7

Inventing Democracy

Chuck Rhoades was the attorney general in New York.1 Harvey “Hap” Halloran was the commissioner of the board of elections there. And in the spring of 2019, they got into it over possibility. Rhoades wanted to allow people to vote over their mobile phones. Halloran was not having it. The US attorney general was objecting, allegedly over “cybersecurity concerns,” and Halloran was, at least at the moment, firmly on his side, telling Rhoades, “I share the attorney general’s objection.”

Rhoades rebutted, “I have an equally pressing concern: Democracy. And inclusiveness. This is why it’s vital we approve mobile voting. This is about more people voting, which is the American ideal. It’s about what’s right, which is your board’s mandate. Hap, let us seize this opportunity to be true patriots to bring a voice to those without one. I propose a pilot program to test the efficacy of mobile voting.”

Halloran turned snarky: “We are all moved. Our hankies are damp.” Then, seeking to end the discussion, he confessed, “We are looking to rule against, but thank you for dropping by.” Later, when Rhoades again reached out to Halloran, the recalcitrant elections commissioner asked, “What do you want?”

Rhoades, restating his case, said, “Open and fair elections.”

Halloran bent toward sarcasm again: “Yeah, well I would like to tap-dance like Gregory Hines, but I have neither the shoes nor the rhythm to pull it off.”

Indignant, Rhoades shot back, “No, you don’t have the will. But I do. Time to get my blockchain voting pilot program out of the gates.”

Halloran was unmoved: “You are not going to get a fucking pilot program.”2

Many of us who’ve labored at one point or another in public service can imagine ourselves as Rhoades in that moment. The fact is that even Rhoades had to imagine himself as Rhoades, because he wasn’t real. He was a television character played by award-winning actor Paul Giamatti on the sensational Showtime television show Billions. How the cast captured the back-and-forth among bureaucrats so granularly (and profanely) is simple to grasp. Just ask any of us who have been there. But how a storyline on the arcana of block-chain voting ended up on a Showtime hit, well, to understand that—and to decide for yourself whether some real-life Rhoades should get a real-life blockchain-enabled mobile-voting pilot to fix our real-life democracy—you have to reach further. You have to go back to an Indian polling place in 1984, to an Afghanistan hillside in 2012, and to the real New York City in 2015.

Nimit Sawhney remembers the first time he thought about protecting the vote. It was not long after India’s Prime Minister Indira Gandhi had been assassinated. “It was a . . . tough time,” he recalled, delicately. “There were elections immediately after, and while I was too young to vote, I did see people forced to vote at gunpoint.”3 Sawhney declined to reveal more about that day to me. But when I read reports of “booth capturing” back then, where armed gangs would attack a polling place and “shoot it out until one party takes command of the ballot boxes,”4 I began to understand why he said, “Maybe someday we’ll find a way to make sure that doesn’t happen again,” and why someday might be someday soon. Sawhney went on to build a career as a software engineer and worked in digital security. He moved to the United States. In 2014, he started prototyping a mobile-phone-based election system with coercion-detection and coercion-prevention capabilities. He began building a platform that would leverage blockchain technologies to make voting tamper-resistant. It was the kind of thing that Giamatti, as Rhoades, was advocating for on Billions, and it was, by 2016, available from the company Sawhney had turned his project into: Voatz.

