Today at the Media Lab, we were joined by Susan Crawford, visiting professor at Harvard Law School and co-director of the Berkman Center for Internet & Society. Susan's last book, Captive Audience, focused on net neutrality. Her most recent book, The Responsive City, focuses on ways that cities are using data to support governance.
(this blog post was written by Nathan Matias and Ed Platt
"The most human technology we have is the Internet," Susan tells us. It gives us the ability to talk to the people we need to, when we need to. "I'm very worried about democracy," she tells us. This past midterm election had the lowest voter turnout in 72 years. At the same time as we have all time lows in participation, citizens are worried about issues of surveillance.
Susan is currently teaching the law of surveillance at Harvard. She argues that we are in a time of great transformation citizens can sense, but we may not always understand. Susan is optimistic that it is possible to draw well-informed courses between the privacy issues and the benefits available for human lives using the data we spin off from our actions. She was moved by Obama’s call for substantive open Internet changes, and calls it a real General Patton moment.
Having spent time at international levels and white house level, Susan is now very excited about cities. In 1950, 30% of people in the world lived in cities. Today, more than half do. By 2050, two-thirds of people will. In that context, Susan identifies three key forces: (a) citizens are becoming accustomed to smartphone interaction, (b) local governments face constrained resources, and (c) there's a trust deficit -- governments need to 'show your work' and do small things well, demonstrating that they're actually working for citizens.
Susan shows us images of the vast challenges that no smartphone app could solve, that only government can take on -- climate change, unemployment, transportation, communications infrastructures, and decaying infrastructure. Susan thinks it's important to see cities as "responsive" rather than "smart" cities-- cities that listen and respond to citizens. To set out this vision, Susan has created a "people magazine" of case studies of people using data and technology to make cities more responsive.
Brenna Berman, the CIO of Chicago, has used A/B testing and statistics to optimise inspection of houses for rats. They demonstrated that by looking in 311 data for correlations of rodent locations, they were able to optimise their rodent response. Susan emphasizes the whole stack of technology, from the fiber that connects data, all the way to the data scientists and city managers who take action on data.
- Open Data
- Open Fiber
In New York City, Mike Flowers has revolutionized city inspections and used data to forge cooperation across different city units who previously weren't talking. ( compared city government to the tribes he saw while serving in Iraq: different groups have their data and are hesitant to share it with others. Mike has worked to overcome that hesitancy and help different government orgs stitch their data together.
In Boston, Nigel Jacob and Chris Osgood at the office of New Urban Mechanics, have pioneered an approach to fixing public issues that creates a conversation between citizen reporters and the city government. Susan reminds us that people aren't voting because they don't think that the government listens or cares. Groups like New Urban Mechanics are changing that, Susan says.
The environmental movement didn't take off, Susan tells us, until we saw a picture of the earth. If we could see the city better, Susan tells us, and understand that we're all in this together, we can start working together on the common good. Lots of screens and visualizations are important, she says. Examples of this include participatory budgeting and 311 platforms. The city of Buffalo is using 311 to identify hot spots in the city that need greater attention. Another example are civic intermediaries like the Smart Chicago Collaborative, led by Dan O Neil. He says "If it doesn’t work for you, it doesn’t work."
She mentioned civic intermediaries could be the glue of civic engagement.
Responsive governments invite citizens to rate their behavior. She talks about Grade.dc, a project that invites citizens to offer feedback on federal agencies-- something she hopes more cities adopt. In Detroit, the Motor City Mapping project asks citizens to report information about blighted properties, helping the city allocate its resources more effectively.
"If you're sick of politics, work on the city level" says Susan. Cities are hiring people from tech companies to drive efforts towards data driven, responsive cities, Susan tells us. Detroit hired Beth Niblock as CIO and Garlin Gilchrist to direct civic technology.
