7

THE NEW WARS OVER THE VOTE

In the years since the Florida debacle of 2000 we have witnessed a partisan war over election rules.… Florida mainly taught political operatives the benefits of manipulating the rules, controlling election machinery, and litigating early and often. Election law has become part of a political strategy.

—RICHARD L. HASEN, 2012

The results that poured in on election night November 7, 2000, showed that George W. Bush was poised to become the first Republican presidential candidate in U.S. history to sweep every southern and border state—except perhaps for one crucial state. At 7:50 P.M., NBC called Florida for Al Gore, which virtually guaranteed him an Electoral College majority. Other major news networks followed NBC’s lead. As election night dragged on, however, the results from Florida looked less certain, prompting the news outlets to recant their call. At 2:16 A.M., Fox News Channel declared Bush the winner in Florida and the next president of the United States; other outlets followed in lockstep. Vice President Gore then conceded the election. By 4:00 A.M. the media for a second time recanted their call after Bush’s lead diminished; by then, Gore had retracted his concession, a first in American history. The recount battle in Florida would soon begin, marking the beginning of what constitutional scholar Richard L. Hasen has called the “voting wars” in the United States.¹

As dawn broke in Florida, less than two thousand votes separated the candidates, a difference of fewer than 0.1 percent of some 6 million votes cast. Under Florida law, a margin of 0.5 percent or less triggers an automatic recount. The following day, November 9, the Gore campaign made a crucial mistake. It requested a hand count of presidential ballots not statewide but in four Florida counties only. The Gore campaign sought the recount because election officials in Florida had rejected many thousands of ballots cast by punch-card voting machines that did not seem to properly record a discernable vote—the so-called undercounted votes. Jeb Bush, Florida’s then-governor and the candidate’s brother, recused himself from the recount process, but Secretary of State Katherine Harris did not, even though she had worked as the co-chair of Bush’s Florida campaign.²

With both sides pursuing litigation in state and federal court to extend or halt the recount, Bush rejected Gore’s new proposal for a statewide recount. Secretary Harris had announced that she would certify the results of the election on November 18, regardless of the progress of the recount. But a state court ruling allowed the hand recount to continue, and the Eleventh Circuit Court of Appeals rejected a motion by the Bush team to stop this recount on constitutional grounds. On November 20, nearly two weeks after election day, the Florida Supreme Court heard arguments on whether Harris should consider hand-recounted ballots before certifying the election results, with Bush still ahead. The court unanimously ruled that state election officials must include the results of manual recounts in Florida’s presidential tally, with a court-ordered deadline of November 26 or early on November 27.³

In majority-Democratic Miami-Dade County, Florida’s largest jurisdiction, a partial recount of slightly more than one-fifth of disputed ballots netted Gore 157 votes. With Bush clinging to a precarious lead of just a few hundred votes, it became clear that a full recount might add enough votes to Gore’s tally to swing Florida and the presidency to him. On November 22, the Bush team executed a brilliant maneuver to halt the Miami-Dade recount. The Bush campaign decided to stop cold the recount in Dade County. Its high command in Austin, Texas, worked with leaders of the county’s conservative Cuban American community to muster 150 demonstrators, including 40 to 60 out-of-state Republicans, to protest the Dade County recount. Al Gore’s supporters were nowhere to be found. Conservatives owned the streets. Led by activists flown into the state, including Ken Mehlman, national field director of the Bush campaign, and employees of national GOP senators and representatives, demonstrators stormed the county’s election office, disrupted the ballot counting, and influenced an intimidated canvassing board to halt the recount. What later become known as the “Brooks Brothers riot,” after the imported participants’ corporate, upscale attire, had worked to perfection.⁴

The Bush lawyers simultaneously pursued a legal strategy to halt the recount. They filed a certiorari petition, which asked the U.S. Supreme Court to reverse the decision of the court of appeals that had denied their motion to stop the recount. On November 26, despite ongoing litigation, Secretary Harris certified George Bush the winner in Florida by 537 votes. Governor Jeb Bush signed the Certificate of Ascertainment designating twenty-five Florida electors pledged to George W. Bush and sent the document to the National Archives as required.

Gore’s lawyers then filed an election contest action under Florida Election Code, challenging the vote counts in Palm Beach, Miami-Dade, and Nassau Counties. State judge N. Sanders Sauls rejected Gore’s claims, but on appeal, the Florida Supreme Court in a four-to-three ruling on December 8 reversed his decision and ordered a manual recount of undervotes that covered most counties in the state. Crucially, the court’s ruling did not include overvotes, where a ballot contained more than one mark for president, even if one of the two marks was a write-in for the same candidate.

On December 9, Florida began its statewide manual recount of the undervoted ballots, while the Florida Supreme Court and Eleventh Circuit Court denied the Bush team its stay to stop the recount. However, a divided U.S. Supreme Court issued a stay, halting the recount. Then on December 12, the U.S. Supreme Court issued its ruling in Bush v. Gore. By a five-to-four decision along ideological lines, the Court overturned the Florida Supreme Court ruling and stopped the manual recount of undervoted ballots. The U.S. Supreme Court ruled, “Because it is evident that any recount seeking to meet the Dec. 12 date will be unconstitutional … we reverse the judgment of the Supreme Court of Florida ordering the recount to proceed.” The decision continued, “It is obvious that the recount cannot be conducted in compliance with the requirements of equal protection and due process without substantial additional work.” With the recount frozen and Bush ahead by 537 votes, he officially won the state of Florida and became president of the United States by a one-vote majority in the Electoral College.⁵

The backstory in Florida struck a familiar theme: the suppression of African American votes. Studies showed that election officials had eliminated allegedly disenfranchised felons from the voter rolls with a flawed list that overcounted African Americans, who voted 93 percent Democratic, and undercounted Hispanics, who usually voted Republican in Florida. Florida officials had also rejected as invalid some 180,000 ballots—3 percent of votes cast—either because a presidential vote could not be discerned or the ballot appeared to register more than one vote. Analysis showed that election officials discarded as invalid more than one out of every ten ballots cast by African Americans, compared to less than one out of every fifty ballots cast by whites, who voted 57 percent for Bush. Two-thirds of the allegedly invalid ballots cast by black voters were not undervotes, where a vote could not be accurately determined, but uncounted overvotes, where more than one candidate seemingly was marked on the ballot. Officials rejected overvotes even if a voter punched in Al Gore and then wrote in Al Gore to assure that there would be no mistaking their intent. A study of overvotes by University of Michigan political science professor Walter R. Mebane, Jr., “estimated the number of true votes that were not counted because they were recorded as overvotes.” He found that Gore lost 46,465 votes, compared to just 10,939 for Bush.⁶

Racial disparities in rejected ballots persisted when controlling for education, income, poverty, literacy, ballot design, and voter technology. If ballots cast by blacks had been rejected at the same rate as ballots cast by whites, more than fifty thousand additional black votes would have been counted in Florida and Gore would have easily won Florida’s election and the presidency.⁷

The postelection mess reflected the legacy of a fragmented, eighteenth-century political system that deferred to the states, and then to more than ten thousand localities, the conduct of elections. Each of Florida’s sixty-seven counties in effect administered its own election, deciding the design of ballots, the technology for casting votes, the training of poll workers, and the procedures for manual recounts. As a result, voters across Florida’s counties did not have an equal opportunity to have their votes counted in the election. Differences in voter technology and procedures partially but by no means fully accounted for racial disparities in rejected ballots. African Americans were more likely than members of other racial groups to reside in counties using defective punch-card technology. However, vast racial disparities persisted within counties using the same voting machines. Technology alone is not the answer to racial disparities in ballot rejection.

Florida’s counties used five distinct voting technologies implemented by twelve different kinds of machines made by seven different manufacturers. Among the sixty-seven counties, twenty-six used optical scanning machines tabulated by precinct, and fifteen used such machines tabulated centrally. Twenty-four counties used punch-card machines; one county used paper ballots, counted by hand; and another used old-fashioned lever machines. The punch-card counties accounted for 60 percent of all votes cast but 89 percent of undervotes and 78 percent of overvotes.⁸

Ballot design, with two major party candidates and eight minor party and independent candidates competing for president along with candidates for state and local offices, daunted county officials. These challenges produced a perfect storm in Palm Beach County. Election officials designed the so-called butterfly ballot, which put the presidential candidates on two separate pages, with the punch holes between the pages, hence the butterfly design of two wings and a spine between them.

Instead of having Bush and Gore as their first two choices, voters, many of them elderly retirees and loyal Democratic voters, confronted a confusing ballot with the punch-card hole for Reform Party candidate Pat Buchanan placed between Bush on top and Gore in the third position. Only by looking at the second page and figuring out the correspondence between names and punch holes could voters discern that punching the second hole was a vote for Buchanan, not Gore. Many voters who clearly intended to vote for Gore voted instead for Buchanan or for both Gore and Buchanan, thereby invalidating their ballots. A postelection analysis by the Palm Beach Post found that the “visually challenging ballot design” in Palm Beach County “turned an estimated 2,800 would-be Al Gore voters into Pat Buchanan voters,” more than enough to reverse the statewide election results in Gore’s favor.⁹

The conclusion is inescapable that Florida’s 2000 presidential election certified the wrong winner and elected the wrong president. Yet the vast racial disparities in rejected ballots, which cost Gore more than fifty thousand votes, remains the great underreported scandal of the twenty-first century. It was only lightly covered in the press, and even civil rights groups like the Urban League and the NAACP failed to publicize the scandal or investigate its cause, beyond citing outmoded voting technology. It was the same old story of blacks in the South denied their full voting rights. Imagine the unending cries of voter fraud that would have issued from conservative pundits and media if Florida officials had rejected more than one out of ten ballots cast by whites, compared to just one out of fifty ballots cast by African Americans.

In the wake of the fiasco in Florida, Congress enacted with overwhelming bipartisan support the Help America Vote Act (HAVA) of 2002. The act established an Election Assistance Commission to provide the states with voluntary guidelines on voting technology and administration. HAVA requires any first-time voter who registers by mail to show current and valid photo or non-photo identification at the polls. It authorizes voters identified by local officials as ineligible, for example because officials could not find them on registration rolls, to cast provisional ballots. Election officials must count the ballots of people later determined as eligible. In the 2016 election, voters cast nearly 2.5 million provisional ballots, 62 percent of which were fully counted and 9 percent partially counted, saving some 1.8 million votes.¹⁰

The act provided federal funds to states for upgrading voting machines, registration databases, voter education, poll worker training, and disabled access. In the decade following HAVA’s enactment, more than three-quarters of the jurisdictions across the nation replaced the voting equipment they had used in 2000. Still, the act’s voluntary guidelines did not impose order or consistency on local jurisdictions, and the initial HAVA funding for the states has faded away. It also did not impact the partisan administration of elections by Republican and Democratic officials in state and local governments, a problem that was so prominently on display in the Florida recount.¹¹

At the state level, other reforms supplemented HAVA. By 2017, reforms to expand access to the vote included early voting prior to election day (33 states), voter registration during the early voting period or on election day (16 states), online registration (38 states), absentee voting without an excuse (27 states), preregistration by individuals aged sixteen or seventeen (18 states), and all-mail elections (3 states).

