Harkenrider v. Hochul ( 2022 )

State of New York OPINION
Court of Appeals This opinion is uncorrected and subject to revision
before publication in the New York Reports.
No. 60
In the Matter of Tim Harkenrider, et al.,
Respondents-Appellants,
v.
Kathy Hochul, &c., et al.,
Appellants-Respondents,
et al.,
Respondents.
Craig R. Bucki, for appellant-respondent Heastie.
Eric Hecker, for appellant-respondent Stewart-Cousins.
Jeffrey W. Lang, for appellants-respondents Hochul et al.
Misha Tseytlin, for respondents-appellants.
Campaign Legal Center et al.; League of Women Voters; Thomas F. O'Mara, et al.;
Jamaal Bowman, et al., amici curiae.
DiFIORE, Chief Judge:
In 2014, the People of the State of New York amended the State Constitution to
adopt historic reforms of the redistricting process by requiring, in a carefully structured
process, the creation of electoral maps by an Independent Redistricting Commission (IRC)
-1-
-2- No. 60
and by declaring unconstitutional certain undemocratic practices such as partisan and racial
gerrymandering. No one disputes that this year, during the first redistricting cycle to follow
adoption of the 2014 amendments, the IRC and the legislature failed to follow the
procedure commanded by the State Constitution. A stalemate within the IRC resulted in a
breakdown in the mandatory process for submission of electoral maps to the legislature.
The legislature responded by creating and enacting maps in a nontransparent manner
controlled exclusively by the dominant political party — doing exactly what they would
have done had the 2014 constitutional reforms never been passed. On these appeals, the
primary questions before us are whether this failure to follow the prescribed constitutional
procedure warrants invalidation of the legislature’s congressional and state senate maps
and whether there is record support for the determination of both courts below that the
district lines for congressional races were drawn with an unconstitutional partisan intent.
We answer both questions in the affirmative and therefore declare the congressional and
senate maps void. As a result, judicial oversight is required to facilitate the expeditious
creation of constitutionally conforming maps for use in the 2022 election and to safeguard
the constitutionally protected right of New Yorkers to a fair election.
I.
Every ten years, following the federal census, reapportionment of the state senate,
assembly, and congressional districts in New York must be undertaken to account for
population shifts and potential changes in the state’s allocated number of congressional
representatives (see NY Const, art III, § 4). Redistricting — which is “primarily the duty
and responsibility of the State” (Perry v Perez, 565 US 388, 392 [2012] [internal quotation
-2-
-3- No. 60
marks and citation omitted]; see Growe v Emison, 507 US 25, 34 [1993]) — is a complex
and contentious process that, historically, has been “within the legislative power . . . subject
to constitutional regulation and limitation” (Matter of Orans, 15 NY2d 339, 352 [1965]).
In New York, prior to 2012, the process of drawing district lines was entirely within the
purview of the legislature,1 subject to state and federal constitutional restraint and federal
voting laws, as well as judicial review.
Particularly with respect to congressional maps, exclusive legislative control has
repeatedly resulted in stalemates, with opposing political parties unable to reach consensus
on district lines — often necessitating federal court involvement in the development of
New York’s congressional maps (see e.g. Favors v Cuomo, 2012 WL 928223 *2, 2012 US
Dist LEXIS 36910, *10 [ED NY, Mar. 19, 2012, No. 11-CV-5632, Raggi, Lynch, and
Irizarry, JJ.]; Rodriguez v Pataki, 2002 WL 1058054, *7, 2002 US Dist LEXIS, *25-27
[SD NY 2002, May 24, 2002, No. 02 Civ. 618, Walker, Ch. J., Koeltl, and Berman, JJ.];
Puerto Rican Legal Defense & Educ. Fund, Inc. v Gantt, 796 F Supp 681, 684 [ED NY
1992]). Among other concerns, the redistricting process has been plagued with allegations
of partisan gerrymandering — that is, one political party manipulating district lines in order
to disproportionately increase its advantage in the upcoming elections, disenfranchising
voters of the opposing party (see generally Rucho v Common Cause, 588 US —, 139 S Ct
2484, 2494 [2019]).
1
A legislative advisory task force on apportionment — created by statute and comprising
lawmakers and staff selected by legislative leaders — conducted studies and proffered
recommendations and proposed maps for the legislature’s consideration (see Legislative
Law § 83-m; L 1978, ch 45, § 1).
-3-
-4- No. 60
By adopting the 2014 constitutional amendments, the People significantly altered
both substantive standards governing the determination of district lines and the redistricting
process established to achieve those standards. Given the history of legislative stalemates
and persistent allegations of partisan gerrymandering, the constitutional reforms were
intended to introduce a new era of bipartisanship and transparency through the creation of
an independent redistricting commission and the adoption of additional limitations on
legislative discretion in redistricting, including explicit prohibitions on partisan and racial
gerrymandering (see Assembly Mem in Support, 2012 NY Senate-Assembly Concurrent
Resolution S6698, A9526 Sponsor Memo, S2107). The Constitution now requires that the
IRC — a bipartisan commission working under a constitutionally mandated timeline — is
charged with the obligation of drawing a set of redistricting maps that, with appropriate
implementing legislation, must be submitted to the legislature for a vote, without
amendment (see NY Const, art III, § 4 [b]; § 5-b [a]).2 If this first set of maps is rejected,
the IRC is required to prepare a second set that, again, would be subject to an up or down
vote by the legislature, without amendment (see NY Const, art III, § 4 [b]). Under that
2
Many other states have also turned to independent redistricting commissions to curtail
partisan gerrymandering (see e.g. Ariz Const, art IV, pt. 2, § 1; Cal Const, art XXI, § 2;
Colo Const, art V, §§ 44 44-48.4; Conn Const, art III, § 6; Haw Const, art IV, § 2; Idaho
Const, art III, § 2; Me Const, art IV, part 3, § 1-A; Mich Const, art 4, § 6; Mont Const, art
V, § 14; NJ Const, art II, § 2; Ohio Const, arts XI, XIX; Va Const, art II, § 6-A; Wash
Const, art II, § 43). In upholding a state constitutional delegation of redistricting authority
to an IRC, the United States Supreme Court has recognized that IRCs “generally draw their
maps in a timely fashion and create districts both more competitive and more likely to
survive legal challenge” and “have succeeded to a great degree [in limiting the conflict of
interest implicit in legislative control over redistricting]” (Arizona State Legislature v
Arizona Independent Redistricting Comm’n, 576 US 787, 798, 821 [2015] [internal
quotation marks and citation omitted]).
-4-
-5- No. 60
constitutional framework, only upon rejection of a second set of IRC maps is the legislature
free to offer amendments to the maps created by the IRC (see NY Const, art III, § 4 [b])
and, even then, a statutory restriction enacted as a companion to the constitutional reforms
precluded legislative alterations that would affect more than two percent of the population
in any district (see L 2012, ch 17, § 3).
II.
Following receipt of the results of the 2020 federal census, the redistricting process
began in New York — the first opportunity for district lines to be drawn under the new
IRC procedures established by the 2014 constitutional amendments. Due to shifts in New
York’s population, the state lost a congressional seat and other districts were
malapportioned, undisputedly rendering the 2012 congressional apportionment —
developed by a federal court following a legislative impasse (see Favors, 2012 WL 928223,
*2, 2012 US Dist LEXIS 36910, *10) — unconstitutional and necessitating the drawing of
new district lines. Throughout 2021, the IRC held the requisite public hearings, gathering
input from stakeholders and voters across the state to inform their composition of
redistricting maps. In December 2021 and January 2022, however, negotiations between
the IRC members deteriorated and the IRC, split along party lines, was unable to agree
upon consensus maps. According to the IRC members appointed by the minority party,
after agreement had been reached on many of the district lines, the majority party
delegation of the IRC declined to continue negotiations on a consensus map, insisting they
would proceed with discussions only if further negotiations were based on their preferred
redistricting maps.
-5-
-6- No. 60
As a result of their disagreements, the IRC submitted, as a first set of maps, two
proposed redistricting plans to the legislature — maps from each party delegation — as is
constitutionally permitted if a single consensus map fails to garner sufficient votes (see NY
Const, art III, § 5-b [g]). The legislature voted on this first set of plans without amendment
as required by the Constitution and rejected both plans. The legislature notified the IRC
of that rejection, triggering the IRC’s obligation to compose — within 15 days — a second
redistricting plan for the legislature’s review (see NY Const, art III § 4 [b]). On January
24, 2022 — the day before the 15-day deadline but more than one month before the
February 28, 2022 deadline— the IRC announced that it was deadlocked and, as a result,
would not present a second plan to the legislature. Within a week, the Democrats in the
legislature — in control of both the senate and assembly — composed and enacted new
congressional, senate, and assembly redistricting maps (see 2022 NY Assembly Bill
A9167, 2022 NY Senate Bill S8196, 2022 NY Assembly Bill A9039-A, 2022 NY Senate
Bill S8172-A, 2022 NY Assembly Bill A9168, 2022 NY Senate Bill S8197, 2022 NY
Senate Bill S8185-A, 2022 NY Assembly Bill A9040-A), undisputedly without any
consultation or participation by the minority Republican Party.3 On February 3rd, the
Governor signed into law this new redistricting legislation, which also superseded the two
3
As one house of the legislature explained during this litigation, in their view “there [was
no] reason for the Democratic super-majorities in both houses of the [l]egislature to seek
‘input or involvement’ from the Republican minorities” regarding the development of these
legislative maps, characterizing such communications as inviting “time-wasting political
theater” (App Div reply brief for respondent-appellant Senate Majority Leader, at 13).
-6-
-7- No. 60
percent limitation imposed in 2012 on the legislature’s authority to amend IRC plans
(Senate Introducer’s Mem in Support, Bill Jacket, L 2012, ch 17, at 11).
That same day, petitioners — New York voters residing in several different
congressional districts — commenced this special proceeding under Article III, § 5 of the
State Constitution and Unconsolidated Laws § 4221 against various State respondents,
including the Governor,4 Senate Majority Leader, Speaker of the Assembly, and the New
York State Board of Elections, challenging the congressional and senate maps. Petitioners
alleged that the process by which the 2022 maps were enacted was constitutionally
defective because the IRC failed to submit a second redistricting plan as required under the
2014 constitutional amendments and, as such, the legislature lacked authority to compose
and enact its own plan. Petitioners also asserted that the congressional map is
unconstitutionally gerrymandered in favor of the majority party because it both “packed”
minority-party voters into a select few districts and “cracked” other pockets of those voters
across multiple districts, thereby diluting the competitiveness of those districts. Petitioners
asked Supreme Court to enjoin any elections from proceeding on the 2022 congressional
map and to either adopt its own map or direct the legislature to cure the infirmities.
Petitioners subsequently sought to amend their petition to include similar challenges to the
state senate map. The State respondents answered that petitioners lacked standing to
challenge most of the districts they claimed were gerrymandered, that the IRC’s failure to
4
Notwithstanding respondent Governor’s contentions to the contrary, any petition
challenging redistricting legislation must be served upon the Attorney-General, President
of the Senate, Speaker of the Assembly and the Governor, who are proper parties to this
proceeding (see Uncons Laws § 4221).
-7-
-8- No. 60
perform its duty did not strip the legislature of its enduring authority to enact redistricting
plans, and that petitioners could not meet their burden of proving that the maps were
unconstitutionally partisan.
A trial ensued, at which petitioners and the State respondents presented expert
testimony regarding the maps. Petitioners’ expert, Sean P. Trende — a doctoral candidate
who has a juris doctorate, a master’s degree in political science, and a master’s degree in
applied statistics, and who has participated as an expert in several redistricting proceedings
in other states — was qualified as an expert in election analysis with particular knowledge
in redistricting, with no objection from the State respondents or any request for a Frye
hearing to challenge the efficacy of his methodology or the basis of his opinion. Trende
testified that a comparison of the enacted congressional map to ensembles of 5,000 or
10,000 maps created by computer simulation revealed that the enacted map was an
“extreme outlier” that likely reduced the number of Republican congressional seats from
eight to four by “packing” Republican voters into four discrete districts and “cracking”
Republican voter blocks across the remaining districts in such manner as to dilute the
strength of their vote and render such districts noncompetitive.
Opposing experts called by the State respondents challenged Trende’s methodology
and asserted that the enacted congressional map actually resulted in more Republican
districts than the simulated maps, although several conceded that they did not analyze the
level of competitiveness of the new districts. Further, the State’s experts defended various
choices made by the legislature as justifiable based on constitutionally required
considerations, contending that the enacted maps were not reflective of partisan intent.
-8-
-9- No. 60
After determining petitioners had standing to challenge the statewide maps,
Supreme Court declared the congressional, state senate, and state assembly maps “void”
under the State Constitution, reasoning that the legislature’s enactment of redistricting
maps absent submission of a second redistricting plan by the IRC was unconstitutional and
that 2021 legislation purporting to authorize the enactment (“the 2021 legislation”) was
also unconstitutional. Further, crediting Trende’s testimony, Supreme Court found that
petitioners had proven that the congressional map violated the constitutional prohibition
on partisan gerrymandering by packing republican voters into four districts while ensuring
there were “virtually zero competitive districts.” Supreme Court declared all three maps
void, enjoined the State respondents from using the maps in the impending 2022 election,
and directed the legislature to submit new “bipartisanly-supported” maps that meet
constitutional requirements for the court’s review by a particular date.
The State respondents appealed, and a Justice of the Appellate Division stayed much
of Supreme Court’s order pending that appeal, including the deadline for submission of
new redistricting maps by the legislature. However, the stay order did not prohibit
Supreme Court from retaining a neutral expert to prepare a proposed new congressional
map, which would have no force and effect until certain contingencies occurred, including
the legislature’s failure to proffer its own new congressional maps by April 30th — 30 days
after the date of Supreme Court’s order.5 Thereafter, in a divided decision, the Appellate
5
Supreme Court also analyzed whether the state senate map was an unconstitutional
partisan gerrymander after granting petitioners’ request to amend the petition to challenge
the senate map but concluded petitioners did not meet their burden of proof on such claim.
Petitioners have not sought review of that determination.
-9-
- 10 - No. 60
Division modified Supreme Court’s order by denying the petition, in part, vacating the
declaration that the senate and assembly maps and the 2021 legislation were
unconstitutional, but otherwise affirmed and remitted, with three Justices agreeing with
Supreme Court that petitioners had met their burden of proving that the constitutional
prohibition against partisan gerrymandering had been violated with respect to the 2022
congressional map, rendering that map void and unenforceable (— AD3d —, 2022 NY
Slip Op 02648 [4th Dept 2022]).6 In reaching that conclusion, the Appellate Division relied
on “evidence of the largely one-party process used to enact the 2022 congressional map, a
comparison of the 2022 congressional map to the 2012 congressional map, and the expert
opinion and supporting analysis of Sean P. Trende” (id. at *3). However, the Court rejected
petitioners’ argument that both the congressional and senate maps were void due to the
failure to adhere to the constitutional procedure, with one Justice dissenting on that point.