Susan points out that we can't and shouldn't be utopian about data. At the same time, citizen unease about the use and collection of data raises question. Susan believes we are beyond fighting collection, it’s going to happen. It's important to develop the forensic capacity to understand who's using data for what purpose, to understand the ethics of the use of data, and to develop forms of accountability -- if a public employee misuses data, they should be fired. So far, not one of those policies has been developed to any sophistication at any level. Unless we get these questions right, the public will not these technologies.
Questions and Answers
Ethan: you're always sharing a mix between hope and appropriate skepticism. I think it's fascinating to start this conversation from the idea that we have a crisis in participation and trust. I love the observation that the city might be the space where we might be civic but not political -- that we might be able to let go along ideological lines and work together. My own town, Pittsfield MA, hasn't had a partisan mayor's race in 50 years. You had to tell apart the candidates based on what they said what they were going to do. I think there's enormous hope around this idea. For Media Lab researchers who are interested to instrument and sense a city - and what it means to have a "big blue marble" view of the city - you're speaking in the language we want to hear. I also thought it was interesting that you brought up issues of creepiness.
Your opening example of data was focused on rats, and we don't worry about trapping rats. But those same technologies are being used around the idea of predictive policing. Do we find ourselves algorithmically deciding to intervene before we actually suspect someone. Where do we draw the line?
Susan: This is where responsiveness is important. Yesterday in the New York Times, there was an article about drones that could carry out attacks without human intervention. What's troubling about that is the absence of human policy or agency. Police departments were the first to use predictive analytics because their business model benefitted from incremental advances in where crime might happen and allocating resources to those places. CompStat was first in New York. The answer is always to support visibility, public understanding (including the disparate impact), the rule of law. Just because technology is capable of being abused doesn't mean it will be. That's why it's important to cross-train people in law, policy, and tech. We don't want to give up on technology because of the risk.
Ethan: You have a vision of a wall of screens that show you the city, built on fiber, sensors, visualizations. Who gets to look at the screen? We presumably don't want a system where an algorithm is making a decision to put more patrols on the streets. We don't even want an algorithm determining by itself where we catch the rats -- we want someone looking at the screen making a humane, policy-informed decision.
One of the problems with the Smart City rhetoric is the idea of the screen view -- which is oriented towards a city manager. We also talk about open data, and the hope there is that somehow citizens are going to be involved in this process. Is that screen for the mayor, the employees of the city, or the citizens?
Susan: it's a false assumption that there is one screen. There are many screens. Some screens could be seen by everybody -- maybe it's just a layer of data on all parts of the city that speaks to us as we need it. The city managers will also have their own screens. Experimenting with this is explosively new within city government, and there's something exciting about it that we should celebrate.
Hal Roberts: you present a picture of happy people working effectively in cities. I'm also afraid about this for two reasons. In the Wire, CompStat is misused by the police. Secondly, in the case of public education, I'm worried that school starts focusing on what you can measure rather than what matters.
Susan: Feedback loops can emerge in these systems. One fear is The Wire. The other side is No Child Left Behind, where education stops becoming about increasing human capacity. I respond that humans are cheerful and resilient. We're still learning how to put data in context, to learn how to use it in a way that is appropriate for humans. Rather than saying "don't do it," there are great benefits, and we should use policy to slog through these issues.
Ethan: I think Hal's reference in the Wire was to the notion of a police department that's already being called out on having a high rate of crime, suddenly has a perverse incentive to ignore a wave of murders. This becomes the main plot point late in the show, because the police department needs to show numbers saying that crime numbers are going down. Might highly responsive cities become a panopticon for the employees of the city -- if you find yourself in a place where you have low trust in government and fear that you're going to be voted out, how do you deal with that issue? How do you make sure that the screen is an honest and fair one and remove the incentive to fudge the numbers? Might transparency make it harder to deal with these issues?
