Moving Beyond the Case Study to Understand How and Why Cities Function

One of the challenges of the twenty-first century is how to create vibrant and productive cities socially, economically, and environmentally. Case studies on specific cities throughout the world have been conducted and provide detailed analysis on how different cities function. For example, the SmartCitiesCouncil (http://smartcitiescouncil.com/smart-cities-information-center/examples-and-case-studies) has compiled such case studies on smart city solutions, such as how the City of Calgary was able to use data to predict and mitigate floods. Case studies cover a range of topics: documenting changes to land use and cover; analyzing the movement and flows of people and information; assessing the role of governance and policy; measuring the impact of technology, infrastructure, and engineering; and determining the influence of economics and finance. They reveal that individual cities possess their own character determined largely by their history, physical environmental, regional to global standing, and—perhaps most importantly—the people living there. Despite the uniqueness of cities across the world, there are common factors and generalized patterns that, when understood together, can lead to a better and more informed urbanism that allows us to solve problems more easily or to potentially avoid them altogether. Local and federal data sources serve as an essential anchor to this process.

Most case studies are driven by underlying data that support city analysis and decision-making. The data identify strengths and weaknesses in the system and provide information on trends over time. More often than not, the data used in these case studies are derived from a combination of primary data collection (e.g., surveys and interviews), authoritative federal data sources, and data acquired through close collaborations with local agencies, including counties and municipalities. Rapid changes to information technology have led to new opportunities to access information, including crowd-sourced data, scraping social media websites, location and time-enabled mobile data collection devices, and volunteered geographic information. This naturally leads to critical questions on how to integrate big data with the case study approach and how to measure the accuracy and validity of such data.

The broad range of access to traditional and new forms of data adds to the challenge of bringing together multiple case studies to form transferable and generalizable knowledge about cities. In some cases, researchers have relied on meta-analytical methods and systematic literature reviews to synthesize case studies. The disadvantages of this approach, which can lead to mixed or biased conclusions, include different data classification schemes (e.g., land use classification), measurement variations (e.g., surface temperature versus air temperature), model parameter choices, and the impact of low quality or inappropriately used data. Alternatively, researchers have also conducted multicity analyses by utilizing the data from the same agency (e.g., federal housing information from Boston and St. Paul) and the same analytical models to identify similarities and differences between cities as the basis for generalizable theories about how cities function. The challenges here include the influence of local and uncontrollable factors (e.g., urban form, policies, climate, social demographics) and uneven access to the required local data sources (e.g., water use, employment, travel patterns).

Recent work by an interdisciplinary team of scholars demonstrates how federal data align with local municipal data sources to form a generalized understanding of cities. In this example, the research focuses on understanding the factors that influence residential water use¹Heejun Chang, Matthew Ryan Bonnette, Philip Stoker, Britt Crow-Miller, and Elizabeth Wentz, “Determinants of Single Family Residential Water Use Across Scales in Four Western US Cities,” Science of the Total Environment 596 (2017): 451-464.²Britt Crow-Miller, Heejun Chang, Philip Stoker, and Elizabeth A. Wentz, “Facilitating Collaborative Urban Water Management Through University-Utility Cooperation,” Sustainable Cities and Society 27 (2016): 475-483.. Studying Austin, Texas, Salt Lake City, Utah, Portland, Oregon, and Phoenix, Arizona, residential water use data (the dependent variable) were assembled from the four separate municipalities, aggregated to monthly usage at the Census Block Group and Census Tract levels. Aggregating to this geographic scale allowed for standardization across individual cities and well as the cross-city comparison. These local data were analyzed in combination with federal data from the National Agricultural Inventory Program (NAIP) for 2011 to classify impervious surfaces (nonvegetative) and vegetation data (NDVI). Findings in all four cities show spatial patterns of water that strongly align with house age and tax-assessed value. This suggests a contribution to the urban theory: aging household water infrastructure and high affluence leads to higher levels of water use. Decision makers could use this knowledge to create targeted incentives and policies aimed at reducing water consumption.

While local decision makers are rarely motivated to theorize on social and environmental issues related to urbanization, they do turn to the experiences of peer cities to help with their own decision-making. For example, they may want to know whether investment in local bicycle infrastructure influences housing prices, human mobility and health, traffic congestion, public perception, bicycle/pedestrian safety, etc. They want insights on what kinds of actions—based on policy, intervention, or incentives—have led to desirable human and environmental outcomes in other locations. Government data provide the anchor that, on a practical level, support these questions and, on a theoretical level, provide the basis for a generalized theory on cities.