals, and it has no names of individuals in it, for example. It is strictly an identification index.
There are three basic purposes that we try to accomplish. One is to look at different crime scenes. Let's say that a rape case occurs in Arlington across the river, and another one occurs in the District. The evidence from those two crime scenes might be collected and the DNA type established. We might be able to indicate, based on that, that the same offender is at work. This would then cause the two jurisdictions to coordinate their investigations and hopefully identify the offender more quickly. We also, as I described earlier, try to make associations with persons previously convicted and released on probation or parole.
In developing this system, now in the pilot phase, we are working with ten different laboratories in seven states across the country. We expect that by this time next year that this will be an operational system. Right now the system has the capability of exchanging data to interpret the significance of DNA test results. It is not now used as an investigative tool, but in about a year, we expect that it will be. There are now about twenty-five or so crime laboratories around the country that have the capability of doing DNA testing, and we've trained about 300 technicians to this point. We expect over the next year or so, you'll see another dozen crime laboratories -- again typically associated with state and local police agencies. We don't work directly with private laboratories with respect to DNA information.
There is a population file, which is an administrative file, and again, there is no personal information in it. It is strictly a collection of types from various population subgroups that's used to develop a statistical estimate of the strength of a particular DNA match. This is an area, as it turns out, that has been subject to so much controversy in the media and some courts -- this statistical interpretation of whether your likelihood of a coincidental match is one in a million, or one in ten million. There has been a lot of dispute about which is significant, one in a million, or one in ten million. We would argue in most courts that one in a million is a pretty convincing figure, particularly when it is associated with other evidence that the jury hears, such as the victim's pointing out that that is the individual that assaulted her, and the physical evidence corroborates this testimony with a high degree of accuracy.
I have already described the other files. The missing person file, the last item, is a technique that we see as potentially of use in law enforcement applications. It could be useful where, at a crime scene, or at a disaster scene such as an aircraft crash, the authorities are going through the process of trying to identify remains. Many times the remains have been mutilated, burned perhaps, so that typical identification techniques, such as fingerprints, can't be carried out. It's possible to take tissue from remains and associate that with blood relatives, at least people who are believed to be blood relatives, and through a paternity-type test establish the identity of the remains. This is one of the primary interests that the military has shown in developing a DNA data file to identify war dead. In fact, it was used in the Desert Storm operation to identify some of the remains that came back.
The kinds of information typically found in a file -- and the reason I stress this is because of some of the issues associated with individual privacy and concerns about how this information might be potentially abused if it got into other than law enforcement hands, perhaps into the hands of an insurer or a potential employer -- are only specimen identification numbers, some information with respect to the particular analytical procedures used, and then the typing data for that particular profile. It typically has a reference back to the particular police department that entered the data. It's at that point that the particular department makes a determination as to whether or not to release the information, based on an inquiry from another police department.
The files will not contain any personal identifying data -- no names, no criminal information such as that typically associated with the fingerprint records that the FBI retains. It will not contain anything that will imply in any way any disease condition or any genetic disorder that may be in the individual. It won't contain any ethnic, racial, or social information. It won't contain suspects. It's simply the information that I described earlier. Just several more minutes and I'll finish up.
Safeguards are built into the system. We believe that we tried to be very specific in how we developed it. We're developing a system, trying to take advantage of exploited technology, to be useful as an investigative tool to help police be more efficient in their operation, to help clear up some of those 50,000 rape cases. There are some who believe that many of those cases can be associated with people who are in custody and perhaps have been convicted of other offenses. The problem is that the victims of those cases don't have the satisfaction of knowing that their assailants are in custody or under control.
The system is accessed only through law enforcement agencies. The accessing agency has to be running a profile itself in order to make that inquiry, so it can't just make an inquiry or order information from the system. It has to actually do a query using its own DNA profile that it has generated, so that in itself is a fairly significant type of control.
We have a number of other system security type things, typical things that you do to try to manage information: some encryption-type systems, and again, limited access to physical space as well as to the information in the computer.
