Fortuity and Forensic Familial Identification

Natalie Ram

Posted in , , , , ,

On July 7, 2010, Los Angeles police arrested a suspect in the Grim Sleeper murders, so-called because of a decade-long series of break-in killings. The critical lead in the case: a genetic profile in the state’s DNA database that was similar, but not identical, to the killer’s DNA. The partial match indicated a possible close genetic relationship to the Grim Sleeper. California’s apparent success in the Grim Sleeper case has intensified interest about policy-making for partial matching. However, little information about existing policies—and the wisdom of those policies—has been available. This Article reports the results of a survey of state policies governing partial matching. Although state policies are in continual flux, especially in response to California’s apparent success in the Grim Sleeper case, this survey is the most complete dataset available. The Article also dismantles a distinction drawn by more than a dozen states between fortuitous partial matches and partial matches that are deliberately sought, exposing this distinction as harmful and illogical.


Existing State Policies for Reporting Partial Matches

For DNA-based investigations, police usually use CODIS, the Combined DNA Index System, an FBI database into which participating states and agencies upload lawfully acquired genetic profiles and search among DNA profiles from other jurisdictions. All fifty states, the District of Columbia, and the federal government collect, store, and share genetic information through CODIS. Under a 2006 interim FBI policy, states may also share partial match information under certain circumstances.

Partial matches in CODIS may be uncovered fortuitously during routine database searches intended to identify exact matches. However, deliberate searches for partial matches (frequently termed “familial searching”) have also been done. Both kinds of searches may lead to relatives of CODIS-profiled subjects, who might not otherwise be suspected of a crime.

States have taken a variety of approaches to partial matching. This survey gathered data from forty-seven state and federal jurisdictions. Forty-one jurisdictions have articulated policies for partial match reporting. Of these, at least nineteen permit or have permitted partial DNA matches for familial investigations.1 Four others—Illinois, North Dakota, Virginia, and West Virginia—reported that policies are currently being formulated. Arkansas and Pennsylvania personnel were unsure as to any policy. Despite a variety of policies, some significant trends are apparent.

Figure 1: How Many States Permit Partial Matching of Any Kind?

First, among the nineteen states that use partial matches, nearly all distinguish between fortuitous and deliberate partial matches, permitting the former while precluding the latter. All nineteen have approved or reported fortuitous partial matches at least once, but at least fourteen states do not permit deliberate familial searching (Figure 2).[2. Alabama, Arizona, Florida, Louisiana, Minnesota, Missouri, Montana, New York, North Carolina, Oklahoma, Oregon, South Carolina, Washington, and Wyoming. Connecticut’s policy is vague with respect to deliberate partial matches, while it clearly permits reporting fortuitous partial matches.] Moreover, even the four states that permit both types of partial matching—California, Colorado, Nebraska, and Texas—often distinguish between the two. For example, California allows deliberate partial match searches only if the unsolved crime presents “critical public safety implications.”

Figure 2: What Kinds of Partial Matching Are Permitted?

Second, many state policies are unwritten or difficult to access. Forty-one responding jurisdictions have partial match policies or practices, but these policies are unwritten in at least eighteen of them. These eighteen mostly have unwritten policies not to search for or distribute partial match information. In contrast, states that permit reporting of fortuitous partial matches generally have written policies on this practice. Of nineteen states permitting partial matching, only four—Louisiana, Montana, North Carolina, and South Carolina—have reported a partial match absent a written policy. California, Colorado, the District of Columbia, Maryland, and New York are marked outliers, having codified their policies in a public memorandum, public policy statement, two statutes, and regulations, respectively.

Figure 3: Are Partial Match Policies Written? Where?

Even written policies may be difficult to locate. Twelve of the states permitting partial matching have recorded their policies only in lab manuals, which may be nearly as inaccessible as unwritten policies. Of the three states that have publicized their policies permitting partial matching—California, Colorado, and New York—only New York adopted its policy through a formal public process. As of this writing, only the District of Columbia and Maryland have enacted statutes regarding partial matches, and these statutes expressly forbid deliberate partial match searches.

Third, states permitting partial matching often impose additional requirements prior to disclosing a partial match. At least fourteen states recommend or require that reporting a partial match to investigators be contingent on additional evidence of a possible familial relationship between the DNA samples being compared.2 Eleven states require that a minimum proportion—generally, at least half—of the partially matched DNA markers be identical.3 The most detailed regulations are found in states that allow both fortuitous and deliberate partial matches. Conversely, the four states that have reported partial matches in the absence of a written policy have fewer barriers to their use.

