Continuing the Conversation:
A Reply to Manfredi and Kelly
Sujit Choudhry and Claire Hunter*
is counter-majoritarian under
In this reply, the authors assert that Professors
Manfredi and Kelly’s response to their original article
either misses the point or is simply mistaken. The
authors clarify the limited purpose of their original
study, which was to assess the extent to which the
Supreme Court
the
Charter. Manfredi and Kelly’s interpretation of the
available data either relies on inappropriate quantitative
measures or draws overly fine distinctions between
highly variable data sets. The burden of proof is on
those who allege that the Court is engaged in judicial
activism, and Manfredi and Kelly have not succeeded
in demonstrating that the null hypothesis has been
disproved.
Dans leur r6plique, les auteurs soutiennent que la
r6ponse des Professeurs Manfredi et Kelly
rate
malheureusement sa cible ou est tout simplement
erron6e. Les auteurs clarifient les vis6es restreintes de
leur 6tude originale, qui s’appliquait A d6couvrir dans
quelle mesure la Cour supreme agit de mani~re anti-
majoritaire en ce qui conceme la Charte des droits et
libertis de la personne. L’interpr6tation des donn6es
disponibles par Manfredi et Kelly se base sur des
mesures quantitatives inapproprides, ou effectue des
distinctions
trop d6licates entre des ensembles de
donndes hautement variables. Le fardeau de la preuve
repose donc sur ceux qui pr6tendent que la Cour
s’adonne A l’activisme judiciaire, et Manfredi et Kelly
nont pas d6montr6 que l’hypoth~se nulle a dt6 r6fut6e.
* The authors thank Ira Parghi for helpful comments.
Sujit Choudhry & Claire Hunter 2004
To be cited as: (2004) 49 McGill L.J. 765
Mode de r6f6rence : (2004) 49 R.D. McGill 765
766
MCGILL LAW JOURNAL / REVUE DE DROIT DE MCGILL
[Vol. 49
Introduction
1. The Quantitative Critique
A. Hypothesis 1
B. Hypothesis 2
C. Hypothesis 3
II. Qualitative Evidence
Conclusion: The Existing Literature, the Goals of Political
Science, and the Value and Limits of Quantitative Methods
767
767
768
769
771
772
776
2004]
S. CHOUDHRY & C. HUNTER- A REPLY TO MANFREDI & KELLY
767
Introduction
In the academic world, criticism is the highest form of flattery. We are delighted
that Christopher Manfredi and James Kelly would so quickly pen a detailed response
to our study2 of government win and loss rates in the Supreme Court of Canada’s
jurisprudence under the Canadian Charter of Rights and Freedoms.’ In parts of their
response, they disagree with the conclusions we draw from our raw data, often by
reinterpreting or further refining
they dispute our alleged
methodological bias toward quantitative evidence and respond to our arguments with
qualitative evidence that we did not generate or squarely address in our study. Finally,
and most seriously, they charge us with misrepresenting the work of political
scientists who write on the Charter and with failing to address questions that are of
interest to them because of our misapprehension of the goals of their discipline.
them. In others,
Although much of what Manfredi and Kelly say is helpful, unfortunately many of
their responses either miss the point of our article or are simply mistaken. In this
reply, we respond to their critiques that rely on quantitative data; counter their
argument that the Court has been more remedially activist since the delegitimization
of the override; describe the academic context within which we situated our original
article; justify why the questions we asked and the methods we employed have been
of interest to political scientists; and repeat the limitations of a strictly quantitative
approach to studying judicial activism that we set out explicitly in our original article.
I. The Quantitative Critique
Manfredi and Kelly accuse us of basing our quantitative definition of judicial
activism on a misunderstanding of the nature of counter-majoritarian judicial review
and promise to offer “an alternative, equally quantitative definition of judicial
activism that better captures the nature and purpose of constitutionally based judicial
review.”4 But what is their alternative definition? It seems that they disagree with
which cases we count for the purposes of calculating government win and loss rates.