I want to finish up by mentioning Rosie Gordon. Those of you who recall the case -- when the crime occurred, one of the investigative approaches they suggested was, because it was a daylight crime in which a ten-year-old had been abducted, that the local jurisdiction might go back and examine other unsolved cases within the jurisdiction which might have had similar circumstances -- a very high risk crime for an offender to engage in. They found four cases where young girls typically of the same stature and same age had been abducted, sexually molested, and then released. When they tested the DNA, they found that the DNA from those four crime scenes matched, indicating the same offender had committed these crimes that occurred over a four-year crime period. When they then went back to the individual jurisdictions, they were able to find pieces of information that helped lead them to a particular suspect. One victim, for example, remembered the number on the inspection sticker in the vehicle. Another victim remembered an unusual tattoo that the individual had. Another of the victims remembered other features. Gathering that information, the investigators came up with a suspect who had a record of offenses up in the northeastern part of the country. He is now serving four life terms for those crimes.
Unfortunately, he has not been associated with the Rosie Gordon case. He adamantly denies being associated with it. If you believe that he could be a viable suspect in that case, I think had this DNA data index been in place, it's conceivable by the time that first rape had occurred, typically not much would have happened beyond what happened at the routine investigation.
By the time the second crime occurred, the police would have been alerted that we had a serial offender at work. By the time the third case occurred, potentially we might have had sufficient information to have identified the suspect before he committed the fourth rape, and perhaps before he assaulted and murdered Rosie Gordon. Thank you. (applause)
POWERS: Our next speaker is Peter Neufeld, who is the chair of the National Association of Criminal Defense Lawyers Forensic DNA Project. He's a member of the New York State Governor's Commission on Implementation of DNA Profiling Forensics. He has litigated many of the key cases in this area of forensic DNA use.
NEUFELD: For those of you who don't know me, I'm just better known as one of those obnoxious lawyers that John Hicks referred to. There are a few points that I want to make with you folks this afternoon. Those points are going to be made from a limited perspective, because I'm not an ethicist or a geneticist. I really don't know a lot about computers, but I have been an observer in the transfer of this technology to the criminal justice system, and I am somewhat familiar, on a first-hand basis, with some of the civil liberties problems that arise.
First of all, what's important is to think about an analytical context for examining these issues and these questions. I think that John Hicks tried to approach you on a very emotional level -- the story about Rosie Gordon. Let me just ask you briefly: who here in this room has been a victim of violent crime? Who here in this room harbors a fear that either you, or someone you love or care about, will be the victim of violent crime? OK. And who here would think that it is totally appropriate for law enforcement to use whatever reasonable means available to identify and apprehend rapists and murderers? When the questions get posed in that way, they invite yes answers.
If, instead, you try to pose the question a little bit differently: how many of you would be in favor of a law that would require each of you to donate a sample of your DNA, your children's DNA, and your loved ones' DNA into a national data bank controlled by police, which would have regulations on who could access that data -- but we all know in terms of our own experiences that access is rarely denied -- and that that access might mean access to employers, insurers, or other social scientists or research scientists. I would guess, without asking you to raise your hands, that if I posed the question that way, that there would be fewer hands than answered affirmatively.
Naturally Mr. Hicks tried to frame the question in a way that would be the most appealing to folks in the acceptance of the need and the necessity for a national data bank of DNA prints. However, I think that the FBI approach to this rests on a series of assumptions, and it is those assumptions that need to be further studied. We have to analyze not only the benefits and the costs, but we also have to analyze the limitations of the technology. Is everything as perfect as it is suggested? Remember, computer science, or DNA typing, is not like the lore of physics. There are human beings involved, and as soon as people interface with some type of science there is plenty of room for screwing up; there is plenty of opportunity for abuse; and the possibilities of violations of privacy and other civil rights are rampant. Those are the kinds of things that we have to consider, before we rush headlong into some national policy which may be difficult to reverse in the near future.
