Big Data, Urban Science and the Search for New Ways of Improving Life in the City
Imagine you had all the data you could possibly want about a city. I’m talking about real time readouts of all inputs and output, along with documentation of public satisfaction levels with all aspects of city life. Further, presume you had access to similar information for other cities as well, so you could see trends and patterns. What might you do with all this information to help improve the quality of life in the city?
There are some urban planners and city administrators who believe that if they had that kind of information they would be able to figure out when and how to spend public money, allocate equipment and personnel as well as restrict and support private efforts most efficiently. They also think they could use the same data to anticipate certain tipping points so that traffic snarls could be avoided before they happen, public works staff (including police and fire) could be deployed where they would be most needed, and infrastructure repairs could be managed with minimal disruption. With such information, the presumption is, water, electricity and other city resources could be priced and deployed on a real-time basis in the most cost-effective way possible. Empty housing and commercial space could be repurposed almost immediately, and priced to match the city’s urban development objectives. Taxes and fees could be collected electronically while feedback from residents could be shared with officials on a continuous basis. With the right kinds of electronic monitoring (including sensors of all kinds), sufficient data collection, investment in high-level analytic capabilities and appropriately trained staff, a city could become a “smart city.” All this digitized data could be displayed in visual form -- across multiple platforms – so that everything would be easy to read and understand.
Of course, knowing what problems a city faces, and even understanding what’s causing them, is not the same thing as being able to respond effectively. The availability of real time data, even with the most advanced application of artificial intelligence, won’t make it clear who ought to do what, in what order, in what way and for whose benefit. These are political choices: “is” does not lead directly to “ought.”
One school of thought, promoted by some economists and engineers, assumes that the goal of city management should be to maximize efficiency – eliminate waste and stretch every dollar as far as possible. They want to make sure that tax revenue, fees and intergovernmental transfers are allocated in the most cost-effective fashion. If the goal is to collect trash, arrest criminals, clean-up air and water pollution, or fight climate change, money shouldn’t be wasted in the process.
A second school of thought, inspired by ecologists and advocates of sustainable development, believes that every dollar of public spending should be used to meet economic, environmental and social needs simultaneously, in ways that account for long-term needs. Efficiency in the short-term isn’t as important to these thinkers as long-term sustainability (which includes meeting the needs of both current and future generations in as fair a way as possible). All the data in the world won’t make it clear what ought to be done. In a democracy, such choices need to be made through a messy process of reconciling conflicting interests and values in which the population participates directly. Efficiency isn’t always the highest priority goal.
My own university, MIT, is thinking about launching a new undergraduate degree program in Big Data and Urban Science. This would bring together faculty from urban planning, information science, electrical engineering, city design and the applied social sciences to prepare undergraduates to build and operate smart cities. Other universities, as described in a recent issue of the Chronicle of Higher Education, are a step ahead. NYU, Northeastern, Carnegie-Mellon, John Hopkins, University of Illinois, University of Rotterdam and others have already launched undergraduate and graduate degree programs that seek to merge teaching about big data and urban studies.
As faculty in all of these programs try to decide what skills and knowledge they want a new generation of urban scientists to master, there are six questions I think they need to confront:
- What do we mean by a city? (Are they talking about activities that take place within a municipal boundary, or will they focus on a larger set of regional, national and international forces that shape urban life more generally?)
- Do they think that privacy is a concern? (Do they assume that any and all information that can be collected, should be collected? And should this information be available to anybody who wants to use it? Can people or organizations opt out and keep information about themselves private?)
- How will we fend off persistent, current, and future cyber-attacks on critical urban infrastructure? (If urban science means greater centralization of information and data management, won’t that increase vulnerability to attack?)
- Who do they think should be in charge of designing and managing big data systems as well as setting the standards used to make judgments about what’s working well and what’s not? (Will this be a managerial task assigned to various government agencies, or will elected officials be accountable for how all this information is collected, analyzed and interpreted? Will new laws be required to ensure that individual and organizational rights are protected? Will this require federal, state or local legislation? Where will enforcement responsibility sit?)
- Will they be working from and toward an idealized model of an efficient city, or will they work to preserve historical and cultural diversity and variation? (I presume that students from all over the world will want to participate in these programs? Won’t the differences in culture, laws, and history require very different ways of applying the new urban science in each country? Should we assume that this is basically a technical education and teach a kind of “one-world view”? Or, would that be a terrible mistake?)
- Are the universities involved aiming to prepare public employees (whose job it is to serve the public interest), or are they training experts who will sell their services to the highest bidder?