In our study, we excluded non-Charter cases and cases that did not involve challenges
to primary legislation. Manfredi and Kelly would include both sets of cases because
doing so “better captures the nature and purpose of constitutionally based judicial
review,” which is the protection of “minorities from any oppressive government
action.”5
1 Christopher P. Manfredi & James B. Kelly, “Misrepresenting the Supreme Court’s Record? A
Comment on Sujit Choudhry and Claire E. Hunter, ‘Measuring Judicial Activism on the Supreme
Court of Canada”‘ (2004) 49 McGill L.J. 741.
2 Sujit Choudhry & Claire E. Hunter, “Measuring Judicial Activism on the Supreme Court of
Canada: A Comment on Newfoundland (Treasury Board) v. NAPE’ (2003) 48 McGill L.J. 525.
3 Part I of the Constitution Act, 1982, being Schedule B to the Canada Act 1982 (U.K.), 1982, c. 11.
4 Manfredi & Kelly, supra note 1 at 744.
51bid. at2.
768
MCGILL LAW JOURNAL / REVUE DE DROIT DE MCGILL
[Vol. 49
There is good reason for constructing the data set as we have. Our goal in the
paper was to assess the extent to which the Court is counter-majoritarian under the
Charter. Why? Because the most serious issue raised by Justice Marshall’s judgment
in Newfoundland (Treasury Board) v. NAPE6 and by critics of the Charter is that the
Court has erected roadblocks in the path of democratic decision-making. We do not
deny that the Court has applied constitutional provisions found outside the Charter to
find legislation unconstitutional, or that the Court has reviewed the constitutionality of
behaviour by government officials under the Charter. But we are not, in this project,
interested in those other features of the Court’s behaviour. Including other cases in the
data set would not enable us to measure the phenomenon of counter-majoritarian
Charter adjudication.
But let us grant Manfredi and Kelly their modified criteria of inclusion. What
difference does it make? As it turns out, very little.
A. Hypothesis 1
As Manfredi and Kelly concede, their modified criteria make almost no difference
to the overall government win rate.7 But Manfredi and Kelly go further, and interpret
their data as proof of the Court has been activist. In our article, we argued that the
question of whether a win rate would be considered high or low is a matter of highly
contextualized judgment and that the burden of proof rested with critics of the Court
We were also very careful not to say whether these rates are high or low in absolute
terms.9 Rather, we simply stated that this figure had to be the starting point of any
informed discussion on the true extent of judicial activism. Manfredi and Kelly
attempt to make such a judgment by comparing Charter win rates to some benchmark.
In this respect, they take up the gauntlet we throw down in our paper, because we agree
that some kind of benchmark is required to rely on this data as proof or disproof of
activism.
The benchmark they rely on is a comparison of win rates under the Charter to
those from the Court’s jurisprudence under the Canadian Bill of Rights” prior to
1982. They correctly note that at first glance, the Court has been more activist under
the Charter. This, however, is a poor comparison to make because the Bill of Rights
was not constitutionally entrenched, and the fact of entrenchment has been identified
by the Court as a reason to interpret the Charter differently than it did the Bill of
granted, [2003] S.C.C.A. No. 45 (QL) [NAPE].
6 (2002), 220 Nfld. & P.E.I.R. 1, 221 D.L.R. (4th) 513, 2002 NLCA 72, leave to appeal to S.C.C.
7 Manfredi & Kelly, supra note 1 at 747.
8 Choudhry & Hunter, supra note 2 at 532.
‘ Ibid. at 545.
‘0 S.C. 1960, c. 44.
2004]
S. CHOUDHRY & C. HUNTER -A REPLY TO MANFREDI & KELLY
769
Rights.1″ Perhaps a better comparison would be the system of constitutional judicial
review that predates the Charter, namely the federal division of powers. Fortunately,
Manfredi and Kelly provide this data elsewhere in their response to argue that
absolute levels of activism are higher under the Charter. But if one compares win
rates (as opposed to absolute levels of wins and losses) under the division of powers
from 1950 to 1984 and under the Charter from 1982 to 2002, the results are roughly
comparable. 12 In division of powers cases, the government loss rate is thirty-five per
cent, while under the Charter it is only one percentage point higher, at thirty-six per
cent (according to Manfredi and Kelly’s coding rules). Manfredi and Kelly’s point is
that the Court’s behaviour under the Charter is dramatically different and more
activist but, on their own data, the Court clearly is not.
