ALISON J. NATHAN, District Judge

Anna Do, Dechert, LLP, Los Angeles, CA, Hilary Bonaccorsi, Dechert, LLP, Boston, MA, Katherine Shorey, Dechert, LLP, Jackson Chin, Puerto Rican Legal Defense and Education Fund, Inc, Jose-Luis Perez, Latinojustice Prldef, Richard Flannery, Tharuni Jayaraman, Neil A. Steiner, Dechert, LLP, New York, NY, Ezra D. Rosenberg, John Powers, Pro Hac Vice, Ryan Snow, Lawyers' Committee for Civil Rights Under Law, Washington, DC, for Plaintiff.

Brian Lee Quail, William J. McCann, Jr., New York State Board of Elections, Albany, NY, Nicholas Robert Cartagena, Lonstein Law Office P.C., Ellenville, NY, for Defendants.

FINDINGS OF FACT AND CONCLUSIONS OF LAW

ALISON J. NATHAN, District Judge: Common Cause brings this action against the New York State Board of Elections, its Co-Executive Directors, and its Commissioners seeking declaratory and injunctive relief for alleged violations of the National Voter Registration Act and the fundamental right to vote contained in the Fourteenth Amendment. Common Cause challenges two of New York's election practices. First, the State regularly removes voters from active status and places them on inactive status if it believes that the voter has moved. The names of these inactive voters, however, are not provided to poll workers at polling locations. Second, the State prohibits inactive voters from voting by regular ballots, and instead requires them to vote using affidavit ballots.

In October 2019, the Court conducted a four-day bench trial on these claims. The Court concludes that the former practice violates the Equal Protection Clause because it burdens all New York voters but serves no legitimate state interest. The latter practice does not violate the Constitution, however, because it advances several legitimate interests. The Court also concludes that Common Cause has identified three violations of the National Voter Registration Act. The Court therefore ORDERS the Defendants to provide the names of inactive voters registered to vote in a particular election district to the poll workers of that election district.

This Opinion and Order constitutes the Court's findings of fact and conclusions of law for purposes of Federal Rules of Civil Procedure 52(a)(2) and 65. To the extent any statement labeled as a finding of fact is a conclusion of law it shall be deemed a conclusion of law, and vice versa.

I. BACKGROUND

A. New York's Voter-Registration Process

States regulate both state and federal elections. "The Times, Places and Manner of holding Elections for Senators and Representatives, shall be prescribed in each State by the Legislature thereof." U.S. Const., art. I, § 4. Though states enact vastly different rules governing elections, one thread is common: all states require voters to vote in the election district in which they reside. Our system of representative democracy is predicated on this rule; only those who reside in a geographic area should be able to choose its representatives. Trial Transcript (Tr.) at 311.

The dispute in this case arises out of a persistent problem faced by states in achieving this goal: voters move, and they do so often. "[M]ore than 10% of Americans move every year." Husted v. A. Philip Randolph Inst. , ––– U.S. ––––, 138 S. Ct. 1833, 1838, 201 L.Ed.2d 141 (2018) (citing Census Bureau data). In designing their voter-registration systems, states must account for this voter movement. But when many residents move, they do not update their voter registration. Indeed, "about 2.75 million people are said to be registered to vote in more than one State." Id. If a voter moves and does not update her registration, she could, in theory, continue to vote at her original address even though she no longer lives there. Both federal and state laws thus require states to use proxies to identify voters who may have moved. Once identified, states often change these likely movers' registration status and require them to take additional steps before casting a regular ballot again. The problem, however, is that those proxies can be overinclusive, capturing voters who have not moved. Both the federal government and states are wary of using too strong a medicine—if states design a system that is overly inclusive, they may disenfranchise voters who remain in the same location.

New York has developed its own procedures to regulate this process. The details of the State's approach are laid out in Section II.A. In short, New York performs this voter list maintenance by using several proxies to determine which voters have moved. It then renders those voters inactive and requires them to cast affidavit ballots (what other states call provisional ballots).

B. Procedural History

Common Cause claims that New York's list-maintenance procedures violate both the fundamental right to vote, contained in the Equal Protection Clause, and the National Voter Registration Act (NVRA). On September 6, 2017, Common Cause filed suit against the New York State Board of Elections and various state officials, seeking both declaratory and injunctive relief. Dkt. No. 1. Two months later, New York moved to dismiss. Dkt. No. 36.

On September 30, 2018, the Court granted the motion to dismiss in part and denied it in part. Dkt. No. 58. The Court's Opinion made several key conclusions. First, the Court held that Common Cause had met its burden at the motion-to-dismiss stage to demonstrate standing. Second, the Court narrowed the scope of Common Cause's statutory claim. It dismissed the facial statutory challenge, finding that New York's procedures as written did not violate the NVRA. However, the Court did not dismiss Common Cause's as-applied statutory challenge, finding that even though New York's regime facially complied with the NVRA, the State could still violate the statute by not following its strictures in practice. The Court did not address Common Cause's constitutional claim. The as-applied statutory challenge and constitutional challenge are thus the remaining claims in this case.

The Court held a four-day bench trial on these two claims in October 2019. Following the bench trial, the parties submitted memoranda of law and proposed findings of fact. Dkt. Nos. 179-84.

II. FINDINGS OF FACT

The Court turns first to its findings of fact from the extensive trial record. The key findings of fact are as follows: (1) tens of thousands of New York voters are improperly registered to vote as inactive, even though they continue to reside at their address of registration; (2) this problem results from errors in two of New York's proxies for voter movement, data from the United States Postal Service and a national registry; (3) affidavit balloting causes substantial delay, for inactive voters specifically and all New York voters generally; (4) the State erroneously rejects some affidavit ballots; (5) a supplemental list of inactive voters kept at polling locations would alleviate some of these problems; and (6) Common Cause diverted resources as a result of the New York laws at issue in this case.

A. New York's Election Apparatus

The Court begins with some background regarding how New York administers elections. The New York State Board of Elections implements state-wide regulations and defines the state's election policies. Plaintiff Reply Findings of Fact (FIF), Dkt. No. 183, at 23-24. Defendants Robert Brehm and Todd Valentine are the State Board's Co-Executive Directors. Defendant Thomas Connolly is the State Board's Director of Operations.

The State Board delegates much of its authority to sixty-two County Boards of Election. Each county in New York has its own Board. The State Board supervises and supports these County Boards. Def. Ex. B., Connolly Decl., at 22. For example, the State Board sets the contours for how County Boards should process affidavit ballots, perform list maintenance, and run a poll site. Id. ; Pl. Reply FIF at 24-25 (discussing guidance provided by the State Board to County Boards). And the State Board supports the County Boards by hosting conferences and providing resources. Pl. Reply FIF at 25. But the County Boards administer elections themselves. The County Boards therefore test and deploy voting equipment, organize training for poll workers, conduct early voting, prepare poll books, and manage voting locations on Election Day. Pl. Ex. 263, Ryan Dep., 19:18-20:19, 41:22-45:24. The County Boards employ more than sixty-thousand poll workers per election; New York City alone employs about 30,000 per election. Pl. Reply FIF at 21; Ryan Dep. at 39:17-40-20. The County Boards have substantial discretion in administering elections, and their policies are far from uniform.

Each of New York's counties are further divided into election districts, which are "the basic political subdivision for purposes of registration and voting." N.Y. Elec. Law § 4-100 ; Pl. Ex. 263, Ryan Dep., 22:25-26:2. New York has about 15,000 election districts. Pl. Reply FIF at 21. Election districts have poll sites, and a voter can vote only at the poll site associated with that voter's residence. Id.

B. New York's List-Maintenance System

All states require voters to register in some way. And as discussed above, all states have procedures to deal with potential mismatches between a voter's actual address and the address of registration. New York residents must register to vote at least 25 days before an election. N. Y. Elec. Law § 5-210. Once they register, they are listed as an active voter. Assuming they continue to satisfy various criteria not relevant here, and the State receives no information suggesting that the voter has moved, the voter will remain in active status. That means the voter will continue to be listed in the poll book kept at polling locations. See Pl. Reply FIF at 21 (defining a poll book as "a list of active voters for the election district with signature exemplars which the voter signs to receive a ballot."). The County Boards regularly send voters several types of mail: Once a year, they send cards to all active voters, notifying them of their polling location and hours. N.Y. Elec. Law § 4-117 ; Tr. 147:8-12, 148:18-20 (Valentine); Connolly Decl. ¶ 27. The statute requires these cards to be sent as non-forwardable mail; if they are not delivered, the Postal Service returns the card and provides the County Board with any forwarding information. Id. ; Tr. 147:8-12; 150:2-5 (Valentine); Connolly Decl. ¶ 27. County Boards also notify voters if their assigned polling locations have changed. Tr. 147:13-16; Pl. Reply FIF at 26. And they notify "voter[s] that his or her affidavit ballot was counted." Tr. 147:17-20 (Valentine); Ryan Dep. 145:17-146:18.

Voters are moved to inactive status when a County Board receives information indicating that a voter may no longer be living at her address of registration. New York relies on five proxies, or triggering events, to move voters to inactive status. First, voters are marked inactive if any mail sent by a County Board to a voter is returned as undeliverable. Tr. 147:4-7 (Valentine). Second, a third-party vendor may match a voter-registration record to a record in the National Change of Address registry, thus indicating the voter has moved. Tr. 112:14-17 (Connolly); Pl. FIF at 16; Def. FIF at 13; Pl. Reply FIF at 25-26. Third, another state may inform New York election officials that a voter has moved. Tr. 143:13-144:8. Fourth, one New York county may inform another New York county that a voter has moved. Id. Fifth, certain information from the Department of Motor Vehicles regarding a change in address can also suffice. Tr. 112:18-20.

As soon as any of these triggering events occur, County Boards take two steps. They send the voter a confirmation notice, advising the voter that the Board thinks she has moved and asking her to confirm her current address. N.Y. Elec. Law § 5-712 ; Tr. 112:21-25; Ryan Dep. 213:13-214:8; Def. Ex. 36 (sample confirmation notice from Schenectady County). The confirmation notice is forwardable and serves as a "double check" on whether the voter has moved. Tr. 113:1-14 (Connolly); Tr. 173:3-10 (Valentine); Pl. Reply FIF at 3-5. At the same time, the County Board moves the voter to inactive status. Tr. 113:15-25, 114:1-4; 132:20-22 (Connolly). New York law states that "When a voter is sent a confirmation notice pursuant to the provisions of this article, the voter's name shall be placed in inactive status." N.Y. Elec. Law § 5-213. To be clear, voters are marked inactive before they have a chance to respond to the double check; a voter is marked inactive as soon as the confirmation notice is sent. Id. ; Ryan Dep. 92:10-93:6. Each of the five triggering events—including a County Board receiving a single piece of returned mail from a voter—are thus by themselves sufficient to move a voter to inactive status.

New York law requires the names of inactive voters to be omitted from the poll books that are issued by most County Boards and used by poll workers at poll sites on election days. N.Y. Elec. Law § 5-213(2) ("The registration poll records of all such voters shall be removed from the poll ledgers and maintained at the offices of the board of elections in a file arranged alphabetically by election district."); accord Tr. 132:20-133:1 (Connolly). Their names are instead kept only in the records of their County Board, and cannot appear in poll books, even if the County uses electronic books. N.Y. Elec. Law § 5-213 ("If such board uses computer generated registration lists, the names of such voters shall not be placed on such lists at subsequent elections ... but shall be kept as a computer record at the offices of such board."); Pl. Reply FIF at 10; Brehm Decl. ¶ 3. Moreover, once a voter is placed on the inactive list, County Boards are no longer required to contact them by mail. Tr. 149:1-4, 150:6-13 (Valentine). County Boards therefore stop sending them notices in the mail of election dates and the location of their polling place. Tr. 149:8-13; 150:11-13 (Valentine). Of about twelve-million eligible voters, about one million New Yorkers are on the inactive list. Tr. 122:5-7 (Connolly); 322:23-323:6 (Meredith); Pl. Reply FIF at 30.

Voters can get off inactive status in a few ways. A voter can reply to the confirmation notice, indicating either that she continues to reside at the same address or that she has moved. Pl. Reply FIF at 5. A voter can also cast a valid affidavit ballot. Pl. Reply FIF at 3-4, 19; Ryan Dep. 177:18-178:22. If an inactive voter fails to vote in two successive federal general elections (a period of four years), the state cancels their voter registration. Ryan Dep. 94:3-18.

C. Affidavit Ballots: The Theory

Inactive voters can still vote—but not by regular ballot. Instead, they must cast an affidavit ballot, or what many states call a provisional ballot. This section summarizes how this affidavit-ballot process is supposed to work.

