There has generally always been a "good faith" admission system in law enforcement, and that's the underlying reason why the ruling on this case went the way it did. The database itself was not, in fact, in error - the error was by the clerk in reading it.

Whenever I run someone and they have a local (in-county) warrant, I can arrest immediately because my dispatcher is looking at our local database and reads the info right off the screen while looking a the judicial bench warrant.

When, however, I run someone and they have a warrant out of a different county or different state, I have to do a "warrant confirmation." That means our 24-hour clerk calls *their* 24-hour clerk, then that person manually looks up the info to confirm the warrant actually exists, then tells our 24-hour clerk, who then tells our dispatcher, who then tells me. This whole process must be completed in under 10 minutes or we have to let the person go.

In the case before the court, the outside agency 24-hour clerk pulled the warrant and saw it was valid, but failed to look on the next section where it showed the suspect had actually already appeared before the judge on it. Typically it only takes about 24 hours to remove a warrant from NCIC, so the manual system exists to cover the interim time. The clerk, in good faith, reported back that the warrant was valid.. The arrest was made and subsequent evidence was located based on the information from the outside agency. After the fact, the clerk discovered she had made the error, then called to relay it, but it was too late.

The ruling on the case hinges on the fact that no deliberate errors (actual police abuse) took place, but only what, if this were not a criminal matter, would be placed in the same purview as a typo in a business document (not my opinion - that's how the ruling is written out). Because it (the arrest and subsequent discoveries) was done in "good faith," it was determined not to be an abuse and left to stand. Had it been deliberate, I am certain they would have ruled the other way.

That said, I think that while good faith follows a letter of the law approach, I think the spirit of the law demands the evidence be suppressed, as Schneier suggests.

For a personal example, I sometimes have had difficulty reading the tiny numbers on our license plate registration tabs and, probably once a year, stop someone for expired tags because the curl on the 9 can look an awful lot like an 8. Most officers will continue the stop and run the driver for wants and warrants, etc. I tell them "I see I made a mistake, have a nice day" and never even identify them. If they are drunk, I will necessarily make them walk or get a ride, but otherwise I don't feel I had the right to invade further, even though the law allows it.

The thing is, there's enough of an issue with the public trust and law enforcement as it is without giving the public more cause to be distrustful, in my opinion.

It seems to me, after some thought, that the SCOTUS is right on the intent of the law.

But the point about the errors that you and Bruce make is also valid.

Thus lawyers will need to prove that no effort is being made to correct errors in the future. If a good faith effort IS occurring to get the data errors out, then...

We'll be back at SCOTUS to get a definition of "good faith effort".