2

Reach Out to Reach Up

On a late afternoon in 2014, Jimmy Chen walked into a food stamps office in Brooklyn, New York.1 He remembers arriving at 3:30, or maybe it was 4:00. Closing was at 5:00. He walked up to the person staffing the entrance and said he wanted to apply for assistance. He remembers that she laughed at him. He recalls her saying that there was already a two- or three-hour wait, that most people wouldn’t be seen today, that if he wanted to be seen the same day, he should come back in the morning. Chen returned the next day at 9:00 a.m., when the office opened, and waited only forty-five minutes, though he expected he wouldn’t qualify for the benefits anyway.

Chen also filled out applications in several other states, where he wasn’t a resident.

Chen asked mothers how to navigate the application process. He went to their living rooms and peered into their lives. He asked them how they secured benefits for their kids. But he had none.

Chen wasn’t eligible for food stamps. But he wasn’t a fraudster.

What was he doing then? “As entrepreneurs, we tend to solve the problems we understand best . . . you need to understand the needs of users,” Chen told me. “You need empathy.” Chen wasn’t waiting in line for food stamps. He was waiting in line for insight.

A few months prior, Chen had left a senior role in Silicon Valley. He had parlayed a degree in software engineering from Stanford into roles at LinkedIn and then at Facebook, where he built technologies that reached hundreds of millions of users around the world. Now he had left San Francisco for Brooklyn, with a vague plan to build a startup that helped people startups weren’t reaching. But he took with him a sense that understanding user experience was the key to building effective products. So when Chen decided to go looking for ideas for helping low-income Americans, he decided to go looking for low-income Americans. New ideas are why Chen was waiting for food stamps he didn’t need and couldn’t get.

While Chen was on Bergen Street in Brooklyn, public officials around the world were pounding the table for new ideas and were mostly looking for them around that same table—but not all of those officials. An emerging type of leader was harnessing a set of techniques to get well beyond the reach of established agencies and even that of so-called experts. Such leaders were turning toward their own citizens, using the kinds of techniques that Chen had learned, to try to identify the right problems to solve and the right solutions. They were looking to the crowd, using the web’s technologies and ethos to take up “wiki government.”2 They were providing pizzas and power outlets for weekend-long hackathons, enlisting weekend warriors in the pursuit of solving public problems. They’d begun to earn both praise and skepticism for these approaches, but nevertheless a new breed of public servant was dumping old, insular traditions. The first-term mayor of St. Paul, Minnesota, Melvin Carter, told me, “My best ideas are in other people’s heads.”3 (He was nothing if not consistent on this front . . . he’d enlisted a hundred volunteer citizens to help him select cabinet chiefs and department heads.4 At his 2018 inauguration, as his community welcomed him with a standing ovation, he told them, “Don’t clap if you’re not going to help.”)5 I asked Carter where new ideas come from, and he replied, “I don’t have a good, finite answer. Other than everyone and everywhere.”6

Everyone? Everywhere?

From Talking to People

Chen and colleagues set out to build a company called Propel to “make America’s safety net more user friendly.”7 Their first product was a benefits-enrollment app. The Propel team pulled this early tool from a bucket of user-engagement techniques that, at the time, were still only lightly utilized by people inside public office. The team had spent time in Supplemental Nutrition Assistance Program (SNAP) offices. They’d downloaded and completed application forms themselves. They had noticed things—like how many people waiting in the same line Chen was were on their mobile phones. He looked into the data, which told him that in New York City only 74 percent of eligible people were using SNAP; also that smartphone use was on the rise among low-income populations and that 45 percent of low-income Americans went online primarily with their phones, compared with 34 percent of Americans in general.8 He had gone to a supermarket in Philadelphia on a December day and found that most of the shoppers were enrolled in SNAP there, and watched a woman (with her permission) as she checked her balance by entering a nineteen-digit number from memory into her phone . . . something she did almost every day. “We had thought enrollment was a major hurdle, but from talking to people, we came to think that there were a lot of issues with the actual delivery of the SNAP program,” Chen said. A shift in insight about what the problem was (from it’s hard to enroll to it’s challenging to use) had prompted a shift in the targeting of the solution. Chen had grown up in a family of immigrants. He’d been through times when finances were tough.9 But then he’d spent time in Silicon Valley, where he felt distant from the problems he was trying to uncover and fix. “A twenty-five-year-old male has technology to meet his every whim. But who is building the technology for the fifty-year-old single mother on food stamps?” he asked. And then he went to meet her, and people facing similar circumstances. He sat in their living rooms and watched them shop, so he could.

