Article Volume 53:1

Regulating Impartiality: Electoral-Boundary Politics in the Administrative Arena

Table of Contents

Regulating Impartiality: Electoral-Boundary

Politics in the Administrative Arena

Ron Levy*

The author canvasses approaches

The author examines impartiality in cases of politically
contentious decision making. Many jurisdictions delegate
decisions over matters such as the establishment of fair election
ground rules to independent bodies. Some of these bodies,
including Canadas Federal Electoral Boundaries Commissions
(FEBCs), attract widespread trust and are by most accounts
substantially impartial. In contrast, commissions empanelled to
draw electoral boundaries in the United States, and to a lesser
extent in certain Canadian provinces, are often plagued by
partisanship.

to controlling
partisanship, relying on a series of interviews conducted with
boundaries commissioners and on interdisciplinary literature on
trust and trustworthiness in governance. Commentators often
favour bolstering formal constraints on FEBC discretion.
However, the author concludes that traditional administrative-
law models favouring such constraints are often inadequate. In
politically sensitive cases these methods frequently catalyze
partisanship. Proposals for more nuanced designdesign
sensitive to the complex interactions between law and
administrative culture in cases where the potential for
partisanship is highare better but rarer. The author focuses in
particular on the use of ambiguity in legal and institutional
design. Although this approach is counterintuitive in light of
rule-of-law assumptions favouring clarity, it has nevertheless
gained traction in commentary and has long been at work in
practice. The author argues that extensively ambiguous design,
as displayed by the complex federal readjustment processes in
Canada, has helped to develop the widely admired impartial
decision-making cultures of the FEBCs.

lectorales

Lauteur tudie limpartialit dans le contexte de la
rglementation dmocratique, qui dfinit les rgles de base de la
gouvernance dmocratique, et sattarde spcifiquement la
redfinition des frontires lectorales. Certaines institutions,
telles
les commissions canadiennes de dlimitation des
circonscriptions
fdrales, bnficient dune
confiance considrable du public et sont pour la plupart
substantiellement impartiales. Au contraire, plusieurs modes de
redfinition des frontires lectorales aux tats-Unis, et dans une
moindre mesure dans certaines provinces canadiennes, sont
entachs de partisannerie.
Sappuyant sur une littrature interdisciplinaire traitant de la
confiance et du mrite de la confiance dans la gouvernance, de
mme que sur des entrevues avec des commissaires canadiens
la dlimitation des circonscriptions lectorales fdrales, lauteur
tudie diverses approches visant contrler la partisannerie.
Plusieurs commentateurs ont rcemment insist sur la ncessit
de rformer les commissions canadiennes de dlimitation des
circonscriptions lectorales
les
contraintes formelles entourant leur pouvoir de discrtion.
Lauteur leur rpond en concluant que les modles traditionnels
de droit administratif permettant dinstaurer de telles contraintes
formelles se rvlent souvent inadquats.
En situation politiquement dlicate, ces modles catalysent

frquemment la partisannerie. Des propositions pour une
approche plus nuanceune approche sensible aux complexits
des interactions entre le droit et les cultures de service civil, o le
potentiel de partisannerie est levsavrent suprieures, mais
plus rares. Lauteur se penche en particulier sur le recours
lambigut dans la planification juridique et institutionnelle.

Bien que cette mthode soit surprenante et contre-intuitive
la lumire des assomptions de primaut du droit favorisant la
clart, elle a nanmoins gagn en importance dans la littrature et
existe depuis longtemps en pratique. Lauteur avance que cette
mthode fortement ambige a contribu dvelopper la culture
dcisionnelle
renomme des commissions
canadiennes de dlimitation des circonscriptions lectorales
fdrales.

impartiale et

fdrales en

renforant

* LL.M. (Columbia), Ph.D. (Osgoode). I researched and wrote this article as a Visiting Scholar at
Yale Law School and with the help of a Social Science and Humanities Research Council Canada
Graduate Scholarship. I am grateful to Federal Electoral Boundaries Commissioners Ritu Khullar,
Ron Landes, David Smith, and others preferring anonymity, and to Jean-Pierre Kingsley, former Chief
Electoral Officer (19902007), for sharing their experiences and insights. I also benefited from
valuable comments by and discussions with (in Canada) Harry Arthurs, Eric Nielsen, Bruce Ryder,
Andrew Sancton, David Smith, and two anonymous referees, as well as (in the United States) Heather
Gerken, Sarah Hammond, Samuel Issacharoff, Reva Siegel, Yale Law School graduate paper series
participants, and (from New Zealand) Andrew Geddis. I can be contacted at ron.levy@griffith.edu.au.

Ron Levy 2008
To be cited as: (2008) 53 McGill L.J. 1
Mode de rfrence : (2008) 53 R.D. McGill 1

MCGILL LAW JOURNAL / REVUE DE DROIT DE MCGILL

2

Introduction

I. Background: Two National Readjustment Systems

A. United States Readjustment
B. Canadian Readjustment
C. Impartiality as the Goal of Readjustment

II. Regulating Impartiality in Traditional Administrative Law

A. Impartiality in Traditional Administrative Law
B. The Case of Democratic Regulation
C. Traditional Administrative-Law Models and Democratic

Regulation

D. Departing from the Constraint Model

III. Regulating Norms of Impartiality in Democratic Regulation

A. Lessig and Ambiguity
B. Forms of Ambiguity

1. Procedural Ambiguity
2. Substantive Ambiguity
3. Role Ambiguity

C. The Uses of Ambiguous Decision-Making Systems

1. Preventing Fixation on Process
2. Broad Impartiality
3. Partisanship Preclusion
4. Developing Informal Decision-Making Norms

IV. The Scope and Limits of Ambiguous Systems

A. Extending Ambiguous Regulation
B. Conditions for Extended Ambiguity

1. Equivalence

a. Equivalence in Readjustment

2. Stability

a. Stability in Readjustment

3. Comprehensiveness

a. Comprehensiveness and Umpired Readjustment
b. Comprehensiveness and Decentralized
Readjustment

Conclusion

[Vol. 53

3

8
8
9
11

13
13
15

23
32
33
34
36
36
36
37
37
37
38
43
45
47
47
48
48
49
51
52
53
53

54

56

2008]

R. LEVY ELECTORAL-BOUNDARY POLITICS

3

Introduction
May 2003 saw a bizarre political development out of Texas as Democratic state
lawmakers boarded a plane to escape the jurisdiction. With a bill to redraw the states
federal congressional districts up for a vote in the Texas Legislature, Governor Rick
Perry had ordered police to bring Democrats in to fill a quorum.1 Partisan motivation
behind the new Republican-sponsored electoral map was evident. The proposed
boundaries concentrated Democratic voters inside relatively few districts. One
revised congressional district, the Texas 25th, extended over a thin and ragged column
from the city of Austin to the Mexican border five hundred kilometres south. The
district united two regions with high concentrations of African Americans, who are
known as dependable Democratic supporters.2 In addition, coming only two years
after the last boundary readjustment,3 the Texas gerrymander departed from the
normal ten-year revision cycle, which is timed to follow the release of census data.4 In
a final remarkable development, the readjustment plans mastermind, the Republican
leader of the United States House of Representatives, Tom DeLay, called on the
Department of Homeland Security and other federal offices to track the progress of
the Democratic jet as it passed into neighbouring Oklahoma.5 With the Texas plans
eventual passage, Republicans took five additional House seats in the 2004 election.6
A coherent historical narrative of American political partisanship is difficult to
trace. But the situation appears to date to the Republics origins, with periods of
relative quiescence and resurgence thereafter.7 The past two decades in particular
have, according to most accounts, been characterized by personal attacks and
polemical discourse.8 Commentators now often picture U.S. electoral politics as a

1 Jeffrey Smith, In Texas Feud, a Plane Tale of Intrigue: U.S. Role in GOP Hunt for Democratic

Lawmakers Is Still Murky Washington Post (7 June 2003) A1.

2 See James T. Campbell, Second Black Seat Tests Uneasy Coalition Houston Chronicle (15

March 2004) A18.

3 The process of redrawing electoral boundaries goes by readjustment in Canada and

redistricting in the United States; for simplicity, I will use only the former.

4 The Supreme Court of the United States recently invalidated selected parts of the plan, but ruled in
favour of the practice of early boundary revision (League of United Latin American Citizens v. Perry,
126 S. Ct. 2594 (2006) [Perry]).

5 Smith, supra note 1.
6 Linda Greenhouse, Supreme Court to Hear Dispute on Redistricting New York Times (13

December 2005) A1.

7 See Adam Cox, Partisan Fairness and Redistricting Politics, Comment, (2004) 79 N.Y.U.L. Rev.
751 at 772-73; Robert Putnam, Bowling Alone: The Collapse and Revival of American Community
(New York: Simon & Schuster, 2000) at 25, 31-47; Daniel Galvin & Colleen Shogan, Presidential
Politicization and Centralization across the Modern-Traditional Divide (2004) 36 Polity 477.

8 See David E. Lewis, Presidents and the Politicization of the United States Federal Government,
1988-2004 (Paper presented at the 2005 Annual Meeting of the American Political Science
Association, Washington, DC) [unpublished], online: Dartmouth College Department of Government
.

[Vol. 53

MCGILL LAW JOURNAL / REVUE DE DROIT DE MCGILL

4

bruising marketplace of political rhetoric led by image framers and computer-
modelling professionals.9

To some, Canada presents a sharp contrast. Much of the history of Canadian
politics, including the striking partisanship of the countrys first decades, belies any
simple comparison.10 Nevertheless, Peter Hoggs recent claim that Canadas political
culture is characterized by civility and moderation plausibly describes late
twentieth-century and early twenty-first-century Canadian politics.11 What, then, has
sent these two countries in sharply divergent directions in recent decades? In contrast
with the present landscape in the United States, federal electoral-district readjustment
in Canada takes place with little public contention and few suggestions of improper
political motivation.12 Organized by Elections Canada,
three
commissionersone set for each provincework mostly out of sight to redraw the
boundaries of federal ridings after each decennial census.13 This obscure
administrative process governs
for
nonpartisanship when it is thought of at all.

readjustment and enjoys a

ten sets of

reputation

9 See David Lublin & Michael McDonald, Is It Time to Draw the Line? The Impact of
Redistricting on Competition in State House Elections (2006) 5 Election L.J. 144 at 154; Mark
Monmonier, Bushmanders and Bullwinkles: How Politicians Manipulate Electronic Maps and Census
Data to Win Elections (Chicago: University of Chicago Press, 2001); Samuel Issacharoff,
Gerrymandering and Political Cartels (2002) 116 Harv. L. Rev. 593 at 624 [Issacharoff,
Gerrymandering]; Gary C. Jacobson, Partisan Polarization in Presidential Support: The Electoral
Connection (2003) 30 Congress & the Presidency 1.

10 See R. MacGregor Dawson, The Gerrymander of 1882 (1935) 1 Canadian Journal of
Economics and Political Science 197; Peter B. Waite, Chartered Libertine? A Case Against Sir John
Macdonald and Some Answers (197576) 3:32 Manitoba Historical Society Transactions, online:
Manitoba Historical Society ;
Terence H. Qualter, The Election Process in Canada (Toronto: McGraw-Hill, 1970); W.E. Lyons,
Legislative Redistricting by Independent Commissions: Operationalizing the One Man-One Vote
Doctrine in Canada (1969) 1 Polity 428 at 432.

11 Peter Hogg, Opening Remarks to Ad Hoc Committee on Supreme Court Appointment (27
February 2006), online: Department of Justice Canada . Hogg may be right about the politics of such tasks as judicial selections and
boundary readjustment. He presumably would distinguish these examples from others, such as debates
on the floors of legislatures, where aggressive and uncivil conduct are inevitably what comes of
politics … in a free society (David Smith, A Question of Trust: Parliamentary Democracy in Canada
Today (Distinguished Researcher Award Lecture, delivered at the University of Saskatchewan, 22
October 2003) at 4, online: University of Saskatchewan ).

12 See John C. Courtney, Commissioned Ridings: Designing Canadas Electoral Districts (Montreal:
McGill-Queens University Press, 2001) at 8, but notice infrequent exceptions at 144-49; Christopher
S. Elmendorf, Representation Reinforcement Through Advisory Commissions: The Case of Election
Law (2005) 80 N.Y.U.L. Rev. 1366 at 1393; Jean-Pierre Kingsley, The Administration of Canadas
Independent, Non-Partisan Approach (2004) 3 Election L.J. 406 at 406-07; Richard L. Hasen,
Beyond the Margin of Litigation: Reforming U.S. Election Administration to Avoid Electoral
Meltdown (2005) 62 Wash. & Lee L. Rev. 937 at 985.

13 Electoral Boundaries Readjustment Act, R.S.C. 1985, c. E-2, ss. 4-6 [EBRA].

5

R. LEVY ELECTORAL-BOUNDARY POLITICS

2008]

This article will look at the problem of democratic regulationthe
interpretation and application of the ground rules of political power in a democracy
and will turn its attention to Canadas electoral-district readjustment regime in
particular. Successes in the Canadian context are striking in light of the pronounced
risk of partisan conflict. Electoral-district readjustment featured vigorous attempts to
gerrymander for a full century before the advent of Canadas Federal Electoral
Boundaries Commissions (FEBCs).14 Overtly partisan decision making was thus
previously the rule and remains so in the United States and, to a lesser extent, in some
provinces with commissions of their own.15 In drawing electoral regions or defining
other rules of politics, we encounter the risk of entrenchment. Lawmakers who
control the rules of their own election can chok[e] off the channels of political
change16 to remain in office against the preferences of their constituents.17
Democratic regulation therefore raises an acute challenge for institutional design:
which systems of decision making will best safeguard impartiality despite the
outsized incentives partisans face to manipulate procedures and lock in their own
authority? The international and provincial comparisons suggest a significantly more
complex set of factors at work than the bare presence of commissions or the
presumptively mild Canadian political culture.
Canadas approach of delegating readjustment to independent administrative

bodies is a subject of rising interest in U.S. legal commentary. Some American
scholars have looked for comparative insights from Canada, the United Kingdom,
and Australiaall countries with independent democratic regulators widely viewed
as impartial.18 Like other authors, I adopt a comparative perspective. However, I am
not primarily interested in further lessons for reforming the U.S. regime of democratic
regulation, but rather focus on what such comparisons recommend for the future of
Canadian democratic regulation. I begin in broad agreement with Hogg. He appears
right to characterize cultures of civility and moderation as among the foremost
safeguards of impartiality in Canadian democratic regulation. However, how
pervasive are these elements of Canadas political culture? As the brief history
outlined above suggests, the impartial readjustment system is a relatively recentand
perhaps tenuousdevelopment. Indeed, a number of social scientists describe
marked Canadian trends of intensifying political contest and declining public trust in

14 See Lyons, supra note 10 at 429.
15 See text accompanying notes 138, 139.
16 John Hart Ely, Democracy and Distrust: A Theory of Judicial Review (Cambridge, Mass.:

Harvard University Press, 1980) at 103.

17 Michael J. Klarman, Majoritarian Judicial Review: The Entrenchment Problem (1997) 85 Geo.
L.J. 491 at 498. See also Sam Hirsch, The United States House of Unrepresentatives: What Went
Wrong in the Latest Round of Congressional Redistricting (2003) 2 Election L.J. 179 at 179; Colin
Feasby, Freedom of Expression and the Law of the Democratic Process (2005) 29 Sup. Ct. L. Rev.
237.

18 See Elmendorf, supra note 12 at 1386-1405; Hasen, supra note 12 at 945, 983-90.

[Vol. 53

MCGILL LAW JOURNAL / REVUE DE DROIT DE MCGILL

6

institutions.19 Do informal norms of political culture then offer a reliable bulwark
against partisanship? Will they continue to do so given the present atmosphere
favouring reform of public administration in general20 and readjustment bodies in
particular?21 These risks take on additional relevance in light of Canadian reformers
inclination to compare and to borrow from the regulatory tools of Canadas very
visible neighbour.22

Responding to these concerns, this article will assess the use of administrative
decision making to solve partisanship problems and will ask two main questions.
First, many recommendations for the reform of readjustment draw on a model of
formal legal constraint23 on partisan behaviour, but is this approach adequate?
Institutional and legal design based on a constraint model subjects decision makers to
a host of processes and formal rules in order to limit discretion and the potential for
its abuse. Oversight bodies therefore closely scrutinize decision making. Additionally,
numerous clear rules lay out conditions for discretions exercise. A question that
remains wide open is how well traditional scholarship on administrative impartiality,
most often rooted in constraint models, applies to democratic regulation. I will assert
that the traditional scholarship must redirect itself in order to address the special
problems raised by administrative bodies as democratic regulators. Proposals for
more nuanced designdesign sensitive to the complex interactions between law and
administrative culture in the context of democratic regulationare better but rarer.

19 David E. Smith, A Question of Trust: Parliamentary Democracy and Canadian Society (2004)
27 Canadian Parliamentary Review 24 at 24-27 (on more polemical Canadian electoral politics); ric
Blanger & Richard Nadeau, Political Trust and the Vote in Multiparty Elections: The Canadian
Case (2005) 44 European Journal of Political Research 121 at 121-25 (on declining Canadian
political trust); Margaret Levi & Laura Stoker, Political Trust and Trustworthiness (2000) 3 Annual
Review of Political Science 475 at 481-83 (citing global declines). Analysts show how trust and
trustworthiness interdepend, the latter in part a consequence of the self-fulfilling effects of the former.
See John Braithwaite, Institutionalizing Distrust, Enculturating Trust, in Valerie Braithwaite &
Margaret Levi, eds., Trust and Governance (New York: Russell Sage Foundation, 1998) 343 at 344-
47.

20 See e.g. Commission of Inquiry into the Sponsorship Program & Advertising Activities, Restoring
Accountability, vols. 1-4 (Ottawa: Public Works and Government Services Canada, 2006) [Restoring
Accountability]; Harper Promises New Era of Accountability CBC News (1 February 2006),
online: CBC News [New Era]; David J. Mullan, Administrative Law (Toronto: Irwin Law, 2001), c. 23.C;
Donald J. Savoie, Breaking the Bargain: Public Servants, Ministers and Parliament (Toronto:
University of Toronto Press, 2003).

21 See Part II, Section C.
22 Lorne Sossin, Boldly Going Where No Law has Gone Before: Call Centres, Intake Scripts,
Database Fields, and Discretionary Justice in Social Assistance (2005) 42 Osgoode Hall L.J. 363 at
402-03 [Sossin, Boldly Going].

23 This term appears across several areas of administration and democratic regulation. See e.g.
Bruce Ackerman, The New Separation of Powers (2000) 113 Harv. L. Rev. 633 at 640 (advocating
the creation of a democracy branch for democratic regulation); Denis Galligan, Discretionary
Powers: A Legal Study of Official Discretion (Oxford: Clarendon Press, 1986) at 20 (on judicial
review of administrative power generally).

7

R. LEVY ELECTORAL-BOUNDARY POLITICS

2008]

Though primarily a work of administrative-law scholarship, this article will address a
special case at the intersection of administrative law and political theory, with
implications that remain underexamined despite their importance.24 Impartial
democratic regulation relies, I will assert, not only on formal constraint rules but also
on informal norms of trust and trustworthiness in a decision-making culture.25
Impartiality, which I will define in detail below, is an instance of trustworthiness.
A second question follows from the first. If informal norms are responsible for
much of the work of political administration, then securing impartiality is particularly
fraught. Such norms do not readily allow for direct legal manipulation. What, then,
are our options for regulating the impartiality of democratic administrators? To
address this question, I will begin with the work of Lawrence Lessig, who some years
ago offered important additions to the tool kit of cultural regulation. Lessig discussed
how lawmakers and institutional designers can shape informal norms to pursue
discrete policy goals.26 But a weakness in Lessigs approach lies in the restricted set
of contexts to which he found his own work applyinga set that seems to exclude
problems as complex as those of impartiality and entrenchment. I will examine one
particular method to which Lessig makes only tantalizingly cursory reference: the use
of ambiguity in law.27 In the rich and growing social-science literature on public trust
and institutional trustworthiness, a subset of works describes types of institutions and
law that have tended to promote trust and trustworthiness, and a number of
contributors have begun examining the role of ambiguity. I will draw together these
disparate contributions and outline how ambiguous institutional and legal design can
be especially useful in the context of democratic regulation and entrenchment.