1. How Voters Cast Affidavit Ballots

When a voter appears at a poll site, the poll worker must first ascertain that they are at the correct polling location. Tr. 155:2-4 (Valentine). Poll workers often start by searching the voter's name in the poll book that they have in print. The poll book contains only the names of active voters. Tr. 150:22-151:3 (Valentine) ("the poll book ... is the list of active registered voters."). If the poll worker finds the voter on the list of active voters, that voter can cast a regular, electronic ballot. Pl. Reply FIF at 21-22.

But a voter may not appear on the list of active voters for several reasons: she may not be registered at all, she may be inactive, or the poll worker may be unable to find her name. Ryan Dep. 89:3-14. If a voter is not listed in the active registry, poll workers cannot look up her registration and are "unable to ascertain the voter's actual registration address." Tr. 150:22-151:7 (Valentine). In other words, a poll worker has no way to know why that voter is not listed as active, or even whether the voter is registered to vote at all. Tr. 151:13-18 (Valentine).

When a voter appears whose name does not appear in the poll book, "poll workers are required to ascertain whether the voter is at the correct polling place ... the poll worker [must] consult a map, street finder (listing of street addresses that indicates the election district and polling place for each address range on the street) or other description of all of the polling places and election districts." Connolly Decl. ¶ 53; Pl. Reply FIF at 27-30; see N.Y. Elec. Law § 8-302(e). Poll workers may also contact the County Boards to obtain the voter's correct polling location. Id. ¶ 54. And the poll worker must provide the voter with a notice-to-voter form, "so that they can decide whether to seek a court order or vote by affidavit ballot." Tr. 155:5-10 (Valentine). These options are explained below. The notice must be presented "in a form prescribed by the New York State Board of Elections." Def. FIF ¶ 89. In other words, even if a voter does not appear in the poll book, poll workers are supposed to offer them a choice between a court order and an affidavit ballot. Tr. 155:5-10 (Valentine); Ryan Dep. 48:8-50:13, 162:19-163:16. The state's policy is that no voter should be turned away because she does not appear on the active list; they should be permitted to vote by affidavit or seek a court order. See Ryan Dep. 162:19-163:16; 174:3-14. The notice-to-voter form explains these two options. Pl. Ex. 172 at 84 (sample notice-to-voter form from Columbia County).

If the voter chooses to proceed by court order, the poll worker informs the voter "where and when a Justice of the Supreme Court or a County Court Judge can be located." Id. Election judges do not sit in polling locations; in some counties, they are in the offices of the County Board. Tr. 238:10-15 (Brehm). A court order allows a voter to cast a regular ballot on a voting machine. A court order is therefore the only way for an inactive voter to cast a regular ballot. N.Y. Elec. Law § 8-302(3)(e) ; Tr. 132:20-133:4 (Connolly). However, voters rarely proceed by court order. Tr. 238:22-239:10 (Brehm), 409:17-19 (Mohr). And even voters who obtain a court order are sometimes required to vote by affidavit, even though the order entitles them to a regular ballot. Tr. 239:1-6 (Brehm); Tr. 409:17-20 (Mohr) ("As a practical matter, there has only been two court orders issued in the 26-plus years that I have been Election Commissioner [of Erie County], and both times [the voters have] been directed to vote an affidavit ballot.").

Inactive voters can also proceed by affidavit ballot. A voter who chooses this method fills out, by hand, (1) the ballot itself, which is identical to that cast by active voters, and (2) a one-page affidavit contained on the ballot envelope. Tr. 155:14-20; Connolly Decl. ¶ 57; Pl. Ex. 259 (sample completed affidavit). Voters then seal the ballot in the envelope and return it to a poll worker, who then places it into a secure location. Tr. 155:14-20; Ryan Dep. 54:21-55:14, 199:2-22. Inactive voters do not use the electronic voting machine. Pl. Reply Facts 14. Among other things, the one-page affidavit asks the voter for the following information: the voter's personal information (such as name and date of birth), the voter's party enrollment, the reason the poll worker was unable to provide a regular ballot, the voter's voting history and information, and the voter's political-party enrollment. See Pl. Ex. 259. It also asks the voter to "swear or affirm that" she is a United States citizen, has "lived in the county, city or village for at least 30 days before the election," and meets all requirements to register to vote in New York. Id. The voter must sign and date the affidavit. The voter must also swear or affirm that "The above information is true, I understand that if it is not true, I can be convicted and fined up to $5,000 and/or jailed for up to four years." Id.

In most counties, the names of inactive voters are maintained in a list available only at the County Boards. Tr. 408:14-25 (Mohr). Two counties, however, do not follow this approach. Both Nassau and Columbia Counties choose to keep a supplemental list of inactive voters at polling sites. In these counties, when a poll worker does find a voter in the active registry, the worker checks the inactive list.

2. How the State Counts Affidavit Ballots

Affidavit ballots are not opened or counted on Election Day. Instead, state officials evaluate them in the days that follow. Ryan Dep. 69:24-82:19. Every affidavit ballot is evaluated, but many are not counted. Id. at 60:22-61:15. The State may conclude that an affidavit ballot should not be counted for several reasons. An affidavit ballot envelope may not be counted if there are errors in its completion. Id. at 195:5-16. For example, a voter may attest that she wished to be registered as a member of a particular political party but then fail to state which party. Id. at 57:10-25. A voter may leave portions of the ballot blank. Id. at 55:18-57:25. Or a voter may fill out an instruction card instead of the envelope itself. Id. at 55:18-56:11. Moreover, affidavit ballots are counted only upon a determination that "the voter is eligible to vote in the election." Tr. 68:5-20 (Martin). When evaluating a ballot, therefore, State officials look up a voter's registration status. Id. There are several possibilities: First, they may determine that the voter was listed in the poll book all along and voted at her correct location. That voter should have been allowed to cast a regular ballot, but a poll worker may have incorrectly instructed her to vote by affidavit because that worker mistakenly overlooked the voter's name in the registry, even though her name was listed there. This vote would be counted. Second, they may determine that the voter voted at her address of registration, but the State had improperly marked her as inactive. In other words, the voter was marked inactive even though she did not move. Tr. 69:17-25-70:8 (Martin). This vote would also be counted, and the voter would be moved back to the active list. Ryan Dep. 178:23-180:3. Third, they may discover that the voter did not vote at her place of registration and stated in her affidavit ballot that she has moved to a new residence. This voter's affidavit ballot would be counted as a vote in her new election district and her voter registration would be updated with the new address. Ryan Dep. 177:18-178:22. Under a new "universal transfer" law, a voter who is registered "anywhere in New York can cast a valid affidavit for the poll site of their new residence anywhere in New York." Connolly Decl. ¶ 63; Pl. Reply FIF at 19. Fourth, they may determine that the voter may have never been registered to vote, either because she never registered or because her registration was canceled. Id. This vote would not count, as an affidavit ballot cannot be used to register in the first instance. Fifth, in a primary election, they may determine that a voter is not registered as a member of the relevant political party. This vote would not count. New York State recently enacted a law that allows the State to count an affidavit ballot if the Board determines that the voter was eligible and that "the voter substantially complied" with election laws. Assembly Bill A1320A (2019-2020 Legislative Session), available at https://www.nysenate.gov//legislation/bills/2019/A1320.

Once this process is complete, each voter who voted by affidavit is mailed a letter informing her whether her vote was counted and, if it was not, the reason why. Ryan Dep. 85:5-86:21. Voters can challenge an adverse determination in state court; if they succeed, the State updates the election tally to include their vote. Id. at 86:22-89:3.

D. Affidavit Ballots: The Reality

The reality, however, is quite different from the theory. In practice, tens of thousands of New York voters are improperly registered to vote as inactive, even though they continue to reside at their address of registration. And affidavit balloting causes substantial delay, in part because of poll-worker error and confusion, increasing voting times for inactive voters and all New York voters.

1. Tens of Thousands of Voters are Placed on the Inactive List Even Though They Have Not Moved

Common Cause has proven, beyond a preponderance of the evidence, that thousands of New York voters are placed on the inactive list even though they have not moved. Examples from voters and statistics analyzed by a voter-theory expert compel this conclusion. And Common Cause has demonstrated the root of this problem: inaccurate data from the United States Postal Service and the National Change of Address registry. In short, the proxies that New York uses to determine voter movement are substantially overinclusive. They erroneously capture tens of thousands of residents who remain at their address of registration and thus lead the State to incorrectly move these voters to inactive status.

a. Voter Examples

Common Cause identified numerous voters who did not move but nonetheless were listed as inactive voters. Robert Holman stated that he has lived in the same home in Erie County since 1965. Pl. Ex. 262, Holman Dep., 11:20-12:19. He testified that he had voted in more than ten elections since 2000, and had no difficulty voting before 2016. Id. 20:12-25, 22:2-4. In 2016, he arrived at a polling location, waited in line, and then was unable to cast a regular ballot because his name was not listed in the active registry. Id. 27:10-19. The poll worker "d[id] [nothing] else to figure out why [he] [was not] in the poll book," and Holman voted by affidavit. Id. 27:20-28:18. Ralph Mohr, one of two commissioners for the Erie County Board of Elections, testified that Holman had been listed as inactive because "an official postcard mailed by our office [was] returned by the United States Postal Service as not deliverable as addressed." Def. Ex. D, Mohr Decl., ¶ 17. The State does not dispute that Holman never moved, and therefore he was incorrectly listed as inactive.

Jacques Fages stated that he has resided in the same New York City apartment since 1980, and has voted in every federal election since. Pl. Ex. 191, Fages Decl. ¶¶ 6-7. When he went to vote in 2017, he found he was not listed in the active-voter registry. He therefore voted by affidavit ballot. Id. ¶ 9-14. Fages stated that he never received any mail from the Board regarding his complaint about being listed as inactive or whether his affidavit ballot was counted. Id. ¶ 19. Robert Brehm testified that Fages had been placed in inactive status because a mail-check card was returned as undeliverable. Brehm Decl. ¶ 55-57. The State does not dispute, however, that Fages never moved from his apartment. Id. Like Holman, therefore, Fages was incorrectly listed as inactive.

Several other voters testified they had similar experiences. When Susan Stewart arrived at her polling location in 2018, she found that her name was not listed in the active registry. Pl. Ex. 202, Stewart Decl., ¶¶ 13-16. Yet she had not moved since she registered to vote. Id. ¶¶ 5-6. After substantial delay, Stewart voted by affidavit. Id. ¶¶ 17-28. Similarly, when Katherine Baldus arrived at her polling location, her name did not appear in the active registry. Pl. Ex. 187, Baldus Decl., ¶¶ 12-13. She had lived at the same residence for five years. Id. ¶¶ 6-8. There were additional examples introduced at trial. See, e.g. , Pl. Ex. 200, Angela Roberts Decl. (listed as inactive even though had not moved); Pl. Ex. 201, Denise Roberts Decl. (same). Michael Ryan agreed that "there's certainly people who you get information from the post office and, in fact, the person still lives where they've always lived." Ryan Dep. 93:7-11. In short, Common Cause introduced undisputed evidence that numerous voters were incorrectly listed as inactive, even though they had not moved from their address of registration.

b. The Statistics

One of Common Cause's expert witnesses, Marc Meredith, analyzed this phenomenon using statistics provided by the State and County Boards. Marc Meredith is a tenured professor of political science at the University of Pennsylvania. Pl. Ex. 264, Meredith Decl., at 2. He has "extensive training in economics, political science, and statistics" and has written dozens of peer-reviewed articles and books. Most importantly for purposes of this case, Meredith specializes in in "examining] information contained in voter registration databases to understand the determinants of voter turnout." Id. His analysis reveals that thousands of New York voters are incorrectly moved to inactive status even though they continue to reside at their address of registration.

The State provides statistics that are probative of this issue. For each election, the State and County Boards release the number of inactive voters who cast a valid affidavit ballot and were then reactivated to active status at their original address of registration. These are voters like the individuals discussed above: they continue to reside at the same location, were improperly designated inactive, and then voted by affidavit. As noted, when an inactive voter casts a valid affidavit ballot, her voter registration is updated and she is moved back into active status. In the State's parlance, these are voters who have been "reactivated."

Analyzing this data, Meredith confirmed that "[s]ome registrants are listed as inactive even though they have continuously resided at their address of registration." Meredith Decl. at 12. He also concluded "that approximately 45,000 affidavit ballots were cast in the 2016 presidential election in New York by inactive registrants who resided at their registration address." Id. ; Pl. Ex. 265, Meredith Supp. Decl., at 2.

Meredith made clear that this number represents a floor of the number of voters improperly marked as inactive. "While these voters represent only a fraction of the total inactive registrants who reside at their address of registration, quantifying the number of affidavit ballots inactive registrants cast at their address of registration provides information on the minimum number of inactive registrants who reside at their address of registration." Meredith Decl. at 14 (emphasis added). In other words, there are voters who continue to reside at their address of registration but did not vote in subsequent elections or were not offered affidavit ballots, and thus were not reactivated. Meredith is unable to quantify that population.

