By Troy Duster

There is abundant evidence to support the claim that there is a systematic racial bias within the criminal justice system.  This begins with decisions by police at the point of stop, search, and arrest, extending through the sentencing guidelines and practices and incarceration.1 There is now also a developing forensic science literature that claims to be able to predict "ethnic-affiliation" from population-specific allele frequencies.

In 1997, Mark Shriver published a paper in which he trumpeted the possibility of using DNA markers to predict race.2 From the very beginning, the forensic science interest in "ethnic estimation" was to permit the police to focus in more narrowly on a specific population of possible suspects.

The states are the primary venues for the prosecution of violations of the criminal law, and their autonomy has generated considerable variation in the use of DNA databanks and storage. Even as late as the mid 1980s, most states were only collecting DNA samples from sexual offenders.  The times have changed quite rapidly. All fifty states now contribute to the FBI's Combined DNA Index System (CODIS). Moreover, there has been rapid change in the inter-linking of state databases. In just two years, the database went from a total of nine states cross-linking "a little over 100,000 offender profiles and 5,000 forensic profiles" to 32 states, the FBI, and the US Army now linking "nearly 400,000 offender profiles, and close to 20,000 forensic profiles".3 States are now uploading an average of 3,000 offender profiles every month.  This may sound daunting, but computer technology is increasingly efficient and extraordinarily fast.  It takes only 500 microseconds to search a database of 100,000 profiles.

A DNA dragnet involves law enforcement officers searching for individuals who match the profile of a criminal suspect, then asking those people to submit a sample of their DNA for analysis.  This approach originated in Britain, and DNA dragnets are still most advanced in Europe and the United Kingdom.  The first DNA dragnet was conducted in Leicester, England in 1987.

While the United States has only conducted about a dozen DNA dragnets, most notable about them is their focus on specific racial groups.  San Diego was among the first jurisdictions to conduct the practice when a serial killer stabbed six persons to death in their homes in the early 1990s.  The suspect was African-American, and more than 750 African-Americans were tested.  In 1994, police in Ann Arbor, Michigan, obtained nearly 200 samples from African-Americans in the hunt for another serial rapist and murderer. In both the San Diego and Ann Arbor cases, the suspect was apprehended and convicted for committing another crime, not as a result of the success of the dragnet.  Then in 2004, Charlottesville, Virginia's racially-driven dragnet generated a controversial response from civil liberties groups, ultimately convincing the police to temporarily abandon the dragnet strategy.4

Miami, Florida was the scene of the most notorious and widespread racial DNA dragnet.  Between September 1994 and January 1995, six women were killed and their bodies were left just outside the Miami city limits on a street known as the Tamiami Trail.  More than 2,300 men were stopped by the police as they drove down streets in the area, each asked to provide saliva samples to determine a possible DNA match.5 While the so-called "Tamiami Strangler" was identified through other means, this dragnet is of particular relevance to the issues raised here because (1) almost all of the men who were asked for DNA samples were African-Americans, and (2) their DNA samples were stored.

Several states in the US have embarked upon data collection of arrestees. The most aggressive programs are in Louisiana and California, but the trendline is clear, so some states are debating whether to include all arrestees. This development carries dramatic implications for the prospect of heavily racialized databases.

First, consider that incarceration rates for blacks and Latinos are now more than six times higher than for whites; 60% of America's prison population is either African-American or Latino.  Just over 20% of black males between the ages of 25 and 44 have served a sentence at some point in their lives, and 8% of black men of working age are now behind bars.6 At current rates, a third of all black males and one-sixth of Latino males will go to prison at some point during their lives, while the figure for whites is one in 17. These are national figures, but depending upon the urban area, things can look even more heavily racialized.

We know African-Americans are being arrested at a rate of at least five times greater than whites for minor violations such as marijuana possession, even though the best available evidence suggests that whites are more likely than blacks to use (and thus possess) marijuana at every age level. The DNA of arrestees is being collected more and more routinely (12 states now collect DNA from those merely arrested); we are witnessing a new kind of convergence with portents for even greater racial disparities in convictions and rates of incarceration.

We have data showing that defendants confronted with the "information" that there is DNA evidence against them are far more likely to see this as "definitive evidence" - and thus more likely to accept a less advantageous plea bargain.7 Murphy (2007) and others have noted the successful "creep" of the CSI Effect on the general public.8 But Prainsack and Kitzberger (2008) have discovered that defendants are perhaps even more susceptible to the CSI Effect, i.e., the tendency to believe that DNA evidence is sufficient to secure a conviction.9 Their work with prosecutors and interviews with defendants in the United Kingdom document just how much the technology of the DNA Mystique10 has become a part of the taken-for-granted features of the zeitgeist. Since prosecutors have become increasingly aware of this, they can and do tell those arrested and accused of a crime that "they have the DNA fit" - whether or not they do! This is legally permitted, and there are also documented cases in the United States where this has been a practice.11

Second, more and more cases will be brought before prosecutors using "cold hits" (that match "known offenders" - which will increasingly include those merely arrested).  We know that those "merely arrested" will be heavily distorted by race.

