In numerous countries, there is considerable debate about establishing or expanding national DNA databases as an investigation tool for criminal law enforcement. The potential benefits of such a database seem enormous in this field: A comprehensive DNA database would not only help to solve crimes more easily, it would also function to prevent future crimes by preventing multiple crimes committed by the same offender and deterring crimes from being committed by other offenders. Finally, DNA evidence can play a large role in exonerating innocent suspects either if the true perpetrator is identified using the "genetic fingerprint" or if the innocent suspects genetic material is not found at the scene of crime.
Many countries have already introduced some forms of collecting DNA samples of criminals. Great Britain established the first criminal DNA database in 1995 and has now the world’s largest of such databases. With great success New Zealand has set up a DNA database including not only coerced samples of convicted felons but also samples of volunteers: Currently, approximately 55% of the crime scene samples match an individual sample present on the database (ESR, The DNA Databank, at http://www.esr.cri.nz/competencies/forensicscience/dna/DNAdatabank.htm, last visited May 14, 2006). In the United States, all states collect samples from certain convicted sex offenders, while some states even require collection of samples from all felons both violent and non-violent. A number of states restrict the types of criminal investigations allowed or require a court order for police access to the database. In addition to the individual state DNA databases, the FBI has created a national DNA database allowing for the exchange of data between the state databases (Combined DNA Index System, CODIS).
The best results in solving and preventing crimes as well as reducing wrongful arrests would arguably be achieved by a comprehensive population-wide DNA database (Kaye, D. H. / Smith, Michael E., DNA Identification Databases: Legality, Legitimacy, and the Case for Population-Wide Coverage, 2003 Wis. L. Rev. 413 (2003)). There should be network effects to large participation in national DNA databases. The more people participate the higher the success rate in solving and preventing crimes. Perpetrators that are part of the database would be found and convicted more easily, and if they are not part of the database the pool of suspects would be at least greatly reduced. Furthermore, people that are registered in the database will be more reluctant to commit crimes in the first place given their higher chances of being apprehended and punished. Moreover, law abiding citizen would not need to worry about being erroneously suspected to have committed a crime since they can quickly clear their name using the genetic data. Finally, it can be argued that the present practice of collecting DNA samples of only convicted persons creates an inherent bias in the system against certain groups. For example, the large number of felons who are black males also increase their abundance of black males in the databases and therefore lead to higher conviction rates in that part of the population (Kaye/Smith, 2003, 454). A nationwide DNA database could avoid such inherent biases.
However, assuming the desirability of a population-wide DNA database, if a government would introduce such a scheme on a voluntary basis, one would expect low participation for numerous reasons discussed below. The concept of libertarian paternalism could be a solution to the participation problem. This concept allows for steering people’s choices in welfare promoting directions without eliminating the freedom of choice (Sunstein, Cass R. / Thaler, Richard H., Libertarian Paternalism is not an Oxymoron, 70 U. Chi. L. Rev. 1159, 1162 (2003)). Applied to DNA databases, the idea is to switch the default rule to mandatory membership in the national DNA database with the right of each individual to opt out of this default. Practically, one could take a DNA sample of each newborn (and each immigrant) and store it safely without government access until the 18. birthday, when each individual receives the right to opt out of the database and have their genetic material deleted. (Possibly there should be an exception from the right to opt out of the database for persons that already committed certain crimes at that point.) In this case, there is no alternative to paternalism (Sunstein/Thaler, 2003, 1180-1182): Once you agree on the desirability of a comprehensive DNA database, the government has to set the default rule one way or the other. Social welfare can be maximized by setting frames and defaults in a welfare promoting way while still respecting the individual choice not to participate.
In this setting, two defaults can be imagined: either having a no mandatory participation in the national DNA database with the option for people to voluntarily opt-in into the database (as used in New Zealand today); or having mandatory registration in the database with the right to opt-out (as suggested by libertarian paternalism). In order to assess whether switching the default rule to mandatory participation with the right to opt out will reach higher participation rates, it becomes necessary to examine the reasons for low participation in a voluntary nationwide DNA database.
Numerous reasons for low participation in a voluntary national DNA database can be found in people’s bounded rationality. Sunstein and Thaler explain the importance of default rules with the psychological phenomena of status quo bias, suggestion, inertia, endowment effect, and ill-formed preferences (Sunstein/Thaler, 2003, 1179-1180). Due to these factors, either default will be "sticky" to some extend. Firstly, status quo bias can play a role in the stickiness of default rules. The existing regime is often robust against changes (Sunstein/Thaler, 2003, 1171; Kahneman, Daniel / Knetsch, Jack L. / Thaler, Richard H., Anomalies: The Endowment Effect, Loss Aversion, and Status Quo Bias, 5 J. Econ. Persp. 193, 197-199 (1991). Given uncertainty about the correct choice, people will often give the default an informational content: the default is seen as something most people do or at least that more informed people do. Moreover, depending on the default setting opting-in or opting-out requires the actions of people. Due to inertia there is a tendency to stick with whatever the default is. The endowment effect is likely to affect people’s valuation of what rights they have making the initial allocation of rights by a default more robust. Finally, many people might lack clear, stable preferences in this area (Sunstein/Thaler, 2003, 1179-1180). For the average law abiding citizen it might not be clear whether participation in the database will make him better or worse off. The above effects of bounded rationality will contribute to the low participation in a voluntary database. These psychological factors play a large role in the low participation of a data base with voluntary opt-in. The same phenomena will support high participation in a national DNA database as soon as the default is switched to mandatory registration with the right to opt-out.
