CGIAR Platform for Big Data in Agriculture The CGIAR Platform for Big Data in Agriculture is a cross-cutting program of the global CGIAR consortium of non-profit research institutes looking into virtually every aspect of food security spanning: genomics, breeding, agroecology, climate science, and the socioeconomic drivers and context of food systems change. The Platform tends to data standards and data sharing, digital innovation strategy and technology transfer, and research into the intersection of digital technologies and agricultural development in emerging regions. CGIAR is a global research partnership for a food secure future dedicated to reducing poverty, enhancing food and nutrition security, and improving natural resources. https://bigdata.cgiar.org/ Citation: Ng M; de Haan N; King B; Langan S. 2021. Promoting inclusivity and equity in information and communications technology for food, land, and water systems. CGIAR Platform for Big Data in Agriculture, Cali Colombia. 64 p. Avalaible at: https://hdl.handle.net/10568/115154 Corresponding author: Simon Langan, Director Digital Innovation, International Water Management Institute (IWMI) S.Langan@cgiar.org Some Rights Reserved. This work is licensed under a Creative Commons Attribution NonCommercial 4.0 International License (CC-BY-NC) https://creativecommons.org/licenses/by-nc/4.0/ © Copyright CIAT 2021. Some rights reserved. Design and layout: Ximena Hiles Photo credits: ©CIAT - ©CGIAR System Organization Photos available on Flickr https://www.flickr.com/photos/cgiarconsortium/ https://www.flickr.com/photos/ciat/ September 2021 Promoting inclusivity and equity in information and communications technology for food, land, and water systems By Michelle Ng1, Nicoline de Haan2, Brian King3, and Simon Langan4 “Technology is always a form of social knowledge, practices and products. It is the result of conflicts and compromises, the outcomes of which depend primarily on the distribution of power and resources between different groups in society.” Judy Wajcman, Feminism Confronts Technology 1 Research Analyst, International Water Management Institute (IWMI) 2 Director, CGIAR GENDER Platform, International Livestock Research Institute 3 Coordinator of the CGIAR Platform for Big Data in Agriculture 4 Director Digital Innovation, International Water Management Institute (IWMI) S.Langan@cgiar.org contents 1 EXECUTIVE SUMMARY 6 2 INTRODUCTION 9 3 WHAT ARE ICTS? 10 4 CONSIDERING INCLUSION AND EQUITY IN ICTS 13 5 LINKING DIGITAL DIVIDES AND DESIGN FRAMEWORKS 29 6 DESIGNING MORE INCLUSIVE, EQUITABLE ICTS FOR FOOD, LAND, AND WATER SYSTEMS 30 7 THE IMPORTANCE OF INCLUSION AND EQUITY WHEN DESIGNING ICTS FOR FOOD, LAND AND WATER SYSTEMS 47 8 CONCLUSION 49 9 REFERENCES 51 Executive Summary Food, land, and water systems underpin the health of societies and the environment, yet they are facing pressure from climate change, population growth, urbanization, and the overexploitation of natural resources. Information and Communication Technologies (ICTs) have the potential to support food, land, and water systems in response to these challenges. Despite the optimism surrounding ICT use in food, land, and water systems, these technologies are not currently being used to their full potential. Low technology adoption rates can, in part, be attributed to issues of inclusivity and equity, including people’s ability to access, use, benefit from, and produce ICTs for food, land, and water systems. Where someone lives within interlocking systems of oppression – such as gender, physical ability, age, and race – play a large role. A feminist approach encourages us to ask three questions about the inclusivity and equity of ICTs for food, land, and water systems: Who benefits from ICTs for food, land, and water systems? Who in this system is either overlooked or actively harmed? Digital divides research has probed the different factors that shape people’s ability to access (first-level digital divide), use (second-level digital divide), and benefit from (third-level digital divide) ICTs. Who is doing the work in designing and developing ICTs for food, land, and water systems and who is not? In addition to who benefits from ICTs for food, land, and water systems, we must also look at the methods of production of ICTs to understand and challenge how the social relationships in our physical world are being reproduced in the digital world. At an individual level, people from dominant groups are more likely to be represented in technical and executive roles at technology companies. It is also worth noting who – and what – is rendered invisible in innovation processes, which speaks to what types of labor are (de)valued: miners and factory workers, and low paid contract workers moderating social media content to name a few. At an 6 Promoting inclusivity and equity in information and communications technology for food, land, and water systems organizational level, institutions such as governments, companies, and universities are best able to harness the benefits of ICTs due to their financial resources, influence, and technical expertise. This means ICTs may be more frequently used for the interests of dominant groups and institutions than for supporting the needs, values, and aspirations of marginalized people. At the national level, many innovation narratives characterize the Global North as an innovation hub, with ICTs diffusing to “peripheral” regions in the Global South; however, it is important to avoid algorithmic and design colonization and recognize that rich digital cultures exist around the world. Whose goals are represented in ICTs for food, land, and water systems and whose are not? Within certain business models, the ICT creators’ goals may be misaligned with those of end users including marginalized groups. ICTs also tend to uphold the status quo (striving “for good”), rather than promote lasting social change (working “toward co-liberation”). Research into the digital divides has helped unveil many disparities in people’s ability to access, use, and benefit from ICTs for development. There are limitations, however, to only thinking about the inclusivity and equity of ICTs within the digital divides framework, which, for example, frames ICTs as deterministic and universal and also characterizes people as consumers of ICTs, rather than producers, critical thinkers, or moral agents. When many recommendations emerging from the digital divides framework are top-down, design can serve as a complementary bottom-up approach for promoting the inclusivity and equity of ICTs for food, land, and water systems – tailoring them, through end user participation and iterative approaches, to better fit people’s values, needs, and aspirations. Design had once been applied only to manufacturable, physical products, but it is now also applied to market opportunities and product strategy. Some of the most popular design methodologies, which are not mutually exclusive, include: design thinking, user-centered design, human-centered design, inclusive design, value-sensitive design, and design justice. Many creators of ICTs for food, land, and water systems focus on heuristic functioning1 and researchers’ perceptions of users’ needs, rather than seeking to engage users during the design process. However, the importance of usability for the adoption of ICTs for food, land, and water systems is well-documented; and user-centered design approaches are shown to improve ICTs’ usability. In particular, research into ICTs for development – including for food, land, and water systems – has shown that education level and literacy, gender, physical ability, age, race, and prior exposure to ICTs all shape people’s ICT design preferences. In light of challenges with the adoption of ICTs for food, land, and water systems, user-centered design is another tool technologists can use to improve their ICTs – particularly to add inclusivity and equity. Of course, people have the right to decide for themselves whether to use ICTs for food, land, and water systems. The goal is not to compel everyone to do so. Instead, the goal is to design ICTs that are as useful, usable, and inclusive as possible, so people can make an informed choice of whether to use them, rather than being excluded by the ICT’s design. To do this, design methodologies need to be adapted to varying 1 A heuristic function, also simply called a heuristic, is a function that ranks alternatives in search algorithms at each branching step, based on available information, in order to ultimately decide which branch to follow. 7 Promoting inclusivity and equity in information and communications technology for food, land, and water systems cultural contexts, address power asymmetries between technologists and users, and avoid being overly prescriptive. In addition, most design methodologies tailor ICTs to the desires of individual end users, rather than considering the wellbeing of society. What might more-than-human design look like? Lack of inclusivity and equity in ICTs for food, land, and water systems has serious implications. It can produce a positive feedback loop in which privileged people benefit from ICT use and become more privileged, while those marginalized people unable to access ICTs are left behind. Inclusive ICTs increase customer engagement, which enables developers to grow a larger customer base. They also promote innovation and differentiation and eliminate any costs that would be incurred were a developer add features to make the product inclusive after the tool or technology is designed. Inclusive ICTs yield better outcomes for all, because everyone’s needs change over time. The Law of Amplification states that ICTs amplify social forces. Innovators have the choice of amplifying inclusion or exclusion, equity or inequities. What social forces should ICTs for food, land, and water systems amplify? 8 Promoting inclusivity and equity in information and communications technology for food, land, and water systems Many governments, international organizations, and donors assert that there is potential for Information and Communication Technologies (ICTs) to improve the lives of people in rural, developing regions worldwide (UN 2015). The World Bank, for example, describes the ways in which “ICTs help businesses become more productive; people find jobs and greater opportunities; and governments deliver better public services to all.” (World Bank 2016). This trend extends to food, land, and water systems – common application areas for ICTs within the sustainable development context. Both agriculture and water underpin the health of sustainable societies and the environment, but these resources face overexploitation and damage from climate change, population growth, and urbanization. In response to these challenges, ICTs can be used to “increase crop yields, reduce food losses and make agricultural supply chains more efficient as well as improve food distribution and retail.” (FAO 2019) These technologies are also expected to support disaster management, weather forecasting, smart cities, land cover mapping, precision agriculture, seasonal forecasting, long- term trend analysis, and the water-energy-food nexus (Sun 2019). In the words of USAID, “the scale of these changes holds out the potential for another agricultural revolution.” (USAID 2018) Despite the optimism surrounding the use of ICTs in food, land, and water systems, they are not currently being used to their full potential. Many smallholder farmers are unable to access ICTs due to poor connectivity, an inconsistent power supply or device unavailability, or unaffordability. Furthermore, people who do have access to ICTs for agriculture do not always choose to use them. Of the 26 million unique registered users in Africa who are reached by at least one digital tool for agriculture, only 11 million remain engaged with at least one tool (CTA 2019). Equally concerning is the number of female digital agricultural solution users: 10% in Senegal, 17% in Ethiopia, 20% in Nigeria, 28% in Kenya, and 30% in Ghana. Low use rates are also reported for digital decision-support tools used by agricultural extension agents (Kragt 2014; Oyinbo 2020) and agronomic modeling (Prost 2012). These numbers suggest that ICTs for food, land, and water systems have room for improvement before people will be able to inclusively and equitably benefit from digitalization. This report aims to (1) unpack how issues related to inclusion and equity characterize the production and usage of ICTs for food, land, and water systems and (2) explore strategies for improving the inclusivity and equity of ICTs for food, land, and water systems. 9 Promoting inclusivity and equity in information and communications technology for food, land, and water systems ICTs are often conceived in terms of a singular technology. This might include a mobile phone, the Internet, blockchain, or virtual reality. But ICTs do not have to be constrained to this tech-centric perspective – and there is value in conceptualizing ICTs more holistically, beyond the technology itself. By engaging with the social, cultural, and organizational contexts in which ICTs operate, for example, technologists are more encouraged to reckon with issues pertaining to privacy, ethics, bias, uncertainty, and trust (Mahyar 2020). This broader conceptualization of ICTs also invites developers to see their value not only for technical innovation, but also for social innovation (Winters 2009). 3.1. Conceptualizing ICTs Several conceptualizations have emerged from different sectors. We provide three examples. 3.1.1. Socio-technical systems One conceptualization of ICTs, called a “socio-technical system,” arose within the context of organizational design. According to the Interaction Design Foundation, socio-technical systems “consider requirements spanning hardware, software, personal, and community aspects.” (Interaction Design Foundation 2020) In other words, this type of system encourages us to view computing through the lenses of the mechanical level of hardware, the information level of software, the cognitive level of the individual, and the economic, legal, political, and social lenses of the community in which a given group of technologies are used. Socio-technical systems emphasize how social and technical factors are interrelated, because each component of a socio-technical system can influence and be influenced by each of the others. When considered as socio-technical systems, ICTs become an interplay of emotion, values, abilities, logistics, and more — thus, decentering technology. 3.1.2. Digital ecosystem A second conceptualization of ICTs, called a “digital ecosystem,” is commonly applied by the private sector (BCG 2021; Deloitte 2017; McKinsey 2020). While the digital ecosystem does not necessarily decenter technology, it does shift the focus away from the technology itself to the networked connections between technologies, organizations, and people. 10 Promoting inclusivity and equity in information and communications technology for food, land, and water systems Within the international development sector, representatives from the United Nations Environment Program have made the case for adopting digital ecosystems thinking, albeit with a technology-oriented focus (Campbell 2019). USAID has also adopted rhetoric around digital ecosystems, which it defines as one that “comprises stakeholders, systems, and an enabling environment that together empower people and communities to use digital technology to access services, engage with each other, and pursue economic opportunities.” (USAID 2021) As part of its digital strategy, USAID is completing Digital Ecosystem Country Assessments (DECA) (USAID 2020a). In these, it looks at three important issues: (1) digital infrastructure, access and use (including digital divides), and last-mile connectivity; (2) digital society and governance (including cybersecurity), misinformation, and e-government, and; (3) the digital economy (including digital trade), digital talent pool, and the tech start-up environment (USAID 2020b). The United Nations Development Program has completed a similar assessment of six countries’ data ecosystems (UNDP 2017). Accenture, a multinational technology services company, advocates for the benefits of digital ecosystems for development, claiming that they “smash siloes” and “scale multipliers in development.” (Accenture 2017) 3.1.3. Distributed cognition Another conceptualization of ICTs, called “distributed cognition,” arose within the discipline of cognitive science. Distributed cognition considers how human cognition is off-loaded into the environment through social and technological means (Hollan 2000; Hutchins 2000). ICTs are often invented to perform some form of cognitive labor. Examples include the creation of Wikipedia for storing information, Microsoft Word for recording our ideas, and calculators for helping us do math. Thinking along these lines, Steve Jobs referred to computers as “bicycles for the mind” – helping us think more efficiently, just as bicycles help humans move more efficiently. At the heart of distributed cognition is the conviction that humans are relying on the Internet for gathering information and using calculators for calculations because they see value in delegating tasks we are bad at – e.g., memorizing lots of information, performing exact calculations quickly – to machines that are good at it, so we can focus on the tasks at which we excel – creativity and problem-solving. In this sense, ICTs are assistive tools, augmenting human abilities. The collaboration can be extremely synergistic. One of the world’s best chess players competed against a mediocre chess player; both were aided by computers that helped them model possible outcomes of each move. The strong human + machine + inferior process was beaten by the weak human + machine + superior process, which demonstrates the power of a collaborative process over the inherent abilities of either the human or the machine alone (Norman 2013). Distributed cognition calls attention to the relational interactions between humans and technology – thinking of ICTs as augmenting people’s abilities to achieve their goals, rather than developing technology for its own sake. 11 Promoting inclusivity and equity in information and communications technology for food, land, and water systems 3.2. ICTs for social change People believe in ICTs’ potential to promote sustainable development – including agriculture – to varying degrees. Four broad attitudes toward the role of ICTs in social change were noted and described in (Toyama 2015). 1 Techno-utopians believe ICTs can effectively solve social issues. This group of people say things such as, “If you want to liberate a society, just give them the Internet.” (That’s a direct quote from Wael Ghonim, an activist and then-Google Marketing Executive.) They tend to focus on the technology itself as solutions to problems. In a 2008 TED Talk about his One Laptop Per Child initiative, Nicholas Negroponte urged, “Think of it as inoculating children against ignorance. And think of the laptop as a vaccine.” 2 The skeptics caution society about how technology impacts equality, human cognition, the social fabric, human rights in repressive regimes, unintended consequences, and carbon emissions. The Internet, which supposedly has the power to liberate a society, according to techno-utopians, is just as likely be used by the state to identify protesters, censor free speech, and spread propaganda, according to skeptics. 3 Contextualists contend the impact of technology depends on context. Of course, this is true, but unfortunately, this theory is not very actionable when the primary conclusion is that such generalizations are unhelpful. 4 Social determinists believe the impacts of technology depend on the people wielding it — in particular the intent of the project leaders, implementers, and beneficiaries. 3.2.1. Social determinism and the Law of Amplification The term “Law of Amplification” describes how social infrastructure determines whether ICTs catalyze positive or negative change. Rather than automatically creating good, ICTs amplify the human factors already at play in any given context (Toyama 2015). An educational ICT introduced to a school with dedicated students and competent teachers can enhance learning in the classroom, but the same educational ICT introduced to a school with less disciplined students and less engaged teachers can distract from learning. In the words of Toyama, “The right people can work around a bad technology, but the wrong people will mess up even a good one.” He argues that, for ICT interventions to be successful at enacting social change, the project leaders, implementers, and beneficiaries must have good discernment, intention, and self-control. Consequently, a key component of an ICT intervention often involves nurturing the personal development of the project’s implementers and beneficiaries. The Law of Amplification begs the question: what social forces ought to be amplified by ICTs for food, land, and water systems? It speaks to the importance of placing an emphasis on inclusion and equity into technology design and development. 12 Promoting inclusivity and equity in information and communications technology for food, land, and water systems ICTs are increasingly serving as access points to economic opportunity, civic participation, and social interactions, and other resources. Food, land, and water systems are no exception. As of January 2020, GSMA had identified more than 700 ICTs for agriculture worldwide (GSMA 2020). These generally fall into five use cases: (1) digital agricultural advisory services; (2) digital agricultural financial services; (3) agricultural e-commerce; (4) digital procurement; and (5) smart farming. There are abundant examples of these technologies and tools. iCow is a digital agricultural advisory service in Kenya, Tanzania, and Ethiopia that sends farmers three SMS messages weekly in their local language about livestock feeding and disease management (iCow 2020). Knowledge about these topics had previously been the farmers’ biggest barrier to increasing production, according to iCow. FarmDrive is a digital agricultural financial service in Kenya that facilitates credit scoring and lending for smallholder farmers (FarmDrive 2021). Frubana is an agricultural e-commerce service in Brazil, Colombia, and Mexico that connects producers to buyers (Frubana 2021). AgriBuddy is a digital procurement service in Bangladesh, Cambodia, and India that improves smallholder farmers’ access to farm inputs, such as seeds and pesticides (AgriBuddy 2021). Hello Tractor is a smart farming platform that enables smallholder farmers in Nigeria to access tractor services, which they may have otherwise been unable to afford (Hello Tractor 2020). A huge challenge to ICTs for agriculture – and food, land, and water systems more widely – is that these technologies are not inclusive and they do not guarantee or provide equitable outcomes. As previously mentioned, the number of female users of digital agricultural solutions is less than a third of all users (CTA 2019). In addition to gender, education level and literacy, physical ability, age, race, and prior exposure to ICTs factor into an individual’s ability to access, use, and benefit from ICTs developed for food, land, and water systems use. 13 Promoting inclusivity and equity in information and communications technology for food, land, and water systems A feminist approach encourages us to question power in the matrix of domination by asking the following questions (D’Ignazio 2020): 1 Who benefits from ICTs for food, land, and water systems? Who is either overlooked or actively harmed? 