Situating the geoweb in political economy

I Introduction

There has been a noticeable shift in the production and dissemination of geographic information since 2005, marked by Google’s launch of Maps, its Web mapping service, and Earth, its virtual globe. In July of the same year, Google made the Maps application programming interface (API) public, allowing users to ‘hook into’ the service from other websites and develop their own applications over the top of it. Google’s activities coincide with and have contributed to the development of the ‘geoweb’, the aggregate of geographically-referenced or ‘marked-up’ ¹ information that is increasingly used to organize and deliver content over the Web (Elwood and Leszczynski, 2010; Haklay et al., 2008; Scharl and Tochtermann, 2007).

Much of this locational-based data is curated by users with no formal background in geography or cartography who ‘voluntarily’ contribute spatial information over the Web. Michael Goodchild (2009) has consequently termed these contributions ‘volunteered geographic information’ (VGI). ² The ‘crowdsourcing’ of geographical information – a term that refers to the practice of harnessing the ‘power of the crowd’ to create collective resources – signifies a profound change within regimes of spatial data authoring and provision, which have historically been the preserve of national mapping agencies and cartographic experts. The geoweb represents a pronounced shift in what constitutes ‘geographic information’ at all levels of data, technologies, and practices, but the impetus for this shift is not geographical per se. Its emergence and development have been overwhelmingly driven outside of academic geography by the Web business and computing communities.

Geographers have turned their attention to these new spatial data technologies and data authoring practices, and have sought to account for their nascence. Early engagements with geoweb phenomena have emphasized the enabling technics (participatory frameworks and interactive functionalities) and pragmatics (business correctives) of ‘Web 2.0’. ‘Web 2.0’ describes the transition away from a Web of the 1990s that served as a data portal and repository to one that functions as a platform of networked services and applications (O’Reilly, 2005). Without these developments, the geoweb would certainly not exist. But geographers have yet to attend to the conditions of possibility that have made for the extension of Web 2.0 technological and commerce innovations to the realm of geospatial information. Beyond being enabled by the technical and business innovations of Web 2.0, I argue that the rise of the geoweb is implicated in contemporary political economic transitions that have allowed for the coming together of information technology and markets, a historically specific articulation of technoscience and capitalist exchange (Sunder Rajan, 2006).

These changes are evidenced, in the North American and European contexts, in the changing role of the state vis-a-vis the ‘cartographic project’ – the entirety of the enterprise of making and circulating geographic information and its representation(s) for boundary-drawing, territory-enforcement, and data provision to a citizen base. Implicated in broader processes of contemporary state restructuring, these reconfigurations are manifesting as pronounced shifts within regimes of spatial data governance – the production, dissemination, and institutionalization of geographic information. Peck and Tickell (2002) characterize these governance shifts in terms of ‘roll-back’/’roll-out’ neoliberalism, whereby as the welfare state is either being eroded or actively rescinding itself from securing certain services and goods in the public sphere (‘roll-back’), their provision is opened up to market regimes, which subsume the delivery of these goods and services to their own logics (‘roll-out’). Drawing upon selective examples, I suggest that as ‘the state’ is partially being rolled back, or actively withdrawing itself, from the public curation and dissemination of geographic information (GI), markedly non-state (market and corporate) regimes of spatial data governance are rolling out in its place. This is not to suggest that the state is no longer present where ‘the map’ is concerned, or that it is being in any way eclipsed by market-based mapping activities. Rather, it speaks to the ways in which the state’s role in relation to the mapping enterprise is altered in significant ways. As the ‘cartographic project’ is increasingly opened to markets and their logics, relationships between citizens, the state, and the market are being simultaneously transformed. This is most immediately discernible in how – and from whom – citizen subjects access and interact with geographic information.

In this paper, I strive to enrich and deepen geographers’ extant accounts of how the geoweb ‘came to be’ by historically situating the rise of the geoweb within the contemporary political economic restructuring of the state and the associated shifts in governance regimes under neoliberalism. By ‘political economy’, I mean that this ‘situating’ operates from the fundamental premise that within any economy, there is a system of (re)distribution, and this (re)distribution is always political. Bringing contemporary theories of governance to bear upon our thinking about the geoweb allows us to begin to flesh out how regimes of the production, regulation, and institutionalization of geographic information – its (re)distribution – are substantively transformed by the emergence of the aggregate of phenomena and practices that collectively constitute ‘the geoweb’. These transformations are rendered invisible if we are left with accounts of the phenomenon as solely technologically and pragmatically driven.

