Indeterminacy and Good Faith: Politics and Legitimacy

Télécharger en PDF

Sommaire de l'article

Brian H. Bix

The paper briefly considers three “moments” of the legal indeterminacy debate, in order to gain an overview of the debate’s issues and themes: (1) Joseph Hutcheson’s discussion of “hunches” in judicial decision-making, with the implications that fact about decision-making may have for legal indeterminacy; (2) Mark Tushnet’s work on advocacy and legal outcomes, showing the importance of focusing on (in)determinacy over the longer term; and (3) the work of various empirical political science theorists constructing an “attitudinal model” of judicial decision-making (along with other related empirical work), displaying how judges’ political inclinations (and other extralegal factors) explain outcomes in a large class of cases. All three topics raise serious concerns of a “rule of law” or “legitimacy” nature.


In the 1980s, the question of legal determinacy or indeterminacy seemed central to legal philosophy.[1] The debate may not now seem as central or as vibrant as it did when there was critical legal studies theorists at one end of the debate (radical legal indeterminacy), and Ronald Dworkin at the other end (the right answer thesis).[2] However, the important issues raised have not gone away simply because we speak about them less often.

Legal (in)determinacy is a simple name for a complicated set of distinct but overlapping issues.[3] There is thus, in a sense, not a single problem of legal indeterminacy, but an array of problems, and it is likely that what solves or “resolves” one element of the array will still leave much of the array still problematic. While conceding the difficulty of offering any simple or comprehensive discussion of legal indeterminacy, this paper will try to offer an overview that will highlight certain themes and possibilities within the debate.

We need to recall some of the confusions common to these discussions. When we make claims about right answers to legal questions, we need to be precise about our assertions: Are we claiming that there are always, sometimes, or never right answers? Do we claim that the answer depends on the difficulty of the legal issue? Must the right answers be unique or can there be more than one right answer (while still claiming that some answers are wrong)?,[4] and so on. Also, can the (right) answer to legal disputes change over time (beyond the obvious point that legal sources can be added, subtracted or modified by legislation and by court decisions, thereby changing citizens’ rights and duties)? Finally, and perhaps most importantly, when we ask about right answers to legal disputes, is this a question about the legal materials, considered in the context of some abstract or ideal decision process; or is it a question regarding the actual or likely decisions of real judges; or perhaps both? If the debate is to focus on actual decisions, is it sufficient for our purposes (whatever those purposes might be) that the decisions be consistent and predictable, even if those decisions are determined in part by factors thought to be “extra-legal” or “legally irrelevant” (e.g., the political inclinations, biases, or socialization of the judges)? Of course, to raise this secondary question in turn raises issues relating to what counts as “legal sources” as against “extra-legal sources” – a debate prominent in Ronald Dworkin’s work. (In a case where a court is authorized, perhaps even required, to look at moral and policy considerations in resolving legal disputes, and those considerations, combined with legal materials traditionally and narrowly understood, lead to “the one right answer,” should we consider this issue to have been legally determined or legally indeterminate? Dworkin and Raz disagree on the question).[5]

It is not surprising, given the complexity of the embedded issues, that discussions of legal determinacy go off in a variety of directions: the (in)determinacy of meaning of individual terms, phrases or sentences (with particular attention to vagueness[6] and “open texture”), [7] the reasons why the meaning and application of legal norms could or should diverge from the norms’ semantic meaning;[8] the problem of conflicting legal norms; the problem of vagueness in the ultimate standard of legal validity; [9] the range of acceptable interpretations of precedent (in Common Law countries); [10] and the general “defeasibility” of law.[11] Terms can be vague, or at least be uncertain in application when confronting novel fact situations. The plain meaning of norms can conflict (in particular applications) with the purpose of the enactment or with the intentions of the lawmakers. One relevant legal norm may conflict with another norm that also seems to apply. And even where the meaning and application of legal norms seems straightforward, the judge may have the legal authority (and perhaps a legal duty) to change the law, where this would improve the law or avoid a clear injustice. Additionally, one regularly comes across instances where the legal sources seem to require one outcome, but the actual case comes out the other way, because of judicial error or bias (more on judicial bias below).[12] Finally, the indeterminacy of legal outcomes may come not from the legal norm(s), but from the facts to which it/they are being applied, in that the facts might be subject to different characterizations (e.g.¸ focusing on different levels of generality or different time frames) which affect the legal outcome,[13] or simply because legal fact-finding is especially prone to error.[14]

Additionally, there are discussions and debates at a higher level, a “meta-level” if you will, regarding the significance and implications of these debates about indeterminacy. Does occasional legal indeterminacy – either in abstract consideration of legal sources or in the actual decisions of judges – undermine the rule of law and create important questions about the legitimacy of law and government? Or is legal indeterminacy (in moderate amounts) both unproblematic and expected?