As in the post–Civil War Reconstruction era, with the voting rights revolution came the reaction. In a book published in 1967, a year before his assassination, Reverend Martin Luther King warned, “For the Negro, there is a credibility gap he cannot overlook.… Each step forward accents an ever-present tendency to backlash.” In our time, it is white Republicans, not white Democrats, who have made voting more difficult, especially for minorities burdened by a history of discrimination and disparities in income, education, housing, health, and access to vehicles and computers. Once again, patterns of racially polarized voting established the political motivations for restricting minority voting. In a reversal of nineteenth-century voting patterns, Republicans are now dependent on older and white voters and Democrats on younger and minority voters. Despite some variations, these voting patterns are remarkably similar in states and localities in all regions of the nation. Thus, it is today’s Republicans, like yesterday’s southern Democrats, who have taken the lead in enacting measures to restrict voting, especially by minorities and young people.¹²

The reaction against expanded voting opportunities began through laws requiring some form of identification for voting at the polls, which could include photo identifications, such as driver’s licenses and passports, or non-photo IDs, such as utility bills or bank statements. From 2001 to 2004, seven states passed voter identification laws, most with governments controlled by Republicans. Then in 2005, Indiana raised the stakes by enacting a strict law requiring government-issued photo-only identification as a prerequisite for voting. With limited exceptions, a voter could not cast a ballot at the polls without presenting such ID. Non-photo identification or photo identification from nongovernmental sources such as private employers did not count. Upon personal application to a Bureau of Motor Vehicles office, voters could obtain a free photo ID card, provided they presented documentary proof of residence and U.S. citizenship. Voters lacking such identification could cast a provisional ballot, which would count only if the voter presented an authorized identification to election officials within ten days after the election. Voters could also sign an affidavit affirming that they could not obtain a photo ID because they are indigent or have legitimate religious objections to being photographed. This was the first state law in U.S. history that barred eligible registered voters from casting a ballot unless they could meet photo identification requirements.¹³

Plaintiffs led by William Crawford, a Democratic member of the Indiana House of Representatives, challenged Indiana’s law prior to its implementation. They did not present evidence that the law discriminated against minorities in violation of the Voting Rights Act or the Fourteenth Amendment. Nor did they present evidence showing that Indiana citizens could not meet the photo identification requirements. Rather, plaintiffs claimed that on its face, the law violated the fundamental right to vote and the free speech clause of the First Amendment.

Given the lack of an explicit right to vote in the Constitution, the Seventh Circuit Court of Appeals and the U.S. Supreme Court declined to subject the law to the strict review required for constitutional rights. In upholding the law, both courts applied a more lenient test that balanced the rights of voters against the interests of the state. The Supreme Court opinion, written by the generally liberal justice John Paul Stevens, upheld the law even in the absence of a showing by the state of instances of the voter fraud that the law was allegedly designed to punish and deter. Stevens wrote, “Because Indiana’s cards are free, the inconvenience of going to the Bureau of Motor Vehicles, gathering required documents, and posing for a photograph does not qualify as a substantial burden on most voters’ right to vote, or represent a significant increase over the usual burdens of voting.” He reasoned that “while the record contains no evidence that the fraud SEA 483 addresses—in-person voter impersonation at polling places—has actually occurred in Indiana, such fraud has occurred in other parts of the country, and Indiana’s own experience with voter fraud in a 2003 mayoral primary demonstrates a real risk that voter fraud could affect a close election’s outcome.” The state, he added, also had an interest in efforts “to improve and modernize election procedures criticized as antiquated and inefficient,” and “in protecting public confidence in elections.”¹⁴

The enactment of voter photo ID laws in Republican states accelerated after the Crawford decision and the election of America’s first African American president, Barack Obama, in 2008. According to exit polls, Obama garnered 95 percent of the black vote nationwide and 67 percent of the Hispanic vote, compared to only 43 percent of the white vote. Blacks and Hispanics accounted for 22 percent of the 2008 presidential electorate, compared to just 15 percent in 1996. Obama won 66 percent of voters aged eighteen to twenty-nine, compared to 49 percent of voters over forty-four.

From the advent of the Obama administration through 2016, eleven states adopted strict photo identification laws. Nine had Republican-controlled legislatures and governors. In Arkansas, Republican legislators overrode the veto of a Democratic governor, and in Mississippi voters enacted the law in a statewide referendum. Most of these states departed from the Indiana precedent by prohibiting the use of various forms of government-issued IDs, such as employee or public college IDs, acceptable in Indiana. Most did not provide for an affidavit alternative to the ID requirement.¹⁵

Advocates justified adoption of voter photo ID laws by raising once again the specter of voter fraud, in this case voter impersonation at the polls. However, in the modern era, studies have debunked the notion of widespread voter fraud corrupting American elections. A nationwide 2011 study conducted by the Republican National Lawyers Association and designed to root out as much voter fraud as possible found only 5 cases nationwide of prosecutions or convictions for voter impersonation since 1997, and fewer than 350 cases of voter fraud of any kind out of many hundreds of millions of ballots cast in primary and general elections across the United States. A more inclusive study by News21, a national reporting project of eleven universities, that considered reported allegations of voter fraud, whether prosecuted or not, uncovered only 2,068 alleged fraud cases and just 10 allegations of voter impersonation from 2000 to mid-2012.¹⁶

Concerted efforts by Republican officials in several states to uncover voter fraud have similarly come up short. Like in Indiana, officials defending photo voter ID laws in Wisconsin and Pennsylvania admitted they could not cite any examples of illegal voter impersonation. Ironically, during the 2016 campaign, candidate Donald Trump charged that Philadelphia was a hotbed of voter fraud. In North Carolina, during debates over the adoption of a voter photo ID law, the state board of elections reported that its study of voter fraud had uncovered only two cases of voter impersonation that “merited a referral to the district attorney’s office,” out of more than 15 million ballots cast from 2000 to 2012.¹⁷

The lack of evidence for voter impersonation or other fraud is not a matter of the difficulty of detection. Three recent political science studies of voter impersonation have relied on methods that do not turn on allegations or prosecutions, thus responding to the claim that such fraud is hard to detect or verify. These studies, which are based respectively on anomalies in voter turnout records, survey methods designed to detect illicit behavior, and the matching of individual voter files with death certificates (to test allegations of voting the graveyards) confirmed that voter impersonation fraud was virtually nonexistent in American elections.¹⁸

Comprehensive postelection investigations have likewise demonstrated a lack of voter impersonation or voter fraud in general. One such investigation took place in the state of Maryland, which has no voter identification requirements and no signature requirement at the polls. In 1994, Republican gubernatorial candidate Ellen Sauerbrey alleged that fraudulent votes accounted for the 5,993-vote victory of Democrat Parris Glendening. As the state of Maryland’s consultant on voting rights, I was asked by Attorney General Joseph Curran to determine whether there was any truth to Sauerbrey’s claims. If there was voter fraud in Maryland, Curran said he wanted to know about it. Although investigation uncovered some unintentional errors by election officials, there was not a single fraudulent vote among the more than 1.4 million ballots cast in the election.

After the presentation by plaintiffs in state district court likewise failed to uncover a single illegal voter, the trial judge, Raymond G. Thieme, who disclosed that he voted for Sauerbrey, dismissed her lawsuit. Her lawyers did not appeal Thieme’s decision. Subsequent investigations by the state prosecutor, the U.S. attorney, and the FBI found no voter fraud, only a “few irregularities” that were “the result of faulty voting machines that didn’t record votes, election judges’ failure to check off voters and people voting in the wrong precinct because they failed to change their addresses.”¹⁹

In Wisconsin, after John Kerry defeated George W. Bush in the state’s 2004 presidential election, Republicans charged that Democrats had benefited from widespread voter fraud. Immediately after the election, Republican assembly speaker John Gard charged that Milwaukee mayor Tom Barrett “has got to be embarrassed about what happened in Milwaukee. You’ve got thousands of addresses they know don’t exist.” The speaker said that “Democrats and Republicans alike should be concerned about the incredible problems we had across this state.”²⁰ After investigating these allegations in Wisconsin, which had no voter ID law at the time, prosecutors charged only 14 out of 3 million voters in the election with voter fraud. None of the charges involved voter impersonation. Ten of the fourteen cases alleged illegal voting by felons, and the remaining four alleged double voting. None of the double-voting cases resulted in a conviction.²¹

At any given moment, there are deceased people whose names have not yet been removed from state registration rolls. However, studies refute the claim by President Donald Trump and others that fraudsters have voted in the name of the dead. In the Maryland election dispute of 1994, we found that all of Sauerbrey’s alleged dead voters were the result of clerical errors or mismatches between live voters and dead people. We knew that Sauerbrey’s allegations had fallen apart when alleged dead voters arose from the grave to walk and talk, including some who said they voted for Sauerbrey.²²

In a 2012 letter to the U.S. Department of Justice, Republican attorney general Alan Wilson of South Carolina claimed that 953 ballots were allegedly cast by voters listed as dead in prior elections during the previous six years. The South Carolina State Election Commission informed Wilson that it had the resources to investigate 207 cases of alleged dead voters from the most recent 2010 election. The commission found that 106 cases were attributable to clerical errors; 56 cases were the result of erroneous data-matching of voters to dead people; 32 cases were “voter participation errors,” including stray marks on lists erroneously indicating that the individuals in question had voted; 3 cases involved absentee ballots filed by voters who later died before Election Day; and 10 cases contained “insufficient information in the record to make a determination.” Not a single case involved a verified example of anyone voting in the name of a deceased person.²³

Given the spate of post-2008 voter photo ID laws enacted almost exclusively by Republican-controlled state governments, Justice Stevens, who wrote the Supreme Court’s opinion in the Crawford case, has since lamented that although the plaintiff’s case was evidentially weak, “we ended up with a fairly unfortunate decision.” He noted that the Court simply did not have before it enough information about the law’s impact, and that if he had applied what he independently knew about photo voter ID laws, he might have reached a different decision. Judge Richard A. Posner, one of the nation’s most respected conservative jurists, who wrote the Seventh Circuit opinion in Crawford, said in 2013, “I plead guilty to having written the majority opinion” in the Crawford case, adding that photo voter ID law is “a type of law now widely regarded as a means of voter suppression rather than of fraud prevention.” He explained, “There’s always been strong competition between the parties, but it hadn’t reached the peak of ferocity that it’s since achieved.… One wasn’t alert to this kind of trickery, even though it’s age old in the democratic process.”²⁴

In his dissenting opinion regarding the Seventh Circuit’s five-to-five vote against holding a full court review of the challenge to Wisconsin’s voter photo ID law, Judge Posner wrote that “there is only one motivation for imposing burdens on voting that are ostensibly designed to discourage voter-impersonation fraud, if there is no actual danger of such fraud, and that is to discourage voting by persons likely to vote against the party responsible for imposing the burdens.” He added, “It is conservatives who support them [photo voter ID laws] and liberals who oppose them. Unless conservatives and liberals are masochists, promoting laws that hurt them, these laws must suppress minority voting and the question then becomes whether there are offsetting social benefits—the evidence is that there are not.”²⁵

States did not limit restrictive measures to voter photo ID laws but also enacted other legislation that restricted access to registration and voting. These measures, primarily but not exclusively enacted in Republican-controlled states, included reduced early voting, limitations on registration opportunities, and tightened laws that barred voting by former felons.

