A report from the Human Genetics Commission has proposed the former. But is the storage of a DNA profile any different from routine storage by the state of date of birth, gender, address or national insurance number which we generally accept?

It is not the same. For a start, the DNA database has helped to solve a lot of criminal cases. In relation to crime detection, civil libertarians should also remember that innocent people are sometimes imprisoned and that a number of them have been set free because of use of DNA profiles. A national DNA database could have saved some of those innocent people from long periods in custody so there can be benefits as well as costs to civil liberties of data being stored by the state.

The present system is seen as unfair because of the opportunistic way in which profiles are added to the DNA database and the fact that they stay there forever. The 4.5 million people whose profiles are on the database are there because either DNA was given voluntarily or they were arrested for a recordable offence. Only about two thirds have been found guilty of an offence but profiles of the innocent are not removed.

A database covering the whole population would be more equitable though also, in the words of one editorial on the Commission's report, "costly and illiberal". The advent of cheap digital storage and fast computers, however, means that the police are bound to use profiles and databases of different sorts. Most people do not object when gender and age are used to narrow down a range of suspects, but many of us would be uneasy if ethnicity, socio-economic group, employment record or choice of newspaper were routinely used to identify likely suspects.

Following the eventual capture of the Yorkshire Ripper almost 30 years ago, West Yorkshire police were criticised for not making better use of the data that they had available to them. Now, police are criticised for collecting and making use of too much data. In the Yorkshire Ripper case, use of a database would probably have speeded up capture of Peter Sutcliffe while a DNA database would almost certainly have saved some of the lives he took.

Therein lies the dilemma; do we face a choice between saving lives or losing liberties? Intelligent use of databases can help improve crime-detection rates enormously. A DNA database is currently the best of all for crime detection because it has such high predictive power and it can save the lives of potential victims of serial killers. So, if the police are going to use databases, why not choose the most accurate predictor of guilt and gradually develop a more complete national DNA database?

This could be done relatively quickly if dentists and doctors were involved in collecting DNA samples routinely, although important sections of the population might be missed (for example, people with poor teeth who are otherwise healthy). I do not seriously propose this but we should think hard about why we reject a practical means of bringing to justice more criminals, in particular, serial rapists and murderers, and preventing more crime.

This may reflect our personal judgment of risk. Hardly any of us expect our family or close friends to become victims of serial rapists or murderers. We may therefore perceive a minimal increase in personal benefit resulting from a national DNA database. In practice, there should be no personal cost either, except for those criminals who more likely to be caught, but there is a general concern about state intrusion in our lives that manifests itself around this issue.

The current situation has been described as a bad compromise. However, it seems to me to be pragmatic and unfair rather than bad. It may be unethical because certain sections of the population are over-represented compared with their involvement in crime, and this is unfair. It is hard to imagine, however, that the Home Office will readily give up any stored DNA profiles when it is both cheaper and more effective to retain them.

I suspect that this pragmatic approach will prevail and that
a universal database will develop, first by stealth and
in due course by public acceptance. In 20 years time, I doubt whether we will be agonising over the issue.


Professor Robert Smith is Dean at Huddersfield University's School of Applied Sciences.