Mac Warner’s first visions of mobile voting came when he was watching a shepherd tending his sheep. Warner was a US Army officer, on one of two tours in Afghanistan, and had had trouble voting. The obstacles that get in the way of voting while abroad—relying on snail mail to remote locations, tracking down fax machines, voting by email without privacy—had bedeviled Warner. He was frustrated. Then he felt frustrated for his kids, all four of whom had followed in his footsteps to serve their country abroad. And then he felt . . . possibility? Warner was on a mission and saw an Afghan shepherd on the side of a remote mountain tapping away on his mobile phone. The logical leap wasn’t too large: If the shepherd could be on his mobile phone out there to take care of his responsibilities, why couldn’t Warner? When the former army officer became the secretary of state for West Virginia in 2017, he told his team to fix overseas voting for West Virginians serving abroad. Mobile phones became the centerpiece of that initiative. “These people are putting their lives on the line. They have skin in the game, and they should be able to vote in our democracy.” A year later, West Virginia paired up with Voatz to try mobile voting for overseas voters for the May 2018 elections, which included the US congressional primary. Two West Virginia counties participated. Thirteen citizens stationed abroad voted from their mobile phones. Rhoades hadn’t yet gotten his “fucking pilot program,” but Warner had gotten his.

Bradley Tusk was doing battle of only the metaphorical sort when he first homed in on the power of mobile phones in voters’ hands. He had, after a career in politics, become Uber’s first lobbyist, its self-described political “fixer.”5 In 2015, New York Mayor Bill de Blasio was backing a plan that would have limited the number of new Uber drivers per year to less than what the company, at the time, was adding each week. Irritated, Uber added a “de Blasio” feature to its app in New York City, so that customers would see what New York might be like, in the company’s view, under de Blasio’s plan: long wait times or no cars at all. And Uber invited its customers, directly from inside the app, to email the mayor and the city council to “say ‘no’ to de Blasio’s Uber.”6 The customers did. De Blasio backed down, and Uber removed the feature.7 But Uber’s use of its app—and of the phones it resided on—didn’t start or end in New York City. Time and again, taking the lead of the company’s cofounder, Travis Kalanick, and with Tusk’s early help, the company activated its customers to text and tweet and talk on Uber’s behalf. Tusk explained, “The reason ridesharing is legal in all the jurisdictions in the US. . . . is because, over time, a couple of million people weighed in with elected officials, and that won the day. What that said to me is that if you give people the tools to advocate from their phones, they will vote.” Tusk dedicated much of his post-Uber time—and some of his post-Uber fortune (he had taken stock in the company as his consulting fee)—to the prospect of mobile voting. Tusk, with Sheila Nix, a political strategist who he’d tapped to lead his philanthropic efforts, would tell anyone who would listen (and many who wouldn’t) about mobile voting and its potential to increase voter participation and, therefore, to tilt democracy away from extreme voters and special interests. Tusk even extolled the virtues of mobile voting to one of the creators of Billions over dinner one night,8 but not before he had made it happen in real life. Tusk had funded the West Virginia pilot.

The Sawhney/Warner/Tusk experiment had come together hurriedly in the spring of 2018, but it had apparently gone off without a hitch. Plans were in place to open it up to overseas voters from all West Virginia counties for the upcoming November election. Tweaks were underway to make it even more robust. Then, on August 6, 2018, Kevin Beaumont, a self-described “EU security tweeter,” tweeting as @GossiTheDog, retweeted a CNN article about Voatz and the upcoming November election with his own commentary: “This is going to backfire.” He accused Voatz of operating with out-of-date security for remote logins. He added, “If a startup (I’m sure they’re nice people btw) with 2m in funding approaches and says they have biometric security and Blockchain it still need[s] independent vetting . . . There needs to be oversight here.” He went on, “I’m a foreign dude with an avatar of a cowboy porg riding a porg dog on Twitter who appears to have done more investigation of the security implications of this than anybody. Bonkers, America.”9

Bonkers, America? Sawhney and his team, Warner and his, Tusk and Nix—they had demonstrated that citizens and leaders, entrepreneurs and elected officials, could together experiment with the apparatus of democracy. But ought we? It’s not just a question on television. Once we’ve come to develop the tools for Possibility Government, what should we use them for? @GossiTheDog had a view on it. What should the rest of us think?

Democratic Recession

Around the time that Sawhney, Warner, and Tusk were making their visions a testable reality, three books came out that sounded the alarm bells on democracy: The People vs. Democracy, by Yascha Mounk; Can Democracy Work?, by James Miller; and How Democracies Die, by Steven Levitsky and Daniel Ziblatt. Their titles foreshadow their conclusions.