In 2013, after Republicans gained unified control over North Carolina’s state government, they enacted along party lines omnibus legislation that bundled together most of these new restrictions. The full legislation, known as the Voter Information Verification Act (VIVA), imposed collectively the most restrictive voter photo identification requirements in the nation, with several other limitations on voting and registration. After the adoption of VIVA, an analysis for the civil rights research group Institute of Southern Studies concluded, “number of states with voting laws that are more restrictive than North Carolina’s: 0.”²⁶

North Carolina adopted its voter restrictions at a time of extreme partisanship and voting that fell along racial lines. According to a 2014 study by the Pew Research Center, in the two decades since the epic Republican victories in 1994, “the partisan divide among the American people [has] widened considerably.” Pew found that “the median Republican is now more conservative than 94% of Democrats, compared with 70% twenty years ago. And the median Democrat is more liberal than 92% of Republicans, up from 64%.” Partisan antipathy has similarly soared. The share of Republicans with “very unfavorable” opinions of Democrats rose from 17 to 43 percent, and the share of Democrats with very unfavorable opinions of Republicans from 16 to 38 percent. Political activity is highest among the most extreme conservatives and liberals. In Congress, as already noted, political polarization has come close to reaching the mathematical maximum in voting in both the U.S. House and Senate.²⁷

Political views are highly correlated with race. Pew Research found that in 2014 party affiliation, Republicans led Democrats by 49 to 40 percent among whites but trailed Democrats by 80 to 11 percent among blacks, 65 to 23 percent among Asians, and 56 to 26 percent among Hispanics.²⁸ Voting closely tracks these partisan affiliations. Presidential exit polls for 2012 showed that Republican Mitt Romney won 59 percent of the white vote but only 6 percent of the African American vote, 26 percent of the Asian American vote, and 27 percent of the Hispanic vote. In 2016, even without an African American Democrat on the ballot, Republican Donald Trump won 57 percent of the white vote but only 8 percent of the African American vote, 27 percent of the Asian American vote, and 28 percent of the Hispanic vote.

Republicans in North Carolina enacted VIVA shortly after the 2012 presidential election, in response to the growing presence of minorities, especially overwhelming Democratic African Americans, in North Carolina’s electorate. From 2004 until passage of VIVA in July 2013, African Americans in North Carolina had increased their share of registered voters from 19.4 to 22.5 percent; members of all minority groups combined increased from 22.1 to 29.0 percent. Voting in North Carolina was even more polarized than nationally. Exit polling data for presidential and other statewide elections from 2004 through 2012 shows that African Americans supported Democratic candidates at near unanimous levels, whereas two-thirds of whites on average backed Republicans. The racial disparity in party voting far exceeded other differences in sex, age, education, and income. In 2008, Democrats had won North Carolina’s presidential election for the first time since 1976, and in 2012 Democrats came within two percentage points of a second consecutive victory. Republicans in North Carolina could not reverse demography and expand the white share of the state’s electorate, but they could attempt to constrain the minority vote.²⁹

VIVA sparked a legal challenge by the NAACP and the U.S. Department of Justice. Rosanell Eaton, a black woman who grew up in Jim Crow North Carolina during the 1920s and 1930s, joined the litigation as a private plaintiff. Eaton had attended segregated schools, drank from colored-only fountains, and one morning woke up to find charred crosses on her front lawn. In 1942, Eaton rode for two miles on a mule-drawn wagon to register to vote at the Franklin County courthouse. Three white male officials confronted her. They ordered her to stand up straight, keep her arms at her side, and recite from memory the preamble to the Constitution. She did so word for word and then passed a written literacy test, becoming one of the few African Americans of her era to vote in North Carolina. She later marched with Martin Luther King and helped several thousand African Americans in North Carolina to register to vote. “I am where I am today,” said President Barack Obama in 2015, “only because men and women like Rosanell Eaton refused to accept anything less than a full measure of equality.”³⁰

More than seventy years after her initial registration, a ninety-three-year-old Eaton had a much harder time qualifying to vote under North Carolina’s new law, which required voters to show specific forms of photo identification at the polls. She had a driver’s license, but the name on her voter-registration card (Rosanell Eaton) did not match the name on her license (Rosa Johnson Eaton). To gain new identification she made eleven trips to different agencies—four trips to the Department of Motor Vehicles, four trips to two different Social Security offices, and three trips to different banks—totaling more than two hundred miles and twenty hours. “You know, all of this is coming back around before I could get in the ground,” she said. “I was hoping I would be dead before I’d have to see all this again.” So, Eaton joined the lawsuit.³¹ North Carolina NAACP v. McCrory produced three years of litigation through a federal district court, to the Fourth Circuit Court of Appeals and ultimately to the U.S. Supreme Court. The North Carolina case is a microcosm of the post-2000 American “voting wars.”³²

The process for enacting VIVA had turned on the U.S. Supreme Court decision in Shelby County v. Holder on June 25, 2013. This ruling relieved North Carolina and all states from preclearing new voter laws or regulations under section 5 of the Voting Rights Act. In the most significant victory for opponents of the Voting Rights Act, an ideologically riven five-to-four Supreme Court ruling struck down as unconstitutional the coverage formula of section 4. The decision, in effect, nullified section 5 preclearance because it eliminated the legal foundation for requiring states or localities to gain approval for new voting laws. The Court held that section 4 trampled on the “equal sovereignty of the states” because the differential treatment of the states is “based on 40-year-old facts having no logical relationship to the present day.” It said that there was little difference in the twenty-first century between covered and noncovered jurisdictions, and that “regardless of how to look at the record no one can fairly say that it shows anything approaching the ‘pervasive,’ ‘flagrant,’ ‘widespread,’ and ‘rampant’ discrimination that faced Congress in 1965, and that clearly distinguished the covered jurisdictions from the rest of the nation.”³³

The Court majority failed to consider the subtler but still effective discrimination of the new voting wars. Just a year earlier, for example, three-judge panels of the District Court of the District of Columbia had ruled unanimously that Texas’s voter photo ID law had the effect of discriminating against minorities and that Texas’s redistricting plans for Congress and the state senate had intentionally discriminated against minorities. In a comprehensive study of voting rights violations, historian J. Morgan Kousser refuted the Court’s claim of minimal recent voting rights violations in covered jurisdictions. Dallas minister Peter Johnson, a civil rights activist since the 1960s, said, “There’s nobody that’s going to shoot at you if you register to vote today. They aren’t going to bomb your church. They aren’t going to get you fired from your job. You don’t have those kinds of overt, mean-spirited behaviors today that we experienced years ago.… They pat you on the back, but there’s a knife in that pat.”³⁴

The Republican sponsors of VIVA introduced the bill as HB 589 at the beginning of the 2013 assembly session, before the Shelby decision. In April, the North Carolina House of Representatives passed a version of HB 589 that included voter photo identification requirements, with no other provisions.³⁵ This bill closely followed the legislation in Indiana by authorizing for voting all forms of government-issued photo identification, including expired IDs. The sponsors of this bill lauded their transparent process and extensive vetting, which included public hearings, testimony from interest groups and authorities in the field, several committee hearings, and two days of debate in the house of representatives. Republican representative Tom Murry, one of the bill’s primary sponsors, said, “The historic vote was the result of a 10-month process that included multiple public hearings, hours of testimony by experts and members of the public, and in-depth analysis of voter ID systems in numerous other states.” Representative David Lewis, chairman of the House Elections Committee, noted, “This has been a fair and open and transparent process, as we committed it would be.”³⁶

With North Carolina still under preclearance requirements, sponsors of HB 589 insisted that the relatively lenient voter photo ID bill balanced concerns for ballot integrity with voter access to the ballot. Then Republican speaker of the house and later U.S. senator Thom Tillis said, “We have arrived at a bill that we believe will stand up in a court of law, address legitimate concerns about voting access, and move North Carolina to a photo identification voting system.”³⁷

The general assembly radically changed course on HB 589 after the June 25 Shelby ruling. Following house passage, the bill sat in the senate for months with no committee or floor action. The Shelby decision, however, prompted the senate leadership to revamp HB 589 into legislation that little resembled the pre-Shelby version. On the day of the Shelby ruling, Republican Tom Apodaca, the chairman of the Senate Rules Committee, said, “I guess we’re safe in saying this decision was what we were expecting.” He added, “So, now we can go with the full bill.” The Shelby decision, Apodaca suggested, eliminated “legal headaches” in adopting the many revised and new provisions of the “full bill.” The ruling, he argued, “should speed things along greatly.”³⁸

The so-called full bill both added to and subtracted from the pre-Shelby version of HB 589. It added provisions that cut the number of early voting days from seventeen to ten and eliminated the option to register during this period. It prohibited the prior practice of partially counting provisional ballots cast in the incorrect precinct for countywide or statewide elections and ended the preregistration of sixteen- and seventeen-year-olds. Simultaneously, the “full bill” subtracted from HB 589 many forms of previously acceptable photo identification, including student IDs, government employee IDs, public assistance IDs, and expired IDs with an exception for registrants aged seventy or more.

