+(,121/,1(

Citation:
Neil MacCormick, On Legal Decisions and Their
Consequences: From Dewey to Dworkin, 58 N.Y.U. L. Rev.
239, 258 (1983)

Content downloaded/printed from HeinOnline

Wed Nov 22 07:51:26 2017

-- Your use of this HeinOnline PDF indicates your acceptance

of HeinOnline's Terms and Conditions of the license
agreement available at http://heinonline.org/HOL/License

-- The search text of this PDF is generated from

uncorrected OCR text.

-- To obtain permission to use this article beyond the scope

of your HeinOnline license, please use:

Copyright Information

Use QR Code reader to send PDF to

your smartphone or tablet device
 NEW YORK UNIVERSITY
LAW REVIEW
VOLUME 58 MAY 1983 NUMBER 2

ON LEGAL DECISIONS AND THEIR

CONSEQUENCES: FROM DEWEY TO DWORKIN*

NEIL MACGORMICK**

Professor MacCormick articulates a view of the proper role of consequentialist

reasoning in judicial decisionmaking. Working from the position that some but not
all of a legal decision'sconsequences bear on its justification, the authorfirst distin-
guishay several types of consequence. He then argues that, in evaluating the conse-
quences of a novel rule of decision, a court should consider primarily what sorts of
conduct the rule would authorize or proscribe, rather than attempt to predict what
human behavior the rule will induce or discourage. Finally, Professor MacCormick
-shows how judicial use of consequentialist reasoning, when limited in the ways
decribed,is compatible with the view that courts are engagedin establishingnorms
of rightness.

It is an interesting and often argued question how far decisions-

and not only legal decisions-can be justified or made right by their
consequences. One can conceive of two extreme positions. On the one
extreme, the only justification of a decision would be in terms of all its
consequences, however remote-in terms, that is, of its productivity
of the greatest net benefit, taking together all consequences and judg-
ing them by some suitable criterion of benefit and detriment. On the
other extreme, the nature and quality of the decision, regardless of
any of its consequences however proximate, would alone be allowed
as relevant to its justification or its rightness. Neither extreme view is
acceptable. The first one excludes the possibility of any rational justifi-
cation of any decision, since the future is unknowable and chains of
consequences may seem to stretch forward into infinity. The second

* This paper is the text of the first Dewey Lecture in Jurisprudence, delivered before the
New York University School of Law on October 5, 1982.
" Regius Professor of Public Law and the Law of Nature and Nations, former Dean of the
Faculty of Law, Edinburgh University. M.A., Glasgow University; M.A., Oxford University;
L.L.D., Edinburgh University.

239

Imaged with the Permission of N.Y.U. Law Review

 NEW YORK UNIVERSITY LAW REVIEW[l [Vol. 58:2,39

one ignores two crucial things. It ignores both the extent to which the
nature and quality of decisions and acts is itself constituted by the
consequences the decider intends, foresees, or hopes to bring about,
and also, more seriously, the extent to which both prudence and
responsibility to one's fellows require that one give serious thought to
the foreseeable outcomes of one's acts and decisions before finally
acting or deciding, the more so the more momentous the act or
decision in view.
So I reject both extremes and entertain only the middle view that
some kinds and some ranges of consequences must be relevant to the
justification of decisions. While not accepting out-and-out consequen-
tialism as a sufficient or acceptable practical philosophy in itself, I
conclude that some element of consequentialist reasoning must be
present in any sound decisionmaking process, in any satisfactory mode
of practical deliberation.
The decisions to which I here address myself are legal, that is,
judicial, decisions. Judges are an important class of decisionmakers,
and their decisions are characterized by a unique process of delibera-
tion aided by the adversarial arguments of counsel. If in general it is
salutary to hold the makers of momentous decisions accountable both
for what they bring about and for at least the intended, the foreseen,
and even the foreseeable consequences and outcomes thereof, the
more so the more thoroughly deliberated the decisions, then it must be
salutary to hold judges so accountable. Conversely, we should expect
the deliberations of such decisionmakers and the justifications they
give for the decisions they make to take account of the consequences
and outcomes of their decisions.
On the face of it, this expectation is actually and amply fulfilled.
Let me cite as but one obvious example the case of Marbury v.
Madison.' There, Chief Justice Marshall backed his ruling that the
Supreme Court has power to decline to implement unconstitutional
acts of Congress by explicit reference to the unacceptable conse-
quences of the alternative ruling:
Those ... who controvert the principle that the constitution is to
be considered, in court, as a permanent law, are reduced to the
necessity of maintaining that courts must close their eyes to the
constitution, and see only the law.
This doctrine would subvert the very foundation of all written
constitutions. It would declare that an act which, according to the
principles and theory of our government,
2
is entirely void, is yet, in
practice completely obligatory.

5 U.S. (1 Cranch) 137 (1803).

Id. at 178.

Imaged with the Permission of N.Y.U. Law Review

May 1983] LEGAL DECISIONS AND CONSEQUENCES

As many writers, myself included, have argued, recourse to such an

argument from unacceptable consequences is an omnipresent feature
of judicial argumentation. In one form or another, consequentialist
reasoning is, and has long been, alive and well in the law.
"In one form or another.. ."-I use those words advisedly. For I
believe that we need to think more about the forms of argument that
may, at least in a broad way, be called "consequentialist." '3 Is there
more than one way in which decisions can have, or be said to have,
consequences? Are some sorts of consequences particularly relevant to
legal deliberation? By what standards of evaluation are consequences
to be weighed and judged better or worse, good or bad, acceptable or
unacceptable? Are there particular standards of evaluation specially
relevant to legal decisionmaking?
To tackle these questions is, I believe, rather appropriate to the
occasion of a lecture named after John Dewey. I am deeply sensible of
the honor which the Dewey Foundation and the Law School have
done me in inviting me to present this lecture here as the first in a
series which bears the illustrious name of Dewey. The only way open
to me to show the depth of my appreciation of this is to address a
theme which has a starting point in Dewey's own work. And so it is
with the theme of legal decisions and their consequences. In his cele-
brated essay entitled "Logical Method and the Law,"'4 Dewey derived
from his pragmatist conception of logic some conclusions of high
importance for lawyers. After surveying the rapidity of social change
in his own times and the resultant strain arising from attempts to fit
new facts into old categories, he came to this conclusion:
[T]o claim that old forms are ready at hand that cover every
case and that may be applied by formal syllogizing is to pretend to
a certainty and regularity which cannot exist in fact. The effect of
the pretension is to increase practical uncertainty and social insta-
bility. Just because circumstances are really novel and not covered
by old rules, it is a gamble which old rules will be declared regula-
tive of a particular case, so that shrewd and enterprising men are

