If it seems like I've written this column before, it's probably because I have. The last time was in June 2000, when a reckless teenager in a stolen car, fleeing a Jefferson County sheriff's deputy at speeds in excess of 100 mph, crashed into a motorcycle and killed two bystanders in Lakewood.

Although there are as many as 200 serious chases involving the Denver Police Department in a typical year, and while the greatest number of injuries are to fleeing suspects, whenever innocent bystanders are hurt or killed, invariably there are impassioned cries for a ban on such pursuits.

Last week, Edith Mack, a tutor at Molholm Elementary School in Jefferson County, was injured when a Denver police detective in an unmarked car ran into her vehicle while pursuing a robbery suspect. Manuel Santistevan, the suspect, was a fugitive believed to be armed and dangerous and a member of the "Hooded Safe Bandits" gang, connected to a spree of armed robberies.

In this case, the chase appears to have been technically in violation of police department policy, which requires pursuit vehicles to be equipped with emergency lights and siren. This unmarked car wasn't so equipped. The brunt of the criticism, however, has been directed at the general practice of police chases. And that's where I disagree with the critics.

Denver police have a stringent policy governing police pursuits that takes pains to restrict, monitor and supervise such chases, minimize risk and balance competing considerations of public safety.

This is a classic example of a public policy trade-off. Those who argue for a ban on police chases - claiming that even if it saved only one life it would be worth it - are only considering part of the picture. Yes, police chases are inherently dangerous, but a flat-out prohibition on pursuits would also carry costs and risks. By allowing a fugitive to escape, we're leaving someone at large who might commit additional crimes.

The victims of those crimes, including rape and murder conceivably, wouldn't have been victims at all if the fugitive had been apprehended as the result of a successful police pursuit.

If the risk of injury to bystanders from high-speed emergency runs is simply intolerable, then why stop at police chases? By that reasoning, we should ban high-speed, flashing-light-and-siren runs for any reason, including assistance calls. And that would apply to ambulances and fire trucks, as well.

Moreover, consider the message that would be delivered to criminals if it were official police policy that those who speed off when a police officer attempts to stop them would not be pursued. It would encourage and reward flight, especially among those with a reason to avoid arrest.

Or try this scenario: A police car signals a drunk driver to pull over. Aware of the no-pursuit policy, rather than stop, be arrested and lose his license (let's say he's a multiple DUI offender), he takes off. Escaping arrest, he abandons his car, takes a cab home, sobers up, reports the car stolen and denies any involvement. From the time that drunk pulls away from the cops, you're placing the public in danger.

Hey, I'm not suicidal. I understand I could be the innocent bystander killed as the unintended result of a police chase. But I believe Denver police have a prudent and effective pursuit policy.

One measure of which is that bystander fatalities are rare in Denver - in some years none, in others one or two. With 2 1/2 million people in the metro area, I assess my practical risk - and yours - as extremely remote. So, rationally, I'll take that risk to secure the benefits, since it's my judgment that prohibiting the police from pursuing fugitives presents a much more tangible danger to public safety.

Mike Rosen's radio show airs daily from 9 a.m. to noon on 850 KOA. He can be reached by e-mail at mikerosen@850koa.com.