Elsevier

Journal of Rural Studies

Volume 84, May 2021, Pages 1-11
Journal of Rural Studies

Growing algorithmic governmentality: Interrogating the social construction of trust in precision agriculture

https://doi.org/10.1016/j.jrurstud.2021.03.004Get rights and content

Highlights

  • Agriculture technology firms socially construct moralistic trust to ensure farmers’ engagement with precision farming.

  • Farmers situate their own identity and knowledge vis-à-vis precision agriculture’s algorithmic rationality.

  • Discursive power can shift farmers' social identities, essential to the reproduction of agritech capital.

Abstract

Precision agriculture (PA) is restructuring farmer livelihoods and identities through a panoply of technologies that generate and process big data to influence agricultural practices. In this paper, we ask the question: How does algorithmic rationality impact farmers' trust in PA? We focus on the modalities of power wielded by agritech firms through PA that socially construct a form of moralistic trust, the politics of knowledge and knowledgeability, and the internalization of new social identities. This research study utilized a mixed methods approach that included focus groups and follow-up surveys with social actors along the PA value chain. We found that agritech firms have successfully positioned their knowledge products as superior to farmers' experiential knowledge, thereby ensuring farmers’ sustained engagement with PA technologies for the purposes of data capture and capital accumulation. Farmers internalize the algorithmic rationality of PA and position themselves along a moral register through governmentalized actions that ostensibly demonstrate moralistic trust in the system. This process has the effect of transforming social identities, interpellating farmers as the architects of their own alienation. Agritech is increasingly adept at digitally abstracting farm knowledge away from farmers. PA is a battleground wherein the politics of agrarian knowledge are contested.

Introduction

Precision agriculture (PA) is an approach to farming that utilizes numerous data-driven technologies that generate localized farm data to assist farmers with decision-making in managing their food production system (Bongiovanni and Lowenberg-DeBoer 2004; Rossel and Bouma 2016). The widescale application of PA technologies, such as in-situ sensors, drones and satellite imagery for collecting data and artificial intelligence (AI) and machine learning algorithms for mining them, are providing farmers with recommendations on when to plant, seed, spray and harvest (Gebbers and Adamchuck, 2010). While these tools can enable a more economically productive and environmentally sustainable farming practice, PA engenders social frictions. In this study, we explore these social frictions and theorize why farmers who are negatively impacted by PA technologies continue to use them. We focus on the social construction of trust, the politics of knowledge and transformations in the social identities of users. In the proceeding subsections, we situate PA as an accumulation strategy within the long genealogy of ecological modernization, whereby agritech firms exercise governmental power to articulate new subjectivities of users through the production of ‘precise’ knowledge.

PA systems create social exclusion that benefit large scale technologically intensive monoculture farming systems over other systems, such as agroecology (Klerkx and Rose 2020). At the same time, agritech firms (e.g., John Deere, Bayer-Monsanto) and state agencies frame PA as a grand technoscientific project seeking to modernize farming in the context of global environmental change by essentializing and capitalizing off of farmers’ ambitions to improve livelihood practices by adopting these digital tools (Kuch et al., 2020; StartupAUS 2016). Proponents of PA in the agricultural technology firms (agritech) seek to frame these technologies as being in the public interest, agricultural innovations that are socially and economically desirable for the public and the environment.

However, such imaginaries are often discursively counterposed against neo-Malthusian tropes that portray non-participating farmers as responsible for declining yields that delimit a burgeoning human population (Godfray et al., 2010). For example, ‘A bevy of new technologies is now starting to pervade the farm that we believe has a central role in ensuring humanity cheats Malthus again over the next 100 years’ (Goldman Sachs, 2016:5). Viewed through this neo-Malthusian logic, so-called ‘early adopters’ of PA are positioned to ‘save the world’ whereas so-called ‘laggards’ are a threat to global food security (Carolan 2020a). Essentially, agritech firms are asking the following question: ‘What makes growers confident enough to turn from “Laggards” into “Early Adopters” or “Early Majority”?’ (AGCO 2016). The following passage is instructive to how agritech firms discursively articulate farmers along a moral register to promote their technologies: ‘Historically, it takes 40 years for a new idea to become fully accepted. In other words, it is a long time from when a Techie first uses a new idea to when a Laggard finally adopts it. And adoption is a process the profitable Early Majority Pragmatists will not take on until the Early Adopter Visionaries do. Likewise, the large Late Majority Conservatives will not until they see that the Pragmatists are successful. However, the pace of adopting new technology is increasing’ (Farm Equipment, 2014). Agritech firms are increasingly interested in psychologically persuading farmers to ensure the firms' capital accumulation.

