06 September, 2023

Mattiucci interviews Carr in ITEM Bookzine di arte e psicoanalisi

Earlier this year, Carr was interviewed by Prof. Cristina Mattiucci Dept of Architecture, University of Naples Federico II. The following is an English-language translation of the original conversation that Mattiucci published in ITEM Bookzine di arte e psicoanalisi N.2 - SI Artificiale - edited by Waiting Room Residency - Giusi Campisi, Sara d’Alessandro Manozzo, Luca Bertoldi - July 2023.

The urban heaviness of the digital / Politics of urban digital infrastructure

by Christina Mattiucci

The Premise: I met Connie (Constance) Carr at INURA - the International Network of Urban Research and Action - which is a network we have shared for many years.

Last year, in June 2022, at the close of the Retreat of the Annual INURA conference held in Luxembourg, she presented her research on - Digital Urban Development - How large digital corporations shape the field of urban governance (DIGI-GOV) - of which she is PI, at the Department of Geography and Spatial Planning, University of Luxembourg. The aim of DIGI-GOV is to explore the role of large digital corporations (LDCs) in digital urban development, how the presence of LDCs in urban planning practice challenge urban governance, and how LDC-led urban development constitutes a new relational geography of digital cities.

I was curious about Carr’s research because it questions dimension of digital urban transformations, and sheds light on 'the weight' of digital dimensions of urban spatial dynamics and in the context of the Urban Question.

Now, almost a year later, I come back to her to try to understand what are the main issues that emerged from that research, beyond the publications resulted from it so far.

Q: It seems to me that your work seeks to understand the on-the-ground politics of urban digital infrastructure. What are the broader questions that have guided your research and what kind of conceptualization of the digital dimension it challenged?

A: The broad aim of DIGI-GOV to examine and explain how large digital corporations such as Amazon or Google influence the development. This is the overarching goal. This research is funded by the Luxembourg National Research Fund. And I say this not only as a logo but also because people often ask me about who funds this research as they are suspicious that it might be Google, or some investor. So, as a small disclaimer, it is important in this context to mention that this is a university research project that is publicly funded and seated at the Department of Geography and Spatial Planning at the University of Luxembourg.

DIGI-GOV grew out and was inspired by a previous project, which looked at Sidewalk Labs in Toronto and what that one happened back then. (see paper[1] about Sidewalk Labs (SL) — a daughter company and urban development arm of Alphabet Inc. and sister to Google LLC— which won the competition to develop 4.9 hectares along Toronto’s shores of Lake Ontario, entering as specific and controversial actor in ordinary urban planning, ndr)

What was interesting about this so-called digital city project was that Sidewalk Labs was a new actor on the local field of urban planning and development. It wasn’t just architects and developers: It was a tech company. Of course, digital technology and urbanization have always gone hand in hand, so in one sense this is not new, but in this case we had a major tech company with enormous capital power, and with access to urban government in ways that were kind of new. This was back in 2017, 2018, and it got massive media attention, and dominated Toronto planning in the port lands until the pandemic hit. Sidewalk was claiming that it would build the most amazing digital city that was the world has ever seen and so on, but what was also remarkable was how it had all levels of Canadian government behind it, which were not only giving their public support, but also coordinating their public messages and appearances. So, we saw the CEO of Alphabet Inc. on stage with the Prime Minister of Canada and the Premier of Ontario. This is not easy to do, actually. So, obviously, there was some concentrated cooperation going on, in addition to the new digital gadgets that Sidewalk wax developing and preparing to sell.

From a research perspective, the next question was: How might this play out in other cities? And so, DIGI-GOV looks at six cities: the Washington Metropolitan Area, Seattle, Toronto, Amsterdam, Luxembourg, and Kiev. It’s a gigantic project -- and there is a fairly large team on it -- and we are currently in various stages of research in all these places.

Q: Let's talk back about DIGI-GOV. Your work also highlights "data matters" through their production/materialization/storage. The graphic you published on data centers in the Washington Metropolitan Area and respective kW needs is very significant in this perspective. It shows an interpretative map, where you show some significant implications. As you wrote, the map provides the visualization of the social spatial distribution of data centers and it points out the five implications you found: data centers are concentrated in metropolitan areas; they have a high demand for energy and water, competing with local residents for these resources; their industry is a state-led niche economy; the uneven distribution of data centers can invoke inter-county competition for tax revenue, in addition to access to the water, power, and land resources they require. In the related paper[2] you stated that ‘data centers present an under explored geography of cyberworlds. By means of that large digital corporations such as Amazon or Google are expanding their role in urban infrastructural development’. What are the main challenges of data centers for urban governance? Then, not forgetting that there are issues of visibility and secretness, what kind of data you were able to spatialize?

A: There are two main vains of research in DIGI-GOV. First, DIGI-GOV looks a symbolic places like Sidewalk Labs or the headquarters of Amazon. Second, the project addresses new kinds of telecommunications infrastructure, data centers in particular. Those are the two key foci. About the maps that we drew: We completed those at the beginning of the project because that was back in 2021 and we were all rather new to the topic of data centers. Actually, no one on the team really knew what a data center even was. Further, it was a rather under-researched infrastructure with most work limited to the domains of engineering and computer science. So, we were pursuing this very basic exploration: What is a data center? What are the varieties of kind of data centers? Where are they? What do they do? We were just exploring some basic facts about what we were dealing with. This is where we discovered, through publicly available sources, where they were, and what the basic characters of these locations were, from which we could extrapolate what this might mean or implicate in spatial terms.

