Skip Navigation

Justin Levitt Before the New York State Bar Association Committee on Attorneys in Public Service

Justin Levitt speaks to the NYSBA on New York’s redistricting process.

Published: February 12, 2010


Justin Levitt, Counsel

Brennan Center for Justice at NYU School of Law


NYSBA Committee on Attorneys in Public Service

The State Legislature and the State Constitution

Focus on the Legislature: Reform and Renewal


January 26, 2010


Members of Congress, state legislators, and many local public officials are elected from districts.  The way that district lines are drawn divides the electorate into groups that determine the character of representation: with the stroke of a pen, those who draw the lines can increase or decrease the likelihood of electing any given representative, or electing a representative responsive to any given community.  Together, the district lines can further shape the composition of the legislative delegation as a whole.

            At least once per decade, following the Census, most of these district lines are redrawn in order to ensure that each district’s population is roughly equivalent, living up to the constitutional command of “one person, one vote.”  See Reynolds v. Sims, 377 U.S. 533, 583–84 (1964).

            There are relatively few federal requirements governing the conduct of redistricting.  As a result, each state’s redistricting system is different, in both design and implementation, with substantial variation around the nation. 

Who draws the lines

In New York, the state legislature has primary control of the redistricting process, both for its own districts and for those of Congress. 

Since at least the 1980 round of redistricting, the legislature has delegated initial responsibility for drafting advisory maps to the Legislative Task Force on Demographic Research and Reapportionment, known as LATFOR, see N.Y. Legis. Law § 83-m; the formal authorization for LATFOR expires on June 30, 2010, but it will likely be reauthorized.  1994 N.Y. Laws ch. 141, § 13.  The Task Force consists of six members: one legislator and one non-legislator appointed by the Temporary President of the Senate, one legislator and one non-legislator appointed by the Speaker of the Assembly, and one legislator appointed by each of the Senate and Assembly minority leaders.   N.Y. Legis. Law § 83-m. 

In 2001, during the last round of redistricting, the Task Force held hearings at various locations around the state to receive input on district bounds.  See, e.g.,  Legislative Redistricting Hearings, at  In practice, many observers note that as with many other legislative processes in New York, the LATFOR tends to implement the will of the legislative leadership.

After give and take with LATFOR, the legislature passes final redistricting legislation as a standard statute, by majority vote and subject to gubernatorial veto.  See, e.g., 2002 N.Y. Laws ch. 35.  The legislature is not bound by LATFOR’s recommendations, but because LATFOR is heavily influenced by the preferences of the legislative leadership, the final redistricting legislation often mirrors LATFOR recommendations closely.   The map for state Assembly districts and state Senate districts must be passed in one single bill.  N.Y. Const. art. III, § 5; Orans v. Rockefeller, 257 N.Y.S.2d 839 (N.Y. Sup. Ct. Special Term 1965).  Historically, this has amounted to something of a tacit agreement between the chambers, in which each chamber determines the lines for its own members independently.

Two other entities are important actors in New York’s redistricting process.  First, the judiciary: any citizen may petition the courts for review of a redistricting plan, N.Y. Const. art. III, § 5, and if redistricting is not completed promptly, or if it is completed in a manner deemed unfair under the criteria below, the map will likely be challenged in court.  Although both federal and state courts prefer to leave the map-drawing process to the state legislature, see, e.g., Wise v. Lipscomb, 437 U.S. 535, 540 (1978) (White, J., joined by Stewart, J.); Bone Shirt v. Hazeltine, 700 N.W.2d 746 (S.D. 2005); In re Below, 151 N.H. 135 (2004), as the time remaining before the proximate election grows shorter, the courts will be increasingly inclined to draw redistricting plans of their own.

