A Cost/Benefit Analysis of Judicial Review

I want to come back to the issue of judicial review ― both of legislation and of administrative decisions ― and deference, about which I wrote earlier this week. In that post, I suggested that our views on deference in judicial review are a function of our deeper beliefs on such principles as democracy and the Rule of Law, as well as on the institutional competence of the various branches of government, and that a coherent set of such beliefs could produce superficially inconsistent views about the degree of deference appropriate in various sorts of judicial review. I would like to expand on that idea, or perhaps just to restate it, in a way that I find helpful to clarify things: by means of a cost/benefit analysis.

Note that the following is likely to be boring, trivial, unoriginal, or simply stupid. It will certainly be exceedingly long. It’s an attempt to think some things through, out loud, and to invite constructive comments from those bold or kind enough to read it. (“This is rubbish” can be a constructive comment in this instance.) The claims I will make are tentative, and the terminology I use even more so. And I’m not sure what, if anything, follows even if everything I say here is right. This may well be a purely intellectual exercise. You’ve been warned!

* * *

To state what should be ― but, I suspect, isn’t always ― obvious, judicial review is costly. Every legal procedure (and indeed any human activity) has costs of course, but judicial review is more costly than other forms of litigation, in a number of ways. Perhaps most fundamentally, all judicial review is in a real sense superfluous. Ordinary litigation is necessary in order to provide the parties with an authoritative determination of their legal position when that position is unknown or contested (though of course the parties may choose to forgo the benefit of an authoritative determination and settle their dispute by themselves, or turn to arbitration instead). But judicial review is not necessary to do this. The legal position of the party or parties involved has already been authoritatively determined, either by the legislature or by an administrative decision-maker. Judicial review is an attempt at overturning an already-existing authoritative determination. We could live without it ― and we do live without it in some areas. Judicial review has other unique costs too, to which I will return shortly.

The costs of judicial review fall into two general categories. I will refer to the first one as that of operational costs. The operational costs of judicial review mostly result from the diversion of resources ― people, their time and energy, and money ― to judicial review from more socially useful operations. (These more useful operations range from the practice of the sorts of law that is actually necessary to determine or to change the legal position of persons to the flipping of burgers, or anything in between.)

The operational costs of judicial review are of two types. Direct operational costs include of the resources actually expended on the litigation of constitutional and administrative cases ― the time of the lawyers who argue these cases and of the judges who decide them (instead of deciding other cases), the clients’ costs, etc. These costs also include the uncertainty about the validity of legal decisions, and thus about the legal position of persons, that persists while a case is being litigated. Indirect operational costs consist of the resources that are expended due to the existence of judicial review, but not on judicial review as such ― for example, the time that I spend on writing this post, and the time you expend on reading it, which might have been spent on more socially productive (if not more enjoyable!) activities if judicial review did not exist.

The other general category of costs of judicial review arises from the courts’ decisions; I will call them the decisional costs. One significant source of such costs is the democratic loss that occurs, most obviously, when courts invalidate legislation enacted by elected representatives of the people, and perhaps also when they invalidate decisions made by administrators whom these elected representatives entrusted with the making of such decisions ― though whether judicial review of administrative action really undemocratic is debatable. A somewhat different source of decisional costs of judicial review is overenforcement: the invalidation of statutes that are not actually unconstitutional or administrative decisions that are not actually ultra vires (or unconstitutional). The converse of overenforcement is arguably the over-legitimation of those statutes or decisions that are not invalidated. Legislators and administrators can see  and present their decisions being upheld as a sign of approval, an imprimatur, even though these decisions may still be not only unwise, but even morally wrong.

We can think of other decisional costs of judicial review as well, but these are more speculative. One example is the indifference to constitutional issues that judicial review is sometimes said to foster among politicians (who simply rely on courts to tell them whether what they do is constitutional). I suspect, as I explained here, that politicians do no better in those areas where judicial review is not available. Another cost of judicial review, according to some, is the erosion of precedent caused by constitutional adjudication. This argument, in my view, is interesting but overstated. Yet another possible but speculative cost is simply judicial self-aggrandizement.

Note, by the way, that almost all of the costs of judicial review are generated in much the same way by review of legislation and that of administrative decisions. It is only those I describe in the previous paragraph that seem to be unique to the constitutional context, and I’m not even quite sure that they are. And of course they might not be real costs to begin with.

* * *

Of course, judicial review has a number of benefits too. The one common to all forms of judicial review is the enforcement of the Rule of Law, in the core sense that “government in all  its actions is bound by rules fixed and announced beforehand” (F.A. Hayek, The Road to Serfdom, Ch. 6). These rules include both the constitution, which binds the actions of the legislature as well as of the executive, and statutes and common law rules that bind the action of the administration. At least so long as the laws and administrative decisions invalidated on judicial review are unconstitutional or illegal more often than not, judicial review increases the government’s overall compliance with the Rule of Law.