In short, Meredith thus found that the problems experienced by voters like Robert Holman and Susan Stewart are not isolated incidents. Just the opposite: they occur to tens of thousands of New York voters. Meredith also explained how so many voters are improperly marked as inactive. Two of the proxies used by the State, the Postal Service and the National Change of Address Registry, are error-prone and cast too wide a net. Id. at 12-13.

c. Problems with the Postal Service

The Court turns first to the State's most important proxy for voter movement—Postal Service data. As discussed, the State and County Boards mail millions of pieces of mail to voters every year. If one of those pieces of mail is returned as undeliverable, the voter is removed from the active registry, deleted from the poll book, and marked as inactive. Data from the Postal Service, however, is an unreliable measure of voter movement.

Michael Ryan is a former Commissioner of the New York City Board of Elections and currently serves as its Executive Director. Ryan Dep. at 11:6-22. The City Board conducts all state and federal elections in the city—it deploys equipment, organizes training for poll workers, and will soon manage early voting. Id. 19:18-20:19. As Executive Director, Ryan is responsible for the City Board's operations and implementing election policy for the city. Id. 17:89-24.

Ryan testified extensively about how Postal Service data is an unreliable proxy for voter movement. Ryan explained that the "process [of determining which voters have moved] is almost exclusively reliant on the United States Post Office ... the quality of the information that we get from the United States Post Office is the linchpin to the success or failure of this inactivity process." Id. 96:8-24. He went on to state that this data is profoundly unreliable. He said that "[the City Board] [has] uncovered over the course of time significant issues with the consistency of the information that we get from the post office vis-[a]-vis who's at this location and who is not at the location. Id. at 99:14-18; see also id. at 104:6-8 (bemoaning "the poor quality and the lack of consistency of the post office"); 205:16-18 (noting his "trepidation about the overall structural integrity of the [Postal Service] process."). In other words, the Postal Service often returns mail as undeliverable even though the voter continues to reside at the same location. And Ryan testified that the Postal Service's coding of undelivered mail, which is supposed to provide the sender an explanation for why the piece of mail was returned as undeliverable, can be arbitrary and cryptic. Id. 100:24-101:20. The Post Office uses "about 18 or 19 different potential categories" to code returned mail, and Ryan expressed confusion about the meaning of these various categories. Id. at 202:11-21; 205:18-206:11; see also 129:6-12 (agreeing that he "still [does not] have clarity" regarding the post office's coding system). Individual postal workers differ in how they understand these categories, further exacerbating the problem. Id. 208:6-209:3; 210:2-211:7 (noting that there are "bins for each type of mail that comes back," and that post office's coding method "switch[es] depending on [which Postal Worker is] throwing the envelopes into the bin."). Ryan explained that "the most disconcerting part of this whole thing is ... it really comes down to, ultimately, the quality of the work of the individual postal worker." Id. 99:23-100:6.

Even the confirmation-notice process, which is supposed to serve as a double check, does not always work perfectly. To start, if the original mail-check notice was returned as undeliverable due to a Postal Service error, then it is possible that the confirmation notice could be too. Indeed, if one piece of mail was returned as undeliverable, and the Board has had no other reason to update the address, it is likely that the second piece of mail will be returned as well. Moreover, Ryan testified that the Postal Service has returned completed confirmation cards from voters several months after the voter sent them. Ryan Dep. 86:2-16. These presumably would have sufficed to reactivate a voter, but because of Postal Service delay, they were received too late and the voter thus remained inactive during an election. Id. Meredith explained that "[s]uch errors may be particularly consequential in a state like New York that does not require that a second piece of mail also be returned as undeliverable for a registration to be moved from active to inactive." Meredith Decl. at 12-13.

The City Board is so concerned with the Postal Service's unreliability that it has taken multiple steps to assuage these problems. First, the City Board met with representatives from the Postal Service to fix these problems, but had little success. Ryan Dep. 133:20-134:4. The City Board, responsible for more than one-million voters, thus continues to use Postal Service data as a proxy for voter movement even though it knows about its high error rate. See id. 149:14-151:16 ("I believe we find ourselves basically in the same position we were [in] when we [the City Board] started" meeting with the Postal Service, more than one-and-a-half years prior). These errors "disproportionately affect people that happen to live in a multi-unit building." Id. 102:2-9; accord 100:7-23. For example, Ryan stated that Bronx County has many multi-dwelling buildings and faced this problem acutely, calling the issue a "systemic failure." Id. 136:11-137:14.

Second, the City Board adopted a policy to confirm Postal Service reports that a voter has moved. Ryan explained that, before listing a voter as inactive, the New York City Board checks whether the voter has voted in a recent election at their address of registration. Id. 125:13-23. Ryan agreed that "every year there are people who fall into that category where the post office is telling you they're not there, but in fact, they [are]." Id. at 125:24-126:9. There is no evidence that any other County Board conducts this sort of confirmation.

Third, the City Board is so concerned about its mail being improperly returned as undeliverable, and thus causing voters to be marked as inactive, that it "tr[ies] not to send casual pieces of mail that aren't necessary to eliminate the return issue ... [because] superfluous communication can lead to an unintended consequence." Ryan. Dep. 108:8-16; accord 98:18-99:9 (noting that "while people want us to communicate with the voters by mail and they think it's a good thing, you can also inadvertently lead to voters becoming inactive"). For example, notices informing voters that their affidavit ballots were counted have been returned as undeliverable, thus rendering those voters inactive—even though they had just confirmed their addresses through the affidavit process. Ryan Dep. 145:17-146:18 (explaining that "this is an ... example of the unintended consequences that we discussed earlier). In short, Ryan and the City Board "have little faith in the overall reliability of the quality of information that we get from the post office." Ryan Dep. 101:21-25. Ryan is "very concerned for the voters on the quality of information that we get from the post office." Id. at 105:4-9. And the upshot is clear: voters are registered as inactive even though they have not moved. Id. 100:25-101:20.

Michael Ryan is not the only election official to testify to this problem. Robert Brehm is the Co-Executive of the State Board. Def. Ex. A, Brehm Decl., ¶ 1. He was also the Commissioner of the Schenectady County Board of Elections for fifteen years. Id. He agreed that Postal Service data is an unreliable proxy. Brehm said he had "seen a few of those" in reference to "voters who went to long-term polling places, hadn't moved, but were wrongly identified as inactive, returned mail check documents apparently as a result of post office error otherwise." Tr. 289:22-290:3 (Brehm). He reiterated that he had experience with voters who were marked inactive but continued to live at their address of registration. Id. 290:16-21 (Brehm). Virginia Martin, the current Commissioner of the Columbia County Board of Elections, made the same point. She agreed that "inactive voters in Columbia County are casting affidavit ballots even if they're appearing at the correct polling place and they have not moved." Tr. 87:22-88:1 (Martin).

In short, the central proxy that the State uses to determine whether a voter has moved has serious problems with its reliability, and multiple State officials have expressed concern about its use.

d. Problems with the National Change of Address Registry

Errors made by the Postal Service are not the only way that the State incorrectly determines that voters have moved and thus marks them as inactive. As noted above, the State Board also receives data from the National Change of Address Registry (NCOA). Several witnesses testified as to the unreliability of that data.

Meredith testified that "[e]rrors in matching registration records to the NCOA registry can also incorrectly identify a registrant as having moved." Meredith Decl. at 13. He explained that the Registry's "primary purpose ... is to support the U.S. Post Office" and it therefore "does not collect certain data, like date of birth, that would help election officials to link the NCOA data to registration records." Id. The State admits that it relies on the Registry even though "[t]he NCOA registry does not collect certain data ... that would help" maintain accurate voter-registration records. Pl. FIF at 16; Def. FIF at 13. Based on data in the NCOA, a County Board "may incorrectly believe that a registrant has moved if, for example, a family member who shares the same name as the registrant is actually the person who moved, rather than the registry." Meredith Decl. at 13. Meredith went on to state that "[a]nother factor that makes it difficult to link NCOA data to registration records is that a single NCOA can apply to an entire family." Id.

Both Michael Ryan and Robert Brehm confirmed this problem with NCOA data. Brehm explained that if an individual within a household has moved, data from the NCOA may lead the Board to mark the entire family as inactive. Tr. 290:3-15 (Brehm). Ryan identified another potential problem with the NCOA: "[t]he problems that we have identified with respect to the NCOA center more on those instances where somebody inadvertently checked the permanent box as opposed to the temporary box and had only meant to temporarily have their mail forwarded to a particular location." Ryan Dep. 225:17-25. This situation could arise if "somebody is temporarily relocated, maybe they care for a loved one." Id. at 226:13-18. In sum, the two proxies used by the State to determine voter movement are substantially overinclusive.

2. Some Inactive Voters Are Turned Away from Voting

Although inactive voters are eligible for affidavit ballots, poll workers often fail to offer them. This is yet another way in which the practice of affidavit voting in New York does not live up to the theory.

Stephanie Goldberg registered to vote around 2010 and has remained at the same address since that time. Pl. Ex. 193, Goldberg Decl., ¶¶ 3-6. When she arrived to vote in 2018, her name was not in the active registry. Goldberg Decl. ¶¶ 15-17. The poll workers did not offer her an affidavit ballot. She then left the polling location, looked up her registration status on her cell phone, discovered she was inactive, returned to the polling location, and showed the poll worker a screenshot of her voter registration status. Even though that screenshot listed her status as inactive and confirmed that she was at the correct polling location, she was still not offered an affidavit ballot. Id. ¶¶ 15-31; Ex. A (screenshot). The State admits that "[h]ad a poll worker provided her an affidavit ballot, Stephanie Goldberg would have cast a ballot that should have counted in the 2018 general election." Pl. FIF at 31, Def. FIF at 23. The State also admits that "[t]he problems experienced by Stephanie Goldberg attempting to vote in the 2018 general election illustrate how an inactive registrant is disenfranchised when a poll worker fails to offer an affidavit ballot." Pl. FIF at 31; Def. FIF at 23.

Susan Stewart testified that she had a similar experience. When she arrived at her polling location in 2018, her name was not listed in the active registry. Stewart Decl. ¶¶ 13-16. She testified that the poll worker "appeared to be confused and overwhelmed" by the situation. Id. ¶ 16. The poll worker did not immediately offer her an affidavit ballot or take any other steps to confirm her registration status or polling location. Instead, Stewart called the County Board herself using her personal cell phone, learned of the affidavit-ballot process, and was then offered one. Id. ¶¶ 17-28.

When Lauren Wolfe arrived at her proper polling location in 2016, she was not listed in the active registry. Pl. Ex. 204, Wolfe Decl., ¶ 11; see also Def. FIF 66-67 (confirming that Wolfe was at her proper voting location). The poll worker was unable to locate her in the active registry, did not offer her an affidavit ballot, and then directed her to another polling location. Wolfe Decl. ¶¶ 10-12. Similarly, when Allison Agro-Paulson went to vote in Brooklyn, she saw several voters whose names were not listed in the poll books leave the polling place without casting an affidavit ballot. Pl. Ex. 186, Agro-Paulson Decl., ¶¶ 27-29.

Another voter with a similar experience is Sandra Copps. Pl. Ex. 189, Copps Decl. Copps has lived in her Bronx apartment since 1998. Id. ¶ 6. When she went to vote in the 2018 primary election, she was informed that she was not listed in the poll book. Id. ¶¶ 13-18. The poll workers did not inform Copps that she could vote using an affidavit ballot or offer her one, and she testified that if she "had been offered the opportunity to cast an affidavit ballot at the primary election held on September 13, 2019, I would have done so." Id. ¶¶ 17-18, 23. The State notes that Copps was not registered as a member of the correct political party to be eligible to vote in the 2018 primary. Def. FIF 64-65. But nothing suggests that the poll worker (or even Copps) knew that at the time. Under the State's policy, when the poll worker was unable to locate Copps' name in the registry, she should have been offered an affidavit ballot—not turned away from the poll site.

Two state officials confirmed that this phenomenon is not limited to just these individuals, but occurs in each of the counties they administer. Virginia Martin testified that she knows that some poll workers "will make a judgment that a voter is not eligible and tell them they cannot vote by affidavit ballot." Tr. 88:25-89:2 (Martin). She confirmed that "based on [her] experience, sometimes poll workers fail to offer affidavit ballots to eligible voters in Columbia County polling places." Tr. 90:9-12 (Martin). Michael Ryan agreed that "as a practical matter, given [his] role and experience as executive director [of the New York City Board of Elections], [he knows] that" "people whose names aren't in the books, in fact, leave the poll site without voting an affidavit." Ryan Dep. 248:11-22.