The second additional entity to consider is the Department of Justice: Kings County, New York County, and the Bronx are subject to a requirement under section 5 of the Voting Rights Act to have districts, like all other voting regulations, “precleared” by the U.S. District Court for the District of Columbia or the U.S. Department of Justice.  As a practical matter, redistricting plans are generally submitted to the Justice Department, to ensure that the plans do not amount to retrogression in the ability of protected racial or language minorities to elect candidates of their choice.  If the Justice Department does find retrogression — or, more accurately, finds that New York has not proved a lack of retrogression — it will object to the district plan, and the challenged redistricting map will have no legal effect.

Where to draw the lines: federal law

            As described above, the legislature has the primary responsibility for redistricting in New York, with much of that power exercised at the behest of the leadership.  But two critical federal principles constrain the legislature’s discretion in determining where to draw the lines.

Equal population

The first federal rule is a constitutional equal population requirement.  In the 18th century, many districts were defined by the borders of towns or counties, or groups thereof; representation was apportioned by periodically assigning a number of representatives to each district, but the districts themselves remained largely fixed.  See Micah Altman, Traditional Districting Principles: Judicial Myths vs. Reality, 22 Soc. Sci. Hist. 159, 180–81 (1998).  Under this system, for example, New York State’s 1777 Constitution assigned nine representatives to New York “city and county,” ten to Albany “city and county,” four to Queens County, two to Kings County, and so on.  N.Y. Const. of 1777, art. IV.

As the population grew, however, it grew unevenly, and few legislatures rigorously kept up with reapportionment — sometimes the neglect was intentional, sometimes not.  The uneven growth meant that some districts became far more populous than others, with the vote of each citizen in the smallest districts worth many times the vote of each citizen in the largest districts.

In the series of “one person, one vote” cases beginning with Baker v. Carr in 1962, the Supreme Court determined  that this sort of population disparity violated the federal constitution.  See Baker v. Carr, 369 U.S. 186 (1962) (population disparity is justiciable); Wesberry v. Sanders, 376 U.S. 1 (1964) (Congressional districts); Reynolds v. Sims, 377 U.S. 533 (1964) (state legislative districts); Avery v. Midland County, Tex., 390 U.S. 474 (1968) (local government districts). 

These cases established that population should be approximately equal for each state and federal district within a state, and each local district within its corresponding jurisdiction.  The standard for congressional districts is set by the Supreme Court’s interpretation of Article I, § 2, and section 2 of the 14th Amendment, both of which directly concern federal representation.   The bar for congressional districts is quite high, with equal population required “as nearly as is practicable.”  Wesberry, 376 U.S. at 7–8.  In practice, this means that states must make a good-faith effort to achieve absolute mathematical equality for each district within the state, with the district size pegged to the mathematical average, or “ideal,” population.  Any differences must be specifically justified by a consistent state policy that requires the deviation.  Karcher v. Daggett, 462 U.S. 725, 730 (1983).  And even consistent state policies that cause a deviation approaching one percent among congressional districts will likely be deemed unconstitutional.  See J. Gerald Hebert et al., The Realists’ Guide to Redistricting 6–8 (2000).

The standard for state and local legislative districts follows the 14th Amendment’s Equal Protection clause, and permits a bit more flexibility.  The population in these districts must be “substantially” equal.  Reynolds, 377 U.S. at 579.  The courts have not defined the phrase precisely, but over a series of cases, it has become accepted that the difference in population between the largest and smallest state legislative districts, the “total deviation,” becomes constitutionally suspect when it exceeds ten percent of the ideal population.  See Brown v. Thomson, 462 U.S. 835, 842–43 (1983).  This is not an absolutely hard line: in some cases, a state may have a compelling reason for drawing districts with a population disparity greater than ten percent, see Mahan v. Howell, 410 U.S. 315 (1973) (justifying a greater population disparity to preserve existing political boundaries); Voinovich v. Quilter, 507 U.S. 146 (1993) (same); and in some cases, an asserted state rationale may be insufficient to justify population disparities less than ten percent.  See Cox v. Larios, 542 U.S. 947 (2004).