2 Who is doing the work in designing and developing ICTs for food, land, and water systems and who is not? 3 Whose goals are represented in ICTs for food, land, and water systems and whose are not? 4.1. Who benefits from ICTs and who is either overlooked or actively harmed? Information about the end users of ICTs for food, land, and water systems is not evenly reported. As a result, this section contains examples that extend to sustainable development more widely. These insights into the promises and pitfalls of ICTs for development provide an opportunity to learn from others’ experiences and to think critically about the future of ICTs’ applications for use in food, land and, water systems. 4.1.1. Digital divides The term “digital divide” refers to disparities between who is able to access (first-level digital divide), use (second-level digital divide), and benefit from (third-level digital divide) ICTs. These tiers define who benefits from ICTs and who does not. 4.1.1.1. First-level digital divide The first-level digital divide describes infrastructure access to ICTs. This is often considered materially, that is, in relation to the Internet and the requisite software or hardware needed. Is the technology or tool broadly available and marketed in a given region? Can people afford it? Do users have the electricity to charge their device? Is their area served by cell towers? Can they access a repair shop for spare parts if something breaks? These considerations affect people disproportionately. At the individual level, for example, women in low- and middle-income countries are 8% less likely to own a mobile phone than men and are 20% less likely to use mobile Internet than men (GSMA 2020). In addition, Aikins (2019) found that Internet use in 37 African countries was influenced by the affordability of broadband services and devices, digital skills, age, and availability of mobile phone subscriptions. In other words, people who are younger and wealthier, possess digital skills, and own mobile phones are the demographic group most likely to access the Internet. At the household level, Mink (2019) found that the number of smartphones owned by households in the Bihar district of India and the Khulna and Chittagong districts of Bangladesh was linked with these individuals’ income and education level. Furthermore, Otioma (2019) found that in Kigali, Rwanda, the first-level digital divide is positively correlated with existing inequalities in urban infrastructure, urban agglomerative strength, and each household’s socioeconomic status. 14 Promoting inclusivity and equity in information and communications technology for food, land, and water systems At the country level, Cruz-Jesus (2018) compared the digital development of 45 countries, finding that income disparities and educational disparities were the biggest drivers of digital asymmetries. In this case, “digital development” was defined as a country’s ICT infrastructure, broadband connection, the number of mobile broadband subscriptions per 100 inhabitants, Internet speed, the number of Internet secure servers, and the portion of the population regularly using the Internet. This suggests that people residing in wealthier countries are more advantaged in ICT access than those residing in less wealthy countries. 4.1.1.2. Second-level digital divide The second-level digital divide assumes individuals have the infrastructure available to access ICTs. Instead, this tier examines users digital skills and use patterns. The “offline” qualities of end users – such as their demographic groups, economic, social, cultural, personal, material attributes, and motivation for accessing technology or digital tools – shapes their digital skills (Scheerder 2017). However, many studies define and measure digital skills differently, largely due to the constantly growing and evolving types of possible digital interactions. van Deursen (2019) describes operational Internet skills, formal Internet skills, information Internet skills, and strategic Internet skills, for instance, while Scheerder (2017) categorizes digital skills as being medium-related (e.g., word processing, file manipulation), content-related (e.g., communication, information, creativity), for ensuring safety and security (e.g., knowledge about online privacy), and for general purposes (e.g., Internet skills, digital literacy). Chetty (2018) proposes a framework that examines five types of digital literacy (information, computer, media, communication, and technology) against three perspectives (technical, cognitive, and ethical). The studies in our survey of the literature do not use standardized methods to assess digital skills and, thus, they cannot be compared. However, these studies demonstrate the large variation in digital skills among different people. Hargittai (2002) found that age is negatively associated with Internet skill and prior experience with ICTs is positively associated with Internet skill. Internet skill, in turn, affected whether and how quickly 54 American participants were able to find information online. In a study of agricultural extension agents using the Farmbook app in Madagascar, Malawi, Zambia, and Zimbabwe, Tata (2016) found that (1) female extension agents reported fewer technical challenges than male extension agents; (2) extension agents less than 35-years-old reported fewer technical problems compared to those who were aged 35 or greater; and (3) extension agents who possessed excellent Internet skills were more comfortable with Farmbook than those who were less able to use the Internet. In conducting a self-assessment of digital skills, Hargittai (2006) found that 100 men and women in the United States did not differ greatly in their online abilities, but women’s self-assessed skill levels were significantly lower than those of men. 15 Promoting inclusivity and equity in information and communications technology for food, land, and water systems People’s digital skills – and perceptions of their own digital skills – may affect their online behavior. Other factors, such as education level, class, and location—residing either in a rural or urban area, may also affect an individual’s online behavior. With regard to education level, Helsper (2009) found that less educated people in four European countries were less likely to use the Internet for learning or financial purposes than were more educated people. Similarly, van Deursen (2015) found that less educated people in the Netherlands were less likely to use the Internet for information or personal development than more educated people, but they were more likely to use the Internet for gaming and social interaction. With regard to class, Internet use among people in Nanjing, China did not significantly differ; however, specific Internet use patterns did differ. Online shopping and social networking were more attractive to lower class groups, while higher social class groups focused their Internet use on capital-enhancing activities (Chang 2016). With regard to differences in Internet use between rural and urban populations, a study of 2,592 high school students in Karnataka state, India, found the most popular Internet uses for rural students included playing games, seeing animal images, and watching sports. In contrast, the most popular Internet uses among urban students were completing school projects, playing games, and completing class assignments (Kumar 2018). 3.1.1.3. Third-level digital divide Even if people have access to ICTs (first-level digital divide) and adequate skills to use them (second-level digital divide), they may not benefit equally from using the ICTs, resulting in unequal returns from ICT use for economic, social, educational, political, and institutional activities (third-level digital divide) (van Deursen 2015). Regarding the real-world outcomes associated with general ICT use, DiMaggio (2008) found that Internet use was positively associated with income growth. Kuhn (2014) found that youth who used the Internet for job hunting between 2005 and 2008 reported that they were employed 25% faster than those who did not. Bhatnagar (2004) suggests that more digitally savvy online shoppers can find products and services at cheaper prices than those who are not. These examples speak to the direct benefits that can result from ICT use. Yet benefits differ amongst ICT users. Mobile phone use for business by female microentrepreneurs in Chennai, India, for example, was not a significant predictor of business growth, but, mobile phone use for business in addition to high entrepreneurial expectations – which were positively correlated with class, caste, and education level – was indeed a significant predictor of business growth (Chew 2011). Attewell (2011) found that the benefits of having a home computer in the United States were greater for (1) children from more affluent and educated families than those from poorer and less educated families; (2) boys than girls; and (3) white children compared to minority children. Teo (2011) found that the ownership and use of home computers in Singapore was linked with students’ computer self-efficacy, which had a significant impact on learning outcomes. Male students reported that they had a greater amount of home computer use for both studies and leisure activities than female students, as well as significantly better computer skills than female students. This contributed to improved learning outcomes for male students. Unfortunately, real-world user outcomes are rarely reported for ICTs for food, land, and water systems. CTA’s 2019 “Digitalisation of African Agriculture Report” was only able to unearth about 50 impact data points. These were primarily yield and income metrics. 16 Promoting inclusivity and equity in information and communications technology for food, land, and water systems 4.1.2. Standpoint theory People’s ability to access, use, and benefit from technology is largely shaped by where they reside within interlocking systems of oppression. Collins (1990) describes how black women occupy a certain intersectional standpoint within the matrix of domination: experiencing or resisting oppression at the individual, cultural, and institutional levels. This influences how a person engages with the world around them, including their ability to access, use, and benefit from ICTs. One’s standpoint also includes their level of ICT knowledge, which may affect how, when, and why they would like to use ICTs (Hartsock 1983; Haraway 1988). In other words, these factors shape which privileges, needs, values, and aspirations people have when considering using ICTs for food, land, and water systems. The standpoint theory challenges the universalizing narratives that often surround ICTs’ use in development. These status quo narratives contend that people will be able to escape from poverty through ICT use (Chan 2014). This assumption underpins a technology’s promise to deliver positive impact at scale; however, as we see from the actual use rates of ICTs for development, these technologies often fail to be adopted widely, suggesting that specific ICTs are not actually universally useful, usable, or inclusive. Thus, an awareness of standpoint theory is a useful starting point for considering the digital divide. Factors that influence people’s ability to access, use, and benefit from ICTs include their educational level and literacy, gender, physical ability, age, race, and prior exposure to ICTs. Three caveats exist: (1) these factors are context specific, meaning they do not consistently generate the same effects for everyone; (2) these are examples of the impacts of single-axis identities2, which may intersect to create different outcomes, and; (3) most research into the digital divides treats gender as binary. 4.1.3. The choice of ICT non-use It is worth noting that some people have the option to access, use, and benefit from ICTs, but choose not to use them. This decision is increasingly becoming a status symbol amongst elites, who have the luxury of limiting their online presence, engaging in a digital detox or shunning social media, and sending their children to Waldorf schools that ban electronics (Dobrinskaya 2019; Toyama 2015). It is also an increasingly important strategy adopted to resist the attention economy3 and reclaim one’s time (Odell 2019). In addition, some people may choose to avoid ICTs developed outside their local context based on “a desire to build an information society from the ‘bottom-up’, on a human scale, based on local circumstances so that it was citizen-oriented.” (Uotinen 2003) This especially makes sense as the rhetoric around ICTs and social change posits that value can come from anyone, anywhere. As ICTs offer new opportunities to extract value from labor, technologies are often used to “colonize the social as a search for investable novelty.” (Irani 2019) For many, whether to use technology is, above all, a question of safety. When technology can be used to surveil individuals and data can be used to oppress, opting out of ICT use may be the safest option. Recognizing and respecting an individual’s decision to opt out of ICT use – for whatever reason – supports this person’s right to self-determination. “Studying non-use can problematize this [digital imperative], calling into question the fundamental premise of both the value and the unavoidability of such technologies. In some ways, this critique may also apply to the narrative of the digital divide, that unequal distribution of technology creates haves and have-nots, and that the best way of ameliorating such inequalities is greater technological saturation and penetration. What if, however, those who do not use a technology do so not from a lack of opportunity but rather from a lack of desire? What if certain individuals or groups prefer to stay on the far side of the digital divide?” (Baumer 2015) 2 A single-axis framework treats race and gender as mutually exclusive categories of experience. 3 Attention economics is an approach to the management of information that treats human attention as a scarce commodity and applies economic theory to solve various information management problems. 17 Promoting inclusivity and equity in information and communications technology for food, land, and water systems 4.2. Who is doing the work in creating ICTs and who is not? In addition to who benefits from the creation and use of ICTs for food, land, and water systems, we must also look at the modes of production to understand and challenge how the social relationships in our physical world are being reproduced in the digital world. We can examine this question at multiple levels: the individual, the organizational, and national levels. 4.2.1. Individual level Our answer to the question of “Who is doing the work in creating ICTs for food, land, and water systems?” largely depends on our definition of innovation. 4.2.1.1. Schumpeterian model of innovation We often think of tech companies as sources of innovation because the dominant narrative of innovation – which can be traced to economist Joseph Schumpeter – requires commercialization. This traditional, Schumpeterian model of innovation views profit-driven producers (businesses) as the primary engine of economic change. It shaped the Organisation for Economic Co-operation and Development’s definition of innovation, which states that “A common feature of an innovation is that it must have been implemented. A new or improved product is implemented when it is introduced on the market.” (OECD 2005) In other words, a new product or service or an improvement on an old product or service only counts as an innovation if it is commercialized. This narrow definition of innovation restricts it to being undertaken by people selling products or services, contributing to the perception that innovations are created by an anointed few in the private sector, rather than viewing innovation as an emergent property of social systems. Within this paradigm of innovation, there are extreme gender imbalances in four of the largest tech companies worldwide: Table 1. Gender representation in four major technology companies. Company-wide Tech positions Executive positions % men % women % men % women % men % women Apple 67.0 33.0 77.0 20.0 81.3 18.7 Facebook 63.1 36.9 77.0 23.0 67.4 32.6 Google 68.0 32.0 76.4 23.6 73.3 26.7 Microsoft 72.4 27.6 78.6 21.4 80.7 19.3 Source: Apple 2020; Facebook 2019; Google 2020; Microsoft 2020. Note: The number of company-wide employees includes those who work in distribution centers and retail outlets in addition to those involved in software/product development. 18 Promoting inclusivity and equity in information and communications technology for food, land, and water systems Single-axis gender representation in four technology companies, all headquartered in the United States, is shown in Table 1. Within these companies, there are also discrepancies in representation by those of different races, people with a lower socio-economic status, people of all abilities, people who identify as LGBTQIA+, and of other similarly marginalized groups. These issues are not at all specific to these four companies or to the private sector, more generally. In American universities, more than 80% of artificial intelligence professors are white, while less than a third of computer science and statistics faculty are women (D’Ignazio 2020). What is the representation of gender, race, sexuality, caste, class, ability, religion, etc. amongst the producers of ICTs for food, land, and water systems? We do not know. There is a significant data gap because much of the research has focused on examining the ICTs themselves instead of the processes used to create them. Worldwide, almost 80% of all patents filed at key intellectual property offices in 2015 came from teams of only men; women represent less than one-third of all start-up founders; and an estimated 11% of start-up founders seeking venture capital backing in 2018 were female (OECD 2018). Of the global artificial intelligence workforce, 78% are male (World Economic Forum 2018). Among the membership of some select online data science platforms, 81.9% of contributors to DS Central are male, 83% of contributors to Kaggle are male, and 92.1% of contributors to Stack Overflow are male (Young 2021). How does the overrepresentation of some populations in innovation processes affect the products and services that are produced? Using gender as an example, in technical positions (i.e., software engineers), male overrepresentation suggests that men are making more design and development decisions about new technologies than females. This could mean programming their own biases into ICTs or creating user interfaces that are more intuitive and beneficial to them – and other men – than to women. In applying this theory to representation among tech firm executives by gender, Williams (2014) suggests that male overrepresentation means men are making disproportionately more business and strategic decisions about the future of technology. In practice, this might be tech employees customizing a demo to appeal to a male executive personally because they need the executive’s approval to continue their work, rather than including the needs, values, and aspirations of female end users. 19 Promoting inclusivity and equity in information and communications technology for food, land, and water systems While these examples focus on gender, the same could be said for the overrepresentation of white innovators, young innovators, able-bodied innovators, straight innovators, innovators from more privileged castes, and innovators from the Global North. Innovation teams need to be diverse to better account for the needs, values, and aspirations of a more inclusive user base. In the United States, tech culture has historically been notoriously inhospitable to females, people of color, LGBTQ+, physically challenged, and older workers. In discussions regarding allegations of inequity inside Google, for example, lawsuits regarding gender bias in pay and promotion, a systematic gender pay gap, sexual harassment, and harassment of diversity advocates will be brought up. In 2018, thousands of Google employees worldwide staged a walkout to protest sexual harassment when The New York Times revealed that Google had paid Andy Rubin, a high level executive, a $90 million exit package and praised him for his creation of the Android mobile software, after concluding that a sexual harassment claim against him was credible (Wakabayashi 2018). Another executive who faced credible claims of sexual harassment, Richard DeVaul, was allowed to keep his high-paying job. Also in 2018, Uber reached a US$10 million settlement agreement in a class-action lawsuit filed by its 483 female employees and employees of color in response to their claims of “incidents of discrimination, harassment, and/or hostile work environment and connecting their experiences to their race, national origin or gender.” (Kennedy 2018) In short, having the power to innovate – at least within American tech companies – is exclusive for two reasons: (1) to secure a job in these firms is predicated on the applicant having a high-quality education and the tech recruitment process provides advantages to middle- and upper-class, straight, able-bodied, young, white men; and (2) a workforce structured as described can create a culture that is unwelcoming, unfair or unsafe for those who identify differently. This creates a power asymmetry not only in who gets to innovate, but also for whom they are innovating. This exclusivity keeps the power to innovate within the hands of the privileged class, to the disadvantage of marginalized groups, thus maintaining the status quo. This theory exists within one definition of innovation. While marginalized groups may not be represented in tech companies (and, thus, counted as “innovators” within a Schumpeterian paradigm), they are still innovating. Alternative models of innovation are useful to better recognize and compensate informal innovators and to think through how to engage more marginalized people in developing ICTs for food, land, and water systems. 4.2.1.2. Grassroots/free models of innovation Alternative models for innovation assert that innovation can come from anyone, anywhere – not necessarily from the private sector actors as in the Schumpeterian innovation paradigm. Gupta (2016) states, “Minds on the margins are not marginal minds.” He started the Honey Bee Network to explore creativity and innovation in rural communities across India, seeking to connect formal and informal knowledge systems and to reduce the transaction costs of innovators, investors and entrepreneurs. The Honey Bee Network’s “doorstep approach” holds that “when somebody solves a problem ingeniously, it is the responsibility of outsiders – the state/market or civil society – to recognize these ideas first in situ, or where they occur.” In addition to acknowledging those innovations that occur on the margins, Gupta notes that the value derived from this type of innovation process can fundamentally differ from those found in commercial spaces. One entrepreneur refused to license a bullock-drawn pesticide sprayer, for example, because it was discovered it could adversely affect the animal’s skin – thus, prioritizing the wellbeing of animals over any potential profits. 20 Promoting inclusivity and equity in information and communications technology for food, land, and water systems Similarly, von Hippel (2016) contrasts the Schumpeterian model of innovation with the household sector innovation model. He defines “free innovation” as a novel product, service, or process developed by a consumer using their own funds and time, which can be acquired by anyone for free. Tens of millions of individuals in the household sector partake in innovation, which both improves social welfare and increases private sector profits. In fact, “about 50 percent to about 90 percent of major consumer innovations commercialized by producers were in fact initially developed by household sector innovators,” he said. Because free innovators are motivated by self-rewards rather than profit, they often pioneer new fields, while private sector producers – who are dependent upon general demand – wait to verify the commercial potential of markets. However, free innovation is rendered invisible and undervalued for two reasons. First, free innovation does not fall into the OECD’s definition of “innovation” which depends on the innovation being “introduced on the market.” (OECD 2005) This means that free innovation is not measured at all in government statistics about innovation, unless the innovation is commercialized by a producer, in which case the credit goes to the producer instead of the free innovator. Second, even if the OECD changed its definition of “innovation,” no transactions document the value of the investments made or any output created by free innovation. This makes it difficult to measure the scope and impact of any household sector free innovations. 