Wednesday, November 5, 2014
Review of The Fires: How a Computer Formula, Big Ideas, and the Best of Intentions Burned Down New York City--and Determined the Future of Cities by Joe Flood (Riverhead, 2011)
From the perspective of my own interest in the present drive to create ‘smart cities’, The Fires provides an excellent analysis of one of the first attempts to systematically apply systems analysis underpinned by computer modelling to the management of city services and logic behind using such an approach. The RAND Corporation were contracted by the city to determine how best to re-organise the fire service to improve its effectiveness whilst saving resources, including identifying what companies to close and where to open new stations. They were employed with the aim of showing the value of displacing Tammany politics through the introduction of rational, technocratic solutions that employed an impartial scientific method to city governance. The findings from the models RAND built were implemented in practice altering the management of the fire service and its day to day operations. Moreover, the models produced earned their developers a host of major international prizes for applied scientific research.
However, subsequent research has demonstrated that the RAND models suffered from four major problems which had severe knock-on effects on the ability of the NYFD to effectively fight fires, which had knock-on effects to the rate and impact of fires. The model suffered from poor and gamed data, an information and cultural gap between the modellers and the domain they were modelling, faulty assumptions concerning how to measure company efficiency and effectiveness and poor model building (for example, comparing companies within seven types of areas but not across them so that they were not computing optimization of coverage across the city but within areas it deemed similar), and political influence shaped how the models were constructed, calibrated and used. In total, 34 of the city’s busiest fire companies were closed, despite warnings from senior fire officers as to the reality of what was going on in neighbourhoods and to the remaining fire companies and suggestions for alternative approaches, and somewhat perplexingly a number of others were opened in sleepy backwaters. Along with the effects of other social, political and economic forces at work, the result was ‘The War Years’ when ‘hundreds of thousands of people in the Bronx, central Brooklyn, Manhattan’s Lower East Side and Harlem neighbourhoods were burned from their homes’ (p. 8). In the Bronx, seven census tracts lost more than 97 percent of their buildings to fire or abandonment (in census tract 2 of 836 residential buildings recorded in the 1970 census only 9 remained in 1980), with 44 tracts (out of 289 in the borough) losing more than 50 percent.
Flood concludes the book with a short overview of how such a systems approach has become increasingly adopted by governments, driven by an audit culture, a political reform agenda, and the adoption of business governance practices. This is the weakest chapter of the book, with the analysis skimming over the rise of new managerialism, neoliberalism, and technocratic approaches to governance. This is clearly not the central mandate of the book, but it does mean that the legacy of the New York RAND experiment is not elaborated.
Despite identifying significant flaws in the systems approach to city governance, Floods’ conclusion is not that systems analysis and city modelling research should be avoided by city administrations, but rather that its use needs to be balanced with other forms of knowledge: rather than episteme (scientific knowledge) and teche (practical instrumental knowledge) replacing phronesis (knowledge derived from practice and deliberation) and metis (knowledge based on experience), the insights drawn from all four need to be considered and debated to formulate policy. In other words, data and algorithms should not be allowed to simply trump reason and experience. Moreover, cities should not be conceived of as easily knowable and manageable systems that can be steered and controlled in mechanical, linear ways -- input data and then pull levers to get an appropriate response -- but rather cities should be understood as complex, contingent, and relational, that often unfold in unpredictable ways. In such a view, models might provide valuable insights, but they are not the only insights which should be inherently and slavishly followed. It’s a conclusion, I think, that holds merit.
Overall, The Fires is an excellent read -- well written, engaging, and insightful. It will provide a fascinating story to anyone who is interested in contemporary urban history, and in my view it’s a must read book for all those presently involved in conceiving and building smart city initiatives.