CHAPTER ◆ 5
Democracy
On December 11, 2000, the Supreme Court held oral arguments, and the eyes of the nation were on Sandra Day O’Connor. The stakes that day could hardly have been higher: the presidency. The election contest between Democrat Al Gore and Republican George W. Bush would be won by whoever received Florida’s electoral votes, and the case the Court was deciding would play a large role in determining who that would be.
The margin in Florida was razor-thin, and there were many irregularities in the state’s results. Among them were tens of thousands of “undervotes,” ballots that had not registered a vote for president, in many cases because of the state’s glitchy voting machines. Both sides rushed into court, and state and federal courts began issuing a flurry of rulings. On December 8, with Bush officially leading by just 537 votes, the Florida Supreme Court ordered a statewide manual recount of the undervotes, which might well have changed the outcome of the election. The Supreme Court halted the recount, however, and scheduled arguments in Bush v. Gore.
The Court that heard Bush v. Gore had five conservative justices and four liberal ones. O’Connor was the swing justice, the most moderate of the conservatives and the one who most often broke ranks with them. If the Court just voted its political preferences, Bush would win 5–4 and he would become president. Many Court watchers, however, thought the Court would not want to resolve one of the most important cases in history by dividing on ideological lines, giving the appearance that it was a partisan body. In other historic rulings, including Brown v. Board of Education and United States v. Nixon, which ordered Nixon to turn over the White House tapes and ended his presidency, the Court had spoken unanimously. If anyone was going to break ideological ranks or lead the Court to a bipartisan solution, O’Connor seemed to be the justice most likely to do it.
At oral argument, however, O’Connor shocked many people in the gallery. In an exchange with Gore’s lawyer over what standard should be used to count undervotes, many of which had the notorious “hanging” chads—caused when the chad in a paper ballot was not completely punched out—O’Connor sounded like the most partisan of the Bush lawyers. Rather than sympathize with voters who had been thwarted by Florida’s flawed voting technology, O’Connor blamed them. “Well, why isn’t the standard the one that voters are instructed to follow, for goodness’ sakes?” she asked testily. “I mean, it couldn’t be clearer.”
It did not appear from the exchange that O’Connor was going to break ranks with the conservative justices, and she did not. On December 12, the Court ruled for Bush by a 5–4 vote, with O’Connor in the majority. The decision handed Bush Florida’s electoral votes, and ultimately the presidency.
The Court that decided Bush v. Gore was a direct heir to the Nixon Court that formed in January 1972, when Rehnquist and Powell arrived. There had been considerable turnover in the intervening twenty-nine years, but the Court’s ideological composition remained what it was then: a conservative chief justice presiding over a conservative majority. There was another point of continuity with 1972: the chief justice was Rehnquist, who, as one of Nixon’s assistant attorneys general, had helped to drive Fortas off the Court and then was plucked from obscurity to be one of the four Nixon justices.
To many observers, Bush v. Gore had every appearance of being a political ruling. The five conservative justices voted to stop the ballot counting in Florida, allowing Bush to win. The four liberal justices voted for Gore. Nothing in the justices’ analysis of the legal issues led a single one to vote in favor of a candidate who did not share his or her ideology.
Another reason the decision appeared to be political was that its legal reasoning made little sense. The majority opinion, which was unsigned, stopped the recount that the Florida Supreme Court had ordered, holding that it would violate Bush’s rights under the Equal Protection Clause. The explanation it gave was that there were not uniform standards from county to county for deciding which ballots counted. Some counties might count a ballot with a dimpled chad while others would not. There was a danger, the majority said, of “arbitrary and disparate treatment” of voters in different parts of the state. That disparate treatment, it said, would violate the Equal Protection Clause.
What the analysis omitted was that the American system of voting has always been rife with disparate treatment, and it is too diffuse, chaotic, and underfunded to expect any kind of uniformity. Voting varied from county to county across the country, not just in Florida. Some jurisdictions used lever machines and others used electronic voting machines. Some areas had long lines to vote on Election Day while others did not. The Court had never said that election mechanics had to be uniform throughout a state, and as a practical matter, it would require herculean efforts and a great deal of money to even approach that standard. The right to uniform election mechanics, it appeared, applied only to presidential recounts in Florida and belonged only to George Bush.
The Court’s equal protection holding was also hypocritical. The conservative Court that formed after Nixon took office had ended the Warren Court’s “rights revolution,” much of which had been based on reading the Equal Protection Clause broadly. When the Burger Court ruled against Linda Williams in Dandridge v. Williams, it insisted that a government policy is not unconstitutional merely because it is “not made with mathematical nicety or because in practice it results in some inequality.” In the years since, skepticism toward equal protection claims had become an article of faith for the conservative justices. One Harvard law professor wrote after Bush v. Gore, with only mild exaggeration, that those justices “almost never find equal-protection violations (except, perhaps, when white people are ‘discriminated’ against by affirmative action).”
The conservative justices all but admitted their bad faith in Bush v. Gore’s most notorious feature: the statement in the majority opinion that the decision was “limited to the present circumstances.” It is a fundamental principle of American law that court rulings have precedential value, and that Supreme Court decisions are binding on the lower federal courts. Scalia, one of the five justices in the majority, had written in an opinion only a few years earlier that the Supreme Court “does not sit to announce ‘unique’ dispositions.” The Court’s “principal function,” he said, was “to establish precedent—that is, to set forth principles of law that every court in America must follow.” In Bush v. Gore, however, the majority announced that its decision would be just the sort of “‘unique’ disposition” that Scalia had disclaimed. To the decision’s critics, its “limited to the present circumstances” language was proof that the Court had, as one commentator put it, “explicitly embraced lack of principle, ad hocery, vulgar partisanship.”
