01 February 2008

Big DNA Evidence Break

The Story

Police claim to have solved a suburban Denver murder-rape from 1976, and nine other rapes in a nine month spree in 2004 and 2005 in Aurora and Denver based upon DNA evidence.

Arrests were made Thursday, although the suspects were not identified in the story.


One suspects that even minimal coroborration that the suspect was present in the vicinity of the places of the crimes and that the perpetrator was not an intimate acquaintance of the victims of the rapes, combined with the DNA evidence, will be sufficient to secure ten rape convictions. Proving the 1976 murder will be more difficult, as intent and proof that the victim was raped and murdered in the same incident by the same person may be difficult to show with 32 year old evidence, although one suspects that it will be charged as a first degree felony murder case. Obviously, with or without a murder conviction, this individual (who must be, at least, in his 40s) will serve the rest of his life in prison - his only real defensive option at this point is to try to avoid the death penalty on the murder charge.

Who should have DNA on file?

The story implies, but does not quite state, that the DNA sample was collected from the suspect subsequent to a felony conviction for some other crime. It will be interesting to see what kind of criminal record the suspects have and how their DNA was collected.

One virtues of a relatively narrow DNA database based upon felony convictions in place now and upon which the arrests reported were probably based is that the risk of a false positives is dramatically reduced.

Consider that even a 99.999% accurate test will generate 46 false positives if compared to the entire population of Colorado, and that about 23 of those false positives, on average, will live in metropolitan Denver and have a number of similarities with the Defendant that arise from their genetic similarity. A false positive based upon DNA evidence is a particularly troubling thing, because DNA evidence is as close as one can come to conclusive physical evidence in a criminal case, nearly assuring a conviction or guilty plea most of the time, because even an innocent defendant knows that a conviction is very likely.

While DNA analysis is increasingly growing more accurate, it will never be perfect, for the very simple reason that crime scene DNA samples are themselves often fragmentary, and that contamination of the crime scene with DNA unrelated to the crime is more and more of a concern as DNA collection methods improve.

But, this risk is far lower with a felony based database because a database of people with felony convictions excludes about 95% of the population, and because recidivism statistics make overhwelmingly clear that people with prior felony convictions are dramatically more likely to commit future felonies, than people without prior felony convictions. The odds of a felon released from prison commiting another crime in the next few years (i.e. the recidivism rate) overall is on the order of 50% plus. The odds of someone who has never been convicted of a felony committing a felony in that same time period are something under 1.5%, more than thirty times smaller. Thus, someone with no prior criminal record is much more likely to be a false positive than someone with a prior criminal record.

To take a more cynical view, call it the kharmic payback view, the odds that a wrongfully convicted person with a felony record escaped conviction for some other equally serious felony by not getting caught are quite high. The overwhelming majority of serious crimes are never solved and the majority of serious crimes are committed by repeat felons. So, even if the particular conviction was unjust, the odds that someone with a prior criminal record will experience "cosmic injustice" as a result of a wrongful conviction are lower than the odds of a wrongful conviction itself. In contrast, the odds that a wrongfully convicted person with no felony record escaped conviction for some other equally serious felony by not getting caught are very low, so the risk of convicting a totally innocent person in that case as a result of a false positive DNA test is quite great.

The law rightly doesn't consider issues like propensity to commit future crimes and the even more nebulous concept of "cosmic injustice" in individual cases, which must be decided based upon the evidence that a particular individual acting in a particular way at a particular time. Doing so would undermine the accuracy of the process and compound errors when they were made by the system. But, that doesn't mean that those considerations are irrelevant to a utilitarian analysis of criminal justice policies at a macro level.

Of course, an alternative to a narrow felony conviction based DNA database would be to decide societally that DNA is so valuable in acquitting the innocent and convicting the guilty, and crime is so great a threat to society, that the risk of false positives is worth the benefits, with the system trusting that non-DNA means will acquit most false positives. The cost of not having someone who commits a crime in a DNA database is real. If the DNA database had been bigger in Colorado in 2004, there is a good chance that most of the rapes in the nine month, nine rape spree by these suspects could have been prevented.

Whether the later rapes could have been prevented by a different DNA collection policy with a much wider net, or was simply a product of the fact that DNA technology has only been affordable to implement widely in recent years (i.e., if the suspects had prior felony records), is relevant to what a case like this means in the public policy calculus.

At its most basic, are there many rapists out there, in cases where the identity of the alleged offender can be reasonably disputed (i.e. in non-acquaintance cases where DNA evidence is most conclusive since consent is rarely a plausible defense), who do not have prior felony records at the time they commit their crimes? If not, a wider DNA database probably doesn't make sense. If most of these rapes are committed by people who have no prior felony record, a wider DNA database might make sense.

Crime And Criminals

Cases like this one, and another recent string of burglary cases solved in Denver based upon DNA evidence, also illustrate vividly a well established fact in criminology: A majority of all serious crimes are committed by something on the order of 3% of the general population, overwhelmingly men.