The other major federal regulation of the redistricting process concerns race.  Many redistricting techniques have been abused in order to dilute racial minorities’ electoral strength.  One such ploy is called “cracking”: splintering minority populations into small pieces of multiple districts, so that their voting power is diluted.  Another tactic is called “packing”: consolidating as many minority voters as possible into a few concentrated districts, leaching the population’s voting power from the surrounding areas.  Others abound.

The federal Voting Rights Act of 1965 was designed to combat practices used to deny minorities the right to an effective vote, including redistricting techniques like those above.  Two sections of the Act are particularly important to New York redistricting: section 2 and section 5.

Section 2 of the Voting Rights Act blocks district lines that deny minority voters an equal opportunity “to participate in the political process and to elect representatives of their choice.”  42 U.S.C. § 1973(b).  It applies whether the denial is intentional, or an unintended end result.  Courts applying the Act in the redistricting context essentially test whether the way that a district is drawn takes decisive political power away from a cohesive minority bloc that has otherwise suffered discrimination in the region.

The predominant test for section 2’s coverage was articulated in Thornburg v. Gingles, 478 U.S. 30 (1986), and the basic components of the test are known as Gingles threshold conditions — requirements for a judicial finding that a jurisdiction has violated section 2.  The first Gingles condition asks whether it is possible to draw a district such that a majority of voters belong to a geographically “compact” racial or language minority community.[1]  See, e.g., Bartlett v. Strickland, 129 S. Ct. 1231, 1241, 1246.  Compactness, discussed further below, has never been precisely defined in this context; it generally refers to a populations that are not particularly “far-flung” from each other, and where the boundaries are moderately regular, without extensive tendrils and the like.  See, e.g., League of United Latin American Citizens (“LULAC”) v. Perry, 548 U.S. 399, 433 (2006).   Overall, the first Gingles condition measures approximately whether a sufficiently sizable minority population lives sufficiently close together to be able to control a district.[2]

The second Gingles condition tests whether the minority population in question votes in cohesive fashion: whether they usually vote for the same candidate.  The third Gingles condition tests the potential competition: whether the majority population in the area votes in cohesive fashion for a different set of candidates, such that the majority population would regularly be able to defeat the preferred candidate of the minority if the minority community’s voting power were not specifically protected.  Together, these conditions are known generally as “racial polarization.”

If demographic and political factors show the presence of all three Gingles threshold conditions, courts then look to the “totality of the circumstances” to determine whether the minority vote has been diluted.  There are many factors involved in the “totality of the circumstances,” as might be expected, but the factor that seems to have the most weight is rough proportionality: whether minorities have the opportunity to elect representatives of their choice in a number of districts roughly proportional to the percentage of minority voters within the population as a whole.  See, e.g., LULAC, 548 U.S. at 436–38; Johnson v. De Grandy, 512 U.S. 997 (1994).  Section 2 does not, it is clear, guarantee proportionality.  But if a minority group with twenty percent of a state’s citizen voting-age population already has the opportunity to elect representatives in twenty percent of the state’s districts, courts will be more hesitant to find a violation of section 2 even if the three Gingles conditions are met; and if the minority group does not have such an opportunity, courts will be more prone to find a violation.

Section 5 of the Voting Rights Act, the other federal provision most applicable to redistricting, works a bit differently.  First, unlike section 2, it is temporary: it was last renewed in 2006, and is scheduled to come up for renewal again in the future.  Fannie Lou Hamer, Rosa Parks, and Coretta Scott King Voting Rights Act Reauthorization and Amendments Act of 2006, Pub. L. No. 109–246, 120 Stat. 577.  Second, while section 2 applies all over the country, section 5 only applies in jurisdictions in which fewer than half of the eligible voters either registered or voted in 1964, 1968, or 1972.  Much of the deep South is made up of “covered jurisdictions” under section 5, but parts of other states outside of the South — including Kings County, New York County, and the Bronx — are also covered.[3]

In a jurisdiction covered under section 5, which will include any New York statewide redistricting map, the government may not implement any change to a voting procedure without first submitting the change to the Department of Justice or the U.S. District Court for the District of Columbia, in a procedure known as “preclearance.”  New district lines will be precleared if the plan is not intended to dilute minority votes, and if it does not result in “retrogression” in minority political opportunity.  A new plan has caused retrogression if it presents a diminished opportunity for minorities to elect their candidates of choice, as compared to the former redistricting map.  See 42 U.S.C. § 1973c.  In order to assess retrogression, it is necessary to assess minority political opportunity given the most recent demographic information available, under both the existing redistricting map (the “baseline”) and the proposed revision.