The dissenting justices strongly objected to the Court’s decision to stop the vote counting and award the presidency to Bush, though they fell into two camps in their legal analysis. Stevens and Ginsburg disagreed with the Court’s holding that the lack of a single standard for counting votes violated the Equal Protection Clause. Souter and Breyer agreed with the majority that there was an equal protection violation, but they argued that there was still time to fix it and that the Court should have sent the case back to Florida courts to establish uniform standards.
What all four dissenters agreed on was that the Court was not only wrong on the law, but guilty of an overreach of historic proportions. Stevens ended his dissent by declaring that “although we may never know with complete certainty the identity of the winner of this year’s Presidential election, the identity of the loser is perfectly clear. It is the nation’s confidence in the judge as an impartial guardian of the rule of law.” Ginsburg seemed most indignant of all. While the other dissenters wrote in their opinions that they “respectfully” dissented, as is customary, Ginsburg wrote only “I dissent.”
Bush v. Gore was met with vociferous criticism. The New York Times, in an editorial, declared that the Court’s decision to stop the recount in Florida “comes at considerable cost to the public trust and the tradition of fair elections.” Six hundred seventy-three law professors signed a statement condemning the Court for using “its power to act as political partisans, not as judges of a court of law.” Many scholars expressed outrage in their own articles and essays. Margaret Jane Radin, a professor at Stanford Law School, declared that Bush v. Gore made it difficult for her to remain in the “intellectual heights” of the academic world. “When I was faced with a gross, bald-faced violation of the rule of law,” she wrote, “I wanted to protest in the streets.”
When the Supreme Court spoke, however, the nation complied. Bush, who lost the popular vote to Gore by more than 500,000 votes, was awarded Florida’s electoral votes, and with them he became the first president since Benjamin Harrison, in 1888, to be elected while losing the popular vote. Exactly twenty-four hours after the Court ruled, Bush addressed the nation from a podium in the Texas House of Representatives and pledged to serve the entire nation “whether you voted for me or not.” Gore, speaking from his vice presidential office in Washington, declared that, while he differed with the Court’s decision, “I offer my concession.”
There was one more winner in Bush v. Gore: the conservative majority on the Court. If Gore had become president and managed to win reelection, he would certainly have chosen a liberal to replace William Rehnquist when he died, in 2005. There would have been a liberal chief justice with a liberal majority behind him—a new ideological orientation that could have lasted for decades. The five conservative justices who stopped the voting were not only choosing the next president—they were ensuring that the conservative Court that Nixon had established in 1972 lived on into the twenty-first century.
Bush v. Gore had enormous significance for American history, for the Supreme Court, and for constitutional law. What was largely lost in the early reaction to the case, however, was what it meant, specifically, for election law. In its rush to stop the Florida recount and put Bush in the White House, the Court had issued a landmark decision on the law governing elections, with a deep ambivalence at its center. In principle, the Court had handed down a bold ruling insisting that election rules had to be highly uniform to meet a very demanding standard of equal protection. Practically, however, it had used that lofty principle to stop validly cast ballots from being counted—and it had told courts never to apply that standard of election fairness again.
The Court’s attitude toward voting was once far less ambivalent. Expanding voting rights was a central part of the Warren Court’s “rights revolution,” starting with legislative districting. In the early 1960s, state legislative districts were not required to be equal in population, and in many cases they were wildly unequal. In Alabama, Jefferson County, with a population of 634,864, had the same number of state senators—one—as Lowndes County, with a population of 15,417. The Warren Court issued a series of landmark rulings, starting with the 1962 case Baker v. Carr, that established the principle of “one man, one vote,” requiring both state legislative and congressional districts to have roughly equal populations. After these rulings, district lines were redrawn nationwide, making state legislatures and Congress dramatically more democratic. When Warren was asked at the end of his career what he considered the most important case in his time on the Court, he said it was Baker, because it “gave to the courts the power to determine whether or not we were to have fair representation in our governmental system.”
The Warren Court also strongly supported Congress’s power to protect the rights of minority voters. In 1966, it upheld the “preclearance” provision of the Voting Rights Act of 1965, which required certain states and localities to get approval from the Justice Department or a federal court before they could make changes in their voting rules that could make it harder for minorities to vote. In the same year, the Court struck down the poll tax, which had long been used to keep poor and minority voters from casting ballots.
It was hard to know how Bush v. Gore fit into this tradition. Even though the Court had stopped votes from being counted and said the decision had no precedential value, its formal holding appeared to help future voting rights claims, since it insisted on a high level of equality in the administration of elections. While the media and most Americans focused on the decision’s impact on the presidential election, some legal scholars and voting rights advocates were considering whether Bush v. Gore could be used to make elections better-run and fairer. It could be “an advancement in voting rights doctrine,” one prominent election law professor wrote in an op-ed piece in The New York Times two days after the ruling. “It has asserted a new constitutional requirement: to avoid disparate and unfair treatment of voters.”
If the Court had been principled about providing all Americans with the exacting level of equal protection in elections that it insisted on for Bush, Bush v. Gore could indeed have been a major “advancement.” Elections are administered in notoriously unequal ways that almost always disadvantage poor and minority voters. Studies have shown that blacks and Latinos are forced to wait in longer lines to vote than whites are—nearly twice as long in the 2012 election, according to a Massachusetts Institute of Technology voter survey. Early voting is often administered unevenly, with some areas having many early voting locations while others have few or none. There are also significant differences in the failure rates of various voting methods, with punch card systems generally failing to record votes more often than newer electronic machines. In theory, Bush v. Gore should have provided a powerful legal weapon for ending these and other voting disparities. It is difficult to see how it could have, however, when the justices who signed on to its broad equality holding did not actually believe in it, and the Court told other courts not to follow it. Bush v. Gore cannot be taken seriously, Richard Hasen, a professor at the University of California–Irvine Law School, has said, “because the Court itself did not take its holding seriously.”