Mounk relays a trove of dire data on democracy. Less than half of young Americans are likely to say “that they take an active interest in politics.”10 Only 48 percent of those born in the 1970s do, and these aren’t kids.11 These are people (like me) with kids. It gets worse. Less than one-third of American millennials believe “that it is extremely important to live in a democracy . . . close to one in four millennials now think that democracy is a bad way of running the country.”12 One in six Americans “believe that army rule is a good system of government.”13 “The good news,” Mounk ventures, “is that the number of people who say that army rule is a good way to run America is indeed smaller than the number of people who hanker after a strongman who doesn’t have to bother with Congress or elections.” If that was the good news, I was afraid to learn what the bad news was, but Mounk gave it to us anyway. “The bad news is that [the number who say army rule is a good way to run America] is rising rapidly.”14

Mounk also reminded us what two renowned political scientists, Martin Gilens and Benjamin Page, had found in their research into the question of who rules? The economically elite, along with narrow special interests, had the most influence, they showed. “The preferences of the average American appear to have only a minuscule, near-zero, statistically non-significant impact upon public policy.”15 Near-zero? No wonder a majority of Americans distrust people in public life.

There’s apathy about our way of life; openness, even enthusiasm, for strongman rule; and antagonism toward our elected leaders. Then there’s antipathy toward our fellow citizens. Levitsky and Ziblatt offered the following observation: “Consider this extraordinary finding: In 1960, political scientists asked Americans how they would feel if their child married someone who identified with another political party. Four percent of Democrats and five percent of Republicans reported they would be ‘displeased.’ In 2010, by contrast, 33 percent of Democrats and 49 percent of Republicans reported feeling ‘somewhat or very unhappy’ at the prospect of interparty marriage.”16 Also bonkers, America?

And bonkers, world? The democratic retrenchment—a “democratic recession,” as political scientist Larry Diamond has called it—is present across the globe. “In much of Europe . . . citizens are less likely now than a few decades ago to believe their elected representatives prioritize the interests of the general public,” Mounk reported.17 He described a 2017 poll that showed rising support of strongmen leaders in Germany (33 percent), France (48 percent), and the United Kingdom (50 percent), double the numbers there from a similar poll in 1999.18

James Miller relayed another depressing, and global, trend:

For many years, the best-known such index was that produced by Freedom House, an American government-funded nonprofit organization dedicated (in its own words) “to the expansion of freedom and democracy around the world.” Ever since the late 1970s, it has graded countries on a scale of one to seven, measuring 10 indicators of political freedom and fifteen indicators of civil liberties, and then sorting 195 countries into three categories: free, partly free, or not free. Its annual report for 2018, titled “Democ racy in Crisis,” found that a total of 71 countries had suffered net declines in political rights and civil liberties, compared with only 35 that registered gains: “This marked the 12th consecutive year of decline in global freedom.”19

If Possibility Government is Mayor Peduto and testing robot cars and “the status quo is the dangerous choice,” then the case for possibility democracy seems, by these measures, open and shut. Doing nothing looks to be awfully risky.

Blinking Red

The problem is, as it almost always is, that doing something is risky, too.