As was required by federal court decisions, VIVA authorized voters to apply for a free state-issued photo ID. But applicants had to visit a Department of Motor Vehicles office in person and provide proof of voter registration, age, and identity; a valid Social Security number; and proof of citizenship and residency, documentation that often eluded poor and minority people and elderly African Americans like Rosanell Eaton, born in the Jim Crow era. Applicants had to attest that they did not already possess an authorized photo ID, which presumed knowledge of specific forms of ID included and excluded under VIVA.³⁹

The senate and house passed the post-Shelby version of HB 589 on July 25, 2013, two days after it emerged from the Senate Rules Committee and just one day before the end of the legislative session, foreclosing any further opportunity for debate or analysis.⁴⁰ The leadership rushed the bill to passage without extended public hearings, testimony by authorities or interest groups, or extensive and open legislative debate. The lack of process given to the full bill departed from the robust scrutiny accorded the pre-Shelby version.⁴¹

In the post-Shelby bill, the legislature eliminated forms of photo identification that were relatively more accessible to minorities and retained forms of identification that were relatively less accessible to minorities. Analyses presented to the federal courts demonstrated that, as compared to whites, African Americans, by far the largest minority group in the state, were significantly less likely to possess unexpired forms of the two most commonly available categories of photo IDs under VIVA: valid Department of Motor Vehicles (DMV) operators’ and nonoperators’ IDs, and U.S. passports. In contrast, analysis showed that African Americans had greater access than whites to higher education IDs, government employment IDs, and public assistance IDs that the “full bill” had eliminated. In North Carolina, 7.3 percent of voting-age blacks were enrolled in public institutions of higher education, compared to 5.7 percent of whites; 10.9 percent of African Americans held government jobs, compared to 9.0 percent of whites; and 27.1 percent of African American households received food stamps / SNAP public assistance, compared to 8.1 percent of white households.

The challenged provisions of VIVA added to the full bill after Shelby similarly discriminated against black voters. African Americans were much more likely than whites in North Carolina to use early voting, which VIVA restricted, and to register during early voting, which VIVA eliminated. In the presidential election of 2012, African Americans accounted for 28.8 percent of early voters, compared to 15.5 percent of other voters, whereas whites accounted for 65.4 percent of early voters, compared to 78.0 percent of other voters. African Americans accounted for 33.7 percent of people registering during the early voting period, compared to 22.3 percent of all registered voters. Whites accounted for 53.7 percent of registrations during the early voting period, compared to 71.6 percent of all registered voters.

African Americans disproportionately cast partially counted provisional ballots and registered as sixteen- or seventeen-year-olds. In 2012, African Americans accounted for 35.0 percent of voters with partially counted provisional ballots, compared to 23.0 percent of other voters, whereas whites accounted for 51.9 percent of voters with partially counted provisional ballots, compared to 70.9 percent of other voters. In March 2013, African Americans comprised 26.2 percent of registered voters who had initially preregistered as sixteen- or seventeen-year-olds, compared to 22.4 percent of all registered voters. Whites comprised 61.9 percent of preregistered voters, compared to 70.7 percent of all registered voters. A recent study by two political scientists “consistently found” that registering potential voters while still in high school “is a viable electoral reform for increasing youth turnout, increasing voting by somewhere between 2 to 8 percentage points.”⁴²

The rationales offered by backers of the post-Shelby bill confirmed the discriminatory intent of the North Carolina legislature and governor. Backers claimed that the photo ID provisions brought North Carolina into line with other states. Senator Robert A. Rucho, chairman of the Joint Legislative Elections Oversight Committee, said, “There are 30 states which currently have a voter ID in place, and 33 states have passed, uh, ID—voter ID laws.” Yet of the states that Rucho cited, most accepted non-photo forms of identification or had nonstrict laws, which enabled voters to cast a regular ballot through alternative means of identification, such as signing a declaration of identity. Even among states with strict photo ID laws, North Carolina became the only one that prohibited government-issued employee, student, public assistance, and expired IDs.

Citing poll data, backers claimed that the people of North Carolina overwhelmingly supported VIVA’s photo identification requirement and electoral restrictions. However, the cited polls refer to support for “some form of government-issued ID,” which would include many forms of identification eliminated in the post-Shelby bill. In addition, a poll conducted in North Carolina by SurveyUSA in April 2013 found that 74 percent of North Carolina’s registered voters agreed that “legislators should show evidence of significant problems, such as real voter fraud, before they pass laws that make voting more difficult.” The polls further demonstrated that North Carolina residents rejected the full VIVA bill. An August 2013 survey by Public Policy Polling found that 39 percent of respondents supported VIVA in its entirety, with 50 percent opposed and 11 percent uncertain. An August 2013 Elon University poll found that only 30 percent of North Carolina residents supported reducing early voting by a week, with 59 percent opposed. These results show that contrary to claims by VIVA backers, the people of North Carolina did not support the panoply of voting restrictions incorporated in the legislation.⁴³

After the state Board of Elections study showed the absence of voter impersonation in the state, VIVA’s sponsors pivoted to the vague notion of increasing voter confidence in the integrity of the ballot as justification for the photo ID requirement. The Supreme Court had cited voter confidence as a legitimate state concern in the Crawford case, but in that litigation, plaintiffs had not attempted to show either a discriminatory effect of the Indiana law or a discriminatory purpose in its enactment. “There is some evidence of voter fraud, but that’s not the primary reason for doing this,” said Speaker Tillis. “We call this restoring confidence in government.”⁴⁴ Just a few months earlier, backers of the photo identification law had with equal conviction affirmed that the pre-Shelby version of the photo ID bill, without the later subtractions, had achieved the goal of upholding ballot integrity and voter confidence. This commonsense measure, Speaker Tillis said of the pre-Shelby bill, “will protect the integrity of the ballot box and restore confidence in our election system.” Yet this affirmation did not deter Republicans in the legislature from imposing much stiffer restrictions in their post-Shelby bill.⁴⁵

Studies show that strict photo voter ID laws fail to increase voter confidence in the integrity of elections. A national study by political scientists Shaun Bowler and Todd Donovan finds that “none of our model specifications yielded evidence that strict photo identification laws had any broad, unconditioned effects on voter confidence. The likelihood that a person perceived voter fraud as common was unrelated to how strict her state’s voter identification laws were.”⁴⁶

Finally, Governor Pat McCrory invoked the need for “registration integrity” to justify the elimination of registration during the early voting period. However, the state Board of Elections report on voter fraud found only one case of registration fraud related to residency after adoption of early voting registration in 2007, compared to two cases before its adoption.⁴⁷

From the post-Reconstruction era through “Operation Eagle Eye,” calls for “election integrity” or “ballot security” have been used in efforts to discourage voting by minorities. More recently, in response to lawsuits in 1982 and again in 1987, the national Republican Party agreed to consent decrees in federal court that prohibited the party from engaging in antifraud activities that targeted minority voters. In a 2009 ruling, New Jersey federal district court judge Dickinson Debevoise denied a Republican National Committee (RNC) motion to end the consent decree. The U.S. Court of Appeals for the Third Circuit upheld his ruling. “Minority voters continue to overwhelmingly support Democratic candidates,” Judge Debevoise wrote in his decision. “As long as that is the case, the RNC and other Republican groups may be tempted to keep qualified minority voters from casting their ballots, especially in light of the razor-thin margin of victory by which many elections have been decided in recent years.” After reviewing all evidence of alleged voter fraud presented by the RNC, he held, “In fact, even a cursory investigation of the prevalence of voter intimidation demonstrates that ballot security initiatives have the potential to unfairly skew election results by disenfranchising qualified voters in far greater numbers than the instances of in-person fraud that may occur during any given race.” After district court judge John Michael Vazquez let the consent decree expire in January 2018, the Democratic National Committee appealed his ruling to the Third Circuit.⁴⁸

The special treatment of mail-in absentee ballots under VIVA further undermined the claim of upholding election integrity. Mail-in absentee ballots, which are disproportionately filed by white people, were exempted from the photo ID requirement. Yet absentee voting is far more susceptible to voter fraud than in-person voting because the perpetrators are nameless and faceless. It is far riskier and less efficient to walk into a polling place under false pretenses than it is to anonymously mail in a fraudulent ballot. Although absentee voting fraud is also rare, it is more common than voter impersonation. The North Carolina State Board of Elections compilation on voter fraud found forty-seven referrable cases of absentee voter fraud. Nationwide data reported in 2011 by News21 found that absentee voting fraud was the most common type of fraud, outnumbering voter impersonation by a ratio of forty-nine to one. A Government Accountability Office report released in October 2014 said that “no apparent cases of in-person voter impersonation [were] charged by DOJ’s [Department of Justice’s] Criminal Division or by U.S. Attorney’s offices anywhere in the United States from 2004 through July 3, 2014.”⁴⁹

North Carolina Republicans noted that the photo ID component of VIVA had a two-year phase-in period during which registered voters could apply for free photo IDs from the DMV. Representative Ruth Samuelson, a primary bill sponsor, said, “I wanted to point out that, once again, the comparison to another state, South Carolina did not have a phase-in period.… The VIVA is going to be an outreach. It’s going to be an outreach to churches. It will be an outreach to 12 community boards of elections. It will be an outreach to schools and other groups, and the kinds of people who are shut in.” However, no other VIVA provision had a phase-in period, and the North Carolina General Assembly eliminated in VIVA many of the public education and outreach provisions of the pre-Shelby version. As of late January 2015, nearly a year and a half after VIVA’s adoption, the state had issued only 771 free ID cards. This stands in stark contrast to the case in Pennsylvania. Before the courts invalidated Pennsylvania’s never-implemented voter photo identification law, the state, which conducted an extensive campaign of public education and outreach, had issued 16,700 free voter ID cards in the first year and a quarter after the law’s adoption.⁵⁰

On the eve of a preliminary injunction hearing on VIVA in June 2015, the North Carolina legislature, fearful of an adverse verdict, modified the photo ID provision of the 2013 law to allow individuals without authorized photo IDs to vote provisional ballots if they signed a declaration indicating a reasonable impediment to obtaining such identification. This put off consideration of that provision until trial on the merits in January 2016. After a trial that lasted several weeks and created a record of 23,000 pages, federal district court judge Thomas D. Schroeder upheld VIVA in its entirety. Judge Schroeder wrote a detailed, 485-page opinion, but on July 29, 2016, a unanimous three-judge panel of the Fourth Circuit Court of Appeals overturned his findings and struck down every challenged provision of the law. Following Supreme Court guidelines for establishing intentional discrimination, the circuit court further ruled that in adopting VIVA, North Carolina had acted with the intent to discriminate against minority voters to achieve political advantages for Republicans.⁵¹