As a paradigm in agricultural production, PA systems represent the latest iteration of capital-intensive ecological modernization of farming. Goldman Sachs' advice on PA to investors is revealing: ‘In a gold rush, sell shovels’ (Goldman Sachs, 2016:10). Previous technological transformations in agriculture, such as the Green and biotechnological revolutions, were framed by agribusiness and governments as a ‘a story of radical and progressive technological change that has been embraced by literally millions of farmers’ against looming crises of food security (Glover 2015:229). Cioffo et al. (2016) argued that many advocates of a new Green Revolution for Africa place the burden of low crop productivity on to farmers' unwillingness to use modern agricultural inputs. Most development programs, in their cost-benefit analyses, narrowly focus on economic benefits of smart-technologies to farmers while ignore its negative impacts on agrarian labor (Kansanga et al., 2019). PA presents a similar narrative of ‘success’ and ‘transformation’ through harvesting the power of high-performance computing, automation, and big data. ‘From the introduction of the steel plow and reaper over 200 years ago to modern advances like large scale planters and genetically modified seeds, man continues to find new ways to feed the world's growing population. As we near the limits of available arable land, a confluence of technologies is driving a new leg of productivity in Precision Farming to enable growers to meet the demand challenge from existing land’ (Goldman Sachs, 2016: 4).

We know from previous technology-driven agricultural paradigms, such as the Green Revolution and biotechnology, that these innovations not only bring economic benefits but also produce social and cultural risks (Bronson 2015). For example, the legal system through patents that protects innovations by large agribusiness corporations on seeds of genetically modified organisms (GMOs) has brought benefits to some farmers from improvements to crop yields but has also expanded power inequities between conventional and organic farmers and bonded stronger dependencies between farmers and powerful agribusinesses (Bronson 2015; Stucke and Grunes 2018). Large agritech corporations, through processes of mergers and acquisitions (e.g. the Bayer-Monsanto merger), are monopolizing the digital agriculture space by consolidating proprietary data and intellectual property from seed and chemical patents and digital platforms to transform farmers from independent business owners to captured users (Bronson 2019; Davidson 2018). PA falls within a genealogy of ecological modernization forged through asymmetrical power relations within a politics of agrarian knowledge.

By wielding discursive elements of ‘accuracy, prediction, and quantification’, PA allows the agritech industry to dominate over existing systems of scientific knowledge (Kuch et al., 2020). ‘Today's farmers have access to a wealth of data. So much data, in fact, they often don't know what to do with it’ (AgFunder 2017). Consequentially, PA is transforming some farmers' ‘knowledgeability’, or everyday social practices, including ways of knowing, doing, and being (Carolan 2020a). Likewise, farming practices are becoming intertwined and transformed through interactions with digital artifacts, spaces, and infrastructures (Orlikowski 2006). Through sustained engagement with PA technologies, farmers begin ‘acting like algorithms’ (Carolan, 2020b). For example, new John Deere's tractors are embedded with ‘digital locks’ that impedes farmers' ability to repair the machinery themselves (Carolan 2017). Farmers must take their equipment to an authorized John Deere dealer, who is the only ‘knowledgeable’ authority to conduct repairs. Digital locks allow firms like John Deere to ‘effectively retain control over aspects of someone's tractor’ (Carolan 2020a:17), while redefining farmers' knowledge of repairing their equipment. These proprietary tractors also use numerous environmental sensors that generate data for the entire farming system, including farmers' preferences and actions, which are protected through intellectual property rights that prevent the farmers from accessing, controlling or possessing this information—an act of dispossession (Carolan 2017; Rotz et al., 2019a). The repeated use of these technologies engenders a social dependency of farmers who ‘get something in return’ for their generation of data (Fraser 2019). Through the production of ‘trusted’ knowledge and equipment, PA systems increasingly mediate a social order in which farmers are becoming dependent on the information and commercial inputs supplied by these firms (Higgins et al., 2017:197). Effectively, ‘trusted’ knowledge is mechanistically configured into a ‘trusted’ PA subject.

