Motivated Legal Bias

The probability of being sentenced to death is much greater if a defendant kills a white or Hispanic victim who is married with a clean criminal record and a college degree, as opposed to a black or Asian victim who is single with a prior criminal record and no college degree. …

“Irrelevant social facts also shape the ultimate state sanction” Phillips says. “In the capital of capital punishment, death is more apt to be sought and imposed on behalf of high status victims. Some victims matter more than others.”

Phillips research is based on 504 death penalty cases that occurred in Harris County, Texas between 1992 and 1999.  Drawing on the same data, Phillips’s previous research demonstrated that black defendants were more likely to be sentenced to death than white defendants in Houston. The racial disparities revealed in the prior paper become even more acute after accounting for victim social status – black defendants were more apt to be sentenced to death despite being less apt to kill high status victims.

More here (HT naz). I expect such patterns to be found in most legal jurisdictions, not just Harris County Texas.  You will find it hard to find any lawyer, judge, or law professor who will go on the record saying these are officially accepted as legitimate considerations in legal sentencing.  Most will say the law “tries” to ignore such considerations.  And yet such patterns have long existed, have long been widely known to exist.

These are motivated biases, not just random accidents of a system trying to be fair but failing to because of limited human mental capacity.  These errors are far more likely to persist than the opposite error.  If the opposite errors were suddenly to become common, enormous concern would be expressed, great resources would be spent, and we’d be willing to consider large institutional changes to eliminate them.

The place such errors enter is of course via “judgment.”  We recoil in horror at the thought of a simple legal system where judges or juries could make any decision they wanted in each case they considered.  But we also recoil at the thought of a legal system with explicit rules which had to be followed exactly in each case.  We instead prefer a legal system with lots of specific rules, where in the end “reasonable” people are allowed to exercise “judgment” about how to “interpret” the rules.   It sure looks like what we want is the appearance of constraining ourselves to follow rules, combined with the practice of arbitrary choice.

GD Star Rating
loading...
Tagged as: , ,
Trackback URL:
  • LetUsHavePeace

    Did the study include the guilty plea bargains made during this period for manslaughter and other killing crimes where the death penalty is not a potential sentence? If it had, you would find that the prosecutions themselves were skewed in favor of the “privileged” classes of victims. Prosecutors have limited resources, and they have to use them as best they can for their central goal – getting re-elected. Which is more likely to help in that goal – a prosecution where the victim was not even registered to vote or one where the victim not only voted but also contributed to political parties and candidates? You are right to see a bias here, but it is likely to be the one in favor of black defendants. Since their victims are overwhelmingly black and poor, the black defendants are more likely to be allowed to plead out rather forced to face the death penalty.
    There is another factor to be considered, and this one does confirm your belief about bias. But the bias is about familiarity, not race. Juries have a clear and strong bias against defendants who kill strangers. Everyone’s fear is a random killing, even though most actual murders involved people who knew one another. White and Hispanic capital crime defendants are not being prosecuted for killing black victims but for killing spouses and relatives and business partners. That goes in their favor and against the black defendants who kill white and brown people they do not know.

  • Phil

    If the victim had a criminal record, it’s more likely that the killing was a result of a criminal dispute; e.g., a drug deal gone bad. In that case, the accused might have a plausible story of how the killing was at least partially mitigated by circumstances.

    In that case, the crime is less “outrageous,” so isn’t it possibly just that the death penalty is less likely then?

    • http://entitledtoanopinion.wordpress.com TGGP

      Commenter Sal Uryasev makes a similar point at the link. I don’t know how many of you watch “The First 48” on A&E, but that’s commonly the case on homicides there. Peter Moskos of “Cop on the Beat” said that pretty much all the homicides he came across as a Baltimore police were linked to the drug trade. I should expect then that the GENDER of the victim would be important, as females are much less associated with such behavior (which is perhaps why Amy Bishop was able to get away with some bizarre behavior for so long).