This term was introduced by G.E.M. Anscombe in Modern Moral Philosophy, 33 Phil. 1
(1958), and is usefully (and critically) discussed in A. Donagan, The Theory of Morality 172-209
(1977), and J. Finnis, Natural Law and Natural Rights 111-18, 131-32 (1980). See also N.
MacCormick, Legal Reasoning and Legal Theory 109-19, 129-51 (1978). Since "consequences" is
an ambiguous term, "consequentialism" is necessarily also ambiguous. I hope that the present
essay helps in clarifying this confusion. I am now inclined to suggest that the intuitive prima
facie appeal of utilitarian consequentialism derives from the proper resort we all have to
"consequences-as-implications" in our practical deliberations, consequences of this sort being
regularly mistaken in moral theorizing for consequences of another kind.
Dewey, Logical Method and Law, 10 Cornell L.Q. 17 (1924-25).

Imaged with the Permission of N.Y.U. Law Review

 NEW YORK UNIVERSITY LAW REVIEW[ [Vol. 58:239

encouraged to sail close to the wind and trust to ingenious lawyers

to find some rule under which they can get off scot free. 5
This, he confessed, was not a specially novel view in itself. But
from it he derived a further implication which he characterized as
"revolutionary," no less. The stated facts of the matter, he said, in
their bearing on legal practice, "indicate either that logic must be
abandoned or that it must be a logic relative to consequences rather
than antecedents, a logic of prediction of probabilities rather than one
of deduction of certainties." 6 My topic, as you see, is a response-only
fifty-eight years late-to this invitation of Dewey's to explore "a logic
relative to consequences rather than to antecedents."
In passing, I suppose we should notice that Dewey's plea for a
consequentialist logic in law is hoist up on consequentialist bootstraps.
In summary form, the argument runs thus: the consequence of a
pretension to syllogistic reasoning is an increase in uncertainty in the
law; uncertainty in the law is undesirable, so we ought to abandon
formalism; since we ought to abandon formalism, we ought to sup-
plant the syllogistic logic of antecedents with a probabilistic logic of
consequences. But when the argument is so summarized, we see that
its conclusion holds good only if the sole available alternative to legal
formalism is legal consequentialism. Leave aside the question of boot-
strapping, and that point of alternatives remains an interesting one.
Much of the most influential modern jurisprudence of the United
States has in fact rested upon the supposition that the alter-
native between formalism and consequentialism is indeed an exclusive
disjunction. That great movement in American juridical thought to
which Professor Robert Summers has given the helpful name of
"Pragmatic Instrumentalism, ' '7 a movement in which Dewey himself
was so significant a figure, has had as a central plank of its platform
the assertion that legal formalism generates substantial irrationality
and serious unpredictability in legal decisionmaking. Both of these
vices, in the pragmatist view, can be cured only by a more plainly and
candidly instrumentalist and policy-oriented approach to the solution
of litigated disputes. For me, the culminating achievement of that
movement is Karl Llewellyn's The Common Law Tradition.8 I take
note that one of Llewellyn's ambitions in propounding the virtues of

Id. at 26.
0 Id. (emphasis in original).
Summers, Pragmatic Instrumentalism in Twentieth Century American Legal Thought-
A Synthesis and Critique of our Dominant General Theory About Law and its Use, 66 Cornell L.
Rev. 861 (1981); see also R. Summers, Instrumentalism and American Legal Theory (1982).
1 K. Llewellyn, The Common Law Tradition: Deciding Appeals (1960).

Imaged with the Permission of N.Y.U. Law Review

May 1983] LEGAL DECISIONS AND CONSEQUENCES

the "Grand Style" over the "Formal Style" in judicial method was to
quell the crisis of confidence in the legal system by demonstrating the
greater calculability or reckonability of decisions in the Grand Style.,
The Grand Style being a consequentialist style, we are still on the trail
blazed out by Dewey.
Be that as it may, the belief that formalism versus consequential-
ism is an exclusive disjunction has provoked more recent criticism.
Robert Summers himself has argued 0 that, over and above the appeal
to formally established authorities in law, there may be not one but
twol1 types of substantive reason advanced in justification of judicial
decisions. On the one hand, there may indeed be "goal reasons" for
giving decision D. D is good because by so deciding we will secure or
promote state of affairs S, and S is a good goal to pursue. Here we
could imagine a case where a court holds that, since busing will
promote educational equality, there must be busing. But on the other
hand, there may be "rightness reasons" for D, reasons which are
backward-looking rather than forward-looking, in contradistinction
to "goal reasons." Given all that has occurred between the parties,
and having regard to some standing "sociomoral norm" about right
conduct in such cases, the right decision as between these parties, in
the relationship which has come about between them, is D. Here we
could imagine a case where defendant has acted in breach of good
faith to plaintiff, and simply on account of that ought to be held
liable.
If Summers is correct about this bifurcation of types of substan-
tive reason-and he has at the very least offered a veritable barrage of
evidence supporting his case-he can raise a dilemma for the thesis
that consequentialist arguments are the most fundamental justifica-
tory reasons available for judicial decisions. The dilemma is this:
either we interpret consequentialist reasoning as coextensive with the
advancing of goal reasons for decisions, in which case the possibility of
rightness reasons as equally fundamental is overlooked; or alterna-
tively we interpret consequentialist reasoning as including both ap-
peal-to-goal reasons and appeal-to-rightness reasons, in which case the

' Id. at 5, 37-38, 401-02.