B. Hypothesis 2
We suggest that the evidence does not show that the Court has become more
activist over time and that there is an absence of an observable trend of either
increased or decreased activism.13 Manfredi and Kelly question this claim through the
use of three different descriptive statistics.
First, they note that of the five years in which the government win rate was
lowest, three of those years (1997, 2001, and 2002) occurred from 1996 to 2002.
They also note that the government win rate declined consecutively in 2001 and 2002.
Their suggestion is that this clustering is indicative of a potential upward trend in
activism. But the problem with their analysis is that the number of cases in each of
those years is quite small. In fact, the numbers are so small that if one government
loss were switched to a win, the government win rates in those years would be very
different. For instance, for 2001 and 2002, the win rates (which are the same
regardless of which coding rules are employed) would increase dramatically from
57.1 per cent to 71 per cent, and 45.5 per cent to 55.5 per cent, respectively. The
lesson is that it is hard to draw reliable conclusions about trends from very small
figures. Indeed, we explicitly note in our paper that the difficulty in finding clear
trends is in part due to the problem of the small numbers of cases in each year.14
Second, they classify the cases according to chief justice and note an increasing
trend of activism by chief justice. While this is an interesting approach to analyzing
the problem, the results do not strike us as particularly telling. While Manfredi and
Kelly note that activism has increased with each chief justice, the differences in the
numbers are too small to be meaningful. The range of government win rates over the
11 See Hunter v. Southam, [1984] 2 S.C.R. 145 at 155, 11 D.L.R. (4th) 641; R. v. Therens, [1985] 1
S.C.R. 613, 18 D.L.R. (4th) 655.
12 Win rates are a preferable measure because they are not skewed by the absolute number of
challenges under each time period.
13 Choudhry & Hunter, supra note 2 at 546.
14 Ibid. at 555.
MCGILL LAW JOURNAL / REVUE DE DROIT DE MCGILL
[Vol. 49
three chief justices since the advent of the Charter is from 66.0 per cent under Chief
Justice Dickson to 63.0 per cent under Chief Justice McLachlin, a difference of only 3
per cent. And the difference between Chief Justice Lamer and Chief Justice
McLachlin is only 0.6 per cent (from 63.6 per cent to 63.0 per cent), which hardly
seems like a difference worth noting.
Finally, Manfredi and Kelly calculate a cumulative win rate to suggest that from
the mid-1990s to 2002 activism has increased. This is an interesting analysis that we
had not previously undertaken, but we do not think that it yields “a different picture”.
The key problem is that the band of variation between the periods of low and high
activism on this statistic is also very narrow, ranging from a low government win rate
of 64.0 per cent in 1989 to a high government win rate of 68.1 per cent in 1996. It
seems to us that it is difficult to distinguish between a trend and random noise in the
data when the variation is so small. Manfredi and Kelly seem to recognize this as well
when they point out that “the data do indicate that judicial activism has reached an
equilibrium point.”15 At the end of the day, even if Manfredi and Kelly are right when
they suggest that, it is not “radically inconsistent” with our data to suggest that the
Court is becoming increasingly activist, it is equally fair to say that it is not radically
inconsistent with our data to suggest that the Court is not becoming increasingly
activist. This is precisely our point. The data do not show the clear trend that would be
necessary for critics of the Court to disprove the null hypothesis that there is no
change in judicial activism.
Next, Manfredi and Kelly probe our government learning hypothesis. They are
right to cite and discuss Kelly’s important work on the topic, which we mention in our
article. 6 They argue that we cannot measure the effect of government learning on
legislative drafting through government success in Charter litigation if the legislation
was drafted before the Charter came into force. Thus, an important variable is the
enactment date of the provision being challenged. One obvious cut off date is 1982,
when the Charter (save section 15) came into force. Another watershed date they
suggest is 1991, when pre-enactment Charter scrutiny was, according to Kelly,
institutionalized within the federal government. Using these cut off dates, Kelly
argues that of the thirty-nine pieces of federal legislation subject to constitutional
challenge which were enacted after 1982, twenty-eight (seventy-two per cent) were
enacted between 1982 and 1990. Kelly (and now Manfredi and Kelly) conclude from
this that one must only look to the post-1991 vetted statutes to find out whether the
government learning hypothesis is correct.
