A few general thoughts about the potential of using big data to better understand cities
How can we more effectively manage cities? How do we know what governmental interventions actually achieve their stated goals? In the not too distant future, today’s ad hoc measurement of public programs will seem as quaint as the 19th century practice of budget-less city appropriations.
Improving public management through measurement is not a new idea, but the recent radical growth in digital data offers hope that we might increase the frequency, robustness and scale of such techniques. Like all things in the future, the ultimate impact of such measurement is uncertain, but we can think deeply about what will and will not change with increased data analysis as a way to understand the scope of this potential and how we might realize it.
Any trip to a town hall meeting will quickly demonstrate that the implications of a data analysis often depends on where you stand. And as Kuhn brilliantly illuminates, even the purest of scientific inquiries are not immune to seeing the world through the lens of their particular position:
“Practicing in different worlds, the two groups of scientists see different things when they look from the same point in the same direction. Again, that is not to say that they can see anything they please. Both are looking at the world, and what they look at has not changed. But in some areas they see different things, and they see them in different relations one to the other. That is why a law that cannot even be demonstrated to one group of scientists may occasionally seem intuitively obvious to another.”
This problem becomes more acute in tackling societal questions. Any scientific researcher or industry analyst or public manager brings a unique set of experiences, values, and other biases.
In conducting such inquiry, data allows us to attempt to “step outside” of our particular position and utilize the “vast and unique man-made imagination machine” that is mathematics. We might build a model to make rigorous the implications of our assumptions about a future scenario. We might quantify observations to synthesize them into a judgment about past performance. Or we might incorporate signals beyond our human sensory limits – say a GIS aerial map – to better understand where we are today. Such abstractions or “maps” serve as incredibly useful tools for navigating the messy “territory” of human cities.
Consider how we might explore basic questions facing any public manager: What are the problems in a community? How might we address them? Any answer remains inextricably linked to time, place, culture and other local contingencies. Moreover, unlike the physical sciences, how we go about answering such questions must engage such contingencies directly. When a rocket scientist conducts an analysis, her subject does not think back at her. A public manager does not analyze in a vacuum but rather engages such questions as part of a larger human community and other individuals with their own perspectives on these issues.
So rather than a new universal theory of cities or yet another attempt at a grand theory of human civilization,  the hope with big data is that we might create something far more useful: nontrivially better maps to guide us through the incredible complexity of human cities. Progress here will be come not merely through analyzing data but testing our new maps in the messy reality of human cities.
Although full of promise, this new inquiry also offers much risk. In our excitement over new tools like machine learning, we might forget to adopt the proper humility towards the thousands of years spent tackling these questions using the tried and true human sort. And in our rush to leverage all this new digital data, we might lose sight of the fact that humans have been making and recording observations – also known as generating data – since the dawn of symbolic communication. One way to describe history is all the data we have collected so far, and we should not forget that old fashioned data collection methods like talking to the people directly affected by our recommendations have not lost their power.
We must also not forget that although we have new tools at our disposal, the problems we are tackling could not be older. How might we as humans find a better way to live together? We need to respect and preserve the wealth of institutional knowledge that’s been accumulated firsthand in addressing that challenge. At the same time, we need to have the courage to follow reason even if it takes us to unexpected places and most importantly not take the current institutional reality of cities as a given. Insofar as these new tools prove useful, we should design our basic public institutions to make use of them and continuously work to align the form of these institutions with the functions informed by our inquiry into big data and cities.
Although aspiring towards the elegance and predictive power of a field like physics, this project must remain humble and retain the lessons of softer fields like the humanities. Like the more general study of measurement and structure or what is commonly called mathematics, this inquiry into big data and cities will remain both an art and a science.
 Thomas Kuhn p. 152 The Structure of Scientific Revolutions
 See Brian Rotman, Mathematics as Sign for an excellent treatment of the foundational question on what is mathematics. For other great takes on the unending tension between objectivity and subjectivity, please see Thomas Nagel’s The View from Nowhere and Quintin Meillasouix’s After Finitude.
 I suppose ultimately this question will hinge on one’s beliefs about the deterministic nature of the universe and individual human free will.
 That doesn’t mean, however, that there aren’t methodological synergies. See The Measurement of Uncertainty: a History of Statistics before 1900 for an incredible chronicle of how techniques developed originally in astronomy (notably the now nearly ubiquitous linear regression) impacted the social sciences.