How much justice is denied? An estimate of unsubmitted sexual assault kits in the United States

Highlights

• Vast amounts of sexual assault kits are never submitted for forensic testing by law enforcement agencies.

• Imputation and estimation model found between 300,000–400,000 unsubmitted sexual assault kits between 2014 and 18.

• County sociodemographic characteristics, crime levels, and state legislation predict county unsubmitted sexual assault kit counts.

• Lack of state data is a major issue for informed decisions to reduce the rape kit backlog

Abstract

Purpose

Prior national, state, and local investigations of law enforcement agencies have revealed large stockpiles of sexual assault kits (SAKs) that were not submitted to a laboratory for forensic testing. The failure to submit these SAKs has resulted in incomplete investigations for sexual assault victims and a lack of accountability for sexual assault offenders. To direct national policy aimed at testing these kits and reducing future stockpiles, it is essential to realize the magnitude of the problem.

Methods

This study uses information on known unsubmitted SAK counts from 911 counties in 15 states and a multitude of county-level covariates to estimate the national number of unsubmitted SAKs during the period 2014–2018.

Results

Based on 95% confidence intervals, there were an estimated 300,000 to 400,000 unsubmitted SAKs in the United States during this period. A county's population was the strongest predictor of whether the county had an unsubmitted SAK count, while county sociodemographic factors and state legislative actions were the strongest predictors of the size of county unsubmitted SAK counts.

Conclusions

To improve future estimates, LEAs should account for SAKs throughout the investigative and laboratory submission stages and states should conduct annual high-quality audits that leverage tracking processes.

Introduction

Sexual assault is widespread in the United States. As many as 1 in 5 (19.3%) women and 1 in 71 (1.7%) men report being raped in their lifetimes (Breiding et al., 2014). After an assault, some survivors choose to seek care from medical providers, where injuries can be treated and a sexual assault kit (SAK) can be collected to preserve physical evidence for potential use within the criminal justice system (Campbell, Patterson, & Lichty, 2005). SAK collection typically comprises a history of the assault, a head-to-toe physical exam, and specimen collection from body surfaces (e.g., swabbing of the vulva, anus, or mouth) (Campbell et al., 2005). This process is often long and invasive for victims, but it is completed with the intention that the SAK will be submitted by the responding law enforcement agency (LEA) to a forensic laboratory for testing and used to aid in the criminal investigation and prosecution of the offender (Campbell et al., 2015).

The benefits of SAK testing are far-reaching. Testing the forensic evidence within a SAK allows for potential extraction of DNA profiles that may then be uploaded into Combined DNA Index System (CODIS). Uploads may then support the identification of previously unknown offenders, confirmation of the identity of known offenders, and/or establishment of linkages to serial sex offenders. Additionally, SAK testing may result in additional sanctions for suspected sexual offenders (e.g., increased likelihood of arrest and conviction; Lovell, Luminais, Flannery, Bell, & Kyker, 2018; Peterson & Sommers, 2010; Wells, Fansher, & Campbell, 2019). Thus, testing the DNA within a SAK can allow law enforcement to identify and prosecute offenders, preventing the opportunity for these perpetrators to commit additional sex crimes (Campbell, Feeney, Goodman-Williams, Sharma, & Pierce, 2019). For victims, choosing to have a SAK submitted for testing and having the decision executed may contribute to feelings of personal agency, trust, and empowerment, which may ultimately support victims' healing trajectories (Sulley et al., 2018; Ullman & Townsend, 2008).

Despite the tremendous utility tested SAKs provide and the hardship victims endure to provide them (Du Mont, White, & McGregor, 2009), reports from jurisdictions across the country reveal that SAKs too frequently go unsubmitted for testing by LEAs (Campbell, Feeney, Fehler-Cabral, Shaw, & Horsford, 2017; Strom & Hickman, 2010). SAKs that law enforcement agencies do not submit for testing are referred to as unsubmitted SAKs, while SAKs that are submitted to crime labs for testing but remain unanalyzed are referred to as untested SAKs. The latter is typically caused by a lack of lab resources, changes in technology, and/or processing delays. For most communities, the former, unsubmitted SAKs, have more complicated origins including rapid changes in DNA technology, resource depletion, and/or endorsement of rape myths by law enforcement (Campbell et al., 2015; HRW, 2009). The focus of the current paper is on unsubmitted, not untested, SAKs. We occasionally refer to an agency's number of unsubmitted SAKs as their rape kit backlog.

While research has yet to establish a current number of unsubmitted SAKs at a national level, the estimate that is most frequently cited by politicians, press, and researchers is one that comes with little justification: 400,000 SAKs (Dickson, 2014; TEST400k, 2016; The White House, 2015). We tracked this estimate's origin back to at least 2004 during the passing of the Justice for All Act (U.S. Senate, 2004), but we were unable to establish a known scientific backing for its development. Instead, the 400,000 number appears to capture a rough estimate of the ever-changing number of previously unsubmitted SAKs and, importantly, reflects an estimate that was projected almost a decade before increased awareness of unsubmitted SAKs encouraged jurisdictions across the country to take stock of and reduce their stockpiles. There is a need for a national estimate that is more recent and based on transparent scientific methods to inform current policy decisions and public discourse. Timely and research-informed information on the prevalence of unsubmitted SAKs can also serve as a key mechanism for accountability and to evaluate how successful the various reform efforts have been toward not only testing SAKs but also actively and consistently pursuing justice for victims in these cases.