Citizen-Centered Design

You can trace Chen’s time spent at a supermarket in 2015 to another supermarket in 1999 and to the core idea that informed his outreach: design thinking. In 1999, Ted Koppel and Nightline placed IDEO; its founder, David Kelley; and his team firmly onto the global map. Millions have since watched the episode as the group went about answering Nightline’s challenge: redesign the shopping cart in five days with their (then) peculiar methodology. IDEO’s engineers and designers wandered Whole Foods Market aisles to observe how shoppers behaved. They cruised local bike stores for tips on design and materials. They papered walls with giant Post-it sheets and prototyped newfangled shopping carts (carts with stackable handbaskets, carts with their own scanners) from materials they’d gathered from the hardware store and welded together.10 From there, IDEO’s design-thinking methodologies—which focus on observing real people and working to uncover their latent needs—made it into best-selling books, course curricula, corporate trainings, and consulting gigs. David Kelley went on to found Stanford’s design school, with these methods as its core. Chen was introduced to design thinking at Stanford as an undergraduate. He’d been to a David Kelley lecture and had been wowed.

By the time Chen was in the SNAP office on Bergen Street, that 1999 version of design thinking had been recast somewhat as human-centered design, and the new label sought not only to emphasize the empathy in design thinking (“to see an experience through another person’s eyes, to recognize why people do what they do,” David and Tom Kelley had written) but also to gesture at its now larger ambitions.11 “The human-centered design process obviously works for products—that’s where it originated—but the beauty of it is that it’s applicable to almost anything you might want to come up with as a challenge,” Diego Rodriguez, an IDEO partner, has said.12 IDEO published a book called The Field Guide to Human-Centered Design, and it opened with a breath of optimism:

Embracing human-centered design means believing that all problems, even the seemingly intractable ones like poverty, gender equality, and clean water, are solvable. Moreover, it means believing that the people who face those problems every day are the ones that hold the key to their answer. Human-centered design offers problem solvers of any stripe a chance to design with communities, to deeply understand the people they’re looking to serve, to dream up scores of ideas, and to create innovative new solutions rooted in people’s actual needs.13

All problems are solvable, they said, if we’d just work hard to deeply understand the people we served.

Across the world, there have been blooming efforts to bring these kinds of practices squarely into the public sector. The UK’s Policy Lab was established in 2014 and has worked to bring “people-centered design approaches to policy making.”14 Andrea Siodmok is the deputy director there. An industrial designer by background, she had been through the same design course as IDEO’s Tim Brown and Apple’s longtime design chief Jonathan Ive. Siodmok and Policy Lab’s head, Vasant Chari, told me their job was to “bring the outside into government” in order to “bridge the gap between evidence and the art of possible.”15 I spoke with them after they’d undertaken an effort to help the United Kingdom’s housing minister address social-housing issues, on the heels of the Grenfell Tower tragedy that had killed seventy-two people. The lab engaged a thousand citizens around the country in conversations about their fears and frustrations and needs. These kinds of efforts have been a large part of the United Kingdom’s other government redesign undertakings, including its Government Digital Service. In the United States, 18F, an agency charged with helping federal agencies better meet user needs, has dozens of blog posts on how to bring design-thinking methods into government.

The increasing prevalence of these practices means there is also a growing body of practical advice for how to design with the public:16

The world was now filling up with ambition and practice and advice for going to citizen “users” for new ideas, so I set out to try these tools myself. I wanted to retrace some of Jimmy Chen’s tracks. I wanted to understand better why he’d gone to that SNAP office and what to make of it and what it might mean amid this wave of citizen-centered design I was witnessing. I decided in the summer of 2019 to go back to the same office where Chen had spent much of his time and to try to do some of the customer discovery that Chen had done.