Other benefits are specific to the various substantive categories of judicial review. Rights-based judicial review (whether of legislation or of administrative decisions) increases the legal system’s respect for individual and minority rights; federalism-based judicial review increases protects federalism (and thus democracy); review of administrative decisions for procedural fairness serves to protect the procedural aspects of the Rule of Law. Judicial review of legislation is also sometimes said to serve a sort of communicative function, educating the people about the constitutional values and commitments of their polity, but I am very skeptical about such claims.

* * *

So how do these benefits compare to the costs of judicial review? Well, as I suggested last week, your answer to this question probably depends on the sort of judicial review we are talking about, because your assessment of both the costs and the benefits of judicial review probably varies depending on the circumstances ― and your legal culture and ideology.

On the cost side, you might think that some forms of judicial review, for example those involving complicated facts, have unacceptably high operational costs. Or you might think that some forms of judicial review are more likely than others to result in over-enforcement. For example, if you accept the post-New Deal consensus view of the “Lochner era” in American constitutional history, you think that overenforcement is likely to happen when courts review laws dealing with economic regulation and social policy. If you tend to view administrative decision-makers as experts and judges as amateurs, you are probably worried about overenforcement resulting from judicial review of administrative decisions. There are, of course, many more examples. And of course your assessment of the costs of judicial review depends on how you interpret and value democracy. If, like Ronald Dworkin, you think that rights-based judicial review is an integral part of a democratic political structure, you won’t view this type of review as having democratic costs; if, like Jeremy Waldron, you value decision-making by elected legislatures very strongly, your estimate of the democratic costs of all judicial review of legislation will be very high.

Your assessment of the benefits of judicial review similarly depends on your interpretation of the various principles judicial review protects, and the importance you attach to them. If you think that the Rule of Law requires legal rules to have a single meaning, and that it further requires courts, rather than decision-makers influenced by politics, to have the last word about this meaning, you will think that judicial review has considerable benefits. If you think that the Rule of Law can accommodate multiple meanings of legal rules, then judicial review is less valuable, quite apart from any concerns about overenforcement. Similarly, if you don’t think that federalism is an important constitutional principle, you will be skeptical of the value of federalism-based judicial review, even if you are not worried about it resulting in overenforcement and are unconcerned about its democratic costs. Again, there are of course many more examples.

How to balance the costs and benefits of judicial review is, of course, a difficult question. People are bound to disagree about it. They might disagree about the balancing even if they generally agree about things like the institutional competence of courts (and the attendant risks of overenforcement) and the interpretation of constitutional principles. But of course people also disagree about each of these things. The existence of widespread agreement about judicial review is almost puzzling ― but it is actually an illustration of legal ideologies at work. An ideology tells its adherents both how to interpret the relevant principles and which values to assign to them, and thus obviates the need for individual balancing ― and prevents disagreement.

The bottom line, despite these complications, is clear: if you think that, for a given type of case, the benefits of judicial review are greater than the costs, you favour judicial review. Otherwise, you want the courts to stay away.

* * *

So how does deference factor into this analysis? It seems to me that deference, whether total or partial, serves to reduce both the costs and the benefits of judicial review. Deferential review is less costly than the non-deferential kind, though the reductions are not equal among the different forms of costs. Clearly, the decisional costs of judicial review go down as the degree of reference to legislatures or administrative decision makers goes up, since overenforcement becomes less likely, and deferential review is less undemocratic. Direct operational costs of deferential review should be lower too, at least in theory, because fewer cases will be brought when potential litigants know they are less likely to succeed due to a higher threshold for judicial intervention. In practice, it may well be the case that litigants will be undaunted at the prospect of a court that defers to government. Certainly the standard of review is not the only consideration that determines whether a litigant will turn to the courts. Indirect operational costs are not reduced by deference at all. Note, by the way, that it is possible that the very existence of a variety of standards of review introduces operational costs of its own, both direct and indirect, as litigants spend their time debating the amount of deference that should apply in their cases.

Deference also reduces the benefits of judicial review. Insofar as judicial review serves to uphold the Rule of Law, individual rights, federalism, procedural fairness, or whatever other substantive value, deferential review presumably accomplishes these objectives less well than aggressive review. Of course, one can think that, in a given area, courts are actually less good than legislatures or administrative decision-makers at upholding these values ― but then you should oppose judicial review in that area, and not merely argue for more deferential review.

Deferential review is thus a compromise. It is less effective, but also cheaper than the non-deferential sort. Importantly, however, there is no reason to believe that costs and benefits will be reduced by the same amount for a type of cases. Conceivably, costs will sometimes go down by a lot whereas the reduction in benefits will be more limited. This will not matter in those cases where the resulting balance of benefits minus costs will still be negative, making judicial review a bad idea. In other cases, however, deferential review will be preferable to the alternatives of either non-review or non-deferential review. Deferential review might also be attractive in those situations where the costs of non-deferential review are seen as intolerably high in an absolute sense, regardless of the benefits they can bring about. Deference can bring the costs down to an acceptable level, even if the benefits go down by as much or more.

* * *

Well, there you are. I said this would boring and long, and so it was. I’ll have more exciting stuff, starting with a review of the federal government’s argument in Aniz Alani’s Senate vacancies case, later this week.