MOBILE PHONES AND SUSTAINABLE AGRICULTURE BASED LIVELIHOODS: A

CASE STUDY OF SMALLHOLDER FARMERS IN THE ASHANTI AND BRONG-

AHAFO REGIONS, GHANA

ByTYLER JAY REYNOLDS

Submitted in fulfilment of the requirements for the degree of Doctor of Philosophy

Centre for Environment and SustainabilityFaculty of Engineering and Physical Sciences

University of SurreyApril 2019

Declaration of Originality

This thesis and the work to which it refers are the results of my own efforts. Any ideas, data, images

or text resulting from the work of others (whether published or unpublished) are fully identified as

such within the work and attributed to their originator in the text, bibliography or in footnotes. This

thesis has not been submitted in whole or in part for any other academic degree or professional

qualification. I agree that the University has the right to submit my work to the plagiarism detection

service TurnitinUK for originality checks. Whether or not drafts have been so-assessed, the University

reserves the right to require an electronic version of the final document (as submitted) for

assessment as above.

Tyler Jay Reynolds

25th April 2019 __________________________________

AbstractThis study investigates the role of mobile phones in the livelihood development of farmers in the

Ashanti and Brong-Ahafo Regions, Ghana. It seeks to uncover how farmers use their mobile phones

and the outcomes associated with that use, as well as the factors that make mobile phones more or

less useful. In order to answer these questions, a grounded theory approach was carried out using

both qualitative interviews (n=48) followed by a manual content analysis, and a quantitative

questionnaire (n=160) that provided for a statistical analysis of 4 communities in the Ashanti region

and 4 communities in the Brong-Ahafo Region, Ghana. In doing so it was possible to identify critical

challenges within important aspects of the production process and evaluate the use and usefulness

of mobile phones in addressing these challenges. Findings show that mobile phones were generally

useful for sending and receiving financial capital, seeking information on pesticides and fertilizers,

organising farm labour, and searching for yam sale pricing information when carrying out temporal

sales arbitrage. Mobile phones were seen as somewhat useful for liaising with extension officers and

less useful for the critical processes of obtaining subsidized fertilizer. It was also found that

individuals carry out the processes of transferring money, organising labour, and obtaining yam sale

pricing information using combinations strategies defined as: (1) heavily focused on the use of the

phone, (2) selective use of the phone, and (3) not making use of the phone. Individuals

implementing strategies heavily focused on mobile phone use tended to have smaller farms and

lower monthly incomes when compared to individuals employing alternative combinations of

strategies. According to these results and subsequent theory building, this research presents a

framework for the ICT related capital required for ICT related strategies to be carried out, as well as

the external structural barriers occurring in the study area, and resulting in a failure to translate the

use of mobile phone within strategies and processes into larger outcomes. While the study group as

a whole is experiencing a progressive trend of decreased vulnerability and positive livelihood

development, the widespread and complex use of the mobile phone has not resulted in a

transformation of livelihood outcomes.

Acknowledgments

I would like to express my profound gratitude towards my two supervisors: Prof Stephen Morse and

Dr Walter Wehrmeyer. You both are the reason why I chose to begin this project, and without you

this would not have been possible. You have my utmost respect. I could not have received better

supervision for this project. I genuinely hope that this work will help to achieve our common goal of

making the developing world a better place.

I would like to graciously thank Dr Kingsley Osei. Planning a project from the UK and then arriving in

Ghana to be met with supreme supervision, commitment, and confidence was absolutely invaluable.

I have been working in the developing world for some time now and I have never seen, let alone had

the privilege to work with, someone with such prowess and strength. I would also like to thank Dr.

Stella Ama Ennin, Prof Emmanuel Otoo, and all of the other research staff at the CSIR Crops Research

Institute (Ghana). I will forever be in debt of your support and accommodation during my time spent

in Ghana.

I would like to thank my parents: Marc Reynolds and Becky Heck. Your appreciation of and interest in

the natural world and all of its inhabitants will always inspire me. Thank you for your never ending

love and support in this endeavour.

I would also like to thank Moira Foster for both your administrative assistance and emotional

support. Finally, I would like to thank all of the participants involved in this research. I hope we may

all continue to participate together in building a better future for this generation and all that follow.

Table of ContentsDeclaration of Originality .................................................................................................................I

Abstract .........................................................................................................................................II

Acknowledgements .......................................................................................................................III

List of Tables .................................................................................................................................VI

List of Figures ................................................................................................................................VI

List of Appendices ........................................................................................................................VII

List of Abbreviations ....................................................................................................................VIII

Chapter 1: Introduction………………………………………………………………………………………………………….……….…

1

1.1 Motivation for research………………………………………………………………………………………………….……….…

1

1.2 ICTs and the developing world…………………………………………………………………..………………….………..…

1

1.3 Research gaps, scope of this research, and research questions…………………………………….………...…

3

1.4 Structure of thesis……………………………………………………………………………………………….………………….…4

Chapter 2: Literature Review – ICT4D Theory and Practice……………………………………………………….………

7

2.1 Introduction, definitions, history, and critique of ICT4D…………………………………………………..…...……

7

2.2 ICT4D Research theories, methods, and ICT and the Sustainable Livelihoods Framework …….…

10

2.3 ICT4D and ICT and agriculture research in Africa…………………………………………………………….….….…

15

2.4 Research aims and objectives……………………………………………………………………………………….….………

20

Chapter 3: Methodology and Methods……………………………………………………………………………….….….……

21

3.1 Research Methodology…………………………………………………………………………………………………….…..…21

3.2 Research design and data gathering methods…………………………………………………………….…….…..…

22

3.2.1 Study site and research ethics…………………………………………………………………………….…………………

22

3.2.2 Phase 1 qualitative exploration, analysis, and limitations…………………………………….…….….……

25

3.2.3 Phase 2 quantitative questionnaire, analysis, and limitations…………………………….……….………

29

3.3.4 Triangulation of results and general limitations………………………………………………….………………

33

Chapter 4: Results and Analysis……………………………………………………………………………………….………………

34

4.1 Demographics and general ICT use………………………………………………………………………….…………….…

34

4.1.1 Vulnerability Context of the villages…………………………………………………………………….…………..…

34

4.1.2 Mobile phone access and adoption………………………………………………………………….…………………

42

4.1.3 General phone uses and use dynamics……………………………………………………………….………………

45

4.1.4 Technoliteracy and challenges of ownership…………………………………………………….……………..…

48

4.2 Financial and social capital in relation to mobile phones……………………………..…………….………….…

50

4.2.1 Managing and transferring financial capital and Mobile Money……………………………….…………

50

4.2.2 Healthcare, Education, Social Events, and Government………………………………………….………..…

59

4.3 Agriculture Production: Inputs, Knowledge Transfer, and Organisation of Labour…….…..…………

62

4.3.1 Agriculture production challenges and processes………………………………….………………..………..…

62

4.3.2 Organising labour………………………………………………………………………………………………………..……..71

4.4 Yam Sale: Pricing and Sale Dynamics………………………………………………………………………….…..……..…

78

4.4.1 Pre-sale context: storage, farmgate vs market………………………………………………..….………………

78

4.4.2 Yam sale challenges and sale dynamics…………………………………………………………….…………..

…...82

4.5 Principle Component Analysis and Cluster Analysis……………………………………………….……..

………….94

4.5.1 Cluster descriptions……………………………………………………………………………………….…………….…….94

4.5.2 Cluster demographics and relation to outcomes………………………………………….

…………………..103

Chapter 5: Discussion……………………………………………………………………………………………………….……………111

5.1 ICT4D Theory building: ICT capitals………………………………………………………………….……..………………

111

5.2 ICT4D Theory building: Processes and outcomes…………………………………………………….

……………..117

5.3 Application of theory and higher level findings………………………………………………….……………….

….121

Chapter 6: Conclusion………………………………………………………………………………………………………….………..125

6.1 Limitations of this research and reflection on methods………………………………………….…….…….

….125

6.2 Summary of findings…………….……………………………………………………………………………….…………….…

127

6.3 Contribution to research………………………………………………………………………………………………….

…….127

6.4 Recommendations for future development and research…………………………………….

………………..129

List of Tables Table 1. Specific study community locations and agro-ecological zones…………………………………………...24Table 2. Total variance explained for livelihood strategies…………………………………………………………….….96Table 3. Rotated Component Matrix (livelihood strategies)………………………………………………………………97Table 4. Reliability analysis for Component 1: Phone Users……………………………………………………………....99Table 5. Reliability Analysis for Component 2: Selective phone users………………………………………………..99Table 6: Reliability Analysis for Component 3: Non-phone users……………………………………………………..100Table 7. Cluster Distribution……………………………………………………………………………………………………….….102Table 8. Cluster profiles ……………………………………………………………………………………………………….………..102

List of Figures Figure 1. The ICT4D Value Chain (Heeks and Molla, 2009) …………………………………………………………………11Figure 2. Sustainable Livelihoods Framework (DIFD, 1999) …………………………………………………………….…13Figure 3. Duncombe (2006) model of information and ICTs within the SLF……………………………………..…14

Figure 4. Study site community locations……………………………………………………………………………………….…24Figure 5 Highest level of education achieved in relation to gender………………………………………………….…37Figure 6. Farm size in relation to gender……………………………………………………………………………………………38Figure 7 Change in farm size over the past five years…………………………………………………………………………38Figure 8. Change in monthly income over the past five years……………………………………………………….……40Figure 9. Monthly income in relation to gender………………………………………………………………………..………40Figure 10. Importance of reasons to send and receive money outside of the community………………..…41Figure 11. Years using a mobile phone in relation to gender…………………………………………………………..…45Figure 12. Importance of purposes for using a mobile phone………………………………………………………….…46Figure 13. Frequency of transferring money to individuals living outside of the community………………50Figure 14. Difficulty of sending and receiving money from people living outside of the community today………………………………………………………………………………………………………………………………………….……51Figure 15. Change in ability to send and receive money from people living outside of the community over the past 5 years…………………………………………………………………………………………………………………..……51Figure 16. Mobile Money kiosk …………………………………………………………………………………………………..……52Figure 17. Importance of strategies for transferring money to and from individuals outside of the community five years ago…………………………………………………………………………………………………………………54Figure 18. Importance of strategies for transferring money today to and from individuals outside of the community at present……………………………………………………………………………………………………..…………55Figure 19. Difficulty of sending and receiving money from people living outside of the community in relation to importance of the strategy: Mobile money to transfer money…………………………………………56Figure 20. Farm size in relation to importance of the strategy: Mobile money to transfer money……...57 Figure 21. Demonstration of production practices in the community…………………………………..……………65Figure 22. Difficulty of organising labour….…………………………………….…………………………………………..……72Figure 23. Seriousness of challenges in organising labour……………………….…………………………………..……73Figure 24. Change in the challenge of organising labour over the past five years…………………………….…74Figure 25. Importance of strategies for organising labour five years ago……………………………………………75

Figure 26. Importance of strategies for organising labour at present…………………………………………………76Figure 27. Strategies for organising labour in relation to the difficulty of organising labour…………..…..77Figure 28. Yam market containing an example of a tractor/trailer used for yam transport…………………79Figure 29. Farm size in relation to ability to carry out temporal arbitrage when selling yam……………….79Figure 30. Average monthly income in relation to ability to carry out temporal arbitrage when selling yam……………………………………………………………………………………………………….…………………………………………80Figure 31. Difficulty of selling yam at present……………………………………………………………………………………83Figure 32. Change in difficulty of selling yam over the past five years……………………………….………………83Figure 33. Significance of challenges of selling yam………………………………………………………………..…………85 Figure 34. Average monthly income in relation to difficulty of obtaining reliable yam sale price information………………………………………………………………………………………………………………………………..……90Figure 35. Importance of strategies for obtaining yam sale price information at present…………………..91Figure 36. Importance of strategies for obtaining yam sale price information five years ago……………..92 Figure 37. Difficulty of selling yam in relation to strategies for obtaining yam sale price information...93Figure 38. Clusters in relation to years using a mobile phone…………………………………………………………104Figure 39. Clusters in relation to frequency of mobile phone use………………………………………………….…104Figure 40. Clusters in relation to purposes of using mobile phones……………………………………………….…105Figure 41. Clusters in relation to farm size (acres) ………………………………………………………………………..…106Figure 42. Clusters in relation to monthly income……………………………………………………………………………107Figure 43. Clusters in relation to challenges in organising labour………………………………………………….…108 Figure 44. Clusters in relation to challenges in selling yam………………………………………………………………109Figure 45. Challenge of being manipulated by middle-women in relation to the importance of…….109Figure 46. Phone, Information, and non-ICT capitals framework………………………………………………….…114Figure 47. Framework for ICT4D processes and strategies………………………………………………………………118

List of Appendices

APPENDIX 1: University of Surrey University Ethics Committee confirmation of ethical opinion…...…141APPENDIX 2: University Ethics Committee Ethics Application Form…………………………………………...…..143APPENDIX 3: Questionnaire…………………………………………………………………………………………..…………….…186APPENDIX 4: Conference Poster: “Impacts of ICT on Agriculture Based Livelihoods in Ghana”…..…....194APPENDIX 5: Conference Poster: “Mobile phones and sustainable agriculture based livelihoods: A case study of smallholder farmers in the Ashanti Region, Ghana” ……………………………………………………….…195

List of AbbreviationsAEA Agricultural extension agents

CA Cluster analysis

CAY-SEED Community action in improving the quality of farmer saved seed yam

CRI Crops Research Institute (Ghana)

CSIR Council for Scientific and Industrial Research (Ghana)

GT Grounded Theory

ICT Information and communication technology

ICT4D Information and communication technology for development

MoFA Ministry of Food and Agriculture (Ghana)

PCA Principal Component Analysis

SLA Sustainable livelihoods approach

SLF Sustainable livelihoods framework

Chapter 1 - Introduction1.1 Motivation for researchThe inspiration for this research occurred during the field work and data collection portion of my

master’s thesis. I was assisting a colleague in interviewing fisherman fishing out of a small

community on the coast of North Sulawesi, Indonesia. These fishermen explained how the adoption

of mobile phones had transformed their fishing practices and in doing so, transformed their

livelihoods. Over just a few years, the fishing community had expanded from just a few small

individually operated boats to more than 25 large boats operated by crews. The mobile phone and

the connectivity it provided made it so that these large boats could be profitable in that fishermen

would only launch and operate their crafts following a phone enabled notification that fish were

schooling from a sentry fisherman located offshore. Furthermore, fisherman used their phones to

contact potential traders who would then arrive at the shore, increasing market efficiency and

decreasing waste. The economic boon that these fisherman were capturing had not only improved

their livelihoods but the small community had experienced an improvement in civil society, social

connectivity, and improved healthcare through Red Cross groups and infrastructure. Throughout my

history of fieldwork I have encountered many examples of top-down introductions of technology

into communities in developing countries, and nearly all of them have been failures. This experience

with the fishermen had a profound effect on me as it was a first-hand account of the transformative

power of a new technology implemented from a bottom-up perspective. I therefore wanted to

design and carry out a PhD project involving other scenarios where mobile phones and other ICTs

were being used without the push of a development initiative. While development initiatives are

important, and can produce positive outcomes, I am also aware that we need to understand how

spontaneous and bottom-up uses of technology are changing the development landscape.

1.2 ICTs and the developing world

1

The use of information and communication technologies as a means by which to achieve poverty

reduction and development is an integral part of the current and future development agenda (Ban,

2014). Despite a period of heavy investment and a string of significant development initiative failures

focused on ICT4D in the early to mid-2000s (Lydon and Williams, 2005), on a private level mobile

phones continue to be adopted (Donovan, 2011) and used to carry out a range of development

strategies across the developing world. Brezis et al.’s (1993) concept that the use of the technology

can produce a leap-frogging effect wherein individual’s skip steps within the traditional development

paradigm has persisted in the discourse around ICT4D investment and research (Van der Boor et al.

2014).

The use of ICTs in development has been shown to decrease individual’s vulnerability to shocks in a

general sense (Skuse and Cousins, 2007) with specific examples of positive effect including an

increased ability for individuals to cope with emergencies (De Silva and Zainudeen, 2007).

Information accessed and communicated using ICTs has been shown to be a means for increasing

social capital (Thapa et al., 2012) by strengthening ties between family members and managing

distance between family members living apart (Nagaska, 2007). This is in part due to the use of

mobile phones as a substitute for travel (Samuel et al. 2005). While aspects surrounding the use of

ICTs in agriculture development have received little attention when compared with other aspects of

livelihoods (Gomez 2013), the use of ICTs in the sale of agriculture products has been shown to

improve farmers’ ability to negotiate sale prices (Minten et al., 2012), broaden trade networks

(Guislain and Qiang, 2006), increase productivity (Lio and Liu, 2006), and production planning (Ali

and Kumar, 2011).

That said, barriers to effective ICT adoption and use persist throughout the developing world

resulting in a phenomenon labelled: the digital divide (Compaine, 2001). Differential access to and

use of ICTs has been shown to be determined by age and education (Tadesse and Bahiigwa, 2013;

Dissanayeke and Wanigasundera, 2014; Agwu et al., 2008). Individuals can also be prevented from

2

accessing and using ICTs due to a lack of technoliteracy wherein they lack the skills to operate

devices (Boyera, 2007).

1.3 Research gaps, scope of this research, and research questionsGaps in the field of ICT4D research are typically centred on a need to connect the use of ICTs and

development processes and outcomes (Urquhart et al., 2008). That said, there is room for

improvement in nearly every aspect within the field of ICT4D. Currently no unified theory exists

defining the ICT4D agenda in terms of frameworks, methodologies, or methods (Walsham, 2017).

Ideas on how to move towards this goal of creating a unifying theory include: adopting holistic

approaches to impact assessment wherein proceeding stages of ICT adoption and use are assessed

(Sein, 2013), being wary of impacts according to the demographics or groups where they occur

(Baumuller, 2018), and being wary of the difference between correlation and causation between ICT

use and impacts (Steyn, 2013). Research is lacking in terms of explanations of ‘what’ and ‘how’ using

methods and subsequent findings that show the ‘why’ a phenomenon may occur (Omland and

Thapa, 2017). Other gaps in research include studies including comparisons of the importance of

factors involved information needs and constraints (Baumuller, 2013). Furthermore, research

including and assessing the specific contexts in which ICT enabled development is occurring

(Avgerou, 2010). Research is lacking in term of including stakeholders’ views of development (Thapa

and Saebo, 2014), and the inclusion of gender dimensions (Agarwal, 2012). Finally, research on how

agriculture extension agents (AEAs) use ICTs for information transfer with stakeholders in

communities is lacking (Duncombe, 2012).

The scope, methodology, and methods included in this research were designed and aimed to fill or

at least contribute to decreasing these gaps in the ICT4D literature. While this research is composed

on a case study of communities in an agroecological region of Ghana, and defined by a snap-shot

data collection method, it was built on a grounded theory approached aimed at theory building

3

through exploratory data collection methods. Theories produced in this research, once uncovered

through analysis, did not produce a unifying theory for ICT4D, but they should serve as a forward

stepping stone towards a unified theory in the future. This research also placed great emphasis on

capturing the vulnerability context within which the study group existed, and both the internal and

external forces and factors that combine to affect how and why ICTs are more or less useful in

contributing to development. In other words the scope of this research aimed to assess forces and

factors both within and outside the control of individuals in the study group. This research largely

relied on stakeholder views of development in framing challenges and assessing outputs.

As such, the central question that this research sought to answer is:

‘What role does the mobile phone play in the development of agriculture based livelihoods of

farmers in the Ashanti and Brong-Ahafo Regions, Ghana?’

In order to answer this main broad question, several research sub-questions must be answered.

These questions are:

‘How do farmers use mobile phones to address the livelihood challenges that they face?’

‘Does the use of mobile phones in addressing livelihood challenges produce positive livelihood

outcomes?’

‘What is the context of and barriers to the mobile phone being more or less useful in addressing

livelihood challenges?’

1.4 Structure of thesis

The structure of this thesis is as follows:

4

Chapter 1, Introduction – This chapter provides the motivation for this research as well as a very

general overview of the interaction between ICTs and development. This is followed by a description

of the current gaps in the field of ICT4D research and how the scope of this research aims to fill

these gaps. The research questions are provided at the end of this chapter.

Chapter 2, Literature Review (ICT4D Theory and Practice) – Chapter 2 describes the definitions

within and history of ICT4D research and initiatives as well as covering some of the theories that

have informed past ICT4D research. This chapter includes an overview of past research on ICT4D in

the context of agriculture in Africa with specific attention paid to research assessing livelihood

processes in the African context that this research will assess.

Chapter 3, Methodology and Methods – This chapter will start by defining the epistemological,

theoretical, and methodological underpinnings of this research. It will then describe the process of

study site location selection and a description of the study site. Next, this chapter will cover the

collection and analysis of the first phase of quantitative data, followed by a description of the second

phase of quantitative data collection and analysis. This chapter will end with a description of how the

two types of data are combined to derive higher level findings as well as limitations involved in data

collection and analysis.

Chapter 4, Results – This chapter is divided into five sections.

Section 4.1, Demographics and general ICT use – this section will cover the vulnerability

context in which the study group persists, demographics and factors of mobile phone access

and adoption, and basic phone uses. This section will conclude with a description of the

state technoliteracy in study group as well as barriers to use.

Section 4.2, Financial and social issues – This section will cover the use and impacts of

mobile money platforms for transferring money. It will also cover uses of the mobile phone

5

in addressing challenges associated with healthcare, education, social events, and

participation in government.

Section 4.3, Agriculture Production: Inputs, Knowledge Transfer, and Organisation of Labour

– This section will address the role that mobile phones play in farmers ability to address

agriculture production challenges with special attention paid to the interaction between

farmers and AEAs using the mobile phone. This chapter will also describe the use and

impacts of the mobile phone in the process of organising labour.

Section 4.4, -Yam Sale: Pricing and Sale Dynamics – This section will describe the context of

yam sale, as well as the use and impacts of the mobile phone in obtaining yam market sale

pricing information.

Section 4.5, Principal Component Analysis and Cluster Analysis – This section will cover the

outcomes associated with combinations of strategies, both making use of, and forgoing the

use of the mobile phone.

Chapter 5, Discussion – This chapter will present theories describing the capital required for effective

phone use in processes and strategies, as well as the theoretical underpinnings of strategies

involving the use of the mobile phone. This chapter will also include a reflection on the methods

used in this research.

Chapter 6, Conclusion – This chapter will explain how this research contributes to the ICT4D field of

research, general conclusions, and recommendations for future research and ICT4D initiatives.

6

Chapter 2 - Literature Review – ICT4D Theory and Practice2.1 Introduction, definitions, history, and critique of ICT4DThis chapter presents a literature review of the field of Information and Communication for

Development, or ICT4D, including theory, methods, and examples of current and relevant research

findings with a particular focus on the African context.

As wireless communication has continued to move towards near universal adoption worldwide it has

diffused into nearly all aspects of economic and social life. Following uptake, individuals and

communities have used and adapted this cross cutting resource to suit their livelihoods needs and

carry out a wide range of activities. This phenomenon has been called the ‘Fourth Industrial

Revolution’ wherein digital technology is “blurring the lines between the physical, digital, and

biological spheres” (Schwab, 2017), or the transition to an information society where technology and

information interact with and transform society (Webster 2014). As this transition and

transformation failed to occur instantaneously and universally around the globe, the concept of the

technology ‘haves and have nots’ spawned what is currently referred to as the Digital Divide wherein

individuals in developing countries have comparatively less access to and use of technology enabled

educational and business related information (Davison et al., 1999). That said, developing countries

such as Ghana now experience mobile penetration rates of greater than 100%1, up from

approximately 50% mobile subscription penetration rate 10 years ago (NCA, 2018). This change in

technology adoption has occurred at the same time that Ghana’s agricultural sector has experienced

significant, albeit not transformative, growth (Breisinger et al., 2011). On a broader scale,

information and communication technologies are growing faster in Africa than any other region

(Chavula, 2014), while on this continent the majority of farmers derive their livelihoods farming less

1 The mobile phone subscription rate in Ghana was 138.88% in December of 2018. This number is greater than 100% because many individuals possess multiple subscriptions to service providers.

7

than 2 hectares (Eastwood et al., 2010). The role that ICTs and specifically mobile phones can play

within this need for agriculture derived livelihood development in developing countries has only

relatively recently received attention from the research and development initiative communities.

This is at least in part spurred on by a continued belief in the potential and opportunities of ICTs as a

means by which to improve agriculture in developing countries in a general sense (Gates, 2016), but

also in specific terms of production, processing, distribution, and marketing of agricultural products

in Sub-Saharan Africa (Maumbe, 2010).

According to continued calls for transparency and clarity regarding definitions within the field of

ICT4D (Brown and Grant, 2010), this research is rooted in the definition of sustainable development

according to Brundtland (1987, p. 8): “meeting the needs of the present without compromising the

needs of future generations”, and identification of development through the emergent definition

provided by Pieterse (2010) wherein development occurs through changes in quality of life,

empowerment, enhanced basic capabilities, equality, and reduced poverty. Information and

communication technologies will be defined by a combination of Marker et al.’s (2002, p. 4)

definition: “technologies that facilitate communication and the processing and transmission of

information by electronic means” and Souter et al.’s (2005, p. 29) expanded definition that includes

“telephony and its derivatives including data communications, the Internet, computer applications

such as those used for e-commerce and e-government, and products and services derived from

these”. When defining the field of research in accordance to Heeks (2006), this study draws upon the

fields of library and information sciences, communication studies, and information systems from

which ICT4D research draws upon, while assuming that the fields of development informatics and

ICT4D are largely interchangeable. ICT4D is defined as “Individual or groups of communication

technologies whose adoption or impact supports ongoing and/or future development aims and

objectives” according to Heffernan et al. (2018, p. 905). This research focuses mainly on two of

8

Walsham and Sahay’s (2006, p. 10) four areas of research within the field of ICT4D: (1) the

contribution of ICT’s to development2, and (2) Local adaptation and cultivation.

The history of the field of ICT4D reaches back to the mid-1980s wherein researchers within the field

of information systems (IS) began studying the “social implications of information systems in

developing countries” (Walsham, 2017, p. 20). By the 1990s the field began to emphasize the

importance of the developing country contexts in which ICTs were being implemented (Robey et al.,

1990), and the outlook that technology is not only a monolithic tool or entity, but instead a

multifaceted and changing force that produces both intended and sometimes unintended

consequences (Sein and Harindranath, 2004). The new millennium ushered in both the more clearly

defined and differentiated field of Information and Communication Technology for Development as

well as a wave of new technological optimism and investment (Wade, 2002). ICTs were seen as

having the transformation potential to subvert hierarchical power structures (Wilson, 2003) and

provide developing countries with the ability to ‘leap frog’ traditional development processes

toward an ‘information economy’ (Heeks, 2002a). The early-2000s also marked the start of the

alignment of the global development agenda and ICT4D initiatives which continues to this day

(Heeks, 2008; UNSDN, 2016). That said, the boom of ICT4D popularity and investment of the early

2000s quickly faded as greater than half of all ICT4D related initiatives were either complete or

partial failures (Heeks, 2002b) resulting in a commitment to the view that ICTs are not a ‘silver bullet’

for development (Aker and Mbiti, 2010). These failings prompted critiques including: a focus on

technology in terms of both research and investment in interventions at the exclusion of

development (Walsham, 2013; Thomson, 2004) or that ICTs may encourage unsustainable trends

(Souter, 2012). Currently, ICT4D research exists in a period of proliferation as both individuals and

developing country governments have invested in ICTs (Duncan-Howell and Lee, 2008).

2 Walsham and Sahay (2006) clarify their category of research on Information and Communication Technologies contribution to development as “sometimes in the context of a specific country”. This research fits that category with its clarification.

9

Despite the examples of ICTs having a negative effect on development, such as failing to return

benefits from costs (Mpogole et al., 2008), exacerbating inequalities (Heeks, 2010), or lacking the

ability to enhance capabilities (Zheng and Walsham, 2008), other research suggests a process in

which ICTs fail to create transformational change but instead result in degrees of change (Donner

and Escobari, 2010) or progressive change (Avergou, 2008) while being sustainably neutral

(Batchelor et al., 2014). Today research on and emphasis of use of ICTs for development tends to be

derived from smaller scale case studies which remain cognisant of the context and potential barriers

to effective ICT use. These barriers are often defined according to access, such as a lack of electricity

infrastructure (Faola and Adewumi, 2012), local content (Boateng, 2012), and use including a lack of

literacy (Furuholt and Matotay, 2011) and a lack of trust in non-face-to-face communication

(Duncombe, 2016). The lessons and experiences described above have led to the need for, and

development of, a wide range of research theories and methodologies which are covered in the

following section.

2.2 ICT4D Research theories, methods, and ICT and the Sustainable Livelihoods Framework “Theory is the critique, revision and summation of past knowledge in the form of general

propositions and the fusion of diverse views and partial knowledges in general frameworks of

explanation” – Jan Nederveen Pieterse (2010, p. 2)

Calls for the need for the use of appropriate theories for research in ICT4D began in the mid-2000s

(Heeks, 2006) and since then interest in theories has continued to grow within the field (Gomez et al.

2012). That said, calls continue for both an underlying unifying construct (Heffernan, 2018), and for

researchers to be more explicit when considering and defining their philosophies and research

paradigms (Gomez and Day, 2013). In response to this, Heeks and Wall (2018) highlight studies in the

field of ICT4D that make use of two paradigms: interpretivism (ex. Alao et al., 2017) and positivism

10

(ex. Erumi Esin and Heeks, 2015), while putting forward critical realism as a third option as an‐

improved approach for addressing causality between the use of ICTs and development.

Here it should be stated that throughout the course of ICT4D research, researchers have mostly

foregone older communication theories in favour of theories for development. Accordingly these

theories often blur the line between methodological approaches and tools for the interpretation of

phenomena. For example, commonly used theories in ICT4D include: actor network theory, Sen’s

(1982) capability approach, and the sustainable livelihoods approach (SLA). Additionally, Heeks and

Molla’s (2009) adapted input process output model deemed ‘The ICT4D Value Chain’ divided the

linked resources and processes required for ICT4D initiatives into four categories: readiness,

availability, uptake, and impact including outputs, outcomes, and development impacts (Figure 1).

Figure 1: The ICT4D Value Chain (Heeks and Molla, 2009)

This ICT4D value chain is helpful in that due to its holistic nature, it can be used to point out gaps in

assessment in the theories stated above, mainly that the actor network theory fails to assess impacts

of technology, the capabilities approach fails to assess causality between technology and ambiguous

capabilities, and the SLA does not address uptake (adoption and diffusion).

Researchers have also sought to identify and define broad categories and discourses within ICT4D.

Donner (2008) proposed the categories: (1) determinates of adoption, (2) impacts of use, (3)

interrelationships between technology and users, and (4) studies largely focused on economic

development. That same year, Avgerou (2008) proposed a set of alternative discourses including: (1)

diffusion, (2) social embeddedness, and (3) transformation. Duncombe (2011) went on to describe

11

the importance of different scales: macro, meso, and micro-level assessments of ICT4D impacts, and

how the scope of theories relates with these scales so as to express impacts (Avgerou, 2017)

supported by appropriate indicators for each scale (Qureshi, 2015).

Furthermore, past failures of ICT4D assessments, particularly impact assessments, have provided

insight into the best practices within the use of theory, methodologies, and specific methods.

Researchers have called for the clear delineation of the technology itself: the mobile phone as an

artefact, the information provided and used (Zheng and Heeks, 2008), and m-services (Nakasone et

al., 2014). Research often also lacks context (Dodson et al., 2013) by not taking into account

motivation and ability to use technologies (Toyama, 2011) and socio-political influences (Cieslik et

al., 2017). From a methods perspective, Duncombe (2011) points out the advantages and

disadvantages of qualitative, quantitative, and mixed methods approaches to studying ICT4D, and

stresses the importance of identifying the direction(s) of causation between the mobile phone

(variable) and other variables. Finally, Grunfeld (2007) points out that most ICT4D research remains

focused on the past, and suggests future research be more forward looking in nature.

Despite the critique of the Sustainable Livelihoods Approach for measuring the impact of ICTs on

development stated above, the use of the SLA has been defined and carried out in the context of

ICT4D research (Chapman et al., 2003; Souter et al., 2005; Duncombe, 2006; Parkinson and Ramirez,

2006).

Chambers and Conway (1991, p. 6) defined sustainable livelihoods as:

“the capabilities, assets (stores, resources, claims, and access) and activities required for a means of

living: a livelihood is sustainable which can cope with and recover from stress ad shocks, maintain or

enhance its capabilities and assets and provide sustainable livelihood opportunities for the next

generation”.

12

The UK’s Department for International Development (DFID) integrated this definition of sustainable

livelihoods with its new policy and goals in the late 1990s into its version of the sustainable

livelihoods framework (Figure 2.).

Figure 2: Sustainable Livelihoods Framework (DIFD, 1999)

The Sustainable Livelihoods Framework (SLF) 3 focuses on how the vulnerability context in which

poor households persist affects access to and use of various forms of capital assets through

transforming structures and processes, and livelihood strategies to produce outcomes (Carney,

1999). The approach is flexible, people centred, and holistic in nature, aiming to capture the context

in which individuals exist as well as the multiple strategies and subsequent, sometimes

unanticipated, outcomes and impacts that they produce (DIFD, 1999). Farrington (2001) points out

that the SLA can be used as (1) guiding development initiatives, (2) a framework for analysis, or (3) a

development objective.

While theories behind the integration of ICTs into the SLA, and the use of the SLA to investigate the

role and impacts of ICTs on livelihoods began in the early to mid-2000s (Albu and Scott, 2001;

Chapman et al., 2003; Souter et al., 2005), Duncombe (2006) built on previous theories with his 3 The Sustainable Livelihoods Framework can also be referred to as the Sustainable Livelihoods Approach (SLA). These phrases will be used interchangeably in this thesis.

13

application of the framework to define both the role that the SLF can play in ICT4D research, and

how ICTs can be applied within livelihood strategies to produce positive outcomes. Duncombe

(2006) first states that information and communication are needs of poor and should be considered

prior to an assessment of applied ICTs, and goes on to makes use of two of the functions defined by

Farrington (2001) above in that: ICTs, but mainly information, captured within the SLF serve an

analytical role, or means by which to access and assess empirical evidence, and a functional role

which focuses on how individuals use information within strategies to produce outcomes (Figure 3.)

Figure 3: Duncombe (2006) model of information and ICTs within the SLF

With regard to the external environmental trends, shocks, and seasonality that make up the

vulnerability context in which individuals exist (DIFD, 1999), Duncombe (2006) claims that ICTs can

play a role in analysing and communicating this information among involved stakeholders so as to be

able to take action when sufficient assets are available. Admittedly, this seems to conflate various

14

elements of the SLF: vulnerability context, assets (ex. Information), and strategies. Aside from

environmental capital, Duncombe (2006) provides a selection of ways that information and ICTs can

be applied to the capital assets included in the SLF:

Human capital: skills, knowledge, and ability to work, as well as information about health, education, and skill acquisition

Financial capital: Monetary resources including savings, remittances, and credit as well as information regarding ability to access and transfer finances

Social capital: Social organization and networks, as well as the trust and communication built and carried out among individuals

Physical capital: Energy, water, transport, shelter, and technological communications infrastructure

Structures and processes involve the information communicated between external organizations,

which can often mediate information, and individuals, and the information required to facilitate the

workings of entities such as markets (Duncombe, 2006). With regard to livelihood strategies, which

are the choices and activities that individuals use to achieve their livelihood goals (DIFD, 1999),

Parkinson and Ramirez (2006) stress how information and ICTs enhance/interact with strategies and

outcomes. Duncombe (2006) puts forth a focus on the categorization of different types of

information used within strategies. He adapted Chapman and Slaymaker’s (2002) differentiation of

information used for short term or long term decision making, and added additional categories

describing if the information was provided via internal informal or external and mediated formal

information channels.

While Duncombe (2006) applied this information and ICT oriented adaptation of the SLF to the

micro-entrepreneur sector in Botswana, Souter et al. (2005) carried out a broader but more directed

approach to assessing how ICT enabled information flows impact livelihoods using the SLF. Using

established indicators of development and largely quantitative methods, unlike Duncombe (2006)

who made use of mainly qualitative methods, Souter et al. (2005) measured use demographics, and

mobile phone usage and strategic value compared with alternative strategies. And finally, Souter et

15

al. (2005) also focused heavily on impacts of the mobile phone use on social, financial, and human

capital.

2.3 ICT4D and ICT and agriculture research in Africa

The use and impacts of ICTs with regard to social capital has been widely covered, with mixed

results, in the ICT4D literature. The greater scope of communication that ICTs provide has led users

to maintain and improve their social networks, particularly within families (Donner, 2007; Sife et al.,

2010). Boateng (2012) states that the possibility of communication between families divided

between rural and urban areas is one of the main drivers of mobile phone adoption in Ghana. This

finding is repeated in Sey (2011, p. 381) who went on to state that the top two reasons for owning a

phone in Ghana, as stated by users, were: “so other people can contact me” and “so I can contact

other people”. That said, despite Sey (2011) finding that Ghanaians tend not to differentiate

between users social and economic uses of the mobile phone, farmers were less likely to use their

phone for farming business activities compared to social communication.

Literature concerning financial capital and the use of ICTs has largely focused on the use of mobile

banking services, or mobile money transfer services. Kuiri et al. (2013) found that the use of mobile

money services increased household income, agricultural input use, and agricultural

commercialization in Kenya. Kikulwe et al. (2014) also found that mobile money users in Kenya

increased their income but this was due in part to increased receiving of remittances. Sekabira and

Qaim (2016) found that significant variables for mobile money adoption included higher education

levels, larger household size, phone ownership, and distance to mobile money agents. Aker and

Wilson (2013) found that mobile money adoption in Ghana was driven by usage of mobile money

services within an individual’s social network. Both Kuiri et al. (2013) and Narteh (2017) measured

the specific reasons, albeit using different variable descriptors, for financial transfers facilitated by

mobile money services. Kuiri et al. (2013) found that Kenyans used the largest proportion of

16

transferred monies for agricultural inputs followed by school fees and then the purchase of food

stuffs. Narteh (2017) found that in Ghana, mobile money services were mainly used for buying

airtime and sending and receiving money, with only a few individuals surveyed using mobile money

to pay bills or manage financial savings.

