Perhaps it would be a good idea to begin with definitions in attempt to distinguish racial profiling from criminal profiling (and/or offender profiling, if one considers these different concepts). In a forensic setting, racial profiling refers to law enforcement officers targeting persons based on their race or ethnicity due to some racial or ethnic stereotype or otherwise a nefarious intent. I'd say it isn't necessary for the LEO to be overtly conscious of this nefarious intent or targeting.
Criminal profiling, on the other hand, is a legitimate investigative tool where the investigator targets persons based on inferences from characteristics of the crime, including known characteristics of the perpetrator (such as his approximate height or that he drove a certain type and color car). Such inferences might legitimately involve race in some way. For example, consider the scenario where there is a Black murder victim in a predominately Black neighborhood, and the murderer is entirely unknown. Statistically there is a great probability that the murderer is also Black. An analogous statistical inference could be made for a white murder victim in a predominately white neighborhood. This does not mean that any person can or should be either included or excluded as a suspect on the basis of his/her race. The greater probability of the race of the perpetrator is merely a bare fact that is inferred from certain characteristics of the crime, namely the race of the victim and the setting. Indeed, the probability regarding the race of the murderer stems partly from the fact that most (~80%) murder victims know their attackers.
DNA phenotyping such as performed by Parabon-NanoLabs, a company that has been used by a number of police departments, begins with a genetic profile of a person, determines the person's sex and assesses “ancestry” according to historical geographical genetic relationships, and calculates probabilities that the person can be included in or excluded from certain categories spanning a range of skin color, eye color, hair color, facial morphology, and the degree of freckling.
It should be noted that even though there can be (but is not always) high confidence about whether a person's genotype is included in or excluded from particular categories of skin color, eye color or hair color in/from which [X]-percent of other people with those genes were included or excluded, the phenotype classifications themselves are not necessarily easily distinguished. Classifications of skin color labeled very fair, fair, light olive, dark olive, light brown, brown, dark brown all fade into the neighboring category, and whether any particular person belongs in one category rather than another is ultimately a subjective judgment. Moreover, apparently Parabon's algorithms most often place a person into 2 rather than a single category (e.g., “blond/brown” hair color; “light olive/fair” skin color). Additionally, hair color, skin color and eye color can easily be changed, unknowingly to the casual observer short of invasive examination; facial morphology can be altered by plastic surgery, and would seem to be outwardly dependent on a person's weight to some degree. The historical geographical categories for classifying “ancestry” are often useless for visually identifying highly mixed-race people, which are becoming a larger percentage of every population. There are few assumptions one can make about a person's historical-geographical make-up just by looking at the person.
To the best of my knowledge, no studies have been published in the peer-reviewed literature demonstrating the reliability of the techniques used by Parabon-Nanolabs.
The ways in which the findings of such DNA phenotyping dovetails into racial profiling seem to be limited only by one's imagination. Consider the scenario where a genetic profile is obtained from semen associated with the body of a murder victim who was sexually active and who knew and regularly came into contact with many people; the DNA does not match any known acquaintance of the victim who is willing to give a DNA sample, and the profile is of someone whose ancestry is an admixture of 70% Southern European and 30% Latino. So, lacking other leads, police tend to ignore the possible suspects whom they judge to be of Northern European, African or Asian ancestry, and concentrate on “brown” men, vaguely losing sight of the fact that this person of interest is not necessarily very “brown” (e.g., a portion of people classified with a Southern European ancestry will have blue or green eyes--and Parabon's probabilities and double categories of eye color, skin color and hair color, in addition to the fact that people can easily alter their eye color, skin color and hair color, may only mislead rather than clarify the appearance of such a mixed-ethnicity person), not to mention the fact that the semen does not necessarily belong to the killer. But a “brown” person of interest without an alibi and with even a speck of seemingly incriminating circumstantial evidence could easily and wrongly become a suspect and subject to an order to unwillingly provide his DNA, whereas he wouldn't have were it not for the DNA phenotyping.
(Just to be clear: the point I wish to highlight here is not the fact that this person can be ruled out as the contributor of the semen if his DNA does not match, or that he should be legitimately considered a suspect and further investigator if he is the contributor of the semen. The issue I am raising here is the wrongful targeting.)
Again, there are just endless such scenarios. It doesn't matter if one assesses the above as an example of racial profiling or criminal profiling--the end product is the same either way. It doesn't matter to what degree one assumes that the investigator was conscious of engaging in wrongly targeting on the basis of racial/ethnic features. One can say that the findings of the DNA phenotyping “channeled” the investigator's attention on the basis of racial/ethnic characteristics, or gave the investigator “tunnel vision”. There seems to be no more common error that occurs among police investigators than “tunnel vision” where they see incriminating circumstantial evidence that isn't really there or isn't really incriminating, and explain away exculpatory evidence.
Indeed, the question I am asking here, and what I am calling a “subtle” form of racial profiling is where the investigator does not perceive him/herself to be profiling on racial or ethnic grounds. I.e., where, as it pertains to race/ethnicity, the investigator perceives him/herself to be merely engaging in criminal profiling rather than racial profiling, but is in fact, for whatever reason, wrongly targeting a person or persons on the basis of his/her/their (perceived) race/ethnicity.
So, might it be that DNA phenotyping encourages subtle forms of racial profiling?
By the way, it should be noted that DNA phenotyping has led to one successful prosecution, where witness reports and criminal or psychological profiling had initially pegged a serial killer in Louisiana as white. But matching DNA samples obtained from several victims were eventually phenotyped and found to indicate a biogeographical ancestry of a man who was 85% African and 15% Caucasian, thereby refocusing the investigation toward Black men. The DNA of Derrick Todd Lee, who had previously been arrested for stalking and voyeuristic crimes, matched that found on 5 of the victims. He was eventually convicted of 2 of 7 murders he was believed to have committed.