History has proven the skeptics right. The Supreme Court for years acted as if Bush v. Gore had never existed. No opinion cited the case until 2013, when Thomas “did what wasn’t supposed to be done,” as one commentator described it. The Court that year struck down an Arizona law that required people to submit proof of citizenship to register to vote, holding that it conflicted with a federal voter registration law. Thomas, in a solo dissent, cited a relatively obscure part of Bush v. Gore: its holding that state legislatures have broad authority in choosing their state’s electors to the Electoral College. He argued that it meant that the Arizona Legislature should be able to require voters to submit proof of citizenship. So the first time Bush v. Gore appeared in a Supreme Court opinion, it was cited as a reason for making it harder for people to vote. Lower federal courts have occasionally cited Bush v. Gore to argue that election procedures should be fairer and more equal, but it is not an idea that has caught on.
Underneath Bush v. Gore’s words about “the equal dignity owed to each voter” was a dark view of elections and democracy. Lani Guinier, a Harvard law professor and former head of the NAACP Legal Defense and Educational Fund’s voting rights project, argued in 2002 that Bush v. Gore’s real holding did not lie in what it said—all of the language about equality and fairness in voting. It lay, rather, in what it did: stop Florida from counting many ballots cast by eligible voters. Florida had hard-to-use voting technology that prevented many Floridians—particularly the poor, the elderly, and racial minorities—from casting ballots that were counted in the final vote tallies. The Court weighed in on the side of the flawed technology, not the thwarted voters. It believed, according to Guinier, that voters “had to pass a test to have their ballots counted, and the implicit suggestion was that only those who passed this test actually deserve to participate in the democratic process.” Its real message, she said, was that democracy is a “domain of governing elites, not robust and engaged citizens.”
In the years since Bush v. Gore, in cases involving voting, the Court has repeatedly sided with “governing elites” over “robust and engaged citizens.” One area where this has notoriously been true is challenges to partisan gerrymandering, the practice of drawing legislative district lines to help one political party win. In addition to tilting elections toward the favored party, partisan gerrymandering takes power away from ordinary voters, because it means that in many cases, elections are all but decided before the voting even begins. Partisan gerrymandering is a new version of the sort of anti-democratic legislative line drawing that the Warren Court struck down in the Baker v. Carr line of cases. The post-1969 Court’s approach, however, has been very different.
The first major test of partisan gerrymandering after Bush v. Gore came in 2004, in Vieth v. Jubelirer. Democrats challenged Pennsylvania’s highly gerrymandered congressional districts, which had been drawn after the 2000 election, when there was a Republican governor and Republicans controlled the legislature. Under pressure from Karl Rove, President Bush’s political adviser, and congressional Republicans, the legislature drew lines designed to create as many districts as possible that would elect Republicans to Congress.
Democrats argued that the plan violated voters’ equal protection rights. They showed that the legislature had gone to extraordinary lengths to draw districts that would elect Republicans. Although there were more registered Democrats than Republicans in Pennsylvania, Democrats were likely to win only seven of the nineteen congressional seats, or less than 37 percent.
The Court ruled 5–4, along the same ideological lines as in Bush v. Gore, that the Democrats had not made a strong enough case. Scalia, writing for himself, Rehnquist, O’Connor, and Thomas, argued that partisan gerrymandering cases were “nonjusticiable,” or not resolvable by the courts, because there was no workable constitutional standard for deciding them. Kennedy, who provided the fifth vote for the majority, wrote a separate opinion saying he believed there could be a case in which the challengers proved that a partisan gerrymander was unconstitutional, and in which the Court identified a workable standard, but this was not that case.
Stevens, in dissent, insisted that the Court had a good basis for striking down Pennsylvania’s district lines but that the justices in the majority simply did not want to. The problem was not the lack of a workable standard, he said, but “a failure of judicial will to condemn even the most blatant violations of a state legislature’s fundamental duty to govern impartially.” Stevens did not invoke Bush v. Gore, but he could well have pointed out that the Court had forgotten the commitment it expressed to “the equal weight accorded to each vote and the equal dignity owed to each voter.”
Challenges to partisan gerrymandering did not end with Jubelirer. The main reason was that Kennedy had held open the possibility that in a future case he might provide the fifth vote to strike down gerrymandered district lines. Another reason the issue stayed alive was that it was a large and growing threat to democracy. As voting data became more voluminous and redistricting software more powerful, legislators were able to draw their lines with greater precision, locking in majorities for their parties more effectively and making voters’ wishes ever less relevant—a phenomenon that was being described as “extreme gerrymandering.”
The Court returned to partisan gerrymandering in 2018. It accepted a pair of cases that presented the issue squarely and on a bipartisan basis, one a challenge to Republican-drawn legislative lines in Wisconsin, the other a challenge to Democratic-drawn lines in Maryland. Many Court watchers expected Kennedy to follow through on the promise of Jubelirer and vote with the liberal justices to strike down partisan gerrymanders. At the oral argument, the focus was on him. When it was over, CNN headlined its story “Anthony Kennedy Doesn’t Tip His Hand in Gerrymandering Case.” After all the buildup, on June 18, the Court found technical reasons in both cases to let the district lines stand, and partisan gerrymandering law remained the same.
