A political ecology of data

Abstract

Conservationists, governments, and corporations see promise in digital technologies to provide holistic, rapid, and objective information to inform policy, shape investments, and monitor ecosystems. But it is increasingly clear that environmental data does more than simply offer a better view of the planet. This special issue makes a single overarching argument: that we cannot fully understand the current conjuncture in global environmental governance without understanding the platforms, devices, and institutions that comprise environmental data infrastructures. The papers draw together scholarship from political ecology and science and technology studies to demonstrate how data has become a significant site in which contemporary environmental politics are waged and socionatures are materialized. We address: (1) the contested practices of utilizing and maintaining data infrastructures; (2) the ways they are governed and the territorial statecraft they enable; (3) the socionatural materiality they arise within but also produce. The papers in this special issue show that, against its dominant representation, data is material, governed, practiced, and requires praxis. Political ecologists could adopt such an approach to make sense of the emerging ways in which data technologies shape environments and their politics.

Keywords

Political ecology digital technology data politics socionatures

Introduction

What are the implications of being able to search for environmental changes from a smartphone in the same way one might scan house prices on Zillow or adopt a puppy on Petfinder? The San Francisco-based company Planet aims to index the surface of the Earth much the same way Google has indexed the web. Planet launches mini-satellites that capture high-resolution images of the Earth’s entire surface and employs machine learning techniques to make sense of the results, representing a paradigm shift in remote sensing towards the privatization and commercialization of satellite imagery (Alvarez León and Gleason, 2017). This imagery is algorithmically labeled with features such as gold mines, deforestation, oil drilling pads, and cargo containers at port. Because its imagery is also temporally precise, Planet and its customers believe they can see real-time economic and environmental activity. Similarly, other entities such as Global Fishing Watch and the World Resources Institute’s Global Forest Watch are harnessing timely satellite imagery as well as location data to shed new light on deforestation and high seas fishing (Goldstein and Faxon, 2020; Toonen and Bush, 2019). The production of such imagery is worth questioning, not only because of the limits to inferring causality with a view from above (Geoghegan et al., 1998; Robbins, 2001; Turner, 2003), but because the drive to produce it stems from value-laden choices, such as centering the economic activities of interest to investors hedging global markets or the regulatory priorities of state agencies.

Digital technologies promise to provide more holistic, rapid, and objective information than ever before, but it is increasingly clear that the data produced and circulated through digital technologies does not simply paint a neutral, more comprehensive picture of the planet. That is because approaches to govern nature with and through the digital are inherently entangled with the governance, politics, and materialization of the digital (cf. Ash et al., 2018). In this special issue, authors’ original empirical and theoretical contributions explore the making, maintaining, and effects of what we call the ‘data infrastructures’ by which conservationists and corporations—as well as development practitioners, scientists, and state planners—generate scaled, uneven, and actionable knowledge about society and nature (Goldstein and Nost, 2022). Drawing from digital geographies and critical data studies, we define data as “captures” of people, places, and their relationships (Drucker, 2011). As such, data may be big or small, qualitative or quantitative it predates digitization. But vast datasets are now produced by digital tools, while analog data is increasingly digitized (Kitchin, 2014). This “datafication” of the world entails its capturing into forms that can be capitalized upon, circulated widely and rapidly, and used as evidence. Such captures are not unmediated snapshots, because how they are “conceived, measured, and employed actively frames their nature … .[they] are situated, contingent, relational … and are used contextually to try and achieve certain aims and goals.” (Kitchin and Lauriault, 2018: 5) Data is commonly thought of as facts and mere substrate for analytical information and contextualized knowledge (Büscher, 2020; Kitchin and Lauriault, 2018). While this hierarchy can be troubled, our contemporary moment is one of rhetorical and practical emphases on data “itself.”

Data-driven environmental governance involves a turn to data as a neutral, objective resource for accountable and transparent decision-making around nature. Confronted with the limitations of existing “knowledge infrastructures” on climate change and conservation (Edwards, 2010), data technologies ranging from drones (Cantrell et al., 2017) to dashboards (Kitchin et al., 2015) are thought to give more direct access to “raw” information and make “invisible problems not only visible but solvable” because the results of management can be more precisely and rapidly measured (Bakker and Ritts, 2018; Krupp, 2018). Conservationists, corporations, and governments—the usual suspects in environmental governance—now find themselves integrating data in portals and platforms alongside newcomers to the scene such as tech giants such as Google, IBM, and Microsoft. For some, data scarcity and irregularity constrain environmental solutions more than political will and institutional capacity or legitimacy. For instance, Fred Krupp (2018), president of the Environmental Defense Fund, argues that we have reached “fourth wave environmentalism” where the vast quantities of big data produced everyday supersede the need for what are increasingly unaccountable and unresponsive governments. For Krupp, innovation, represented by new air pollution sensors or blockchain records, promises new kinds of productive partnerships between environmentalists and corporations. Meanwhile, states are building out new software platforms and altering existing data repositories that seek to extend control over land, populations, and natural resources (Boucquey et al., 2019; Lin, 2020). Businesses and other non-state actors are developing data products that direct investments in renewable energy (McCarthy and Thatcher, 2019), commodify nature conservation (Büscher, 2020; Büscher et al., 2017; Dempsey, 2016), and support sustainability initiatives (Freidberg, 2017).