The use of mobile phones within smallholder African agriculture production has also received ample

attention from both the research and development initiative communities with a particular focus on

the implementation of ICT enabled agriculture extension services. Issahaku et al. (2018) found that

mobile phone users in Ghana increased their agriculture productivity at least in part due to access to

extension services. That said, farmers still access information on inputs and best practices through

other means of communications such as pre-existing face-to-face interactions with individuals in

social networks (Molony, 2007; Lokanathan and Kapugama 2012), or through input suppliers (Clark,

2012). Despite these alternatives strategies for information communication, as well as the barriers to

ICT enabled extension communication such as a lack of perception of mobile functionality for

extension access (Anaglo, 2014) and possible ICT enabled exclusion due to a lack of literacy (Perez et

al., 2010), ICT enabled extension service initiatives have garnered the attention of governments and

agencies in hopes of replicating previous initiative positive outcomes including increased speed and

quality of information delivery in India (Fu and Akter, 2011) and China (Zheng et al., 2016). Munthali

et al. (2018) carried out an extensive assessment of ICT use within the agriculture extension system

in Ghana including the use of both private and public platforms as well as formal and informal ICT

enabled communication among extension personnel and farmers. They found that despite the main

function of both platforms: SmartEx (Grameen Foundation) and E-extension (MOFA), being farmer

registration and production data record keeping, these platforms largely failed to achieve their

ambitious goals of providing farmers with improved access to information including weather,

transport, market prices, best practices, and problem diagnosis. Additionally, MOFA’s E-extension

platform served as an ICT enabled upgraded version of Ghana’s Fertilizer Subsidy Programme

wherein registered farmers receive an identification code that can be used to allocate and track their

17

purchase of subsidized fertilizer from private input suppliers. Munthali et al. (2018) found that the

initiatives suffered from extension officers inability to populate, maintain, and capitalize on the

platform’s databases due to a lack of funds to travel to communities, afford internet credit bundles

for their devices, and platform data collection requirements that extended beyond the traditional

role of extension officers. Farmers made little use of the weather information provided by the

platforms due to illiteracy (inability to read SMS text message) and the greater timeliness of weather

information provided by radio services. Due to these barriers and challenges, extension officers

often fell back on traditional means of provision of subsidized fertilizer, and communicating with

already established farmer networks wherein face-to-face communication dominated the process of

extension. That said, Munthali et al. (2018) found that informal ICT enabled communication

platforms such as WhatApp emerged, at least in a limited capacity, to connect officers and other

stakeholders along the agriculture value chain.

Research on the role of ICTs in organising hired labour for agriculture production has been extremely

limited. Ogutu et al. (2014) perhaps provides one of only a few research projects that have been

carried out on this subject. They found that participants in an ICT-based market information services

(MIS) initiative produced greater labour productivity but lower labour usage than non-participants.

Balasuriya and de Silva (2011) assessed a computer and mobile phone facilitated database for local

labour supplies in rural Sri Lanka. They found that almost no farmers made use of this service due to

a lack of trust in information communicated via the phone. Currently, no research appears to have

taken a nuanced approach to assessing the variety of strategies, including those that involve the use

of ICTs, which farmers employ in order to organize labour for agriculture production. This dearth of

research comes at a time when demand for labour currently outstretches supply in Ghana (Ngeleza

et al., 2011), and yam production expansion is currently limited by increasing scarcity and high cost

of hired labour (Mignouna et al., 2015).

18

While research has not covered ICT enabled labour organization, the provision of information

regarding market prices accessed via mobile phones has received a significant level of attention in

the ICT4D community. The use of ICTs in agricultural marketing has reduced search costs and

increased market efficiency (Aker and Mbiti, 2010), reduced price dispersion between markets (Aker,

2010), and improved decision making with regard to planting (Baumuller, 2015) and harvesting times

(Lokanathan et al., 2011). The issue of bargaining power as a function of access to market

information is currently a topic of debate in the ICT4D literature. Baumuller (2012) showed that

informed farmers did not have a significant bargaining advantage, while Krone et al. (2015) states

that the use of mobiles can increase the number of possible business partners thereby improving

individuals’ ability to bargain. Furthermore, barriers to the use of market information for improving

the sale of agricultural products include a lack of ability to carry out market arbitrage due to a

dependency on the nearest market (Fafchamps and Minten, 2012), the perishability and lack of

storage facilities for produce (Tadesse and Bahiigwa, 2015), and products that require visual

inspection prior to sale (Fafchamps and Hill, 2005). Zanello and Srinivasan (2013) assessed the

various means (strategies) by which grain farmers in Ghana access market price information in terms

of both quantity and reliability of information related to differences between expected and agreed

upon sale prices. They found that face-to-face communication with an ‘informant’ (‘word of mouth’)

was the most common price search strategy followed by the use of mobile phones to contact an

‘informant’. Both of these strategies yielded an increased quantity of price information which

contributed to an increase in matching of expected and realized prices. They also found that if a

seller informs a buyer of intention to sell prior to sale, and trust between the buyer and seller,

increases the probability that sale price expectations match actual selling prices.

Hildebrant et al. (2015) took an assessment of the effects of farmers access of market price

information via an SMS text based initiative applied to yam sale in Ghana a step further by including

additional effects on, and activities of, traders as well as farmers experiencing ‘spill over effects’ of

increased information in the market system. They found that treatment farmers receiving text

19

messages containing sale price information received higher sale prices via improved bargaining with

buyers when compared with untreated control farmers who did not receive texts containing pricing

information. This in turn changed the bargaining of traders who struggled to differentiate between

treatment and non-treatment farmers, and subsequently offered higher prices when bargaining so

as to improve the ability of non-treatment group farmers to access higher prices. Findings were

controlled for communication between treatment and non-treatment group farmers. Hildebrant et

al. (2015) also found that due to financial constraints, farmers tended to sell closer to harvest as

opposed to having the ability to carry out extensive temporal arbitrage and access higher out of

season prices. Finally, they also found that farmers often complain about being cheated by traders,

and sell, on average, to 3-5 different traders per season.

2.4 Research aims and objectives

This study will attempt to capitalize on the previous research that has been described above as well

as fill several gaps and answer several questions that previous researchers have asked. This research

will both try to strike a balance between a focus between the use technology and development while

being cognizant of all of the challenges of assessing interactions and causality between the two

variables. Despite the criticisms and limitations the SLF stated above, this study will attempt to show

that the SLF remains as a viable theory and framework for assessing ICT4D, and attempt to improve

on previous attempts at applying the SLF to ICT4D research. In the tradition of the application of SLA

to development research, this study will attempt to capture of the nuance and diversity of livelihood

strategies that individuals employ. Weaknesses of the SLF will be addressed by incorporating aspects

of Heeks and Molla’s (2009) ICT4D value chain. Furthermore, the scope of this study will include

both the ‘what’ and ‘how’ ICTs affects users’ livelihoods but also attempt to explain the ‘why’

question via an assessment and analysis of context of, and barriers to, use. Power structures and

gender issues including empowerment have been included, at least to some extent, throughout the

process of this research and analysis. Currently research is severely lacking with regard to the

20

specifics of how small holder farmers in Africa access the critically needed input: hired labour. This

research will attempt to provide the first nuanced assessment of the strategies, including the use of

the mobiles phones, which farmers employ in order to access this resource. The summation of these

goals and subsequent activities carried out in order to achieve this study should provide a significant

contribution to answering the ever evolving question: ‘What is the role of ICTs in sustainable

development?’

Chapter 3 - Methodology and Methods

3.1 Research MethodologyThis section outlines the methodology and methods used in this research. As stated above, research

in the ICT4D field has often failed to clearly define its epistemological, theoretical, and

methodological underpinnings. This is no doubt due, at least in part, to the complex nature of the

interaction between technology, humans, and development itself. This problem is further

exacerbated by researchers’ failings to differentiate between frameworks as theories in themselves,

or as methods by which to conduct research. Furthermore, this lack of epistemological and

theoretical basis can trickle down so as to mire assumptions and limitations of studies in obscurity.

For these reasons, this section will attempt to maximize transparency, including potential

weaknesses, in how this research was built and carried out.

This research was largely built on an inductive process. This was so much the case that when trying

to answer ‘what role do mobile phones play in livelihood development?’, it was never assumed that

mobile phones play a role at all in livelihood development. This perspective was maintained

throughout the entire process of the project design and implementation via continuous cross

checking and comparison of factors related to mobile phones with factors unrelated to mobile

phones. From the beginning, the purpose of this research was to discover and bring together

evidence, observations, and a posteriori knowledge to form a contextualized and nuanced view of

the relationship(s) between variables. The epistemological basis for this research largely employs

21

constructivism with a theoretical underpinning of interpretivism. According to these perspectives,

the truths within the world that this research represents were given meaning and value by the

subjects in it (Walsham, 1995). As this research will show, individuals can and do hold different and

sometimes contradictory views of the world, while maintaining validity. For example, from the

outset mobile phones and the access to information that they provide, were not viewed as

inherently valuable or having an intrinsic meaning.

As such, elements of the highly flexible sustainable livelihoods approach were applied to a grounded

theory approach (Glaser, 1978) to provide the methodological basis for this research. Here it should

be stated that while the SLA can serve as a theory of development in itself, for the purposes of this

research it was mainly used as an inspiration for exploring the context and elements of individuals’

livelihoods. Grounded theory provided the central methodology for uncovering emergent theories

and findings via layers of meaning (Charmaz, 1994) that are less explicit in the SLA.

As stated above, Duncombe’s (2006) qualitative application of ICTs with the SLA, and Souter’s (2005)

quantitative assessment of ICT uses within livelihoods in relation to widely established development

indicators both served as inspirations for this research. While these studies did not explicitly state

their epistemological, theoretical, or methodological perspectives, they did provide a basis for

appropriate research methods. Souter et al. (2005) in particular, provided direction in that they

carried out an extensive consultation process with in-country experts wherein they designed their

questionnaire with considerations of the SLA. This study employed a similar design in that it occurred

largely through two phases. The first phase was largely exploratory in nature involving qualitative

interviews. This was followed by a content analysis and informed questionnaire design. The second

phase involved a qualitative questionnaire and subsequent statistical analysis of that data. While

these two phases of data collection occurred approximately one year apart, this study is defined by a

cross-sectional or ‘snapshot’ approach. These sources of data as well as in-field experience and

observation were aimed at facilitating methodological and analytical triangulation. As such this

22

research can be defined as employing a mixed methods approach. The following sections will detail

the design and implementation of the specific methods, as well as how those methods were

informed by the higher level epistemology and theory described above.

3.2 Research design and data gathering methods3.2.1 Study site and research ethics

The study location was selected and accessed via consultation and in collaboration with the Crops

Research Institute (CRI), part of the Council for Scientific and Industrial Research of Ghana (CSIR)

which operates under the Ghanaian Ministry of Food and Agriculture (MoFA). During the time of this

research, CRI was carrying out the Community action in improving the quality of farmer saved seed

yam (CAY-seed) initiative. This initiative provided the research infrastructure to make this project

possible as both researchers from CRI, and MoFA agricultural extension agents (AEAs) had access to,

and extensive familiarity with, the communities from which study participants could be drawn from.

The selection of communities was also based on the expertise of Dr. Kingsley Osei4 who assisted in

selecting communities that were similar in economic status, organisation, mobile phone coverage,

and access to markets. Secondary reasons for selecting communities were: accessibility via car or

motorbike and the jurisdiction of specific participating AEAs. The latter justification was important

because the local AEAs would serve as assistants and translators for both the interview and

questionnaire phases of this research. The local AEAs with the best research and translation abilities

were selected. The research communities were also selected based on the fact that approximately

every community member grew yam (Dioscorea rotundata), for both home consumption and for

sale, making up the majority of their income. Finally, both the general study area and specific

communities therein, were selected based on a general idea that nearly everyone in the community

either currently owned a mobile phone, had at one time recently owned a mobile phone, or had

4 Dr. Osei is the head of the department of the Plant Health Division at the CRI. He has expert knowledge of the study area and served as an ‘in-country supervisor’ for this project. Dr. Osei provided significant assistance in terms of planning, liaising, and logistics in the field.

23

access to, and knowledge of, mobile phone use. This selection idea arose during the selection

process with Dr. Osei and other officers with detailed knowledge of the potential study group5.

Eight communities were selected as specific data collection locations: four communities in the Ejura-

Sekyedumase District of the Ashanti Region, and four communities in the Atebubu-Amantin District

of the Brong-Ahafo Region. Figure 4 provides a map of the study site communities.

Figure 4: Study site community locations

The study site communities and the agro-ecological zones that they inhabit are provided in table 1.

Region District Community Agro-ecological zone

Ashanti Ejura-Sekyedumase Mesuo Humid forest

Kramokrom Humid forest

Nyinasei Humid forest

Bisiw 1 Humid forest

Brong-Ahafo Atebubu-Amantin Watro Humid forest

5 The results chapter bares out this study site section variable. Approximately every individual in the study area communities community either currently owned a mobile phone, had at one time recently owned a mobile phone, or had access to and knowledge of mobile phone use.

24

Asanteboa Humid forest

Mem Derived savanna

Arbour Humid forest

Table 1: Specific study community locations and agro-ecological zones (Mignouna et al., 2014, p.34)

While Table 1 provides specific agro-ecological zones, the study area can also be characterized as the

transitional zone (EPA, 2011), which due to the removal of forest cover in favour of agriculture

production can also be viewed as derived savannah (Codjoe, 2006). Here it should be noted that

Table 1 shows that the community: Mem falls under the agro-ecological zone: ‘derived savanna’

while all other communities are defined as ‘humid forest’. While this is technically true, every

community in the study area had approximately the same climate and agriculture production

conditions. Every community exists on a boundary between the two agro-ecological zones: derived

savanna and humid forest, and therefore this differentiation of community descriptors was

negligible in practice.

Prior to undertaking any empirical research, a full ethical review and approval was carried out.

Following the completion of the first phase of this research, the initial approved ethical review

application was amended and expanded before being approved for the second phase of this

research. This final ethics approval and ethical review application are included in Appendix 1 and 2.

A full ethical review was required because the study population was deemed to be vulnerable,

largely as a function of high levels of illiteracy in the study area. Without exception, this study abided

by the research and ethical protocol defined in the ethical review. Data from this study is currently

being handled according to the protocol defined in the APPENDIX 2: UNIVERSITY ETHICS COMMITTEE

ETHICS APPLICATION FORM, and can be found at the University of Surrey Research Insight Open

Access repository (https://epubs.surrey.ac.uk/). It is my hope that the data derived from this study

may be used by future researchers so as to compare and show how ICT4D continues to evolve and

interact with development.

25

3.2.2 Phase 1 qualitative exploration, analysis, and limitations

The initial phase of data collection occurred between May 30th and June 19th 2016. 48 qualitative

interviews (6 in each community) were carried out with the assistance of a translator (a local AEA).

Only two interviewees spoke English well enough for the interview to be conducted without the

need for a translator. All other interviewees required translators to communicate through both Twi

and local languages. Interview audio was recorded and each interview lasted for approximately one

hour. Participants were selected according to the following protocol: upon entering a community a

random initial participant would be solicited for interview according to the approach and consent

protocol defined in the ethics review. Following the initial interview, and in line with the exploratory,

open ended nature of this project, subsequent participants would be identified and selected based

on a snowball sampling technique (Goodman, 1961). This snowball sampling technique influenced

the gender balance of the interviews as 30 men and 18 women were interviewed. No specific gender

balance target played a role in the design of the sampling process. The robustness of the snowball

sampling technique took precedence over seeking a perfect gender balance and representation.

Proceeding from individual to individual via the snowball sampling technique provided an increased

level of comfort for potential participants when introduced to the project.

The first phase of this research was designed to be largely exploratory. Outside of general trends of

mobile phone adoption and ownership proposed in the planning stage in coordination with the

supervisors above, little was known about the specific uses of mobile phones within livelihood

development by farmers in the study area. While potential interview questions were generated via

past ICT4D research and my own experience successfully implementing the SLA during my Master of

Development Practitioner degree field work, interviews were largely unstructured and exploratory.

Interviews were unstructured in the sense that very few specific questions were repeated

throughout the interview process, and if a participant expressed extensive knowledge of, and

enthusiasm for, a particular topic, interview lines of questioning would attempt to derive as much

rich and detailed information as possible. Additionally, throughout the interview process, trends and

26

patterns emerged when discussing vital livelihood processes and strategies. I attempted to capitalize

on these processes and strategies that involved the use of the mobile phone, while striking a balance

and not specifically avoiding those elements of livelihoods that did not involve the mobile phone.

This was intentional in that apart from being informed that this research was investigating mobile

phones during the pre-interview consent process, interviewees were not ‘primed’ to view the

interview solely in terms of the use of the mobile phone. Interview lines of thought and questions

began with vague prompts about livelihood challenges and strategies, and only following initial

responses would I bring up the mobile phone and its possible relevance. This interview strategy was

aimed at identifying mobile phone use amongst the matrix of alternative strategies and outcomes,

and derived from my experience in interviewing participants in the past in that when a new

technology frames the conservation, interviewees are often quick to extol the virtues of the

technology on a relatively surface level without a deeper consideration for alternative, and

potentially more realistic and effective, strategies. Grunfeld (2011) also considered the potential that

participants may overstate the benefits of technology, in her case referring to a formal technology

based intervention, in her research methods and analysis.

Here it should also be noted that while working in the field I had extensive opportunities to collect

observational data. I participated in multiple extension workshops both in the community and in

yam fields. Of particular note was the interaction between farmers and AEAs both making use of

mobile phones and using traditional communication methods. I was also able to observe yam

markets on multiple occasions. I took extensive field notes of observations while working in the field.

These notes were all technical in nature so as to avoid the bias of involving personal meaning within

observations (De Laine, 2000).

Following the completion of the interviews, the audio of each interview was manually transcribed. I

then carried out a manual content analysis of the interview transcripts wherein findings were

classified, described, and connected (Dey, 2003). Steps in this process followed the direction of Flick

27

(2014) in that the text was reduced and bundled according to common and relevant themes and

concepts, followed by a recognition and possible reconciliation of contradictions, and then finally the

identification of structures and scales within the material. In practice, this analysis was defined by

the following: (1) descriptions of the vulnerability context of the study group with special attention

to technoliteracy, (2) descriptions and classifications of livelihood processes and the various

strategies in which individuals produce outcomes with special attention paid to strategies involving

the mobile phone, and (3) Identification of higher level structures and their effect on and interaction

with the use of the mobile phone in addressing livelihood challenges and producing livelihood

outcomes. I did not seek independent verification of my ‘coding’.

Several research limitations occurred while phase 1 data was being collected and during the analysis

of data following the interview process. The first of these potential limitations involved the use of

AEAs as translators. Interviewees that were familiar with the AEAs may have incentives to either

appear to be achieving in their farming activities either better or worse than what is actually

occurring, better in that they may be attempting to build a relationship, or worse in hopes of gaining

greater attention from the AEA. This is also a potential bias of this study as a whole. Individuals in the

study area have experienced a history of contact with and support (with mixed success) from a wide

variety of government, aid agency, and research organizations. For this reason nearly every person

interviewed expressed a question of what sort of aid this project would provide. I attempted to

mitigate any possible bias that may arise from these circumstances by clearly stating at the

beginning of interviews that the AEAs translating were not functioning as AEAs during the interview,

and that this project was not associated with any source of direct aid. The only other source of bias

that interviewees expressed occurred as a result of being influenced by other individuals lurking

around the interview area. This occurred in only two instances where a male community member

sought to influence the answers that a female interviewee by exerting an unspoken power dynamic

via watching and listening to the interview at a distance. In both instances I was able to immediately

28

identify the problem and have the problem person removed. The small number of responses from

interviewees prior to my identification of this problem were omitted from analysis.

A second limitation of the first phase of date collection involved the snowball sampling technique

and the time of day that individuals were interviewed. On average, 4 interviews were carried out per

working day with interviews beginning in the late morning and wrapping up before sundown. This

meant that interviews carried out in the morning were less likely to capture individuals working in

the field on that particular day. Individuals interviewed in the afternoon and evening had, without

exception, returned to the community from working in the field that morning. All of this is to say

that during particular times of the day when interviews were taking place, the potential population

of respondents was somewhat variable. That said, this potential bias was largely offset by attempts

to interview farmers on ‘taboo days’ when they would remain in the community throughout the day.

Taboo days could vary between communities and among religious groups within communities.

While not persisting necessarily as a limitation of the analysis of the interview data, it is worth noting

that a manual content analysis was carried out in favour of the use of computer assisted qualitative

data analysis software such as NVivo. The first reason for this was the translation and subsequent

transcription process. Translators would often not translate an interviewee’s response verbatim and

instead provide a description of what the interviewee was saying or explaining. The translators

stated that the reason for this was that interviewees’ responses often contained words, phrases, or

explanations that were difficult to translate verbatim. Furthermore, interviewees varied widely in

their willingness to expand in their responses. Responses could be as short as a single word. These

kinds of responses required special attention to the context in which they were provided in order to

produce further meaning. The interviews were also wide ranging in terms of topics and the pace of

questions would vary both between interviews and within interviews. A topic could be exhausted

quickly or take up the majority of an interview. This meant that data could not effectively be

analysed quantitatively, invoking the weakness of ‘’counting’’ that Glaser (1998, p146) describes as

29

an effect of reliance on software. These variables within the structure and nature of the qualitative

data set produced in phase 1 of data collection meant that a manual content analysis provided the

most appropriate and best method for data analysis.

3.2.3 Phase 2 quantitative questionnaire, analysis, and limitations

The second phase of data collection began with the design of the qualitative questionnaire. The

selection of topics and variables to be measured in the questionnaire was largely informed by the

findings of the first phase. Apart from demographic and general mobile phone ownership and use

variables, the first phase of data collection showed that the use of mobile phones played a role in

three major livelihood processes: (1) transferring money, (2) organising labour, and (3) obtaining

yam market sale price information. The alternative strategies, not involving the use of the mobile

phone, for addressing these livelihood challenges, revealed in interviews, were also included in the

design of the questionnaire. Additionally, questions measuring outcome variables: (1) extent of

particular challenges, (2) income, and (3) farm size, were kept separate from questions measuring

the importance of particular strategies. These questionnaire design features were aimed to improve

upon work carried out by Souter et al. (2005) and avoid the potential bias explained above, that

interviewees may be quick to overstate the importance of mobile phones and other technologies in

addressing their livelihood challenges when these variables are linked in a single question.

Furthermore, alternative strategy variables were included in the questionnaire so as to avoid any

bias resulting from an unhelpful excessive focus on the use and impacts of technology alone

(Walsham, 2013).

Data collection using the questionnaires was carried out between April 10 th and May 26th 2017. 20

individuals from each of the 8 communities included in the first phase of this research, 160 in total,

completed the questionnaire. Questionnaire respondents were selected in the same manner that

interview respondents were selected as described above. Here it should be noted that a small

number of individuals that were interviewed in the first phase of data collection also completed the

30

questionnaire in the second phase of data collection. Both interviews and questionnaires were

carried out in a manner in which the specific identity of the participant remained anonymous, but I

can confirm via observation in the field during data collection, that no fewer than a single individual

and no more than three individuals from any given community participated in both phases of data

collection. This was consistent across all communities. Due to widespread illiteracy in the study

group, questionnaires had to be translated in real time using AEAs 6 as translators. I was present

during the majority of the time that questionnaires were being completed, but due to time and

resource constraints, I was not able to be present at all times. The AEAs did not cite any challenges

of data collection.

The quantitative data collected using the questionnaires was analysed using SPSS. Analyses largely

consisted of comparing error bars representing confidence intervals of responses regarding the

importance of individual strategy variables to outcome variables such as ‘extent of challenge’,

‘income’, and ‘farm size’. In order to gain a more nuanced view of how individuals in the study group

address livelihood challenges using combinations of strategies, I carried out a Principal Component

Analysis (PCA) followed by a Cluster Analysis (CA) on livelihood strategy variables. The PCA was used

to reduce the livelihoods strategy variables into components (Wehrmeyer and Parker, 1995), and the

CA was used to identify groups (Anderberg, 2014) and apply cluster identification variables to

individuals employing independent combinations of livelihood strategies. These collections of

individuals defined according to strategy clusters were then compared in relation to the outcome

variables stated above. Prior to the cluster analysis, a Cronbach’s Alpha reliability test was carried

out on the variables contained in components. The inclusion of a component with less than

significantly consistent variable data (α < 0.7) suggests careful and cautious treatment of the results.

The other limitation of the second phase of data collection occurred according to the nature of many

of the response variables in the questionnaire. Many of the questions contained in the questionnaire

6 The AEAs that assisted in data collection were all MoFA extension officers with at least an undergraduate degree and extensive knowledge of both the study area and research techniques including questionnaire based data collection.

31

inquired about the ‘significance’ of a particular sub-strategy carried out in order to achieve a higher

level livelihood process. These kinds of questions set out to assess the relative importance of

particular sub-strategies in relation to alternative strategies as well as in relation to outcome

variables. Several of these outcome variables measured the difficulty of carrying out a higher level

livelihood process. All of these kinds of questions relied on an individual’s perception of the

importance of particular sub-strategies and the difficulty of particular higher level livelihood

processes, as opposed to alternative methods and/or indicators of importance or implementation of

a particular strategy or challenge of carrying out a livelihood process. The strengths and weaknesses

of dealing with individual perceptions and how they reflect reality poses a potential limitation of this

kind of research (Black, 1993). That said, due to the interpretivist nature of this research and

consideration of this potential weakness in data collection and analysis, findings in this study largely

represent a reality constructed by its participants. The value of these results, including the value of

the technology being assessed, is presented through the lens that is the outlook and worldview of

the participants in this study. Questions included in the interviews and questionnaires that were not

based on perceptions, such as income and farm size variables, were designed so as to be compared

to questions based on perceptions specifically so that questions of significance and value of

challenges and strategies could be compared to higher level outcomes not based on perceptions.

These methods afforded the opportunity for an analysis of the interaction of these two types of

variables to potentially form a basis for higher level grounded theory construction.

Within the context of this research the variable: ‘cost of monthly living expenses’ was viewed as

synonymous with ‘monthly income’. The questionnaire did not include a question addressing

‘monthly income’ directly as it was deemed that this question may be too intrusive and respondents

may have either refused to respond or provided an untruthful response. Considering that individuals

in the study group often have little to no savings, ‘monthly expenditure’ should serve as an accurate

proxy for ‘monthly income’, and these concepts will be used interchangeably in this study.

32

Finally, it should be reinforced that the literacy of the study group provided a small limitation to the

process of data collection. Translators translated spoken word in the interviews and the questions in

the questionnaire as precisely as possible and without additional superfluous explanation. That said,

on rare occasions translators were required to add additional explanation and clarification of

questions and responses so as to find an appropriate middle ground between research protocols and

local context and customs. An example of this was the translator having to repeat questions or

clarify the meaning of English words as they were translated to respondents.

3.3.4 Triangulation of results and general limitations

The mixed methods approach described above was designed with the intention that phase 1 findings

would be used to inform phase 2 methods. The final higher level analysis, following the completion

of sub-analyses of each individual phase, employed a triangulation of findings between the

qualitative, quantitative, and observational data. The quantitative data was designed and used to

draw statistically supported conclusions about the larger study population, while qualitative data

and observations were used to provide nuance, context, and the connective tissue among concepts

and variables. Duncombe (2011) states that quantitative studies in ICT4D research often fail to

establish both the definition and directionality of causation between independent and dependent

variables. While the method used in this research views that problem as a possible limitation, the

qualitative data collected in phase 1 provides significant insight into establishing causation between

independent (mobile phones) and dependent variables (outcomes) included in phase 2 design and

analysis.

It should also be stated that while this research provided a ‘snap-shot’ of reality during the time of

data collection, and this should be seen as a limitation, questions in both phases included questions

about the state of livelihoods and mobile phone uses in the past. These responses were based on

memory and perceptions and will be treated as such in the analysis of findings.

33

Finally, the higher level limitations that this research faces involve its balance between a specific

focus on mobile phone technology and livelihood development not necessarily directly related to the

mobile phone. Phase 1 of data collection yielded a significant understanding of the livelihood

challenges, processes, and outcomes that individuals in the study group possess. The most critical of

these issues will be at least provided as context in which livelihoods processes that do involve the

mobile phone persist, as well as examples of areas where the mobile phone is more or less useful.

Chapter 4 – Results and Analysis

4.1 Demographics and general ICT use

4.1.1 Vulnerability Context of the villages

Before narrowing this analysis and discussion to specific livelihood elements related to mobile

phones and farming in later chapters, general livelihood challenges and context warrant

examination. Interviewees were asked about general challenges that their communities face, and

responses mirrored many of the basic challenges seen across the developing world. The most basic

of these challenges were access to water, roads, housing, and sanitation. The results presented in

Chapter 4 are produced via a combination of phase one and phase two data analysis.

Every community in the study group has at least one working borehole which can be used to collect

clean water for various uses. Unfortunately, a single borehole does not meet the demand for water

in a community resulting in long wait times that can disrupt farm work and other day to day

activities. Every community has at least one additional borehole, but in several of these communities

the additional borehole is either completely or partially defective. Participants indicated that this

defectiveness can stem from a variety of problems including: a fluctuating water table and lack of

rain, and initial construction problems. The methods by which individuals overcome the challenge of

lack of water access have varying degrees of effectiveness and safety. Participants stated that in

34

some communities, the inhabitants have attempted to generate collective funds to pay for borehole

repairs, and community leaders have attempted to contact local government officials asking for aid.

Neither of these pursuits have proved successful at overcoming this problem in the past. Other

methods by which individuals attempt to overcome water access challenges include collecting water

from local streams, springs, and dams. This water can at times be polluted by chemical and sewage

run off. Participants indicated that they had some understanding of these risks when collecting

water from outside sources, but many community inhabitants treat these sources of water as they

would the clean water produced using the borehole.

Road access and quality is a constantly changing challenge that the communities in the study group

face. Only two communities in the study group have access via a paved road, and the rest are

located along an unpaved road that consists of a mix of sand and clay. These unpaved roads are

subject to washing out and sometimes complete inaccessibility during the rainy season, as well as

increasing rutting due to increasing traffic from logging trucks and heavy equipment. Community

inhabitants have no ability to address these road access challenges on their own, and rely entirely on

government support for providing resources such as graders to repair these roads. Here it should be

noted that poor roads generally provide a basic inconvenience for individuals to travel to and from

nearby towns because these trips are typically made via bicycle or motorcycle, and these forms of

transportation can more easily traverse a road that is in poor condition. Significant problems arise

when farmers need to transport their crops to market using a car or tractor. It is likely that these

road access problems have contributed to the rise in popularity of transporting products using motor

trikes which are more nimble on poor roads than cars and tractors.

Sanitation and housing conditions were a common topic when discussing community challenges

during the interview phase. No community in the study group had what community inhabitants call a

‘formal toilet’. This refers to at least some kind of building and plumbing infrastructure. Currently,

communities either have informal ‘pit toilets’ or, and in most cases, individuals have no toilet access

35

and make use of the surrounding bush. Understanding the broader social and environmental impact

of these practices was beyond the scope of this project. Housing construction types in the

communities in the study group mainly consisted of stick and mortar walls and tin roofing.

Participants indicated that maintaining and improving their houses was a constant process, and

houses are mainly damaged as a result of extreme wind on occasion. Repairs and improvements to

houses occurs when funds are available, typically directly following the sale of crops. Less than half

of the communities in the study group had access to electricity, and one community (Watro) was

wired for electricity, but it had not been connected to a source. For communities that do have access

to electricity, the main uses and advantages of that electricity is light for cooking at night, charging

mobile phones, and improving security and safety. Only in a few instances were community

inhabitants using electricity for other ventures such as selling cold drinks from a refrigerator, or

running other electronics such as televisions.

Following general community development challenges, general individual development challenges

as identified in the interviews and in no order, largely consist of (1) access to education, (2) access to

healthcare, (2) access to land, and (3) living expenses/income dynamics. The majority of the study

group (56.3%) had no education, and 80.6% of the study group had no education above the primary

school level. The individuals that fall into these categories are largely illiterate. That said, access to

education, especially junior high school and high school, is improving due to a range of factors.

Interview participants indicated that a cultural shift has occurred where parents place greater value

on maintaining their children’s school attendance and selecting school location. Many parents in the

study group have relocated their children to nearby towns where the children live with family and

friends so that they have better access to higher quality schools in town. Participants indicated that

paying school fees is one of their most important expenditures. Furthermore, questionnaire data

shows that income is not a determining factor for access to education. That said, gender does play a

role in access to education and the highest level of education achieved. Women in the study group

36

are significantly less likely to achieve higher levels of education than men, and no woman in the

study group had accessed a high school level education or higher (Figure 5).

Figure 5: Highest level of education achieved in relation to gender (n=160)7

Farm size and access to land pose significant challenges for the study group, and data indicate

several seemingly conflicting trends. Interview participants state that land is increasingly difficult to

access due to increasing population pressure from outside the community and inside the

community. Every individual in the study group traces their family roots to the Northern Region of

Ghana. These individuals emigrated from the dry Northern Region in search of more productive farm

lands in the study area, and this movement of peoples still continues today. Furthermore, it is a

cultural norm for patriarchs to break up their farm land and provide fields for their children once

their children can farm on their own, and at times to provide space for their wives to farm their own

crops. Married women in the study group generally only access farm land through their husband and

this practice is likely a contributing factor for why women in the study group have significantly

smaller farm sizes on average than their male counterparts (Figure 6).7 It should be noted that when considering levels of education, ‘MSLC’ is the now outdated classification approximately equivalent to Junior High School, but these classifications should remain separate due to differentiating characteristics of educations levels over time.

37

Figure 6: Farm size in relation to gender (n=160, error bars represent 95% confidence interval (CI))

That said, questionnaire data indicates that the majority of farmers perceive their farm sizes to have

increased over the past 5 years (Figure 7).

Figure 7: Change in farm size over the past five years (n=160)

38

That said, it should be noted that interviewees stated that they are traveling further distances to

their farm lands than in the past. These trips from the home to the field can be as long as 45 minutes

on foot, which can pose a logistical challenge when considering input and harvest output, and

transport logistics. Farmed land is almost never contiguous, but instead broken up into several

parcels.

The final individual livelihood challenge that participant’s identify as significant to their development

is income expenditure and management. As stated in the methodology, this study will approximately

equate monthly expenditure with monthly income on the basis that individuals in the study group

exist in a state of vulnerability whereby they do not consistently save income. In other words, it is

assumed that individuals in the study group have little or no savings, and little to no liquid assets. For

this reason, ‘income’ and ‘monthly expenditure’ will be used interchangeably, although this

assumption is admittedly simplistic. In other words, questionnaire data showed that the average

monthly expenditure for individuals in the study group is 375 CEDIs per month. When equated to

income, individuals in the study group live on approximately 2.60 USD per day. 98.1% of the study

group indicated that their living expenses has ‘somewhat increased’ or ‘significantly increased’ over

the past five years (Figure 8).

39

Figure 8: Change in monthly income over the past five years (n=160)

That said, women in the study group have significantly less monthly living expenses on average than

men (Figure 9).

Figure 9: Monthly income in relation to gender (n=159, error bars represent 95% CI)

40

Here it should be noted that the relationship between income and farm size is complicated. While

the main source of income for the study group is the sale of agricultural products, many individuals,

both men and women, sell and produce other products such as charcoal and petty items. These

additional economic ventures can bring in supplemental income apart from traditional farming

practices. That said, in some cases, income can be a limiting factor for farm size. For instance, if a

farmer cannot afford sufficient inputs such as fertilizer and pesticide for their current farm size, they

will be forced to farm a smaller area of land. These are just a few of the factors at play in the

potential relationship between farm size and income. In addition to purchasing farm inputs,

participants state that paying ‘school fees’ and paying for ‘health care (for themselves or for their

family)’ are significantly more important than other types of expenditures. Paying for ‘petty items’,

which include food stuffs not produced on the farm, and clothing and supplies, is on average

considered to be an ‘important’ expenditure for individuals in the study group (Figure 10)

Figure 10: Importance of reasons to send and receive money outside of the community (n=160, error

bars represent 95% CI)

41

4.1.2 Mobile phone access and adoption

While mobile phones have been available in Ghana for more than 10 years, it has only been in the

last 5 years that wide-spread adoption has occurred. The average time that individuals in the study

group have had direct access to mobile phone use is 4.96 years. This adoption rate is due to a

number of reasons including coverage, financial capital to purchase phones, and technoliteracy

required to use them. This section will investigate these first two variables of adoption and a later

section will more closely examine technoliteracy. Also, the question of ‘motivation for mobile phone

purchase’ is best answered by describing general phone uses in the following section. Participants in

the study struggled to differentiate the question of ‘what was your motivation for purchase?’ from

the question of ‘what do you use this phone for?’. That said the general responses for the motivation

question all involved the themes: connectivity with family and the use of the mobile to substitute for

travel aimed at connecting with family. These concepts should remain separated for the purpose of

this research because the data show that family connectivity, produced by substituting travel

otherwise required without the use of the mobile phone, using mobile phones exceeds the

connectivity produced by travel in the absence of mobile phone use. For instance, the most common

response to the motivation for acquiring a phone is ‘to keep in contact with family’ which is typically

accomplished by regular calls from one family member to another one as a kind of ‘greeting’. These

‘greeting’ calls typically involve general polite salutations, but at times these general conversations

can lead to more specific questioning and inquiry. In general, regardless of the content of these calls,

this contact would be far less likely to occur were it not for the connectivity provided by the mobile

phone. That said, for research purposes, disentangling greeting calls made for the purpose of non-

directive polite salutations from calls that begin in that fashion and morph into directive inquires,

and those calls that are placed for the sole purpose of directed information inquiry and information

transfer, pose a significant challenge when exploring individuals’ motivations for acquiring a mobile

42

phone. Nevertheless, participants state that the primary motivation for acquiring a mobile phone is

basic connectivity with family.