Nine days later, Kennedy announced his retirement. President Trump nominated Brett Kavanaugh, a conservative judge on the U.S. Court of Appeals for the D.C. Circuit, to replace him. Kavanaugh, who had worked for the independent counsel investigating President Bill Clinton and for the Bush campaign in the 2000 Florida recount, was a favorite of movement conservatives. They expected that he would be a solid fifth vote for the Court’s conservative bloc, rather than an unreliable, at times wavering one, as Kennedy had been. The confirmation battle over Kavanaugh was one of the most hard-fought in history, with Dr. Christine Blasey Ford, a psychology professor at Palo Alto University, testifying that Kavanaugh had attempted to rape her when they were teenagers, a charge the nominee vigorously denied. Despite the controversy, on October 6, the Senate approved Kavanaugh by a 50–48 vote. His confirmation fulfilled “a long-held dream of conservatives,” The New York Times reported.
Within six months of Kavanaugh’s arrival, the Court heard arguments in another partisan gerrymandering case. Rucho v. Common Cause was a challenge to congressional district maps in North Carolina and Maryland. North Carolina was roughly evenly divided between Democrats and Republicans, but the Republican-controlled legislature drew lines that led to Republicans winning about 77 percent of the state’s congressional seats. In Maryland, Democrats redrew the lines of a Republican congressional seat, moving hundreds of thousands of voters into or out of the district, until it became a solid Democratic seat.
The arrival of Kavanaugh made all the difference: the conservatives finally had the five votes they needed to hold that partisan gerrymandering was nonjusticiable. Roberts, who wrote the Court’s opinion in Rucho, declared that the case presented “political questions beyond the reach of the federal courts.” The Court finally abandoned Kennedy’s search for an appropriate standard to use in partisan gerrymandering challenges. “There are no legal standards discernible in the Constitution for making such judgments,” Roberts said, “let alone limited and precise standards that are clear, manageable, and politically neutral.”
Kagan, writing for the four liberals, delivered a blistering dissent, which she read from the bench. Kagan castigated the majority for deciding that the legal issues presented by partisan gerrymandering were beyond the Court’s reach. “Of all times to abandon the Court’s duty to declare the law, this was not the one,” she wrote. “The practices challenged in these cases imperil our system of government.” The Court had a duty, Kagan said, to defend the “foundations” of American government, and “none is more important than free and fair elections.” Kagan said that she dissented with “deep sadness.” Once again, the Court had reduced voting to what Guinier called “an existential gesture,” in which “ordinary people” were limited to “the act of casting a ballot that may not count.”
The Court has also deferred to “governing elites” on another issue that is fundamental to the working of American democracy: voter ID laws. In 2008—after Vieth v. Jubelirer and before the Wisconsin and Maryland gerrymandering cases—it decided a challenge to Indiana’s voter ID law. Indiana’s Democratic Party and groups representing elderly, disabled, poor, and minority residents argued that the law, which was one of the strictest in the nation, violated the Equal Protection Clause. Once again, the “robust and engaged citizens” challenging the law could not find a majority on the Court to take their side.
Crawford v. Marion County Election Board came at a pivotal time: the run-up to the 2008 election, which would choose the successor to President George W. Bush. It was no secret that voter ID laws were a part of the Republicans’ strategy for keeping their hold on the White House. The Republican voting base—older, whiter, wealthier, and less urban—was more likely to have driver’s licenses and other government-issued ID than the Democratic base. One study found that 25 percent of black citizens of voting age lacked current government-issued photo ID, compared with 8 percent of white voting-age citizens. According to the same study, at least 15 percent of voting-age citizens who earned less than $35,000 a year lacked valid government-issued photo ID, more than twice the percentage of those earning more than $35,000.
National Republican strategists, including Karl Rove, had worked to put laws like Indiana’s in place. They had help from the American Legislative Exchange Council, a business-supported group that lobbies state legislatures. Indiana’s law, which was part of this national effort, was pushed through the Indiana General Assembly in 2005 by the Republican leadership without winning a single Democratic vote.
Advocates for voter ID laws argue that they are needed to prevent voter fraud. When they speak more honestly, however, they often admit that their actual purpose is to suppress voting among Democratic constituencies, particularly black voters. From time to time these statements, generally made behind closed doors, become public. One aide to a Republican state legislator in Wisconsin spoke about attending a Republican caucus meeting on a pending voter ID bill where “GOP Senators were giddy about the ramifications and literally singled out the prospects of suppressing minority and college voters.”
The Crawford of Crawford v. Marion County Election Board was William Crawford, an Indiana House of Representatives member whose Indianapolis district was one of the poorest in the state. He heard often from his constituents that they did not have the ID they needed to vote. Until the new law was enacted, Indiana voters only had to sign their names at the polls. Afterward, they needed a valid photo ID issued by the federal or Indiana government—more than was required in many other states, which accepted a wider variety of ID, including employee ID, or did not require ID at all. The Indiana Legislature expected most voters to use a driver’s license or a state-issued non-driver ID. To get either one, however, a prospective voter had to produce a birth certificate, a passport, or similar documents, which many did not have. Fees for birth certificates, the document many Indianans would have to use, could be up to $20 or more, which made the law resemble a poll tax.
