Jonathan Kimmelman holds a PhD in molecular biophysics and biochemistry. He lives in Montreal, where he is writing a book on the political dimensions of biotechnology.
Nothing is more galling to scientists than outsiders questioning their
DNA testing can convict the guilty; it can also destroy the privacy of millions.
In New Mexico, communists who fail to register their party affiliation with the state commit a felony. Under New Mexico's DNA databanking law, if they are caught they are required to submit a DNA sample to the department of public safety. In Idaho, consensual sodomy with a partner other than your spouse constitutes a sex-crime felony. Those unfortunate enough to be caught in the act are similarly required by law to submit a tissue sample to the state's DNA databank for the purposes of preventing future sex crimes. And if Governor George Pataki is successful in the next legislative session, New York will begin collecting genetic material from any person convicted of a misdemeanor, such as resisting arrest or disorderly conduct as a result of peaceful civil disobedience.
In an age of biotechnology and computers, we are all but a needle-stick away from disclosing hereditary-disease susceptibilities, familial relationships and identifying information. Anyone who values privacy should therefore be concerned that US law-enforcement agencies are amassing ever larger portions of the general population's DNA while neglecting to implement measures that would protect the privacy and presumptive innocence of citizens. And because DNA evidence is currently enjoying an unprecedented degree of bipartisan enthusiasm, these gradual developments have tended to be sheltered from the criticism that might otherwise confront such policies.
Not that DNA evidence's celebrity isn't well deserved. It is many rape victims' best hope for identifying their assailants and law enforcement's most penetrating method of apprehending serial offenders. It can be credited with triggering a re-examination of the nation's capital punishment system by exonerating eight death-row inmates. Like its predecessor, the fingerprint, DNA profiles are a reliable means of identifying individuals (except in the case of identical twins). But glib analogies to fingerprints obscure important differences. DNA samples can reveal far more information than fingerprints, including sensitive medical conditions, traits or a person's biological parentage. In addition, while fingerprints are unique to every individual, genetic profiles are partially shared among blood relatives. Thus, databanks contain identifying information on nonoffending relatives of people explicitly covered by databanking statutes. Finally, because we shed our genetic calling cards in a trail of hair follicles, skin flecks, saliva aerosols and finger smudges, DNA can also provide a trace of our activities.
DNA databanks are premised on statistics indicating that individuals convicted of a serious violent offense often commit other violent offenses that leave behind incriminating DNA. Tissue samples, usually in the form of a blood sample or cheek swab, are thus collected from offenders covered by their state's databank laws and are analyzed using a technique called "profiling," which detects genetic variations among individuals that, at least as currently understood by geneticists, have no biological function. The resulting data are then computerized so that profiles produced from crime-scene samples can be compared with those already in the database, allowing authorities to eliminate certain suspects or target those whose profiles match. In effect, databanks provide a means of genetically frisking anyone who has ever committed a covered offense for any crime in which DNA has been recovered.
As of June 1998 all fifty states had enacted statutes authorizing state and local law-enforcement agencies to operate criminal DNA databases and to pool their DNA profiles into a national FBI-operated database called CODIS (Combined DNA Identification System). Though the earliest laws targeted convicted violent sexual felons, civil libertarians looked to the history of Social Security numbers, fingerprinting and drug-testing to warn of an inevitable migration of the technique from convict to suspect terrain. A decade later, as many states have passed laws to cover new offender categories, the Cassandras appear to have been vindicated. Delaware, for instance, requires submission of genetic samples for all those who have committed offenses against children, which include selling tobacco or tattooing minors without the consent of a guardian. Twenty-three states cover certain categories of misdemeanors, and seven states have enacted legislation that would require DNA submission for any felony, which extends DNA databanking into realms such as perjury, larceny, bribery and fraud. Thus, in addition to New Mexico's statute covering unregistered communists, Alabama's code covers tax evaders and Virginia's targets people who deface brands or marks on timber. Experts like CODIS program director Steve Niezgoda have predicted that all states will eventually amend their statutes to cover all felonies; four states have already done so, and another three have recently considered or will consider such an expansion in their next legislative sessions. Among these three, New York's proposal stands out as by far the nation's most comprehensive, targeting all convicted felons and class-A misdemeanants.