The parties now cross appeal as of right (see CPLR 5601 [b] [1]), challenging certain
aspects of the Appellate Division order.
III.
As a threshold matter, relying on common law standing principles, the State
respondents assert that petitioners lack standing to challenge many of the districts that they
claim reflect unconstitutional partisan gerrymandering because none of the individual
6
Supreme Court, as permitted by the stay, has procured the services of a neutral
redistricting expert “to serve as special master to prepare and draw a new neutral, non-
partisan [c]ongressional map” and has established a schedule by which the parties and other
interested persons may submit commentary and proposed redistricting plans for
consideration prior to a planned hearing. Petitioners and several interested parties have
already proffered submissions to that court.
- 10 -
- 11 - No. 60
petitioners reside in those districts. Even absent the procedural challenge applicable to all
districts, this contention is unavailing because standing is expressly conferred by
constitution and statute. Article III, § 5 of the New York Constitution provides that “[a]n
apportionment by the legislature, or other body, shall be subject to review by the supreme
court, at the suit of any citizen, under such reasonable regulations as the legislature may
prescribe” (NY Const art III, § 5 [emphasis added]; see 3 Rev Rec, 1894 NY Constitutional
Convention at 987; see Matter of Dowling, 219 NY 44, 50 [1916]; Schieffelin v Komfort,
212 NY 520, 529 [1914]). Moreover, statutes may identify the class of persons entitled to
challenge particular governmental action, relieving courts of the need to resolve the
question under common law principles (see Matter of Mental Hygiene Legal Serv. v
Daniels, 33 NY3d 44, 50 n 2 [2019]; Society of Plastics Indus. v County of Suffolk, 77
NY2d 761, 769 [1991]; Wein v Comptroller of State of N.Y., 46 NY2d 394, 399 [1979]; see
e.g. State Finance Law § 123) and, here, Unconsolidated Laws § 4221 likewise authorizes
“any citizen” of the state to seek judicial review of a legislative act establishing electoral
districts. We therefore turn to consideration of the merits of petitioners’ challenges to the
2022 redistricting maps.
Petitioners first assert that, in light of the lack of compliance by the IRC and the
legislature with the procedures set forth in the Constitution, the legislature’s enactment of
the 2022 redistricting maps contravened the Constitution. To conclude otherwise,
petitioners contend, would be to render the 2014 amendments — touted as an important
reform of the redistricting process — functionally meaningless. We agree.
- 11 -
- 12 - No. 60
Legislative enactments, including those implementing redistricting plans, are
entitled to a “strong presumption of constitutionality” and redistricting legislation will be
declared unconstitutional by the courts “‘only when it can be shown beyond reasonable
doubt that it conflicts’” with the Constitution after “‘every reasonable mode of
reconciliation of the statute with the Constitution has been resorted to, and reconciliation
has been found impossible’” (Matter of Wolpoff v Cuomo, 80 NY2d 70, 78 [1992], quoting
Matter of Fay, 291 NY 198, 207 [1943] [internal quotation marks omitted]; see Cohen v
Cuomo, 19 NY3d 196, 201-202 [2012]). Nevertheless, invalidation of a legislative
enactment is required when such act amounts to “‘a gross and deliberate violation of the
plain intent of the Constitution and a disregard of its spirit and the purpose for which
express limitations are included therein’” (Cohen, 19 NY3d at 202, quoting Matter of
Sherrill v O’Brien, 188 NY 185, 198 [1907]).
To determine whether the legislature’s 2022 enactment of redistricting legislation
comports with the Constitution, our starting point must be the text thereof. “In construing
the language of the Constitution as in construing the language of a statute, . . . [we] look
for the intention of the People and give to the language used its ordinary meaning” (Matter
of Sherrill, 188 NY at 207; see White v Cuomo, — NY —, — 2022 NY Slip Op 01954, *
5 [2022]; Burton v New York State Dept. of Taxation & Fin., 25 NY3d 732, 739 [2015];
Matter of Carey v Morton, 297 NY 361, 366 [1948]). Upon careful review of the plain
language of the Constitution and the history pertaining to the adoption of the 2014 reforms,
it is evident that the legislature and the IRC deviated from the constitutionally mandated
procedure.
- 12 -
- 13 - No. 60
From a procedural standpoint, the Constitution — as amended in 2014 — requires
that, every ten years commencing in 2020, an “independent redistricting commission”
comprising 10 members — eight of whom are appointed by the majority and minority
leaders of the senate and assembly and the remaining two by those eight appointees —
shall be established (see NY Const, art III, § 5-b [a]). The members must be a diverse
group of registered voters and cannot be (or recently have been) members of the state or
federal legislature, statewide elected officials, state officers or legislative employees,
registered lobbyists, or political party chairmen, or the spouses of state or federal elected
officials (see NY Const, art III, § 5-b [b], [c]).
Under the Constitution, the IRC must make its draft redistricting plans available to
the public and hold no less than 12 public hearings throughout the state regarding proposals
for redistricting, ensuring transparency and giving New Yorkers a voice in the redistricting
process (see NY Const, art III, § 4 [c]). After considering public comments and working
together across party lines to compose new redistricting lines, the IRC must submit its
approved plan and implementing legislation to the legislature no later than January 15th in
a redistricting year (see NY Const, art III, § 4 [b]), with the caveat that, if the IRC is unable
to muster the requisite number of votes for a single plan, it must provide the legislature
with each plan that “garnered the highest number of votes in support of its approval by the
[IRC]” (NY Const, art III, § 5-b [g]). If the legislature rejects the IRC’s first plan, the
Constitution requires the IRC to go back to the drawing board, work to reach consensus,
and “prepare and submit to the legislature a second redistricting plan and the necessary
implementing legislation” to the legislature within 15 days and in no case later than
- 13 -
- 14 - No. 60
February 28th (NY Const, art III, § 4 [b]). “If” the legislature fails to approve the second
plan without amendment, the Constitution then directs that “each house shall introduce
such implementing legislation” — a clear reference to the IRC’s second plan — with any
amendments each house of the legislature deems necessary (NY Const, art III § 4 [b]). As
a further safeguard against one party dominating redistricting, the Constitution dictates that
the number of votes required for the IRC and legislature to approve a plan differs depending
on whether the legislature is controlled by one political party or control of the houses are
split between the parties (see NY Const, art III, §§ 4 [b] [1] – [3]; 5-b [f] [1], [2]).
The Redistricting Reform Act of 2012, legislation enacted in conjunction with the
2012 constitutional resolution, further provides as a matter of statutory law that “[a]ny
amendments by the senate or assembly to a redistricting plan submitted by the [IRC] shall
not affect more than two percent of the population of any district contained in such plan”
(L 2012, ch 17, § 3). As the sponsor of the legislation explained, “[i]f the [IRC’s] second
plan [was] also rejected . . . , each house may then amend that plan prior to approval except
that such amendments . . . cannot affect more than two percent of the population of any
district in the commission’s plan,” a limitation designed to “provide reasonable restrictions
on the legislature’s changes to the commission’s plans” (Senate Introducer’s Mem in
Support, Bill Jacket, L 2012, ch 17, at 15 [emphasis added]).
The plain language of Article III, § 4 dictates that the IRC “shall prepare” and “shall
submit” to the legislature a redistricting plan with implementing legislation, that IRC plan
“shall be voted upon, without amendment” by the legislature, and — in the event the first
plan is rejected — the IRC “shall prepare and submit to the legislature a second
- 14 -
- 15 - No. 60
redistricting plan and the necessary implementing legislation,” which again “shall be voted
upon, without amendment” (NY Const, art III, § 4 [b] [emphasis added]). “If” and only
“if” that second plan is rejected, does the Constitution permit the legislature to introduce
its own implementing legislation, “with any amendments” to the IRC plans deemed
necessary that otherwise comply with constitutional directives (NY Const, art III, § 4 [b]
[emphasis added]).
“In the construction of constitutional provisions, the language used, if plain and
precise, should be given its full effect” and “[i]t must be presumed that its framers
understood the force of the language used and, as well, the people who adopted it” (People
v Rathbone, 145 NY 434, 438 [1895]). Our Constitution is “an instrument framed
deliberately and with care, and adopted by the people as the organic law of the State” and,
when interpreting it, we may “not allow for interstitial and interpretative gloss . . . by the
other [b]ranches [of the government] that substantially alters the specified law-making
regimen” set forth in the Constitution (Matter of King v Cuomo, 81 NY2d 247, 253 [1993]).
Article III, § 4 is permeated with language that, when given its full effect, permits
the legislature to undertake the drawing of district lines only after two redistricting plans
composed by the IRC have been duly considered and rejected. 7 Moreover, the text of
section 4 contemplates that any redistricting act ultimately adopted must be founded upon
a plan submitted by the IRC; in the event the IRC plan is rejected, the Constitution
7
Indeed, the description on the 2014 ballot informed voters considering whether to support
the constitutional amendments that “the legislature may only amend the redistricting plan
. . . if the commission’s plan is rejected twice by the legislature.”
- 15 -
- 16 - No. 60
authorizes “amendments” to such plan, not the wholesale drawing of entirely new maps
(NY Const, art III, § 4 [b]; see NY Assembly Debate on Assembly Bill A9557 Mar. 15,
2012 at 39 [“The Constitutional amendment allows the (l)egislature to amend the plan
submitted by the independent redistricting commission if the (l)egislature has twice
rejected submitted plans” (emphasis added)]).8
Despite clear constitutional language, the State respondents posit that it is wrong to
interpret the 2014 constitutional amendments as requiring two separate IRC plans as a
precondition to the legislature’s exercise of its longstanding and historically unbridled
authority to enact redistricting legislation.9 They further rely on the 2021 legislation
authorizing the legislature to move forward on redistricting even if the IRC fails to submit
maps as permissibly filling a purported gap in the constitutional design. However, in
8
Judge Rivera’s contention that the IRC process was not violated because two sets of maps
were simultaneously submitted by the IRC in the first round — one by the Democratic
delegation and one by the Republican delegation — is remarkable. Under her view, this
was the functional equivalent of the successive presentations required by the Constitution.
Aside from being directly contrary to the text of the constitution, the intent of the People
who adopted the 2014 reforms, and the relevant legislative history, such contention has not
been advanced by any party before this Court, a reflection of its total lack of merit.
9
In a reply brief submitted in the Appellate Division, one of the State respondents candidly
acknowledged that the constitutional process was not followed here, asserting that
“[e]veryone agrees” that the Constitution requires two rounds of IRC recommendations
“and that the [l]egislature vote up or down on each Commission proposal without
amendment before exercising its authority to make any amendments”; and “that nobody
suggests that ‘the process’ is optional” (App Div reply brief for respondent-appellant
Senate Majority Leader, at 2-3). Despite acknowledging the constitutional violation,
however, they essentially view it as irrelevant because the legislature could ultimately have
adopted its own maps through the amendment process following a properly completed IRC
procedure. This view ignores the fact that procedural requirements matter and are imposed
precisely because, as here, they safeguard substantive rights.
- 16 -
- 17 - No. 60
addition to being contrary to the text of the Constitution as we have explained, the State
respondents’ arguments are also belied by the purpose of the 2014 amendments and the
relevant legislative history — including the legislature’s own statements regarding the
intent and effect of the 2014 constitutional reform effort.
Indeed, the State respondents studiously ignore events that gave rise to the 2014
amendments. During the previous redistricting cycle in 2012, the New York legislature
was unable to reach agreement on legislation setting the congressional district lines and, as
a result, a federal court ordered the adoption of a judicially-drafted congressional
redistricting plan (see Favors, 2012 WL 928223, *2, 2012 US Dist LEXIS 36910, *10).
While the 2012 legislature did agree on state senate and assembly maps, the proposed maps
were widely criticized as a product of partisan gerrymandering, prompting the then-
Governor to threaten to veto the plans absent a concrete legislative commitment to
redistricting reform (see Micah Altman & Michael P. McDonald, A Half-Century of
Virginia Redistricting Battles: Shifting from Rural Malapportionment to Voting Rights to
Public Participation, 47 U Rich L Rev 771, 829 [2013]; Thomas Kaplan, An Update on
New York Redistricting, NY Times, March 7, 2012; Thomas Kaplan, An Update on New
York Redistricting, NY Times, March 9, 2012). Thus, as we have discussed, in conjunction
with enactment of the 2012 redistricting acts (see L 2012, ch 16), the legislature affirmed
its commitment to redistricting reform by passing the Redistricting Reform Act of 2012
(see L 2012, ch 17) and the first of the two concurrent resolutions proposing the
constitutional amendments creating the IRC process (see 2012 NY Assembly Bill A9526
[Mar. 11, 2012]). Characterizing the legislature’s 2012 senate and assembly district lines
- 17 -
- 18 - No. 60
as “significantly flawed,” the Governor nevertheless approved the redistricting legislation
that year in light of the legislature’s demonstrated agreement to “permanent[ly]” and
“meaningful[ly]” reform the redistricting process for future years and “provide
transparency to a process [otherwise] cloaked in secrecy and largely immune from legal
challenges to partisan gerrymandering” (Governor’s Approval Mem, Bill Jacket, L 2012,
ch 17 at 5; 2012 NY Legis Ann at 12-13).
As the surrounding context and history of the 2014 amendments illustrate, the
constitutional amendments adopted by the two consecutive legislatures and the voters —
from the provisions detailing the composition of the IRC to those setting forth the voting
metrics — were carefully crafted to guarantee that redistricting maps have their origin in
the collective and transparent work product of a bipartisan commission that is
constitutionally required to pursue consensus to draw district lines. The procedural
amendments — along with a novel substantive amendment of the State Constitution
expressly prohibiting partisan gerrymandering, discussed further below — were enacted in
response to criticism of the scourge of hyper-partisanship, which the United States
Supreme Court has recognized as “incompatible with democratic principles” (Arizona
State Legislature v Arizona Independent Redistricting Comm’n, 576 US at 791 [internal
quotation marks, punctuation and citation omitted]).