The challengers also showed that there was no need for the voter ID law. There was no evidence that in-person voter fraud occurred to a significant extent in Indiana or anywhere else in the country. Having people who are not eligible vote in person is an ineffective way to steal an election, far harder to pull off on a mass scale than other tactics, such as absentee ballot fraud. One study released after the Indiana case was decided, by professors at the University of Wisconsin and Stanford University, concluded that the proportion of people reporting voter impersonation in a presidential election was “indistinguishable from” the proportion reporting “abduction by extraterrestrials.”
On April 28, 2008, the Court in Crawford upheld Indiana’s voter ID law, by a 6–3 vote. It divided along ideological lines, with the conservatives in the majority and the liberals in dissent, with one exception. Stevens—a Ford nominee who had become a reliable part of the liberal bloc—gave the conservatives a sixth vote. Writing what amounted to the opinion of the Court, he said the law advanced Indiana’s legitimate “interest in protecting the integrity and reliability of the electoral process.” On the other side of the balance, he said, the challengers had not shown that the law “imposes ‘excessively burdensome requirements’ on any class of voters.” Stevens left open the possibility that if the Court was presented with a stronger factual record at some point, it could come out the other way.
In dissent, Souter said there was “no reason to doubt” that many Indianans would be discouraged or prevented from voting. Against that harm, he said, the state’s interests were minor. There was, he said, simply “no evidence of in-person voter impersonation fraud in” Indiana. The case, Souter maintained, was similar to the 1966 case in which the Court struck down the poll tax. In both circumstances, he said, the laws at issue wrongly made wealth a consideration in eligibility to vote.
It took little time for Stevens to be proven wrong in his belief that the law did not impose “‘excessively burdensome requirements’ on any class of voters.” Just over a week after the ruling, on primary day, twelve nuns in their eighties and nineties, including one who was ninety-eight, were prevented from voting at the polling place at Saint Mary’s Convent, in South Bend, because they did not have a federal or state photo ID. None of the nuns had driver’s licenses, but some brought outdated passports. “Indiana’s Voter ID law applies to everyone,” Indiana’s secretary of state said when the nuns were turned away. “From all accounts that we’ve heard, the sisters were aware of the photo ID requirements and chose not to follow them.”
Critics of the Crawford decision predicted that it would encourage more states to adopt restrictions like Indiana’s, and that is, in fact, what happened. Since 2008, voter ID laws have been adopted in Wisconsin, Texas, Pennsylvania, Missouri, Alabama, Mississippi, Arkansas, Iowa, Kansas, Rhode Island, South Carolina, Tennessee, Virginia, Idaho, North Carolina, West Virginia, Oklahoma, Utah, New Hampshire, and North Dakota. Some of the laws are stricter than others, but all make it more difficult to vote.
There is evidence that voter ID laws are doing their job, the one they were really enacted for: suppressing the vote from certain parts of the electorate. After the 2016 election, a study of two large Wisconsin counties estimated that the state voter ID law may have kept nearly 17,000 voters from the polls in those counties alone—in a state Donald Trump won by 22,748 votes. It also found that black registered voters were more than three times as likely as white registered voters to be deterred from voting by the law, and that registered voters earning less than $25,000 a year were almost eight times as likely to be deterred as registered voters making $100,000 or more. The Wisconsin study was part of a growing body of research showing that voter ID laws disproportionately disenfranchise racial minorities and the poor. These studies confirm what Souter said in his Crawford dissent: that Indiana’s voter ID law “crosses a line” because it “targets the poor and the weak.”
In 2016, Stevens, who was then retired, discussed the Crawford decision with Kagan on a panel at a judicial conference in Chicago. He said he had learned of many problems with voter ID outside of the record in the case, but he felt he had to limit himself to the facts that the parties had presented to the Court. The result, Stevens said, was “a fairly unfortunate decision.” Judge Richard Posner, who had written the opinion for the Chicago-based U.S. Court of Appeals for the Seventh Circuit upholding the law before it got to the Supreme Court, was more direct. Posner, a prominent conservative nominated to the court by President Reagan, said a few years earlier that his court and the Supreme Court had gotten the case wrong. Indiana’s voter ID law, Posner said, is “a type of law now widely regarded as a means of voter suppression rather than of fraud prevention.”
In 2013, the Court had another chance to side with “robust and engaged citizens” who wanted to vote, when it heard a challenge to a key part of the Voting Rights Act of 1965. The act, one of the most important civil rights laws in history, was adopted after Martin Luther King Jr. led a campaign across the South protesting the intimidation, rigged literacy tests, and other tactics being used to prevent blacks from registering and voting. Demonstrators marching from Selma to Montgomery to demand voting rights were violently attacked by Alabama state troopers on the Edmund Pettus Bridge on March 7, 1965, which came to be known as “Bloody Sunday.” After the violence appeared on the network news, Congress passed the Voting Rights Act, and President Johnson signed it on August 6, with King and Rosa Parks, the heroine of the Montgomery Bus Boycott, looking on.
The Voting Rights Act allowed individuals to sue when state or local officials interfered with minority voting, but Congress understood how difficult it was to bring a lawsuit. They included a provision, Section 5, that was designed to avoid the need for litigation. It required states and localities to preclear possibly discriminatory changes in rules and procedures with the Justice Department or a federal court before they could take effect. Another provision, Section 4, set out which parts of the country would be subject to the preclearance requirement. It brought most of the South under the act, along with other jurisdictions outside of the South, based on a formula that took into account an area’s history of erecting obstacles to voting and its past registration and voting rates.
Congress provided that the Voting Rights Act would expire and have to be reauthorized, to ensure that the burdens it placed on states and localities were not imposed needlessly. It extended the act every time it expired, and presidents of both parties signed the extensions. President Reagan signed an extension in 1982, declaring that “the right to vote is the crown jewel of American liberties, and we will not see its luster diminished.”