I begin by examining how the geoweb has been engaged within the geographic literature. I argue that while the descriptive emphases of early engagements are appropriate given the nascent nature of developments, they do not capture the ways in which the geoweb is politically and economically produced beyond being technically enabled. Situating the geoweb within political economy reveals the phenomenon to be immediately implicated in, and made possible by, emergent forms of governance that are best described as ‘neoliberal’. As evidence of new regimes of spatial information, and a changing balance of power between the state, markets, and individual actors in the governance of geographic information, I draw upon the recent activities of corporate entities such as Google; citizen activities and movements centered around OpenStreetMap (OSM), an open-source Web-based mapping application crowdsourced by users; and state-based actions around GI, including responses of individual states to the geoweb. I broadly organize this evidence, discerned largely from a North American and European context, in terms of Peck and Tickell’s (2002) ‘roll-back’/’roll-out’ dynamics, suggesting that as ‘the state’ is rolling back (or is being undermined) from select aspects of the ‘cartographic project’ writ large, the rolling out of market-based regimes of geographic information governance is signaled in terms of the neoliberal imperatives of free labor, corporate/private ownership, and supra-accountability.

II Writing the geoweb

From a data perspective, the geoweb consists largely of information ‘volunteered’ or generated by lay cartographers over the Web. While much of this locational data curated by users is certainly geographical, little of it is map-based in the traditional sense, consisting instead of geotagged blog posts, georeferenced digital photos, GPS waypoints, locational updates to social networking sites, and other texts. Much of this information is largely quotidian (Elwood, 2009), but these contributions are nevertheless valuable. Attributing locational referents to content allows geographic details to be used as a primary way of indexing, or organizing, the Web such that, for example, search result relevancy is based on the proximity of a user’s location (Crampton, 2009; Haklay et al., 2008). Google has accordingly termed the crowdsourcing effect the ‘geoindex’ – the notion that all information may be organized, and is therefore searchable, by location (Crampton, 2009: 95). Even more importantly, as per Haklay et al. (2008), the geographical annotation of Web content provides this information, which is otherwise dominantly abstract, with context. The ascendance of locational data has alternatively been referred to as ‘Where 2.0’ (Newitz, 2006) because the ability to ‘volunteer’ GI is premised upon and enabled by Web 2.0 functionalities and applications. Thinking of the geoweb as an outgrowth of Web 2.0 accurately captures the extent to which it is pragmatically and technically driven outside of disciplinary geography, and this understanding is reflected in early disciplinary engagement of the geoweb.

Initially, geographers largely engaged the geoweb descriptively. The descriptive bent of the nascent literature is consistent with the novelty of the phenomenon and our disciplinary attempts at coming to terms with what, exactly, the geoweb ‘is’ by itemizing that which is new and unique about it. Early engagements of the geoweb chronicle its genesis and enabling conditions and the forms that this phenomenon is assuming in relation to GIS (Crampton, 2008a, 2009; Elwood, 2009; Goodchild, 2007; Haklay, 2008; Haklay et al., 2008; Sui, 2008). Elsewhere, researchers have sought to describe volunteered contributions in terms of their validity and/or credibility (Crampton, 2008a; Flanagin and Metzger, 2008; Goodchild, 2009; Haklay, 2010). Whereas some authors emphasize the modes of VGI contribution, others describe particular instances of mapping services such as OpenStreetMap (OSM) or Google Maps, or detail specific deployments of geoweb applications such as Google Earth (Crampton, 2008a, 2008b, 2009; Haklay, 2008; Haklay et al., 2008; Miller, 2006; Nourbakhsh et al., 2006; Sui, 2008; Yarbrough and Easson, 2005).