There are several more aspects determining low participation in a voluntary database that are affected by a switch to default participation with an opt-out option:
Some people do not voluntarily join a criminal DNA database because they feel it infringes their right of privacy. These people will not be affected by the switch in the default because they can readily opt out of the database.
Another reason for the reluctance to join voluntarily can be explained by a free-riding effects. Citizen can already profit from the decrease of crime linked to the use of a DNA database in investigating crimes without having to join themselves. Default membership in the database can to some extend avoid such free-riding on the public good of general safety: The stickiness of participation as a default could balance out some free-riding tendencies. However, free-riding on reduced crime rates is still possible by opting out of the default database.
Other people may have an incentive to hold out as long as possible with their participation. For example, a person with criminal intention will be better off not being a part of the database as long as the participation rate in the database will not be significantly above 50% of the population. It is only then that holding out without being registered can result in being in the smaller pool of potential suspects. Switching the default rule would rid this group of the hold out option. Only if the majority of the population opted out of the database, it would be safe for these people to also have their data deleted. However, such a large opt-out rate is unlikely given the psychological phenomena described above.
Another obstacle to joining a voluntary DNA database could be the stigma attached to be in the "database of criminals". This effect is likely to disappear after switching the default for two reasons. First, the suggestive power of the new default will give database participation a better social reputation. Second, the increased participation in the database will make it hard to see the participants as high-potential criminals. A stigma of having something to hide will rather be attached to people opting out of the database while people that are part of the database make a contribution to overall safety.
Finally, many people might feel reluctant to join a database because of the risk of being erroneously suspected of a crime. There is a fear that one might find oneself entangled in investigation because of a mistake in the handling of the DNA. However, this risk assessment of people might be inaccurate. British authorities estimate that the likelihood of a match occurring at random is only one in thirty-seven million (Willing, Richard, Mismatch Calls DNA Tests Into Question, USA Today (Feb. 8, 2000)). Furthermore, there should be benefits offsetting the increase in the risk of error: Current false arrests and convictions could often be prevented if there was genetic evidence to exonerate the suspect. There are often cognitive errors involved in people’s assessment of the risk of error. Rare cases of false indictment based on DNA material will be more salient, triggering the availability heuristic (Tversky, Amos / Kahneman, Daniel, Judgment under Uncertainty: Heuristics and Biases, Introduction, 11-14). Also, people frequently have difficulties comparing risks and benefits of a new technology (Slovic, Paul, in: Tversky, Amos / Kahneman, Daniel, Judgment under Uncertainty: Heuristics and Biases, Chapter 5, 129-131). Setting the default rule on general participation in a DNA database can help overcome such judgment errors by balancing it with the stickiness of the default that works in the opposite direction.
As shown above, employing libertarian paternalism in the context of a national DNA database has numerous positive effects. Several aspects of bounded rationality, such as the status quo bias, suggestion, inertia, endowment effect, and ill-informed preferences, will make the new default rule sticky resulting in higher participation rates that allow the desired network effects in the reduction of crime. Switching the default to mandatory registration with the right of the individual to opt-out, is likely to solve collective action problems connected to voluntary participation, such as free-riding and hold-outs. In the face of the above collective action problems, it seems to leave society better off to switch the default rule automatic membership in the DNA database. Furthermore, the negative stigma of being part of such a database will be turned to a positive contribution to crime reduction. Moreover, people’s frequently ill-informed assessment of the risk of error can be balanced out to some extend by switching the default rule.
The above analysis does not, however, answer two questions:
First, the desirability of nationwide DNA databases to fight crime needs to assessed in a cost-benefit analysis based on empirical data. It is well imaginable that including the large law abiding majority of citizens in such a database will cause enormous costs of acquiring and administering the genetic material that will outweigh the benefits of increased conviction and deterrence (For a positive assessment of a population-wide database see: Kaye/Smith, 2003, 450-459).
Second, the introduction of a national DNA database can be challenged on constitutional grounds. The protection against unreasonable searches and seizures in the 4th amendment can pose a problem (Kaye/Smith, 2003, 442-446). Furthermore, freedom from self-incrimination protected by the 5th amendment could be infringed (Betzel, Margaret, Biometrics: Privacy Year in Review: Recent Changes in the Law of Biometrics, 1 ISJLP 517, 540). These questions need serious examination. However, by establishing a nationwide DNA database using the concept of libertarian paternalism, as argued by this paper, the constitutional problems should rather be diminished that increased: The non-discriminatory sampling of genetic material at birth and the right to opt-out respect the rights of the individual as much as possible.