How to study the smart city?

Over the past years, the notion of ‘smart city’ has become firmly entrenched in a variety of domains, including urban planning, policymaking, and science and technology studies, to name a few. Many of those working on smart cities stress the highly convoluted nature of the adjective ‘smart’, which contains a plurality of more and less commensurable perspectives and values. The ‘smart city’ point is a two-sided phenomenon that could bring about potential dangers as well as benefits. How can discussions on urban governance gain more traction on the notion of ‘smart’ and be aligned with developments taking place on the cusp of the world-wide smart city?

The rise of data-driven urbanism

On May 25th 2016, PBL Netherlands Environmental Assessment Agency invited two scholars working on smart cities to clarify potential benefits and dangers of smart cities. Professor Rob Kitchin (Maynooth University, Ireland) kicked off the debate with his talk entitled ‘Smart Cities: Realizing the promises whilst minimizing the perils’. As Kitchin pointed out, most discussions on smart cities encompass different dynamics, i.e. instrumentation and the regulation needed to ensure socially acceptable forms of data collection, implications of smart cities on policy and economy, and social innovation in the realm of civic engagement. These dynamics each put different emphases on different aspects of smart cities.

Smart cities typically imply that cities are becoming more and more data-driven, which is due to the increasing ubiquity of various technologies used for data gathering (e.g. sensors, cameras, and mobile phones), and the subsequent intertwining of this data in various aspects of urban governance (e.g. government, security and emergency services, transport, energy, waste processing, management of buildings and homes, and civic engagement, see figure 1). Publicly and privately generated data is uniquely indexical due to unique identifiers of various devices, the most dramatic example being one’s mobile phone that has a unique ID and network address. Together, increasingly ubiquitous technologies used for data gathering and uniquely indexical devices produce data about citizens and places in real-time, leading to a ‘data deluge’ that can be combined, analyzed, and acted upon. Although more and more data is available, the challenge is to turn this data into information that can be acted upon.

Figure 1. Representations of sensor networks and ICT infrastructures in smart cities often adopt a bird’s eye diagrammatic approach, through which the various functionalities of technologically augmented environments are illustrated.

The data-driven forms of urbanism that result from the aforementioned developments have become a global phenomenon, and have established networks of interlinked cities that have become ‘knowable’ and ‘controllable’. As a result, the operational governance of city services is becoming highly responsive to a form of networked urbanism in which big data systems prefigure and set the urban agenda, persistently driven by the promise of smart people, governance, mobility, sustainability, and cutting-edge innovation.

Critiquing data-driven urbanism

Kitchin countered this promise of by outlining the following critiques of the smart city.

Many practices around the smart city presuppose that cities are knowable, rational, and steerable machines. As a result, operational governance can supposedly be performed using a set of mechanistic principles, underpinned by a belief in instrumental rationality, key performance indicators (KPIs), and analytics, all of which drive new forms of managerialism. However, cities appear to be fluid, open, complex, multi-leveled, contingent and relational systems, which is an insight that does not bode well for the viability of the new managerialism accompanying thinking about smart cities.

Smart city discourses often present data-driven urbanism as an objective, neutral, and non-ideological approach to urban governance, thereby delivering a politically benign and commonsensical image of itself. However, data-driven urbanism (and other forms of urban governance like it) is highly situated, contingent, relational, framed, and implemented in order to achieve certain aims and goals. Failing to recognize this leads to an inability to understand the profound social, political, and ethical effects of data-driven urbanism, which are due to surveillance and the ensuing erosion of privacy, issues regarding the ownership and access to data, and ‘control creep’ – a process whereby a technology being used for one purpose, slowly becomes adopted in another area of urban governance beyond its originally intended use, where it ends up delivering new forms of control and surveillance. What is more, smart cities are created by visions that betray vested interests, which may end up reinforcing power geometries and inequalities. Through control and regulation, certain constituencies rather than others are serviced, leading to the marginalization and dispossession of certain populations.