Geographers have also begun to attend to the functionalities – and subjectivities – that the geoweb enables. In particular, they emphasize the participatory and interactive frameworks that make the ‘volunteering’ of GI possible (Budhathoki et al., 2008; Crampton, 2008a; Guptill, 2007; Haklay et al., 2008; Sui, 2008). The ease of use of geoweb applications and services is seen to signal the ‘death of the expert’ cartographer or GIS technician (Crampton, 2008a; Goodchild, 2007, 2009; Newitz, 2006; Sui, 2008). The rise of the ‘citizen sensor’ as simultaneously the active producer and end-user of GI is associated with new subjectivities: the user as ‘produser’ or ‘prosumer’ (producer/consumer) of geographic information (Goodchild, 2007; Haklay et al., 2008). These emergent subjectivities are related to a radical shift in the production of spatial data (Crampton, 2009; Elwood, 2008a; Flanagin and Metzger, 2008; Haklay et al., 2008). Contrary to a model where GI is produced by cartographic experts and disseminated by them to passive end-users, VGI efforts represent a distinctly different regime, whereby it is private citizens – increasingly addressed as consumers – who create spatial data over the Web (Budhathoki et al., 2008; Elwood, 2008a, 2008b, 2009; Goodchild, 2007).

While changes to user identities signal changes in modes of geographic data production, geographers question the extent to which these practices constitute a form of voluntarism (Blamont, 2008; Crampton, 2008a; Elwood 2008a, 2008b, 2009; Haklay et al., 2008; Tulloch, 2008). These challenges have, however, been raised largely with respect to the degree to which new technologies fully disclose ‘participation’ to end-users – for example, are users aware that their GPS-enabled devices are ‘calling home’ and reporting data to the geoweb (Tulloch, 2008)? While these are valid concerns, questions of voluntarism are effectively reduced to concerns with how ‘participation’ is engineered into applications and hardware. These hardware/software objects are themselves subsequently engaged as natural extensions of Web 2.0 technologies and functionalities to the geographical domain. This is evident in the isolation of, for example, the development of the AJAX group of programming languages which enabled ‘mashups’ by allowing the embedding of content across browsers, or the rise of wiki architectures that support the editing of content by multiple users (Crampton, 2008a; Guptill, 2007; Haklay, 2008; Haklay et al., 2008; Malik, 2005; Sui, 2008).

This reading of the geoweb as technics and pragmatics on its own would suggest that the geoweb is but a natural outgrowth of Web 2.0. This Web 2.0-centric narrative inadvertently reduces the geoweb to be the autonomous outcome of linear technological progress that is inevitable, natural, and cannot be stopped. This leaves us with a socially and politically ahistorical account that fails to contextualize the very technological and business transitions underwriting the geoweb’s emergence. Moreover, geographic information, software/hardware, and data praxes associated with the geoweb do not represent a radical historical rupture with their GIS antecedents. As Haklay et al. (2008) point out, many geoweb functionalities parallel existing capabilities built into ‘old’ GIS architectures. While new developments may not represent a fundamental change in terms of how geographic information is delivered over the Internet, geographers agree that they do represent a radical shift in its production (Crampton, 2009; Elwood, 2008a; Flanagin and Metzger, 2008; Haklay et al., 2008). These changes to production cannot, however, be explained solely in terms of enabling technologies that transfer data-authoring capabilities to everyday cartographers. Rather, they constitute fundamentally new regimes of production that make the ‘prosumption’ of geographic information valuable. These regimes of data production themselves have not emerged out of the technological ether, but are instead historically coincident with, and implicated in, political economic shifts associated with changes in practices of governance (Singh, 2002).

III Neoliberal governance and the state

Sparke (2006) defines ‘governance’ as top-down forms of political decision-making by social elites. Elites may govern through securing hegemony, or willful acquiescence – the ‘sensual-cum-socialized forms of political control’ (Sparke, 2006: 357). Governance constitutes a ‘structural’ thesis of power, which suggests that power assumes particular ‘forms’, be they political (governments), economic (markets), or social (e.g. patriarchy) These forms materialize in processes of ‘regulation’, or the exercise of authority, over persons and territory (Rose, 1999), and also over data (Singh, 2002).

‘The state’ is one such material ‘form’ of governance. This does not mean that the state is a monolithic entity that secures hegemony by monopolizing the use of force. Theories of the modern state maintain that the state has been governmentalized, such that the mechanisms by which it secures itself are not necessarily state-based (for example, sustained by social hierarchies or the control of sexuality; Foucault, 1980). Governmentality represents the location of ‘governing’ beyond the state in the sense that power relations exceed the available categories and apparatuses of the state. Although ‘the state’ does not exhaust the field of power relations, Rose (1999) argues that the governmentalization of the state is that which has allowed it to endure within the contemporary power matrix, where it functions as a connective force that brings together ‘a diversity of authorities and forces … within a whole variety of complex assemblages’ (Rose, 1999: 18). It is precisely in this way that ‘the state’, though governmentalized, remains a meaningful category. It persists as an instrument of governance that functions to impose ‘fixity (and hence stability)’ onto the politico-economic order via regulatory and other mechanisms (Harvey, 1990: 109).