Now to be absolutely clear, this is not actually a critique of our article, but rather,
an attempt to refine a hypothesis we advance but do not explore because our article
was limited in its examination of judicial behaviour. It is just the right question to ask.
There are, however, two problems with the way Kelly has suggested answering it.
15 Manfredi & Kelly, supra note 1 at 750.
16 Choudhry & Hunter, supra note 2 at 548, n. 84.
2004]
S. CHOUDHRY & C. HUNTER -A REPLY TO MANFREDI & KELLY
771
First, his statistics only rely on challenges to federal laws. To provide a complete
answer to the question, he would need to generate comparable data for challenges to
provincial laws, adjusted to take into account the date of the institutionalization of
pre-enactment Charter review (if any) on a province-by-province basis. We
understand that Kelly is currently conducting the research that will enable him to
answer this question, and we look forward to the publication of his findings. But there
is a much more serious problem-the statistic that Kelly relies on does not enable him
to answer the question of whether pre-enactment scrutiny has lead to greater
government success in Charter litigation. To answer this question, he should have
compared the win rates for Charter challenges to statutes enacted between 1982 and
1990 and those enacted from 1991 onward. Only if these rates were significantly
different would there be proof of government learning. But instead, Kelly argues we
should compare the absolute numbers of statutes enacted between 1982 and 1990 and
from 1991 onward that were found unconstitutional. The problem is that this statistic
could be skewed by differences in the numbers of statutes under challenge from each
time period. Moreover, it could also reflect the fact that Charter challenges to federal
legislation enacted from 1991 onward have not yet reached the Court; for example,
the Court has just held the hearing in a Charter challenge to a ministerial decision
taken in 1994.17
C. Hypothesis 3
In our original article, we attack the claim that section 1 is a site of activism and
suggest that section 1 win rates have to be understood in their proper doctrinal
context-that is to say, when disaggregated by Charter right, win rates are a function
of whether the right in question contains internal limits. In response, Manfredi and
Kelly run two lines of argument. The first is that we ought to have further
differentiated within subsection 2(b) cases between those involving “core” free
speech that the Court subjects to stringent review under section 1 and those involving
“peripheral” speech that is subject to deferential section 1 review. According to
Manfredi and Kelly, our failure to do so means that we “may also be guilty of
designing a quantitative analysis around a deficient doctrinal understanding of Charter
adjudication,” which presumably masks higher government loss rates under section 1.
This is a good suggestion. Unfortunately, as Manfredi and Kelly say, it is a hypothesis
that their own anecdotal examples do not either prove or disprove. However, they
claim that this does not matter because they are making a larger point-that we
assume section 1 is applied in a “neutral” and “technical” manner by the Supreme
Court and does not involve “the discretionary choices of the judiciary.”‘ 8 That is a
provocative and important claim that has been the subject of an extensive legal
17 Hodge v. Canada (Minister of Human Resources Development), [2003] 1 EC. 271, 214 D.L.R.
(4th) 632, 2002 FCA 243, leave to appeal to S.C.C. granted, [2002] S.C.C.A. No. 345 (QL) (hearing
date 18 March 2004).
18 Manfredi & Kelly, supra note 1 at 753.
MCGILL LAW JOURNAL / REVUE DE DROIT DE MCGILL
[Vol. 49
literature, but it is not one we address in our paper. Our point was much narrower and
modest-namely, to demonstrate that descriptive statistics alone cannot prove the
claim made by Kelly in previously published work (as he and Manfredi acknowledge
in their response) that adjudication at the section 1 stage is a measure of judicial
activism.”9
In sum, even if we use the counting rules Manfredi and Kelly suggest, the
quantitative criticisms levelled at our paper leave it largely unscathed. Furthermore,
the alternative interpretations of the data they put forward are based on drawing
overly fine distinctions between rates that are very similar. It bears repeating that the
goal of our paper was to show that the quantitative measures of judicial behaviour that
have been used in the political science literature could not be relied on as proof of the
four hypotheses advanced by critics of the Court. The burden of our argument was
negative. In our view, the criticisms put forward in the Manfredi and Kelly comment
do not displace our general conclusion that the null hypothesis in each case has not
been disproven.