Farmers have expressed concerns from opaque data sharing agreements and recommendations made by intelligent decision support systems. A recent survey conducted by the American Farm Bureau Federation showed that almost 70% of farmers are concerned that corporations and the government may unfairly access, use and sell their data or use it for marketing products and services and enforcing regulations for clean soil and water quality (American Farm Bureau 2016). Notwithstanding these concerns, through repeated engagement with PA systems, many farmers internalize the algorithmic rationality of PA technologies (Miles 2019). Farmers begin to ‘trust’ the data-driven recommendations and accept the ‘disciplinary directives offered by algorithmic authority’—subject to a form of ‘governance by algorithm’ (Carolan, 2020b)—thereby ensuring the cultural hegemony of agritech firms and PA technologies (Lupton, 2015:104). Farmers internalize the new logics of privacy or data sharing and modify their livelihood practices to fulfill the policies and ambitions of the state and agriculture tech providers, becoming subjects of the PA system itself (Cheney-Lippold 2011; Fraser 2019). This process can be conceived of as governmentality, exercising technologies of power that seek to govern behaviors and influence social identities through subject-making (Foucault 1991).

Digital representations and algorithms represent an increasingly sophisticated ‘technology of government,’ modalities of knowledge production wielded for the purposes of population management by both state and non-state actors. Rose-Redwood (2006) identifies geocoding as a necessary prerequisite of constituting ‘governable subjects’ that could be enumerated for census-taking. Wilson (2011) studies governmental practices of geocoding within a citizen engagement campaign in Seattle which consisted of residents normatively mapping the built environment along a value-laden continuum of ‘deficits’ to ‘assets’. Jefferson (2017) asserts the Chicago Police Department's GIS-based CLEARmap application is a surveillance apparatus that constitutes Black and Latinx populations as racialized subjects for policing and carceral discipline and refracts public perceptions of crime through the prism of geospatial rationality. Effectively, CLEARmap is an example of how “cartographical production of subjects intersects with racialized carceral power” (Jefferson 2017: 9). Absent algorithms and cartography, subjects are also articulated within environmental registers. Indeed, there exists a long history of subjectification to new systems of environmental management through political economy (see Agrawal 2005; Birkenholtz 2009; Cavanagh 2018; Moulton and Popke 2017). For example, Bose et al. (2012) traces a genealogy of the colonial and postcolonial Indian state that created administrative categories of identity for indigenous groups for the purposes of forest management. The Bhil indigenous group strategically internalized state-created categories of ethnic identity to ensure access rights to the forest. The ‘forest governmentality’ of this case enabled the perpetuation of state control over forest lands and the domination of tribal groups.

Agritech firms wield discursive power to create new social categories of farmers. These categories influence PA farmers and their social identities. Their internalization is accomplished through the social construction of trust. PA's algorithmic rationality that mediates user-technology relations interpellates1 a PA farmer subject with governmental responses who increasingly performs the role of a technology savvy farmer, an early PA adopter (Carolan 2020a). Hence, PA serves as a moralistic designation of a ‘modern’ farmer who ‘trusts’ scientific knowledge and technology to ‘improve’ farming practices (Kuch et al., 2020; StartupAUS, 2016). In other words, farmers come to perceive and behave towards PA systems through a moralistic form of trust (Uslanar 2008). Farmers in their pursuit of becoming a ‘successful’ and ‘modern’ farmer come to see PA to belong to their ‘moral community’ (Uslaner 2008:4). They consider big data, machine learning algorithms and algorithmic recommendations as an ethical and social practice, not through a careful assessment of their risks and benefits, but through cultural transmission of idealized values of what it means to be a ‘good farmer’, in this case an early PA adopter (Higgins et al., 2017). The development and use of PA interpellates a farmer subject whose interactions with the technology are socially constructed as ‘trust’ in the system. Against this background, we ask the question: How does algorithmic rationality impact farmers' trust in PA?