As reflected in the legislative record, the IRC’s fulfillment of its constitutional
obligations was unquestionably intended to operate as a necessary precondition to, and
limitation on, the legislature’s exercise of its discretion in redistricting. The legislative
record shows that the 2012 legislature — the drafters of the constitutional amendments —
- 18 -
- 19 - No. 60
intended to “comprehensively” reform and “implement historic changes to achieve a fair
and readily transparent process” to “ensure that the drawing of legislative district lines in
New York will be done by a bipartisan, independent body” — rather than entirely by the
legislature itself (Assembly Mem in Support, 2012 NY Senate-Assembly Concurrent
Resolution S6698, A9526; Sponsor’s Mem, 2013 NY Senate Bill S2107). As the sponsors
explained, the reforms were designed to “substantively and fundamentally” alter the
redistricting process, allowing “[f]or the first time, both the majority and minority parties
in the legislature [to] have an equal role in the process of drawing lines,” with these “far-
reaching” constitutional reforms touted as a template “for independent redistricting
throughout the United States” (Assembly Mem in Support, 2013 NY Senate-Assembly
Concurrent Resolution S2107, A2086).
The Senate debate indicates that the constitutional provision allowing the legislature
to amend the second redistricting plan submitted by the IRC only after twice voting on and
rejecting IRC plans was intended to encourage bipartisan participation by the legislature in
the redistricting process. The Senate sponsor explained that “[o]n the third enactment,
there could be amendments under this provision. But again, it would be the third time –
not first time, not the second time, but the third time in order to get ultimately a product
produced” (NY Senate Debate on AB2086, January 23, 2013 at 222). In other words, “[i]f
there cannot be agreement, if the Governor vetoes the provision twice, . . . that third time
the Legislature would be acting. But not until that time” (id. at 224) because “the intent of
th[e] resolution [wa]s to have the Legislature act and vote on . . . a [second] plan” before
undertaking any amendments of its own (id. at 226). Answering a charge that the IRC
- 19 -
- 20 - No. 60
would essentially be only “an advisory commission” since the legislature could ultimately
reject both sets of IRC maps, the Senate sponsor explained that the IRC process was
intended, in part, to impose consequences on the legislature for rejecting plans developed
through a bipartisan process by forcing it to take a public position refusing to adopt district
lines that were developed with an “enormous amount of citizen input” and effort (id. at
228).
It is no surprise, then, that the Constitution dictates that the IRC-based process for
redistricting established therein “shall govern redistricting in this state except to the extent
that a court is required to order the adoption of, or changes to, a redistricting plan as a
remedy for a violation of law” (NY Const art III, § 4 [e] [emphasis added]). Contrary to
the State respondents’ contentions, the detailed amendments leave no room for legislative
discretion regarding the particulars of implementation; this is not a scenario where the
Constitution fails to provide “specific guidance” or is “silen[t] on the issue” (Cohen, 19
NY3d at 200, 202). Under the 2014 amendments, compliance with the IRC process
enshrined in the Constitution is the exclusive method of redistricting, absent court
intervention following a violation of the law, incentivizing the legislature to encourage and
support fair bipartisan participation and compromise throughout the redistricting process.10
10
The State respondents and Judge Rivera assert that giving force to the constitutional
language risks gamesmanship by minority members of the IRC, claiming such members
could potentially derail the redistricting process by refusing to participate. In giving effect
to the constitutional reforms endorsed by the People of this state, our decision does not
leave the legislature hostage to that body as Judge Rivera contends. Legislative leaders
appoint a majority of the IRC members and, in the event those members fail either to appear
at IRC meetings or to otherwise perform their constitutional duties, judicial intervention in
the form of a mandamus proceeding, political pressure, more meaningful attempts at
- 20 -
- 21 - No. 60
That the IRC process was intended to operate as a limitation on the legislature’s
power to compose district lines is further underscored by the Redistricting Reform Act of
2012 (see L 2012, ch 17). That legislation, adopted simultaneously with the 2012
constitutional resolution, instituted the two percent limitation on the legislature’s authority
(see L 2012, ch 17, § 3). In describing this particular reform, the Sponsor of the bill
explained that “[i]f the legislature fails to pass” the IRC’s second plan “it may then amend
such plans and vote upon them as amended. However, any such amendments shall be
limited . . . to affect no more than two percent of the population of any district in such plan”
(Senate Introducer’s Mem in Support, Bill Jacket, L 2012, ch 17, at 11). Thus, although
the legislature retains the ultimate authority to enact districting maps upon completion of
the IRC process, the constitutional reforms were clearly intended to promote fairness,
transparency, and bipartisanship by requiring, as a precondition to redistricting legislation,
that the IRC fulfill a substantial and constitutionally required role in the map drawing
process.11
compromise, and possibly even replacement of members who fail to faithfully perform
their duties, are among the many courses of action available to ensure the IRC process is
completed as constitutionally intended. The IRC may not be a panacea, but to accept the
crabbed description of that body proffered by the State respondents and Judge Rivera
would be to render the body nothing more than “window dressing” masquerading as
meaningful reform.
11
In 2022 — the very first time that the legislature had occasion to implement the IRC
procedure and the two percent rule (L 2012, ch 17, § 3) — that provision was disregarded.
The legislature wholly superseded the two percent rule by prefacing the 2022 redistricting
legislation with language indicating that such districts were enacted as provided therein
“notwithstanding any other provision of law to the contrary” and providing that the new
legislation “shall supersede any inconsistent provision of law including but not limited to”
- 21 -
- 22 - No. 60
Indeed, recent events suggest that the legislature itself recognized that the
Constitution did not permit it to proceed with redistricting absent compliance with the
bipartisan IRC process. Apparently forecasting that the IRC would not comply with its
constitutional obligations, in the summer of 2021 — before the IRC had even been given
a chance to fulfill its constitutional role — the legislature attempted to amend the
constitution to add language authorizing it to introduce redistricting legislation “[i]f . . . the
redistricting commission fails to vote on a redistricting plan and implementing legislation
by the required deadline” for any reason (2021 NY Senate-Assembly Concurrent
Resolution S515, A1916). After New York voters rejected this constitutional amendment
(among others) — and with the first redistricting cycle since the 2014 amendments on the
horizon — the legislature attempted to fill a purported “gap” in constitutional language by
statutorily amending the IRC procedure in the same manner (see L 2021, ch 633). In this
Court, the State respondents attempt to rely on the 2021 legislation to justify the deviation
from constitutional requirements. Needless to say, the bipartisan process was placed in the
State Constitution specifically to insulate it from capricious legislative action and to ensure
permanent redistricting reform absent further amendment to the constitution, which has not
occurred. The 2021 legislation is unconstitutional to the extent that it permits the
legislature to avoid a central requirement of the reform amendments (see Matter of King,
81 NY2d at 252 [“The (l)egislature must be guided and governed in this particular function
by the Constitution, not by a self-generated additive”]).
the two percent rule (L 2022, chs 13, 14, 15, 16). Despite this attempted end run, however,
the 2012 redistricting reform legislation provides relevant evidence of the drafters’ intent.
- 22 -
- 23 - No. 60
In sum, there can be no question that the drafters of the 2014 constitutional
amendments and the voters of this state intended compliance with the IRC process to be a
constitutionally required precondition to the legislature’s enactment of redistricting
legislation. In urging this Court to adopt their view that the IRC may abandon its
constitutional mandate with no impact on the ultimate result and by contending that the
legislature may seize upon such inaction to bypass the IRC process and compose its own
redistricting maps with impunity, the State respondents ask us to effectively nullify the
2014 amendments. This we will not do. Indeed, such an approach would encourage
partisans involved in the IRC process to avoid consensus, thereby permitting the legislature
to step in and create new maps merely by engineering a stalemate at any stage of the IRC
process, or even by failing to appoint members or withholding funding from the IRC.
Through the 2014 amendments, the People of this state adopted substantial redistricting
reforms aimed at ensuring that the starting point for redistricting legislation would be
district lines proffered by a bipartisan commission following significant public
participation, thereby ensuring each political party and all interested persons a voice in the
composition of those lines. We decline to render the constitutional IRC process
inconsequential in the manner requested by the State respondents, a result that would
“violat[e] . . . the plain intent of the Constitution and . . . disregard [the] spirit and the
purpose” of the 2014 constitutional amendments (Cohen, 19 NY3d at 202 [internal
quotation marks and citation omitted]).
- 23 -
- 24 - No. 60
IV.
Having addressed the procedural violation, we turn to the substantive partisan
gerrymandering claim. As a threshold matter, despite our invalidation of the maps on
procedural grounds, we nevertheless must determine on the State respondents’ cross appeal
whether the courts below properly declared that the congressional map was also
substantively unconstitutional.12
In addition to the procedural amendments, in 2014, the People also amended the
New York State Constitution to include certain substantive limitations on redistricting,
including an express prohibition on partisan gerrymandering, commanding that “[d]istricts
shall not be drawn to discourage competition or for the purpose of favoring or disfavoring
incumbents or other particular candidates or political parties” (NY Const, art III, § 4 [c]
12
While we agree with Judge Troutman that this Court should not issue advisory opinions,
her suggestion that no actual case or controversy is presented by the State respondents’
appeal — here as of right on the substantial constitutional question of whether the
Appellate Division erred in invalidating the congressional map on the ground of partisan
gerrymandering — is quite extraordinary. Even if the State respondents were not otherwise
entitled to review of the declaration that the apportionment legislation was infected by such
invidious intent, there are substantial arguments before this Court concerning the proper
remedy in the event of a constitutional violation — arguments that turn, in part, on whether
the violation involved procedural or substantive constitutional provisions. The question of
whether the congressional map amounts to a partisan gerrymander is also relevant to the
issue of whether the primary election should be permitted to proceed on the maps drawn
by the legislature, despite the determination of procedural unconstitutionality. Moreover,
given our conclusion that new maps must be drawn in light of the procedural violation —
a conclusion with which Judge Troutman agrees — resolution of the issue is critical to
provide necessary guidance to inform the development of a new congressional map on
remittal.
- 24 -
- 25 - No. 60
[5]).13 This amendment was made in recognition that the practice of partisan
gerrymandering “jeopardizes [t]he ordered working of our Republic, and of the democratic
process” and, “[a]t its most extreme, the practice amounts to ‘rigging elections,’” which
violates “the most fundamental of all democratic principles — that ‘the voters should
choose their representatives, not the other way around’” (Gill v Whitford, — US — , 138
S Ct 1916, 1940 [2018], quoting Arizona State Legislature, 576 US at 824).
In this case, petitioners asserted that, along with being procedurally flawed, the 2022
congressional map enacted by the legislature violates the constitutional provision
prohibiting partisan gerrymandering. To prevail on such claim, petitioners bore the burden
of proving beyond a reasonable doubt that the congressional districts were drawn with a
particular impermissible intent or motive — that is, to “discourage competition” or to
“favor[] or disfavor[] incumbents or other particular candidates or political parties” (NY
Const, art III, § 4 [c] [5]). Such invidious intent could be demonstrated directly or
13
The 2014 constitutional amendments also forbid racial gerrymandering, in a provision
that similarly prohibits an invidious intent or motive, requiring that district lines “shall not
be drawn to have the purpose of, nor shall they result in, the denial or abridgement of” the
voting rights of racial or minority language groups (NY Const, art III, § 4 [c] [1]). Other
requirements added that year directed certain results, namely, that redistricting, to the
extent possible, maintain cores of existing districts, pre-existing political subdivisions —
such as counties, cities, and towns — and communities of interest (see NY Const, art III,
§ 4 [c] [5]). These requirements supplement the longstanding constitutional constraints on
redistricting embodied in the State Constitution requiring, to the extent practical, that
districts “contain as nearly as may be an equal number of inhabitants,” “consist of
contiguous territory,” and be “as compact in form as practicable” (NY Const, art III, § 4
[c] [2] – [4]), and those required by federal law — such as conformity with the “one person,
one vote” principle (Abrams v Johnson, 521 US 74, 98 [1997]; see Wesberry v Sanders,
376 US 1, 8 [1964]) and with the federal Voting Rights Act (see generally 52 USC §
10301).
- 25 -
- 26 - No. 60
circumstantially through proof of a partisan process excluding participation by the minority
party and evidence of discriminatory results (i.e., lines that impactfully and unduly favor
or disfavor a political party or reduce competition).
Here, at the conclusion of the non-jury trial, Supreme Court — based on the partisan
process, the map enacted by the legislature itself, and the expert testimony proffered by
petitioners — found by “clear evidence and beyond a reasonable doubt that the
congressional map was unconstitutionally drawn with political bias” to “significantly
reduce[]” the number of competitive districts. The Appellate Division affirmed, similarly
drawing an inference of invidious partisan purpose based on “evidence of the largely one-
party process used to enact the 2022 congressional map, a comparison of the 2022
congressional map to the 2012 congressional map, and the expert opinion and supporting
analysis of Sean P. Trende,” finding that “the 2022 congressional map was drawn to
discourage competition and favor democrats” (— AD3d at —, 2022 NY Slip Op 02648, *
4).
We reject respondents’ assertion that the evidence was legally insufficient to
establish an unconstitutional partisan purpose. Viewing the evidence in the light most
favorable to petitioners and drawing every inference in their favor, there is a “valid line of
reasoning and permissible inferences” which could possibly lead [a] rational [factfinder]
to the conclusion reached by the [factfinder] on the basis of the evidence presented at trial”
(Cohen v Hallmark Cards, 45 NY2d 493, 499 [1978]). Moreover, where, as here, this
Court is presented with affirmed findings of fact in a civil case, our review is limited to
whether there is record support for those findings (see Matter of Rittersporn v Sadowski,
- 26 -
- 27 - No. 60
48 NY2d 619 [1979]). There is record support in the undisputed facts and evidence
presented by petitioners for the affirmed finding that the 2022 congressional map was
drawn to discourage competition. Indeed, several of the State respondents’ experts, who
urged the court to draw the contrary inference, concededly did not take into account the
reduction in competitive districts. Thus, we find no basis to disturb the determination of
the courts below (see Matter of Rittersporn, 48 NY2d at 619).14
14
Although purporting to treat the question as an issue of law, Judge Wilson impermissibly
performs a weight of the evidence analysis, largely parroting the points in the State
respondents’ briefs. Tellingly, however, Judge Wilson repeatedly acknowledges that an
inference of intent could rationally be drawn from proof in the record. Determining
whether to draw such an inference when multiple inferences are possible is a quintessential
function of a finder of fact and, here, the courts below — which, unlike this Court,
possessed fact-finding authority — credited Trende’s testimony. Contrary to Judge
Wilson’s contention, the burden of proof was not impermissibly shifted to the State
respondents. As noted, respondents did not seek exclusion of Trende’s testimony on the
basis that his methodology or the computer algorithm on which he relied — drafted by a
recognized expert and, according to Trende, a “state of the art” program repeatedly
accepted by other courts — was insufficiently reliable. Although Trende did observe that
the State respondents completely failed to refute any of his simulations with simulations of
their own, he also responded substantively to the criticisms of his methodology. Trende
explained that his map ensemble “perform[ed] comparably to the enacted plan in terms of
compactness,” “minority-majority districts,” and county lines. He ran additional
simulations, freezing municipalities kept intact by the enacted plan, freezing district cores,
freezing every “ability-to-elect district,” and even conceding the split in southeast
Brooklyn to respondents. Trende testified that even when the simulations were run in a
manner “incredibly generous” to the State respondents by “ced[ing] to [respondents] . . . a
third of the districts drawn in New York,” the simulations produced “the same basic
output,” showing the same cracking and packing patterns in the enacted maps. As even a
short rendition of just some of the proof presented by petitioners demonstrates, Judge
Wilson refuses to apply the proper standard of review, which — even in cases where the
legal standard is proof beyond a reasonable doubt — requires that the evidence be viewed
in the light most favorable to petitioners, the prevailing party at trial.