Further evidence comes from an audit conducted by New York City's Office of the Comptroller. The Office audited the City's Board of Elections in 2017 and produced a report summarizing its finding. See Pl. Ex. 55. The parties stipulated to the report's admission into evidence, and the State did not raise any objection to it. Dkt. No. 157 (stipulation); Dkt. No. 130, Joint Pretrial Report, at 26 (no objection to this document). The Report noted that "[a]t one site, rather than following federal and State law and offering an affidavit ballot to a voter whose name was not found on the registration rolls, the coordinator told the information clerk that the voter was not able to vote at all. This error could have disenfranchised the voter, whose eligibility could have been verified had the voter been allowed to vote with an affidavit ballot." Pl. Ex. 55 at 11-12.

The State does not track how many voters appear at the correct polling place, are eligible to cast an affidavit ballot, yet leave because they are not offered one. Connolly testified that the Board did not know, and did not keep any statistics on, how many inactive voters fell into this category. Tr. 129:3-6, 130:4-6 (Connolly). Valentine stated that "there is no way to measure how many inactive voters are turned away from the poll without being offered an affidavit ballot." Tr. 156:14-17 (Valentine). He also later testified that there is no way to know how many voters arrive at the polling site but then choose not to vote, perhaps because of delay. Tr. 165:11-168:7, 174:11-21 (Valentine). The State does not "have a systematic procedure with respect to retaining or logging voter complaints." Tr. 166:9-15 (Valentine). The Court then asked Valentine how the State would even know if they have a systemic issue, as it does not "regularly log or retain complaints." Tr. 167:7-8 (Valentine). Valentine claimed that the Board would know through word of mouth, calls from the County Boards, and media reports. Tr. 167:7-169:12 (Valentine).

3. Affidavit Voting Causes Delay

Affidavit ballot also causes substantial delay at polling locations throughout New York. Inactive voters often face delays in receiving and casting ballots, and these delays have ripple effects for all New York voters.

a. It Takes More Time to Receive an Affidavit Ballot than a Regular Ballot

Common Cause describes the problem as follows: "The Voting experiences of New York voters who attempted to vote at the correct polling place but whose names could not be located in the poll book were complicated, involved multiple and often redundant interactions with poll workers, and took a relatively long time." Pl. FIF at 19. The evidence supports this conclusion.

When a voter is not listed in the poll book, New York law requires poll workers to attempt to find their correct polling location. "Whenever a voter presents himself or herself and offers to cast a ballot, and he or she claims to live in the election district in which he or she seeks to vote but no registration poll record can be found for him or her in the poll ledger ... a poll clerk or election inspector shall consult a map, street finder or other description of all of the polling places and election districts within the political subdivision in which said election district is located and if necessary, contact the board of elections to obtain the relevant information and advise the voter of the correct polling place and election district for the residence address provided by the voter to such poll clerk or election inspector." N.Y. Elec. Law. § 8-302(3)(e) ; accord Connolly Decl. ¶ 53; Def. Ex. D, Mohr Decl., ¶ 7 ("It is the policy of the Erie County Board of Elections that if a poll worker cannot find the name of a voter in the poll book, the poll worker must verify that the voter is in the proper polling location and election district by utilizing the street listing which is provided to each election district with their supplies; that poll workers are instructed that persons appearing at the wrong location must be directed to the correct location and provided with directions."). This process takes substantially longer than check in for active voters, who often can cast ballots within minutes of reaching the poll worker.

Virginia Martin testified that poll workers take even more steps for inactive voters. She stated that when an inactive voter "arrives at the polling place" and "the poll inspector cannot find her name," the poll worker "will question [the voter] as to whether she has had a name change, will question her as to the spelling of her name." Tr. 84:1-16 (Martin). Poll workers are also "instructed ... [to] just call the Board of Elections if [they] have any questions." Tr. 85:21-25 (Martin) (emphasis added). The State thus requires poll workers to jump through a few hoops before offering inactive voters affidavit ballots. And the State does not make the process seamless. For example, the State does not provide landlines to all poll workers, and workers trying to make calls or use the internet from personal cell phones may face connectivity issues. Trial Tr. 92:12-20 (Martin) (noting that poll sites do not have landlines and poll workers "rely on their cell phones in almost every instance, if they can get connection."). Various voters testified that when they were not found in the poll book, the poll workers proceeded to give them an affidavit ballot; these voters did not mention the poll workers conducting any independent search about their polling location. For example, when poll workers failed to locate Stephanie Goldberg in the poll book, they expressed confusion and did not look up her address of registration. Goldberg Decl. ¶¶ 15-20, see also Wolfe Decl. ¶¶ 11-20. The same was true for Robert Holman, who testified that poll workers told him he was not listed in the active registry but took no further steps, such as calling the County Board or using Voter Search, to confirm his registration status or polling location. Holman Decl. ¶¶ 11-14; Holman Dep. 27:20-31:13. And more than 35,000 voters who were not listed in the poll books in the 2016 election cast ballots in the wrong polling location. Meredith Supp. Decl. ¶ 5.

Voters not appearing in the poll book should, under New York law be provided with a notice-to-voter form, explaining that they can either seek a court order or vote by affidavit ballot. However, no voter reported that poll workers actually provided this form. See, e.g. , Ex. P186, Agro-Paulson Decl.; Ex. P187, Baldus Decl.; Ex. P189, Copps Decl.; Ex. P190, Edelman Decl.; Ex. P191, Fages Decl.; Ex. P193, Goldberg Decl.; Ex. P194, Goldblum Decl.; Ex. P195, Holman Decl.; Ex. P196, Johnson Decl.; Ex. P197, Lavnick Decl.; Ex. P198, Matika Decl.; Ex. P200, A. Roberts Decl.; Ex. P201, D. Roberts Decl.; Ex. P202, Stewart Decl.; Ex. P203, Wilson Decl.; Ex. P204, Wolfe Decl. Similarly, voters who cast affidavit ballots should be informed at a later date whether their affidavit ballot counted. However, several voters did not receive these notifications. Ex. P262, Holman Dep., 49:14-16, 50:14-51:22, 54:12-13 (no confidence that his affidavit ballot was counted); Ex. P186, Agro-Paulson Decl., ¶ 34 (does not know whether her affidavit ballot was counted); Ex. P187, Baldus Decl., ¶ 25 (same); Ex. P202, Stewart Decl., ¶ 38.

Inactive voters thus face delay even if all goes according to plan. But the testimony of numerous voters reveals that New York's poll workers do not uniformly follow this process, are confused about the affidavit-ballot process, and often give voters incorrect directions. Taken together, voters who do not appear on the active registry can spend significant amounts of time just getting the affidavit ballot to which they are entitled.

The record is replete with testimony to this effect. To start, Allison Agro-Paulson arrived at her polling location during the 2018 election, but was not listed in the active registry. Agro-Paulson Decl. ¶¶ 14-16. She spent time with the poll workers "double-check[ing] the poll book to confirm [she] was not listed, and checked all possible iterations of my last name, such as ‘Agro Paulson,’ ‘AgroPaulson,’ ‘Agro,’ and ‘Paulson,’ for example. None of the possible permutations of my name were listed in the poll book." Id. ¶ 16. She also spent time checking whether her husband's name was listed in the poll book, and the poll worker confirmed that he had already voted and had signed the poll book. Id. ¶ 17. The poll workers directed Agro-Paulson to check with various other workers, causing her to spend additional time walking around the poll site. Id. ¶¶ 19-21. The poll workers did not attempt to contact the State or County Boards. Id. ¶ 22. The poll workers eventually gave her an affidavit ballot, which she filled out. Id. ¶¶ 23-24. The State represents that her ballot was counted, but provides no explanation for why she was not listed in the active registry, even though she had not moved. Def. FIF at 67. All in all, Agro-Paulson spent approximately one hour attempting to vote. Agro-Paulson Decl. ¶ 27. Stephanie Goldberg had a similar experience. When she arrived to vote in 2016, her name was not in the active registry. Pl. Ex. 193, Goldberg Decl. ¶¶ 15-17. The poll workers were unsure how to proceed, spoke with each other, and ultimately did not offer Goldberg an affidavit ballot. Id. ¶¶ 18-21. She then left the polling place, returned to her car, used her cell phone to search her registration information, and discovered she was inactive. Id. ¶¶ 20-25. She then returned to the polling place and showed the poll workers a screenshot of her registration. Id. ¶¶ 27-29. She was still not offered an affidavit ballot, returned to her car again , called the County Board, and emailed the State Board. Id. ¶¶ 30-37. She then left the polling location to attend university classes. Id. ¶¶ 37-38. The State Board eventually informed her that she could vote by affidavit, but she did not have enough time after her classes to return to the polling location. Id. ¶¶ 38-42. Goldberg does not state how long she spent at her polling location, but it is clear that she was not offered an affidavit even after multiple interactions and after showing the poll worker her registration.

Other voters faced similar delays in getting their affidavit ballots. As discussed above, when Susan Stewart arrived at her polling location in 2018, her name was not listed in the active registry. Stewart Decl. ¶¶ 13-16. The poll worker did not immediately offer her an affidavit ballot or take any other steps to confirm her registration status or polling location. She said that a poll worker tried to call the Board, but "was taking a long time and his phone did not seem to be working properly." Id. ¶ 18. Stewart thus took matters into her own hands, called the County Board herself using her personal cell phone, learned of the affidavit-ballot process, and was then offered one. Id. ¶¶ 17-28. When Lauren Wolfe arrived at her polling location in 2016, she too was not listed in the active registry. Pl. Ex. 204, ¶ 11. The poll worker directed her to another polling location. When she reached that location, the worker there told her she was not in the book, and that she should go back to the original location. She was given an affidavit at the original location, and it is not clear how long this back-and-forth took. Id. ¶¶ 11-16. Robert Holman testified as to his experience not being listed in the active-voter registry, despite not having moved in decades. When he arrived at the polling location, the poll workers were unable to locate his name in the poll book. Holman Dep 27:20-28:18. They took five to ten minutes to offer him an affidavit ballot, even though they did nothing else to confirm his address or proper polling location (such as calling the County Board or using Voter Lookup). Holman Dep. 28:5-29:4.

The Audit Report further supports this conclusion. The Report states that "[w]e also observed an apparent lack of knowledge on the part of numerous poll workers as evidenced by their failure to follow proper procedures mandated by law, including ... An inspector at one site did not know how to proceed after being unable to find a voter's name on the list of registered voters." Pl. Ex. 55 at 16. The Report also confirmed that poll workers provided "misinformation or no information ... at various poll sites that, at a minimum, caused confusion, and that increased the risk that ... voters would not be able to vote or that their choices of who to vote for might be influenced by the misinformation provided." Id. at 16. For example, "[a]t one site a coordinator was not able to answer a poll worker's question regarding affidavit ballots and sent the poll worker to other poll workers for the answer." Id. The Report provides another example in which "the workers' apparent lack of knowledge resulted in a chain of missteps that lasted thirty minutes ." Id. at 17 (emphasis added). There, the poll worker was unable to find the voter's name in the active registry, and then "asked the voter for an ID (prohibited by federal Election Law) rather than offering an affidavit ballot, as required." Id. The coordinator overheard this exchange and offered the voter an affidavit, "but failed to instruct the voter to fill it out a privacy booth." Id. The inspector then "incorrectly instructed [the voter] to scan the affidavit ballot at the scanner rather than to return the ballot to the table inspector, who was required to store all completed affidavit ballots in a separate envelope." Id. at 17-18. The scanner did not accept the ballot, the voter returned it to the inspector, who then "required the assistance of another table inspector in order to properly store it." Id. at 18. In sum, the affidavit-ballot process confuses poll workers and voters alike, and New Yorkers spend significant chunks of time receiving the affidavit ballot to which they are entitled.

b. It Takes Time for Voters to Cast an Affidavit Ballot

It also takes longer to cast an affidavit ballot than a regular ballot, thus placing an additional burden on this subset of voters. Brehm testified that, in his experience as an election administrator, completing an affidavit ballot takes "two to three extra minutes" more than a regular ballot. Tr. 285:7-13 (Brehm). Valentine agreed that "filling out the affidavit" will "take more time" than casting a regular ballot. Tr. 154:17-23 (Valentine). Ryan likewise noted that affidavit ballots "provide a level of inconvenience to the voter in that [they] cause[ ] their process to take longer." Ryan Dep. 51:12-19. For example, completing the affidavit ballot took Holman an additional "[f]ive, ten minutes." Holman Dep. 31:2-4.