II. Qualitative Evidence
At numerous points, Manfredi and Kelly marshal qualitative evidence to refute
some of our hypotheses. In framing these arguments, they have misunderstood the
narrow and limited purpose of paper. As we said:
It is one of the great shortcomings of Canadian legal scholarship that the
legal academy has largely failed to engage with political scientists who make
empirical claims about judicial activism…. To be sure, the use of qualitative
evidence … is one way to respond. But detailed case studies should be
supplemented where possible with quantitative evidence of a systematic nature
to give a more comprehensive picture of the actual practice of Canadian
constitutionalism. Anecdote will no longer do.20
In other words, we never claimed that quantitative methods could ever tell the whole
story about the Court’s record under the Charter. And precisely because we are legal
scholars who study judicial reasoning, we spent several paragraphs explaining the
limitations of a strictly quantitative approach to the issue, at the end of which we
recognized that “it is important to not fall into the reductionist trap … to privilege
variables that can be measured with relative ease from the entire set of Charter cases
… ,21 But as Manfredi and Kelly’s own response demonstrates, the fact that numbers
alone do not tell the entire story does not mean that they do not matter at all. As we
said, “[j]ust as qualitative evidence on its own can provide only an incomplete
understanding of Canadian constitutional practice, so too is quantitative evidence
‘9 Ibid. at 758.
20 Choudhry & Hunter, supra note 2 at 530.
21 Ibid. at 534.
2004]
S. CHOUDHRY & C. HUNTER – A REPLY TO MANFREDI & KELLY
773
limited in its ability to shed light on the Court’s constitutional jurisprudence. ’22 Legal
scholars in Canada, however, have thus far worked largely in a qualitative mode. Our
desire was to participate in this debate using methods that most legal scholars
generally find alien.
But there is one argument made by Manfredi and Kelly relying on qualitative
evidence to which we feel compelled to respond-the claim that there is a link
between the delegitimization of the override and the level of judicial activism on the
Court. The delegitimization of the override is a familiar story. Like others, Manfredi
has traced the delegitimization of the override to the political outcry over its use by
Quebec in Bill 17823 to respond to the Court’s judgment in Ford v. Quebec (Attorney
General).24 Manfredi’s novel contribution
is to argue that the
delegitimization of the override has resulted in increased levels of judicial activism by
the Court. His argument is that the Court is a strategic actor whose judgments are
dependent on the external political environment, including the anticipated responses
of legislatures. Manfredi’s claim is that while the threat of override led the Supreme
Court to be cautious in its judgments in order not to provoke a legislative reversal of
its rulings prior to Bill 178, the delegitimization of the override has altered the
this external
institutional balance between courts and legislatures by removing
restraint on judicial review.
to this debate
fundamentally misunderstood Manfredi’s argument, because
In our study, we observed that there was no discernible trend in activism pre- and
post-Ford as measured in strictly quantitative terms. Manfredi and Kelly respond that
these
we have
descriptive statistics mask other ways in which activism has been on the rise. The
powerful comparison Manfredi (and now Manfredi and Kelly) deploys is the striking
contrast between “two cases involving divisive moral issues,”’25 Vriend v. Alberta26
and R. v. Morgentaler.27 In Morgentaler (a pre-Bill 178 case) the Court struck down
the Criminal Code’s abortion provisions28 on narrow procedural grounds that left
considerable room for Parliament to respond through more carefully tailored criminal
legislation and a reformed administrative apparatus. In Vriend (a post-Bill 178 case),
by contrast, “the Court left the Alberta legislature with almost no room to manoeuver
by reading sexual orientation into the province’s human rights legislation. ”
29
22 Ibid.
23 Bill 178, An Act to amend the Charter of the French Language, 2d Sess., 33d Leg., Quebec,
1988, cl. 10 (assented to 22 December 1988), S.Q. 1988, c. 54.
24 [1988] 2 S.C.R. 712, 54 D.L.R. (4th) 577 [Ford].