Data is not free-floating in the world, patiently waiting for these conservationists, governments, and corporations to come along and plug it into environmental management. Instead, infrastructures—place and time-specific networks of funding, standards, rules, technologies, and environments—structure data as capture, its organization, analysis, and dissemination, and, ultimately, its use in governing people and nature. Data infrastructures themselves are governed, repaired, and reproduced by specific actors in specific places; they are socio-material. Despite the rhetoric of ethereal “cloud” data infrastructures (see Amoore, 2020; Furlong, 2021; Mattern, 2016), data-driven environmental governance does not unfold in abstract space (Bauch, 2015). For instance, many conservation NGOs and tech start-ups alike require data storage and analytical capacities beyond their own, so they turn to energy-intensive server farms in data centers. These globally distributed facilities rely on access to cheap water and energy (Lally et al., 2019) and are networked by international undersea cables (Starosielski, 2015). The cables are mostly suspended in the ocean but must surface on coastlines in order to link to mainland grids, where because of climate change shores are eroding and destabilizing interconnection stations and other parts of internet infrastructure. In a cruel irony, sea level rise and coastal erosion are due at least in part from the greenhouse gases emitted in the use of fossil fuel-powered data technologies themselves (Durairajan et al., 2018). And just as oceanic cables undergird the proliferation of data portals and platforms, rare-earth metals and other minerals such as lithium comprise much of the digital realm’s hardware, including the batteries that enable the mobility of our phones. The world’s voracious appetite for lithium batteries, to power devices that store, transmit, and display data is contributing to longstanding worries over resource scarcity: of scarce water in the Bolivian salt flats mined for lithium (Aronoff, 2019), of violence in Congolese cobalt mines (Ochab, 2020). The extraction of digital-enabling materials has further engendered a variety of socio-environmental consequences around the world, as the residents of one Tibetan village attested to when they protested a lithium mine by dumping fish killed by mine-polluted water (Katwala, 2018).

Not only is data-driven environmental governance produced through material environments, it is practiced by people whose labor is organized by political economic structures and who are subjected to socio-cultural institutions and norms. Data infrastructures necessitate the vital work of maintenance, which includes managing “cybersecurity” issues. For instance, in 2016, hackers broke into the software that a flood control agency outside New Orleans, Louisiana uses to operate flood gates and pumping stations (Marshall, 2016). While staff were able to quickly resolve the hack, for many local residents, it raised the possibility that the multi-billion dollar flood infrastructure that keeps the city dry might be susceptible to risks of digital failure, compounding the more visible political and economic sources of vulnerability in New Orleans. Similar cybersecurity breaches have affected small-scale municipal infrastructures across the US and echo global incidents where the data infrastructures behind the circulation of resource commodities, including those of Maersk, the world’s largest maritime shipping firm, and Colonial Pipeline, owner of the US’s third-largest pipeline carrying gasoline and jet fuel from Texas to New York, have been targeted in cyberattacks (Greenberg, 2018; Sanger et al., 2021).

Towards a political ecology of data

How can key concepts and tools across political ecology and science and technology studies (STS) help us make sense of the ways digital data technologies intersect with environmental politics? What, in turn, might need reworking? There are four key themes in political ecology that we see as applicable to understanding environmental data infrastructures. First, political economy: political ecologists know that environmental change and adaptation is not a matter of individual choice much less some abstract, homogenized notion of “culture,” but is mediated by socially-differentiated access to and control over resources. Political economy refers to the power-laden choices about what gets produced and distributed, protected and extracted, and by whom. Thus, in a world centered on making all sorts of decisions with data, it is worth asking who makes and controls data. How are data-centered platforms like Global Forest Watch and Planet mobilized not just towards better understanding of deforestation but for different forms of cultural gain, power gain, and profit for specific actors? (cf. Büscher, 2020) What does it mean when the US state of Louisiana turns to the Rand Corporation to help it process data to inform its climate adaptation decisions? Or when North American farmers are asked to pay for Monsanto's Climate FieldView app to see their yield data and get recommendations on what to plant?

Second, political ecologists know that the dynamics by which expertise is funded, from citizen science projects to patent-seeking neoliberalized universities and outsourced state science (Lave, 2015), shape the production of knowledge. While political ecologists have often been skeptical of elite forms of environmental expertise—contra local and indigenous knowledge—in many contexts, “hybrid” forms of knowledge application are in play where land managers draw on both their own understanding of the land and are informed by data technologies as they make decisions. This may be so when, for instance, West African farmers use smartphone apps to guide when and what to plant (Abdulai et al., 2019) or when North American farmers rely on John Deere’s data platform to recommend appropriate fertilizer applications (Miles, 2019). A turn towards more holistic and precise data as the currency of environmental expertise is both facilitated by and bolsters the standing of data technology companies from Google to Amazon to Facebook, not to mention traditional actors in environmental governance who are becoming data brokers, such as John Deere and Walmart (Friedberg, 2017). These companies and their respective platforms now mediate access to much of the world’s data about itself; the most profitable part of Amazon is its data center division (Novet, 2020). This raises concerns about access when these providers are enmeshed in varying for-profit (e.g. John Deere), state or civil society-led, non-profit (e.g. Global Forest Watch), or colonial knowledge regimes (Fairbairn and Kish, 2022; Fraser, 2018). With a turn to “data science” at the expense of existing formalized knowledge regimes, it is more important than ever to build on this scholarship and look more closely at the making of data and the work it performs. While data as the “raw facts” is often presumed to lead to better information and knowledge, the growth of platforms driven less by truth and more by advertising revenues and grants enabled by user activity suggests otherwise. A political ecology of data is, perhaps, not far from “a political ecology of truth” (Büscher, 2020).

Third, materiality is central to how political ecologists understand the relationships between the natures and societies in their research. Here we make two interrelated points. First, political ecologists know to go beyond simply calculating environmental impact in terms of tons of carbon dioxide released or acres of land degraded, instead seeing impact in relational terms—as a moment in a process (e.g. the hydro-social cycle; Linton and Budds, 2014). This should be true in understanding data’s environmental and climate “footprint” as well. Critical responses to the idea that data is immaterial have taken several shapes, including enumerating the carbon, energy, and water footprints of the data centers that serve Facebook, Netflix, Microsoft, and so on (e.g. Mytton, 2021; Strubell et al., 2019). This is vital work. But we also need to further understand the many ways data is imbricated in socio-natures and their governance (see also Hogan, 2015; Lally et al., 2019; Levenda and Grabowski, 2022; Levenda and Mahmoudi, 2019; Pasek, 2019; Pickren, 2016). Second, while refuting any suggestion that environments determine society, political ecologists recognize that the non-human world—be it water or lawns—comes with properties that mold how it is extracted, commodified, or otherwise made use of (Bakker, 2005; Robbins, 2007). This manifests geographically, such as in the siting of data server farms in locations where water to cool computers is abundant. A political ecology of data likewise situates data infrastructures materially in terms of where, and from what, they are derived as well as what, and whose, natures they imprint.