Meredith testified that "the amount of time it would take to fill out an affidavit ballot would depend heavily on someone's reading comprehension ... the type [on the affidavit ballot] is small." Tr. 316:25-317:12 (Meredith). In other words, "filling out an affidavit ballot [is not] equally burdensome for all voters," and the time involved will depend on reading level, eyesight, and whether the affidavit ballot is offered in the voter's preferred language. Tr. 352:1-10 (Meredith). And although one portion of the affidavit ballot is optional, in theory reducing delay, Meredith testified that "based on [his] own visual inspection of affidavit ballots that even" voters fill out this section do so anyway because "they don't realize they don't have to do that." Tr. 351:15-21 (Meredith).

c. This Additional Time Spent Causes Ripples of Delay, Thus Creating Delay for All New York Voters

The delays faced by inactive voters creates delay for all New Yorkers. Meredith testified that "excluding inactive voters' name from poll books ... increases the wait times for anyone who shows up to vote in New York." Tr. 350:1-12 (Meredith). He reached that conclusion based on the limited bandwidth of poll-workers. "[W]e know poll worker resources are scarce on Election Day, that poll workers have to deal with a lot, and any time that poll workers are spending doing lookups on where someone, if someone is in the correct election district or not, is taking away resources that could be applied to other tasks that they have to do on Election Day including checking-in other voters." Tr. 350:13-22 (Meredith).

These delays also cause some voters to not vote all together. "Reneging is when someone gets into a line in a polling place and then leaves that line without casting a ballot." Tr. 318:6-8 (Meredith). Meredith testified that "the best peer-reviewed evidence" suggests that the longer someone has to wait to vote, the more likely she is to renege. Tr. 353:9-17; accord Meredith Decl. at 7. Meredith thus concluded that the affidavit-ballot process and the State's decision not to list inactive voters in poll book both increase delay and therefore reneging.

Common Cause's other expert witness agreed with this conclusion. Troy Grayson served as Kentucky's Secretary of State for seven years, and he was thus the state's "chief election officer" during that period. Pl. 238, Grayson Decl., ¶ 2. He also served as the President of the National Association of Secretaries of State and was appointed by President Obama to the Presidential Commission on Election Administration. Id. ¶¶ 3-5. Grayson testified that "[v]oters on the active list, for example, face longer lines because of the extra time that inactive voters require at check-in because of the absence of the inactive voter list in the precinct." Id. ¶ 43.

d. These Problems are Made Worse by Limited Poll-Worker Training and Bandwidth

A lack of poll-worker training and understaffing contribute to this delay. Poll workers receive little training on how to process inactive voters. County Boards are responsible for providing poll worker training. Def. Ex. B, Connolly Decl., ¶ 16 ("The training of these poll workers is provided at the county level."). Connolly testified that "[t]he duration of poll worker training sessions ranges from one and half hours to six hours, depending on the county." Id. Most poll workers work only one to three times a year, although some poll workers may work for multiple years. Connolly Decl. at ¶ 4. In 2016, more than 7,000 of the poll workers in New York City were first-time poll workers. Meredith Decl. at 24. The State agrees that "[t]he temporary nature of poll work also increases the likelihood of poll workers failing to adhere to proper protocols." Pl. FIF at 25; Def. FIF at 19. And although the State Board provides the base curriculum that County Boards are supposed to use in these trainings, which briefly covers inactive voters, it does not evaluate the extent to which the trainings provided by the County Boards follow its curriculum. Tr. 123:8-22 (Connolly); Pl. FIF 23; Def. FIF at 17.

In New York City, most poll workers receive training once a year. Ryan Dep. 41:22-44:20 (testifying that poll workers in New York City are typically trained once a year). Ryan testified that poll workers receive a basic training course of four hours. Id. 159:24-160:15. The New York City board scaled back its training recently, reducing it from six hours to four. Id. 172:3-12. And at this limited training, affidavit ballots are just one of many topics covered. Ryan stated that instructions regarding affidavit ballots are "an element of the training ... I couldn't tell you how much it's stressed in the moment." Ryan Dep. 166:21-9; see also Tr. 123:20-22 (Connolly) ("I cannot say with any certainty" how much time "is spent on training poll workers regarding inactive voters and affidavit ballots.").

State officials testified that poll workers thus make mistakes. Grayson Decl., ¶ 39, Ex. P263, Ryan Dep., 45:3-24 (acknowledging that New York City poll workers are "human beings" and "occasionally make mistakes"); Connolly Decl. ¶ 5; (admitting that "there will inevitably be problems to address"); Tr. 98:8-17 (Mohr); see also Tr. 106:24-107:3, 107:13-108:7, 111:2-112:6 (Connolly).

Polling locations may also be understaffed. Meredith testified that "[i]n New York City, at least 3,000 poll worker positions went vacant, suggesting no one qualified could be found who wanted the job. And it is not just New York City that struggles to find poll workers in New York. Responding to the question of how easy or difficult is it to obtain poll workers for the general election, 46 of the 56 counties outside of New York City reported it was at least somewhat difficult, with 20 reporting it was very difficult." Meredith Decl. at 22 (citing reports from the State and New York City Boards); Pl. FIF at 22; Def. FIF at 16. Meredith further explained that "we know poll worker resources are scarce on Election Day, that poll workers have to deal with a lot, and any time that poll workers are spending doing [a task] is taking away resources that could be applied to other tasks that they have to do on Election Day including checking-in other voters." Tr. 350:14-20 (Meredith). Understaffing compounds the problem of limited poll-worker bandwidth. The State agrees that "[j]urisdictions cannot be particularly selective about whether potential poll workers possess qualities that would help them do a more effective job when they cannot even find enough poll workers to fill all of their available positions." Pl. FIF at 25; Def. FIF at 19.

The Audit Report confirms the staffing problem, at least with respect to New York City. The Comptroller "found staffing deficiencies at 76 percent of the poll sites [it] visited," such as "specific roles that were not filled by the BOE, poll workers who did not arrive for work on Election Day, and inadequate assignment of interpreters for required languages." Pl. Ex. 55 at 18. The Report made clear that "[a] lack of poll workers at each site can result in longer lines, less assistance for voters and greater confusion in the voting process, all of which would frustrate individuals' attempts to vote and ultimately depress the total numbers of votes cast." Id.

To be sure, training poll workers is no easy task. These individuals work only a few days per year, and some do not work every year. Def. FIF at 16. Yet they are effectively responsible for ensuring that millions of New Yorkers exercise their franchise. And increased training strains the resources of the County Boards, which already spend substantial resources administering the elections. The State is correct that "[w]hen the conduct of an election involves a one day work force of over 60,000 people administering an election at which, in the case of the 2016 election, 7.8 million people participate, there will inevitably be problems to address." Connolly Decl. ¶ 21. The Court does not question the State's good faith and the substantial efforts it has taken to improve poll-worker training. Nor does the Court question the good faith of the poll workers. As Michael Ryan noted, many individuals work as poll workers because "[t]hey want to serve their community, they want to be part of an important piece of our democracy." Ryan Dep. 45:7-24. Yet the record nonetheless compels the Court to conclude that poll workers receive minimal training about the affidavit-ballot process, leading some voters to not receive affidavit ballots and causing delay for all New York voters, and that these problems are compounded by understaffing.

4. The State Erroneously Rejects Some Affidavit Ballots

Another problem with affidavit balloting arises in the State's review process. To be sure, the State has legitimate reasons for rejecting affidavits, like if a voter fails to complete them. But the State also admits that it has improperly rejected some affidavit ballots, thereby disenfranchising those voters.

Denise and Angela Roberts were both listed as inactive even though they continued to reside at the same address. When they attempted to vote in 2016, they arrived at the correct polling location and cast affidavit ballots. But a few months later, they learned that the State had not counted their ballots. Pl. Ex. P261, D. Roberts Dep., 10:16-12:5, 46:16-47:12, 51:5-17; Ex. P201, D. Roberts Decl., ¶¶ 11-21; See Ex. P200, A. Roberts Decl., ¶¶ 16-35. The State cannot identify any legitimate reason why these ballots were not counted. Bernadette Tooms, the Commissioner of Elections for Tioga County, admitted that these affidavit ballots should have been counted and could not identify why this mistake occurred. Pl. Ex. 230, Toombs Decl. ¶ 10. He stated that "it appears that [Denise Roberts'] affidavit ballot cast in 2016 was not counted. It should have been. This was an error and we are very sorry that it occurred." Id. He further explained that the "ballot was incorrectly researched by a former Commissioner and it got by [his] review as well." Id. He said he was not aware of other similar instances, but could not "state this never happened on any other occasion." Id. ¶ 12 (emphasis in original). Connolly admitted that these affidavit ballots should have been counted, and agreed that Denise Roberts "was improperly disenfranchised." Tr. 110:23-111:1 (Connolly).

E. A Supplemental List of Inactive Voters Would Alleviate Some of These Problems

Two New York counties, Columbia and Nassau, offer two lists of voters at polling locations. The first list is the registry of active voters, in accordance with state-wide practice. But they also offer a second list: the names of all inactive voters registered to vote in that election district. The Court concludes that providing this supplemental inactive list alleviates some of the problems discussed above.

Virginia Martin testified to the benefit having such a list provides. She stated that when an inactive voter "arrives at the polling place" and "the poll inspector cannot find her name," the poll worker "will question [the voter] as to whether she has had a name change, will question her as to the spelling of her name." Tr. 84:1-13 (Martin). She explained "[s]ometimes poll inspectors have a little trouble finding names in the poll book and hopefully they will look for the challenge inactive list and see if the voter is on that list. And if the voter is on that list and they have already ascertained that the voter is at the same address, as you have said, then they would -- then they would offer the voter the notice to voter's list where the voter could either request a court order or could ask to vote by affidavit ballot." Id. In other words, if a voter is not listed in the active registry, poll workers in Columbia County can check the inactive list. If the voter appears on that list and states they continue to reside at the same address, the poll worker can offer that voter an affidavit ballot and be confident that the voter is at the correct polling site. Martin confirmed that this is the case: "if the voter is on the inactive voter list and is at the same address as the voter says he lives at, then that indicates to the inspector that that voter is in the proper poll site at the proper district, therefore that's the proper district for the voter to vote from." Tr. 84:23-85:3 (Martin). The Court asked Martin whether "based on [her] experiences, having the supplemental poll list available at the polling place is a helpful tool for the poll workers and identifying inactive voters?" Tr. 84:14-16. Her answer was unequivocal: "Yes, it is." Tr. 84:17 (Martin).

Meredith agreed that "if inactive registrants were in the poll book at their address of registration, that would simplify the process that they would have to go through relative to what it is now." Tr. 348:11-20 (Meredith). And because "the process of identifying whether that person was in the right spot would be simplified," "the total burden on voters" would be reduced. Id. 348:21-25 (Meredith).

The State contends that providing the inactive list "could bog down the voting process." Pl. Reply FIF at 13. But they present no evidence to support this proposition. As discussed, Common Cause has proven that the opposite is true: the lack of supplemental lists at the poll sites causes confusion and delay. And the experience of Columbia and Nassau Counties discredits this argument. Martin testified that, in her many years of experience, she was "not aware of any complaints about Columbia County's placement of inactive voters on [a supplemental list] from voters, poll workers, election officials, or anyone else." Martin Decl. ¶ 25. Valentine further testified that he was not "aware of Nassau County having problems with implementing supplemental poll books at the polling place." Tr. 163:24-164:1. In sum, providing the list of inactive voters at polling locations helps assuage many of the problems associated with New York's affidavit-balloting procedures.

F. Common Cause Has Diverted Resources Because of New York's Practices

As a result of the New York laws at issue in this litigation, Common Cause diverted various resources from its usual activities.

Susan Lerner, the Executive Director of Common Cause's New York chapter, testified to the resources expended by the organization because of these issues. Common Cause deploys Election Day poll monitors throughout the State, and their attention and efforts have frequently been diverted because of issues related to New York's inactive voters, including making sure that voters who are not listed in the poll books are allowed to cast affidavit ballots. See Tr. 203:13-204:3, 204:10-25 (Lerner); Ex. P237, Lerner Decl., ¶ 12. During and after the 2016 election cycle, Common Cause staff spent substantial staff time and resources reaching out to, educating, and assisting voters who complained that they were not listed in the poll book on Election Day. Lerner Decl. ¶¶ 10-12; Tr. 203:13-204:3, 204:10-25 (Lerner). Common Cause has also conducted public education and outreach campaigns, public advocacy at hearings, press conferences, meetings with the New York State Board of Elections, and legislative advocacy. Lerner Decl. ¶¶ 13-18; Tr. 205:3-206:9, 209:14-211:18 (Lerner).