If democracy in 2018 was atrophying where it resided, it was also under assault from its adversaries. “Today, the digital infrastructure that serves this country is literally under attack,” the US director of national intelligence, Dan Coats, warned in July of that year. “I’m here to say the warning lights are blinking red,” he said, using the same language that had been deployed, unsuccessfully, ahead of the 9/11 attacks on the United States. He named several “offenders”—China, Iran, and North Korea—but he said that Russia was the “most aggressive foreign actor, no question. And they continue their efforts to undermine our democracy.”20 Continue . . . because what was widely understood, if not universally acknowledged, was the Russian intervention in the 2016 presidential election. The colloquially named Mueller Report was actually a “Report on the Investigation into Russian Interference in the 2016 Presidential Election,” and the conclusions on Russia’s role were damning. “The Russian government interfered in the 2016 presidential election in sweeping and systematic fashion,” reads the report’s second sentence.21 Then, in great detail, that sweep is spelled out. Coats was worried in 2018 about what the Russians had done in 2016 and were still doing and were likely to do. And the fear extended beyond presidential elections and particular campaigns. US intelligence agencies had also concluded that Russia had hacked state election systems. Three days before @GossiTheDog wondered if anyone in America was paying attention to the cyberthreats to mobile voting, the Wall Street Journal had reported that the American security apparatus was paying attention to cyberthreats to voting writ large. General Paul Nakasone, director of the National Security Agency (NSA), said US Cyber Command and the NSA were “tracking a wide range of foreign cyber adversaries and are prepared to conduct operations against those actors attempting to undermine our nation’s midterm elections.”22

Given these vulnerabilities, @GossiTheDog wasn’t the only one sounding the alarm about the West Virginia pilot, now slated to expand for the November general election. Computer scientists, legal scholars, and other observers wondered aloud whether the time was really ripe to be experimenting on the vote. A Vanity Fair article gathered some of the building objections and wrapped them up thusly: “Should those votes be compromised, or should the app glitch, it would have a non-zero impact on overseas voters. Furthermore, other, similar, companies are testing their own versions of blockchain-backed mobile voting . . . These companies argue that their systems could help increase voter turnout. But an atmosphere in which the ‘warning lights are blinking red’ seems a less than ideal testing ground.”23 The article cited Joseph Lorenzo Hall, the chief technologist at the Center for Democracy and Technology, who called Voatz “a horrifically bad idea.” He went on to say, “Imagine if you’re a uniformed military serviceman stationed abroad, excited to be able to cast a ballot in, say, the West Virginia primary, where they plan on using a remote blockchain voting system . . . then imagine that in 20 years, the entire contents of your ballot are decryptable and publicly available . . . It’s not something we should throw to the [venture capital] wolves or allow bleeding-edge technologies to mess with, without serious and deep inquiry and interrogation.”24

Resistance to mobile voting was a contemporary strain of already established resistance to electronic voting and internet voting. In 2017, fourteen countries used some form of internet voting. Estonia was the first to introduce permanent national internet voting.25 But political scientists worried about all the risks. “Security threats to internet voting exist in the form of internal actors (i.e., voters, election officials, service providers) or outside actors (i.e., individuals, organizations, or even hostile countries),” one warned.26 These actors could launch denial-of-service attacks to prevent access to voting websites, they could imitate other voters, they could hack into systems to steal or otherwise modify votes, they could dupe voters via phishing attacks, they could infect the voting software with viruses or malware. Any or all of this could break confidentiality of the vote, compromise its integrity, make the vote unavailable, or make it impossible to know who had voted. It could swing elections. A thousand things could go wrong, some of them cataclysmic.

Imagining, Trying, and Scaling

In light of the risks of doing nothing to fix our democracy and the risks of doing something, Tusk came down firmly on one side. “It’s not that the cybersecurity people are bad people per se. I think it’s that they are solving for one situation, and I am solving for another. They want zero technology risk in any way, shape, or form. So, paper maybe solves the problem for them. But in my view, when you can’t resolve the issues on guns, on climate, on immigration, because the middle 70 percent doesn’t participate in primaries—where I don’t see a world where the U.S. still exists as one country because we are this polarized—well, if you don’t want that to happen, you have to solve the dysfunction a lot, lot faster. I am solving for the problem of turnout. If turnout were actually 82 percent and going to paper-only ballots would reduce it to 80 percent, well okay, that’s not an unreasonable tradeoff. But when turnout in a primary election is really only 12 percent, I argue that even if you win the battle, you lose the war.” He held his view about the potential of mobile voting not only strongly but also optimistically: “At a certain point, the future always wins.”