4.2.1.3. Collaborative, community-led models of innovation The Schumpeterian model of innovation contends that businesses are the producers of innovation. The grassroots and free models of innovation hold that individuals in the household sector can produce innovations. Another approach to innovation is more collaborative and community-led. Rather than placing the focus on individual technical mastery or creation, this model prioritizes the formation of inclusive design spaces and partnerships between communities, researchers, designers, and technologists to create products and services that address community challenges (Costanza-Chock 2020). This model is more politically and socially engaged and places a greater emphasis on the real-world outcomes of innovations. This model of innovation is the cornerstone of design justice, which places equity at the center of its community-led design process and its outcomes (Costanza-Chock 2020). Escobar (2018) has also written extensively about a new model for design, which favors “more collaborative and place-based approaches” and “questions of environment, experience and politics” over pure capitalist production. Norman (2020), too, has called for “community-led design.” (Norman 2020). 21 Promoting inclusivity and equity in information and communications technology for food, land, and water systems Visions of collaborative, community-led innovation have led to the creation of public innovation spaces, such as those in the Fab Lab network (Fab Lab 2020). Fab labs or “fabrication laboratories” are workshops equipped with machinery and tools that enable people to pursue their own projects. They often offer training and actively reach out to their surrounding communities. Mz* Baltazar’s Lab in Vienna, Austria, for example, is a self-described “collective and feminist hackerspace” that seeks to ensure that makerspaces are more inclusive to counter the homogeny of most makerspaces. This particular fab lab states that it is “intended as a safer space for people who have traditionally been excluded from or have felt unsafe in spaces where science is taught, or technology is being used, and invite those people (e.g., women and trans individuals) to participate or give workshops that bring together technology, art, and have a critical understanding of social structures. Our exhibitions and events are open to all audiences, and are intended to support women in the broad sense of the political terms, and those who work on feminist issues, empowerment, and overturning patriarchy.” (Mz* Baltazar’s Lab 2020) 4.2.1.4. Who doesn’t count? Many informal innovation and contributions to innovation processes are never recognized, valued, or credited. Designers, software developers, and engineers usually receive credit for new products and services. Rarely acknowledged or fairly compensated, however, are the invisible laborers who make those same products and services possible. The essay “Anatomy of an AI System” seeks to rectify this by highlighting the extensive contributions, including human labor, required to produce an Amazon Echo (Crawford 2018). This work starts long before the designers, software developers, and engineers get involved – and often occurs far away from corporate offices and labs. Many people and processes build the foundation for manufacturing an Amazon Echo: miners excavating lithium in Bolivia, divers extracting tin in Indonesia, and Malay, Chinese, and Dayak workers collecting latex in Malaysia. Only the uppermost triangle of the pyramid of labor, however, is credited for the design and development of an Amazon Echo: software developers, marketing professionals, and Jeff Bezos. The labor that occurs at the bottom of this production pyramid is crucial, as are other types of emotional and affective labor4 distributed throughout the pyramid, but these workers are not typically considered part of the innovation prwocess. As a result, most of the material inputs and human labor that underpins ICT production is erased and, thus, under waged. With regard to material inputs specifically, ICTs exact a heavy toll on the environment. This issue has created some tension in using ICTs for sustainable development, when they are not sustainable themselves (van der Velden 2018). Technology devices, for example, negatively impact the environment as a result of the resource extraction needed to make them as well as the processes associated with manufacturing, the device’s energy consumption, and the end-of-life e-waste created (exacerbated by designed obsolescence). Strubell (2019) quantified the amount of CO2 emitted when training a Natural Language Processing model: 626,155 pounds of CO2. That is comparable to the lifetime emissions of five cars or the lifetime emissions of 59 humans. Belkhir (2018) estimated that the tech sector will contribute between 3.0% and 3.6% of global greenhouse gases by 2020 – comparable to that of the airline industry. The heavy environmental costs of ICTs are particularly problematic when this unsustainable ecosystem surrounding digitization is exported to low-income countries – the same people who are seen as being the primary beneficiaries of sustainable development. The labor of women – especially black women – has also been erased from the history of innovation, particularly in the United States. During World War II, the first computers were women (Light 1999) – “computer” was literally a job description. When NASA started to work with electronic computers in the 1950s, the job was considered “women’s work.” Thus, the first programmers were women. As described in the book and movie Hidden Figures: The American Dream and the Untold Story of the Black Women Mathematicians Who Helped Win the Space Race, black women at NASA played a critical role in sending 4 Affective labor is work carried out that is intended to produce or modify emotional experiences in people. This is in contrast to emotional labor, which is intended to produce or modify one’s own emotional experiences. 22 Promoting inclusivity and equity in information and communications technology for food, land, and water systems humans to space (Shetterly 2016). Data shows that women at NASA received lower pay and were less often promoted than male colleagues, and, when people think back to the heyday of the American space race, they tend to picture male engineers, male executives, and male astronauts. Furthermore, the challenges of social media moderation are becoming increasingly visible with much of the labor borne by low-income women in the Global South. Facebook, Twitter, and other social media sites have not historically automated content moderation, partially because they have not had great values or principles guiding their moderation. As a result, they initially outsourced content moderation to click workers, creating a shadow industry in the process (Dwoskin 2019). These click workers – often low-income women in the Philippines – manually review posts flagged by the online community, exposing themselves to violent content and hate speech. Many workers report having been traumatized after watching videos of torture, sexual assault, and other unspeakable acts, yet they receive little pay and are almost entirely unrecognized for their critical role in ensuring the safety of other social media users. Our current global model of innovation enables social media companies to receive profits and credit, while low-paid workers are rendered invisible and exploited, causing great personal harm by their exposure to extreme content. It is worth noting, however, that these companies are rapidly evolving their content moderation strategies. Invisible user labor is instrumental for the technical performance and business models of many ICTs. With regard to technical performance, user labor – such as selecting what to watch on Netflix or YouTube, or searching on Google – helps Netflix, YouTube, and Google refine their recommendation algorithms. With regard to business models, users need only browse their newsfeed to support the likes of Facebook and Twitter, since a user’s attention is monetized for ad revenue and this data can be sold to other third-party companies. To summarize, power imbalances shape who and what counts (or does not count) as an ICT innovator. “Showing the work,” said D’Ignazio, “is crucial to ensure that undervalued and invisible labor receives the credit it deserves, as well as to understand the true cost and planetary consequences of data work.” (D’Ignazio 2020) 4.2.2. Organizational level Cutting-edge ICTs are expensive to use, build, and maintain. As a result, the power to develop ICTs – especially those technologies with the potential to be the farthest-reaching and most powerful – is limited to those organizations with the most money and influence: namely governments, corporations, and universities, rather than individual farmers and local NGOs. This means that ICTs are more often used for surveillance, selling, and science, rather than developed for the benefit of ordinary people (D’Ignazio 2020). Corporations have unfettered access to our data every time we use the Internet, for example, whether we like it or not. Furthermore, those in control of the technology tend to be some of the most privileged members of society – which supports the prior discussion on how innovation is more accessible and welcoming for some than for others. 23 Promoting inclusivity and equity in information and communications technology for food, land, and water systems 4.2.3. National level While technologies and tools are created all over the world, the Global North often perceives itself as an innovation hub, with its innovations “diffusing” to “peripheral” regions in the Global South (Chan 2014). This Euro-centric perspective leads many ICTs that are created in the Global North to be exported to the Global South. Algorithmic colonization5 calls attention to the parallels between the desires of Western tech monopolies, which are motivated by their corporate agendas, and traditional colonizers, which are motivated by government forces, to “dominate, control, and influence social, political, and cultural discourse” in the Global South (Birhane 2020). In Africa specifically, Birhane describes how both tech companies and colonizers came to mine the continent for profit, whether in the form of customers, data, or natural resources. Furthermore, the rhetoric of Western tech monopolies is often reminiscent of the historic justifications given for colonialism and imperialism, evoking the specter of white savior-ism. They describe Africa as “data-poor” and, thus, promise to “connect the unconnected” (Gibbs 2017) and help “the next billion users” to “unlock economic opportunity.” (Google 2020) Yet rich digital cultures around the world boast their own innovations (Chan 2014; Dye 2018; Arora 2019; Irani 2019; Lindtner 2020). Narratives around ICT usage and innovation in economically less developed countries “unsettle the unspoken presumption that a single, universal narrative could adequately represent the distinct digital futures and imaginaries emerging from local sites today.” (Chan 2014) In short, widely acknowledging a variety of technological innovations – regardless of where they were developed – helps decenter the Global North in innovation narratives and credit more innovators worldwide. 4.3. Whose goals are represented in ICTs (and whose are not)? Lack of data makes it difficult to determine whose goals are specifically represented in ICTs for food, land, and water systems; however, we can look to technologies and tools in other sectors for examples of whose goals are often represented and extrapolate what this might imply for ICTs for food, land, and water systems. 4.3.1. ICT creators over end users Sometimes the goals of ICT creators are in tension with those of end users. This asymmetry manifests as privacy violations, corporate surveillance, and disinformation/misinformation campaigns, to name a few. In addition, many tech companies profit from user engagement by selling users’ attention to advertisers. They consequently design dark patterns into their ICTs with the goal of psychologically manipulating users to constantly engage, view more ads, and make them more money. Furthermore, the buzzword around ICTs is often “innovation.” The focus on creating new, bigger and better ICTs – rather than attend to the repair, maintenance, and care of existing ICTs – is not only patriarchal, but also may enhance the status of creators rather than support the technology’s utility to end users (Wajcman 2010). 4.3.2. The status quo over social change Technologies are often produced to maintain the status quo rather than to trigger or enact lasting social change. Corporate interests often produce new ICTs, for example, to encourage consumption “rather than production or critical analysis.” (Shade 2003; Flanagan 2009) These entities, thus, treat users as “consumers” rather than “producers,” “moral agents,” or “citizens.” 5 Algorithmic colonialism, driven by profit maximization at any cost, assumes that the human soul, behavior, and action is raw material free for the taking. 24 Promoting inclusivity and equity in information and communications technology for food, land, and water systems Furthermore, ICTs are often portrayed as neutral, value-free, and apolitical. Yet, it is useful to examine the ways in which this image of ICTs being “objective” has been constructed. Data visualizations are expected to be “clean” to highlight their objectivity – prioritizing reason over emotion, embodiment, or decoration, as recommended by Edward Tufte’s design principles (D’Ignazio 2020). With regard to creating new ICTs, design challenges typically scope problems so participants can skip to the solutions stage; however, these types of challenges are often stripped of discussions of power, equity, or politics, but are framed in terms of a vague “social good.” The deliberate separation between innovation and politics indulges designers and leads them to erroneously believe they are doing good for the world without creating any actual opportunities to meaningfully resist systems of oppression. Costanza-Chock (2020) questions, “How do institutions frame and scope ‘problems’ for designers to ‘solve’ in ways that systematically render structural inequality, history, and community resistance invisible?” When we think of innovation, many people conjure up images of hackathons, makerspaces, and innovation hubs. Yet, hack labs were originally explicitly politicized spaces, where activists and geeks joined forces. Only now have they been co-opted by universities, municipalities, and other entities as purportedly apolitical sites of neutral technology production. Design challenges facilitate the generation of solutions, but, typically, the winning solutions are only those solutions that fall within the acceptable limits of the status quo – drawing the attention and abilities of designers away from questioning the status quo itself. The resulting ICTs consequently tend to uphold the prevailing political, economic, and social systems and power structures. ICTs can be evaluated on the basis of their “ethics” or their “justice.” It is worth noting that justice can have different meanings in different contexts, e.g., global versus local justice (Sikor 2017). They can be created “for good” or “for co-liberation.” These differing goals have important implications. 25 Promoting inclusivity and equity in information and communications technology for food, land, and water systems Much of the work on attempting to improve the inclusivity and equity of ICTs focuses on reducing bias. While undoubtedly important, bias is but one type of algorithmic harm. A range of other harms can also be enacted by ICTs and their modes of production, including misinformation, political polarization/ radicalization, exploitation of workers, energy consumption, and surveillance. Is “unbiased” too narrow an ideal to precipitate a debate on equitable outcomes? The “reduce bias” or “increase fairness” approach places blame on individual technologies and suggests that cases of bias or exclusion are “isolated incidents.” The response is often a technical fix (Hoffman 2019). Fixing the technical problem does not address the underlying social problems that led to the introduction of bias in the technology. “Many of these ethical and technical approaches define the problem space very narrowly, neither contending with the historical or social context nor providing mechanisms for public accountability, oversight, and due process.” (Whittaker 2018) A researcher may find that a dialogue system trained on real-world conversations, for example, shows significant prejudices towards different genders and races. The researcher may recommend that the developers of the dialogue system reduce bias by cleaning their training data, or, in other words, removing biases from the real-world conversation data. While admirable that the authors are committed to reducing bias in dialogue systems, this technical fix fails to address why the biases exist in the real-world conversation data in the first place. Their recommendation is a Band-Aid. Adopting the suggestions may reduce bias in these particular dialogue systems, but there has been no recognition of or addressing of the root of this issue, which will continue to embed itself in other dialogue systems. Within this paradigm, ICTs may be designed to be less biased one-by-one, but a truly inclusive world – either physical or digital – is not possible with the technologies produced. The difference between “data ethics” and “data justice” is illustrated in Table 2 (D’Ignazio 2020). Table 2 DATA ETHICS DATA JUSTICE Concepts that secure power because they Concepts that challenge power because they locate the source of the problem in individuals acknowledge structural power differences or technical systems and work toward dismantling them Ethics Justice Fairness Equity Accountability Co-liberation Transparency Reflexivity Understanding algorithms Understanding history, culture, and context Similarly, the goal of creating “ICTs for good” introduces a power differential because the task is framed as being charity work, by which innovators seek to benefit “others.” (Irani 2019) It also raises the question “good for whom?” On the other hand, the goal of creating “ICTs for co-liberation” works to free both dom- inant groups and minority groups from oppressive systems. This distinction is illustrated in Table 3 (D’Ig- nazio 2020): 26 Promoting inclusivity and equity in information and communications technology for food, land, and water systems Table 3 DATA FOR GOOD DATA FOR CO-LIBERATION Leadership by members of minority groups working in a x community Money and resources managed by members of minority x groups Data owned and governed by the community x Quantitative data analysis “ground trothed” through a x participatory, community-centered data analysis process Data scientists are not rock stars and wizards, but rather x facilitators and guides Data education and knowledge transfer are part of the x project design Social infrastructure—community solidarity and shared x understanding— is built as part of the project design Of course, this is not to say that we should stop caring about ICTs’ ethics or ability to do good. Approaches for improving the inclusivity and equity of ICTs, however, should not make us lose sight of how we can also use them to promote inclusion and equity offline. Otherwise, technologies will merely uphold the status quo. 4.3.3. Dominant groups over marginalized groups ICTs tend to privilege dominant groups over marginalized groups for three reasons. There are many data gaps regarding marginalized groups. Considering gender, while some datasets fail to take gender into account, other data collection efforts do not adequately seek out female representation. Both situations render women “invisible.” (Criado-Perez 2019). The treatment of men as the “default human” – with women treated as “small men” (Goldberg 2003; Sims 2019) – results in the development of technologies and tools that better serve the needs of men than women. The desire to scale products provides an incentive for technologists to appeal to the average user and dismiss so-called edge cases (Wachter-Boettcher 2017). A/B testing is one common strategy that identifies and upholds the goals of the average user, who most likely belongs to a dominant group, and provides business justification for disregarding the preferences of people deemed edge cases, who most likely belong to marginalized groups. Perhaps an interface with a red button and a green button is well received by the majority of users, but causes confusion for people with red-green color blindness. Or, perhaps the majority of visitors to a website are men, so it saves them time if the default setting in a form about gender identity is “male,” but this creates extra work for female and gender non-binary visitors. The desire to appeal to average users shapes an ICT’s user base and determines who most benefits from a specific technology. The people creating the ICTs are likely be in the dominant group. As discussed in Section 4.2.1, this may mean that they are unable to recognize biases in their ICT, they are more likely to create user interfaces that are more intuitive to them or to customize pitches that appeal to their boss (who is even more likely to belong to a dominant group) than they would create a technology that would be useful for a diverse set of end users. 27 Promoting inclusivity and equity in information and communications technology for food, land, and water systems The inclusivity and equity of ICTs are being interrogated and improved by a number of approaches. 5.1. Within the digital divides framework The inclusivity of ICTs for food, land, and water systems is often discussed in terms of the digital divides: people’s ability to access, use and benefit from ICTs. This has led to both research and efforts toward bridging the digital divides. A rural telecenter program in Vietnam sought to connect scientific information with farmers (Gorannson 2016), and public libraries were found to be go-to places for people to access ICTs in South Africa and Nigeria (Lediga 2018; Ngozi Anasi 2018). Recommendations for bridging the digital divides are usually top-down. The suggested approaches may include reducing the cost of mobile data, providing digital skills training, or changing policies. There are limitations to solely thinking about the inclusivity and equity of technologies within a digital divides framework. For one, it frames ICTs as deterministic and universalistic. It does not ascribe people agency when calling for the transition from a “traditional” society to an “information” society – it makes this change seem inevitable, as “naturalized progress” without questioning whether people would like to use ICTs. The second-level digital divide, in particular, assumes that people do not use ICTs because they lack digital skills, rather than entertaining the possibility that the ICTs do not fit a user’s values, needs, or aspirations. The digital divides framework also creates a deficit narrative. It describes people in terms of their lack of access, rather than their assets or aspirations for ICT use. Because wealthy people will always be able to afford more ICTs and more advanced ICTs than poor people, this deficit narrative risks characterizing marginalized people indefinitely (Toyama 2015). The digital divides framework positions people as consumers of ICTs, rather than producers, critical thinkers, or moral agents. They exist to access, use, and benefit from ICTs when they become available, not to question them or make their own technologies. This creates power asymmetries in two types of one-directional ICT interactions: (1) the emphasis on ICT-to-user communication, rather than two-way communication (underplaying the value that can come from the users themselves), and (2) the emphasis on user-to-ICT attention, rather than any support being provided to users by the ICT (thus, ignoring the attention deserved or required by users to improve their experiences). Most importantly, the digital divides framework treats ICTs as fixed. The digital divides are described as “a social and political problem, not a technological one” (van Dijk 2005). While this made sense when the term “digital divide” only referred to Internet access in a binary “yes or no” fashion, there are now a range of devices, platforms, content, and standards each of which have the potential to leave people behind (Burrell 2018). This requires contending with the inclusivity and equity of each individual type of technology. 