In this article, then, I will look to electoral-boundary readjustment in practice to
suggest how ambiguity helps sustain impartiality in Canadas federal readjustment
commissions. To be sure, regulatory ambiguity is only one of several plausible
explanations. However, this underexamined approach appears to have played a
leading role in the successes of Canadian administrative readjustment. A decision-
making culture characterized by substantial impartiality developed quickly after the
commissions entry into the Canadian political landscape.28 Readjustment thus serves
as a rich setting in which to explore the regulation of impartialitywith cautionary
American and provincial comparisons outlining the risks that attend often well-

24 Elmendorf expresses surprise at a dearth of scholarship on this issue (supra note 12 at 1376).
25 For the sake of simplicity, I will use trust and trustworthiness to stand in for several similar or
related terms in the literature, such as civility and moderation and social capital. See generally
Hogg, supra note 11; Jason Mazzone, The Social Capital Argument for Federalism (2001) 11 S.
Cal. Interdisciplinary L.J. 27 at 57-58; Putnam, supra note 7 at 19 (social capital refers to
connections among individualssocial networks and the norms of reciprocity and trustworthiness that
arise from them); Richard H. Pildes, The Destruction of Social Capital Through Law (1996) 144
U. Pa. L. Rev. 2055.

26 Lawrence Lessig, The Regulation of Social Meaning (1995) 62 U. Chicago L. Rev. 943.
27 Ibid. at 1015.
28 See Lyons, supra note 10 at 458-59.

[Vol. 53

MCGILL LAW JOURNAL / REVUE DE DROIT DE MCGILL

8

intentioned reforms. In one commissioners words, the Canadian systems recognized
impartiality is premised on ambiguity and on the mythology … of the neutral fair
system.29 But this state of affairs is fragile and [t]he whole premise can fall apart
pretty quickly.30

The article will proceed as follows. Part I will provide an overview of
readjustment procedures in the United States and Canada. Part II will critique the
reliance on constraint that prevails in proposed reforms of electoral-boundary
readjustment in recent years. The next two parts will then look more closely at the
two national readjustment systems, examining how different institutional models
impact impartiality. Part III will outline alternatives to constraint in institutional
design, focusing in particular on ambiguity. Finally, Part IV will consider the limits of
ambiguity approaches, detailing the range of cases in which these methods can
improve the impartiality of democratic regulation.

I. Background: Two National Readjustment Systems

A. U.S. Readjustment

In the United States, systems for drawing electoral districts for the federal House
of Representatives vary from state to state. State legislatures most often readjust
federal boundaries through normal acts of legislation.31 However, some states give
commissions control over readjustment. These commissions are often politicized:
membership is usually split between commissioners affiliated with the United States
two main political parties, either equally or in proportion to the parties legislative
representation, and commissioners typically promote the interests of their party.32 A
few apparently non-partisan state commissions offer intriguing exceptions.33

There is federal administrative oversight of readjustment to the extent that the
Voting Rights Act of 196534 gives a group of lawyers in the United States Justice
Department the task of preclearing proposed electoral-rule changes.35 The lawyers vet
the changes, looking for deleterious impacts on the voting patterns of historically

29 Interview of Ritu Khullar, Federal Electoral Boundaries Commissioner (Alberta) (28 September
2006) [unpublished, on file with author]. See also Interview of Ron Landes, Federal Electoral
Boundaries Commissioner (Nova Scotia) (24 July12 September 2006) [unpublished, on file with
author]; Interview of David Smith, Federal Electoral Boundaries Commissioner (Saskatchewan) (9
August 2007) [unpublished, on file with author]; Interviews of four Federal Electoral Boundaries
Commissioners preferring anonymity (200607) [unpublished, on file with author].

30 Khullar, ibid.
31 See generally Michael P. McDonald, A Comparative Analysis of Redistricting Institutions in the

United States, 200102 (2004) 4 State Politics and Policy Quarterly 371.

32 Ibid.
33 See ibid.
34 42 U.S.C. 1973c (2000) [VRA].
35 Ibid., s. 5.

9

R. LEVY ELECTORAL-BOUNDARY POLITICS

2008]

disadvantaged groups.36 The preclearance process applies to nine states in their
entirety, most of which are in the South, and to parts of seven more states scattered
across the country.37 Justice Department lawyers can approve the proposed changes or
refer cases deemed suspect for final determination by the courts. Under a separate
process of the VRA, litigants in all states can refer cases to the federal courts without
Justice Department involvement.38

Litigation occurs often.39 The Supreme Court of the United States laid down new
gerrymandering tests in every decade from the 1960s to the 2000s. Beginning with
the establishment of justiciability40 and the rule of one person, one vote,41 the court
repeatedly augmented or replaced earlier rules. Newer rules have included a strict test
for declining numbers of elected representatives from minority groups,42 a prohibition
on any district bizarre on its face,43 and a test for declining minority-group
representation in light of the totality of the circumstances.44 However, the case law
and the power to enforce the VRA rest on the equality protections of the
Constitutions Fourteenth Amendment45 rather than any wider-ranging guarantee of
impartiality.46

B. Canadian Readjustment

Canada has adopted a different set of legal solutions. At the federal level, the
Electoral Boundaries Readjustment Act,47 certain constitutional provisions,48 a
handful of court decisions,49 and assorted customary rules50 govern readjustment. An

36 See ibid.
37 28 C.F.R. 51, App. (2007).
38 Supra note 34, s. 2.
39 See Pamela S. Karlan, The Fire Next Time: Reapportionment After the 2000 Census (1998) 50

Stan. L. Rev. 731 at 762 [Karlan, Fire].
40 Baker v. Carr, 369 U.S. 186 (1962).
41 Reynolds v. Sims, 377 U.S. 533 (1964).
42 Beer v. United States, 425 U.S. 130 at 208ff. (1976).
43 Shaw v. Reno, 509 U.S. 630 at 644 (1993).
44 Georgia v. Ashcroft, 539 U.S. 461 at 480 (2003).
45 U.S. Const. amend. XIV.
46 See Issacharoff, Gerrymandering, supra note 9 at 598, 600.
47 Supra note 13.
48 See Constitution Act, 1867 (U.K.), 30 & 31 Vict., c. 3, s. 8 (requiring a census of the population of
Canada in 1871 and every ten years thereafter), s. 40 (setting out initial ridings at Confederation), s. 51
(requiring readjustment of House of Commons seat totals, and apportionment to the various provinces,
following the decennial census; therefore implicitly mandating electoral map adjustment), s. 51A
(requiring that every province have no fewer House of Commons seats than it has Senate seats), s. 52
(granting Parliament power to expand the total number of House of Commons seats), reprinted in
R.S.C. 1985, App. II, No. 5.

49 For a discussion of court decisions, see text accompanying note 143.

there

[Vol. 53

MCGILL LAW JOURNAL / REVUE DE DROIT DE MCGILL

10

administrative body, Elections Canada, has jurisdiction over most electoral issues,
from readjustment to the oversight of campaigns.51 The thirty members of the FEBCs
do the work of readjustment across the country, guided at arms length by Elections
Canada. For every province
is one commission made up of three
commissioners. The provinces chief justice appoints a judge to chair the commission,
while the Speaker of the House of Commons appoints the other two members from
among the provinces residents. Once constituted, each commission prepares
proposals while drawing on Elections Canada resources. Notices invite public and
parliamentary participation in open hearings after publication of the proposals. This
period sees (a) the Speaker of the House of Commons table the proposed ridings in a
committee of the House of Commons, (b) a thirty-day period of availability for
review by MPs and the filing of objections, (c) a further thirty-day period of evidence
collection and review by the committee, (d) the return of the proposal to the Chief
Electoral Officer and thereafter the provincial commissions, and (e) a final thirty-day
period during which the commissions render their final decisions independently of,
but with attention to, Parliament and any public input that was received.52 At the
conclusion of hearings commissioners usually modify
their proposals. The
commissions then enact the final boundaries by regulation.53
While a handful of court decisions in Canada address boundary drawing, few
spell out substantive criteria for readjustment.54 The number of all readjustment cases
is small in comparison with American litigation, even allowing for differences of size
between the two countries. The leading Canadian case is the Supreme Court of
Canadas Reference Re Prov. Electoral Boundaries (Saskatchewan),55 which
considers the right to vote under Section 3 of the Canadian Charter of Rights and
Freedoms56 and outlines criteria for electoral-boundary drawing. Criteria include
geography, community history, community interests and minority representation,57
as well as practical considerations like the difficulty of legislators in fulfilling their
ombudsman problem-solving roles across vast northern ridings.58 This list is not
closed,59 and the deeper set of values that govern the process include effective

50 See e.g. Louis Massicotte, Establishing the Rules of the Game: Election Laws in Democracies
(Toronto: University of Toronto Press, 2004) at 97-98 (outlining customs such as joint selection of the
chief electoral officer by all political parties).

51 Canada Elections Act, S.C. 2000, c. 9, ss. 13-21 [CEA]. The CEA creates the offices of the Chief
Electoral Officer and his or her staff (ibid.). Much of the remainder of the CEA details their
responsibilities.

52 EBRA, supra note 13, ss. 3-6, 14, 19-25.
53 Ibid., ss. 24-25.
54 See text accompanying notes 141-43.
55 [1991] 2 S.C.R. 158, (sub nom. Reference re Electoral Boundaries Commission Act) 81 D.L.R.

56 Part I of the Constitution Act, 1982, being Schedule B to the Canada Act 1982 (U.K.), 1982, c. 11

(4th) 16 [Carter cited to S.C.R.].

[Charter].

57 Carter, supra note 55 at 184.
58 Ibid. at 187-88.
59 Ibid. at 184.

11

R. LEVY ELECTORAL-BOUNDARY POLITICS

2008]

representation,60 as well as respect for the inherent dignity of the human person,
commitment to social justice and equality, respect for cultural and group identity, and
faith in social and political institutions that enhance the participation of individuals in
society.61 The EBRA lays out some similar substantive guidelines for federal
readjustment. The rules to be followed include the community of interest or
community of identity in or the historical pattern of an electoral district in the
province … and a manageable geographic size for districts in sparsely populated, rural
or northern regions of the province.62 These rules are both general readjustment
criteria and the criteria that govern departures from the statutes one person, one vote
default where the commission considers such a departure necessary or desirable.63

C. Impartiality as the Goal of Readjustment

From these background notes, a distinction between the aims of each national
system begins to emerge. Impartial judgment is a primary goal of electoral-boundary
readjustment schemes in both Canada and the United States.64 However, each country
leans toward one of two separate notions of impartiality.65 The first notion is one of
balance. Impartiality in U.S. democratic regulation is most often understood as the
product of finely tuned tensions, between the two pre-eminent political parties66 or
between contending social factions, especially white versus black or Latino.67 An
impartial decision is therefore generally an average or a compromise between
opposing partisan groups. This generalization appears to hold true in most but not all
states.68

60 Ibid. at 172.
61 Ibid. at 188.
62 Supra note 13 at s. 15(2)(a, b).
63 Ibid. Except in exceptional circumstances, no district should range beyond plus or minus 25 per
cent of the average size of all districts. Note that in practice in a given district, deviation from the
average is the rule (Landes, supra note 29).

64 Issacharoff notes that the gerrymandering doctrines elaborated by the Supreme Court of the
United States, while premised on constitutional anti-discrimination doctrines, are driven by the Courts
broader but inchoate sense of the unfairness of entrenchment (Issacharoff, Gerrymandering, supra
note 9 at 598, 600). Partisan entrenchment is an extreme example of partiality.

65 This is a useful but simplifying dichotomy necessarily omitting other potential definitions.
66 Often we see assertions such as that of OConnor J. in her concurring opinion in Davis v.
Bandemer: There is no proof before us that political gerrymandering is an evil that cannot be
checked or cured by the people or by the parties themselves (478 U.S. 109 at 152 (1986)
[Bandemer]). Thus even when impartiality appears for the moment to have been abandoned in
practice, the opposing party or electorate will perhaps even the score in time. The success of
Democrats in the 2006 mid-term election offers some support for this perspective.

67 See Issacharoff, Gerrymandering, supra note 9 at 598, 600.
68 See generally McDonald, supra note 31. As I will show, impartiality as balance pervades most
American scholarship on readjustment. However, readjustment in certain exceptional states is in
practice more in line with the second notion of impartiality.

[Vol. 53

MCGILL LAW JOURNAL / REVUE DE DROIT DE MCGILL

12

A second understanding of impartiality is more robust, if harder to conceptualize
and put into practice. Impartiality according to this definition means a lack of
attachment to or predisposition toward an interest, ideology or social or political
faction.69 In addition, impartiality in this sense is usually broader-ranging. It implies a
lack of partiality not only along one or a few dimensionssuch as Democratic and
Republicanbut along many. For example, Canadas FEBC process apparently
avoids no particular form of partiality, but is rather geared toward decision making
that is generally without favourfor example, to a given party, ideology, region, or
ethnic or linguistic group.70

The latter of the two notions is not only more robust, but perhaps also truer to
what we usually mean in speaking of impartiality. One might question whether
decisions emerging from highly polarized decision-making bodiesbodies where
partial decision makers serve as mutual counterweightsare properly understood as
impartial at all.71 In these cases, individual decision makers are certainly not
conceived of as impartial, though their collective decisions sometimes are. The notion
of a compromise between factions suggests a negotiated result. In contrast, the ideal
of broad impartialityand it is surely only an ideal72implies rational judgment,
not arbitrary agreement between partial decision makers.73 On the broader of the two
notions, impartial decision making takes into account most of the relevant

69 There are many formulations along these lines, citing impartiality as avoidance by decision
makers of politics, policy, partisanship, and personal preference. See e.g. Hasen, supra note 12 at 979,
982; Jeffrey Rosen, The Most Democratic Branch: How the Courts Serve America (Oxford: Oxford
University Press, 2006) at 8, 10 [Rosen, Democratic Branch]; Kenneth Kernaghan, Political Rights
and Political Impartiality: Finding the Balance Point (1986) 29 Canadian Public Administration 639;
Osborne v. Canada (Treasury Board), [1991] 2 S.C.R. 69, 82 D.L.R. (4th) 321; Lublin & McDonald,
supra note 9 at 156. Note that neutrality often connotes something similar.

70 This calls for some further explanation. The major point is that broader-ranging impartiality
avoids favouring not one or two, but several or many social groupings. However, in Canada, regions
and ethnic and linguistic groupings are in fact important in the decision-making process; as will be
seen, they are often kept intact when electoral boundaries are drawn. But the regime gears toward
preserving such groups generally, rather than favouring one group over others.

71 An example is the Supreme Court of the United States, whose members frequently organize into
two camps separated by a wide ideological gulf (Frank I. Michelman, Suspicion, or the New Prince
(2001) 68 U. Chicago L. Rev. 679 at 680 [Michelman, Suspicion]).

72 The FEBCs are inevitably imperfect. However, as discussed in this article, having chosen the
more idealistic of two potential meanings of impartiality, Canadas FEBCs are substantially impartial
in practice.

73 In concrete terms, evenly divided partisan commissioners commonly draw boundaries that
maintain the status quo in the legislature (e.g. 40 Democrats, 40 Republicans). See generally
McDonald, supra note 31. Alternatively, a slight majority of one party on a commission can push
through significant changes (now e.g. 45 Democrats, 35 Republicans). Such variability demonstrates
how decision making by partisan balance can be arbitrary, in the sense that community history or
other rational readjustment criteria play a reduced role.

13

R. LEVY ELECTORAL-BOUNDARY POLITICS

2008]

perspectives and substantive factors bearing on an issue.74 An ethic of broadly
impartial decision making entails, at a minimum, that decision makers commit to
hearing and rationally considering the full range of relevant facts and arguments.
Unless otherwise indicated, I will use the broader meaning of impartiality from here
on.
Neither of these definitions is uncontroversial.75 However, they provide some
context as we turn to the approaches Canada and the United States have adopted in
efforts to ensure impartial readjustment.

II. Regulating Impartiality in Traditional Administrative Law
Administrative law and scholarship recognize several safeguards for impartiality.
These safeguards appear substantially unaltered in much of the literature on
readjustment administered by commissions. But as I will demonstrate in this part, the
traditional safeguards are insensitive to or muddle the distinctive problems of
administration in the context of democratic regulation.

A. Impartiality in Traditional Administrative Law

Administrative law and commentary commonly make reference to at least six
impartiality safeguards. (1) Transparency. Open views of the decision-making
processes of agencies, commissions, and the like are thought necessary for
impartiality.76 Abuses of power might be less likely under the light of scrutiny.77 And
revealing internal procedures can lead to more rational decision making.78 (2) Public
Participation. A right of members of the public with relevant interests to speak at
hearings is understood in part as an impartiality safeguard. The contributions of
individuals and groups from civil society can enhance the impartiality of procedures

74 See Jennifer Nedelsky, Embodied Diversity and the Challenges to Law (1997) 42 McGill L.J.
91 at 107; Baker v. Canada (Minister of Citizenship and Immigration), [1999] 2 S.C.R. 817 at paras.
47-48, 174 D.L.R. (4th) 193 [Baker cited to S.C.R.].

75 Of the various kinds of partialities that broad impartiality, as noted above, might lack, ideology is
the most difficult and contentious. On a basic level, ideology is unavoidable because laws express
ideologies, as do the character and the constitutionin all senses of the wordof a state. However,
impartiality means avoiding, in particular, narrow or rigid ideology. The primary ideology to which
impartial decision makers should be committed is intellectual openness (Nedelsky, ibid.).

76 See Keith Hawkins, The Use of Legal Discretion: Perspectives from Law and Social Science in
Keith Hawkins, ed., The Uses of Discretion (Oxford: Clarendon Press, 1992) at 11; Restoring
Accountability, supra note 20; Baker, supra note 74 at paras. 35-44; Mullan, supra note 20, c. 23.

77 See Northwestern Utilities Limited v. Edmonton (City of), [1979] 1 S.C.R. 684 at 706, 89 D.L.R.

(3d) 161; Baker, supra note 74 at para. 38.

78 See Williams v. Canada (Minister of Citizenship and Immigration), [1997] 2 F.C. 646 at para. 38,
147 D.L.R. (4th) 93 (C.A.) [Williams cited to F.C.]; Baker, ibid. at para. 39; Mullan, supra note 20 at
c.13.O.3.

[Vol. 53

MCGILL LAW JOURNAL / REVUE DE DROIT DE MCGILL

14

by broadening the range of views available to decision makers.79 Public participation
also supports transparency.80 (3) Representativeness. According to some views,
decision makers should reflect in their own ranks the distribution of perspectives or
identities in the society they serve. Broadly representative bodies accommodate
interests beyond those of the mainstream or elites.81 (4) Independence. Impartiality is
commonly understood as requiring administrators to decide in independent settings,
without fear of sanction or other interference from influential partisans.82 In addition,
the rule that no one should be a judge in their own cause disqualifies decision makers
who have an interest in [a] case through a social, economic, or political relationship
with one of the parties to the dispute.83 (5) Reconsideration and Review. Recourse to
further decision making by the same or different decision makers, allowing
consideration of claims of bias or denial of other procedural guarantees, is a further
impartiality safeguard.84 (6) Rational Organization, Simplicity, and Clarity. The rules
guiding administrative discretion should, a final argument goes, be rationally
organized, simple, and clear; power should be centralized and coordinated and there
should be a framework of constraining principles85 at work. In contrast, relatively
chaotic administration obscures the channels of, and responsibility for, decision-
making. Clarity and coherence, cost-benefit analyses, and top-down control produce
rational arrays of rules and accountable decision-making.86

79 Participatory rightsfor example, rights to notice and an opportunity to respondcan ensure that
decision makers hear from all affected individuals. In courts and some administrative tribunals, the
natural justice rule of audi alteram partem requires, inter alia, on notice, an open hearing before the
decision-maker (Hon. Hugh F. Landerkin, Custody Disputes in the Provincial Court of Alberta: A
New Judicial Dispute Resolution Model (1997) 35 Alta. L. Rev. 627 at 654). See also Mullan, ibid. at
675; Frank I. Michelman, Brennan and Democracy: The 1996-97 Brennan Center Symposium
Lecture (1998) 86 Cal. L. Rev. 399 at 423 [Michelman, Brennan].