Which brings us to a crucial distinction between science and forensic science.  One of the most essential elements of science is replication of findings by an independent investigator.  If a researcher claims to have discovered some empirically derived finding (think of cold fusion), s/he must make available the method of investigation and open up for scrutiny the procedures so that other scientists can determine whether the finding was spurious, unique, doctored, a fluke, etc.  Not so with empirical evidence on DNA matches in a court of law.  The crime labs are routinely held proprietarily, where the government agency refuses to permit independent laboratory work by "outsiders" who could use the same "scientific methods" to either corroborate or refute a finding of a DNA match.12 This barrier to comparative laboratory analysis is not science - but it is the current state of forensic science.  Much more is at stake than the scientific reputation of some principal investigator; an error in forensic analysis could result in wrongful imprisonment or the death sentence for an innocent person.

Thus we can begin to get a glimpse into the future to see how these various forces (racialized dragnets, expanding offender databases to include arrestees, and the CSI effect - on both prosecutors and defendants) can and will further distort the racial bias in the criminal justice system.  The vast majority of young persons of African-American and Latino descent who are brought into the criminal justice apparatus cannot afford representation by private attorneys, and are thus doubly victimized by a system that dramatically over-selects them at point-of-arrest by ratios from 5 to sometimes even 8 to 1.13 The seemingly inexorable move now gaining momentum to include those merely arrested in the national DNA database will only increase the current disparities, for the full range of reasons chronicled above. What is to be done? The two initial tasks are to a) start to rollback the "function creep" of adding those merely arrested into a database that now totals nearly six million, and b) use this as an opportunity to explain to an unsuspecting public that "cold hits" are not nearly what they seem to be on their favorite crime television programs.

Troy DusterTroy Duster is Silver Professor of Sociology and Director of the Institute for the History of the Production of Knowledge at New York University, and he holds an appointment as Chancellor's Professor at the University of California, Berkeley. From 1996-98, he served as member and then chair of the joint NIH/DOE advisory committee on Ethical, Legal and Social Issues in the Human Genome Project (The ELSI Working Group).  He is past-president of the American Sociological Association (2004-2005), a member of the Board of Advisors of the Social Science Research Council, and in 2003-2004 served as chair of the Board of Directors of the Association of American Colleges and Universities. 


1. Mauer, Marc. Race to Incarcerate, New York: New Press, distributed by W.W. Norton. (2006)
2. Shriver, Mark D., Michael W. Smith, Li Jin, Amy Marcini, Joshua M. Akey, Ranjan Deka, and Robert E. Ferrell. "Ethnic-Affiliation Estimation by Use of Population-Specific DNA Markers, American Journal of Human Genetics, 60:957-964. (1997)
3. Gavel, Doug. "Fight Crime Through Science," Harvard Gazette, November 30. (2000)
4. Glod, Maria. "Police in Charlottesville Suspend 'DNA Dragnet.'" Washington Post. April 15, B01. (2004)
5. Pan, Philip P. "Prince George's Chief has used Serial Testing Before" Washington Post, January 31,  B1. (1998)
6. Austin, J., T. Clear, T. Duster, D.F. Greenberg, J. Irwin, C. McCoy, A. Mobley, B. Owen, J. Page . Unlocking America: Where and How to Reduce America's Prison Population, Washington, DC: The JFA Institute. (2007)
7. Prainsack, Barbara and Martin Kitzberger. "DNA behind bars - 'other' ways of knowing forensic DNA technologies,"  Social Studies of Science (in press) (2008)
8. Murphy, Erin. "The New Forensics: Criminal Justice, False Certainty, and the Second Generation of Scientific Evidence," California Law Review, 95, June, 1-81. (2007)
9. Prainsack and Kitzberger (2008).
10. Nelkin, Dorothy and M. Susan Lindee. The DNA Mystique: The Gene as a Cultural Icon, Ann Arbor: University of Michigan Press. (2004)
11. LaDuca, Rocco. "Fake Report puts Focus on Police Tactics," Observer-Dispatch,  (November 3, 2007) Accessed 26 July 2008.
12. Murphy (2007)
13. Austin, et al (2007)

Search: GeneWatch
The Gene Myths series features incisive, succinct articles by leading scientists disputing the exaggerations and misrepresentations of the power of genes.
View Project
The purpose of the Genetic Bill of Rights is to introduce a global dialogue on the fundamental values that have been put at risk by new applications of genetics.
View Project