The appeals court found that “in holding that the legislature did not enact the challenged provisions with discriminatory intent, the [district] court seems to have missed the forest in carefully surveying the many trees. This failure of perspective led the court to ignore critical facts bearing on legislative intent, including the inextricable link between race and politics in North Carolina.” As in the earlier Garza case, the court firmly rejected the claim that burdens placed on minority voters can be justified under the Constitution by political motivation. It noted that “the totality of circumstances … cumulatively and unmistakably reveal that the General Assembly used [VIVA] 2013-381 to entrench itself. It did so by targeting voters who, based on race, were unlikely to vote for the majority party. Even if done for partisan ends, that constituted racial discrimination.”⁵²

The appeals court stated that provisions of the post-Shelby version of photo ID were deliberately designed to “target African Americans with almost surgical precision, they constitute inapt remedies for the problems assertedly justifying them and, in fact, impose cures for problems that did not exist. Thus, the asserted justifications cannot and do not conceal the State’s true motivation.” The court further found that “before enacting that law, the legislature requested data on the use, by race, of a number of voting practices. Upon receipt of the race data, the General Assembly enacted legislation that restricted voting and registration in five different ways, all of which disproportionately affected African Americans.”⁵³

A finding of intentional racial discrimination under the equal protection clause of the Fourteenth Amendment is rare and requires plaintiffs to surmount a high barrier of proof, which they did in this litigation. Despite the Supreme Court’s invalidation of section 4 of the Voting Rights Act in the Shelby decision, this finding of intentional discrimination authorized the appeals court under section 3(c) of the act to “bail in” North Carolina to preclearance coverage. However, the court chose not to exercise this option.⁵⁴

Even with this victory for black voting rights, the North Carolina case highlights the consequences of losing the preclearance provision of the Voting Rights Act. Black plaintiffs succeeded in overturning VIVA only after protracted, expensive, and complex litigation, with the full power and resources of the U.S. Justice Department behind them and a liberal appeals court in the Fourth Circuit, not a set of circumstances likely to be replicated for other instances of discriminatory state and local practices. Litigation under section 2 of the Voting Rights Act and the equal protection clause of the Fourteenth Amendment requires considerable resources, legal expertise, and the testimony of expert witnesses.

After the adverse VIVA ruling, for example, Republicans in North Carolina, unshackled by preclearance requirements, acted again to burden African Americans voting in the 2016 elections by reducing early voting locations in seventeen counties during the first week of early voting. These counties included the four counties with the largest African American populations. In some nine counties, local Republican election officials opened only one polling location for early voting’s first week. Turnout during this period declined substantially in these seventeen counties as compared to the remainder of the state. The state also engaged in a flawed purge of registered voters that disproportionately impacted African Americans. Shortly before the election, U.S. District Court judge Loretta Copeland Biggs ordered the state to restore the purged voters. “Voter enfranchisement cannot be sacrificed when citizens through no fault of their own have been removed from the voter rolls,” she ruled. Six months after the 2016 election, on May 15, 2017, the U.S. Supreme Court declined to take up an appeal by North Carolina on the VIVA case, leaving in place the appeals court’s ruling.⁵⁵

North Carolina was not alone in implementing after Shelby such seemingly small but consequential discriminatory measures. A 2016 study of a limited sample of 381 counties in formerly covered states found that since Shelby, 43 percent had closed polling places, piling up a total of 868 closures. The study found that 67 percent of counties in Alabama reduced the number of polling places, 61 percent in Louisiana, and 53 percent in Texas. In Arizona, nearly every county had closed polling places, erasing 212 polling locations. Although there may be “justifiable reasons” for closing polling places, the study concluded, “the loss of Section 5 means that there is no process to ensure that reductions are disclosed to the public, are conducted with the input of impacted communities, and do not discriminate against voters of color.”⁵⁶

In Arizona, for the 2016 presidential preference primary, Maricopa County reduced the number of polling places from more than two hundred in 2012 to sixty in 2016. Maricopa is the largest county in Arizona and the fourth largest in the United States, with a population of 4.2 million, including a 31-percent-Hispanic and 44-percent-minority component. The county’s chief election official, Republican Helen Purcell, authorized this reduction despite having been warned ahead of time by county supervisor Steve Gallardo that sixty polling places was inadequate for the county. A study of polling place locations by the Arizona Republic found that “while both rich and poor areas were hurt by a lack of polling sites this year, a wide swath of predominantly minority and lower-income areas in west Phoenix and east Glendale, along with south Phoenix, were particularly lacking in polling sites compared with 2012.” Most Arizona counties had about 2,500 eligible voters per polling place, compared to 21,000 per polling place in Maricopa County.⁵⁷

In 2011, Alabama enacted a photo identification requirement for voting in person or absentee. It did submit the law for preclearance but delayed its implementation after the Shelby decision. Although driver’s licenses are one of the limited forms of identification authorized under the law, in 2015 Alabama proposed to close thirty-one driver’s license offices, mostly in rural areas of the state’s black belt. After protests from civil rights groups, Alabama agreed to keep the offices open one day a week, still severely restricting access for residents of these substantially black counties.⁵⁸

For eight years from 2011 to 2018, Texas defended patently discriminatory redistricting plans for congressional seats, also facilitated by the end of preclearance. The plan led to multiple lawsuits. From the 2000 to the 2010 U.S. Census, the population of Texas soared by 21 percent, which netted the state four new congressional seats, raising its apportionment to thirty-six seats. No other state gained more than two new districts. African Americans and Latinos accounted for 79 percent of this population growth, and all minorities combined accounted for 89 percent. Texas had become a majority-minority state, in both total and voting-age population. African Americans and Latinos comprised 39.5 percent of the state’s citizen voting-age population, which translated into 14.2 congressional districts. All minorities combined comprised 43 percent of the citizen voting-age population, which translated into 15.5 districts, compared to 20.5 districts for whites.⁵⁹

It was not easy, but by fragmenting and in some cases packing enough minority communities, the Republican legislature adopted a 2011 plan that created only ten districts in which minority voters had an effective opportunity to elect candidates of their choice. That left twenty-six districts in effect controlled by whites, over-representing whites by five and a half districts. The cracking and packing of minorities is illustrated by the districts that the Republican legislature drew in the heavily populated Dallas and Tarrant Counties of north Texas. From 2000 to 2010, in the region known as DFW, because it includes the cities of Dallas and Fort Worth, the Latino population grew by 84 percent (more than 440,000 people), the African American population grew by 34 percent (more than 179,000), and the population of other minorities by 12 percent (more than 63,000). The white population declined by 30 percent (more than 156,000). By 2010, minorities comprised 57 percent of the population in DFW, 53 percent of the voting-age population, and 40 percent of the citizen voting-age population. In defiance of these trends, the 2011 plan did not create a single additional minority opportunity congressional district in DFW, with the result that only one of eight districts (12.5 percent) that fell primarily within DFW provided minorities such an opportunity.

Republicans in the state legislature crammed additional African Americans into the already-packed congressional district 30, which had overwhelmingly elected a black candidate in every prior election. The plan upped the minority population of district 30 to 86 percent. The legislature then cracked or fragmented many minority communities in the region and created several bizarrely shaped districts controlled by whites. The legislature separated the Latino neighborhoods in Fort Worth from nearby African American neighborhoods by drawing a notorious “lightning bolt” district that shot into Tarrant County from white-controlled Denton County. The plan left many hundreds of thousands of minority voters stranded in white-dominated districts in which they had no ability to elect candidates of their choice to Congress or even influence election results.

The plan dismantled a minority opportunity district in Travis County, one of the state’s most liberal counties, encompassing the city of Austin. The legislature cracked Travis County into five separate districts, all controlled by white Republicans. In Harris County, which encompasses the city of Houston, nearly all the substantial population growth was minority. Yet the legislature failed to create an additional minority opportunity district there even though minority advocacy groups presented alternative plans for the county that incorporated one.

In congressional district 23 in southwest Texas, the legislature used a different maneuver to dilute Hispanic votes. In 2006, the U.S. Supreme Court had found a voting rights violation in the prior version of this district and ordered a bolstering of the Hispanic population. To mask another voting rights violation, the legislature knowingly and intentionally reduced the effectiveness of the Hispanic vote without reducing the Hispanic voting-age percentage in the district. The map designers craftily switched out high-voting Democratic Hispanic areas with Hispanic areas that have lower turnout and moved in high-turnout white Republican areas. To achieve this end, the plan moved more than 600,000 people in and out of district 23 and split thirty-nine voting precincts. Texas’s own redistricting expert, Professor John Alford of Rice University, conceded that district 23 “is probably less likely to perform than it was, and so I certainly wouldn’t count and don’t [and] haven’t counted the 23rd as an effective minority district in the newly adopted plan.”⁶⁰

In defense of the plan, Texas attorney general Greg Abbott (who later became governor) admitted that it disparately burdened minority voters but raised the familiar, though often rejected, claim that the goal of advancing Republican interests immunized the plan from a constitutional violation of intentional discrimination, despite “incidental effects on minority voters.” He observed, “DOJ’s accusations of racial discrimination are baseless.… In 2011, both houses of the Texas Legislature were controlled by large Republican majorities, and their redistricting decisions were designed to increase the Republican Party’s electoral prospects at the expense of the Democrats. It is perfectly constitutional for a Republican-controlled legislature to make partisan districting decisions, even if there are incidental effects on minority voters who support Democratic candidates.”⁶¹

The plan’s defenders also dusted off the states’-rights plea by claiming that Texas should be free to formulate its redistricting plan without challenge from the Department of Justice. In an op-ed article in the Washington Times entitled “Obama’s Scheme to Take Over Texas,” Abbott wrote, “The Constitution makes elections the states’ business, not the federal government’s,” and argued that the Department of Justice should not be challenging decisions made by the Texas legislature. Republican U.S. representative Blake Farenthold tweeted that the Department of Justice’s “Attorney General Holder is trying to wage a war on Texas & states’ rights,” and “DOJ’s push to overturn #TX voting laws: the govt’s latest attempt at trying to tell states what they can & can’t do.”⁶²

Abbott and Farenthold should have known that elections are not just the business of the states; both the Constitution and the Voting Rights Act have long protected minorities from discrimination by the states. In effect, Abbott and other Republican leaders in Texas have constructed an internally self-justifying rationale for intentionally discriminating against minorities. In their view, it is legitimate for Texas’s Republican majority to discriminate against minorities because that is in the partisan interests of the Republican Party and consistent with states’ rights. But it is apparently not legitimate for the U.S. Department of Justice to defend minority voting rights because that is in the partisan interests of the Democratic Party and interferes with state business. Yet the intent of the federal government is to enforce the Voting Rights Act and the Fourteenth Amendment to assure that minorities have the opportunity to participate fully in the political process and elect candidates of their choice. Change the names of the parties and the claims made by Texas Republicans in the twenty-first century replicated the arguments made by Texas Democrats in the nineteenth century against enforcing the Fourteenth and Fifteenth Amendments.