80 See Jean-Luc Bilodeau, La participation des administrs llaboration des normes de porte
gnrale dans le cadre de la rgulation du march des valeurs mobilires (1999) 33 R.J.T. 449 at para.
30.

81 See Elena Kagan, Presidential Administration (2001) 114 Harv. L. Rev. 2245 at 2264-66. Note

that representativeness may conflict with independence and the rule of nemo judex in parte sua.

82 See Lorne Sossin, Speaking Truth to Power? The Search for Bureaucratic Independence in
Canada (2005) 55 U.T.L.J. 1 at 2, 6-15, 19-24 [Sossin, Truth to Power]; International
Woodworkers of America, Local 2-69 v. Consolidated-Bathurst Packaging Ltd., [1980] S.C.R. 282 at
332-33, 68 D.L.R. (4th) 524.

83 K.D. Ewing, A Theory of Democratic Adjudication: Towards a Representative, Accountable and
Independent Judiciary (2000) 38 Alta. L. Rev. 312 at 314 (on the rule of nemo judex in parte sua).
This is also a rule of administrative justice commonly known as the rule against bias (Landerkin,
supra note 79 at 654).

84 See Mullan, supra note 20 at 675.
85 Denis Galligan, Discretionary Powers: A Legal Study of Official Discretion (Oxford: Clarendon
Press, 1986) at 20. See also Kenneth C. Davis, Administrative Law Text, 3d ed. (St. Paul, Minn.: West,
1972) at 226.

86 See Sossin, Truth to Power, supra note 82 at 2 (civil service rules, principles and conventions
should be more fully elaborated); Richard H. Pildes & Cass R. Sunstein, Reinventing the
Regulatory State (1995) 62 U. Chicago L. Rev. 1 at 8.

2008]

R. LEVY ELECTORAL-BOUNDARY POLITICS

15

B. The Case of Democratic Regulation

The impartiality safeguards noted above, all based on a model of formal
constraint of discretion, apply in a straightforward manner to much of the normal
business of administrative regulation, such as standard making or adjudication on the
environment, immigration and refugee claims, health, and much more. This changes
when administrative decision makers are not just another set of organs elaborating
and implementing the policies of government, but are regulators of government itself.
It is important to consider exactly how thinking about impartiality should change
when administrators serve in this way as branches of government; though we
commonly call them governmental branches,87 administrators fully assume this status
only when they are integral units in a system of mutual regulationcapable of
directing power and not merely receiving direction.88 By regulating politics,
administrators cease being discrete decision makers hived off from more senior power
holders and instead share space at the peak of governmental hierarchy.89
How, then, should the constellation of impartiality safeguards outlined above
differ in the case of democratic regulation? As already noted, some observers view
informal norms of trust and trustworthiness as the indispensable safeguards of
impartiality in democratic regulation. Trust at its most basic level is an expectation of
good conduct by decision makers.90 The expectation can be held by an individual or
in common by members of the public.91 Trust in the context of electoral-boundary
readjustment is an expectation of impartiality, and impartial processes of democratic
regulation are trustworthy processes.92 A useful resource in the design of impartial
institutions is therefore the sizeable literature characterizing norms of trust and
trustworthiness. Such norms are informal, richly complex, and widely varied. They
are informal in the sense that no ceremony marks their creation or evolution, and
nowhere are they set down in writing. The literature describes public trust in and
trustworthiness of political institutions as taking the forms of rules, values, and

87 See e.g. Retail, Wholesale and Department Store Union, Local 580 v. Dolphin Delivery Ltd.,

[1986] 2 S.C.R. 573 at 598, 33 D.L.R. (4th) 174, 1 W.W.R. 577 [Dolphin Delivery cited to S.C.R.].

88 This is similar to the dialectical power relations among the traditional branches: executive,

legislative, and judicial.

89 Though decision making entrusted to democratic regulators is ostensibly limited to boundary
drawing or other discrete political rule making tasks, influence over electoral outcomes represents an
open-ended power to steer general policy in certain directions. Administrative democratic regulators
are perhaps weaker than other branches of government, given that the former exercise an indirect and
intermittent power. However, the regulators substantive influence is potentially broad at these
intermittent junctures. Their influence also persists and is generally irreversible over several electoral
cycles; on this temporal scale, their power surpasses that of legislatures, executives, and sometimes
courts.

90 See Karen Jones, Trust as an Affective Attitude (1996) 107 Ethics 4 at 5.
91 See Philip Pettit, Republican Theory and Political Trust in Braithwaite & Levi, supra note 19,

295 at 296-99.

92 See Tom Tyler, Trust and Democratic Governance in Braithwaite & Levi, supra note 19, 269 at
269-70. To be sure, there are several alternative definitions of trust that are relevant in other contexts.

[Vol. 53

MCGILL LAW JOURNAL / REVUE DE DROIT DE MCGILL

16

symbols,93 as well as social meanings94 and agreements,95 and social bonds, networks,
organizations, and loyalties.96 These norms help determine which forms of politics
will gain public favour or disfavour: civil versus polemic; moderate versus extreme;
impartial versus partisan.
Although their origins are obscure and their nature is diverse and intangible,
norms of trust and trustworthiness may steer decision makers toward impartiality
more powerfully than other kinds of rules. This suggestion appears implicitly but
strongly in the Supreme Court of Canadas approach of hands-off judicial review in
Carter,97 the Courts last decision on boundary drawing and one that now largely
belongs to a bygone era. Indeed, the notion that impartial democratic regulation
primarily relies on trust and trustworthiness seems to be on the wane in Canada.
Reforms increasingly fixate on the more tangible methods of regulating politics,
especially those subjecting discretion to clear and rational arrays of rules. As I outline
in Section C, most newer works on administrative approaches to readjustment feature
the conventional raft of impartiality safeguards. Authors focus more on constraint
tools such as transparency, accountability, and democratic scrutiny than on trust and
trustworthiness.98 The Canadian author Mark Carter has asserted a need for a
Charter jurisprudence to elaborate rules for and more clearly restrict the
management of legislative readjustment according to a more consistent vision that
promises to structure and place limits upon the scope of interpretive discretion.99
The courts have generally not fulfilled his wish. However, as discussed below, in the
years since Carter, courts have toyed with tightening the decisions loose effective

93 See Piotr Sztompka, Trust: A Sociological Theory (Cambridge: Cambridge University Press,

94 See ibid.; Lessig, supra note 26.
95 See Daniel Fitzpatrick, Evolution and Chaos in Property Rights Systems: The Third World

1999) at 2.

Tragedy of Contested Access (2006) 115 Yale L.J. 996 at 1001, 1003.

96 Trust and trustworthiness are, in a word, polynormative (ibid. at 1001). On the widely varied
forms into which trust and trustworthiness solidify, see e.g. Levi & Stoker, supra note 19; Francis
Fukuyama, Trust: The Social Virtues and the Creation of Prosperity (New York: Free Press, 1995) at
26; Putnam, supra note 7 at 19-22; Charles Tilly, Trust and Rule (Cambridge: Cambridge University
Press, 2005) at 39. With some simplification, the rubric of norms captures these many forms.

97 Carter, supra note 55 at 185 (adopting a flexible effective representation language and
assuming fair readjustment by provincial riding redistribution commissions, even absent strict judicial
oversight).

98 See e.g. Restoring Accountability, supra note 20; New Era, supra note 20; Mullan, supra note
20; Savoie, supra note 20; Sossin, Boldly Going, supra note 22 (citing other works); Michael
Ignatieff, The Lesser Evil (Princeton: Princeton University Press, 2004) at 51-52; Mark Carter,
Reconsidering the Charter and Electoral Boundaries (1999) 22 Dal. L.J. 53 at 59-60, 71; Canada,
House of Commons, Minutes of Proceedings and Evidence of the Standing Committee on Procedure
and House Affairs, vols. 1-4 (Ottawa: Public Works and Government Services Canada, 1995)
[Milliken Report]; Royal Commission on Electoral Reform and Party Financing, Reforming Electoral
Democracy, vol. 1 (Ottawa: Supply and Services Canada, 1991), c. 7 [Lortie Report].

99 Carter, ibid.

17

R. LEVY ELECTORAL-BOUNDARY POLITICS

2008]

representation standard. And a series of proposals and studies on readjustment have
applied pressure for tighter constraints.

The main assumptions behind these efforts are twofold. First, administrators
tasked with regulating politics will tend to hoard power for themselves or their allies.
Second, these administrators must therefore be subject to limits on their authority.
These assumptions should be uncontroversial at a basic level. And indeed they enjoy
a celebrated pedigree: rooted in the liberal tradition, the constraint model has long
informed the design of political institutions such as those of the United States,
England, and Canada.100 But the approach is problematic if we push it too far, and
especially if it crowds out alternative approaches to regulation.
There are at least three reasons why constraint can be inadequate in the context of

democratic regulation. First, focusing on constraint rather
trust and
trustworthiness has often simply been futile. Three of the most compelling American
critics of constraint tools in democratic regulation are Samuel Issacharoff, Pamela
Karlan, and Gerald Rosenberg. Each shows that constraint approaches have not
worked or will likely cease to work.101 Issacharoff and Karlan use the metaphor of
hydraulics to evoke the tendency of electoral-campaign financing or partisanship, for
example, to flow out of the formal legal channels that are designed as constraints.102
Issacharoff goes so far as to recommend against the renewal of the VRAs Justice
Department review provisions,103 which had been set to expire in 2007.104 In his view,
the most serious exclusions of African Americans and other minority groups from
political participation are now mostly at an end. Issacharoff therefore argues that the
system of VRA scrutiny from above now only stunts the development of a mature

than

100 For early literature on the two constraint assumptions see e.g. Alexander Hamilton, James
Madison & John Jay, The Federalist Papers, ed. by Isaac Kramnick (London: Penguin, 1987), No. 48
(Madison) at 309 [The Federalist] (noting the encroaching spirit of power); ibid., No. 47 (Madison)
at 303. The Federalists were animated by a dark vision of the nature of man (ibid., No. 10
(Madison) at 124). See also Thomas Hobbes, De Cive or The Citizen (New York: Appleton-Century-
Crofts, 1949); David Hume, Of the Independency of Parliament in Essays: Moral, Political and
Literary (London: Oxford University Press, 1963) 40. Even Magna Carta originally came with an
oversight council of 25 barons (Magna Carta, 1215, c. 61, translation reprinted in J.C. Holt, Magna
Carta (Cambridge: Cambridge University Press, 1965) at Appendix IV). See also, more recently,
Ignatieff, supra note 98 at 52.

101 The authors use varying terminologies suggesting the notion of constraint. See e.g. Issacharoff,
Gerrymandering, supra note 9 at 612 (advocating a rescue of U.S. doctrines on gerrymandering
from constricting language).

102 Samuel Issacharoff & Pamela Karlan, The Hydraulics of Campaign Finance Reform (1999) 77
Tex. L. Rev. 1705 at 1713-15. See also David Butler & Bruce Cain, Congressional Redistricting:
Comparative and Theoretical Perspectives (New York: Macmillan, 1992) at 149-50.

103 VRA, supra note 34, s. 5.
104 Congress renewed the VRA in 2006 (Fannie Lou Hamer, Rosa Parks, and Loretta Scott King
Voting Rights Act Reauthorization and Amendments Act of 2006, Pub. L. No. 109-246, 120 Stat. 577).

[Vol. 53

MCGILL LAW JOURNAL / REVUE DE DROIT DE MCGILL

18

Southern politics that transcends race.105 For his part, Rosenberg collects a number of
studies on judicial control of electoral-boundary readjustment that cover the period in
the Warren Court era when readjustment became a justiciable issue and judges
stepped forcefully into the field. Rosenberg concludes that there was at best a
spotty record of actual reform.106 Only some states experienced increases in
legislator turnover and saw their assemblies undergo shifts in party distribution.107
Most famously, some legislatures, stripped of the power to map out unequal districts,
turned instead to gerrymandering within the new constraints imposed and enforced by
the courts.108

Constraint can be especially ineffective at governing democratic regulators, for
whom entrenchment is a constant temptation. Entrenchment is a concern for
boundary readjustment, campaign-financing regulation, constitutional amendment,
judicial selections, and any other area where decision makers can both make and
benefit from significant changes to procedural rules.109 Additionally, impartiality in
democratic regulation markedly defies clear definition,110 which is a requirement for
any workable formal legal rule against partisanship in democratic regulation.

105 See Samuel Issacharoff, Is Section 5 of the Voting Rights Act a Victim of Its Own Success?
(2004) 104 Colum. L. Rev. 1710 at 1714 [Issacharoff, Section 5]. With the VRAs early successes in
enforcing African-American voting rights, new black voters became an easily identified population
that was overwhelmingly Democratic, and race became a marker enabling party-based
gerrymandering.

106 Gerald N. Rosenberg, The Hollow Hope: Can Courts Bring About Social Change? (Chicago:
University of Chicago Press, 1993) at 298. The Supreme Court of the United States first found the
readjustment issue to be justiciable, under selected conditions, in Baker v. Carr (supra note 40).

107 See Rosenberg, ibid. at 296-303. See generally Richard G. Niemi & Laura R. Winsky, The
Persistence of Partisan Redistricting Effects in Congressional Elections in the 1970s and 1980s
(1992) 54 Journal of Politics 565 at 571.

108 Rosenberg, ibid. at 301.
109 Changes that are significant are those that are long-lasting, self-perpetuating, or otherwise
important. Electoral-boundary manipulation is a good example of all three: it is long-lasting because
readjustment normally occurs on a ten-year cycle; it is self-perpetuating because a political party
elected with help from gerrymandered districts can be well positioned to continue manipulating
boundaries; and it is important because general elections determine the policy directions of the state.
Note also that some decisions benefit not the decision makersat least not directlybut their
political party.

110 See Peter Schuck, The Thickest Thicket: Partisan Gerrymandering and Judicial Regulation of
Politics (1987) 87 Colum. L. Rev. 1325 at 1345-48; Issacharoff, Gerrymandering, supra note 9 at
596, 602-05. Some rules constrain governmental abuse that is not political but rather personal, such as
accepting bribes or taking government funds for personal use. See e.g. F.C. DeCoste, Political
Corruption, Judicial Selection, and the Rule of Law (2000) 38 Alta. L. Rev. 654 at 672-73
(discussing personal versus public corruption). These forms of abuse are relatively easy to define, if
not always crystal clear. But we are harder pressed to define the conduct that a constraint rule against
political partisanship would disallow. A straightforward legal prohibition against partisanship can have
some success: judges might easily spot the most egregious results of partisanship in decision making,
such as the outline of the 25th congressional district of Texas. But most cases are harder cases. Courts
in the United States at one time attempted to fashion direct rules against partisan gerrymandering, but

19

R. LEVY ELECTORAL-BOUNDARY POLITICS

2008]

How does the problem of ineffective constraints apply to readjustment
commissions? Authors proposing new commissions generally expect the bodies
themselves to serve as constraints on governmental powers.111 Like the constraints
described by Issacharoff, Karlan, and Rosenberg, commission proposals place
excessive faith in the ability of new decision-making organs to readjust electoral
boundaries impartially.112 Some authors recognize that new decision-making bodies
face the same risks as the old, and that domination of commission membership by
partisans may still occur.113 But examples below will suggest how reform projects
favouring new readjustment bodies typically reflect faith in more, rather than better,
procedure to thwart entrenchment and partisanship. Even authors who declare
themselves wary of entrenchment on commissions make this error. I do not mean to
overstate the tendency toward governmental abuse of discretion; indeed, my own
focus on trust solutions to entrenchment problems suggests that the fear of such abuse
in Canada is often overestimated. The point of the present criticism is that new
decision-making branches per se do not secure impartiality or avoid entrenchment,
notwithstanding the popularity of this assumption. To be sure, entrenchment in a
system of more, rather than fewer, branches of government takes more effort, as it is
harder to spread entrenchment across many branches. But perhaps equally, more
branches also present more potential points of origin of entrenchment.114

have all but given up trying after almost two decades of effort. See Vieth v. Jubelirer, 541 U.S. 267,
124 S. Ct. 1769 (2004) [Vieth]. Not long before Vieth, the Supreme Court of the United States had
ruled in Bandemer that the issue of partisan gerrymandering was justiciable (supra note 66). But the
same court in Vieth considered ensuing efforts to define workable partisanship tests unsuccessful.
Nevertheless, the court allowed that judges could keep trying.

111 This point requires some clarification. Commissions inevitably constrain lawmakers by
depriving the latter of powers to decide on the rules that govern, for example, their own elections. This
is constraining or limiting a power, rather than eliminating or replacing it, because lawmakers
typically retain some role in the processappointing members, making submissions before
commissions, or attempting to influence readjustment commissions.

112 See e.g. Jeffrey Rosen, Divided Suffrage (1995) 12 Const. Commentary 199 [Rosen,

Divided] at 201.

113 Elmendorf, supra note 12; McDonald, supra note 31 at 385-91. Note that entrenchment,
defined above as the manipulation of rules by political insiders to avoid electoral defeat, is a matter of
degree. In Canada and the United States, for all the risks that electoral-system manipulation poses, we
are still far from outright domination of politics by a single party. The risk is rather one of significant
mismatch between popular preferences and electoral outcomes, due to the conscious manipulation of
political-process rules.

114 Domination of any single branch can spread to some or all of the other branches. In the recent
history of the United States, we have seen much creeping entrenchment of this kind: from a majority
Republican Senate, which confirmed conservative judges (based on the advice and consent power to
vet judicial nominees); from the judiciary (and governorships) to the presidency (e.g., in the
partisanship of the 2000 Florida presidential recount and the subsequent Supreme Court of the United
States opinion); from the presidency to the judiciary (e.g., at the Supreme Court of the United States);
and from the presidency and judiciary to the House of Representatives (e.g. in the increased
partisanship of the Department of Justice and in a weakening gerrymandering jurisprudence, both of
which give freer reign to Congressional district gerrymandering). These chains of influence represent

[Vol. 53

MCGILL LAW JOURNAL / REVUE DE DROIT DE MCGILL

20

The second problem with constraint is that the approach is not only commonly
ineffective, but can also diminish trust. Trust and trustworthiness in governance may
be incompatible with law and procedure strongly symbolizing distrust and the
inevitability of partisanship.115 Some legal constraints will always be necessary.116 But
while governmental abuse is inevitable, the relevant question is how prevalent
incidents of abuse are, and what therefore the nature and level of the response should
be. On sober examination we should find that in fashioning constraints that are not
sensitive to the degree of risk, we craft remedies that are worse than the risks
themselves. This tendency appears to be strongest after revelations of significant
abuses of governmental powers, as in the case of Canadas federal-unity sponsorship
program of 19952003.117 Constraint solutions, taken too far, can assume away laws
potential to produce impartial decisions.118 Examples in the next section will illustrate
this point. In a system dominated by constraint and symbolizing distrust, there is little
room for the complex and intangible norms of trust and trustworthiness that can
sustain impartiality.119

Third and finally, the elaboration of new procedural rules based on the constraint
model often steers political decision making and discourse away from substantive

the translation of even transient majoritiesfor example, in the Senate, courts or governorshipsinto
majorities, in other branches, that resist easy reversal. See e.g. Jack M. Balkin & Sanford Levinson,
Understanding the Constitutional Revolution (2001) 87 Va. L. Rev. 1045 at 1085-86. See also
Rosen, Democratic Branch, supra note 69 at 4. Reversing a course toward entrenchment can require
extraordinary shifts in popular party support, exceeding thresholds set artificially high by
gerrymandered districts or other obstacles in election laws (Lublin & McDonald, supra note 9 at 156).

115 See Braithwaite, supra note 19 at 351.
116 Methods relying on constraint and trust in institutional design often coexist. For example, in
Carter, the Supreme Court of Canada does not abandon all commitment to reviewing readjustment
commissions for abuses of power, but rather commits only to giving the commissions considerable
leeway (supra note 51).

117 This corruption scandal prompted changes in the regulation of the conduct of government. See
Restoring Accountability, supra note 20. Another case is the now-defunct Special Prosecutor law
passed in the United States after Watergate. During the Clinton administration, Congress allowed the
law to quietly expire, having seen it bring a culture of scrutiny in U.S. government to regrettable
extremes. Special Prosecutor Kenneth Starrs investigation of the so-called Whitewater affair set up
the conditions for Clintons unsuccessful impeachmenta high-water mark of partisan political
combat in the United States. See 28 U.S.C. 599; Mark Tushnet, The New Constitutional Order and
the Chastening of Constitutional Aspiration (1999) 113 Harv. L. Rev. 29 at 59-61 (analyzing the
impeachment and its partisan origins).