28 Promoting inclusivity and equity in information and communications technology for food, land, and water systems 5.2. The role of design The digital divides framework asserts that inequalities can manifest in people’s different abilities and means to access, use, and benefit from ICTs. Yet, inequalities can also manifest in biases directly embedded into the ICTs themselves. Understanding these inequalities requires examining the ICTs – treating them not as fixed, but as malleable objects. After all, they are designed. Commercial facial recognition software is 44 times more likely to correctly classify lighter-skinned males than darker-skinned women because it was trained on 78% male faces and 84% white faces (Buolamwini 2018). Someone who is non-literate or low-literate will likely be better served by a voice-based or pictorial interface than a text-based interface. An older person with vision loss might be better able to see and press bigger buttons on a user interface. Each of these are examples of the design choices that could be made by the creators of ICTs. To improve the inclusivity and equity of ICTs for food, land, and water systems, ICTs need to be considered as products that can be changed and altered rather than fixed in their current design. This questioning should begin at the project’s inception rather than later in the process and should include discussion on inclusivity and the equitable allocation of benefits and risks. If the design team is not diverse, its members may struggle to recognize how people unlike themselves may be harmed by particular design choices. There is also a perceived meritocracy in technology. People sometimes assume the smartest people are in the room, so they do not always feel the need to consult anyone else (Wachter-Boettcher 2017). ICTs are often created for the “average person” – in other words, the majority, rather than considering any possible “edge cases.” A/B testing is employed to determine the design preferred by the majority of users, which is then adopted to maximize end user satisfaction, even if some users were marginalized or harmed by this design element. Of course, examining the inclusivity and equity of technologies is not constrained to the technology itself. We should also consider technologies in social, institutional, and organizational contexts. Cooperative work is needed, for example, to integrate ICTs in communities (Christensen 2019; O’Neill 2017). There is a human infrastructure that supports community-run digital services (Jacobs 2020) and the business models for them. Businesses are generally less incentivized to target the most marginalized people because it is more difficult and costly to reach them. This returns us to the social and political problems of the digital divides model. In short, improving the inclusivity and equity of ICTs requires considering them holistically: their contexts, their modes of production, and the values encoded directly into ICTs. Design is a missing piece of the puzzle within the more popular discussion around the digital divides. It allows us to interrogate the inclusivity and equity of the tools and technologies themselves. 29 Promoting inclusivity and equity in information and communications technology for food, land, and water systems While many efforts emerging from the digital divides framework are top-down, design can serve as a complementary bottom-up approach for promoting the inclusivity and equity of ICTs for food, land, and water systems. It aims to improve ICTs themselves – tailoring them, through end user participation and iterative approaches, to better fit people’s values, needs, and aspirations. Design becomes an especially important component considering the relatively low adoption rates of ICTs for food, land, and water systems. Of the 26 million unique registered users in Africa who are reached by at least one digital tool for agriculture, only 11 million remain engaged with one of these (CTA 2019). In other words, people have access to ICTs for agriculture; however, they try them and conclude these tools are not worth their time. This suggests that the issue is not only access, but also the ICTs themselves. Of course, people have the right to decide for themselves whether to use ICTs for agriculture, which may address food, land, and water systems. The goal is not to compel everyone to do so. Instead, the goal is to create ICTs for agriculture that are as useful, usable, and inclusive as possible, so people can choose whether to use them, rather than being excluded by the design of the ICT itself. Since “design shapes our ability to access, participate in, and contribute to the world,” designing inclusive ICTs for agriculture empowers people to choose to access, participate, and contribute as they see fit (Holmes 2018). 6.1. What is design? Today’s design methodologies – including, but not limited to, design thinking, user-centered design/ human-centered design, inclusive design, value-sensitive design, and design justice – have their roots in industrial design (Kuang 2019). While it is difficult to trace the varied histories of design around the world, two turning points stand out. The first turning point was the transition from a focus on products, whether physical or digital, to a focus on users. The term “user-centered design” was coined by Rob Kling in 1977 and popularized by Don Norman in the 1980s. Norman’s book The Design of Everyday Things, in particular, articulated the importance of not only product functions, but also the communication of those functions to users through the use of signifiers, mapping, and feedback (Norman 2013). In other words, good design can remove friction in interactions between products and users by prioritizing the needs of the user and embedding intuitive mental models in them. Doing so successfully requires engaging with users during the design process. 30 Promoting inclusivity and equity in information and communications technology for food, land, and water systems The second turning point was the expansion of design from manufacturable, mechanical, and physical products to intangible items or concepts such as market opportunities and product strategy. In the early 2000s, the United States designed its own products and outsourced its manufacturing overseas, where there was cheaper labor (Irani 2018). Then, globalization policies and Chinese designers began to threaten the business of American design firms. The Stanford d.school and IDEO began developing the concept of design thinking to argue for the value of American creative industries. According to Irani, this involved aligning design with creativity, empathy, aesthetic judgments, and cosmopolitanism – asserting that right- brain thinking can uniquely generate insights that can open new markets. It also portrayed designers as people who can go from building rapport with consumers to rubbing shoulders with corporate executives. Following this line of reasoning, which evokes many class, racial, and gendered stereotypes, the qualities of good designers began to mirror those of wealthy, white men. In short, rather than competing, American businesses reinvented design to “protect North American difference” – claiming that left-brain tasks like math and programming could be outsourced to China and India, but design was a task better suited to creative, right-brain types: wealthy, white, American men. This shift devalued the labor of manufacturing workers, miners, industrial designers, etc. while elevating the empathetic, social, creative labor of design. These two turning points, among other factors, expanded the purview of design: from a focus on products to users, from an output of products to corporate strategies. This wider scope of design – as a core part of innovation – in turn, revolutionized how people think about social change. Ethnographic research at a design firm in New Delhi, India, for example, suggests that for the new generation of Indians, positive social change is perceived as occurring through innovation (Irani 2019). “Entrepreneurial citizenship,” a term coined by Irani, promises that citizens can simultaneously construct markets, produce value, and conduct nation-building work. The upside of this model is that innovation is intimately tied to social good. Entrepreneurial citizenship, thus, subsumes people’s hope for better worlds into capitalist production. This line of thinking is echoed in the rationale behind hackathons, which “orient toward Silicon Valley for models of social change” and “favor quick and forceful action with socially similar collaborators over the contestations of mass democracy or the slow construction of coalition across difference” (Irani 2015). In addition, as citizens begin to play greater roles in technology development, a new creative class (consisting of those who count themselves as innovators) govern, guide, and employ others, forming new social hierarchies (Irani 2019). A variety of design principles and traditions are employed across sectors and ideologies. The following six design methodologies, which are not mutually exclusive, serve as illustrative examples. 31 Promoting inclusivity and equity in information and communications technology for food, land, and water systems 6.1.1. Design thinking Design thinking was originally popularized by David Kelley – co-founder of IDEO and the Stanford d.school – as a process that could deliver packaged innovation to big businesses. Design thinking was also purported to hold the key to creativity by considering social issues. Design thinking has achieved a number of good things. Its simple, five-step process – (1) empathize with the user; (2) define the problem; (3) ideate solutions; (4) prototype solutions, and; (5) test and refine solutions – made creativity and innovation feel possible for anyone. It broke design out of its silo, expanded its scope from product to strategy, embedded design concepts within industry, non-profits, and government organizations, and demonstrated how design can be applied to important social issues. Perhaps most importantly, it asserted the value of listening to users, rather than relying on “rationalistic, impersonal, and quantitative forms of corporate knowing.” (Irani 2018) On the flip side, design thinking can be prescriptive. The history of design reveals the tension between focusing on problems versus process. Far from a methodlessness that creates an “epistemic freedom” for “a multiplicity of critical voices batting a problem around unknown terrain until it formed itself, or not, into some kind of resolution,” as once imagined by designer Horst Rittel, design thinking lays out a five-step process (Gram 2019). Yet, how can one method “fix” wicked or challenging social problems that are context- specific, ambiguous, and lack universal means of resolution? In addition, by embracing design thinking, we are embracing design thinkers. This often results in problems being “resolved” by inviting outsiders in, as opposed to employing internal consensus-building. In this way, design thinking threatens to devalue older, customary, local, blue-collar, unionized approaches to problem solving. 6.1.2. User-centered design Coined by Rob Kling in the late 1970s, user-centered design was popularized by Don Norman in the 1980s. According to Gould (1985), key principles of a user-centered approach to product usability are an early and sustained focus on engagement with users, empirical product use measurements, and an iterative development process that includes prototyping and testing. Sturm (2019) expounds, writing that “(User- centered design) enables the creation of useful and usable products by significantly involving users throughout an interactive software development process. The approach focuses on the users’ needs, wants and goals, as well as their perception and responses before, during and after using a product, system or service.” User-centered design employs a number of strategies at different stages in the development process to optimize products based on users’ preferences. While exploring user needs, a designer may engage in participant observation, interviews, diary studies, and/or experience mapping. The designer might synthesize and humanize user insights through employing personas. To refine a design, a designer creates a series of prototypes. These change form over the course of the design process. Prototypes may include paper prototypes, Wizard of Oz prototypes6, video prototypes, and mock-ups. Since each prototype tests a hypothesis about a design decision, it is usually most efficient to front-load the hardest questions about use and utility into the earliest prototypes. 6 Wizard of Oz prototyping is a design methodology used in rapid product development to improve the user experience. Once the prototype has been created, developers use role playing to test how end users will interact with the product. 32 Promoting inclusivity and equity in information and communications technology for food, land, and water systems To evaluate a prototype, a designer may plan usability studies (such as think-alouds, card-sorting, eye- tracking, first-click testing7, and task scenarios), surveys and interviews, focus group discussions, feedback from experts (such as peer critique, and dogfooding8), comparative experiments (also known as A/B testing9), participant observation, and simulation or formal models in an assessment. These evaluation methods could be moderated or unmoderated, take place remotely or in-person, and conducted for explorative, evaluative, or comparative purposes. Each method has its benefits and shortcomings. During interviews, for example, there is often a significant difference between what people self-report in a lab and how they actually engage with an ICT when they are not being observed. In general, quantitative metrics (including use data, error rates, task completion time, number of log-ins per week and session duration – especially over time) are helpful for understanding how people interact with a tool and qualitative data are helpful for contextualizing and understanding why they are using it. Some other factors to consider when selecting methods are their reliability, generalizability, realism, comparison, and the amount of time and work involved for participants. After testing each prototype, usability issues are often categorized as critical, serious, or minor. The results inform the design of the next iteration of the prototype. It is worth mentioning some limitations and critiques of user-centered design. Many of these information- finding activities are criticized for mining end users, extracting as many insights as possible for power/ knowledge (Foucault 1977). Under the guise of “improving user experience,” for example, A/B testing increases the decision-making power of product managers, who learn how to manipulate users and who have motives (i.e., ensuring a profit) that extend beyond customer satisfaction (Costanza-Chock 2020). This has the ripple effect of shaping the user base, which determines who reaps the benefits of the ICT. In addition, user-centered design assumes that users are unmarked (stripped of their intersectional identities) and somehow universal. In its constant averaging of data, it builds tools that privilege the majority of users – usually the dominant groups – at the expense of people on the margins, who are considered edge cases. 6.1.3. Human-centered design Human-centered design, which was popularized by IDEO.org, can be considered both a mindset and a methodology. It is frequently applied in the private sector and its use has started to gain traction in the international development sector as well. The standard ISO 9241-210: 2019 describes the main principles of human-centered design as: (1) understanding user needs and providing a use context; (2) involving users; (3) including evaluation by users; (4) iterating development solutions; (5) addressing the whole user experience, and; (6) employing a multi-disciplinary design team (ISO 2019). Human-centered design advocates for a design process that starts with people, addresses a problem, and designs with, not for, end users. There are three phases of human-centered design: inspiration, ideation, and implementation. It also requires thinking about what is at the intersection of desirability, feasibility, and viability. One distinction between human-centered design and user-centered design is, obviously, the name. By calling people “humans” instead of “users,” human-centered designers aim to encourage empathy by recognizing humanity, rather than treating people as mere parts within the technical system (Baron-Cohen 2011). Recognizing the humanity of users also reminds designers to examine a user’s needs in a broader context. It is one thing to examine the needs of ICT users, which will likely result in the development of a technical solution to meet that need. It is another thing all together to examine the needs of humans, for whom a solution to a particular challenge may not be an ICT at all. 7 First Click Testing examines what a test participant would click on first on an interface to complete an intended task 8 “Dogfooding” refers to the use of a newly developed product or service by a company’s staff to test it before it is made available to customers. 9 After A/B testing digital agricultural extension services in Ethiopia, for example, the share of users who accessed agricultural content increased from 52% to 63% (Kremer 2020). 33 Promoting inclusivity and equity in information and communications technology for food, land, and water systems 6.1.4. Inclusive design In User Friendly: How the Hidden Rules of Design Are Changing the Ways We Live, Work, and Play, Cliff Kuang writes, “The commonalities in design that technology has been driving toward, in an effort to make things easy to use, have finally run aground on the truth that we’re not all the same person.” (Kuang 2019) Inclusive design was developed around the kernel of this very idea. It enables and draws on the full range of human diversity, with principles that include: (1) recognizing exclusion; (2) learning from diversity, and; (3) solving for one, extending to many (Microsoft 2020). In particular, inclusive design emphasizes a one-size-fits-one approach. Rather than trying to create one design that works for everyone, it produces designs that can be used in a diversity of ways by a diversity of people. “The signature trait of an inclusive solution is how it adapts to fit each unique person,” wrote Holmes (2018). The United Kingdom’s National Autistic Society’s website, for example, has an option for visitors to toggle between “vivid” and “calm” color modes (National Autistic Society 2020). Personalization serves as an important part of inclusive design, with the automation of personalized interfaces considered to have some potential (Findlater 2004; Kuang 2019). 6.1.5. Value-sensitive design Value-sensitive design is a design methodology that makes the wellbeing of humans and the natural world the most prominent or important feature (Yoo 2013; Friedman 2019). The concept was developed by Batya Friedman, who defines values as “what is important to people in their lives, with a focus on ethics and morality.” Her premise rests on the fact that values are embedded within the features of a tool or ICT. A cookie embedded on a webpage, for example, values the convenience of personalization over privacy. Cell phones value 24/7 access to family, friends, and work over being left alone. Parenting software that tracks teenagers’ whereabouts values child safety and adult responsibility over teenagers’ autonomy and independence. Electronics that are designed for obsolescence value profit and a growing economy over product durability and environmental sustainability. Note that value-sensitive design itself values universal applicability. It is technology-agnostic and value- agnostic so designers can apply it to any domain and can use it with stakeholders who hold any values. Stakeholders can be direct (end users) or indirect (people affected by the technology), including non- human actors such as heritage buildings, mountains, and non-human species. Value-sensitive design has already been applied to domains as diverse as care robots in health settings (van Wynsberghe 2013), empowerment and marginalization in crowd-work (Deng 2016), cross-cultural design with a transnational population, (Alsheri 2019) and in a transition to solar energy (Mok 2018). While value-sensitive design places the values of the most affected stakeholders at the center of the process, value-sensitive design is sometimes critiqued for not taking a stronger stance on certain values, such as justice and sustainability. It examines what values stakeholders hold, as opposed to what values stakeholders ought to hold. Nevertheless, value-sensitive design serves as an example of the types of design methodologies that might become more common in light of the growing pushback against harmful technologies, as evidenced by movements such as #TechWontBuildIt. The current modus operandi of design is to optimize for content and engagement – in other words, revenue. In the future, more attention might be paid to methodologies that design for the wellbeing of users, societies, and the environment, rather than for greater page views and longer session durations. 34 Promoting inclusivity and equity in information and communications technology for food, land, and water systems 6.1.6. Design justice Design justice is a design methodology that places those people who are often marginalized in the process of designing new and just ICTs to address social challenges at the center of the process (Costanza-Chock 2020). A group of 30 people gathered at the Allied Media Conference in Detroit in 2015 to generate its shared principles. This led to the finalization of the principles in 2016, the formalization of the Design Justice Network in 2018, and the publication of Design Justice: Community-Led Practices to Build the World We Need in 2020. The key principles are as follows (Design Justice Network 2018): 1. We use design to sustain, heal, and empower our communities, as well as to seek liberation from exploitative and oppressive systems. 2. We center the voices of those who are directly impacted by the outcomes of the design process. 3. We prioritize design’s impact on the community over the intentions of the designer. 4. We view change as emergent from an accountable, accessible, and collaborative process, rather than as a point at the end of a process. 5. We see the role of the designer as a facilitator rather than an expert. 6. We believe that everyone is an expert based on their own lived experience and that we all have unique and brilliant contributions to bring to a design process. 