Lerner also personally took various steps as a result of New York's conduct. She created a survey to send to New York State residents to learn about their experiences attempting to vote in 2016. Ex. P237, Lerner Decl., ¶ 14. She personally spoke with voters who said that they were not listed in the poll book on Election Day. Lerner Decl. ¶ 14. She helped affected voters by contacting county and state election officials to try to ascertain their voter registration status and whether or not their affidavit ballot might be counted. Id. And during and after the 2016 election cycle, Lerner developed a public-information campaign to raise awareness and educate voters concerning the circumstances faced by the many voters who arrived at the polling place but find that their names are not listed. Lerner Decl. ¶ 13. "Common Cause's efforts included alerting voters to potential problems at the polls and encouraging them to be proactive if they experienced problems with voting or their registration status." Id; Tr. 205:3-206:9, 209:14-211:18 (Lerner).

During the 2018 election cycle, Common Cause also hired a contractor, Katherine Hawkland, who was primarily devoted to voter engagement. Ex. P237, Lerner Decl., ¶ 16. Hawkland spent substantial time reaching out to inactive voters, who were at the risk of being disenfranchised. Id. Hawkland was compensated at a rate of $5,000 per month. Lerner Decl. ¶ 16. In short, Common Cause spent substantial monetary and dedicated non-monetary resources as a result of the New York laws at issue in this litigation. III. COMMON CAUSE HAS STANDING TO BRING THESE CLAIMS

The Court turns next to its conclusions of law. As always, the Court must first ensure that it has subject-matter jurisdiction. See Ruhrgas AG v. Marathon Oil Co. , 526 U.S. 574, 583, 119 S.Ct. 1563, 143 L.Ed.2d 760 (1999) ("Article III generally requires a federal court to satisfy itself of its jurisdiction over the subject matter before it considers the merits of a case."). Federal courts may hear only "Cases" and "Controversies." U.S. CONST. art. III, § 2. Courts have interpreted this requirement to mean that a party invoking a court's jurisdiction must have standing. See Friends of the Earth, Inc. v. Laidlaw Envtl. Servs. (TOC), Inc. , 528 U.S. 167, 188, 120 S.Ct. 693, 145 L.Ed.2d 610 (2000). In order to have standing, a plaintiff must establish three elements. A plaintiff must have: "(1) suffered an injury in fact, (2) that is fairly traceable to the challenged conduct of the defendant, and (3) that is likely to be redressed by a favorable judicial decision." Spokeo, Inc. v. Robins , ––– U.S. ––––, 136 S. Ct. 1540, 1547, 194 L.Ed.2d 635 (2016). The standing dispute in this case focuses exclusively on the first factor. To establish injury in fact, a plaintiff must demonstrate an injury that is "concrete, particularized, and actual or imminent." Clapper v. Amnesty Int'l USA , 568 U.S. 398, 409, 133 S.Ct. 1138, 185 L.Ed.2d 264 (2013) (internal quotation marks omitted). The Second Circuit has repeatedly described the injury-in-fact requirement as a "low threshold." John v. Whole Foods Mkt. Grp., Inc. , 858 F.3d 732, 736 (2d Cir. 2017) (quoting Ross v. Bank of Am., N.A. , 524 F.3d 217, 222 (2d Cir. 2008) ); see also WC Capital Mgmt., LLC v. UBS Sec., LLC , 711 F.3d 322, 329 (2d Cir. 2013). "The plaintiff, as the party invoking federal jurisdiction, bears the burden of establishing these elements." Spokeo , 136 S.Ct. at 1547.

The Supreme Court and the Second Circuit have recognized diversion of resources as an injury in fact that may be sufficient to establish organizational standing. In Havens , the Supreme Court held that an organization has standing to sue on behalf of its members if the defendant's illegal acts impair its ability to engage in its projects by forcing the organization to divert resources to counteract those illegal acts. Havens Realty Corp. v. Coleman , 455 U.S. 363, 378-79, 102 S.Ct. 1114, 71 L.Ed.2d 214 (1982). The Court determined that these injuries were sufficiently concrete to be more than the "abstract social interests" not cognizable under Article III. See id. at 379, 102 S.Ct. 1114 ; accord Nnebe v. Daus , 644 F.3d 147, 157 (2d Cir. 2011) ; Ragin v. Harry Macklowe Real Estate Co. , 6 F.3d 898, 905 (2d Cir. 1993). The diversion of resources need not be monetary. See Mid-Hudson Catskill Rural Migrant Ministry, Inc. v. Fine Host Corp. , 418 F.3d 168, 174-75 (2d Cir. 2005) (organization spending time to locate, recruit, manage, train, and supply volunteers sufficient to confer standing). Additionally, when a defendant's actions impede an organization's ability to carry out its daily responsibilities, the plaintiff has suffered an injury in fact. See Centro de la Comunidad Hispana de Locust Valley v. Town of Oyster Bay , 868 F.3d 104, 110 (2d Cir. 2017).

The Courts of Appeals are uniform in recognizing standing in these circumstances. The Seventh Circuit recently explained "that a voting law can injure an organization enough to give it standing by compelling [it] to devote resources to combatting the effects of that law that are harmful to the organization's mission." Common Cause Indiana v. Lawson , 937 F.3d 944, 950 (7th Cir. 2019) ; see also Fla. State Conference of N.A.A.C.P. v. Browning , 522 F.3d 1153, 1164-65 (11th Cir. 2008) (finding standing because "[t]he organizations reasonably anticipate that they will have to divert personnel and time to educating volunteers and voters on compliance with [voting law] and to resolving the problem of voters left off the registration rolls on election day"); Common Cause/Georgia v. Billups , 554 F.3d 1340, 1350 (11th Cir. 2009) (NAACP had standing to challenge a photo ID law based on diversion of resources from regular activities to educating voters about the new requirements and helping them get IDs); Scott v. Schedler , 771 F.3d 831, 836-39 (5th Cir. 2014) (NAACP had standing to challenge failure to provide registration forms to persons visiting benefit offices because NAACP spent additional time on registration drives as a result); Nat'l Council of La Raza v. Cegavske , 800 F.3d 1032, 1040 (9th Cir. 2015) (organizations had standing based on additional resources spent assisting people who should have been registered through state public assistance offices with voter registration); Northeast Ohio Coal. for the Homeless v. Husted , 837 F.3d 612, 624 (6th Cir. 2016) (organization that helped homeless voters had standing to challenge a change in law that required it to overhaul the focus of its voter-education and get-out-the-vote programs); OCA-Greater Houston v. Texas , 867 F.3d 604, 612 (5th Cir. 2017) (upholding organizational standing for non-profit based on injury resulting from extra time spent educating voters about a new voting law).

Common Cause argues it has standing because it has diverted resources as a result of the New York election laws at issue here. As noted above, Common Cause held trainings, spent substantial energy assisting voters, hired new staff, and conducted a large-scale survey directly in response to New York's laws. Common Cause has demonstrated that it would not have conducted these activities but for the State's practices. And it has shown that absent the laws at issue here, Common Cause would have dedicated these resources to other issues. These facts are sufficient to confer organizational standing, and Common Cause has thus met its burden under Article III.

IV. THE ANDERSON-BURDICK FRAMEWORK

The Court turns next to Common Cause's constitutional challenge. Voting is a fundamental right, and is protected by the Fourteenth Amendment's guarantee of "equal protection under the laws." U.S. CONST. AMEND. XIV ; see e.g., Reynolds v. Sims , 377 U.S. 533, 561-62, 84 S.Ct. 1362, 12 L.Ed.2d 506 (1964) ("Undoubtedly, the right of suffrage is a fundamental matter in a free and democratic society."). The right to vote is the fount from which all other rights flow. "Especially since the right to exercise the franchise in a free and unimpaired manner is preservative of other basic civil and political rights, any alleged infringement of the right of citizens to vote must be carefully and meticulously scrutinized." Reynolds , 377 U.S. at 562, 84 S.Ct. 1362.

A balancing test governs the constitutionality of laws regulating the right to vote. The Supreme Court laid out this framework in Anderson v. Celebrezze , 460 U.S. 780, 788-89, 103 S.Ct. 1564, 75 L.Ed.2d 547 (1983), and Burdick v. Takushi , 504 U.S. 428, 433-34, 112 S.Ct. 2059, 119 L.Ed.2d 245 (1992). Anderson - Burdick scrutiny is a flexible test that aims to balance citizens' constitutional right to vote against states' legitimate interests in regulating elections. It requires courts to "weigh ‘the character and magnitude of the asserted injury to the ... rights that the plaintiff seeks to vindicate’ against ‘the precise interests put forward by the State as justifications for the burden imposed by its rule,’ taking into consideration ‘the extent to which those interests make it necessary to burden the plaintiff's rights.’ " Burdick , 504 U.S. at 434, 112 S.Ct. 2059 (quoting Anderson , 460 U.S. at 789, 103 S.Ct. 1564 ); see also Green Party of N.Y. State v. N.Y. State Bd. of Elections , 389 F.3d 411, 419 (2d Cir. 2004).

Under this framework, election laws that impose no burden on the right to vote are subject to rational-basis review. See Ne. Ohio Coal. For the Homeless v. Husted , 696 F.3d 580, 592 (6th Cir. 2012). Laws that severely burden the fundamental right to vote, such as a poll tax, trigger strict scrutiny, and must be "narrowly drawn to advance a state interest of compelling importance." Norman v. Reed , 502 U.S. 279, 289, 112 S.Ct. 698, 116 L.Ed.2d 711 (1992). And "[f]or the majority of cases falling between these extremes, [courts] apply the ‘flexible’ Anderson - Burdick balancing test." Coal. For the Homeless , 696 F.3d at 592 (quoting Burdick , 504 U.S. at 434, 112 S.Ct. 2059 ). The Second Circuit summarized this framework as follows:

Under this balancing test, the State's reasonable and nondiscriminatory restrictions will generally be sufficient to uphold the statute if they serve important state interests. Review in such circumstances will be quite deferential, and we will not require elaborate, empirical verification of the weightiness of the State's asserted justifications. Nonetheless, where the burden imposed by the law is non-trivial, we must weigh the State's justification against the burden imposed.

Price v. New York State Bd. of Elections , 540 F.3d 101, 109 (2d Cir. 2008) (internal quotation marks and citations omitted).

The Supreme Court has squarely rejected "the erroneous assumption that a law that imposes any burden upon the right to vote must be subject to strict scrutiny." Burdick , 504 U.S. at 432, 112 S.Ct. 2059. Indeed, "[s]tates may, and inevitably must, enact reasonable regulations of parties, elections, and ballots to reduce election- and campaign-related disorder." Timmons v. Twin Cities Area New Party , 520 U.S. 351, 358, 117 S.Ct. 1364, 137 L.Ed.2d 589 (1997) ; see also Clingman v. Beaver , 544 U.S. 581, 593, 125 S.Ct. 2029, 161 L.Ed.2d 920 (2005) (same). Since election laws inevitably impose some degree of burden on the right to vote, it follows that "not every limitation or incidental burden on the exercise of voting rights is subject to a stringent standard of review." Bullock v. Carter , 405 U.S. 134, 143, 92 S.Ct. 849, 31 L.Ed.2d 92 (1972).

V. NEW YORK HAS VIOLATED THE RIGHT TO VOTE BY NOT PROVIDING THE INACTIVE LIST, AND MUST DO SO

Common Cause alleges that two of New York's policies violate the fundamental right to vote: (1) the State's refusal to maintain the inactive list at polling locations, and (2) the State's requirement that voters who are not listed on the active list must vote by affidavit ballot. The Court applies the Anderson - Burdick framework to each of these policies. It concludes that both policies burden voters. The former however does not advance any legitimate state interest, while the latter advances several. The State's refusal to provide the inactive list therefore violates the Fourteenth Amendment.

A. New York's Refusal to Provide the Inactive List at Polling Sites Burdens Two Sets of Voters

To start, it is important to identify the two classes of voters relevant to the Court's analysis. The first is those inactive voters who continue to reside at their address of registration; in other words, voters whom the State incorrectly marked as inactive. As noted, this population consists of tens of thousands of New Yorkers. Mark Meredith testified that it contains 45,000 voters at a minimum. Meredith Decl. at 11-14; Meredith Supp. Decl. at 2. And the reason is clear: the core proxies that New York uses to identify inactive voters, data from the Postal Service and the National Change of Address registry, are both overinclusive, capturing thousands of voters who have not moved. The second burdened population relevant to the Court's analysis subsumes the first and is even larger—all New York voters.

Both populations are burdened by New York's refusal to provide the inactive list. The Court begins with the former. In counties that do not provide the inactive list, inactive voters are sometimes not offered affidavit ballots entirely, even though New York law gives them the right to vote by affidavit. Indeed, these voters satisfy all the requirements to vote. They are registered. They are at the correct polling location. They identify themselves to poll workers. Yet because of the overinclusivity of New York's proxies, they find that they are not in the poll book. And because of poll-worker error and confusion, they are not offered affidavit ballots. These individuals are therefore improperly disenfranchised—and thus suffer perhaps the greatest burden a state can impose on a voter.

