Kansas v. Glover and Conditional Irrelevance

Posted in: Criminal Procedure

In 2001, in the North Carolina Law Review, I published an article that identified the concept of “conditional irrelevance.” An evidentiary exhibit is conditionally irrelevant when the presence of some fact reduces or eliminates the relevance of that evidence. The concept is useful for understanding the arguments before the Supreme Court in Kansas v. Glover. In this column, I will elaborate my theory of conditional irrelevance (only 19 years after the fact) and then explain the utility of conditional irrelevance in understanding what divides petitioner from respondent in Glover.

Conditional Irrelevance

To understand conditional irrelevance, it is useful to first consider the existing concept of conditional relevance, described in Federal Rule of Evidence 104(b) (“Relevance that Depends on a Fact”). When a piece of evidence is relevant only if some preliminary fact is true, then we say that the evidence is conditionally relevant, with the truth of the preliminary fact serving as the condition. For instance, say you are on trial for murder, and you want to put on an expert to talk about the impact on aggressiveness of having the low activity form of the MAOA gene. The expert’s testimony is relevant only if you have the indicated form of the gene in question. You can perhaps argue, then, that your gene made you do it and that you are warlike by nature and could not control yourself. Assuming there is a legal defense for this situation, it would have something to offer you only if you had this warlike gene. Otherwise, the MAOA evidence has no bearing on your guilt.

In the case of conditional irrelevance, the reasoning is similar. Some evidentiary exhibit is highly relevant (or at least somewhat relevant) to a case. Then another fact comes to light, and the original evidence loses some or all of its relevance. One illustration would be evidence that a particular vehicle was moving faster than the speed limit from the direction of a hit and run. That would look like incriminating evidence against the driver of the speeding vehicle. But if we learned that the vehicle at issue was an ambulance, then the apparent relevance of the evidence would vanish.

More controversially, I argued in my North Carolina piece that a woman’s sexual history is irrelevant to whether she consented to a man that she claims raped her. The rape shield law generally excludes such evidence, but most of the law’s defenders cite the policy harm and prejudicial impact of sexual history evidence rather than arguing irrelevance. (The policy harm is the shockingly low reporting rates for sexual assault relative to other violent crimes, and the prejudicial impact is the juror or jurors who hear about a victim’s active sex life and think “she’s a slut who probably asked for it so let’s acquit.”).

In my article, I proposed that sexual history is ordinarily relevant to whether a person decided to sleep with someone on a given occasion. Imagine, for instance, that your neighbor is trying to put odds on whether you slept with anyone exactly two weeks ago. If you are someone who tends to sleep with a lot of people, that fact makes it more likely that your neighbor should bet “yes.” People do tend to be creatures of habit and, to some degree, creatures of character and inclination as well. So what happens at a rape trial? Please note that I will here speak of the victim as female and of the perpetrator as male, not because I harbor the belief that the sexes always align in this way, but out of a recognition that they ordinarily do and that it is easier and makes for clearer writing to assume that they do for purposes of a column (rather than offering countless “if he or she rapes him or her” and other such inelegant in-the-alternative statements).

What happens at a rape trial is that we have an additional piece of information, the fact that the victim has accused the defendant of rape. For those readers becoming angry, let me be clear. I am not treating the accusation as establishing the defendant’s guilt. I am suggesting instead that the fact of the accusation is itself important because it reduces to nearly zero the apparent relevance of promiscuity evidence. People who are promiscuous may be more likely to sleep with someone on any given day, but they are no more likely than a less sexually experienced person to invent a false accusation against an innocent man. Indeed, some have argued that because a promiscuous woman is more often in situations that make her vulnerable to assault, she is more likely to be raped and accordingly more likely to be telling the truth when she claims that she was raped. I reject this view along with its opposite. Once we know of the rape accusation, the relevant variable becomes credibility and not promiscuity.

Kansas v. Glover

In Glover, police learned that a car on the road belonged to a person whose license had been revoked. Based on this information, the police pulled over the driver, Charles Glover, who was subsequently charged with driving as a habitual violator. The question before the Court was whether knowing that the owner of a car has had his license suspended or revoked gives police reasonable suspicion to stop the vehicle. At argument, Glover’s side acknowledged that it is generally correct to conclude that a driver is steering her own car. If all you needed to know was whether the owner was driving his car, you would have reasonable suspicion upon seeing the car. Things change, however, when the owner’s license has been revoked or suspended. At that point, it may no longer be fair to assume that if you see a moving vehicle, the owner of that vehicle will be at the wheel.

The argument here is one of conditional irrelevance. Ordinarily, knowing that X’s car is on the road gives you reason to conclude that X is the person driving the car. You might therefore think that if X’s license is revoked, then when police see X’s car on the road, they have reasonable suspicion to believe that X is driving without a license. But the relevance of X’s car’s presence on the road to X’s presence at the wheel is conditioned on X being licensed to drive. The moment we learn that X does not currently have a valid license, the moving vehicle may no longer suggest the presence of the driver at the wheel. People without a license are much less likely to drive than people with one.

Basic Commonsense

None of this is rocket science. New facts can make an item of evidence seem relevant when it previously did not, and new facts can make an item of evidence appear irrelevant when it seemed relevant before. We know, for example, that a police officer can get a search warrant based on probable cause but then learn something during the lawful search that terminates probable cause and the authority to continue the search authorized in the warrant. My warrant says I can look for a missing child in your house, but other officers find the child elsewhere and radio me. Despite what my warrant says, I am no longer authorized to search your home.

However obvious, though, we tend not to think in terms of conditional irrelevance. If owners usually drive their own cars, then seeing a car on the road signals that the owner is likely at the wheel. And if the owner is not allowed to drive, then we have reasonable suspicion to stop the car. But oddly, the very incriminating fact, that the owner has no license at the moment, is the fact that reduces the validity of the original relevance proposition, that owners tend to drive their own cars, rendering reasonable suspicion far less clear. And yet there is something counterintuitive about that conclusion.

When I have suggested that an alleged victim’s promiscuity is virtually irrelevant to whether she consented to an alleged rapist, people have found that counterintuitive as well. How can one’s inclination to consent to sex be irrelevant to whether one consented to sex on a specific occasion? The answer, here too, is conditional irrelevance. Once we know that someone has made a rape accusation about a particular occasion, the fact that she is promiscuous no longer makes consensual sex more likely. The important dimension has moved from sexual experience to dishonesty. And likewise, the relevant dimension in Glover is, at least arguably, the inclination to violate the law by driving without a license. If one accepts this framing of the case, one that the advocates did not fully develop, then it would appear that owning a car would in no way bear on the likelihood of driving that car in the absence of a license.