118 See Khullar, supra note 29 (so long as people believe in the myth [of impartiality], and govern

themselves accordingly, then the myth becomes a reality).

119 Braithwaite claims that [t]here are grave dangers in following the advice of Thomas Hobbes
and David Hume and designing institutions that are fit for knaves, based on distrust. The trouble with
institutions that assume that people … will not be virtuous is that they destroy virtue (supra note 19 at
351 [endnotes omitted]). Interestingly, Hume was also an early theorist of trust, viewinglike many
modern writerstrust as facilitating complex forms of social and state action. In a culture where
significant distrust prevails, we do not risk such undertakings (David Hume, A Treatise on Human
Nature, ed. by David Fate Norton & Mary J. Norton (Oxford: Oxford University Press, 2000) at 189).
See also Pildes, supra note 25.

21

R. LEVY ELECTORAL-BOUNDARY POLITICS

2008]

considerations, such as
the geography, community history, and minority-
representation criteria for electoral-boundary readjustment. Instead, constraint
engenders a fixation on process. New constraint rules are meant to limit the leeway
for partisan decision making and entrenchment. But because political combatants, as
noted, manage to steer around constraints, elaborating new rules of constraint often
does little more than shift the locus of political battles. Conflict resumes under these
new and often narrower procedural rules. Manipulating procedural constraints can
then become a main focus of political energies.120 This was starkly illustrated in the
United States as President George W. Bush revealed his intentions to appoint
ideological partisans to federal appellate courts in 200405. Once vacancies opened
on these courts, partisan conflict on the matter began, centering first around the
possibility of a vote on the Senate floor and moving quickly to the issue of minority
filibustering. In response to the possibility of filibusters, Republican senators, who
were in the majority at the time, threatened changes to Senate filibuster rules.
However, a bipartisan Gang of Fourteen centrist senators then agreed to vote
against changes to filibuster procedures except in extraordinary circumstances
thereby laying down a rule governing rules governing filibusters governing judicial
appointments.121 Later the contest over judicial appointments edged toward a fifth layer
of rule making, as debate turned to the boundaries of extraordinary circumstances.122
In debate among politicians and in broader public discourse, the appointments
controversy now focused on this discrete procedural problem.123

In this judicial-appointments episode, new rules merely replaced existing ones;
new constraints did not stop, but rather shifted, battles over entrenchment. At each
step the parties fought under the rule laid down most recently. While elaborating ever-
more precise rules of constraint restricted the leeway for such battles, the process also
narrowed and focused the resulting conflict. This pattern can introduce clarity to a
process and thereby facilitate exploitation of the process. In such cases, the
momentum of partisan conflicts over democratic regulation does not halt, but rather
increases. The benefit of elaborating new constraint rules is often temporary at best:
there is a chaos period, or a lag phase, during which parties adapt to shifting and still-
uncertain rules. Impartiality might well thrive in this brief calm, but the drawbacks
are considerable. Periods of lag and calm in the United States have declined as
partisan efforts have been systematized and professionalized, escalating turnover rates
for new rules that are then exploited and rendered obsolete.124 This rapid turnover and
the narrowing of battlegrounds for procedural dominance engender a focus on

120 See e.g. Rosenberg, supra note 106 at 300; Issacharoff, Section 5, supra note 99.
121 Carl Hulse, Compromise in the Senate: The Nominees; Many Republicans Are Already Eager

to Challenge Agreement on Filibusters The New York Times (25 May 2005) A18.
122 Ibid. This cycle ended with the takeover of the Senate by Democrats in 2006.
123 See Charles Babington & Susan Schmidt, Filibuster Deal Puts Democrats In a Bind

Washington Post (4 July 2005) A1.

124 See Pamela Karlan, The Rights to Vote: Some Pessimism About Formalism (1993) 71 Tex. L.

Rev. 1705 at 1726-37 [Karlan, Rights to Vote]; Hasen, supra note 12 at 949.

[Vol. 53

MCGILL LAW JOURNAL / REVUE DE DROIT DE MCGILL

22

process.125 These problems appear intractable: given a cycle of partisan rule making,
legislators or judges seeking a way out most often turn to new rules of constraint. The
hope is that a new rule will wrest decision making from the cycle.126 In a partisan
system, new rules frequently falter just like their predecessors. Nevertheless, we see
the approach followed determinedly cycle after cycle.
As noted above, impartial decision making is receptive to the full spectrum of
relevant substantive concerns. Impartial decision makers therefore decide without
fixating on the rules of process. They are aware of and follow, but do not focus on or
attempt to manipulate, the rules of procedure that govern their work. On the
impartiality ideal, decision makers are attentive to appropriate substantive criteria
rather than strongly conscious of process or focused on the downstream effects of
their decisions on particular parties.127 In the case of impartial boundary readjustment,
decision makers would especially ignore the effects of their decisions on the
performance of political parties in elections. However, regulating in favour of such
ideal decision making presents a Catch-22: in aiming to reduce behaviour fixated on
process, straightforward attempts to deploy new rules of process to constrain abuse
including new regulatory branches or the traditional safeguards of administrative
impartialitycan be self-defeating by drawing further attention to process.128 In order
to prevent abuse of process, we must be subtler. A rule against process exploitation
should appear natural, unconnected to any partisan camp, authored by no one in
particular, and thus separate from or above normal politics.129 Such rules would
command adherence without opening further avenues for rule exploitation. As I will

125 Part of the problem is that substantive argument becomes processs rhetorical proxy. As Karlan
notes, groups battling for procedural dominance over decision-making systems invoke ostensibly
impartial substantive reasons for rule changes. She predicts that [j]ust as the political parties learned
to use one person, one vote, the Voting Rights Act, and the Shaw principle [applying strict scrutiny to
readjustment geared to increasing minority-representation] as stalking horses for pursuing partisan
ends, so too they will learn to use [the newer legal standards of] Growe and Lawyer (Fire, supra
note 39 at 762).

126 In the Senate judicial-appointments case, with each of five procedural manoeuvres, the parties
did not appear to predict counter-attacks, such as that the threat of filibuster could provoke the so-
called nuclear option of changing filibuster rules.

127 See Lorraine Weinrib, Appointing Judges to the Supreme Court of Canada in the Charter Era: A
Study in Institutional Function and Design in Appointing Judges: Philosophy, Politics and Practice:
Papers (Toronto: Ontario Law Reform Commission, 1991) 109 at 132.

128 See examples in Part IV.
129 Roberto Mangabeira Unger has noted that even though such norms are of course human
artefacts, they suggest origins in the underlying natural order (Social Theory: Its Situation and Its
Task (Cambridge: Cambridge University Press, 1987) at 1). In Lessigs words, an informal norm does
not appear contingent or contested but rather feels natural and is accept[ed] or taken for
granted (supra note 26 at 958-59 [emphasis in original]). See also Harold Hongju Kohs remarks in
Stephen J. Toope et al., Contemporary Conceptions of Customary International Law (Panel
discussion at the Annual Meeting of the American Society of International Law, Washington D.C., 2
April 1998), (1998) 92 Am. Soc. Intl L. Rev. 37 at 38.

R. LEVY ELECTORAL-BOUNDARY POLITICS

2008]

show further below, rules based on informal norms of political culture can have these
features.

23

C. Traditional Administrative-Law Models and Democratic Regulation

In Part IV, I will consider democratic-regulation models that can foster informal
norms of impartial decision making. But to close out the present part, I will first show
that traditional administrative-law approaches based on the constraint model
predominate in works on reforming electoral-boundary readjustment. Three main
contributions present practical proposals backed by sustained theoretical accounts.
These include two from American authors who are concerned with the polarized and
dysfunctional state of politics in that country. Richard Hasen and Christopher
Elmendorf provide thoughtful and creative solutions in the form of administrative
bodies that are meant to address the entrenchment problem in readjustment. Hasen
proposes a set of commissions reined in by elaborately articulated standards for
impartial conduct. Elmendorf advocates the creation of an advisory commission
whose lack of binding authority mitigates the risk of entrenchment. The Canadian
author Mark Carter addresses commission oversight much more as a critic. From a
position firmly rooted in the constraint model, he suggests a shift toward more
searching judicial involvement. In addition, I will give attention to the popular
advocacy of an American organization, Common Cause, whose membership is
largely drawn from the academic community. I will also reference certain Canadian
reports, studies, and comments, each of which, to some degree, promotes reform.130
These various sources
intellectual dominance of constraint
assumptionsthough tentative gestures toward alternatives are made in Hasen and
Elmendorfs works.

following familiar administrative-law safeguards:
(1) Transparency. Mark Carter criticizes the room in which to hide that remains

for gerrymandering after the Carter decisions loose standards of review of
commission decision making.131 Additionally, a House of Commons committee report
by former Speaker Peter Milliken advocates opening up the selection process for
FEBC commissioners. Potential commissioners should, Milliken argues, nominate
themselves for membership by applying to the Speaker of the House of Commons.132
And in another recommendation designed to keep influential partisans from
exercising power behind closed doors, Common Cause proposes the creation of new
state readjustment commissions whose members would be prohibited from all ex-

In proposals for changes to administrative-readjustment procedures, we see the

illustrate

the

130 In addition to the sources cited in this section, several sources call in passing for commission
solutions to readjustment problems. See Rosen, Divided, supra note 112; Klarman, supra note 17 at
529 (offering, briefly, one of the earliest proposals for an independent readjustment commission);
Issacharoff, Gerrymandering, supra note 9 at 644.

131 Carter, supra note 98 at 58.
132 Milliken Report, supra note 98, vol. 3 at 33:17.

[Vol. 53

MCGILL LAW JOURNAL / REVUE DE DROIT DE MCGILL

24

parte communications with elected officials and lobbyists.133 Commission hearings
would also be publicized, and all records of debates and of information that is relied
on would be widely distributed.134
(2) Public Participation. Opportunities for public participation are key to

Elmendorfs approach.135 They are similarly central to the Common Cause
Guidelines, which call for public comments and questions. These recommendations
view public participation as a means of scrutinizing and constraining discretion to
avoid its abuse.136
(3) Representativeness. Some approaches see impartiality safeguards split the

composition of an agency between or among political parties, either in equal numbers
or in proportion to their legislative representation. Representativeness of this kind is a
common approach to political balance in the United States.137 Some Canadian
commissions charged with drawing ridings for the provincial legislatures employ
party-representation techniques.138 These bodies have sometimes run less smoothly
than their federal counterpart; for example, members of one rancorous bipartisan
commission in Alberta in the 1990s were unable to reach agreement over final
boundaries.139 Elmendorf suggests the creation of a readjustment commission for the
United States that comprises an ideologically representative cross section of the
citizenry.140 He
to an approach mandating bipartisan
membership.141 Similarly, Common Cause cites bipartisanship as a possibility.142 In
these assorted recommendations, we see the often self-fulfilling assumption,
described in the previous section, that individual decision makers will render partisan
decisions, and that the process will feature procedural manoeuvring.143

is also sympathetic

133 Common Cause, California Common Cause Redistricting Guidelines (August 2005), s. 3
[Common Cause, Guidelines], online: Common Cause .

134 Ibid.
135 Supra note 12 at 1376.
136 Supra note 133, s. 3. An alternative rationale for public participation can be enhancing the

democratic legitimacy of the process.

137 State electoral commissions generally split representation evenly between the two parties or offer
a numeric advantage to the states majority party (Hasen, supra note 12 at 974-76). In another
example, the six commissioners of the Federal Elections Commission (in charge of implementing
federal election rules, including campaign-finance legislation) include three Republicans and three
Democrats. See e.g. Gerard Clark & Steven Lichtman, The Finger in the Dike: Campaign Finance
Regulation After McConnell (2006) 39 Suffolk U.L. Rev. 629 at 656.

138 Alberta, Newfoundland, Nova Scotia and Prince Edward Island have employed party-

representation methods. See Courtney, supra note 12 at 107-10, 293.

139 See ibid. at 111-12.
140 Supra note 12 at 1407.
141 Ibid. at 1408-10.
142 See e.g. supra note 133, s. 1. See also John Fund, Stealing Elections: How Voter Fraud

Threatens Our Democracy (San Francisco: Encounter, 2004) at 147.

143 See Hasen, supra note 12 at 989.

25

R. LEVY ELECTORAL-BOUNDARY POLITICS

2008]

The Common Cause Guidelines also recommend more complex forms of

representativeness, reflecting the geographic, racial, ethnic, gender, and age
diversity of a jurisdiction.144 We may read this benignly as a means of including
otherwise marginalized perspectives, or alternatively as an expression of the
constraint model, with representatives of various social factions meant to hold each
other in mutual check and to battle out compromises.145
(4) Independence. Most proposals, including those of Hasen,146 Elmendorf147 and

Common Cause,148 assume the necessity of commission independence. This position
should be uncontroversial at a basic level. But in the context of regulating political
power, independence safeguards face particularly acute troubles. American examples,
such as most of the existing state readjustment commissions, demonstrate that
partisans can almost always degrade independence by bypassing constraint
safeguards.149 The traditional legalistic formulae for administrative independence are
less important to impartial outcomes than is a culture of trust and trustworthiness.
Indeed, the separation of certain governmental powers is not a feature of some
systems that have generally been successful in avoiding entrenchment. The
Westminster model, which fuses the executive and legislature, is an example of
this.150
(5) Reconsideration and Review. Mark Carters proposals contemplate pervasive

judicial involvement in the administrative readjustment process.151 In addition, the
Common Cause Guidelines allow for judicial review of plans and reconsideration
by the commissions.152 We have already seen that judicial involvement extends

144 Supra note 133, s. 1.
145 The latter suggests the weaker of the two conceptions of impartiality introduced above.
146 Supra note 12 at 984.
147 Supra note 12 at 1380, 1405-06, 1408, 1412-14.
148 Supra note 133, s. 1.
149 There are two general cases exemplifying this point. First, to bring greater independence to a
process, we can create new bodies to carry out contentious tasks, such as electoral-boundary drawing.
The problem then becomes how to appoint to the body independent-minded members rather than
partisans. (This is a problem of the regress of partisanship from one point to another.) Second, a
solution to the first problem is often thought to be bipartisan selections to the independent body.
However, this solution generally produces bipartisanrather than non-partisanbodies, which as
noted are often dysfunctional. To be sure, some multipartisan selection processes work well, such as
Canadas method of all-party parliamentary selection of the Chief Electoral Officer (see Massicotte,
supra note 50 at 97-98). But the deciding factor here seems to be that the selection process is already
characterized by broad trust and trustworthiness. This, then, brings us back to the problem with which
we startedthat the strongest safeguard against partisanship is a well-developed set of norms of trust
and trustworthiness.

150 See Peter Hogg, Constitutional Law of Canada (Toronto: Carswell, 1992) at para. 7.3(a). The
Cabinet holding sway over Parliament can manipulate political rules and seek selective party benefit;
while this has happened often in Canada, it has not produced long-term single-party entrenchment.
See supra note 10.
151 Supra note 98.
152 Supra note 133, s. 5.

[Vol. 53

MCGILL LAW JOURNAL / REVUE DE DROIT DE MCGILL

26

throughout readjustment in the United States. In Canada, some recent cases suggest
an increasing judicial willingness to intervene. Recently in Rache v. Canada
(A.G.),153 the Federal Court of Canada subjected the work of the FEBC for New
Brunswick to judicial review. The court applied a standard of simple reasonableness,
having considered adopting the lower patent-reasonableness threshold.154 At the same
time, the case affirmed a substantively flexible approach; the FEBCs error, in the
courts view, was in limiting itself to electoral-district population variations of no
more than plus or minus 10 per centin contrast with the more generous plus-or-
minus 25 per cent range in the governing legislation.155 Other courts have affirmed the
longstanding flexible framework of Carter. But as in Rache their support is
sometimes grudging or mixed; courts appear ready to step in more often to review
readjustment, especially should the Supreme Court of Canada revisit Carter.156
(6) Rational Organization, Simplification, and Clarity. Under this final heading,

the key assumptions of each author become clearest. Does a plan place hope in the
creation of a new administrative branch per se to solve the entrenchment risk, without
sufficient regard to specifics? Does the plan treat impartiality as a value that is
directly manipulable under a regime of ever more elaborate formal rules? Each of the
surveyed proposals, being premised on a wariness of power and anticipating an
attitude of mutual distrust among decision makers, ultimately calls for a constraint
approach. Among these proposals there is general agreement that lines of exercise of
power should be (a) simple, clear, and rationally organized, (b) centralized,
coordinated and coherent, and (c) subject to cost-benefit analyses and top-down
control.

That constraint appears in the various U.S. proposals might not be surprising,
given the persistence of this model in the American intellectual tradition.157 But Mark
Carters constraint proposals for Canada are the most extreme. In addition, John
Courtney, the leading observer of Canadian readjustment, laments, more mildly to be
sure, that Parliament and the courts can only do so much to offer clarification in
statutes and guidelines.158 The Milliken Report and the report of an inquiry headed
by Pierre Lortie both outline reforms to tighten controls on the boundary-drawing

153 2004 FC 679, [2005] 1 F.C.R. 93, 120 C.R.R. (2d) 133 [Rache].
154 Ibid. at paras. 54-65.
155 Ibid. at paras. 67-72, 82; EBRA, supra note 13, s. 15(2)(b).
156 Cases following the flexible standard include City of Charlottetown v. Prince Edward Island
(1998), [1999] 168 Nfld. & P.E.I.R. 80, 168 D.L.R. (4th) 79 (P.E.I.S.C. (A.D.)); East York (Borough
of) v. Ontario (1997), [1998] 36 O.R. (3d) 733, 45 C.R.R. (2d) 237 (C.A.); Friends of Democracy v.
Northwest Territories, [1999] N.W.T.J. No. 28 (N.W.T.S.C. (A.D.)); Reference Re Electoral
Boundaries Commission Act (Alberta) (1991), 120 A.R. 70, 86 D.L.R. (4th) 447 (C.A.) [Boundaries
Comission cited to A.R.]. In Boundaries Commission in particular, the court appeared reluctant to
apply the precedent (ibid. at 454).

157 See supra note 100.
158 Courtney, supra note 12 at 259. Courtney generally believes that the Canadian commissions

merit trust in handling the elusive and imprecise standards that remain (ibid.).

27

R. LEVY ELECTORAL-BOUNDARY POLITICS

2008]

process in Canada. Reforms would include narrower population-variance allowances
and, most notably, more explicit definitions of standards such as communities of
interest.159 And in a recent study, Michael Pal and Sujit Choudhry echo these
sentiments, criticizing the lack of discipline imposed by the Charter [case law] on
the drawing of electoral boundaries.160 The various works thus present a modest
inversion. American authors are increasingly seeking alternative solutions. These
authors look with frustration at the escalating dysfunction of their own national
system, a development that presents cautionary lessons as the constraint model
captures more Canadian imaginations. Hasen and Elmendorfs solutions take
tentative steps past rationalization and constraint. Their plans deserve serious
consideration. But despite their intentions, neither author effects a complete break,
and their remaining reliance on constraint gives reason to doubt the likely efficacy of
their plans.
Within each proposal, then, a project of constraint predominates. As noted, Mark
Carter believes there is a need for a Charter jurisprudence to elaborate rules for
readjustment and more clearly restrict[] its management.161 The notion of a more
consistent vision that promises to structure and place limits upon the scope of
interpretive discretion is an example of rationalization and rule making elaborated to
curtail ambiguity.162 Common Cause similarly employs the language of clear
process.163 But Hasens work provides the most prominent example, proposing rules
tweaked toward perfection as a means of limiting discretion.164 He recommends
[r]emoving the opportunity for partisan election officials to make discretionary
decisions,165 in part by implementing periodic election law audits, thereby

159 Milliken Report, supra note 98, vol. 3 at 33:25; Lortie Report, supra note 98 at 150, 157-58.
160 Michael Pal & Sujit Choudhry, Is Every Ballot Equal? Visible-Minority Vote Dilution in
Canada (2007) 13:1 Choices 1 at 14-16. The authors phrase this as a general fault in the Charter case
law. But their primary concern is minority-vote dilution: the electoral votes of visible and other
minorities concentrate within urban ridings, each of which encompasses more voters than does the
average rural riding. I am sympathetic to their more particular critique. Carter justified ruralurban
constituency size disparities primarily in light of MPs difficulties in serving physically vast rural
ridings. However, I wonder whether, in the fifteen years since Carter, developments such as
unprecedented advances in information technology now weaken this justification. Smaller urbanrural
disparities therefore make sense generally. However, it remains important that FEBCs have the
flexibility to draw boundaries in light of the many substantive considerationssuch as the cohesive
communities of Rache (supra note 140)that keep the commissions busy for up to two years of
deliberations. Some of these considerations would necessarily become difficult to implement given
narrower population variances.