The office Chen had gone to had closed. I looked into why. The press coverage I found said it was because of the advent of technology: more people were applying online, so there was less need for offices. The irony of this was the first thing that struck me. People thinking like Chen did had made it easier to access benefits from a computer but harder if you needed to do so in person.

Then a second thought occurred to me: whether the site had been closed, in part, because it had become politically toxic. Between when Chen had gone there and I tried to, something quite sad had happened.17 In the winter of 2018, a mother had been sitting on the floor there with her crying infant, trying to comfort him. She was told she wasn’t allowed to sit on the floor. She insisted that this was the only way to comfort the child. Eventually, the security team there tried to lift her off the floor and ultimately pulled her screaming baby out of her arms. The video went viral, and it’s incredibly hard to watch. New York’s mayor, Bill de Blasio, had to apologize. I wondered whether that also had something to do with why the office on Bergen Street was gone.

I also thought about the alternative universe the IDEO types were trying to conjure up, one where when a mother is sitting on the floor with a crying baby, it isn’t security guards who approach her but instead design thinkers, who sit down next to her and begin to ask, respectfully, Why? I wondered whether—if given the opportunity to ask—they would have generated a whole list of things to try that would have had nothing to do with dragging her out of there. After a conversation with her and others like her, maybe they would have arrived at a list of new problems to tackle and new ideas for doing so. Maybe the problem to address isn’t seats at SNAP offices but childcare, which is prohibitively difficult to access in the United States. Or maybe we should have toys for kids at the office. Or a TV showing Curious George, like a dentist’s office might. Maybe we should create a special unit that enrolls people for SNAP at their own houses, or a mobile team that helps them just after they’ve dropped their kids at day care. Or maybe we should create different kinds of seats. Bleachers. Or lofts, like in a dorm room. Or maybe we should create a mobile app (like Chen did). Or maybe none of these things, but something else that makes applying for SNAP easier and that someday has a hope—and I realize this is a giant leap from better seats and forms—of lifting people out of poverty. But that’s the exercise, and I was struck by how different that would be from what governments normally do, how generative it would be. If that’s what the human-centered designers were aiming to bring about, I was warming to the idea.

Desire Lines

There are, in fact, several potential logics for designing with users.

There is historical evidence validating the approach. “The idea that novel products and services are developed by manufacturers is deeply ingrained in both traditional expectations and scholarship,” Eric von Hippel wrote in his Democratizing Innovation in 2005. “Nonetheless, there is now very strong empirical evidence that product development and modification by user firms and users as individual customers is frequent, pervasive, and important.”18 He’d provided a lot of that evidence—as had other social scientists—including in one study where von Hippel had found that “users were the developers of about 80 percent of the most important scientific instrument innovations.”19 He also noted that there was a long tradition of user-driven innovation. He pointed out that no less than Adam Smith had documented so in his seminal The Wealth of Nations. In 1776, the same year American colonists declared their independence from a government that wasn’t working for them, Smith wrote that many of the machines enabling a transformation in productivity driven by division in labor “were originally the invention of common workmen, who, being each of them employed in some very simple operation, naturally turned their thoughts towards finding out easier and readier methods of performing it.”20

There is also a potential physio-logic to working with users. “Why fight human instinct?” Tom Kelley, an IDEO partner and David Kelley’s brother, noted while explaining IDEO’s penchant for “going to the source” for ideas rather than relying on focus groups and market research. He told a story about a customer—a medical-device company—insisting on a new device that could be used with one hand in the operating room, and of the reality that IDEO’s team saw when they themselves went to the operating room: “Although the current product could theoretically be used with one hand, it really worked that way if you had a hand the size of Michael Jordan’s. In actual practice, medical technicians used both hands with the device.” Why not design the new device for two hands? Which is precisely what they did. By closely observing the end user, they could see how users really experienced products and services and could then come up with better ones. It was also why they’d come up with a trackball for younger kids on computers, who were confused by a traditional mouse’s limited range and the need to pick the device up and reset it on a mouse pad. The point was to uncover what comes naturally to the user.21