DNA databanking laws are furthermore part of the ferment that is corroding the century-old juvenile justice system that treats minors as a category of offenders separate from adults. More than half of all states authorize inclusion of DNA profiles collected from juveniles in their databanks. In contrast to the convention of sealing or erasing juvenile criminal records after a period of time--a practice grounded on a rehabilitative ideal--none of the statutes require states to remove juvenile DNA profiles from their databanks, and one (Arizona's) expressly prohibits their removal. Several states have revised their original legislation to cover juvenile offenders as well. The spread of DNA databanking to minors is especially troubling when considered against the racial inequities that plague the juvenile justice system. According to Vincent Schiraldi, president of the Center on Juvenile and Criminal Justice, "When you control for poverty, white and black [teens] commit the same amount of violent crime, [but] blacks are arrested at four times the rate of whites and imprisoned at seven times the rate of whites. So don't think for a second this databank will be race-neutral. This policy will grossly overrepresent criminal behavior by blacks and exacerbate disparities in incarceration because [databanks are] going to be used against people."
An indirect consequence of expanding DNA databanks is their partial coverage of a larger proportion of nonoffending relatives as well. Because individuals share portions of their DNA with biological relatives--half in the case of siblings, parents and children--an incomplete match between a databanked person's profile and that of a crime-scene sample might lead investigators to question an individual's immediate family. The effect of such profiling by proxy is that identifying information about nonoffenders is present in criminal databank systems as well; in effect, if you have a relative whose profile has been databanked, you're likely to be partially genetically frisked as well.
A critical unresolved question about current databanking practices concerns what law-enforcement agencies actually do with their frozen vials of human tissue. The human genome contains approximately 100,000 different genes, many of which are associated with specific illnesses. Though DNA profiles have few applications beyond linking individuals to biological specimens, the actual tissue samples submitted by offenders could in principle be analyzed for genetic traits ranging from sickle-cell anemia to schizophrenia. Since evolving typing techniques may one day outmode profiles currently being entered into computers, more than half of US states are authorized or required by law to archive their samples so they can be retested. This sustains the possibility that samples may eventually be used for purposes other than profiling.
Most statutes restrict sample use to "law enforcement"--a term whose broadness in this context can only be described as oceanic. Twenty states allow law-enforcement agencies to use samples for research on improving forensic techniques, which could mean searching banked DNA samples for genetic predictors of recidivism, pedophilia or aggression. One Massachusetts legislator publicly advocated such a use, and Tom Callaghan, program manager of the FBI's Federal Convicted Offender DNA Database, refused to rule out such a possibility when pressed at a National Institute of Justice Symposium in September 1999. Moreover, tissue repositories created by databanks would provide genetics researchers with congenial waters in which to trawl for genes thought to be involved in criminal behavior. Alabama's databanking law brushes perilously close to this by authorizing release of anonymous DNA population data collected by law-enforcement authorities to "assist in other humanitarian endeavors including, but not limited to, educational research or medical research or development."
Experimenting with offender DNA in this way would violate basic tenets of biomedical ethics by using tissues that were not obtained by consent for purposes that arguably run counter to the interests of the research subject. "If [law-enforcement authorities] want to do research," argues Boston University bioethicist George Annas, "they should follow the same rules everyone else has to follow in terms of informed consent and privacy.... Criminals have privacy rights like everyone else." As such, using databanked samples for research without consent also runs counter to recommendations by the American College of Medical Genetics.
Such research authorizations are especially troubling in light of this nation's checkered history of experimentation on prisoners. In 1875 social reformer and prison inspector Richard Dugdale wrote his famous study of the Jukes family after he noticed a disproportionate number of inmates with that last name. The availability of banked criminals' tissues may prove a valuable resource should society's interest in genetic explanations for social ills be renewed.