7. We share design knowledge and tools with our communities. 8. We work towards sustainable, community-led, and -controlled outcomes. 9. We work towards non-exploitative solutions that reconnect us to the earth and to each other. 10. Before seeking new design solutions, we look for what is already working at the community level. We honor and uplift traditional, indigenous, and local knowledge and practices. Design justice embeds inclusivity, the goal of co-liberation, and reflexivity about designers’ power in its design process, particularly when considering the relationship between the designer(s) and the affected community. It encourages designers to think about who has power in the design process and what bene- fits or harms technology use may bring to users as well as who stands to benefit or be harmed by its use. 6.2. Inclusive design considerations of ICTs for food, land and water systems As mentioned previously, standpoint shapes how people are able to access, use, and benefit from ICTs. This includes their ability to engage with different designs. Five factors – education level and literacy, gender, physical ability, age, race, and prior exposure to ICTs – influence people’s preferences for the designs of ICTs. While described separately here, many of these factors overlap and intersect in practice. These design considerations are important to take into account when designing ICTs for food, land, and water systems, in particular, due to the diverse target audience for these technologies. 35 Promoting inclusivity and equity in information and communications technology for food, land, and water systems While the provided examples are from areas other than food, land, and water systems, they offer us the opportunity to learn from other experiences as the digitalization of food, land, and water systems progresses, ideally inclusively and equitably. 6.2.1. Education level and literacy Neuroplasticity describes how brains are malleable, not fixed. Brains differ and alter in both structure and physiology in response to how we use them (Paul 1972; Elbert 1995; Pascual-Leone 1995; Sadato 1996; Maguire 2000; Doidge 2007). Attending school and learning to read have profound impacts on the brain (Ostrosky-Solis 2004; Wolf 2007), resulting in differences between how literate and non-literate people understand language, process visual signals, reason, and form memories (Carr 2010). The impact of literacy is reflected in different spheres of cognitive functioning, and learning to read reinforces and modifies certain fundamental abilities, such as verbal and visual memory, phonological awareness, and visuospatial and visuomotor skills. (Ardila 2010). Ardila wrote that “Functional imaging studies are now demonstrating that literacy and education influence the pathways used by the brain for problem-solving. The existence of partially specific neuronal networks as a probable consequence of the literacy level supports the hypothesis that education impacts not only the individual’s day-to-day strategies, but also the brain networks.” This, in turn, results in differences between how literate and non-literate people interact with ICTs. Understandably, non-literate people and those with lower rates of literacy struggle to interact with text- heavy user interfaces. This leads people with less literacy to avoid the more complex functions of phones, such as texting or engaging with apps – using them predominantly to make voice calls (Medhi Thies 2015). A lack of education and/or literacy has effects that span beyond an individual’s ability to read. A study of 400 low-literacy, low-income people in India, the Philippines, and South Africa identified more nuanced issues that mediate how users interact with computing technologies. These included cognitive difficulties, collaboration, cultural etiquette, experience and exposure, intimidation, mediation, motivation, pricing, power relationships, and social standing, among other factors (Medhi 2010). An example of such a cognitive difficulty is the degree to which non-literate people have more trouble naming two-dimensional representations (such as in drawings or photographs) of common objects compared to literate people (Reis 2001) and how those people who are non-literate or who have a low-level of literacy have more trouble navigating hierarchical menus (Medhi 2010). An example of cultural etiquette as a factor would be that users without any formal education rarely criticize technology projects – even in participatory settings and even if the systems are intentionally designed to be obviously bad, which results in a survey with participant response bias (Dell 2012). Intimidation, in this context, occurs when people with a low literacy rate and who have low incomes experience a greater degree of discomfort when using a technology (e.g., a cellular phone) that looks more expensive than one they might be able to afford (Medhi 2010). An example of “experience and exposure” would be the difficulty that people with low literacy levels experience when interacting with touch screens. This may be, according to researchers, from a lack of fine motor skills, which can be developed by writing with a pen or pencil (Katre 2010). The preferences of people who have varying levels of education and literacy must be taken into account in the design of ICTs if these new technologies are to be accessible to these populations. A few relevant areas of research, so far, have evaluated the inclusion of pictorial interfaces, conversational and voice-activated interfaces, linear menus, and icons designed to reflect local meaning in making technology more accessible (Pejovic 2019). 36 Promoting inclusivity and equity in information and communications technology for food, land, and water systems 6.2.2. Gender Throughout the history of industrial design, the implicit assumption has been that the default human user is male. Standard hand tools, such as wrenches, are the perfect size for a man’s hand, but are too big for a woman’s hand to grip tightly. An A1 architect portfolio fits nicely under most men’s arms, while most women’s arms can’t reach around it. Personal protective equipment – from eye masks to body armor – is reported by 95% of women surveyed to be so ill-fitting that it hampers their work (Criado-Perez 2019). The lack of consideration for women in design has important implications. Not only for women’s ability to participate in and have a sense of belonging in the world, but also for their safety. Women have historically been considered “small men” in industrial design (Goldberg 2003; Sims 2019), despite anatomical differences that extend beyond size, such as their muscle mass distribution, bone density, vertebrae spacing, and their potential for pregnancy. Since the practices of industrial design laid the foundation for other design disciplines, this failure to design for women now persists in not only the design of the physical world, but also in the design of the virtual world, including software and hardware development. Smartphones are too big for a woman’s hand, for example, and smart watches are too big for a woman’s wrist. Also, speech emotion recognition models are more likely to accurately recognize male speech samples than they are female speech samples (Gorrestieta 2019). Beyond anatomical differences, research shows that different genders perceive and interact with ICTs differently. While gender exists across a spectrum, most of the following studies examine the behaviors of women and men as binary. Motivations for using technology. Men and women tend to use ICTs for different reasons – and the motivational divergence is discernable from a young age. As early as third grade, boys view computers as toys for play and girls view computers as tools for accomplishing a task (Cassell 2002). When asked why he was attending an inner city after-school computer program, one boy responded, “It’s fun. I mean, there are all these computers for me to play with.” When asked the same question, one girl responded, “Well, I really think this is a good opportunity for me to better my situation in life and I believe that I can get a better job if I know how to use a computer.” These attitudes seem to persist through adulthood, with implications for the types of ICTs created for boys and girls. “When educators with software design experience were asked to design software specifically for boys or for girls,” Cassell wrote, “they tended to design learning tools for the girls and games for the boys. When they were asked to design software for generic ‘students,’ they again designed games — the type of software that they had designed for boys.” 37 Promoting inclusivity and equity in information and communications technology for food, land, and water systems Cognitive styles. Several cognitive styles are linked to gender, which shape how men and women engage with ICTs (Burnett 2016; Vorvoreanu 2019): Confidence about using unfamiliar technology. Women tend to rate their confidence when using new computers or technologies lower compared to that of their overall peer group and lower than men tend to rank their own confidence level. Attitude toward risk when using technology. Women tend to be more risk-adverse when using ICTs than men, who are more risk-tolerant. Information processing styles. Women tend to prefer gathering information comprehensively to solve problems, whereas men gather information selectively. Learning styles for new technology. Women tend to prefer process-oriented learning, whereas men tend to prefer learning by tinkering--sometimes to excess. Attitudes toward technology and features. The differences between women and men’s cognitive styles also manifest as differing attitudes toward ICTs and digital features. In a study of 3,000 programmers (Burnett 2010): Women reported using fewer applications (feature sets) at work than men. They are more feature conservative, agreeing more with survey statements such as “I only learn the technology I have to know to perform my duties.” Women were more enthusiastic about avatars that provide tutorials or assistance within a piece or suite of software than men. This could stem from women’s inclination toward a process-oriented learning style, whereas men indicated more interest in tinkering and independently exploring new features. Women were less enthusiastic than males about piloting new technologies, agreeing less with statements like “I enjoy piloting/dogfooding next-generation technology.” Women were neutral in their responses about learning technology only if needed, but men tended to disagree with this statement. Notably, among male and female hobbyists, who by definition enjoy exploring technologies for fun, women request more beginner-level features and starter kits than do men. Overall, the female programmers were less confident than male programmers in their expertise. This has a profound impact on people’s engagement with ICTs, since confidence is significantly correlated with one’s propensity to explore new technologies. Note that this was a study of programmers – demonstrating that gendered differences in attitudes toward ICTs exist not only among users, but also among technologists. Preferences for user interface. Men and women may have different preferences for user interfaces. Female and male designers produce different-looking user interfaces, for example. Female and male end users tend to prefer designs created by designers of their own gender (Moss 2006). This speaks to the importance of female representation in user interface design, when most digital tools are created by men. 38 Promoting inclusivity and equity in information and communications technology for food, land, and water systems Language. Female designers tended to be more conversational and informal in their use of language than male designers, who used more professional and assertive language. Color. Female designers tended to use more colors and brighter colors, whereas male designers gravitated toward fewer colors and favored using grayscale. Layout and data structure. Female designers tended to create organic-looking, amorphous shapes that fill the screen, whereas male designers employed more rigid, sharp edges in a condensed layout. Online and mobile behaviors. Women and men exhibit difference in their online behavior. Maudlin (2020) used eye-tracking to compare how male and female foresters in the United States engaged with a climate decision-support system. She found that male visual fixation patterns were focused on the center map in the application, while female visual fixation patterns focused predominantly on a variety of the tool’s features, such as dropdown menus and map layers, rather than the center map. Before the study started, Maudlin found men were more likely to look at the map; women were more likely to read the text on the application’s homepage. On Pinterest, for example, one study found women were more likely to post reciprocal social links than men. They also tended to curate generalist content, while men tended to curate specialist content (Ottoni 2013). Another example of these differences in interactions, within one recipe website, a greater percentage of female users were found to scroll up and down the page, browse through different categories of food recipes at the top of the screen, and, on average, visit six recipe pages or more (Margalit 2014). This suggests that women visiting the recipe site were open to exploration and were patient in searching. On the other hand, men visited a maximum of three pages, concentrated on the ingredients in a given recipe and how to prepare the dish. They also tended to read more comments from other visitors. This suggests that men visiting the recipe site had a very focused intent and perhaps had trouble interpreting the recipe. In examining mobile device interactions, a study of the gender gaps in engagement with smartphone features in India found that men and women differ in the ways they make calls, receive calls, use SMS, access the Internet, use social media, enjoy entertainment, and use applications (Barboni 2018). In short, women and men may interact with and experience ICTs differently due to different physical, cognitive, sensory, and societal factors. Because men are sometimes viewed as the “default human,” women’s preferences risk being marginalized by ICT design. 6.2.3. Physical ability It is also important to take into account the preferences of people of all abilities during the design process. Most video game controllers, for example, require two hands to play. Using the small keyboard and buttons on a touchscreen requires an individual to have good fine motor skills, which are almost universally lost with age. A data visualization featuring reds and greens requires people to have trichromatic vision, when approximately 8% of Caucasian men and 0.5% of Caucasian women experience some form of color perception deficiency. The social-relational model of disability distinguishes between “impairment” and “disability.” (Oliver 1990) Impairment refers to the attributes of a person’s body (similar to the medical model of disability). Disability refers to a mismatch between the design of the environment and the needs of a person with an impairment (Holmes 2018). Bad design “actively disables people through unnecessary exclusion.” (Costanza-Chock 2020) 39 Promoting inclusivity and equity in information and communications technology for food, land, and water systems The tireless advocacy of many disability activists has led to the creation of some standards to address these issues, such as the Web Content Accessibility Guidelines 2.1 (Web Content Accessibility Guidelines 2018). ICTs are required to meet – but will hopefully go beyond – these standards, ensuring that websites can be used by people of all abilities. Inclusive design advocates for a one-size-fits-one (not one-size-fits-all) approach. A truly inclusive design, thus, accommodates many different means of engaging with a technology for many different end users. 6.2.4. Age Age shapes how people interact with ICTs in numerous ways. First, there are motivational differences. Younger agricultural extension agents, for example, were more eager to learn about the Farmbook application than were older extension agents (Tata 2016). For youth engaging in the digital economy, intrinsic motivations include passion, enjoyment, creativity, self-expression, meaning, progress, and skill development, while extrinsic motivations include financial and social rewards (Lombana-Bermudez 2020). There are also age-based differences in digital skills. Younger people have been found to demonstrate greater digital skills proficiency than older people (Hargittai 2002; Soomro 2020). Some attribute this finding to the increased difficulty of learning how to use ICTs when one is older. Others suggest that older people may have started using ICTs only recently and, because late adopters are often older, they have had less time in which to practice their digital skills. There are use discrepancies also based on age. The larger the population of people between the ages of 15 and 64, the greater the percentage of African households who use the Internet (Aikins 2019). In Nanjing (China), Internet usage was similarly linked to age (Chang 2016). Younger people tended to use the Internet more than older people. Youth also tend to use the Internet in a greater variety of ways. ICTs offer new capital-enhancing opportunities. From video blogging to music production to art creation to podcasting to coding to gaming, youth are more frequently economic actors in networked, online communities (Lombana-Bermudez 2020). These “prosumers” not only consume digital content, but also produce it (Toffler 1980). Their user-generated content creates a new form of capitalism – reoriented away from scarcity and toward abundance, away from paid labor and toward offering products at no cost (Ritzer 2010). And far from launching their content into the void, youth prosumers curate others’ content and also form relationships with communities of influencers and followers. Design considerations related to age are largely linked with physical ability. We will all experience vision loss, hearing loss, and loss of motor control as we age. Worsening vision may require larger font sizes for older users just as hand tremors may require larger buttons and forgiving, clear methods for error correction. Older users may also benefit from training when introduced to new technologies. 6.2.5. Race There are countless examples of how ICTs uphold white supremacy – coined “the New Jim Code.” (Benjamin 2019) As mentioned previously, Buolawmwini (2018) demonstrated how commercial facial recognition software is predominantly trained on male (78%) and white (84%) faces. This is one of many racial biases encoded into facial recognition software that researchers have found. Buolawmwini also found lighter- skinned males are up to 44 times more likely to be correctly classified than darker-skinned women. States and companies may use and benefit from facial recognition software, but this technology has the potential to misidentify suspects, prevent citizens from getting passports or crossing borders, or prevent certain people from unlocking their own phones. In light of Buolamwini’s research and the Black Lives Matter movement in the United States, IBM, Amazon, and Microsoft have all announced some temporary rollbacks in deploying their facial recognition software (Kaye 2020). 40 Promoting inclusivity and equity in information and communications technology for food, land, and water systems While this report primarily focuses on end users (direct stakeholders), ICTs also have indirect stakeholders: people who do not use the ICTs themselves, but who are affected none the less. Northpointe’s Correctional Offender Management Profiling for Alternative Sanctions (COMPAS) algorithm, for example, was used in the United States criminal justice system to predict recidivism, which is the likelihood of a defendant committing another crime. Unfortunately, “black defendants were far more likely than white defendants to be incorrectly judged to be at a higher risk of recidivism, while white defendants were more likely than black defendants to be incorrectly flagged at low risk.” (Larson 2016) Noble (2018a; 2018b) writes about an ICT that hits much closer to home: Google. In 2009, Noble typed “black girls” into Google’s search engine. The first page of results was rife with sexually explicit content, including a pornography site that was returned as the top hit. In 2013, she found similar trends when using the search terms “Latina girls” and “Asian girls,” but not “white girls.” Noble describes how Google – among other search engines, which control representation – are co-opting Black identities, reinforcing both racism and sexism (Noble 2018a). Google has since tweaked its search algorithm, but search results continue to perpetuate stereotypes. In 2016, the search for “three black teenagers” returned mugshots whereas the search for “three white teenagers” returned wholesome, all-American photos (Noble 2018b). In an ethnographic study of three American schools, digital skills were perceived and rewarded differently by teachers (Rafalow 2018). At a private school serving primarily wealthy white youth, teachers applauded students’ digital work, translating their digital skills into cultural capital and treating these skills as essential for success. At a public school serving mainly middle-class Asian youth, teachers viewed students’ digital skills as a threat to their school, citing concerns of hacking and cheating. They introduced surveillance software and monitored students’ digital connections – even disciplining students based on their text messages. At a public school serving mainly working-class Latinx youth, teachers perceived digital skills as irrelevant to school, instead prioritizing basic skills that could get them factory jobs. Needless to say, these different responses to digital use have profound consequences for social mobility. While these examples mostly describe white supremacy, ICTs can encode the biases of racial hierarchies against any marginalized race. For instance, “people of color” is a common term in the United States, but there are different racial hierarchies in other countries and contexts, which may not define race in relationship to whiteness at all. 6.2.6 Prior exposure to ICTs People without prior exposure to ICTs have a steeper learning curve when adopting a new technology. They might not, for example, recognize visual symbols such as the “play” or “pause” buttons commonly used on videos. They may also be more afraid of making errors or breaking the ICT. Furthermore, user-centered technologies are typically designed to reflect people’s mental models of the physical world. Apple – after a visit to Xerox PARC – modeled the screens of its personal computers after the concept of a physical desktop, for example (Kuang 2019). An Apple user could stack 2D windows atop each other as they would place papers in a pile on their desk in the physical world. Thus, Apple leveraged useful metaphors such as “folders” and “trash” in their products, which were intuitive to users who organized their papers in folders and threw stuff in the trash already. People who do not have much prior exposure to ICTs, however, and especially those who have not personally engaged with physical items such as desktops and folders, would need not only to learn the steps for navigating the new technology itself, but would also need to learn seemingly abstract and arbitrary information organizational concepts in the process. When it comes to user involvement in the design process, those people who have not had much exposure to ICTs may not have the vocabulary to comment on user interfaces’ utility and may have more difficulties imagining alternative designs (Pejovic 2019). 