I had seen what came naturally. A few years ago, Harvard Business School installed a sign just outside of my office, at one of the entrances to the school (see figure 2-1). Not a day went by without somebody being photographed standing near the sign. Newly admitted students. Graduating ones. Tourists. Executives on campus to address a class or take part in one. Pumpkins showed up around Halloween, and banners did around New Year’s Eve. I’d noticed that soon after the sign was installed, dirt patches appeared to the sides of it, where the grass had been worn away. The cement strips that looked meant for standing on went relatively unused, because next to the sign (not in front of or behind it) was where people wanted to be captured beaming for Instagram. And then I noticed that the patches got resodded, and then got worn down again. In eighteen months, I watched this process repeat itself three times.

FIGURE 2–1

I approached Andy O’Brien, the school’s amazing chief of operations, and asked him about the sign and the sod. He laughed good-naturedly and acknowledged that the cement strips—which now looked less than ideally placed—had been the product of a lot of meetings with a lot of experts. Yes, he said smiling, “There was a study.”22 I asked him why not just move some cement blocks to where the sod had been worn out? Why not put them where users seem to want to stand?

The idea of doing this—of finding “desire lines,” paths created by human usage—was not one I had invented. It had been around for a while in transportation planning. Janette Sadik-Khan, New York City’s transportation commissioner under Mayor Mike Bloomberg, brought desire lines to life in New York City. In one instance, her team put a new crossing between Sixth and Seventh Avenues in New York where hundreds of people crossed midblock each day to access a pedestrian arcade.23 “Desire lines are the native operating code for a new approach to urban design,” she explained. “Instead of asking why people aren’t following the rules and design of the road, we need to ask ourselves why the rules and design of the road aren’t following people . . . Desire lines are a road map of opportunity.”24

Mayor Carter of St. Paul was one of those leaders who looked for “desire lines.” He cited Sadik-Khan’s influence on his thinking but, being from the Midwest, leaned toward the label “cow paths” when telling the story of his old high school and the debate over locating a sidewalk on its campus. “There’s a cow path across the grass where students have been walking for a long time. If we ask 2,500 students to come to a public hearing on proper sidewalk alignment, most would roll their eyes. Maybe only two would show up. Most would say, I don’t know what you are asking me. Or I don’t belong. Meanwhile, all 2,500 know it makes more sense to cut across the grass, where the cow path is, than to go where the sidewalk is. So often we accept testimony out of a city council at 5:30 on a Wednesday instead of acknowledging that all day, every day, people are voting with their feet to tell us what is working for them.” The metaphor, he said, illustrates that we should “stop forcing people to make their behavior meet our geometric sidewalk shapes or whatever public problem we are talking about [and instead ask] how do we transform to meet theirs?”25 Carter then carried through on the metaphor in many places. He ended library fines for the St. Paul Public Library. And when he explains why, you almost forget that you thought at first that the idea sounded crazy. “Think about it,” he says. “We were penalizing people for doing exactly what we wanted them to do: read books.” St. Paul unfroze 42,000 library cards. And after a few months they had 43,000 materials checked out on those cards. “We fundamentally redefined our library. And it’s akin to if we had opened a new branch, for $200,000 in late fees.” A front-line employee had suggested the change, and Carter’s library chief (one selected by the citizen process) liked it, and so did Carter.

Nothing about Us without Us

There’s also justice in designing with users, the logic of fairness. It respects a “nothing about us without us” sensibility, gleaned from the disability-rights movement.26 The mantra has spread well beyond that movement. I first heard it in a session on criminal-justice reform, from Anthony Ray Hinton, who had been incarcerated on death row before ultimately being exonerated. I later heard Mayor Carter say America is great “when the ‘We’ in ‘We the People’ truly means all of us.”27 If “nothing about us without us” seems obvious or even cliché, think about how many meetings you’ve attended about homelessness without any homeless people in attendance; about education without any students in attendance; about workforce training without any workers in attendance.