Even if these voters do get an affidavit ballot, they face substantial delay in doing so. The State requires a lengthy colloquy between a poll worker and a voter when the former cannot find the latter in the poll book. N.Y. Elec. Law. § 8-302(3)(e) ; accord Connolly Decl. ¶ 53; Def. Ex. D, Mohr Decl., ¶ 7. And poll workers often tack on additional questions of their own or call the County Boards. Tr. 84:1-16, 85:21-25 (Martin). Moreover, poll workers are often confused by voters who do not appear in the active registry. These voters can spend significant amounts of time—up to an hour—receiving the affidavit ballot to which they are entitled. See Agro-Paulson Decl.; Goldberg Decl.; Stewart Decl.; Pl. Ex. 55 at 15-18 (Audit Report). It also takes these voters time to complete affidavit ballots. And that time varies with the voter's reading comprehension and eyesight.

The State asserts that poll workers can mitigate these burdens on inactive voters by consulting a map or street finder to look for voters' correct polling place, posting flyers concerning affidavit ballots, giving voters a notice-to-voter form, and calling County Boards. Def. Br. 18-19, 32. As noted, however, few voters testified that poll workers actually used these tools. Valentine testified that there is no requirement that poll workers call election officials to verify voters' registration status when they cannot find them in the poll book. Tr. 150:22-151:18 (Valentine). Indeed, workers may not have the ability to make phone calls at all. See Tr. 92:12-20 (Martin). And these tools, including the street finder, do not assuage the confusion experienced by voters who have not moved from their address of registration, are appearing at the correct polling place, and do not find their name in the poll book. See Meredith Decl. at 21-22; Wolfe Decl. ¶¶ 10-14; Holman Decl. ¶¶ 14-22.

These burdens then have ripple effects, burdening the second relevant group—all New York voters. The idea is simple: delay for some creates delay for all. And Common Cause has demonstrated this occurs across New York's polling sites. As poll workers struggle to process inactive voters, they are unable to assist other voters, whether they appear on the active registry or not. These ripple effects increase the time everyone spends at the ballot box attempting to vote. And they increase the prospect of voters reneging, as this delay will dissuade certain prospective voters from voting entirely. In short, the bandwidth of poll workers is limited, and the opportunity cost of the time spent sorting out the affidavit-ballot process is that other voters wait longer and may not cast a ballot at all. The State's decision not to provide the inactive list therefore imposes substantial burdens.

B. New York Demonstrates No State Interest Justifying Its Policy

The imposition of a burden, however, is not enough for the Court to strike down this practice. Even severe restrictions on the right to vote can be justified by strong state interests. The State points to several purported interests in not providing the inactive voter list at polling locations. The Court addresses these in turn, and concludes that no state interest is advanced by this policy.

First , the State argues that not having the inactive list ensures that people vote in the location in which they are registered. See Def. Br. 21-24. This is certainly an important state interest. Preventing mismatch between where voters vote and where they reside is critical to representative democracy. But the Second Circuit has made clear that states must do more than merely recite important interests to satisfy Anderson - Burdick scrutiny; they must show how their conduct actually advances that state interest. "[T]he fact that the defendants asserted interests are ‘important in the abstract’ does not necessarily mean that its chosen means of regulation ‘will in fact advance those interests.’ " Lerman v. Bd. of Elections in City of New York , 232 F.3d 135, 149 (2d Cir. 2000) (quoting Turner Broadcasting Sys., Inc. v. F.C.C. , 512 U.S. 622, 664, 114 S.Ct. 2445, 129 L.Ed.2d 497 (1994) ).

The State relies heavily on the following theory: voters are "creatures of habit," and so even when they move, they often return to their previous location. Ryan Dep. 90:2-6. New York, however, does not persuasively explain why not listing inactive voters at the polls ensures voters are at their correct polling site. To see why New York's policy does not further this interest, the Court walks through a hypothetical. Assume that Voter A moves from Nassau County to Westchester County. Now consider the following scenarios:

• Voter A is on the active list. First assume that Voter A does not update her voter registration, and none of the State's proxies is triggered. Voter A thus remains on the active registry in Nassau County. On Election Day, Voter A drives back to Nassau County and attempts to vote. She will still be on the active registry in her original Nassau County election district. And she will be able to cast a regular ballot in that district—no matter if the poll worker has the inactive list or not. With or without the inactive list, the result is the same: Voter A votes where she is not supposed to.

• Voter A is on inactive status. Now assume that Voter A's move did trigger a proxy, and the State listed Voter A as inactive. Once again, the absence or presence of the inactive list makes little difference.

• The State does not provide an inactive list. In this scenario, assume that Voter A arrives in Nassau County and finds that she is not on the active list. If there is no supplemental inactive-voter list, the poll worker should ask Voter A where she lives. When Voter A reveals that she has driven an hour from her home, the poll worker should direct her back to Westchester and tell her to vote by affidavit in her new election district. Under the State's universal-transfer policy, that vote will be counted, reactive her as active,

and update her registration with the new address. As discussed above, however, all sorts of problems can arise at this step, delay and confusion among them. It is possible that when the poll worker does not find Voter A in the active registry, she expresses confusion, spends time talking with other poll workers, calls over an inspector, calls the County Board, directs Voter A to the wrong location, or perhaps even offers Voter A an affidavit ballot (which would not be counted).

• The state provides an inactive list. Now assume that the State provides an inactive list. Once again, the poll worker does not find Voter A's name in the active registry. She then checks the inactive list, and finds Voter A there. By this point, the poll worker should know that Voter A lives in Westchester, and direct her back to her new home. With or without the inactive list, the result for Voter A is therefore the same. The State however is concerned that when the poll worker discovers Voter A's name in the inactive registry, she will let her cast an affidavit ballot in Nassau County. But that presupposes that the poll worker did not ask Voter A her new address, which goes against the policies of the State and County Boards. And it presupposes that the poll worker was not trained on how to use the inactive registry or on universal transfer. And even if Voter A casts an affidavit ballot, it would not be counted, thereby not diluting the votes of Nassau County residents. True enough, Voter A's voice would not be heard, but it is not clear that the chance of her receiving an affidavit is any greater in this scenario than if there were no inactive list.

In short, the State has shown no justification for not providing the inactive list benefits a voter who has moved and returns to her polling location. Whether that voter is inactive or active, she receives no benefit from not having the list, and perhaps is harmed by the additional delay and confusion that its absence causes.

Second , the State argues that its affidavit-ballot process is necessary to comply with the National Voter Registration Act. And true enough, the NVRA requires that states engage in list-maintenance activities. The statute requires "each State" to "conduct a general program that makes a reasonable effort to remove the names of ineligible voters from the official lists of eligible voters by reason of ... a change in the residence of the registrant." 52 U.S.C. § 20507(a) ; Def. Br. at 3. But the NVRA merely lays out that requirement and then some outer boundaries of what states cannot do in maintaining their lists. Like many federal election laws, the NVRA leaves substantial discretion to the states. And the statute says nothing about the precise restriction at issue here: not keeping a list of inactive voters at the polling site. Nothing in the NVRA prohibits states from keeping such a list. Indeed, dozens of states in the nation do provide this supplemental list. See Pl. Br. App. A, B (providing states that list inactive voters in poll books on Election Day). The State thus cannot rely on vague assertions about NVRA compliance to state a legitimate interest.

Third , the State argues that not providing the inactive list in fact increases efficiency at the polls "because it gives poll workers, who have limited bandwidth, a binary choice when a voter votes: 1) the voter is on the list; give them a [regular ballot]; or 2) the voter is not on the list; ensure they are in the correct poll site and, if so, offer them an affidavit ballot." Def. Br. 1-2. This argument fails. Assume that a County provides the inactive list at polling locations. Poll workers would still have a limited set of options: If the voter is on the active list, she should be given a regular ballot. If the voter is on the inactive list, she should be given an affidavit ballot. And if the voter is on neither list, she should either be redirected to another polling location or given an affidavit ballot (as the lists may be erroneous or the poll worker may have failed to locate her name). The State's theory about efficiency therefore does not hold up. Moreover, the evidence adduced at trial suggests that the State has it backwards: providing the list in fact benefits poll workers by reducing delay and confusion. Indeed, Martin testified at length to the positive benefits Columbia County has received by providing the inactive list. The limited bandwidth of poll workers, therefore, does not counsel against providing the inactive list—it counsels in favor of it.

The legitimacy of the State's interest is further undercut by evidence from other states. Dozens of other states list inactive voters in the poll books used at precincts on Election Day. PL Br. App. A, B. The State does not dispute that New York is an outlier in this regard, or claim that it faces problems regarding voter movement any different from its sister states. The fact that most states have rejected New York's approach suggests that excluding inactive voters is not required to achieve the State's purported aims. In short, the State does not demonstrate that its policy of not providing a list of inactive voters advances any state interest.

C. New York's Refusal to Provide Inactive Lists at Polling Sites Therefore Fails Scrutiny

Having identified the burdens imposed by, and the interests in favor of, New York's law, the Court can now apply the Anderson - Burdick framework. Under the test, the applicable tier of scrutiny is determined by how heavily the right is burdened. The parties thus spend substantial energy debating the tier of scrutiny that applies here. Common Cause argues that New York's refusal to provide the inactive list disenfranchises some voters, and thus imposes a severe burden and requires application of strict scrutiny. Pl. Br. at 4. The State counters that its policy is a reasonable and non-burdensome regulation, and therefore requires either "rational basis review or, at most, lenient middle-tier review (well towards the deferential end of the spectrum)." Def. Br. at 19.

Given that New York's prohibition on providing the inactive list burdens its voters, rational basis does not apply. But the Court need not decide whether strict or some form of intermediate scrutiny governs, because the law cannot withstand any level of scrutiny. New York's policy burdens voters, and the State provides no legitimate interest to justify that burden. Indeed, when pressed at trial to provide a legitimate interest, the State was repeatedly unable to do so. See, e.g. , Tr. 474:3-12. New York then argued that "the state doesn't necessarily need an interest for not providing a particular piece of information," i.e. the names of inactive voters. Tr. 474:17-20. The State is incorrect. Election decisions that burden the fundamental right to vote must be justified by legitimate state interests. None is provided here. The Court therefore concludes that New York's policy of not providing the inactive voter lists is unconstitutional.

D. The Appropriate Remedy is for New York to Provide the Inactive Voter List

Because New York's refusal to provide the inactive list violates the Fourteenth Amendment, the appropriate remedy is for New York to provide the list. The State may choose whether it wishes to do so by including the names of inactive voters in the registered-voter poll book or by following the example of Nassau and Columbia Counties and creating supplemental lists made available to poll workers at the poll site. Either way, the result is the same: poll workers will be able to identify inactive voters, inform them of their inactive status, and determine where they should vote and whether they should cast a regular or affidavit ballot.

As noted, the record makes clear the benefits that derive from provision of the inactive list at polling locations. The first step poll workers must take when a voter enters a polling site is to determine whether that voter is at the correct location. Without including inactive voters in a supplemental poll book, poll workers are unable to determine whether a potential voter is not listed in the poll book because she (1) is in the wrong polling place, (2) is an inactive voter who is eligible to vote at that polling place, (3) is not registered at all and therefore cannot vote at any polling place, or (4) for some other reason. Trial Tr. 151:13-18 (Valentine) (agreeing that under the State's approach "poll workers have no way of knowing whether a voter arriving at a polling place is not in the poll book because he or she is in inactive status or because he or she never registered to vote, or he or she is not in the poll book for some other reason, correct); Pl. FIF at 27; Def. FIF at 20. Having a supplemental poll book will remove the second category. Poll workers will thus be more likely to offer an inactive potential voter who is at the correct polling site an affidavit ballot. Providing the list will also prevent some affidavit ballots from being rejected for being cast at the wrong location. For example, if Lauren Wolfe was on an inactive list, she likely would have cast her ballot at the right voting location.

The State will face a reduced burden in implementing inactive poll books, as State Boards are increasingly starting to use electronic poll books. Def. FIF at 15. And the Court encourages the State to provide the list of all inactive voters on electronic poll books. Such a list could help not only those voters who arrive to vote in their correct election district, but also those who are inactive and go to the incorrect polling location. And this list would be especially useful given the universal-transfer law; if a poll worker finds that a voter has moved and appears to vote at her new location, the worker can inform the voter that her affidavit ballot in the new location will be counted and update her registration. But the Court does not order the State to do so, because the record suggests that not all County Boards have switched over to electronic poll books, and a printed registry of the State's one-million inactive voters would be impossible to print or use effectively. Def. FIF at 15, Tr. 364:14-365:15 (Meredith) (noting that such a list would be enormous and have "too much information for a poll worker to handle," but stating that the "advent of electronic poll books" may allow the State "to provide a broader list").