This article will briefly consider three “moments” of the legal indeterminacy debate, in order to gain an overview of the debate’s issues and themes. Those three are: (1) Joseph Hutcheson’s discussion of “hunches” in judicial decision-making, with the implications that fact about decision-making may have for legal indeterminacy; (2) Mark Tushnet’s work on advocacy and legal outcomes, showing the importance of focusing on (in)determinacy over the longer term; and (3) the work of various empirical political science theorists constructing an “attitudinal model” of judicial decision-making (along with other related empirical work), displaying how judges’ political inclinations (and other extralegal factors) explain outcomes in a large class of cases. As will be discussed, all of these topics raise serious concerns of a “rule of law” or “legitimacy” nature.[15]

In what follows, Part I looks at hunches and indeterminacy; Part II examines at indeterminacy over time; Part III describes positive political theory and indeterminacy; and Part IV then adds some reflections and conclusions.


I. Hunches and Indeterminacy


In 1929, Judge Joseph Hutcheson wrote an influential article about the judicial process,[16] in which he argued against the conventional view of how judges decide cases. Judges had been portrayed (and portrayed themselves) as coming to their decisions through deduction from general premises. The general premise might be: “When facts A, B & C are present, defendant has committed crime Y.” Then the court would note that facts A, B & C are present, and would conclude that the defendant was thus guilty of crime Y. Hutcheson argued that this picture of courts analyzing cases as syllogisms was a false picture of how judicial decisions – and, indeed, most practical decisions, legal or non-legal – are made.[17]

Hutcheson claimed that judges generally start with a “hunch” about how the case before them should come out, and then try to write an opinion consistent with that hunch. The great American pragmatist philosopher, John Dewey, writing in the Cornell Law Quarterly, offered a similar position. Dewey wrote:

[W]hile the syllogism sets forth the results of thinking, it has nothing to do with the operation of thinking. […]

[…] As a matter of fact, men do not begin thinking with premises. […] [W]e generally begin with some vague anticipation of a conclusion […], and then we look around for principles and data which will substantiate it […].[18]

Assuming that Hutcheson’s (and Dewey’s) descriptions of decision-making are largely accurate (and they are consistent with recent work by behavioral psychologists)[19], there remains a question of whether this fact is problematic.

On one hand, if there is a single unique answer to all (or nearly all) legal questions (as Dworkin argued),[20] and we can be confident that competent judges acting in good faith will find that right answer, then we need not worry about hunches. Wherever judges begin in their deliberations, in due course they will reject all incorrect answers and find their way to the correct answer.

On the other hand, if there are regularly more than one correct answer – more than one answer that is equally legitimate and defensible given the facts and the relevant legal norms – then the initial “hunch” a judge has can be quite relevant. If one could justify an outcome for either the plaintiff or the defendant, then which way one decides the case will likely be determined by which way one’s initial hunch went. If one starts off believing that the plaintiff should win, one will then try – and succeed – in writing an argument justifying the decision in favor of the plaintiff, and one will have little reason to change one’s mind and write the decision instead for the defendant. Similarly, if both a “conservative” and a “liberal” resolution of some question of constitutional interpretation would be justifiable, then we would expect conservative judges to have conservative hunches, and see no reason later to reject that hunch, and liberal judges to have liberal hunches, and similarly stay with that initial position. (We will revisit the importance of political inclination for judicial decision-making in Part III.)

To summarize, if judicial decision-making starts with a hunch (and this seems likely), that hunch can affect the outcome of the decision if there is more than one defensible conclusion. Thus, it is indeterminacy that makes hunch-based decision-making significant. If the starting point of the decision (the hunch) is affected by political (or other) bias, then it is those biases rather than the legal materials that is determining the outcome for those cases where there is more than one defensible outcome. In other words, the hunch aspect of decision-making does not make law more indeterminate – it does not explain indeterminacy – but it does help us to understand why (and when) indeterminacy might be troubling. The importance of hunches is exacerbated by the context (especially in the United States): where there is frequently a great deal at stake in judicial decisions (especially, though not exclusively, in constitutional law cases), and judges have a wide variety of backgrounds and values.

Another wrinkle to the issue that we do not have time to consider at length here: Duncan Kennedy has argued that the extent to which the legal materials “resist” the outcome set by a “hunch,” or desired by a judge (for ideological or other reasons), should be seen as variable, a function of the judge’s ability, motivation, and available time, and not merely something that is “objectively in” those legal materials.[21]


II. Indeterminacy Over Time


Most discussions of indeterminacy are relative to a particular legal system at a particular time. Often, this context is unstated, but nonetheless, the legal determinacy is assumed to be this legal system, now. Some new questions and issues arise when one considers how the determinacy of certain questions can change over time.