The technocratic approaches to urban governance that often accompany perspectives on smart cities can establish ‘solutionism’, which is the idea that all aspects of a city can be treated as technical problems and solved through technical interventions. This approach to urban governance presupposes complex issues can be broken down into neatly identified and solvable puzzles, all of which can be fixed using computational techniques. Those who subscribe to solutionism emphasize the value of data and algorithms, and may thereby end up disqualifying other forms of knowledge, such as phronesis (knowledge derived from practice and deliberation) and metis (knowledge based on experience). Kitchin emphasizes the value of careful deliberation and inclusion of various kinds of knowledge in order to remedy potentially harmful effects of solutionism. Failing to adopt a diverse portfolio of knowledge can establish solutionism as an ahistorical, aspatial, and homogenizing ‘one size fits all approach’, in which cities are treated as generic ‘markets’. This runs counter to the aforementioned idea that cities are complexly relational entities.

More generally, data-driven urbanism relies heavily on technological augmentation of urban governance. This reliance on networked technologies may render cities buggy, brittle, and hackable. Smart cities often appear to be built on neoliberal political economies that promote the corporatization of governance, effectively reducing government functions to market opportunities and hollowing out of the state. However, when cities are administered for profit only, corporate path dependencies and technological lock-ins may be the result.

Balancing ‘the positive’ and ‘the negative’

In order to deal with the foregoing issues, Kitchin proposes to “get smarter about smart cities”. This entails realizing the promises of smart cities while minimizing potential harms, in other words, a matter of balancing the positive and the negative. Rather than abandon the notion of smart cities, we need to re-imagine and reframe them in various ways. One way of doing so is reframing goals related to smart cities by asking for whom and what they are important. Are they aimed at establishing new markets and forms of profit, state control and regulation, or aimed at citizens and improving their quality of life? The process of reframing smart cities could also benefit asking what kind of cities ‘we’ want to live in (more on this later). Another way of reframing smart cities is reframing the notion of ‘city’ itself. Thinking about smart cities needs to recognize the value of a more nuanced and relational understanding of cities that abstains from the neocybernetic frame in which cities are equivalent to machines, and improving cities becomes a matter of turning cogs. In addition, reframing smart cities requires the reframing of management and governance. One way of doing so is by creating new forms of civic engagement It would be wise for cities to avoid becoming accidental smart cities and make plans or roadmaps to ‘smartness’ instead. Finally, Kitchin stresses the importance of reframing epistemology: rather than taking recourse to realist epistemologies and instrumental forms of rationality, our understanding or smart cities could benefit from an acknowledgment of situatedness, positionality, contingencies, uncertainties, and assumptions.

The Datapolis

In the second talk of the debate, prof. Albert Meijer (Utrecht University, The Netherlands) argued that an understanding of the various notions of ‘smart’ don’t really help us to know what’s going on in the area of smart cities. Typically, smart cities are approached from an engineering perspective, which contributes to the view of cities as systems that can be optimized using technological interventions in ways outlined above. Another perspective on smart cities is the so-called ‘power struggle perspective’, which sees smart cities as an arena involving multiple agents that all attempt to bring their own ideals and visions on smartness into existence. Meijer proposes a governance perspective that combines both of the aforementioned perspectives, and argues that smart cities involve both ‘puzzling’ and ‘powering’. Meijer urges us to reconsider the imaginaries used to think through potential designs and implications of smart cities. A striking example is that smart city imaginaries often prominently feature high-tech buildings while people are rarely shown, if at all.

Meijer’s notion of the ‘datapolis’ is meant as an alternative imaginary for thinking about the repercussions of smart cities. Its basic structure consists of the state, civil society, and the market. Each of these components is important, but should not become dominant in order to avoid issues. A dominant state could establish a panoptic control society, a dominant civil society could establish forms of civic engagement that feature exclusion, and a dominant market may lead to ‘selling out’.

What appears most fruitful about the notion of the datapolis is that it acknowledges the tensions involved with smart cities and refuses an easy solution. Rather than finding the best design in some kind of revised utilitarian process of weighing pros and cons, the datapolis is replete with conflict that we cannot assume to solve in any straightforward manner due to the interests involved. The datapolis features three challenges in which the state, civil society, and the market each have a role of importance. The inclusion challenge, is that cities could benefit from being an inclusive city, e.g. by enabling civic engagement by as many social groups as possible, but that too much focus on inclusion will result in missed opportunities, e.g. when decision-making is time-consuming due to the previously mentioned inclusion of social groups. The multi-level challenge recognizes that synergy between local initiatives is important, but also acknowledges that too much emphasis on synergy kills the power of local innovation, e.g. when innovations benefit from relative seclusion. Finally, the information challenge argues that good data are valuable, but also that these data may need to be ignored when confronted with contradictory experiential data.