Not always. Sometimes, on some issues, the future takes next to forever in arriving or never does. And sometimes it arrives, but dangerously or unfairly. The techniques of Possibility Government should be for occasioning the future and for helping to get it right. By that bar, how did Tusk, Sawhney, and Warner do?

Ideas don’t get actors to play them in Showtime dramas, but if they did, the Sawhney version of mobile voting would have needed one right out of Possibility’s central casting. It was born in Sawhney’s own experience as a citizen, when he was troubled by what he had seen in India. It was kindled, at a hackathon at South by Southwest, the film, media, and music festival in Austin, Texas, in 2014. It was there that Sawhney and his brother brainstormed areas around creating tamper-resistant technology via blockchain and landed on voting, walking away with $10,000, one bitcoin, and thirty minutes with a venture capitalist as their winning prize. The venture capitalist then promptly told them not to pursue it: It was too hard of an industry. They had too little knowledge of US elections or elections anywhere. Sawhney spoke to people across the technology industry. “Interestingly, almost every person we talked to said, ‘Don’t do it. It will be a giant waste of your time.’ Everyone kept telling us there were so many hurdles. Legislative issues, political challenges, regulatory matters . . .” Unswayed, the team spoke to politicians, to town and city clerks, to election auditors across the country. And they met with voters, and they thought like them—pondering the obstacles they faced to vote and the lines they waited in—and they made the leap. Create choices before making choices. Widen the aperture for ideas. Design with the users. Every problem is solvable, and the people who face them hold the key to their solutions.27 If “government that can imagine” is the first step toward Possibility, here, in the form of hackathons and techniques of design thinkers, was that first step.

How Voatz built from that idea, and how it unfolded into the West Virginia test, had many of the hallmarks of “government that can try new things,” too. Before the Voatz team rolled out mobile voting for congressional primary elections, they tested it on college campuses in student-council elections and in the Democratic and Republican party conventions in Massachusetts. These venues had similar(ish) characteristics, but lower stakes. Build, measure, learn. And when primary time did come, Warner’s team had limited the test to only two counties and without too much fanfare. “We didn’t push for the military agencies to send out email blasts; we didn’t have the Federal Voting Assistance Program advertise on our behalf; we didn’t do social media,” said Donald “Deak” Kersey, Warner’s general counsel. “We had no problem pulling the plug; we weren’t too bold or too proud to stop doing what we are doing just because it provides a solution to a problem. We can always find other solutions.” The results of the primary had been reassuring. The thirteen ballots were cast; none appeared tampered with. “The safeguards worked. We felt a lot more confident to move forward to the general,” Kersey explained. His team and Voatz planned to put in “more stuff” for the November election. (For example, the West Virginia election officials wanted voter-verifiable ballots emailed to the voter and the state without sacrificing the voter’s anonymity.) Build, measure, learn, again.

The question that lingered—lingers for us—now that blockchain-enabled mobile voting had been thought up and tried was, How should it be scaled? Should it be scaled? The November election, really? . . . The congressional election? And if that, then what? Then where? More than just overseas voters? Voters with disabilities, who have a harder time getting to the polls? All voters? (Voters during a pandemic?) Other cities and states? (Voatz had been in discussion with several states.) The 2020 presidential? How big could it get, and how big should it get? I was told that India’s Prime Minister Narendra Modi might be interested in mobile voting. Close to a billion voters all on their mobile phones someday? What do we think about someday soon?