When Congress reauthorized the Voting Rights Act in 2006, it knew the Court had become more hostile to voting rights. To strengthen the act for the next challenge, it created an extensive record to show why its protections were still needed. It held extensive hearings, which produced a more than fifteen-thousand-page record reviewing the ways in which minorities were still discriminated against in voting. The reauthorization passed the Senate 98–0 and the House 390–33, and George W. Bush signed it into law.
The case the Court heard in 2013, Shelby County v. Holder, challenged the Voting Rights Act’s preclearance process. Shelby County, Alabama, which brought the case, focused on Section 4, arguing that its formula for determining which jurisdictions should be covered was unconstitutional. If the Court struck down the Section 4 formula, it would effectively end the whole preclearance requirement, because there would be no valid list of which jurisdictions were required to seek preclearance.
In Shelby County, the Court, by a 5–4 vote, struck down Section 4’s coverage formula. Roberts, writing for the conservative justices, said that when Congress applied preclearance to some states and not others, it violated “the fundamental principle of equal sovereignty” among the states. That put a heavy burden on Congress to justify its decisions about which states to include, he said. Section 4 failed to meet this burden, according to Roberts, because it relied on “40-year-old data” on voting discrimination, from when the Voting Rights Act was first adopted, rather than information on current conditions.
Roberts couched his concerns in the language of equal treatment of the states, but he also made it clear that he did not believe discrimination in voting was still a very serious problem. There were, he noted, no more literacy tests or poll taxes. “Our country has changed,” he said. “Congress must ensure that the legislation it passes to remedy that problem speaks to current conditions.” He was not impressed or dissuaded by the fact that Congress had compiled a fifteen-thousand-page record of evidence to prove just this point.
Roberts conceded that declaring Section 4’s formula unconstitutional was an extreme step. “Striking down an Act of Congress” was “‘the gravest and most delicate duty that this Court is called on to perform,’” he said, quoting Justice Oliver Wendell Holmes. His concession, however, understated the enormity of what the Court was doing. It was not merely striking down any act of Congress. The Court was invalidating a law that was a crowning achievement of the civil rights movement, and one that had won overwhelming bipartisan support in Congress and from presidents for decades. The law also regulated a subject, voting in federal elections, that the Constitution specifically and repeatedly assigned to Congress. Article I of the Constitution states that Congress shall regulate the “times, places, and manner” of federal elections, and the Fifteenth Amendment, which bars racial discrimination in voting, says that Congress “shall have power to enforce this article by appropriate legislation.” The senators who heard Roberts say at his confirmation hearing that he intended to be an “umpire” on the Court, simply calling “balls and strikes,” could not have been expecting a ruling like this.
Making matters worse, the legal principle Roberts invoked to strike down Section 4 was a made-up one. This was not merely the view of the dissenters and the ruling’s many liberal critics. Judge Richard Posner of the U.S. Court of Appeals for the Seventh Circuit, the same judge who came to regret his opinion on voter ID, stated flatly after Shelby was decided that “there is no doctrine of equal sovereignty.” It was, he said, “a principle of constitutional law of which I had never heard—for the excellent reason that . . . there is no such principle.” Posner, a Reagan nominee and a prominent conservative, insisted that the whole majority opinion in Shelby County “rests on air.”
The other part of the Court’s reasoning that lacked a basis in reality was its suggestion that preclearance was no longer needed because the nation had changed so much. Ginsburg, in a dissent for the four liberal justices, insisted that a major reason there was less discrimination in voting was precisely because of the preclearance requirement, which prevented it. “Throwing out preclearance when it has worked and is continuing to work to stop discriminatory changes,” she wrote, in a memorable line, “is like throwing away your umbrella in a rainstorm because you are not getting wet.”
In fact, there was still a great deal of racial discrimination in voting, and Ginsburg provided examples. In Kilmichael, Mississippi, the white mayor and the all-white board of aldermen canceled the 2001 town election when an “unprecedented number” of black candidates announced and it appeared that power in the town would shift from whites to blacks. The Department of Justice ordered Kilmichael to hold an election, and the town elected its first black mayor and a majority-black board of aldermen.
When the Court struck down Section 4’s coverage formula, Section 5 was effectively nullified and preclearance ended. Congress could, in theory, enact a new Section 4, with a different formula for choosing covered jurisdictions, but it would be extremely difficult for members of Congress to agree on which parts of the country to include. Even if it did adopt a new formula, as long as it did not cover every jurisdiction in the country, it would be vulnerable to another challenge under the Court’s invented doctrine of “equal sovereignty.”
Supporters of voting rights were deeply disturbed by Shelby County’s brazenness. John Lewis, the Georgia congressman who had been beaten as a young man on the Edmund Pettus Bridge on Bloody Sunday, said that the Court had stuck “a dagger into the heart of the Voting Rights Act.” President Obama said he was “deeply disappointed,” noting that “for nearly 50 years, the Voting Rights Act—enacted and repeatedly renewed by wide bipartisan majorities in Congress—has helped secure the right to vote for millions of Americans.” He promised that his administration would “continue to do everything in its power to ensure a fair and equal voting process.”
Shelby County had an immediate impact, as jurisdictions across the country were emboldened to change their election procedures, freed from the requirement to get approval first. Five years after the decision, The New York Times noted the ways in which Alabama had made it harder to vote. Within twenty-four hours of the ruling, the state announced that it would start requiring photo ID, reviving a plan that had been blocked by the Voting Rights Act. Alabama then announced that it would close thirty-one driver’s license offices, one of the main places where voters obtained ID that could be used to vote. There was a clear racial pattern: eight of the ten counties with the highest percentages of black voters lost offices, compared with just three of the ten counties with the lowest percentages.
