When confronted with a dispute about a statute, people will have policy
preferences about how the case should come out. They also will have
opinions about what the statute means just as a matter of English and
whether it’s ambiguous. The questions that interest me and my
co-authors (Ward Farnsworth and Dustin Guzior) involve the relationships
between those issues. Do policy preferences affect peoples’ judgments
about what a statute means and how clear it is? If so, is there any way
to reduce that influence?
We examined these issues by surveying over 1,000 law students at various schools. We showed them statutes and facts that the statutes might apply to; the problems were based on real Supreme Court cases. We asked our respondents what outcome they preferred as a matter of policy preference. Then we asked them to say whether they thought the statute was ambiguous, or what reading of the statute’s text they thought was best.
The first result is that peoples’ judgments about ambiguity and meaning are closely entwined with their policy preferences. People with strong preferences about how a case should come out tend to find that the statute at issue is unambiguous, and that it means what they want it to mean. This is true even when we instruct them to put the statute’s text aside when stating their preferences, and to put their preferences aside when stating their views about the text. They generally can’t do it.
But something interesting happens when we change the questions. Instead of asking our respondents whether they thought the statute was ambiguous, sometimes we asked them whether they thought ordinary readers would likely agree about the statute’s meaning. Or instead of asking them what they thought the statute meant as a matter of ordinary English, we asked them what they thought ordinary readers would think it meant. The answers were then remarkably different. They weren’t biased. When saying what ordinary readers would think, the respondents, as a group, were able to give answers that were independent of their preferences about how the case should come out.
Here’s an example. In Chapman v. United States, a statute applied a harsh sentence to anyone who distributed more than one gram of a "mixture or substance containing a detectable amount” of LSD. The question was whether, in sentencing the defendant, the court should weigh the blotter paper onto which his LSD had been sprayed, or just the LSD itself. We asked our respondents whether the statute was ambiguous, but we put the question to them in different ways. The results are shown in this graph:
The vertical axis measures judgments about ambiguity. The horizontal axis measures policy preferences. The green and orange lines show the results when we ask people whether they think the statute is “ambiguous” or whether multiple readings of it are “plausible”. The black line shows the results when we ask whether ordinary readers would likely agree about the statute’s meaning. Notice that the black line does not vary with policy preferences nearly as much as the other lines do.
Here is another example. In Smith v. United States, the defendant was arrested after trying to trade his machine gun for a bag of cocaine. A statute imposed a long sentence on anyone who “uses” a firearm in relation to a drug trafficking crime. The question was whether the defendant “used” the gun by trying to trade it. We asked our respondents what they thought the statute meant. But again we put the question in different ways. The results have not yet been published, but they are illustrated here:
Subjects who wanted the government to win were much more likely than others to say that the government’s reading of the text was best as a matter of ordinary meaning. They also were likely to say that the government’s position was most consistent with the intent of the statute’s drafters. But when they are asked what ordinary readers would think the statute means, the result is the nearly flat line: their preferences and their judgments become two different things.
These findings raise many interesting questions. Why do these questions produce such different answers, and with such different relationships to policy preferences? What, if anything, might they suggest about how a judge or lawyer should think about a statute’s meaning? Of course we can’t generalize too aggressively from student survey takers to lawyers and judges who have more experience, more time to think, and more resources to bring to bear on their interpretive problems. But the basic cognitive issue we find here may well extend to those other populations. Everyone knows that some judges tend to read criminal statutes in the government’s favor more often than others do; yet the judges speak and often claim to feel as though they are just arguing about law. These studies suggest how it is that this happens. But of course it would be better to survey practicing lawyers and judges (it’s just harder to get access to those populations).
And the surveys have other limitations, and produce other findings, and raise other questions. But I can get to that later, since this post is long enough. Let me invite initial thoughts, questions, suggestions, criticisms, etc., from our other participants.
We examined these issues by surveying over 1,000 law students at various schools. We showed them statutes and facts that the statutes might apply to; the problems were based on real Supreme Court cases. We asked our respondents what outcome they preferred as a matter of policy preference. Then we asked them to say whether they thought the statute was ambiguous, or what reading of the statute’s text they thought was best.
The first result is that peoples’ judgments about ambiguity and meaning are closely entwined with their policy preferences. People with strong preferences about how a case should come out tend to find that the statute at issue is unambiguous, and that it means what they want it to mean. This is true even when we instruct them to put the statute’s text aside when stating their preferences, and to put their preferences aside when stating their views about the text. They generally can’t do it.
But something interesting happens when we change the questions. Instead of asking our respondents whether they thought the statute was ambiguous, sometimes we asked them whether they thought ordinary readers would likely agree about the statute’s meaning. Or instead of asking them what they thought the statute meant as a matter of ordinary English, we asked them what they thought ordinary readers would think it meant. The answers were then remarkably different. They weren’t biased. When saying what ordinary readers would think, the respondents, as a group, were able to give answers that were independent of their preferences about how the case should come out.
Here’s an example. In Chapman v. United States, a statute applied a harsh sentence to anyone who distributed more than one gram of a "mixture or substance containing a detectable amount” of LSD. The question was whether, in sentencing the defendant, the court should weigh the blotter paper onto which his LSD had been sprayed, or just the LSD itself. We asked our respondents whether the statute was ambiguous, but we put the question to them in different ways. The results are shown in this graph:
The vertical axis measures judgments about ambiguity. The horizontal axis measures policy preferences. The green and orange lines show the results when we ask people whether they think the statute is “ambiguous” or whether multiple readings of it are “plausible”. The black line shows the results when we ask whether ordinary readers would likely agree about the statute’s meaning. Notice that the black line does not vary with policy preferences nearly as much as the other lines do.
Here is another example. In Smith v. United States, the defendant was arrested after trying to trade his machine gun for a bag of cocaine. A statute imposed a long sentence on anyone who “uses” a firearm in relation to a drug trafficking crime. The question was whether the defendant “used” the gun by trying to trade it. We asked our respondents what they thought the statute meant. But again we put the question in different ways. The results have not yet been published, but they are illustrated here:
Subjects who wanted the government to win were much more likely than others to say that the government’s reading of the text was best as a matter of ordinary meaning. They also were likely to say that the government’s position was most consistent with the intent of the statute’s drafters. But when they are asked what ordinary readers would think the statute means, the result is the nearly flat line: their preferences and their judgments become two different things.
These findings raise many interesting questions. Why do these questions produce such different answers, and with such different relationships to policy preferences? What, if anything, might they suggest about how a judge or lawyer should think about a statute’s meaning? Of course we can’t generalize too aggressively from student survey takers to lawyers and judges who have more experience, more time to think, and more resources to bring to bear on their interpretive problems. But the basic cognitive issue we find here may well extend to those other populations. Everyone knows that some judges tend to read criminal statutes in the government’s favor more often than others do; yet the judges speak and often claim to feel as though they are just arguing about law. These studies suggest how it is that this happens. But of course it would be better to survey practicing lawyers and judges (it’s just harder to get access to those populations).
And the surveys have other limitations, and produce other findings, and raise other questions. But I can get to that later, since this post is long enough. Let me invite initial thoughts, questions, suggestions, criticisms, etc., from our other participants.
Comments
You can follow this conversation by subscribing to the comment feed for this post.