Summers, Two Types of Substantive Reasons: The Core of a Theory of Common Law
Justification, 63 Cornell L. Rev. 707, 716-22 (1978); cf. Summers & Kelley, "Economists"
Reasons for Common Law Decisions-A Preliminary Inquiry, 1 Oxford J. Legal Studs. 213
(19s1).
I, Summers introduces a third type of "substantive reason," namely "institutional reasons,"
Summers, supra note 10, at 722-27. I pass over this third type here, not out of disrespect, but
because the distinction between "rightness" reasons and "goal" reasons, if well made, is sufficient
for the present case.

Imaged with the Permission of N.Y.U. Law Review

 NEW YORK UNIVERSITY LAW REVIEW (Vol. 5S.-2:39

category of consequentialist reasoning is a confused and ambiguous

one. So either consequentialist reasoning is only a part, and not neces-
sarily a fundamental part, of legal justification or it is a confused and
ambiguous category. This, I confess, is an acute dilemma for me,
since my own stance has hitherto been one of representing consequen-
tialist arguments as the most fundamental ones in legal justification-
and I should add that Professor Summers has, in friendly discussion
though not in print, taxed me with being impaled on that horn of the
dilemma which is confusion and ambiguity. Since I am fond neither
of being taxed nor of being impaled, I clearly have to do or say
something about this.
In Professor Ronald Dworkin's work12 we find a different though
not utterly dissimilar critique of some contemporary ideas about con-
sequentialism in the law. At the heart of Dworkin's central thesis
about law, his "rights thesis," lies (I believe) an objection not so much
to the view that good legal decisions must be oriented to the achieve-
ment of political aims as to the view that political aims are all identi-
cal in kind. Some of the aims which we think it good for our political
community to pursue may be envisaged as collective goals or advan-
tages, like an adequate defensive capability or an efficient postal
service or an increase in Gross National Product. But others are envis-
aged as goods realizable only through the participation of individuals
in them, like free speech or the enjoyment of private property or of
some tolerable minimum standard of personal welfare. Most funda-
mental among this latter class of goods is, in Dworkin's view, the good
for any individual of being treated with the same degree of concern
and respect as any and every other individual. The only morally
acceptable form of human community is one in which every individ-
ual member receives from every other and from collective institutions
of government the same equal degree of concern and respect, indeed
that full concern and respect which is properly due to moral agents as
such.
The distinction between these two types of social aim is mirrored
in and definitive of the distinction which Dworkin stipulates as be-
tween policies and principles.' 3 This stipulation enables him to distin-
guish between arguments of policy, which support a decision calcu-
lated to further some intrinsically collective good or advantage, and
arguments of principle, which support a decision on the ground of the

'2 See R. Dworkin, Taking Rights Seriously (rev. ed. 1978) (Chapters 2-4 and the Appendix
are especially useful.).
13 Id. at 90. For criticism of the stipulation, see N. MacCormick, supra note 3, at 259-64.

Imaged with the Permission of N.Y.U. Law Review

May 1983] LEGAL DECISIONS AND CONSEQUENCES

rights secured both to individuals at large and thus to each and every
qualifying individual. The proper task of the distinctively legal agen-
cies of government and in particular of courts of law is to ascertain
and vindicate rights. The rights to be vindicated are not every back-
ground right we might think of, but those rights which are grounded
in the political principles best geared to justifying the community's
institutions, whether or not these principles happen to be at any given
moment concretized in explicitly formulated rules of statute or case
law. It is not a proper task of courts to pursue or implement policies,
except where decisions of policy have been duly enacted into legisla-
tion by authorized legislators in a manner consistent with fundamen-
tal constitutional principles of individual right.
One criticism of Dworkin's thesis, offered by Professor Kent
Greenawalt, 14 has been that Dworkin's insistence on a radical differ-
entiation between questions of policy and matters of right is refuted
by the prevalence of consequentialist arguments in justifications of
legal decisions. If consequentialist reasons for decisions are good rea-
sons, policy must have a fundamental role in legal decisionmaking;
and hence the rights thesis fails as a description of the legal system we
actually have.
To this, of course, Dworkin's reply 15 has been that it all depends
on the kinds of consequences and the grounds of their evaluation. The
securing of rights is one kind of aim of a political community; hence
the argument that D ought to be the decision in this case in order to
secure or promote aim A may just as likely be an argument of princi-
ple as an argument of policy. In fact, he claims, legal consequentialist
arguments are characteristically, and ought characteristically to be,
arguments which test decisions and their consequences against princi-
ples rather than against pure policies.
So in the debate between Dworkin and Greenawalt, we are again
alerted to the possibility of some hidden ambiguities in the very notion
of a consequentialist argument. Everybody seems sure that some kind
of consequentialism has some part to play. But it's much less clear
what kind, and what part.
It does, therefore, seem that even a short conspectus of American
juristic debate from Dewey to Dworkin reveals a need for unhesitating
acceptance of Dewey's invitation to us all to get into a "logic relative
to consequences" and see what we come up with. Thus, the conse-
quence of this first part of my lecture is, if nothing else, a vindication
of the title "Legal Decisions and their Consequences: From Dewey to

14 Greenawalt, Policy, Rights and Judicial Decision, 11 Ga. L. Rev. 991 (1977).
' See R. Dworkin, supra note 12, at 294-330.

Imaged with the Permission of N.Y.U. Law Review

 NEW YORK UNIVERSITY LAW REVIEW (Vol. 58:S39

Dworkin," a vindication that I hope you have not found too inconse-
quential.
II

A. Kinds of Consequences
We have detected signs of ambiguity in the very notion of a
consequentialist argument. This suggests that our very idea of conse-
quences may be a somewhat vague and woolly one. To search out help
in curing vagueness, and even at the risk of merely going a-woolgath-
ering, I suggest that we scrutinize some ways of clarifying our confu-
sion by adopting or making some distinctions of terminology. I will
first examine a possible distinction between "results" and "conse-
quences;" then I will examine a distinction within the latter category
as between causal consequences and ulterior outcomes; finally I will
further complicate matters by reflecting on logical consequences, or
"consequences-as-implications." I aim to show that it is the last
named, consequences-as-implications, which are of the greatest mo-
ment in respect of today's topic.