Individuals in the study group acquire phones through a variety of means, ranging from being gifted

a phone by a family member, to having someone purchase a phone on their behalf, to personally

shopping for and purchasing a phone in the nearby town. Financial issues aside, the specific act of

purchasing a phone serves as a barrier to entry for some in the study group. Several individuals are

intimidated by the phone selection process. These individuals have trouble differentiating both

hardware and software, as well as brands and pricing. This ability to judge phones in a market falls

on a scale between: no ability, to knowing specific features sought but not being able to recognize

them in person, to an extensive understanding of phone quality and construction. Individuals that

struggle with this process will often leave the selection of phones for purchase up to a more

knowledgeable family member to purchase on their behalf. Commonly sought features when

selecting phones include good quality batteries and small size.

The process of purchasing a phone typically occurs in tandem with the purchase and registration

with a network provider. Individuals in the study group generally have stronger feelings and

knowledge of network providers than they do of specific phones. Individuals select providers for a

range of characteristics including: network access (coverage), popularity, credit scratch card

denominations, and sense of brand ‘strength’. The way these variables interact to persuade an

individual to adopt a provider is highly complex and interrelated, but if any through-line can be

drawn it is defined by affordability of use. Providers that offer lower scratch card credit

denominations have an advantage in the market, but most providers offer denominations as low as

one or two CEDIs in phone credit. Every individual in the study group purchases phone credit using

‘pay-as-you-go’ scratch cards which are purchased in town during weekly trips for petty items, but

these cards are increasingly becoming available in communities as individuals become credit vendors

as a means by which to generate supplemental income. Individuals also select providers based on

43

popularity, which roughly translates to affordability in that it is more affordable to make calls to

individuals that are on the same network. This popularity may contribute to the perceived ‘strength’

of a provider, but this variable is also influenced by provider advertising, promotion, and presence.

This results in a kind of brand loyalty and confidence among customers. Some phones used by the

study group have the capability to make use of up to two SIM cards. This allows these uses to switch

between networks/providers in order to manage both: the cost of calling individuals based on

network as stated above, but also accessing different networks based on coverage. Participants state

that in general, network coverage has improved over time, but remains patchy in most communities

and rural areas. Participants overcome this challenge by identifying locations in the community

where network service is perceived to be greater. Individuals will occupy these locations in the

community when attempting to improve their opportunity to make and receive calls. Finally, it

should be stated that while the study group has on average had access to mobile phones for the past

five years, women have been using mobile phones for significantly fewer years on average than men

(Figure 11). This disparity may be due to the fact that women have less income, and therefore

decreased ability to purchase phones, than men. It may also be due, at least in part, to cultural

norms of female access to communication which will be described in section 4.1.4 below.

44

Figure 11: Years using a mobile phone in relation to gender (n=160, error bars represent 95% CI)

4.1.3 General phone uses and use dynamics

This section should serve as an introduction and framing discussion of how and why individuals use

their phones among their livelihood processes. It will also begin to differentiate how to disentangle

phone use in terms of information searching and communication, as well as additional phone

features and functions. Many of these examples of phone uses will be discussed in greater detail in

following sections, but this discussion will provide a general perspective for how crosscutting phone

use is throughout individuals’ livelihoods.

As stated above, a major motivator for phone purchase and subsequent use is ‘greeting friends and

family’. Questionnaire data show that this ‘greeting’ practice, and contacting ‘friends and family to

discuss health issues’ are the two most important functions of mobile phone use (Figure 12).

45

Figure 12: Importance of purposes for using a mobile phone (n=160, error bars represent 95% CI)

These two functions play out on the ground in a number of ways. The first, and most common

circumstance is that children no longer stay with their parents in the community throughout the

week. These children stay in town with friends and family in order to access higher quality schools.

When this occurs, parents, living and farming in the study group communities, typically make at least

one call per day to the children or their caretaker, typically a family member or friend, to check on

their condition. Participants view this daily call as a critical process to maintaining their connection

with their children as well as their children’s health and schooling. Interviewees state that specific

examples of information on these topics often include the state of paying school fees and making

sure that the children are healthy enough to attend school. The second kind of ‘greeting’ call is

typically placed to an individual’s parents or family that resides in the Northern Region of Ghana.

These calls are typically casual salutations, but they often include a check up on the healthcare of

aging parents. The majority of the study group has family living outside of the community, and

sometimes the family is spread out across Ghana. Interviewees state that prior to the advent of the

mobile phone, maintaining these familial connections posed a significant challenge in terms of both

46

time and financial capital. Communication could only occur by traveling to relatives at significant

financial cost or the unreliable practice of sending messages though other individuals traveling

around the country. These calls to relatives living across Ghana often serve the important purpose of

notifying the family of family related events such as births, marriages, and funerals. This

communication allows individuals to assess whether or not they should travel to attend these

events. In the event that they decide to travel, the mobile phone is then used to communicate travel

progress as well as the logistics of arrival and meet up.

As stated above, these non-directed ‘greeting’ calls often morph into more specific discussions of

means by which to improve livelihoods. Interviewees state that these phone conversations

commonly move toward topics such as weather patterns, crop yields, and crop varieties grown. This

communication functions as a means by which agricultural information diffuses throughout Ghana,

though the impacts of this kind of non-directed information transfer are extremely difficult to assess.

Aside from ‘greeting’ calls, individuals in the study group also use their mobile phones for directed

information communication and inquiry. These kinds of calls are placed in order to accomplish a

specific targeted task, typically related to agricultural production, or in some rare cases they are

placed during emergencies. Examples of these kinds of calls include: checking crop price information,

contacting extension officers for assistance, and organising the purchase of petty items or inputs via

someone with immediate access. Examples of emergency calls include: calling transport for a woman

in labour, or calling for assistance when stranded while traveling. These kinds of calls increase

individuals’ ability to carry out production logistics, and improve safety and security.

Outside of information communication and transmission, individuals in the study group use various

features and functions on their phones for both work and pleasure. The majority of phones owned

and used by the study group have applications such as calculators and calendars that can be used

when managing input costs and sale prices. Individuals living in communities that do not have

47

electricity place great value on their phones’ torch function which aids in cooking and traveling at

night. Participants also use their phones to play music and video games for leisure.

4.1.4 Technoliteracy and challenges of ownership

This section will cover challenges of phone use and ownership, as well as gender based access

barriers. For the purposes of this discussion, challenges of use will be limited to technical

technoliteracy challenges. Use challenges relating to individual and structural information

communication and inquiry will be discussed in a later chapter.

In general, the study group possess a low level of technoliteracy as related to the use of their mobile

phones. More than 75% of the study group is illiterate and therefore cannot use SMS text based

phone communication on their own. Individuals with poor eye sight and a lack of access to

corrective lenses also makes texting impossible, and operating the buttons on the phone is highly

challenging. Many of the individuals in the study group overcome this texting challenge by enlisting

family and friends who are literate to provide assistance. For these reasons, individuals in the study

group largely prefer the voice call function for phone based communication. In many instances, the

only function that individuals can operate on their mobile phones is the voice call function. At least

three individuals interviewed were unable to input and save phone numbers using the ‘contact’

function, they overcome this challenge using several methods including: having friends and family

input numbers on their behalf, recalling numbers by identifying and selecting them from the ‘recent

calls’ record in the phone, or writing down contact numbers on paper.

Many individuals in the study group possess a sort of fear in exploring the functions on their phones.

They rely on friends and family to assist them in learning phone operations, but many individuals

retain a fear that they will break their phone by pressing a wrong button or using a feature. There is

a sense of risk aversion when it comes to using the phone that leads individuals to a conservative

mind set of only using voice call functions.

48

Apart of technoliteracy related challenges to phone use, participants report numerous instances of

phone loss, breakage, and operation failure. This is partly due to the nature of individuals’ work and

life environment. At least three interviewees reported that phones often become ‘spoiled’ due to

water damage from being caught in a rain storm while working in the field. Phones are also dropped

in puddles and in soup. Other common reasons for phone breakage include phones being stepped

on or broken by children. Several individuals stated that they have had phones lost or stolen, and

that batteries and torches become weak and unusable over time. These electrical problems may be

due to an unstable electrical grid and problems when charging. In the event that a phone breaks,

individuals in the study group will take the phone to one of the numerous phone repair shops in the

nearest town where on they will receive a quote for the cost of repair. Interviewees reported that

they often weigh this cost of repair against the cost of purchasing a new phone. The average life

span of phones in the community varies, but on average individuals state that a phone will be in

working condition for around two years.

In communities without electricity, individuals charge their phones by taking them into town on their

weekly trips to purchase petty items. Upon arrival in town they will leave their phone at a ‘charging

kiosk’ where the phone will be charged for a cost of around one CEDI. These individuals will reclaim

their charged phone upon their departure back to the community.

The final challenge of phone access and use relates to social and gender norms in the community.

Many individuals interviewed did not own a mobile phone at the time of the interview. For many,

their phone had been broken or lost, and they did not have to funds to immediately purchase a new

phone. These individuals stated that they were averse to asking other people in the community to

use their phones. They were afraid that they might break other peoples’ phones, that paying for and

applying credit would be a challenge, and that they were worried that their calls would not be

private.

49

It should also be noted that several of the women interviewed, owned personal phones, but they

were restricted by their husbands in terms of who they could call. These women were typically

limited to calling only close family and married men that the husband approved of. This

circumstance poses a significant challenge for information access and searching.

4.2 Financial and social capital in relation to mobile phones4.2.1 Managing and transferring financial capital and Mobile Money

This section will cover how farmers in the study group manage and transfer financial capital using a

range of strategies aimed at improving their resilience to shocks and stresses. Interviewees stated

that they tend to possess only small amounts of liquid financial capital at any given time, so the

ability to receive remittances and payments from friends, family, and business partners plays an

important role in both agriculture production and in support of family health and education. While

financial transfers are typically small and occur between ‘a few times a year’ and ‘a few times a

month’ (Figure 13), receiving these funds can be a determining factor in accessing healthcare and

agricultural inputs at critical times.

Figure 13: Frequency of transferring money to individuals living outside of the community (n=149)

50

As stated in the previous chapter, the three most important reasons for transferring money are for

paying children’s school fees, purchasing agricultural inputs, and paying for healthcare. That said, in

a general sense, the majority of the study population stated that transferring money is currently

‘somewhat easy’ or ‘very easy’ (Figure 14), and the process of transferring money has become

‘somewhat easier’ or ‘much easier’ over the past five years (Figure 15).

Figure 14: Difficulty of sending and receiving money from people living outside of the community today (n=149)

51

Figure 15: Change in ability to send and receive money from people living outside of the community over the past 5 years (n=149)

This change in ability to transfer financial capital has occurred simultaneously with the adoption of

the mobile phone and subsequent use of Mobile Money mobile phone applications and services. The

following sections will explain the technical process of using Mobile Money, adoption, and a closer

examination of the use of these mobile based applications in context with alterative complementary

and substitutive strategies for transferring money and associated outcomes.

Transferring financial capital using mobile money, and adoption of mobile money platforms

The process of transferring money using mobile money platforms is relatively straight-forward and

simple. A ‘sender’ travels to a mobile money kiosk (Figure 12.), which are typically located along the

sides of roads in small towns and urban areas, and indicates to the kiosk attendant the phone

number of the intended ‘receiver’ and the amount to be transferred.

52

Figure 16: Mobile Money kiosk (in yellow next to an agrochemical dealer)

The sender then deposits that amount with the kiosk attendant who then registers the transfer and

sends an SMS text notification (including a confirmation code) to the receiver’s phone. Upon receipt

of this text message, the receiver can travel to any mobile money kiosk, present the phone and text

message code, and the kiosk attendant will allocate the receiver the amount of the transfer. Here it

should be noted that the closest Mobile Money kiosks that individuals in the study group could

access were in the nearby towns. Any mobile money transfer could only be accomplished via either a

personal trip to town or an individual could send their phone to town with a friend or family

member to complete a transfer on their behalf.

The adoption of mobile money services has not occurred on a universal personal basis in the study

group, but many individuals who have not registered for the service personally still make use of it by

transferring money using phones owned by friends and family. That said, awareness of Mobile

Money services is universal in the study group as the result of TV and radio advertisements, and

recommendations from friends and family members. Study participants stated that motivations for

53

adopting the service included ‘saving time and money otherwise spent traveling to transfer money

face to face’, ‘increased safety (not having to carry cash while traveling on roads where robberies

occur)’, and ‘logistics of agricultural product sale and payment from middlemen’. Individuals in the

study group were typically educated on the process of transferring money using Mobile Money

services by a friend or family member.

Mobile Money in context with alternative strategies for transferring money

Prior to the widespread adoption of the mobile phone and Mobile Money platforms over the past

five years, individuals in the study group transferred money either by traveling to a recipient and

transferring face-to-face, sending the money via a friend or family member that was traveling to an

area in close proximity to the intended recipient, or carrying out the process of a bank transfer or

courier service. These processes and strategies are still used today, but they require more time and

financial cost, or technical access and logistics of making use of a formal institution such as a bank.

For these reasons, and the motivations for adopting mobile money stated above, individuals in the

study group have largely transitioned from transferring money by ‘traveling and transferring money

face to face’ and ‘sending with a friend or relative who is traveling’, to placing great importance on

the use of Mobile Money platforms for transferring money (Figures 17 and 18).

54

Figure 17: Importance of strategies for transferring money to and from individuals outside of the community five years ago (n=149, error bars represent 95% CI)

Figure 18: Importance of strategies for transferring money today to and from individuals outside of the community at present (n=149, error bars represent 95% CI)

Impacts of Mobile Money use

55

Individuals in the study group that place greater emphasis on the use of Mobile Money platforms

when transferring money also tend to state that transferring money is ‘somewhat easy’ or ‘very

easy’ (Figure 19).

Figure 19: Difficulty of sending and receiving money from people living outside of the community in relation to importance of the strategy: Mobile money to transfer money (n=149)

Additionally, while the data show that as individuals place greater emphasis on using Mobile Money

platforms to transfer money they tend to have larger farm sizes (Figure 20), connections between

variables such as these should be viewed with reservation.

56

Figure 20: Farm size in relation to importance of the strategy: Mobile money to transfer money (n=149, error bars represent 95% CI)

The advantages of Mobile Money use in relation to higher level outcomes should be carefully

parsed. Interview data show that the benefits of Mobile Money transfer over alternative strategies

generally fall into two categories: (1) saving the time and cost of transferring money using less

efficient means, and (2) due to the efficiency of Mobile Money transfer, financial emergencies have a

greater chance of being effectively addressed. The savings in terms of time and cost to transfer

financial capital via Mobile Money platforms as opposed to alternative strategies would seem to be

marginal, but not entirely insignificant considering the average income level of the study group. That

said, it is possible that for individuals transferring money at higher frequencies, these marginal

savings on transfers could add up to a significant avoided cost over longer time scales. An example of

this described in the interviews involved a charcoal producer who regularly sold small amounts of

charcoal throughout the week as a strategy to obtain supplemental income. The charcoal buyers

would often pay for the product in small instalments via Mobile Money transfer. The comparative

efficiency and security of this system, as opposed to alternative payment and transfer strategies,

could result in significant savings over time. The second benefit of Mobile Money use: dealing with

57

financial emergencies, acts as a means by which to increase the resilience of individuals but it is

unclear if this increased resilience would contribute to a higher-level livelihood outcome such as

farm size. For example, interview data show that if an individual requires pressing medical attention

they may ask a family member to send a small remittance in order to assist in paying medical fees.

This process can play an important role in maintaining the health of individuals, but it remains

unclear if these are rare instances of transfers contribute to higher level outcomes such as farm size.

That said, several other advantages of Mobile Money use were brought to light in the interviews.

Participants stated that transferring money using Mobile Money platforms was easier and faster

than transferring money through a bank. Participants also stated that transferring money using

Mobile Money was a safer option than alternative methods. During the times of the year when high

quantities of yam are being sold there have been instances of robberies of cash along the roads as

farmers and middlemen move between the market and the communities. Transferring money using

Mobile Money platforms avoids the need to carry large quantities of cash in person for purchase or

sale of agricultural products. Interview participants perceive this as a method by which to decrease

the risk of robbery. Finally, several interviewees stated that transferring money using Mobile Money

platforms is a more secure means by which to transfer money. These individuals expressed distrust

in a system of sending money via courier or other strategies that do not involve face-to-face

transfer.

Despite the majority of the study group adopting and using Mobile Money platforms, several

barriers to entry and use of Mobile Money persist. Two interviewees stated that the only system of

money transfer that they trust is through a face-to-face interaction. This sentiment is shared by a

few other interviewees who had a general sense of distrust in both the technology of Mobile Money

and other new societal and cultural changes. This distrust is made worse by the spread of

disinformation about Mobile Money or the outlying examples of problems with the Mobile Money

system. Participants in the study group are generally warry of financial scams, some of which have

58

occurred through mobile phone use, and financial systems such as banks and savings opportunities

cheating them or failing. The other major barrier to entry into the Mobile Money system involves

gender and social norms. It is uncommon for women in the study group to have bank accounts in

their name despite the fact that their husband may have a bank account. In some instances these

men may deposit profits produced by their wives into these personal accounts. This form of male

dominated financial management has likely become manifest as a barrier to Mobile Money entry

for some women, and these women have to rely on their husbands for transferring money using

Mobile Money., Hence, the husband will transfer money using Mobile Money on his phone on his

wife’s behalf. That said, it is not uncommon for individuals in the study group to transfer funds using

Mobile Money on behalf of another friend or family member. This typically occurs during the time

between when an individual loses or breaks their phone and they are able to purchase a new phone.

Finally, with regard to the advantage of Mobile Money transfers as a means by which to garner

emergency funds, it should be noted that a portion of the study group has access to and has

benefitted from a process that does not involve the mobile phone. Several of the communities in the

study group have ‘women’s agriculture groups’ which often possess and lend micro-loans to women

in need. These groups are supported by donations and group farming efforts and provide a

significant and fast acting social and financial safety net that can provide financial capital faster and

more efficiently than mobile money in some cases.

4.2.2 Healthcare, Education, Social Events, and GovernmentHealthcare

In terms of healthcare infrastructure and presence of local clinics and hospitals, the study group has

relatively good access to healthcare. That said, for many people in the study group financial capital is

a limiting factor for healthcare access, so much so that interviews revealed examples of individuals

unable to purchase sufficient agriculture inputs due to healthcare expenditures. Knowledge of and

participation in the National Health Insurance Program is mixed and limited in the study group.

Families often cannot afford to insure all of their children in the scheme and must prioritize those in

59

the most need. Commonly cited maladies in the interviews included: fever (malaria), boils, joint pain,

and unknown sickness in children. Most participants possess almost no medical knowledge and

some individuals rely on traditional medicines and treatments. Participant’s struggle to differentiate

and identify different qualities of healthcare facilities and treatments. Healthcare emergencies are

not dealt with using a mobile phone. Individuals requiring immediate healthcare are sent to the

hospital in the nearest town via a motorized tricycle or motorcycle volunteered by any of the several

individuals owning such transport in each community. No individual interviewed stated that they

had called a medical professional using a mobile phone. Women in labour typically give birth in the

community with assistance from a community member who has received training as a midwife.

That said, checking in on and discussion of the state of friends and family health is one of the main

reasons that individuals state for owning a phone. Individuals in the study group regularly call their

family members living around Ghana to check their health and these calls become more frequent if a

friend or family member is suffering from a health problem. While most individuals will visit a clinic

or hospital at the first sign of a health condition, several individuals interviewed stated that they

would call a family member in advance to discuss whether they should go to a clinic or not, and one

interviewee stated that he would discuss possible healthcare costs if he were to seek treatment.

Parents also reported discussing whether they should take their sick child to seek treatment or not

with their partners. It is likely that this discussion involves the cost of treatment, and whether

enough funds are available, and if those funds should be dedicated to treatment. Individuals also

reported using their mobile phones to aid in the logistics of seeking healthcare treatment such at

informing family members of their travel and station at a healthcare facility.

Education (children)

Much like with the case of healthcare, the main function of mobile phone use in the study group in

relation to children’s education is to check on the general status of children attending school. This

call is typically made from one parent to another, or a family member where the child is staying, to

60

make sure that the children are out of the house and attending school. This call is more important

and common when a child, typically attending secondary education that is only available in towns, is

living away from the parents while attending school. Only one individual in the study group stated

that they use their mobile phone to more specifically investigate the state of their child’s education.

This individual would make calls to the school to enquire about school fees and supplies, as well as

the child’s performance and participation. Another individual in the study group stated that he has

been using his mobile phone and contacts to look for opportunities of, and preparation for, higher

education. That said, in general, parents in the study group have very little understanding of their

children’s education and differentiating between different education opportunities (schools). The

most commonly sought aspects of schools that parents identified were second language (English or

Arabic) learning opportunities.

Events: Funerals, Marriages, and other social events

The most important social events that occur in the communities in the study group are funerals and

marriages. In the past these events could only be organized and potential attendees invited by either

traveling and inviting individuals in person or sending written invitations through friends and family

traveling around Ghana. The majority of the study group has at least one close family member living

in the Northern Region of Ghana and interviewees stated that it is common for individuals in the

study group attend at least one wedding or funeral in the Northern Region or host a family member

from the Northern Region for a wedding or funeral in the community each year. The advent of the

mobile phone has made planning weddings and funerals for individuals in the study group easier

when compared to the traditional methods stated above. Individuals check for weddings and

funerals while making regular ‘greeting’ calls and individuals discuss potential attendance at such

events. An additional aspect of these kinds of social events is that they often involve the outsourcing

of music systems and computers brought in to provide entertainment during ceremonies. This

61

provides individuals who own phones with memory cards to import new music and recordings of

religious sermons on their phones for later enjoyment.

Government

Individuals in the study group have little or no ability to influence their local government outside of

voting in local elections. This is evidenced by the fact that no individual interviewed stated that local

government had responded to requests made by community leaders. Interviewees stated that they

make voting decisions based on development promises made by candidates, and attempt to

leverage incumbents seeking re-election by threatening to withhold votes if past development

promises are not fulfilled. No interviewee had ever contacted their local government on an

individual basis through any means including using a mobile phone. One interviewee stated that he

knew individuals in the community that had the contact numbers of individuals working in the local

government but it was unclear if anyone had made phone calls contacting local government officials.

This may be due to cultural norms whereby any concerns that community members have must be

communicated to the local government through community elders. One interviewee stated that this

communication might take the form of a written letter. Finally, it should be stated that the majority

of interviewees, when asked, struggled to describe how the local and national government of Ghana

is organized and functions.

4.3 Agriculture Production: Inputs, Knowledge Transfer, and Organisation of Labour4.3.1 Agriculture production challenges and processes

This section will cover the agriculture production challenges that farmers face and the methods that

some farmers are using to address those challenges. It is difficult for respondents to rank production

challenges due to the range of challenges in terms of acute challenges such as a new plant disease

and general challenges such as declining soil fertility. Furthermore, production challenges are often

linked, interdependent, or nested. For instance, respondents state that weeds are an increasing

challenge but also state that purchasing herbicides poses a significant challenge. These challenges

62

are made worse by farmers’ lack of technical agricultural training and limited access to high quality

sources of information used to address these challenges. That said, over the past few years the study

group has enjoyed increased access to local extension officers and the advent of the mobile phone

has, in some cases, further improved this connection and subsequent ability to address challenges

and acquire appropriate inputs.

Individuals in the study group report that common production challenges include: declining soil

fertility, competition with weeds, insect attack, and plant disease. There is a general sense in the

study group that soil fertility is declining but testing the accuracy of these claims using any kind of

soil testing analysis was beyond the scope of this study. That said, one farmer described how

farmers in the area have transitioned from a slash and burn land acquisition and preparation

technique to a current mode of increasing dependence on the use of synthetic fertilizers. That same

farmer noted that the presence of earthworms had been declining in the soil and that he believed

this was due to an over application of herbicides. This idea of external negative effects on production

was also expressed by two farmers in relation to the problem of weeds. Crop competition with

weeds poses persistent and significant challenge to the study group which is commonly addressed

using manual weeding techniques, but increasingly farmers are turning to synthetic herbicides in

order to accomplish the task of removing weeds. Two individuals in the study group believed that

the increased use of these herbicides has been counterproductive and contributed to an increase in

the presence of weeds. Responses such as these should serve as examples of the challenge that

farmers face in understanding how their farms are changing with changes in production practices

and the introduction of new synthetic inputs. Examples of information deficiencies were also

exposed via interview questions regarding insect attack and plant disease. Farmers in the study

group struggle to treat insect attacks which are mainly comprised of yam seed damage and

subsequent germination failure due to nematodes. The same can be said about plant disease,

farmers recognize plant disease via visual inspection of plants, but possess very little ability to

specifically identify diseases and undertake an appropriate course of action to remedy these

63

problems. Here it should be stated that, in certain cases, farmers do not take action to address

insect attacks and plant diseases because they are not empowered to do so. These circumstances

will be discussed in greater detail in the next chapter.

Knowledge of agriculture and the role Extension officers

This section will focus on knowledge of agriculture practices and knowledge of how to deal with

agriculture production challenges as a form of human capital critical to the maintenance of

livelihoods within the study group. No participants in this study have received any form of formal

agricultural education and every individual learned the basics of agriculture production from family

members beginning at a young age. Only over the past five to ten years have farmers had access to

sources of agriculture related information outside of their close family and colleagues. Today

farmers can access agricultural information through new sources such as extension officers, chemical

dealers, television, radio, and in some cases, the ease and speed of this information transfer has

been increased via the use of the mobile phone. That said, challenges in communication and transfer

of accurate information persist in the study group.

Information is researched, transmitted, and retained through are variety of systems and processes in

the study area and within and outside of the study group. This study revealed a dichotomy in these

information systems that can be defined as passive and active information transmission. Passive

information transmission can be described as an undirected diffusion of information. On the ground

this takes several forms such as farmers relaxing after a day’s work and having a causal conversation

about how they are carrying out their activities which may uncover differences in practices and

subsequent discussions regarding differential outputs. On the other hand, active information

transmission typically occurs in order to address a particular production challenge and involves a

specific targeted information query. An example of this would be a phone call put in to an extension

officer upon noticing that a plant is discoloured and may have a disease. That said, underlying the

layer of human capital that is agriculture knowledge is that of a motivation and ability to access

64

knowledge. This applies to both types of information transmission. If a farmer is anti-social and

therefore does not engage in casual conversations about farming practices they may miss out on

otherwise useful information, and if a farmer cannot conceive of the usefulness of placing a phone

call to a family member who may be knowledgeable about a certain production problem,

information flow will not occur.

Extension officers serve to aid farmers via both kinds of information transfer and in a general sense,

the knowledge they provide is of a higher quality and standard than any other information source

that is currently accessed by farmers in the study group. As a consequence of involvement in the

CAY-Seed project (CRI), communities in the study group have received regular visits by extension

officers offering demonstrations and outreach over the past several years.

Figure 21. Demonstration of production practices in the community

65

These regular visits have increased the accessibility and social connection between farmers and

extension officers, as well as among the farmers themselves. Over the past few years more farmers

in the community have obtained the mobile phone contacts of the extension officer and calls have

increasingly been exchanged in order to organise logistics surrounding meetings and

demonstrations, as well as for checking in on demonstration plots and project progress. The

increased strength of these social bonds and trust have led several individuals in the study group to

view extension officers as a resource for addressing agriculture production challenges outside of the

scope of the CAY-seed project. Six of the interviewees spoke about how they had called an extension

officer in order to address agriculture challenges such as dealing with an insect attack. These calls

typically involve a request for the extension officer to travel to the farmer’s farm in order to inspect

the problem first hand. Alternatively three farmers stated that if they encounter a problem they will

forego the phone call and travel to the extension office in search of information, or to schedule a

meeting on the farm to inspect the problem. In almost all cases, a trip to or from the community in

to town is involved, meaning that the phone only serves as a mode of accomplishing logistics and

not necessarily a direct line of transmission used to solve a specific problem in real time. That said,

the use of the mobile phone to contact the extension officer provides a connectivity advantage in

that a trip to the extension office may be unsuccessful in the event that the extension officer is not

present in the office upon the farmer’s inquest. In some cases this may be due to farmers’ inability

to correctly identify and assess the problems that they face and must therefore rely on first hand

examination carried out by the extension officer. One participant noted that in the past they have

transported a sample of diseased crops to the extension office so as to avoid a trip required by the

extension officer.

Here it should be noted that while extension officers may serve as highly accurate source of

agricultural information, at times farmers address challenges and access information through other

means. Furthermore at least seven interviewees stated that they had not requested assistance or

called an extension officer in a time of need. Reasons for this lack of connection ranged from being

66

unaware that extension officers could serve as a source of helpful information to a belief that the

farmer may know better than the extension officer. In a small number of cases (n=2) female

interviewees stated that they could not call extension officers due to moratoriums on calling

individuals outside of their family which were instituted by their husbands. This may also be due to a

cultural norm whereby if a female encounters a production challenge, they must bring the issue up

with their husband and he may take the initiative to seek out the appropriate solution. Additionally,

this cultural norm may play out during demonstrations held by extension officers (see vignette

below). A small number of interviewees also expressed a sense of personal pride in solving

production challenges on their own or by contacting close friends, and one farmer remarked that

this pride may be enhanced in the younger generation who may possess less reverence for

institutions such as the extension service.

Akuba, age 44, grows groundnut, cowpea, and maize to support her children while her husband is

working outside of the community. She has been attending demonstrations held by extension officers

in the past but she has not felt empowered to contact an extension officer outside of the

demonstrations. Furthermore during the demonstrations she has not spoken up when she has a

question due to the fact that men are present and hold position and power in that setting. She states

that if the demonstrations were divided by gender, women in the community would feel freer to

exchange ideas and improve their farming methods. Unfortunately she does not feel empowered to

make such a request for separate demonstrations.

Aside from demonstrations, extension officers fill two very specific roles in assisting farmers in

dealing with their production challenges. These are (1) recommendations of and assistance in

obtaining appropriate pesticides, herbicides, and fungicides, and (2) the provision of access to the

national fertilizer subsidy scheme. As stated above, if a farmer notices that a plant disease or insect

67

attack is affecting their crops they will often contact an extension officer to schedule a farm visit or

they will take an example of the affected plants to the extension officer for examination. At this time

the extension officer will indicate the appropriate chemical treatments that the farmer should

purchase. These purchases are made in the nearby town from agricultural input retailers.

Interviewees also stated that they learn about new chemical inputs from television and radio

advertisements and broadcasted agricultural education programs, or while discussing their practices

in person in the community or during phone calls with friends and family. That said, over the past

few years a trend has arisen whereby when farmers encounter a disease or infestation they by-pass

seeking assistance from an extension officer and instead call or go directly to input retailers for

advice and subsequent purchase of inputs. Interviewees struggled to identify why this change has

occurred but it may be due to the convenience of the input retailers or the social and economic

relationship that many farmers have with particular retailers. At least four of the interviewees

explained a situation where they feel committed to a particular retailer because that retailer will

offer the purchase of inputs on credit or other financing schemes. One of these interviewees felt

that if they were to purchase inputs from another dealer and the dealer that had offered the

purchase of inputs on credit in the past had found out, the option to purchase on credit in the future

may expire. Two of the farmers interviewed stated that they did not feel this pressure and in turn

make use of their mobile phone to call multiple input retailers and friends and family in town that

have knowledge of input prices, in order to carry out a kind of arbitrage in search of the lowest input

prices on offer. This kind of task requires that a farmer and their colleagues have the ability to

identify, recall, and communicate the specific chemical inputs that they require. Considering that a

large portion of the study group remains illiterate, many farmers face a significant challenge in

identifying and recalling these specific inputs, and therefore heavily rely on assistance when

purchasing provided by family member who is literate or extension officers.

68

Aba, 46, was working in her groundnut field last year when she noticed a significant insect attack on

her crop. She called her local extension officer and explained the problem. At the time the extension

officer indicated which insecticide she should purchase but she was unable to understand and recall

the specific insecticide. The extension officer then requested that she travel to the extension office so

that they could meet. During the meeting the extension officer offered to write down the specific

insecticide but instead found a used container of said insecticide. Aba then took this container as an

example to a local retailer where she was able to purchase the correct chemical and treat her crop.

Finally, extension officers also provide the direct link between farmers and access to The National

Fertilizer Subsidy Programme (MoFA). Access to fertilizer is one of the limiting factors for livelihood

development and also a significant production expense. Fertilizer prices and availability on the open

market can fluctuate significantly throughout the year, and many farmers can only access fertilizer if

they can establish an ability to purchase on credit with a retailer. Several of the most vulnerable

individuals in the study group stated that due to a lack of access to fertilizer they have had to reduce

their farm sizes which has further decreased their ability to deal with shocks and stresses over the

past few years. For reasons like this, the Ministry of Food and Agriculture began The National

Fertilizer Subsidy Programme in 2008. At a local level this programme has been rolled out and

maintained by local extension officers registering farmers and issuing subsidized fertilizer quotas to

farmers using a ‘passbook’ system. On the ground this process looks like the following: a farmer

69

must meet in person with an extension officer each year in order to register, the officer then

denotes in a passbook the amount of fertilizer that that farmer can purchase from the programme at

a subsidized (reduced) price, the farmer then travels to a programme office or retailer and presents

the passbook where in the retailer denotes the amount purchased and acquired. Opinions of this

programme are mixed in the study group. They range from individuals who are unaware that the

program exists, to individuals who have used the program in the past but not regularly due to

challenges of access, to individuals who make full use of the program every year and believe it has

greatly improved their ability to produce a livelihood. Farmers who are not making consistent use of

the programme state that this is mainly because they have a level of distrust in the system. They

state that they do not have good access to extension officers required to register, and that

subsidized fertilizer is either unavailable or available at inopportune times of the year.

Kwasi, 54, registered for the fertilizer subsidy programme last year and received his passbook from

the extension officer. He then travelled to the input retailer that he prefers in town in hopes of

purchasing his quota of ten bags of fertilizer. Upon arrival at the retailer, he was told that the retailer

could only offer him one bag of fertilizer at the subsidized price. Later, Kwasi checked back with this

retailer in hopes that the stock had been replenished. At that time the retailer indicated that no

fertilizer was available at the subsidized price. Dejected, Kwasi purchased the remaining fertilizer

that he required at the much more expensive open market price. Today he has a negative outlook on

the fertilizer subsidy programme and he has not inquired about how he might solve his fertilizer

access challenges in the future.

During the time of this research the Ghanaian government was in the middle of a roll out of an

updated version of the fertilizer subsidy programme. This new version would eliminate the use of

the passbook system and replace it with a mobile phone based system. Under this new system,

70

extension officers have been registering farmers using the farmer’s mobile phone numbers. The

proposed plan in the future is that farmers will be issued a code in the form of a text message that

they will then present to fertilizer retailers when subsidized fertilizer is available. At this time of

purchase the retailer will call the MoFA central office and verify the quota and purchase using the

farmer’s code. In theory this system should be more efficient and provide better opportunities for

MoFA to manage and record data and practices involved in fertilizer distribution and purchase.

While many of the interviewees had been registered with this new phone based subsidy system, no

farmer indicated that they had received a code via text to date. That said, many interviewees

expressed trepidation toward this new system. These worries included: a fear that retailers would

not be able to contact the central office in Accra, that their phone numbers were different from their

personal identities, that they have trouble operating their mobile phone and reading text messages,

and that they may be eliminated from the system if their phone were lost or broken. It should be

stated that according to the new phone based system protocols, in the event that a phone is lost or

stolen, the owner of that phone should contact an extension officer who would then re-register that

farmer according to a new number (replacement phone) or the original number applied to a new

SIM in a replacement phone. This of course does not account for the possibility that in the event of a

lost or broken phone, a farmer may not have the financial capital to replace said phone for an

extended period of time. Currently it is unclear how this new phone based subsidy system will be

accessible to farmers that do not or have not ever owned or used a mobile phone. At least one

agriculture extension officer suggested that some form of paper based system would continue to

function. I was unable to verify the validity of this statement.

4.3.2 Organising labour

Due to the physically challenging nature of yam production, farmers organise and hire labourers

from outside of the community to assist in land preparation (‘stumping’, ‘stamping’, ‘mounding’),

planting, and weeding. These labourers are typically high school aged and travel from the North of

71

Ghana to the study area in search of income during breaks in the school year that coincide with

production cycles. The majority of the study group states that hiring labour is expensive and

represents a significant input cost in agriculture production. Farmers often have to provide housing

and some provisions of food for labourers, and hiring costs have increased in unison with increasing

fuel and transport costs. In addition to the cost of labour, the logistics of organising labour also

represents a significant challenge to a large portion of the study group (Figure 22).

Figure 22: Difficulty of organising labour (n=159)

A breakdown of this challenge reveals several sub-challenges, the most significant of which are the

cost of labour and a situation where at the time that labour is required, labourers are working for

other farmers. This is followed by the challenge that labourers are not present in the community

when they are required (Figure 22).

72

Figure 23: Seriousness of challenges in organising labour (n=159, error bars represent 95% CI)

This is due to the fact that labourers are not only a scarce resource but they are required at precise

times during the production schedule for each farmer. This means that farmers not only need to

attract labourers to the community and surrounding areas in a general sense so that they have the

potential to become available for work, but also coordinate schedules and logistics so that the

labourers can be hired at specific appropriate times.

That said, nearly half (n=79) of the study group (n=159) state that organising labour has become

‘somewhat easier’ over the past five years (Figure 24).

73

Figure 24: Change in the challenge of organising labour over the past five years (n=159)

Farmers in the study group implement a range of strategies for organising labour which produce

differential outcomes when addressing this challenge. These strategies have changed over the past

five years and it is likely that these changes have contributed to farmers’ perceptions of the change

in difficulty of organising labour over that time period.

In the past farmers simply waited for labourers to arrive in the community at which point the

farmers would approach the labourers and negotiate work face to face. At this time it was also

common for farmers to organise future labour hire during trips to see family in the North of Ghana.

In other words, a farmer might travel to the North of Ghana for a wedding or funeral at which time

they would interact with labourers and indicate the dates which they anticipated that they would

require hired labour in the future. Both of these processes are imprecise and rely heavily on both the

farmer’s ability to predict their future work schedule and luck according to weather patterns which

play a significant role in production schedules. Figure 25 below shows the ubiquity of these two

strategies for organising labour five years ago.