Democracy has been the architecture of one of the world’s great scaling experiments: How might we allow the people to self-govern? In Athens, self-rule was invented and then spread. (Though some scholars have said it might also have existed in India, China, and Japan; later in Africa and Iceland; and among the native populations of Australia and America.)28 It has spread inconsistently and imperfectly, but it has spread nonetheless. Perhaps sixty thousand citizens lived in Athens in 431 BC, when democracy made its “surprising appearance.”29 Today, more than four billion people live in democracies.30 Government as a platform thinking offers insights into how that happened. At its best, democracy passes our “does an additional user make me better off?” test; at its best, democracy creates giant and positive network effects. The more people who come together, who vote, who participate, who rule, the better our society. Out of difference, strength. This was the central insight behind America’s unofficial motto: “E pluribus unum.” Out of many, one.31

The question of whether and how mobile voting should scale thus becomes a platform-architecture question. Voatz had made a series of hardware choices (e.g., limiting its offering only to more-modern phones with certain identity-confirming biometrics) and software choices (e.g., leveraging a public-permissioned blockchain) that aimed to preserve the platform. With West Virginia, the team arrived at a nine-step process (application, authorization, authentication, voting, transmission, voter verification, preparation, tabulation, post election auditing) in accordance with rules (e.g., the system is intuitive, it works for users with screen readers, it does not allow mismarking of ballots, it protects personal data, etc.) to protect each vote’s integrity. The question is, At scale, will these components—the software, hardware, rules, and process—hold up? Sawhney is confident that in time they will. Alex Halderman, a University of Michigan computer science professor who made a name for himself by successfully hacking into voting machines, remained a skeptic. He worried about the risks to voter anonymity. As the November elections creeped closer, Halderman called blockchain voting “mostly hype.”32 Possibility Government imagines, tries, and then scales. Sawhney, Warner, and Tusk had availed themselves of its tools. Time will tell us how well.

Democracy Entrepreneurship

There are many ways of answering the question of whether mobile voting should exist. We can weigh the risk of not doing versus doing. We can assess, as we just have in part, the way of going about it. We can also ask, as some mobile-voting skeptics have, about alternatives. With so many pieces to an election, why not focus elsewhere? Somewhere safer? Why not registration? Or why not some other angle on participation? (Why not a voting holiday?)33 Maybe work on making democracy more open, more representative. Change who runs. Reinvent redistricting. Expand the franchise. But do you have to experiment on the vote itself?

This, broadly speaking, was the view of Larry Schwartztol. He was counsel at an organization called Protect Democracy, which had been watching Voatz with considerable concern. “Tech and voting?” I had asked him, “Is there any combination of those two things that you would be okay with?” He replied, “I have a hard time seeing a vision for voting and technology that doesn’t increase the attack space for interference. I don’t know that that will be the governing principle for all time, but for right now, that should have some weight.”34 (“Right now” was his version of “the warning lights are blinking red.”) He continued: “There is an intense threat from Russia and others. It’s not clear that the federal government is taking sufficient steps to mitigate those. The states are a patchwork of voting systems, which means there are multiple pathways of attack but no unified defense. This is important because attacking any one state could undermine the country’s confidence in election outcomes more broadly.” As for Voatz in particular, he felt that the biometrics and blockchain gave a false sense of security. And he worried, as others had, about the risk of even false claims of intrusion. “Voting is like currency,” he elaborated. “Its value depends on the confidence of the electorate. So when confidence takes a hit, democracy declines.”

What made Schwartztol’s reaction so fascinating to me, and a little confounding as I tried to decide what I thought about Voatz, was that although Schwartztol and his organization objected to this experiment, they didn’t object to the notion of testing and experimenting on democracy. In fact, they built their whole organization around that very core. I learned about Protect Democracy in the first place from Justin Florence, who had cofounded the startup in 2017. We become friends after he had moved to Boston from Washington, DC, where he had been in the Office of White House Counsel. He had seen up close how rules and norms in the executive branch protected the democracy, and he worried they were in retreat. Florence and Ian Bassin, who had also been a White House lawyer, started Protect Democracy to halt that slide. But they didn’t start it in the mold of the traditional not-for-profit. They injected novelty and innovation into their culture. “We default toward action and we are willing to fail,” their principles read in part. “We do unconventional things to address unconventional challenges . . . We would rather try ten things and fail on seven . . .”35 It was a culture of possibility. Bassin told me how this was playing out in 2019. “We have set up an entire stream of work . . . little beta projects. They are meant to be lean. They have gating tests. We give them limited time and limited resources. And we have set targets, so we will have clear signals on did we hit them or not. And if we do, we invest more. The goal is that we will come up with twelve of these. We will test five or six. And we will have two that show enough traction that we will invest in next year.”36