After the marathon attacks, when we set out to establish a new fund that would be a conduit for the world’s generosity, people immediately and rightly wanted to know . . . what will the money be for? Ken Feinberg, whom the mayor had asked to help administer the One Fund’s payouts after the marathon attacks, advised those of us shaping the fund to reserve it for the survivors and the families of the victims. There was talk of supporting the businesses affected on Boylston Street (workers would end up losing substantial wages) as well as the wider community. Feinberg was adamantly against those ideas, having been chastened by his experience administering other funds after other tragedies. “No matter how much money you raise, it will never be enough,” he told us. Focus it on the survivors and the victims’ families. It was consistent with a message we’d been hearing since the street exploded on Monday: survivors first. Friends and colleagues who’d been involved in 9/11 recovery efforts had emailed or called, having some sense of the flood of decisions that would be headed in our direction and some sense of how to wade through that flood. Their advice was unambiguous: prioritize the needs of those most affected by this tragedy, and not the needs of the agencies helping them.

Feinberg also said that once you give survivors the money, don’t tell them what to do with it. If they need it for job training, let them use it for that. If they need to reconstruct their house, they can use it for that. If they want to get away and take their families to Disney World, that’s their choice. The money is a gift for them to do with what they want. For the well-meaning bureaucrat, the instinct is to decide: not only to allocate the money for survivors but also what the survivors themselves should allocate it for. For scholarships? For psychological support? For housing reconstruction? We resolved that the money would be for them to do with as they chose.

A similar kind of thinking is at the heart of the direct-giving movement in international disaster relief and poverty alleviation. The organization at the center of these efforts is GiveDirectly, and it explains its outlook and findings pretty clearly. “We give cash directly to people living in poverty . . . We believe people living in poverty deserve the dignity to choose for themselves how best to improve their lives . . . Donors have given over 140 million dollars to people in need . . . and no, people don’t just blow it on booze . . . in fact, research finds people use cash in impactful and creative ways . . . if you think about it, doesn’t giving directly make sense? Cash allows individuals to invest in what they need, instead of relying on aid organizations and donors thousands of miles away to choose for them. Isn’t this what you would prefer?”28 The experts say it will take at least a decade to really know what to make of these efforts, in particular whether direct giving does help lift people out of poverty and whether it does so more effectively than more-typical efforts, which involve in-kind types of aid (training, livestock, food, etc.) designed, in many cases, by insiders and “experts” for the benefit of people in the world who have the least, but not always (or even often) designed with them.29 It’s no coincidence that when Jimmy Chen and his team wanted to support their customers faced with massive and pressing needs during the Covid-19 crisis, Propel partnered with GiveDirectly to raise and distribute cash relief to users of its Fresh EBT app.30

More Ideas Plus Different Ideas Equals Better Ideas

Then there’s a third logic to looking for ideas with citizens, and it has to do with the math. In design thinking, an exploration phase with users (“What challenges are we trying to address?”) is meant to give way to the concepting phase (“How might we address those challenges?”).31 Problem finding lays the groundwork for solution finding. Now that we know what latent needs the users have, the questions turn to how might we solve them?

IDEO’s favored technique at this point was brainstorming. The team would target a hundred ideas in a sixty- or ninety-minute session.32 (This strikes me as more new ideas than some public agencies might have in a year.) IDEO avoided rules, generally, but had rules for this:

  1. Defer judgment.
  2. Encourage wild ideas.
  3. Build on the ideas of others.
  4. Stay focused on the topic.
  5. One conversation at a time.
  6. Be visual.
  7. Go for quantity.

Most of the first six rules were in service of the seventh.33

IDEO and its ilk seemed to favor almost anything that would get you more ideas. I observed a simulation once in Copenhagen that an innovation-training company had designed with IDEO. The setup was that a city in California was trying to raise the participation rates in its recycling programs. The task was to come up with ideas to help. So one of the first jobs was to look through pictures (since it was a simulation, the participants couldn’t go into people’s homes themselves and meet with them) to try to develop some insight into what the problem was. Were people not recycling because they didn’t care? Didn’t know? Didn’t have enough room to store recyclables? After some thoughts were synthesized, the challenge was to brainstorm some new ideas. And just in case you got stumped, there was a card deck with some prompts. One of those prompts actually read, “What ideas could get the mayor kicked out of office?” How was this supposed to help?