            Thus, section 2 of the Voting Rights Act protects many populations of minority voters large enough to form a district’s majority, and section 5 prevents backsliding to take away opportunities that minority voters had before.  Beyond those two limitations, the legislature may include race in the factors considered when determining where to draw district lines — but without a reason sufficiently compelling to withstand strict scrutiny, race may not be the “predominant” reason for a district’s shape.  See Bush v. Vera, 517 U.S. 952, 958–59, 962–64 (1996) (plurality opinion).  Courts will generally assess the impact of race by first looking at the degree of irregularity in a district’s shape, and then attempting to determine whether race, or other factors, best explain the irregularities.  The more a district’s overall shape can be traced to other goals like those reviewed below, the less likely it is that voters’ race will be found to be the impermissibly driving force behind that district’s lines.


Where to draw the lines: state law


Even after accounting for the requirements of the Voting Rights Act, there remain countless ways to divide New York into districts of roughly equal population.  For Congressional lines, the remaining choices are completely up to the legislature: other than the requirement that each federal district be drawn for a lone Representative, see 2 U.S.C. § 2c, there are no additional limitations under federal or state law on how the lines are to be drawn. 

For the state legislature, on the other hand, the state constitution provides a few additional constraints governing the legislature’s discretion.




            New York’s constitutional provisions on redistricting require that state legislative districts be composed of “contiguous territory.”  N.Y. Const. art. III, §§ 4–5.  A contiguous district is one in which it is possible to travel from any point in the district to any other point in the district without crossing the district boundary: a district, that is, in which the portions of the district are “touching, adjoining and connected.”  Matter of Sherrill v. O’Brien, 188 N.Y. 185, 207 (1907).   Most observers understand contiguity to require portions of a district to be connected by more than a single point, but there is no further general agreement that a district be connected by territory of a certain area. 

Few concepts in redistricting are absolute, and contiguity is no exception.           Water, for example, creates a special circumstance for contiguity.  In most cases, districts divided by a waterway are contiguous if a common means of transport connects the two sides of the district: a bridge, for example, or a ferry route.  Several New York districts take advantage of waterways and islands to connect two portions of a mainland district that would otherwise be noncontiguous. 



The New York Constitution also requires that state legislative districts be “in as compact form as practicable,” although the term “compactness” is not further defined.  N.Y. Const. art. III, §§ 4–5.  Most courts and commentators understand  compactness to refer to a district’s geometric shape: generally, a district in which constituents generally live near each other is usually considered more compact than one in which they do not, and a district with a relatively geometrically regular convex shape is usually considered more compact than one with multiple extended tendrils.   The Supreme Court, however, has construed compactness, at least in some circumstances, to indicate that residents have something in common: some sort of cultural cohesion, or lack of severe cultural dispersion, among the citizens in a district.  See, e.g., LULAC, 548 U.S. at 432–33 (addressing the difference between compactness for Equal Protection purposes and compactness for Voting Rights Act purposes). 

More than thirty different measures of compactness have been proposed, but the concepts can generally be sorted into three categories.  See Richard H. Pildes & Richard G. Niemi, Expressive Harms, “Bizarre Districts,” and Voting Rights: Evaluating Election-District Appearances after Shaw v. Reno, 92 Mich. L. Rev. 483, 553 (1993).  The first set elevates the importance of contorted boundaries: a district with smoother boundaries will be more compact, and one with more squiggly boundaries will be less compact.  The second set focuses on the district’s dispersion, or the degree to which it spreads from a central core: a district with few pieces sticking out from the center will be more compact, a district with pieces sticking out farther from the district’s center will be less compact.  The third set of measures elevates the importance of citizen housing patterns in relation to the district’s boundaries, generally granting more flexibility for district tendrils in more sparsely populated areas. 