Fourth, context—inclusive of history, place, and scale—matters, in contrast to the universality and novelty often associated with data technologies. Political ecology as a field has always been place centric. In order to counter the apolitical, neo-Malthusian discourses explaining environmental degradation in a region, often made in reference to some other place or solely as part of a global problem, researchers needed to deeply understand the region and its history (Hecht, 1985; Peluso, 1994). When it comes to data-driven environmental governance, a focus on history means wrestling with the layers of knowledge regimes that have sedimented over time in any place and charting what difference, if any, data makes. As Shannon Mattern (2016) has declared, “If ours is the Information Age, it is not the first.” Such skepticism of the novelty of data technologies is immensely valuable (see also Barnes, 2013). At the same time, political ecologists’ is not a narrow, insular focus on place. “Chains of explanation” start with environmental change or knowledge production in one place and contextualizes them alongside broader regional, national, and international patterns (Robbins and Monroe Bishop, 2008). Political ecologists’ emphasis on context may seem at odds with our digital world—seemingly placeless, revolutionary, and without history—but we can show that this is indeed not the case.

A political ecology of data thus asks of something like Planet: who funds this? Who has access to its tools? Who benefits from it? It poses questions concerned with materiality: what routes does Planet’s data take to get to its end users? What servers is it stored on and how are these imbricated in particular socio-environments? It questions the context of Planet’s data production or application: starting from a site where Planet’s imagery has tracked economic activity or informed investment decisions or where its data is stored and processed, a researcher might ask, how did the change here occur? Questions about data politics and the social construction of Planet’s knowledge regime abound: why are Planet and other remote sensing-based efforts focused on the global scale? Are its estimates of economic change accurate? Even if not, how are they productive? How do they perform competence and promise to Planet’s investors? Finally, political ecologists might seek to work with this data itself; though Planet’s profit may be derived from satellite imagery’s utility to market investors, in what ways is it also valuable to civil society groups and educators tracking the environmental change caused by mining, deforestation, and so on? With a lens fashioned from political economy, attention to materiality and knowledge politics, a focus on historical and geographical context, and a desire to be relevant, we can see that environmental data is—against widespread beliefs in it as immaterial and as a resource ready and waiting to drive decision-making—material, practiced, governed, and demanding of our praxis. Our aim is not to circumscribe a political ecology of data—to say it is this and not that—but to suggest its conceptual and methodological starting points.

Data is …

The papers in this special issue examine a wide range of technologies that generate, process, and disseminate data, from GPS data loggers (Blair, 2019) and satellite sensors (Vurdubakis and Rajão, 2020) to server farms (Lally et al., 2019), blockchain records (Ritts and Bakker, 2019), modeling software (Nost, 2020), and interactive maps (Goldstein and Faxon, 2020). Many of the papers start from common entry points into environmental conservation (is it effective and who benefits?) and deepen this by examining how digital systems intertwine with resource access and control (Blair; Vurdubakis and Rajão; Goldstein and Faxon). In line with the proliferation of STS work in political ecology over the past two decades that has furthered our understanding of the social construction of environmental science (Forsyth, 2004; Goldman et al., 2010), our authors bring in a wide range of STS scholarship beyond those commonly found at the intersection of political ecology and STS, such as boundary objects and centers of calculation, by incorporating concepts such as infrastructuring (Star and Ruhleder, 1994) and technological zones (Barry, 2006).

Political ecologists take care to refute grandiose apolitical ideas of environmental change and solutions. Data technologies ranging from big data to algorithmic machine learning are said to provide the precision, holism, speed, transparency, and objectivity that will help us save the planet (Gilpin, 2014; Herweijer, 2018). In this, data is considered immaterial, a resource waiting to be found (the new “oil”), disruptive, and already sufficient to solve society’s environmental challenges. But when we look closely at any data infrastructure—from web platforms displaying forest cover in community territories (Padilla Calderon and Vásquez Ruiz, 2019), to server farms in Chicago’s southside (Pickren, 2016)—it is hard to avoid noticing data infrastructures’ material imbrication with nature; the contested practices of making and maintaining data infrastructure; and the governance of data—the rules and norms around which data is accessed, used, and territorialized. As such, the special issue’s contributing authors emphasize the process by which data infrastructures come into being, rather than taking them, their context, and their meaning for granted.

… Material

The ubiquity of concepts like “cloud” computing and “virtual” environments run against political ecological commonsense (Kinsley, 2014; McLean, 2019). As McLean (2020) notes, the distinction often made between the material and the virtual is problematic: “We might think of the digital as intangible or immaterial and sideline the environmental impacts that our digital lives have if they are deemed to be less than real.” Using the US Pacific Northwest as a case study, Lally et al. provide a unique perspective in the context of the special issue, discussing the materiality of computation. Going beyond an accounting of the environmental impacts of the cryptocurrency Bitcoin (cf. Joshi, 2021), they highlight the historical, economic, and ecological conditions in which data center infrastructures come to be embedded, and the contemporary local politics that arise around them. While Bitcoin may be a special case of data infrastructure given the intensity to which server farms are put to use to compute or “mine” coins rather than store or distribute data, it nonetheless shows us how data centers mediate information flows with material environmental and space-making effects. In other words, data is a moment in socio-ecological metabolism (Pickren, 2016), one that should be understood, as Lally et al. argue, in political economic rather than functionalist terms: where and how data flows are routed is ultimately a product of the ways in which ecologies are mobilized.

… Practiced

Here, the papers seek to overcome the idea that data is just “out there” and ready to be applied anywhere. Part of the problem with “data technologies will help us save the planet” discourses is that they rarely consider how data will actually be developed, deployed, maintained, or otherwise circulated (or only in the most technical of senses). Instead, we see data as made available and at-hand through practice (Gabrys, 2016; Mattern, 2018). After all, there is no such thing as raw data, as Lisa Gitelman (2013) and colleagues remind us. We explore how data is instead always “cooked” as it is brought together and circulated in infrastructures. If data is to drive, what work does it take to put it in the driver’s seat?

