Are you a drunken driver after you stop driving?

I can't be the only one who thinks they should bring back the original "Law & Order." If NBC did, the show's first case should be one that went on trial this week in a local court in Mineola, New York. A man has been charged with homicide in the death of a police officer who was hit by an SUV.

The twist is that, when the crash happened, the defendant was leaning against the guardrail. He had been driving home from a night of drinking, got involved in a minor accident and was pulled over. The policeman was hit by a different car while investigating the crash. Is the homicide charge justified? Ask Jack McCoy -- or really, if you want to be a hard-core "L&O" fan (and I am), ask Ben Stone.

This case is begging to be ripped from the headlines by TV writers. In real life, the Nassau County district attorney filed charges against James Ryan in the death of police officer Joseph Olivieri. A New York state appeals court had already blessed the charges.

Back in February 2015, the appellate court issued a brief ruling allowing the charges in the October 2012 crash to go forward. The court explained that, according to the grand jury, Ryan "caused two collisions" while driving under the influence of alcohol. Then, "in the immediate aftermath of those collisions, the defendant's stopped vehicle was in the eastbound High Occupancy Vehicle lane, facing perpendicular to the direction of traffic." Officer Olivieri, who had gotten out of his vehicle to investigate, "was struck and killed when the driver of a sport utility vehicle traveling in the HOV lane did not see him or the defendant's stopped car in time to avoid hitting them."

The legal question comes down to whether Olivieri's death was a foreseeable consequence of Ryan's criminal act of driving under the influence. Between five and 10 minutes elapsed from the time when Ryan stopped after causing two relatively minor collisions to the time when the passing SUV hit Olivieri.

The New York trial court initially dismissed the charges against Ryan in Olivieri's death. It said the accident that killed the officer wasn't part of a "continuing chain of events" set in motion by Ryan.

But the appeals court reversed that decision. It said that, under New York law, it was enough for Ryan's act to have "contributed" to Olivieri's death. "The defendant's actions," the court added, "need not be the sole cause of death and, indeed, the defendant need not have committed the fatal act to be liable."

The only relevant legal question, according to the appellate court, was "whether it may be reasonably foreseen that the defendant's actions would result in the victim's death."

As a statement of New York law, the appellate court's ruling can't be contradicted. Claire Kincaid (a graduate of Harvard Law School, where I teach) couldn't have said it better. But the legal rule should also take into account the chain of causation between the defendant's act and the death of the officer. Foreseeability is a very loose concept. It's all well and good to say that a reasonable person foresees the probable consequences of his actions. But reciting this legal formula doesn't tell you very much about the real world.

When a court finds reasonable foreseeability, what it really means is that it's making a policy judgment to assign liability. Would a person driving under the influence reasonably foresee that he would be pulled over after a fender bender? Well, sure. Would he reasonably foresee that a passing car might hit the police officer investigating the accident? That's hardly a question susceptible to a "reasonable" answer. It could happen. Then again, most times it wouldn't. Who's to say what counts as foreseeability?

The New York appellate court was in fact assigning responsibility to a driver under the influence for any bad consequences that might follow. At this moment in history, we treat drunken driving as a moral wrong of the highest order, so it's easy to understand what the court was thinking. You take a huge risk if you drive while impaired, according to the court. Whatever consequences flow from that act, expect to be held criminally liable.

It seems to me that the New York appellate court got it wrong, and that the Nassau County prosecutors shouldn't have brought charges. Drunken driving is a bad act, to be sure. It's a horrible tragedy that a police officer like Olivieri should've paid the ultimate price while investigating an accident caused by a driver who had too much to drink.

But the truth is that Olivieri could've been in the exact same situation had the underlying accident been the result of random chance or of poor driving by completely sober drivers. The cars were aligned in a certain configuration. The SUV driver didn't read the situation correctly. None of this happened particularly because the underlying accident resulted from alcohol.

There's no doubt that Ryan contributed to Olivieri's death. But he would've contributed to the same extent if the crash hadn't derived from alcohol. For this reason, it was a mistake for the New York courts to hold that Ryan could be criminally liable for the police officer's death.

In minute 58 of "Law & Order," Ryan would reach a plea bargain with the police. He would get a punishment worse than if he'd simply driven drunk, but not as extreme as he would've if held responsible for Olivieri's death. The trial's begun, but there's still time for life to imitate art.

Comment by clicking here.

Noah Feldman, a Bloomberg View columnist, is a professor of constitutional and international law at Harvard University and the author of six books, most recently "Cool War: The Future of Global Competition."