41 Promoting inclusivity and equity in information and communications technology for food, land, and water systems 6.3. Current trends in ICTs for food, land, and water systems Many ICTs for food, land, and water systems focus on the heuristic functioning of ICTs and researchers’ perceptions of users’ needs rather than seeking to engage users during the design process (Prost 2012; Ditzler 2018; Kragt 2014; Hewitt 2017). In a review of digital citizen science tools, for example, Andrachuk (2019) found that only 15% of developers reported that they involved users in the design process. Skarlatidou (2019) wrote, “In citizen science, digital technologies are often developed without (human-computer interaction) principles and methodologies in mind. Thus, it is not surprising that many citizen science applications fail or cause problems for researcher and users. These problems can impact adoption, continuous participation, data quality, and other aspects.” These problems can also result in mismatches between “the context in which a technology is created and the context in which it is meant to be used” (Pejovic 2019). Dodson (2013) similarly concludes that these “top-down, technology-centric, goal-diffuse approaches to (ICTs) contribute to unsatisfactory development results,” since these tools and technologies are often developed without understanding the local context or involving end users. This has serious implications for the usefulness, usability, and inclusivity of ICTs in food, land, and water systems and, consequently, their adoption. The importance of usability for the adoption of ICTs for food, land, and water systems is well-documented (Rose 2017; Oyinbo 2020). User-centered design approaches are shown to improve the usability of ICTs (Andreasson 2015) since they enhance a technology’s ability to meet its’ users’ needs and to be integrated into institutional structures, as opposed to a one-size-fits-all or “systems requirements” approach to design (Ssozi-Mugarura 2017; Champanis 2012). User involvement in the design process, in particular, helps to improve design at all levels – including perception and cognition, interface and task, platform and social network (Houghton 2019) – as well as overall system success (Bano 2015; Ssozi-Mugarura 2017; Sturm 2019). User-centered design has begun to gain traction in the creation of ICTs for food, land, and water systems. Within the field of digital agricultural extension applications, user-centered design (specifically co-design) is one of two cornerstones of its innovation agenda (Steinke 2020). A user-centered design process was used to develop a system called Ushauri, for example, which provides farmers in Tanzania with agricultural advice about groundnuts (Ortiz-Crespo 2020). 42 Promoting inclusivity and equity in information and communications technology for food, land, and water systems For climate information services, Bouroncle (2019) found that co-produced ICTs in Colombia and Guatemala received higher usability scores than those produced by single organizations. Muller (2020) examined the non-technical factors that are shaping the non-use of climate information services in Guatemala, including attitudes toward climate change, bureaucratic structures, centralized budgeting/planning, renegotiation due to political instability, and failure avoidance. In the area of citizen science, van de Gevel (2020) outlines how citizen science is conducive for participatory research, which offers a potentially fertile space for promoting user-centered design and participation in agricultural research. These applications exemplify the interest in and potential of user-centered design and other design methodologies, notably co-design/participatory design, when creating ICTs for food, land, and water systems. As ICTs are designed to fit local contexts, they can also be linked to other localized ICTs, thus enabling larger-scale collaboration and collective action. Awareness of inclusivity and equity during the design process is critical for producing more inclusive and equitable ICTs for food, land, and water systems. 6.4. Considering inclusion and equity when designing ICTs for food, land, and water systems 6.4.1. Inclusion and equity in the design process Many ICTs for food, land, and water systems are designed without involving the end user or placing her at the center of the design, which negatively impacts adoption rates. This suggests a need to transition from waterfall, expert-led design processes toward more inclusive, iterative processes inviting end user participation. This is particularly important when many designers of ICTs for food, land, and water systems come from outside the communities of target end users. Even if there is increased user participation, much of the decision-making power risks lying outside the group that will be most affected by the ICT. Traditional software design methodologies (such as agile) may need to adapt production roles to address power asymmetries between technologists and users and to more ethically distribute decision-making power (Dearden 2010). Just as ICTs are not always the solution to a problem (tech-solutionism), applying a certain design methodology is not always the best means of addressing a problem. While certain design principles may help ameliorate an issue, they should not be privileged over alternative or local forms of problem solving. One critique is that design methodologies are often assumed to be universal – hence their export or imposition from the Global North to designers in the Global South. This is a particular problem in light of how design has historically been wielded to secure capital for American design firms when faced by competition from designers abroad (Irani 2019). This has resulted in calls to decolonize both design and computing (Ansari 2019; Ali 2016; Irani 2010). In the words of one interviewee, “[Human-computer interaction for development] is about breaking out of these rich, white, male, U.S. systems into all kinds of other systems. What would a tropical computing environment look like?” (Dell 2016) More technically, there are also challenges associated with applying design methodologies to rural regions and different cultural contexts. While many ICTs are designed, developed, and tested in a lab, ICTs for food, land, and water systems must be designed and evaluated in real-world settings (in-situ) to ensure usability. In rural areas, however, traditional user-centered design methods – such as focus groups, think-alouds and shadowing – can be problematic (Sturm 2019). Dearden (2018) explains how design methodologies from the Global North “rely upon certain assumptions about how users and developers can interact and discuss ideas that are not necessarily valid in developing world settings.” He cites, for example, how the spoken word is privileged in design processes. Designers often transcribe these interviews with potential users. In contrast, designers are rarely instructed in how to interpret non-verbal signals, especially those that occur in a cross-cultural exchange or environment. 43 Promoting inclusivity and equity in information and communications technology for food, land, and water systems A growing body of literature examines how to design and evaluate ICTs in-situ. (In-situ is sometimes referred to as “in the wild” despite criticism for using such a biased phrase (Ssozi-Mugarura 2016). One recommendation for improving the usability of ICTs for food, land, and water systems is to expand the education and purview of a technologist beyond a narrow, technical focus (Blake 2006) and toward a facilitation role. Another suggestion is for technologists to create spaces in which value-sensitive design (Yoo 2013) or co-design with local communities can occur (Ssozi-Mugarura 2017; Hewitt 2017). One of the most common recommendations is to adapt the design/facilitation process – in addition to the ICT itself – to the context. When designing within a patriarchal society, for example, it is important to understand the systemic challenges that women face, since design occurs against this backdrop (Sultana 2018). In these cases, women in conservative societies may be more comfortable participating in design activities if there is at least one female facilitator present (Anokwa 2009). Also, although verbal hypotheticals are frequently used in usability testing, they may be more effective if they are adapted and represented using physical prototypes (Vitos 2017). ICTs could also be better contextualized as socio-technical systems with a wider range of focus than the technology itself and a wider range of actors than just end users (Kim 2011; Andrachuk 2019). A critique of many common design methodologies – such as user- and human-centered design – is that there is a focus on the preferences of individual end users (Kuang 2019). It is difficult to enact lasting social change – such as the United Nations Sustainable Development Goals – through ICTs when they cater to the current preferences of individual users, rather than the current needs of or future aspirations for society. How do we weigh the convenience of a technology for an individual user against the wellbeing of society? Furthermore, user- and human-centered design privileges not only individuals, but also humans as a species. What would more-than-human design look like, taking into account the wellbeing of non-human species and non-living things? (Wright 2020) 6.4.2. Inclusion and equity in the resulting ICTs It is possible for exclusion and inequities to be embedded directly into ICTs for food, land, and water systems. This can occur for a number of reasons. The ICT’s business model and user interests may be misaligned. The former, for example, usually demands scaling and growth. In the case of many social media platforms, the desire for growth resulted in the business decision to make them free to use – instead, they rely on ads for revenue, while monetizing users’ clicks and attention. Unfortunately, this ad-tech business model means that social media platforms are motivated to design for users’ constant engagement. They may embed dark patterns – such as alerts, quantification numbers and eye-catching colors – to “[hijack] our psychological vulnerabilities” and manipulate users to compulsively use the platform (Thomson 2019). These techniques have become especially prevalent since the rise of smartphones. One interviewee observed, “There’s a fundamentally adversarial relationship between the goals of the coders and designers and those of their users.” While this is not true of all business models, it speaks to the manipulation that can arise when a business model does not align with users’ interests. Even if an ICT has certain affordances10, standpoint shapes whether they are perceptible or available to a given end user (Costanza-Chock 2020). Someone who had access to a formal education will likely be better able to access a text-heavy user interface than someone who has a low level of literacy or is non- literate. Someone who has spent more time in urban or technology-intensive settings may be more familiar with symbols such as the “send” arrow or a “done” checkmark than someone with less exposure to the iconography of globalization. Singular ICT designs are not universal, which risks generating inequitable outcomes. 10 Affordances are properties of objects which show users the actions they can take. Users should be able to perceive affordances without having to consider how to use the items. A button on an application, for example, can be designed to look as if it needs to be turned or pushed. 44 Promoting inclusivity and equity in information and communications technology for food, land, and water systems ICTs can be extractive. Compared to earlier development paradigms, ICTs for development create a “newly intensified role for individual participants.” (Chan 2014) They promise to extract productivity not only from urban or elite people, but also from rural and indigenous populations, resulting in many “encounters” between local and global (Tsing 2004). “Subjects who had once been held as the abject other to modern universalizing projects could now be given the chance to emerge as model global participants instead.” (Chan 2014) While increased participation in development can be positive in certain cases, whether it is indeed actually positive relies on how meaningfully people are able to participate and whether their participation yields benefits or harms. It is one thing if one is empowered to make decisions about their community’s future through ICT use, but it is another thing all together if that individual is having their activities mined for data or used to generate ad revenue. 6.5. Visions for designing ICTs for food, land, and water systems Design processes can be made more inclusive and equitable in a number of ways. These could, in turn, improve the inclusivity and equity of ICTs for food, land, and water systems. Within the design process of ICTs for agriculture, designers hold a lot of power. They are sometimes even called “wizards,” “unicorns,” “ninjas,” or “rock stars” in their job descriptions (Wachter-Boettcher 2017). Even in participatory design processes, participants may lack the confidence or digital skills to articulate their needs, preferences, and aspirations. User-centered design processes could be improved if the role of a designer shifts away from being a “rock star” and moves toward being a facilitator. This serves two purposes. First, it makes tech more welcoming to more diverse types of people, including some who may not consider themselves as special and superior as those lofty job descriptions seem to suggest. It also redistributes power within the design team, which decreases the authority of professional designers and increases the authority of end users. In the words of Digital Democracy Director Emily Jacobi, “At Digital Democracy, we try to fight the superhero narrative. We are sidekicks rather than superheroes.” How would a “sidekick” designer collaborate with local communities to co-create ICTs for food, land, and water systems? Also, within the design process of creating ICTs for food, land, and water systems, design teams could actively seek out diversity. A more diverse design team reduces the likelihood that the ICT will suffer from embedded biases and cause harm. It may also help stakeholders feel more comfortable when interacting with the design team. Design teams can also redefine their metrics for success. Many ICTs are built within the logic of surveillance capitalism. That is, they link social good to benefiting a corporate system based on profits and a search for investable novelties (Irani 2019). Other design teams emphasize engagement over wellbeing. They may mistake efficiency for effectiveness, often dismissing useful ambiguities or slow processes, such as trust- and consensus-building, to save time. In the 1980s, the evaluation of system success shifted from tech-oriented metrics (e.g., ‘is it working?’) to more user-oriented business metrics, such as number of registered users and session duration. Now, there is a need to shift to real-world, outcome- oriented metrics, such as indicators for quality of life and environmental sustainability. In the words of artist Jenny Odell, “Living between the mountains and this hyper accelerated, entrepreneurial culture, I can’t help but ask the question: What does it mean to construct digital worlds while the actual world is crumbling before our eyes?” (Odell 2019) 45 Promoting inclusivity and equity in information and communications technology for food, land, and water systems Regarding design methodologies and applications more generally, design – at its core – is about imagining the future(s). Decolonizing design requires welcoming more people into collaborative design processes, as per Ezio Manzini’s vision for “cosmopolitan localism.” (Manzini 2015) It also entails designing “a world where many worlds fit,” as the Zapatista movement in Mexico described it, and “heterogeneity of times, heterogeneity of worlds,” rather than reproducing the “one-world” project of neo-liberal globalization by imposing certain ways of knowing on individuals or privileging certain types of knowledge (Garcia Canclini 1995; Escobar 2018; Mignolo 2011). How might the design of ICTs for food, land, and water systems create room for many avenues of social change, while crediting designers/innovators from more diverse backgrounds, and building people’s self-efficacy?11 Finally, it is important to remember the bigger picture. As Toyama (2015) described in his Law of Amplification, ICTs amplify social forces. He argues that we ought to view social issues less as problems to be solved – which could lead to creating technocratic, expert-driven solutions – and more as people and institutions to be nurtured, which could lead to ICT-enabled personal connections and collective action. What social and planetary forces should ICTs for food, land, and water systems amplify? The answer to this question largely comes down to the people who will be affected by these technologies. As ICTs have become increasingly pervasive, there are also more opportunities to link them across localized contexts to facilitate larger-scale collective action for creating and fostering transformative food, land, and water systems. 11 Self-efficacy reflects confidence in the ability to exert control over one’s own motivation, behavior, and social environment. 46 Promoting inclusivity and equity in information and communications technology for food, land, and water systems Lack of inclusivity and equity in ICTs has serious implications, as technologies are increasingly mediating access to information, jobs, communication, and – in this case – food, land, and water systems. A few arguments about why inclusivity and equity matter when designing ICTs for food, land, and water systems. 7.1. Preventing a positive feedback loop that reproduces inequalities Digital divides research has demonstrated that people have different abilities to access, use, and benefit from ICTs. If ICTs for food, land, and water systems are not inclusive, they risk creating an unintended positive feedback loop. Privileged people who can access the ICTs will reap the benefits, which will make them increasingly more privileged. Marginalized people who cannot access the ICTs will not be able to use or benefit from them; thus, they will become more marginalized. Of course, this argument assumes that using a given technology will indeed yield real-world benefits. 7.2. Social and psychological impacts of ICT usage An interactional theory of ICTs posits that “human beings acting as individuals, organizations, or societies shape the tools and technologies they design and implement; in turn, those tools and technologies shape human experience and society.” (Friedman 2019) In other words, people shape ICTs and, then, they shape us in return. The social and psychological impacts of ICT use are far-ranging. As a small sample, search engines such as Google reinforce racism and sexism (Noble 2018). Misinformation and disinformation campaigns threaten democratic elections (Allcott 2017). ICT use drives cultures to become more individualistic (Salehan 2018). Recommendation algorithms, such as those found on YouTube, steer viewers toward more inflammatory content, which may play a role in online radicalization (Ribeiro 2020). Artificial Intelligence takes a static view of human nature based on its training data – thus, “codifying the past,” including all of its inequities, rather than creating a more equitable future (O’Neil 2016). The consistent use of ICTs literally rewires our brains due to neuroplasticity. The alterations to this use range from shortened attention spans (Carr 2010) to lower levels of cognitive construal (Kaufman 2016). The lack of embodiment during digital experiences contributes 47 Promoting inclusivity and equity in information and communications technology for food, land, and water systems to a loss of sense of place (Meyrowitz 1986). Within international development specifically, the digital divides can cause rural partnerships to suffer and produce outcomes that are unsustainable, lopsided, and non- participatory (Erdiaw-Kwasie 2014). The digital divides also pose challenges to disseminating information about the SDGs across technologically disparate environments (McEntee-Atalianis 2018). ICTs galvanize positive social and psychological impacts, too. Marginalized people have historically lacked access to elite media spaces, for example; however, social media sites, especially Twitter, have helped individuals from these groups in sharing news and organizing (Jackson 2020). Video games improve visual- spatial skills and Internet use improves our ability to multi-task (Carr 2010). The downsides are worth mentioning, however, because they are less commonly discussed and often dismissed as “unintended consequences” – a term frequently used to dismiss the political and ethical concerns of ICTs (Parvin 2020). We must seriously consider the inclusivity and equity of ICTs for food, land, and water systems during the design process because they have profound impacts on both the psychology of individuals and of society. 7.3. Business justifications In addition to the moral arguments for improving the inclusivity of ICTs for food, land, and water systems, there are also numerous business justifications (Holmes 2018). More inclusive ICTs increase customer engagement and contribution, which results in more meaningful use per user. They also help businesses and organizations grow a larger customer base, as technologies appeal to more potential users. They promote innovation and differentiation, both of which serve as comparative advantages over competitors. They also avoid the high costs of retrofitting inclusion. Retrofitting inclusion can be quite costly. Beyoncé’s management company, Parkwood Entertainment, was sued for discrimination because her website failed to comply with The Americans with Disabilities Act (Tariq 2019). The class action lawsuit claims that her website denies equal access to users with visual impairments. Domino’s was also sued because its website and mobile app were inaccessible to users with visual impairments (Hurley 2019). These are only two in a string of lawsuits against high-profile websites that do not comply with Web Accessibility Standards. 7.4. Better outcomes for all Designing for inclusivity supports not only marginalized end users, but also all end users throughout the different stages of their lives. Consider how physical ability changes over time. One can suddenly break a leg, suffer a concussion, or develop Carpal Tunnel Syndrome. Virtually everyone loses their hearing and vision with age. These physical changes can make certain ICTs unusable permanently, temporarily, or situationally (Microsoft 2020). Designing ICTs for inclusivity can help accommodate users as their bodies and circumstances evolve over the years, which is especially important in light of the aging global population. This would have benefits for all users – from the parent carrying a baby, who only has one arm free, to the person preoccupied or running late, who can hold fewer items in their working memory. Able-bodied users regularly use many design features originally created for users with disabilities. High- contrast screens were first designed for people with vision loss, for example, but are now widely used on all mobile phones because they can adapt to whether the user is standing in shade or direct sunlight. Closed captioning on videos was originally created for people with hearing impairments. The technology is now widely used for social media videos, because so many people scroll through their newsfeeds in public places where they might not be able to hear any audio. Web accessibility standards require that websites be navigable by keyboard for users who are unable to operate a mouse (Web Content Accessibility Guidelines 2018). This led to the creation of the tab shortcut that many people now use to jump from button to button on a webpage. 