Data practices we elaborate on in this issue include forms of abstraction that political ecologists will be familiar with because of their role in crafting legibility for state and capital. For instance, Blair illustrates how scientists collect and analyze data for an environmental impact statement (EIS) on proposed oil extraction in South America’s Falkland Islands (Malvinas) using several techniques and technologies for sensing (such as GPS trackers) that are circulated in models that gloss over localized impacts and whose uncertainties get enrolled into simplified categories. He also highlights practices of contextualization and maintenance, mentioning the importance and challenge of data cleaning to ensure accurate counts of penguins for the EIS. Nost shows how existing data on coastal Louisiana has to be made available to decision-makers through various techniques—such as maintaining and curating those pre-made data sets so that they can be incorporated into models or adjusting off-the-shelf model parameters for local conditions—that are always practiced in reference to political economic factors. These practices of working with existing data stand against the discourse of novelty and the hype around new data collection methods. Regardless, Nost illustrates how decisions about how much data and which data to infrastructure are also decisions about how much time and money to invest and are therefore consequential for the politics of climate adaptation. Vurdubakis and Rajão, meanwhile, illustrate the sedimentary dimension of data infrastructure by outlining the history of knowledge systems behind detecting Amazonian deforestation. From the desire of Brazil’s military to settle and secure the region, to more recent NGO-influenced sustainability efforts, they show how changing data practices build on each other over time. For instance, while the collection of deforestation evidence was initially a localized affair involving local agency officials with relatively little training, the centering of satellite imagery for accountability and anti-corruption purposes has, at least in name, led to more data-driven enforcement practices (i.e. the images point foresters to investigate certain areas). But Vurdubakis and Rajão caution against the “enchantment” of data infrastructures that puts blind faith in data artefacts without recognizing the practices that produce and circulate them. Instead, they illustrate the “street-level” practices of forestry agents who must nonetheless contextualize and interpret the data and do so in ways that make sense in their bureaucracy (see also Ghosh, 2019).

… Governed

Beyond how data is practiced—that is, made relevant and resourceful for end users—data is fundamentally enmeshed in contexts of statecraft and governance. Data is governed in the sense that rules and norms circumscribe access to and use of it (Kitchin and Lauriault, 2018; Scassa, 2019). So too data is governed in the sense that it is a product of social, political, cultural, and economic circumstance (boyd and Crawford, 2012; Dalton and Thatcher, 2014). In these ways, data is not as much “disruptive” as it is mediating of existing patterns and trends. As states establish infrastructures for managing environmental data, they build on older systems and reconfigure—if not drastically alter—relations between states and other actors in the process (Ascui et al., 2018).

Contributing authors complicate the idea that data—as something set apart from politics, economics, and, especially, emotion (cf. Verma et al., 2017)—drives and disrupts. In the eyes of many thought leaders (though not necessarily data practitioners themselves), data appears out of nowhere and spurs new forms of partnership and action on longstanding challenges. For instance, as noted earlier, Fred Krupp of the EDF suggests emerging kinds of data-generating sensor technologies may, in the context of partisan politics and government gridlock, spur new types of multilateral agreements between civil society and corporations to manage so-called “stealth” releases of methane from oil and gas production facilities. But to what extent is it that data drives and disrupts, how does it do so and what does it disrupt?

The case studies in this special issue illustrate how data reinforces rather than disrupts the practice of statecraft, enrolled as it often is to evaluate the quality and quantity of territorial resources. In this, data is not only a product of but enters into governance regimes that are inseparable from historical and place-based context. Goldstein and Faxon situate data-driven forest conservation alongside the history of surveillance in a phase of recent attempted democracy and ongoing militarization of Myanmar. Goldstein and Faxon review online apps, such as Global Forest Watch, which international donors fund with the aim of increase transparency in monitoring deforestation and thus good environmental governance. Actors’ understanding of how transparency, very much a thrust of contemporary environmental governance, contributes to improved forest conservation outcomes depends on their historical relations with the militarized state, with many domestic civil society organizations opting out of sophisticated data platforms in favor of offline qualitative data out of fear of ongoing state surveillance. In Brazil, Vurdabakis and Rajão not only note the 50-year history of knowledge infrastructures for detecting Amazonian deforestation, but the contradictions of recent moves to set up a voluntary land ownership database for enforcing forest protections. This involved asking landowners to register their property, but few did because this would expose them to penalties if deforestation was detected on their land. Satellite data may be able to show deforestation, but it actually does little without other governance measures, like requiring property registration in the case of the Amazon or actual shifts away from military control of resources in the case of Myanmar. Data does not drive on its own as much as it is something all actors—including landowners accused of deforestation as well as forestry officials—point to alongside other evidence. What it disrupts is less the outcomes of forest conservation and more its internal workings: what kinds of staff are hired, the reports they produce, and their on-the-ground operations.

In the Falkland Islands (Malvinas), Blair shows how the historical geopolitics of territory and sovereignty inform environmental data regimes. Penguin tracking data, among others, reinforces certain sovereignty claims in the context of the EIS. It is impossible to understand the work of data without acknowledging this context, Blair argues. Blair also discusses the governance context in which this data is profitable—the local science center has positioned itself to produce rather than review these EIS because there is more money to extract from oil companies in this way. Ritts and Bakker pick up the thread that the sociality of everyday life and governance is not something we need to isolate technics from. They cautiously engage with novel forms of environmental governance centered around interactive and playful data technologies and methods including hackathons and data-centric art experiments. They call this emerging modality the “Anthropocene Festival” given a focus on the question of unprecedented human modification of earth system processes. Ritts and Bakker provide us with an overview of dozens of instantiations of this Festival—in art exhibits, university panels, and even commercial events—and focus on two: first, efforts to make climate change data sensible, especially through auditory means; and second, a practical art experiment proposing a blockchain-based infrastructure by which forests could own and manage themselves. The Anthropocene Festival has both “progressive and regressive” elements and “fetishism” towards data should be avoided, Ritts and Bakker argue. In doing so, they open up the question of how digital technologies might be practiced to support forms of sociality that rework boundaries between the human and more-than-human.