Dismantling the Distinction Between Fortuitous and Deliberate Partial Matching

At least fourteen states distinguish fortuitous and deliberate partial matches, permitting the former but not the latter. This review suggests that this distinction is supported by neither logic nor principle. States should treat fortuitous and deliberate methods similarly.

A. The Fortuitous/Deliberate Distinction Imposes Structural Costs

Distinguishing between fortuitous and deliberate partial matching may encourage strategic behavior by laboratory personnel regarding what constitutes a “routine” database search. CODIS software supports three search stringency levels: high, moderate, and low. Several states have already adopted moderate stringency searches as routine. When only fortuitous partial matches are allowed, lab personnel may routinely set the stringency of their searches lower than necessary so that more partial matches “fortuitously” occur. This statistical adjustment may thereby circumvent a state’s decision to eschew deliberate partial matching.

The fortuitous/deliberate distinction also encourages strategic behavior by policymakers interested in taking advantage of the crime-solving potential of partial matching without bearing the costs of public scrutiny. Congress has made no law regulating partial matching, leading Thomas Callaghan, the former CODIS Unit Chief, to comment, “The FBI would be more comfortable with congressional authorization to conduct familial searches.”4 Even if there is reason for the public to worry about deliberate partial matching, the theory goes, such concerns need not be visited upon its benign, fortuitous cousin. As argued below, however, fortuitous partial matching is not as distinct from deliberate partial matching as policymakers have assumed.

B. Arguments Favoring and Opposing Partial Matching Apply Broadly

Indeed, the generally accepted reasons for allowing fortuitous partial matching also justify deliberate searches for partial matches. The chief justification usually given for fortuitous partial match reporting is that more crimes can be solved. Deliberate partial matching, however, also advances that goal. Both methods of partial matching use genetic data to incorporate otherwise non-included individuals into the scope of investigation.

Similarly, arguments against partial matching apply equally to both fortuitous and deliberate partial matches. In a recent analysis, Erin Murphy argues against use of partial matching because it unjustly distinguishes between individuals related to databased offenders and individuals who are not; aggravates racial disparities already embedded in offender-based DNA databases; invades privacy; makes offenders “involuntary ‘genetic informants’”5; and subverts democratic values since partial matching rarely has statutory authorization or public knowledge.  These arguments share a basic objection to otherwise “presumed innocent” individuals becoming implicitly suspect through mere “biological happenstance.”6 This problem of “biological happenstance” would, however, apply equally to both fortuitous and deliberate partial matching. Every database search accomplishes implicit inclusion regardless of whether, how often, or how deliberately such matches are found. States therefore cannot rely on these arguments for using fortuitous partial matches differently from partial matches obtained through deliberate familial searches.

C. Arguments Favoring a Fortuitous/Deliberate Distinction Are Insufficient

The most intuitive distinction between fortuitous and deliberate partial matching turns on intent. The FBI distinguished its interim policy on partial matching from “deliberately trolling the database looking for relatives.”7 But fortuitous partial matching is not something that “just happens”; it requires intentional conduct beyond the initial database search itself. Most states, in fact, require additional evidence before partial match information may be disclosed, and disclosure to investigators must be authorized by specific authorities, such as lab personnel or state attorneys general. Moreover, partial matches, whether fortuitous or deliberately sought, are, at best, weak evidence that a particular individual has engaged in wrongdoing. Partial match information thus does not, absent follow-up analysis, provide sufficient reason to suspect someone of wrongdoing.

Another cluster of arguments about partial matching relates to the limited resources of laboratories and law enforcement units, asserting that fortuitous matches are functionally “better” than deliberately identified ones for reasons of quality or cost. But there is no sound basis to believe that fortuitous partial matches are of higher quality or lower overall cost than deliberately uncovered ones. This is especially so in view of the fact that CODIS is not well designed to identify familial relationships and the pioneering of more sensitive database software in California, Colorado, and elsewhere. There is also little reason to believe that, insofar as DNA evidence may monopolize investigative efforts while other critical investigative information is lost in the interim, that allure is any less harmful in the context of fortuitous partial matches than deliberate ones. If partial matching is to be pursued, more principled lines need to be drawn in response to concerns about skewed investigations than the one between fortuitous and deliberate partial matches. For instance, states might restrict partial matching to those cases in which other investigative leads have been exhausted. To date, only California, Nebraska, and Texas have implemented such a rule, and each permits both forms of partial matching.