This paper contains five additional sections. The next section explores the interrelations between social construction of trust in PA, the politics of knowledge and the reconfiguration of social identities. Afterwards, we discuss the methodological approaches used to conduct this research, as well as the study sites where fieldwork was conducted. Following this, we discuss our research findings on moralistic trust, the politics of knowledge and social identities. This section is followed by a discussion on the implications of PA for food production systems. This paper concludes by asserting the importance of algorithmic governmentality in sustaining digital agriculture.

Section snippets

Defining trust

Conceptualization of trust in the social sciences have differed substantially but can be categorized into two broad analytical dimensions: strategic trust (Cook et al., 2005; Hardin 2002; Robbins 2016) and moralistic trust (Dinesan 2011; Uslaner 2002)2. Strategic trust relies on the beliefs about others' trustworthiness that are formed through personal experiences and subsequent evaluation of others' competence, goodwill, or benevolence (Coleman, 1990; Hardin 2002). This form of trust ‘reflects

Methods

Fieldwork for this research was carried out in South Dakota (SD) and Vermont (VT) between October and December 2019. The research used a mixed methods approach that included six focus group discussions (FGDs) and a follow-up survey with 52 FGD participants. The participants included represented different sections of the PA space, including 1) software and hardware developers, 2) state and county extension specialists, 3) non-profit and government agencies and 4) crop, livestock, and dairy

Results

In this section, we explore power relations and subjectivities within PA between agritech firms and research participants. Specifically, we focus on modalities of power wielded by agritech firms that socially construct a form of moralistic trust, the politics of knowledge and knowledgeability, and the internalization of new social identities. In doing so, we hope to illuminate the social and political effects of this current wave of technological innovation in agriculture. In response to our

Discussion

In order to sustain engagement through a moralistic form of trust in PA technologies, agritech socially constructs their knowledge products as more precise and accurate than farmers' knowledge. Over time, this creates a form of ‘knowledge lock-in’ that erodes farmers' analogue knowledge (Carolan, 2020b). Effectively, this politics of knowledge produces subjectivities of trust by gradually shifting users' social identities. Users generate data from the food production system that is vital to

Conclusion

This paper is animated by an intellectual commitment to explore the social and political effects of PA through participatory, deliberative approaches that include perspectives of multiple and diverse stakeholders across the food system. Precision agriculture is restructuring farmer livelihoods and identities through a panoply of technologies that generate and process big data to influence agricultural practices. However, the sociopolitical effects and farmers' motivations remain unclear. This

Notes

1. Althusser explains interpellation thusly: ‘We shall go on to suggest that ideology 'acts' or 'functions' in such a way as to ‘recruit’ subjects among individuals (it recruits them all) or 'transforms' individuals into subjects (it transforms them all) through the very precise operation that we call interpellation or hailing. It can be imagined along the lines of the most commonplace, everyday hailing, by (or not by) the police: ‘Hey, you there!’’ (Althusser 2014:190).

2. Affective-based trust

Author statement

No potential conflict of interest was reported by the authors.

Funding details

National Science Foundation: Award No. 1929814 and No. 2026431.

Acknowledgments and credits

The paper is based upon work supported by the National Science Foundation under Grant Numbers (1929814 and 2026431). This paper greatly benefitted by including useful comments made by the reviewers for the journal. We would like to thank all focus group and survey participants for their valuable gift of time and input. Any opinions, findings, and conclusions or recommendations expressed in this material are those of the authors and do not necessarily reflect the views of the National Science

References (91)

  • Leanne Wiseman et al.

    ‘Farmers and their data: an examination of farmers’ reluctance to share their data through the lens of the laws impacting smart farming

    NJAS - Wageningen J. Life Sci.

    (2019)
  • ‘Increasing Precision Agriculture Technology Adoption.’ AGCO Corporation

    (2016)
  • ‘What is precision agriculture?’ AgFunder news

  • Arun Agrawal

    Environmentality: Technologies of Government and the Making of Subjects

    (2005)
  • Louis Althusser

    On the Reproduction of Capitalism: Ideology and Ideological State Apparatuses

    (2014)
  • American Farm Bureau Federation Farm Bureau Survey

    Farmers Want to Control Their Own Data American Farm Bureau Federation

    (2016)
  • Wiebe E. Bijker et al.