Susan: City employees often point out that if people tell them about issues, they would have to fix them, and that without public trust, they don't have the resources. Susan says that her co-author Stephen Goldsmith would argue that it *is* possible, and that trust is the most important part of this. You can avoid corruption by showing transparency. That's why she focuses on trust.
Jason Haas: Is there a degree to which operationalize the way city workers do their jobs might put them in a bound?
Susan: In the beginning of the 20th century, we were worried about corruption and created lots of rules and roles-- these were created by Teddy Roosevelt. People forgot the reason for those rules, they outlived their usefulness. The pendulum that swings back and forth; we need to break through that. By giving people data, it's possible to make available information about the context to empower people to act in context. At the moment, we're so far on the constraint side that I'm not worried about employees losing agency. We need to reform civil service rules and procurement -- issues that are both addressed in the book.
Saul Tannenbaum: Who chooses the sensors, and who chooses the algorithms? All the analytics in the world won't address big issues like incarceration rates or the prison industrial complex. Saul refers to Million Dollar Blocks, a project by Laura Kuran to show areas of New York City where we are spending more than a million dollars a year to incarcerate people. Her argument is that by doing this visualization, we can argue a policy point. Ethan: deciding what to instrument and display could be very different depending on what your goals are.
Susan: I agree and seek opportunities to make plain what was invisible before. To take advantage of projects like Million Dollar blocks, we need data informed policy maker. We need more people informed about policy and data and social justice. Who decides what sensors go out and what data is collected are matters of public policy. In Chicago, each sensor is subject to public deliberation, are swapped out every six months, and automatically contribute to an open data portal. You need leaders who are not afraid to talk about the politics of data, listen to the public, and articulate the benefits.
Matt Carroll: Many cities don't have the resources to do this. What about the small cities?
Susan: I have never run into a more collaborative sector in my live. Chicago, funded by Bloomberg Philanthropies, is developing an open source smart city platform that can take data inputs, map things against a shared map interface, and use that data to make policy decisions, and it has some analytics above it. Chicago wants that tool to be used by every city in America. Chattanooga is a great example of this work: they have fiber, a library that's hosting the data, and the mayor wants to make analytics-informed decisions. Since there are so many open source candidates - you don’t need to buy softwares from big companies - The main gap is in the data scientists who can support this work. With a fiber + data combination, small cities are going to make the leap first. I see an optimistic story for small cities if there's a leader and maybe even a student intern who is excited about these issues.
Participant: How do you propose we solve the problem of not having enough civic engagement?
Susan: If something is more visible to you and you're involved in it, you will care. Cities actually need to respond to pulses coming in from various places -- the worst thing would be if cities said, "we want to co-create things with you" and then didn't actually listen.
Rahul Bhargava: He wants to probe the "engaging communities" part of the subtitle of her book. How are communities engaging with data being collected? Can we move beyond just mapping potholes?
Susan: In Austin - when school data is shared, people’s motion slowed down when they are looking at the shared screen.
Carey Nadeau: Once we get citizen and government participation, that doesn’t directly translate into policy impact. Are there examples of best practices?
Susan: We don’t have enough of that yet. Mapping things like blight do influence policy, but these are very primitive stories. There are some examples - such as mapping toilets in a community in India - where the feedback loop has been tracked, but these examples are rare.
Participant: Asks about lack of responsiveness from the governments themselves.
Susan: Younger people are used to faster response times. Some cities like Barcelona are adjusting to this well. In some cases, the problem becomes the opportunity.
Erhardt: What is the role of private industry in creating a responsive city?
Susan: I like to see more small business bring about flexibility to this field. For now, big giants are driving the crossbow. Private industry, small and local companies are experiencing crisis and opportunity.
Lily Bui: The book discusses defining roles and using data to contextualize those roles. What are the roles that need to be assigned?
Susan: Groups that we see develop can have different roles, and might change over time. Who is the project manager, who is the scheduler, etc. These are roles that could be held by many different people in the real world, and can manifest in a persistent way in a virtual world.