Our sympathies for the plight of the lab analyst who discovers, but cannot report, a fortuitous partial match should be no more persuasive. Such a perspective once again falsely treats fortuitous partial matches both as data points that “just happen” during routine database searches and as information that can simply be reported to investigators. As the preceding discussions have made evident, extra steps are necessary wherever partial matches are in play, whether fortuitous or deliberate. This argument thus collapses back into an argument about the relevance of intent with regard to identifying partial matches.

Finally, the possibility of a near-perfect match is not a sound basis for justifying the partial matching policies states have put in place. No state has in place so limited a policy permitting partial matching. Moreover, such matches are extremely unlikely to occur. If arguments about privacy, discrimination, or family integrity are at all persuasive, they remain persuasive as a general matter and ought not be breached for the sake of a policy that will virtually never yield its intended benefits. And if these arguments are not persuasive, then there is no sense in only responding to near-perfect matches when they arise fortuitously.

Conclusion

Very little is publicly known about state practices regarding partial matching, whether fortuitously or deliberately discovered. Policymaking in this arena has largely taken place behind closed laboratory doors, with little or no public knowledge. But public interest in and concern about partial matching is growing rapidly, thanks to California’s evident success in identifying the Grim Sleeper killer.

This survey reports that at least nineteen states have already released partial match information to investigators or have a policy in place for doing so. At least fourteen of these states draw a clear line between fortuitous and deliberate partial matches.

But this line creates a distinction without a difference, and policymakers should reject it. States ought to treat the two forms of partial matching similarly. The reason is simple: public oversight matters. The distinction between fortuitous and deliberate partial matching obscures what partial matching does—it provides genetically-based investigative leads to possible perpetrators, but at the same time makes numerous innocent individuals potential investigatory targets simply by virtue of their genetic relationship with a past offender or arrestee. This is not the ordinary course of genetic identification or the traditional use of DNA databases. If we are to accept such a change, we ought to do so forthrightly and transparently. Hiding this fact behind the fortuitous/deliberate distinction disserves the public.

Acknowledgments:

Copyright © 2011 Stanford Law Review.

About the Author: Natalie Ram is a Law Clerk to the Honorable Stephen G. Breyer, Associate Justice, U.S. Supreme Court.

Citation: Natalie Ram, Fortuity and Forensic Familial Identification, LEGAL WORKSHOP, Nov. 16, 2011, https://legalworkshop.org/2011/11/16/fortuity-forensic-familial-identification.

Based on: Natalie Ram, Fortuity and Forensic Familial Identification, 63 STAN. L. REV. 751 (2011).

  1. Alabama, Arizona, California, Colorado, Connecticut, Florida, Louisiana, Minnesota, Missouri, Montana, Nebraska, New York, North Carolina, Oklahoma, Oregon, South Carolina, Texas, Washington, and Wyoming. As this article was going to press, two states underwent significant changes in policy. Virginia completed its draft policy and validated software for deliberate partial matching. That state now conducts at least deliberate searches for partial matches. In Minnesota, draft legislation was introduced that would permit both fortuitous and deliberate partial matching. These new data are not reflected in the Article’s analysis, but they do not change its reasoning or conclusion.
  2. Alabama, Arizona, California, Colorado, Connecticut, Florida, Minnesota, Missouri, Nebraska, New York, Oklahoma, Texas, Washington, and Wyoming.
  3. Arizona, California, Colorado, Connecticut, Florida, Nebraska, North Carolina, Oklahoma, Oregon, Texas, and Washington.
  4. Ellen Nakashima, From DNA of Family, a Tool to Make Arrests, WASH. POST, Apr. 21, 2008, at A1 (quoting Thomas Callaghan).
  5. Erin Murphy, Relative Doubt: Familial Searches of DNA Databases, 109 MICH. L. REV. 291, 320 (2010).
  6. Michael Seringhaus, The Problem Child: Forensic DNA Databases, Familial Search, and A Call for Reform 70 (May 5, 2010) (unpublished manuscript), available at http://digitalcommons.law.yale.edu/ylsspps_papers/50/.
  7. Nakashima, supra note 4, at A1 (quoting Thomas Callaghan).

Post a Comment (all fields are required)

You must be logged in to post a comment.