While Foucauldian approaches emphasizing governmentality help us to understand power as relational, positive, and productive, Foucault’s emphasis on what Jessop (1990: 238) calls the ‘dispersed multiplicity of “micro-powers”’ does not connect the praxis of power to the particular ‘interests’ that are served, or satisfied, in any exercise of power. Foucault impressed upon us the capillary nature by which power circulates through the social fabric. Ontologically, Foucauldian theses see power as a priori in the Nietzschean sense of the will to power rather than as originating with particular subjects, institutions, or regimes (‘interests’). Asking ‘how’ power circulates but not ‘why’ it is exercised in particular ways (and by whom) is a useful exercise, but it does not present us with an entry point for the critique of power.

There is an established geographic tradition – often attributed to the pioneering work of the late J.B. Harley – of understanding spatial information and its (cartographic) representation as inherently implicated in practices of securing and exercising power. In other words, the ‘cartographic project’ has always served particular interests (of power). ‘Interests’, per Jessop (1990), ‘presuppose a certain level of structural determination’ (p. 238). If we wish to attend to changes in regimes of regulating the production, distribution, and consumption of geographic information, a theoretical lens that allows us to link particular exercises of power to interests is required such that we may engage ‘the use of institutions, structures and processes of collaboration to allocate [geographic information] resources and … coordinate or control [mapping] activity in society or the economy’ (Christophers, 2009: 87). Barnett (2003) argues that the nature of these political economic transitions – which constitute an ‘institutional shift of government from the state to the market’ (p. 98) – necessitate a theory of governance precisely because it makes visible the ability of both the state and the market (as structural forms) to accrue and exercise power.

‘Governance’ is, ultimately for Sunder Rajan (2006), a question of political economy. Late capitalism has facilitated a particular co-articulation of markets, technology, and science that he terms ‘technoscientific capitalism’. In the wake of the 1970s energy crisis, this took the form of radical experimentation with modes of social and political regulation introduced as correctives to Fordist modes of production (Cooper, 2007; Haraway, 1997; Harvey, 1990; Sunder Rajan, 2006). In the USA, this radical restructuring of labor and production was realized by transitioning away from manufacturing to an ‘innovation-based economy, one in which the creativity of the human mind (a resource without limits) would replace the mass-production of tangible commodities’ (Cooper, 2007: 17–18). What is unique to technoscientific capitalism is this very capitalization upon the creative potential of human intelligence (Sunder Rajan, 2006; Terranova, 2004). It is exactly the kind of creativity that is latent in the crowdsourcing of geographical content over the Web.

Technoscientific capitalism raises questions about the changing balance of powers between the state and private sectors. Fluctuations in state/market dynamics are consonant with the restructuring of the state under neoliberalism. Neoliberalism does not exclusively refer to changes in state dynamics, but it is associated with the emergence of new modes of political economic governance that may be engaged in terms of policy (the planned transference of public services to the public services to the private sector and the deregulation of markets), ‘Ideology (the hegemonic maxim that human wellbeing is maximized by enacting the policy triumvirate of privatization, liberalization, and deregulation), and governmentality (the transference of the onus of responsibility for social welfare to individuals themselves, and the attendant formation of the hyper-responsibilized, entrepreneurial, self-actualizing subject) (Foucault, 2008; Larner, 1999; Mann, 2010; Navarro, 2007; Peck, 2004; Read, 2009; Rose, 2007).

Whether understood as policy, ideology, or governmentality, neoliberalism is nevertheless fundamentally associated with the ‘restructuring of welfare state processes’ (Larner, 1999: 6). The introduction of novel modes of political economic governance may be evidenced in the ‘rolling back’ of the state (Peck and Tickell, 2002). Theories of state ‘roll-back’ are not, however, tantamount to the hollowing out of the state (Peck, 2004). Neoliberalism, much like capitalism, is always incomplete and contradictory (Larner, 2000). Furthermore, it assumes unique forms in different political, sociocultural, and geographic contexts (Harvey, 2005; Larner, 2000). It less signals the disappearance of the state than a reconstitution of state power vis-a-vis the market. As per Sunder Rajan (2006), governance, neoliberal or otherwise, is ‘always already a melding of “state” and “corporate” forms and rationalities’ (p. 80).

