Tracking a Killer
Police Use of DNA and Forensic Genetics
Many of us love exciting investigative dramas like Bones and Hawaii Five-0. The twists and turns, momentary dead ends, and dramatic reveal of new clues keeps us on the edge of our seats. But we also know that real life is different. Most of the time.
Late last month, California investigators arrested Joseph James DeAngelo, who they suspect was the notorious killer and rapist known by several monikers including the Golden State Killer and the East Area Rapist. The details that emerged of the decades-long investigation sounded more like a Hollywood script than real life. In a genius move, officers used a DNA sample from one of the original crime scenes and a free database meant for genealogy enthusiasts to come up with a list of suspects, leading to the arrest of DeAngelo at his home near Sacramento.
Even though bringing someone who had committed many horrible atrocities against innocent citizens would be a major benefit to society, numerous articles worrying about the privacy implications of the investigation appeared in its aftermath. For example, some questioned whether police ought to be allowed to use a third party’s personal information without their consent (or even knowledge) without a warrant or other oversight, even if such information is freely available.
Again, bringing criminals to justice is a good thing and we should thank all law enforcement involved in this arrest. But as science unlocks more personal secrets encoded in our DNA, we ought to reevaluate how we allow government to use this extremely personal information.
The story of DeAngelo’s arrest begins in the 1970s. Between 1974 and 1986, the Golden State Killer terrorized ten California counties, committing multiple rapes, murders, and burglaries. The stories victims tell of these crimes emphasize just how sadistic and vile this serial criminal is. But in 1986, the crime spree stops.
Almost ten years later, an investigator and forensic scientist named Paul Holes discovers the Golden State Killer case file and becomes obsessed.
A colleague tells Holes about a case he was working on in which a kidnapping victim was identified by using a sample of her DNA and a genealogy website. Holes thinks he might be able to do the same thing in the Golden State Killer case.
After some research, Holes finds a genealogy service called GEDmatch, a free, open website where enthusiasts can upload raw genetic data in order to identify potential relatives. Holes obtains a sample of the Golden State Killer’s DNA from a duplicate rape kit placed in a freezer by a pathologist working on one of the murders in 1980. A lab sequences the sample and converts it to a format suitable to upload to the GEDmatch database.
The database provides several possible distant relatives of the killer, most about the equivalent of third cousins. Holes and his team then work backwards to construct a family tree that includes these matches. If he goes back far enough he can pinpoint an ancestor that all the matches, and the killer, shared in common. Once there, he can move forward in time to attach a name and a face to the killer.
Holes is eventually able to find that common ancestor. It was a set great-great-great grandparents who lived in the early 1800s. Then he uses publicly available information like census data and newspaper clippings to find members of the intervening generations. As he gets closer to the present, he turns to police databases and tools like LexisNexis. This eventually results in a huge maze of family trees that branch out from those great-great-great grandparents. He begins looking for individuals that match the Golden State Killer’s age and gender, and has connections to the locations of the crimes. Eventually he lands on DeAngelo.
The police place DeAngelo under surveillance. To test if their genealogy work has been correct, officers pick up an item containing DeAngelo’s DNA that he discards in public. This DNA is sequenced and compared to the DNA from the original 1980 crime scene. The two match. Just for good measure, investigators repeat this process with another discarded item. This evidence is enough for a judge to sign an arrest warrant, which investigators carry out on April 24.
Use of GEDmatch Database
It seems that there are a couple key aspects of the investigation that could lead privacy conscious observers to worry.
The first is the use of the GEDmatch database by Holes and his team.
GEDmatch is a free database open to both amatur and professional genealogy enthusiasts that provides DNA and genealogical analysis tools. Users upload their raw, autosomal DNA (autosomes are the numbered chromosomes as opposed to the X and Y chromosomes) that has already been tested by a company like 23andMe or Ancestry. Since the database can read data from multiple testing companies, users can locate possible relatives from a much broader group of people.
Once the user uploads their data to the database, GEDmatch searches for sections of the user’s chromosomes that match other users in the GEDmatch universe. The more matching sections, the more likely the two users are related. The service will then provide the user with a spreadsheet that includes the usernames of matches, an estimation of how closely the two are related, how much DNA is common to both profiles, and any contact information that the match has included. The purpose of the service is to spur collaboration between two relatives who want to explore their family’s history and create a family tree.
The policy also specifies that users uploading raw DNA data agree that the raw data is one of a number of options. One of those options is as follows:
DNA obtained and authorized by law enforcement to either: (1) identify a perpetrator of a violent crime against another individual; or (2) identify remains of a deceased individual
It appears that this option was added after DeAngelo’s arrest, though a lawyer affiliated with the site stated that the actions of the police still did not technically violate the site’s policies. Either way, the addition of this option explicitly legitimizes future investigative uses.
And the police have noticed. Investigators in Washington State used GEDmatch to arrest a suspect in a double murder from 1987.
As indicated above, the privacy section (which was included prior to the Golden State Killer episode) explicitly warns users of the open nature of the database:
In today’s world, there are real dangers of identity theft, credit fraud, etc. We try to strike a balance between these conflicting realities and the need to share information with other users. In the end, if you require absolute privacy and security, you agree that you will not provide your personal information, Raw Data, or Genealogy Data to GEDmatch. If you do not agree and you have already provided your personal information, Raw Data, or Genealogy Data, you agree to delete it immediately.
In light of these policies, Holes and his colleagues did not need a warrant to upload the DNA data from the 1980 crime scene to the GEDmatch database, since that DNA was already legally in police custody.
While it seems no part of the use of the database was illegal, it is concerning for several reasons.
Police could link DeAngelo to the crime even though he did not personally upload his data to GEDmatch. Instead, his distant relatives uploaded their own data, something DeAngelo could never expect to control. Using modern technology and DNA from a family member (even a distant one), police can have access to personal information which they never received permission to have and use. And considering that a random American of European background has a 90% chance of finding a third cousin in GEDmatch, and a 94% chance of finding a second cousin in the Ancestory.com database, the police could use these methods to discover personal information of many Americans, whether they committed a crime or not.
Furthermore, the genealogy approach used by investigators in this case is not perfect. In 2015, a filmmaker named Michael Usry opened his front door to find three police officers asking him to accompany them down to the station. He learned that he was a suspect in an almost twenty year-old murder of a young girl. Police had focused on Usry after they used a publicly available version of Ancestry's database and discovered that the DNA of Usry’s father (who was in the Ancestry database) bore similarities to the DNA found at the crime scene.
After taking a sample of Usry’s DNA and comparing it to the sample found at the crime scene, he was cleared of all suspicion since the samples did not match. But Usry had to wait for thirty three tense days before the sample could be processed.
These false-positives are not uncommon, as Erin Murphy, a New York University law professor notes in a Wired article about the Usry incident. This can create life-ruining consequences for innocent citizens, especially in a digital age where news travels quickly and social media outcries can pressure organizations to take hasty actions. That same Wired article invites readers to imagine if Usry was a school teacher under suspicion of murdering a young girl. If this was leaked to the media during that thirty three day waiting period, Usry would likely face irreparable damage to his career and reputation, even though he was completely innocent.
Law enforcement has already dramatically increased its efforts to use this genealogy method to track down more criminals. Working with several law enforcement agencies, a Virginia company called Parabon NanoLabs acknowledged that have uploaded DNA data from about one hundred crime scenes into GEDmatch. In about twenty of these cases so far, the company says the database has found relatives that are third cousins or closer.
It is extremely important that proper oversight of these efforts is in place to protect the innocent that will inevitably be implicated in these cases. As is, the police do not need any warrants to use the GEDmatch site. States and localities could require that police consult a judge or expert council of some kind at each stage of the investigation, especially as specific citizens are identified and come under suspicion.
Obtaining of DeAngelo’s DNA
The second key aspect of the investigation that could raise concerns is the obtaining of DeAngelo’s DNA by Holes and his team.
After the genealogy work led Holes to DeAngelo, he needed evidence that he was on the right track. The police placed DeAngelo under surveillance and picked up a couple items that contained his DNA that he discarded in public. They compared this discarded DNA to the DNA sample found at the scene of the original 1980 crime. The samples matched, giving Holes enough evidence to get a judge to sign an arrest warrant for DeAngelo.
The Fourth Amendment provides broad parameters for law enforcement’s gathering of evidence. The Amendment reads,
The right of the people to be secure in their persons, houses, papers, and effects, against unreasonable searches and seizures, shall not be violated, and no Warrants shall issue, but upon probable cause, supported by Oath or affirmation, and particularly describing the place to be searched, and the persons or things to be seized.
The goal of this Amendment is to protect privacy and property rights against unreasonable actions of government. To determine whether a search or seizure is reasonable or unreasonable, a court must decide if the citizen had a reasonable expectation of privacy. For example, courts have afforded great protections against warrantless searches of private homes because Western society has placed an emphasis on the privacy and sanctity of the home.