74

Figure 25: Importance of strategies for organising labour five years ago (n=159, error bars represent 95% CI)

Over the past five years, and with the adoption of the mobile phone, the set of strategies that

farmers use to organise labour has changed significantly. While a large portion of the study group

still relies on the traditional practice of organising labour by waiting for labourers to arrive in the

community, the dominate strategy for organising labour is now to contact labourers on the phone

directly, inquire about their availability, and request work (Figure 26).

75

Figure 26: Importance of strategies for organising labour at present (n=159, error bars represent 95% CI)

This can be a request for immediate work, or a phone call to assist in scheduling and logistics for

work in the near future. The mobile phone offers a faster and more efficient means for contacting

labourers and organising hire than alternative methods. Farmers are incentivised to request and

retain labourer contact numbers, and maintain a social network with labourers throughout the year

but particularly before and during the peak times of the year when labourers are required.

Individuals in the study group also use their mobile phones to contact both friends and family, and

even business partners such as middlemen, to assist in organising labour on their behalf. This greatly

speeds up the ability for farmers to organise labour and avoid the cost of traveling to nearby towns

or to the North in search of labourers to work in the immediate or relatively distant future. The use

of mobile phones in the process even apply to the labourers themselves. For example, a farmer in

the community may call a relative in the North asking for the assistance of multiple labourers on a

particular date. That family member may then call a labourer, who might then call additional

colleague labours, all resulting in the accomplishment of a logistical task in a matter of minutes and

76

at virtually no cost that would have required multiple days travel and the significant cost of field

work days forgone, in the past.

Figure 27 shows that farmers who state that organising labour is ‘somewhat easy’ or ‘very easy’

today, also state that calling labourers on the mobile phone directly is a significantly more important

strategy for organising labour than the traditional strategy of waiting for labourers to arrive in the

community and negotiating hire face to face.

Figure 27: Strategies for organising labour in relation to the difficulty of organising labour (n=159 , error bars represent 95% CI)

That said, challenges in organising labour persist in the study group. Farmers are increasingly finding

themselves in competition with one another to offer the best amenities and services such as food

and other incentives for labourers to assist them. And due to the high one off cost of hiring

labourers, farmers that have access to credit groups, such as women’s credit groups, have a better

opportunity to acquire the funds in a timely fashion to hire labour at specific times of the year. And

farmers who have trouble creating and maintaining phone contact lists due to a lack of literacy and

technoliteracy may be falling behind as more technoliterate and socially connected farmers

differentially access labourers.

77

Finally, farmers that implement strategies for organising labour using their mobile phones do not

have statistically significantly greater incomes or farm sizes. This may be due to the same factors

that apply to other livelihood processes that involve the mobile phone. The mobile phone serves to

increase the speed and efficiency, and decrease the cost of processes, but these gains do not appear

to have a significant impact on larger livelihood and development outcomes.

4.4 Yam Sale: Pricing and Sale Dynamics4.4.1 Pre-sale context: storage, farmgate vs market

This section will cover the context in which farmers sell yam including decision making, selling

processes, and challenges. Following the harvest of yam, farmers are faced with the decision to sell

immediately or store their produce. If the latter, then they must decide upon the process of yam

storage. Yam market prices fluctuate over the course of the year but also over the course of a single

day in the market. Farmers who can afford to store yam will carry out a kind of temporal arbitrage

whereby they will attempt to hold off sale until prices become more favourable. This practice stands

at odds with the pressure to sell immediately and derive income that can be used to make payment

on the living costs that were covered in section 4.1.1. It should be noted that a portion of the yam

produced is used for home consumption. Approximately 7% of the study group (n=160) reported

that they never store yam and sell immediately following harvest. Approximately 33% respondents

stated that they can always store yam for long periods of time until prices improve and 61% of

respondents stated that they can sometimes store yam until prices improve. The latter of these

groups will at times be forced to sell stored yam due to immediate need for cash income regardless

of longer term yam sale price fluctuations. Respondents who stated that they can store yam for long

periods of time until prices improve had significantly larger farms and higher average monthly

incomes than farmers who stated that they can only sometimes store yam until prices increase

(Figures 29 and 30).

78

Figure 28: Yam market containing an example of a tractor/trailer used for yam transport

Figure 29. Farm size in relation to ability to carry out temporal arbitrage when selling yam (n=149, error bars represent 95% CI)

79

Figure 30: Average monthly income in relation to ability to carry out temporal arbitrage when selling yam (n=149, error bars represent 95% CI)

Farmers in the study group who store yam all do so using a technique of stacking harvested yams on

the ground and then covering them using yam vines, palm fronds and other various plant material.

This storage technique (a yam ‘clamp’) can lead to crop damage if the yams are in contact with

ground that is wet, but the potential trade-offs between the possible cost of crop damage versus the

benefit of selling at an opportune time were beyond the scope of this study.

Before embarking on a closer examination of yam selling dynamics with special attention paid to

yam market dynamics, the relatively recently adopted practice of selling yam at the farm gate

deserves recognition. Some 15% of the study group (n=160) stated that they sell the majority of their

yam crop at the farm gate, meaning directly from the farmer, rather than transport them to a

market. That said, all farmers in the study group sell at least some of their yam in the markets which

are located in the nearest town. Furthermore, farmers that live in Mem, Watro, and Masuo sell

mainly in the Atebubu market, and farmers that live in Nyinasie, Bisiw #1, and Kramokrom sell

mainly in the Ejura market. The farmers that live in these communities do not sell in any other

80

markets due to the prohibitive nature of additional transport costs, such as crop damage occurring

during transport, required to transport crops to other markets. Asanteboa and Arbour are located

approximately between Ejura and Atebubu, and farmers living in these communities do at times sell

in the Ejura or Atebubu markets. Transport cost to these markets and the average yam sale prices in

these markets are approximately the same, and therefore while farmers living in Asanteboa and

Arbour may undertake a kind of spatial arbitrage, the value in doing so is likely nominal. That said,

over the past few years some farmers have begun selling yam at the farm gate. The process for

accomplishing this task can be carried out in two ways. Typically, when a farmer is ready to sell at

the farm gate they call their preferred middle-woman using a mobile phone and request that the

middle-woman travel to the farm. Upon arrival in the community and in-person inspection of the

yam ready for sale, the farmer and middle-woman will negotiate in person. If a sale price is settled

on, the middle-woman will organize the transport of the yam to the nearest town. The cost of this

transport will have been factored into the purchase price. This practice of deducting the cost of

transport from the farm gate sale price confused one individual interviewed who stated that he

preferred to sell at the farm gate because it allowed him to forgo the cost of transporting yam to

market on his own.

81

Kobi, 47, prefers to sell his yam at the farm gate. When it is time to sell he calls his middle-woman on

his phone and states his preference to sell. He does not talk about potential sale prices on the phone

with his middle-woman, but he regularly discusses the current state of yam sale prices with his

brother who lives in the nearby town. When the middle-woman arrives in the community, he inspects

Kobi’s crop and they begin to negotiate prices. Currently Kobi has become disenchanted with the yam

market in town. He has grown tired of the complexities, costs, and pressures of the market. He knows

that if the prices that the middle-woman is offering are not to his liking, he can simply refuse sale

and his yam will remain in storage on his farm as opposed to the nearly overwhelming cost of

transporting yam back from the market in the event that sale cannot be achieved there. He sees this

position as a strength in his negotiating process, while also knowing that selling at the farm gate

does not allow him the advantage of knowing real time fair market prices like he might observe in a

live market setting in town. He weighs these trade-offs in his head when arriving at a price with the

middle-woman, and as such believes that he accrues no additional net profit when selling at the farm

gate as compared to selling in the market in town.

4.4.2 Yam sale challenges and sale dynamics

When posed with the question: ‘In general how difficult is it to sell yam today?’ the majority of the

farmers in the study group stated that yam was either ‘somewhat easy’ or ‘very easy’ to sell (Figure

31).

82

Figure 31: Difficulty of selling yam at present (n=149)

When asked if ‘yam has become easier or more difficult to sell over the past five years’,

approximately 38% of respondents stated that yam had become ‘somewhat more difficult’ to sell

and approximately 34% of respondents stated that yam had become ‘somewhat easier’ to sell

(Figure 32).

83

Figure 32: Change in difficulty of selling yam over the past five years (n=149)

Of the respondents that stated that yam was ‘very easy’ to sell, 38 individuals (approx. 24% of the

study group as a whole) stated that yam had become ‘somewhat more difficult’ to sell over the past

five years. While these responses may exactly reflect reality on the ground in a general sense, a

technical view of the Likert scale wherein farmers responding that yam is currently ‘very easy’ to sell

would indicate an extreme ease in selling and therefore if those same farmers indicate that yam had

become ‘somewhat more difficult’ to sell over the past five years, these seemingly contradictive

responses could indicate some degree of cognitive dissonance. In other words, technically speaking

how could yam currently be ‘very easy’ to sell, while also having become ‘somewhat more difficult’

to sell over the past five years? That said, a finding such as this can be viewed from several different

perspectives. Firstly, the finding that selling yam has become somewhat more difficult for the same

individuals who state that yam is very easy to sell should serve an indicator of a worrying trend

wherein farmers who state that yam is very easy to sell today may find it increasingly more difficult

in the future. Secondly, this finding expresses a strength and possible weakness of this research

design. The vast majority of farmers in the study group stating that yam is easy to sell does not

84

signify selling yam as a significant development challenge, while many of these farmers state that

selling yam is becoming more difficult. That said, the interviews revealed that in many cases

individuals would indicate that a livelihood process was not challenging in a general sense, but

following further deeper questioning, they could describe significant challenges within the larger

process. This finding is mirrored in the questionnaire data. Interviews revealed five major challenges

for selling yam and questionnaire respondents indicated that the three most serious challenges for

selling are ‘being manipulated by middle-women’, ‘obtaining reliable price information’, and

‘transporting yam to market’ (Figure 33)

Figure 33: Seriousness of challenges of selling yam (n=149, error bars represent 95% CI)

Thus, when combining these findings: current challenge of selling yam, change in challenge of selling

yam over the past five years, and the significance of challenges within the larger challenge of selling

yam, we find that while yam is generally easy to sell for most of the study group today, a slight

majority of the study group states that yam sale has become more challenging over time and serious

sub-challenges persist. For farmers that state that selling yam has become ‘somewhat more difficult’

over the past five years, ‘being manipulated by middle-women’ is a greater sub-challenge than any

85

other sub-challenge and ‘obtaining reliable price information’ is a greater challenge than

‘transporting yam to market’, ‘the length of time between arriving in the market and yam being

sold’, and ‘understanding the protocols of the market’.

With that in mind, these sub-challenges deserve a closer examination. The process of transporting

yam to market is typically characterized by a farmer contacting, often using a mobile phone, an

individual or team who has access to or owns either a tractor/trailer, motorized tricycle, or at times

a car. Once contact is made the farmer and vehicle owner/operator schedule a time for the yam to

be loaded, taken to market, and then unloaded. The decision between the use of a tractor and the

use of a motorized tricycle is typically based on the size of the load of yam to be transported, with

larger loads typically requiring a tractor and the employment of additional individuals who load and

unload the yam. That said, a motorized tricycle can transport the same amount of yam as a tractor

by making multiple trips between the farm and the market. One interviewee stated that the recent

advent of these motorized tricycles has brought with it increased speed and decreased cost of

transport as these motorized tricycles do not require the hiring of additional individuals who load

and unload yam. Interviewees stated that the main challenges around yam transport involve cost

and access. In the past, farmers could only access transport vehicle operators by traveling to them

on foot or on a bicycle. While this practice still occurs today in some cases, many farmers now use

their mobile phones to contact operators in order to schedule and work out the logistics of

transport. The speed and accessibility that the mobile phone provides in this scenario greatly

improves farmers’ ability to carry out this task. During peak harvesting times and market days, the

supply of operators will often not meet demand for transport, so farmers that can use their mobile

phones to contact sometimes multiple operators in search of availability have an advantage over

those farmers who are not effectively using mobile phones. One interviewee stated that tractor

operators may not always be dependable when it comes to the time they are scheduled to pick up

products and in some cases operators may shirk their commitment to one farmer in favour of

another farmer who is closer in their business network and/or willing to pay a premium on top of

86

established transport costs. Considering that farmers who use their mobile phones to regularly

contact and get updates from operators in the days leading up to and most importantly the day that

yam is to be transported, it is likely that these mobile phone using farmers have a greater ability to

maintain access to and management of this transport process.

The speed of transport organisation and transport itself plays into another of the challenges of

selling yam: ‘the length of time between deciding to sell and when the yam is actually sold’. This is

because yam prices can fluctuate significantly throughout a single market day with prices typically

dropping into the afternoon and evening. As such, farmers typically obtain price information while

located in the community which gives them a general sense as to what they should expect as a fair

sale price in the market, decide to sell, and then from that point in time they generally work to get to

the market with their product as quickly as possible. That said, market forces are out of the control

of a single farmer and therefore a price drop in the market between the time that the decision to sell

is made and when a farmer can arrive in the market with his crop is largely an unavoidable challenge

of selling yam.

Once a farmer arrives in the market with his or her crop, they are faced with the challenges of

‘understanding the protocols of selling in the market’, ‘the length of time between arriving in the

market and when the yam is finally sold’, and ‘being manipulated by middle-women’. While there is

some level of interplay between these challenges, unpacking them one at a time will show the

individual processes that they represent. Understanding the protocols of the market was uncovered

in the interview stage on this research as many interviewees struggled to describe the complex

market protocols in any detail and often with significant confusion. Only a few of the individuals

interviewed were able to not only accurately describe the protocols of the market, but conceptualize

the challenges and possible opportunities that they present.

87

Adom, 55, worries about his farming colleagues and their lack of understanding of how supposed

market protocols may be eating into their profit margins. He states that the payments to the District

Assembly to enter the market and to the ‘Market Queen’ to assume a location in the market are

required and accepted, but recently a new, sometimes obscured, and possibly problematic layer to

the market structure has emerged. Some individuals in the market have taken to calling themselves

‘agents’ and serve the purpose of connecting farmers to middle-women. These agents are effectively

lower level middlemen between farmers and higher level, buyer connected middlemen (middle-

women). Agents may even seduce farmers by offering them food which the farmer may unwittingly

be paying for because once their yam has been sold they must pay a commission to the agent. This

fee comes on top of the two fees described above and the commission for sale paid to the middle-

woman. Even more recently, Adom has seen these agents actually purchase yam themselves, often

at prices well below fair value, and then in turn sell that yam to middle-women. Adom is worried

about how farmers may be taken advantage of for having far less detailed knowledge of market

protocols and sale prices than these agents and middle-women.

The challenge of the ‘length of time spent in the market between arrival and the sale of yam’ arises

from several factors, one of which involves some possible manipulation by middle-women, but also

stems from a general uncomfortableness with being in the market/town, the management of

changing sale prices, and the fear of not selling at all. As stated above, most farmers spending most

of their time in the community express a level of discomfort at spending time in the nearest town.

88

They often feel as if the towns are less safe and more chaotic. These feelings may be intensified in

the markets which can be very busy and overwhelming at times, and farmers feel the constant need

to keep watch over their crop for fear of petty theft. Physical discomfort aside, farmers also worry

about how the prices they are being offered may be changing throughout the day. They must

attempt to find a balance between waiting for prices to possibly go up while trying to avoid the

possibility that prices may drop as the day continues. In the event that negotiations and hold outs

continue for an increasing amount of time, farmers begin to increasingly feel a pressure to sell for

fear of not making a sale on the market day. Almost all of the farmers in the study group do not have

financial capital to be able to transport unsold yam back to their farm. While it is sometimes possible

that farmers can store their yam in the market overnight, under the watch of more or less vigilant

guards and the increased threat of petty theft, and a small number of farmers in the study group

have houses or access to houses in town where they can store unsold yam, most farmers see the

amount of time spent in the market as an unwanted inconvenience at the least.

The way that middle-women manage the stress that farmers feel to sell throughout the day is one of

the means by which farmers feel that they may be being ‘manipulated by middle-women’. In many

instances middle-women hold greater power than farmers throughout the negotiation/sale process.

Middle-women have better access to accurate market price information and typically do not

compete directly amongst themselves for access to buyers or for connection to farmers; they often

make arrangements amongst themselves to avoid this. Middle-women also manipulate farmers by

giving preferential treatment or attention to farmers who are willing to provide them with food or

other perks during the market day. Some middle-women provide farmers with inputs or loans in

return for the farmer’s loyalty, or in some rare cases forced commitment, to sell through them at

unfair prices. That said, interviews revealed the complete range of commitment(s) to middle-

women, from complete commitment and trust in a particular middle-women, to farmers that

manage multiple middle-women in hopes to creating competition, to a high level of distrust in

middle-women and near constant turnover of which middle-woman a farmer may sell through.

89

Finally, middle-women may manipulate farmers by providing inaccurate market information such as

sales price information during the period of time when farmers are deciding whether to transport

their yam to market for sale or not.

This research placed a greater focus on understanding the challenge of obtaining yam sale price

information because not only is this challenge one of the greatest in terms of the larger goal of

selling yam, but also because farmers are increasingly using their mobile phones to accomplish this

task. Obtaining reliable price information in order to make more informed decisions about when to

sell, or temporal arbitrage, is also one of the few means by which farmers may significantly increase

their profits throughout the selling process. That said, farmers who state that obtaining reliable price

information is ‘very easy’ have smaller farmers than farmers who state that obtaining reliable price

information is ‘somewhat difficult’ or ‘neither easy nor difficult’ (Figure 34).

Figure 34: Average monthly income in relation to difficulty of obtaining reliable yam sale price information (n=149, error bars represent 95% CI)

This may suggest that while obtaining reliable price information poses both a challenge and

opportunity, currently the opportunity for increased profits remains either somewhat negligible or

90

offset by other challenges in terms over overall income, or the study group is currently not fully

capitalizing on the potential increased profits derived from using price information to carry out

advantageous temporal sales arbitrage.

Today, the most important strategy for obtaining reliable price information involves farmers calling

middle-women using their mobile phones (Figure 35).

Figure 35: Importance of strategies for obtaining yam sale price information at present (n=149, error bars represent 95% CI)

These calls are used to discuss the general trends in market prices, and not to negotiate a specific

sale price of any particular load of yam. This is due to the variable nature of yam products and

subsequent sale prices, therefore yam sale prices can only be settled upon following an in-person

examination by the middle-woman/buyer. As such farmers are calling to inquire about the general

nature of the yam market and either specific prices that yam is being sold at or possible future

market trends that the middle-women is expecting. Farmers also accomplish the task of the

obtaining yam sale price information by calling other individuals, such as friends and family, typically

living in town or visiting the market in order to get current pricing information. These kinds of

information searches are active means by which farmers collect information. Alternative strategies

91

for obtaining price information are more passive and involve either farmers waiting for middle-

women to call them by phone and indicate favourable market selling conditions or farmers inquiring

face to face about sale information from fellow colleague farmers, typically in an informal fashion

while relaxing in the evenings after working on the farm during the day. Prior to the advent of the

mobile phone, this information transmission system of conversing face to face with colleague

farmers in the community was the main way in which farmers gathered pricing information. Five

years ago the main alternative strategy, and one that could be characterized as more active, would

be for farmers to travel to the market on market days and inspect the nature of the sales conditions

in person (Figure 36).

Figure 36: Importance of strategies for obtaining yam sale price information five years ago (n=149, error bars represent 95% CI)

This practice of traveling to town, typically via bicycle or on foot, can provide farmers with an

accurate, in-person assessment of the market conditions, but requires both the time and physical

exertion to make the trip between the community and the market and then back. This extension of

time may contribute to the challenge described above where prices may change between the time

when the decision to sell is made and the time when the farmer actually arrives in the market with

92

his or her crop. Today, farmers generally prefer strategies that do not involve traveling to the

nearest market, actively searching for information using their phone and passively obtaining

information through informal conversations while sitting in the community. The mobile phone has

provided the advantages of decreased travel time and connectivity that the mobile phone provides

as expressed by increased used of the mobile phone as compared with prominent strategies used

five years ago.

The reliance on middle-women for price information and market conditions may appear

contradictory to the significant challenge of being manipulated by middle-women described above.

Interviewees struggled to conceptualize this seeming contradiction of relying on a possibly

manipulative middle-woman for the information to be used to make the critical decision of when to

sell or not sell. Furthermore, farmers struggled to describe how they had been cheated or

manipulated by middlemen if this had occurred in the past. That said, farmers who state that yam is

‘very easy to sell’ rely on calling their middle-women for pricing information significantly more than

alternative strategies (Figure 37).

Figure 37: Difficulty of selling yam in relation to strategies for obtaining yam sale price information (error bars are 95% CI) (n=149, error bars represent 95% CI)

93

Furthermore no statistical relationships exist between the importance of any particular strategy for

obtaining price information and a farmer’s average monthly income or farm size. This finding serves

to highlight the complexity of how the livelihood processes described above contribute to higher

level outcomes such as income and farm size.

In summary we find that selling yam has in a general sense become easier for some and somewhat

more difficult for others. This change occurred at the same time that farmers where changing their

strategies, in particular a move towards an increased use of their mobile phone and reliance on

phone calls to middle-women, for accomplishing a component of selling yam: obtaining reliable yam

sale price information. But despite these specific improvements, challenges such as being

manipulated by middle-women persist, and while in a localized setting the use of the mobile phone

provides for increased speed and efficiency of information transmission, decision making, and

potential profits, the larger contributions to higher level outcomes remain obscured.

4.5 Principal Component Analysis and Cluster Analysis4.5.1 Cluster descriptions

As described in the methods chapter, this research included a more advanced set of statistical

analyses in an attempt to gain a more nuanced view of the complex nature of farmers’ livelihoods.

This involved carrying out a PCA and then a CA on the data assessing the importance of particular

strategies involved in transferring money, organising labour, and obtaining yam sale price

information. While the results chapters above have presented the analysis of these variables

separately, this section will assess them together as part of a combination of strategies which

individuals employ to achieve livelihood outcomes.

The PCA produced four components explaining approximately 68.75% of the variance in the strategy

response data (Table 2.). While the first four components were selected according their eigenvalues

(λ>1), subsequent reliability tests revealed that component 4 was significantly unreliable as it

94

contained only two factor loadings greater than .3 (Table 3). A Varimax rotated component matrix

was used in analysis as it provides a modest simple structure aimed at increasing the number of

near-zero loadings defined by smaller clusters of intercorrelation (Kaiser, 1958). Factor coefficients

below .3 have been suppressed in the component matrix for easier interpretation and coding.

According to the factor loadings for the components, they have been coded as such: (1) Heavy

phone users, (2) Selective phone users, and (3) Non-phone users.

Principal Component Analysis: Total Variance Explained (Livelihood strategies)

Total Variance Explained

Component Initial Eigenvalues Extraction Sums of Squared Loadings Rotation Sums of Squared Loadings

Total % of Variance Cumulative % Total % of Variance Cumulative % Total

1 4.502 32.160 32.160 4.502 32.160 32.160 4.122

2 2.350 16.786 48.946 2.350 16.786 48.946 2.495

3 1.722 12.299 61.245 1.722 12.299 61.245 1.761

4 1.052 7.512 68.757 1.052 7.512 68.757 1.248

5 0.752 5.372 74.129

6 0.655 4.678 78.807

7 0.627 4.479 83.286

8 0.543 3.878 87.164

9 0.524 3.742 90.906

10 0.455 3.250 94.156

11 0.286 2.046 96.202

12 0.226 1.612 97.813

13 0.170 1.213 99.027

95

14 0.136 0.973 100.000

Extraction Method: Principal Component Analysis.

96

Table 2: Total variance explained for livelihood strategies

Rotated Component Matrixa

Questions Component1 2 3 4

14 Travel and transfer money face to face 0.78 Send and receive money through friends and relatives traveling to and from community 0.775 Mobile money transfer -0.373 0.788Bank transfer 0.601 0.479 Hired car/courier to transfer money 0.489 0.746

18 Walk up and ask labourers face to face in the community to hire -0.591

Call labourers on your mobile phone directly to hire 0.781 -0.314 Call a friend or relative on your mobile phone to organize labour for you 0.836 Travel to labourers outside of the community to ask them face to face before hiring 0.734

25 Travel to the market and look at prices/talk to people in the market 0.377 0.713Call middle-woman about price information 0.809 Wait for middle-woman to call and provide price information 0.804 Call people in the market and ask them about pricing information 0.66 0.447 Talk to colleague farmers in the community about pricing information 0.795 Extraction Method: Principal Component Analysis. Rotation Method: Varimax with Kaiser Normalization.a

a. Rotation converged in 5 iterations.

97

Table 3: Rotated Component Matrix (livelihood strategies)

98

Reliability tests were carried out on the variables with factor loadings greater than .3 within each

component. Here it should be stated that the strategy variable: ‘wait for labourers to call before

hiring’ was not included in the PCA or subsequent CA. This variable was highly correlated with

enough other variables so as to result in only one component having a Cronbach’s Alpha score above

.7, and all other components having Cronbach’s Alpha scores below .3. For these reasons, that

variable was omitted from the PCA and CA, serving as a limitation of this analysis. Regardless, only

one component (Component 1) had a Cronbach’s Alpha score of greater .7. Component 4 was

omitted because it had a Cronbach’s Alpha score below .3. These low Cronbach’s Alpha scores for

the components included in the CA show reasonable reliability for Components 1 and 2. The

Cronbach’s Alpha scores for the first three components are provided in tables 3, 4, and 5 below.

99

Questions Item-Total Statistics (Component 1: Heavy Phone users)

Cronbach's Alpha

Scale Variance if Item Deleted

Corrected Item-Total Correlation

Cronbach's Alpha if Item Deleted

14 Bank transfer

0.775

27.962 0.593 0.732Hired car/courier to transfer money 32.222 0.591 0.756

18 Walk up and ask labourers face to face in the community to hire

40.815 -0.421 0.868

Call labourers on your mobile phone directly to hire 28.23 0.519 0.744Call a friend or relative on your mobile phone to organize labour for you

26.313 0.719 0.71

25 Call middle-woman about price information 24.881 0.696 0.707Wait for middle-woman to call and provide price information

23.483 0.695 0.705

Call people in the market and ask them about pricing information

24.551 0.679 0.71

Table 4: Reliability analysis for Component 1: Phone Users

Questions Item-Total Statistics: (Component 2: Selective phone users)

Cronbach's Alpha

Scale Variance if Item Deleted

Corrected Item-Total Correlation

Cronbach's Alpha if Item Deleted

14 Bank transfer

0.635

10.587 0.602 0.497Hired car/courier to transfer money 12.819 0.787 0.537

18 Call labourers on your mobile phone directly to hire 14.291 0.053 0.707Travel to labourers outside of the community to ask them face to face before hiring

12.891 0.276 0.624

25 Call people in the market and ask them about pricing information

9.028 0.598 0.477

Talk to colleague farmers in the community about pricing information

12.611 0.230 0.647

Table 5: Reliability Analysis for Component 2: Selective phone users

100

Questions Item-Total Statistics (Component 3: Non-phone users)

Cronbach's Alpha

Scale Variance if Item Deleted

Corrected Item-Total Correlation

Cronbach's Alpha if Item Deleted

14 Travel and transfer money face to face

0.398

2.804 0.267 0.272Send and receive money through friends and relatives traveling to and from the community

2.414 0.345 0.162

Mobile money transfer 4.62 -0.102 0.49425 Travel to the market and look at prices/talk to

people in the market2.561 0.273 0.262

Table 6: Reliability Analysis for Component 3: Non-phone users

101

Following the identification and reliability testing of the components produced by the PCA, I carried

out a CA on the remaining components. Three clusters were produced with a silhouette measure of

separation and cohesion greater than .5. The first cluster contained 78 individuals or 52.7% of the

study population (n=148)8 (Table 7).

Cluster characteristics were defined according to the mean scores for each component. Higher

means are highlighted in blue and lower means are highlighted in orange in Table 8 below. These

defining characteristics facilitated the labelling of the clusters: (1) Non-phone based strategists, (2)

Phone based strategists, (3) Face to face strategists with limited phone use.

8 12 cases were excluded due to non-responses.

102

Two-Step Cluster (Cluster Distribution) N % of Combined % of TotalCluster

1 - Heavy Phone users 78 52.70% 48.80%2 - Selective users 41 27.70% 25.60%3 - Non-phone users 29 19.60% 18.10%Combined 14

8100.00% 92.50%

Excluded Cases 12 7.50%Total 16

0100.00%

Table 7: Cluster Distribution

Cluster Profiles (Centroids) Heavy phone users Selective phone users Non-phone usersCluster Mean Std. Deviation Mean Std. Deviation Mean Std. Deviation

1 (n=78) -0.0979062 0.36093818 -0.7847410 0.52881324 0.2193948 1.154536372 (n=41) 1.1930914 0.35758296 0.7925385 0.58417278 -0.1722890 0.527938643 (n=29) -1.4234504 0.70670622 0.9901973 0.58542595 -0.3465154 0.94036768Combined (n=148) 0.0000000 1.00000000 0.0000000 1.00000000 0.0000000 1.00000000

Table 8: Cluster profiles

103

104

Characteristics of individuals that shape each cluster are as follows:

(Cluster 1) Non-phone based strategists: Individuals that transfer money using traditional means

based on face to face exchange or reliance on friends and family who are traveling, these individuals

also hire labour face to face by either meeting labours in the community or traveling to labourers to

hire them. On rare occasion these individuals may use their phone to hire labour or gain yam pricing

information.

(Cluster 2) Phone based strategists: Individuals who focus largely on the use of their mobile phone

to carry out their livelihood strategies. These individuals use their mobile phone and social networks

to organize labour and gain yam price information by calling labourers, middle-women, and friends

and family directly in order to accomplish their livelihood processes.

(Cluster 3) Face to face strategists with limited phone use: These individuals organize labour in-

person and prefer to obtain yam sale price information through the passive means of talking with

other farmers in the community. That said, they may also call friends and relatives in order to assist

in organising labour and gaining pricing information.

4.5.2 Cluster demographics and relation to outcomes

Clusters do not significantly differ between any of the basic demographic variables: age, gender, or

education. That said, clusters do differ in terms of basic mobile phone use variables. ‘Phone based

strategists’ have been using a mobile phone for more years than ‘non-phone strategists’ and ‘face to

face strategists with limited phone use’ (Figure 38).

105

Figure 38. Clusters in relation to years using a mobile phone (n=149, error bars represent 95% CI)

Furthermore, ‘phone based strategists’ use their phone more frequently than ‘non-phone

strategists’ and ‘face to face strategists with limited phone use’. Interestingly ‘non-phone strategists’

use their phones more frequently than ‘face to face strategists with limited phone use’ (Figure 39).

Figure 39. Clusters in relation to frequency of mobile phone use (n=149, error bars represent 95% CI)

106

The reason for this difference is difficult to parse, particularly because of all of the ‘purposes’ of

using a mobile phone, ‘non-phone strategists’ only rank ‘calling middle-women to organize sale of

crops’ significantly higher than any purpose ranked by ‘face to face strategists with limited phone

use’ (Figure 40). These findings express the difference between the value of the phone in terms of

the purpose that it serves and the frequency of use. In other words, a phone can be seen as having

great value for a specific purpose, but this may not translate into a higher frequency of use.

Figure 40: Clusters in relation to purposes of using mobile phones (n=149, error bars represent 95% CI)

This analysis shows that while ‘face to face strategists with limited phone use’ largely value their

mobile phone for its various utilities, they tend to either be more selective about making phone calls

or face an unknown barrier to frequent calling. Figure 3 also shows that ‘phone based strategists’

see great value in their devices for achieving processes not directly related to a specific goal of

transferring money, organising labour, or obtaining price information. Furthermore, despite the

differences in the importance of the mobile phone between the clusters, ‘non-phone strategists’

rank the purposes: ‘greeting friends and family’, ‘calling friends and family about financial issues’,

107

‘calling friends and family about health issues’, and ‘calling middle-women to organize sale of crops’

as greater than ‘neither important, nor unimportant’. This points out the difficulty in coding clusters

in that ‘non-phone strategists’ value their phones for processes not necessarily directly related to

agriculture production and income generation.

With regard to higher level livelihood outcomes, ‘non-phone strategists’ have larger farms and

higher monthly income than ‘phone based strategists’ (Figures 41 and 42).

Figure 41: Clusters in relation to farm size (acres) (n=149, error bars represent 95% CI)

108

Figure 42: Clusters in relation to monthly income (n=149, error bars represent 95% CI)

This provides yet more evidence that mobile phones can be useful in accomplishing specific tasks

such as transferring money, organising labour, or obtaining pricing information, but even when

individuals are viewed as complex actors employing combinations of strategies to achieve livelihood

processes, combinations of strategies largely involving the use of the mobile phone are less effective

at achieving higher level outcomes. Possible explanations for this disparity between outcomes

between ‘phone based strategists’ and other cluster groupings may be found in the differential

challenges that these groups face in achieving the goals of organising labour and selling yam. ‘Phone

based strategists’ rank the challenge ‘labourers are not in the community when I need them’

significantly higher than the other clusters (Figure 43).

109

Figure 43. Clusters in relation to challenges in organising labour (n=149, error bars represent 95% CI)

This finding cuts against the finding in section 4.3.2 that strategies involving the use of the mobile

phone in comparison to non-phone based strategies, decreases the general challenge of organising

labour. From a practical sense, this may indicate calling labourers directly helps decrease the

challenge of organising labour in a general sense, by may be ineffective at luring labourers to or

retaining them in, the community so as to be available for hire.

Phone based strategists also rank the challenge: ‘being manipulated by middle-women’ higher than

other clusters with regard to challenges in selling yam (Figure 44).

110

Figure 44: Clusters in relation to challenges in selling yam (n=149, error bars represent 95% CI)

This finding is particularly difficult to parse because ‘phone based strategists’ ‘call people in the

market and ask them about pricing information’. This specific strategy is loosely associated with a

decreased challenge of being manipulated by middle-women (Figure 45).

Figure 45. Challenge of being manipulated by middle-women in relation to the importance of strategies for obtaining pricing information (n=149, error bars represent 95% CI)

111

In other words, this CA suggests that calling people in the market to obtain yam sale pricing

information, among other phone based strategies for obtaining price information may not play an

important role in addressing the challenge of being manipulated my middle-women. This in turn

contributes a lack of comparative ability to achieve higher level outcomes such as greater income

and farm size.

112

Chapter 5 - Discussion‘What role does the mobile phone play in the development of agriculture based livelihoods of farmers in the Ashanti and Brong-Ahafo Regions, Ghana?’

‘How do farmers use mobile phones to address the livelihood challenges that they face?’

‘Does the use of mobile phones in addressing livelihood challenges produce positive livelihood outcomes?’

‘What is the context of and barriers to the mobile phone being more or less useful in addressing livelihood challenges?’

5.1 ICT4D Theory building: ICT capitalsAs stated in the methods chapter, this research sought to use grounded theory via a content analysis

of the qualitative data and a statistical analysis of the quantitative data to derive emergent theories

and layers of meaning. Not unsurprisingly, this practice arrived at several similar theories as other

research, while modifying and expanding on these theories as well as creating entirely new ones.

This discussion section will also link the findings above to these theories in a more detailed and

expanded fashion compared to past research. Also it should be stated that development and

explanation of these theories will employ aspects of the SLA. This is in part due to the influence of

the SLA in creating questions for both the interviews and questionnaires, but may also serve to

highlight some advantages and disadvantages of applying SLA language and elements to theory

building in ICT4D. Also, in terms of answering the research questions, the question of ‘how’

individuals are using mobile phones will be presented according to a theory of processes and

strategies, the ‘result’ of using mobile phones will be framed according to a discussion of outcomes,

and the question of what makes mobile phones more or less useful will begin with a theory of the

capital(s) required for use and later will invoke a description of barriers to effective use.

113

For the purposes of this discussion, the difference between ‘capital(s) required for mobile phone

related strategies to be carried out’ and ‘barriers to outcomes produced using mobile phones’

deserves attention. The reason for this is driven by the problem in ICT4D research wherein ‘barriers

to outcomes’ are sometimes defined by ‘a lack of capital(s)’. These are typically ‘internal’ barriers

such as a lack of technoliteracy, but they can also be ‘external’ such as a lack of phone coverage.

While this is true in a sense, and can be useful in framing results and analysis in a large portion of

past ICT4D research, I suggest a theory of all of the ‘capital required for processes and outcomes’ to

be viewed as separate from ‘barriers to outcomes that involve a lack of capital’. In other words, a

‘lack of capital required’ serves as a non-starter in livelihood development and data analysis, which is

different from a ‘barrier to outcome(s)’ wherein an individual can possess the capital required to

carry out a livelihood process and the process is carried out, but due to an external barrier, the

process does not produce an outcome. This is different from ‘a lack of capital required’ in that the

process never begins, dooming the outcome from the start.

This ‘theory of capitals required’ for processes is designed so as to capture, in a general sense, the

‘capitals required for ICT use’ in that ICT use plays at least some role in a larger process. Capitals that

are not directly related to ICT use are included in the theory but only as additional capital that, in

theory, might be used to carry out the process regardless of ICT involvement. Furthermore, this

theory places great importance on the analytical capital required for effective ICT use. Many of the

concepts involved in this ‘analytical capital’, which would largely be defined as human capital, are

valid without the technology component of the capitals, but are included as they are required for

effective technology use. These ‘analytical capital(s)’ are largely based on abilities to interpret

challenges and potential opportunities. This is somewhat similar to Sen’s (1985) capabilities

approach wherein individuals with increased information and understanding have increased decision

making ability. That said, my focus on understanding the role of the mobile phones requires a scope

wherein the mobile phone and subsequent capitals required for mobile phone use are combined

with what I call ‘information capital(s)’, as well as additional non-ICT related capitals so that

114

processes can be carried out. Here it should be stated that the ‘information capitals’ required for

processes vary slightly depending on if the subsequent processes are ‘active’ or ‘passive’. This

differentiation will be covered in greater detail in the next section. A diagram of a framework for the

‘phone’, ‘information’, and ‘non-ICT related capitals’ required for processes is provided in Figure 46.

115

Figure 46: Phone, Information, and non-ICT capitals framework

116

This framework not only shows the capitals required for processes involving ICT to occur, but it also

expresses a hierarchy of sub-capitals. For example, network coverage is required before phone

credit, because without coverage credit is useless. The presence of, and focus on, ‘information

capital’ highlights the idea that phones are not intrinsically valuable, apart from their re-sale value

and financial capital therein, but must be paired with informational capital in order to be useful. For

example, a phone is not useful for collecting information on agriculture production if it is not

interpreted by its owner as having the potential and functionality to accomplish such as task. This

theoretical framework expresses the complexity and challenge of defining its aspects in terms of the

SLA capital categories. Phone access is considered to be physical capital when the phone is owned by

the user, and should be viewed as more closely related to social capital when phones are accessed

only through a friend or relative who owns a mobile phone. Furthermore, within the category of

information capital, the sub-capitals are largely described as abilities, and in this case largely

abstracted abilities involving conceptualization and motivation. These variables are very difficult to

capture in the data collection process. Assessing these abilities was based on individuals’ abilities to

describe challenges and the roles that mobile phones play in accomplishing those challenges.

Interviewees were sometimes then presented with scenarios where the mobile phone could be

more useful9 and a small number of individuals responded ‘that they had not considered that

strategy or possible use of the mobile phone’. That said, it should be noted that respondents did not

necessarily frame the use of their phone, to collect price information for example, for ‘information

searching’ per se, but that the mobile phone was useful in ‘collecting price information.’ This

phenomenon in itself expresses an important aspect of information capital: that individuals see their

phone as having the ability to accomplish a specific task, and not necessarily as an adaptable and

multifunctional tool. In other words, the mobile phone was not seen as providing access to all

available information, but instead as a precise instrument for collecting specific information. This

highlights one of the most momentous challenges in theory building in ICT4D research: the concept

9 These scenarios where the phone was more useful were derived from data in previous interviews

117

of ‘quality or degree of use’. A phone user can possess degrees of the sub-capitals provided in the

framework. Furthermore, when the sub-capitals under Information capital are activated in a ‘sub-

process’, they can be activated according to degrees as well. This will be covered in a section below.

That said, these information capitals should be viewed and assessed on a continuum. Only a few

individuals interviewed fully conceptualized how they could use their mobile phone to connect with

AEAs and in turn solve their production problems. At the other end of that spectrum, many

individuals interviewed had never considered the idea of calling an AEA to solve a production

problem. These individuals did not possess the information capital to make use of their mobile

phone in addressing this livelihood challenge.

The issue of internal and external information sources is adapted from Duncombe (2006), in that

internal sources involve informal indigenous knowledge from experience and external sources are

formalized and mediated through formal structures. In this research, I suggest altering these

definitions to some extent. Internal information sources should be defined according to an

individual’s social network (including business contacts such as middle-women) wherein information

is acquired both actively and/or passively. External sources of information are those that are

formally provided through structures and institutions, but unlike Duncombe (2006), this information

can be accessed by the illiterate.

The issue of AEAs transmitting information can appear to fall between these two definitions. I

suggest the differentiation occurs according to the nature of the communication. Formal phone calls

aimed at, and facilitating, the scheduling of extension programs is an example of external

information, while individual farmers calling extension officers in search of potential solutions for

production challenges can in practice be more informal and should be regarded as an internal source

of information. These differentiators also involve the nature of the relationship between actors and

institutions. Farmers often have a friendly and sociable relationship with AEAs. Finally, all of these

aspects of capitals, leads into the next section covering processes and strategies, but first it should

be stated that the capitals included in the framework above largely apply to ‘analytical processes’.

118

This concept comes from Duncombe (2006) as well in that ICTs serve both a functional and analytical

role: functional in that they facilitate action or a strategy in themselves, and analytical in that they

facilitate the collection of evidence. The use and impact of mobile money received a great amount of

attention in the results chapter above. The capitals required to use mobile money are generally

similar to those provided in the framework above if one substitutes the concept of information or

information searching for ‘transferring money using mobile money’. In other words, individuals need

to conceptualize the process of mobile money, and understand their potential interactions with the

system before they can use it. This shows that the use of the capital framework can be transferable

in terms of identifying and capturing the capitals for the highly different ICT based processes such as

calling a middle-woman about market prices, and transferring money using mobile money.

5.2 ICT4D Theory building: Processes and outcomes

Following the identification and conceptualization of required capitals, this section will present a

theory of the processes and strategies10 that individuals use in conjunction with mobile phones to

produce outcomes. The mobile phone aspect is key in this section because the theory presented

describes processes of information collection and communication that can be carried out in the

absence of the mobile phone. As such the application of the theory to multiple strategies aimed at

achieving a larger process allows for comparison. That said, strategies should not necessarily be

viewed individually as farmers often employ a variety of strategies in order to carry out a higher level

process. This theory has been built with the intention that it is flexible enough to account for a

singular or more nuanced view of strategies and how they contribute to processes. The framework

of the theory is provided in Figure 47.

10 The definitions of processes and strategies can overlap, but for the purpose of this section I will refer to the way the variables are organized and interact in the results section. For example, the use of mobile money is a strategy for carrying out the higher level process of transferring money. Money is transferred in many ways, strategies should be viewed both individually and in combination with other strategies to contribute to processes.

119

Figure 47: Framework for ICT4D processes and strategies

At the heart of many of the strategies that individuals in the study group employed was a

differentiation between active and passive information communication and collection. Examples of

each side of the spectrum include: ‘waiting in the community for labourers to arrive before hiring

them’ (passive), and ‘calling friends and relatives to organise labour on your behalf’ (active). That

said, individuals in the study group also collected information via passive methods such as the

spontaneous communication of information that could be used to improve agricultural production

during a phone call that was placed for simple greeting purposes. One could argue that this kind of

‘passive communication’ is not a strategy at all, but this kind of communication plays an important

role in producing, or not producing, outcomes. ‘Waiting for middle-women to call and provide

120

pricing information’ is a passive strategy of information collection that was part of a combination of

strategies that produced comparatively poor outcomes as shown in section 4.5. Furthermore,

respondents in the study group, define this seeming non-action: waiting for labourers or middle-

women, as a strategy.

The timeframe of strategies is adapted from Chapman and Slaymaker (2002) who differentiate the

timeframes based on the role that information plays in strategies. For example, with regard to this

research, farmers accessing yam market sale prices by calling middle-women would be considered a

short term process according to the lifespan or duration of usefulness of that information. A farmer

calling an AEA regarding information on a new agriculture production technology such as planting

yam in rows instead of mounds could have a lifelong impact on the farmer.

As stated above, Duncombe (2006) defines the use of ICTs in terms of their analytical and functional

roles. This differentiation can mire the analysis if not clearly defined. For example, the strategy of

collecting yam sale price information using a mobile phone can be seen as analytical in that it is

simply the collection of data, or it can be seen as functional in that the collection of this data is an

action in itself. That said, I would argue for a clarification on Duncombe’s (2006) delineation in that

analytical strategies are those that provide information that is then used in a separate or subsequent

process. In other words, calling a potential labourer in order to schedule a date when labour can

occur is analytical in that only informational data is produced and transferred. Using a mobile phone

to make a mobile money transfer involves the use of the phone to carry out a direct action that is in

itself a process: transferring money.

The assessment of whether a strategy involving a mobile phone is substitutive/complementary/or

exclusionary plays an important role in understanding how the advent of mobile phones and their

use has interacted with alternative strategies (Hodge, 2005). While this study did not find any

evidence that mobile phones are exclusionary: that the use of mobile phones prevented users from

employing alternative strategies, this study did find evidence of substitution and complementation.

The overwhelming adoption of mobile money based platforms for transferring money represents a

121

substitution of alternative strategies. The use of mobile phones for obtaining yam sale price

information serves as a complementary strategy with alternatives. Farmers still make use of

strategies that do not involve the mobile phone in combination with strategies that involve the

mobile phone to accomplish the larger process of selling yam. That said, when it comes to obtaining

yam market price information, farmers make use of the mobile phone through multiple phone based

strategies. These can be viewed as complementary, and they are in a sense, but when viewing the

mobile phone as a critical variable between combinations of strategies, section 4.5 suggests that

combinations of strategies only involving the mobile phone do not produce comparatively positive

outcomes. This provides more evidence for viewing the phone as a variable when assessing the

substitutive or complimentary value of the mobile phone within strategies.

And finally, the delineation that the introduction of mobile phones has produced progressive or

transformed new strategies is based on an adaption of a combination of both Avergou’s (2008)

description of ICT based outcomes, covered below in greater detail, and Heeks and Molla’s (2009)

definition of ICT4D ‘outputs’ within their ICT4D value chain. Avergou (2008) argues that

transformation has occurred when underlying mechanisms have changed and Heeks and Molla

(2009) state that changes in communication patterns are an output which should be seen as an

impact. For the purposes of this discussion, I will define a progressive strategy according to an

adaptation of Heeks and Molla’s (2009) definition of an output in that new communication patterns

may be progressive in terms of change, while not altering the underlying structure of the strategy or

process. These definitions are somewhat difficult to apply to the processes included in this research

because they rely heavily on somewhat subjective sub-definitions of ‘change’ and ‘structures’. While

it may be help to link these definitions directly to outcomes, I suggest that they do not need to be

connected as such. For example, a strategy can be defined as transformative in that it is highly novel

according to a number of characteristics: new contacts, greatly increased speed of information

transfer, or new information, while not producing transformative outcomes due to external barriers.

For this reason, I would argue that the use of mobile money is a transformed strategy in that it relies

122

on a novel structure. On the other hand, while it could be argued that ‘calling a middle-woman’

using a mobile phone relies on a novel structure, the mobile phone in this case, the degree of

difference between calling a middle-women on a mobile phone and traveling to the market to obtain

yam sale prices is less than that of the degree of difference between transferring money using

mobile money and traveling to transfer mobile money.

Finally, the theoretical basis for identifying outcomes comes from the descriptors provided by

Avergou (2008): progressive or transformative outcomes. The results sections of this research have

defined higher level outcomes mainly as changes in farm size and monthly income, with lower level

outcomes defined by the relationship between strategies and the extent of challenges associated

with processes. What transformation would look like in terms of any of these variables is difficult to

define in that this research was largely unable to show what that would look like. It should also be

stated that transformation in terms of higher level outcomes could occur as a function of many

progressive strategies. Section 4.5 provides the best evidence that this is not occurring in the study

population, indicating that transformation of outcomes could be occurring as a function of variables

not measured in this research or that external barriers are preventing individuals from using

transformed and/or progressive strategies to achieve transformative outcomes. As stated at the

beginning of this chapter, barriers within this research are largely seen as external structures that

prevent successful strategies from being converted into higher level outcomes. The best example(s)

of a structural barrier is the combination of yam sale only occurring after visual inspection and in a

market structure where current real sale prices are largely obscured. This means that farmers can

maximize their price gathering strategies prior to a decision to sell and still face changing market

conditions and manipulation by middle-women. That said, the next section will apply the strategies

and processes theory above to the main findings provided in the results chapters, and in doing so

provide greater insight into the issues surrounding the lack of transformative outcomes uncovered in

this research.

5.3 Application of theory and higher level findings123

The role that mobile phones play in transferring money in the study population provides the best

example included in this research of a transformed strategy that produces positive outputs and is

associated with a positive outcome. Over the past five years a significant portion of the study

population has been able to conceptualize and adopt the use of mobile money platforms. This has

brought with it a new level of technoliteracy for individuals that use the platform directly, and an

increased reliance on a social network for those who rely on friends and family to transfer money

using the platform on their behalf. The use of mobile money is an action centred and functional

strategy that can occur both over the short and long terms. Individuals in the study group use mobile

money over the short term for sending and receiving small payments required for critical uses such

as payment for healthcare or agricultural inputs such as fertilizer. While these transfers have been

carried out for these purposes, the speed and decreased cost involved in addressing these

sometimes highly timely challenges have made mobile money a substitute for alternative strategies

as well as the closest example of a transformed strategy involving the use of a mobile phone found

in this research. The use of mobile money is associated with a greater ease in the challenge of

transferring money and with the higher level outcome: farm size. The low cost and increased safety

of transfers, as well as no significant drawbacks, indicate that the use of mobile money is a

sustainable strategy with sustainable outcomes including a decrease in individual vulnerability.

The role of the mobile phone in organising labour takes two forms: ‘calling labour directly’ and

‘calling a friend or relative to organise on your behalf’. These are both active strategies that are

aimed at achieving the short term goal of attracting and hiring labour for a particular day or set of

days. In this case the phone plays an analytical role in that farmers are using them largely for

scheduling purposes though there is limited evidence suggesting that these calls may also have a

functional role in creating and maintaining social networks. Strong social networks are particularly

important when farmers call friends and relatives for assistance with organising labour on their

behalf. Section 4.3 suggests that the use of phone based strategies in organising labour serve as a

compliment to alternative strategies in that while phone based strategies were unpopular five years

124

ago and are now the most popular strategies, the alternative passive strategy: ‘walk up to labourers

in the community’ remains popular in the study group. The use of mobile phones in organising

labour should be seen as a progressive strategy that produces marginally progressive outputs.

Furthermore, the use of mobiles in this way is not associated with higher level outcomes. These

strategies are not transformed mainly as a function of the persistence of alternative strategies and

continued barriers to outcomes. Individuals who place significantly greater value on the strategy of

‘calling labourers directly’ find that organising labour is easier, but barriers such as ‘labourers are not

in the community’ and ‘labourers are working for other farmers’ when needed persist in the study

group. This appears to suggest a larger structural problem that was not uncovered in this research. A

disconnect persists between the specific activity of organising labour using the mobile phone and

the event in time and place when a labourer is meant to begin work. While the use of the mobile

phone may marginally reduce the vulnerability of farmers, the question of sustainability remains.

This question is further complicated by an increasing demand and decreasing supply of labourers as

farms expand and the younger generations that make up the hired labour force look for alternative

options of employment. Climate change may also play a role in the sustainability of this system in

that changing climates will be associated with changing growing seasons and greater variability of

when labourers and needed and when they are available.

The process of obtaining yam market sale pricing information and eventual sale of yam is achieved,

at least partially, via three strategies involving the use of the mobile phone: ‘calling middle-women

directly’ (active), ‘waiting for middle-women to call’ (passive), and ‘calling people in the market to

inquire about prices’ (active). All of these strategies are short term in nature and highly analytical.

The use of these strategies can be viewed as either complementary in that many farmers continue to

include the passive strategy: ‘talk to colleague farmers in the community’ as part of a repertoire of

strategies including the use of the mobile phone, or as substitutive in that the active strategy: ‘travel

to the market and look at prices’ was a prominent strategy five years ago and significantly less

important when compared to alternative strategies. This suggests that strategies involving the use of

125

the phone are neither transformed as strategies, nor transformative in terms of outputs and

outcomes within the study group as a whole. Mobile phone use is only vaguely related to a decrease

in yam selling difficulty and not related to higher level outcomes. That said, an example from the

interviews on the use of mobile phones in obtaining price information in relation to sale and

outcomes highlights the concept of degrees of capital input and degrees of strategies mentioned

above. One of the individuals interviewed expressed a significantly greater conceptualization and

understanding of how yam prices change both over long term and short term scales than any other

individual interviewed. This individual had a strong social network with actors in urban and local

markets, calling these contacts regularly and often in order to check pricing information. These

circumstances are indicative of both a comparatively increased degree of information capital and

increased degree of carrying out a strategy for accessing price information. And yet, this individual

stated that the structural barriers that persist traveling to and within the market mitigate any

advantage gained using his increased degree of inputs and processes. These structural barriers

included the cost of transport to market and manipulation by middle-women. All of this is to say that

obtaining yam market sale pricing information using the mobile phone only marginally increases the

resiliency of farmers in the study group. That said, yam prices continue to increase as threats of

climate change and land degradation also play a role in the sustainability of selling yam.

126

Chapter 6 - Conclusion

6.1 Limitations of this research and reflection on methods

The methods used in the research present several limitations that are worth noting. The first of

these limitations involves the nature and scope of this research in relation to its applicability to

ICT4D as a larger field. This research assessed communities of yam farmers using a snap-shot

technique of data collection. While this research did include data collection addressing changes in

strategies and outputs over time by asking farmers about the state(s) of these variables five years in

the past, this is a weaker method for measuring change over time than a long term longitudinal

study. These responses relied not only on individual’s perceptions but also on their memory which

should serve as a limitation in assessing their validity. That said, large portions of the data collected

in this study were based on individual’s perceptions of reality: the importance of strategies and the

extent of challenges. While great value was placed on these perceptions according to the

epistemology used in this research, these strategies and challenges could have been measured using

objective indicators according to a positivist view of reality. This may have increased the

reproducibility of the results presented in this study.

Another possible limitation to this study involves the greater challenge in ICT4D of identifying the

direction(s) of causality between strategies and outcomes. This study purposefully divided questions

of surrounding strategies and outcomes in an attempt to avoid potential respondent bias derived

from an over estimation of the effectiveness of the technology. Questions regarding strategies that

did not involve the use of the mobile phone also sought to decrease the potential for this bias while

also facilitating a comparative and complimentary analysis of these strategies. This meant that

causation between quantitative variables had to be derived from an analysis of the qualitative data.

This method may be limited in its complexity and nuance. Furthermore, in both the qualitative and

quantitative methods, the resources spent and amount of data collected regarding strategies,

127

processes, and outcomes not necessarily related to ICTs could be viewed as a limitation in that the

overall focus of this research was on the role of ICTs. Finally, while this research did include a gender

dimension to both its methods and analysis, the finding that women sometimes have decreased

access to mobile phones as a function of power structures and societal norms showed that this

variable may have played a greater role in determining outcomes. Despite findings showing that

gender did not play a significant role in determining many of the strategies, processes, and

outcomes presented in this research, the lack of greater focus on gender issues should serve as a

limitation for this research. The final limitation of this research involved the selection of the

processes and subsequent strategies and outcomes that received attention in this research. While

interviewees stated that the processes: transferring money, organising labour, and obtaining yam

sale price information, were critical to their livelihoods and posed at least some challenge to be

overcome, this study did not place greater emphasis on other processes that may play a significant

role in determining higher level outcomes. Examples of these processes that individuals provided,

and were covered at least to some extent in the results chapter, included: access to fertilizer and

access to healthcare. Fertilizer is a limiting factor for farm size and healthcare is a limiting factor for

‘days spent in the field’ which can play a significant role in productivity. These processes were not

covered in greater detail due to resources available for this research and because the interviews

revealed that the mobile phone did not play a significant role within strategies or in determining

outcomes. An extreme example of a process that may be playing a significant role in determining

higher level outcomes in the study group that was not covered significantly in this research, but

should be viewed as a limitation, was that of weather. During the first phase of data collection

farmers’ production cycles were being heavily restricted due to a lack of seasonal rains. The effect of

this phenomenon was beyond the scope of this study. Farmers have no ability to influence the

weather, and the advantages of weather information accessed via the use of the mobile phones are

useful only to the extent that farmers can manage daily or weekly weather trends, as opposed to

larger climatic forces. While these limitations were considered in the design of the methods, and the

128

analysis and interpretation of the results, this research retains value according to its contribution to

the larger field of ICT4D described in the next sections.

6.2 Summary of findings

This research shows that currently the mobile phone plays a limited role in the development of the

agriculture based livelihoods of farmers in the Ashanti and Brong-Ahafo Regions, Ghana. Farmers

have adopted the use of the mobile phone both as a substitute or a compliment to strategies that do

not involve the use of the mobile phone within critical processes aimed at producing outcomes. This

research showed that the use of mobile money platforms serves to transform the process of

transferring money, which is linked to a higher level outcome: larger farm size. Furthermore, this

research showed that mobile phones are more useful when users possess higher degrees of

‘information capital’ that when combined with the capital required to use the phone and additional

non-ICT related capitals, can be used to implement strategies. In the case of organising labour and

obtaining yam pricing information, even with a maximisation of these capitals and strategies,

external barriers prevent a translation of capital via strategies and processes into higher level

outcomes.

6.3 Contribution to research

This research contributed to the field of ICT4D research in terms of both theories, methods, and

findings. This approach effectively applied a grounded theory approach so that higher level theories

could be derived from the data. These theories described and arranged the different capital(s)

required for mobile phone use as part of strategies and processes, and how to view the use of

mobile phones as part of strategies, and comparing strategies that do not involve the use of mobile

phones, that contribute to processes. While these theories are not extensive enough to serve as a

unifying theory in the field of ICT4D, how to link ‘ICT’ to ‘D’, they build upon past theories. The

application of data collected in this research to the theories developed as a result of this analysis was

more detailed and extensive than what can be found in a large portion of past research.

129

This research put into practice many of the calls for improved methodologies and methods. Unlike a

large portion of the ICT4D field of research, this study clearly defined and applied an epistemological

and theoretical approach. This was followed by implementing a mixed methods approach and a level

of triangulation of results aimed at avoiding past problems associated with assessing causal

relationships between variables and/or over reliance on anecdotal findings. Furthermore, this

employed a more holistic approach to assessing ICT4D in that it focused on the use and impacts of

using ICTs for development, while also assessing strategies that did not involve the use of ICTs so as

to afford both comparison of these strategies and a perspective that strategies using ICTs can be

used and assessed as part of combinations or sets of strategies. This provides a large step forward in

viewing the use of ICTs as a potential part of a portfolio of strategies that complex actors employ.

This nuanced assessment of complexity was furthered by a significant focus on the context in which

individuals in the study group persist. Finally, this research relied on a view of development defined

by the individuals that it assessed as opposed to relying on a top-down traditional definition and

agenda of development. As such, this method placed value on individuals’ perceptions of challenges

in addition to value placed on traditional indicators of development.

With regard to findings, this research contributed to the growing body of research surrounding the

use of mobile money platforms, but this research is one of only a few studies that have assessed the

organisation of labour in a context of small holder African agriculture, and possibly the only study

that includes an assessment of the use of mobile phones as a strategy for organising labour. This is a

particularly valid contribution to research in that the topic of organising labour arose organically in

the exploratory interview phase of this research. With regard to obtaining yam sale pricing

information and the process of selling yam, this research contributes to the larger body of ICT4D

research in that it exposes individual’s conceptualization and understanding of market structures

and how mobile phones may or may not be used to overcome barriers and challenges associated

with these external structures. In doing so this research avoided a techno-centric, tech-

130

deterministic, or unhelpful focus on hypothetical potential uses and outcomes that the use of

phones could produce.

Many of the findings in this research are applicable to agriculture systems outside of yam under the

correct circumstances. As stated above, the strength of the mobile phone was derived mainly from

the advantages that it provided farmers in terms of production and sale logistics. Therefore, these

finding can and should be applied to other crops, but particularly those crops and products where

the sale price is negotiated and determined on an in-person basis. I suggest, based on observations

during data collection, that mobile phones are used for the sale of livestock in a manner similar to

the sale of yam. Furthermore, the findings in this research related to the organisation of labour

should extend, at least to some extent, to the organisation of mechanized labour such as the use of

tractors for both yam and other crops. Finally, the barriers and challenges associated with both

illiteracy and a lack of technoliteracy shown in this research should be seen as representative of

problems across the developing world. Phone designs and user interfaces are largely universal at

present and generally inaccessible to illiterate individuals and those lacking technical phone use

skills.

6.4 Recommendations for future development and research

The aim of this research was to assess the ‘how’ (use), ‘what’ (impact), and ‘why’ (factors of

usefulness) and not to assess an ICT4D initiative or the potential uses of ICT. As such, this research is

limited in terms of the recommendations that it can provide with regard to future initiatives and

policy surrounding ICT4D. That said, this research can provide insight into how Ghana, and possibly

other developing countries, can move forward with the use of ICTs in extension and education

services. This research shows that many individuals in the study group do not possess the capital

required to access and use the new SMS based fertilizer subsidy program that is described in Section

4.3.1. The use of SMS text messaging in this new scheme has a significant potential to exclude

131

illiterate farmers from accessing greatly needed subsidized fertilizer. If this programme continues to

move towards widespread implementation, action should be taken to accommodate farmers that

are unable to make use of mobile platforms. That said, this research shows that the use of the

mobile phone can be effective in linking farmers with AEAs so as to address agriculture production

challenges. AEA’s should seek to capitalise on the use of the mobile phone by promoting its potential

connectivity and transfer of information by promoting its use as part of extension programmes and

workshops.

This research also indicates several areas of future research regarding ICT4D both in the study area

that this research covers and moving forward in ICT4D in general. This study showed that only a very

small portion of the study group has any understanding of the internet and even less have access to

the internet. Only one person in the study group had used the internet, and that was in a highly

limited capacity. That said, 4G network coverage is already available in all of Ghana’s regional capital

and larger cities, and smart phones are available for purchase in the study area. It is highly likely that

individuals in the study group will leapfrog internet access via fixed landline in favour of mobile

based internet platforms. The initial forays into internet use covered in this research suggest that

users will make use of messaging applications such as WhatsApp. A framework for assessing the role

of the internet in individuals’ livelihoods could adapt frameworks used in the past to assess

individual’s readiness, availability, uptake, and impact. Other recommendations for future research

are drawn from several subtle findings uncovered in this research that did not receive greater

attention as they fell somewhat outside the scope of this study. Potentially the most critical of these

findings involved evidence that the use of mobile phones and other communication technologies

such as the internet may have the potential to play a role in what may be a societal shift in the study

area. Future research should investigate how ICTs are interacting with changing community and

social structures wherein younger generations may be more likely to move back and forth between

communities and nearby towns where they can organise and maintain social networks in both

132

locations. This may have implications for the resilience of communities as well as changes in power

structures and social dynamics.

Both the findings presented in this research and the recommendations for future research highlight

the crossing cutting, nuanced, and ever changing landscape of ICT4D. This study serves as yet

another stepping stone in what should be a never ending trail towards understanding how

individuals can reduce their vulnerability and increase the sustainability of their livelihoods through

the process of development. Mobile phones are currently playing a role in this process but

continued and improved use as well as a change of, or ability to overcome, external structural

barriers, will be required before greater development outcomes can be achieved.

133

Reference List

Agarwal, B. (2015). Food Security, Productivity and Gender Inequality. In R. J. Herring (Ed.), The Oxford Handbook of Food, Politics, and Society (pp. 273-301). New York, NY: Oxford University Press.

Agwu, A. E., Ekwueme, J. N., & Anyanwu, A. C. (2008). Adoption of improved agricultural

technologies disseminated via radio farmer programme by farmers in Enugu State, Nigeria. African Journal of Biotechnology, 7(9), 1277-1286. doi: 10.5897/AJB

Aker, J. C. (2010). Information from markets near and far: Mobile phones and agricultural markets in Niger. American Economic Journal: Applied Economics, 2(3), 46-59. doi: 10.1257/app.2.3.46

Aker, J. C., & Mbiti, I. M. (2010). Mobile phones and economic development in Africa. Journal of Economic Perspectives, 24(3), 207-32. doi: 10.1257/jep.24.3.207

Aker, J. C., & Wilson, K. (2013). Can mobile money be used to promote savings? Evidence from Northern Ghana. (SWIFT Institute Working Paper No. 2012-003). doi: 10.2139/ssrn.2217554

Alao, A., Lwoga, T.E., & Chigona, W. (2017) Telecentres Use in Rural Communities and Women Empowerment: Case of Western Cape. In J. Choudrie, M. S. Islam, F. Wahid, J. M. Bass, & J. E. Priyatma (Eds.), Information and Communication Technologies for Development: Proceedings of the 14th IFIP WG 9.4 International Conference on Social Implications of Computers in Developing Countries (pp. 119-134). Cham, Switzerland: Springer.

Albu, M., & Scott, A. (2001). Understanding livelihoods that involve micro-enterprise: markets and technological capabilities in the SL framework. Warwickshire, UK: Intermediate Technology Development Group (ITDG).

Anderberg, M. R. (2014). Cluster analysis for applications: probability and mathematical statistics: a series of monographs and textbooks (Vol. 19). New York, NY: Academic Press.

Avgerou, C. (2008). Discourses on innovation and development in information systems in developing countries’ research. Journal of Information Technology, 23(3), 133-146. doi: 10.1093/oxfordhb/9780199580583.003.0029

Avgerou, C. (2010). Discourses on ICT and development. Information Technologies & International Development, 6(3), 1-18.

134

Avgerou, C. (2017). Theoretical framing of ICT4D research. In J. Choudrie, M. S. Islam, F. Wahid, J. M. Bass, & J. E. Priyatma (Eds.), Information and Communication Technologies for Development: Proceedings of the 14th IFIP WG 9.4 International Conference on Social Implications of Computers in Developing Countries (pp. 10-23). Cham, Switzerland: Springer.

Balasuriya, A., & de Silva, H. (2011). Connecting to work: non-agricultural livelihood opportunities for rural wage labour in Sri Lanka. In D. J. Grimshaw & S. Kala (Eds.), Strengthening rural livelihoods: The impact of information and communication technologies in Asia . (pp. 71-87) Ottawa, Canada: Practical Action.

Ban, K. M. (2014). The road to dignity by 2030: ending poverty, transforming all lives and protecting the planet. Synthesis report of the Secretary-General on the post-2015 sustainable development agenda. New York, NY: United Nations.

Batchelor, S., Scott, N., Valverde, A., Manfre, C., & Edwards, D. (2014). Is there a role for Mobiles to support Sustainable Agriculture in Africa?. In M. Höjer, P. Lago, & Wangle, J. (Eds.), ICT4S 2014: Proceedings from the 2nd International Conference on ICT for Sustainability (pp. 272-280). Paris, France: Atlantis Press.

Baumüller, H. (2012). Facilitating agricultural technology adoption among the poor: The role of service delivery through mobile phones (ZEF Working Paper Series 93). http://dx.doi.org/10.2139/ssrn.2237987

Baumüller, H. (2013). Enhancing smallholder market participation through mobile phone-enabled services: The case of M-farm in Kenya. In: Horbach, M. (Ed.), INFORMATIK 2013 – Informatik angepasst an Mensch, Organisation und Umwelt (pp. 1730-1738). Bonn, Germany: Gesellschaft für Informatik e.V.

Black, T. R. (1993). Evaluating social science research: An introduction. London, UK: Sage.

Boateng, M. S. (2012). The role of information and communication technologies in Ghana’s rural development. Library Philosophy and Practice, 871, 1-22.

Boyera, S. (2007, May). The mobile web to bridge the digital divide. Paper presented at IST-Africa Conference, Maputo. doi: 10.1145/1242421.1242433

Brezis, E. S., Krugman, P. R., & Tsiddon, D. (1993). Leapfrogging in international competition: A theory of cycles in national technological leadership. The American economic review, 83(5), 1211-1219.

Brundtland, G. H. (1987). United Nations world commission on environment and development: our common future. Oxford, UK: Oxford University Press.

Carney, D. (1999). Livelihoods approaches compared: A brief comparison of the livelihoods approaches of the UK Department for International Development (DFID), CARE, Oxfam and the United Nations Development Programme (UNDP). London, UK: The Department for International Development.

135

Chapman, R., & Slaymaker, T. (2002). ICTs and Rural Development: Review of the Literature, Current (ICTs and Rural Development: Review of the Literature, Current (Working Paper 192). London, UK: Overseas Development Institute.

Chapman, R., Slaymaker, T., & Young, J. (2003). Livelihoods approaches to information and communication in support of rural poverty elimination and food security . London, UK: Overseas Development Institute.

Chambers, R., & Conway, G. (1992). Sustainable rural livelihoods: practical concepts for the 21st century (IDS Discussion Paper 296). Brighton, UK: Institute of Development Studies.

Charmaz, K. (2014). Constructing grounded theory (2nd ed.). London, UK: Sage.

Chavula, H. K. (2014). The role of ICT in agricultural production in Africa. Journal of Development and Agricultural Economics, 6(7), 279–289. doi: 10.5897/JDAE2013.0517

Cieslik, K., Leeuwis, C., Dewulf, A. R. P. J., Feindt, P. H., Lie, R., Werners, S. E., ...Struik, P. C. (2017, July). Addressing socio-ecological development challenges in the digital age: environmental virtual observatories for connective action. Paper presented at 2nd International EVOCA Workshop, Accra, Ghana.

Clark, J. (2012). The Effectiveness and sustainability of the input supplier model . (ALINe Working Paper 9). London, UK: Institute of Development Studies

Codjoe, S. N. A. (2006). Migrant versus indigenous farmers. An analysis of factors affecting agricultural land use in the transitional agro-ecological zone of Ghana, 1984–2000. Geografisk Tidsskrift-Danish Journal of Geography, 106(1), 103-113.

Compaine, B. M. (Ed.). (2001). The digital divide: Facing a crisis or creating a myth?. Cambridge, MA: MIT Press.

Davison, R., Harris, R., Vogel, D., & Jan de Vreede, G. (1999). Information Technology in Developing Countries: Closing the Digital Divide. Journal of Global Information Technology Management, 2(3), 1-4. https://doi.org/10.1080/1097198X.1999.10856255

de Laine, M. (2000). Fieldwork, participation and practice: Ethics and dilemmas in qualitative research. London, UK: Sage.

Dey, I. (2003). Qualitative data analysis: A user friendly guide for social scientists . London, UK: Routledge.

De Silva, H., & Zainudeen, A. (2007, December). Teleuse on a Shoestring: Poverty reduction through telecom access at the ‘Bottom of the Pyramid’. Paper presented at the Centre for Poverty Analysis Annual Symposium on Poverty Research, Colombo. doi: 10.2139/ssrn.1555597

DFID. (1999). Sustainable Livelihood Guidance Sheets. London, UK: The Department for International Development. Retrieved from www.dfid.gov.uk/

Dissanayeke, U., & Wanigasundera, W. A. D. P. (2014). Mobile Based Information Communication Interactions among Major Agriculture Stakeholders: Sri Lankan Experience. The Electronic

136

Journal of Information Systems in Developing Countries, 60(1), 1-12. doi: 10.1002/j.1681-4835.2014.tb00422.x

Dodson, L. L., Sterling, S., & Bennett, J. K. (2013). Considering failure: eight years of ITID research. Information Technologies & International Development, 9(2). 19-34. doi: 10.1145/2160673.2160681

Donner, J. (2007). The use of mobile phones by microentrepreneurs in Kigali, Rwanda: Changes to social and business networks. Information Technologies & International Development, 3(2), 3-19. doi: 10.1162/itid.2007.3.2.3

Donner, J., & Escobari, M. X. (2010). A review of evidence on mobile use by micro and small enterprises in developing countries. Journal of International Development, 22(5), 641-658. https://doi.org/10.1002/jid.1717

Duncan-Howell, J. A., & Lee, K. T. (2008, September). Policy catch up: developing nations and developing ICT policy documents. Paper presented at the British Educational Research Association (BERA) Annual Conference, Edinburgh.

Duncombe, R. (2006). Using the livelihoods framework to analyze ICT applications for poverty reduction through microenterprise. Information Technologies & International Development, 3(3), 81-100. doi: 10.1162/itid.2007.3.3.81

Duncombe, R. (2011). Researching impact of mobile phones for development: concepts, methods and lessons for practice. Information Technology for Development, 17(4), 268-288. doi: 10.1080/02681102.2011.561279

Duncombe, R. (2016). Mobile phones for agricultural and rural development: A literature review and suggestions for future research. The European Journal of Development Research, 28(2), 213-235. doi: 10.1057/ejdr.2014.60

Eastwood, R., Lipton, M., & Newell, A. (2010). Farm size. In R. Evenson, & P. Pingali (Eds.). Handbook of agricultural economics (pp. 3323-3397). Burlington, VT: Academic Press.

EPA. (2011). GHANA STRATEGIC INVESTMENT FRAMEWORK (GSIF) FOR SUSTAINABLE LAND MANAGEMENT (SLM) (2011-2025). Accra, Ghana: EPA.

Erumi-Esin, R., & Heeks, R. (2015, May). e-business adoption and use among African women-owned SMEs: an analytical study in Nigeria. In ICTD 2015: Proceedings of the Seventh International Conference on Information and Communication Technologies and Development (pp. 11-21). Singapore: ACM.

Fafchamps, M., & Hill, R. V. (2005). Selling at the farmgate or traveling to market. American journal of agricultural economics, 87(3), 717-734.

Fafchamps, M., & Minten, B. (2012). Impact of SMS-based agricultural information on Indian farmers. The World Bank Economic Review, 26(3), 383-414.

Falola, A., & Adewumi, M. O. (2012). Constraints to use of mobile telephony for agricultural production in Ondo State, Nigeria. Journal of research in forestry, wildlife and environment, 4(2), 52-63.

137

Farrington, J. (2001). Sustainable livelihoods, rights and the new architecture of aid. London, UK: Overseas Development Institute.

Flick, U. (2014). An introduction to qualitative research (5th ed.) London, UK: Sage.

Fu, X., & Akter, S. (2011). The impact of ICT on agricultural extension services delivery: evidence from the rural e-services project in India (TMD Working Paper Series No. 046). Oxford, UK: Department of International Development.

Furuholt, B., & Matotay, E. (2011). The developmental contribution from mobile phones across the agricultural value chain in rural Africa. The Electronic Journal of Information Systems in Developing Countries, 48(1), 1-16. doi: 10.1002/j.1681-4835.2011.tb00343.x

Gates, B. (2016, January 20). Why the future is bright for the world’s poorest farmers. [Web Blog Post] Retrieved from https://www.gatesnotes.com/Development/The-Future-of-Farming

Glaser, B. G. (1998). Doing grounded theory: Issues and discussions. Mill Valley, CA: Sociology Press.

Gomez, R., Baron, L. F., & Fiore-Silfvast, B. (2012, March). The changing field of ICTD: content analysis of research published in selected journals and conferences, 2000--2010. ICTD ’12: Proceedings of the fifth international conference on information and communication technologies and development (pp. 65-74). Atlanta, GA: ACM.

Gomez, R., & Day, S. A. (2013, May). Research Questions, Paradigms and Methods in ICT for Development: Content Analysis of Selected ICTD Literature, 2000-2010. Paper presented at the 12th International Conference on Social Implications of Computers in Developing Countries, Montego Bay, Jamaica.

Goodman, L. A. (1961). Snowball sampling. The Annals of Mathematical Statistics, 32(1), 148-170.

Grunfeld, H. (2007, December). Framework for evaluating contributions of ICT to capabilities, empowerment and sustainability in disadvantaged communities. Paper presented at the In CPRsouth2 Conference on Empowering rural communities through ICT policy and research, Chennai, India.

Grunfeld, H. (2011). The contribution of information and communication technologies for development (ICT4D) projects to capabilities, empowerment and sustainability: a case study of iREACH in Cambodia (Doctoral dissertation, Victoria University). Retrieved from VU Library http://vuir.vu.edu.au/19359/

GSM Association. (2011). Women and Mobile: A Global Opportunity. London, UK: GSMA. Retrieved from https://www.gsma.com/mobilefordevelopment/resources/gsma-women-mobile-a-global opportunity/

Guislain, P., & Qiang, C. Z. W. (2006). Foreign direct investment in telecommunications in developing countries. In 2006 Information and Communications for Development: Global Trends and Policies (pp. 15-40). Washington, DC: The World Bank

138

Heeks, R. (2002a). i-development not e-development: special issue on ICTs and development. Journal of International Development, 14(1), 1-11. https://doi.org/10.1002/jid.861

Heeks, R. (2002b). Information systems and developing countries: Failure, success, and local improvisations. The information society, 18(2), 101-112.

Heeks, R. (2006). Theorizing ICT4D research. Information Technologies & International Development, 3(3), 1-4.

Heeks, R., & Molla, A. (2009). Impact assessment of ICT-for-development projects: A compendium of approaches (Development Informatics Working Paper 36). doi: 10.13140/RG.2.2.13011.53282

Heeks, R., & Wall, P. J. (2018). Critical realism and ICT4D research. The Electronic Journal of Information Systems in Developing Countries, 84(6), 1-11. doi: 10.1002/isd2.12051

Heffernan, C., Lin, Y., & Thomson, K. I. M. (2016). Drawing from development: Towards unifying theory and practice of ICT4D. Journal of International Development, 28(6), 902-918. doi: 10.1002/jid.2882

Hildebrandt, N., Nyarko, Y., Romagnoli, G., & Soldani, E. (2015). Price Information, Inter-Village Networks, and Bargaining Spillovers: Experimental Evidence from Ghana. Unpublished manuscript, New York University, New York, NY.

Hodge, J. (2005). Tariff structures and access substitution of mobile cellular for fixed line in South Africa. Telecommunications Policy, 29(7), 493-505. doi: 10.1016/j.telpol.2005.05.001

Kaiser, H. F. (1958). The varimax criterion for analytic rotation in factor analysis. Psychometrika, 23(3), 187-200. https://doi.org/10.1007/BF02289233

Krone, M., Dannenberg, P., & Nduru, G. (2016). The use of modern information and communication technologies in smallholder agriculture: Examples from Kenya and Tanzania. Information Development, 32(5), 1503-1512. doi: 10.1177/0266666915611195

Lio, M., & Liu, M. C. (2006). ICT and agricultural productivity: evidence from cross country‐ data. Agricultural Economics, 34(3), 221-228. https://doi.org/10.1111/j.1574-0864.2006.00120.x

Lokanathan, S., & Kapugama, N. (2012). Smallholders and Micro-enterprises in Agriculture: Information needs and communication patterns. http://dx.doi.org/10.2139/ssrn.2309313

Marker, P., McNamara, K., & Wallace, L. (2002). The significance of information and communication technologies for reducing poverty. London, UK: The Department for International Development.

Maumbe, B. M. (2010). E-agriculture and E-government for Global Policy Development: Implications and Future Directions. Hershey, NY: Information Science Reference.

Mignouna, D. B., Abdoulaye, T., Alene, A., Akinola, A. A., & Maroya, N. (2014). Baseline protocols: the case of Yam Improvement for Income and Food Security in West Africa (YIIFSWA) project.

139

(YIIFSWA working paper series, No. 4). Ibadan, Nigeria: International Institute of Tropical Agriculture.

Mignouna, D. B., Abdoulaye, T., Akinola, A. A., Alene, A., & Nweke, F. (2015). Factors influencing the use of selected inputs in yam production in Nigeria and Ghana. Journal of Agriculture and Rural Development in the Tropics and Subtropics (JARTS), 116(2), 131-142.

Minten, B., Stifel, D., & Tamru, S. (2012, May). Structural transformation in Ethiopia: Evidence from cereal markets. (IFPRI-ESSP II Working Paper 39). Addis Ababa, Ethiopia: International Food Policy Research Institute.

Molony, T. (2006). "I don't trust the phone; it always lies": Trust and information and communication technologies in Tanzanian micro-and small enterprises. Information Technologies & International Development, 3(4), pp-67. doi: 10.1162/itid.2007.3.4.67

Munthali, N., Leeuwis, C., van Paassen, A., Lie, R., Asare, R., van Lammeren, R., & Schut, M. (2018). Innovation intermediation in a digital age: Comparing public and private new-ICT platforms for agricultural extension in Ghana. NJAS-Wageningen Journal of Life Sciences, 86, 64-76. https://doi.org/10.1016/j.njas.2018.05.001

Nagasaka, I. (2007). Cellphones in the rural Philippines. In R. Pertierra (Ed.), The social construction and usage of communication technologies: Asian and European experiences , (pp. 100-125). Quezon City, Philippines: The University of the Philippines Press.

Nakasone, E. (2013, August). The role of price information in agricultural markets: experimental evidence from rural Peru. Paper presented at the 2013 Annual Meeting of the Agricultural and Applied Economics Association, Washington, DC.

National Communications Authority (NCA). (2018). Mobile Voice Subscription for December 2018. Retrieved from https://www.nca.org.gh/assets/Uploads/Industry-Information-December-2018-Final2.pdf

Ngeleza, G. K., Owusua, R., Jimah, K., & Kolavalli, S. (2011). Cropping practices and labor requirements in field operations for major crops in Ghana: What needs to be mechanized? (IFPRI Discussion Paper 1074). Washington, DC: The International Food Policy Research Institute

Ogutu, S. O., Okello, J. J., & Otieno, D. J. (2014). Impact of information and communication technology-based market information services on smallholder farm input use and productivity: The case of Kenya. World Development, 64, 311-321. https://doi.org/10.1016/j.worlddev.2014.06.011

Omland, H. O., & Thapa, D. (2017, May). Methodological approach for identifying mechanisms in ICT4D: A critical realism perspective. In In J. Choudrie, M. S. Islam, F. Wahid, J. M. Bass, & J. E. Priyatma (Eds.), Information and Communication Technologies for Development: Proceedings of the 14th IFIP WG 9.4 International Conference on Social Implications of Computers in Developing Countries (pp. 182-193). Cham, Switzerland: Springer.

Parkinson, S., & Ramirez, R. (2006). Using a sustainable livelihoods approach to assessing the impact of ICTs in development. The Journal of Community Informatics, 2(3). Retrieved from http://ci-journal.net/index.php/ciej/article/view/310

140

Pieterse, J. N. (2010). Development theory. London, UK: Sage.

Qureshi, S. (2015). Are we making a better world with information and communication technology for development (ICT4D) research? Findings from the field and theory building. Information Technology for Development, 21(4), 511-522. https://doi.org/10.1080/02681102.2015.1080428

Samuel, J., Shah, N., & Hadingham, W. (2005). Mobile communications in South Africa, Tanzania and Egypt: results from community and business surveys. In Vodaphone, The Vodafone Policy Paper Series, (pp. 44-52). Birkshire, UK: Vodaphone.

Schwab, K. (2017). The fourth industrial revolution. New York, NY: Crown Business.

Sein, M. K., & Harindranath, G. (2004). Conceptualizing the ICT artifact: Toward understanding the role of ICT in national development. The information society, 20(1), 15-24. https://doi.org/10.1080/01972240490269942

Sekabira, H., & Qaim, M. (2016). Mobile money, agricultural marketing, and off-farm income in Uganda (Global Food Discussion Paper 82). Göttingen, Germany: Global Food. https://doi.org/10.1111/agec.12360

Sey, A. (2011). ‘We use it different, different’: Making sense of trends in mobile phone use in Ghana. New Media & Society, 13(3), 375-390. https://doi.org/10.1177/1461444810393907

Sife, A. S., Kiondo, E., & Lyimo Macha, J. G. (2010). Contribution of mobile phones to rural livelihoods‐ and poverty reduction in Morogoro region, Tanzania. The Electronic Journal of Information Systems in Developing Countries, 42(1), 1-15. https://doi.org/10.1002/j.1681-4835.2010.tb00299.x

Skuse, A., & Cousins, T. (2007). Managing distance: rural poverty and the promise of communication in post-apartheid South Africa. Journal of Asian and African Studies, 42(2), 185-207. https://doi.org/10.1177/0021909607074867

Souter, D., Scott, N., Garforth, C., Jain, R., Mascarenhas, O., & McKemey, K. (2005) The Economic Impact of Telecommunications on Rural Livelihoods and Poverty Reduction: a study of rural communities in India (Gujarat), Mozambique and Tanzania (Report of DFID KaR Project 8347). London, UK: The Department for International Development

Souter, D. (2012). ICTs, the Internet and Sustainability: A discussion paper. Winnipeg, Canada: International Institute for Sustainable Development.

Steyn, J., Hayes, N., Kirlidog, M., & La Rovere, R. L. (2013). A brief critique of the standard view of ICT4D. Paper presented at the 12th International Conference on Social Implications of Computers in Developing Countries, Ocho Rios, Jamaica.

Tadesse, G., & Bahiigwa, G. (2015). Mobile phones and farmers’ marketing decisions in Ethiopia. World Development, 68, 296-307. https://doi.org/10.1016/j.worlddev.2014.12.010

141

Thapa, D., Sein, M. K., & Sæbø, Ø. (2012). Building collective capabilities through ICT in a mountain region of Nepal: where social capital leads to collective action. Information Technology for Development, 18(1), 5-22. https://doi.org/10.1080/02681102.2011.643205

Thompson, M. (2004). Discourse, ‘development’ & the ‘digital divide’: ICT & the World Bank. Review of African Political Economy, 31(99), 103-123. https://doi.org/10.1080/0305624042000258441

Toyama, K. (2011, February). Technology as amplifier in international development. Paper presented at the 6th iConference, Seattle WA.

United Nations. (2016, June 7). To leave no one behind, innovation is key. [Web Blog Post] Retrieved from http://unsdn.org/2016/06/07/to-leave-no-one-behind-innovation-is-key/

Van der Boor, P., Oliveira, P., & Veloso, F. (2014). Users as innovators in developing countries: The global sources of innovation and diffusion in mobile banking services. Research Policy, 43(9), 1594-1607. https://doi.org/10.1016/j.respol.2014.05.003

Wade, R. H. (2002). Bridging the digital divide: new route to development or new form of dependency. Global governance, 8(4), 443-466.

Walsham, G. (1995). The emergence of interpretivism in IS research. Information systems research, 6(4), 376-394.

Walsham, G. (2013). Development informatics in a changing world: Reflections from ICTD2010/2012. Information Technologies & International Development, 9(1), 49-54.

Walsham, G. (2017). ICT4D research: reflections on history and future agenda. Information Technology for Development, 23(1), 18-41. https://doi.org/10.1080/02681102.2016.1246406

Webster, F. (2014). Theories of the information society. London, UK: Routledge.

Wilson, M. (2002). Understanding the international ICT and development discourse: assumptions and implications. The African Journal of Information and Communication, 3, 80-93. doi: 10.23962/10539/19833

Zanello, G., & Srinivasan, C. S. (2014). Information sources, ICTs and price information in rural agricultural markets. The European Journal of Development Research, 26(5), 815-831. doi: 10.1057/ejdr.2014.1

Zheng, Y., & Heeks, R. (2008). Conceptualising information culture in developing countries. (Development Informatics Working Paper 34). Manchester, UK: Development Informatics Group.

Zhang, Y., Wang, L., & Duan, Y. (2016). Agricultural information dissemination using ICTs: A review and analysis of information dissemination models in China. Information processing in agriculture, 3(1), 17-29. doi: 10.1016/j.inpa.2015.11.002

142

APPENDICIESAPPENDIX 1: University of Surrey University Ethics Committee confirmation of ethical opinion

143

144

145

APPENDIX 2: UNIVERSITY ETHICS COMMITTEE ETHICS APPLICATION FORM

Ethics Application Form

Section A Guidance1. Title of project

Livelihood Development and the Impacts of Mobile Phone on Yam Production in Ghana

Please provide a summary of the projectOver the past 10 years the use of information and communication technologies (ICTs) has been linked with improvements in livelihoods. Preliminary research focused on adoption of technology and barriers to use and only recently have researchers began to investigate behaviour patterns associated with use and impacts. Yet, the crosscutting and complex role that technology plays within individual’s livelihoods has proved difficult to fully understand and analyse. For this reason, research into ICTs and agriculture production has been limited to focusing on specific use and impacts. My aim is to gain a richer and more comprehensive understanding of how farmers use ICTs, in this case mobile phones, and how that use impacts their livelihoods.In order to achieve this, I will use a mixed methods approach involving a qualitative analysis to identify types of use and usage behaviours, and then a quantitative analysis to assess the impacts of mobile phone use on livelihoods. The study locations will be composed of several small farming villages outside of the Ejura, Ghana. Livelihoods in these villages are mainly derived from the production of yams and significant development challenges currently persist. It is our hope that that this kind of research can identify effective strategies for livelihood development involving the use of mobile technology and identify ways in which farmers not exploiting the potential that these technologies provide might increasingly do so in the future.

Approximately 500 words, must be in lay termsThis must not be taken from your protocol.

2. Name of person submitting applicationTyler Jay Reynolds

Main contact for any correspondence

3. Level of research:

PhD ☒ Staff ☐

Other:

Faculty:FEPS

UG and PGT applicants must contact their Faculty Ethics Committee

4. Is this project a collaboration with an external body?Kingsley Osei, a senior scientist at the Crops Research Institute (CRI), has agreed to be an ‘in country’ supervisor for this project. An invitation letter provided by the Crops Research Institute parent body: The Council for Scientific and Industrial Research (CSIR) is attached

Yes

☐

No

☒

If yes, please state the collaborators in the space providede.g. another Higher Education Institution (HEI), NHS, Pharmaceutical company

5. Is this research funded?This research is self-funded by the principle researcher

Yes

☐

No

☒

If yes, please provide details of the funding in the space provided

146

6. Where will the research be carried out?Ejura, Ghana

Main research site:Ejura, GhanaOther research site/s:Villages around Ejura, Ghana

e.g. University of Surrey, NHS site/s, other HEI

7. Proposed research start date: March 25, 2017Expected research end date: May 25, 2017

Section B – Recruitment information Guidance1. Will you be recruiting participants? Yes

☒No☐

2. If yes, from where will you be recruiting participants?In the villages around Ejura, Ghana

e.g. UoS staff and/or students, email, posters, online survey

3. Estimated number of participants:160

4. Details of reimbursement to participants:Participants will be reimbursed with a monetary payment equal to typical wages, approx. 2 pounds,

e.g. Travel expenses, entry into prize draw

5. Does your research involve contact with children or vulnerable adults?

Yes

☐

No

☒

If yes, please provide details

6. What are the potential benefits to research participants?The outcomes of this research might identify livelihood strategies that produce positive livelihood outcomes. The identification of these strategies could be used by development agencies and extension worker in promoting means by which livelihoods could be improved. This promotion could serve to inform participants and methods by which they could improve their lives and would therefore provide them with benefit. If this study identifies policies, institutions or market processes that serve as barriers to development, this information could be used to recommend policies that may reduce these barriers and therefore provide benefits to participants.

You should state hereany potential benefits tobe gained by the researchparticipant through taking part in the research either now or in future. However, don’tover-emphasise the benefits. In some cases there may be no apparent benefit.

7. Will you be collecting participant’s personal data during the project?Personal data that will be collected will be: age, gender, and level of education

Yes

☒

No

☐

If yes, please provide details.e.g. email addresses, personal addresses, telephone numbers, postcodes

8. I confirm I will retain research data for at least 10 years, in line with the University of Surrey’s Code on Good Research Practicehttp://www.surrey.ac.uk/about/corporate/policies/code_on_good_research_practice.pdf

☒

Section C – Document checklist Guidance1. Protocol ☒ Must include version

number and date

2. Participant Information Sheet ☒ Must include UoS logo, version number and date

147

3. Consent Form ☒ Must include UoS logo, version number and date

4. Questionnaire (s) ☒ If relevant, please include questionnaires for all participant groups

5. Interview Schedule ☒ If relevant

6. Recruitment email/advert/poster ☐ Please note that posters should only be put up on dedicated poster boards. Where applicable, please ensure you obtain permission from the appropriate authority.

7. Risk Assessment (s) ☒ Please provide justification if a risk assessment has not been included

8. Evidence of agreement with other collaborators ☒ If relevant

9. Relevant insurance proforma ☒ If unsure of the relevant documentation please contact ethics@surrey.ac.uk

10. Short Curriculum Vitae for the research team ☐ This should be no longer than 2 pages, and is an optional addition to your application.

11. Other, please state: ☐ e.g. DBS confirmation, debriefing statements.

Section D Guidance1. I confirm that I have read and understood the Ethical Principles

and Procedures for Teaching and Research ☒2. Please indicate the process the application is being submitted for:

Full review ☐ Proportionate review ☒

3. Names and signatures of all investigators (This must be an electronic or wet [handwritten] signature. We cannot accept a typed name in place of a signature)

Chief Investigator: Tyler Jay Reynolds

For doctoral level research the CI should be the student

For PhD projects

I, the undersigned, confirm that I have read and approved the attached documentation for submission to the University Ethics Committee.

Academic Supervisor:

Co-supervisor :

For staff

Co-investigator:

148

Co-investigator:

Co-investigator: Co-investigator:

Final checks

Please ensure that your documentation has been proof-read before submitting it for the consideration of the University Ethics Committee (UEC). If spelling errors are identified in the initial checks, all documentation will be returned to the researcher. Failing to proof-read study documentation can cause significant delays in the UEC reviewing process

All study documentation must show the version number and date Please ensure that you have this application form signed by all relevant people, we cannot

accept a typed name in place of a signature Please include the relevant insurance proforma.

If you have any questions please contact us:

ethics@surrey.ac.uk , 01483 68 9103 or 01483 68 2051

Project Protocol (version 2, 15/02/2016)

Livelihood Development and the Impacts of Mobile Phone on Yam Production in Ghana

Principle Investigator: Tyler Jay Reynolds

149

Background on the study: Over the past 10 years the use of information and communication technologies (ICTs) has been linked with improvements in livelihoods. Preliminary research focused on adoption of technology and barriers to use and only recently have researchers began to investigate behaviour patterns associated with use and impacts. Yet, the crosscutting and complex role that technology plays within individual’s livelihoods has proved difficult to fully understand and analyse. For this reason, research into ICTs and agriculture production has been limited to focusing on specific use and impacts. My aim is to gain a richer and more comprehensive understanding of how farmers use ICTs, in this case mobile phones, and how that use impacts their livelihoods.

In order to achieve this, I will use a mixed methods approach involving a qualitative analysis to identify types of use and usage behaviours, and then a quantitative analysis to assess the impacts of mobile phone use on livelihoods. The part of the study involving a quantitative analysis will be the subject of this ethics application. This ethics application only covers the qualitative phase of this study. The only quantitative data collected during the qualitative phase of this study will be age, gender and education. The study locations will be composed of several small farming villages outside of the Ejura, Ghana. Livelihoods in these villages are mainly derived from the production of yams and significant development challenges currently persist. These challenges consist of everything from water access to improper farming techniques. It is our hope that that this kind of research can identify effective strategies for livelihood development involving the use of mobile technology and identify ways in which farmers not exploiting the potential that these technologies provide might increasingly do so in the future.

Objectives of the Study: The objectives of this study are to collect a small amount of ethnographic information (age, gender, and level of education), and information regarding the use of mobile phones and other information and communication technologies, and how that use is translated into livelihood outcomes. The research framework is made up of a combination of the sustainable livelihoods framework (Carney et al. 1999) and an ICT-for-development input-process-output model developed by Heeks (2010). Accordingly, questions will be geared toward gaining an understanding of the vulnerability context in which people live, the capital assets which they possess and the strategies by which they transform those capitals into livelihood outcomes. Subsets of questions within each of those aspects of livelihood development will focus on individual’s readiness to use ICTs, the availability of ICTs, and the use of ICTs by which individuals achieve livelihood outcomes. This study will not test a hypothesis as it will implement a grounded theory approach.

Selection of Participants: Participants will be selected based on a snowball sampling technique. Participants will be selected from the villages surrounding the town of Ejura, Ghana and locations where yam production serves as a substantial local economic driver of livelihood development. Attempts, based on visual observation, will be made to gain a diverse selection of participants in terms of age, gender, and level of education. Participants with no access to mobile phones will be excluded from the study.

Number of Participants: Approximately 160 individuals will be recruited for this study. This number was decided upon because it is a representative sample size that is suitable for statistical analysis of the data.

Recruitment methods: Recruitment methods will consist of a verbal request of the principal researcher in the field.

In some cases it is possible that participants will be recommended by local/village leaders. In these cases the local/village leaders will serve as gate keepers to the community. Establishing a relationship of trust among these gate keepers will be critical to the success of this study and will be developed over the necessary and appropriate amount of time. This relationship will be aided by the introduction of the researcher to the village leaders by a member of the Crop Research Institute

150

(CSIR), an organization which has a long history of contact with the village leaders and community members. That said, if a village leader does not deem the research appropriate at any time, the research will cease to proceed in that location.

It is likely that the village leader will introduce the researcher to possible participants. These recommendations will be respected and likely provide the jumping-off point for the snowball sampling method for selecting participants. If the village leaders asks for criteria for recommending possible participants, the researcher will request a variety of participants based on age, gender, level of education, and known mobile phone use experience.

This recruitment will likely look like the following according to the principle researcher’s past field experience and phase one of this project. In local villages such as the ones that will be included in this study, the arrival of an outsider/visitor, and especially a researcher or aid worker, will mark a sort of occasion in the village. It is typical that the village leader or an assistant of a village leader will greet a visitor as the village leader will be informed of the arrival of the visitor, typically via mobile phone prior to the arrival of the visitor. The news of a visitor coming, will often spread through a village by word of mouth very quickly. Upon arrival, it is common that any inhabitants of the village that are not directly involved in a pressing economic or social activity will come to investigate the visitor. This often results in a crowd, typically in a central meeting location in a village. At this point it is typically customary and culturally appropriate for the village leader or an assistant to introduce the visitor where-by increasing the level of comfort that inhabitants of the village have. The village leader and the any inhabitants that arrive will be briefed by the principle researcher on the reasons for the visit and the potential activities that will be taking place. From here the principle researcher will need to manage the recruitment process with great care. The first step in this process will be to ask the village leader if they have any suggestions for participants based on the priority of having a variety of participants based on age, gender, and level of education. The village leader will have an immense knowledge of the people in his or her village with regard to these variables. It is also here that the principle researcher will reinforce the idea that the study is seeking individuals with at least some experience using mobile phones. It is also here that the principle researcher will inform the village leader that as part of the protocol of this project, the principle researcher retains the right to decline a recommendation for a particular person as a participant at any time if the principle researcher feels that any individual feels undue pressure to participate. It is often the case that following this exchange, the village leader will provide recommendations for participants by calling to them if they are present or providing their contact if they are not present but may return and hence be available to participation at a later time. It is also customary that a village leader may ask for volunteers to participate. In most cases this action will result in a resounding agreement to participate and in doing so potential participants will often form a line or queue. This will have to be managed by the principle researcher by informing the village leader and the inhabitants of the village of what this process might look like. The principle researcher will make it clear to the village leader and the villagers that they will be subject to a consent process which they will be given ample time to consider. The principle researcher will not openly place a defined time limit on the consideration process in that doing so places greater pressure on a potential participant than if they are informed that they will have ‘ample’ time. Even if the principle researcher would state than any potential participant would have 45 minutes or even one hour to consider participating, regardless of the simplicity of the consent process, this defined time frame would be comparatively more stressful and possibly even confusing in that even the thought of a consent process that could possibly take such an extended period of time would be strange and unlike any research they had participated in previously or even heard of. The participants will not be informed that the principle researcher will cease to participate in the consent process if it takes longer than 45 minutes for a participant to decide if they would like to participate. The consent process described below, in a worst case scenario of extended questions from the participant aimed at gaining a better understanding of any aspect of the consent process, should take no longer than 45 minutes. That is

151

reasonable and in most cases the consent process will take no longer than 10 minutes. So to be clear, the recruitment process may take, but is not limited to, approximately 20 minutes, the consent process may take approximately 45 minutes, and the interview process may take approximately 1 hour. Due to the voluntary nature of the recruitment process described so far, it is unlikely that participants will feel pressured to participate. The study population in this study will have experience with researchers and many will know that they have an option to participate in a study of any kind. If any pressure on them participate exists, it will likely be the result of the village leader’s recommendation and more likely the result of the economic incentive to participate as a result of the compensation for lost time monetary payment for participation. While the people in this study will almost surely make less than $2 USD per day, they will not be so pressured by this compensation for lost time monetary payment so as to face undue pressure or harm in participation. The majority of participants will be yam farmers, the stress and pressure of farming practices would far outweigh any stress or pressure they would face being interviewed. It is for this reason that many participants may see this study as the easiest income they can make, so much so that they may try to take advantage of the opportunity in a cynical way. This circumstance is examined in more detail later in this report, but should serve as an indicator that the stress and pressure of participating this this study as a result of the potential for monetary compensation is both comparatively and actually very low. According to the World Bank (2016), the average income of Ghanaians is approximately $4.35 USD per day. The study area is not located in the poorer northern regions of the country and it is expected that individuals in the study area will earn between $3-$4 USD per day. While this number may seem small, it is well above the extreme poverty line of less than $1 USD per day and with the nature of the livelihoods of the participants in this study being subsistence farmers, food security is not so bad as to be a life and death challenge. In other words, no potential participant in this study will deem the monetary compensation payment received for participation as anything more than a small boon which was both reasonable and fair.

During the recruitment process, the potential participant will be briefed on the project in terms of the types of information that the principle researcher is seeking. This will involve a short explanation of aspects of the project such as livelihood development challenges that the potential participant may be facing, capital assets, the use of strategies to address challenges, and changes in the significance of the challenges over time. They will also be briefed on the questionnaire process in terms of number of questions and the time frame (approx. 0.5 hour). This process of informing potential participants will be conducted in an informal manner so as to not cause a potential participant stress. A certain level of ‘small talk’ will also be involved in this process as a means by which to gain a potential participant’s trust and raise their level of comfort. This is no way is a means by which to deceive potential participants into a false sense of security. It is fairly typical for inhabitants of villages to take great interest in a visitor, especially one from the west. It would not be out of the ordinary to carry on a casual conversation for more than 20 minutes before even beginning the recruitment process. The best results in social science are gained when the participant and the researcher build rapport with one another. The major outcome of this is not only to ease in the interview process but to improve communication. Difference in cultures in terms of communication must be overcome on a learning curve for both parties and casual and honest conversation may serve at the best means to achieve this. If at any time during the briefing of the project in the recruitment stage the participant has specific questions about the project or seems confused, the principle researcher will inform the potential participant that the principle research has in his possession, a participant information sheet (v3. 10/01/2017) which provides information about the project in great detail. It should also be noted that any discussion involving potential livelihood development challenges will be intensely moderated by the principle researcher so as to not plant unwarranted information or topics in the mind of a potential participant. This conversation of topics and issues will remain vague and only intend to provide a potential participant with a general idea of the concepts behind the research. The principle researcher will also make hard copy paper versions of the participant information sheet available for any potential participant to keep.

152

The principle researcher will also state the he is available to read aloud any and all information contained in the participant information sheet at the request of the potential participant. If the potential participant reads, or has the participant information sheet read to them by the principle researcher, and the potential participant still requests further information or explanation, the principle researcher will provide additional information and explanation to the best of his ability. It can be difficult to predict how potential participants will conceptualize and retain knowledge about the project. Overcoming this, in circumstances where the participant information sheet is too complicated or difficult for a participant, sufficient explanation will fall on the shoulders of the principle researcher. This may require a level of creativity that facilitates effective communication and learning while at the same time retains the information, principles and requirements that this ethics application provides. If at any time the principle researcher feels that the task of communicating information about the project, either on the participant information sheet or otherwise, has become too difficult, the principle researcher will end the recruitment, consent, or interview process. This circumstance should serve to show the implications of providing information on the participant information sheet either in writing, verbally, or in discussion. It should be clarified here what is meant by discussion as a means by which to communicate the information on the participant information sheet. The word ‘discussion’ is used here because in the event that the potential participant requests that the principle researcher communicate any and all information on the participant information sheet verbally, the principle researcher will undertake a system of reading parts of the sheet and then pausing to ask the potential participant if they understand what has been read and if they have in further questions for explanation or discussion. This will serve to decrease the stress that an individual might have in dealing with the level of information contained in the participant information sheet, consent form, or any other aspect of the project. One other possible implication for this form of information provision is the possibility that a potential participant may become distracted by the multitude of information possibly being brought to their attention. In order to overcome this, the principle researcher will seek to review past information that may have been lost throughout the process. The goal of the recruitment and consent process will not only be a specific one in terms of critical points of understanding, but also a holistic one where a potential participant should be able to conceptualize the project as a whole, at least to a sufficient extent.

In instances where a queue for participation does not form, a snowball method for gaining participants will be implemented. This assumes that at least one person that the village leader recommends participates in the survey and that no other inhabitants of the village that could be potential participants are present in the area where the principle researcher is received by the village leader. If this is the case, the principle researcher will ask the participant if they could recommend any other individuals as participants and if assistance in making contact with the potential participants could be provided in some manner or another (typically verbal introduction). Again, it is here that any potential participants will be informed that they will be given ample time to consider participation in the study.

In the event that a trail of a snowball runs dry, the principle researcher will seek assistance from the village leader for more recommendations or for a blessing to walk around and ask people in the village if they would like to participate in the study. In the villages that will be studied most of the housing structures are relatively open, they rarely have typical doors or windows and almost all of them will have dirt floors. In places like these, especially small villages, people tend to sit or do house work outside their houses. This makes walking up and introducing one’s self very easy. In these cases the inhabitants of the village will be aware of the principle researcher’s presence long before the principle researcher will be aware of theirs as a result of the way word spreads through small villages. This is often facilitated by the inevitable groups of children that move around the villages outside during the day. If an inhabitant is not entirely distracted with some kind of work, it is very likely that they will at least come to their door or stand outside their house to greet a visitor or

153

in the very least look into who that visitor might be. Again, this makes approaching people very easy and with little stress because they have essentially made themselves available to you.

In cases where the researcher perceives any form of undo pressure placed on a participant recommended by a village leader, that potential participant will not be selected for an interview and other participants will be sought. This situation can arise when for whatever reason a potential participant is not provided sufficient opportunity to decline participation in the study due to intentions of a village leader.

The presence of the village leader in the approximate location of an interview has the potential to have adverse effects on the extent to which participants answer questions honestly. It is often the case that in the presence of a local leader that is often charged with governing and providing opportunities for the village to grow and develop, participants have incentives to either respond in a manner that downplays the livelihood related challenges they face for fear of challenging the leader or expressing disapproval which might have adverse effects, or participants might up play or exaggerate the challenges they face in their livelihood development if they perceive the study to be a tool in which any failings of the local leader, whether real or fanciful, can serve as leverage for change or even to serve some selfish purpose.

The probability of having to address these sorts of adverse effects from village leaders is low due to long history of community members having interactions with research and development agencies such as the CSIR. Community members and village leaders are accustomed to the presence of researchers and development practitioners, and are typically versed in the motives and positive outcomes that these interactions have.

Kingsley Osei, a senior scientist at the CSIR, has agreed to be an ‘in country’ supervisor of this project. Kingsley is not an official supervisor for this project. He has an in-depth knowledge of the research area, village leaders, and community members. Kingsley will serve as a conduit between myself and the village leaders. This will greatly decrease the risk of stress on the gate keepers and community members that trust Kingsley and any research, such as my own, that is associated with the Crops Research Institute. It may also be the case that the local police should be contacted about my operating in the area. This contact will likely be liaised via the CSIR. It is often the case that local police should be informed of research activity. The choice to connect with the local police will be made on recommendation for doing so by the CSIR. The principle researcher will rely on the CSIR for advice regarding the contacting the local police as this activities varies between and within in countries.

Method of taking informed consent: The process of taking informed consent will begin with the principle researcher asking the potential participant if they would like to read the Written Consent Form (v3. 10/01/2017) or have the Verbal Consent Form v3. 10/01/2017) read to them. This will most likely be a fluid process with the possibility that a participant may at first ask to read the Written Consent Form and then decide that the information is too cumbersome and they would like the Verbal Consent Form read to them. The principle researcher will make it clear to the participant that the Written Consent Form and the Verbal Consent Form serve the same purpose. It is possible that this decision making process will in itself be cumbersome for potential participants and the principle researcher will err on the side of the Verbal Consent Form if a potential participant indicates that they are struggling to comprehend the Written Consent Form.

In either case of using the Verbal Consent Form or the Written Consent Form, the principle researcher will offer the participant a paper copy of either consent form and a copy of the Participant Information Sheet (v3. 10/01/2017). If the participant has any questions about the Verbal Consent Form, the Written Consent Form, or the research project, the principle researcher will refer to the Participant Information Sheet. No information contained in the Participant Information Sheet

154

would affect the ability for participants to answer questionnaire questions honestly and to the best of their ability.

The Written Consent Form (v3. 10/01/2017) has been modified from the templates provided by the University of Surrey so as to make it less cumbersome for potential participants to digest. It is likely that in instances where participants are able to read and write, it is unlikely that their skills will be sufficient to comprehend the formality and legal-like language provided in the consent form templates provided by the University of Surrey. An example of a change made by the principle researcher to the Written Consent Form is the omission of the acknowledgement of data being held and processed in accordance with the UK Data Protection Act (1998). This information will largely be meaningless to the study group and only serve to unnecessarily increase the level of stress they feel throughout the consent process.

In cases where verbal consent is given, the researcher will indicate that consent was given on the Verbal Consent Form (v3. 10/01/2017) by initialling next to the line of script contained on the form that have just been agreed to. This process is justified in accordance with the UCL Research Ethics Committee Guidance Note 2: The Ethics of Research Related to Healthcare in Developing Countries (https://ethics.grad.ucl.ac.uk/forms/guidance2.pdf) which states:

“The 'Guidelines for the Conduct of Health Research involving Human Subjects in Uganda' note that a research participant's wish not to execute a written informed consent form should be honoured but the investigator must obtain oral informed consent and document such. NBAC (2001) recognises that this rejection stems from Uganda's past experience of torture and persecution of individuals found to be associated with particular enterprises and that individuals may consequently be reluctant to sign a form which associates them with certain activities. In some cultures, participants' only experience of signing forms may be in relation to tax documents or court proceedings. Thus, signing a consent form is likely to have negative connotations, making otherwise willing participants less likely to take part. In one research trial examining the consequences of domestic violence, it was considered inappropriate to ask female participants to sign a consent form before enrolling them in the research because of their concerns that signing a form would mean that a record of victims of domestic violence would be kept and this might lead to them suffering more harm. If requesting that participants sign consent forms is inappropriate, other means of recording their genuine consent to participation in research is required to protect them from being enrolled in research that they have not consented to. In many circumstances, the research worker who is informing the participant will sign a form stating that the appropriate information was given and verbal consent received.”

It is unlikely that participants in my study will be literate and in cases where participants are able to read and write they will be presented with the option of reading and signing the Written Consent Form (v3. 10/01/2017). If the participant requests the Written Consent Form and then decides they would rather be subject to the verbal consent process, then the verbal consent process will begin.

The issue of providing potential participants with enough time to make a decision with regard to the Verbal Consent form is reconciled by the insertion of breaks in the Verbal Consent Form script. These breaks include a questions asking the participant if they understand what is being said, if what is being said makes sense, and if they have any questions. The script also includes the line “Take as much time as you need to decide.” The time involved in gaining consent does play a role in the ability of the principle researcher to carry out the research. An approximate timeframe of 30 minutes has been set for questionnaire completion. While the consent process described above is comprehensive, the limited attention span of the participants will mean that a degree of expeditiousness will be sought so as to capitalise on the time available for questionnaire completion. As stated above, the consent process can take approximately 45 minutes before the principle researcher will end the process. This expeditiousness will not be sought at the expense of the participant’s allowance for an appropriate amount of time to consider whether consent will be

155

given. The length of participant’s attention span is difficult to predict and it is expected that it will vary, possibly significantly. Due to the nature of this project largely focusing on the use of mobile phones, if a participant only rarely uses mobile phones, or is entirely unable to describe or conceptualize their use of mobile phones, this finding will be included in the data set but could result in an abbreviated interview at the request of the researcher or participant.

The monetary payment for compensation of time spent during the interview will be determined by a combination of factors. Advice from the CSIR will be sought on this matter in the field as their researchers will have the best knowledge and experience when it comes to defining what an appropriate payment is. It is important that payment for participation in this research does not significantly exceed payments for past research as that could set an unhelpful precedent for other research that is going on and conducted by other ethical research organizations in the area. A payment that is much higher or much lower than other research organizations would likely affect the ability for participants to answer questions honestly and to the best of their ability. While it is often the case that participants will receive a decreased amount of compensation if they fail to complete the questionnaire, this action will only be taken if it seems as if the participant is leveraging or taking advantage of the system by ending the questionnaire process early without a valid reason. In other words, a participant might conceive a situation where they could make some quick and easy money by agreeing to complete the questionnaire and then quickly ending the questionnaire process under the false pretence that they would be paid the same amount as a participant that completed the questionnaire. For this reason, the wording on the Verbal Consent Form is such: “If you decide to participate in this study you will be compensated for your time and inconvenience.” In practice, the highest amount any participant will receive is highly unlikely to exceed the equivalent of 2 pounds. This is because the average hourly income for participants is unlikely to exceed the equivalent of 2 pounds. It is very unlikely that any participant would receive less than the standard maximum payment, set according to the recommendations of the CSIR, for participation in the study. A decreased payment amount would only occur according the circumstance described above.

Several differences between the written and verbal consent forms deserve explanation:

The written consent form indicated that the participant information sheet has been provided, but the verbal consent form included wording that indicates that a participant information sheet can be provided if necessary. The verbal consent form does not require that a potential participant be provided the participant information sheet because the sheet contains written language which is assumed to be inaccessible to participants on the assumption that the inaccessibility of written language is the basis on which consent is being given verbally. The indication on the written consent from, that a potential participant has been provided the participant information sheet is different from an indication that they have read and understood the participant information sheet. This difference in meaning should serve as nearly equal to the intention of the verbal consent form indicating that a participant information sheet is available if a potential participant should so choose. It should also been noted that the use of the word ‘provide’ in the verbal consent form was intended to mean that in the case that a potential participant requests a participant information sheet, the principle researcher would give the potential participant the option of reading that sheet, or more likely having that sheet read to them by the principle researcher according to the protocol described in the recruitment methods described above.

The written consent form includes the wording ‘data to be used for this study/future research that will have all relevant legal, professional, and ethical approvals’ and the verbal consent script does not. The omission of this wording from the verbal consent form was motivated by the fact that it is highly unlikely that potential participants would have any knowledge of the legal, professional and ethical approvals that are involved in research and this kind of language would only

156

result in the potential participant experiencing stress and possibly misinterpreting in the intention behind this wording. Again, this omission was driven by the likely distrust in legal and government institutions. It is likely that the statement that I am a researcher from the University of Surrey in the verbal consent form will serve as sufficient evidence that this is a professional study without the added stress of legal pressure. It is very unlikely that any potential participant that is being subjected to the verbal consent form would have any knowledge of an ethics approval process and the very mention of the words ‘ethics approvals’ may insight distrust, and the notion that this project may be in fact unethical in some way. While this may seem counterintuitive it is likely that any potential participant would place greater trust in the legality, ethics, and professionalism of this project as determined by the principle researcher (someone standing in front of them), than as approved by some higher power that could be manipulating the principle researcher in some malicious manner.

The same circumstance is the case with the omission of the requirement for data storage for monitoring and auditing purposes for 10 years in the verbal consent script. It is extremely unlikely that any potential participant that is subject to the verbal consent process would have an understanding of the definition, process, or intention of monitoring and auditing. It is also very unlikely that any potential participant would be able to understand why data would need to be protected and saved for 10 years. This information may serve to intimidate potential participants in the sense that they would be put on a watch list or that their information be saved for the purpose of manipulating them in some malicious manner in the future. This worry may only be increased by the reinforcement that this data would be saved in the UK under the UK Data Protection Act (1998) as indicated on the written consent form and omitted on the verbal consent form. The notion that data would be protected and stored in accordance to the ethics that are intended by the more specific wording used in the written consent form is provided by the verbal consent form’s inclusion of the language: “I will do everything I can to protect your privacy and information you provide.” This wording places emphasis on the concept that the principle researcher will be responsible for the protection of the data as opposed to a body that the potential participant is unlikely to be familiar with.

The verbal consent form does not indicate that in the event that a participant remove themselves from the study that their legal rights would not affected as is indicated on the written consent form. The reason behind this omission is that the potential participants that will be subject to the verbal consent process are unlikely to have any understanding of their legal rights as they apply to this research. The mention of any form of legal rights would likely be seen to imply that this study is somehow infringing on their legal right and incentivize them, in an unwarranted and unintended fashion, to decline to participate in this study.

The verbal consent form script does not include the requirement for participants to agree to participate to the best of their ability as is indicated on the written consent form. With regard to the verbal consent process, the inclusion of this requirement might be interpreted as binding legal jargon in a manner in which a potential participant’s responses might be used against them in some kind of malicious manner in the future. In other words, that they would be bound by this document (the verbal consent form) in some way that might cause them undue stress and might affect their ability to answer honestly and to the best of their ability. This may seem like a sort of two-way street. One direction involving a scenario where potential participants would be needlessly facing a disincentive as a result of a misinterpretation of the intention of the words, or the other direction, that they would lose their ability to answer questions honestly and to the best of their ability and as such answer questions dishonestly in a manner in which they perceived that the answers they provided would not facilitate some malicious use of that information against them by some bad actor.

Experimental design and methods used:

157

The experimental design is to use quantitative methods involving questionnaires to collect data on livelihood development and the use of ICT that can be statistically analysed. Variables and indicators included in questionnaire questions were determined using outcomes from the interviews in the first phase of this project. Statistical analysis will be carried out using SPSS software. Special attention will be paid to instances where ICT use as a strategy produces positive livelihood outcomes. Inquiry will also be made into possible correlations between demographics and possession of physical/financial capital with ICT use.

Information on collaboration with an external body and Consideration for working with other institutions: The relationship between this project and the Crops Research Institute does not completely meet the definition of project collaboration. That said, the nature of this relationship does deserve attention. As stated above, Kingsley Osei has formally agreed to be an ‘in country’ supervisor for this project. It is likely that he and other members of the CSIR will continue to aid in several ways. Already, the Deputy Director of the CSIR, Dr. Emmanuel Otoo, has provided a letter of invitation which will be used in the travel visa application process. Once in Ghana the CSIR will likely assist in providing or recommending appropriate accommodation in Kumasi and in Ejura. CSIR will be providing transportation to and from the field and will be providing safe and secure storage places for the data that is collected. This assistance from the CSIR will greatly reduce the risks of traveling and staying in Ghana. The employees of CSIR are uniquely familiar with the area and the people and will reduces risks of challenging logistics associated with doing research in the rural areas around Ejura.

An employee of the CSIR will be present during the questionnaire completion process. This will further put the participants at ease as the participants will have a level of trust with the CSIR. It is extremely unlikely that the presence of an employee of the CSIR would have any adverse effects on the participant’s ability to answer interview questions honestly, to the best of their ability, and without stress. If any adverse effects arise as a result of the presence of the CSIR during the interview, the principle researcher will ask the employee of the CSIR if the questionnaire process can be done in private. If the CSIR employee does not grant this wish and the participant remains unable to answer questions honestly, to the best of their ability, and without stress, the questionnaire process will not proceed.

The circumstance of a member of the CSIR being present during the questionnaire process may seem counterintuitive based on the issues described above whereby poor rural farmers are likely to have a level of distrust with large government and/or institutional bodies. A justification of a member of CSIR being present during questionnaire process requires significant experience working with such organizations, experience which the principle researcher has. Research and intervention agencies such as the CSIR have a very different relationship with inhabitants of villages than government bureaucrats or lawyers. This is due to the nature of the work that CSIR conducts. Central tenants of the CSIR mission include capacity building and livelihood development. They accomplish these goals through research and extension services. A large portion of the CSIR’s research into livelihoods development is very similar to the research being conducted in this study. This should serve as another aspect of the case for potential participants in this study already having some understanding for how research is carried out. The extension services that the CSIR provides typically include educational opportunities for farmers to learn ways in which to improve their agricultural production and hence, their livelihoods. For this reason, it is very likely that inhabitants of villages will see the CSIR in a very positive light: as bringers of positive impacts. For this reason, it may be the case that participants will feel more comfortable with the presence of a member of CSIR in the audience of an interview and may even request that a member of CSIR be present with the knowledge that the member of CSIR may be interested in what the participant is expressing. This emphasis on the presence of a member of CSIR as an audience member during the questionnaire process, at the request of a participant is likely the result of two scenarios, one of which serves to

158

express a possible bias that could result from the presence of a member of the CSIR. The first scenario aligns with the idea above, that a participant would both feel more comfortable with someone they know, or know of, from a body that they trust, and that they feel like a member of CSIR would find it useful to be privy to the information being expressed. The second scenario results from a phenomenon in the development intervention discourse where individuals that could be or are being subject to an intervention exaggerate the challenges they face or the negative state they are in, in hopes of drawing comparatively greater aid attention and assistance. An example of this occurred during the selection phase of the international Millennium Villages project (http://millenniumvillages.org/), but was not reported. The principle researcher observed this directly while working on agroforestry in Malawi with The World Agroforestry Center. During this work, the principle researcher experienced individuals clearly lying about the states of their livelihoods in an attempt to gain attention and development assistance from the principle researcher. This should serve as an extreme example of this bias and is extremely unlikely to replicate itself in this project. The reason for this is that the practices of the principle researcher and the CSIR vary in ways that downplay this possible bias. For instance, the majority of the extension work that the CSIR carries out is based on the idea of bottom-up empowerment involving providing people the skills so that they can improve their livelihoods on their own. This differs from past development interventions and projects such as the Millennium Villages project where organizations simply gifted material goods to individuals facing livelihood development challenges. This gifting of goods incentivizes individuals to appear needier than they actually are and creates discrepancies in aid attention at small spatial scales due to the limited amount of material resources that development agencies such as the one conducting the Millennium Village project possess. In the case of the CSIR and the study location where this study will take place, all of the villages in this study receive the same amount of attention from the CSIR. This means that no individual in any village will have an incentive to dishonestly play up the challenges that they face because they will already be aware of the fact that the CSIR does not preferentially dedicate more resources to one village or another. All of that said, it is still only remotely possible that a participant would assume that playing up the challenges that they face and pandering to the member of CSIR in the event that they are present for the questionnaire process. The principle researcher will be attentive to any overt or even covert emphasis that a participant would place on the information they are providing in a context where a participant would seek to portray their situation in hopes of achieving preferential treatment from the CSIR. If the principle researcher suspects that this may be the case, the principle researcher will request the leave of the member of the CSIR that is present during a questionnaire process. If the member of the CSIR is unwilling to void their presence from the questionnaire process, the questionnaire process will be terminated by the principle researcher.

It should also be said that the member of the CSIR that will be assisting in the form of acting as a liaison between the principle researcher and the village leader, will at the very least have a deep understanding of the scientific and research process and will likely be cognizant of the possible research biases described above. That said, the interaction between the member of the CSIR and the village leader as a gate keeper deserves attention. In the overwhelming majority of African villages, the village leaders serves as the gate keeper for the village. This means that for anyone to enter the village (spatial area) they must first gain the blessing of the village leader. This practice is in place for the protection of the village and its inhabitants in areas where law enforcement may be lacking. In the context of this project, law enforcement in the study area is of a sufficient standard so as to warrant safe travel and working conditions for the principle researcher but the concept of a village leader protecting their village is still relevant. Gaining the blessing of a village leader can be difficult in circumstances where a potential visitor does not have a liaison or some actor with enough trust instilled in them by the village leader so as to have the ability to recommend that a visitor be granted safe and welcomed entry to the village by the village leader. For this study, that liaison will be a member of the CSIR: an organization and its members having pre-established good quality relationships with all the village leaders in the area in a manner of mutual respect and appreciation.

159

No village leader will feel undue pressure from the CSIR to allow the principle researcher to enter their village. All of the village leaders in the study area are sufficiently empowered to reject the request for the principle researcher’s entry knowing that they will not face any negative repercussions as a result of any decision that they make. The process of requesting the entry of the principle researcher in a particular village will be composed of either a phone call or a face-to-face conversation between first the member of the CSIR and the village leader and possibly a continuation of that conversation between the principle researcher and the village leader. The makeup of that conversation will look largely like that of the conversation involved in the recruitment process whereby the project and methodology will be described. It is possible that logistics will also be discussed with the village leader in terms of the provision of food and accommodation for the day. This will likely involve a number of scenarios. It is possible that the village leader may request some small monetary or material compensation for entry to the village. This is not a bribe but could very well be custom in the area. It is also possible, and somewhat likely that the village leader will offer their house or the house of one of their substituents as a place where a meal provided by the village may be consumed, and/or interviews may be conducted, and/or the principle researcher may seek an area to rest or spend time with the inhabitants of the village. This is all part of the process of immersion in a village that when abided by typically produces the best results in terms of participants being able to participate in the study honestly and to the best of their ability. This process also decreases the stress that members of the community may experience as a result of the arrival of a visitor in a community. That said, it is typical that any visitor to a village would serve as a particular level of entertainment value and a positive and worthwhile disruption in an area where little such variety in day to day life might persist.

The members of the CSIR have a deep respect for the scientific process and have an understanding that any biases that they might insight would only weaken their ability to carry out their own work in the area. The purpose and role of the member of CSIR that will be assisting with this project should be clear. They will act as a liaison between the principle researcher and the village leaders first and foremost. This practice is essential to the success of this project, as in-country assistance from local actors such as the CSIR will serve as the best and most ethical means for gaining access to local villages. The alternatives being that the principle researcher would seek the approval of village leaders on his own are not viable and would most certainly result in a greater probability of receiving dishonest responses from participants. Following the function of serving as a liaison, the member of the CSIR will not technically play a role in this research but it should be noted that during the time spent in the field, the member of the CSIR will always be available to the principle researcher in case the principle researcher would need assistance in the form of transport, medical assistance in case of medical problems, and possibly doing a small amount of facilitative translation with members of the villages albeit not during the recruitment, consent or interview process. Again, the member of the CSIR will serve as an in country fixer whereby they will assist in logistics and planning in order to make the research process as smooth and safe as possible. It is likely that the member of CSIR would be present during the questionnaire process as a result of interest in the content of the project or even boredom. If at any time the principle researcher feels like the member of the CSIR is interfering with the project in any manner in which the ethics or scientific integrity of the project were being compromised, the principle researcher will terminate field activities. Contingency plans for this problem, in the case that it arises, involve a request by the principle to the CSIR for an alternate member of CSIR to serve in the intended manner that a member of the CSIR should serve as described above.

Here it might be constructive to provide some background on the Crops Research Institute (CRI) and the role that its employees will play in this project. The term Crops Researcher Institute can be used interchangeably with the term The Council for Scientific and Industrial Research (CSIR). These two bodies are one in the same and the terms describing them are often used interchangeably. The CSIR acts as a scientific arm of the larger Ghanaian Ministry of Agriculture (MOFA). CSIR is one of the

160

premier agricultural research bodies in the study area and was selected for assistance with this project for this reason. The CRI website can be found at: http://www.cropsresearch.org/ and should serve as evidence of the professionalism in science that the CRI embodies. The CRI will not seek to influence this project in any way. The link between the principle researcher and the CRI was instigated by the principle researcher’s supervisor, Prof. Stephen Morse, who regularly works with the CRI on research and intervention projects. Prof. Morse provided the contacts between the principle researcher and the members of the CRI. The CRI has provided the principle researcher with an invitation letter that the principle researcher will use in obtaining a travel visa to Ghana. The wording used in the letter of invitation, signed by Dr. Emmanuel Otoo (Deputy Director of the CRI) states: “I wish to invite Mr. Tyler Jay Reynolds to Ghana as part of his PhD studies under my supervision.” In this case the use of the word ‘supervision’ does not refer to supervision in terms of the formal relationship between a PhD student and one of his or her supervisors. The intention of the word ‘supervision’ in the letter of invitation was that of protection and support. The role of members of CRI in this project will be that of ‘fixers’, individuals and bodies that provide logistical support and assistance to the principle researcher. This support will include arrangement of accommodation, provision or organization of travel by car to and from the field and accommodation, provision of office space where work can be carried out, and to serve as a liaison between the principle researcher, village leaders and inhabitants of villages.

The vast majority of the study group will have almost no ability to speak or write English and they will not be able to write in their local language. The principle researcher will have the questionnaire included in this application translated into local language by a CRI employee. In the field a CRI employee will assist participants that are unable to read and write by reading the questions for the participant and the CRI employee will record the responses. Visual aids displaying simple images such as faces with expressions according to questions with responses that fall on a likert scale may also be used. The process of the CRI employee asking the participant the questionnaire questions is dependent on the principle researcher’s inability to communicate in the local language effectively even when the questionnaire has been translated. The local language that participants speak involves sounds and intonations that the principle researcher in unable to replicate. The principle researcher will not be recording responses to interview questions because the principle researcher will largely be unable to understand the language that participant’s speak. Due to complex nature of some of the questions in the questionnaire, there is a distinct possibility that participants may require some clarifications when trying to understand the questions and trying to respond. Overcoming the threat of bias in responses that result from clarifications to questions will fall on the principle researcher to communicate effectively with the CRI employee. The CRI employee will be required to inform the principle researcher of any questions or comments that the participants’ have regarding the questions and the principle researcher will provide appropriate clarifications or responses (or no clarification in the event that the answer might introduce some inappropriate bias) that will then be translated/transmitted to the participant via the CRI employee. The principle researcher will closely monitor the questionnaire completion process in the presence of the participant and CRI employee. Following the first phase of research for this project in which 48 interviews, each an hour in length, were carried out, the principle researcher has a keen sense for when a participant does not understand a question.

No data collected in the project will be made available to the CRI, currently no plans exist to collaborate with the CRI in any fashion greater than what is likely to be a small amount of translation. Members of CRI will not be involved in any data or findings, write up or analysis. The CRI has not played a role in the design of this project in any way. The motivation of the CRI to participate in the form of providing support and protection stems from an inclination to gain a greater understanding of the use of ICT in agriculture development via publications that are produced as a result of this project. This emphasis on understanding ICT in agriculture is driven by a host of aid organizations such as The Gates Foundation, which are openly expressing a need for greater

161

emphasis to be placed on technology in development (https://www.gatesnotes.com/Development/The-Future-of-Farming).

Information on the collection, retention, use and disposal of research data and the measures in place to ensure the confidentiality of personal data: It is unlikely that a consistent internet connection will be available in the field. Data, which will largely be in the form of hard copy questionnaires and excel (data imported to an electronic document) files will be stored under lock and key in a drawer at the principle researcher’s accommodation (for the hard copy questionnaires) and on the principle researcher’s password protected computer and backed up on a password protected external hard drive in folders that are password protected. The principle researcher will not be using his computer to connect to the internet (a smart phone that will not contain data will be used to access the internet) and therefor a need for encryption is diminished. In the event that any encryption is needed, the principle researcher will consult the Faculty IT at the University of Surrey. The principle researcher is extremely well versed when it comes to methods of cyber security and encryption. The principle researcher regularly takes measures to increase his cyber security that go beyond what the University of Surrey provides in the work environment. These electronic storage devices will also remain under lock and key in a drawer at the principle researcher’s accommodation in the field. Consent Forms will be stored under lock and key in a separate drawer or safe at the principal researcher’s accommodation. Upon arrival back at the University of Surrey, all electronic data will be transferred to the principle researcher’s password protected secure file (file only accessible by principle researcher and the principle researcher’s supervisors) in the University of Surrey’s S: drive shared projects folder and removed from the principle researcher’s field computer. The data backed up on an password protected external hard drive and hard copy questionnaires will be locked in the researcher’s office at the University of Surrey. Consent forms will be stored in a separate locking drawer in the principle researcher’s office at the University of Surrey. Excel data will remain confidential by anonymizing hard copy questionnaire data (one excel sheet per questionnaire prior to analysis) and keeping them in password protected folders on the principle researcher’s folder in the secure S: drive shared projects folder at the University of Surrey. The principle researcher is aware of University of Surrey Faculty IT and will seek advice from them if any data protection problem arises. Data from this project will not be placed on any servers outside of the S: drive at the University of Surrey and therefore the encryption offered by the S: drive should be sufficient to keep the data safe. Following the completion of this project and the departure of the principle researcher from the University of Surrey, the external hard drive containing data from this project will be turned over to the principle researcher’s supervisor, most likely Walter Wehrmeyer, or an appropriate member of CES at the University of Surrey for safe keeping over the required 10 year storage time. The principle researcher’s supervisor will delete any data from the project stored on the secure S: drive shared projects folder following the 10 year requirement for storage.

In case of auditing, a method for proving that consent forms were filled out prior to each individual questionnaire will be integrated into this project. This method will consist of numbering consent forms using a random number 1 through 160 while not repeating numbers after they have been used. Excel sheets with questionnaire data will have their file names changed to “questionnaire_data” and then a number random number between 161 and 320 while not repeating numbers after they have been assigned. A spreadsheet including consent form numbers and the excel sheets that they correspond to will be created and stored on the principle researcher’s computer in a password protected folder that will only contain this spreadsheet. This file will be moved to the principle researcher’s desktop computer upon return to the University of Surrey following the completion of this field work. An example of this spread sheet is provided here:

Consent form number Corresponding excel data file name 5 questionnaire_data_20017 questionnaire_data_311

162

12 questionnaire_data_16121 questionnaire_data_287

The principle researcher will not be collecting any data in the form of video or images of any of the participants. Hard copy questionnaires will make up the only form of media collected in this study.

Ethical issues: The only foreseeable ethical issue involves the use of verbal consent. It is extremely likely that a more information intensive consent method would be highly difficult for respondents to understand, would significantly lower participation rates, and could possibly affect the reliability of responses. Increasing the formality of the consent process often has a chilling effect on the participants and effectively disempowers them to speak honestly about the challenges that they face in deriving their livelihoods. While it might sound counterintuitive, increased formality often leads illiterate individuals feeling less protected and as if they might be held accountable for the information that they provide.

Nowhere in the reporting of results in this study will the combinations of ethnographic information including age, gender, level of education be made available to the public. This is to prevent anyone from having the ability to ascertain the identity of a participant due to the small numbers of individuals living in any particular village. Data will not be broken down by specific villages, but it may be analysed in terms of ‘distance to market’ as this variable has proved to be important in ICT4D studies in the past (Muto and Yamano, 2009). The key here will be that nowhere in analysis will any of these ethnographic factors be cross-referenced, regressed, analysed or made available in such a way as so that they could be combined or collected so as to gain an ability to identify a participant. The purpose of collecting these ethnographic factors is to gain an ability to break down the data set of vulnerability context, capital assets, livelihood strategies, and development outcomes into subsets for analysis as these ethnographic factors have proved to be significant indicators of a range of use and outcomes in the past (Scott et al. 2005). Here it should be noted, that due to the nature of this research examining patterns within the data and attempting to expose the relevance of particular patterns of vulnerability context, capital assets and livelihood strategies as they relate to outcomes could have the unintended possibility that a participant may be identified. The risk of this is extremely low due to the fact that multiple villages will be included and the extreme extent to which someone would have to work to gain enough knowledge of the research outcomes so as to identify a participant, and the possibility of malicious effects of using that knowledge. There will always be some risk of this in social science, it is the nature of work in understanding people’s lives. If someone were to ascertain a participant’s responses to the questions in this study, that information would be nearly useless in inflicting harm on a participant. The principle researcher estimates that in a worst case scenario, a bad actor with this kind of information would at a maximum gain a slight power advantage with the knowledge of someone’s vulnerability context and assets. That said, the type of information collected in this study is such that the most sensitive types of information will involve the importance (state) and change in access to financial services. It is hard to imagine how this type of information (the most sensitive) could be used in a genuinely malicious manner that might inflict harm on a participant. If a bad actor sought to target an individual using the information contained in this study, in the unlikely event that they would be able to achieve such a feat, their time would be much better spent on far easier means of targeting individuals for exploitation or harm. The goal of this study is not to expose the weakness in individuals, it is the opposite. The best research outcomes will involve positive development impacts and outcomes. If this information were used for malicious purposes, those activities would pale in comparison to the malicious and disempowering activities that may already be occurring. The point of this argument should not be that this study is the lesser of two evils, or that the existence of evil justifies possibly more evil potentially caused by this study, it is meant to provide a picture of reality where the

163

elimination of all potential threats is impossible. In other words, if a bad actor where to gain access to all the information in this study, anonymous or not, the amount of harm they could inflict on another individual with that information is negligible. If anything, this information would provide a bad actor with a road map for improving their own livelihood – a central goal of this study.

It should also be noted that in the verbal consent form, the phase “I will do everything I can to protect your privacy and information you provide.” will serve the same purpose as the phrase “I understand that all project data will be held and may be looked at for monitoring and auditing for at least 6 years and all research data for at least 10 years in accordance with University policy and that my personal data is held and processed in the strictest confidence, and in accordance with the UK Data Protection Act (1998)”. It is unlikely that any of the participants making use of the verbal consent form will be able to understand the concepts of monitoring and auditing in a manner which will be appropriate and without undue stress. Participants will be completely unaware and likely unable to access information on the UK Data Protection Act (1998) and this will only serve to increase their stress and distrust of the project to an extent where they would be unlikely to answer interview questions honestly and to the best of their ability. Participants will be unfamiliar with any ‘university policy’ and the word ‘policy’ would likely be associated with governments and in such might result in a level of discomfort and distrust in the project. It is wholly likely that participants would interpret the idea of their data being stored for such purposes that they would not understand in a negative light as opposed to the intention of the line being that this is a positive protocol with their best intensions in mind. Poor rural people often have a level of distrust in governments, especially if the government is made up of a political party that they do not support or a political party made up of a different ethnic group, and the phasing of the verbal consent form is such that it was not meant to imply or include any notion of a connection to a government or legal body. The phasing was intended to empower participants by placing their trust in the principle researcher who is not a lawyer or a politician. Even though the principle researcher is part of a research body, the University of Surrey, it will be the individual personal trust between the principle researcher and the participant that will yield the best results in a manner in which the participant feels the most comfortable while at the same time understanding the intentions of the consent and research process and protocol. Poor rural farmers typically have more confidence in personal face-to-face communication and are weary of more formal actors and processes (Batchelor et al. 2014). They will place greater confidence in the principle researcher to protect their data than an unknown body at a university mandated by an Act that they have never heard of and have little ability to understand regardless of its name. The act of taking information from poor rural people can even sometimes be interpreted as taking a part of them, or more importantly, a process by which they give a part of themselves to you. This is only accomplished in a manner in which the responses are honest if an honest and trusting relationship can be built between the principle researcher and the participant. The language in the verbal consent form is not such that a participant would interpret it as the principle researcher being subject to or controlled by higher powers which may have intentions more insidious than those of the principle researcher that have been communicated face-to-face, in the flesh, as having the best intentions in mind. In other words, the wording in the verbal consent form is such that the principle researcher is unlikely to be seen as a shill for a government body or higher institution that a participant may see as untrustworthy or so powerful so as to be irreproachable in the event that a participant may develop any qualms regarding the project or the data that they provided, in the future. If a potential participant has questions about the protection and storage of data as a result of being subject to either the written or verbal consent process, the principle researcher will provide the potential participant with the participant information sheet either in paper copy or provided orally or both. The process of explaining and communicating the protocol of data collection and storage could possibly be extensive and the principle researcher is cognizant of that possibility. During this process of communication, the principle researcher will begin with and err on the side of providing the underlying broader issues associated with data storage and protection, and check the potential participant for cognition and understanding

164

throughout the communication process. The legality and communication of specific data protection acts will be provided in the event that a potential participant signifies that they would like more information provided. This method is not meant to serve as a cover up of the specific drivers of the language provided in the consent process, it is meant to ease a potential participant into an understanding of the ethics behind the consent process prior to the legal issues behind it. This process is ordered due to the possibility, that if the principle researcher were to lead the conversation with information on the institutions and legality behind the language, it is likely that a participant would assume distrust and be unwilling or unable to comprehend the more pertinent intensions of the language in the consent form.

Consideration for working overseas: The Ghana travel advice provided by the UK government (GOV.UK) states the following:

“Around 90,000 British nationals visit Ghana every year. While most visits are trouble-free, there has been a recent increase in petty street crime; violent crime can occur at any time. Be particularly vigilant in public areas, and take care when travelling by road. See Crime

It’s mandatory for all foreign nationals resident in Ghana to register with the National Identification Authority (NIA) of Ghana and get a non-citizen Ghana card. See Local laws and customs

Localised outbreaks of civil unrest can occur at short notice, particularly in the north. If this does happen, local police may impose curfews to contain the situation. See Local travel

There is a general threat from terrorism. You should be vigilant after recent attacks in Cote d’Ivoire, Mali and Burkina Faso. Attacks could be indiscriminate, including in places visited by foreigners. See Terrorism

The Overseas Business Risk service offers information and advice for British companies operating overseas on how to manage political, economic, and business security-related risks.

Take out comprehensive travel and medical insurance before travelling.”

The primary researcher is well aware of the possibility of local civil unrest in the study area. It should be clear that cases of local civil unrest are more likely to occur in the Northern part of Ghana. While the ethic group that comprises the study group is the same as what can be found in northern Ghana, the drivers of civil unrest are typically commodity shortages driven by a lack of rainfall. The study location where this project is taking place is not subject to the same climate as can be found in the north. For this reason the possibility of civil unrest in the study area is decreased as compared to other places in Ghana. Civil unrest can also occur at political rallies during election times. Elections will not be occurring while the principle researcher is in the field and even if they were occurring, the principle researcher would avoid them.

That said, the principle researcher is well aware of the possibility of both civil unrest and terrorist attack. The principle researcher did not see any signs of fundamentalism or conflict between religions or ethnic groups in the study area during the first phase of research. If anything, religious groups cooperate in the study area much better than many places that the principle researcher has worked in the past.

The primary researcher has experience conducting research in Cameroon and Malawi for extended periods of time. The areas where field work has been taking place in Ghana have a long history of interaction with researchers and development practitioners. Areas where research will be taking place are relatively safe and the principle researcher will be accompanied by someone attached to

165

the CSIR during field work and travel. The principle researcher is relatively familiar with cultural norms in the study population having worked in similar agriculture based communities in Malawi and Cameroon. The villages in Ghana will largely be Catholic or Christian communities and no research or interviews will take place on Sundays or around the Easter Holiday in accordance with religious and cultural norms. The principle researcher is also accustomed to eating practices such as only eating with your right hand when it is custom to eat only using your hands. The principle researcher is also prepared to deal with cultural norms relating to gender and the use of technology and data collection related to this issues will be treated as sensitive despite the fact that addressing these issues is unlikely to cause increased stress for participants. In the Catholic and Christian societies in Ghana women are typically more empowered than in the Muslim societies in Cameroon where the principle researcher has worked in the past. Several studies have documented differential use and empowerment to use technology with respect to gender. For instance there are examples where African women have comparatively less access to technology due to differential incomes which serve as a limiting factor (Batchelor et al., 2014) and there have been cases in India where women need permission from their husbands in order to use mobile phones (Seshagiri et al., 2007). This study does not intend to focus on or even endeavour to gain an in-depth understanding of the causes or dynamics within gender inequalities with regard to technology use if they do indeed exist in the study population. For instance a question such as “do you need permission from someone to use your mobile phone?” will not be included in this study. Any discussion of gender inequality included in the results of this study will be based on an a sort of a holistic analysis of factors not specifically related to gender but analysed by comparing factor sets divided by gender in the same way factors will be compared via age and level of education.

In the case of other unforeseen cultural and societal norms that must be followed, the principle researcher will rely on information provided by the CSIR which will serve as a valuable resource for this type of information having had experience working in the villages that will comprise the study group. Prior to the commencement of the field work, the principle researcher will inquire with hosts at the CSIR of any pertinent information regarding protocols for abiding by appropriate norms while in the field. Currently there are no foreseeable cultural practices or norms that conflict with the principle researcher’s morals, ethics, or regular mode of operating. The study population practices a modern and western view on human rights in accordance with their Catholic and Christian religions. The study population will most likely be made up of a single ethic group: the Akan and is positioned squarely within the Ashanti Administrative Division or region of the country. For these reasons it is unlikely that cultural and societal norms will vary in the study locations. Information regarding the ethnic groups, religions, or political allegiances will not be collected in this study and information on these subjects will only be used by the principle researcher to inform appropriate behaviour in the field.

Study evaluation and statistical analysis: Data collected in this study will be quantitative and collected using a questionnaire. Data will be analysed using SPSS software. Specific statistical tests are subject to change but the current plan is to conduct a factor analysis.

Conflicts of interest: There are no foreseeable conflicts of interest regarding this project.

Other issues associated with working in the field and risk assessment information: As this study will be addressing issues related to farming, it is likely that some of the research will take place in a farming/agriculture setting. It is distinctly possible that a participant may want to provide first hand demonstrations of farming practices in the field and if this is the case the principle researcher will oblige. The principle researcher comes from a farming family and has conducted research in the past in farming setting in 8 different countries across the world. As such, the principle researcher has an in-depth knowledge of safety precautions related to farming practices. That said, there is always a certain level of risk working around tools, farm implements, and while moving around the farming

166

landscape. In the case of this project the risks of conducting research in a farming setting will be extremely low. If for any reason an accident should occur, the principle researcher or participant that is injured will be immediately transported by car to the nearest town: Ejura which will be no more than 20 Kms away. If the injury is so serious so as to need more than first aid, the principle researcher will immediately be transported by car to the city of Kumasi which is 2-3 hours away from the study area. The town of Ejura has basic modern hospital services and the city of Kumasi has extensive modern hospital services. A car for transport to medical services will always be available to the principle researcher because private automobile transport will be the only method used for traveling to study villages. Farming practices in the study area are carried out mainly with hand tools and it is extremely unlikely that the researcher will come in contact with any large farming machinery. Hand tools mainly consist of hoes, axes, and knives and there is always the very low probability of stepping on a tool that has been left out. The principle researcher will not be taking part in the farming practices themselves but the principle researcher is distinctly familiar with the proper use of all of these farming tools. The principle researcher is also entirely versed in the use and safety aspects of being around agricultural inputs such as a wide range of fertilizers, pesticides and herbicides. Again, the principle researcher will not be taking part in the farming process but it is likely that the researcher will come in contact farming inputs in the field, though not directly as direct contact will be entirely avoided. While the health risk of being in a farming area such as the study area is very low, precautions to mitigate any risks will consist of the standard practice of having appropriate attire. While in the field the principle researcher will always wear heavy work boots and sturdy trousers. Agricultural land can sometimes be difficult to navigate on foot, boots will be worn to mitigate this problem. Long sleeve shirts may be worn if there is a threat of sun exposure. It is also common practice for the principle researcher to wear a baseball style hat in order to avoid sunburn on the head.

With regard to risk of infections and diseases the principle researcher will take several precautions. The principle researcher will avoid the risk of malaria that is present in the study area by taking malaria medication (doxycycline) which the principle researcher has used effectively in the past. The principle researcher will also avoid mosquito borne infections by wearing long trousers and long sleeve shirts at night and sleeping in a mosquito net. The principle researcher has spent years living and working in areas where malaria is present and has never contracted the disease following the protocol above. A slight risk of water and food borne diseases and infections will also be present. In the case of water, either water will be boiled and filtered before use or bottled water will be consumed for hydration purposes or for the purpose of preparing food. Risk of infection or disease from food borne pathogens will also be mitigated by proper preparation of food and proper hygiene in the places where food is prepared. The principle researcher will aim to prepare his own food and exceptions to this protocol will be made when it is culturally appropriate to accept food that is offered by members in the community. It is somewhat likely that hosts in the villages will offer food to the principle researcher and it would be disrespectful and culturally insensitive to decline. The principle researcher does not have any food or other types of allergies. The principle researcher is not currently taking any medication, does not suffer from any chronic health problems and is in good physical condition. The principle researcher has experience with suffering from food and water borne infections in the past (ex. e.coli) and protocol when the possibility of being infected arises will include the immediate contact of a medical professional in the town of Ejura, and the immediate use of the Cipro (ciprofloxacin hydrochloride) antibiotic which the principle researcher has successfully used in the past to overcome infections. The risk of skin infections is also present in areas where farming takes place often do to the presence of livestock and composting practices which involve the culturing of bacteria. If a skin infection occurs, professional health treatment will be sought immediately. The principle researcher will carry with him a supply of both antibiotic tablets (Cipro) and a topical antibiotic cream. The principle researcher has suffered from skin infections such as staph in the past from working in agricultural settings. The risk of suffering from skin infections will also be decreased by the proper first aid treatment of any open wounds that may occur in the field.

167

A protocol of proper cleaning and dressing of wounds will be followed in the case that an injury occurs. The fact that the principle researcher has experienced infections of any kind in the past is not indicative of any extra susceptibility toward infections. The principle researcher has a long history of working and researching in remote tropical environments in developing countries and instances of infections have been few. The principle researcher has received all appropriate vaccinations that are required for travel by the Ghana High Commission in London and recommended for travel by the UK government. Documentation of having received these vaccinations will be carried by the researcher at all times. In case of emergency the researcher will also carry relevant emergency contacts, proof of insurance cover, and passport with identification at all times. The principle researcher will also carry enough cash currency to cover the cost of travel to a medical centre or transport or in any other case that travel is required. Overall, the risk of disease or infection for this field work will be low.

The mental and physical state of the principle researcher will be reported by the principle researcher to the PhD supervisor Walter Wehrmeyer via email on a weekly basis during the period of the field work. Other information that will be contained in these weekly email updates will include: research itinerary, in country host contact information, any changes to the fieldwork activities, in country mobile telephone (which will be carried at all times) number(s), and completed work. If an internet connection is not available, a phone call will be used for the same purpose.

Following the completion of this field work and upon return to the University of Surrey the principle researcher and the associated PhD supervisors will allow time to reflect on the adherence to the guidelines included in this ethics protocol and application. Any difficulties experienced in the field work will be expressed and factored into future ethics applications for research of this nature and around this study site. If any incidents occur during the field work these will be reported and addressed as soon as possible in the field and further managed upon arrival back at the University of Surrey through the supervisors associated with this study and through the University of Surrey’s health and well-being centres.

(Gill et al., 2008, Carney, 1999, Heeks, 2010, Sey, 2011)

References

BATCHELOR, S., SCOTT, N., MANFRE, C., LOPEZ, A. V. & EDWARDS, D. Is there a role for Mobiles to support Sustainable Agriculture in Africa. ICT4S, 2014.

CARNEY, D. 1999. Approaches to sustainable livelihoods for the rural poor, Overseas Development Institute ODI Poverty Briefing, Brighton, UK.

GILL, P., STEWART, K., TREASURE, E. & CHADWICK, B. 2008. Methods of data collection in qualitative research: interviews and focus groups. British dental journal, 204, 291-295.

HEEKS, R. 2010. Do information and communication technologies (ICTs) contribute to development? Journal of International Development, 22, 625-640.

MUTO, M. & YAMANO, T. 2009. The impact of mobile phone coverage expansion on market participation: Panel data evidence from Uganda. World development, 37(12), 1887-1896.

SCOTT, N., GARFORTH, C., JAIN, R., MASCARENHAS, O. & MCKEMEY, K. 2005. The economic impact of telecommunications on rural livelihoods and poverty reduction: a study of rural communities in India (Gujarat), Mozambique and Tanzania.

SESHAGIRI, S., AMAN, S. & JOSHI, D. Connecting the bottom of the pyramid: an exploratory case study of india's rural communication environment. Proceedings of the 16th international conference on World Wide Web, 2007. ACM, 855-862.

SEY, A. 2011. ‘We use it different, different’: Making sense of trends in mobile phone use in Ghana. New Media & Society, 13, 375-390.

168

WORLD BANK, WORLD DEVELOPMENT INDICATORS. 2016. GNI per capita, Atlas method (current US$). Retrieved from http://databank.worldbank.org/data/reports.aspx?source=2&country=GHA&series=&period=

Participant Information Sheet v.3 10/01/2017

Livelihood Development and the Impacts of Mobile Phone on Yam Production in Ghana

Introduction

I would like to invite you to take part in a research project. Before you decide you need to understand why the research is being done and what it will involve for you. Please take the time to read the following information carefully and ask questions about anything you do not understand. Talk to others about the study if you wish.

What is the purpose of the study?

The purpose of this study is to ascertain the role that mobile phones play in the development of livelihoods derived from agriculture. I aim to gain an understanding of how people use mobile phones as part of agriculture and livelihood activities. I also aim to examine the outcomes and impacts of mobile phone use the different types of capital that make up an individual’s livelihood.

Why have I been invited to take part in the study?

You have been invited to take part in this study because you derive your livelihood, at least in part, from agriculture practices.

About 160 participants in Ghana will take part in this study.

Do I have to take part?

No, you do not have to participate. There will be no adverse consequences if you decide not to participate or withdraw at a later stage. You can withdraw your participation at any time. You can request for your data to be withdrawn until publication of the data without giving a reason and without prejudice.

If you withdraw from the study this will mean the following for your participation and data:

All identifiable data would be withdrawn from the study. Data which is not identifiable to the research team may be retained because we cannot trace this information back to you. No further data would be collected or any other research procedures would be carried out on or in relation to you.

What will my involvement require?

169

If you agree to take part, we will then ask you to sign a consent form or provide verbal consent. If you do decide to take part you will be given this information sheet to keep and a copy of your signed consent form if you sign the written consent form. This research will last 8 weeks but your involvement would only be approximately 1 0.5 hours. During this time, you will be asked to participate in an interview complete a questionnaire.

What will I have to do?

The questionnaire will contain questions mainly focused on the livelihood development challenges that you face and the strategies that you use to overcome those challenges. Specific attention will also be paid to your mobile phone use. If you are inclined and able to complete the questionnaire on your own you will be afforded with that opportunity. If you are unable to read and write, a translator (member of CRI) will read the questions and possible responses to you and then record the answers that you provide. In the event that you require a translator but are unwilling to use one, you will be excluded from the study.

What will happen to data that I provide?

Research data are stored securely for at least 10 years following their last access and project data (related to the administration of the project, e.g. your consent form) for at least 6 years in line with the University of Surrey policies.

Personal data will be handled in accordance with the UK Data Protection Act (1998).

What if there is a problem?

Any complaint or concern about any aspect of the way you have been dealt with during the course of the study will be addressed; please contact Tyler Jay Reynolds, Principal Investigator on [t.j.reynolds@surrey.ac.uk or UK Phone: 4401483689559]. You may also contact my supervisors: Stephen Morse [s.morse@surrey.ac.uk] or Walter Wehrmeyer [w.wehrmeyer@surrey.ac.uk].

The University of Surrey holds insurance policies which apply to this study. If you experience harm or injury as a result of taking part in this study, you will be eligible to claim compensation. This does not affect your legal rights to seek compensation.

If you are harmed due to someone's negligence, then you may have grounds for legal action. Regardless of this, if you wish to complain, or have any concerns about any aspect of the way you have been treated during the course of this study then you should follow the instructions given above.

Will my taking part in the study be kept confidential?

Yes. Your details will be held in complete confidence and we will follow ethical and legal practice in relation to all study procedures. Personal data [name, audio recordings questionnaire data] will be handled in accordance with the UK Data Protection Act 1998 so that unauthorised individuals will not have access to them.

Your personal data will be accessed, processed and securely destroyed by Tyler Jay Reynolds. In order to check that this research is carried out in line with the law and good research practice, monitoring and auditing can be carried out by independent authorised individuals. Data collected during the study, may be looked at by authorised individuals from the University of Surrey or from regulatory authorities, where it is relevant to your taking part in this research. All will have a duty of confidentiality to you and we will do our best to meet this duty. We will anonymise any documents or records that are sent from the University of Surrey, so that you cannot be identified from them.

170

The data you provide will be anonymised and your personal data will be stored securely and separately from those anonymised data. You will not be identified in any reports/publications resulting from this research and those reading them will not know who has contributed to it. With your permission we would like to use anonymous verbatim quotation in reports.

In certain exceptional circumstances where you or others may be at significant risk of harm, the researcher may need to report this to an appropriate authority, in accordance with the UK Data Protection Act 1998. This would usually be discussed with you first.

Examples of those exceptional circumstances when confidential information may have to be disclosed are:

- The researcher believes you are at serious risk of harm, either from yourself or others

- The researcher suspects a child may be at risk of harm- You pose a serious risk of harm to, or threaten or abuse others- As a statutory requirement e.g. reporting certain infectious diseases- Under a court order requiring the University to divulge information- We are passed information relating to an act of terrorism

How will I be compensated for taking part in this study?

If you agree to take part in this study you will be compensated via a monetary payment of approximately an amount that is based on typical wages for the same time spent working in this area. In circumstances where Tyler Jay Reynolds determines that an interview is proceeding in a sufficiently unethical manner, Tyler Jay Reynolds may decide to end an interview early and in these circumstances, a decreased monetary payment will be provided as compensation. In the event that you decided to end an interview early, for whatever reason, Tyler Jay Reynolds may decide to provide a decreased monetary payment as compensation.

Full contact details of researcher

Tyler Jay Reynolds

Phone (Ghana): TBD

Phone (UK): 4401483689559

Email: t.j.reynolds@surrey.ac.uk

Who is organising and funding the research?

This research is organised by the University of Surrey and funded by Tyler Jay Reynolds. The funder is has no conflict of interest.

Ethical Review

This research has been looked at by an independent group of people, called an Ethics Committee, to protect your interests. This study has been reviewed by and received a favourable ethical opinion from the University of Surrey Ethics Committee.

Thank you for taking the time to read this Information Sheet.

171

Written Consent Form v.3 10/01/2017

Livelihood Development and the Impacts of Mobile Phone on Yam Production in Ghana

*Please note that consent can be given verbally as some participants will be unable to read and write

Please initial each boxConsent form number …...

I have read and understood the Information Sheet provided (version 3, date 10/01/2016) and I have been given a full explanation by the investigators of the nature, purpose, location and likely duration of the study, and of what I will be expected to do.

I agree for my anonymised data to be used for this study / future research that will have received all relevant legal, professional and ethical approvals.

I understand that all project data will be held and may be looked at for monitoring and auditing for at least 6 years and all research data for at least 10 years in accordance with University policy and that my personal data is held and processed in the strictest confidence, and in accordance with the UK Data Protection Act (1998).

I understand that I am free to withdraw from the study and have my data withdrawn at any time until publication without needing to justify my decision, without prejudice and without my legal rights being affected.

I acknowledge that in consideration for completing the study I shall receive compensation for my time and inconvenience.

I confirm that I freely consent to participating in this study and I while participating, and do so to the best of my ability. I have been given adequate time to consider my participation.

Name of participant (BLOCK CAPITALS) ……………………………………………….

Signed ……………………………………………….

172

Date

Name of researcher/person taking consent ……………………………………………….(BLOCK CAPITALS) Signed ……………………………………………….

Date ……………………………………………….

Verbal Consent Form v.3 10/01/2017

Livelihood Development and the Impacts of Mobile Phone on Yam Production in Ghana

Consent form number……….

As you know, I am a researcher from the University of Surrey in the United Kingdom. I am conducting a study on the use of mobile phones in agriculture production, and I would like to ask you some questions about that. I have here a questionnaire that I would like you to complete with the help of this translator. If you choose to complete the questionnaire, your answers will be anonymous. This means that I will not use your name in this study. Do you understand?

Understanding was indicated (initialed by researcher)

If at any time during our talk you feel uncomfortable and you do not want to answer a question or stop the questionnaire completion process entirely, you may do so. If at any time you wish to have the information you provide withdrawn from the study, please tell me and I will destroy the data. Does that make sense?

Understanding was indicated (initialed by researcher)

I will do everything I can to protect your privacy and information you provide. If you decide to participate in this study you will be compensated for your time and inconvenience. Do you have any questions so far?

Any questions were answered and understanding was indicated (initialed by researcher)

If you would like more information on consent, I can provide you with a more detailed Information Sheet (version 3, date 10/01/2017). Now I would like to ask you if you agree to participate in this study. Do you agree to participate and complete this questionnaire? Take as much time as you need to decide.

Verbal Consent was provided (initialed by researcher)

173

Name of researcher/person taking consent ……..............................................(BLOCK CAPITALS)

Signed ....................................................

Date .………………………………………………..

Questionaire Schedule for Livelihood Development and the Impacts of Mobile Phone on Yam Production in Ghana

Principle Researcher: Tyler Jay Reynolds

The questionnaire schedule between the dates that this study intends to be carried out (March 25, 2017 to May 25, 2016) will largely be opportunistic and will depend heavily on the availability of employees and fixers from the CSIR. These individuals will provide transport to the field sites (villages), aid in relationship building with gate keepers (village/community leaders), and members of the community. This means that at any time that these individuals are available for assistance, the opportunity will be seized and questionnaires will be completed. It is likely that no more than three six questionnaires will be completed on any given day due to time restraints associated with conducting any particular questionnaire.

With regard to question and questionnaire design, a face to face approach will be applied. The question topics will be derived by applying the Sustainable Livelihoods Approach (SLA) (Carney et al. 1999) and Heeks (2010) ICT4D model to the context of Ghanaian farmers with some level of access to ICT. The basic structure of the framework for generating questions and hopefully arriving at specific and holistic conclusions is provided in Figure 1.

174

The first sets of questions in the questionnaire will be used to gain a general understanding of the vulnerability context within which individuals live. In the case of this research the use of the word vulnerability refers directly to the aspect within the SLA known as the vulnerability context that does not imply or even directly state that individuals participating in this study are assumed to be vulnerable according to definitions found in the psychological, sociological, or social work sense. No participants in this project could be defined as vulnerable according to any British Common Law definition. The vulnerability context as used in this study will be based on the definitions provided by DFID’s Sustainable Livelihoods Framework in that the vulnerability context in which people live plays an important role in their ability to derive livelihoods (DIFID, 1999). DFID explicitly states that some trends that occur in the vulnerability context that an individual exists in are not always negative (DFID, 1999). This makes up one of the key strengths of the SLA, that it enables researchers to gain a comprehensive and complex understanding of how individuals derive their livelihoods (Heeks and Molla, 2009). The reason behind this being, that on outward and simple appearance and individual may appear to face significant development challenges, especially compared to developed societies. Upon closer examination and research, individuals may effectively derive livelihoods using seemingly limited assets transformed through complex and innovative livelihood strategies. This is not meant to downplay any real vulnerabilities and livelihood development challenges, perceived by any actors or otherwise, but lowering the bar for development with soft expectations and pandering attitudes towards the world’s poor is something this study and many others will avoid. This study aims to address this issue in research to an even greater extent than previous studies that make use of the SLA in that this study contains an increased focus on the vulnerability context in which individuals exist. This is one of the innovative and experimental aspects of this study and depending on the results that this study produces may serve as both a critique of previous studies and building block on which future studies making use of the SLA may do better to more realistically assess individual’s vulnerability context and hence livelihood development. More specifically this study will seek to ascertain data on an individual’s vulnerability context by inquiring about an individual’s perception of the importance that an individual places on any particular livelihood development challenge. These livelihood development challenges are often defined in the literature, typically for good reason because they serve as significant challenges in previously studied populations, but it is entirely possible that the reliance of researchers on these livelihood development challenges as subjects of study and more importantly the likelihood that researches assume that because a particular livelihood development challenge is found in the literature than any particular individual operating in a developing country would be subject to significant negative impacts of said challenge, serves as a significant failure on the part of the researcher to gain a more realistic picture of both reality and the perceptions of reality held by research subjects and researchers. This study will grapple with this problem of individual’s perceptions of livelihood development challenges and how that data relates to capital assets possessed, livelihood strategies employed, and livelihood development outcomes. The vulnerability context that the participants in this study is likely to be is made up of challenges that are addressed in the examples of questionnaire questions provided below. That said, it is extremely unlikely that any individual living in the study area would be considered extremely vulnerable. While the people that will serve as participants in this study would be considered to be poor by western standards, they are not so vulnerable so as to face any inordinate stress or harm as a result of participating in this study. This is exemplified by the fact while one of the main goals of CSIR research in the study area is to improve food security, and yet no individuals in the study area are threatened by starvation. If the individuals in the study area were vulnerable to the extent that participating in a study with as little impact as this one would cause them undue harm or stress, international development agencies would be focusing on more extreme challenges to livelihood development in the area such as directly addressing starvation through direct food aid. In areas where significant development challenges exist, it is often times the case that study participants find these challenges common place and either downplay their significance as a result of not being able to imagine a challenge being reduced to an extent which

175

might be achieved in the developed world and/or that is it easy to express themselves with regard to the challenge as the stigma surrounding the admission and expression of a development challenge is comparatively less in an area where it is common place than in an area such as the developing world where taboos around the admission of struggle as a result of inequality and unequal power structures exist. That said, the principle researcher is keenly aware of the livelihood development challenges that people in the developing world face. A line of questioning regarding a particular potential development challenge that a participant shows any level of uncomfortableness in addressing, the principle researcher will stop the questionnaire completion process. The point of this research is not necessarily to highlight the failures of individual’s actions to effectively address their development challenges. While that information would contribute to the world’s collection of knowledge on what does and not work in terms of livelihood development, the main goal of this study is to discover what does work in terms of livelihood development in whatever vulnerability context an individual exists in. Finally, as stated in the design of this project, the principle researcher is seeking participants which have access to mobile phone technology. This access is in itself an indicator that an individual is comparatively less vulnerable than individuals without access.

The questionnaire will also include general questions about an individual’s access and ability to use ICTs. This access to ICTs maybe subject to policies and institutions such as the phone service providers and as such will serve as a means by which to address that aspect of the SLA. After these general areas have been covered, capital assets, livelihood strategies, and impacts on livelihoods (outcomes) will be addressed on an individual specific development challenge basis. That said, even within each line of sight addressing each development challenge, questions regarding the vulnerability context in which that challenge exists will be included. The basic logic behind this is expressed in Figure 2.

Figure 2.

The Basic structure of the questionnaire is as follows:

a. Vulnerability Context

1. Ethnographic factors

2. Physical/financial capital

b. Aspects of livelihood development

1. How significant is livelihood development challenge X?

2. How do you address livelihood development challenge X?

3. How has the significance of livelihood development challenge X changed over the past few years?

176

4. What is/are the drivers of the challenge of livelihood development challenge X?

c. Mobile phone demographcis

The logic behind the selection of this set of questions regarding each livelihood development challenge is expressed through an analysis of the different combinations of responses which provide a relatively rich picture of the dynamics associated with identifying and prioritizing challenges and methods by which they can be addressed. For instance, if a livelihood development challenge is not seen as significant this might be because it was significant and it is no longer or the capacity to address it is such that it is not seen as a challenge, or both. It is also possible that the challenge is not actually a challenge regardless of capacity to address it or how its significance has changed. One possible downside to the structure of this questionnaire is the somewhat single dimensionality of the lines of questioning being divided among specific livelihood development challenges. This mode does not easily lend itself to discovering and addressing cross cutting effects among challenges or the effect of feedback within an analysis of challenges. Mitigating this weakness will involve careful reflection of possible interactions and dynamics within the system as a whole and more advanced statistical testing techniques.

It should also be noted that the classification of various livelihood development challenges according to one of the five types of capital assets should not serve as a hard and fast rule for classification as some livelihood development challenges could be described as affecting more than one type of capital asset. For instance, the challenge of gaining knowledge of new farming techniques could have implications for both human and financial capital as knowledge is traditionally described as human capital and the application of that knowledge with regard to farming practices would result in higher production and amount sold, hence financial capital. This issue plays out in the selection of aspects of development that the questionnaire includes. It may appear that the questionnaire focuses too heavily on aspects related to farming but this is intentional. The participants in the study group are mainly farmers and as such, if they are unable to farm, they will be unable to develop aspects of their livelihoods outside of farming such as improving their health and maintaining their social circles. The aspects of development included in the questionnaire were also selected on the basis that they came up as important issues or challenges during the first qualitative stage of this study. In some ways, this second phase of qualitative data collection is intended to build upon and support the findings from the first phase of this study by measuring and analysing them using statistical analysis.

It should also be noted that the questions concerning aspects of development included in the questionnaire are heuristic in nature and this study design is somewhat experimental. While this design is rooted in the SLA, the specific nature of logical inferences that may be drawn from the results of the sets of questions each related to a specific aspect of development is largely new. It is our hope that through a yet to be determined statistical analysis we may be able to test the reliability of this proposed method/design while at the same time drawing conclusions about the state of development and its connections to ICT described above and derived in the first phase of this project. We are well aware of the possibility that this kind of post-hoc analysis is somewhat undesirable and depending on the results, might call into question our findings. The limited timeline of this project has determined this situation. In the event that these problems come to fruition, data produced from questions outside of the sections on aspects of development will still be of use in terms of correlations that can be drawn (or fail to be drawn) between the vulnerability context and participants’ use of ICT.References

DFID. 1999. Sustainable Livelihood Guidance Sheet Section 2. DFID, London

177

http://www.eldis.org/vfile/upload/1/document/0901/section2.pdfHEEKS, R., GALPERIN, H., MARISCAL, J., GILLWALD, A., STORK, C., ADEYINKA, T., AJIBOYE, J. O.,

EMMANUEL, A. O. & WOJUADE, J. I. 2009. Beyond subscriptions: actual ownership, use and non-use of mobiles in developing countries. ICTs for Development Blog.

Questionnaire Draft (v2 10/01/2017)

Demographics and basic indicators

Village:_______________________

Age:______________

Sex:______________

Level of Education:______________

What type of roof do you have on your house?-all thatch-partial thatch/partial zinc-all zinc

What transport options do you have access to (you or someone in your immediate family owns)?-no transport (walking)-bicycle-motorcycle-motorized tricycle (moto king)-tractor

How large is your farm (acres)?: ______________

Has the size of your farm changed over the past five years?-increasing-no change-decreasing

Production Issues

How often do you change the land you are farming on?

178

-Change land every five years+-Change land every two years-Change land every year-Never change land

How easy or difficult is it for you to organize paid labour?-Very easy-Somewhat easy-Neither easy nor difficult-Somewhat difficult-Very difficult

What is the method by which you organize paid labour? (rank top 2)-Walk up and ask labourers face to face in the community-Call labourers on your mobile phone directly-Call a friend or relative on your mobile phone and have them organize the labour for you-Travel to labourers outside of the community and ask them face to face-Wait for labourers to call you on your mobile phone-Wait for labourers to approach you face to face in the community

To what extent has organizing paid labour become easier or more difficult over the past five years?-It has become much easier-It has become somewhat easier-It has not changed-It has become somewhat more difficult-It has become much more difficult

What is the most difficult aspect of organizing paid labour? (rank top 2)-labourers are not in community when I need them-Labourers are working for other farmers when I need them-labour is too expensive-Labour does poor quality work-I pay labour and they fail to do/complete the work

How easy or difficult is it for you to obtain high quality chemicals (weedicide, pesticide, fertilizer)?-Very easy-Somewhat easy-Neither easy nor difficult-Somewhat difficult-Very difficult

If you need chemicals, what do you do? (rank top 2)-Talk face to face to a friend or relative who can organize the purchase of chemicals for you-Call a friend or relative who can organize the purchase of chemicals for you-Call an Agric extension officer-Talk to an Agric extension officer face to face the next time they come to the community-Travel to town and talk to a chemical dealer

If you notice an insect, weed, disease infestation on your field, what do you do? (rank top 2)-Talk face to face to a friend or family member in the community-Wait for the Agric extension officer to come to the community and talk to them face to face

179

-Do not talk to anyone, you know how to deal with infestations-Call a friend or family member on the phone and discuss it with them-Call an Agric extension officer on the phone-Call a chemical dealer on the phone-Travel to a chemical dealer and talk to them face to face

To what extent has obtaining chemicals become easier or more difficult over the past five years?-It has become much easier-It has become somewhat easier-It has not changed-It has become somewhat more difficult-It has become much more difficult

What is the most difficult aspect of obtaining chemicals/fertilizer?-You don’t know what type to buy-Counterfeit/weak chemicals-You don’t have enough money to buy chemicals-Chemical dealers do not have enough supply of chemicals- other

Are you registered for the fertilizer subsidy program?-Yes-No

How easy or difficult is it for you to organize a tractor for ploughing (maize/cowpea)?-Very easy-Somewhat easy-Neither easy nor difficult-Somewhat difficult-Very difficult

How do you organizing a tractor for ploughing?-Travel to tractor owner and speak face to face in another community or in town-Travel to tractor owner and speak face to face in this community-Call tractor owner on your mobile phone-Call a friend or relative on your mobile phone and ask them to organize a tractor for you

To what extent has organizing a tractor for ploughing become easier or more difficult over the past two years?

-It has become much easier-It has become somewhat easier-It has not changed-It has become somewhat more difficult-It has become much more difficult

What is the one most difficult aspect of organizing a tractor for ploughing?-Tractor operators are unavailable when you need them-Tractor operators do not come to your field when they say they will-Ploughing is too expensive-Tractor operators do a poor job of ploughing-It is difficult/expensive to travel to tractor operators to talk to them

180

How easy or difficult is it for you to control weeds and/or do weeding on your farm?-Very easy-Somewhat easy-Neither easy nor difficult-Somewhat difficult-Very difficult

What is your typical method for controlling weeds on your farm?-Hiring paid migrant labour to spray weeds or manually weed-Asking a friend/relative/work group member to help spray weeds-Asking a friend/relative/work group member to help manually weed-I spray weeds on my own-I manually weed on my own

To what extent has weeding become easier or more difficult over the past five years?-It has become much easier-It has become somewhat easier-It has not changed-It has become somewhat more difficult-It has become much more difficult

What is the most difficult aspect of weeding?-The number of weeds are increasing-The chemicals that I spray on weeds are becoming weak/ineffective-I cannot afford or find labour to help with weeding-There are new and different weeds that have come into my farm-I cannot afford weedicide-I do not have enough strength to weed

Agricultural product (yam) sale issues

Where do you most often sell your yam?-farm gate-market

What is your ability to store yam to sell when the price is better?-I can store my yam for long periods while I wait for the price to go up-I sometimes store yam waiting for the price to go up-I never wait for the price to go up, I sell immediately after I harvest

How easy or difficult is it to sell yam?-Very easy-Somewhat easy-Neither easy nor difficult-Somewhat difficult-Very difficult

How to do you obtain yam selling prices/market information? (rank top three)-I don’t obtain selling prices before selling

181

-I travel to the market and look at prices/talk to people in the market-I call my middle-woman-I wait for my middle-woman to call me-I call people in the market and ask them about pricing information-I talk to colleague farmers in the community about the prices they have just sold at

To what extent has the selling yam become easier or more difficult over the past five years?-It has become much easier-It has become somewhat easier-It has not changed-It has become somewhat more difficult-It has become much more difficult

What is the most difficult aspect of selling yam? -Obtaining reliable price information-Being manipulated by middle-women-Transporting yam to market-Yam prices change from the time I decide to sell to the time I arrive in the market-The length of time between arriving in the market and when my yam is finally sold-Understanding the protocols of selling in the market-Other

Other issues outside of producing and selling agriculture products

Are you a part of the national health insurance program?-Yes-No

How easy or difficult is it for you to access medical treatment?-Very easy-Somewhat easy-Neither easy nor difficult-Somewhat difficult-Very difficult

If you have a health problem what do you normally do?-Travel to the nearest health clinic/hospital-Buy medicine from the mobile medicine dealers-Use traditional medicines-Pray-Call friends and family for help-Do nothing

To what extent has access to medical treatment become easier or more difficult over the past five years?

-It has become much easier-It has become somewhat easier-It has not changed-It has become somewhat more difficult-It has become much more difficult

182

What is the most difficult aspect of accessing medical treatment? (Rank top 2)-Traveling to a hospital or clinic is too difficult/expensive-Hospitals/clinics do not provide enough medication-Hospital/clinic treatment is too expensive-Hospitals/clinics make you wait a long time before you are served-You don’t have enough information about health problems and treatments-Other

How important is it for you to have many friends?-Very important-Somewhat important-Neither important or unimportant -Not very important -Not at all important

State of friends in the community?-I have many friends in the community (I am happy with them)-I have a few friends in the community-I have almost no friends in the community

State of friends outside the community?-I have many friends in the community (I am happy with them)-I have a few friends in the community-I have almost no friends in the community

Change in number of friends in the community over the past five years?-I have more friends in the community than I did five years ago-I have the same amount of friends in the community than I did five years ago-I have fewer friends in the community than I did five years ago

Change in the number of friends outside of the community over the past five years?-I have more friends outside of the community than I did five years ago-I have the same amount of friends outside of the community than I did five years ago-I have fewer friends outside of the community than I did five years ago

Method of communicating with friends? (Rank top 2)-I mostly talk to my friends face to face in this community-I mostly travel to my friends outside this community and talk to them face to face-I mostly call my friends in this community on the phone-I mostly call my friends outside this community on the phone

Mobile phone demographics

What is the main purpose of the mobile phone?-Greeting friends and family-Calling friends and family about financial issues-Calling friends and family about health issues-Calling middle-women to organize sale of crops-Calling friends and family to learn how to improve the household

183

-Calling friends and family to improve agriculture practices

Functions that you can use? (Tick all that apply)-None-Voice call-SMS text-Mobile money-torch-storing numbers-calculator-voice recorder-camera-games-Other

Frequency of phone use (separate making and receiving calls)?-make at least one call a day-make at least 2 calls a week-make less than 2 calls a week-more than 2 calls per day

How many years have you been using a mobile phone?

What is the main cause of your phone to spoil?-battery fails-dropping it on the ground or in water (rough nature of work)-water damage from rain-interference from someone else pressing it (using it)-stolen

Mobile phone coverage?-anywhere in the community-changes from day to day-only in particular places in the community-overall poor coverage everywhere I the community

How do you use mobile money?-Don’t use it-A friend/relative uses it for me-I use it

Main person that helps with phone use when assistance is needed-Child-Friend-Adult family member-Phone kiosk/network officer worker-Other (middle-woman?, hired labour, etc.)

184

Consideration Person at Risk

Scale of Risk

Existing Protocols Additional

Mechanisms

Vulnerability due to economic state

Participant and / or Researcher

Low / Medium / High

What is currently in place to mitigate this risk? Is there anything in addition to the existing protocols that can be done to mitigate this risk?

Vulnerability due to economic state

Participant Low Participants will be told in advance the length of the questionnaire process. Any cost of economic activities forgone for the case of the questionnaire process will be offset with a monetary payment approximately covering the cost of work forgone.

When possible the questionnaire process will be conducted outside of working hours

Vulnerability due to gender

Participant Low The only gender specific question included in the questionnaire is ‘What is your gender?’

Stress from not being able to read and write

Participant Low Participants will be assured that their personal identities will be protected and never made public. They will be given ample time to consider the information provided to them during the consent process. They will be informed that if they have questions regarding

185

details of the project not included in the verbal consent script, that those details will be provided for them verbally according to the project information sheet. If the participant visibly shows signs of discomfort at any time, they will be reminded that they can choose to end their participation in the study and this will not have negative consequences for them.

Sharing information in the community and outside influences on the participant

Participant Low Participants will be reassured that the information produced by completing the questionnaire will not be shared with anyone in the community. Ideally completing the questionnaire will be one on one but in many cases in the field other individuals will sit around the participant. If this is the case, the researcher will ask the participant if they would prefer to complete the questionnaire in private. The researcher will also ask if the participant accepts the fact that other individuals are present during the time spent completing the questionnaire. If during the time that a participant is completing the questionnaire an individual not completing the questionnaire speaks or in any way affects the ability for the participant to speak freely and honestly, the researcher will request that that the interrupting individual remain quiet and not to interrupt the questionnaire completion process. At this time the researcher will also ask the participant if they would like to continue the questionnaire process. If the researcher suspects that the participant is not responding to question in an honest manner because the participant is being affected by another individual that is present, the researcher may decide to end the questionnaire process.

Sun exposure Researcher Low The researcher will use sun screen cream and a hat if sun exposure is expected to be an issue.

Operating in an agricultural setting

Researcher Low Boots and trousers will be worn at all times in the field. In general, sturdy clothing will be worn. The principle researcher will be wary of his surroundings where tools, machinery, and farm inputs may be present.

Traveling by automobile

Researcher Low Only traveling in automobiles that would be deemed safe by UK standards will be used for travel. The proper inflation of tired will be checked as roads in Ghana are likely to be of low quality and contain hazards such as holes. Cars will need to contain fire extinguishers and a spare tire with jack. Night driving will be avoided when possible. The principle researcher

186

will not be driving himself while in Ghana. Cars in which the principle researcher will be travelling will be driven by an employee of CSIR or a person trusted and recommended by the CSIR. Seat belts will be worn at all time while cars are in operation and Ghanaian traffic laws will be followed.

Threat of disease or infection

Researcher Low Appropriate medication will be carried by the researcher at all times as will emergency numbers and a basic first aid kit. More information can be found in the Project Protocol sheet under the heading ‘Other issues associated with working in the field and risk assessment information’

Existing medical conditions or disability

Researcher Low The principle researcher has no pre-existing medical conditions or disabilities. The principle researcher is not on any medication and is in good physical condition.

Deep-vein thrombosis

Researcher Low Flight time to the study location is around 5 hours and broken up between two days. The research has no history of circulation problems and is experienced in stretching and moving about regularly on plane flights.

Personal attack/kidnap

Researcher Low The principle researcher will be accompanied by an employee of the CSIR during the time spent in the field. The researcher will not travel alone at night in the town of Ejura. The principle researcher has some training in mixed martial arts self-defence such as Brazilian jiu-jitsu. The FCO website has been checked

Social/Political unrest

Researcher Low The FCO website has been checked and the area that the field work will be conducted is not prone to social/political unrest.

Potential for data loss

Researcher Low Once an internet connection is made available in the field the principle researcher will upload data from the principle researcher’s personal computer to the principle researcher’s personal folder on the University of Surrey’s S: drive shared projects folder. In the event that data is able to be uploaded to the S: drive, the principle researcher will contact the principle researcher’s supervisors and confirm that the data has been uploaded. Once this data has been confirmed as uploaded to the University of Surrey’s S: drive by the principle researcher’s supervisors, the principle researcher will delete the data from the principle researcher’s personal computer.

The FCO does not advise against

187

traveling to this location.

Protocol Submission Proforma: Research Insurance

The University holds Zurich Municipal Public Liability insurance which provides cover in the event that the University of Surrey is held legally liable for negligent harm caused to a research subject.

Ethics Number

Title and brief description in lay terms Livelihood Development and the Impacts of Mobile Phone on Yam Production

Department FEPS

Location of Research Ejura, Ghana

Nature of Research Qualitative research on mobile phones and agriculture

Expected Start Date March 25, 2017

Expected End Date May 25, 2017

Lead Researcher Tyler Jay Reynolds

Externally Funded? No

Name of Sponsor Self-funded

Projected/Cumulative Number of Subjects 160

Research involving the following require special consideration and the insurer’s prior approval must be sought:

Any pregnant research subjects? No

Any research subjects under 5 years of age? No

Is this an overseas trial? Yes

More than 5,000 subjects? No

188

* Assign to one of the following categories:-

P- Pharmaceutical PS - Pharmaceutical, externally funded,

NP - Non-pharmaceutical NPS - Non-pharmaceutical, externally funded,

Q - Questionnaire/interview/observation only

APPENDIX 3: QUESTIONAIRE

Questionnaire - ICT and SL: Yam farmers in Ghana

General Survey Data (to be filled out prior to the start of the interview)

1. Entry number ___________________

2. Date of the Interview _________________________

3. Name of the Interviewer (translator) _______________________

4. Name of the supervisor ______________________________

5. Name of the village______________________________

Survey questions

1. Age:______________

2. Sex:______________

3. Level of Education:__________________________________________

4. What transport options do you have access to (you or someone in your immediate family owns – tick all that apply)?

189

no transport (walking)

bicycle

motorcycle

motorized tricycle (moto king)

tractor

5. How large is your farm (acres)?:

______________

6. Has the size of your farm changed over the past five years?

significantly decreased some decrease no change some increase significantly increased

7. What is the cost of your monthly living expenses? ________________

8. How have your living expenses changed over the past five years?

significantly decreased some decrease no change some increase significantly increased

9. How often do you send or receive money from people living outside of the community?

never

a few times a year

a few times a month

190

1

a few times a week

more than once a day

10. How easy or difficult is it for you to send or receive money from people living outside of the community today?

very easy somewhat easy neither easy nor difficult somewhat difficult very difficult

11. How significant are the following reasons for sending and receiving money to people living outside of the community? (score each from 1-5, 5 being very significant)

School fees

Petty items (including small food stuffs)

Wedding expenses

Agricultural inputs (seeds, chemicals, tools)

Receiving payment for agricultural products

Healthcare (for you or a loved one)

Funeral expenses

Other

(if other, please state _____________________________________________)

12. To what extent are the following methods by which you send or receive money from people living outside of the community five years ago significant in your life? (score 1-5, 5 being highly significant and 1 being little/no significance)

Travel and transfer face to face

Send and receive through friends and relatives traveling to and from

Mobile money transfer

191

Bank transfer

Hired car/courier

Other

(if other, please state _____________________________________________)

13. In general, has your ability to send and receive money from people living outside of the community changed over the past five years?

much easier somewhat easier no change somewhat more difficult much more difficult

14. To what extent are the following methods by which you send or receive money from people living outside of the community today significant in your life? (score 1-5, 5 being highly significant and 1 being little/no significance)

Travel to and transfer face to face

Through friends and relatives traveling to and from

Mobile money transfer

Bank transfer

Hired car/courier

Other

(if other, please state _____________________________________________)

15. How easy or difficult is it for you to organize paid labour? (Today)

very easy somewhat easy neither easy nor difficult somewhat difficult very difficult

16. To what extent are the following methods by which you organized paid labour five years ago significant in your life? (score 1-5, 5 being highly significant and 1 being little/no significance)

192

Walk up and ask labourers face to face in the community

Call labourers on your mobile phone directly

Call a friend or relative on your mobile phone to organize labour for you

Travel to labourers outside of the community to ask them face to face

Wait for labourers to call you on your mobile phone

Other

(if other, please state _____________________________________________)

17. Has organizing paid labour become easier or more difficult over the past five years?

much easier somewhat easier no change somewhat more difficult much more difficult

18. To what extent are the following methods by which you organize paid labour today significant in your life? (score 1-5, 5 being highly significant and 1 being little/no significance)

Walk up and ask labourers face to face in the community

Call labourers on your mobile phone directly

Call a friend or relative on your mobile phone and have them organize labourers for you

Travel to labourers outside of the community and ask them face to face

Wait for labourers to call you on your mobile phone

Other

(if other, please state _____________________________________________)

19. What are the most difficult aspects of organizing paid labour today? (score 1-5, 5 being very significant challenge)

193

Labourers are not in community when I need them

Labourers are working for other farmers when I need them

Labour is too expensive

Labourers do poor quality work

Labourers fail to do/complete the work

Other

(if other, please state _____________________________________________)

20. Where do you sell most of your yam?

farm gate

market

21. Can you store yam to sell when the price is better?

I always store my yam for long periods while I wait for the price to go up

I sometimes store yam waiting for the price to go up

I never wait for the price to go up, I sell immediately after I harvest

22. How easy or difficult is it to sell yam today?

very easy somewhat easy neither easy nor difficult somewhat difficult very difficult

23. To what extent are the following methods by which you obtain yam selling prices/market information five years ago significant in your life? (score 1-5, 5 being highly significant and 1 being little/no significance)

194

I don’t obtain selling prices before selling

I travel to the market and look at prices/talk to people in the market

I call my middlewoman

I wait for my middlewoman to call me

I call people in the market and ask them about pricing information

I talk to colleague farmers in the community about the prices they have just sold at

Other

(if other, please state _____________________________________________)

24. Has the selling yam become easier or more difficult over the past five years?

much easier somewhat easier no change somewhat more difficult much more difficult

25. To what extent are the following methods by which you obtain yam selling prices/market information today significant in your life? (score 1-5, 5 being highly significant and 1 being little/no significance)

I don’t obtain selling prices before selling

I travel to the market and look at prices/talk to people in the market

I call my middlewoman

I wait for my middlewoman to call me

I call people in the market and ask them about pricing information

I talk to colleague farmers in the community about the prices they have just sold at

Other

(if other, please state _____________________________________________)

26. What are the most difficult aspects of selling yam? (score 1-5, 5 being very significant challenge)

195

Obtaining reliable price information

Being manipulated by middlewomen

Transporting yam to market

Yam prices change from the time I decide to sell to the time I arrive in the market

The length of time between arriving in the market and when my yam is finally sold

Understanding the protocols of selling in the market

Other

(if other, please state _____________________________________________)27. What is the most important purpose of the mobile phone? (score each from 1-5, 5 being very important)

Greeting friends and family

Calling friends and family about financial issues

Calling friends and family about health issues

Calling middle-women to organize sale of crops

Calling friends and family to learn how to improve the household

Calling friends and family to improve agriculture practices

Other

(if other, please state _____________________________________________)

28. Frequency of phone use?

at least one call a day

at least 2 calls a week

196

less than 2 calls a week

more than 2 calls per day

29. How many years have you been using a mobile phone?

______________________

197

APPENDIX 4: Conference Poster Presentation: “Impacts of ICT on Agriculture Based Livelihoods in Ghana”

198

APPENDIX 5: Conference Poster Presentation: “Mobile phones and sustainable agriculture based livelihoods: A case study of smallholder farmers in the Ashanti Region, Ghana”

199