- 27 -
- 28 - No. 60
V.
Based on the foregoing, the enactment of the congressional and senate maps by the
legislature was procedurally unconstitutional, and the congressional map is also
substantively unconstitutional as drawn with impermissible partisan purpose, leaving the
state without constitutional district lines for use in the 2022 primary and general elections.15
The parties dispute the proper remedy for these constitutional violations, with the State
respondents arguing no remedy should be ordered for the 2022 election cycle because the
election process for this year is already underway. In other words, the State respondents
urge that the 2022 congressional and senate elections be conducted using the
unconstitutional maps, deferring any remedy for a future election.16 We reject this
invitation to subject the People of this state to an election conducted pursuant to an
unconstitutional reapportionment.
15
Inasmuch as petitioners neither sought invalidation of the 2022 state assembly
redistricting legislation in their pleadings nor challenge in this Court the Appellate
Division’s vacatur of the relief granted by Supreme Court with respect to that map, we may
not invalidate the assembly map despite its procedural infirmity.
16
The State respondents’ reliance on the federal Purcell principle is misplaced (see Purcell
v Gonzalez, 549 US 1 [2006]). The Purcell doctrine cautions federal courts against
interfering with state election laws when an election is imminent (see Republican National
Committee v Democratic National Committee, 589 US — , — , 140 S Ct 1205, 1207
[2020]) and does not limit state judicial authority where, as here, a state court must
intervene to remedy violations of the State Constitution. Indeed, most recently the
principle was cited to justify the United States Supreme Court’s decision not to disturb a
state court order requiring alteration of North Carolina’s existing congressional maps for
the upcoming 2022 primary (Moore v Harper, 595 US —, 142 S Ct 1089, 1089 [2022,
Kavanaugh, J., concurring in denial of application for stay]).
- 28 -
- 29 - No. 60
“The power of the judiciary of a State to require valid reapportionment or to
formulate a valid redistricting plan has not only been recognized by [the United States
Supreme] Court but appropriate action by the States in such cases has been specifically
encouraged” (Scott v Germano, 381 US 407, 409 [1965]; see Growe, 507 US at 33).17
Indeed, our State Constitution both requires expedited judicial review of redistricting
challenges (see NY Const, art III, § 5) — as occurred here — and authorizes the judiciary
to “order the adoption of, or changes to, a redistricting plan” in the absence of a
constitutionally-viable legislative plan (NY Const, art III, § 4 [e]). Where, as here,
legislative maps have been determined to be unenforceable, we are left in the same
predicament as if no maps had been enacted. Prompt judicial intervention is both necessary
and appropriate to guarantee the People’s right to a free and fair election.
We are cognizant of the logistical difficulties involved in preparing for and
executing an election — and appreciate that rescheduling a primary election impacts
administrative officials, candidates for public office, and the voters themselves. Like the
courts below, however, we are not convinced that we have no choice but to allow the 2022
primary election to proceed on unconstitutionally enacted and gerrymandered maps. With
judicial supervision and the support of a neutral expert designated a special master, there
17
A number of other state courts have been called upon to intervene in redistricting just
this year (see League of Women Voters of Ohio v Ohio Redistricting Commn., 2022-Ohio-
789, — NE3d — [2022]; Harper v Hall, 2022-NCSC-17, ¶ 6, 868 SE2d 499, 510 [2022];
Johnson v Wisconsin Elections Commn., 2022 WI 19, ¶ 3, — NW2d — [2022]; Carter v
Chapman, 270 A3d 444, 450 [Pa 2022]).
- 29 -
- 30 - No. 60
is sufficient time for the adoption of new district lines.18 Although it will likely be
necessary to move the congressional and senate primary elections to August, New York
routinely held a bifurcated primary until recently, with some primaries occurring as late as
September. We are confident that, in consultation with the Board of Elections, Supreme
Court can swiftly develop a schedule to facilitate an August primary election, allowing
time for the adoption of new constitutional maps, the dissemination of correct information
to voters, the completion of the petitioning process, and compliance with federal voting
laws, including the Uniformed and Overseas Citizens Absentee Voting Act (see 52 USC §
20302).
Finally, the State respondents’ protest that the legislature must be provided a “full
and reasonable opportunity to correct . . . legal infirmities” in redistricting legislation (NY
Const, art III, § 5). The procedural unconstitutionality of the congressional and senate
maps is, at this juncture, incapable of a legislative cure. The deadline in the Constitution
18
Delaying a remedy until the next election would substantially undermine the People’s
efforts to temper partisan gerrymandering. Here, the legislature enacted maps within one
week of the IRC’s abdication—which itself came more than a month before the
Constitution’s outer end date for the IRC process—and petitioners commenced this
proceeding on the same day. If there is insufficient time to order a remedy for the 2022
primary election under these circumstances, it is unlikely there would ever be sufficient
time to challenge a redistricting plan and obtain relief before an upcoming primary election.
Such a conclusion would be contrary to the Constitution, which contemplates that the IRC
process may not be completed until February 28th (to be followed by legislative action)
but nevertheless expressly authorizes expedited judicial review and modification or
adoption of redistricting plans by the courts. Delaying a remedy in this election cycle —
permitting an election to go forward on unconstitutional maps — would set a troubling
precedent for future cases raising similar partisan gerrymandering claims, as well as other
types of challenges, such as racial gerrymandering claims.
- 30 -
- 31 - No. 60
for the IRC to submit a second set of maps has long since passed.19 Although the State
respondents assert that, even following a constitutional violation, the legislature possesses
exclusive jurisdiction and unrestricted power over redistricting, the Constitution explicitly
authorizes judicial oversight of remedial action in the wake of a determination of
unconstitutionality — a function familiar to the courts given their obligation to safeguard
the constitutional rights of the People under our tripartite form of government. Thus, we
endorse the procedure directed by Supreme Court to “order the adoption of . . . a
redistricting plan” (NY Const, art III, § 4 [e]) with the assistance of a neutral expert,
designated a special master, following submissions from the parties, the legislature, and
any interested stakeholders who wish to be heard.20
19
To the extent the 2022 redistricting legislation, which we invalidate here, purported to
render any court order “tentative” for a period of 30 days (L 2022, ch 13, § 3, [5] [i]) such
a limitation on judicial authority appears inconsistent with (among other things) the
constitutional provision authorizing judicial review without limitation and requiring
“disposition” of the claim by Supreme Court within 60 days. The Constitution does not
contemplate an advisory order. In any event, here, due to the procedural constitutional
violations and the expiration of the outer February 28th constitutional deadline for IRC
action, the legislature is incapable of unilaterally correcting the infirmity.
20
While accusing this Court of “step[ping] out of its judicial role” (Troutman, J. dissenting
in part op, at 2), Judge Troutman crafts a remedy that is neither consistent with the
constitutional text nor requested by any of the parties to this proceeding. She proposes that
the legislature should be directed to adopt one of the two plans submitted by the IRC and
already rejected by the legislature (although she does not specify which one). Judge
Troutman’s position is incongruous; she agrees that the legislature lacked authority to enact
redistricting legislation absent a second submission from the IRC but, paradoxically, she
suggests that we should now order the legislature to enact redistricting legislation despite
their inability to cure the procedural violation. Moreover, although Judge Troutman posits
that the People would not approve of a court-ordered redistricting map that is, in fact,
exactly what the People have approved in the State Constitution as a remedy by declaring
that the IRC “process . . . shall govern . . . except to the extent that a court is required to
order the adoption of, or changes to, a redistricting plan as a remedy for a violation of law”
- 31 -
- 32 - No. 60
Nearly a century and a half ago, we wrote that “[t]he Constitution is the voice of the
people speaking in their sovereign capacity, and it must be heeded” (Matter of New York
El. R.R. Co., 70 NY 327, 342 [1877]). Thirty years later, we relied on that fundamental
principle to conclude that “[a] legislative apportionment act cannot stand as a valid exercise
of discretionary power by the legislature when it is manifest that the constitutional
provisions have been disregarded . . . [because] [a]ny other determination by the courts
might result in the constitutional standards being broken down and wholly disregarded”
(Matter of Sherrill v O’Brien, 188 NY at 198). Today, we again uphold those constitutional
standards by adhering to the will of the People of this State and giving meaningful effect
to the 2014 constitutional amendments.
We therefore remit the matter to Supreme Court which, with the assistance of the
special master and any other relevant submissions (including any submissions any party
wishes to promptly offer), shall adopt constitutional maps with all due haste. Accordingly,
the Appellate Division order should be modified, with costs to petitioners, in accordance
with this opinion and, as so modified affirmed.
(NY Const, art. III, § 4 [e]). Just as puzzling, Judge Wilson begins his dissent with a
nonsensical advisory opinion, indicating that although he concludes no violation of the
constitution occurred, he nonetheless agrees with Judge Troutman’s proposed remedy – a
solution to a problem that, in his view, does not exist.
- 32 -
TROUTMAN, J. (dissenting in part):
I agree with the majority that petitioners have standing, and I further agree with the
majority’s holding that the 2022 congressional and state senate redistricting plans (2022
plans) were not enacted by the legislature in compliance with the constitutional process.
-1-
-2- No. 60
However, I dissent as to the majority’s advisory opinion on the substantive issue of whether
the plans constitute political gerrymandering and as to the remedy.
The majority correctly concludes that sections 4, 5, and 5-b of article III of the State
Constitution, as ratified by the citizens of the State, provide the exclusive process for
redistricting (see NY Const, art III, § 4 [e]). This process requires, among other things,
that any redistricting plan to be voted on by the legislature must be initiated by the
Independent Redistricting Committee (IRC) (see § 4 [b]). Once this Court holds that the
2022 plans were unconstitutionally enacted and must be stricken on that threshold basis, it
should not then step out of its judicial role to further opine on the purely academic issue of
whether the 2022 congressional map failed to comply with the substantive requirements of
section 4 (c) (5). The 2022 plans, which the majority concludes are void ab initio, are no
longer substantively at issue, nor can the majority seriously claim them to be so.
Furthermore, although the majority purports to provide “necessary guidance to inform the
development of a new congressional map on remittal” (majority op at 24 n 12), the
majority’s opinion provides no such guidance. Its conclusion, based on affirmed findings
of fact that the congressional map was drawn with partisan intent, is not illuminating in the
least because the majority does not engage in the kind of careful district-specific analysis
that might provide any practical guidance to an actual mapmaker, nor could it on this record
(cf. Wilson dissenting op at 12-25). By opining on this academic issue, the majority renders
“an inappropriate advisory opinion” by “prospectively declar[ing] the [redistricting]
invalid on additional . . . constitutional grounds” (T.D. v New York State Off. of Mental
Health, 91 NY2d 860, 862 [1997]; see Self-Insurer's Assn. v State Indus. Commn., 224 NY
-2-
-3- No. 60
13, 16 [1918] [Cardozo, J.] [“The function of the courts is to determine controversies
between litigants . . . They do not give advisory opinions. The giving of such opinions is
not the exercise of the judicial function”]).
Given the procedural violation flowing from the breakdown in the constitutional
process, we must fashion a remedy that matches the error.1 The Constitution contemplates
that a court may be “required to order the adoption of . . . a redistricting plan as a remedy
for a violation of law” (NY Const, art III, § 4 [e]). In so ordering, where a court finds that
redistricting legislation violates article III, “the legislature shall have a full and reasonable
opportunity to correct the law’s legal infirmities” (§ 5). Consistent with these provisions,
this Court should order the legislature to adopt either of the two plans that the IRC has
already approved pursuant to section 5-b (g). Those plans show significant areas of
bipartisan consensus among the IRC commissioners. The boundaries of the districts of
Upstate New York, in particular, are nearly identical between the two plans and similar to
those in the procedurally infirm plan enacted by the legislature (see Matter of Harkenrider
v Hochul, — AD3d —, 2022 NY Slip Op 02648, *7 [4th Dept April 21, 2022] [Whalen,
P.J. & Winslow, J., dissenting in part]). Given the existence of these IRC-approved plans,
there is no need for a redistricting plan to be crafted out of whole cloth and adopted by a
court. Rather, the legislature should be ordered to adopt one of the IRC-approved plans on
a strict timetable, with limited opportunity to make amendments thereto. As part of our
judicially crafted remedy, we could order that any amendments to either plan “shall not
1
The majority seems unwilling to grasp this concept (majority op at 31-32 n 20).
-3-
-4- No. 60
affect more than [2%] of the population of any district contained in such plan” (Legislative
Law former § 94). In other words, the legislature would be bound by its own self-imposed
restrictions, which were in effect at the time these plans were first presented for legislative
approval.
Such a remedy not only adheres more closely to the constitutional redistricting
process, but it discourages political gamesmanship. Throughout this proceeding,
respondents have asserted that the legislature has near-plenary authority to adopt a
redistricting plan, whereas petitioners have sought to take the process out of the hands of
the legislature and to place it into the hands of the judiciary. It is of course disputed why
the constitutional process broke down, but it is readily apparent that the IRC’s bipartisan
commissioners failed to fulfill their constitutional duty. None of the parties is entitled to
the resolution that he or she seeks.
In addition, this remedy allows the legislature to enact a plan that minimizes the
impact on the reliance interests of both the voters and candidates. Petitions have been
circulated, citizens have contributed monetary donations to the candidates of their choice,
and eligible voters have had the opportunity to educate themselves on the candidates who
are campaigning for their votes, all in reliance on the procedurally infirm redistricting plan
enacted by the legislature. Of course, entrenched candidates have the party apparatus to
support them in the event that further redistricting causes excessive upset to the current
plan. In such a circumstance, outside candidates, upstart candidates, and independent
candidates, who lack the resources of the well-heeled, will be disadvantaged most, leaving
-4-
-5- No. 60
the voters who support them without suitable options. The legislature, duly elected by the
citizens of this State, is in the best position to take these considerations into account.
Yet, the remedy ordered by the majority takes the ultimate decision-making
authority out of the hands of the legislature and entrusts it to a single trial court judge.