What are these assumptions that he relies upon? The first assumption which was brought out in the slide show is that this national data bank will be limited to certain descriptive numbers. I believe that he used the analogy of the social security number, so that none of the critical genetic information that we don't want to share with other people, necessarily, unless we can control it, will be part of that data bank. I submit to you that that misses a key point in the process by which they arrive at these coded numbers. It doesn't just happen; they don't just pull these numbers out of thin air. Right now, the FBI takes the position that it cannot be the one responsible for doing the individual DNA testing on all the folks whose DNA profiles are going to comprise their database. They are leaving that responsibility to various state and local law enforcement agencies throughout the country. The states are not simply doing the DNA testing and creating this numerical code that will feed into this national database. One of the reasons that they're not doing this is that, unlike other computer applications, the scientists have not decided what methods are the most effective -- cost-efficient -- and have the most long-term usefulness for doing DNA typing. The methodology that the FBI is currently using for its forensic investigations is not the methodology that the FBI believes that they will be using two or three years from now. Mr. Hicks will be the first to admit that. Just from what you read in the newspapers every day, the whole area of genetics and molecular biology is a rapidly-evolving technology. So the FBI doesn't want to be caught investing a tremendous amount of money into antiquated methods; consequently what they are doing is actually storing the DNA samples. A lot of people who are concerned with the civil liberties implications and ethicists feel that it is inappropriate to retain someone's DNA sample. The opportunity for others to have access to the DNA information is greater if you do the test, get the anonymous identifying information (That's an odd term, "anonymous identifying information," but what they mean is that it can help identify a person by name, but can't identify a person by the genetic aspects of their makeup), and then discard the DNA sample. To a very large extent that concern is minimized, but that's not what's being done. You now have states -- I believe Mr. Hicks said there are now 17 states that have legislation -- that will start accumulating DNA samples from all these people in jail. What they're doing is they're accumulating the samples, but not even testing them. The reason they're not testing them is because they have not decided what methodology to use yet. For instance, California has already collected 25,000 DNA samples. Given their present people power -- forget their budget -- it would probably take them about seven or eight years, given current technology, to type those people and not even investigate new cases. So, they are waiting. Meanwhile, these samples from people which can potentially contain all that private and privileged information that we are referring to are just sitting there.
At this point in time, there is no meaningful control over what the states and local law enforcement agencies are going to do with DNA samples. The legislation varies from state to state. There has been an attempt by the federal government to intercede in this matter and set some kind of national standard, but at this point in time, there is no federal legislation on that front. Not only do most of the states have no rigid prohibition against what the DNA samples can be used for and who can have access to this data, but, more importantly, the laws use this kind of broad, very vague language saying that the information can only be used for law enforcement purposes. Well, one law enforcement purpose is, yes, to identify the perpetrator of a crime, but certainly in the broader sense it has been argued in the past in other contexts that it is a valid law enforcement purpose to investigate the causes of crime. It wasn't too long ago that people were running around saying that those folks who had an extra Y-chromosome had a propensity for violence. And that certainly is something that has always tickled the fancy of social scientists who try to deal with criminal justice issues in this country.
Several weeks ago in the newspaper, the Department of Justice came out with a study indicating that a majority of juvenile criminals, juvenile delinquents, and adult committers of crime come from families where there were criminals. And so the earlier notion we had that crime was learned from friends is not really true -- crime is learned from family members. Now it doesn't take a huge quantum leap for those folks in that area to say that it's not that they learned it from their family, it's that they inherited it from their family. Why don't we start doing linkage studies to see if there are certain genetic traits that generate a predisposition for violent behavior? Indeed, maybe we should monitor those people who possess those genetic traits that generate a predisposition for particular violent behavior. Now of course this is not something that is happening right now, and the technology is not there right now to define all these traits, but it is moving very, very, very rapidly. The question is, do you want to have the DNA samples hanging around and the access to them available to create that potential for abuse? That's what we are talking about.
The second assumption is, we're impinging on the privacy rights only of convicted criminals. You know everyone talks about our democracy. What makes the United States so unique is not that we are a democracy -- democracy has been floating around for about 2,000 years. What makes the United States unique amongst all other societies is, quite frankly, the Bill of Rights. And it's that Bill of Rights that protects everybody, even against the sense of the majority, even against the notions of democracy that the majority rules. And one of those concerns in the Bill of Rights is the right to be free of unreasonable searches and seizures. Another concern raised by the Bill of Rights is that fundamental right of privacy. Others will address those rights in greater detail, but to think that we are only going to look at criminals, when in fact that DNA might give us some kind of insight into tomorrow's criminals is something that we have to consider. Indeed there are some states right now, because there is no national control over this, which are saying they will collect samples from everyone accused of committing certain violent crimes with nothing built into the legislation to require the expungement of that data should the individual be exonerated at trial. Those are the kinds of privacy concerns and civil rights concerns that we need to be vigilant about.