The City As a Place vs. the City as an Idea
Most efforts to model urban dynamics assume that a city can be described as a series of “stocks” and “flows” within a set of boundaries. While there are always important “feedback loops,” many of which are likely to cause unexpected consequences, most modeling (and forecasting) efforts begin by postulating a set of boundaries. But, what if cities, as many urbanists contend, are largely a product of a great many extra-territorial (even global) forces? Capital or data flows originating in other parts of the world may have as much of an effects as economic and social forces originating in the city. Moreover, if we say that the operation of many of the sub-systems in a city reflect the ways that groups of people or institutions think about things – their perceptions -- how do we include these in the models we teach students to build? The city is a physical space affected by global geological and ecological forces. It is also an idea shaped by millions of individual perceptions. Will it be possible to make sufficiently simplified models of the city to generate useful insights and predictions?
What’s Confidential and What’s Not?
Assume we can answer the first question, and we know which data are required to make a city substantially smarter. Gathering some of these data will require tapping into otherwise secure or private data sources. And, even if we argue that individual identities will be scrubbed, should people and organizations be forced to give up their privacy in the name of the greater good? In the United States, we are currently watching a version of this question play out as a National Commission on Elections and Voting demands that all 50 states to turn over voting data that may reveal how specific individuals voted. When it comes to financial records (including tax returns and credit card expenditures), even if these data are crucial to modelling how a city is doing, should individuals have a right to keep such personal data confidential? What ethical obligations will we impose ona new generation of big data analysts or urban scientists?
Cyber-Security in the Urban Realm
Urban infrastructure is already under attack from hackers who seek to hold energy, medical, water, sewage and other systems hostage. Each new layer of encryption added in response just ups the ante. These systems are vulnerable because individuals are not as conscious of their cyber-security responsibilities as they should be. Solutions will require further technological innovation and investment, but that won’t be sufficient. What will we teach a new generation about cyber-security and how to maintain it? And, given the vulnerability of critical urban infrastructure to global attack, should that affect what we guarantee with regard to privacy and control over personal data?
Knowledge, Power and Authority
Let’s say a city has put together a comprehensive data gathering, analysis and visualization operation. Who will have access to the raw data? Who will have the right to publish analyses of the information that has been collected? Will the city be willing to share assessments of things that residents think are going badly? Will those who want to challenge current office holders be allowed access and permitted to publish any analysis they like? Who will make decisions about how data should and should not be interpreted? It’s my assumptions that managers of smart cities will have “to do” lists that far exceed their resources. Setting priorities (often in real time) will require quick decisions, faster than the public can follow. If all big data about cities were open sourced, would that allow more citizens to be involved in helping to make decisions that are going to affect them? While real time referenda might be possible, is that how cities should set priorities and make judgments?
Is There an Ideal City?
Urban planners are very place-oriented; data scientists are not. Urban planners want to preserve the special historical and cultural features of each city. Data scientists, on the other hand, are looking for rules of thumb to describe the most efficient ways of delivering goods and services in general. They might inclined to disregard inefficiencies that are a by-product of local history, culture or values. There’s no point collecting data if there’s no intention to use it, but putting all these data to use means measuring how things are going compared to some benchmarks. Should benchmarks be unique to each community? Or, is the goal of merging big data and urban science to create “ideal benchmarks” (based on studies of many cities over time)? Should urban science be practiced differently in different cities, let alone different countries?
Public Sector vs. Private Sector Careers
Urban planning education in North America has been provided by major colleges and universities for more than 80 years. The majority of graduates of such programs aspire to work in the public sector or in civil society (e.g. NGOs or public interest organizations). This is true regardless of where students originate. Of course, some graduates find private sector jobs, either temporarily or permanently in consulting firms or corporations. Whether they are headed to the public or private sector, students studying urban planning tend to focus on ways of meeting the needs of the poor and the disadvantaged; they start with a theory of market failure and look for ways of using public-private partnerships, regulation, public investment or political advocacy to meet the needs of those for whom the market tends to fail. The engineers and scientists likely to be drawn to these new urban science programs may not be so public sector oriented. Will the new urban scientists/big data managers who graduate from these programs be public sector/civil society or private sector oriented?
I see the emergence of interdisciplinary urban science programs around the world as a good sign. Merging the capabilities of scientists and engineers with applied social scientists, designers and urbanists interested in the life of urban residents would be a positive development. We need to provide all the help we can to people in cities trying to make adjustments and reforms that reflect a clear-headed awareness of the complex dynamics they face. I worry, though, that some universities moving in this direction may pay too much attention to the advice of economists and management gurus obsessed with numerical trends, who are willing to focus on correlation because they don’t have the tools to understand causal dynamics. I hope the applied social scientists and urban planners will succeed in ensuring progressive values like concerns about fairness and sustainability are at the core of the training of a new generation of urban scientists. I’m certainly glad that most of the people involved in these new efforts appear to be committed to blending schools of thought that have operated separately for too long.