VI. NEW YORK HAS NOT VIOLATED THE FUNDAMENTAL RIGHT TO VOTE BY USING AFFIDAVIT BALLOTS

Common Cause also challenges the State's use of affidavit ballots as unconstitutional. The Court rejects this challenge. Although affidavit ballots impose burdens on voters, the State has multiple, weighty interests in using affidavit ballots and not permitting all voters who present themselves at the polls on Election Day to vote by regular ballot.

To be sure, as the Court has discussed repeatedly, affidavit ballots burden voters. They take longer to complete than regular ballots. Tr. 316:20-37 (Brehm); Tr. 154:17-23 (Valentine); Ryan Dep. 51:12-19. For example, completing the affidavit ballot took Holman an additional "[f]ive, ten minutes." Id. 31:2-4. Meredith testified that "the amount of time it would take to fill out an affidavit ballot would depend heavily on someone's reading comprehension ... the type [on the ballot] is small." Tr. 316:25-317:12. In other words, "filling out an affidavit ballot [is not] equally burdensome for all voters," and the time involved will depend on reading level, eyesight, and whether the affidavit ballot is offered in the voter's preferred language. Tr. 252:3-10. And this extra time can create delay for all voters and lead voters to renege.

The State however asserts several legitimate interests to justify its policy of using affidavit ballots. First, the State argues that "[r]equiring inactive voters to vote via affidavit ballot causes the inactive voter to affirm that they are living at the same address and gives election administrators an opportunity to validate the voter's information." Def. Br. 21. As noted, New York has an important interest in ensuring that only voters who are properly registered to vote in a location vote in that location. And it has an important state interest in prohibiting voters from voting in locations where they do not reside. Indeed, these dual interests are two sides of the same coin. Michael Ryan stated that this is "the best function of the affidavit process, because people do move around and they should vote at their new location." Ryan Dep. 89:23-27. He explained that "people, being creatures of habit often don't vote at their new location and go back to their old location" and cast a regular ballot. Ryan Dep. 90:2-6. As discussed, states have a strong interest in ensuring voters vote where they live. And as the State points out, allowing non-residents to vote in an election district would dilute the vote of those who actually reside there. Def. Br. 22-23.

Once again, it is useful to return to the hypothetical Voter A. Assume first that Voter A moved from Nassau County to Westchester County without updating her registration. If she attempts to vote at her new home, she will not be on the active or inactive lists. She can then inform the poll worker that she recently moved and cast an affidavit ballot. Under the State's universal-transfer policy, that affidavit ballot will be counted and update her registration. And if Voter A attempts to vote in Nassau County, the poll worker should not offer her an affidavit ballot. But even if the worker does, the State will reject her affidavit, as she no longer resides in Nassau County. In either situation, a regular ballot would not serve the State's interest. If she voted in Westchester by regular ballot, she would remain inactive in Nassau County and her registration would not be updated to active status with her new address. And if she voted by regular ballot in Nassau County, her vote would count, diluting the vote of that election district's residents. Affidavit ballots therefore advances the State's legitimate and important interest in ensuring that voters who move do not cast valid ballots in their old election districts.

Second, affidavit ballots also serve as a check on poll-worker error. Several state officials testified that poll workers sometimes fail to locate active voters in the poll book, even though that voter is properly registered and in the correct location. Tr. 88 (Martin); Ryan Dep. 50:14-51:19 (noting that this "is a common error"). Poll workers are directed to give such voters an affidavit ballot, and their vote should be counted. Id. 51:12-19. In the absence of an affidavit ballot, however, a voter erroneously not found in the poll book may not be permitted to vote at all. Affidavit ballots thus serve as an error-correction mechanism and allow such individuals to cast a ballot. Ryan explained that affidavit ballots are necessary in these "circumstances of poll worker error ... because it provides [these voters] an avenue to exercise the franchise." Ryan Dep. 126:13-127:7.

Third, affidavit ballots also help New York effectuate its state interest in regulating the primary system. New York has closed primaries, meaning that only registered voters who have indicated they are members of a particular party can vote in that party's primary. Tr. 312:13-18 (Meredith). Yet voters at times attempt to vote in the primaries for parties for which they are not registered. Tr. 313:11-16 (Meredith); Ryan Dep. 175:3-177:17. The voter will not appear in the poll book, but if she insists on voting, she is to be given an affidavit ballot. Ryan Dep. 48:8-50:13. The affidavit ballot can thus be a tool to sort out which voters are properly registered; poll workers can provide voters who do not appear on the party's list with a ballot, that voter can cast an affidavit ballot, and then the State can later determine whether the voter is in fact a registered member of the relevant party. Tr. 313:17-22 (Meredith).

Fourth, affidavit ballots help the State comply with federal law. The NVRA "erect[s] a complex superstructure of federal regulation atop state voter-registration systems." Arizona v. Inter Tribal Council of Ariz., Inc. , 570 U.S. 1, 5, 133 S.Ct. 2247, 186 L.Ed.2d 239 (2013). The statute has two primary goals: increasing voter registration and removing ineligible persons from the States' voter registration rolls. 52 U.S.C. § 20501(b) (listing these purposes).

The Supreme Court analyzed the NVRA at length in Husted v. A. Philip Randolph Institute , ––– U.S. ––––, 138 S. Ct. 1833, 201 L.Ed.2d 141 (2018). The plaintiffs in that case challenged Ohio's process for identifying voters who may have moved and removing them from the rolls as violative of the NVRA. The Court explained that Congress passed the NVRA out of a concern that states were removing voters who they thought had moved from the rolls without providing those voters any written notice. Id. at 1838. The NVRA thus lays out certain requirements states must satisfy before removing a voter from the rolls because of a perceived change in residence. Id. §§ 20507(b), (c), (d). The Court explained that "the provision of the NVRA that directly addresses the procedures that a State must follow before removing a registrant from the rolls on change-of-residence grounds, provides that a State may remove a registrant who ‘(i) has failed to respond to a notice’ and ‘(if) has not voted or appeared to vote ... during the period beginning on the date of the notice and ending on the day after the date of the second general election for Federal office that occurs after the date of the notice’ (about four years)." Id. at 1841 (quoting 52 U.S.C. § 20507(d)(1)(B) ). Voters who do not respond to the confirmation notice can be removed from the rolls only if they fail to vote in two general federal elections and do not otherwise inform the state they have not moved. "Only if the registrant fails to vote during that period and does not otherwise confirm that he or she still lives in the district (e.g., by updating address information online) may the registrant's name be removed." Id. at 1839-40 (citing §§ 20507(d)(2)(A); (d)(1)(B), (3)."

The Court recognized that "States take a variety of approaches" in complying with the NVRA's requirements. Id. at 1839. In other words, the NVRA sets the floor for how states conduct list maintenance; states have discretion to choose how precisely to follow its requirements. See id. But the Court made clear that compliance is not optional: the NVRA requires states to "conduct a general program that makes a reasonable effort to remove the names" of voters who are ineligible "by reason of" death or change in residence. § 20507(a)(4). And the statute states that "[n]ot only are States allowed to remove registrants who satisfy these requirements, but federal law makes this removal mandatory ." § 20507(d)(3) (emphasis added).

After the NVRA was passed, New York enacted its affidavit-balloting scheme to bring the state into compliance. The NVRA's legislative history expressly contemplates the use of provisional ballots for voters who do not respond to states' mailings, like inactive voters in New York, as a mechanism for compliance. The House Report provides: "However, while [the NVRA] sets out where an individual may vote, it is silent as to how that individual may be permitted to vote. Under certain circumstances it would be appropriate, and in compliance with the requirements of this Act, to require that such a person vote by some form of provisional ballot ." H.R. REP. 103-9, 17-18, 1993 U.S.C.C.A.N. 105, 121-22 (emphases added). Indeed, another federal statute, the Help America Vote Act, directly envisions states using provisional ballots and provides various requirements these ballots must comply with. See 52 U.S.C. § 21082 (laying out "Provisional voting requirements"). Federal law thus imposes a floor for how states can conduct list maintenance and permits the use of provisional ballots. New York's affidavit-ballot process is supported by a state interest in following these federal directives.

Under the Anderson - Burdick framework, the State burdens voters through its use of affidavit ballots. But New York has weighty interests justifying that practice: affidavit ballots ensure voters are in the correct polling location, comply with federal law, and help regulate its closed-primary system. No matter what tier of scrutiny applies, those interests are sufficient for the law to survive review. In sum, New York has not violated voters' fundamental right to vote by requiring some voters to use affidavit ballots.

VII. COMMON CAUSE HAS ALSO PROVEN SEVERAL DE FACTO NVRA VIOLATIONS

Section 8 of the NVRA addresses the procedures a state must follow before it may remove an eligible voter from the official list of registered voters. 52 U.S.C. § 20507. In relevant part, Section 8 provides that a state may not remove a registered voter from the official list of registered voters based on the belief that a voter has changed residence unless (1) the voter confirms in writing that he or she has moved to a new jurisdiction or (2) the voter has failed to respond to a notice seeking confirmation that the voter continues to reside in the jurisdiction and the voter fails to vote in two consecutive general elections for federal office. 52 U.S.C. § 20507(d). The first half is not relevant here. As applied to New York, the second half captures inactive voters. Under the State's regime, most inactive voters have failed to respond to confirmation notices; if they had, they would have been relisted as active. The State therefore cannot disenfranchise these voters until they fail to vote in two consecutive federal general elections. Id.

Common Cause argues that New York effectively removes inactive voters from the official list. In its earlier Order, the Court identified several allegations that, if proven true, could support a de facto removal claim: (1) state poll workers "routinely" informed inactive voters that they were not registered to vote; (2) state poll workers failed to inform inactive voters that they may vote by affidavit ballot; (3) state poll workers failed to offer inactive voters affidavit ballots; (4) state poll workers offered voters affidavit ballots only at their "insistence." See Dkt. No. 58 at 25-26.

Common Cause has identified three voters as to whom the State violated its obligations under the NVRA. Stephanie Goldberg was listed as inactive and should have been offered an affidavit ballot at the polling site, but poll workers failed to provide her one. See Goldberg Decl. ¶¶ 15-21. Goldberg was thus disenfranchised before she failed to vote in two federal general elections, thereby violating the NVRA. Moreover, Angela and Denise Roberts were listed as inactive and cast affidavit ballots, but the State rejected their ballots for no legitimate reason. See D. Roberts Dep. 51:5-17; D. Roberts Decl., ¶¶ 11-20; A. Roberts Decl., ¶¶ 25-35; Tr. 107:13-108:7 (Connolly). State officials provided no explanation for why these affidavit ballots were not counted, and thus these voters were improperly disenfranchised in violation of the NVRA. See Ex. P230, Toombs Decl. ¶ 12 (stating "it appears that [Denise Roberts'] affidavit ballot cast in 2016 was not counted. It should have been. This was an error and we are very sorry that it occurred."). In reference to Stephanie Goldberg, Denise Roberts, and Angela Roberts, the State concedes that "Plaintiff does submit three declarations where there was clear error." Def. Br. at 33.

The Court thus grants Common Cause's request for a declaration that "Defendants have violated Section 8 of the NVRA by denying eligible voters the right to vote based on a purported change in residence without following the procedures set forth in 52 U.S.C. § 20507" as to these three individual voters. The Court does not address Common Cause's further request for injunctive relief and monitoring under the NVRA, as any such relief is redundant with the relief ordered under the Fourteenth Amendment.

VIII. CONCLUSION

The Court concludes that New York's refusal to provide inactive lists at polling locations violates the Equal Protection Clause. The State is therefore ordered to provide the names of inactive voters registered to vote in a particular election district to the poll workers of that election district. The Court further concludes that the State's affidavit-ballot process does not violate the Fourteenth Amendment. And the Court identifies three discrete voters as to whom the State violated the National Voter Registration Act.

In its Amended Complaint, Common Cause requested the award of "the costs and disbursements incurred in connection with this action, including, without limitation, their reasonable attorneys' fees, expenses, and costs." Dkt. No. 72 at 18. If Common Cause continues to seek this award, it is ordered to detail its expenses and submit letter briefing justifying its request no later than three weeks from the date of this Order. The State may submit its briefing regarding Common Cause's request no later than three weeks after that. Common Cause may then submit a reply no later than one week after that.

The Court retains jurisdiction to monitor the State's compliance with the terms of this Order. The Clerk of Court is respectfully ordered to enter judgment. This also resolves Dkt. No. 144.

SO ORDERED.