1. Acts and Results

Anthony Kenny has made a suggestion about distinguishing
"results" from "consequences."' 6 This distinction arises in the context
of his elucidation of the notion of an act. If my squeezing my finger on
the trigger of a gun results in a gunshot, we call this the act of firing a
gun. To perform an act is to engage in action that has a specific result.
This is exhibited in the case of what we call the act of firing a gun.
That very act can, of course, be envisaged as having consequences
beyond its necessary result, for example, if the bullet hits a man and
causes his death. "Fred fired his gun and as a consequence Bill died,"
is one way of putting that. Of course, if in this case Bill's death was
what Fred meant to bring about by firing his gun, we might prefer to
give the business a different act-description. We would say "Fred
killed Bill" or "Fred shot Bill dead." So there is relativism in act-
description-act is action relative to specific result, as firing results in
a gunshot or killing in death. And what may be a consequence relative
to the act of firing-that a bullet hits Bill-is simply the built-in result
of the act of shooting Bill.
Let us apply this idea to the case of a legal decision. The kind of
decision a judge or court makes when it decides a case at law is a

16 A. Kenny, Will, Freedom and Power 54-59 (1975).

Imaged with the Permission of N.Y.U. Law Review

May 19831 LEGAL DECISIONS AND CONSEQUENCES

public act. Since it is an act it is defined by its issuing in a specific type

of result, in this case a legally valid order or decree condemning or
absolving a party, declaring a right, interdicting or enjoining some
misconduct or whatever. Notice, in our case, that if the decider of a
controversy fails to produce a valid order determining the issue in
controversy, then necessarily he has failed to decide it, although he
may purport to have decided and may even be believed to have
decided it; producing a valid order is the result definitive of the legal
act of deciding a case.
But since this result is simply definitive of the act of deciding, it is
the very thing which has to be justified, not a factor in the justification
process. So we must move on to reflection on consequences as ulterior
to results.

2. Causal Consequences and Ulterior Outcomes

Under the usage I am describing, "consequences" go beyond
"results." If the decision of a case results in-"takes the form of"-an
order to defendant to pay large damages to plaintiff, the new legal
state of affairs may be a cause of further changes, for example,
defendant's despair at having to find half a million dollars, or plain-
tiff's rejoicing and triumph at being awarded the half million. These
are in an obvious sense (which I borrow from Professor Alan Dona-
gan) 17 causal consequences of the judge's decision. We can reasonably
and intelligibly say that the judge by his decision caused the defend-
ant's despair and the plaintiff's exultation. In general terms, whatever
is caused by the result of an act is the causal consequence of the act.
Out of the despair and the exultation there may be ulterior
outcomes of which the judicial decision is a necessary precondition.
Defendant may decide that the only way he can pay the judgment
debt is by withholding a large payment he had intended to make to a
cancer research charity, and in turn the trustees of the charity may for
want of that payment cancel an intended research program, throwing
various people out of work. Plaintiff may go on an enormous binge,
squander his gains, become an alcoholic and die on skid row. All these
events are indeed outcomes of the judge's decision, but we would not
say that they are caused by it even though we assign it a place in their
history as a necessary condition of their coming about; even, indeed,
though we may think it right to trace two tragedies back to this one
common starting point.
They are not caused by it according to the eminently sound legal
doctrine of novus actus interveniens as helpfully expounded and ex-

17 A. Donagan, supra note 3, at 42-46, 160-64.

Imaged with the Permission of N.Y.U. Law Review

 NEW YORK UNIVERSITY LAW REVIEWol [Vol. 58:239

panded by Hart and Honor6, 8 and by Alan Donagan. 9 If we believe

in human agency at all, we must believe in a difference between my
causing a state of affairs S and my causing a prior state of affairs in
which another agent has some opportunity, temptation, ground or
means to bring about S. That other agent causes S,and I don't cause it
even though he would not have done so but for the opportunity,
temptation, ground, means or whatever furnished by me. Opportuni-
ties may be necessary for thefts, but exceptional cases of compulsive
behavior apart, they do not necessitate people to steal. I don't cause
the theft of my briefcase by leaving it in an unlocked car; and that is
true even though my loss of it by theft in such a case may be my own
fault.
There is a significant difference between ulterior outcomes of
which a decision (or other act) was among the necessary conditions or
preconditions, and the events or states of affairs of which the decision
is itself the cause. Let us signify this as a difference between conse-
quences as outcomes and causal consequences. (But do not let this
distinction make us suppose that only consequences but never out-
comes can be imputable to deciders, bizarre chains of consequences
can be unforeseeable, as in the Palsgrafcase, 20 while outcomes can be
readily foreseeable, as in the English case of the painter who went off
for his lunch leaving the house unlocked, which incautious act fur-
2
nished an unidentified third party with the opportunity to burgle it.1
Foreseeability is, in this sphere, more significant than causation.2)
Be that as it may, we can already see that consequences in a wide
sense comprehend at least results, causal consequences, and conse-
quences as outcomes of an act or decision. And we have seen that legal
decisions must have a result and can have both other sorts of generic
"consequences." But while we remain focused on the particular deci-
sion and its particular results, consequences and outcomes, I fear we
still have not reached a point that helps us in clarifying a role for
consequentialism in legal justification of decisions. For it is a truth as

18 H.L.A. Hart & A.M. Honor6, Causation in the Law 69-70, 94 (1959).
10A. Donagan, supra note 3, at 42-46.
20 Palsgraf v. Long Island R.R., 248 N.Y. 339, 162 N.E. 99 (1928).
21 Stansbie v. Troman, [1948] 2 K.B. 48.
22 It should be noticed that although, as in Stansbie, id., (noncausal) outcomes of incautious

acts often are quite foreseeable, outcomes of outcomes are not commonly so, since at each stave
some human agent is choosing how to respond to a new situation created by another. Hence,
there are not "chains of outcomes" in the same way as there can be "chains of consequences" in
the causal sense. Even though some causal consequences are unforeseeable at the moment of an
act, given the agent's knowledge, the problems of foresight are different in the two cases.