Third assumption: since everyone's DNA is unique, there is no danger of the data bank prosecuting an innocent person. Well, forget what you all know about computer glitches and computer screw-ups and how those kinds of things can happen. There are fundamental problems and limitations to the technology which can lead to innocent people being prosecuted and being convicted. There is this notion out there when they do a DNA profile for forensic purposes, they're going to be typing all 3 billion base pairs that make up the human genome in any given person. That's not what happens. Rather, what they are doing is looking at three or four, maybe five, different regions of the human genome where there is some degree of variation between people. They are not looking at those regions, for instance, which determine eye color, or hair color, because there is not that much variation, but if you could imagine them looking at regions of the genome which determine height, for instance -- even though it's not height, I'm just using that for illustrative purposes -- then you have a sense of the regions they are studying. It's not all the genes, it's just a few of them. If that's the case, what is the likelihood, especially if you're dealing with forensic specimens where you might not be able to look at 5, 6, or 7 different genes, because they're degraded samples -- there's bacteria in there, there's all kinds of stuff that affects the quality of the sample. Then you run that through a database -- what happens? One thing that we know is there is a much greater likelihood that relatives and people who are related more closely to one another are more likely to share similar genetic patterns. Of course they are going to be unique if you study 50,000 genes, but you don't. You're only looking at three or four genes. We know we have data, for instance, on certain Indian communities where they look at a village of 55 Indians, and they find that two-thirds of all the Indians have three genes that are identical to one another. So what are you going to do when you start thinking about crime, and you start to realize that very often the victims of crime today live in urban communities where their relatives may be the alleged perpetrators; where they themselves who are the victims today may indeed be the perpetrators tomorrow. The likelihood of people being incorrectly identified in that kind of situation is not trivial.
The last point I want to make -- I'm skipping over some, but hopefully they will come up in the question and answer period -- is the fourth assumption. The FBI wants you to believe that they are the entity best equipped to guarantee the highest standards for use of this DNA technology, and that they are the best equipped to insure that privacy rights will be scrupulously protected. And I submit to you folks that they have not demonstrated the capacity for either at this time. At this point in time, there is no formal, external laboratory inspection of even the FBI's own DNA laboratory. They have been reluctant to participate in ongoing blind external proficiency tests at their own DNA laboratory. When 225 FBI agents were retested using their own DNA analysis, sixteen of them did not match themselves, so that deals with reliability.
The second issue, of course, is privacy. The FBI position on that is "Trust me," but you know that, as a matter of public position they have taken before, they are in favor of linking up all these different databases. They are in favor of sharing intelligence information between many different agencies, as long as they are the ones who select who the agencies are. More recently, in matters where privacy and confidentiality were very important, the FBI willingly received data that was supposed to be subject to those rules of confidentiality and privacy. Mr. Hicks himself, in fact, quite recently was the recipient of data from the National Academy of Science investigation, which was data that by their own internal rules was not supposed to be leaked to anyone. Yes, it's true that Mr. Hicks didn't leak the data, he was just on the receiving end. Did he give it back? Did he refuse to analyze it because it was inappropriate? No! Instead he disclosed it to other people who were not permitted to see that data, because he thought that it was expedient and appropriate. These are serious questions, because we have to decide whether or not this is the appropriate entity to take on this very considerable and very serious obligation. One thing we do know is that in somewhat of an analogous situation, with the NCIC index, there is legislation which allows the FBI to terminate the service for states that do not rigorously uphold the privacy concerns. It's been pointed out by the Office of Technology Assessment of the United States Congress that because the FBI has no active enforcement process, compliance with these privacy provisions was mostly voluntary.
These are some of the questions that you are going to have to consider. Hopefully they will generate some questions from you folks, and we'll take it up then. Thank you. (applause)
POWERS: Thank you, Peter. I knew there wouldn't be any disagreement here today. (laughter) The next speaker is Paul Mendelsohn. He is the president of the Neurofibromatosis Corporation. Did I get that out right? You bet that I wouldn't -- it's close. It's a voluntary organization representing the consumer viewpoint on questions of this sort. Professionally, he is involved in the Medicaid program in charge of quality assurance in the n-stage renal disease program. Paul?
MENDELSOHN: Well, first of all, I'm on this panel as the consumer and I didn't realize that what that would mean is that I would have to present myself as the felon. I'm not here to represent the debate between these two gentlemen, but to present you a totally different perspective to "Who's in Your Genes?" With that provocative topic, I can continue.
