One reason is that the database for England and Wales includes 8% of the population. (Scotland and Northern Ireland are in a separate database.) Everyone arrested, even for a minor offense, must give a sample, and the DNA profile stays in the database permanently. In the U.S., most jurisdictions test only convicted felons, although some states are moving to a broader requirement. The total database is about 2% of the population. Some in England want to broaden their database even further. "A senior appellate judge, as well as former prime minister Tony Blair, have called for the database to ultimately include DNA profiles of every U.K. resident."

Another way to get more results is to go beyond exact matching of samples in the database. When the exact-match search comes up empty, a search for partial matches turns up a list of people who might be close blood relatives. This list can be long, though, and refining the list can involve interviews casting suspicion on innocent people. The article tells of one investigation where interviewed family members reacted in horror, "Are you suggesting one of my relatives is a rapist?" Most of them weren't, but one was, and the authorities caught and convicted a rapist who had gotten away with it for over a decade.

A third advance is to analyze DNA in ways that the sample itself yields information about the perpetrator, such as racial group, hair color, or eye color. Until recently, a convincing answer to privacy objections was that DNA profiling only involved "junk" DNA, bits of genetic code unconnected to any known characteristics of the person. In that circumstance, DNA was like a fingerprint in that it was useful for identification but nothing else. When databases include more personal information, the possibility of misuse becomes a significant issue.

The number of crimes that goes unsolved is an atrocity. In California in 2005, for example, half of all homicides and 57% of forcible rapes were not cleared. Solving more of these crimes is an imperative. The privacy concerns that arise from expanding databases and from new techniques are legitimate and need to be addressed. We should not, however, go the extremes of banning the new techniques or unduly restricting the databases. We need to move forward with reasonable safeguards.