48 Promoting inclusivity and equity in information and communications technology for food, land, and water systems 8. conclusion Information and communications technologies (ICTs) – conceptualized broadly as “socio- technical systems” – have ample potential to support food, land, and water systems as well as the Sustainable Development Goals; however, concerns about the inclusivity and equity of the technology merit attention, including who benefits, who controls their modes of production, and whose goals are served. This is especially pressing in light of technology’s low rates of adoption for use in food, land, and water systems. The inclusivity and equity of ICTs for food, land, and water systems can be improved in several ways. Recommendations emerging from the digital divides framework are often top-down, including suggestions such as offering digital skills training or improving the affordability of mobile data and treating the technology as fixed. ICTs are neither neutral nor apolitical. This means that technologies – including those for food, land, and water systems – are more accessible, usable, and beneficial to some people compared with others, which translates into inequitable outcomes. Design can serve as a bottom-up strategy for improving the inclusivity and equity of technologies by building technological systems that place people’s needs, values, and aspirations as central to their creation and use. Design processes raise their own questions of inclusivity and equity. These range from how to equitably distribute decision-making power within the design team to how to tailor design methodologies, many of which emerged from the Global North, for different cultural contexts, many of which are rural settings in the Global South. Furthermore, users can be exploited if the business model of the technology is misaligned with a user population’s own interests; and some users may not perceive or be able to use certain aspects of a platform if it is not designed inclusively. Lack of inclusivity and equity in information and communications technology risk creating a positive feedback loop, in which privileged people become more privileged and marginalized people become more marginalized. Technologies also have a range of social and psychological issues, which could – if these ill-constructed technologies continue to be developed without 49 Promoting inclusivity and equity in information and communications technology for food, land, and water systems addressing these issues – further perpetuate inequalities. Business justifications for supporting design inclusivity include increasing customer engagement, growing a larger customer base, promoting innovation and differentiation, and avoiding the high costs of retrofitting innovation. Inclusive designs benefit not only marginalized people, but also all end users since everyone’s needs and preferences evolve over the course of their lives. While the rhetoric around ICTs is often steeped in determinism and inevitability, it is important to respect people’s self-determination with regard to technology use. This may mean supporting people in building their own platforms or being accepting in those instances in which people do not want to use technology at all. After all, technology is not the answer to all problems. Information and communications technologies offer opportunities not only to personalize designs for individual users, but also to create links across these personalized contexts that facilitate large-scale collective action to transform food, land, and water systems. Responsible innovation and governance are necessary to ensure that technologies better serve a wide diversity of end users and promote inclusivity and equity in rapidly digitizing societies. 50 Promoting inclusivity and equity in information and communications technology for food, land, and water systems references Accenture. 2017. Digital disruption: development unleashed. https://www.accenture.com/_acnmedia/pdf- 40/accenture-digital-disruption-development-unleashed.pdf. AgriBuddy. Home. https://www.agribuddy.com/. Accessed 22 February 2021. Aikins, Stephen Kwamena. 2019. Determinants of digital divide in Africa and policy implications. International Journal of Public Administration in the Digital Age. DOI: https://doi.org/10.4018/ IJPADA.2019010104. Ali, Syed Mustafa. 2016. A brief introduction to decolonial computing. ACM XRDS. DOI: https://doi. org/10.1145/2930886. Allcott, Hunt and Matthew Gentzkow. 2017. Social media and fake news in the 2016 election. Journal of Economic Perspectives. DOI: https://doi.org/10.1257/jep.31.2.211. Alsheri, Tag, Norah Abokhodair and Patrick Olivier. 2019. Understanding Culture with a Transnational Population: A Value-Sensitive Approach. SocArXiv. DOI: https://doi.org/10.31235/osf.io/kpz2r. Andrachuk, Mark, Melissa Marschke, Charlotte Hings, Derek Armitage. 2019. Smartphone technologies supporting community-based environmental monitoring and implementation: a systematic scoping review. Biological Conservation 237 (2019) 430-442. DOI: https://doi.org/10.1016/j.biocon.2019.07.026. Andreasson, Rebecca, Jessica Lindblom and Peter Thorvald. 2015. Towards an increased degree of usability work in organizations. Procedia Manufacturing 3 2015 5739-5746. DOI: https://doi.org/10.1016/j. promfg.2015.07.814. Anokwa, Yaw, Thomas N. Smyth, Divya Ramachandran, Jahanzeb Sherwani, Yael Schwartzman, Rowena Luk, Melissa Ho, Neema Moraveji and Brian DeRenzi. 2009. Stories from the field: Reflections on HCI4D experiences. Information Technologies and International Development 5(4) 2009. http://www.cs.cmu. edu/~jsherwan/pubs/stories-itid09.pdf. Ansari, Ahmed. 2019. Decolonizing design through the perspectives of cosmological others: arguing for an ontological turn in design research and practice. ACM XRDS. DOI: https://doi.org/10.1145/3368048. Apple. ‘Inclusion and Diversity.’ https://www.apple.com/diversity/. Accessed 5 June 2020. Ardila, Alfredo, Paulo H. Bertolucci, Lucia W. Braga, Alexander Castro-Caldas, Tedd Judd, Mary H. Kosmidis, Esmeralda Matute, Ricardo Nitrini, Feggy Ostrosky-Solis and Monica Rosselli. 2010. Illiteracy: The Neuropsychology of Cognition Without Reading. Archives of Clinical Neuropsychology 25(8) 2010: 689-712. https://doi.org/10.1093/arclin/acq079. Arora, Payal. 2019. The Next Billion Users: Digital Life Beyond the West. Harvard University Press. Attewell, Paul. 2001. The first and second digital divides. Sociology of Education. DOI: https://doi. org/10.2307/2673277. Bano, Muneera and Didar Zowghi. 2015. A systematic review on the relationship between user involvement and system success. Information and Software Technology 58 (2015) 148-169. DOI: http://dx.doi. org/10.1016/j.infsof.2014.06.011. Barboni,  Giorgia, Erica Field, Rohini Pande, Natalia Rigol, Simone Schaner, Charity Troyer Moore. 2018. ‘A Tough Call: Understanding barriers to and impacts of women’s mobile phone adoption in India.’ Harvard Kennedy School: Evidence for Policy Design.  https://epod.cid.harvard.edu/sites/default/ files/2018-10/A_Tough_Call.pdf.  51 Promoting inclusivity and equity in information and communications technology for food, land, and water systems Baron-Cohen, Simon. 2011. Zero Degrees of Empathy: a New Theory of Human Cruelty. Allen Lane. Baumer, Eric, Jenna Burrell, Morgan Ames, Jed Brubakar and P. Dourish. 2015. On the importance and implications of studying technology non-use. ACM Interactions. DOI: https://doi.org/10.1145/2723667. Belkhir, Lotfi and Ahmed Elmeligi. 2018. Assessing ICT global emissions footprint: trends to 2040 & recommendations. Journal of Cleaner Production. DOI: https://doi.org/10.1016/j.jclepro.2017.12.239. Benjamin, Ruha. 2019. Race After Technology: Abolitionist Tools for the New Jim Code. Polity. Bhatnagar, Amit and Sanjoy Ghose. 2004. Segmenting consumers based on the benefits and risks of Internet shopping. Journal of Business Research, Volume 57, Issue 12, p. 1352-1360. DOI: https://doi. org/10.1016/S0148-2963(03)00067-5. Birhane, Abeba. 2020. Algorithmic Colonization of Africa. scripted. DOI: https://doi.org/10.2966/ scrip.170220.389. Boston Consulting Group. The disruptive power of digital ecosystems. https://www.bcg.com/capabilities/ digital-technology-data/digital-ecosystems. Accessed 22 February 2021. Bouroncle, Claudia, Anna Muller, Diana Giraldo, David Rios, Pablo Imbach, Estuardo Giron, Fernando Portillo, Alejandra Boni, Jacob van Etten and Julian Ramirez-Villegas. 2019. A systematic approach to assess climate information products applied to agriculture and food security in Guatemala and Colombia. Climate Services. DOI: https://doi.org/10.1016/j.cliser.2019.100137. Buolawmwini, Joy  and  Timnit  Gebru. 2018. Gender Shades: Intersectional Accuracy Disparities in Commercial Gender Classification. Proceedings of Machine Learning Research 81:1–15, 2018 Conference on Fairness, Accountability, and Transparency. http://proceedings.mlr.press/v81/buolamwini18a/ buolamwini18a.pdf. Burnett, Margaret, Scott D. Fleming, Shamsi Iqbal, Gina Venolia, Vidya Rajaram, Umer Farooq, Valentina Grigoreanu and Mary Czerwinski. 2010. Gender differences and programming environments: across programming populations. Proceedings of the 2010 ACM-IEEE International Symposium on Empirical Software Engineering and Measurement. DOI: https://doi.org/10.1145/1852786.1852824. Burnett, Margaret, Simone Strumpf, Jamie Macbeth, Stephann Makri, Laura Beckwith, Irwin Kwan, Anicia Peters and William Jernigan. 2016. GenderMag: a method for evaluating software’s gender inclusiveness. Interacting with Computers 28(6). DOI: https://doi.org/10.1093/iwc/iwv046. Burrell, Jenna. 2018. Thinking relationally about digital inequality in rural regions of the U.S. First Monday. DOI: http://doi.org/10.5210/fm.v23i6.8376. Campbell, Jillian and David Jensen. 2019. The case for a digital ecosystem for the environment. Medium. https://medium.com/@davidedjensen_99356/building-a-digital-ecosystem-for-the-planet- 557c41225dc2. Carr, Nicholas. 2010. The Shallows: What the Internet is Doing to Our Brains. W.W. Norton and Company. Cassell, Justine. 2002. Genderizing  HCI. MIT Media Lab.  http://www.justinecassell.com/publications/ gender.hci.just.pdf.  Champanis, Michael and Ulrike Rivett. 2012. Reporting water quality – a case study of a mobile phone application for collecting data in developing countries. ICTD’12. DOI: 10.1145/2160673.2160688. Chan, Anita Say. 2014. Networking Peripheries: Technological Futures and the Myth of Digital Universalism. MIT Press. Chang, Enyu, Feng Zhen and Yang Cao. 2016. Empirical analysis of the digital divide from the perspective of Internet usage patterns: a case study from Nanjing. International review for spatial planning and sustainable development. DOI: http://doi.org/10.14246/irspsd.4.1_49. 52 Promoting inclusivity and equity in information and communications technology for food, land, and water systems Chetty, Krish, Liu Qigui, Nozibele Gcora, Jaya Josie, Li Wenwei and Chen Fang. Bridging the digital divide: measuring digital literacy. Economics: the Open-Access, Open-Assessment E-Journal. DOI: https://doi. org/10.5018/economics-ejournal.ja.2018-23. Chew, Han Ei, P. Vigneswara Ilavarasan, Mark R. Levy. 2011. When there’s a will, there might be a way: The economic impact of mobile phones and entrepreneurial motivation on female-owned microenterprises. Proceedings of the Sixth International Conference on Information and Communication Technologies and Development. DOI: https://doi.org/10.1145/2516604.2516605. Christensen, Lars Rune, Hasib Ahsan and Badal Kumar Das. 2019. Are You Magicians? The Collaborative Work of an Agricultural Information Service. ICTD ‘19. DOI: https://doi.org/10.1145/3287098.3287109. Collins, Patricia Hill. 1990. Black Feminist Thought: Knowledge, Consciousness, and the Politics of Empowerment. Routledge. Costanza-Chock, Sasha. 2020. Design Justice: Community-led practices to build the world we need. The MIT Press. Crawford, Kate and Vladan Joler. 2018. Anatomy of an AI System. https://anatomyof.ai/. Criado-Perez, Caroline. 2019. ‘The deadly truth about a world built for men – from stab vests to car crashes.’ The Guardian. https://www.theguardian.com/lifeandstyle/2019/feb/23/truth-world-built-for-men- car-crashes. Criado-Perez, Caroline. 2019. Invisible Women: Data Bias in a World Designed for Men. Harry N. Abrams Press. Cruz-Jesus, Frederico, Tiago Oliveira, Fernando Bacao. 2018. The global digital divide: evidence and drivers. Journal of Global Information Management. DOI: https://doi.org/10.4018/JGIM.2018040101. D’Ignazio, Catherine and Lauren F. Klein. 2020. Data Feminism. The MIT Press. Dearden, Andy, Haider Rizvi and Subodh Gupta. 2010. Roles and responsibilities in agile ICT for development. Electronic Workshops in Computing, 1-12. https://pdfs.semanticscholar. org/92ce/6e1a07cb87a4acf601228696ea3130f0ca5f.pdf. Dearden, Andy. 2008. User-centered design considered harmful (with apologies to Edsger Dijkstra, Niklaus Wirth, and Don Norman). Proceedings of the International Conference on Information and Communication Technologies and Development 2008. DOI: https://doi.org/10.1162/itid.2008.00013. Dell, Nicola and Neha Kumar. 2016. The Ins and Outs of HCI for Development. Proceedings of the Conference on Human Factors in Computing Systems (CHI’16). DOI: https://doi.org/10.1145/2858036.2858081. Dell, Nicola, Vidya Vaidyanathan, Indrani Medhi, Edward Cutrell and William Thies. 2012. ‘Yours is Better!’ Participant Response Bias in HCI. Proceedings of the Conference on Human Factors in Computing Systems (CHI’12). https://www.microsoft.com/en-us/research/wp-content/uploads/2016/02/CHI2012- Dell-ResponseBias-proc.pdf. Deloitte Southeast Asia. 2017. Accelerating digital ecosystem development through strategic alliances. https://www2.deloitte.com/content/dam/Deloitte/sg/Documents/strategy/sea-strategy-monitor- deloitte-accelerating-digital-ecosystem.pdf. Deng, Xuefei (Nancy), K.D. Joshi and Robert D. Galliers. 2016. The duality of empowerment and marginalization in microtask crowdsourcing: giving voice to the less powerful through value sensitive design. The Management Information Systems Research Center Quarterly Vol. 40 No. 2. https:// dl.acm.org/doi/10.5555/3177617.3177620. Design Justice Network. ‘Read the Principles.’ https://designjustice.org/read-the-principles. Accessed 31 October 2020. DiMaggio, Paul and Bart Bonikowski. 2008. Make money surfing the web? The impact of 53 Promoting inclusivity and equity in information and communications technology for food, land, and water systems Internet on the earnings of U.S. workers. American Sociological Review. DOI: https://doi. org/10.1177/000312240807300203. Ditzler, Lenora, Laurens Klerkx, Jacqueline Chan-Dentoni, Helena Posthumus, Timothy J. Krupnik, Santiago Lopez Ridaura, Jens A. Andersson, Frederic Baudron and Jeroen C.J. Groot. 2018. Affordances of agricultural system analysis tools: a review and framework to enhance tool design and implementation. Agricultural Systems 164 2018 20-30. DOI: https://doi.org/10.1016/j.agsy.2018.03.006. Dobrinskaya, D.E. and T.S. Martynenko. 2019. Defining the digital divide in Russia: key features and trends. Monitoring of Public Opnion: Economics and Social Changes. DOI: https://doi.org/10.14515/ monitoring.2019.5.06. Dodson, Leslie, S. Revi Sterling and John K. Bennett. 2013. Considering failure: Eight years of ITID research. Information Technologies and International Development 9(2) 2012 19-34. Doidge, Norman. 2007. The Brain that Changes Itself. Penguin Books. Dwoskin, Elizabeth, Jeanne Whalen and Regine Cabato. Content moderators at YouTube, Facebook and Twitter see the worst of the web — and suffer silently. The Washington Post. 25 July 2019. https://www. washingtonpost.com/technology/2019/07/25/social-media-companies-are-outsourcing-their-dirty- work-philippines-generation-workers-is-paying-price/. Dye, Michaelanne, David Nemer, Josiah Mangiameli, Amy Bruckman and Neha Kumar. 2018. The human infrastructure of El Paquete, Cuba’s offline internet. ACM interactions. DOI: https://doi. org/10.1145/3289487. Elbert, Thomas, Christo Pantev, Christian Wienbruch, Brigitte Rockstroh and Edward Taub. 1995. Increased Cortical Representation of the Fingers of the Left Hand in String Players. Science. DOI: https://doi. org/10.1126/science.270.5234.305. Erdiaw-Kwasie, Michael O and Khorshed Alam. 2014. Toward understanding digital divide in rural partnerships and development: a framework and evidence from rural Australia. Journal of Rural Studies. DOI: https://doi.org/10.1016/j.jrurstud.2015.12.002. Escobar, Arturo. 2018. Designs for the Pluriverse: Radical Interdependence, Autonomy, and the Making of Worlds. Duke University Press. Fab Lab. https://www.fablabs.io/. Accessed 13 July 2020. Facebook. ‘2019 Diversity Report.’ https://diversity.fb.com/read-report/. Accessed 5 June 2020. FarmDrive. Home. https://farmdrive.co.ke/. Accessed 22 February 2021. Findlater, Leah and Joanna McGrenere. 2004. A comparison of static, adaptive, and adaptable menus. Proceedings of the Conference on Human Factors in Computing Systems (CHI’04). DOI: https://doi. org/10.1145/985692.985704. Flanagan, Mary. 2009. Critical Play: Radical Game Design. MIT Press. Food and Agriculture Organization of the United Nations and International Telecommunication Union. 2019. E-Agriculture in Action: Big Data for Agriculture. Bangkok, Thailand. http://www.fao.org/3/ ca5427en/ca5427en.pdf. Foucault, Michel. 1977. Discipline and Punish (English translation). Pantheon Books. Friedman, Batya and David G. Hendry. 2019. Value Sensitive Design: Shaping Technology with Moral Imagination. The MIT Press. Frubana. Home. https://www.frubana.com/. Accessed 22 February 2021. Garcia Canclini, Nestor. 1995. Hybrid Cultures: Strategies for Entering and Leaving Modernity. University of Minnesota Press. 54 Promoting inclusivity and equity in information and communications technology for food, land, and water systems Gibbs, Colin. 2017. ‘Facebook takes aim at ‘connecting the unconnected’. Fierce Wireless. https://www. fiercewireless.com/wireless/facebook-takes-aim-at-connecting-unconnected. Accessed 31 October 2020. Goldberg, Nieca. 2003. Women Are Not Small Men: Life-Saving Strategies for Preventing and Healing Heart Disease in Women. Google. ‘2020 Annual Diversity Report.’ https://diversity.google/. Accessed 5 June 2020. Google. ‘Next Billion Users.’ https://blog.google/technology/next-billion-users/. Accessed 31 October 2020. Gorannson, Bo and Doan Dai Ngoc Diep. 2016. Bridging the digital divide: the role of the scientific and technological information stations in rural areas of Dong Nai Province. VNU Journal of Policy: Policy and Management Studies. Gorrestieta, Christina, Reza Lotfian, Kye Taylor, Richard Brutti, John Kane. 2019. Gender de-biasing in speech emotion recognition. INTERSPEECH 2019. DOI: http://doi.org/10.21437/Interspeech.2019-1708. Gram, Maggie. 2019. On Design Thinking. n+1: issue 35. https://nplusonemag.com/issue-35/reviews/on- design-thinking/. GSMA. 2020. Connected Women: The Mobile Gender Gap Report 2020. London, United Kingdom. GSMA. 2020. Digital Agriculture Maps: 2020 State of the Sector in Low and Middle-Income Countries. https://www.gsma.com/r/wp-content/uploads/2020/09/GSMA-Agritech-Digital-Agriculture-Maps.pdf. Gupta, Anil. 2016. Grassroots Innovation: Minds on the Margins Are Not Marginal Minds. Penguin. Haraway, Donna. 1988. ‘Situated Knowledges: The Science Question in Feminism and the Privilege of Partial Perspective.’ Feminist Studies 14.3 (1988): 575-93. DOI: https://doi.org/10.2307/3178066. Hargittai, Eszter. 2002. Second level digital divide: differences in people’s online skills. First Monday. DOI: https://doi.org/10.5210/fm.v7i4.942. Hargittai, Eszter. 2006. Differences in actual and perceived online skills: the role of gender. Social Science Quarterly. https://doi.org/10.1111/j.1540-6237.2006.00389.x. Hartsock, Nancy C.M. 1983. The Feminist Standpoint: Developing the Ground for a Specifically Feminist Historical Materialism. Discovering Reality. DOI: https://doi.org/10.1007/0-306-48017-4_15. Hello Tractor. Home. https://hellotractor.com/. Accessed 31 October 2020. Helsper, Ellen Johanna and A. Galacz. 2009. Understanding the links between social and digital exclusion in Europe. In The World Wide Internet: Changing societies, economies and cultures, p. 146-178. https:// www.researchgate.net/publication/284994185_Understanding_the_links_between_social_and_ digital_inclusion_in_Europe. Accessed 29 June 2020. Hewitt, Richard J. and Christopher J.A. Macleod. 2017. What Do Users Really Need? Participatory Development of Decision Support Tools for Environmental Management Based on Outcomes. Environments 2017, 4, 88. DOI: https://doi.org/10.3390/environments4040088. Hoffman, Anna Lauren. 2019. Where fairness fails: data, algorithms, and the limits of antidiscrimination discourse. Information Communication and Society 22(7). DOI: https://doi.org/10.1080/136911 8X.2019.1573912. Hollan, James, Edwin Hutchins and David Kirsh. 2000. Distributed cognition: toward a new foundation for human-computer interaction. ACM Transactions on Computer-Human Interaction. DOI: https://doi. org/10.1145/353485.353487. Holmes, Kat. 2018. Mismatch: How Inclusion Shapes Design. The MIT Press. 55 Promoting inclusivity and equity in information and communications technology for food, land, and water systems Houghton, Robert J., James Sprinks, Jessica Wardlaw, Steven Bamford and Stuart Marsh. 2019. A sociotechnical system approach to virtual citizen science: an application of BS ISO 27500:2016. Journal of Science Communication 18 (2019). DOI: https://doi.org/10.22323/2.18010201. Hurley, Lawrence. U.S. Supreme Court rejects Domino’s bid to avoid disabilities suit. Reuters. 7 October 2019. https://www.reuters.com/article/us-usa-court-dominos-pizza/u-s-supreme-court-rejects- dominos-bid-to-avoid-disabilities-suit-idUSKBN1WM1P1. Hutchins, Edwin. 2000. Distributed cognition. IESBS Distributed Cognition. https://arl.human.cornell.edu/ linked%20docs/Hutchins_Distributed_Cognition.pdf. iCow. Home. https://icow.co.ke/. Accessed 31 October 2020. Interaction Design Foundation. What are socio-technical systems? https://www.interaction-design.org/ literature/topics/socio-technical-systems. Accessed 5 June 2020. Irani, Lilly, Janet Vertesi, Paul Dourish, Kavita Philip and Rebecca E. Grinter. 2019. Postcolonial computing a lens on design and development. Proceedings of the Conference on Human Factors in Computing Systems (CHI’10). DOI: https://doi.org/10.1145/1753326.1753522. Irani, Lilly. 2015. Hackathons and the Making of Entrepreneurial Citizenship. Science, Technology, and Human Values. DOI: https://doi.org/10.1177/0162243915578486. Irani, Lilly. 2018. ‘Design Thinking’: Defending Silicon Valley at the Apex of Global Labor Hierarchies. Catalyst: Feminism, Theory, Technoscience, 4(1), 1-19. https://catalystjournal.org/index.php/catalyst/ article/view/29638/html. Irani, Lilly. 2019. Entrepreneurial Citizenship and the Subsumption of Hope. UCSD Design Lab: Design@ Large. https://www.youtube.com/watch?v=zeWd6gc6slg. Irani, Lily. 2019. Chasing Innovation: Making Entrepreneurial Citizens in Modern India. Princeton Studies in Culture and Tech. ISO 9241-210: Ergonomics of human-system interaction — Part 210: Human-centred design for interactive systems. 2019. https://www.iso.org/standard/77520.html Jackson, Sarah J., Moya Bailey and Brooke Foucault Welles. 2020. #HashtagActivism. MIT Press. Jacobs, Abigail Z. and Michaelanne Dye. 2020. Internet-human infrastructures: lessons from Havana’s StreetNet. https://arxiv.org/pdf/2004.12207.pdf. John D. Gould and Clayton Lewis. 1985. Designing for usability: key principles and what designers think. Communications of the ACM, March 1985. DOI: https://doi.org/10.1145/3166.3170. Katre, Dinesh. 2010. One-Handed Thumb Use on Smart Phones by Semi-literate and Illiterate Users in India. International Federation for Information Processing 2010.  https://link.springer.com/content/ pdf/10.1007%2F978-3-642-11762-6_16.pdf. Kaufman, Geoff and Mary Flanagan. 