Roll-back/roll-out dynamics are a useful way of thinking through this shift in the balance of power between state and private actors. As categories, they begin to capture – albeit at a very coarse level – the ways in which welfare state processes are being recoded in the image of neoliberalism. The emergence of the geoweb signals a shifting relationship between the state and market entities such as corporations in accordance with neoliberal rationalities that are uniquely structuring the production and dissemination of geographic information.

IV Rolling back …

Mapping has always been intimately bound up with the state. The very notions of ‘state’ and ‘map’ ‘developed together and in reference to each other’ (Edney, 2009: 12). Maps and mapping emerged not only as tools for state-making, but also as instruments of governance (Pickles, 2004). The establishment of governmental order via cartographic systems such as cadastres not only make a land and its people taxable, but render them legible and thereby governable (Mitchell, 2002; Pickles, 2004; Scott, 1998). As a complex of the ways in which geographic information is produced, circulated, and represented, the ‘cartographic project’ is one of a series of ‘state projects’: sets of practices that give ‘the state’ form, delineating it as an internally cohesive entity and endowing it with substance (an exclusive territory, population) (Jessop, 1990).

Historically, mapping was not only the responsibility but indeed the preserve of the state or sovereign. The dissemination of maps was strictly controlled, not only for reasons of geopolitical security or strategy, but also because the knowledge they contained was considered democratizing and empowering, and therefore subversive (Hannah, 2000; Harvey, 1990). The state governed GI in almost all aspects of its production, dissemination, and usage. This historically tight coupling between the state and the enterprise of mapping is, however, no longer exclusive.

For Michael Goodchild (2009), the rise of VGI and other geoweb practices signals the end of what he has termed the ‘modernist era of mapping’. This era was characterized by the authoritative production of ‘the map’ by the state, which acted as the purveyor of geographic information and the guarantor of its (in)accuracy (Goodchild, 2009). In the public provision of mapping and cartographic services, the state is now beginning to look to private actors not only to supplement the curation and dissemination of geographical information, but also to take over responsibility for it in instances of, for example, natural disasters such as Hurricane Katrina or the more recent Haitian earthquake. As the state is partially ‘rolling back’ from the mapping enterprise, distinctly market regimes of spatial data governance are simultaneously ‘rolling out’. This is not to imply that the state no longer maintains any role in the provision of GI. In the West (the USA and UK in particular), rather, the state’s role is changing from that of sole purveyor of geographic information and arbiter of cartographic truth to that of one of many producers and facilitator or institutional body of oversight. Corporations, non-state actors, and private citizens are now performing and fulfilling functions that were long the exclusive preserve of state mapping organizations.

VGI is not the same as the data authored within formal institutional contexts (Elwood, 2008a). It diverges from the Spatial Data Infrastructure (SDI) model by which formal agencies made information available through state geodata portals, which secured the ‘credibility’ of data as vested in the authority of the state producers and cartographic experts (Budhathoki et al., 2008; Elwood, 2008a; Flanagin and Metzger, 2008). The state, however, is actively undermining itself as the seat of cartographic legitimacy by abdicating from serving as the arbiter of geographical truth. Instead, geographical ‘truths’ are now being created by millions of users. As per Michael Jones, Google Earth’s Chief Technologist, the hope is that the dissemination of geospatial tools will result in a ‘big number of users converging on a truth’ (cited in Crampton, 2009: 97). This is guided by the premise that people in a place know the ‘local’ better than an institutional map agency.

It is under this premise that the state is actively looking to VGI as a ‘[resource] to augment, update, or complete existing spatial databases’ (Elwood, 2008a: 176). Goodchild (2007) describes this in terms of a ‘patchwork’ approach to curating and maintaining spatial fabrics, where users supplement existing grids where ‘public funds and staffing cannot generate a common data set’, as in, for example, the annual Christmas Bird Count (Elwood, 2008a: 176). But the more salient examples are those that demonstrate an increasing reliance of the state upon the mapping practices of its citizens to fulfill functions that it can no longer sustain economically. In the USA, spatial data have long been ‘free’ under the auspices that they are collected using public revenues. The inability of the state to recoup its costs for spatial data has translated into the state recruiting its citizens to ‘patch’ the national geographical grid. Although now defunct, the United States Geological Service (USGS) began a program soliciting the time of volunteers who, through the National Mapping Corps, would contribute spatial information to the US National Map to spare agency resources (Sui, 2008). The US Census Bureau, which is one of the largest public producers and distributers of public spatial data in the USA, now uses volunteers to collect information on the homeless (Sui, 2008).