25 Manfredi & Kelly, supra note 1 at xxx.
26 [1998] 1 S.C.R. 493, 156 D.L.R. (4th) 385 [Vriend].
27 [1988] 1 S.C.R. 30,44 D.L.R. (4th) 385 [Morgentaler].
28 Criminal Code, R.S.C. 1970, c. C-34, s. 251
29 Manfredi & Kelly, supra note 1 at 760.
MCGILL LAW JOURNAL / REVUE DE DROIT DE MCGILL
[Vol. 49
One of us has critiqued this more sophisticated version of Manfredi’s argument in
a book review cited by Manfredi and Kelly.3″ Unfortunately, since they do not
describe this review in their response, we are forced to repeat and elaborate upon this
critique so that readers can draw their own conclusions as to the cogency of
Manfredi’s argument. As he and Kelly present it here, Manfredi’s argument is about
the rise of remedial activism on the Court, focusing on reading-in, the remedy
employed by the Court in Vriend. At first glance, reading-in does appear to be a very
intrusive remedy because it permits the Court to enter into the domain of legislative
drafting. By contrast, a declaration of invalidity (employed in Morgentaler) leaves the
framing and enactment of constitutionally compliant legislation
to legislatures.
Moreover, even though legislatures are free to depart from the Court’s remedy of
reading-in, it could be argued that judicially amended legislation enjoys the burden of
legislative inertia, which substantially increases the chance that it will not be
modified.
Reading-in has been used by the Court on three occasions-in Vriend, Miron v.
Trudel,3″ and R. v. Sharpe3 2-all of which were handed down post-Bill 178. This
evidence provides superficial support for Manfredi’s story. But a closer examination
of the legal and political context surrounding these cases reveals that reading-in does
not always reflect judicial aggressiveness. The best example for Manfredi is Vriend,
which held that the omission of sexual orientation as a prohibited ground of
discrimination in Alberta’s Individual Rights Protection Act33 was unconstitutional.
The three possible outcomes were: (a) a human rights code enumerating sexual
orientation as a prohibited ground of discrimination; (b) no code; and (c) a code
omitting sexual orientation, protected by the override. If we assume that the Court’s
desired outcome was (a), Manfredi is right to note that the Court’s remedy of reading-
in could be evidence of strategic behaviour, on the basis of a calculation that the
prospect of (a) was greatest when (a) was the status quo because of the burden of
legislative inertia (versus (b) and (c)), the need for some kind of human rights code
(versus (b)), and the delegitimization of the override (versus (c)).
But now compare Miron. In Miron, the Court read common law partners into the
definition of “spouse” in a provincial law prescribing that all automobile insurance
contracts provide benefits to the spouse of the policy holder, with “spouse” defined as
a husband or wife. In principle, there were three possible outcomes: (a) a definition of
spouse that included common law partners; (b) no law mandating accident benefits
for spouses; or (c) a law mandating accident benefits only for spouses in a marriage,
protected by the override. Manfredi could argue by analogy with Vriend that (a) was
30 Sujit Choudhry, Book Review of Judicial Power and the Charter: Canada and the Paradox of
Liberal Constitutionalism by Christopher P. Manfredi (2003) 1 Int. J. of Const. L. 379 at 385-89.
31 [1995] 2 S.C.R. 418, 124 D.L.R. (4th) 693 [Miron].
32 [2001] 1 S.C.R. 45, 194 D.L.R. [Sharpe].
3′ R.S.A. 1980, c. 1-2, now called the Human Rights, Citizenship and Multiculturalism Act, R.S.A.
2000, c. H-14.
