STEPHEN GOLDSMITH, the Daniel Paul Professor of the Practice of Government and former mayor of Indianapolis, is focusing much of his current work on data and cities. Last year he launched Data Smart Cities Solutions at the Ash Center for Democratic Governance and Innovation. He is co-author of the soon-to-be-released book titled "The Responsive City."
Q: How does data have the potential to transform cities in the future?
Increasingly, people are living in cities – there is much more congestion, a lot more people, but at the same time there is a lot more data being generated. There are sensors, there are cell phones, there are all kinds of information being generated. And so if you think about government needing to individualize its responses to problems and people, it is now very possible because there are huge amounts of data that can be relatively easily mined and analyzed, while also paying close attention to privacy issues. The information can be used by local government leaders to think through how to better maintain their cities and help city residents. So, as we think about the tension between an overabundance of municipal challenges and too much data, if you bring those together you can dramatically improve the quality of city services.
Q: Which cities have been the leaders in harnessing data to improve public services?
There is a wide range of ways you can use data. We have featured a good amount of time in our writing, in the book I recently did, and on our website to the Boston experience, asking how do citizens engage their community, how do they use their iPhone, Android, to report a problem? How do they get back a response? But that’s just really the start. New York City determines which building’s going to burn down next, not which building has burned down, but which one’s going to burn down next. Police departments around the country are dramatically changing the way they police, based on data. But, Indiana’s using a case that says, we have a lot of kids that are in trouble, we have a high infant mortality rate, how can we bring down the infant mortality by looking at lots of data – medical data, school data, family data, all with attention to privacy – and then figuring out what interventions work in what sort of situations. So, there are massive amounts of data that we see being used around the country and around the world to change the way social services, the hard services, are being delivered.
Q: Are there cities outside of the U.S. that are also serving as leaders in this space?
The space is developing in all sorts of places. In Europe, Barcelona and other cities in Spain have been leaders. Of course, Singapore is a leader. In South America, we see work in Brazil, San Paolo and Rio, as well. There are a couple of projects in Buenos Aires, Argentina. There are activities going on around the globe.
What’s interesting about the movement is that it sometimes begins in the private or non-profit sector and sometimes it begins in the government sector. Sometimes the government leads by saying “we’re going to take all this data and put it together, and figure out how to do services better.” For instance, Smart Chicago in Chicago, funded in part by MacArthur and the Chicago Community Trust, is looking at ways that they’re taking the data in the open data transparency movement. Now, data sets are out there, so a non-profit that is sophisticated in technology tools can help show how services are being delivered in my neighborhood, how they should be delivered in my neighborhood, how my neighborhood is performing compared to others, and how can we fix problems. So we see data being used across various sectors and various places to derive value.
Q: What are the barriers that cities and other local governments face when attempting to leverage the ever growing amount of data available to improve governance?
The barrier question is an interesting one for a couple of reasons. Until recently, most people may have thought that the barriers were technological – we have separate data systems; this system doesn’t talk to that system; there’s too much data to look at; some of the data is structured data; some is unstructured on YouTube or on other video websites. But technology tools today are so sophisticated. That makes it easier to reach in and grab some data, and to determine how that data might relate to other data in a totally different set. So, the technology barriers are falling dramatically.
The real important barriers, though, are people barriers. You need a leader – be that a public official, an agency leader, a mayor or governor, or an agency inside of a federal office – that can demonstrate the value in reaching across silos to leverage relevant data. There has to be leadership at the top that gets over those barriers, that gets those people to work together. So, that’s one issue. The second issue is even more difficult, and one we’ve been spending a fair amount of time on. Bureaucracies that run public sector organizations in highly advanced civil service organizations have very defined work functions. Employees work within very narrow boxes with lots of rules. The view is that you can’t give public employees too much discretion or they’ll abuse the discretion; that’s the only way to hold them accountable. What the data tools do is very critical. They allow us to give our employees more discretion to be responsive to our citizens. We know in real-time what they’re doing, and we can hold them accountable with the same data that we’re allowing them to use. So, this last hurdle comes down to how the government is structured. How does it hire? How does it procure? What discretion does it give its employees? So, we have data tools, leadership issues, and the structure of government. When they come together, the responses are really quite dramatic.