161 Carter, supra note 98 at 59-60.
162 Ibid. at 71. Note that Carter addresses and ultimately rejects critiques of rights-based judicial
oversight of readjustment. But he narrowly addresses and shoots down only the critical-legal-theory
view that the legal system operates so as to maintain unequal and oppressive social relations (ibid. at
62).

163 Supra note 133, s. 5.
164 Hasen, supra note 12 at 944.
165 Ibid. at 983.

[Vol. 53

MCGILL LAW JOURNAL / REVUE DE DROIT DE MCGILL

28

progressively eliminating potential ambiguities from the law.166 To Hasen,
ambiguities are failures of vigilance, presenting decision makers with irresistible
opportunities for partisanship and entrenchment. A reliance on the perfectability of
formal law to control conduct animates his approach.

But Hasen himself acknowledges that his proposed set of legal constraints upon
discretion would occasionally let partisan abuses slip past. He therefore offers a
second, institutional solution: a readjustment commission for each state. Citing the
track records of Canada and Australia, Hasen proposes to import their commission
solutions.167 It is, from the outset, an instance of faith in the self-evident wisdom or
common-sense168 of creating new institutions per se. However, to his credit,
Hasens solution of a new administrative apparatus goes somewhat beyond the bare
adoption of a new branch and into the specifics of design. Hasen even touches on the
question of how trust develops in decision-making institutions. Here he offers his
most compelling ideasmeant explicitly to restore some public trust in the process
of election administration.169
A first idea, and the weaker of two, looks much like his earlier formal-constraint
proposal. He reasons
that the fundamental principles of neutral election
administration are not subject to serious debate;170 by laying these out in law, it is
possible to develop trust and trustworthiness within a new commission. But there are
problems with this approach. Hasen proposes that trust be legislated by laying out in
law the features of an ideal impartial decision maker. However, the familiar Catch-22
applies: because part of the challenge of impartiality is getting decision makers to
fixate less on process, elaborating new process rules that call directly on decision
makers to act impartially often falters, as we saw in the previous section.
Hasen seems aware of the difficulty of legislating impartiality. He therefore
writes that laws should set out the features of trustworthy decision makers and not
decision making: [N]eutral election administration is easier to achieve than neutral
redistricting principles.171 But can we separate the decision makers impartiality from
concrete meanings or manifestations of impartiality? How does a decision maker
acting as free from fraud as possible172 approach a given boundary-drawing
problem? These are not quibbles over the meanings of words that are indeterminate at
their margins, but rather problems of basic terminological emptiness. Some parts of
Hasens proposed checklist for impartial administrators are reasonable and workable.

166 Ibid. One method is to articulate a consensus set of election administration principles that
nonpartisan administrators could apply (ibid.). Note that with the language of limiting ambiguity
here, Hasen advocates an approach almost explicitly opposite to the ambiguation methods proposed
below.

167 Ibid. at 945, 983-90.
168 Ibid. at 974.
169 Ibid. at 983.
170 Ibid. at 988.
171 Ibid.
172 Ibid.

29

R. LEVY ELECTORAL-BOUNDARY POLITICS

2008]

For instance, he provides that a commissioner should not co-chair a presidential
committee.173 But most parts are problematic: what does it mean to say that
administrators should not do anything to favor one candidate, party, or issue?174
These are features not of impartial administrators, but impartial administration, and
the routine slide from the former to the latter seems inevitable. Almost anything a
decision maker doesdrawing a black-majority township into or out of an electoral
district, for examplefavours one candidate, party, or issue. While Hasen therefore
means to avoid defining impartial decision making in elaborate detail, being aware of
the pitfalls of such efforts, his focus on impartial decision makers is not likely to be
more effective.
A second trust-generating design feature is more elegant and more plausible,
though still primarily intelligible as a form of constraint. Hasen proposes to select
commissioners through 75 per cent super-majority votes in the state legislatures, thus
effectively requiring support from both political parties.175 The plan is a constraint
proposal in that it assumes distrust between the parties and systematizes and
reinforces the same. However, the proposal is also intended as trust-generating,
conceivably installing non-partisan decision makers to begin rehabilitating the
American readjustment system. Nevertheless, this part of Hasens plan also appears
flawed. Legislators must divine in advance which selectees will act above
politics176selecting perhaps for inoffensive but middling administrators. Other
authors propose institutions that promote good decision making irrespective of the
particular identities of the decision makers. On these approaches, the success of the
system depends less on getting the selection process right at the outset. But the
foremost problem with Hasens plan remains the formidable constraint presented by
the super-majority requirement. The use of constraint mechanisms to build trust is a
creative, if fraught, approach. Would kicking off a process with a method of member
selection premised heavily on mutual party animus undercut the main subsequent aim
of inculcating trust? Super-majority requirements potentially recast each individual
legislator as a powerful swing votera formula for intensified politicization and
gridlock.177
Christopher Elmendorfs alternative administrative solution to the problem of

readjustment is an Advisory Commission (AC) that would draw new district maps
and urge legislators to adopt them.178 Elmendorf cites two key rationales for creating a
body whose decisions are advisory rather than formally binding. First, the AC would
need to work for its trust. Because advisory decisions can be ignored, the plan sets up

173 Ibid.
174 Ibid.
175 Ibid. at 984.
176 Ibid.
177 Individual dissenters can wield inordinate power to dictate terms by threatening to withhold
votes. See John O. McGinnis & Michael B. Rappaport, Supermajority Rules and the Judicial
Confirmation Process (2005) 26 Cardozo L. Rev. 543 at 550.

178 Elmendorf, supra note 12 at 1371, 1380.

[Vol. 53

MCGILL LAW JOURNAL / REVUE DE DROIT DE MCGILL

30

a competition for authority between the AC and legislators.179 Like Hasen,
Elmendorf explicitly means for this dynamic to engender trust. The AC would be
authorized to place its concerns on the legislatures agenda or a referendum ballot,
and positioned to compete with legislators for the voters trust.180 Elmendorfs plan
aims to build trust and trustworthiness as the competing authorities demonstrate that
they are trustworthy: that they habitually issue impartial decisions apparently driven
by appropriate substantive criteria for readjustment.181 A second reason for the ACs
advisory function is that if partisans nevertheless do assume control of the
commission, the bodys non-binding decisions may safely be ignored.182

The plan has a mark in its favour given its goal of pushing democratic regulation
beyond constraint and toward trust solutions. However, apart from its advisory role,
the ACs structure is undistinguished and would likely still engender partisanship
problems. The AC does little else to improve on the constraint model, placing too
much faith in the solution of adopting a new decision-making body per se. We see
some of this misplaced faith in Elmendorfs proposal of bluntly combining trust and
constraint approaches. He imagines trust developing in a milieu where constraint now
strongly dominates. As we saw, the momentum of constraint can be difficult to halt.
This momentum formalizes democratic regulation: constraintsformal, direct,
explicit, and premised on distrustbreed more of the same. In American democratic
regulation, we have thus seen processes increasingly centred around formal rules.
These are rules whose authority is relatively clear: explicit agreements between
political parties, enforceable legislation, regulation, and court judgments.183 In such a
formalizing process weaker informal norms of institutional culture are lost, along
with their richness, subtlety, intermediacy, flexibility, and comparative efficacy.
Effectively regulating informal norms of trust would mean adopting procedures in
which trust is stable because it is not susceptible to the push and haul of politics and
the constraint model. Decision making premised on trust and trustworthiness should
remain largely separate from any constraint rules. How regulation can be designed in
this way will be a major concern of Parts III and IV.

Looked at more concretely, there are two specific problems with Elmendorfs
AC. First, the AC is meant to enjoy trust and exert sway as a matter of subtle degree.
The AC would be expected to bind in proportion to its demonstrated impartiality. But
such a fine balance appears unlikely within a context of robust constraint. This

179 Ibid. at 1382.
180 Ibid. at 1371.
181 Ibid. at 1382-83.
182 Ibid. at at 1436.
183 A recent example was the United States Attorney Generals firing of eight leading federal
prosecutors, apparently for their weak loyalty to the President and the Republican party. Responding
to the scandal, the President stated that U.S. attorneys serve at his pleasure. See Michael Abramowitz,
Bush Asserts Increased Confidence in Gonzales Washington Post (24 April 2007) A3. While
perhaps true as a matter of formal law, the observation dramatically devalues the unwritten principle
of prosecutorial impartiality.

31

R. LEVY ELECTORAL-BOUNDARY POLITICS

2008]

particular hybrid of trust and distrust is likely to be unstable as trust norms yield to
encroaching formal rules. A second concrete problem follows from a distinction
between decisions meant for selective adoption and full adoption. Consider first a
kind of commission that issues numerous recommendations from which legislators
may pick and choose. Some such recommendations are impliedly optional, or are
aspirational and await a time when governments can afford to adopt them. For
example, the American 9/11 Commission proposed more changes to national-security
procedures than can realistically be implemented.184 Generally, advice subject to
second-guessing defeats the purpose of trust in advice, except in the limited sense that
the advice provides helpful information to guide other decision makers.185 At the other
extreme is a strong notion of trust: we may trust the whole decision and adopt all its
parts. Given the potential for entrenchment in the democratic-regulation context, it is
generally preferable for the whole of a decision to be binding.186 It may often be
reasonable for legislators to second-guess other kinds of administrative decisions, but
selective trust poses special risks for democratic regulation. A process that subjects
trust in democratic regulation to competition encourages legislators to challenge the
very trust in commissions that can keep legislators from entrenching their own power.

Elmendorf predicts success for his plan by relying on foreign examples for
empirical backing, noting that trust in democratic regulation has survived and
flourished outside the United States.187 However, in most of these cases regulators did
not have to develop trust in an environment of robust formal constraint. Even if
applied to contexts such as Canadian readjustment, where constraint has yet to
become the regulatory tool of choice, Elmendorfs innovation would probably present
an unstable marriage of trust and distrust. The plan presents a contradiction of
concepts: trust is meant to emerge within an institution designed to enhance …
political contestation,188 with powers counterpoised to prompt constraint, a struggle
for dominance, and a hawkish mutual distrust between decision makers.189

184 U.S., The 9/11 Commission Report: Final Report of the National Commission on Terrorist

Attacks Upon the United States (New York: W.W. Norton, 2004), cc. 12-13.

185 See Joseph Raz, The Morality of Freedom (Oxford: Oxford University Press, 1986) at 28-31
[Raz, Morality] (utterances may provide information about the balance of reasons as they exist [in
favour of a particular action] separately and independently of such utterances at 29).

186 A wholly binding AC would effectively not be advisory and would lack the asserted benefits of

the latter.

187 Elmendorf, supra note 12 at 1372, 1386-90 (referencing Australia, Canada, Germany, and the

United Kingdom).
188 Ibid. at 1371.
189 See e.g. ibid. at 1383-84. Analysts generally characterize decision making based on trust as
rational and cooperative, not combative. See e.g. Denise Scheberle, Federalism and Environmental
Policy: Trust and the Politics of Implementation, 2d ed. (Washington D.C.: Georgetown University
Press, 2004); Robert Putnam, Making Democracy Work: Civic Traditions in Modern Italy (Princeton,
N.J.: Princeton University Press, 2002) at 167.

32

MCGILL LAW JOURNAL / REVUE DE DROIT DE MCGILL

[Vol. 53

D. Departing from the Constraint Model

We have already seen two alternatives to constraint. One simple alternative is a
laissez-faire approach granting wide decision-making latitude to democratic
regulators. Issacharoffs opposition to renewing parts of the VRA is an example. As
already noted, he views constraints as straightjackets: remove them, and perhaps a
mature Southern politics that is more impartial and less reliant on administrative and
judicial intervention will develop. Equivalently in Canada, keep the straightjackets
off, and let existing norms of impartiality in democratic regulation continue to work
unhindered. In contrast with the laissez-faire model, a second alternative sees
institutions actively designed to engender norms of trust and trustworthiness. As we
saw above, both Hasen and Elmendorf nod toward this approach.
Attempting deliberately to generate, through law, norms mediated outside of
lawto regulate informal cultural norms of trust and trustworthinessis among the
thorniest challenges of institutional design. Few forms of regulation are more certain
to produce unintended consequences than those functioning at the interface between
law and culture.190 And most such active regulation must usually begin with formal
rule making, the primary tool available to governments. However, looking back at a
number of cases in which governments regulated cultural practices, Lawrence Lessig
finds a number of successful challenges to existing normsfor example, norms
against interracial mixing and norms favouring duelling or gender discrimination.191
Indeed, I will outline further examples in Parts III and IV, including examples from
the readjustment context. Ambiguity in Canadas FEBC readjustment procedures
helps to account for their early and sustained successes in developing impartial
decision making.192 Governments can in some cases change or create informal
institutional norms, even if such results are seldom guaranteed.

190 Lessig himself notes that [g]overnments, as other institutions, are inept; changes are very often
not as intended (supra note 26 at 957). See also Richard H. Pildes, The Unintended Cultural
Consequences of Public Policy (1991) 89 Mich. L. Rev. 936 at 938-40; Issacharoff & Karlan, supra
note 102; Rosenberg, supra note 107; Cass R. Sunstein, Radicals in Robes: Why Extreme Right-Wing
Courts Are Wrong for America (New York: Basic, 2005) at 100-01.

191 Lessig, supra note 26 at 964-72, 989-91.
192 Like any history, the fullest account is one of multiple concurring influences, worth bearing in
mind even as we isolate and focus on apparent primary causes. There may be several parallel
explanations, all of which inevitably interrelate. Courtney cites a set of three causes: first, a 1950s60s
shift toward a politics regarding the interests of the country as a whole; second, numerous unstable
minority Parliaments over this same period; and third, a mood for change developed in Quiet
Revolution-era Quebec, site of one of Canadas first readjustment commissions (supra note 12 at 44-
52, 55). To Courtneys history we might add, fourth, the broadly-acknowledged differences in
structure as between parliamentary and presidential democracies; but these, though relevant, may not
explain why Canada diverged in the 1960s. Fifth, Canadas multi-party system might encourage
polarization less than does the American two-party system; but this difference, too, held before the
1960s. Sixth, French marginalization inside Quebec subsided in the 1960s, forcing a national politics
of engagement and accommodation. Finally, the United States experienced a more dramatic series of
political scandals in recent decades. See generally David Farrell, Comparing Electoral Systems

33

R. LEVY ELECTORAL-BOUNDARY POLITICS

2008]

In any case, regulating to promote trust and trustworthiness is an approach worth
trying, or that we must try in light of the limitations of regulation by constraint. As I
have argued, informal norms of trust and trustworthiness control partisanship at least
as well as formal constraints do. Indeed, it is often the case that only the latter are able
to diminish partisanship. But active regulation is also preferable to laissez-faire
approaches. In a United States South that is now heavily politically polarized, it is
doubtful that leaving readjustment in the hands of legislators would bring
impartiality. As we saw, boundary readjustment through normal legislation can
seldom be seen as impartial.193 Readjustment must be carried out by third-party
commissions. But the status quo for federal readjustment in Canada may also be
inadequate. At the outset of this article, I listed the stresses facing the existing system,
including declining trust and a rising culture of scrutiny of government. These
coincide with frequent studies and committee reports advocating reforms to
readjustment according to the constraint model.194
In the next two parts, I will focus on understanding the institutional features of

the Canadian readjustment process that have helped to generate its trustworthiness
and impartialityfeatures that therefore merit preservation or even extension. I will
rely in part on interviews with individual FEBC commissioners, who spoke to me
about the strengths and faults of their commissions. I will also draw on the
interdisciplinary literature describing specific institutional forms that have previously
seen the emergence of trust and trustworthiness in democratic regulation. Actively
designing institutions to build trust and trustworthiness is at once one of the most
promising models of democratic regulation and the least explored in legal
scholarship.

III. Regulating Norms of Impartiality in Democratic Regulation
Authors have often sought to explain the conditions under which legal change
can bring about desired social change.195 Lawrence Lessig offers ideas that are among
the most detailed, insightful, and distinctively relevant to impartial democratic
regulation. Further on, I will bring in other authors and other ideas not contemplated
by Lessig. However, Lessigs work is a helpful stepping-off point. His approach

(London: Prentice Hall, 1997); Paul R. Abramson, Political Attitudes in America: Formation and
Change (San Francisco: W.H. Freeman, 1983).

193 Supra notes 1-10, 14 and accompanying texts. This has been true of readjustment by legislatures

in both the United States and Canada.

194 Supra notes 19-22 and accompanying text.
195 Some authors bring to bear the tools of economic analysis. See e.g. Rosenberg, supra note 107.
Other authors isolate the democratic legitimacy of institutions as essential for successful social
regulation. See e.g. William N. Eskridge Jr., Pluralism and Distrust: How Courts Can Support
Democracy by Lowering the Stakes of Politics (2005) 114 Yale L.J. 1279 at 1285-87, 1312-13; Pettit,
supra note 91 at 296-99; Rosen, Democratic Branch, supra note 69 at 14; Mullan, supra note 20, c.
8(A); Tom R. Tyler, What is Procedural Justice? Criteria Used By Citizens to Assess the Fairness of
Legal Procedures (1988) 22 Law & Socy Rev. 103 at 129.

MCGILL LAW JOURNAL / REVUE DE DROIT DE MCGILL

34

touches briefly on a general and promising set of solutions to problems of democratic
regulation. This part introduces ambiguity as a tool of democratic regulation by
beginning with Lessigs ideas on the subject and then considering several additional
rationales for ambiguous legal and institutional design. In Part IV, I will detail the
limitations and potential scope of ambiguity methods.

[Vol. 53

A. Lessig and Ambiguity

Lessig describes four methods by which governments deliberately reshape social
norms. Two of these methods, which he labels behavioral techniques, change the
meanings associated with particular actions indirectly by inducing certain behavior
that, over time, will affect these meanings.196 One behavioural technique is ritual.
Ritual sees a behaviour mandated by law fortify a particular social norm. For
example, American schoolchildren daily salute the flag; this state-mandated policy,
which the Supreme Court of the United States has affirmed as constitutionally
permissible, aims to inculcate patriotism.197 Another behavioural technique is that of
inhibition, which is rituals opposite. Following this approach, laws repress a
particular behaviour. Lessig gives the example of segregation, which is both an
instance of racial harm and a behavior that reinforces the social meaning of
inequality.198 As he notes, Prohibiting segregation is a way of undermining practices
that reinforce social meanings of stigma and inequality.199

Two more methods are semiotic. These change[] meaning[s] directly, by
interfering with existing meanings.200 The first such method is that of tying. Tying
is an attempt[] to transform the social meaning of one act by tying it to, or
associating it with, another social meaning.201 Sanctioning an act by criminalizing it
or advertising its ills, in either case with the intention of stigmatizing and deterring
the conduct, is one form of tying open to legislators. Examples include attempts to
control duelling in the old American South (discussed below) and cigarette smoking
more recently. Both demonstrate the complexity of social regulation since neither was
entirely successful.

right to view this one as the most interesting.202 Here the social-norm

Last on Lessigs list is the semiotic technique of ambiguity. Lessig is probably

architect tries to give the particular act, the meaning of which is to be regulated,
a second meaning as well, one that acts to undermine the negative effects of the
first. In this sense, while tying is about establishing that X is like Y,

196 Lessig, supra note 26 at 1008. Lessig cites social meanings throughout, but for present

purposes differences between this term and social norms are not significant.