Texas officials knew that the 2011 congressional redistricting plan would not gain Justice Department preclearance. Instead, they sought a judgment in favor of the plan from a three-judge panel of the District Court for the District of Columbia. In a 2012 decision, the district court unanimously ruled against the state. It held that the plan had both the effect and intent of discriminating against minorities. The court found that “the parties have provided more evidence of discriminatory intent than we have space, or need, to address here.”⁶³

As part of the tangled knot of litigation created by Texas’s redistricting decisions, civil rights groups and Democratic interests sued the state in 2011, charging violation of section 2 of the Voting Rights Act and the Constitution. The Department of Justice later joined on the lawsuit as a plaintiff. Before issuing a final ruling, a three-judge court in San Antonio adopted an interim plan for the 2012 elections, closely based on the 2011 plan and agreed to as part of a compromise between the state and some, but not all, of the plaintiffs. The San Antonio court did not have the benefit of the District of Columbia court’s ruling at the time that it implemented the plan, and its order made clear that the plan was interim only; it did not include any final ruling on any pending voting rights or constitutional claims. The interim plan fixed some but not all deficiencies of the 2011 plan.

In March 2017, the San Antonio court issued a final ruling on the 2011 congressional plan, agreeing with the DC court, whose ruling had no legal standing after the Shelby decision, that the plan had both the effect and intent of discriminating against minority voters. The court rejected the state’s claim “that they did not engage in intentional vote dilution of minority voting strength, but only of Democrat voting strength.” For example, in DFW, it found that “while there is certainly an overlap between cracking and packing Democrats and cracking and packing minorities, the Court finds that Plaintiffs have satisfied their burden of showing that intentional minority vote dilution was a motivating factor in the drawing of district lines in DFW and that mapdrawers intentionally diluted minority voting strength in order to gain partisan advantage.” Across the state, the court held that “the record indicates not just a hostility toward Democratic districts, but a hostility to minority districts, and a willingness to use race for partisan advantage.”⁶⁴

Meanwhile, in a special session of the state legislature in 2013, Texas adopted the interim congressional plan as the official, permanent state plan. Although the legislative leadership ruled cosmetically that the legislature could make changes in the interim redistricting plans, it followed closely the initial call of Governor Rick Perry for the special session to adopt the interim plan with no changes. The Republican majority in the Texas legislature pushed the interim plan through along party lines with no changes, despite objections from minority advocacy groups and minority members of the legislature.⁶⁵

In a subsequent August 2017 decision on the state’s enactment in 2013 of the interim congressional plan, the San Antonio court again found that the state had engaged in intentional discrimination in certain parts of the state. It did not find intentional discrimination in DFW, not because it absolved the state of such intent but because it ruled that any further remedy of the plan’s intentional discrimination would fail. It found that an additional single-race majority district could not be drawn in DFW and that a combined majority-minority district of African Americans and Hispanics failed to meet legal requirements because blacks and Hispanics diverged in their choice of candidates in Democratic primary elections. This decision is currently on appeal to the U.S. Supreme Court.⁶⁶

Civil rights advocates achieved other victories in the courts. In 2013, federal district court judge Nelva Gonzales Ramos found that Texas’s strict voter identification law violated section 2 of the Voting Rights Act and intentionally discriminated against minorities. Like North Carolina’s VIVA law, the Texas statute included IDs more accessible to whites than minorities, notably licenses for the concealed carry of firearms, but excluded IDs more accessible to minorities, such as student, public assistance, and government employee IDs. A three-to-zero decision by a panel of the Fifth Circuit Court of Appeals and a nine-to-six en banc decision by the Fifth Circuit upheld Judge Ramos’s finding that the Texas ID law violated the Voting Rights Act. It neither upheld nor overturned the finding of intentional discrimination in the enactment of the law but remanded the issue back to Judge Ramos for further consideration.⁶⁷

The U.S. Supreme Court declined on January 23, 2017, to hear an appeal on the Fifth Circuit’s ruling against the voter ID law brought by the state of Texas. Chief Justice John G. Roberts noted that the “issues will be better suited for certiorari review” after Judge Ramos issued a final order.⁶⁸ Upon remand from the Fifth Circuit, Ramos entertained briefings on the question of intentional discrimination. She again found discriminatory intent, ruling that actions of the state legislature “revealed a pattern of conduct unexplainable on nonracial grounds, to suppress minority voting.”⁶⁹

She subsequently rejected as inadequate a state revision of the ID law that changed none of its discriminatory requirements. The revision only enabled voters lacking an authorized ID to present an alternative photo or non-photo identification and sign an affidavit attesting to a “reasonable impediment” to obtaining an acceptable one. The new law subjected individuals who allegedly lied on the affidavit to felony prosecution with potential penalties of six months to two years in jail. Ramos found that “not one of the discriminatory features of [the old law] is fully ameliorated by the terms” of the new law. The revised law, she ruled, “trades one obstacle to voting with another—replacing the lack of qualified photo ID with an overreaching affidavit threatening severe penalties for perjury.”⁷⁰

Combined with the decision of the District of Columbia District Court and two decisions from the San Antonio court on redistricting, federal courts have ruled four times since 2012 that the state of Texas had intentionally discriminated against minority voters. So too did the appeals court in North Carolina. A district court in Louisiana also found in August 2017 that the state of Louisiana had intentionally discriminated against African American voters by maintaining an at-large system for electing judges in the state’s Thirty-Second Judicial District.⁷¹ These intent findings in three jurisdictions, which pre-Shelby had been covered under the preclearance provision of the Voting Rights Act, provide further evidence that contrary to the reasoning in the Shelby decision, racial discrimination has not faded away in previously covered jurisdictions.

More than fifteen years after the Page v. Bartels litigation, the Supreme Court recognized the discriminatory effects of packing minorities into districts, common practice for Republican-controlled legislatures. By confining minority voters to a relatively few districts, packing impedes their ability to elect candidates of their choice—typically Democrats—in additional districts. The threshold percentage of minorities for a packed district is a factually intensive inquiry and is usually much higher for Hispanics than for African Americans. Even in districts with less than a voting-age majority, black voters can still elect candidates of their choice in partisan elections. They have done so in many instances, by winning in Democratic primaries, with white voters primarily participating in Republican contests, and then drawing enough crossover votes from whites to elect their chosen candidates in general elections.

During its 2017 term, the Supreme Court upheld anti-packing decisions in Virginia and North Carolina. In Virginia, a three-judge federal court found that Republicans in the legislature had unduly packed minorities into congressional district 3, represented by African American Democrat Bobby Scott. In 2016, with unanimous approval from the Supreme Court, the lower court implemented a new plan that unpacked district 3 and enabled African Americans to elect candidates of their choice in neighboring district 4. In North Carolina, three-judge courts found that the state’s plans for both Congress and state legislatures had impermissibly used race to pack African Americans into certain districts. The decisions covered two congressional districts, nineteen state house districts, and nine state senate districts. The Supreme Court upheld both the congressional and state legislative findings of discrimination.⁷²

In reaching these findings on excessively packed minority districts, the courts drew on Supreme Court precedent from the 1990s that prohibited jurisdictions from drawing districts predominantly based on race, absent a showing that the districts were narrowly drawn to meet a compelling state interest, such as complying with the Voting Rights Act. The packing of districts to create wasted minority votes would clearly not qualify as a legitimate state interest of any kind. Writing for a court majority in the North Carolina congressional case, Supreme Court justice Elena Kagan rejected the state’s defense that it drew district lines based on politics, not race. “The sorting of voters on the grounds of their race remains suspect even if race is meant to function as a proxy for other (including political) characteristics,” she wrote.⁷³

One of the ironies of voting rights jurisprudence is the unexpected consequences of the racial gerrymandering prohibitions of the 1990s that advocates initially believed would thwart the ability to create minority opportunity districts. Instead, as the recent rulings indicate, these 1990s restraints have checked the packing of minorities by Republican legislatures intent on confining minorities to racially packed districts and expanding the political power of Republicans elsewhere across redistricting maps.

In January 2018, a three-judge court in North Carolina also turned against the state of North Carolina its claims of political gerrymandering as a defense against racial gerrymandering. In what the media dubbed a “bombshell ruling,” the court struck down as an unconstitutional political gerrymander the congressional redistricting plan that North Carolina had adopted to remedy its nullified racial gerrymander. Never before in American history had a federal court strode boldly into the “political thicket” to invalidate a congressional redistricting plan as an unconstitutional partisan gerrymander. The court found that the plan violated the equal protection clause because the North Carolina General Assembly enacted it “with intent of discriminating against voters who favored non-Republican candidates.” The plan violated the First Amendment “by unjustifiably discriminating against voters based on their previous political expression and affiliation,” and “Article I, section 2’s grant of authority to ‘the People’ to elect their Representatives.” The court denounced that partisan gerrymandering for “undermining the right to vote—the principle [sic] vehicle through which the public secures other rights and prevents government overreach.” It said that the state had no legitimate interest in “the drawing of legislative district lines to subordinate adherents of one political party and entrench a rival party in power.” The Supreme Court stayed implementation of a new redistricting plan and may ultimately review this ruling as part of its unprecedented roster of partisan gerrymandering cases.⁷⁴

Even before the North Carolina case, the Supreme Court was reconsidering the legality of partisan gerrymandering, with potentially far-reaching effects on the balance of party power in Congress and state and local legislatures. Although political gerrymandering recognizes no party boundaries, Republican victories in gubernatorial and state legislative elections in 2010 gave the party control of far more state governments than Democrats, including such large swing states as Pennsylvania, Michigan, Florida, Ohio, and Wisconsin. Republicans used their control of state governments to enact political gerrymanders during the redistricting that followed the 2010 U.S. Census. Republican control of the redistricting process followed from a carefully planned effort to gain control of government in key states during the 2010 elections. According to a report by the Republican State Legislative Committee (RSLC), “As the 2010 Census approached, the RSLC began planning for the subsequent election cycle, formulating a strategy to keep or win Republican control of state legislatures with the largest impact on Congressional redistricting as a result of reapportionment.” The Republican “REDistricting MAjority Project (REDMAP), focused critical resources on legislative chambers in states projected to gain or lose Congressional seats in 2011 based on Census data.”⁷⁵