Imaged with the Permission of N.Y.U. Law Review

May 1983] LEGAL DECISIONS AND CONSEQUENCES

old as David Hume 23 and as young as Graham Hughes24 that, with

the exception of things like sentencing decisions in criminal trials and
decisions about discretionary remedies, neither the particular proba-
ble consequences nor the particular probable outcomes of a judicial
decision between parties are relevant grounds for its justification. The
mean exultation of the plaintiff and the understandable consternation
of the defendant may indeed by utterly regrettable consequences of an
award of damages; they may be compounded by the likely outcomes
of the debauchery of the plaintiff and the disablement of the defend-
ant from continuing a lifetime's practice of charitable good works,
both these outcomes being to the considerable net detriment of the
community at large; but however certain such consequences and out-
comes might seem at the final moment of deliberation, they would not
in themselves count as sound justifications for a decision against
awarding damages. (They might be pressed rhetorically by counsel,
and might furnish a judge with a strongly tempting motive for bend-
ing the law in the defendant's favor. But they could neither be pre-
sented nor accepted as justifications in themselves for a decision. That
is of course the point of the old saying about hard cases making bad
law; they do so, as Graham Hughes observes, 25 when such a tempta-
tion is not resisted. Here it occurs to me in an aside to hope that my
audience is not thinking that at least in the good old days hard cases
made only bad law, not bad jurisprudence as well. To avoid this
reproach, let me now focus on "consequences-as-implications.")

3. Consequences-as-Implications,Alias "JuridicalConsequences"
If we are to find how consequences may be relevant to justifying
decisions, we must then look beyond and away from particular causal
consequences and outcomes. In fact, the very activity of justifying
decisions by giving reasons for them will prove to be the source of the
kinds of consequences that count. To justify a decision in law is to
present in its support universalized or universalizable reasons. Good
reasons for a decision cannot be ad hoc and for-this-case-only. Where
we are not dealing with a decision which is an open-and-shut applica-

11See D. Hume, A Treatise of Human Nature 497 (L.A. Selby-Bigge & P.H. Nidditch, 2d
ed. 1978) (lst ed. London 1739-40). ("A single act of justice is frequently contrary to public
interest .... Nor is every single act of justice, consider'd apart, more conducive to private
interest, than to public.") The proper extent to which individualizedcircumstances, objectives or
goals may affect, for example, sentencing decisions or decisions about discretionary remedies is
overlooked by Hume.
" See Hughes, Rules, Policy and Decision Making, 77 Yale L.J. 411 (1967-68) (also published
as chapter 4 of Law, Reason and Justice (G. Hughes ed. New York & London, 1969)).
" Hughes, Law, Reason and Justice, supra note 24, at 117.

Imaged with the Permission of N.Y.U. Law Review

 NEW YORK UNIVERSITY LAW REVIEW [Vol. 58,23.9

tion of a pre-established rule, the court must explicitly or at least

implicitly make and act upon some ruling on issues of law in dispute
between parties and any such ruling has to be universal or generic in
its terms. Both because of the institutional nature and setting of the
practice of adjudication and because such a practice is properly gov-
erned at least by the principle of formal justice, to treat like cases
alike, justified judicial decisions presuppose universalizable reasons or
rulings in law that "cover" the particular decision justified. And only
26
well justified rulings can in turn justify decisions.
A necessary part of the justification of such a ruling is to show
that it does not contradict validly established rules of law, 27 and a
further necessary part is to show that it is supported by established
legal principles or by reasonably close analogy with established rules
of law where some statable principle sustains the relevancy of the
analogy. 2 But such grounds of justification, although always neces-
sary, are by no means always sufficient or conclusive in favor of but
one possible ruling in a case. The conclusive or clinching point of
argument when a case still stands open after such testing for consist-
ency and coherence is an argument about consequences in a somewhat
different sense than any we have yet considered. 29
Do you recall the English case of Regina v. Dudley & Stephens?0
The case concerned the prosecution for murder of two shipwrecked
mariners who saved their own lives by killing the cabin boy and eating
him after they had been adrift in the lifeboat eight days beyond the
end of their food supply and six since they had finished their last
drinking water. The defense tried to set up a principle of necessity as a
justification for killing in order to save one's own life. But said Lord
Chief Justice Coleridge:
It is not needful to point out the awful danger of admitting the
principle which has been contended for. Who is to be the judge of
this sort of necessity? By what measure is the comparative value of
lives to be measured?... It is plain that the principle leaves to him
who is to profit by it to determine the necessity which will justify
him in deliberately taking another's life to save his own.... [I]t is
quite plain that such a principle once admitted might be made the
legal cloke for unbridled passion and atrocious crime. 3'

26 See N. MaeCormick, supra note 3, at 73-82, 100-08.

27 Id. at 126-28, 195-228.
28 Id. at 119-26, 152-94. The principles of law explicated and deployed in this type of
argument seem to me to comprehend a good deal of what Summers, supra note 10, treats as
"rightness reasons."
29 N. MaeCormick, supra note 3, at 128-51.
30 [1884] 14 Q.B.D. 273.
3' Id. at 287-88.