The statistics seem to indicate that almost everyone in this class of criminals is incarcerated at some point,at least in the United States.

More than 5.6 million Americans are in prison or have served time there, according to a new report by the Justice Department released Sunday. That's 1 in 37 adults living in the United States, the highest incarceration level in the world. . . . If current trends continue, it means that a black male in the United States would have about a 1 in 3 chance of going to prison during his lifetime. For a Hispanic male, it's 1 in 6; for a white male, 1 in 17.

The notion that some people are "criminals" as opposed to merely people who happened to commit a particular crime, and that many kinds of serious crime are to a great extent committed by people who are "criminals" is empirically true.

The strong links between crime and the economic circumstances, race and social class of the offender, however, also suggest that the lion's share of "criminals" would not offend if they were in better economic circumstances. Few "criminals" are beyond redemption. Indeed, rehabilitation may be easiest for felons who have grown up in poverty and are commiting crimes for reasons relating to those circumstances. Those felons, it should follow, have a good chance of going straight, if they can escape the substance abuse problems they so often suffer from, treat mental illnesses that they often have, and develop the skills and habits necessary to hold down a regular job and manage their own finances responsibly.

The percentage of population who commit serious crimes primarily because of who they are, because of their overwhelming bad character, unaggravated by economic circumstances, is very likely well under the one in seventeen men who are incarcerated in a lifetime among whites (since white criminal defendants too grossly disproportionately come from poor socio-economic circumstances).

It is a fair guess that poverty and discrimination in the criminal justice process drives the criminal experiences of something in excess of 27 percentage points of the general population of African-American men (the excess of the lifetime incarceration rate for black men less the lifetime incarceration rate for white men) and something in excess of 10 percentage points of the general population of Hispanic men (by the same reasoning, and notably, with native born Hispanics considerably more likely than foreign born Hispanics to commit crimes, other factors being equal). Thus, one would suspect on this basis alone that more than five out of six black men incarcerated and on the order of two out of three Hispanic men incarcerated commit crimes rooted in poverty and disparate treatment by the criminal justice system.

If one assumes that at least half of crimes committed by whites are rooted in poverty (a number which is order of magnitude reasonable given statistics on poverty, high school dropout rates, unemployment and the like that we are aware of concerning white felons), the numbers are more like eleven out of twelve cases of black men with a history of incarceration, and seven out of eight cases of Hispanic men with a history of incarceration, and by definition half of the cases of white men with a history of incarceration. Thus, the facts we know suggest that at least 70% of people who are ever incarcerated have a real likelihood of rehabilitation if their health, substance abuse and economic circumstances are addressed.

While deeply complicated by the impact of the screening process for new military recruits (which weeds out most felons, high school dropouts and people who simply have fragile mental health), the statistics on court martial rates which are much lower than felony conviction rates for members of the general civilian population, are also suggestive of the power of improved secure economic circumstances to reduce crime rates.

[T]he prosecuted crime rate in the Army is less than 5% of the crime rate in the general population, for individuals of comparable age and gender.

It is also worth noting that the U.S. Army has a far larger percentage of African-American men than the comparable military age men demographic in Colorado does, because military aged African American men are charged with felonies and misdemeanors at a far higher rate than the general population in civilian life.

Now, to be clear, poverty and race cannot excuse crime, even though they help to explain it. A large majority of black men who grow up poverty never end up in prison. And, victims of crime are no less victimized because the offender who did it faced pressures in his or her own life. Poverty is not destiny, and neither, in most cases, are genes for that matter.

A modern geneticist using the narrative of genetic-environment interaction in vogue now would say something like this:

Different people are born with, and have based upon their childhood environments which are beyond their control, different degrees of ability to resist a temptation to commit a crime or get into a situation that could lead to criminal activity, call it strength of character if you will. Poverty influences the intensity of that temptation. Discrimination influences the likelihood that one will be in poverty, and also influences the likelihood that law enforcement officials and the courts will choose to response to something that an individual is suspected of having done with a sentence of incarceration.

Thus, a middle class or rich person is far less likely to encounter an intense temptation to commit a crime that will test his or her character. Many middle class and rich people with only a weak ability to resist a temptation to commit a crime will never be tested. And these same people are far less likely to receive harsh treatment if suspected of having committed a crime. Some will be given the benefit of the doubt by law enforcement officers investigating a crime, by district attorneys deciding how to prosecution a case identified by law enforcement, by juries assessing whether there is reasonable doubt that a crime was committed, by judges assessing an appropriate sentence, and by parole boards determining a release date.

In contrast, a poor young man in a ghetto is almost certain to be have his character tested intensely, unless society steps in to intervene vigorously on behalf of the well being of at risk young men, and many people in that situation will be found wanting in resolve without intense societal assistance.

No comments: