Welcome

1. Welcome

2. Comparative Analysis of smallholder farmers in Kenya, Zambia and Tanzania Comparing results from FinScope Tanzania 2017 and AFA benchmark studies in Kenya and Zambia ANDREW KARLYN Director for Strategy and Learning

4. Comparative Analysis of smallholder farmers in Kenya, Zambia and Tanzania Executive Summary (1) Agricultural livelihoods • SHFs in all three countries are poor, and though they often engage in multiple income-generating activities the majority of their income comes from agriculture. • Productivity varies greatly between the study countries, with Zambia in particular lagging behind. • While this productivity gap is likely driven by many factors (infrastructure/access to markets etc) there may be opportunities here for digital financial inclusion and programmes like Agrifin to reduce this gap.   Gender • There is a clear gender gap in terms of income across all three countries, and interestingly this is most pronounced in Kenya where farmers have the highest comparative income • Although the gender gap is closing in mobile money in Kenya, it remains significant in Tanzania and Zambia. • Women continue to lag in terms of their access to formal financial services (with the exception of insurance). Financial Inclusion Levels • Financial exclusion is by far the highest in Zambia. Exclusion in Tanzania is much lower though SHFs tend to rely more on informal services than their counterparts in Kenya who mainly use mobile money. • Kenya leads the three countries in terms of uptake of nearly all formal services (mobile/bank accounts/insurance).  • Mobile money uptake varies in in each country, underlining their different progress in terms of market evolution. • Some clear opportunities emerge for savings and credit products considering the proportion of farmers who save at home/informally. Specifically in Tanzania there is an opportunity to further deepen mobile money usage.

5. Comparative Analysis of smallholder farmers in Kenya, Zambia and Tanzania Executive Summary (2) Digital preparedness and Mobile Money Readiness index • Mobile phone penetration is well progressed in all three countries, but in Zambia, this has not led to an uptake of mobile money. Key barriers may lay elsewhere (penetration of agents; awareness etc). • Applying a mobile money readiness index shows that readiness is low in Zambia and high in Kenya, which is reflected in uptake. • In Zambia and Tanzania there seems an opportunity among SHFs that show high mobile money readiness but do not use mobile money yet. Information, advisory services and training • There is not a consolidated channel through which SHFs access agricultural advisory services and many do not access any services at all. • Of the services used, government officers seem to have the greatest penetration followed by peers/cooperatives. • In terms of how information is accessed, this is predominantly through demonstrations and field days.  Resilience and threats to livelihoods • The resilience of farmers and the greatest threats to their livelihoods are not as easily comparable across the three countries. • We do however find that many farmers remain vulnerable and experience shocks that have a significant impact on their household income with many lacking appropriate coping strategies. • Many of these shocks are agriculture related and issues such as pests/weather and loss of harvest are important across different markets. • Insurance does not provide resilience for agricultural shocks: despite a growing uptake of insurance products, this is very rarely ag-related. Mostly health and life insurance is taken-up.

6. Comparative Analysis of smallholder farmers in Kenya, Zambia and Tanzania Contents 1 Research framework and methodology 2 Segmentation of by smallholder farmers and comparability 3 Research findings 3.1: Profile of Smallholder Farmers 3.2: Access and usage of financial services 3.3: Digital Services 3.4: Information, advisory services and training 3.5: Resilience 4 Discussion points 5 Annex

7. 1 Research framework and methodology This research framework is designed to generate deeper insights on smallholder farmers in Tanzania, Kenya and Zambia. For Tanzania, data from FinScope’s 2017 survey is used. For Kenya and Zambia, data from AgriFinAccelerate’s benchmark studies are used. These were administered in 2017 by Research Solutions Africa (RSA). Please see a summary of each survey below:

8. 1 Research framework and methodology Defining Financial Inclusion across countries * See the Annex of this presentation for an extensive list of questions used to define each category.

