Forensic DNA Surveillance Tool

A critical question in the development of state DNA index system (SDIS) databases has been weighing the additional crime-solving and deterrence benefits of expansion against concerns that SDIS databases reinforce racial disparities in criminal justice involvement and become, in effect, genetic surveillance tools.[1] To help better understand this issue, we developed an online resource for estimating how different policies have shaped the racial composition of SDIS databases.

How to Use the Tool

How to Use the Tool

Left Panel

The left panel of the tool shows current state DNA database policies. Specifics on SDIS policies were gathered from the National Conference of State Legislatures' Forensic Science Laws Database,[2] the Justia website on state statutes,[3] and the website of forensic DNA lobbying firm Gordon Thomas Honeywell Governmental Affairs.[4]

Middle Panel

The middle panel allows the user to alter the current policies to increase or decrease the scope of the SDIS, as well as whether three common criminal offenses are typically included in the database. Laws pertaining to simple assault, larceny, and marijuana possession can vary significantly from state to state and from case to case and thus have the potential to significantly affect state DNA database population.[5]

In the lower middle panel, users can adjust the time frame. The main reason for changing the default time span of 2006-2015 is that the tool treats current state DNA database policies as having been in place over the entire period covered by the model. This is not the case, and in fact some laws did not take effect until after 2015, the latest year for which arrest data were available.[6] This oversimplification can be mitigated by changing the default policies and restricting the time frame as needed. The bottom of the middle panel asks users to choose values for three important state-level factors that impact the population that can be identified through the state DNA database. The re-arrest rate is the annual percentage of arrestees that are prior arrestees in order to offset the database population by not double-counting recidivists. As a benchmark, the latest Bureau of Justice Statistics (BJS) report on felony defendants in the 75 largest counties found that in 2009, 62 percent of felony arrestees in these jurisdictions had a prior felony arrest, although this is likely an overestimate because it does not include felony arrestees who were never charged.[7] The average conviction rate is an important variable for establishing whose DNA is added to the SDIS when conviction is a prerequisite or expungement is automatic absent a conviction. The BJS report on felony defendants puts the overall felony conviction rate at 66 percent, although again, this percentage excludes those not charged.[8] California criminal justice data on felony arrestee case processing, which includes law enforcement releases and dismissals, indicate a slightly higher rate.[9] Available data, which are scant, suggest that the misdemeanor conviction rate is comparable to the felony conviction rate, although this likely varies by state and subpopulation. Between 2008 and 2017, for example, 95 percent of adult misdemeanor arrestees in California were referred for prosecution, virtually identical to the rate for adult felony arrestees over the same period.[10] Finally, the family member average is the mean number of first-degree relatives per databased offender that can potentially be identified through familial searching. Sophisticated demographic analysis might have yielded defensible estimates for this number. However, in light of the uncertainty around this and other model inputs, we opted to allow users to define this value, bearing in mind this number should be set to zero in any state that does not conduct familial database searches. In addition, a familial database search policy using Y-STR (patrilineal DNA testing) to confirm kinship would encompass offenders' fathers, sons, and brothers, whereas a familial database search using either Y-STR or mtDNA (matrilineal DNA testing) to confirm kinship would encompass offenders' fathers, mothers, sons, daughters, brothers, and sisters, so in theory the family member average would be roughly double that for Y-STR testing alone.

Only the racial disparity in arrest rates is built into the model. These rates are drawn from agency-level arrest data from the Uniform Crime Reports (UCR), which include race, gender, and age of arrestees, compared with American Community Survey state population estimates.[11] The resource also relies on UCR arrest data as the starting point for determining the size of each state database. The UCR data feature offense codes from which the racial breakdown for felony and misdemeanor arrests could be approximated, as previous researchers have done.[12]

Right Panel

The third panel compares the average state population, the state arrestee population excluding recidivists, and the searchable database population under model-provided and user-provided assumptions.


Let's use California as an example for how the tool works. Selecting California from the dropdown menu (or clicking the state of California on the map), the databasing policy description indicates DNA is collected from all convicted offenders and some misdemeanor convictions, including juveniles; all felony arrestees and some misdemeanor arrestees, excluding juveniles; conducts familial database searches involving the entire convicted offender database (male and female). With all three of the common offenses set to misdemeanors, the annual re-arrest rate set to 65 percent, the felony and misdemeanor conviction rate set to 80 percent, and a family member average of 0.2, the population that can be potentially identified through the SDIS database is 2,521,000 people. The racial breakdown of this population is 78.0% White, 19.6% Black, 0.4% Native American, and 1.9% Asian, compared to a state population of 30,675,000 that is 75.6% White, 7.4% Black, 0.9% Native American, and 16.1% Asian.


  1. Machado, Helena and Susana Silva, "What Influences Public Views on Forensic DNA in the Criminal Field? A Scoping Review of Quantitative Evidence," Human Genomics, Vol. 13, No. 23, 2019. Return to content
  2. National Conference of State Legislatures, "Forensic Science Database," November 17, 2014. As of August 5, 2019: Return to content
  3. Justia, homepage, undated. As of August 5, 2019: Return to content
  4. Gordon Thomas Honeywell Governmental Affairs, "DNA Resource: Forensic DNA Policy," undated. As of August 5, 2019: Return to content
  5. As of 2019, ten states and the District of Columbia, had legalized recreational marijuana use. However, this is a relatively recent trend, and it is still a crime for minors to possess, for persons to possess over a certain amount, or to possess in proximity to a school, for example. Return to content
  6. Utah, for example, did not begin collecting DNA from all felony arrestees until 2014. Utah Code § 53-10-404.5, online at:; Florida's felony arrestee DNA collection law (Florida Statutes § 943.325) did not take ef-fect until January of 2019; because Florida does not participate in UCR, its database cannot be es-timated with our online resource. Return to content
  7. Reaves, Brian A., "Felony Defendants in Large Urban Counties, 2009—Statistical Tables," U.S. Department of Justice, Bureau of Justice Statistics, NCJ 243777, Table 8, December 2013, p. 11. As of August 5, 2019: Return to content
  8. Reaves, 2013, p. 24. Return to content
  9. California Department of Justice, Interactive Crime Statistics Tables, "Arrest Dispositions," un-dated. As of August 5, 2019: Return to content
  10. California Department of Justice, undated. Return to content
  11. National Archive of Criminal Justice Data, "Types of UCR Data Available at NACJD: Agency-Level UCR Data: Arrests by Age, Sex, and Race, monthly reports," undated. As of August 5, 2019: Return to content
  12. Stevenson, Megan and Sandra Mayson, "The Scale of Misdemeanor Justice," Boston University Law Review, Vol. 98, 2018, pp. 731–777. As of August 5, 2019: Return to content