It turns out that “what ideas could get the mayor kicked out of office” is of a piece with a swelling body of research that suggests that the volume of ideas plus the diversity in those ideas (and the people providing them) can be key in arriving at good new ideas.34 One strain of this research leans on something called extreme-value theory.35 And the point is that traditional methods of innovation tend to get us traditional kinds of solutions, close to our current practices and capacities. But that some alternative methods for going after ideas, including the increasingly popular reliance on contests and competitions that enlist “the crowd” (sometimes a highly curated one) for ideas, are likely to produce much more novel ideas, some of which might, possibly, be extremely valuable. The point for government is that while it’s true that conventional ways of arriving at new public services or new policy proposals might come up with better ideas, on average, than these new methods, contests and competitions and even the kinds of activities behind a robust brainstorm might come up with a small number of super ideas. And if what you really care about is the quality of the best idea—because that’s the one you can run with—rather than the average quality of all the ideas, then whatever you can do to get to those extremely positive outcomes would be worth the effort (see figure 2-2).

FIGURE 2–2

This kind of thinking isn’t brand-new to government. It was behind an oft-cited episode in the 1700s when the British government offered a prize for solving the problem of tracking longitude at sea in order to save its Navy from getting lost. The financial award eventually went to a clockmaker who had solved the problem, but only after debate about whether someone unexpert in celestial bodies could really have come up with the winning idea.36 The winner (and the debate over it) highlights the role of the outsider here. New ideas, from new fields . . . they probably won’t work, but they might. I’d been involved in more-mundane efforts along these lines. In Boston, we’d partnered with the company InnoCentive to crowdsource an algorithm for a pothole-sensing app. Beth Noveck, New Jersey’s first chief innovation officer, had documented dozens of episodes like this, using contests and crowd collaboration to solve public stuff. She’d been intimately involved in an effort to crowdsource solutions for limiting the spread of Zika in South America. Yochai Benkler, the legal scholar and expert on peer production, had first turned me on to the potential of the crowd with his story of NASA’s Clickworkers project. A decade later, when I was wondering about the wider merits of reaching out, NASA was using contests to source new ideas on everything from solar flares to space suits.37 It had been searching for extreme values for ten years, and getting them.

One Bright Spot

The math of more ideas plus different ideas equals better ideas was behind another strategy that was gaining attention in the years since I’d left government.38 I’d seen it up close when Annie Rittgers, a native Ohioan just out of graduate school, used it to look for new ideas for one of the newest and biggest problems. “The worst public health crisis in our lifetime,” Mike DeWine, then Ohio’s attorney general, had called it in 2017. “A human tragedy of epic proportion.”39 (He’d later win praise as governor of that state during its response to the Covid-19 pandemic, another epic human tragedy.) Heroin and synthetic opioids were killing ten people a day in Ohio. The question of the moment was whether the scourge was at its peak or just a new peak. That it had come to Rittgers’s hometown of Cincinnati was indisputable. A May 2017 Bloomberg Businessweek article called Ohio’s third-largest city “a center of the crisis.” In one six-day stretch in 2016, 174 people had overdosed. “Everybody and their mom sells drugs these days,” a local detective had observed.40 Rittgers had been a student of mine, and I was rooting for her effort, for possibility, but even I had to face the brutal facts when I arrived at Talbert House, a large Cincinnati human-service provider. People were being brought there every weekend on death’s door, I was told. The nighttime parking-lot staff were now de facto first responders. Everybody in the city was feeling heroin’s effects.

Rittgers decided to wade into this horrible problem. She had grown up thirty miles north of Cincinnati. Her parents were lawyers in her small town. She had left for Dartmouth and Harvard, for Barclays and McKinsey. Now she wanted to return. “I was bothered by my distance,” she said. Rittgers had spent a prior summer at a venture-capital firm that had asked her to research the addiction-recovery space. She found a few solutions to the addiction epidemic out there, but nowhere near enough. She wanted to generate some more ideas.