Moreover, it may ultimately subject the citizens of this State, for the next 10 years, to an
electoral map created by an unelected individual, with no apparent ties to this State, whom
our citizens never envisioned having such a profound effect on their democracy. That is
simply not what the people voted for when they enacted the constitutional provision at
issue. Although the IRC process is not perfect, it is preferable to a process that removes
the people’s representatives entirely from the process. The majority states that it
“decline[s] to render the constitutional IRC process inconsequential in the manner
requested by the State respondents” (majority op at 23); however, the majority does just
that by crafting a remedy that cuts the legislature out of the process. The citizens of the
State are entitled to a resolution that adheres as closely to the constitutional process as
possible. By ordering the legislature to enact redistricting legislation duly initiated by the
IRC, this Court could afford the legislature its “full and reasonable” opportunity while
honoring the constitutional process ratified by the people.
-5-
WILSON, J. (dissenting):
I agree with Judge Troutman that Article III, Section 5 of the Constitution means
that the majority’s referral of this matter to a special referee is not allowable, and I further
agree that her proposed solution of requiring the Legislature to act on the Independent
-1-
-2- No. 60
Redistricting Commission (“IRC”) maps that have been submitted, though novel, would
be acceptable in the unusual circumstances presented here. I also fully concur in Judge
Rivera’s dissenting opinion, and I do not view Judge Rivera’s opinion as necessarily
inconsistent with Judge Troutman’s proposed remedy. Therefore, I address the merits of
the claim that the 2022 redistricting itself violates the Constitution. It does not.
The burden a plaintiff must meet to overturn legislative action as violative of the
New York Constitution is extraordinarily high. We have often (though not always)
described that burden as proving unconstitutionality “beyond a reasonable doubt” (Matter
of Wolpoff v Cuomo, 80 NY2d 70, 78 [1992]; but see Matter of City of Utica, 91 NY2d 964
[1998] [upholding a state statute’s constitutionality without reference to the beyond a
reasonable doubt standard]; Matter of Sherrill v O’Brien, 188 NY 185, 198 [1907] [“A
legislative apportionment act cannot stand as a valid exercise of discretionary power by the
legislature when it is manifest that the constitutional provisions have been disregarded”];
Matter of Whitney, 142 NY 531, 533 [1894] [upholding the apportionment of Kings County
into assembly districts because, although flawed, “the division has seemed to us a
reasonable approach to equality, and under all the circumstances of the case a substantial
obedience to the writ”]). Both Supreme Court and the Appellate Division described the
test that way. Thus, to prevail, the petitioners need to have proved beyond a reasonable
doubt that the Legislature’s 2022 Congressional and State Senatorial districts were “drawn
to discourage competition or for the purpose of favoring or disfavoring incumbents or other
particular candidates or political parties” (NY Const, art III, § 4 [c] [5]). It is important to
-2-
-3- No. 60
pay close attention to the wording of the Constitution. It does not prohibit the creation (or
maintenance) of districts that are highly partisan in one direction or the other. Indeed, both
in New York and around the rest of the nation, voters tend to cluster in geographic areas
that reflect party affiliation. As a simple example, rural areas in New York and in the
United States generally tend to have much higher concentrations of Republican voters than
do urban areas. What the Constitution prevents is purposefully drawing districts to
discourage competition or favor particular parties or candidates.
After a review of the record, I am certain that the petitioners failed to satisfy the
“beyond a reasonable doubt” standard. By that, I do not mean to say that I know the
Legislature did not draw some districts in a way that violated our State Constitution; rather,
the evidence here does not prove that to be the case at the level of certainty required to
invalidate the 2022 redistricting as unconstitutional. Perhaps with a different record,
petitioners could make such a showing, but they have failed to do so here.
The question before us, then, is whether the petitioners introduced sufficient
evidence to discharge their very high burden of proving that the Legislature adopted
gerrymandered district lines in violation of the Constitution. That is unequivocally a
question of law, and thus within the heartland of our Court’s power of review (see
Glenbriar Co. v Lipsman, 5 NY3d 388, 392 [2005]; see also People v Jin Cheng Lin, 26
NY3d 701, 719 [2016] [noting that whether “the proof (does not meet) the reasonable doubt
standard” is “a matter of law” (alterations in original)]; People v Tarsia, 50 NY2d 1, 13
[1980] [evaluating “the total evidence” as to whether “the proof was insufficient as a matter
-3-
-4- No. 60
of law to support the affirmed findings that defendant’s inculpatory statements . . . were
voluntary”]; People v Anderson, 42 NY2d 35, 39 [1977] [“(W)hether the proof met the
reasonable doubt standard at all is a matter of law”]; People v Leonti, 18 NY2d 384, 389
[1966] [“(W)hether the evidence adduced meets the standard required is one of law for our
review”]). The majority incorrectly treats this as an unreviewable question of fact,
characterizing Supreme Court’s finding that the 2022 congressional map was drawn to
discourage competition as a factual “determination” that has “record support” and thus
should not be “disturb[ed]” (majority op at 26-27)—a distinct, and here inapt, standard (see
Stiles v Batavia Atomic Horseshoes, Inc., 81 NY2d 950, 951 [1993]).
Indeed, it is remarkably inaccurate to suggest that our Court is without power to
review the Appellate Division’s ruling on the partisan gerrymander claim. This case is
before us as an appeal as of right based on CPLR 5601 (b). This case satisfies the
conditions for an appeal as of right because the question presented—whether a
congressional map, i.e., a legislative enactment, is constitutionally invalid—is a question
of law that is reviewable by this court (see Cayuga Indian Nation of New York v Gould, 14
NY3d 614, 635 [2010] [“(A) query concerning the scope and interpretation of a statute or
a challenge to its constitutional validity” is a “pure question of law”]).
Petitioners’ evidence falls into three basic categories. First, petitioners primarily rely on
the testimony of Sean P. Trende, an elections analyst and doctoral candidate at Ohio State
University. At best, Mr. Trende’s results are incomplete and inconclusive, but they are
also legally insufficient to meet the above standard. Second, petitioners rely on the
-4-
-5- No. 60
projected loss of four Republican Congressional seats (out of eight that currently exist).
The difficulty with that proof is that it assumes that factors unrelated to how the districts
were drawn have not caused the result. Third, petitioners contend that the 2022
redistricting was accomplished through the complete exclusion of Republican members of
the Legislature from the process and a failed attempt by Democrats to further amend the
Constitution, followed by the enactment of a statute. I view that as their best argument in
support of their gerrymander claim but one that, without more, does not meet the high bar
for invalidating the Legislature’s 2022 redistricting plan.
I
The petitioners, Supreme Court, and the Appellate Division plurality each relied
heavily on the testimony of Mr. Trende. Mr. Trende’s testimony is based on simulations
in which a computer algorithm uses demographic data, takes parameters set by the user,
and draws districting maps for the region (in this case, New York State) specified by the
user. This is the first time Mr. Trende has testified in a case in which he prepared
redistricting simulations of any kind. Instead of using the Markov Chain Monte Carlo
simulation algorithm, which has been regularly used in redistricting cases, Mr. Trende used
a new simulation algorithm developed by Dr. Kosuke Imai, a Harvard professor, along
with publicly available political and demographic data at the census block and precinct
levels. Dr. Imai’s new algorithm appeared in an unpublished paper that had yet to be peer-
reviewed. In that paper, Dr. Imai reported that he had tested the reliability of his new model
-5-
-6- No. 60
by applying it to a 50-precinct map and running 10,000 simulations. By comparison, New
York State has more than 140,000 precincts; uncontroverted evidence (including from Mr.
Trende) establishes that the complexity of producing a working algorithm increases as the
number of precincts increases.
In brief, Dr. Imai’s algorithm draws possible maps, starting from a blank page, but
taking into account parameters the user sets. For example, a user can specify to avoid
splitting a county (or city) into different districts, though sometimes splitting is inevitable
and may be accomplished in myriad ways. By running thousands of simulations and
comparing them to what the Legislature has done, the model allows for measurement of
the difference in party breakdown between the collection of simulated maps and the
legislatively drawn map. The model can produce summary statistics showing, for example,
that, when compared to the legislative map, the simulated maps distribute voters of one
party or another (here, Republicans) in a way that concentrates a lot of them into some
districts where Republicans would likely have won elections anyway, thus removing them
from districts where Democrats might have faced a close election. In simple terms, Mr.
Trende concluded that the legislative map consolidated Republican voters into a few
Republican-leaning districts and spread Democratic voters in an efficient fashion. Of
course, the model cannot tell you why the Legislature drew the districts that way, but,
provided that a scientific method is proven to be reliable, the data entered is of good quality,
the parameters chosen are correct, and the results are robust (i.e., not susceptible to material
swings in output when parameters are varied within reasonable ranges for those
-6-
-7- No. 60
parameters), the law allows intent to be inferred from results in a variety of areas (e.g.,
People v Guzman, 60 NY2d 403, 412 [1983] [discriminatory intent inferred from
underrepresentation in Grand Jury selection]; 303 W. 42nd St. Corp. v Klein, 46 NY2d 686,
695 [1979] [discriminatory intent inferred from “a convincing showing of a grossly
disproportionate incidence of nonenforcement against others similarly situated in all
relevant respects save for that which furnishes the basis of the claimed discrimination”]).
Again, Article III, Section 4 of the Constitution states that “[d]istricts shall not be
drawn to discourage competition or for the purpose of favoring or disfavoring incumbents
or other political candidates or other political parties” (emphasis added). The prohibition,
then, is against drawing maps with the intention to discourage competition or favor or
disfavor incumbents, political candidates, or political parties. In other words, if a given
map ends up discouraging competition or favoring a political party, that map does not
necessarily run afoul of the Constitution’s prohibition. Instead, an intent to discourage
competition or to favor that political party must be shown for the map to violate the
Constitution.
Staten Island provides a good example to keep in mind, one to which I will return
later. Staten Island is traditionally Republican. It does not have quite enough people in it
to constitute an entire congressional district, but it forms the vast portion of Congressional
District 11, both in the 2010 districting and the Legislature’s 2022 districting, with the
added voters coming from Brooklyn. No one suggests that, by keeping Staten Island intact
within a single congressional district instead of splitting it across two districts with more
-7-
-8- No. 60
Brooklynites, the Legislature in 2010 or 2022 did so with the intent to advantage
Republicans. If you split Staten Island into two different congressional districts and added
enough Brooklynites to fill out those districts, each of the districts would have more
Brooklynites than Staten Islanders, and the strength of the Republican voting of Staten
Island would be diluted. The two new districts might be more competitive—i.e., closer to
50/50 than District 11 is or has been—but it is sufficient, to reject a claim of intent to
advantage Republicans by keeping Staten Island whole within a single district, to say that
it is an island and people there live in communities that are distinct from those in Brooklyn.
Again, the why is important, not the what.
Mr. Trende’s testimony and analysis were legally insufficient to bear on the question
of intent for three reasons. First, the New York Constitution requires the consideration of
several specifically identified factors when creating congressional districts, with some
additional factors required for State Senatorial districts. Thus, Mr. Trende’s results at most
show that if we amended the New York Constitution to strike out those factors, he could
conclude the Legislature acted with intent to disfavor Republicans or reduce competition.
Second, close examination of districts in the real world, as compared to those hidden in
thousands of hypothetical unseen maps, further exposes the unreliability of Mr. Trende’s
conclusions. Finally, the novelty of Dr. Imai’s algorithm and the opacity of Mr. Trende’s
implementation of it create very substantial doubt as to his conclusions. The method is
novel and not peer reviewed. Mr. Trende did not attempt the established Markov Chain
Monte Carlo simulation to compare it to his results, nor did he provide the model, inputs,
-8-
-9- No. 60
data sets, or output maps that formed the basis for his analysis. Indeed, neither he nor
anyone has seen the algorithm-produced maps underlying his analysis. We are being asked
to determine unconstitutionality based on shadows.
New York’s Constitution requires that the following factors be considered when
drawing congressional districts:
1. Compliance with “the federal constitution and statutes” (NY Const, art III, § 4 [c]);
2. “whether such lines would result in the denial or abridgement of racial or language
minority voting rights, and districts shall not be drawn to or have the purpose of,
nor shall they result in, the denial or abridgement of such rights” (id. § 4 [c] [1]);
3. “Districts shall be drawn so that, based on the totality of the circumstances, racial
or minority language groups do not have less opportunity to participate in the
political process than other members of the electorate and to elect representatives of
their choice” (id.);
4. “Each district shall consist of contiguous territory” (id. § 4 [c] [3]);
5. “Each district shall be as compact in form as practicable” (id. § 4 [c] [4]);
6. “Districts shall not be drawn to discourage competition or for the purpose of
favoring or disfavoring incumbents or other particular candidates or political
parties” (id. § 4 [c] [5]);
7. Consideration of “the maintenance of cores of existing districts” (id.); and
8. Consideration of the maintenance of “pre-existing political subdivisions, including
counties, cities and towns, and of communities of interest” (id.).
-9-
- 10 - No. 60
For senatorial districts, the Constitution adds requirements that “senate districts not divide
counties or towns, as well as the ‘block-on-border’ and ‘town on border’ rules” (id. § 4 [c]
[6]).
Mr. Trende admittedly did not attempt to have his simulations account for several
of the constitutionally required factors listed above. For that reason alone, his simulations
do not provide evidence of the Legislature’s intent to disfavor Republicans or reduce
competition. Putting aside all other methodological and implementation problems, a
proper comparison would ask: what would an unbiased mapmaker (the algorithm) do if
given the same constitutional requirements as the Legislature has? Instead, Mr. Trende has
attempted to answer a different question: what would an unbiased mapmaker do if it lacked
some of the constitutional requirements the Legislature is required to follow?
This is not merely a conceptual problem, which is readily seen by identifying the
constitutional factors Mr. Trende omitted. First, under the Equal Protection Clause and the
federal Voting Rights Act (“VRA”), the composition of congressional districts must not
discriminate on the basis of race or color (52 USC § 10301; US const, amend XIV, § 1).
New York’s Constitutional requirements, listed as items 2 and 3 above, represent similar
protections not just on the basis of race, but language as well. Mr. Trende gave no
instruction to his algorithm to take any consideration of those constitutional requirements
for drawing districts. Mr. Trende noted that his “simulated maps are not drawn with any
racial data available to the simulation”—that is, the simulation could not even take race
into account in drawing districts if Mr. Trende had specified that as a parameter. Likewise,
- 10 -
- 11 - No. 60
nothing in the record suggests that Mr. Trende’s simulation used any data concerning the
language of inhabitants, and he made no claim to have done so.