Towards a new methodology for studying smart cities

Meijer’s talk, though briefly discussed here, illustrated a landscape of dilemmas in which contradictions between ‘right’ and ‘wrong’ are not easily made. Such ambiguities deserve recognition with regard to Kitchin’s presentation, which argued positives need to be embraced and negatives need to be avoided as possible. The problem is that there are perhaps multiple ‘goods’ and ‘bads’ in the pluralistic landscape of smart cities, where partially commensurable values and perspectives compete for attention. It is not self-explanatory that some of the ‘goods’ attributed to smart cities are beneficial to their inhabitants. Yes, current imaginaries of smart cities are value-laden, but that applies to imaginaries that contest these established imaginaries as well. For example, enhanced mobility of the workforce is sometimes presented as a ‘good’, involving independent and young entrepreneurial spirits that roam about the city, congregate using social media, and come up with innovative products, all while working on a laptop and sipping on a soy latte. Less often mentioned is that this workforce is effectively becoming a vulnerable ‘precariat’, since labor rights and job security are eroding at rapid speed.

A plea for different imaginaries that yield alternative understanding of what smart cities are and perhaps will be could help in this regard: by offering up new imaginaries (cf. figure 2), established ways to understand smart cities and their impact can be offered – but who is to say these imaginaries will not fall upon deaf ears? The sheer popularity of the term ‘smart city’ and the fact that relatively very little institutions are contesting these claims should be sufficient reason to doubt the existence of a willing audience, waiting for new imaginaries that they will receive with open arms and minds. This is not to say it is impossible for alternative imaginaries to acquire currency in, for example, the realm of policymaking. However, it does mean that science-policy interfacing needs to take place in order for policy measures to be taken up in policymaking and decision making. What is perhaps necessary is a form of institutionalized critique on smart cities that persistently counters vested interests and viewpoints. It is hard to imagine the viability of these established forms of critique without some integration in the process in policymaking through institutional renewal. In other words, a plea for a new epistemology invokes reconsiderations in the realms of institutions, politics, and normativity as well.

Figure 2. A typical smart city imaginary: high-tech, green, clean, and largely devoid of humans.

‘Smart city’ is but a hype, but one with an effect. It is a labelling term deployed in order to mobilize attention and resources. In order to truly unleash the power attributed to alternative imaginaries, research on smart cities would do well to align themselves with concrete manifestations of ‘smart’. This idea is not new, and explained in more detail and more eloquently in Shelton et al. (2015), whose work on ‘actually existing smart cities’ proposes a more grounded and empirical approach: “it is only through a grounding of our analysis in the actually existing cities, territories and relationalities where these policies are being constructed and implemented that we can understand both the promise and the peril of the smart city model.” (Ibid. p. 22)

As of late, the ‘experimental city’ is a popular manifestation of thinking about ‘smart’ forms of urban governance (see for example Evans et al. 2016). By studying how experiments are conducted in an urban setting and how they end up shaping cities and drive changes throughout society, a focus on the experimental city has the ability to cater to those wishing to implement technological innovations in order to establish novel forms of urban governance and those wishing to adopt a more ‘critical’ approach. In terms of the datapolis, ‘puzzling’ and ‘powering’ could be combined in an empirically grounded and pragmatic approach that intertwines studious examination of ‘actually existing smart cities’ with concrete interventions.

Such interventions could again fall prey to hegemonic approaches to urban governance, implying it is unsafe to assume a space where experimentation is free from external influence. The ability of disciplines like Science and Technology Studies to deliver exhaustive descriptions truly shines here, for it is such eloquent descriptions that can ward off a rekindled desire for straightforward urban renewal along the lines described above.


Evans, James, Andrew Karvonen, and Rob Raven, eds. The Experimental City. London: Routledge, 2016.

Shelton, Taylor, Matthew Zook, and Alan Wiig. “The ‘Actually Existing Smart City’.” Cambridge Journal of Regions, Economy and Society 8 (2015): 13-25.