Imaged with the Permission of N.Y.U. Law Review

May 1983] LEGAL DECISIONS AND CONSEQUENCES

Notice that the "awful danger" apprehended here is an awful

danger concerning what will follow logically from the principle if
adopted. The only statable way of framing a suitable exception for
necessity in the law of murder will be a ruling in law, or in Lord
Coleridge's term a principle, which necessarily authorizes people
whose lives are in pressing danger to judge whether they should kill
another innocent victim of the same danger in order to save them-
selves. The alarming or unacceptable quality of the principle is exhib-
ited by attending to its logical implications as a possible governing
norm for future cases.
What are in view then are consequences in the sense of logical
implications. Why they matter depends on the context in which they
arise, for here the court could not justify an acquittal unless it were to
commit itself to a ruling in law which allows this defendant the
defense he has pressed. Professor Bernard Rudden, in a valuable
critique of my Legal Reasoning and Legal Theory has suggested call-
ing such consequences-as-implications "juridical consequences," 32 and
well so, for to declare a particular right (us dicere) is necessarily, in
the role of an impartial judge, to declare it as available in every like
case. Hence, as a person of prudence and forethought any judge must
look across the range of possible situations which will have to be
covered by this ruling in point of right. Such consideration of a range
of possible cases cannot but be necessary to adequate assessment of the
acceptability of the decision entertained in the present case.
Here, however, contrary to Dewey's notion of the logic of conse-
quences being a logic of probabilities rather than certainties, we have
in Dudley all the certainty and necessity that is involved in any case of
logical implication. Questions of doubt or probability focus only on
whether the proposed necessity defense requires a statement of princi-
ple in such a form as to have those implications for hypothetical test
cases, that is, on whether we have exhausted all the reasonable possi-
bilities in the way of presently formulable defenses.
Of course, what is not certain is whether the hypothetical test
cases will arise at all, whether passing judgment with reference to
them will alter or affect behavior patterns in the community. Such
questions about the outcomes consequent on adoption of a given
ruling, questions about what Rudden has called "behavioral conse-
quences," ' 33 are indeed open only to conjectural answers. There are

32 Rudden, Consequences, 24 Jurid. Rev. 193, 197-99 (1979). It seems to me that what
Rudden distinguishes as "inbuilt consequences," id. at 199-201, are simply a special case of
"consequences-as-implications."
33 Id. at 194-97.

Imaged with the Permission of N.Y.U. Law Review

 NEW YORK UNIVERSITY LAW REVIEW [Vol. 58:239

plenty of studies, although many more in respect of statute law than

case law, of past legal changes that indicate how difficult and disput-
able it is to establish what have been consequences and outcomes of,
or responses to, any given change in the general rules of law. All the
more conjectural must be forecasts about the likely impact of novel
rules or rulings on the way people will conduct themselves. 3 4 No
wonder, then, if judges rest but lightly on such predictions.
Twice over is this noticeable in Lord Coleridge's opinion in Dud-
ley & Stephens. Recall his objection to admitting the so-called neces-
sity defense into the law of murder. Recall his description of it as a
potential "cloke" for "unbridled passions and atrocious crime." There
is here a hint of a prediction that the consequence of such a general
exception to prohibitions on killing will actually be more passions less
bridled and more crimes more atrocious. But I doubt if such probabi-
listic prediction is much at issue; the point seems rather to be that
there will be a legal cloak for certain such things if they occur, and
hence whatever the influence of the law as a disincentive, which we
do not know, the disincentive will be diminished. The second caveat
against too probabilistic an interpretation of the opinion arises from
Chief Judge Coleridge's candid admission that the defendants' temp-
tation is one to which he himself might likely have given way in their
position:
It must not be supposed that in refusing to admit temptation to be
an excuse for crime it is forgotten how terrible the temptation was;
how awful the suffering; how hard in such trials to keep the
judgment straight and the conduct pure. We are often compelled to

3 See, e.g., David Nelken's forthcoming book, The Limits of the Legal Process: A Study of
Landlords, Law and Crime (1983). At pages 36-37, there is a brief discussion of the unresol%ed
controversy over the question whether exploitation of tenants by landlords was a consequence of
the Rent Act of 1957 (enacted under a Conservative government), or was a consequence not of
the 1957 Act but of a housing shortage. On this topic, see also M.J. Barnett, the Politics of
Legislation (1969). The main burden of Nelken's book is an attempt to establish what have been
the consequences and outcomes of the enactment of the Rent Act of 1965, from which may be
gathered how difficult it is to answer such questions with reasonable certainty at any point in
time, much less through a continuity of changing times and circumstances. Moreover, the study
certainly shows that, as is now officially admitted, the Act has not brought about the conse-
quences that were predicted at the time of legislation as being likely to flow from it. By contrast
with the case of the 1957 Act, this failure of prediction cannot be ascribed to a lack of care in
investigating the problem and in attempting to make well-grounded predictions. The 1965 Act
was preceded by the Milner Holland Report (Report of the Committee on Housing in Greater
London, Cmnd. 2605 (1964-65)), a report that might be considered perhaps the most thorough-
going British social policy report of this century. It appears that retrodiction and prediction of
consequences-as-outcomes is sandy ground on which to build the House of Justification.
I am deeply indebted to Dr. Nelken for permission to cite his book before publication, and
for the evidence he has afforded me from it, all the more so since he does not fully agree with the
case I attempt to build from this evidence.

Imaged with the Permission of N.Y.U. Law Review

May 1983) LEGAL DECISIONS AND CONSEQUENCES

set up standards we cannot reach ourselves, and to lay down rules

which we could not ourselves satisfy. But a man has no right to
declare temptation to be an excuse, though he might himself have
yielded to it ....35
There could scarcely be a more candid (or, I think, honorable) admis-
sion of the uncertainty that in extreme cases a rule one way or the
other affects behavior at all. Yet again, the judgment is obviously that
if the law can have an influence, then in so far as it does, it ought to be
on the side of not yielding even to extreme temptation.
I am inclined to think that it is almost always in this somewhat
hypothetical way that courts examine possible behavioral conse-
quences or outcomes of rulings one way or the other on a difficult
point. Even in the cases that Bernard Rudden cites as relatively clear
examples of "behavioral consequences," this is so. 36 He alludes to a
number of admiralty cases involving rescues and salvage at sea, such
as the case of Scaramangav. Stamp 37 in which Chief Justice Cock-
burn made the following point:
The impulsive desire to save human life when in peril is one of the
most beneficial instincts of humanity, and is nowhere more salu-
tary in its results than in bringing help to those who, exposed to
destruction from the fury of winds and waves, would perish if left
without assistance. To all who have to trust themselves to the sea, it
is of the utmost importance that the promptings of humanity in this
respect should not be checked or interfered with by prudential
considerations as to injurious consequences, which may result to a
ship or cargo from the rendering of the needed aid. It would be
against the common good, and shocking to the sentiments of man-
kind, that the shipowner should be deterred from endeavouring to
save life by the fear, lest any disaster to ship or cargo, consequent
38
on doing so, should fall on himself.
Yes, indeed. But we notice that Chief Judge Cockburn does not com-
mit himself to a vtiew on the probable effects of such a deterrent. It is
enough that it would be bad if the law constituted such a deterrent,
however people might actually respond to it.
Common experience suggests that the greatest weight should be
given to outcomes in the way of probable behavioral changes in

3 [1884] 14 Q.B.D. at 288.

a It occurs to me that my somewhat sketchy response to Professor Rudden's observations on
this point may do less than justice to an interesting and carefully drawn distinction. But my point
stands that judicial reasoning in such cases appears to remain somewhat hypothetical rather than
to rest on probabilistic predictions of changes in behavior.
17 [1880] 5 C.P.D. 295, 304.
31 Id.