Police may seize property that is in plain view, in an open field, or has been abandoned. They do not need a warrant to take property in these circumstances because courts have found that the owner does not have a reasonable expectation of privacy.
The Amendment treats persons differently than it does property. Police cannot seize someone just because she is walking through an open field. In most situations, an arrest warrant is needed to detain or transport someone against her will. Seizing someone is only justified when the officer has probable cause and believes there are exigent circumstances that necessitate immediate action. Exigent circumstances could include harm to others or the destruction of relevant evidence. Unless these conditions are met, officers may not detain someone against their will without getting a warrant.
As law professor Albert Scherr explains in the Georgia Law Review , current DNA gathering jurisprudence creates an “all-or-nothing dynamic.” DNA abandoned in public places falls on the “all” side of the equation. Police are free to take and use DNA that has been discarded with no controls imposed by the Fourth Amendment, whether a person is a suspect or not. Considering the vast amount of deeply personal information that is contained in DNA, this is concerning.
Scherr reasonably proposes that courts recognize that DNA belongs to a middle category. Police may still pick up and use discarded property that bears cells with DNA, just as before. But in order to mine the information contained in the DNA, police must get a warrant supported by probable cause.
The Constitution represents a contract between the citizens of the United States and the government of the United States. In this contract, the citizens agree to cede certain natural rights to the government, in order that the public benefits named in the Preamble might be achieved more efficiently.
Perhaps the most consequential right that the citizenry has bestowed upon its government is the authority to coerce wrongdoers. This granting of coercive power absolutely must be accompanied by strict limitations on its use in order to protect the innocent. The authors of the Constitution placed some of these limitations explicitly in the first ten amendments, and the Fourth Amendment is especially important.
The recent arrest of DeAngelo thanks to forensic genetics, and the reports of rapid increases in these techniques, make it clear that courts must rethink their application of the Fourth Amendment. Legislatures at all levels ought to also consider how they might implement new protections that go beyond current jurisprudence.
A government, even one that is well-meaning, can irreparably damage the lives of innocent citizens when it has both unchecked coercive power and access to such intensely personal information as DNA. Strong, clear controls on how police collect and use DNA are needed now more than ever.
June 13, 2018 Update
On June 1, Judge Michael Sweet of the Superior Court of Sacramento County released redacted versions of both the arrest warrant and the search warrant in the Golden State Killer case.
The release came in response to a motion filed by a group of local media companies seeking to unseal two warrants pertaining to the case. DeAngelo’s public defender sought to block the motion for fear that the information could poison potential jurors and deny DeAngelo’s right to a fair trial. Judge Sweet tried to balance DeAngelo’s rights and the public’s right to open judicial proceedings by personally going through each page and redacting potentially problematic information before releasing the documents.
The warrants fill in several details about how investigators obtained DNA samples from DeAngelo to link him to the crimes committed by the Golden State Killer.
After investigators begin to suspect DeAngelo based on the results of the GEDmatch search and the construction of a family tree, they place DeAngelo under surveillance. On April 18, the investigators observe DeAngelo get into his car and drive to a Hobby Lobby in Roseville, CA, and park the car in the public lot. While DeAngelo is inside shopping, officers swab the driver’s side door handle of the vehicle to collect DNA samples.
The results come back on April 20. DNA from three people is found on the door handle. The DNA of one of these people matches DNA found at several of the original crime scenes.
Immediately, officers begin to watch DeAngelo’s home in Citrus Heights around the clock. They see that he is the only male at the home during that time. On the night of April 22, DeAngelo places his garbage can on the street in front of his house for pick up. The next morning investigators take several items from the garbage can in order to test them for DNA. One item, a used tissue, provides a readable sample. This sample also matches the crime scene DNA.
Investigators took this DNA information, as well as a multitude of other evidence, and incorporated it into requests for a warrant to search DeAngelo’s person and property and a warrant to arrest DeAngelo. On April 24, the officers submitted these requests to the Superior Court. A judge then signed the warrants and they were executed immediately.
These unsealed documents reveal that the vast majority of this investigation was carried out without any warrant. Importantly, the DNA collection took place without a judge’s approval. This is perfectly acceptable in current Fourth Amendment jurisprudence.
It remains to be seen if DeAngelo’s lawyers will raise these issues in their arguments or if the judge will use this opportunity to reconsider the issue of surreptitious DNA sampling.
Whether they do or not, the facts of this case that are being revealed make it clear that the treatment of DNA under the Fourth Amendment does, in fact, need to be reconsidered.