Legal challenges of DNA database laws have generally failed and are therefore unlikely to stem their widening sweep. Practices in Britain, the first country to enlist DNA in its crime-fighting cavalry, may portend dramatically widened use of databanking in the United States. Britain's Forensic Science Service is authorized to collect DNA samples from anyone questioned about or suspected of any offense for which a person could be detained. As of July 1999, England had collected 547,000 DNA samples; the effort was projected to reach 30 percent of British men eventually. In addition, England has conducted at least eighty "intelligence-based screens"--the official term for what is colloquially called a "genetic sweep"--in which the general population is asked to submit DNA samples to help police investigate a particular crime. Although samples are provided voluntarily, social pressures, heavy media coverage and the concern that failure to submit a sample may itself invoke police suspicion undermine the notion of submissions being truly consensual. Other countries, including Canada and Germany, have conducted similar sweeps, and while some argue that the Fourth Amendment would probably bar such practices in the United States, privacy watchdogs like New York Civil Liberties Union's executive director Norman Siegel caution that "Fourth Amendment challenges [of databanks] have not been successful; these are the only reference points we have [for predicting how courts will rule on genetic sweeps], and they're not promising."
The next battle between civil libertarians and law-enforcement authorities concerning DNA databanking is likely to concern the leap from profiling convicted felons to arrestees. Former NYPD chief Howard Safir has championed arrestee profiling, and US Attorney General Janet Reno has begun to explore the implications of such a policy by querying a National Institute of Justice commission. Arrestee profiling would dramatically broaden the reach of DNA databanking and, if not subject to careful restrictions, would empower law-enforcement authorities to arrest people for minor offenses, collect a tissue sample and search their databases for a match between the arrestee's profile and another crime-scene sample. Despite widespread enthusiasm in law-enforcement circles, profiling on such a scale isn't likely to be implemented anytime soon, given the backlog of tissue samples awaiting profile analysis and the high costs (at least $100 per sample). Nevertheless, one state (Louisiana) profiles arrestees for sexual offenses, and advancing automation technologies are likely to erode these fiscal barriers.
Though this is reason for despair among privacy advocates, there are a few hopeful signs among the various statutes. Twenty-seven states (and the federal government), for example, prohibit disclosure of genetic materials or information to unauthorized third parties. Wisconsin requires that law-enforcement authorities eliminate DNA samples of convicted persons after profiling is complete, and six states (Indiana, New Hampshire, Rhode Island, Texas, Vermont and Wyoming) restrict what authorities can do with collected DNA by prohibiting analysis of genetic mutations that could predict a person's traits. But in an environment where the political leaders most likely to raise objections to such policies are often silenced by a fear of appearing to be soft on crime, the stability of these protections remains to be seen.
Imagining a fair and protective system for using DNA evidence in the criminal justice system isn't all that difficult. People claiming innocence should be given opportunities to volunteer DNA to clear their name. For them--and more broadly for the credibility of the criminal justice system--DNA forensic technology may be the only life vest within reach. Upon overturning a conviction, volunteered DNA samples and profiles should be promptly destroyed, preserving the individual's presumptive innocence. For people convicted of serious violent offenses and beyond the reach of such exculpatory evidence, however, the trade-off between privacy and public interest may tilt toward favoring a DNA databanking system with strong privacy protections, including sample destruction after profiling and prohibitions on uses other than comparing profiles with those collected from crime scenes. And finally, to protect the presumptive innocence of convicted offenders' family members, states should impose stringent requirements for when a match between a crime-scene sample and a databanked profile can trigger an investigation.
Privacy is a zero-sum entity: The extension of law-enforcement authorities' genetic gaze comes directly at the expense of an individual's power to withhold such information. Where most human DNA handling once occurred in medical clinics and research laboratories--institutions that are generally subject to public oversight and cautious (if imperfect) ethical review--DNA has now entered a territory not particularly distinguished for its ethical circumspection. States are not providing many reasons for the public to be confident that they are taking these concerns seriously; perhaps of even greater concern, negligence in protecting the privacy of offenders and criminal suspects may acclimate a public to weak protections of genetic materials. As the predictive powers of genetic technologies are refined, this could have grievous consequences for everyone.