In the swing state of Pennsylvania, REDMAP put “critical resources” into elections to assure a GOP majority in the state’s general assembly in the 2010 elections and corresponding control over the 2011 redistricting process. This effort paid off with one of the nation’s most consequential political gerrymanders. Barack Obama won Pennsylvania with 52.7 percent of the two-party vote in 2012, and Democrats won 51 percent of the statewide, two-party congressional vote. Yet under a plan drawn by a Republican-controlled legislature, Republicans in Pennsylvania won thirteen U.S. House seats (72%) and Democrats only five (28%). Republicans held this two-thirds majority of seats through 2016. In its report on the REDMAP initiative, the Republican State Legislative Committee bragged about the importance of party gerrymandering in securing a nationwide congressional majority:

Voters pulled the lever for Republicans only 49 percent of the time in Congressional races, suggesting that 2012 could have been a repeat of 2008, when voters gave control of the White House and both chambers of Congress to Democrats.… Instead, Republicans enjoy a 33-seat margin in the U.S. House seated yesterday in the 113th Congress, having endured Democratic successes atop the ticket and over one million more votes cast for Democratic House candidates than Republicans.… All components of a successful Congressional race, including recruitment, message development and resource allocation, rest on the Congressional district lines, and this was an area where Republicans had an unquestioned advantage.⁷⁶

The Pennsylvania State Supreme Court, voting along party lines in February 2018, struck down the state’s congressional redistricting plan as a violation of the state constitution. After the Republican-controlled state legislature and the Democratic governor deadlocked, the court drew its own plan. The Pennsylvania ruling shows the value of litigating state constitutions, which, unlike the U.S. Constitution, usually mandate affirmative voting rights. See also Martin v. Kohls, 2104 Ark. 427, striking down Arkansas’s voter ID law. On March 19, 2018, a panel of three federal judges and the U.S. Supreme Court rejected petitions by Republican legislators to halt implementation of the court’s plan.⁷⁷

Earlier, in a challenge to the redistricting plan for the Wisconsin State Assembly, a federal court for the first time struck down as unconstitutional a statewide redistricting plan. Unlike the North Carolina case, where the court found that plaintiffs had presented sufficient social science evidence using multiple methodologies to prove its case, the three-judge court in Wisconsin proposed to fill the gap left by the Supreme Court’s Vieth ruling, which failed to establish a standard for identifying an unconstitutional partisan gerrymander.

The Wisconsin court relied on the so-called efficiency gap, a measure of the relative number of wasted votes for each party that resulted from a redistricting plan. An excessive number of wasted votes for the disadvantaged party equal to 7 percent or more of total turnout in elections would establish a prima facie case for a partisan gerrymander of legislative redistricting plans. The Wisconsin case is pending before the U.S. Supreme Court. The Court also agreed to hear a case in the state of Maryland, where Republicans have alleged that Democrats shifted a single congressional district from Republican to Democratic in violation of Republican voters’ rights of free expression under the First Amendment. The Court took up the case although the lower court had not yet conducted a trial.⁷⁸

As part of its new wave of voting rights litigation, the Supreme Court has revisited the issue of purging voters from registration rolls. In Ohio, the Sixth Circuit Court of Appeals found that the state’s “Supplemental Process” for purging individuals from the registration rolls violated the provision of the National Voter Registration Act that prohibits a state from using procedures that “result in the removal of the name of any person from the official list of voters registered to vote in an election for Federal office by reason of the person’s failure to vote.” Under its supplemental process, Ohio purged registrants who had not voted in two years, failed to respond to a state-issued mailing, and did not vote for another four years.

The circuit court noted that registrants are automatically removed under this supplemental process in Ohio “even if he or she did not move and otherwise remains eligible to vote.” In the North Carolina VIVA litigation, scrutiny of the state’s online voter files for 2012, for example, disclosed that 37,489 North Carolina citizens registered as of 2002 first voted in 2012 after a ten-year gap. The court ruled in the Ohio case that “it is clear that the Supplemental Process does include a trigger [two years of nonvoting], and that that trigger constitutes perhaps the plainest possible example of a process that ‘result[s] in removal of a voter from the rolls by reason of his or her failure to vote.’ ” The case is now before the Supreme Court on appeal by the state.⁷⁹

The question of whether not voting for a period of time should be a basis for denying the right to vote at a future time is fundamental to the franchise for potentially millions of voters nationwide. Seventeen states, all controlled by Republicans, filed an amicus brief in the Supreme Court to support the Ohio law. Twelve states, mostly controlled by Democrats, filed an amicus brief in opposition to the law.⁸⁰

With the Supreme Court deciding in January 2018 to also review the lower court’s finding of intentional discrimination in the formulation of Texas’s congressional redistricting plan, it has taken more important voting cases than at any time in recent history. The fate of American voting rights for the foreseeable future may well depend on decisions reached by a closely divided court ideologically. Justice Anthony Kennedy will likely provide the critical swing vote as he has done in prior voting rights cases that split the other eight justices four to four, across a conservative / liberal divide.

In the presidential election of 2016, an unprecedented threat to the vote in America emerged from a hostile foreign power. On October 7, 2016, the Department of Homeland Security and Office of the Director of National Intelligence issued a joint statement concluding that “the U.S. Intelligence Community (USIC) is confident that the Russian Government directed the recent compromises of e-mails from US persons and institutions, including from US political organizations.… These thefts and disclosures are intended to interfere with the US election process.” On January 6, 2017, the U.S. intelligence agencies released a second report that pinned this interference on Russian president Vladimir Putin and said that its purpose was to aid in the election of Donald Trump, not just to disrupt America’s democracy. This interference had three interlocking components: the hacking of Democratic emails; propaganda from state-controlled media, Twitter, and Facebook ads; and the use of internet trolls and bots to poison political debate.⁸¹

Later, in 2018, intelligence officials indicated that Russian hackers had tried to penetrate registration rolls—though not the voting machines—in some twenty-one states, and likely succeeded in several. Officials have not reported that the hackers altered the registration rolls to block citizens’ ability to vote in the election. This hacking again exploited the fragmented American election system where localities are responsible for the conduct of elections and a lack of national standards or funding for technological improvements continues to exist. The hacking has the potential to corrupt the vote in America by adding or deleting names from registration rolls, requesting and submitting phony absentee ballots, and even altering or deleting the votes cast in voting machines. Michael Bahar, a former Democratic staffer for the House Intelligence Committee, warned that “the integrity of the entire system is in question. So you need the system to push back and find out what happened and why, so it never happens again.”⁸²

Russian manipulation of the 2016 elections extended beyond the presidential contest to aiding Republican congressional candidates in selected districts. They did so by releasing detrimental information to Republican operatives on Democratic candidates that Russian hackers had stolen from the Democratic Congressional Campaign Committee. The Russians did not assist any Democratic candidates in 2016 or release any information from Republican sources that their hackers had breached prior to election day.⁸³

Former acting CIA director Michael Morell and former Republican representative and chair of the House Intelligence Committee Mike Rogers have warned that Russian manipulation of American elections did not end in 2016. “The United States has failed to establish deterrence in the aftermath of Russia’s interference in the 2016 election,” they wrote. “We know we failed because Russia continues to aggressively employ the most significant aspect of its 2016 tool kit: the use of social media as a platform to disseminate propaganda designed to weaken our nation.” In February 2018, President Trump’s intelligence officials confirmed that Russia will continue to target American elections in the 2018 midterms and beyond.⁸⁴ Hacking into the voting machinery to alter or scramble results may be next on the Russian agenda. At a cybersecurity conference held in Las Vegas during July 2017, attendees easily hacked into thirty voting machines of various types. According to Fox News, “With every machine successfully breached in less than a day, the conference proved the devices are not up to par with modern technologies.”⁸⁵

Donald Trump and his administration have shown no inclination to protect the vote from manipulation by the Russians or another malevolent foreign power. Instead, he has harped on the bogus issue of voter fraud, claiming that, contrary to all evidence and plain commonsense, he would have won the popular vote had it not been for 3 to 5 million illegal voters. These phantom millions presumably all voted for Hillary Clinton—Trump said, “none of ’em come to me”—and then somehow disappeared without detection, including by the Republicans who administered elections in most states. It would have been the most extraordinary conspiracy in the history of the United States. Yet Trump’s outlandish charge gained some traction. According to a Morning Consult / Politico poll released on July 26, 2017, 47 percent of Republicans said they believed that Trump won the 2016 election’s popular vote. Even 25 percent of independents and 12 percent of Democrats professed this belief.⁸⁶

In fact, Trump lagged Hillary Clinton by 2.9 million popular votes, equal to 2.1 percent of the votes cast in the election. No Electoral College winner has ever lost the popular vote by a larger numerical margin, and only in the disputed election of 1876 did the Electoral College winner lose the popular vote by a larger percentage margin.

Rather than deal with the foreign threat to our democracy, President Trump launched a new voter fraud commission with the potential to suppress voting in the United States. Kansas’s Republican secretary of state Kris W. Kobach, co-chair with Vice President Mike Pence of the Presidential Advisory Commission on Election Integrity, has asked for the registration records of every American, including personal information such as addresses, dates of birth, political affiliation, military service, voting history, criminal convictions, and the last four digits of social security numbers. Kobach’s data-mining request met considerable resistance from state officials including Mississippi’s Republican secretary of state Delbert Hosemann, who said, “They can go jump in the Gulf of Mexico and Mississippi is a great state to launch from.”⁸⁷

Kobach has used false claims of widespread voter fraud to justify making voting more difficult, especially for the minorities and young people that primarily vote Democratic. In 2011, Kobach drafted and the Republican-controlled Kansas state legislature adopted the Secure and Fair Elections Act. It requires that voters at the polls present photo identification, absentee voters provide their driver’s license numbers and have their signatures verified, and all new registrants present documentary proof of American citizenship.