New York courts have generally shunned any of these more specific measures, focusing more on the gestalt.  See Bay Ridge Community Council, Inc. v. Carey, 479 N.Y.S.2d 746, 748 (N.Y. App. Div. 1984).   Absent districts reflecting a “complete departure” from any understanding of compactness, Schneider v. Rockefeller, 31 N.Y.2d 420, 430 (1972), the courts seem likely to give great deference to the legislature’s choices.

Political boundaries


            The remaining requirement in New York is the textual command to heed county and municipal boundaries.  As mentioned above, New York was once apportioned entirely based on representation per town or county, and the state constitution continues to reflect that history.  N.Y. Const. art. III, §§ 4–5.  However, the federal “one person, one vote” cases overrode this political boundary mandate: as the New York Court of Appeals declared in 1965, “it is patent that under the new rules the integrity of all the counties in these respects cannot be complete.”  In re Orans, 15 N.Y.2d 339, 351 (1965). The court continued, finding that “the historic and traditional significance of counties in the districting process should be continued where and as far as possible,” id. at 352, but that “historic and traditional significance” has seldom taken center stage.  As with compactness, the courts have showed extremely broad deference to the legislature.  They have been particularly forgiving of legislative plans that violate county boundaries in the name of limiting population deviation, even below the 10% deviation threshold generally permitted by the federal constitution.  See, e.g., Wolpoff v. Cuomo, 80 N.Y.2d 70, 79 (1992); Schneider v. Rockefeller, 31 N.Y.2d 420, 428 (1972).

Reform on the horizon

Several legislators, advocates, and observers of New York’s redistricting process have noted room for reform, from state constitutional amendments to amendments to the procedures used by LATFOR and the legislature in passing redistricting legislation.  Changing the state constitution  is certainly the most durable method, but consequently the most difficult to achieve.  Two successive legislatures must vote for a proposed constitutional amendment, which must then be ratified by popular vote, N.Y. Const. art. XIX, § 1; in the alternative, the legislature may call for a constitutional convention, and if the public approves, elected delegates must have their handiwork again approved by the public before it becomes binding.  N.Y. Const. art. XIX, § 2.

There is some debate about whether a statutory change could effectuate change, particularly if it delegated control of the redistricting process to a body other than the legislature, given the state’s constitutional command that the districts be redrawn “by law.”  If changing the decisionmaker is the goal of statutory reform, the safest legal route would be to delegate primary control of the process, with a requirement that the legislature ratify the result, with or without amendment, or with amendment constrained to a certain quantum of change.  Even though such a statutory change could be repealed by subsequent legislatures, that repeal would still be subject to gubernatorial veto, with focused political consequences for a change of course.

Many of the recent reform proposals have focused on the identity of the decisionmakers, seeking a form of greater independence in those who draw the lines and/or attempting to ensure bipartisanship in the process.  See, e.g., A.5279, 2009–2010 Reg. Sess. (N.Y. 2009); S.6240, 2009–2010 Reg. Sess. (N.Y. 2009); Ass’n of the Bar of the City of N.Y., Comm. on Election Law, A Proposed New York State Constitutional Amendment to Emancipate Redistricting from Partisan Gerrymanders (2007), at  Others have focused to a greater extent on modifying the criteria governing where the lines should be drawn.  See, e.g., A.6721, 2009–2010 Reg. Sess. (N.Y. 2009). 

