The exoneration of Lynn DeJac, reported in the New York Times, brings good news and bad news.  First the good:  For the first time, DNA evidence has exonerated Lynn for the murder of her 13 year old daughter, Crystallynn.  She was convicted in 1994, but her attorney, Andrew C. LoTempio, never gave up.  This raised the number of exonerations to 209, according to the Innocence Project.

Some murder accusations are worse than others.  While Lyn DeJac was not, according to the story, a model mother by any stretch, being accused (not to mention convicted) of murdering her own daughter is about as bad as it gets.  The trial consisted largely of evidence by Lynn’s neighbors about what a lousy human being she was and how much they didn’t like her.  I’m sure the judge allowed it as “background” or some less than cogent Molineux (other crimes) theory.  (Aside: Yet another opportunity to opine about how Molineux, which was held to serve as a shield to protect defendants from evidence to show propensity to commit crimes, is used routinely as a roadmap to circumvent its fundamental purposes by judges who just refuse to comprehend the basic rule).

So bad girl Lynn DeJac became the cops immediate target.  The closest family member is always the person who committed the murder.  Just ask any cop, they’ll tell you.  This is where  inductive reasoning comes in. 

Normal people assume that the police first gather the evidence, sift through it, investigate it and then, upon the completion of these tasks, identify the person they believe committed the crime.  First they have evidence.  Then they have a target.  But that’s merely what normal people think.  Not cops.

Cops decide who the perp is.  They then collect evidence to prove themselves right.  Evidence that disproves their theory is rejected, explained away or just deep-sixed.  After all, it’s bad evidence since they already know who did the deed.  They ignore leads that take them in a direction away from the perp, because that’s just a waste of their precious time.  They embrace witnesses who support their theory, and convert witnesses who don’t to the truth. 

Why do cops know who the perp is before the evidence is gathered?  They just do.  They’re cops.  They have a “sixth-sense,” street-wise experience, whatever phrase newspaper writers like to attribute to the flea-bitten detective who can small a criminal half-way down the block.  It’s a very romantic notion. 

Other times, cops just go with the odds, as courts invariably do.  If a wife is murdered, 90% of the time it’s the husband.  Therefore, arrest the husband and you’re right 90% of the time.  What about the other 10%?  Well, it’s not a perfect system, you know. 

It’s no different for the judge, except the odds aren’t as clear.  Let’s say 97% of all defendants are guilty, at least of something.  So if a judge finds the police testimony credible by default, then he’s right 97% of the time.  That’s a pretty good statistic, and will keep him off the front page of the New York Post as the Worst Judge in New York.  But don’t bother asking any judges about this.  Most are in denial. 

So a case as built against Lynn DeJac, heavy drinker and really bad neighbor.  No doubt it was based on “compelling evidence,” as prosecutors love to say.  Sure, it’s meaningless, but that phrase always ends up in the papers and inexplicably makes reporters and readers alike think it must be true.  The only problem is that she didn’t do it. 

Now for the bad news.  Lynn DeJac’s boyfriend, Dennis P. Donahue, was the murderer.  DNA is a two-way street, you see.  So why is this bad news?  Ah, there’s the rub.

In their zeal to convict the heinous Ms. DeJac (who cops just knew was the culprit), Donahue (who doesn’t get a “Mr.” in front of his name because he’s undeserving of such gratuitous formality) was called to testify against her in the grand jury.  Now, for those of you unfamiliar, while the feds give stinky little testimonial immunity for grand jury testimony, New York gives broad transactional immunity.  Since Donahue was called by the prosecution before the grand jury to testify, and was questioned about the murder, he gets to walk.  The prosecution gave Donahue a free ride on murder.

Before anyone comes up with the brainstorm that the problem is transactional immunity, bear in mind why Donahue gets a pass while DeJac gets hosed.  This, not immunity, is the lesson of this case:  The rush by police to decide who dunnit, and then the effort to collect evidence to support their induction.  It would strike any thinking person, even a jaded criminal defense lawyer, that the cops really ought to take a look at the potential killers, and consider the full significance of the potential evidence, before picking one and framing them.  But no, that’s just not effective law enforcement work.  We’re the police.  We know what we’re doing.  Trust us.

And as long as we’re on the subject, consider the importance of what the neighbors thought of Lynn DeJac.  After all, being some drunken party-girl is almost the same thing as being a murderer, right?  The taint of allowing evidence to come before a jury with no real purpose other than to besmirch a defendant’s character is strong stuff.  People love to hate people who others love to hate.  Show us some bad character and we’ll happily attribute all sorts of evil to them.  A little two dimensional, perhaps, but we’re too busy to think hard about things, so a rush to judgment (like we do in our everyday lives, as the jury instruction goes) makes for swift justice.

Not quite the way you want the system to work, you say?  Then judges, give it some thought before you bend over backwards to find a Molineux exception to let irrelevant bad acts evidence in through the back door to smear a defendant.  It wouldn’t kill you to make the prosecution actually prove it’s case directly, now would it?

Welcome back the world, number 209.