There are some obvious ways in which the resolution of legal issues changes over time, ways which do not raise theoretical concerns. Law can change because the legislature passes new statutes changing the law on certain matters. Courts and administrative agencies can also come down with important new decisions that appear to change, or at least clarify, the law, even though the decision-makers may purport only to be declaring what the law already is (how to characterize decisions that seem to change the law even while purporting only to declare existing law, is a theoretical issue of its own, that I will not explore further here).

What I want to focus on is Mark Tushnet’s argument, in his 1996 article, “Defending the Indeterminacy Thesis,”[22] that accepted understandings of authoritative legal texts can change over time in response to legal and social advocacy. Tushnet’s point is that which legal questions we consider “easy” – and, its correlate, which legal arguments we consider “clearly wrong,” “off the wall,” or “nonsensical” – can change over time, and often as the result of purposeful action by activists.[23] Advocates for particular positions, connected with certain ideologies or interest groups, work to change our understanding and application of particular legal texts.[24] There are numerous examples of issues in American constitutional law where a position went from being dismissed as silly to its being seriously considered or even adopted, this move happening in the wake of intense scholarly and social advocacy. Such results from right-wing advocacy include the view that government regulations could constitute “takings” that require compensation under the 5th Amendment to the U.S. Constitution,[25] and the view that the 2nd Amendment to the Constitution protects individual gun ownership[26]; results of left-wing advocacy include the view that the 14th Amendment to the U.S. Constitution requires the inclusion of same-sex couples in state marriage laws.[27]

At one level, there is nothing strange about shifts in law through advocacy. It is commonplace for the law to be changed through concerted advocacy that results in new legislation or regulation. Of course, we may not be happy about the direction of such changes, and we might raise questions of legitimacy if we think that certain groups (in particular, the wealthy or some other group) seemed to have disproportionate influence, but the basic idea of legal change through lobbying is central to our understanding of democratic governance. One might assert that the sort of legal change Tushnet describes is simply a variation of this kind of normal, democratic change.

The response would be that lobbying a legislature is different than lobbying a court. Courts are supposed to be (in Ronald Dworkin’s terminology) “forums of principle.”[28] Judges are supposed to apply the existing law, not make up new law (while pretending that they are only applying existing law). Of course, we lawyers and academics know (however much judges may protest to the contrary) that judges do legislate, they do make new law. However, the conventional view is that judges only legislate (or only should legislate) to fill in gaps, to deal with the “open texture” of legal provisions.[29] The sort of change in legal meaning Tushnet is describing is not merely interstitial gap-filling; it is a significant change in the central meaning of important legal texts (does the Second Amendment protect individual gun ownership?, does the Fourteenth Amendment require the inclusion of same-sex couples in state marriage statutes?, and so on). The general understanding is that constitutional provisions, statutes, and administrative regulations have set meanings, with perhaps some need for judicial clarification or filling in on borderline cases or for unforeseen fact situations. However, the kind of changes Tushnet is describing go far beyond that – making settled law unsettled, or changing the settled law entirely from one conclusion to its opposite. From a rule of law perspective, it is just as bad (if not worse) for the legal materials to lose or change their settled meanings as it is for those materials to leave disputes undetermined.


III. Positive Political Theory and Indeterminacy


Political theorists have done extensive empirical work on judicial decision-making, and have found what appear to be explanations of judicial behavior that are distant from legal sources and legal doctrine. Best known, perhaps is the “attitudinal model” of judicial behavior, based upon data which appear to show a strong correlation, at least for appellate court cases, between outcomes of cases and the political inclinations of the judges (or, a close proxy, the party-political affiliation of the official who appointed the judge).[30] Other recent work has found a correlation between outcomes and other extra-legal factors, like the political inclinations of the other judges on a multi-judge panel,[31] or even whether a judge has a daughter.[32]

To be sure, claims of the determination of legal decisions by politics (and other extra-legal factors) are often overstated: even the strongest versions of these theories do not claim that all decisions are determined by these factors. There are easy cases, where judges of different political inclinations and different circumstances would agree as to the outcome. Further, there are many disputes that, because they are easy as a matter of law, because different judges would agree on the outcome, never get litigated at all.[33]

Cases whose outcomes are determined and explained by political affiliation or similarly extraneous factors are instances of “indeterminacy,” at least in one of the senses in which that term was used by the American legal realists and critical legal studies (CLS) theorists: that these are cases where it is not the legal materials themselves, but something else, that is determining the outcome. As noted earlier, this claim of indeterminacy is consistent with claims that these politically-determined cases are fully predictable. We can (and often do) predict (accurately) that (e.g.) a court dominated by Republican-appointed judges will decide a controversial case in a “conservative” way, but that is different from concluding that this is the “right” outcome according to the relevant legal materials.

