Embracing the Data Revolution for Development: A Data Justice Framework for Farm Data in the Context of African Indigenous Farmers.

• Author: Dagne, Tesh W.

Introduction

Agricultural systems are increasingly depicted as digital landscapes as the digitalisation of agriculture becomes one of the significant features of technological transformations in the twenty-first century. The collection, aggregation, and processing of data from multiple data sources in the digital landscape of agriculture brings data governance questions that affect the organisation and management of agricultural production while at the same time raising intricate concerns regarding the ownership, privacy, and safety of farm data.

The digitalisation of agriculture is one of the significant features of technological transformations in the twenty-first century. Agricultural systems are increasingly depicted as digital landscapes, as shown by such expressions as 'smart' (Guerrini 2020) and 'predictive' agriculture (Food and Agricultural Organization 2018), 'precision farming' (Rasmussen 2016), 'farming 4.0' (Adam 2016) and the 'fourth industrial revolution' in agriculture (Nijhuis and Herrmann 2019).

The new digital landscape in agriculture rests on the collection, aggregation, and processing of data from multiple data sources by multiple actors. Thus, data governance strategies are needed to guide the important shifts that digitalisation brings regarding the organisation and management of agricultural production while at the same time addressing the intricate concerns that have arisen regarding the ownership, privacy, and safety of farm data. This article examines the challenges that the digitalisation of agriculture in Africa brings with respect to ownership and control of data and proposes a framework for governing the allocation of rights in data and for ensuring control over data from the perspective of African indigenous farmers.

The digitalisation of agriculture in Africa is an aspect of the data revolution that holds potential for development and sustainability. African indigenous farmers can realise the potential of the data revolution if inequalities in access to and over utilisation of data are systematically addressed to support development endeavours. Predominant regimes for the allocation of rights in data favour exclusive data ownership by such intermediaries as data collectors, aggregators, processors, and users. As originators of data, African indigenous farmers contribute to farm data that later becomes a subject of proprietary control. African indigenous farmers face the challenges of inequality in access to data and of unfair utilisation of data. These challenges hold negative repercussions for African countries' development aspirations as proprietary control of data restricts the countries' ability to control the transborder flow of data. This article proposes the development of an Africa-wide data governance framework in which the challenges on access to data and unfairness in its utilisation are addressed in a manner consistent with the continent's aspirations for intra-regional relations.

To accomplish this objective, this extended article is structured as follows. Sections 2 and 3 set the background by discussing the phenomena of the data revolution and digital agriculture in Africa, respectively. The discussion in Section 2 creates an understanding of the 'data revolution' and its relation to development. Section 3 maps out the ecosystem of digital agriculture in Africa, identifying general trends, key players, types, and features of digitalisation of agriculture in Africa that form the cornerstone of data utilisation and governance. The discussion identifies aspects of digitalisation that are driven by the capabilities of mobile and network infrastructure on the one hand, and higher-level digitisation supported by data infrastructures capability, on the other. Section 4 identifies African indigenous farmers as originators of data, whereas Section 5 situates farm data as a constitutive element of traditional knowledge systems of agricultural production that is subjected to datafication.

Section 6 explores the challenges to African indigenous farmers in the face of the increased datafication of traditional agricultural systems. The challenges of access to data are outlined as resulting from technological barriers and due to exclusivity of proprietary control of data. Similarly, unequal utilisation of data is discussed as posing a challenge to the survival of traditional agricultural systems in light of the emergence of the data marketplace in which data are shared with and transferred to global actors. Given the exploitative aspects of such inequality in the utilisation of data, this section also analyses the implication of the data revolution for development. It highlights the increased shift of power to private corporations in the collection and processing of data and sheds light on development imperatives that necessitate better control of data flows.

In Section 7, predominant frameworks for the governance of farm data are discussed. The discussion demonstrates the insufficiency of a privacy framework to regulate access and control of farm data from the perspective of African indigenous farmers as data subjects. Strategies for collective management of farm data as data commons under open data and creative data licensing regimes and under an emerging framework of data philanthropy are also identified as providing a model of governance for data. Given the inadequacy of these frameworks and models to address the challenges identified, Section 8 proposes data justice as a conceptual framework for an Africa-wide governance of farm data. A data governance framework focused on instrumental perspective aims at controlling the impact of data irrespective of claims of rights underlying the data. Section 9 discusses how such perspective supports African countries' interest to data sovereignty through data localisation schemes. A distributive rights-based perspective to data justice addresses the challenges of inadequate access and unequal utilisation of data through recognition of rights and by defining such rights' contents. Section 10 outlines the basis for the recognition of African indigenous farmers as rights holders and elaborates how such rights are consistent with emerging personal data economy models and are necessary to ensure indigenous farmers' control of access to their data. Section 11 is the Conclusion.

Section 2: The Data Revolution and Development

According to the United Nations (UN), the 'data revolution' is a phenomenon that marks a unique departure from the past, when 'a relatively small volume of analog data was produced and made available through a limited number of channels', to the generation and flow of data from various sources and through different channels with a markedly different 'speed and frequency' (Letouze 2016: 8). Such flow of data is coupled with 'the rise in the number and variety of sources from which it emanates' (Letouze 2016: 8). In this context, the data revolution explains the vitality of 'big data' and 'small data' in a data-driven economy in which individuals and firms use data to create new goods and services and to solve complex problems (Aaronson 2019). It is noted that 'big data is revolutionising 21st century business without anybody knowing what it actually means' (Emerging Technology from the arXiv, 2013). Understanding the phenomenon of data revolution entails, therefore, a brief discussion of what 'big data' and 'small data' are, and of how the two are related.

There is presently no working definition of the term 'big data' (Hu 2015: 794). The classic definition of big data comes from a 2001 Gartner report that anchored the definition on several data-specific characteristics called the 'three Vs' of big data: volume, velocity, and variety (Laney 2001). The report proposed that volume refer to the amount of data, velocity to how rapidly data are produced, and variety to diversity of the data formats (Laney 2001). From a technological point of view, the 'three Vs' definition of big data is taken as 'high-volume, high-velocity and high-variety information assets that demand cost-effective, innovative forms of information processing for enhanced insight and decision making' (Richards and King 2014: 394). Later, the concept was expanded to include a fourth V, veracity, which refers to 'the level of reliability associated with certain types of data' that brings issues of trust and uncertainty with regards to data and the outcome of analysis of the data (Jung and Kim 2014: 54).

Big data is not to be understood merely in terms of size. According to Cukier and Mayer-Schoenberger, 'big data is also characterised by the ability to render into data many aspects of the world that have never been quantified before..."datafication''' (Cukier and Mayer-Schoenberger 2013: 29). Datafication is commonly understood as putting a phenomenon 'in a quantified format so it can be tabulated and analysed' (Cukier and Mayer-Schoenberger 2013: 78). Distinguished from digitisation, i.e. 'turning analogue information into computer readable format' (Gattiglia 2015: 115), datafication is a process in which data are standardised through systemic classification or categorisation to be aggregated, processed, and analysed computationally (Ambrose 2015). As an aspect of big data, datafication is manifest in a variety of forms and can also, but not always, be associated with sensors/actuators and with the Internet of Things (Ambrose 2015). The National Institute of Standards and Technology explains that big data is data which 'exceed(s) the capacity or capability of current or conventional methods and systems' and as such 'the notion of "big" is relative to the current standard of computation' (Emerging Technology from the arXiv, 2013). Similarly, the OECD notes that 'big data represents large and complex, often unstructured, datasets that are difficult to work with using conventional tools and techniques' (OECD 2016: 48). This description of big...