Criminal profiling, on the other hand, is a legitimate investigative tool where the investigator targets persons based on inferences from characteristics of the crime, including known characteristics of the perpetrator (such as his approximate height or that he drove a certain type and color car). Such inferences might legitimately involve race in some way. For example, consider the scenario where there is a Black murder victim in a predominately Black neighborhood, and the murderer is entirely unknown. Statistically there is a great probability that the murderer is also Black. An analogous statistical inference could be made for a white murder victim in a predominately white neighborhood. This does not mean that any person can or should be either included or excluded as a suspect on the basis of his/her race. The greater probability of the race of the perpetrator is merely a bare fact that is inferred from certain characteristics of the crime, namely the race of the victim and the setting. Indeed, the probability regarding the race of the murderer stems partly from the fact that most (~80%) murder victims know their attackers.
DNA phenotyping such as performed by Parabon-NanoLabs, a company that has been used by a number of police departments, begins with a genetic profile of a person, determines the person's sex and assesses “ancestry” according to historical geographical genetic relationships, and calculates probabilities that the person can be included in or excluded from certain categories spanning a range of skin color, eye color, hair color, facial morphology, and the degree of freckling.
It should be noted that even though there can be (but is not always) high confidence about whether a person's genotype is included in or excluded from particular categories of skin color, eye color or hair color in/from which [X]-percent of other people with those genes were included or excluded, the phenotype classifications themselves are not necessarily easily distinguished. Classifications of skin color labeled very fair, fair, light olive, dark olive, light brown, brown, dark brown all fade into the neighboring category, and whether any particular person belongs in one category rather than another is ultimately a subjective judgment. Moreover, apparently Parabon's algorithms most often place a person into 2 rather than a single category (e.g., “blond/brown” hair color; “light olive/fair” skin color). Additionally, hair color, skin color and eye color can easily be changed, unknowingly to the casual observer short of invasive examination; facial morphology can be altered by plastic surgery, and would seem to be outwardly dependent on a person's weight to some degree. The historical geographical categories for classifying “ancestry” are often useless for visually identifying highly mixed-race people, which are becoming a larger percentage of every population. There are few assumptions one can make about a person's historical-geographical make-up just by looking at the person.
To the best of my knowledge, no studies have been published in the peer-reviewed literature demonstrating the reliability of the techniques used by Parabon-Nanolabs.
The ways in which the findings of such DNA phenotyping dovetails into racial profiling seem to be limited only by one's imagination. Consider the scenario where a genetic profile is obtained from semen associated with the body of a murder victim who was sexually active and who knew and regularly came into contact with many people; the DNA does not match any known acquaintance of the victim who is willing to give a DNA sample, and the profile is of someone whose ancestry is an admixture of 70% Southern European and 30% Latino. So, lacking other leads, police tend to ignore the possible suspects whom they judge to be of Northern European, African or Asian ancestry, and concentrate on “brown” men, vaguely losing sight of the fact that this person of interest is not necessarily very “brown” (e.g., a portion of people classified with a Southern European ancestry will have blue or green eyes--and Parabon's probabilities and double categories of eye color, skin color and hair color, in addition to the fact that people can easily alter their eye color, skin color and hair color, may only mislead rather than clarify the appearance of such a mixed-ethnicity person), not to mention the fact that the semen does not necessarily belong to the killer. But a “brown” person of interest without an alibi and with even a speck of seemingly incriminating circumstantial evidence could easily and wrongly become a suspect and subject to an order to unwillingly provide his DNA, whereas he wouldn't have were it not for the DNA phenotyping.
(Just to be clear: the point I wish to highlight here is not the fact that this person can be ruled out as the contributor of the semen if his DNA does not match, or that he should be legitimately considered a suspect and further investigator if he is the contributor of the semen. The issue I am raising here is the wrongful targeting.)
Again, there are just endless such scenarios. It doesn't matter if one assesses the above as an example of racial profiling or criminal profiling--the end product is the same either way. It doesn't matter to what degree one assumes that the investigator was conscious of engaging in wrongly targeting on the basis of racial/ethnic features. One can say that the findings of the DNA phenotyping “channeled” the investigator's attention on the basis of racial/ethnic characteristics, or gave the investigator “tunnel vision”. There seems to be no more common error that occurs among police investigators than “tunnel vision” where they see incriminating circumstantial evidence that isn't really there or isn't really incriminating, and explain away exculpatory evidence.
Indeed, the question I am asking here, and what I am calling a “subtle” form of racial profiling is where the investigator does not perceive him/herself to be profiling on racial or ethnic grounds. I.e., where, as it pertains to race/ethnicity, the investigator perceives him/herself to be merely engaging in criminal profiling rather than racial profiling, but is in fact, for whatever reason, wrongly targeting a person or persons on the basis of his/her/their (perceived) race/ethnicity.
So, might it be that DNA phenotyping encourages subtle forms of racial profiling?
By the way, it should be noted that DNA phenotyping has led to one successful prosecution, where witness reports and criminal or psychological profiling had initially pegged a serial killer in Louisiana as white. But matching DNA samples obtained from several victims were eventually phenotyped and found to indicate a biogeographical ancestry of a man who was 85% African and 15% Caucasian, thereby refocusing the investigation toward Black men. The DNA of Derrick Todd Lee, who had previously been arrested for stalking and voyeuristic crimes, matched that found on 5 of the victims. He was eventually convicted of 2 of 7 murders he was believed to have committed.