2016. High-low split: divergent cognitive construal levels triggered by digital and non-digital platforms. Proceedings of the Conference on Human Factors in Computing Systems (CHI’16). DOI: https://doi.org/10.1145/2858036.2858550. Kaye, Kate. ‘IBM, Microsoft, and Amazon’s face recognition bans don’t go far enough.’ Fast Company. 13 June 2020. https://www.fastcompany.com/90516450/ibm-microsoft-and-amazons-face-recognition- bans-dont-go-far-enough. Kennedy, Merrit. ‘Details of Uber harassment settlement released.’ NPR. 22 August 2018. https://www.npr. org/2018/08/22/640900988/dozens-sued-uber-for-harassment-heres-what-they-re-set-to-receive. 56 Promoting inclusivity and equity in information and communications technology for food, land, and water systems Kim, Sunyoung, Christine Robson, Thomas Zimmerman, Jeff Pierce, Eben M. Haber. 2011. Creek Watch: Pairing Usefulness and Usability for Successful Citizen Science. ACM Computer-Human Interaction Conference May 2011. DOI: http://dx.doi.org/10.1145/1978942.1979251. Kragt, M.E. and R. Llewellyn. 2014. Using a Choice experiment to improve Decision Support Tool design. Applied Economic Perspectives and Policy, 36(2): 351-371. DOI: https://doi.org/10.1093/aepp/ppu001. Kremer, Michael. 2020. ‘Farmers on the frontline: Making digital agriculture work for small scale farmers in a post coronavirus world.’ Hosted by the International Fund for Agricultural Development. 12 June 2020. https://www.youtube.com/watch?v=t_dgnG3r5a0&. Kuang, Cliff. 2019. User Friendly: How the Hidden Rules of Design Are Changing the Way We Live, Work, and Play. MCD. Kuhn, Peter and Hani Mansour. 2014. Is Internet job search still ineffective? The Economic Journal, Volume 124, Issue 581, p. 1213-1233. DOI: https://doi.org/10.1111/ecoj.12119. Kumar, B.T. Sampath and S.U. Shiva Kumara. 2018. The digital divide in India: use and non-use of ICT by rural and urban students. World Journal of Science, Technology and Sustainable Development. DOI: https://doi.org/10.1108/WJSTSD-07-2017-0021. Larson, Jeff, Surya Mattu, Lauren Kirchner and Julia Angwin. ‘How we analyzed the COMPAS recidivism algorithm.’ Propublica. 23 May 2016. https://www.propublica.org/article/how-we-analyzed-the- compas-recidivism-algorithm. Lediga, Mamathole Margaret and Madeleine C. Fombad. 2018. The use of information and communication technologies in public libraries in South Africa as tools for bridging the digital divide: the case of the Kempton Park public library. Public Library Quarterly. DOI: https://doi.org/10.1080/01616846.2018.1471 964. Light, Jennifer S. 1999. When Computers Were Women. Technology and Culture Vol. 40, No. 3, pp. 455-483. https://www.jstor.org/stable/25147356. Lindtner, Silvia. 2020. Prototype Nation: China and the Contested Promise of Innovation. Princeton University Press. Lombana-Bermudez, Andres, Sandra Cortesi, Christian Fieseler, Urs Gasser, Alexa Hasse, Gemma Newlands and Sarah Wu. 2020. Youth and the digital economy: exploring youth practices, motivations, skills, pathways, and value creation. Youth and Media, Berkman Klein Center for Internet & Society. https:// dash.harvard.edu/handle/1/42669835. Accessed 31 October 2020. Maguire, E.A., D.G. Gadian, I.S. Johnsrude, et al. 2000. Navigation-related structural change in the hippocampi of taxi drivers. Proceedings of the National Academy of Sciences of the United States of America 97(8): 4398-4403. DOI: https://doi.org/10.1073/pnas.070039597. Mahyar, Narges, Mahmood Jasim and Ali Sarvghad. 2020. Designing technology for sociotechnical problems: challenges and considerations. IEEE Computer Graphics and Applications, vol. 40, no. 6, pp. 76-87. 2020. DOI: https://doi.org/10.1109/MCG.2020.3017405. Manzini, Ezio. 2015. Design, When Everybody Designs: An Introduction to Design for Social Innovation. MIT Press. Margalit,  Liraz. ‘Gender-oriented design in light of the extreme male brain model.’ UX Matters. 7 July 2014.  https://www.uxmatters.com/mt/archives/2014/07/gender-oriented-design-in-light-of-the- extreme-male-brain-model.php.  Maudlin, Lindsey, Karen McNeal, Heather Donin-Aldridge, Corey Davis, Ryan Boyles and Rachel Atkins. 2020. Website usability differences between males and females: an eye-tracking evaluation of a climate decision support system. Weather, Climate, and Society Vol 12: Issue 1. DOI: https://doi.org/10.1175/ WCAS-D-18-0127.1. 57 Promoting inclusivity and equity in information and communications technology for food, land, and water systems McEntee-Atalianis, Lisa J. 2018. ‘Leave no one behind’: linguistic and digital barriers to the dissemination and implementation of the United Nations Sustainable Development Goals. Language Problems and Language Planning. DOI: https://doi.org/10.1075/lplp.00001.mce. McKinsey. 2020. Ecosystem 2.0: climbing to the next level. https://www.mckinsey.com/business-functions/ mckinsey-digital/our-insights/ecosystem-2-point-0-climbing-to-the-next-level. Medhi Thies, Indrani. 2015. User Interface Design for Low-literate and Novice Users: Past, Present and Future. Foundations and Trends in Human-Computer Interaction. DOI: https://doi.org/10.1561/1100000047. Medhi, Indrani, Ed Cutrell and Kentaro Toyama. 2010. It’s not just illiteracy. https://www.microsoft.com/en- us/research/wp-content/uploads/2016/02/Medhi_IndiaHCI2010.pdf. Meyrowitz, Joshua. 1986. No Sense of Place: The Impact of Electronic Media on Social Behavior. Oxford University Press. Microsoft. ‘Diversity and Inclusion Report 2019.’ https://query.prod.cms.rt.microsoft.com/cms/api/am/ binary/RE4aqv1. Accessed 5 June 2020. Microsoft. ‘Microsoft Inclusive Design’ https://devblogs.microsoft.com/premier-developer/microsoft- inclusive-design/. Accessed 31 October 2020. Mignolo, Walter. 2016. The Darker Side of Western Modernity: Global Futures, Decolonial Options. Duke University Press. Mink, Annemarie, B.A. Hoque, Sufia Khanam and Doris van Halem. 2019. Mobile crowd participation to root small-scale piped water supply systems in India and Bangladesh. Journal of Water, Sanitation and Hygiene for Development. DOI: http://doi.org/10.2166/washdev.2019.117. Mok, Luisa and Sampsa Hyysalo. 2018. Designing for energy transition through Value Sensitive Design. Design Studies Vol. 54 (162-183). DOI: https://doi.org/10.1016/j.destud.2017.09.006. Moss, Gloria, Rod Gunn and Jonathan Heller. 2006. Some men like it black, some women like it pink: consumer implications of differences in male and female website design. Journal of Consumer Behavior. DOI: https://doi.org/10.1002/cb.184. Muller, Anna, Claudia Bouroncle, Ada Gaytan, Estuardo Giron, Andrea Granados, Vesalio Mora, Fernando Portillo and Jacob van Etten. 2020. Good data are not enough: understanding limited information use for climate risk and food security management in Guatemala. Climate Risk Management. DOI: https://doi.org/10.1016/j.crm.2020.100248. Mz* Baltazar’s Laboratory. ‘About Us.’ http://www.mzbaltazarslaboratory.org/about/. Accessed 13 July 2020. National Autistic Society. ‘Home.’ https://www.autism.org.uk/. Accessed 31 October 2020. Ngozi Anasi, Stella, Clement Chinemerem Ukangwa and Abimbola Fagbe. 2018. University libraries: bridging digital gaps and accelerating the achievement of sustainable development goals through information and communication technologies. World Journal of Science, Technology and Sustainable Development. DOI: https://doi.org/10.1108/WJSTSD-11-2016-0059. Noble, Safiya Umoja. 2018. Algorithms of Oppression: How Search Engines Reinforce Racism. NYU Press. Noble, Safiya. ‘Google has a striking history of bias against black girls.’ Time Magazine. 26 March 2018. https://time.com/5209144/google-search-engine-algorithm-bias-racism/. Norman, Donald. 2013. The Design of Everyday Things: Revised and Expanded Edition. Norman, Donald. 2020. ‘Community-based, human-centered design’ on LinkedIn. https://www.linkedin. com/pulse/community-based-human-centered-design-don-norman/. Accessed 14 July 2020. O’Neil, Cathy. 2016. Weapons of Math Destruction: How Big Data Increases Inequality and Threatens Democracy. Crown. 58 Promoting inclusivity and equity in information and communications technology for food, land, and water systems O’Neill, Jacki, Anupama Dhareshwar and Srihari H. Muralidhar. 2017. Working digital money into a cash economy: the collaborative work of loan payment. Computer Supported Collaborative Work. DOI: https://doi.org/10.1007/s10606-017-9289-6. Odell, Jenny. 2019. How to Do Nothing: Resisting the Attention Economy. Melville House. Oliver, Mike. 2013. The social model of disability: thirty years on. Disability and Society. DOI: https://doi.org /10.1080/09687599.2013.818773. Organisation for Economic Co-operation and Development. 2018. ‘Bridging the Digital Gender Divide: Include, Upskill, Innovate.’ http://www.oecd.org/internet/bridging-the-digital-gender-divide.pdf. Organisation for Economic Co-operation and Development. 2005. Oslo Manual: Guidelines for collecting and interpreting innovation data. DOI: https://doi.org/10.1787/9789264013100-en. Ortiz-Crespo, Berta, Jonathan Steinke, Carlos F. Quiros, Jeske van de Gevel, Happy Daudi, Majuto Gaspar Mgimiloko and Jacob van Etten. 2020. User-centred design of a digital advisory service: enhancing public agricultural extension for sustainable intensification in Tanzania. International Journal of Agricultural Sustainability. DOI: https://doi.org/10.1080/14735903.2020.1720474. Ostrosky-Solıs, Feggy, Miguel Arellano Garcıa, and Martha Perez. 2004. Can learning to read and write change the brain organization? An electrophysiological study. International Journal of Psychology 2004 39(1), 27-35. http://www.feggylab.mex.tl/imagesnew/7/0/4/8/6/ostrosky-solis(2)-e-print.pdf. Otioma, Chuks, Ana Mafalda Madureira, Javier Martinez. 2019. Spatial analysis of urban digital divide in Kigali, Rwanda. GeoJournal. DOI: https://doi.org/10.1007/s10708-018-9882-3. Ottoni, Raphael, Joao Paulo Pesce, Diego Las Casas, Geraldo Franciscani Jr., Wagner Meira Jr., Ponnurangam Kumaraguru and Virgilio Almeida. 2013. Ladies first: analyzing gender roles and behaviors in Pinterest. Proceedings of the Seventh International AAAI Conference on Weblogs and Social Media. Retrieved from https://ojs.aaai.org/index.php/ICWSM/article/view/14438. Oyinbo, Oyakhilomen, Jordan Chamberlin and Miet Maertens. 2020. Design of Digital Agricultural Extension Tools: Perspectives from Extension Agents in Nigeria. Journal of Agricultural Economics: Vol 73 Issue 3. DOI: https://doi.org/10.1111/1477-9552.12371. Parvin, Nassim and Anne Pollock. 2020. Unintended by design: on the political uses of ‘unintended consequences.’ Engaging Science, Technology, and Society. DOI: https://doi.org/10.17351/ests2020.497. Pascual-Leone, A., D. Nguyet, L.G. Cohen, et al. 1995. Modulation of muscle responses evoked by transcranial magnetic stimulation during the acquisition of new fine motor skills. Journal of Neurophysiology. DOI: https://doi.org/10.1152/jn.1995.74.3.1037. Paul, Ronald L., Herbert Goodman and Michael Merzenich. 1972. Alterations in Mechanoreceptor Input to Brodmann’s Areas 1 and 3 of the Postcentral Hand Area of Macaca mulatta after Nerve Section and Regeneration. Brain Research. DOI: https://doi.org/10.1016/0006-8993(72)90782-2. Pejovic, Veljko and Artemis Skarlatidou. 2019. Understanding interaction design challenges in mobile extreme citizen science. International Journal of Human-Computer Interaction 2019. DOI: https://doi. org/10.1080/10447318.2019.1630934. Prost, Lorene, Marianne Cerf and Marie-Helene Jeuffroy. 2012. Lack of consideration for end-users during the design of agronomic models. A review. Agronomy for Sustainable Development, 2012, 32 (2), pp.581-594. DOI: https://doi.org/10.1007/s13593-011-0059-4. Rafalow, Matthew H. 2018. Disciplining play: digital youth culture as capital at school. American Journal of Sociology. DOI: https://doi.org/10.1086/695766. Reis, Alexandra, Karl Magnus Petersson, Alexandre Castro-Caldas and Martin Ingvar. 2001. Formal schooling influences two- but not three-dimensional naming skills. Brain and Cognition 47, 397-411. DOI: https:// doi.org/10.1006/brcg.2001.1316. 59 Promoting inclusivity and equity in information and communications technology for food, land, and water systems Ribeiro, Manoel Horta, Raphael Ottoni, Robert West, Virgilio A.F. Almeida and Wagner Meira. Auditing radicalization pathways on YouTube. Proceedings of the 2020 Conference on Fairness, Accountability, and Transparency. DOI: https://doi.org/10.1145/3351095.3372879. Ritzer, George and Nathan Jurgenson. 2010. Production, Consumption, Prosumption: The nature of capitalism in the age of the digital ‘prosumer’. Journal of Consumer Culture. DOI: https://doi. org/10.1177/1469540509354673. Rose, David C., Caroline Parker, Joey Fodey, Caroline Park, William J. Sutherland and Lynn V. Dicks. 2017. Involving stakeholders in agricultural decision support systems: Improving user-centred design. International Journal of Agricultural Management 6:3/4 2017. DOI: https://doi.org/10.5836/ijam/2017- 06-80. Sadato, Norihiro, Alvaro Pascual-Leone, Jordan Grafman, Vicente Ibañez, Marie-Pierre Deiber, George Dold and Mark Hallett. 1996. Activation of the primary visual cortex by Braille reading in blind subjects. Nature 380 526-528(1996). https://www.nature.com/articles/380526a0. Salehan, Mohammad, Dan J. Kim and Jae-Nam Lee. 2018. Are There any Relationships Between Technology and Cultural Values? A Country-Level Trend Study of the Association between Information Communication Technology and Cultural Values. Information and Management. DOI: https://doi. org/10.1016/j.im.2018.03.003. Scheerder, Anique, Alexander van Deursen and Jan van Dijk. 2017. Determinants of Internet skills, uses and outcomes. A systematic review of the second- and third-level digital divide. Telematics and Informatics. DOI: https://doi.org/10.1016/j.tele.2017.07.007. Shade, Leslie Regan. 2003. Whose Global Knowledge?: Women Navigating the Net. Society for International Development. DOI: https://doi.org/10.1177/1011637003046001649. Shetterly, Margot Lee. 2016. Hidden Figures: The American Dream and the Untold Story of the Black Women Mathematicians Who Helped Win the Space Race. William Morrow Paperbacks. Sikor, Thomas, Poshendra Satyal, Hari Dhungana and Gyanu Maskey. Brokering justice: global indigenous rights and struggles over hydropower in Nepal. Canadian Journal of Development Studies. DOI: https://doi.org/10.1080/02255189.2018.1540344. Sims, Stacy. 2019. Women are Not Small Men: a paradigm shift in the science of nutrition. TEDxTauranga. https://www.ted.com/talks/stacy_sims_women_are_not_small_men_a_paradigm_shift_in_the_ science_of_nutrition. Skarlatidou, Artemus, Marisa Ponti, James Sprinks, Christian Nold, Muki Haklay and Eiman Kanjo. 2019. User experience of digital technologies in citizen science. Journal of Science Communication. DOI: https://doi.org/10.22323/2.18010501. Soomro, Kamal Ahmed, Ugar Kale, Reagan Curtis, Mete Akcaoglu and Malayna Bernstein. 2020. Digital divide among higher education faculty. International Journal of Educational Technology in Higher Education. DOI: https://doi.org/10.1186/s41239-020-00191-5. Ssozi-Mugarura, Fiona, Edwin Blake and Ulrike Rivett. 2017. Codesigning with communities to support rural water management in Uganda. Special issue on participatory design in an era of participation. CoDesign 13 2017. DOI: https://doi.org/10.1080/15710882.2017.1310904. Ssozi-Mugarura, Fiona, Thomas Reitmaier, Anja Venter and Edwin Blake. 2016. Enough with ‘In-the-Wild’. AfriCHI’16. DOI: http://doi.org/10.1145/2998581.2998601. Steinke, Jonathan, Jacob van Etten, Anna Muller, Bertz Ortiz-Crespo, Jeske van de Gevel, Silvia Silvestri and Jan Priebe. 2020. Tapping the full potential of the digital revolution for agricultural extension: an emerging innovation agenda. International Journal of Agricultural Sustainability. DOI: https://doi.org /10.1080/14735903.2020.1738754. 60 Promoting inclusivity and equity in information and communications technology for food, land, and water systems Strubell, Emma, Ananya Ganesh and Andrew McCallum. 2019. Energy and Policy Considerations for Deep Learning in NLP. Proceedings of the 57th Annual Meeting of the Association for Computational Linguistics. Accessed at https://www.aclweb.org/anthology/P19-1355.pdf. Sturm, Ulrike and Martin Tscholl. 2019. The role of digital user feedback in a user-centred development process in citizen science. Journal of Science Communication 18 (2019). DOI: https://doi. org/10.22323/2.18010203. Sultana, Sharifa, Francois Guimbretiere, Phoebe Sengers and Nicola Dell. 2018. Design within a patriarchal society: opportunities and challenges in designing for rural women in Bangladesh. Proceedings of the Conference on Human Factors in Computing Systems (CHI’18). DOI: https://doi. org/10.1145/3173574.3174110. Sun, Alexander Y. and Bridget R. Scanlon. 2019. How can Big Data and machine learning benefit environment and water management: a survey of methods, applications, and future directions. Environmental Research Letters 14 073001. DOI: https://doi.org/10.1088/1748-9326/ab1b7d. Tariq, Imran. Beyonce’s official website got sued – and yours could be next. Entrepreneur. 3 October 2019. https://www.entrepreneur.com/article/339556. Tata, Joyous S. and Paul E. McNamara. 2016. Social factors that influence use of ICT in agricultural extension in Southern Africa. Agriculture 2016, 6, 15. DOI: https://doi.org/10.3390/agriculture6020015. Teo, Hock-Tai, Hock Chuan Chan and Bernard C.Y. Tan. 2011. Conceptualizing and testing a social cognitive model of the digital divide. Information Systems Research. DOI: https://doi.org/10.1287/isre.1090.0273. The Technical Centre for Agricultural and Rural Cooperation (CTA) and Dalberg Advisors. 2019. The Digitalisation of African Agriculture Report 2018-2019. https://www.cta.int/en/digitalisation- agriculture-africa. Thomson, Clive. 2020. Coders: The Making of a New Tribe and the Remaking of the World. Penguin Books. Toffler, Alvin. 1980. The Third Wave. William Morrow. Toyama, Kentaro. 2015. Geek Heresy: Rescuing Social Change from the Cult of Technology. Public Affairs. Tsing, Anna Lowenhaupt. 2004. Friction: An Ethnography of Global Connection. Princeton University Press. United Nations Development Programme. 2017. Data ecosystems for sustainable development. https:// www.undp.org/content/undp/en/home/librarypage/poverty-reduction/data-ecosystems-for- sustainable-development.html. United Nations. 2015. The Sustainable Development Agenda. https://www.un.org/sustainabledevelopment/ development-agenda/. Uotinen, Johanna. 2003. Involvement in (the Information) Society - The Joensuu Community Resource Centre Netcafé. New Media & Society. DOI: https://doi.org/10.1177/14614448030053003. USAID and the Grameen Foundation. 2018. Digital Farmer Profiles: Reimagining Smallholder Agriculture. https://www.usaid.gov/sites/default/files/documents/15396/Data_Driven_Agriculture_Farmer_ Profile.pdf. USAID. 2020a. Digital Strategy 2020-2024. https://www.usaid.gov/sites/default/files/documents/15396/ USAID_Digital_Strategy.pdf. USAID. 2020b. Digital Ecosystem Country Assessment: Kenya. https://www.usaid.gov/sites/default/files/ documents/DECA_Report_KENYA_EXTERNAL_2OCT2-2-_2.pdf. USAID. 2021. Digital Ecosystems. https://www.digitaldevelopment.org/digital-ecosystems. Accessed 22 February 2021. 61 Promoting inclusivity and equity in information and communications technology for food, land, and water systems van de Gevel, Jeske, Jacob van Etten and Sebastian Deterding. 2020. Citizen science breathes new life into participatory agricultural research. A review. Agronomy for Sustainable Development. DOI: https://doi. org/10.1007/s13593-020-00636-1. Van der Velden, Maja. 2018. Digitalisation and the UN sustainable development goals: what role for design. Interaction Design and Architecture(s) Journal. Accessed at https://www.duo.uio.no/bitstream/ handle/10852/71632/37_8.pdf?sequence=1&isAllowed=y. van Deursen, A.J.A.M. and J.A.G.M. van Dijk. 2009. Improving digital skills for the use of online public information and services. Government information Quarterly. DOI: https://doi.org/10.1016/j. giq.2008.11.002. Van Dijk, J.A. 2005. The Deepening Divide: Inequality in the Information Society. Thousand Oaks: Sage Publications. Van Wynsberghe, Aimee. 2013. Designing Robots for Care: Care Centered Value-Sensitive Design. Sci Eng Ethics 19, 407–433 (2013). DOI: https://doi.org/10.1007/s11948-011-9343-6. van Deursen, Alexander J.A.M. and Ellen Johanna Helsper. 2015. The third-level digital divide: who benefits most from being online? Communication and Information Technologies Annual (Studies in Media and Communications, Vol. 10). Emerald Group Publishing Limited, p. 29-52. DOI: https://doi.org/10.1108/ S2050-206020150000010002. Vitos, Michalis, Julia Altenbuchner, Matthias Stevens, Gillian Conquest, Jerome Lewis and Muki Haklay. 2017. Supporting collaboration with non-literate forest communities in the Congo-Basin. ACM Computer- Supported Collaborative Work Conference 2017. DOI: https://doi.org/10.1145/2998181.2998242. Von Hippel, Eric. 2016. Free Innovation. The MIT Press. Vorvoreanu, Mihaela, Lingyi Zhang, Yun-Han Huang, Claudia Hilderbrand, Zoe Steine-Hanson and Margaret Burnett. 2019. From Gender Biases to Gender-Inclusive Design: An Empirical Investigation. Proceedings of the 2019 CHI Conference on Human Factors in Computing Systems. DOI: https://doi. org/10.1145/3290605.3300283. Wachter-Boettcher, Sara. 2017. Technically Wrong: Sexist Apps, Biased Algorithms, and Other Threats of Toxic Tech. W.W. Norton & Company. Wajcman, Judy. 1996. Feminism Confronts Technology. Penn State University Press. Wajcman, Judy. 2010. Feminist theories of technology. Cambridge Journal of Economics. DOI: https://doi. org/10.1093/cje/ben057. Wakabayashi, Daisuke and Katie Benner. ‘How Google protected Andy Rubin, the ‘Father of Android’’. The New York Times. 25 October 2018. https://www.nytimes.com/2018/10/25/technology/google-sexual- harassment-andy-rubin.html. Web Content Accessibility Guidelines 2.1. 2018. https://www.w3.org/TR/WCAG21/. Whittaker, Meredith, Kate Crawford, Roel Dobbe, Genevieve Fried, Elizabeth Kaziunas, Varoon Mathur, Sarah Myers West, Rashida Richardson, Jason Schultz, Oscar Schwartz. 2018. AI Now 2018 Report. https://ainowinstitute.org/AI_Now_2018_Report.pdf. Williams, Gayna. 2014. Are you sure your software is gender-neutral? ACM Interactions. DOI: https://doi. org/10.1145/2524808. Winters, Niall and Kentaro Toyama. 2009. Human-Computer Interaction for Development: Mapping the Terrain. Information Technologies and International Development: Vol (5) No. 4. https://itidjournal.org/ index.php/itid/article/viewFile/419/187. 62 Promoting inclusivity and equity in information and communications technology for food, land, and water systems Wolf, Maryanne. 2007. Proust and the Squid. Harper Perennial. World Bank. 2016. Digital Dividends. World Development Report 2016. http://documents.worldbank.org/ curated/en/896971468194972881/pdf/102725-PUB-Replacement-PUBLIC.pdf. World Economic Forum. 2018. ‘The Global Gender Gap Report 2018.’ http://www3.weforum.org/docs/ WEF_GGGR_2018.pdf. Wright, Thomas. 2020. ‘More-than-human design: rethinking agency and sustainable practices.’ UX Collective. 22 October 2020. https://uxdesign.cc/more-than-human-design-rethinking-agency-and- sustainable-practices-926d580d5311. Yoo, Daisy, Alina Huldtgren, Jill Palzkill Woelfer, David G. Hendry and Batya Friedman. 2013. A value sensitive action-reflection model: evolving a co-design space with stakeholder and designer prompts. ACM Computer-Human Interaction Conference 2013. DOI: https://doi.org/10.1145/2470654.2470715. Young, Erin, Judy Wajcman and Laila Sprejer. 2021. Where are the Women? Mapping the Gender Job Gap in AI. Policy Briefing: Full Report. The Alan Turing Institute. https://www.turing.ac.uk/sites/default/ files/2021-03/where-are-the-women_public-policy_full-report.pdf. 63 Promoting inclusivity and equity in information and communications technology for food, land, and water systems 64 Promoting inclusivity and equity in information and communications technology for food, land, and water systems