These examples, though few, are indicative of the state’s devolution of (cartographic) responsibility to individual citizens (Rose, 2007). Elsewhere, the state is also actively looking to the corporation as both content provider and spatial delivery mechanism. The much-cited aftermath of Hurricane Katrina is illustrative in this regard. On the day after Hurricane Katrina made landfall on 29 August 2005, one of the masters of a popular blog commented on previous posts advocating the use of Google Earth to analyze US National Oceanic and Atmospheric Administration (NOAA) imagery that had started being flown earlier that day (Jardin, 2005). Specifically, the blogger was suggesting using Google Earth to visually map out areas that had been affected, and to mark the extent of the inundation as a means of helping residents identify whether or not their properties had been flooded. NOAA flew high-resolution aerial photo non-stop for 10 days in the immediate aftermath of the disaster, taking over 8000 images (Nourbakhsh et al., 2006; Yarbrough and Easson, 2005). NOAA then passed the images on to the Carnegie-Mellon University-based Global Connection Project (GCP) to process the imagery. GPC members calculated the proper alignment of the images, georeferenced them against Google Earth, and disseminated them as KML files to be overlaid with Google Earth such that they could be used by emergency responders on the ground (Nourbakhsh et al., 2006; Yarbrough and Easson, 2005). This was deemed a much more efficient means, over the use of OGC ³ standards, of making the most current and relevant information available for crisis response (Haklay, 2008).

While the state (NOAA) was the primary supplier of data (aerial imagery), it turned to private citizens (GCP) to process that information and relied on a proprietary technology, Google Earth, as a platform for not only disseminating those data, but also rendering them meaningful – both useful and useable. NOAA’s use of Google Earth and GPC evidence an increasing incidence of state reliance upon lay citizen mapping and the use of proprietary Web-based geovisualization technologies. More importantly, it represents a political economic reconfiguration of state the state vis-a-vis the cartographic project.

Although state power is being reconstituted in relation to mapping, the state retains a presence in the ‘brave new world’ (Schuurman, 2009) of the geoweb. This is demonstrated by, for example, recent state-level attempts to regulate Google Street View, a perspective available in Maps and Earth that allows users to view high-resolution panoramic images taken at street level along roads in major cities throughout the world (Elwood and Leszczynski, 2010). The governments of several industrialized countries, including those of Canada, Japan, Switzerland, Greece, and Germany have charged that the technology, and the inconsistency of blurring algorithms for eliding identity (faces and automobile license plates), constitute serious privacy violations (BBC News, 2009; Schmidt, 2009; Siddique, 2009; Timmer, 2009; Williams, 2009; Wong, 2009). These states have accordingly attempted to restrict, with various degrees of success, Google’s Street View imaging and data handling activities. Mandated privacy ‘fixes’ include improvements to blurring algorithms, advanced public notification of imaging, lowered camera heights, and finite raw image data retention periods (BBC News, 2009; Capper, 2009; Kirk, 2009; Paul, 2009; Schmidt, 2009; Timmer, 2009; Wong, 2009; Zafra, 2009). The state is still very much present with respect to geoweb activities, but its role is fundamentally changed from that of sole or dominant provider of geographic information to its citizen base to, in the case of Street View, a regulatory body of oversight and an intermediary between citizens and corporations (Elwood and Leszczynski, 2010).

Following Peck (2004), the neoliberal state continues to intervene, albeit in an entirely different capacity. Rather than intervening on the behalf of labor, it functions to create conditions for private enterprise (Peck, 2004). The reduction of the role for the state in the enactment of neoliberal policy agendas, as well as economic pressures which further underwrite the partial withdrawal of the state from the ‘cartographic project’, are creating conditions for the entry of non-state entities into the provision, dissemination, utilization, and visualization of geographic information. As they secure an increasing presence in the enterprise of public spatial data provision, these entities are subsuming geographic information to distinctly market-based regimes of governance.