We learned that it was a booming business, that their input needs (such as land) were expanding rapidly. We also found – and this was surprising at the time – that data centers were concentrating in metropolitan areas. I had gone into this thinking that data centers would be a rural phenomenon, which was not only totally wrong, it was predictable according to the urban studies literature, as telecommunications infrastructure have always concentrated in urban areas. So, if you look at publicly available maps (e.g. Baxtel.com), you'll find that the data centers are usually in big cities like Amsterdam, Frankfurt, London, Paris, Washington, Seattle. They're concentrating in the metropolitan areas.

We also noticed a certain set of institutions, carving out their economic positions. The one that really stuck out to us, of course, was the prevalence of Amazon particularly in the Washington Metropolitan Area. Of course, Amazon just does not disclose anything, actually, but they have 50 or so data centers in the WMA. We also know that they have the largest and most modern data centers with huge data input, storage and processing needs, so they must be enormous. But we can't access this in specific terms.

Q: As an exploratory work, I imagine the maps started to speak to your project. If you had to imagine to integrate your maps at the end of the project, do you think that are other elements that should be made visible or just the power of seeing the located data centers works in itself?

A: I don't know yet. On one hand, this was not supposed to be solely a story about location. But on the other hand, it is definitely interesting to think about how data centers are changing urban and regional landscapes. We did find that they are near waterways, so this is a territorial question. And, they're also in well-to-do neighborhoods (another surprise). Whether this should be ‘mapped’, per se, I don’t know. We can also illustrate with text.

Q: Going back to the challenges of visibility and secretness…

A: For us, secrecy was and remains the biggest problem. There is a lot of information out there about the massive amounts of electricity and water that data centers need. There's a lot available on industry websites about where data centers are and what they are willing to reveal about electricity consumption. There is also a lot of discussion about improving efficiencies. This is of course very important. But what we find is that we cannot really access what is behind these processes, which is also an interesting phenomenon. So, for example, there are a lot of engineers working on improving energy efficiencies, but very little about actual input needs. It's one thing to be efficient, but if your absolute input continues to grow then there's still an issue about availability of resources. So, that is an area that is not really clarified. And then of course, the issue of what data is being stored where, by what company etc., and this is all super-secret. There are of course good reasons for secrecy (e.g. security), but this also creates a situation where there is no room for public input and certainly no room for public debate. Further, it is worth mentioning that protests against data centers are becoming commonplace. So, there is a need on one hand for public conversation about these, but there is also a strong need for secrecy, which is driven by security concerns and, we cannot forget, corporate secrecy as is practiced in profit driven enterprises.

Q: It seems to me that it means looking at a kind of materialization of the data in the city. What would you say are the main challenges of this material dimension of data? And, let's think too about some of the political implication of your research questions. That is: What does this work bring out about the neoliberal directions of urban transformations?

A: There is more to explore in terms of neoliberal urbanism, and what that means when for-profit urbanism is driven by big tech that prioritize their agendas, under the veil of secrecy. This, I think is really interesting.

Q: What do you think are the "exportable" themes of your research, which can be a reference for a critical reading of the digital dimension in other urban contexts as well, where for instance processes related to resource consumption or to financialization are somehow less evident?  

A: Hard questions! <laugh> Okay, what we can learn from, which we would say as interesting? It's funny because I think maybe, maybe digital urbanization is a better term than smart cities or even digital cities, because for me urbanization implies a set of processes which then expose how cities form, produce and constitute each other. This refers to the urban theoretical concept of relationality: that cities are not atomized, particulate places, but mutually producing one another. This is a very broad field of urban studies research, which talks about urban comparison, how to conceptualize urban spaces as part of international networks of spaces and flows of many kinds. There is a lot there, and there are better urban theorists than me that discuss this. But here I can give you a simple but rather extreme example: I just got back from Washington DC where I observed that there were lots of protests about data centers. The repeating narrative was - and this is incredible if it's true - is that 70% of the internet goes through Virginia. If that is true, that's insane! Ok, because of secrecy we cannot actually verify it, but if true, it is not only extreme, it also shows how places are interconnected and involved in digitalization processes. Our (online, ndr) conversation here, the one between you and me, is going through another place, completely different, far away, in a different jurisdiction, and the spatial manifestation of both places – in this case, data center sprawl in Virginia and office development in Europe – define and shape one another. I think that this is very significant.

[1] Carr, C, Hesse, M (2020). When Alphabet Inc. Plans Toronto’s Waterfront: New Post-Political Modes of Urban Governance. Urban Planning, Volume 5, Issue 1, p. 69–83. DOI: 10.17645/up.v5i1.2519
[2] Desmond Bast, D, Carr, C, Madron, K and Syrus, AM (2022). Four reasons why data centers matter, five implications of their social spatial distribution, one graphic to visualize them. EPA: Economy and Space, p. 1–5 . DOI: 10.1177/0308518X211069139