197 Ibid. at 1014.
198 Ibid. at 1013.
199 Ibid.
200 Ibid. at 1008.
201 Ibid. at 1009.
202 Ibid. at 1010, 1015.

2008]

R. LEVY ELECTORAL-BOUNDARY POLITICS

35

ambiguation is about establishing that X is like Y or Z. It simply adds a link
without denying an existing link, and thereby blurs just what it is that X is.203

Returning to the case of duelling, Lessig writes that straightforward prohibitions
failed more often than the alternative of ambiguous regulation.204 He notes that what
held dueling together was solidarity among an elite class of Southern gentlemen. 205
Simply banning dueling, he claims, would not necessarily challenge that
solidarity.206 A man who declined a duel on grounds of its illegality and who was
consequently viewed as a coward risked his place at the heights of the social ladder.
Lessig draws on histories illustrating another, somewhat more effective regulatory
option: barring duellers from public office. Since holding such office was a duty of
the elite, the message of duelling thereby came to include not only the dishonour of
declining a challenge but also the dishonour that would result from accepting it.207
Lessig explains:

The states action here served to ambiguate a gentlemans duty, and thereby
facilitated the transformation of the social meaning of dueling itself. Against the
background that the state has reconstructed, to choose to duel would be to
choose to serve private interests over collective duty.208

Lessigs categories are fascinating in themselves, though his examples illustrate
only part of what makes ambiguity as interesting as he suggests. In exploring how
Lessigs basic outline of norm-regulation techniques can relate to political decision
making, I wish to diverge from his work and to build further. What is disappointing in
Lessigs recognition of the broader implications of ambiguity is the narrow range of
cases in which he imagines this technique applying. The example of duelling is not
one that demonstrates the benefits of ambiguous social meaningsof meanings that
are multiplied and rendered inconsistent. It is simply a case of meaning Y (dishonour)
negating and replacing meaning X (honour). The aim of the regulation barring access
to public office for duellers was to turn a social meaning of honour into one of
dishonour. What then does ambiguous regulation do that simpler techniques negating
or creating meaning, such as tying, cannot? Good examples of the full use of
ambiguity would see ambiguity itself, rather than ambiguity as a by-product or
transition stage of negation, put to work as a tool of legal regulation. Such uses do
exist and display some of the real promise of obscuring what was clear that Lessig
inchoately identifies. 209 The tool can be a powerful one, not least in the context of
democratic regulation.

203 Ibid. at 1010.
204 Ibid. at 970.
205 Ibid. at 971.
206 Ibid.
207 Ibid.
208 Ibid. at 972.
209 Ibid. at 1015.

MCGILL LAW JOURNAL / REVUE DE DROIT DE MCGILL

36

In the remainder of this part, I first outline the ways in which the FEBC system is
an ambiguous one. I then examine four rationales for designing ambiguities into law
and institutions. All are ways in which ambiguity can help to build trust,
trustworthiness, and impartiality. Throughout, I offer examples from an assortment of
past projects of institutional design. But the major examples will be the Canadian
FEBCs. Extensive ambiguity appears to have helped develop the strong culture of
impartial decision making of the FEBC system.

[Vol. 53

B. Forms of Ambiguity

Ambiguity in law and institutional design can be a result of numerous and diverse
(1) procedures, (2) substantive criteria, and (3) decision-maker roles. The
extraordinarily complex FEBC system illustrates each of these types.210

1. Procedural Ambiguity

From the description of FEBC readjustment above, we see that decision making
runs an obstacle course of diverse and redundant stages. There is a mixture of
advisory and direct influences: parliamentarians and members of the public make
recommendations, and commissioners finally decide. There is a combination of
partisan and generally impartial contributions. There are open and in camera sessions.
The stages of the Canadian process are numerous and their total duration extensive.
Together, they typically exceed the minimum consultation periods defined by statute
and typically last two years or more.211

2. Substantive Ambiguity

The FEBC system features a diversity of broad and sometimes incommensurable

substantive criteria for commissioners to consider.212 The principles set out in Carter
and in legislation are pitched at different levels of specificity and breadth, as well as
clarity and vagueness. The practical problems of representing a large rural riding are
concerns, but so are human dignity and social justice. Many principles overlap or
conflict with others even in the abstract: geography, community history, community
interests, practical problems of representation, numeric parity among ridings, and
effective representation.213 Others overlap in their application, as where multiple

210 Courts have made note of the complexity of the electoral-boundary readjustment process

(Rache, supra note 153 at para. 32). See also Carter, supra note 51.

211 Elections Canada, Representation in the House of Commons of Canada (Ottawa: Elections

Canada, 2002) at 14.

policies).

212 Rache, supra note 153 at para. 32 ([t]he commissions are required to balance conflicting

213 The most common conflict of principles is that between numeric parity of ridings (the one
person, one vote default) and the principle that large rural ridings should have populations well below
the average, given the difficulty of representing vast under-populated areas.

R. LEVY ELECTORAL-BOUNDARY POLITICS

2008]

communities with distinct histories occupy the same region. The rules for numeric
apportionment of voters among ridings alone are elaborately complex.214 In contrast,
[n]o equivalent rabbit-warren of representational protection exists in the United
States.215 There is in general a marked indeterminacy in the correct application of
readjustment criteria.216 The substantive guidelines of the process can be applied in
many alternative ways,217 and have an elusive and imprecise quality about them.218

37

3. Role Ambiguity

Commissioners are drawn from a diversity of professional cultures with
surprisingly little in common. The most recent round of readjustment saw judges and
political scientists, who still predominate, sitting with members who included a social
worker and an RCMP officer.219 Furthermore, the extensive opportunities for public
participation noted above add a key additional dimension of role ambiguity; as I will
note below, public submissions before the FEBCs are often highly persuasive.

C. The Uses of Ambiguous Decision-Making Systems

There are several reasons why institutional architects might adopt ambiguous
design features such as the above. In what follows, I outline four reasons that are
particularly relevant to impartial democratic regulation.220 Although the reasons
overlap at several points, each brings some precision to the insight that ambiguity
help[s] promote a fairer system.221

1. Preventing Fixation on Process

We previously saw that a problem for impartial regulation is keeping decision
makers from focusing on process, and that constraint solutions often aggravate this

214 Courtney, supra note 12 at 29-31.
215 Ibid. at 29.
216 An anonymous commissioner, supra note 29.
217 Elections Canada observes that applying these substantive criteria is an enormous task, which
requires a delicate balancing act that must take into account human interests as well as geographic
characteristics (supra note 210 at 13). See also Rache, supra note 153 at para. 32 (balancing
readjustment criteria is not an exact science).

218 Courtney, supra note 12 at 259.
219 See Federal Elections Boundaries Commission, online: Elections Canada .

220 Among other notable contributions is Cass Sunsteins theory of incompletely theorized
agreements. Sunstein notes that some judicial decisions avoid aggravating political tensions by
issuing one-off judgments while keeping underlying reasons ambiguous (Cass R. Sunstein, Legal
Reasoning and Political Conflict (Oxford: Oxford University Press, 1996) at 37-38, 102-06, 108-10
[Sunstein, Legal Reasoning]).

221 Khullar, supra note 29. The commissioner colourfully describes the FEBCs ambiguity as part of

a recurring approach: Some might say it is the Canadian way of muddling through (ibid.).

[Vol. 53

MCGILL LAW JOURNAL / REVUE DE DROIT DE MCGILL

38

problem. Constraints draw attention to process and often catalyze cycles of partisan
procedural manipulation. A first reason for preferring ambiguous design, then, is that
it can help avoid the traps of regulation by constraint. Lessig writes that ambiguous
regulation functions not by clarifying, but by blurring.222 In short, ambiguity
confuses.223 [W]hile we ordinarily think of law as functioning to clarify
obligations and norms, he notes, here it functions by obscuring what was clear.224
One application that Lessig does not contemplate is the use of ambiguous rules of
decision making to avoid the hazards of constraint. Commissioners note that the
thoroughly ambiguous FEBC rules provide no overriding theory or central
authority.225 The rules therefore present little of the clarity and focus that elsewhere
prompt decision makers to compete and try to win changes to procedure that are
favourable to themselves. By avoiding rule clarity and the partisanship it catalyzes,
ambiguity can be a practical means of promoting impartiality.

Importantly, ambiguous rules can also go some way toward solving a difficult
problem: avoiding the clear rules that catalyze partisanship but keeping commissions
nonetheless bound by rules. In this way, ambiguous rules potentially improve on the
laissez-faire approach that Issacharoff and others promote. Issacharoff, as we saw,
advocates withdrawing key constraint rules under the VRA to encourage the
development of informal norms of impartial democratic regulation.226 That constraint
rules often catalyze partisan conflict is a powerful argument against them. But it does
not necessarily commend an alternative laissez-faire strategy of withdrawing most
rules. Decision making ought in general to remain rule-bound, in part for the
straightforward reason that democratic regulation should not be arbitrary. A
commission should not, for example, select boundaries without a stipulated set of
governing criteria and procedures.227 Regulation should pursue rational ends such as
allowing cohesive social groups to vote together as constituents in a single riding. A
laissez-faire solution may or may not see decision makers develop boundary-drawing
practices that are based on rational criteria. In contrast with laissez-faire approaches,
ambiguous regulation does not eliminate rules, but rather maintains them and makes
them more numerous, complex, and conflictual.

2. Promoting Broad Impartiality

Avoiding arbitrary decision making is just one reason why keeping an ambiguous
set of rules in place can be advantageous. Ambiguous rules may also promote
impartial judgment in accordance with the broader of the two models of impartiality

222 Lessig, supra note 26 at 1010.
223 Ibid. at 1016.
224 Ibid. at 1015.
225 See Part IV.B.3.
226 Issacharoff, Section 5, supra note 105.
227 The most extreme laissez-faire model would return carriage over readjustment to legislatures

acting through regular acts of legislation.

39

R. LEVY ELECTORAL-BOUNDARY POLITICS

2008]

outlined in Part II. Recall that decision making following the broad model draws on
diverse and even incompatible sources: facts and arguments of numerous kinds and
from numerous contributors, heard by various decision makers, and embodied in
complex networks of informal norms. Decisions following the alternative, balance
model of impartiality are products of compromise between mutually opposed groups
or interests.
Ambiguous rules are more consistent with broad impartiality than with
impartiality as balance for at least three reasons. First, ambiguous rules do not
artificially simplify decision making as both constraint and balance impartiality tend
to. Broad impartialitynuanced and richly variedis inherently ambiguous. Clearly
articulating what would make a decision maker more impartial in a given context is
often not feasible. One commissioner says of the FEBCs, for example, that their
criteria inevitably cannot be clearer.228 Considerations thought essential to the
process include defining discrete communities and fitting them within electoral
boundaries. These considerations conflict with a host of other criteria, depend on
judgments about the value of assorted communities, and yield no obvious single
line.229 Indeed, ambiguous and contested criteria and procedures
boundary
predominate in almost any context of democratic regulation.230 Legislation therefore
seldom sets out with any precision the standards that inform broadly impartial
decision making. More commonly, indistinct informal customs embody these elusive
standards and their [c]uriosities, anomalies and contradictions.231 Yet there is a
powerful inclination among reformers to systematize and simplify. From time to time
we see calls for clarifying language, as in Millikens and Lorties proposals to give
more definite meaning to communities of interest,232 and Hasens preference for a

228 An anonymous commissioner, supra note 29 ([i]t just cannot happen). Courtney similarly
acknowledges that clarification of statutes and guidelines of the FEBCs is not likely to remedy
their imprecision and elusiveness (supra note 12 at 259).

229 For example, FEBC commissioners generate computer models to consider how a slight
boundary shift affects district demographics. See Interview of Jean-Pierre Kingsley (5 July 2007)
[unpublished, on file with author]. Based on seemingly simple hard numbers, this process is actually
profoundly complex when several ethnic, linguistic, socio-economic, or other groups occupy and
effectively compete over one geographic space, especially in major urban centres. Other factors are
still less deterministic and less amenable to numeric analysis, such as local history and considerations
of social justice.

230 See e.g. judicial appointments (Weinrib, supra note 127 at 110) and defining election campaign
rules (Issacharoff & Karlan, supra note 102). More generally and across many contexts, some
commentators now view considerable ambiguity as inevitable and doubt whether rule of law
principles of clarity and predictability are realistic. See e.g. Lon L. Fuller, The Morality of Law, rev.
ed. (New Haven, CT: Yale University Press, 1969) at 33-94 (the noted rule of law principles);
Sunstein, Legal Reasoning, supra note 220 at 102-06 (casting doubt on the principles).

231 Elizabeth Longford, Prelude in Lord Longford, A History of the House of Lords (Thrupp,

U.K.: Sutton, 1999) 10 at 10 (informal norms of governance).

232 Milliken Report, supra note 98, vol. 3 at 33:25.

[Vol. 53

MCGILL LAW JOURNAL / REVUE DE DROIT DE MCGILL

40

seamless web of impartial criteria for regulating politics.233 Impartiality as balance
similarly elucidates and refines questions up for determination. But it thereby
arranges complex policy considerations along an artificially small number of clear
poles, such as Democrat versus Republican, or white versus black or Latino.234 In its
tendency to polarize and distort complex issues into reductive and binary terms, 235
this form of decision making is thus incompatible with the nuance of broad
impartiality.236
Ambiguous rules do not constrain decision making or subject it to a clear series
of conditions. The rules are written down, but beyond this basic clarity they are
markedly open-ended. They set out criteria and procedures that only guide decision
making at the outset, without clarifying how the rules should interact with one
anotheror how they should apply to redistricting.237 Says one commissioner, While
we were aware we were to consider community of interest or identity and history
pattern, how we would do that was left up to us.238 Ambiguous rules, even though
they are written, are therefore most similar to the contradictory, vaguein short,
ambiguousinformal norms of democratic regulation. The former, then, avoid the
artificial simplifications of both balance impartiality and constraint.

Second, ambiguous rules also appear actively to promote broad impartiality. They
can improve on most of the traditional strategies of formal regulation, which define
impartiality in ever-more precise terms, counterbalance decision makers of opposite
views, or simply exhort commissioners to be impartial. A laissez-faire regulation
strategy does little to ensure that decision making will develop following the broader
impartiality model. And a bare exhortation to be impartial may be little more likely to
ensure broad impartiality than laissez-faire.239 In contrast, ambiguous rules can go a
step further and lead decision makers through numerous processes of deliberation, as
is typical of broad impartiality. Complex sets of rules like those of the FEBCs can
direct decision makers to range broadly in their work. Each substantive rule pushes

233 None of these reformers goes the next step and states a full range of specific rules. Note also that
these are examples of the constraint approach to regulation. Balance impartiality is at root a variety of
constraint, in light of the similarities outlined in this section.

234 See Pildes, supra note 25 at 968. For the example of the U.S. Federal Elections Commission, see

supra note 137.

235 See Jacobson, supra note 9 at 27; Michelman, Suspicion, supra note 71 at 680.
236 See Pildes, supra note 25 (noting how formal laws of democratic regulation can displace more

nuanced norms of trust and trustworthiness).

237 On the distinction between guiding rules and constraining rules see e.g. Joseph Raz, The Rule
of Law and Its Virtue (1977) 93 Law Q. Rev. 195 at 198; T.R.S. Allan, The Rule of Law as the Rule
of Reason: Consent and Constitutionalism (1999) 115 Law Q. Rev. 221 at 225-27.

238 An anonymous commissioner, supra note 29.
239 Some observers, such as Canadas former long-serving Chief Electoral Officer, Jean-Pierre
Kingsley, believe in the effectiveness of such exhortations (Kingsley, supra note 228). Note that the
EBRA does not explicitly state impartiality is its goal, though the point is certainly implicit. But in this
implicit sense all readjustment commissions call on decision makers to be impartial, including some
that are in practice impartial and also others that are not.

41

R. LEVY ELECTORAL-BOUNDARY POLITICS

2008]

deliberation in a new direction and each procedural step pushes decision makers into
discourse and cooperation with other participants. Going beyond a vague exhortation
to be impartial, FEBC rules instead direct decision makers to deliberate and interact
broadlyto act like broadly impartial decision makers.

Broad impartiality avoids focusing on procedural exploitation, unduly rigid
ideology, or narrow sets of social interests. Open therefore to any cogent and relevant
argument, it is in this sense relatively rational.240 In practical terms, being exposed to
the full blast of … sundry opinions,241 and to each other, decision makers should be
more sensitive to perspectives that they otherwise would not encounter. Analysts
often make this suggestion, albeit more typically in the context of the broader public
arena.242 A rational model of decision making should, in turn, be particularly
egalitarian because it is amenable to all helpful submissions of fact and argument,
including public submissions. As Joseph Raz describes this ideal, fact and argument
persuade according to their inherent informational or rational value, and not as a
function of the contributors formal power.243 Indeed, in comments that Courtney
broadly affirms,244 FEBC commissioners report being strongly influenced by citizen
submissions regarding facts on the ground in a given community.245 One
commissioner reports, We took very seriously the objections and proposals
presented to us and tried to accommodate community sentiment as we understood it
from the hearings. We extensively redrew our initial map in response to public
submissions.246 In fact, another commissioner notes that [b]y and large we
welcomed the submissions, more so from the public than the politicians.247 Still

240 The literature contrasts partisanship with impartial, principled or reasoned decision on an
issue on its merits (Hasen, supra note 12 at 974-79, 982; Kari Palonen, Four Times of Politics:
Policy, Polity, Politicking, and Politicization (2003) 28 Alternatives 171). See also Nedelsky, supra
note 74. Economic notions of rationality based on self-interest offer another, very different
understanding.

241 Michelman, Brennan, supra note 79.
242 Mazzone, supra note 25 at 51 ([A] person who, because of her participation in civic networks,
is accustomed to negotiation and compromise may be less demanding and less selfish in her daily
transactions with others than someone who has never learned similar cooperative habits); Jon Elster,
The Market and the Forum: Three Varieties of Political Theory in James Bohman & William Rehg,
eds., Deliberative Democracy: Essays on Reason and Politics (Cambridge, Mass.: MIT Press, 1997) 3
at 12 (describing similar effects in the context of democratic discourse).

243 Morality, supra note 185 at 29.
244 Supra note 12 at 136.
245 There are of course exceptions. One commissioner reports learning that four unusually
knowledgeable constituents who made submissions had been coached by members of an affected
political party (An anonymous commissioner, supra note 29).

246 An anonymous commissioner, supra note 29. Two commissioners recall public submissions, on
issues such as ease of transportation in ridings and keeping communities more intact, being highly
influential (Two anonymous commissioners, supra note 29). See also Kingsley, supra note 229
(public submissions significantly impact the process).

247 An anonymous commissioner, supra note 29 ([t]he politicians submissions were more political
and were received as such). Parliamentarians submissions are influential primarily when echoed by

[Vol. 53

MCGILL LAW JOURNAL / REVUE DE DROIT DE MCGILL

42

another commissioner relates that the submissions were [a]bsolutely crucial and led
to huge changes. … Public submissions are key inputs into the whole thing.248 These
reports suggest that at least some ambiguous processes satisfy the prediction of
relatively rational decision making sensitive to cogent public contributions.

Third and finally, an ambiguous system makes balance impartiality difficult to
sustain. A complex set of incommensurable and unpredictable rules scrambles the
delicate balances of power on which this form of impartiality is premised. With a
chaotic multiplicity of rules and players such as that of the FEBCs, it may be harder
to sustain conceptual simplifications, clear polarities, and counterbalanced interests.
To illustrate this point, I will begin with the case of the Supreme Court of the United
States. Observers typically recognize the polarization of its deliberations between two
political blocs: conservative and liberal.249 The neatness of this distinction is
striking, imposed as it is on a much more plural range of ideologies held by members
of the court and the broader public. The United States Senates two-party polarization
seems to bring order to an otherwise chaotic array of libertarians and moral
conservatives, liberals and class-conscious leftists, originalists and textualists, and
many others.250 In comparison, polarization is less likely given a set of poles whose
number and relative positions are obscure and ever-changing. When ambiguous
arrangements disrupt conceptual simplifications, a decision makers personal
alignment with a single umbrella ideology is less tenable. Justices of the Supreme
Court of Canada evidently maintain an intelligible but relatively broad set of
ideological perspectives. The set is large and chaotic enough to frustrate, at least in
part, alliances and ideological cohesion among the judges. Indeed, Canadas highest
court issues decisions that are generally less predictable in their results than those of
the Supreme Court of the United States.251

If decision makers cannot sustain conceptual simplifications, they may be more
likely to address the substance of an issue relatively free of predisposition. For this
reason it again follows that ambiguous systems should produce relatively rational
judgment. Additionally, because ambiguity scrambles alliances and policy
simplifications, decision making should be relatively cooperative. This is a
characteristic often ascribed to trusted, trustworthy, and impartial decision making.252
In comparison with balance impartiality and constraint approaches, it should be
uncommon to see decision makers compete as mutually distrustful partisans.253 In

community input (ibid.). Some had useful observations, but many just wanted to make the
Commission the target of attacks, and that was not helpful (ibid.).