V … and rolling out

The ‘rolling out’ of corporate regimes of spatial data governance is consonant with new, post-dot.com logics of free labor, private ownership, and supra-accountability that may be framed in terms of the imperatives of technoscientific capitalism. Free labor, private ownership, and unaccountability are not exclusive to, nor sufficiently definitive of, technoscientific capitalism, nor of neoliberalism. Rather, they constitute three dimensions of contemporary political economy – which we may call neoliberal – in which the market regulation of the geoweb is particularly (and increasingly) evident.

VI Relationships, changed

The emergence of ‘Google governance’ is simultaneously implicated in, and contingent upon, a reconfiguration of the state’s relationship to the ‘cartographic project’. The partial withdrawal of the state from the public provision of GI opens the production and dissemination of geographic information to the market, where private actors subsequently engage ‘mapping’ as an opportunity for capital accumulation. Profit is by no means the only incentive driving the proliferation of geographic information, applications, and services over the Web. Nor are corporations the only force motivating the diminished role of the state – in particular capacities – from mapping. Citizens are active participants in driving a reconstitution of the cartographic project beyond the state. OSM was started in the UK in 2004 as a citizen response to the lack of publicly available geographic information. It furthermore organized campaigning on the part of citizens that resulted in the recent public release of Ordnance Survey (OS) data in April 2010 (Stratton, 2009), a profound shift to historical regimes of spatial data governance in Britain where the OS had maintained geographic information as the preserve of the state via strict copyright and steep cost for access to and use of the data.

The online presence of OSM has not in any way replaced state mapping activities. The OS continues to provide authoritative, high-resolution data accurate to within a meter (Atkinson, 2010). It is a dominant GI producer, generating enormous amounts of data with up to 5000 edits to the national MasterMap a day (Atkinson, 2010). The ‘roll-back’ of the state from the ‘cartographic project’ is partial, and is discernible most immediately in the ways in which ‘netizens’ (Internet users) both access and interact with geographic information, rather than the wholesale displacement of state mapping by corporations and other non-state entities. Net citizens now access data from non-state sources that may be considered ‘market’ to differing degrees. The production and dissemination of GI to the public is increasingly controlled by these non-state entities through their geospatial services and applications, even when the data they serve are ‘state’ in origin. This is the case in the American context, where state-produced data comprise a significant portion of Google Maps’ US base geometry (Dobson, 2009). Google, however, brands these data as its own (© Google). This is possible given a culture where GI collected by the state using tax dollars is publicly available and may then be leveraged at no cost by market actors who subsequently own any ‘value added’ to the data (Dobson, 2009). This is similarly true of Bing Maps, though Microsoft also continues to use commercial data providers for the USA.

The prolific successes and public nature of these online presences change the relationship of citizens to GI. These changes in how citizens relate to ‘the mapping enterprise’ have subsequently transformed interactions between the state, non-state mapping entities, and geographic information. In late 2010 the Winchester, MA, planning department requested local council funding to update its GIS (Petrishen, 2010). This was denied, in light of budget shortfalls, due to the misconception, held by council members, that aerial imagery would not provide anything that could not already be acquired from Google Earth (when indeed Google Earth imagery was often not high enough in resolution, or had too much tree cover to be useful for modelling or planning; Petrishen, 2010). The consequences of the esteem of applications and services such as Google Earth or Bing Maps in the public consciousness are much more serious when they are geopolitical. In November of the same year, the Nicaraguan military used erroneous Google Maps geometry of the final stretches of the Rio San Juan – which demarcates the Nicaragua/Costa Rica border – showing a portion of Costa Rican territory as lying within Nicaraguan as an excuse to invade the area, which Nicaragua maintains rightfully belongs to it despite international recognition otherwise (Carr, 2010; Geens, 2010). Costa Rica subsequently appealed to both the Organization of American States (OAS) and Google to intervene in the dispute (boz, 2010): a clear indication of the emergence of market-based regimes of geographic information governance, even at the international scale. Google, however, claimed that it was not responsible – i.e. unaccountable for any fallout – citing the source of the data as the US State Department (Geens, 2010).