The special issue’s papers distill one key normative assumption behind data driving—that what is needed to ensure environmental protection of forests or penguins or anything else is transparency, so that data can be mobilized as evidence for all to see and act upon (Goldstein and Faxon; Blair; Vurdubakis and Rajão). They question the practical value of transparent data in environmental governance. Indeed, Blair discusses this as a performance, one that, rather than disrupting anything, in fact permits the continued extraction of fossil fuels by appeasing audiences through airs of rigor and engagement with uncertainty, all while leaving blind spots (potential localized and more widespread impacts from oil spills). Similarly, the papers question the assumption that states are actually seeking more and more actionable data about the world, especially in a so-called post-truth context. Recently, political ecologists have rethought the politics and limits of legibility (Ghosh, 2019; Lin, 2020; Martin, 2019). In this vein, Vurdubakis and Rajão discuss data-driven “un-knowledge”—looking but not seeing—and what they call the “irony of transparency.” None of this negates the fact that the state has traditionally relied on abstractions in the form of data to make subjects and resources legible; indeed, it only contributes to such understanding. Data may not disruptively illuminate for all as much as it selectively spotlights for the state.

Horizons

What do we see as future research trajectories beyond this special issue? A political ecology of data has many possible horizons. For one, we see a need to strengthen connections between political ecology and digital geographies. Most of the authors in this special issue ground their work in political ecology and STS, making sense of data technologies through and for these existing fields. But there are many fruitful points of contact to be made with urban geography and critical GIS colleagues working under the umbrella of digital geographies, and to learn from their conceptual approaches to the spaces, subjects, and economies of data (Ash et al., 2018; Thatcher et al., 2018). Likewise, we hope a political ecology of data could illustrate why digital geography researchers should pay attention to environmental politics (Machen and Nost, 2021).

One such point of contact is further attention to the materiality of data technologies. Not only have digital geographers already wrestled with the stubborn materiality of data centers (Kinsley, 2014), they explain how the digital mediates spatiality more generally (Leszczynski, 2018). How are farm, forest, ocean, and other spaces being reconfigured around data production (Gabrys, 2020) and platform economies (Wang 2020)? Answering this may involve figuring out what methodologies, particularly collaborative ones, might draw together political ecology’s emphasis on site-specific ethnography and digital geography’s approach to data as an object of inquiry (cf. Duggan, 2017; Leszczynski, 2017). Another important point of contact is the political economy of data collection and analysis. Given the configuration of environmental governance and politics today, data technologies are often driven by transnational NGOs, local civil society organizations, and governments in partnership with tech start-ups and tech giants alike. Digital geographers often emphasize the extraction and accumulation that motivates tech companies (e.g. Thatcher et al., 2016), but what exactly do they have to gain from these partnerships? Finally, digital geographers have imagined speculative technological presents and futures, notably in the recent volume, How to Run a City like Amazon (Graham et al., 2019). Likewise, political ecologists have both evaluated visions of socio-ecological futures and crafted them themselves (Martin and Sneegas, 2020). There may be some productive middle ground here. To return to the question we began the paper with, what would it mean to index and search nature the way you look up dinner recipes through Google? What would it mean to manage wetlands the way Amazon runs retail and distribution? What does an “Airbnb for the environment” do? (Jayachandran, 2017).

On the horizons of a political ecology of data is also the opportunity to imagine collaborative work between political ecologists and tech practitioners. If political ecology is a “trickster science” that embraces yet critiques its objects of inquiry (Robbins, 2015), then we might pursue research where political ecologists pair with software developers and study climate change, deforestation, environmental justice, and so on within the framework of “data science.” This might involve work to “uncover” previously hidden patterns in social datasets and ecological ones, repurposing these to demonstrate inequalities and injustices, and making them “open” for public consumption, while questioning not just how the data were cooked but the very assumption that more data and information translates into awareness and action (e.g. Walker et al., 2018). Critical physical geography provides us with a model here, as it has created space for political ecologists and physical geographers alike to productively explore how the concepts, objects, and practices of environmental science are socially constructed and could be done differently while telling us better truths about the world (Lave et al., 2014). Indeed, calls to bring together political ecology and data analytics are not necessarily new and, in many ways, extend ongoing engagements between political ecology and sustainability science (Goldstein et al., 2020; Kelley, 2018), not to mention critical GIS and cartography (Foo, 2019; Weiner et al., 1995). But such partnerships ought to be renewed in the context of, for instance, modeling climate futures and taken in new directions (Colven and Thomson, 2018).

Finally, while several papers in this issue (Goldstein and Faxon; Vurdabakis and Rajão; Blair) offer ethnographic insight into case studies from the Global South, research on data and the digital realm still skews heavily towards the Global North. While this may reflect the technology sector’s past and ongoing development of digital technologies in and for the world’s wealthiest populations, that is rapidly changing as countries including India, Indonesia, and, of course, China build software intended for users within their own and neighboring countries. Without intending to reify such North-South divisions, there is nevertheless much room for new empirical work on a political ecology of data from and in the Global South in places that have traditionally been the purview of political ecology. A focus on data and data infrastructures within and across these contexts may in fact illuminate new ways of seeing ‘development’. We live in a moment where data technologies are simultaneously centered as solutions to global health, economic, and climate crises, downplayed when politically inconvenient, and sites of controversy themselves (Everts, 2020; Shelton, 2020). Yet, they also offer, in complicated ways, an opportunity to create more equitable and livable futures for many people and the environments they depend on. To confront all of this and more, and to chart our way out of these crises, we need a political ecology of data.

Highlights

The digital realm has become a significant site in which contemporary environmental politics are waged: governing nature with and through the digital have become entangled with the governance, politics, and materialization of the digital.

Papers in this special issue show that, against its dominant representation, data is material, governed, practiced, and requires praxis.

Four themes in political ecology are applicable to understanding environmental data infrastructures: political economy, the production of knowledge, materiality, and contexts of history, place, and scale.

This article and special issue points a conceptual way forward for further work in a political ecology of data.

Footnotes

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

References

Abdulai

A-R

Duncan

Fraser

(2019) How digital technologies can help Africa’s smallholder farmers. Available at: http://theconversation.com/how-digital-technologies-can-help-africas-smallholder-farmers-119952 (accessed 14 April 2021).