Kobach bragged at the time in a Wall Street Journal op-ed that “although a few states, including Georgia, Indiana and Arizona, have enacted one or two of these reforms, Kansas is the only state to enact all three.” A federal court temporarily enjoined the proof of citizenship requirement for voters using the federal form, and a trial on the merits of the full law commenced in March 2018. At trial, Kobach’s expert witness, Jesse Richman, a political science professor at Old Dominion University, said he was unaware of any research that supported Trump’s claim that millions of illegal votes cast ballots in the 2016 presidential election.⁸⁸

In June 2017, a federal magistrate sanctioned Kobach, fining him one thousand dollars for his “deceptive conduct and lack of candor” in response to requests for the discovery of information in his possession. The court found that he had made “patently misleading representations to the court about the documents” requested. In upholding the sanction, federal district court judge Julie A. Robinson noted that Kobach had exhibited a “pattern” in federal litigation of “statements made or positions taken by Secretary Kobach that have called his credibility into question.”⁸⁹ Kobach’s claims of rampant voter fraud to justify the omnibus 2011 law were equally misleading. The Republican National Lawyers Association’s study found one prosecution for voter fraud in Kansas from 1997 through 2011. When Kobach claimed lax enforcement of voter fraud, the legislature in 2015 made him the only secretary of state in the nation authorized to investigate and prosecute cases of voter fraud.

Kobach has tacitly endorsed President Trump’s claim of millions of illegal votes cast nationwide, saying, “We will probably never know the answer to that question” of who won the popular vote. In support of his fraud claims, in September 2017, on Breitbart, Kobach claimed that “now there’s proof” that illegal voters turned the 2016 New Hampshire election to the Democrats, because some 5,313 voters lacked either a New Hampshire driver’s license or a registered vehicle in the state. New Hampshire’s Democratic secretary of state Bill Gardner, a commission member, debunked Kobach, because New Hampshire law authorizes temporary residents, such as college students, with out-of-state driver’s licenses and no vehicles registered in the state to vote legally.⁹⁰

In Kansas, as of late August 2017, the zealous fraud hunter Kobach has charged just twelve people for fraud in all prior elections out of millions of ballots cast. Ten cases involved people who double-voted. Only two involved noncitizen voting, despite Kobach’s claim that in his state, “the total number [of noncitizens] could be in excess of 18,000 on our voter rolls.”⁹¹ Despite the absence of significant voter fraud, the strict voter photo ID law in Kansas resulted in a 1.9 to 2.2 percent decline in voter turnout in Kansas according to a study by the nonpartisan U.S. Government Accountability Office. Black turnout dropped by 3.7 percentage points more than white turnout, and turnout by eighteen-year-olds declined by 7.1 percent more than among people aged forty-four to fifty-three.⁹²

This finding for Kansas is consistent with the results of nationwide studies of how voter photo ID laws have impacted turnout. A study published in the April 2017 Journal of Politics using validated voter data demonstrates that photo ID laws functioned as intended: “The analysis shows that strict identification laws have a differentially negative impact on the turnout of racial and ethnic minorities in primaries and general elections. We also find that voter ID laws skew democracy toward those on the political right.” The authors conclude that voter ID laws are only one of the ways in which states are restricting the opportunity to vote:

What makes voter ID laws more disturbing is that they are just one of the many different ways in which the electoral system is being altered today. Shortened early voting periods, repeal of same-day voter registration, reduced polling hours, a decrease in poll locations, and increased restrictions on voting by felons are all being regularly implemented at the state or local level, and all have been cited as having the potential to skew the electorate and American democracy.⁹³

Kobach has advocated the crosscheck of voter registration rolls with other lists, for example, of felons, double-registered, and deceased people. Such crosschecks lead to voter suppression because they generate far more false than true positives. A study by professors at Harvard, the University of Pennsylvania, and Stanford found that purging registration rolls using the crosscheck system “would eliminate about 300 registrations used to cast legitimate votes for every one registration used to cast a double vote.” As applied by advocates like Kobach, matching has relied on flawed lists biased against minority voters as in Florida’s 2000 election. In 2004, a public outcry stopped Florida officials from using a similarly flawed purge list.⁹⁴

Kobach has stacked the commission with like-minded members, one of whom, Hans Von Spakovsky, argued against appointing any Democrats or “mainstream Republicans” to the commission, because such membership “would guarantee its failure.” Had it continued, the commission might well have recommended crosscheck purges, strict voter photo ID laws, proof of citizenship requirements, and other restrictive measures. One invited witness at a commission hearing went so far as to propose making every voter pass the background check used for gun purchases, which would disenfranchise millions of otherwise eligible voters. The commission has been plagued with more than fifteen lawsuits against its data requests and allegedly secret practices, including a successful case filed by one of the commissioners, Matthew Dunlap, Maine’s Democratic secretary of state, charging that he had been denied access to internal documents and commission activities.⁹⁵

On January 3, 2018, President Trump abruptly dissolved the fraud commission by executive order. A White House statement said that the president had terminated the commission because “many states have refused to provide [it] with basic information relevant to its inquiry” and because the president sought to avoid “endless legal battles at taxpayer expense.” However, the statement reiterated the claim that there is “substantial evidence of voter fraud,” even though the fraud hunters on the commission had uncovered no such evidence in its eight months of work. The statement indicated that the failed fraud-hunting of the commission could reemerge under the auspices of the Department of Homeland Security, with no input from Democrats or any independent review. An indictment by Special Counsel Robert Mueller of thirteen Russian nationals and three companies charged that Russian trolls in 2016 echoed Trump’s false claims of rampant voter fraud by Democrats. Republican senatorial candidate Roy Moore, who lost a December 2017 special election for U.S. Senate in heavily Republican Alabama, struck a similar theme, blaming fraudulent voting for his defeat by more than 21,000 votes. However, a state judge dismissed Moore’s effort to block certification of the election, and Alabama’s Republican secretary of state, John Merrill, who supported Moore, found not a single instance of any type of election fraud.⁹⁶

The advent of the Trump administration has threatened Justice Department backing for voting rights litigation. Attorney General Jeff Sessions reversed his department’s position on the Texas voter identification law, now taking the stand that the law did not intentionally target minority voters. He similarly reversed the department’s opposition to Ohio’s voter purging law, now asserting that the state’s practice is consistent with federal law, and he dropped the department’s support for the ruling of a three-judge court that Texas’s congressional redistricting plan intentionally discriminated against minorities. In 1986, a Republican Senate had rejected Sessions’s nomination by President Ronald Reagan for a federal judgeship, because of credible allegation of the nominee’s racism. He was only the second nominee in fifty years to be denied a federal judgeship by the Senate.⁹⁷

With America standing nearly alone among nations in lacking a constitutional right to vote, our democracy finds itself at a crossroads, with crucial decisions pending in courtrooms and legislatures. The U.S. Supreme Court has an opportunity in pending cases cited above in Wisconsin, Maryland, and North Carolina for the first time to strike down redistricting plans as illicit partisan gerrymanders. The Court is reviewing a congressional redistricting plan that a lower court found had intentionally discriminated against minorities and a voter purge law in Ohio in which nonvoting triggers a process for removing people from the registration rolls.

Pending in the legislatures of most states are bills to either expand or restrict opportunities to register and vote. Expansive measures include legislation to automatically register citizens who apply for driver’s licenses, to expand early voting, to adopt independent redistricting commissions, and to authorize registration on election day. Restrictive measures include voter photo ID laws, documentary proof of citizenship for registration, onerous residency requirements, and cutbacks in early voting. At least a dozen states are poised to adopt stringent voter purge legislation if the Supreme Court sanctions the Ohio law. With positions on voting rights mostly falling along party lines, the nation may increasingly become bound to a dual system of political rights, one for blue states and another for red states. Americans’ right to vote is in danger of becoming dependent on their place of residence, just as it once depended on their property-holding, their race, or their sex.

An international assessment reflects the fragile state of democracy in the United States. The toxic mix of low turnout, partisan gerrymandering, and rising cynicism in government has cost the United States its leadership place as the world’s beacon of democracy. The 2017 report of the Economic Intelligence Unit in Great Britain, which tracks the status of democracy among 167 nations across the world, classified only nineteen nations, with 4.5 percent of the world’s population, as “full democracies” and fifty-seven nations, with 44.8 percent of the population, as “flawed democracies.” It classified the remaining nations with slightly more than half the world’s population as “authoritarian” or “hybrid” regimes (with a mix of democratic and authoritarian features). Among democracies, the United States had for the first time in 2016 fallen from a full to a flawed democracy, a classification that remained unchanged in 2017, with a ranking of just twenty-first, tied with Italy.⁹⁸

The United States has declined in its rank among democracies because of partisan gerrymandering and public cynicism about government, fueled in part by repeated, unsubstantiated claims of voter fraud. According to a September 2016 AP-NORC poll, “35 percent of respondents said there is a great deal of election fraud in the United States, 39 percent say there is some election fraud. This poll found that only 24 percent say there is hardly any voter fraud, which is consistent with the outcome of multiple studies and court findings.”⁹⁹

Freedom House, another international monitor, found that the election of Donald Trump accelerated the decline of democracy in the United States. It reported that the Trump administration’s lack of transparency, disregard for ethical standards, and failure to safeguard U.S. elections from foreign manipulation posed unprecedented threats to American democracy. It found that 2017 “brought further, faster erosion of America’s own democratic standards than at any other time in memory.” America’s decline as a beacon of freedom and democracy, the study found, contributed to the erosion of democracy worldwide and the emboldening of authoritarian rulers. “Democracy is facing its most serious crisis in decades,” warned Freedom House’s president, Michael J. Abramowitz. “Democracy’s basic tenets—including guarantees of free and fair elections, the rights of minorities, freedom of the press, and the rule of law—are under siege around the world.”¹⁰⁰

White House spokesperson Sarah Huckabee Sanders indicated on March 20, 2018, that the Trump administration had abandoned America’s historic role as a champion of democratic elections worldwide. When asked if Vladimir Putin’s sham reelection as Russian president was “free and fair,” Sanders responded, “We’re focused on our elections. We don’t get to dictate how other countries operate. What we do know is that Putin has been elected in their country, and that’s not something that we can dictate to them—how they operate.”¹⁰¹

America can reclaim its place in the front ranks of the world’s democracies. The United States, which once inspired emerging democracies across the world, now has much to learn from the experiences of other nations and its own history. Simple, practical reforms are within reach to enhance access to the vote in America, end discriminatory practices, and help assure that people’s votes will count effectively in the election of public officials. But reform will come only if the American people are willing to see past the smokescreen of voter fraud, and demand real change to expand opportunities to register and vote. Despite the lack of a constitutional guarantee, the right to vote remains the bedrock of all other rights in a democratic nation.