Sunday, April 14, 2013

High Money Bail and Prediction

In a recent story out of Aspen, we have what the headline reads as a “huge bond reduction,” leaving everyone scratching their heads why.

The article states: 

“In a bizarre twist Tuesday, a local man accused of a brutal attack on his former girlfriend had his bond reduced from $400,000 to $2,500.

“Peter Nardi, 50, who lives on Highway 82 near Smith Hill Way, was expected to be released from jail Tuesday afternoon, said his attorney, Arnold Mordkin of Snowmass Village.

“Chief Deputy District Attorney Andrea Bryan was mum on why she agreed to the vast reduction. She told Judge Gail Nichols of Pitkin County District Court that she did not want to state her reason for agreeing to the lower amount for fear of jeopardizing an “extensive, ongoing investigation.”

“But Bryan said the same felony charges remain pending against Nardi, a former restaurateur in Snowmass Village.

The only thing everyone does know, it seems, is that $400,000 kept the defendant in jail, but $2,500 allowed him out. Which begs the question – did the judge mean for him to be initially detained pretrial? If so, the first high bond amount raises constitutional issues not only concerning the judge’s reason (setting bail to detain in a right-to-bail state like Colorado is constitutionally suspect) as well as to its excessiveness under the 8th Amendment to the U.S. Constitution.  The judge apparently mentioned the factual allegations as a basis for the initial amount, but the problem with basing financial conditions of bond on the charge or the allegations is that it starts to resemble punishment for something the government has yet to prove.    

Given that the only two constitutionally permissible reasons for limiting pretrial freedom are public safety and court appearance, it would have been nice if the judge had said something about them. More specifically, given that no amount of money in Colorado can be forfeited due to breaches in public safety, it would have been nice and actually more appropriate for the judge to explain exactly how the money was necessary to address risk of flight. That’s really the only reason why you should use money in Colorado. Money has nothing to do with public safety. It can’t be forfeited for new crimes or other breaches in public safety. And the research shows that money doesn’t make anyone safer. Using money to make people safe when it doesn’t work and you can’t even forfeit it for new crimes is irrational, which means it’s probably unlawful no matter which constitution you are looking at.   

In the story, the defense attorney said, “It’s clear to me that there was something that was originally [in the investigation] that is no longer there – I assume it is something to do with the woman’s story and that something significant has changed.” So what could have changed in the woman’s story that would have altered the defendant’s risk for flight when the charge remains the same? Nothing, most likely, even though that’s how bail is supposed to work. One of the toughest applications of the presumption of innocence at bail is for judges to note the charge and an affidavit (usually a sad tale of events that always supports the charge), but to look beyond those things to evaluate the only two constitutionally valid reasons for limiting pretrial freedom.

And here’s the kicker: Colorado has recently developed an empirical risk assessment instrument designed to determine which factors actually predict pretrial failure based on these two constitutional reasons. In the past, present charge and affidavit would have guided most decisions, as it perhaps still did in this Aspen case. But guess what factors are actually predictive of pretrial failure? There are twelve: (1) having a home or cell phone; (2) owning or renting one’s residence; (3) contributing to residential payments, (4) past or current problems with alcohol; (5) past or current mental health treatment; (6) age at first arrest; (7) past jail sentence; (8) past prison sentence; (9) having active warrants; (10) having other pending cases; (11) currently on supervision; (12) history of revoked bond or supervision. That’s right, no “current charge at arrest” or “what the police said he did in the affidavit so that a judge can find probable cause.” Not having the current charge as predictive of pretrial failure is so counterintuitive to our current thinking that that someone actually told the State Crime Commission recently to add it to the assessment tool, “even though it would make the tool less predictive.” Of course, the group decided not to do so. Adding a non-predictive element, no matter how logical it may seem, to a predictive tool to make us feel better would be no different than adding “type of shoes worn at arrest.” It would be worthless and irrational.      

Looking at this case from a distance, time will tell if this defendant was the sort of person who was so likely to flee that only $400,000 (an amount virtually assuring his detention) was the only solution. If not, then there must have been some other reason for the initial high financial condition. Let’s hope it wasn’t just some judge’s subjective opinion about the seriousness of the charge. That’s not on the predictive list.