VII Conclusion

‘The state’ and ‘the market’ are coarse categories, and the shifting power dynamics between these two entities are not as cut-and-dried as ‘roll-back’/’roll-out’. In the first instance, the ‘roll-out’ of market-based regimes of geographic information governance do not signal the eclipsing of the state by the market. The issue, as evidenced throughout this paper, is one of complementarity – of a co-occurrence, in many cases, of both state and market-based geographic information governance regimes. Evidenced by the rise of non-state entities in the provision and dissemination of GI and a reconfiguration of the role of the state in certain of its mapping capacities, this coexistence is not always a happy one. Market actors are not simply picking up where the state leaves off. Corporations are aggressively challenging – and competing against – state control of the cartographic project. Google and the Ordnance Survey are embroiled in legal action over Google’s desire to include high-resolution OS data in its (Google’s) map fabric of the UK (Atkinson, 2010). The OS objects to Google’s licensing clauses, which stipulate that Google reserves the right to license – i.e. effectively owns – any downstream uses of Google Maps data, regardless of who the original data producer is (here, the OS; Atkinson, 2010). Google, on the other hand, is citing favoritism as OS data are available through both Bing and Yahoo Maps (Atkinson, 2010).

Similarly, geoweb praxes are not ‘market’ by virtue of being non-state. In the era of the geoweb, this boundary is not clearly discernible. Online presences such as OSM present new hybrids that do not fit a neat ‘state’/’market’ dichotomy. OSM is freely volunteered; this does not, however, mean that it exists entirely outside of market logics. Bing Maps recently began including OSM as a data layer (Jeffries, 2010). According to a popular blogger (Fee, 2010), the marriage of Bing and OSM allows Microsoft to compete with Google where Google data are comparatively less detailed or do not have as established a presence (western Europe, for example). OSM in turn gets to use Bing’s high-resolution imagery.

Moreover, the state makes the difference. Governments are not responding uniformly to geoweb presences. Individual states’ responses to Street View have been distinctly unique, with states seeking the instatement of different privacy mediation measures (Elwood and Leszczynski, 2010). The privacy apparatuses of these countries are furthermore distinct. Whereas most European countries and Canada have privacy commissioners or data protection branches, in America privacy is considered a legal entity, and claims against Street View privacy violations proceed through the courts (Elwood and Leszczynski, 2010). The free curation and circulation of geographical information characteristic of many of the instances cited here are not a worldwide reality. In China, mapping and surveying are activities restricted to the State Bureau and subject to arrest, a fate met by a group of Chinese students collecting VGI in 2008 (Ball, 2009a). Here, the state and the cartographic project remain tightly coupled. Even in places around the world where there is evidence that the state is rolling back from the public face of mapping, where it seems to disappear in one capacity (as the sole provider and ‘governor’ of geographic information), it reappears elsewhere – e.g. militarily.

The extension of state mapping activity to the military is nothing new. Many geographic information technologies (e.g. GPS, satellite) are military in origin. Goodchild (2009) however notes that the US military is actually ramping up spatial data production for strategic and other military purposes. These data, though funded by public monies, are not in the public domain (Harvey, 2007). On the one hand, the state is no longer acting as the singular source of geographic information, seeming to partially ‘roll back’ from sole responsibility for the cartographic project; on the other, it is strengthening its grip on the cartographic project in ways that further remove particular aspects of mapping from the public sphere.

These caveats notwithstanding, ‘roll-back’/‘roll-out’ and ‘state’/‘market’ are useful for thinking through the ways that the geoweb signals, and is heralded by, a neoliberal recoding of the production, distribution, and institutionalization of geographic information. As argued in this paper, the increasing marketization of spatial data regimes may be evidenced in how GI is regulated in terms of labor, ownership, and accountability. Geographers such as Zook and Graham (2007) and Harvey (2007) have begun to consider the changing political economy of geographic information. Yet without an attention to political-economically driven shifts to regimes of how spatial data are produced, distributed, structured, and institutionalized, Zook and Graham’s (2007) nod to ‘Google governance’ on its own suggests that the ability of Google to capture geoweb market share is based on its corporate prowess or the soundness of Google as a business model (for an extreme example, see Lee, 2010). I have sought to build upon these initial accounts by situating the geoweb within a political economic context that allows us to understand these new online presences and attendant practices as implicated within a much broader restructuring of governance regimes affecting state processes and projects that include, but are not limited to, the cartographic project.

This is significant for our thinking about the geoweb because it enriches our narratives of geoweb genesis beyond simply the enabling technical developments and business solutions of Web 2.0. Such a narrative suggests that the geoweb is the result of autonomous technological change, eliding the ways in which technology is always socially produced, contingent, and contested. As socially produced, the geoweb is coincident with real-world political economic processes that structure the forms that geoweb phenomena assume, whether as data, technological media, or praxis.