I represent a voluntary health organization made up of families and professionals dealing with the genetic group of conditions known as Neurofibromatosis. Most of you have never heard that word before, but many of you may have heard the term "Elephant Man" and the "Elephant Man's disease." It got some popularization through a dramatic play a number of years ago. Furthermore, it seems that NF was chosen as a popular topic for situation dramas. Several years ago on "L.A. Law," they ran an episode about a doctor who had Neurofibromatosis and was denied employment; and "St. Elsewhere" had a series of shows about a patient who endured much abuse and went through an enormous amount of surgery because she had Neurofibromatosis.
To tell you why I am here, it is because modern technology has created and is creating a growing database -- databases that contain genetic information. Because of what our society has been, is now, and will continue to be in the future, this information can jeopardize the lives of people and place them in vulnerable positions which they do not deserve. On the other hand, and we have to present a balanced perspective, we have to balance out the society's need to know against the individual's right to privacy. Let me say in hearing these forensic discussions, that the genetic databases that exist today, unlike any other databases that I know, without anyone touching them, will be voluminous in the future as we develop more and more genetic information technology to analyze those samples. So Mr. Hicks' FBI database may be charting on 12, 15, even 100 aspects of the human genome. In the future, the computer may be able to take that same tissue database that he has and make better than three billion different identifications. In fact, most of you are already in a genetic database. Anyone who was born in the last 25 years -- unfortunately I can't put myself in that group -- was probably as an infant given a PKU test. They prick the heel of an infant when it is born to assure that it does not have a very, very rare genetic condition called phenylketonuria (See, I have to deal with Neurofibromatosis, I've gotten that, but PKU is harder). Those blood spots still are maintained by most states and can be used to do genetic modelling.
I talked about the Elephant Man and the Elephant Man's disease -- there's a lot of argument questioning whether the Elephant Man, who died 100 years ago, had Neurofibromatosis. Now that's only an intellectual argument, but in fact his skeletal remains still exist at the London Hospital, and as soon as we have a reliable genetic marker for NF, we will be able to confirm whether or not the Elephant Man had Neurofibromatosis. I am just demonstrating to you how significant the database is, and how powerful a tool it is. The other thing that I say about Mr. Hicks' database as a model of others -- and I'm not here to attack Mr. Hicks ... God forgive me, attack the FBI? -- the issue that exists is that the information in that database is not only the government's, possibly, is not only the individual's whose information is in it, but it also belongs to the descendents of that individual. I have a partial property right, I believe, to the genetic information of my parents, because that is mine, that is my genetic history, and it says something about me.
All right, I want to get down to absolutes and I'm here to present the real life situation. I was involved with Neurofibromatosis because my wife developed in her mid-twenties a condition called Neurofibromatosis Type 2. She was perfectly healthy up until that time, had no idea what was going to happen as that, too, is generally a late-onset disease. My wife is now totally deaf, legally blind, in a wheelchair and significantly debilitated. That doesn't mean that she is not functioning, because, by golly, she is clearly a very, very functioning individual who does a lot of individual advocacy herself. What could now happen with NF2 modelling -- which does now exist with diseases like Huntington's -- is that an employer could deal with that genetic information and determine that they would not employ her. An educational school system might look at genetic information to see where it could put this child in a school system or an educational program. There are both good and bad balances. People who are in jeopardy of NF are denied health insurance and life insurance. They are frequently denied credit. Much of this is from derived databases, not actual databases. Data that is not currently being protected in any great way.
The thing that is so interesting about NF and why it's used to characterize the problem is that NF has quite a different variation in penetrance, in other words, how it affects individuals. Many people have NF. You may not have heard of it, but it affects as many as one in 2,000 births. Fifty percent of the cases of NF are spontaneous, meaning their parent did not have it; but once someone has it, they have a 50 /50 chance of passing it on. So it is genetically and statistically very significant. Many of the people who have NF have such mild cases that they don't even know it. In fact, it's interesting, this morning I was at the coffee break, and I saw an individual here -- I'm not a physician -- who I think clearly has NF and doesn't know it. He asked me what NF was. He had all of the characteristics that generally symbolize the symptoms of someone who has NF, and the doctors tell us that you only need two of these. So by visually looking at someone, I can establish a database of whether a person is likely to have a disease or a disorder.