Imaged with the Permission of N.Y.U. Law Review

 NEW YORK UNIVERSITY LAW REVIEW[l (Vol. 58:2:39

respect to novel rulings, in those areas where it is particularly likely

that people will explicitly ground their actions in the law as it is laid
down (interalia) by courts. Tax law, insurance law, and conveyanc-
ing are such fields. No doubt, in such fields in which people and
companies are expected to act after informing themselves, or being
professionally advised, about the law, it is highly probable that the
outcome of rulings on the law will be behavior that either conforms to
it or takes advantage of opportunities offered by it or otherwise adjusts
affairs and practices to allow for it. And in such fields indeed rather
more account is, quite properly, taken of probable outcomes in the
way of future responses to legal rulings. But even here it is worth
remembering that the law and rulings in law are not causes of behav-
ior; rather, they are grounds for choice by people, and how people
will choose to respond is always in some degree an open question.
Probably the only conduct one can see as actually necessitated by a
ruling in law, and here the necessity is that of obligation rather than
causation, is the conduct of a judge who adheres faithfully to prior
rulings and seeks to act in their spirit. And that sort of "behavioral
consequence," if such it be, is already covered by our discussion of
consequences-as-implications. So, in the main, what I call consequen-
tialist reasoning law is focused not so much on estimating the proba-
bility of behavioral changes, as on possible conduct and its certain
normative status in the light of the ruling under scrutiny.

B. Evaluation of Consequences
And surely it is in this very way that behavioral consequences and
outcomes matter to us. This follows from the point that responsibility
attaches for the foreseen and foreseeable consequences and outcomes
of one's actions. We cannot conceivably speculate on all the things
that will or might possibly happen if people react in some way or
another to a new ruling in law, but we can at least realize that they
are entitled to take the law to be as a court has ruled it to be. People
are supposed to act conformably to the law, and when they do act on
the law as the court has ruled, the judges at least would be debarred
from saying that they hoped it would not be so. This is really just to
come back yet again in another way to the kind of thing that Chief
Justice Coleridge meant when he spoke about not creating a cloak for
unbridled passions or atrocious crimes, or when Chief Justice Marshall
apprehended the subversion of the written Constitution as an outcome
of holding acts of Congress to be above judicial review.
But to see this is to see how we proceed from scrutiny of juridical
consequences to their evaluation. What is at issue is whether the
conduct that the law would tolerate or permit is acceptable when or

Imaged with the Permission of N.Y.U. Law Review

May 1983] LEGAL DECISIONS AND CONSEQUENCES

if, in reliance on the law's permission, people engage in it; whether it

is acceptable that the law stigmatize as wrongful or invalid deeds that
citizens may have contemplated committing even in the face of such a
ruling as is in view. But what grounds this notion of acceptability?
What is it that reflection on juridical consequences via such hypotheti-
cal test cases enables us to evaluate, and how does this help us to reach
fairly confident conclusions about what is or is not acceptable? By
what standards of value do we do this?
As a matter of fact, judges often allude to justice, to public policy
and/or the common good of the community, to legal expediency or
convenience and to common sense as ostensibly different grounds or
criteria of evaluation which they apply to the juridical consequences
and possible ulterior outcomes of possible rulings in contested cases. 39
That they do so at least helps us to recognize that legal evaluation
operates on a plurality of values, rather than on some single standard
such as so-called "utility," but beyond that all is somewhat opaque.
What is more, to take but one of the concepts mentioned, "justice"
hardly seems to be the name of a simple and unitary standard of
evaluation. Justice itself has many aspects, and the problem is under
which of its aspects it bears upon particular problems. When a judge
tells us, as for example did Chief Justice Vanderbilt in the New Jersey
case of Greenspan v. Slate,40 that the implications of a particular
ruling are contrary to justice-the rejected ruling would have allowed
parents of an infant child to refuse payment to a doctor for urgently
needed medical treatment given to the child without the parents' prior
onsent-the appeal to justice seems conclusory rather than argumen-

' See N. MacCormick, supra note 3, at 105, 111-12, 149-50. The following discussion in the
present paper deals with "justice" only, and seeks to explicate the internal complexities of such
judgments of justice. Insofar as "public policy" is a distinct head of evaluation, contrasted with
justice, it appears to bear more upon the "behavioral" aspects of consequences-as-implications.
That is, it returns to the question whether it would be on some ground desirable or undesirable
that the law tend to promote or encourage or give some motive or ground for a certain sort of
conduct, supposing the law to have a certain influence on conduct. Arguments of "convenience"
or "expediency" are a special version of this sort of argument, dealing with such possible
behavioral outcomes in their bearing on legal processes themselves (of. cases of resort to "flood-
gates" arguments). As for "common sense," I have nothing for the moment to add to the brief
discussions in N. MacCormick, supra.
41 12 N.J. 426, 97 A.2d 390 (1953) (Vanderbilt, C.J.).
[A]ny such artificial basis for a fundamental doctrine as "inferences from slight evidence"
is not only unsound in principle but ineffective in operation, because it does not reach the
cases where no express promise exists and where there is no "slight evidence" from which to
infer a promise, and the cases not so reached are the ones where in simple justice a legal
right and an adequate remedy are most needed.
Id. at 432, 97 A.2d at 392.

