The National Institute of Justice, the agency tasked with research for the Justice Department, realized the scope of the problem with untested kits and funded four cities
-- Los Angeles in 2006, and Detroit, New Orleans and Houston in 2011 -- to process part of their backlog to help answer two main questions: How much of this backlog should be tested, and if only some of the backlog should be tested, which kits should get prioritized?
As it turns out, these studies resulted in more questions. In reports released in 2012
, researchers analyzing the data gathered from two of these cities, Los Angeles and Detroit respectively, came to divergent conclusions: the Detroit researchers recommended testing all kits in the backlog, while the Los Angeles researchers recommended only testing kits associated with a stranger assault (one in which the victim does not know the offender) and not testing kits from non-stranger assaults (also known as acquaintance rapes).
The Los Angeles researchers called for an analysis of the costs and benefits associated with testing the backlog, and -- despite the eloquent moral arguments that all kits should be tested -- emphatically state
that these issues should not be dictated by community pressure from victim groups.
Read a certain way, research might on first glance support such a conclusion. DNA can be recovered from only half the kits on average, and about 60% of the recovered DNA
generate a hit in CODIS. When New York City tested its entire backlog, only 3.7% of CODIS hits led to convictions. This is because, according to
data from the National Center for Victims of Crime, around 60% of hits led to cases that were beyond the statutes of limitations, 20% of hits led to offenders who were already arrested, and 10% of hits were associated with victims who were either missing, unwilling to move forward with the case, or deemed unreliable.
The CODIS hit rates for stranger vs. non-stranger kits were qualitatively similar for all four
NIJ-funded cities, but the studies in each city followed different methodologies. The Los Angeles study, unlike those in Detroit and the other cities, looked downstream from the hits, and found that the hits in non-stranger cases ultimately led to very few arrests, charges and convictions.
But the Los Angeles study's conclusions are also misleading, when read in context with an in-depth collaborative study
(by a different team of researchers, published in 2012) of Los Angeles' handling of sexual assaults during this same period. That study revealed some troubling observations (that have been documented
in many other cities in the United States
as well) and that illustrate why testing non-stranger kits might lead to fewer convictions in the end. The researchers interviewed sexual assault detectives, some of whom never arrested in non-stranger cases, and victims found that many detectives were skeptical of their claims in those cases.
Meanwhile, deputy DAs based their charging decisions on the probability of conviction, which in turn reflected jurors' preconceived notions of what constitutes rape, and prosecutors varied in the extent to which they emphasized DNA as relevant in non-stranger assaults. For the most part, prosecutors in Los Angeles did not file charges in "he said/she said" non-stranger assaults, so even in cases where a kit was tested, it would seem most unlikely that a non-stranger rape kit would lead to a conviction.
At least in hindsight, the Los Angeles study seems to suffer from circular reasoning. Historical biases helped lead to fewer charges, arrests and convictions from non-stranger hits -- and their recommendation, if adopted, would perpetuate these historical biases.
Stanford graduate student Can Wang and I used the Detroit data to develop a mathematical model that allowed us to perform a conservative (where possible, overestimating the costs and underestimating the benefits) cost-benefit analysis of testing a kit.
that emerges from our study
is that testing a sexual assault kit is similar to playing a high-stakes lottery with very favorable odds. It costs $1,600 to test a kit, but if you get a conviction from testing the kit, the lottery's jackpot is $11.4 million: In the absence of kit testing, an active offender commits another 26 assaults (on average) before either being convicted or voluntarily stopping, and the cost of each sexual assault is approximately $435,000, which includes both victim costs (such as treatment, court costs, and lost work productivity) and a willingness to pay (the maximum amount an individual would be willing to sacrifice to avoid the assault).
And while a strict cost-benefit analysis doesn't incorporate other benefits to testing, those are still worth noting: closure for victims, the exoneration of falsely accused people, potential prevention of other crimes committed by sexual offenders, and the fact that populating the CODIS database with more kits can help solve future crimes and incentivize other victims to report sexual assaults, one of the most underreported violent crimes.
Taken together, these calculations lead to one conclusion: a large one-time investment in testing the nation's backlog and investigating the resulting hits will pay off. The 2016 Sexual Assault Survivors' Rights Act
takes the important first steps of granting victims the right to collect and preserve sexual assault kits, but stops short of requiring testing.
The recent reauthorization of the Sexual Assault Forensic Evidence Reporting Act increases funding
for untested kits by 35%. While individual states and cities have gotten creative in trying to fund the testing of kit backlogs and communities can apply for federal grants through the Bureau of Justice's Sexual Assault Kit Initiative, this issue is too important to leave to crowdfunding, grant applications or modest budget increases. Now that there is an economic argument
to buttress the moral argument, there are no excuses: The backlog must be tested.