Schwartztol said to me, “I think we should ask: What’s the problem Voatz is trying to solve? Is it participation? Well, are there other ways to solve it?”37 He wasn’t just asking hypothetically. He and his colleagues were actively experimenting with other ways to protect democracy. They had engaged software engineers (unusual for an organization that was otherwise made up mostly of lawyers, with Democratic and Republican pedigrees) to develop VoteShield. The idea behind the technology was to layer some data analytics on top of the voter-registration lists made public by the states each month and to monitor for aberrant changes: fifty thousand address changes in one month, say, instead of a more typical one thousand.

Bassin remarked,

People more typically think of the vulnerabilities of voting machines. But when we asked about their biggest fears on elections security, the voter-registration databases kept coming up. Voter-registration files are online all year long . . . the vulnerability is that much greater. And they are a centralized file containing every latent vote. Unlike voting machines, where you might get access to just one or a few . . . access the voter file and you now have access to the entire state. You could make very small changes in them that would be very hard to detect, and maybe then only months later. What would happen is someone’s address might have been changed from 100 Main Street to 1000 Main Street. On election day, they will be told they are at the wrong polling place. They will be told to go across town. And maybe they will, or maybe they will have to go back to their job before their lunch break ends and miss the chance to vote. And if a particular demographic were targeted, it could change the outcome of the election.38

VoteShield could compare voter files from month to month and track the differences. Using machine learning, it could compare those differences to all other historical differences and find the changes that stood out. Protect Democracy built VoteShield the same way it built the organization. “We said, ‘Let’s do a beta,’” Bassin explained. “We said, ‘Let’s pick one state.’ It was lean startup. ‘Let’s do the smallest possible version to see if it works.’” In less than four months—between September 2017 and January the following year—the team had a beta for Ohio that “absolutely spotted when there was a change in a voter file that went outside of standard deviations.” While Sawhney was piloting his software with Mac Warner, Bassin and team were experimenting on theirs in Ohio, and later with Iowa’s Secretary of State Paul Pate, and others.

“Are there other ways to solve it?” Schwartztol had asked. There is a small wave of democracy startups trying to answer just that. For-profit and not-for-profit. With tech and otherwise. On voting and more broadly. You can find more than a handful of new organizations working on democracy, even if we take a minimalist view of what democracy is: where “almost all adults have the right to vote; almost all adult citizens are eligible to hold office; political leaders have the right to compete for votes; elections are free and fair; all citizens may form and join political parties and other kinds of associations; diverse sources of information about politics and public policies exist and are legally protected; and government policies depend on votes, or other expressions of public opinion.”39 David Moss, a historian and a colleague of mine, started an effort that I find most hopeful: the Case Method Institute.40 He and his team were training teachers to teach democracy by the case method all across the country. He and collaborators wrote twenty-two cases that “each explores a key decision point in the history of American democracy,” and they invited the students to step into the role of president, senator, judge, citizen.41 Over the question of the federal negative. During the battle for women’s suffrage. Amidst the debate over referendum government and jury trials. “Faith in the democracy and in the essential principles of self-government has long been what united most Americans—what most bound them together. Yet sustaining and nourishing that faith, particularly in the face of unending crises, scandals, and transgressions, has always been part of the grand struggle of democracy,” Moss wrote in an introduction to a compilation of those cases.42 Each invites younger (and all) Americans to struggle. I’m astonished by what the early results (achievement test scores, but also measures of willingness to participate in democracy and even run for office) say about the ability of this effort to lift up teaching and learning, and I am made optimistic about our democracy. I believe Moss is onto something invaluable when he says engagement with the democracy, in the deepest and most active sense, is what we all need. Through each case, in each session, students are asked to reinvent a bit of America. Not always differently, but again.