2004]
S. CHOUDHRY & C. HUNTER – A REPLY TO MANFREDI & KELLY
775
the Court’s desired option and that reading-in was the way to secure it. But the history
of the provision under challenge tells a different story. The provision under challenge
was enacted in 1980. In 1990-after the accident that prompted the litigation took
place-the legislation was amended to expand the definition of spouse to include
common law spouses. The 1990 amendments were relied on explicitly by the majority
as a reason to read-in, as opposed to strike down. As Justice McLachlin, as she then
was, explained in her plurality judgment:
the fact that in 1990 the Ontario Legislature amended the eligibility criteria in a
way which would include the appellants, thus giving an indication of what it
would do if the matter were remitted to it anew. While this does not meet
concerns that the social and legislative picture may have changed further in the
years since 1990, or resolve the problem for the other Ontario statutes
containing similar provisions, it does offer reasonably conclusive evidence of
how the Legislature would have remedied the 1980 legislation had it been
required to do so when the appellants’ claim arose.34
In other words, the Court did not read-in on the basis of a calculation that the
delegitmization of the override would allow
it to outmanoeuvre the Ontario
legislature. Rather, it read-in because it was reasonable to assume that the legislature
would simply amend the legislation to include common law partners (i.e., that it
would have arrived at the same end point as the Court).
Finally, consider Sharpe, where the Court read-in defences to a law prohibiting
the possession of child pornography which it found unconstitutional because of
overbreadth. There were three possible outcomes: (a) a criminal prohibition narrowed
by defences that would survive Charter scrutiny; (b) no special criminal prohibition
on child pornography; and (c) an overbroad criminal prohibition, protected by the
override. Manfredi would argue that Sharpe is on all fours with Vriend, because (a)
was the Court’s desired option. But this analogy is simplistic. The use of the override
was actively discussed throughout the litigation culminating in the Sharpe decision.
There was a great deal of pressure on the federal government to invoke it had the
Court adopted the “deferential” remedy of striking down the provision in its entirety.
Ironically, it may well have been fear of the override that lead the Court to turn to the
superficially aggressive remedy of reading-in. To be sure, reading-in was likely a
strategic choice by the Court, but it was one made to allow Parliament to save face by
claiming the prohibition had been left largely intact without the need to invoke the
override.
So where does this leave us? On the broader issue of whether the Court acts
strategically in respond to the external political environment, we do not necessarily
disagree. Indeed, we think there is a lot to this view. One of us argues in a work-in-
progress that the Supreme Court of Canada acted strategically in the Quebec
Secession Reference to craft a road map to secession in face of the profound failure of
34 Miron, supra note 31 at 509. Justice McLachlin (as she then was) spoke for four members of the
Court. Justice L’Heureux-Dubd concurred with this aspect of her reasons (at 481).
MCGILL LAW JOURNAL / REVUE DE DROIT DE MCGILL
[Vol. 49
federal political institutions to do so.3 5 But on the narrow issue of whether the Charter
remedy of reading-in could be understood as a consequence, and as proof of, the
delegitimization of the override, the evidence turns out to be quite equivocal. To
reiterate a point we made in our article, we do not infer from all of this that the
override has not been delegitimized. As we carefully said in our article with respect to
our quantitative finding:
the absence of such a trend does not prove that the override has not been
delegitimized. In fact, this result is consistent with a number of different
hypotheses. Perhaps … the override has not been delegitimized. It is interesting,
for instance, to note that the override was used by the Alberta legislature in
March 2000
from
constitutional challenge. On the other hand, perhaps the override has been
delegitimized and replaced by other constraints to judicial activism in the
Canadian constitutional system, such as public opinion, to which the Court is
responding strategically. A poll taken in May 2002 reported that fifty-four
percent of respondents opposed the very idea of the override. Finally, perhaps
the override was and remains irrelevant to the adjudication of constitutional
cases.
the heterosexual definition of marriage
to shield
36
Conclusion: The Existing Literature, the Goals of Political Science,
and the Value and Limits of Quantitative Methods
We acknowledge that political scientists have a variety of views on the Charter
and Charter adjudication. Manfredi and Kelly are two cases in point. But in
recognizing the diversity of opinions among political scientists, one must not gloss
over the fact that many political scientists have made very negative assessments of the
Court’s record under the Charter, and have energetically and persistently advanced
these views in public debate. The leading example here is Ted Morton. Morton is a
well-known academic critic of the Court’s record under the Charter. But he has also
been very active politically and has been centrally involved in making judicial
activism a concern of the Conservative Party of Canada (and the Reform and then
Alliance Parties preceding it).37
Manfredi and Kelly suggest that we have constructed a straw man and proceeded
is constructed from
to knock him down. However, the so-called straw man
scholarship and commentary
in this area. Throughout the paper, we link our
hypotheses to claims found in the literature. The claim we examine in our first
hypothesis, that judicial activism is high in absolute terms, can be found in Morton
35 Sujit Choudhry, “Constitutional Moments in Translation? Ackerman’s Higher Lawmaking and
the Quebec Secession Reference”, which is part of a book project, tentatively entitled Searching for
Legitimacy through Legalism: The Constitutional Politics of Quebec Secession.