People are so afraid of society today because of the limitations of having a stigma -- this is becoming clearer with the AIDS crisis. We now have people who don't even want to involve themselves with the diagnosis. They are afraid. We send out newsletters to some families in brown paper wrappers, because they don't want other family members to know that this disease exists. Huntington's disease, which is another late-onset disease, now has a genetic marker, and there's a better than 90% accuracy rate in identifying whether the person will get the disease or not. Yet most of the people who are at risk for Huntington's disease are currently refusing the test, partly because of psychological factors, but also because of the sociological factors of what it will mean to them in employment, social decisions, and progressing in life in general in society. A public abuses this information in general. It is dangerous. We are very cautious of our mailing list, which is a database.
Just to show you, I'll carry it on to other types of conditions. NF2, as I said, has a deafness relationship. California, a number of years ago, in trying to help the deaf population, issued special license plates that identified a person as deaf. What happened was that a group of criminals found out that these people were more vulnerable -- they could creep up on them to steal purses, etc. So it actually jeopardized these people by trying to help them, by trying to identify them. So even society, working with good intentions, makes serious mistakes. We talk about national health insurance as an answer, and yet national health insurance will probably mean rationing. With that I have to make a side comment -- my credentials, in terms of the fact that I am a federal employee, were given and it gives credibility to what I am saying as well as just the consumer responsibility. The comments I am making are in no way representative of federal opinion, and I am required to say that. I am here on leave. I want to keep my job, because my wife can't get employment.
What I want to emphasize to you is that I think the battle on privacy is to be lost, because I think the economic public need will be the driving force, and therefore we must shift our focus to how we accommodate the problems that create the need to protect individual privacy.
In thinking about what I was going to say, this morning I was invited to an eight o'clock breakfast in the Caucus Room. I went to the wrong Caucus Room and ended up at a conference of the National Association of Funeral Directors. (laughter) I thought that that reinforced my issue that maybe privacy was a dead issue, but I want to bring it down to your level. Not only do I say that you are at risk of a genetic disease, but the federal government has funded, through the National Institute of Health and the Department of Labor, the Human Genome Project, which means that within the next twenty years, every individual could be categorized on a microchip. THIS IS YOU! What that means to society is overwhelming. We're not talking about a national issue -- the gene pool is an international gene pool, so what the United States does may have impact. The fact is, if you don't get international cooperation on this issue, you've only scratched the surface of the problem. I thank you. (applause)
POWERS: I believe by my watch that we have about six or seven minutes time left. Questions for our various panel members, please?
AUDIENCE MEMBER: I'm from the American Association for the Advancement of Science. I'd like each of the panelists to reflect -- I know you don't have too much time -- you have brought up some serious concerns. To what degree are they unique and ones that we haven't had to deal with before, problematic as they are? Just as an example, it isn't only because we have genetic data that there is information on your wife and other people having NF2. Presumably other people were diagnosed through other measures with other diseases, before doctors' records -- hospital records have existed for a long time. There have been control questions, privacy questions, etc. So these are not brand new things to think about. They've been problems and we have been thinking about them with sometimes better and sometimes poorer resolutions. I think that obviously the FBI collected data before we even understood much of anything about DNA. Again, these are not entirely new things, and I don't think that we should treat them as though they are entirely new, and we have to re-invent the wheel again. I would just like to invite the panelists to comment on that.
NEUFELD: One of the things that is actually very, very different about the storage of genetic matter, as opposed to other types of information, even more importantly is the type of discrimination that can ensue from access to that information. It has been my experience that ordinarily there is discrimination based on if it's the commission of a crime, that's one thing. If the insurance companies want to discriminate, because we choose to abuse ourselves by smoking or alcohol, that's something else. If we don't get a job because we did something wrong in the past, that's also something else. Now we have a situation where there is going to be discrimination, and lots of different contacts for reasons that you as an individual can't control, that you absolutely had nothing to do with, simply because you were born with certain genes. That raises a specter that is fundamentally offensive on a moral level that doesn't exist for other kinds of discrimination.
AUDIENCE MEMBER: I don't think that's new here.
MENDELSOHN: You're right in terms of the fact that the problem has been there, but I think that American society in particular has chosen to generally ignore -- I use this as an example -- how mental illness was handled in this country for so many years. Even the Kennedy family with the daughter that had mental retardation, in the biography that they sponsored, indicated that she had accepted a religious vocation and that's why she was not involved with the family. So there was an attempt to cover it up. No longer are we going to be able to do that because we are not in control of our own information, and we have been until recently. The computer databases in themselves have allowed this information to be readily accessible, and that did not exist before.