Faced with criticism that he had omitted consideration of factors 1 through 3 above,
Mr. Trende responded generally that, “every one of Respondent’s experts could readily
demonstrate that . . . fixing the purported omissions might lead this Court to arrive at
different conclusions,” which, as explained below, attempts to shift the burden of proof
onto respondents. He then explained his omission on the ground that “there is no evidence
proffered by any party of racially polarized voting in New York City or in particularized
boroughs, nor is there any evidence that any single minority group can form a reasonably
compact majority in a district.” Besides lacking any evidentiary support, his assertion is
patently and commonly understood to be wrong. Looking just to last year’s New York
City mayoral election, Curtis Sliwa, the Republican nominee, “scored 44% of the vote in
precincts where more than half of residents are Asian — surpassing his 40% of votes in
white enclaves, 20% in majority-Hispanic districts and 6% in majority-Black districts”
(Rong Xiaoqing et al., Chinese Voters Came Out in Force for the GOP in NYC, Shaking
Up Politics, The City [Nov 11, 2021],
https://www.thecity.nyc/politics/2021/11/11/22777346/chinese-new-yorkers-voted-for-
sliwa-gop-republicans). In the same election, now-Mayor Eric Adams “dominated” the
“Black Bloc,” a “63 percent non-Hispanic Black and 23 percent college-educated swath of
Brooklyn and Queens,” where Adams grew up and where he won “63 percent of first-place
votes” (Nathaniel Rakich, How Eric Adams Won The New York City Mayoral Primary,
- 11 -
- 12 - No. 60
FiveThirtyEight [Aug 25, 2021], https://fivethirtyeight.com/features/how-eric-adams-
won-the-new-york-city-mayoral-primary/).
Mr. Trende attempted to make some account of the omission of the federal and state
protections for racial minority voting rights by “freezing” certain census blocks in nine
districts to remove them from his analysis, explaining that those districts are “plausible
candidates for protection under the VRA or the State Constitution.” Even assuming that
his choice of districts is sound, his results demonstrate the importance of his omission of
constitutionally required factors: his “frozen” simulations produced results that “ma[ke]
Petitioner’s case more difficult.” Specifically, those “plausible” protections for minority
voters produced results that “accept[] the Legislature’s decision to pair Yorktown with
Yonkers in the Sixteenth District, and to crack Republican-leaning areas in Midwood and
Sheepshead Bay between the Ninth and Eighth districts.” In other words, by including
even a rough proxy for protection of minorities, he admits that some of what he described
as gerrymandering is explainable instead by protection of minority voting rights. Mr.
Trende’s utter lack of consideration of the constitutional requirement to consider protection
of non-English language groups inherently means his simulations do not show what an
unbiased mapmaker would do if that constitutional command mattered.
Likewise, Mr. Trende completely neglected considering keeping “communities of
interest” together (item 8 above), as the Constitution requires. Keeping in mind that
differences in party affiliation within a district do not matter unless they were created with
the intent to disadvantage a party or candidate or to reduce competition, Mr. Trende ignored
- 12 -
- 13 - No. 60
that the IRC—composed in equal parts of persons appointed by Democrats and
Republicans—reached agreement on keeping together many communities of interest. For
example, both sets of IRC maps (one produced by the Democratic faction and the other by
the Republican faction) agreed that the Southern Tier of New York should be unified in a
district. The Southern Tier is a strip of eight counties along upstate New York’s southern
edge, the part of the state that shares a border with Pennsylvania.1 Those counties are
grouped as a region in New York State’s materials on economic development (see New
York State, Empire State Development: Southern Tier,
https://esd.ny.gov/regions/southern-tier [last accessed Apr 26, 2022]). Indeed, the region
has a storied history of being a manufacturing powerhouse, though the region also faced
struggles within the past decade due to a decline in manufacturing and uncertain economic
development (Susanne Craig, New Yorkʼs Southern Tier, Once a Home for Big Business,
Is Struggling, NY Times [Sept 29, 2015],
https://www.nytimes.com/2015/09/30/nyregion/new-yorks-southern-tier-once-a-home-
for-big-business-is-struggling.html). Those counties are more Republican than
Democratic; in a show of how culturally distinct the region is, hundreds of residents in the
Southern Tier in 2015 rallied in support of seceding from the state of New York (id.). One
Republican lawmaker even applauded the fact that the maps proposed by the Democratic
and Republican commissioners to the IRC both kept the Southern Tier intact (Rick Miller,
1
The Southern Tier has long been recognized as a cohesive political unit (see Warren
Moscow, GOP Held Strong in Southern Tier, NY Times [Oct 16, 1946],
https://timesmachine.nytimes.com/timesmachine/1946/10/16/107146657.html?pageNum
ber=31).
- 13 -
- 14 - No. 60
Southern Tier Congressional District Essentially Maintained in NY Redistricting Maps,
Olean Times Herald [Jan 4, 2022], https://www.oleantimesherald.com/news/southern-tier-
congressional-district-essentially-maintained-in-ny-redistricting-maps/article_56c5d543-
6c8a-55d3-a3de-e662bdb0f6dd.html). For upstate New York, the Democratic
Commissioners and the Republican Commissioners agreed that there should be three
Republican-leaning districts: one uniting the Southern Tier, one uniting the North Country,
and one by Lake Ontario. The Commissioners from the two parties also agreed that there
should be Democratic-leaning districts in the four urban areas in upstate New York: in and
around Albany, Syracuse, Rochester, and Buffalo. The result of those bipartisan decisions
by the IRC demonstrates that those districts (broadly, all of upstate New York, about which
the IRC had no substantial disagreements) should have been excluded from Mr. Trende’s
simulations. But even though the Southern Tier and the other upstate counties and cities
were bipartisanly districted as “communities of interest,” Mr. Trende made no effort to
keep the Southern Tier, or other communities of interest, intact in his model. Indeed, Mr.
Trende “didn’t pay any attention to what any of those [IRC] commissioners [had] done in
their proposals,” had not read any of the testimony before the IRC, and did not know
whether there was any testimony before the IRC about communities of interest.
Instead, he told Supreme Court that such communities are too difficult to code, even
though he also acknowledged that in a redistricting exercise he undertook for Virginia, he
and his co-researcher accounted for communities of interest. Mr. Trende did not do any
sort of proxy analysis as he did for race, and because neither he nor anyone else ever looked
- 14 -
- 15 - No. 60
at the 10,000 maps his simulation drew, he has no idea what his algorithm did to the
Southern Tier or any other upstate areas. But Dr. Imai’s own data provides some insight.
Mr. Trende used Dr. Imai’s model and data. The record includes four sample maps
from a set of 5,000 simulations for New York prepared by Dr. Imai himself. Two of the
sample maps from Dr. Imai’s simulations broke up the North Country. All three of the
sample maps broke up the Southern Tier. None of Dr. Imai’s sample maps maintained
Democratic-leaning districts around all of Albany, Syracuse, Rochester, and Buffalo.
Those samples strongly suggest that Mr. Trende’s conclusions about intentional
gerrymandering depend on comparison to maps that would have broken up congressional
districts arrived at by bipartisan consensus. Of course, had Mr. Trende looked at his own
maps, or even turned them over for respondents to examine, we would be able to know
how many of his “less gerrymandered” simulations were incompatible with districting
actually arrived at bipartisanly, with regard for the constitution’s directions.2 Instead, it is
2
Mr. Trende’s decision not to examine his own maps and not to permit anyone else to see
them poses a separate reliability issue. Dr. Imai’s algorithm generates huge numbers of
redundant maps, which should be weeded out before analysis is conducted. Mr. Trende
himself did so when working on a redistricting map for Maryland. There, he completed
three sets of 250,000 simulations. He then eliminated the duplicates, which ranged from
220,000 to 160,000 for each of his sets—that is, 64% to 88% of the maps produced were
duplicates that he discarded (Szeliga v Lamone, Nos. C-02-CV-21-00173, Slip Op at 99,
102-104). Furthermore, New York State is significantly larger than Maryland; whereas
Maryland only has 8 congressional districts, New York has 26 congressional districts. Mr.
Trende acknowledged that that the more precincts are involved, the more complicated it
becomes to accurately use redistricting simulations to draw conclusions. Yet, in spite of
acknowledging that using simulations for New York would be more difficult than for
Maryland, Mr. Trende inexplicably generated only 10,000 simulations for New York and
subsequently failed to check even that small set for duplicates.
- 15 -
- 16 - No. 60
clear that, just as with the racial and language protections in the constitution, Mr. Trende’s
exclusion of communities of interest has made his analysis legally irrelevant: at most, it
answers what an unbiased mapmaker would do if that mapmaker was told to disregard
protection of racial minorities, language minorities and communities of interest.
One final example from Dr. Imai’s work illustrates the unsoundness of Mr. Trend’s
conclusions. His conclusions are based on comparing the algorithm-drawn simulated
districts, which purportedly are “less gerrymandered,” against the Legislature’s
redistricting plan. Because neither we nor Mr. Trende knows what his “less
gerrymandered” maps look like, we cannot know whether they are sensible maps that
should be included in such a comparison. But because Dr. Imai, using the same data and
same model, displayed some sample maps, we can observe the kind of maps Mr. Trende
has relied on for his conclusions. Sample Plan 1 from Dr. Imai’s simulation placed
Schuyler County and Franklin County into the same congressional district. Schuyler
County is near upstate New York’s southern border with Pennsylvania, and Franklin
County is one of the northernmost counties in New York, on the border with Canada—that
is, those two counties are on opposite sides of upstate New York. Their county seats are
262 miles away via highway (Google, Google Maps Driving Directions for Driving from
Watkins Glen, New York to Malone, New York, https://perma.cc/L3KH-DN5B [last
accessed Apr 26, 2022]). In essence, what Mr. Trende is showing is that the partisan
imbalance of some congressional districts could be reduced by radically rejiggering them
- 16 -
- 17 - No. 60
in a way that no human mapmaker (or resident of either of those counties) would think
remotely sensible. Interesting though it may be, it is legally irrelevant.
Apart from the omitted constitutional requirements, the creation of districts requires
balancing among the different constitutional requirements. Some are relatively
inflexible—such as districts of equal population (see Baker v Carr, 369 US 186 [1962]),
compliance with the VRA or, for senatorial districts, the “block-on-block” rule; others,
such as compactness or protection of communities of interest, allow for an exercise of
judgment in how to balance them. Mr. Trende made no explicit decision in how to balance
the factors he did include, was uninformative about what balance was implied, and did not
vary the relative weights of his parameters to determine the robustness of his conclusions.
For instance, Mr. Trende included a parameter for the compactness of districts, which the
constitution instructs should be considered. When asked how he valued compactness, he
testified to selecting a value of “1” in Dr. Imai’s model because he knew that “the other
choices don’t work well.” He agreed that the compactness parameter could be set at less
than 1, or more than 1, but provided no explanation for what the settings meant, how much
priority a change in setting gave to compactness versus any other factor, or even what was
meant by other values not working well—which may simply mean that when he tested for
robustness of the parameter, he found that changing the relative weight given to
compactness resulted in statistics that did not support his conclusions or that the model
ceased to function, neither of which should give us confidence sufficient to hold the
redistricting unconstitutional.
- 17 -
- 18 - No. 60
Similarly, Mr. Trende said that Dr. Imai’s model allowed an “on” or “off” switch
on whether to split counties. He put that switch “on,” even though New York map drawers
must balance county preservation with other considerations—effectively meaning he gave
county integrity a superpriority over other constitutional factors. Nothing in the
Constitution requires the Legislature to prefer county integrity over any other factor, or
even to give the same priority to county integrity for every county. Rather, the Constitution
gives the Legislature flexibility in weighting many of the required considerations
differently in different circumstances, but Mr. Trende implicitly assigned fixed and
universal relative weights to every one of those that he included. Faced with the potential
for differently weighting parameters, responsible modelers alter the parameters within
reasonable bounds to see whether the alterations make a difference. When the difference
is not great, models are robust; when they are great, models are lacking in probative value
(see, e.g., Amariah Becker et al., Computational Redistricting and the Voting Rights Act,
20 Election L J 407, 430 & n 31 [2021]). When nobody tests for robustness, invalidating
districts as unconstitutional beyond a reasonable doubt is sheer guesswork.
Respondents pointed out the many deficiencies in Mr. Trende’s model. In addition
to the examples explained in detail above, Mr. Trende repeatedly and improperly answered
in a way that attempted to shift the burden of proof from petitioners onto respondents. For
instance, in response to respondents’ assertion that his failure to consider all the relevant
constitutional considerations undermined the validity of his methodology, Mr. Trende
asserted that “[e]very one of Respondents’ experts is more than capable of either re-running
- 18 -
- 19 - No. 60
the relevant simulation algorithm that I employed or executing a competing algorithm” and
“[i]f there are indeed important communities of interest to be protected, however, any of
Respondents’ experts could program a simulation that respected those communities of
interest and potentially harm Petitioners’ case.” On cross-examination, he reiterated that
“if there is something that [the respondents’] experts believe . . . is missing that makes a
difference -- they think makes a difference, they can do it.”
The lower courts erroneously acceded to Mr. Trende’s burden shifting, which itself
is a legal error requiring reversal (Harkenrider v Hochul, --- AD3d ---, 2022 NY Slip Op
02648, *7 [4th Dept 2022] [Whalen, P.J., dissenting]).3 Proof beyond a reasonable doubt
is an exacting standard: a party bearing that burden must remove all reasonable doubt,
which is not met by saying that the opponent has the ability to disprove an assertion.
Faulting the respondents for the petitioners’ failure to account for constitutionally required
redistricting criteria improperly reverses the burden of proof; it is the petitioner’s burden
to prove unconstitutional partisan intent beyond a reasonable doubt.
In short, the factors set out in the Constitution must be considered during
redistricting with flexibility in the relative weighting on a case-by-case basis. Maintaining
the Southern Tier as a community of interest may be powerfully important; maintaining
the Upper West Side as one may not be. Mr. Trende acknowledged that his algorithm
3
For example, Supreme Court noted that Mr. Trende “did not include every constitutional
consideration”—which should render his evidence legally insufficient. Supreme Court
explained away that deficiency by saying that “[n]one of Respondents’ experts attempted
to draw computer generated maps using all the constitutionally required considerations,” a
clear example of improper burden shifting.
- 19 -
- 20 - No. 60
cannot undertake that balancing, and to his credit explained that “the more that you
adequately control all of the variables that the actual mapmakers actually used, the more
you can infer intent, and the less you adequately control for those variables, the less you
can infer intent” to gerrymander. Because Mr. Trende’s analysis omitted constitutionally
required factors and fixed implicit weights for others without allowing for flexibility, all
his analysis demonstrates, at best, is that if our Constitution read very differently, he could
find an intent to gerrymander. That conclusion is orthogonal to the issue here.4
II
Apart from Mr. Trende’s opinion, the Appellate Division plurality concluded that
the “‘application of simple common sense’ from the enacted map itself and its likely effects
on particular districts” supports petitioners’ argument that the legislative districts were
intentionally created to disfavor a party or candidate or render certain districts less
competitive (2022 NY Slip Op 02648, *5 [citations omitted]). There are three significant
problems with that conclusion. First, as noted above, for the great majority of
4
The error in the majority’s sole, footnoted response, contending that I have performed a
weight of the evidence analysis (majority op at 27 n 14), can be illustrated as follows: Mr.