Imaged with the Permission of N.Y.U. Law Review

 NEW YORK UNIVERSITY LAW REVIEW [Vol. 5S:239

tative. It states the result of an evaluation without showing the work-

ing of it.
Could we go deeper than this? I am not certain, but I think we
can. Just as there are many virtues, so there are many values-things
or states of affairs or qualities in life and interpersonal dealings on
which we set value. Our notion of rightness and wrongness in action
maps onto these values in the sense that conduct is reprobated as
wrong when it falls below some minimal degree of, say, humanity or
good faith or honesty or carefulness, or, a fortiori, when it manifests
an active disregard for them or their like. 41 One acts rightly though
not perhaps excellently or even well, provided one exhibits the mini-
mum required regard for such values. Individuals and groups may
vary, of course, in where they draw the line, and even in the things on
which they draw lines at all. But legal orders are in central part
authoritative orders of line drawing between right and wrong, the
lines so drawn having application in common to all persons within a
jurisdiction. Justice, or at least justice according to law, is a rather
complex function of the lines between rightness and wrongness, and it
ranges across the whole set of relevant values.
Different branches of law focus on different values or clusters of
values. For example, tort law focuses mainly on respect for the integ-
rity of persons and their possessions; contract law on the liberty of
persons to pursue their own ends, balanced against fidelity to under-
takings, and mutual trust and good faith; all criminal law to some
extent and some parts of criminal law exclusively, focus on public
peace and order as a context of personal security and integrity; the
law of trusts focuses obviously enough on trust, and so on. These
examples are of course all too gross and crude, but I hope they
sufficiently indicate what I mean. The justice at which legal systems
aim includes centrally the preservation of adequate respect for these
values in dealings between persons and the enforcing of due remedies
and penalties when they are breached.
To see this is to take a little further an answer to the question
about grounds of evaluation of juridical consequences. For legal prob-
lems arise in legal settings and new rulings in hard cases build out
from established bodies of law. Established bodies of law focus on
given values or complexes of value. It is against those very values that
we test and eliminate rival rulings in hard cases. In considering the
juridical consequences of a ruling by way of its implications for hypo-
thetical cases we discover whether a ruling commits us to universally
treating as right deeds that subvert or fail of sufficient respect for the
41 For a full statement of the argument as to how judgments of rightness and wrongness map
on to judgments of value, see N. MaeCormick, H.L.A. Hart 50-54, 61-65 (1981).

Imaged with the Permission of N.Y.U. Law Review

May 1983] LEGAL DECISIONS AND CONSEQUENCES

values at stake, or to treating as wrong forms of conduct which

include no such subversion or failure. Either consequence is unaccept-
able because it wreaks injustice, that is, fails of the proper virtue of
legal institutions; but the present analysis perhaps shows why and
how such a conclusion of injustice would hold, rather than merely
asserting it.
If I am right about that, the question about the particular values
proper to legal deliberation almost answers itself. The values against
which we properly test juridical consequences are those which the
branch of law in question makes relevant. And if that is too abstract
an answer, recall the place that constitutionalism or respect for a
constitution as the written foundation of the body politic had in
Marbury v. Madison;42 that which respect for life had both in Dudley
& Stephens43 and in the maritime rescue case that I quoted; 44 that
which proper parental care for children and reciprocation of benefits
conferred played in Greenspan v. Slate. 45 If these references are still
too sketchy and abstract for comprehension, alas that time has pre-
cluded my condescending to particulars with proper richness of detail.
What I finally derive from these reflections is a view of legal
consequentialism which allows me a confession and avoidance in
response to Professor Summers' charge that I conflate rightness reasons
and goal reasons. 46 I confess to a certain conflation, but plead that
there is here no confusion. We may envisage the values pursued
through and upheld by legal rules as standing aims of and for the law.
It must then count in favor of a ruling in law that it promotes or
supports a certain value, and this can be expressed in a sort of goal
reasoning. But since judicial reasoning proceeds under the constraint
of universalizability, the matter is also one of setting norms of right-
ness, of providing rightness reasons. In this context and at this level
47
rightness and goal orientation are two sides of the same coin.
That recognition perhaps puts me in something like the same
camp as Dworkin after all. But I remain uneasy about his stipulative
distinction of policy and principle. 4 Are not the values of the law

425 U.S. (1 Cranch) 137 (1803); see text accompanying notes 1-2 supra.
[1884] 14 Q.B.D. 273; see text accompanying notes 30-31, 35 supra.
44 See text accompanying notes 36-38 supra.
4 12 N.J. 426, 97 A.2d 390 (1953); see text accompanying note 40 supra.
4, See text accompanying notes 11-12 supra.
17 But this should not obscure the possible differences of evaluation of consequences as
between "'public policy" evaluations and "justice" evaluations, where these are conceived to be in
conflict.
41 See text accompanying notes 12-13 supra. The confusion again arises between the general
policies of the law, which include the policy of promoting justice, and "public policy" as a
ground of evaluation of relevant consequences, in those situations in which "public policy" is
properly to be contrastedwith justice.

Imaged with the Permission of N.Y.U. Law Review

 NEW YORK UNIVERSITY LAW REVIEW [Vol. 58:239

capable of being considered a kind of standing policy? That judges

pursue them under the constraint of universalizability entails indeed
that their decisions must always be decisions of principle. But this does
not mean that such decisions require pre-identificationof rights as a
route to finding principles. I should rather say that legal rights are
consequential upon than presupposed by the decisions of principle we
make in law. And, for now, that is as far as my reflection on Dewey's
problem about the logic of consequences has brought me.
There is always far more to be said than lecture lets one say. Even
at the end I have had to raise a doubt rather than commit myself to a
certainty. But in this at least I hope to be echoing that spirit of inquiry
which so animated John Dewey's work. If I were thought to have
echoed it convincingly and with conviction I could be content that I
had fulfilled the task you set me, and that in doing so I had paid you
49
the thanks which I owe.

4' In addition to those thanks, I am deeply grateful to Michael J. Machan for telling and
helpful criticisms of an early draft of this lecture, and to Robert Summers and David Lyons for
observations on the relevance of goal reasons and policy.

Imaged with the Permission of N.Y.U. Law Review