POWERS: John, do you want to speak briefly about this?
HICKS: I guess I agree, I think, with the comment that I don't believe that discrimination is entirely new. It's true that it's a new technology and some different kinds of opportunities, but the fact is that we've been confronted with other technologies in times past. I think that through responsible use of the technology, putting it through appropriate social processes to determine whether or not socially we want to apply this technology in particular kinds of ways -- controlled and very specified kinds of ways -- the technology can be structured in such a way that these protections are in place.
MENDELSOHN: Let me address one point that's what so different about this. When you leave your chair to go out of this room, I can go to your chair and take genetic droppings, analyze them and have information on you that is my property, because you left it there.
AUDIENCE MEMBER: You don't have my fingerprints.
MENDELSOHN: Your fingerprints don't tell me what your health status might be, or what your future might be, and I can predict your future with some genetic markers.
POWERS: It's clear that we've only got a small number of you that we can get the questions from, but let me take one from each side here at least.
KEN STONE: My name is Ken Stone I didn't actually catch your name on the medical side of things.
MENDELSOHN: I'm sorry, I didn't hear you.
STONE: I didn't get your name, I'm sorry.
MENDELSOHN: My name is Paul Mendelsohn.
STONE: What I was wondering is, are you against the establishment of such a database at all? What I see the primary conflict being is the insurance companies versus the individual. Perhaps I'm fantasizing about this, but we're heading for the day when we are going to have national health care anyway. Might this solve the problem?
MENDELSOHN: It may actually exacerbate the problem. I'll give you an example with renal disease. In England, which has national health insurance, if you have renal disease and the onset is after 55, you don't get treatment. I use that as an illustration that national health insurance with an economic shortage of resources may lead to additional discrimination, a denial of treatment, because your future is so unpredictable as to make it uneconomical for society to expend resources on you -- and health insurance is only one small aspect of this problem. I do believe, on the other hand, you did ask me if I wanted to establish a database. No, I said there were balances here. We really want this database. It's the only way we can identify, diagnose, and treat the disorder and hopefully come up with some cures.
POWERS: Question on my right?
RICH NEUMEISTER: My name is Rich Neumeister and I'm from Minnesota. I'm a privacy lobbyist as a citizen of Minnesota for years. In 1989, the use of DNA for criminal enforcement came. It was just the public defender from Hennepin County and myself lobbying against some of the issues that you bring out three years later. We were testifying against making some changes and this goes to my next comment. I talked to Art Kaplan, who some of you may know, a well-known ethicist from the University of Minnesota. I see him quite often at various conferences. I was also a participant at a genetic conference, a conference that state legislatures had last year. I raised some issues -- what can we do, people like myself in the trenches at the state level who want to do things, but then we hear sometimes from Mr. Kaplan or some folks, that "Well, it's expanding technology, you can't do anything." Or else people who may represent criminal defense lawyers are out there helping out some people on the issues of privacy or due process or whatever. When will there be a piece of model legislation for something that some national group or whatever does, so that folks like me in Minnesota can take it to legislators and get groups to coalesce, because it is a big issue and it is a concern?
MENDELSOHN: Congressman John Conyers sponsored some legislation last year to address the maintenance and protection of databases, genetic bases, and tissue banks. However, the problem when it was analyzed by the various components, it was found to be unenforceable the way it was written, and very, very difficult to write a law that would be enforceable. It's a paradox right now.
NEUFELD: On a related front, Congressman Edwards also introduced legislation last year to try and put certain controls over state and local law enforcement laboratories doing DNA profiling, to assure limited access to the data, but that legislation also did not go through. I mean, I don't know, I think the answer is to become more involved with the democratic process and lobby your own representatives and others and educate others to try and push for that kind of meaningful legislation.
POWERS: I'm on strict marching orders to bring this to a halt and I apologize that we can't get to as many questions as we want, as we did get started late. I do invite you, if you have additional questions, to speak with some of the panel members individually. Thank you all for attending and thank you to the panelists as well. (applause)
Return to CPSR conferences page.
Return to the CPSR home page.
Send mail to webmaster.