Trende uses a Ouija board to determine that the districts have been gerrymandered, and,
when communicating with the spirits in the netherworld, directs them to the provisions in
North Carolina’s constitution instead of New York’s. The lower courts rely on that
evidence to hold that the New York Legislature has engaged in gerrymandering.
According to the majority, the New York Court of Appeals could not conclude an error of
law has been made. The majority is right about one thing: I disagree that my job is so
limited.
- 20 -
- 21 - No. 60
congressional (and senatorial) districts, the Republican and Democratic factions of the IRC
substantially agreed as to the district boundaries, and the legislative plan does not deviate
materially in the case of those districts. Of course, that does not resolve the question for
districts on which the IRC factions disagreed or for which the Legislature’s plan was
materially different, but it should remove most districts from the dispute.
Second, the Appellate Division relied on the following observation: “under the 2012
congressional map there were 19 elected democrats and 8 elected republicans and under
the 2022 congressional map there were 22 democrat-majority and 4 republican-majority
districts” (2022 NY Slip Op 02648, *3). The majority acknowledged that, standing alone
or even in conjunction with the lack of Republican input into, or vote for, the 2022 map,
the evidence would not be strong enough to surmount the high standard for invalidating
the 2022 redistricting as unconstitutional. However, the mere change in the number of
majority Democratic and Republican districts says nothing about why those changes
occurred or about intent. The inference that the change is nefarious ignores important
undisputed data.
The 2012 districts are obsolete and not a relevant source of comparison. Population
and registration shifts demonstrate that New York’s voting populace has changed in the
Democrats’ favor. In the past ten years, Democratic voter registration has outstripped
Republican voter registration ten-to-one: Democratic voter registration increased by more
than one million people statewide between April 2012 and February 2021, whereas
Republican voter registration increased by less than 100,000 people during the same
- 21 -
- 22 - No. 60
period. Similarly, over the decade, Democrat-leaning counties have increased in
population, whereas Republican-leaning counties have decreased in population. It is
unsurprising that such drastic shifts would occur in just a ten-year time horizon; that’s why
the Constitution requires decennial redistricting (NY Const, art III, § 4 [a]).
The characterization of the outgoing 2012 map as having 19 Democrat-leaning and
eight Republican-leaning districts—in comparison to the four Republican-leaning districts
in the 2022 map—is misleading because it disregards the changes of the last decade. To
start, it is undisputed that one Republican seat under the 2012 map, former District 22, was
eliminated due to substantial population shifts and New York’s loss of a congressional
seat. But more importantly, it is undisputed that, based on the 2020 census data, the 2012
map would also produce only four Republican-leaning districts.
Third, and most importantly, it is undisputed that the 2022 legislative redistricting
was slightly more favorable for Republicans than the array of simulated “unbiased” maps
produced by Mr. Trende’s simulation. The Appellate Division contended that, by “boldly
asserting” that the Democratically created 2022 plan tended to favor Republicans more
than Mr. Trende’s supposedly neutral maps, “respondents have created a further inference
that they acted with a partisan purpose favoring democrats” (2022 NY Slip Op 02648, *4).
That claim confuses intent with effect. I return to Staten Island to illustrate the point.
Staten Island has historically been treated as a community of interest and not split
into different congressional districts. If Staten Island is to be kept that way (wholly within
District 11), it needs to include voters from somewhere else because Staten Island does not
- 22 -
- 23 - No. 60
have enough people to make up a full congressional district. Because of contiguity
requirements, that must be Brooklyn. The 2012 map of District 11 included all of Bay
Ridge (which is just north of the Verrazano Bridge) and Bath Beach, a few blocks of
Bensonhurst, and Gravesend (all south of the bridge). The Legislature’s 2022 redistricting
keeps Bay Ridge to the north (itself a community of interest) with Staten Island, but instead
of then going south, it drops out Bath Beach, the bit of Bensonhurst and Gravesend, and
goes north and incorporates Sunset Park and a small bit of Park Slope.
Among the thousands of comments sent to the IRC after it publicly released its draft
report for comments, looking just at the Richmond and Kings County submissions
(https://nyirc.gov/storage/archive/Kings_Richmond_Redacted.pdf), numerous letters
asked the IRC to keep various groups together. Among those is a letter from OCA-NY
(formerly known as the Organization of Chinese Americans), a “non-profit, non-partisan
organization dedicated to protecting the rights of Asian Americans in New York
City.” That letter urged the IRC that, with regard to District 11, which contained Staten
Island, “Bensonhurst and Bath Beach should NOT be with Staten Island. . . . Staten Island
does not share a similar concentration of Asians, nor the culture of Asian businesses as
Bath Beach/Bensonhurst, nor do residents in Bath Beach/Bensonhurst travel on a regular
basis to Staten Island and vice versa.” Justin Wood, a Staten Islander, asked the IRC to
“counter decades of artificial gerrymandering” by “extend[ing] NY11 northward into Bay
Ridge and Sunset Park to unify linguistic and ethnic communities with shared
interests.” Karen Zhou, the past president of Homecrest Community Services, wrote the
- 23 -
- 24 - No. 60
IRC noting that “Sunset Park, Bensonhurst, Homecrest, Sheepshead Bay, Dyker Heights,
Bath Beach and Gravesend . . . [have] an interconnection bounded by common culture,
language and socioeconomic factors,” further requesting that Bensonhurst and Homecrest
be “together in one Congressional district . . . [to] ensur[e] communities of interest are not
ignored or neglected.”
District 11 has been made less Republican by paying attention to unifying Asian
American communities (which relates to the racial, language and community of interest
requirements in the Constitution), for which the comments to the IRC were uniformly
supportive. Because of contiguity requirements, there was nowhere to go but further north.
The Appellate Division’s observation that the reduction in Republican-leaning districts (or
in the strength of the Republican lean) demonstrates an intent to gerrymander rather than
an attempt to pay attention to the Constitution is unsupportable. Data tells you effect only.
But the record before the IRC shows that various members of the Asian American
community—and one Staten Islander—urged the IRC to go north instead of south
specifically to serve the ends of the VRA and the constitutional provision requiring weight
be given to communities of interest. The algorithmic comparators on which the lower
courts relied, by omitting considerations required by the Constitution, gave zero weight to
those considerations, effectively saying that the Asian American community does not
matter. That, in turn, leads to an unfounded inference that the 2022 redistricting was
intended to disadvantage Republicans, when, in the case of Staten Island, it was intended
- 24 -
- 25 - No. 60
to protect Asian American voting rights and community interests, as the Constitution
requires.
III
The remaining evidence on which petitioners rely to demonstrate that the 2022
redistricting was done with intent to disfavor Republicans or make certain districts less
competitive relates to procedural issues concerning the 2021 legislation, a failed 2021
constitutional amendment, and the creation of the 2022 districts in a three-day period after
the IRC failed to deliver a revised report. Unlike the prior two factors, these are not legally
irrelevant. As the Appellate Division concluded, however, as to petitioners’ arguments on
the process pursued to enact the 2022 map and its projected loss of Republican seats:
without more and even with every reasonable inference taken in petitioner’s favor, they do
not meet the standard to declare the 2022 redistricting plan unconstitutional (2022 NY Slip
Op 02648, *3).
First, petitioners claimed that Democrats unilaterally drafted the 2022 redistricting
map without any input or involvement from Republicans. The Appellate Division plurality
further pointed to the “largely one-party process used to enact the 2022 congressional map”
as partial support for its conclusion that petitioners met their burden of proving an inferred
intent to favor the Democratic party (2022 NY Slip Op 02648, *3). That the process was
dominated by one party, however, is a result of the current political reality of the
Legislature. Put another way, the Legislature reflects the current choice of the people as
- 25 -
- 26 - No. 60
to who will best represent their interests. Indeed, even had the IRC not shirked its duty,
the Democratic supermajority in both houses could have rejected all IRC plans and then,
consistent with the Constitution, adopted a plan without any Republican support. That
result would be “partisan” in a sense, but not in the sense that would be necessary to show
an intent to violate the Constitution. That the vote was along party lines could just as well
suggest that the Republicans wanted to prevent a redistricting map that corrected past
gerrymandering favoring Republicans (or an electoral shift that diminished their chances)
as it could that Democrats sought to exclude Republicans for their party’s benefit.
Next, petitioners contend that the (Democratically controlled) Legislature, in June
2021, passed legislation providing for the possibility that the IRC might not vote on any
redistricting plans, which the Governor signed in November 2021, and that the statute
provides evidence of partisan intent to gerrymander because it provides that the Legislature
will conduct the redistricting in that eventuality. As with the above claim, the statute’s
adoption is not particularly probative as to intent. It is equally possible that the Legislature,
seeing the possibility of electoral chaos in the event that the IRC failed to act as required,
clarified that the outcome would be the same as if the IRC produced plans that the
Legislature rejected. The fact that the statute was passed without Republican support might
suggest a future intent by Democrats to gerrymander. It might suggest an intent by
Republicans to oppose any measures that would correct existing imbalances. Or it might
suggest that legislators simply sought to provide for something not contemplated by the
Constitution.
- 26 -
- 27 - No. 60
Finally, petitioners point to a failed attempt by Democrats to further amend the
Constitution as supporting an inference that the Democrats intended to favor a political
party through the 2022 map. In November 2021, the Legislature proposed a constitutional
amendment to the voters. Under that proposed constitutional amendment—if the IRC
failed to vote on any redistricting plan or plans by the date required—the Commission
would submit to the Legislature all plans in its possession, completed and in draft form,
and the data upon which those plans were based (2021 SB 515 § 5-b [g-1]). If the IRC so
failed in voting and had to submit its plans to the Legislature, that failure would require the
Legislature to create its own redistricting plan, to be enacted by the Governor (id. § 4-b).
The proposed constitutional amendment also included other changes, including increasing
the number of state senators (id. § 2), establishing a timeline for 2022 redistricting (id. § 4
[b]), and requiring that incarcerated people be re-numerated to their last place of residence
for the purpose of drawing redistricting lines (id. § 4 [c] [6]). On one hand, the petitioners
argue that the voters’ rejection of the amendment shows that the voters would also have
disapproved of the statute, and that both the failed amendment and statute were part of a
plan by Democrats to bypass the IRC. On the other hand, as with the statute, it is perfectly
feasible that Democrats worried that the IRC process would break down and wanted to
clarify what should occur in that instance for the sake of election efficiency and integrity.
Taking all of this together, and taking every inference in favor of petitioners, one
could colorably believe that the Legislature was attempting to position itself to be able to
draw legislative districts unfettered by the IRC if the IRC deadlocked. As the Appellate
- 27 -
- 28 - No. 60
Division concluded, however, that evidence, standing alone, does not prove intent to
gerrymander beyond a reasonable doubt (2022 NY Slip Op 02648, *3).
IV
I agree with the principles underlying the majority’s opinion. Election districts
should not be created for the purpose of disadvantaging political opponents. Nor should
they be created to disadvantage racial or ethnic minorities, or constructed in ways that
minimize the responsiveness of elected officials to their constituents by, for example,
splitting cities or communities of interest apart. I also do not rule out that, with a sound
analysis, these plaintiffs or others could prove that the 2022 legislative plan violated the
Constitution, at least in some districts. My disagreements are threefold:
• I read the constitutional provision as Judge Rivera does—leaving the redistricting
authority ultimately in the hands of the Legislature;
• I am convinced these petitioners have not adduced legally sufficient evidence to
demonstrate gerrymandering; and
• given my first two disagreements, I believe the majority’s remedy inappropriately
strips from the Legislature the right clearly provided in Article III, Section 5: “In
any judicial proceeding relating to redistricting . . . [i]n the event that a court finds
such a violation, the legislature shall have a full and reasonable opportunity to
correct the law’s legal infirmities.” This case is such a proceeding. As the majority
- 28 -
- 29 - No. 60
says, “[t]he Constitution is the voice of the people speaking in their sovereign
capacity, and it must be heeded” (majority op at 32, quoting Matter of New York El.
R.R. Co., 70 NY 327, 342 [1877]). Why, then, does the majority not heed the
Constitution’s command that the Legislature must be given a “full and fair
opportunity” to address the legal infirmities identified in this judicial proceeding?
- 29 -
RIVERA, J. (dissenting):
I would reverse the Appellate Division judgment because petitioners failed to
establish that the legislature violated the state’s redistricting procedures or constitutional
mandates. The legislature acted within its authority by adopting the redistricting legislation
-1-
-2- No. 60
challenged here after the Independent Redistricting Commission (IRC) chose not to submit
a redistricting plan by the second constitutional deadline. Thus, there is no procedural error
rendering the redistricting legislation void ab initio. Petitioners’ claim of a substantive
violation based on gerrymandering is also without merit as their evidence fell far short of
proving that the legislature’s congressional map was unconstitutional beyond a reasonable
doubt.
I.
In interpreting a constitutional provision, the primary role of this Court is to give
effect to its unambiguous text and the intent of the People in adopting the provision (see
White v Cuomo, — NY3d —, —, 2022 NY Slip Op 01954, *5 [2022]). This appeal requires
that we interpret Article III, §§ 4 and 5 of the New York Constitution. Under section 4, the
IRC shall prepare decennially a redistricting plan to establish State Assembly and Senate
and federal congressional districts and submit such plan and implementing legislation to
the legislature for its consideration, without amendment (see NY Const, art III, § 4 [b]). If
the legislature fails to approve the proposed legislation, the IRC shall prepare and submit
a second redistricting plan and necessary implementing legislation for consideration (see
id.). If the legislature fails to approve the second plan, the legislature shall approve its own
implementing legislation (see id.). Section 4 (e) acknowledges that the redistricting
procedure may not be followed where “a court is required to order the adoption of, or
changes to, a redistricting plan as a remedy for a violation of law.” Section 5 further
provides that upon a judicial finding that a redistricting law violates Article III, such law
shall be “invalid in whole or in part,” and that “the legislature shall have a full and
-2-
-3- No. 60
reasonable opportunity to correct the law’s legal infirmities.” Here, the IRC initially
submitted two redistricting plans by the first deadline. The legislature failed to approve
either. When the IRC chose not to make another submission by the second deadline, the
legislature drafted and approved redistricting implementing legislation which the Governor
signed.1
Petitioners, residents of several New York districts, claim that the legislature
avoided the exclusive redistricting process set forth in sections 4 and 5 by enacting
redistricting legislation in the absence of an IRC submission by the second deadline,
because a second IRC submission is a constitutional requirement that triggers the
legislature’s authority to act. Petitioners further claim that the redistricting legislation is the
product of intentional gerrymandering by the democratic members of the State legislature,
in violation of section 4 (c) (5) of article III of the Constitution. As I discuss, petitioners
are wrong as a matter of law on their procedural challenge and have failed to prove their
gerrymandering allegation.