“Are there other ways to solve it?” Schwartztol had asked about repairing our democracy, and Yordanos Eyoel was underway supporting a slew of organizations exploring just that. Eyoel led Civic Lab, itself a new effort as of 2019, and its work to support “democracy entrepreneurs.”43 A first cohort of seven organizations ran the gamut from promoting new models of civic organizing to supporting a new generation running for public office.44 Their leaders were people like Katie Fahey, who had successfully led a grassroots campaign to prohibit partisan gerrymandering in the state of Michigan. Fahey was now attempting at her new organization, The People, to bring together Americans across deep political divides and “from all walks of life,” to empower them to get more involved together in the political process.45 Civic Lab was also supporting Darrell Scott, who’d cofounded PushBlack. Among the organization’s aims: “to transform Black voting rates by building deep, daily relationships with millions of African-Americans through their phones.”46 By the summer of 2020, PushBlack was sending daily black-history stories and news stories via mobile platforms to more than five million subscribers.47 Scott and his team were “huge believers in experimentation, constantly testing, innovating, and changing,” Eyoel said. “That’s what’s made them nimble and successful.”48

What does Eyoel say to those who tell her none of this can work, that democracy is too delicate for trying new things? “The thing I would say is that we can’t afford not to try,” she told me. “Our democracy is very delicate. And it cannot be left up to the political parties, the political elite, the academic elite, the economic elite, to define how our democracy works and who it works for. Every single person has a legal and moral obligation to contribute to refining and developing our democracy . . . And let’s not forget, this entire republic we have”—she said, about the one her family had brought her to from Ethiopia at thirteen—“has always been an experiment. And when only 17 percent of people say they trust their government, there is something fundamentally flawed. We have to use everything at our disposal, including experimentation and innovation, to make this an inclusive and responsive and healthy democracy. I just don’t see any alternative options.”

Not all of the democracy-startup efforts will succeed. We know statistically that most of them won’t. Not all of them will have the right idea. Most will try to make things better; some might make it worse. Ian Bassin says that generally speaking, and given the scope of the challenge, “the more the merrier.” Eyoel has called on Americans to match every dollar they give to political campaigns with twice as many for democracy entrepreneurs, and she has called for philanthropists to invest in civic culture at the scale they invest in things like education, workforce training, and public health.49

In June 2020, as Americans took to the streets to protest in support of Black lives and against racial injustice, and as people around the world did the same, I asked Eyoel how she saw democracy entrepreneurship in that context. “One reason Civic Lab exists is because of my own personal experience,” she replied, referring to the Boston Women’s March she had helped organize in January 2017. The march ultimately brought out more than 175,000 people in Boston, as marches like it brought out millions across the world in support of human rights.50 “I saw a lot of people who had never marched before; they’d never done anything ‘public’ besides vote. They marched and said, ‘What do we do now?’

REVISITING PART THREE

Government That Can Scale

Possibility leaders

Possibility citizens

  • Look to create complementary services to government platforms (e.g., a teen basketball league as part of a community public-safety effort).
  • Be watchful for who owns the platforms: government or business. In either case, retain democratic governance; you are the platform sponsor.
  • Look for privacy, equity, autonomy, incumbency threats of platforms. Platforms can fail in many ways; they can concentrate power.
  • For govtech entrepreneurs who want to help scale solutions with new products and services: selling is buying. Learn to break down the building blocks of government buying.
  • Develop careers as trisector entrepreneurs. Rotate in and out of public service. Learn differences across sectors and how ideas translate across them. Build integrated personal networks. Then leverage these relationships and your contextual intelligence with integrity, not greed.

What everyone can do

  • Work to solve problems at “population scale.”