248 An anonymous commissioner, supra note 29.
249 Jacobson, supra note 9 at 27.
250 Ibid.
251 See generally Peter McCormick, Follow the Leader: Judicial Power and Judicial Leadership on
the Laskin Court, 19731984 (1998) 24 Queens L.J. 237 at 268; Birds of a Feather: Alliances and
Influences on the Lamer Court 19901997 (1998) 36 Osgoode Hall L.J. 339.

252 See supra note 189.
253 Part of the intentand certainly the effectof balancing opposed decision makers off of each
other is to have them hold each other in mutual check. Each is watchful for offences by the other, and

R. LEVY ELECTORAL-BOUNDARY POLITICS

2008]

practice, many commissioners report not perceiving each other as competitors, but
rather deciding cooperatively through collective deliberation and even consensus. As
one commissioner notes, Decision making was highly cooperative. We listened to
one anothers suggestions and ideas and made decisions based on what seemed to
work best. It was consensus reasoning.254 The same appears to have been the case in
several FEBCs.255

43

3. Precluding Partisanship

Under the previous two headings, an assumption was that decision makers intend
in good faith to make impartial decisions when possible.256 An alternative assumption
is more cynical, and more common. The bulk of writers who analyze ambiguous
decision making systems start from the premise that decision makers are partisans.
These analysts assert that the efforts of partisans are less effective amid ambiguity.
Given ambiguous procedures and legal norms, partisan decision makers might be less
able to predict how manipulating procedures will affect their own fortunes.257
A helpful example comes out of a recent work by Adam Cox, who presents one
of several new contributions characterizing ambiguity as a tool of legal regulation.258
His contribution is also a rare application of ambiguity scholarship to readjustment in
particular. Cox praises the longstanding rule mandating a long period of time, a
temporal floor, between rounds of readjustment. One intention is to curb the
effects of partisan gerrymandering by promot[ing] beneficial uncertainty in the
redistricting process.259 Cox explains that [w]hile redistricting authorities can make
some predictions about voting behavior, … the accuracy of those predictions
decreases as one moves further in time from the point of prediction.260 The
observation is a clever one. It identifies a familiar element of readjustment as a source
of ambiguity. Readjustment is at root a response to the changing demographics of
electoral districts. Long periods between the drawing of district maps and their use in
elections can produce marked unpredictability; most North-American electoral
districts undergo sizeable demographic shifts over a ten-year timeframe.261 Until
recently in both Canada and the United States, readjustment occurred on a fixed

each seeks to block and counter the attempts of others to gain the upper hand in democratic regulation.
See e.g. Steven Calabresi & Kevin Rhodes, The Structural Constitution: Unitary Executive, Plural
Judiciary (1992) 105 Harv. L. Rev. 1153 at 1156; The Federalist, supra note 100, No. 48 (Madison)
at 311.

254 An anonymous commissioner, supra note 29.
255 See Khullar, supra note 29 (the three of us worked well together, and I guess also shared some

fundamental values of how to approach the task). See also Kingsley, supra note 228.

256 They will seize the opportunity to manipulate rules in their favour perhaps only if opponents

otherwise will do so.

257 Issacharoff calls this ends-oriented manipulation (Gerrymandering, supra note 9 at 595).
258 Supra note 7.
259 Ibid. at 769.
260 Ibid. at 769-70.
261 See generally Pal & Choudhry, supra note 160 at 4-5.

[Vol. 53

MCGILL LAW JOURNAL / REVUE DE DROIT DE MCGILL

44

decennial timetable. However, over the past few years, Texas and four other states
have sought to readjust districts outside the traditional cycle.262 In its recent ruling on
the Texas readjustment controversy, the Supreme Court of the United States held that
a decennial floor rule is not constitutionally required, potentially weakening this
temporal-ambiguity feature in the United States.263 But the floor rule fortunately
remains a constitutional requirement in Canada.264
Offering another intriguing example, Jason Mazzone identifies how some forms
of federalism generate ambiguities precluding partisan decision making:

An ambiguous division of power … creates uncertainty regarding which
government, national or state, will eventually make the decisions on a particular
matter. This ambiguity in turn casts doubt on the merits of pursuing one avenue
of influence rather than another.265

Finally, in a growing subset of ambiguity cases, commentators elevate an insight
of John Rawls from thought experiment to practical tool of regulation. In Rawls
celebrated hypothetical case, founders of a political community, being oblivious to
their own identities, elaborate its features without self-interest.266 A number of later
authors imagine concrete applications. These mostly appear in contexts where
entrenchment is a live risk, such as at a time of constitutional amendment. In one
recent effort to rewrite the founding document of a developing state emerging from
conflict, international advisors called for a Rawlsian moment.267 The countrys first
elections were to be held only after passage of the new constitution, which would
spell out the powers of incumbent governments.268 To the advisors disappointment,
the elections went ahead before completion of the constitution, and the elected party
wrote, as predicted, wide governing powers into the new document. Several works
address analogous effects in other contexts, ranging from the determination of health
care rights to the design of basic constitutional rules.269 In Canadian electoral-
boundary readjustment, John Courtney notes that the perennial minority governments
of the 1960s probably helped prompt Parliament to create the FEBCs.270 With that
decades uncertain electoral prospects for political parties, incumbents could not rely
on holding power for long, and could count even less on controlling the next

262 See Cox, supra note 7 at 751-52.
263 See Perry, supra note 4.
264 Constitution Act, 1867, supra note 48, s. 51. However, it should be noted that Parliament
breached the rule, apparently for partisan purposes, with a two-year delay in the early 1990s
(Landes, supra note 29).

265 Supra note 25 at 57-58.
266 A Theory of Justice (Cambridge, Mass.: Harvard University Press, 1971).
267 For this account I am indebted to John Barker of the University of Cambridge, a member of the

Malawi advisory group.

268 Ibid.
269 See e.g. Russell Korobkin, Determining Health Care Rights from Behind a Veil of Ignorance
(1998) U. Ill. L. Rev. 801; Adrian Vermeule, Veil of Ignorance Rules in Constitutional Law (2001)
111 Yale L.J. 399.

270 Supra note 12 at 44-52.

45

R. LEVY ELECTORAL-BOUNDARY POLITICS

2008]

readjustment. The parties unpredictable futures apparently provided some incentive
to place readjustment in the hands of independent commissions, and to bring more
impartiality to readjustment.

In Part IV, I will discuss how the FEBC systems elaborate set of ambiguous
features generates unpredictable outcomes and, among other effects, precludes
partisanship. It is sufficient for now to note that the unpredictable system would
confound attempts to draw partisan lines. Most of the time, we dont even know the
effect anyway, one commissioner states, referring to how readjustment impacts
election outcomes.271

4. Developing Informal Decision-Making Norms

An important benefit of ambiguous systems is their potential to promote the
development of informal norms of trust, trustworthiness, and impartiality. Formal
rules, however cleverly designed, generally leave open loopholes that committed
partisans may seek and exploit: the unusual shapes of gerrymandered districts are
prominent examples. However, the FEBCs ambiguous design apparently promotes
impartiality that is reinforced in norms of trust and trustworthinessor, in the words
of one commissioner, in the mythology [of] the neutral fair system. … The ambiguity
and mythology work[] in concert in this way.272 Indeed, we might expect at least
some of the effects outlined aboveending process fixation, promoting broad
impartiality, and precluding partisanshipto produce corresponding changes in the
attitudes and habits of decision makers. Sustained over time, these aspects of
impartial decision making may solidify into informal norms as members of the public
and decision makers themselves come to expect impartial decision making.273 The
idea of discernible institutional cultures developing over long periods of time is
widespread.274 Lessigs own illustration of behavioral regulation,275 as we saw,
raises the examples of racial-segregation prohibitions and policies mandating a daily
salute to the flag in schools. Both regulated behaviours inculcate attitudes that in turn
reinforce desired behaviours by developing lasting informal norms. It must be noted,
however, that Lessig does not explore the behavioural side of ambiguity methods, but
rather places them within his semiotic category.276 Nevertheless, if ambiguous

271 An anonymous commissioner, supra note 29. Another anonymous commissioner adds that

decisions [do] not appear to be … geared to manipulating electoral outcomes (supra note 29).

272 Khullar, supra note 29. In addition, an anonymous commissioner cites the importance of

developing symbolic legitimacy (supra note 29).

273 As noted, trust and trustworthiness are premised on decision makers and others expectations of
good decision making (Braithwaite, supra note 19 at 344-47; Jones, supra note 90). Note the
alternative possibility that partisans will find and exploit loopholes before trust develops.

274 See e.g. Heather Gerken, Second-Order Diversity (2006) 118 Harv. L. Rev. 1099 at 1147.
275 Lessig, supra note 26 at 1008. See also e.g. the remarks of Koh in Toope et al., supra note 129 at
38 (describing a process of norm internalization, driven by legal process [emphasis in original]);
Mazzone, supra note 25 at 57-58.

276 Lessig, ibid.

[Vol. 53

MCGILL LAW JOURNAL / REVUE DE DROIT DE MCGILL

46

systems produce decision making that is at first impartial in effect, the behavioural
insight suggests that cultures of impartiality should develop over time.

Less straightforwardly, ambiguous systems can facilitate the development of
complex relations among decision makers that characterize trust, trustworthiness, and
broad impartiality. As we saw, commentators commonly describe trust and
trustworthiness as sustained in dense networks of rules, values, symbols, social
meanings, agreements, social bonds, organizations, and loyalties. An ambiguous
complex of rules and participants can lay the groundwork for developing these
informal norms. That groundwork can in turn put decision makers in a position to
begin cooperating and deciding rationally.277 This idea has precedents in a number of
contexts. For example, Mazzone points to how Progressive Era reformers … sought
to increase [trust in the public arena] in the United States by creating networks of
voluntary associations.278 Mazzone and another author, Denise Scheberle, creatively
and persuasively reread the dynamics of governance in the United States by
suggesting how dispersing power into complex forms can engender trust, and by
concentrating on the development of trust amid U.S. federalism.279 Theories of urban
planning provide a further, and better known, example out of a loosely analogous
context. Some authors call for municipalities
to adopt relatively complex
arrangements of physical structures, businesses, public services, and demographics.
The aim is to mimicand thereby perhaps to seedthe dense networks of
interpersonal interactions and trust, and the richness and variety of advantages, that
usually characterize communities that are safe, economically viable, and appealing.280

Returning to the FEBCs, we have seen that these bodies feature a host of
complexities, such as broad consultation, diverse decision makers, and a richly
convoluted substantive decision-making framework. By engaging its assorted
participants together in this circuitous process for over two years, the FEBC system
may have encouraged the development of informal networks of trust among its
participants. These participants include the commissioners, organized civil-society
groups such as farmers representatives, municipal-government representatives (who
are among the most common participants in the federal readjustment process),

277 We have already seen a number of examples of informal rules, values, symbols, and social
meanings in references to cultures of cooperation; to impartiality as a distinct, if tenuous, mythology,
to public perceptions of FEBC trustworthiness, and to egalitarian ethic regarding submissions to the
FEBCs, which in turn invites robust public involvement. Under the present heading, I will focus more
on agreements, social bonds, organizations, and loyalties among decision-making participants.

278 Supra note 25 at 36.
279 Denise Scheberle suggests that one of the many effects of complex federalism can be enhancing
relations of trust. Looking at case studies in environmental regulation, Scheberle traces the
development of trust amid networks of interactions among federal and state governmentssometimes
finding surprising levels of cooperation (Scheberle, supra note 176). To be sure, neither author
suggests that trust pervades at every level of U.S. governance.

280 See e.g. Peter Katz, The New Urbanism: Towards an Architecture of Community (New York:

McGraw-Hill, 1994).

47

R. LEVY ELECTORAL-BOUNDARY POLITICS

2008]

representatives of First Nations governments, individuals speaking as members of
identity groups such as the New Brunswick Acadians, other interested individuals,
and MPs affected by proposed changes.281 A simpler decision-making system might
see just a few decision makers exercising clear decision-making control, thereby
overshadowing the roles of diverse public and private participants.282 The example of
the FEBCs suggests that in some cases of democratic regulation a valuable design
strategy may be to avoid unambiguous lines of top-down control. Relatively
numerous and variegated power arrangements can prompt the development of the
complex normative networks that can in turn help solidify trust, trustworthiness, and
impartiality in decision making.

IV. The Scope and Limits of Ambiguous Systems

A. Extending Ambiguous Regulation

We have seen that ambiguity methods address the risk of entrenchment by
counterintuitively blurring the lines within which decision makers exercise power. An
important question is how far we can push this strategy. Going beyond Lessigs
conception, decision making can feature not one but many sources of ambiguity to
yield the benefits outlined in the previous part. However, some writers assume that

281 See supra notes 227-32, 237-39. See also Federal Elections Boundaries Commission for
Alberta, Part III: Summary of Submissions and Post-Hearing Decisions in Report of the Federal
Electoral Boundaries Commission for Alberta (15 January 2003), online: Elections Canada
(citing involvement of
municipal representatives); Federal Elections Boundaries Commission for Manitoba, Report of the
Federal Electoral Boundaries Commission for Manitoba (15 January 2003), online: Elections
Canada (citing involvement of
Manitobas Southern Chiefs Organization); Federal Elections Boundaries Commission for New Brunswick,
Report of the Federal Electoral Boundaries Commission for New Brunswick (10 January 2003), App. C,
online: Elections Canada (citing involvement of Patton MacDonald, a representative of New Brunswick farmers).
282 Note that this suggests a further form of evidence of robust public influence in the
ambiguous FEBC processes. Adopting reasoning apparently in the mode of rational-choice
analysis, Mazzone argues that trust develops when dividing power between the national
government and the states provides greater opportunities for citizen groups to influence politics
and for individual citizens to participate in public life (supra note 25 at 27). Mazzone argues
that federal divisions of power thereby engender trust, cooperation, and social justice.
However, evidence for this particular claim appears to be mixed. Mazzone claims that [g]roups
like the American Association of Retired Persons (AARP) … are likely to dominate in a
[unitary] system (ibid. at 44). In contrast, federal systems require such large groups to focus on
many sites of influencefor example, on fifty state governmentsand federalism provides
opportunities for smaller, weaker organizations to compete for influence and pursue their
agendas. In a federalist system, it is easier for smaller organizations to mount opposition to even
powerful groups because [larger groups] resources will be more diffused (ibid. at 45). Whether
Mazzone is right depends on a number of particulars, such as the actual size, and therefore the
state-level influence, of groups such as the AARP.

MCGILL LAW JOURNAL / REVUE DE DROIT DE MCGILL

48

extensive ambiguity and unpredictability produce poor functionality.283 I address such
concerns in this part. I outline limiting conditions for the extensive use of ambiguity
as a tool of democratic regulation and provide examples illustrating these conditions
both generally and more specifically within the contexts of readjustment in Canada
and the United States. Successful applications of the ambiguity methods I have
outlined
and
comprehensiveness.

equivalence,

conditions:

to meet

[Vol. 53

stability,

appear

three

B. Conditions for Extended Ambiguity

1. Equivalence

Ambiguous systems should firstly fulfill an equivalence condition: roughly
speaking, they must function equally as well as clearer and more predictable
alternatives.
Democratic-regulation cases often meet this condition. In scholarship on such
cases, there has been a longstanding recognition that no single model of government
accommodates all the factors that potentially contribute to good democratic design.284
For example, a perennial question is how to ensure that voters will express their own
substantive electoral preferences, rather than preferences influenced by irrational
considerations such as group allegiances and polls.285 A common practice in Canada
is the polling blackout as elections near.286 However, an alternative can be to adopt a
circuitous route from individual votes to electoral outcomes.287 This approach is less
likely to raise free expression concerns and is somewhat common in voting systems
in other countriesfor example, in mixed systems in which voters simultaneously
vote via majoritarian and proportional voting models.288 An ambiguous voting

283 See e.g. Karlan, Fire, supra note 39 at 733-35, 741, n. 7; Sunstein, Legal Reasoning, supra
note 220 at 102-06. In one interesting work, M. Elizabeth Magill analogizes the confusing American
case law on the separation of legislative, executive, and judicial powers to the chaos of the three-
body problem known to physicists (M. Elizabeth Magill, The Real Separation in Separation of
Powers Law (2000) 86 Va. L. Rev. 1127 at 1128-29). In another context see generally Fitzpatrick,
who discusses chaotic informal norms of property use generating failure in property-rights regimes
(supra note 95).

284 See Kenneth Arrow, A Difficulty in the Concept of Social Welfare (1950) 58 Journal of

Political Economy 328 at 328-331.

285 See e.g. Andr Blais & Robert Young, Why do People Vote? An Experiment in Rationality
(1999) 99 Public Choice 39 at 39-40; Guy Lachapelle, Polls and the Media in Canadian Elections:
Taking the Pulse, vol. 16 (Toronto: Dundurn Press, 1991) at 13-14.

286 See J. Andrew Fraser, The Blackout of Opinion Polls: An Assault on Popular Sovereignty

(1995) 4 Media and Communications Law Review 365 at 379-80.

287 See Richard Johnston et al., Letting the People Decide: Dynamics of a Canadian Election

(Montreal: McGillQueens University Press, 1992) at 197-98 (factors controlling strategic voting).

288 New Zealand, Italy, and Israel, among others, employ mixed systems. See Pippa Norris,
Choosing Electoral Systems: Proportional, Majoritarian and Mixed Systems (1997) 18 International

49

R. LEVY ELECTORAL-BOUNDARY POLITICS

2008]

system of this kind can make the connection between a vote and electing a particular
government more unpredictable, in turn diminishing considerations of process and
strategy. Citizens are more likely, then, to vote based on their substantive policy
preferences. Importantly, the more unpredictable process still leaves unimpeded the
right to cast a vote;289 the example illustrates how different options for voting-system
design can function equally well.

Some ambiguity techniques will not satisfy the equivalence condition. For
example, certain regulatory problems cannot be approached piecemeal. There are
many cases where ambiguous systems prevent coordination or institute through
discrete steps decisions that should be made at the level of the complex whole.
Indeed, ambiguity is often a burden for federal systems. One example is the problem
of urban revitalization in Canada and the difficulties of reaching agreement among
the three levels of government. Redevelopment plans can stall for years as
ambiguously divided powers frustrate any coherent plan.290 But the full story of
federalism is complex and, as already noted, federalism can have positive effects such
as precluding partisanship.

a. Equivalence in Readjustment

How might a readjustment regime as ambiguous as Canadas be less effective
than simpler systems, even if it is more impartial? There are at least two
possibilities.291 The first and primary consideration should be whether the FEBCs
follow appropriate substantive readjustment criteria as well as other systems do. By
nearly all accounts the FEBCs diligently follow the rules that govern them.292 As in
the case of voting, there are rules producing unpredictable outcomesdue here to the
elaborate and conflictual criteria and procedures as well as the diverse participants
involved. The FEBCs nevertheless consistently readjust electoral maps by following
the substantive criteria set out in cases and legislation. As noted in the previous part,
rules continue to guide readjustment, even if the rules are ambiguous as a whole.
However, we cannot indefinitely continue adding substantive criteria to a

Political Science Review 297. Note however that mixed systems can also be simple, for example if
they distribute legislative seats based primarily on the proportional vote.

289 Note however that there is arguably democratic value in voting strategicallyand generally in
the power not only to express a position but also to exert concrete influence over which political party
triumphs in an election. There is no correct answer in this debate; where one comes out in it depends
on how much weight one ascribes to the different democratic values in conflict. However, the debate
illustrates the broader point about the equivalence of several design options.

290 The city of Toronto has faced this problem since basic plans for a renewed waterfront were
announced in 2000. Antagonistic federal, provincial, and city governments have attempted in vain to
coordinate and have made at best only halting progress. See Jennifer Lewington, Waterfront
Revitalization Gets New Lease on Life Globe and Mail (28 July 2004) A9.