36 Choudhry & Hunter, supra note 2 at 554-55.
37 See e.g. E L. Morton, “Dialogue or Monologue?” Policy Options (April 1999) 23.
2004]
S. CHOUDHRY & C. HUNTER -A REPLY TO MANFREDI & KELLY
777
and Knopff’s The Charter Revolution and the Court Party.38 The third hypothesis,
that judicial activism is located in section 1, was a primary focus of Justice Marshall’s
reasons in NAPE. While Manfredi and Kelly are right to note that Justice Marshall did
not cite political scientists for this claim, we nonetheless thought it important to place
his claim in the context of the critical literature to illustrate how Justice Marshall’s
statements are part of a wider critique of the Court’s record under the Charter. Finally,
Manfredi himself has made the claims we examine in the second and fourth
hypotheses, namely that judicial activism is increasing over time and that increased
levels of activism are tied to the delegitimization of the override.39
So we did frame our response to political scientists who have been critical of the
Charter. And in framing our response, we addressed precisely the types of issues that
political scientists have found important. Unfortunately, Manfredi and Kelly stridently
disagree. They state that we “attempt … to disprove claims that, in the final analysis,
are at the periphery of our concern with the Supreme Court and the Charter.” We must
confess that we find this statement deeply puzzling because it contradicts Manfredi
and Kelly’s own definition of the basic concern of their discipline–“the institutional
relationships between actors that exercise power and to evaluate how this power is
used.””0 The power we are examining is the power of judicial review. The institutional
in
relationship we are examining is the interaction of legislatures and courts,
particular, the Supreme Court of Canada. In this interaction, many things matter. Ex
ante strategic behaviour is very important, and in our learning hypothesis, we suggest
as much. But surely the Court’s judgments that hold legislation to be unconstitutional
count as well-because political outcomes ought to count, and count centrally, for
political scientists. Put simply, it should matter whether governments are frustrated in
achieving their policy objectives by the courts, especially the highest court in the land.
We can look to the political science literature itself to see that these issues do in fact
matter to political scientists. In fact, we developed the idea for the original study not
only in response to NAPE but also after reviewing the critical literature. The principal
quantitative studies of the Court’s behaviour have been done only by political
scientists themselves.41 And we relied on studies by political scientists not only to
identify which questions to pose, but also to determine which methods to employ in
answering those questions. Indeed, the most recent study that uses quantitative
methods is by Kelly himself.42
38 F.L. Morton & Rainer Knopff, eds., The Charter Revolution and the Court Party (Peterborough,
Ont.: Broadview Press, 2000). See also Choudhry & Hunter, supra note 2 at 530, n. 25 and
accompanying text.
39 Choudh-y & Hunter, ibid. at 530, n. 26, 535, n. 41 and accompanying text.
40 Manfredi & Kelly, supra note 1 at 763.
41 EL. Morton, Peter H. Russell, & Michael J. Withey, “The Supreme Court’s First One Hundred
Charter of Rights Decisions: A Statistical Analysis” (1992) 30 Osgoode Hall L.J. 1.
42 James B. Kelly, “The Charter of Rights and Freedoms and the Rebalancing of Liberal
Constitutionalism in Canada, 1982-1997” (1999) 37 Osgoode Hall L.J. 625.
778
MCGILL LAW JOURNAL / REVUE DE DROIT DE MCGILL
[Vol. 49
One objective of our article was to spark a much-needed dialogue with political
scientists on the Supreme Court of Canada’s record under the Charter. The fact that
Manfredi and Kelly would respond to our article suggests that we were partially
successful. Unfortunately, we clearly have a long way to go before we engage each
other directly. We look forward to continuing the conversation.