Alvarez León

Gleason

(2017) Production, property, and the construction of remotely sensed data. Annals of the American Association of Geographers 107: 1–15.

Amoore

(2020) Cloud Ethics: Algorithms and the Attributes of Ourselves and Others. Durham, NC: Duke University Press.

Aronoff

(2019) Climate Change’s Great Lithium Problem. The New Republic, 16 November. Available at: https://newrepublic.com/article/155753/climate-changes-great-lithium-problem (accessed 14 April 2021).

Ascui

Haward

Lovell

(2018) Salmon, sensors, and translation: The agency of big data in environmental governance. Environment and Planning D: Society and Space 36(5): 905–925. DOI: 10.1177/0263775818766892.

Ash

Kitchin

Leszczynski

(2018) Digital turn, digital geographies? Progress in Human Geography 42(1): 25–43.

Bakker

(2005) Neoliberalizing nature? Market environmentalism in water supply in England and Wales. Annals of the Association of American Geographers 95(3): 542–565.

Bakker

Ritts

(2018) Smart earth: A meta-review and implications for environmental governance. Global Environmental Change 52: 201–211.

Barnes

(2013) Big data, little history. Dialogues in Human Geography 3(3): 297–302.

10.

Barry

(2006) Technological zones. European Journal of Social Theory 9(2): 239–253.

11.

Bauch

(2015) Bodies in the cloud: A geography of electronic health data. In: Yoo

Blanchette

(eds) Regulating the Cloud: Policies for Computing Infrastructure. Cambridge, MA: MIT Press, pp.258–278.

12.

Blair

JJA

(2019) Tracking penguins, sensing petroleum: “Data gaps” and the politics of marine ecology in the South Atlantic. Environment and Planning E: Nature and Space: DOI: 10.1177/2514848619882938.

13.

Boucquey

Martin

Fairbanks

, et al. (2019) Ocean data portals: Performing a new infrastructure for ocean governance. Environment and Planning D: Society and Space 37(3): 484–503.

14.

boyd D and Crawford

(2012) Critical questions for big data: Provocations for a cultural, technological, and scholarly phenomenon. Information, Communication & Society 15(5): 662–679.

15.

Büscher

(2020) The Truth about Nature: Environmentalism in the Era of Post-Truth Politics and Platform Capitalism. Oakland, CA: University of California Press.

16.

Büscher

Koot

Nelson

(2017) Introduction. Nature 2.0: New media, online activism and the cyberpolitics of environmental conservation. Geoforum 79: 111–113.

17.

Cantrell

Martin

Ellis

(2017) Designing autonomy: Opportunities for new wildness in the anthropocene. Trends in Ecology & Evolution 32(3): 156–166.

18.

Colven

Thomson

(2018) Bridging the divide between human and physical geography: Potential avenues for collaborative research on climate modeling. Geography Compass: 13(2): e12418.

19.

Dalton

Thatcher

(2014) What does a critical data studies look like, and why do we care? Society & Space. Available at: www.societyandspace.org/articles/what-does-a-critical-data-studies-look-like-and-why-do-we-care (accessed 26 December 2017).

20.

Dempsey J (2016) Enterprising Nature: Economics, Markets, and Finance in Global Biodiversity Politics. John Wiley & Sons.

21.

Drucker

(2011) Humanities approaches to graphical display. Digital Humanities Quarterly 5(1): 20.

22.

Duggan

(2017) Questioning “digital ethnography” in an era of ubiquitous computing. Geography Compass 11(5): e12313.

23.

Durairajan

Barford

(2018) Lights Out: Climate Change Risk to Internet Infrastructure. New York: ACM Press, 2018, pp.9–15.

24.

Edwards

(2010) A Vast Machine: Computer Models, Climate Data, and the Politics of Global Warming. Cambridge, MA: MIT Press.

25.

Everts

(2020) The dashboard pandemic. Dialogues in Human Geography 10(2): 260–264.

26.

Fairbairn M, Kish Z (2022) “A poverty of data”? Exporting the digital revolution to farmers in the Global South. In The Nature of Data: Infrastrutures, Environments, Politics (Goldstein J and Nost E, eds) University of Nebraska Press.

27.

Foo

(2019) Visualizing undisciplined environments: An introduction. Geoforum 102: 230–234.

28.

Forsyth

(2004) Critical Political Ecology: The Politics of Environmental Science. Milton Park: Routledge.

29.

Fraser

(2018) Land grab/data grab: Precision agriculture and its new horizons. The Journal of Peasant Studies: 46(5): 893–912.

30.

Freidberg

(2017) Big food and little data: The slow harvest of corporate food supply chain sustainability initiatives. Annals of the American Association of Geographers: 107(6): 1389–1406.

31.

Furlong

(2021) Geographies of infrastructure II: Concrete, cloud and layered (in)visibilities. Progress in Human Geography 45(1): 190–198.

32.

Gabrys

(2020) Smart forests and data practices: From the Internet of Trees to planetary governance. Big Data & Society 7(1): 205395172090487.

33.

Gabrys

(2016) Practicing, materialising and contesting environmental data. Big Data & Society 3(2): 1–7.

34.

Geoghegan

Pritchard

Ogneva-Himmelberger

, et al. (1998) ‘Socializing the pixel’ and ‘pixelizing the social’ in land-use and land-cover change” in people and pixels: Linking remote sensing and social science. In: Liverman

Moran

Rindfuss

, et al. (eds) Committee on the Human Dimensions of Global Environmental Change, National Research Council. Washington, DC: National Academy of Science Press, pp.51–69.

35.

Ghosh

(2019) Appetite for imprecision: The role of bureaucracy in implementing a pay-for-performance program. Annals of the American Association of Geographers 109(4): 1208–1225.

36.

Gilpin

(2014) 10 big data projects that could help save the planet. Available at: www.techrepublic.com/article/10-big-data-projects-that-could-help-save-the-planet/ (accessed 2 March 2018).

37.

Gitelman

(2013) Raw Data Is an Oxymoron. Cambridge, MA: MIT Press.