Q: How does this data revolution impact the average city resident?
There are a couple ways that the average resident of a city will be affected: One is that he or she wants more in services than there are resources available. So, for every dollar that a city or state spends, it has to spend that money better. So, instead of spending a dollar on a hundred different projects, you would want to do some sort of 80/20 role, right? Where are the 20 percent of the problems where I can make the most difference, and how do I use data to make a difference there? Not every problem is the same. Not everybody applying for a permit is the same – some are good actors, and some are bad actors. Let’s distinguish between those. Let’s fast-track those who want to get permits, who have a history of good operations, and let’s look more carefully at those who don’t. So, if you’re an average small business person, data will help you get your licenses and permits faster. If you’re a contractor or you have a project, it will help you do things better. But if you’re a bad actor then you’ll face more scrutiny. That’s one issue.
Another issue is figuring out how citizens can get those services on a more personal level. I have a personal relationship with Amazon. Maybe not so personal to them, but it’s pretty personal to me. There’s no reason why that same relationship can’t develop between a citizen and his or her government. What are the traffic problems in my neighborhood? When is the park going to open up in my neighborhood? When do I have to get my dog permit again? How do I apply for my benefits one time instead of 20?
So, the personalization of these tools is dramatic. It allows me to have a more effective voice through social media. I have a more personalized relationship through these tools, and I have more effective government. All these things are inherently possible if we have innovative leadership and pursue changes in the ways in which we govern.
Q: You touched on privacy earlier in our discussion. How would you reassure the citizen who has that concern when they hear about the increasing use of data by the government?
I think privacy questions are very serious and very significant. It’s a mistake to assume them away, but it’s also a mistake to assume that they’re so significant we can’t use data. So, there’s a couple ways to think about this: One is there’s a lot of data that can be anonymized. If you’re driving in midtown Manhattan, and you are generating EZ-Pass information or GPS information, then that information can be utilized without any identifiers in order to set the timing on the street lights. So, we know how fast 10,000 cars are moving. We take that information, and we connect it to the street light cycles so that traffic flows more smoothly. Nobody keeps a record of who it is that was driving. So, if the controls are right, that helps.
The second issue is a little bit more difficult. Let’s take a child welfare case for example. We may have a family that’s involved with child protective services, and may also be involved with other government agencies. So, certain individuals who are in trusted relationships with that family – case worker relationships – should have access to that information under carefully guarded conditions.
We are increasingly aware that our mobile phones generate GPS information about where we are all the time. Our buying habits are collected by a lot of folks, and the government has access to some of that information, as well. I think the short answer to this very difficult question is that we need to be very intentional about when and how government agencies have access to different levels of information. We need to have very thoughtful policies in place. And we need to ensure that people who abuse these privacy and opportunity rights are punished.
Q: If I’m a mayor in Cambridge or Cincinnati or some small city and I want to get on this bandwagon, how do I do it? Where do I start? How do I think about these challenges?
What’s interesting today is that we think about the high costs of sophisticated technologies, but if we look at the pieces of exiting low-cost technologies, they add up in a really remarkable way allowing for small and mid-sized municipalities to use them. We all have cell phones, and most people have smart phones, even the majority of folks in low income communities. Increasingly, public employees have smart phones so that when they’re in the field they know where to go, how to solve a problem, and how to get advice. The cost of those phones has dropped significantly in recent years.
Cloud computing allows a small city outside of Boston, or outside of Indianapolis, or outside of New York to acquire many of these services in much less expensive ways than when you had to buy an enterprise license for software. Cloud computing allows the smaller and mid-size cities to buy the best access to technology at very affordable prices.
So, the issue now becomes hiring the right people to leverage the data and the technologies to which we have access. Data scientists are highly sought-after. They’re difficult to secure. So, where does one get good advice? I think that if you have somebody in your government who is responsible for performance and for predictive analytics, they don’t have to be an amazing technology person, but they do need to know how to secure these services and how to procure the right guidance. Increasingly, we have witnessed two, three, four, five cities start adopting the same software, and it therefore becomes more affordable and much more available in a shared service environment. The opportunities to buy right and buy smart, and deliver services for mid-size and even small cities are better today than ever before.