  • JGWeissman

    “It sure looks like what we want is the appearance of constraining ourselves to follow rules, combined with the practice of arbitrary choice.”

    I don’t dispute that that is what we have. Did I miss the part where you show that we have what we want?

  • http://www.johnson.cornell.edu/faculty/profiles/bl Robert Bloomfield

    You will find it hard to find any lawyer, judge, or law professor who will go on the record saying these are officially accepted as legitimate considerations in legal sentencing.

    By these, it seems that you include education and marriage. I doubt anyone would support race, and I don’t seem much institutional support for race-based sentencing. But in the case of education, employment and marriage, isn’t it standard for sentencing hearings to include statements from the victim’s survivors? If the murdered person has 100 people who say “we depended on them”, isn’t a more severe sentence appropriate? Doesn’t our court system in fact incorporate that notion explicitly?

  • Eric Falkenstein

    I think LetUsHavePeace makes a profound point. When lower class people kill upper class people, it’s more random. People think random crimes are worse than raping or killing your brother or friend. Black and white people have different distributions among the upper and lower class. A classic omitted variables bias.

  • tom

    I am not sure you are serious here. Are you really accepting the premise of the article so uncritically? Do you know that there is a body of these studies specifically designed as Supreme Court bait to encourage 8th Amendment Constitutional claims? Don’t you think that there are other explanations for the different results?

    This is one area where I think you are naively loving too much your idea that we create signals of things like fairness.

  • mike

    “But we also recoil at the thought of a legal system with explicit rules which had to be followed exactly in each case. We instead prefer a legal system with lots of specific rules, where in the end “reasonable” people are allowed to exercise “judgment” about how to “interpret” the rules.”

    The problem with explicit rules is that they are brittle and complex. A good comparison is software, where explicit rules are executed by unbiased silicon.

    Such software is subject to vulnerabilities, where a gap exisits between the spirit and letter of the rules it was programmed with. The exploitation of such a gap can undermine and subvert the process, and sometimes even the entire system.

    Software also often fails completely when placed in environments and situations that were not fully expected by those that created the rules.

    If we are ever able to develop software that did not have those properties, and apply those lessons to the legal system, then I think there would be more willingness to remove judges that are allowed to ‘interpret’ the rules.

  • http://t-a-w.blogspot.com/ Tomasz Wegrzanowski

    The most obvious fix of mandatory sentencing moves control from judges, juries, and lawyers to politicians, and it turned out to work far worse and be even more biased.

  • michael vassar

    How would it be possible, in practice, to *not* leave choices up to judges, juries, or some equivalent. You can always claim that there are rules, but someone always has to actually execute them, or not, as they see fit.

  • Joe
  • A.S.

    We have a legal apparatus that generally favors wealthy defendants. How motivated is this legal bias? Let’s assume that an individual can make social contributions both outside the criminal justice system (in the form of labor, consumption…) and inside the CJS (in the form of stimulating employment…). Then, those who “contribute” less outside the CJS than inside it could be expected to be biased against. I wonder if there are any numbers to suggest that the likelihood of entering/escaping the CJS pivots around the level of contributions that one could make while imprisoned.

  • Tyrrell McAllister

    It sure looks like what we want is the appearance of constraining ourselves to follow rules, combined with the practice of arbitrary choice.

    It would look more like that if anyone could come up with a purely rule-based way to decide legal cases that did not have obvious and undeniable problems that made it blatantly unacceptable to everyone. I’m referring to problems that are far more undeniable than racial bias, such as frequently jailing people universally acknowledged to be innocent of wrong-doing.

    Since no one has been able to do that, the hypothesis remains plausible that our legal system is approximately as rule-based as we are able to make it without failing in such manifest ways.

  • tndal

    So it is a more serious crime to kill someone who is socially well-adapted than someone who is not?

    Is this not OK? Is it not absolutely normal and good? What else would one expect of a group’s behavior? I fail to see any problem here.