At the national level, a Pew Charitable Trusts report found that, five years after Shelby County, nearly a thousand polling places had been closed, many of them in black communities in the South. In Indiana, the Republican secretary of state eliminated 170 voting precincts in Lake County, where the state’s largest Latino and second-largest black communities were located. Reducing the number of polling places is an effective way of reducing voter turnout—and as a result of the Court’s ruling, it is happening a lot more.
In 2018, the Court handed down another ruling on the mechanics of democracy, this one on voter roll purges, and once again it upheld the right of “governing elites” to make it more difficult for ordinary citizens to vote. Voter roll purges have been a sensitive subject since 2000, when Florida conducted an error-filled purge before the presidential election that may have made the difference in George W. Bush’s victory. Republican elected officials hired a private firm to remove felons from the rolls and, owing to mistakes in how it was done, as many as twelve thousand voters who were wrongly labeled felons were taken off. In the end, Bush won Florida by just 537 votes. The U.S. Commission on Civil Rights, which investigated what went wrong with the voting in Florida in 2000, estimated that 4,752 black Gore voters were wrongly prevented from voting by the pre-election purge, and it concluded that the purge was “outcome-determinative.”
The justification for voter roll purges is that they promote election efficiency and integrity by removing people who have moved, died, or become ineligible. In many purges, however, voters are removed even if they are still alive, still living at their address, and still eligible to vote. They have simply failed to meet bureaucratic eligibility requirements that “governing elites” have put in place—requirements that often disenfranchise “low-interest” voters, who choose not to vote in some elections.
In other cases, voters are removed by mistake. Voter roll purges are often conducted haphazardly, and they frequently end up, as in Florida in 2000, with voters being wrongly disenfranchised. In 2016, the Arkansas secretary of state notified county clerks of more than fifty thousand voters who were convicted felons and directed that they be removed from the rolls. It later turned out that four thousand of the people on the list did not actually have felony convictions, and as many as 60 percent of the others flagged for purging had actually had their voting rights restored.
As with voter ID laws, the strongest supporters of expansive voter roll purges are Republican elected officials and political operatives, who seem convinced that more Democrats than Republicans will be taken off the rolls. There is reason to believe they may be right, and that the purges may be removing more minority and poor voters. A national study by the Brennan Center for Justice found that people were removed from the rolls at much higher rates in counties with a history of voter discrimination. It also reported that between 2016 and 2018, at least seventeen million voters were removed.
The 2018 Supreme Court case was a challenge to Ohio’s voter roll purges, which were among the most aggressive in the nation. In Ohio, voters who had not voted in two years—who had missed just one federal election—received mailed notices from the board of elections. If they did not respond and then did not vote in the next four years or have other interactions with the election system, such as signing a nominating petition, they were purged, even if they still met the legal requirements to vote.
Larry Harmon, the lead plaintiff, was a Navy veteran and software engineer living near Akron. When he went to vote in 2015, he was told that his name was no longer on the rolls. Harmon had voted in 2004 and 2008 but not in 2010, 2012, or 2014. Election officials said they sent a notice in 2011 asking him to confirm that he was eligible to vote, but Harmon said he did not recall receiving it.
The Cincinnati-based U.S. Court of Appeals for the Sixth Circuit ruled for Harmon, holding that the purge violated the National Voter Registration Act of 1993 (NVRA), a federal law that imposes conditions on how states register voters for federal elections. The NVRA has rules for purges, including a requirement that they not result in anyone being removed “by reason of the person’s failure to vote.” The Sixth Circuit said Ohio’s purge of Harmon, which was triggered by his not voting in some elections, violated the NVRA’s failure-to-vote clause.
On June 11, 2018, in Husted v. A. Philip Randolph Institute, the Supreme Court upheld Ohio’s purge by a 5–4 vote. Alito, writing for the five conservative justices, said the NVRA’s failure-to-vote provision prohibited a state from removing a voter from the rolls solely because of his or her failure to vote. It did not apply to the Ohio purge, he said, because failure to vote was only part of the process that led to his being removed.
In his dissent, Breyer agreed with the Sixth Circuit that the purge violated the NVRA’s failure-to-vote provision because it was failing to vote that got Harmon removed. He went on, however, to make a larger point about how Ohio’s purge undermined American democracy. The purpose of election processes, he said, was “not to test the fortitude and determination of the voter, but to discern the will of the majority.” Sotomayor, in a solo dissent, tied Ohio’s purge to larger “efforts to prevent minorities from voting and to undermine the efficacy of their votes,” which, she said, were “an unfortunate feature of our country’s history.”
Husted was important in part because of Ohio’s traditional role as a swing state with enormous influence in presidential elections. Its implications, however, went beyond any single state. Before the decision, Kristen Clarke, executive director of the Lawyers’ Committee for Civil Rights Under Law, said she was “deeply concerned” that if Ohio won, “we will see states taking a copy-cat approach” and working to “gut the voter rolls across the country.” In fact, those efforts were already under way. Even before the ruling came down, the Trump Justice Department had written to forty-four states asking questions about what they were doing to keep their voter rolls up to date. A head of the Justice Department Civil Rights Division under President Obama called the letter “virtually unprecedented” and warned that it was “a prelude to voter purging.”