Still other reform proposals have addressed a particular portion of the redistricting process: the counting of incarcerated populations.  The New York Constitution provides that “For the purpose of voting, no person shall be deemed to have gained or lost a residence . . . while confined in any public prison.”  N.Y. Const. art. II,  § 4.  Yet in calculating the population of legislative districts, the State does precisely the opposite, deeming incarcerated persons to reside in the prison facility in which they are confined.  Districts are thus constructed on the backs of “ghost voters,” packing in prisoners who count toward the district size but who are not permitted to vote, and who are not connected to the other residents of the district.   This inflates the political power of voters in prison districts, and deflates the vote of citizens elsewhere, so that votes in prison districts are worth far more than others.  One bill would correct the skew by counting prisoners for state legislative redistricting purposes based on their residence prior to incarceration.  S.1633, 2009–2010 Reg. Sess. (N.Y. 2009).

Among the calls for reform, the Brennan Center has endorsed four basic principles that we believe should be part of any meaningful redistricting proposal.

First, an independent process.  When legislators are intimately involved in drawing their own district lines, there arises an irresistible temptation to conflate the public interest with personal or partisan gain.  The authority responsible for redistricting in New York State – and just as important, the staff supporting that process – should be meaningfully independent from undue legislative influence: free from obligation, and possibly even free from ex parte contact.  This does not simply mean bipartisanship, though bipartisanship may be desirable as well.  Nor does it mean a process devoid of politics, or one that eliminates entirely politicians’ roles.  The difference is that, in a body with independence, those with a particular incentive to lock out competent challengers are not given unfettered access to the keys.

Second, a diverse representative body.   The need to reconcile the competing and complementary interests involved in the redistricting process means that to gain the confidence of the public, the redistricting body must be meaningfully diverse.   Those responsible for drawing district lines should reflect ample geographic, racial, ethnic, and political diversity, so as to prevent charges of self-dealing similar to those that have found a foothold in the current system, but on a group level rather than an individual level. 

Third, meaningful redistricting criteria.  There are many available guiding principles.  Some present affirmative requirements, such as the mandate to further the representation of discrete communities of interest.  Others are negative injunctions, such as the obligation to avoid drawing lines in order to disadvantage a particular incumbent or challenger.  One stands out as particularly important: given a commitment to the principle of majority rule, it is beneficial to pay some attention to the likely partisan balance of a redistricting map, so that a minority of the state’s population does not reliably and durably control the majority of the legislature.  The need for clear governing criteria should not be confused with a demand that the criteria in question dictate a particular result.  Rather, the criteria should retain enough flexibility to allow trusted decisionmakers — the diverse and independent redistricting body mentioned above — to apply overall state priorities to peculiar local circumstances, sensibly and in the broader public interest.

Fourth, meaningful transparency.  At the moment, most citizens are excluded from the redistricting process, which concerns not merely public policy, but the aggregation of group interests that are the foundation of all policy discussions.   Communities are splintered and electoral fortunes tailored, by and large, without meaningful opportunity for input.  A commitment to basic transparency requires not only public hearings, but the opportunity to submit draft maps, and the opportunity to respond to drafts before they are enacted. 

[1] At the moment, many presume that section 2 of the Voting Rights Act also protects multiracial coalitions voting in concert, where any one racial group alone would not represent a district majority.  See, e.g., Bartlett v. Strickland, 129 S. Ct. 1231, 1242–43 (2009) (reserving the question).

[2] There is some debate about the extent to which section 2 of the federal Voting Rights Act protects minority voters  who are geographically dispersed, beyond the scope of this CLE.  See, e.g., Steven J. Mulroy, The Way Out: A Legal Standard for Imposing Alternative Electoral Systems as Voting Rights Remedies, Harv. C.R.-C.L. L. Rev. 333, 364–79 (1998).  

[3] Coverage is not eternal: in a procedure known as “bailout,” after ten years of steps to improve opportunities for minority voting, a covered jurisdiction can petition the federal trial court in Washington, D.C. to be released from Section 5.  Seventeen jurisdictions have thus far successfully taken advantage of the “bailout” provision.  Northwest Austin Municipal Utility District No. One v. Holder, 129 S. Ct. 2504, 2516 (2009).