291 A third possibility is that ambiguous systems are ineffective given the room in which to hide
that flexibility and ambiguity might give partisans aiming to manipulate procedures (Carter, supra
note 98 at 58). However, the main thrust of this article aims to refute that argument.

292 See Courtney, supra note 12.

[Vol. 53

MCGILL LAW JOURNAL / REVUE DE DROIT DE MCGILL

50

readjustment system; the criteria must play rational roles, such as directing
commissioners to consider the locations of cohesive groups. Given the expansiveness
of the current set of substantive criteria for readjustment, the list of criteria cannot
likely become longer without veering toward irrelevance.
A number of years ago Issacharoff interestingly called for the randomization of
readjustment.293 The computer-generated maps he contemplated would have little
relation to the locations of cohesive communities and other traditional readjustment
factors. This approach would sacrifice most other ends of readjustment to the cause of
impartiality. The merits of the approach cannot be dismissed entirely, given the
partisan state of much American readjustment. But circumstances in Canada do not
call for such radical solutions. The Rache case, for one, demonstrates the importance
to certain cultural minorities of unified representation and of members voting together
as coherent units in elections.294 In the controversy leading up to Rache, many New-
Brunswick Acadians forcefully argued against their transfer to an Anglophone
riding.295
A second possible consideration affecting equivalence is the democratic
legitimacy of different readjustment systems. Legislatures arguably bring larger
measures of democracy to the task of readjustment. Robert Post argues that only
elected representative bodies have the legitimacy to define the basic ground rules of a
democracy.296 The FEBCs are appointed bodies, and their ambiguous procedures
effectively block internal decision making from public view and in this sense prevent
public accountability. Nevertheless, the commissions democratic legitimacy appears,
if anything, to be stronger than that of legislatures. One important consideration is
that representative democracy is not the only way in which a democracy can function.
Taking citizens submissions directly, where it is feasible to do so, can be more
effective. As we saw in the previous part, the ambiguous FEBC system places all
contributors, from MPs to lay citizens, on roughly equal footing. The commissions
democratic bona fides are in this respect arguably at least as strong as Parliaments.
We also previously saw a more straightforward democratic argument for FEBCs.
Throughout much of Canadas history, Parliament often did not write the ground rules
of the democratic process in a fair manner. There is no legitimate democratic

293 Samuel Issacharoff, Judging Politics: The Elusive Quest for Judicial Review of Political

Fairness (1993) 71 Tex. L. Rev. 1643 at 1647, 1693-95.

294 Supra note 153. The desire for joint representation raises intriguing questions about the function
of a democratic vote. When a community votes as a discrete unit, there is no guarantee that members
will vote together for any given party. Nevertheless, the voters involved in litigation and other forms
of advocacy, including the 2656 Acadians who signed a petition sent to the provinces FEBC,
apparently saw voting together as a way to preserve their community.

295 Ibid. at paras. 11-15.
296 Robert Post, Democracy, Popular Sovereignty, and Judicial Review (1998) 86 Cal. L. Rev. 429

at 432.

R. LEVY ELECTORAL-BOUNDARY POLITICS

2008]

prerogative permitting legislators to gerrymander unfettered, and thereby to remain in
power contrary to public wishes.297

51

2. Stability

The stability and prospectivity of rules are facets of the rule of law.298 There is a

risk that an unusually complex process will fall apart over time and stop guiding
governance. On the other hand, even chaotic systems can be stable systems. A
decision-making process can persist over time, consistently yielding a given type of
decision even if the details of decisions are unpredictable.

I return first to the example of voting systems. These can function in a
dependable manner, establishing governments according to a reliable election cycle
even if an election result cannot be determined in advance. Ambiguity as a tool for
regulating politics must, in general, be similarly stable because entrenchment is a
constant risk, not a fleeting possibility. Merely transient ambiguities such as those
described by Lessigintermediate states that dissipate once cultural change has
occurredtherefore have limited value. Thus in the judicial-selection conflict at the
United States Senate, ambiguities that arose during lag periods when parties adjusted
to unfamiliar rules were quickly resolved and clarity ensued. But other cases offer
examples of stable ambiguity in democratic regulation. An early and interesting
example is that of the French administrative divisions, or dpartements, which were
designed by the Abb Siyes during the Revolutionary era and persist to the present
day. These geographic subunits violated the territorial integrity of [existing]
provinceswhich were far fewer in number and reflected regional loyaltieswith
the deliberate intention of reducing cleavages within the nation.299 In the words of a
modern commentator, the plans supporters adopted this ambiguous strategy of
decentralizing the administrative system in order to regenerate the state.300 The
system did clarify things in one sense: Through this hierarchical organization of
space, the rights of citizens and the unity of the nation would be guaranteed.301 But
the new structure also deliberately cut across geographic lines and historic
communities, dividing the country into several-hundred units in order to blur
divisions and durably construct French national unity.

297 See Klarman, supra note 17 at 498 (noting that legislatures tend to make anti-majoritarian

attempts to entrench their own power).

298 Fuller, supra note 230 at 33-94.
299 Ted Margadant, Book Review of La formation des dpartements : la reprsentation du territoire
franais la fin du 18e sicle by Marie-Vic Ozouf-Marignier, (1991) 63 Journal of Modern History
396 at 397.

300 Ibid.
301 Ibid.

52

MCGILL LAW JOURNAL / REVUE DE DROIT DE MCGILL

[Vol. 53

a. Stability in Readjustment

Though readjustment in Canada is ambiguous in the ways we have seen, it is by
most accounts stabler than its American counterpart. On a simple level, the well-
established pattern of the FEBCs sees districts take shape on a ten-year cycle, albeit
with some delays.302 We have already seen that the ten-year cycle has been
jeopardized in some American states.
More importantly, the American commitment to constraint in readjustment
appears, as in the context of judicial selections, to have generated unstable patterns of
chaotic lag periods characterized by ambiguity, followed by resolution and clarity.
U.S. readjustment is complex but not lastingly ambiguous. Courts and other bodies
have frequently elaborated new rules that produce periods of procedural ambiguity.
But these periods have been transient. As Issacharoff and others relate, an initial
period of gains in black voting in the United States was followed by a rise of VRA
litigation by partisans,303 as political parties moved forcefully into the judicial arena to
combat one another. Tests centred on racial representation increasingly became mere
points around which broader political contests focused. The centrality of race in
conflicts under the VRA has thus diminished since the 1960s and the civil-rights
era.304 Partisan contest has spread from the courts to the VRA administrative-oversight
process. Here we are now seeing the VRA enforcement branch displace its long-
standing corps of trusted Justice Department lawyers.305 Observers had recognized in
these lawyers an ethic of impartial decision making, which rested on a well-
established civil-service culture.306 Though there is a cost in political capital to be
paid for transgressing informal rules of impartiality, the recent political history of the
United States has seen these costs diminish, with transgression itself being

302 See Courtney, supra note 12 at 144-48.
303 Issacharoff notes that the VRAs early success initially succeeded in breaking up single-party
dominance in the South, but that this success was in part responsible for the later partisan tinge of
VRA intervention, as the results of intervention in what then became a two-party system carried
significance for the electoral performance of each party (Section 5, supra note 105).

304 Even Republicans, whose electoral successes among the African-American community are
limited, have brought VRA challenges to readjustments, ostensibly for their minority representation-
diluting effects. See Page v. Bartels, 144 F. Supp. 2d 346 (N.J. 2001). Additionally, causes of action
under [the] one person, one vote doctrine have had the incongruous result of producing plaintiffs
whose motivation for litigation is almost entirely divorced from the doctrinal basis for their suit
(Karlan, Fire, supra note 36 at 736).

305 See Mark Posner, The Politicization of Justice Department Decisionmaking Under Section 5 of the
Voting Rights Act: Is it a Problem and What Should Congress Do? (2006) at 13-15, online: American
Constitution Society for Law and Policy (describing several forms of politicization, including precluding career staff from making s.
5 recommendations and elevating the influence of political staff in their place). On the more recent attorney
firing controversy in the United States, see also text accompanying note 188.

306 Posner, ibid. at 6. Posner served for several years at the Department of Justice on VRA

preclearance submissions.

R. LEVY ELECTORAL-BOUNDARY POLITICS

2008]

normalized.307 When the political will to breach impartiality norms developed, the
system failed to respond and protect itself. Partisan manipulation became common
and helped to rewrite the norms of democratic regulation.308

53

3. Comprehensiveness

Ambiguity as an effective tool of democratic regulation appears to require not
one or two, but rather many ambiguous rulesoutstripping Lessigs limited scenarios
of two norms in conflict. Thus Coxs floor rule, mentioned above, did not end
gerrymandering on its own in the United States, nor even in Canada before the 1960s.
More useful would be a broadly complex and unpredictable process, of which a floor
rule, for example, might be one contributing part. Most importantly, there should be
few if any exceptional sources of clarity. These can sustain partisan conflict even in
an otherwise ambiguous system. Comprehensiveness appears to mark a key
distinction between the Canadian readjustment scheme and its largely dysfunctional
American counterpart.

a. Comprehensiveness and Umpired Readjustment

There are mixed features of ambiguity and clarity in U.S. readjustment. As we
saw, the American case law is now complex and even incoherent;309 some substantive
rules are markedly ambiguous, such as the test for bizarrely shaped districts.310 On the
other hand, legislative majorities control most readjustment, including readjustment
by commission, leaving little mystery as to which political party will come out on top
in the process. In addition, we saw that the grounds for court and administrative
review remain conceptually narrow, focusing on minority disenfranchisement rather
than on fairness at large.
Most importantly perhaps, the readiness of a central umpire to settle readjustment
disputes focuses and clarifies the disputes. The administrative-review mechanism of
the VRA is statutorily required to be expeditious and is substantially faster, simpler
and cheaper than the judicial alternative.311 It is also the venue of choice in the
overwhelming majority of cases in states covered by the VRA.312 But even the judicial
process provides substantially more simplicity, clarification, and central coordination
than most features of the Canadian system. Courts review federal Canadian

307 More generally in recent years, breaking a culturally based rule of impartiality on one hand has
carried with it a meaning of political impropriety, while on the other hand it has often led to a
retroactive redefinition of the rule as having been weak, optional, or outmoded. See Hasen, supra note
12 at 957-58.

308 Ibid. See also Karlan, Fire, supra note 39 at 736.
309 See ibid. at 733-35, 741, n. 7.
310 See supra note 43.
311 Posner, supra note 305 at 7.
312 See ibid.

[Vol. 53

MCGILL LAW JOURNAL / REVUE DE DROIT DE MCGILL

54

readjustment infrequently, and reverse FEBC decisions rarely.313 Moreover, the
FEBCs are radically decentralized administrators, as I discuss below. The U.S. model
of umpiring by a central body can render a system inescapably clear and open to
exploitation. In the present decade, litigation over election disputes has burgeoned
dramatically,314 a trend that Karlan believes will continue as the proliferation of
constraints on the reapportionment process moves political contests into the
courts.315 Waiting on a court or administrative body to resolve disputed electoral
ground rules presumably adds some unpredictability to the system. However, a
central arbiter lies ready to take over decision making in all instances. Such clear and
ready litigation provides a focus for partisan conflict.

b. Comprehensiveness and Decentralized Readjustment

ten provincial

Models of decentralized power appear prima facie to improve on umpiring and its
effect of aggravating partisan conflict. The Canadian FEBC process is comparatively
decentralized in two important respects. First, there is little countrywide coordination.
Few cases in the courts produce nationally-prominent disputes over electoral
boundaries. The
readjustment commissions work highly
independently of one another.316 We were not, reports one commissioner, aware
of the specifics of [other FEBCs] proposals nor the approaches they were taking in
their decision making.317 Input by parliamentarians provides a potentially unifying
national perspective. However, most input and objections filed are in practice
province- or riding-specific.318 Additionally, Elections Canada is a passive
procedural overseer of readjustment with neither a dispute settlement function nor
any other substantive decision-making responsibility.319

The process is decentralized by design. Debates in Parliament at the inauguration
of the FEBCs in 1964 show that the government of Lester B. Pearson was aware of
the confusion ten disparate commissions would bring to readjustment.320 Further, in
the early days a single national representation commissioner authored first-draft maps
for the FEBCs to work from, but critics warned against centralized control over all
the maps321 and the office came to an end in 1979.322 In comparison, frequent

313 See Part III.2.
314 See Hasen, supra note 12 at 958.
315 Fire, supra note 39 at 735.
316 An anonymous commissioner, supra note 29.
317 Ibid. See similarly Khullar, supra note 29.
318 See Elections Canada, supra note 211.
319 Landes, supra note 29.
320 House of Commons Debates, Vol. 2, 1964 (16 April 1964) at 2261-64, 2266-67 (various
speakers). Opposition Conservatives opposed a process that would see many distinct provincial
interpretations applied to the readjustment criteria of the new act.

321 Norman Ward, A Century of Constituencies (1967) 10 Can. Pub. Admin. 105 at 113.
322 Harvey Pasis, Achieving Population Equality among the Constituencies of the Canadian House,

19031976 (1983) 8 Legislative Studies Quarterly 111 at 115, n. 4.

55

R. LEVY ELECTORAL-BOUNDARY POLITICS

2008]

litigation in U.S. federal courts under the VRA and the Fourteenth Amendment
generates a more or less continuous national discourse over readjustment plans in the
various states.323 In the Texas gerrymandering battle, the United States Senate
majority leader openly masterminded the plan, and in litigation before the Supreme
Court of the United States, the administration of President George W. Bush
intervened in favour of the plan.324

Second, there are different levels of central oversight inside each provincial or
state process. While state legislative majorities generally control readjustment, in
each provincial FEBC no person, and no group of persons with common interests,
exercises top-down control. To be sure, the commissioners number only three per
provincea total that should ideally increase to reduce chances of partisan
manipulation within a given FEBC.325 However, extensive contributions from the
public and parliamentarians effectively expand the ranks of decision makers.
Decentralization can produce procedural ambiguity, but there are exceptions. For
example, the judicial-selections system in the United States Senate, which we saw
previously, has no ultimate umpire (other than, perhaps, the vague and contested
dictates of popular sentiment). Partisans elaborate rules of procedure by agreement, a
method that has produced the many layers of rules that we saw. These rules are not
ambiguous in effect, however, because they respond to balances of power in the
Senate and evolve along the shifting contours of a power struggle. For example, as
Republicans raised the threat that they would change filibuster rules, a bloc of
bipartisan moderates fashioned the noted rule of extraordinary circumstances.326
Better rules would be in place in advance of partisan conflict and would have a
greater chance of precluding it. A problem with rules that respond to balances of
power and specific controversies is that they tend to replace, rather than augment,
earlier rules. Much partisan Senate debate in late stages of Justice Samuel Alitos
nomination centred around whether
raised extraordinary
circumstances.327 The same procedural confines even channelled public rhetoric.328
The extraordinary-circumstances rule was, as we saw, only one of several layers of
rules. Each had been prominent in a previous stage in the debate, though now
extraordinary circumstances were key. We can expect similar phenomena in other
cases in which decision makers who are already partisan develop the rules of their
own process. A complex system generated out of partisan balance is ultimately rather

the nomination

323 American journals produce a continuous flow of new works on readjustment, most proposing

doctrinal tweaks on the constraint model.

324 See Smith, supra note 1; Charles Lane, White House Defends Texass GOP Remapping Plan to
Justices Washington Post (2 February 2006) A3. The court used the opportunity to reassert the role of
judges as the ultimate arbiters of readjustment (Perry, supra note 4).

325 But note that Landes cites a risk that larger Commissions will increase the demand for partisan

representations (supra note 29).

326 Hulse, supra note 121.
327 Babington & Schmidt, supra note 123.
328 Ibid.

[Vol. 53

long associated with constraint assumptions

that are common

MCGILL LAW JOURNAL / REVUE DE DROIT DE MCGILL

56

clear.329 In contrast, the collection of procedures described above makes the Canadian
system thoroughly and comprehensively ambiguous, with few clarifying exceptions;
in Canada, complex rules were set in place ahead of partisan rule manipulation.

Finally, it should be noted that when authors such as Mazzone and Scheberle
contend that complex institutions can generate informal norms of trust, they go
against the grain of prevalent reasoning about democratic regulation, especially in the
American tradition. Decentralizing power into many hands is an institutional-design
strategy
in
commentaryfrom early notes on institutional design in The Federalist Papers, to
the works of many current writers.330 In these works, notions opposite to those of
ambiguity animate calls to distribute powers widely: an aim is to maintain tensions
among counterbalanced power holders, each of whom acts to oppose and
[restrain]331 the others in a system of institutionaliz[ed] conflict.332 To this end,
institutions maximize the clarity of the setting in which a decision maker acts,
enabling watchfulness over, and prompting distrust of, other power holders. But a
seemingly small distinctionbetween rules responding to partisanship and rules laid
down in advance of partisanshipmay be one key to making complexity
comprehensive, and therefore useful as a means of building trust. This can help to
explain what is, to many, the FEBCs surprisingly developed impartiality.

move beyond constraint, and can also shape political cultures:

Interestingly, the Federalists contemplated but eschewed the notion that law can

There are two methods of curing the mischiefs of faction: the one, by

removing its causes; the other, by controlling its effects. …

It is in vain to say that enlightened statesmen will be able to adjust …
clashing interests and render them all subservient to the public good.
Enlightened statesmen will not always be at the helm. …
The inference to which we are brought is, that the causes of faction cannot

be removed and that relief is only to be sought in the means of controlling its
effects.333

Conclusion

Predictability and clarity are deeply prized, almost axiomatic principles of legal
design, especially in the context of regulating political power.334 Intentionally

329 As noted, some provincial-level readjustment commissions have been substantially partisan and
dysfunctional after appointing commissioners on political-party lines (Courtney, supra note 12 at 111-
12).

330 Supra note 259.
331 The Federalist, supra note 100, No. 48 (Madison) at 311.
332 Calabresi & Rhodes, supra note 252 at 1156.
333 The Federalist, supra note 100, No. 10 (Madison) at 123, 125.
334 See Karlan, Fire, supra note 36 at 733-35, 741.

R. LEVY ELECTORAL-BOUNDARY POLITICS

2008]

increasing ambiguity as a tool of administrative regulation runs counter to the
classical narrative of administrative delegation as a response to the exploding reach
and complexity of the business of the state. For example, Michael Taggart writes of
Great Britain in the nineteenth and twentieth centuries that

57

extensive delegated powers sprung, not from some sinister plot hatched by
bureaucrats, but from pressing necessity. The state was doing more. … The
night-watchman state was being replaced quite quickly by the welfare state. If
the state was to look after its subjects from the cradle to the grave which was
the wish of the voters … then delegation was as inevitable as was the growth
of the bureaucracy to implement Parliaments wishes. Parliament had neither
the time nor the capacity to attend to all the necessary details and tasks.335

Administrative apparatuses, which distance government from public accountability,
were (and are) necessary evils. Administrative decision making to some extent
preserves rationality and accountability in the face of the enormous scope,
pervasiveness, and impact of the various commitments of the state. But in a twist on
the traditional narrative, we might also design administrative regulation to increase,
rather than accommodate, complexity. We have seen works showing how ambiguous
rules can be deployed creativelythrough administrative forms especiallyto
regulate informal norms generally and political impartiality in particular.
A handful of readjustment authors are urging approaches to democratic regulation
favouring neither accreting new strata of constraining rules nor passively relying on
existing informal norms of impartiality. In the context of the American experience
and its perennial problems, these writers advocate a third option of actively regulating
informal norms of impartiality. And in a simultaneous but largely separate trend,
authors in various disciplines have explored the uses of ambiguity as a tool of
regulation. The ambiguity methods outlined in this article can help address problems
of democratic regulation. A chief purpose of this article has been to show how the use
of ambiguity has worked for Canadian readjustment. A broader goal has been to
outline a theoretical framework for continuing the study of the use of ambiguity and
its conditions, as applications continue to surface in the literature on democratic
regulation.

335 Michael Taggart, From Parliamentary Powers to Privatization: The Chequered History of

Delegated Legislation in the Twentieth Century (2005) 55 U.T.L.J. 575 at 584-85.

in this issue Patrolling the Borders of Sexual Orientation: Bisexual Refugee Claims in Canada

related content