II.
1
Contrary to the majority’s view, the IRC was not required to submit a different set of
second plans. Indeed, the lead Republican IRC Commissioner noted that the Republican
members of the IRC had considered agreeing to submit the same plans during the second
round, but he concluded that “he would prefer for the Legislature to begin its process then
postpone it one week with presumably voting down maps that he claims have not changed”
(Joshua Solomon, Independent Redistricting Commission Comes to a Likely Final Impasse,
Times Union [Jan. 24, 2022], https://www.timesunion.com/state/article/Independent-
Redistricting-Commission-comes-to-a-16800357.php).
-3-
-4- No. 60
There is no procedural error of constitutional magnitude warranting invalidation of
the legislature’s redistricting implementing legislation. That conclusion is supported by
either of two analytic paths.
A.
By one view, the process followed by the legislature here does not violate the text
or purpose of article III because the IRC in fact submitted two plans, albeit all at once, in
furtherance of the purpose of section 4, and, in any case, the legislature is not bound to
approve an IRC plan as drafted.2 Under that view, the legislature acted appropriately on
the unique facts of this case. First, the Constitution does not mandate legislative adoption
of any IRC-proposed implementing legislation; the legislature may opt to reject the IRC
submissions and proceed to draft implementing legislation, which would then be submitted
to the Governor for action (see NY Const, art III, § 4 [b]).3 That is exactly what happened
here. Second, the Constitution requires that in the event that more than one draft plan
receives an equal number of IRC member votes for approval, above the votes garnered for
any other plan, the IRC must submit all of those plans to the legislature in accordance with
section 4 (b) of article III of the Constitution (see id. § 5-b [g]). Thus, if the IRC fails to
2
The majority incorrectly asserts that the legislature’s alleged violation of the
constitutional procedure is undisputed (see majority op at 2). In fact, respondents have
maintained that the IRC, not the legislature, is at fault here.
3
Several of the states cited by the majority (see majority op at 4 n 2) have adopted
redistricting commissions which are not subject to legislative approval (see e.g. Cal Const,
art XXI, § 2; Colo Const, art V, § 48; Mich Const, art 4, § 6; see generally Loyola Law
School, All About Redistricting: National Summary, https://redistricting.lls.edu/national-
overview/?colorby=Institution&level=Congress&cycle=2020 [last visited Apr. 27, 2022]).
-4-
-5- No. 60
garner a majority vote, the IRC is empowered to submit more than one redistricting plan
and implementing legislation for the legislature’s consideration. That is also what
happened here. Third, nothing in the Constitution expressly prohibits the legislature from
acting if the IRC chooses not to submit yet another plan after the legislature has considered
and failed to approve all the plans with the highest number of IRC votes. The Constitution
is simply silent on how to address the IRC’s choice to forego submission of a redistricting
plan and implementing legislation before the second deadline. Nor does the constitutional
framework command that the legislature remain idle in the face of an IRC decision not to
submit a plan despite section 4 (b)’s mandatory language setting forth deadlines for
submission. The Constitution requires the legislature approve redistricting legislation,
upon consideration of one IRC plan and, if necessary, a second plan. The legislature did
exactly that, reviewing two IRC plans and determining not to approve either, but instead
adopting legislation which it maintains wholly comports with the Constitution.4 The
majority’s decision leaves the legislature hostage to the IRC, and thus incentivizes political
gamesmanship by the IRC members—the exact scenario the majority claims it avoids by
interpreting the second IRC submission as a mandatory predicate to legislative action (see
majority op at 20).
4
The majority, in claiming that my view ignores the constitutional text and purpose (see
majority op at 16 n 8), ignores that under the unique facts here, we must harmonize the
constitutional process with the overriding intent of the amendment—to create a process for
public, bipartisan input in redistricting to provide the legislature with background data and
options for redistricting. The majority view rests on a distinction without a difference; had
the IRC merely submitted the competing plans in succession, and if the legislature had not
approved either, the majority would conclude, as I do, that there was no procedural error.
-5-
-6- No. 60
The majority claims that upholding the legislative action here would undermine the
redistricting process adopted by the 2014 constitutional amendment and thwart the purpose
of the amendment (see id. at 23). That is only true if we ignore the salutary aspects of the
entire redistricting process and how it informs the legislature’s decisions. Under the
Constitution, the IRC is tasked with drafting proposed districts that are contiguous,
compact, and equipopulous, while considering the maintenance of cores of existing
districts and political subdivisions, and avoiding line-drawing that denies or abridges the
rights of communities of interest, including racial and minority language groups, or the
formation of districts that favor or disfavor political candidates or parties (see NY Const,
art III, § 4 [c]). The goal of fair, non-gerrymandered line drawing is furthered, in part, by
a robust public hearing and comment process that allows the IRC to consider diverse
viewpoints when preparing its redistricting plan (see id.). In turn, the legislature benefits
from this same process when it considers the IRC’s draft plan. Here, in accordance with
the Constitution, the legislature considered both of the plans submitted by the IRC, fully
aware of the public process that preceded the approval of both plans by a concededly split
IRC membership. Unfortunately, like the IRC, the legislature could not agree on only one
of those plans. When the IRC chose not to make a submission by the second deadline—of
a plan that would be subject to legislative amendment, unlike the two plans submitted by
the first deadline—nothing in the Constitution prohibited the legislature from drafting and
approving redistricting legislation that it determined was in compliance with the
constitutional mandates set forth in article III.
-6-
-7- No. 60
The majority also concludes that the legislature may only may “amend[]”
redistricting plans submitted by the IRC (see majority op at 14, quoting NY Const, art III,
§ 4 [b]). The extent of the legislature’s authority to redraw the IRC’s proposed maps,
however, is not before us since that did not occur here. Moreover, the majority’s
interpretation ignores that legislative plans may include “any amendments” that are
“deem[ed] necessary” (NY Const, art III, § 4 [b]), giving the legislature significant
discretion to reject the IRC’s proposals. Likewise, the two percent rule—which the
majority seems to interpret as a constitutional requirement (see majority op at 21 n 11)—
is also not properly before us, and in any case, that statutory rule applies only when the
IRC submits a plan by the second deadline, which concededly it did not do. In sum, the
majority is incorrect that the legislature’s authority to approve redistricting legislation is
subject to the two percent rule after it decides not to approve the first IRC plan as drafted
because that legislative authority can only be triggered after the IRC submits a plan
pursuant to the second deadline.
Even assuming the majority is correct that the Constitution provides the legislature
with express and exclusive choices—either approve, as drafted, the IRC implementing
legislation submitted by the first or the second constitutional deadlines, or don’t approve
either and amend and approve bicamerally the second submission which is then presented
to the governor for action—the majority correctly concedes that the legislature is not
required to adopt, without change, the IRC recommendations (see majority op at 13-14).
Instead, the legislature must exercise its constitutional duty to ensure that New York’s
-7-
-8- No. 60
district lines comply with the constitutional factors set forth in Article III and do not
otherwise violate federal or state law (see NY Const, art III, § 4 [c]; Voting Rights Act of
1965, 52 USC § 10101 et seq., as added by Pub L 89-110, 79 US Stat 437). As this Court
has made clear, redistricting is a complex and intricate task, involving a “[b]alancing” of
“myriad requirements imposed by both the State and the Federal Constitution,” which is
ultimately “entrusted to the legislature” (Matter of Wolpoff v Cuomo, 80 NY2d 70, 79
[1992]; see Matter of Schneider v Rockefeller, 31 NY2d 420, 431 [1972] [“The
gerrymandering is . . . rather deep in the ‘political thicket’”]). Thus, and contrary to the
majority’s conclusion (see majority op at 18-19), the legislature was not required to ignore
its constitutional duty because the IRC “abandon[ed] its constitutional mandate” (id. at 23).
And, despite the majority rhetoric about redistricting reform—that the IRC process was
designed to “incentiviz[e] the legislature to encourage and support fair bipartisan
participation and compromise throughout the redistricting process” (id. at 20)—it is the
majority’s interpretation of the Constitution that effectively places the redistricting process
at the mercy of the IRC, which cannot be what the People of the State of New York
intended when they approved the amendment and even though the Constitution does not
mandate legislative approval of any IRC plan. Indeed, recognition that the legislature
retains the ultimate authority to enact a redistricting plan does not, as the majority posits,
“render the 2014 amendments . . . functionally meaningless” (id. at 11); it merely confirms
that the legislature must step in when the IRC fails in its task.
B.
-8-
-9- No. 60
Even if the plain text of the Constitution did not support the legislative action taken
here, there is an alternative analytic basis for rejecting the petitioners’ procedural argument.
The constitution is silent as to how to respond when the IRC does not submit a plan in
accordance with Article III, as in this case where the IRC chooses not to make a second
deadline submission. Notably, petitioners did not sue the IRC to secure compliance with
what they and the majority maintain is the “exclusive method of redistricting” (majority op
at 20). Nor have petitioners requested the courts to adopt either of the IRC plans even
though petitioners, like the majority, claim that the IRC’s submissions are a constitutional
predicate to legislative action (see id. at 21).
However, the legislature anticipated just such a failure in the IRC process by
passage of an amendment to the Redistricting Reform Act of 2012 (L 2012, ch 17), which
provides that “if the commission does not vote on any redistricting plan or plans, for any
reason, by the date required for submission of such plan and the commission submitted to
the legislature . . . all plans in its possession, both completed and in draft form, and the data
upon which such plans are based, each house shall introduce such implementing legislation
with any amendments each house deems necessary”(see Redistricting Reform Act § 3 [c],
-9-
- 10 - No. 60
as amended by L 2021, ch 633, § 1).5 That statute, having been properly enacted, controls
and provided the legislature with the authority to act as it did here.6
III.
Turning to petitioners second claim, that the legislative plan is an unlawful
gerrymander, we review this challenge, like other constitutional attacks on redistricting
plans, de novo and not, as the majority suggests, under a deferential standard of review
(see Matter of Wolpoff, 80 NY2d at 78 [“(W)e examine the balance struck by the
(l)egislature in its effort to harmonize competing Federal and State requirements”]; Matter
of Schneider, 31 NY2d at 427 [“Our duty is . . . to determine whether the legislative plan
substantially complies with the Federal and State Constitutions”]). Thus, petitioners are
held to the highest burden in our law—one generally enshrined in criminal law—proof
beyond a reasonable doubt:
“A strong presumption of constitutionality attaches to the
redistricting plan and we will upset the balance struck by the
Legislature and declare the plan unconstitutional ‘only when it
can be shown beyond reasonable doubt that it conflicts with the
fundamental law, and that until every reasonable mode of
reconciliation of the statute with the Constitution has been
5
The majority’s discussion of the legislative history of the 2014 amendment is incomplete
(see majority op at 18-20). Several legislators and commentators recognized, prior to
adoption, that—contrary to the views of its sponsors—the amendment did not guarantee
that the IRC would follow the constitutional process (see e.g. NY Senate Debate on
Assembly Bill A2086, Jan. 23, 2013 at 252 [warning that an evenly-divided IRC might
“foster gridlock”]).
6
The statute’s two percent rule would also control. If failure to comply with that rule were
the sole alleged problem with the legislature’s redistricting plan, the courts could mandate
compliance as a targeted and narrow remedy rather than reject the entire redistricting plan
as the majority does, thus creating confusion for candidates and their supporters, and
necessitating the adoption of new deadlines (see majority op at 29-30; Troutman dissenting
op at 4).
- 10 -
- 11 - No. 60
resorted to, and reconciliation has been found impossible’”
(Matter of Wolpoff, 80 NY2d at 78, quoting Matter of Fay, 291
NY 198, 207 [1943]; accord Cohen v Cuomo, 19 NY3d 196,
201-202 [2012]).
Upon review of the record before us, I conclude that petitioners failed to meet their
heavy burden. As three justices concluded below, and as Judge Wilson explains, other than
the petitioners’ expert analysis alleging gerrymandering, the petitioners’ other evidence
cannot satisfy their burden of proof (see Matter of Harkenrider, — AD3d at —, 2022 NY
Slip Op 02648, *4 [plurality]; Wilson dissenting op at 25-28).7 I have already discussed
why there was no constitutional procedural violation, but even if there had been, the
legislature’s approval of a redistricting plan in the absence of a second IRC submission
does not establish intentional gerrymandering. This case does not rest on “the credibility
issue routinely seen in battle-of-the-experts cases,” but rather turns on petitioners’ expert
evidence and its “probative force . . . regardless of respondents’ opposition” (id. at —, 2022
NY Slip Op 02648, *8 [Whalen, P.J., and Winslow, J., dissenting in part]). For reasons
discussed at length in Judge Wilson’s thorough and compelling analysis of petitioner’s
evidence and gerrymandering claim, which I fully join, petitioners failed to carry their
burden. In sum, petitioners relied on an expert who failed to account for several
7
With respect to one of those alleged grounds, the majority is incorrect to the extent that it
suggests that the legislature did not consider Republican views (see majority op at 6 n 3).
As Judge Troutman and Judge Wilson explain in their dissents, the legislature enacted a
plan that includes similar Upstate boundaries as the two IRC plans actually submitted to
the legislature (see Troutman dissenting op at 3; Wilson dissenting op at 12-14). As for the
other ground—that the legislature’s redistricting differs from the 2012 district lines—the
purpose of redistricting is to address demographic changes and so it is no surprise that
population shifts in New York State would result in a different redistricting map in
accordance with constitutional requirements (see Wilson dissenting op at 21-22).
- 11 -
- 12 - No. 60
constitutional requirements and who used an untested, unverified algorithm (see Wilson
dissenting op at 5-6; cf. People v Wakefield, — NY3d —, —, 2022 NY Slip Op 02771,
*15-19 [2022, Rivera, J., concurring in result]). No district line drawer could do so and still
comply with the Constitution.
I dissent.
Order modified, with costs to petitioners, in accordance with the opinion herein and, as so
modified, affirmed. Opinion by Chief Judge DiFiore. Judges Garcia, Singas and Cannataro
concur. Judge Troutman dissents in part in an opinion, in which Judge Wilson concurs in
part in a dissenting opinion, in which Judge Rivera concurs in part. Judge Rivera dissents
in a separate dissenting opinion, in which Judge Wilson concurs.
Decided April 27, 2022
- 12 -