Rittgers had the gall, beginning in about February 2017, to think that she could do better than the existing solutions to this evolving problem, that she could triumph over the pessimistic sense of “been there, tried that.” And her big idea for big ideas? A hackathon. A weekend-long event that would engage Cincinnati’s tech community to spur new thinking about the city’s addiction epidemic.

Rittgers reached out. She enlisted the help of a local city councilor and old friend, P. G. Sittenfeld, and his director of community affairs, Colleen Reynolds. They enlisted Emily Geiger, a managing director at a local health-care-innovation lab. Together the group ventured further into the Cincinnati community to secure the hackathon’s participants. They hoped to attract a broad group: software engineers, designers, and health-care professionals; experts but also (especially?) novices. They recruited undergraduates from the University of Cincinnati, IT professionals from Procter & Gamble, and the chief technology officer of a fashion-app startup. They also invited judges and wooed sponsors. They fanned out in the community to try to understand what the roadblocks were for solving the problem in Cincinnati and to try to narrow them to eight challenges the solvers would work on during the two-day event. They sat with city officials and service providers, including the ones I eventually met at Talbert House, and also individuals in recovery and family members. “There will continue to be a lot of preventable deaths and wasted potential if the opioid crisis continues unabated,” Rittgers told me in early spring. “The biggest threat with a crisis like this is the increasing feeling of hopelessness around solving it . . . Bright spots and positive momentum matter.”

What was Rittgers thinking? Would this effort prove to be one of those bright spots? Or would the weekend prove that enlisting college students and weekend warriors for two days to generate “new ideas” was no way to address a problem of epic proportions?

The view on public hackathons is mixed. There is some evidence that hackathons can speed the creation of new ideas.41 One expert estimated that well-organized hackathons could reduce the time for companies to bring new products or services to market by 25 to 50 percent.42 And there was hope that in the public sector the results could be similar. Advocates said that one of the outcomes of hackathons is simply bringing people together who otherwise wouldn’t have been; that useful relationships get built. Others argued it would save the public money: Products developed during a 2015 hackathon for the US General Services Administration were estimated to have saved the organization more than $500,000.43 Other supporters pointed to the civic value of public hackathons, that they built a culture of cooperation among people and their governments.

Critics argued that civic hackathons produced products that were often inherently incompatible with underlying systems.44 They also questioned whether the products generated would be sustainable. One 2011 study found that of all the apps created during New York City’s BigApps Challenge, just 35 percent were operational after a year.45 (This actually strikes me as quite encouraging. I would have thought the number would be lower.) Other critics argued that hackathon participants were not representative of broader societal demographics. Past studies indicated that at the average hackathon, for example, 80 percent of hackers were male and most were either White/Caucasian (44 percent) or Asian/Pacific Islander (42.2 percent).46 Finally, hackathons had elicited some strong negative reactions when proposed as a solution for deeply entrenched, complex social problems. In March 2017, a Chicago-based tech worker suggested to a local paper that the city might consider organizing a hackathon to tackle violent crime.47 In a response titled “A Citywide Hackathon? Please. Technology Alone Can’t Solve Chicago’s Murder Problem,” three Chicagoans argued that solutions to this stubborn problem required systems-level thinking and problem solving as well as engagement from a wide variety of stakeholders.48

Rittgers hoped to avoid many of the possible critiques while making the most of the weekend. She’d taken steps to balance some of the tensions in setting up a productive hackathon:

In the end, Hacking Heroin generated many ideas, not all of them good. A few of them were, though, and one of those grew into a company that was making a real difference in Cincinnati three years later. It was a platform that provided real-time treatment-capacity information to patients in need and to providers. Rittgers called it “just one cool bright spot” and said it was “random and lucky.” She was not precisely correct about that. When contests or competitions or reaching out broadly in other ways is done well, the bright spots aren’t exactly matters of happenstance. They’re a matter of knowing that there are lots of potential solutions out in the universe, along a spectrum from truly bad to really good. And if you invite enough entries and get enough diversity, you are likely to get draws from all along that spectrum, including one bright spot.

