Credibility Proxies: Final Thoughts

In this series of posts I’ve outlined a number of credibility proxies currently in use in the justice system.  In order to assist them in the search for truth, fact-finders may learn of witnesses’ prior acts from shoplifting to prostitution to the possession of illegal drugs.  At the same time, they are often not told that a witness has violence in his or her past.  

Courts often offer scant explanation for why certain crimes are or are not related to credibility.  They suggest that credibility is related to deliberate violation of basic standards, to scheming, to taking for oneself at the expense of others.  They contrast these notions with acting on provocation or out of passion, which they posit has little to do with a person’s credibility. 

The series of posts has tried to find coherence in the doctrine, but what stands out is that these explanations do not add up to much.  The courts have maintained a status quo in which many crimes are admissible on the question of credibility but in which precedent has far more power in predicting which crimes will be admissible than does principle. 

Accepting, for the moment, that proxies are useful in assisting the fact-finder to make accurate credibility determinations, it is hard to justify informing the fact-finder about prostitution but not about assault.  Similarly, having serious traffic violations in ones’ past seems like it might be just as relevant to truthfulness as jumping a subway turnstile.  Also perplexing is the idea that a jury will glean something informative about the defendant’s credibility from being told that he or she committed a crime in the x degree without any of the context, such as information about the circumstances or even the type of crime.  There is no data that I can find to explain or justify this type of line drawing.  Nor does intuition about what seems more relevant to credibility get me very far in an effort to understand it.

 What might best explain the contours of credibility proxy doctrine is history.  As I have explored in other work on moral turpitude as a legal standard, the shape of that standard had to do with beliefs about moral rectitude that were widely held in the early nineteenth century. The honor code at the time when our evidentiary rules were first becoming entrenched condemned deceptive business practices and dishonesty in men and lack of chastity in women. Violent behavior, in contrast, was treated more forgivingly. Credibility proxy doctrine evolved in tandem with the moral turpitude standard (which itself is still in place in the evidentiary codes of California and Texas).  Both modern moral turpitude jurisprudence and modern credibility proxies to a large extent mimic a nineteenth-century system of values. 

Using proxies that don’t correlate with data or common sense distinctions seems to raise a red flag about what we are doing with credibility proxies.  As others have pointed out, admitting prior felonies has a disproportionate impact on minority communities while not necessarily offering much useful information, given the frequency with which defendants plead guilty to charges arrived at through negotiation rather than because they best align with the underlying conduct (see an excellent recent article by Anna Roberts for more on this).  While women are far more likely to have prostitution convictions, men are more likely to have been convicted of violent crimes (or any crime, for that matter).  If we aren’t bringing ourselves closer to the truth with these proxies, then are we instead amplifying and entrenching certain social hierarchies? 

One solution to this quagmire is to eliminate the line-drawing altogether and use something akin to Minnesota’s “whole person” doctrine.  If anything is relevant to credibility because the fact-finder needs to see the whole person, then the only question is one of balance.  Would the information would be more probative than prejudicial?  Where do we want the scales to tip?  Of course, placing total discression in the hands of the trial judge has perils of its own, not to mention opening the door to a limitless store of prior acts that it could – and no doubt would – be argued are relevant to impeachment.

More radical, but possibly more effective if we believe that lies are context-specific (and therefore prior lies have little predictive power in new situations), is the idea of eliminating credibility proxies altogether.  We could continue to use evidence of bias and prior inconsistent statements to impeach but preclude the possibility of introducing other credibility-related evidence. This would change a number of incentives, removing a major impediment to defendants testifying in their own defense while also creating a more efficient trial practice by eliminating a whole category of evidence and a basis for numerous appeals.

-JSK