  • Philo

    The most striking sentence in the post: “It sure looks like what we want is the appearance of constraining ourselves to follow rules, combined with the practice of arbitrary choice.” But this is only half right. We certainly want to appear to be following rules but, in fact, we actually do want to be following rules. We don’t want, and we don’t have, “arbitrary choice.” We just don’t want to articulate the rules that we are using, and that we want to use. Partly, they are too complicated; partly we don’t want to see them articulated because we don’t want to have to acknowledge them.

  • http://osteele.com Oliver Steele

    Commentator Philo writes “[W]e actually do want to be following rules. […] We just don’t want to articulate the rules that we are using, and that we want to use. Partly, they are too complicated; partly we don’t want to see them articulated because we don’t want to have to acknowledge them.” The first case being that we might not have declarative access to the rules that we trust to make judgements (just as most language users don’t have access to the rules of grammar); the second being that there might be at least two different systems that both have to be satisfied.

    You could explode the second category into two subcategories depending on whether the contradictory systems are within a person, or within a society (i.e. on how you distribute “want” over “we”). Each member of a society individually might not want to acknowledge the rules, or the society as a whole might not want to acknowledge because this would alert some members that their system isn’t being used.

    In this second subcategory, where different constituents want different rules, attempting to formalize them would surface (and necessitate a response to) the conflict. Formalizing them prior to legislation might prevent us from having a body of law. Recovering the implicit rules from the behavior, via statistical analysis, avoids putting this resolution on the critical path towards creating a legal system, and makes it easier to avoid the conflict — since then resolving the conflict isn’t any critical path; and since the recovered rules don’t have the same ontological status (it’s not clear whether the methods of any particular study are valid, as some of the comments illustrate); and since they don’t have the same epistemic status (many people don’t believe anything that touches statistics, as evinced by the arguments about statistical sampling for the census, or about the evidence about climate change).

    Maybe these reasons apply to the intrapersonal case of “we don’t want to have to acknowledge them” too.

    • http://rationalmechanisms.com DWCrmcm

      Sounds like you are suggesting that a higher status will kill a lower status to save face, and by extrapolation juries will empathize and act accordingly.

  • http://rationalmechanisms.com DWCrmcm

    My bias alarm went off when I realized the it was a University of Denver analysis of Texas legal phenomena, and the article linked back only to University of Denver website.
    Let us assume indifference to the above for a simpler context. Let us assume further that the victims’ status juxtaposed with the convicts’ status is the prime factor. In model-speak –

    [victim status l:l convict status].

    If this juxtaposition precipitates one’s ambivalence, then, with your permission, I submit that our model asserts you are one in search of the attributes of greater truth. Justice should be such an undertaking.

    Is it bias when we allow high status citizens greater behavioral discretion, than low status citizens? Is it bias when we allow juries discretion?

    [famous actor murder conviction l:l famous obstetrician murder conviction]

    Status is bias; that is its value.

    Maybe we need to focus more on a question of when to overcome bias.

  • mjohns2
  • Jeff

    If the opposite errors were suddenly to become common, enormous concern would be expressed, great resources would be spent, and we’d be willing to consider large institutional changes to eliminate them.

    How do you know that?

  • http://kanBARoo.blogspot.com Stephen R. Diamond

    We instead prefer a legal system with lots of specific rules, where in the end “reasonable” people are allowed to exercise “judgment” about how to “interpret” the rules. It sure looks like what we want is the appearance of constraining ourselves to follow rules, combined with the practice of arbitrary choice.

    Why is your inference that we seek merely the constraint’s appearance more plausible than the straightforward alternative: we most prefer mild constraints? Telling against yours— imperfections of constraint don’t prove its absence.

  • Pingback: Law, Power, and Status « azmytheconomics

  • Pingback: Capital Punishment (26): The Probability of Capital Punishment in the U.S., by Race, Ctd. | Cosmologically Insignificant