We Need More Bad Ideas

When Rittgers told me of her plans to host a hackathon on the opioid epidemic, I winced. When she told me she planned to call it Hacking Heroin, I begged her not to. I thought the activity seemed all too specious and that the name just called attention to that fact. I now see why I was wrong and why she was right to pursue it. I now know what Jimmy Chen was really doing in a Brooklyn SNAP office, even if his first idea didn’t turn out to be a winner, or his second. If we are going to truly solve our biggest problems, we are going to need to go after them in ways that generate extreme-value ideas, and we are going to need outsiders to help us. And since such ideas are a product of these processes, and since they come along with all manner of ideas, in order to solve our biggest problems, we are going to have to go after more bad ideas.

When Netflix offered a $1 million prize to get outsiders to enhance its movie-recommendation algorithm, the company received more than 44,000 ideas. More than two-thirds of them did worse than Netflix’s own tools.50 That means close to 30,000 bad ideas. But Netflix was unfazed. The one entry that eventually won the competition is what mattered to the company—and that entry was 10 percent better than what the company had come up with on its own. In governing, on our most pressing issues, we’re likely to want to do more than 10 percent better. (Though we might take even that.) But the point is that to get there, we’ll have to be okay with processes that generate plenty of ideas that are worse.

There are other reasons, too, that in the pursuit of good new ideas we need to tolerate bad new ones. Foremost, as the famed sociologist and organizational theorist James March wrote, “Inspiration and lunacy are not ex ante so clearly distinguishable.”51 At the outset of new projects, we’ll have to take the bad with the good, lest we too quickly mistake the good for the bad. Moreover, March described how a competency trap seduces us into sticking with existing practices, because we have become relatively skilled at them, and this keeps us from pursuing new ones that will take us some time to master. Even good new policies and programs can look poor until we get the hang of them. New ideas look bad at first and get better over time.

March encouraged organizations to be “impatient with old ideas, which tend to be relatively good on average, but patient with new ideas, which tend to be relatively poor on average.”52 Only after meeting Jimmy Chen in Brooklyn and watching Annie Rittgers in Cincinnati and seeing citizen-centered and citizen-fueled brainstorms and government-sponsored contests did I understand the real power in them and what working to get beyond average ideas could mean in public life.

Sometimes March wrote about this patience for new ideas and the embracing of serendipity and inconsistency as “the technology of foolishness.”53 And sometimes people bought the argument generally but begged him not to recommend it where they were. “We have enough foolishness,” one colleague told him, hoping that March might not show up where he lived and worked.

We have enough foolishness these days in our politics. We have bad ideas that are bad because we disregard the truth. We have bad ideas that are bad because ideology causes us to hold on to them. I see the foolishness in making the case for more bad ideas. But we can tolerate some more bad ideas—in fact, we need them—as long as we can, ultimately, figure out a way to pursue the good ones that come alongside. We have enough foolishness these days. But also not enough.

REVISITING PART ONE

Government That Can Imagine

What possibility leaders can do

  • Reframe missions around problems to be solved not resources we have.
  • Build diverse teams—pick “not ducks.”
  • Enlist outsiders and the crowd, via contests, crowdsourcing, hackathons, and more.
  • Engage users—find those for whom the need is important, and where and when it is important.
  • Look for “desire lines.”
  • Foster psychological safety—make clear the need for people to speak up; create forums for input; listen; ask good questions; and communicate appreciation for new ideas.
  • Multiply options—use brainstorming (invite “wild” ideas) and other techniques to generate more choices before selecting among them.

What possibility citizens can do

  • Hold officials accountable for their missions, not just their budgets.
  • Demand participation at the outset of policy-generating processes; object when public opinions are solicited only after recommendations have been made.
  • Allow and even invite observation of your experiences with government services—but insist on privacy protections and accountability for follow-up enhancements.
  • Participate in idea-generating activities like community hackathons and provide follow-up engagement.
  • Make known your own needs and the needs of others. Develop empathy for theirs.
  • Push officials to generate lists of possibilities before settling on one approach.

What everyone can do

  • Reject best practices when best isn’t good enough.
  • Be impatient with old ideas. Be patient with new ones.