38.

Goldman

Nadasdy

Turner

(2010) Knowing Nature: Conversations at the intersection of political ecology and science studies. Chicago, IL: University of Chicago Press.

39.

Goldstein

Faxon

(2020) New data infrastructures for environmental monitoring in Myanmar: Is digital transparency good for governance? Environment and Planning E: Nature and Space: DOI: 10.1177/2514848620943892.

40.

Goldstein

Graham

Ansori

, et al. (2020) Beyond slash‐and‐burn: The roles of human activities, altered hydrology and fuels in peat fires in Central Kalimantan, Indonesia. Singapore Journal of Tropical Geography 41(2): 190–208.

41.

Goldstein

Nost

(2022) The Nature of Data: Infrastructures, Environments, Politics. Lincoln: University of Nebraska Press.

42.

Graham

Kitchin

Mattern

, et al. (2019) How to Run a City Like Amazon, and Other Fables. London: Meatspace Press.

43.

Greenberg

(2018) The untold story of NotPetya, the most devastating cyberattack in history. Wired. Available at: www.wired.com/story/notpetya-cyberattack-ukraine-russia-code-crashed-the-world/ (accessed 14 April 2021).

44.

Hecht

(1985) Environment, development and politics: Capital accumulation and the livestock sector in eastern Amazonia. World Development 13(6): 663–684.

45.

Herweijer

(2018) 8 ways AI can help save the planet. Available at: www.weforum.org/agenda/2018/01/8-ways-ai-can-help-save-the-planet/ (accessed 10 May 2019).

46.

Hogan

(2015) Data flows and water woes: The Utah Data Center. Big Data & Society 2(2): 205395171559242.

47.

Jayachandran

(2017) Using the Airbnb model to protect the environment. The New York Times, 29 December. https://www.nytimes.com/2017/12/29/business/economy/airbnb-protect-environment.html (accessed 26 August 2021)

48.

Joshi

(2021) Bitcoin is a mouth hungry for fossil fuels. Available at: https://ketanjoshi.co/2021/03/11/bitcoin-is-a-mouth-hungry-for-fossil-fuels/ (accessed 17 March 2021).

49.

Katwala

(2018) The spiraling environmental cost of our lithium battery addiction. Wired UK, 5 August. Available at: www.wired.co.uk/article/lithium-batteries-environment-impact (accessed 14 April 2021).

50.

Kelley

(2018) The politics of uneven smallholder cacao expansion: A critical physical geography of agricultural transformation in Southeast Sulawesi, Indonesia. Geoforum 97: 22–34.

51.

Kinsley

(2014) The matter of ‘virtual’ geographies. Progress in Human Geography 38(3): 364–384.

52.

Kitchin

(2014) The Data Revolution: Big Data, Open Data, Data Infrastructures & Their Consequences. Thousand Oaks, CA: Sage Publications.

53.

Kitchin

Lauriault

(2018) Data and data infrastructure. In: Ash

Kitchin

Leszczynski

(eds) Digital Geographies. 1st ed. Thousand Oaks, CA: Sage Publications, pp. 83–94.

54.

Kitchin

Lauriault

McArdle

(2015) Knowing and governing cities through urban indicators, city benchmarking and real-time dashboards. Regional Studies, Regional Science 2(1): 6–28.

55.

Krupp

(2018) Welcome to the fourth wave: A new era of environmental progress. Available at: www.edf.org/blog/2018/03/21/welcome-fourth-wave-new-era-environmental-progress (accessed 14 November 2018).

56.

Lally

Kay

Thatcher

(2019) Computational parasites and hydropower: A political ecology of Bitcoin mining on the Columbia River. Environment and Planning E: Nature and Space: DOI: 10.1177/2514848619867608.

57.

Lave

(2015) The future of environmental expertise. Annals of the Association of American Geographers 105(2): 244–252.

58.

Lave

Wilson

Barron

, et al. (2014) Intervention: Critical physical geography. The Canadian Geographer/Le Géographe canadien 58(1): 1–10.

59.

Leszczynski

(2017) Digital methods I: Wicked tensions. Progress in Human Geography 42(3): 473–481.

60.

Leszczynski

(2018) Spatialities. In: Ash

Kitchin

Leszczynski

(eds) Digital Geographies. 1 ed. Thousand Oaks, CA: Sage Publications, pp. 13–23.

61.

Levenda A, Grabowski Z (2022) Data, colonialism, and the transformation of nature in the Pacific Northwest. In The Nature of Data: Infrastructures, Environments, Politics (Goldstein J and Nost E, eds) University of Nebraska Press.

62.

Levenda

Mahmoudi

(2019) Silicon forest and server farms: The (Urban) nature of digital capitalism in the pacific northwest. Culture Machine 18: 1–14.

63.

Lin

(2020) How to make a forest. Available at: www.e-flux.com/architecture/at-the-border/325757/how-to-make-a-forest/ (accessed 16 April 2020).

64.

Linton

Budds

(2014) The hydrosocial cycle: Defining and mobilizing a relational-dialectical approach to water. Geoforum 57: 170–180.

65.

McCarthy

Thatcher

(2019) Visualizing new political ecologies: A critical data studies analysis of the World Bank’s renewable energy resource mapping initiative. Geoforum 102: 242–254.

66.

McLean

(2019) Changing Digital Geographies: Technologies, Environments and People. 1st ed. London: Palgrave Macmillan.

67.

McLean

(2020) McLean on remaking digital geographies. Available at: www.palgrave.com/gp/blogs/social-sciences/mclean (accessed 3 March 2020).

68.

Machen

Nost

(2021) Thinking algorithmically: The making of hegemonic knowledge in climate governance. Transactions of the Institute of British Geographers 46(3): 555–569.

69.

Marshall

(2016) After flood officials ransomed their data, colleagues say protection is in place. Available at: https://thelensnola.org/2016/05/26/after-flood-officials-paid-a-ransom-to-recover-data-colleagues-say-protection-is-in-place/ (accessed 14 April 2021).

70.

Martin

(2019) Between scylla and charybdis: Environmental governance and illegibility in the American West. Geoforum 123: 194–204.