Months after Husted, purges became an issue in the Georgia governor’s race, when it was revealed that the Republican candidate, Secretary of State Brian Kemp, had purged as many as 1.5 million voters between 2012 and 2016, and another 665,000 in 2017. More than 100,000 of the voters purged in 2017 were removed for not voting—the issue in Husted. Critics charged that Kemp had conducted the purges to increase his own chances of winning the gubernatorial election. He won the governor’s race by about 55,000 votes. In her concession speech, his Democratic challenger, Stacey Abrams, charged that Kemp had used “deliberate and intentional” voter suppression to win.
Voter roll purge rates have been increasing. The seventeen million voters removed from the rolls between 2016 and 2018 represented a sharp increase from 2006 to 2008, when about twelve million voters were purged, according to the Brennan Center for Justice. Partisan organizations have been suing states and localities to conduct more purges. One suit by a conservative group demanded that Noxubee County, Mississippi, a poor, majority-black county, purge its rolls more aggressively. “They went after minority counties who didn’t have the financial resources to push back,” a Noxubee County election commissioner said.
With the Court’s green light in Husted, aggressive voter roll purges are likely to become even more common, and more eligible voters will be removed. When the Brennan Center released a follow-up study after Husted, in the three states it focused on—Florida, Georgia, and North Carolina—it found “cause for alarm.” In her Husted dissent, Sotomayor said that “communities that are disproportionately affected by unnecessarily harsh registration laws should not tolerate efforts to marginalize their influence in the political process.” It was the right advice, but what was less clear was what those communities could do to stop it.
Bush v. Gore did usher in a new era for the law of democracy, but it was not one of fairer elections, as its equal protection holding seemed to promise. Instead, the Court all but banished the equality principles of the 2000 ruling from its collective memory and it built further on its true holding: that governing elites should be allowed to make it more difficult for ordinary citizens to vote. It was the exact opposite of the Warren Court’s approach to voting rights cases, which had always made voters’ rights paramount.
The Court’s election law decisions starting with Bush v. Gore and its campaign finance rulings starting with Buckley v. Valeo are a potent anti-democratic combination. The campaign finance decisions have expanded the rights of wealthy individuals and corporations to use their money to gain influence over government. The election law decisions on partisan gerrymandering, voter ID, the Voting Rights Act, and voter roll purges have diminished the ability of those with less money to use the one thing they have at their disposal to win influence over government: their votes.
The Court’s election law rulings since 2000 have made a real difference in whose votes count and who is elected. With partisan gerrymandering, the outcomes of many elections are now virtually foreordained. In North Carolina, after the 2010 census, Republicans drew lines designed to elect Republicans from ten of the state’s thirteen congressional districts. That is precisely what happened, even though the state’s electorate was almost evenly divided between Democrats and Republicans. When voting does not matter, or hardly matters, the poor and the middle class, who do not have much to offer elected officials except their votes, have little leverage.
The Court’s other significant rulings—on strict voter ID laws, the Voting Rights Act, and voter roll purges—have also weakened the influence of lower-income Americans. They have, as a group, helped to make the electorate wealthier and whiter. By shifting the demographics of voters in that way, the Court’s rulings have created an electorate that is more likely to favor policies that help the well-off, like lowering the top income tax rate, and less likely to demand ones that help the poor and middle class, like increasing the earned income tax credit for low-income workers.
There is also an unmistakable partisan slant to the Court’s major election law rulings since 2000: like Bush v. Gore, they have favored Republicans. In the case of partisan gerrymandering, while both parties engage in it, Republicans do it more. The Associated Press analyzed the 2018 election, when Democrats took a majority of the House of Representatives, and concluded that Democratic candidates would have won about sixteen more seats if the Republicans had not had a structural advantage resulting from how district lines were drawn. Not all of that structural advantage was due to partisan gerrymandering, but much of it was.
Strict voter ID laws also have a pro-Republican bias, since racial minorities and poor people—two core constituencies of the Democratic Party—are so much less likely to have acceptable ID. Republican political strategists certainly believe that strict voter ID laws will help their party, and that is why they have worked so hard to get them enacted. Weakening the Voting Rights Act makes it easier for election officials to suppress the votes of blacks and other minorities, which is also likely to help Republican candidates.
Making government more Republican hurts the economic standing of lower-income Americans. The Republican Party promotes a wide array of policies that harm the poor and the middle class, including opposing increases in the minimum wage, fighting labor unions, and resisting more generous social welfare programs. Larry Bartels, a political scientist at Vanderbilt University, studied how different economic classes fared under presidents of each major party. He found that since the late 1940s, the real incomes of middle-class families rose more than twice as fast under Democratic presidents as under Republicans, and the real incomes of working poor families rose ten times as fast under Democratic presidents. “Escalating inequality is not simply an inevitable economic trend,” Bartels said. A lot of the economic inequality in the United States today, according to Bartels, “is specifically attributable to the policies and priorities of Republican presidents.” There is every reason to believe Republican Congresses, governors, and state legislatures have similar effects.
The Court could have chosen a different kind of American democracy. In Bush v. Gore, it could have insisted that the primary constitutional value in elections is that all valid votes should be counted. In the cases that followed, it could have decided that partisan gerrymandering and strict voter ID laws violate equal protection, that aggressive voter roll purges violate the National Voter Registration Act, and that the entire Voting Rights Act is constitutional. It could have adhered to the Warren Court’s guiding principle that, as Earl Warren himself said, what matters most in election law cases is ensuring that there is “fair representation in our governmental systems.” Instead, the Court has taken an approach to election law that defers to the decisions of elites about which voters should be allowed to participate in democracy and whose votes should count. In doing so, it has created an electorate that is less inclined to support policies that help poor and working-class Americans—and a nation that is less likely to get them.