Research suggests that SAK backlogs were created over decades because of multiple interwoven processes. First, rapid changes in DNA technology and misguided perceptions of SAK utility resulted in many LEAs viewing forensic testing as a futile activity for sexual assault investigations (Lovrich, Pratt, Gaffney, et al., 2004; Strom & Hickman, 2010). For many jurisdictions, lack of critical resources (e.g., time, personnel, expendable finances) left LEAs prioritizing the investigation of cases that did not require forensic testing (Campbell et al., 2015; Human Rights Watch, 2009). Finally, rape myth acceptance among stakeholders and decision makers resulted in choices to not invest in the safety of their most vulnerable citizens via SAK testing, particularly to the detriment of marginalized groups (Shaw & Campbell, 2013). Moreover, only recently have state legislators begun enacting laws that establish how local and state LEAs must handle SAK testing. These laws often require one or more inventories of unsubmitted SAKs, the testing of all old and/or new kits, and a process for tracking kits from collection through forensic testing (endthebacklog.org, 2020). Without state laws like these catalyzing changes to LEA policy on SAK testing, it is possible jurisdictions would allow for SAK backlogs and their consequences to remain indefinitely.

For some jurisdictions, a “perfect storm” (Campbell et al., 2015) of these circumstances resulted in incredibly low rates of SAK submission. For example, in the late 1990's, researchers found that up to 50% of SAKs collected at a women's health center went unsubmitted (Parnis, 1997). More recent research found jurisdictional rates ranged from 20% to 60% of SAKs submitted for testing (Human Rights Watch, 2009, Human Rights Watch, 2010; Patterson & Campbell, 2012). In New York, the first city to report a backlog of unsubmitted SAKs in the late 1990s, approximately 16,000 SAKs were sitting in police storage (Bashford, 2013). Over a decade later in 2009, Human Rights Watch reported that there were 12,669 unsubmitted SAKs in Los Angeles County (Human Rights Watch, 2009). More recent reports have documented between 200 (Nashville) and 12,000 (Memphis) unsubmitted SAKs in numerous cities across the United States (Campbell et al., 2017; Joyful Heart Foundation, 2019). This problem extends far beyond cities, with individual states conducting audits to reveal similarly large numbers from 2014 through 2018 (e.g., North Carolina: 15,610 in 2018; Oregon: 4902 in 2015; Texas: 2138 in 2017) (Joyful Heart Foundation, 2019).

Individual agencies, cities, and states have conducted inventories to measure the scale of “justice denied” in their areas. Attempts to scale up measurement to a national level have faced many challenges, due in part to a lack of adequate tracking mechanisms and inventorying among LEAs. In one attempt at producing a national number of unsubmitted SAKs, Lovrich et al. (2004) surveyed a nationally representative stratified random sample of approximately 3400 LEAs and estimated that in 1982–2002 there may have been as many as 169,000 unsolved sexual assault cases with untested biological evidence in the United States. In a follow-on study, Strom and Hickman (2010) found that 18% of unsolved sexual assault cases from 2003 to 2007 (an additional estimated 27,595 cases) contained forensic evidence that was not submitted to a crime laboratory for analysis. While these studies were critical in demonstrating the extent of the national problem, survey respondents sometimes had to approximate the presence of untested forensic evidence in unsolved cases, as they did not have the data systems or capacity to physically count evidence.

The Joyful Heart Foundation, a national advocacy group pushing for the inventorying and testing of previously unsubmitted SAKs through their End the Backlog campaign, has identified over 100,000 previously untested kits across selected states (Joyful Heart Foundation, 2019). This figure is surely an underestimate, because it does not account for unsubmitted SAKs from the 12 states that have not formally inventoried or reported their counts. Furthermore, because the number of unsubmitted SAKs is constantly changing and states use varying definitions for what constitutes an untested and an unsubmitted SAK, any national estimate must be based on clear, replicable criteria. When reviewing past audits from some of the states included in the Joyful Heart Foundation's count, we noticed state variation in both the definitions used for an unsubmitted SAK and the audit response rate. Because there is no methodology report associated with their count, it is impossible to know how the Joyful Heart Foundation handled these issues when publishing their state counts. The current study builds on their important work by estimating a national number for all 50 states and the District of Columbia that is based on unsubmitted SAK counts taken in states with consistent and correct definitions of an unsubmitted SAK and high response rates to the state audit.

The problem of shelving kits indefinitely has been referred to as “justice denied” by Strom and Hickman (Strom & Hickman, 2010, p382) because it eliminates the opportunity for the SAK to be used in criminal investigations and court proceedings. This leaves victims without a thorough exploration of their cases or the option for prosecution and offenders with the opportunity to re-offend. In addition, leaving SAKs unsubmitted for testing prevents law enforcement from making connections that can help identify repeat and serial sex offenders and exonerate wrongfully arrested or convicted individuals (Campbell et al., 2019, Campbell et al., 2019). For these reasons, the testing of SAKs has become a national political issue (The White House, 2015) and much federal funding has been directed to testing SAKs through the Sexual Assault Kit Initiative (SAKI; SAKI, 2019). To inform future national legislation and funding decisions in this area, it is imperative that policy makers have an accurate, evidence-based understanding of the problem's scope. Unfortunately, due to a lack of national data on unsubmitted SAKs, policy makers have relied on an estimate that is not based on any clear science. To remedy this, we use county-level inventories collected from 2014 through 2018 from 911 counties in 15 states along with multiple county-level covariates to impute the remaining counties' unsubmitted SAK counts and develop a national estimate. Although our national estimate relies heavily on imputed county counts, it is constructed from known information on many predictors of unsubmitted SAK counts and is based on transparent, consistent, and replicable procedures. Our findings describe the national extent of unsubmitted SAKs and the factors that contribute to the stockpile. This information can enable a grounded, research-driven discussion on the problem of unsubmitted SAKs, including its dynamic scope and possible causes and solutions. We also hope this study can highlight the need for a reliable and accepted process for routinely monitoring and tracking the collection, submission, and testing of SAKs at the national, state, and local level moving forward.