71.

Martin

Sneegas

(2020) Critical worldbuilding: Toward a geographical engagement with imagined worlds. Literary Geographies 6: 15–23.

72.

Mattern

(2018) Maintenance and care. Places Journal. https://placesjournal.org/article/maintenance-and-care (accessed 26 August 2021)

73.

Mattern

(2016) Cloud and field. Places Journal. https://placesjournal.org/article/cloud-and-field/ (accessed 26 August 2021)

74.

Miles

(2019) The combine will tell the truth: On precision agriculture and algorithmic rationality. Big Data & Society 6(1): 205395171984944.

75.

Mytton

(2021) Data centre water consumption. npj Clean Water 4(1): 11.

76.

Nost

(2020) Infrastructuring “data-driven” environmental governance in Louisiana’s coastal restoration plan. Environment and Planning E: Nature and Space: 251484862090972. DOI: 10.1177/2514848620909727.

77.

Novet

(2020) Amazon cloud revenue jumps 29%, in line with expectations. Available at: www.cnbc.com/2020/10/29/amazon-cloud-revenue-jumps-29percent-in-line-with-expectations.html (accessed 14 April 2021).

78.

Ochab

DEU

(2020) Are these tech companies complicit in human rights abuses of child cobalt miners in Congo? Available at: www.forbes.com/sites/ewelinaochab/2020/01/13/are-these-tech-companies-complicit-in-human-rights-abuses-of-child-cobalt-miners-in-congo/ (accessed 14 April 2021).

79.

Padilla Calderon

Vásquez Ruiz

(2019) Mapa Oaxaca platform puts the power of data into the hands of Mexican communities. Available at: https://blog.globalforestwatch.org/people/mapa-oaxaca-platform-puts-the-power-of-data-into-the-hands-of-mexican-communities (accessed 14 April 2021).

80.

Pasek

(2019) Managing carbon and data flows: Fungible forms of mediation in the cloud. In: Culture Machine. Available at: http://culturemachine.net/vol-18-the-nature-of-data-centers/managing-carbon-and-data-flows/ (accessed 26 March 2019).

81.

Peluso

(1994) Rich Forests, Poor People: Resource Control and Resistance in Java. Oakland, CA: University of California Press.

82.

Pickren

(2016) ‘The global assemblage of digital flow’: Critical data studies and the infrastructures of computing. Progress in Human Geography 42(2): 225–243.

83.

Ritts

Bakker

(2019) New forms: Anthropocene festivals and experimental environmental governance. Environment and Planning E: Nature and Space: DOI: 10.1177/2514848619886974.

84.

Robbins

(2001) Fixed categories in a portable landscape: The causes and consequences of land-cover categorization. Environment and Planning A: Economy and Space 33(1): 161–179.

85.

Robbins

(2007) Lawn People: How Grasses, Weeds, and Chemicals Make Us Who We Are. Philadelphia, PA: Temple University Press.

86.

Robbins

Monroe Bishop

(2008) There and back again: Epiphany, disillusionment, and rediscovery in political ecology. Geoforum 39(2): 747–755.

87.

Robbins

(2015) The Trickster science. In: Perreault

Bridge

McCarthy

(eds) The Routledge Handbook of Political Ecology. Milton Park: Routledge, pp. 89–101.

88.

Sanger

Krauss

Perlroth

(2021) Cyberattack forces a shutdown of a top US pipeline. The New York Times, 8 May. https://www.nytimes.com/2021/05/08/us/politics/cyberattack-colonial-pipeline.html (accessed 26 August 2021)

89.

Scassa

(2019) Ownership and control over publicly accessible platform data. Online Information Review 43(6): 986–1002.

90.

Shelton

(2020) A post-truth pandemic? Big Data & Society 7(2): 205395172096561.

91.

Star

Ruhleder

(1994) Steps towards an ecology of infrastructure: Complex problems in design and access for large-scale collaborative systems. In: Proceedings of the 1994 ACM conference on Computer supported cooperative work. New York: ACM, 1994, pp.253–264.

92.

Starosielski

(2015) The Undersea Network. Durham, NC: Duke University Press.

93.

Strubell

Ganesh

McCallum

(2019) Energy and policy considerations for deep learning in NLP. arXiv:1906.02243 [cs]. Available at: http://arxiv.org/abs/1906.02243 (accessed 14 June 2019).

94.

Thatcher

Bergmann

Ricker

, et al. (2016) Revisiting critical GIS. Environment and Planning A 48(5): 815–824.

95.

Thatcher

Eckert

Shears

(2018) Thinking Big Data in Geography. Lincoln: University of Nebraska Press.

96.

Thatcher

O’Sullivan

Mahmoudi

(2016) Data colonialism through accumulation by dispossession: New metaphors for daily data. Environment and Planning D: Society and Space 34(6): 990–1006.

97.

Toonen

Bush

(2018) The digital frontiers of fisheries governance: Fish attraction devices, drones and satellites. Journal of Environmental Policy & Planning 22: 1–1–13.

98.

Turner

(2003) Methodological reflections on the use of remote sensing and geographic information science in human ecological research. Human Ecology 31: 255–279.

99.

Verma

van der Wal

Fischer

(2017) New technological interventions in conservation conflicts: Countering emotions and contested knowledge. Human Ecology 45(5): 683–695.

100.

Vurdubakis

Rajão

(2020) Envisioning Amazonia: Geospatial technology, legality and the (dis)enchantments of infrastructure. Environment and Planning E: Nature and Space: DOI: 10.1177/2514848619899788.

101.

Walker D, Nost E, Lemelin A, et al. (2018) Practicing environmental data justice: From DataRescue to Data Together. Geo: Geography and Environment 5(2): e00061

102.

Wang

(2020) Blockchain Chicken Farm: And Other Stories of Tech in China’s Countryside. New York: Farrar, Straus and Giroux.

103.

Weiner

Warner

Harris

, et al. (1995) Apartheid representations in a digital landscape: GIS, remote sensing and local knowledge in Kiepersol, South Africa. Cartography and Geographic Information Systems 22(1): 30–44.