We can also run the analysis in the other direction: not from political determination to legal indeterminacy, but rather from legal indeterminacy to a suspicion of political determination. Where legal sources do not determine the outcome, then the outcome must be explained (at least in part) by other factors, which often will include unconscious biases of one kind or another.

There is a sort of second-level indeterminacy issue lurking here as well. To conclude that decisions must be attributed to extra-legal as contrasted with legal factors, one needs a fairly clear sense of when the legal materials are sufficient (or at least necessary) to explain the outcome. And that remains a matter of significant controversy. Critical scholars, from the American legal realist, European Free Law, and critical legal studies movements often spoke of judges who portrayed themselves as bound (to a single legal answer) when they were in fact not bound (there was more than one legally justifiable response).[34] The implication, at least sometimes, was that these judges were hiding policy choices under the guise of legal constraint, but I think that we should take seriously the possibility that the judges sincerely thought that they were constrained, and simply did not see that more than one outcome was justifiable.

There is an obvious connection between this third topic (the attitudinal model) and the first one (the role of the hunch in judicial decision-making). As mentioned, hunches are problematic primarily when there is more than one tenable (defensible) resolutions of a legal dispute and which resolution one is directed to by one’s hunch is determined by non-random factors, of which political/ideological tendencies may be both the most obvious and the most problematic.

Again, all of the above analysis assumes that judges are acting in good faith, attempting to apply the law. There are, of course, corrupt judges, and judges who (some of the time or much of the time) seek only to reach particular outcomes, without any regard for what the law is, but from my experience, both working for and observing judges,[35] such cases are rare.


Reflections and Conclusions


As stated at the start: “indeterminacy” is many different things, and these many different things have many different causes and implications. It is not easy to find what binds all of them together. However, one might start from the concerns that motivate the inquiries.

One important concern (associated with Ronald Dworkin) is that where the legal sources do not determine the outcome, then judges will be making decisions affecting liberty, property, and perhaps even life, based on a retroactive application of new (judge-made) law. It is only if there are right answers to legal disputes, answers based on legal sources – the actions of legal officials – that had occurred prior to the action(s) being judged, that a criminal or civil defendant can be assured that decisions are grounded on what the law actually was, rather than on what a judge has decided to make it. (Dworkin was not much troubled by the fact that the legal “right answers” might not be demonstrable, that reasonable judges acting in good faith could, and in the harder cases likely would, disagree on what the law is. What was important to him was that judges focus on trying to determine what the law is, rather than turning to their own views about what the law should be.)

Another worry (relating to the one just mentioned) is that indeterminacy undermines the ideal of “governance of law, not of [people].” There is a hope that judges would simply apply the decisions already made by other legal officials, where “apply” was understood as something mechanical,[36] or at least neutral. If the legal sources need to be supplemented by other sources – whether the idiosyncratic preferences of judges or moral or political truth (on which judges will inevitably differ) – then the ideal of transparency or neutrality is undermined.[37] Where law is indeterminate, the decisions made in the name of the law can be criticized as arbitrary,[38] inconsistent, and unconstrained.[39]

Questions of determinacy are also often equated with issues of objectivity.[40] However, this is as likely to create confusion as clarity, as the concept of objectivity is at least as controversial and multi-faceted as that of determinacy.[41] Sometimes “objectivity” in law is equated with there being one right answer to interpretive questions, sometimes to the further claim that this right answer is grounded in realist or natural kinds analysis.[42]

One last point: if there are now fewer publications on the issue(s) of legal (in)determinacy, it may reflect just the coming and going of “fashion” for different topics. However, whether the issue is in or out of “fashion,” the determinacy of legal disputes raises a series of important issues that warrant our continuing attention.


Brian H. Bix

Brian H. Bix is the Frederick W. Thomas Professor of Law and Philosophy at the University of Minnesota. He received his J.D. (law degree) from Harvard University and his D. Phil. (doctorate) from Oxford University. He is a Member of the Bar of the States of Minnesota, Massachusetts, and Connecticut, and also a Member of the American Law Institute. He writes primarily in Jurisprudence, Contract Law, and Family Law. His publications include Jurisprudence: Theory and Context (7th ed., Sweet & Maxwell, 2015); A Dictionary of Legal Theory (Oxford, 2004); Contract Law: Rules, Theory, and Context (Cambridge, 2012); and The Oxford Introductions to U.S. Law: Family Law (Oxford, 2013).

Remonter en haut de page