    The Social Construction of Technological Systems: New Directions in the Sociology and History of Technology

    (2012)
  • Trevor Birkenholtz

    ‘Groundwater governmentality: hegemony and technologies of resistance in Rajasthan's (India) groundwater governance

    Geogr. J.

    (2009)
  • Rodolfo Bongiovanni et al.

    Precision agriculture and sustainability

    Precis. Agric.

    (2004)
  • Cennydd Bowles

    Future Ethics

    (2018)
  • Kelly Bronson

    Responsible to whom? Seed innovations and the corporatization of agriculture

    Journal of Responsible Innovation

    (2015)
  • Kelly Bronson

    ‘A digital ‘revolution’ in agriculture?: critically viewing digital innovations through a regenerative food systems lens.’

  • Rob J. Burton

    ‘Seeing through the 'good farmer's' eyes: towards developing an understanding of the social symbolic value of 'productivist' behaviour

    Sociol. Rural.

    (2004)
  • Michael Carolan

    Publicising food: big data, precision agriculture, and Co-experimental techniques of addition

    Sociol. Rural.

    (2017)
  • Michael Carolan

    Automated agrifood futures: robotics, labor and the distributive politics of digital agriculture

    J. Peasant Stud.

    (2020)
  • M. Carolan

    Acting like an algorithm: digital farming platforms and the trajectories they (need not) lock-in

    Agric. Hum. Val.

    (2020)
  • David W. Cash et al.

    Knowledge systems for sustainable development

    Proc. Natl. Acad. Sci. Unit. States Am.

    (2003)
  • Conner J. Cavanagh

    Political ecologies of biopower: diversity, debates, and new frontiers of inquiry

    Journal of Political Ecology

    (2018)
  • John Cheney-Lippold

    A new algorithmic identity: soft biopolitics and the modulation of control

    Theor. Cult. Soc.

    (2011)
  • G.D. Cioffo et al.

    Modernising agriculture through a ‘new’ green revolution: the limits of the crop intensification programme in Rwanda

    Rev. Afr. Polit. Econ.

    (2016)
  • Karen S. Cook et al.

    Cooperation without Trust?

    (2005)
  • Melinda E. Cooper

    Life as Surplus: Biotechnology and Capitalism in the Neoliberal Era

    (2011)
  • Jason Davidson

    ‘Bayer, Monsanto and Big Data: Who Will Control Our Food System in the Era of Digital Agriculture and Mega-Mergers?’ Medium

    (2018)
  • del Mármol et al.

    Socio-economic transitions and everyday life changes in the rural world: pyrenean households and their contemporary economic history

    J. Agrar. Change

    (2018)
  • Thomas Dietz et al.

    Support for climate change policy: social psychological and social structural influences

    Rural Sociol.

    (2007)
  • Peter T. Dinesen

    Where you come from or where you live? Examining the cultural and institutional explanation of generalized trust using migration as a natural experiment

    Eur. Socio Rev.

    (2011)
  • Farm Equipment

    Technology for Profit: How Farmers, Dealers Adopt New Technology

    (2014)
  • Luciano Floridi

    Establishing the rules for building trustworthy AI

    Nature Machine Intelligence

    (2019)
  • Martin Ford

    Rise of the Robots: Technology and the Threat of a Jobless Future

    (2015)
  • Michel Foucault

    ‘Governmentality’ Pp. 87–104

  • Alistair Fraser

    Land grab/data grab: precision agriculture and its new horizons

    J. Peasant Stud.

    (2019)
  • Morten Frederiksen

    Trust in the face of uncertainty: a qualitative study of intersubjective trust and risk

    Int. Rev. Sociol.

    (2014)
  • Maaz Gardezi et al.

    ‘Examining the social and biophysical determinants of US midwestern corn farmers’ adoption of precision agriculture

    Precis. Agric.

    (2020)
  • Robin Gebbers et al.

    Precision agriculture and food security

    Science

    (2010)
  • Frank W. Geels

    ‘The dynamics of transitions in socio-technical systems: a multi-level analysis of the transition pathway from horse-drawn carriages to automobiles (1860–1930)

    Technol. Anal. Strat. Manag.

    (2005)
  • Cited by (0)

    View full text