9. 1 Research framework and methodology Our analysis is structured in the following sections and research questions: RQ1 What is the profile of an average (or median) farmer in Tanzania, Kenya and Zambia? 1. Profiles of SHFs What is the uptake and usage of financial services by SHFs in Tanzania, Kenya and Zambia? RQ2 2. Access and usage of financial services How do SHFs save and borrow in Tanzania, Kenya and Zambia? RQ3 3. Digital Services RQ4 How prepared are SHFs to use digital services in Tanzania, Kenya and Zambia? RQ5 What is the uptake and usage of digital financial services in Tanzania, Kenya and Zambia? RQ6 How do farmers currently access information, advisory services or trainings on agriculture or finance? 4. Information, advisory services and training RQ7 What do farmers perceive as the largest threat to their livelihoods? What are the coping strategies? 5. Resilience

10. 1 Research framework and methodology Analytical methods applied Segmentation and Indices • Segmentation approaches identify clusters of households with similar characteristics. • One way to segment the sample is to develop an index to rank households from high to low index values. In this analysis we develop a mobile money readiness index for RQ4. Summary Statistics Analysis • Summary statistics are used to make comparisons in nearly all research questions. We provide statistical measures such as: • Mean, median and mode; • Maximum, and minimum; and • Frequency analysis and cross-tabulation analysis. Graphical Analysis • Graphical analysis visualises differences in data groups. All research questions make use of this method. Multiple Regression Analysis • Regression analysis is used to identify any relationship between a dependant variable and one or more independent variables.

11. Comparative Analysis of smallholder farmers in Kenya, Zambia and Tanzania Contents 1 Research framework and methodology 2 Segmentation of by smallholder farmers and comparability 3 Research findings 3.1: Profile of Smallholder Farmers 3.2: Access and usage of financial services 3.3: Digital Services 3.4: Information, advisory services and training 3.5: Resilience 4 Discussion points 5 Annex

12. Segmentation of smallholder farmers and comparability 2 Defining Smallholder farmers across Tanzania, Kenya and Zambia Note that there are methodological differences with respect to how our data sources define SHFs in Tanzania as opposed to Kenya and Zambia. These are highlighted below: SHF screening questions for Zambia and Kenya RSA’s benchmark studies in Zambia and Kenya only interviewed households in rural areasthat passed the following set of screening question: • Screening Q1: Does your household practice any form of agriculture? • Screening Q2: Has agriculture been among the major income sources for your HH in the past 12 months? • Screening Q3: Will your HH continue with Agriculture? • Screening Q4: Do you consider yourself as a farmer? • Screening Q5: The HH cultivates less than 10 acres of land. SHFs segmentation FinScope Tanzania* FinScope Tanzania is a general population survey. SHFs are therefore a sub-set of the overall sample that was collected. To define SHFs the following criteria was applied to the overall sample: • Condition 1: Income from agriculture (either crops or livestock) contributes a significant amount to overall income (>40%) • Condition 2: HH sells the produce that it grows [not just trading]. • Condition 3: The HH cultivates less than 10 acres of land. • Condition 4: Only SHFs in rural areas are considered. * See the Annex of this presentation for more detail on the FinScope Tanzania segmentation.

13. Segmentation of smallholder farmers and comparability 2 Comparability of key variables: It is important to also note differences in the way variables are defined in FinScope vis-à-vis RSA’s benchmark studies. Below we list differences in the definition of key variables.

14. Segmentation of smallholder farmers and comparability 2 Comparability of key variables (continued)

15. Segmentation of smallholder farmers and comparability 2 Comparability of key variables (continued)

16. Comparative Analysis of smallholder farmers in Kenya, Zambia and Tanzania Contents 1 Research framework and methodology 2 Segmentation of by smallholder farmers and comparability 3 Research findings 3.1: Profile of Smallholder Farmers 3.2: Access and usage of financial services 3.3: Digital Services 3.4: Information, advisory services and training 3.5: Resilience 4 Discussion points 5 Annex

17. Comparative Analysis of smallholder farmers in Kenya, Zambia and Tanzania • Profile of Smallholder Farmers 3.1 Research Question 1: What is the profile of an average (or median) farmer in Tanzania, Kenya and Zambia?

18. Research Question 1: What is the profile of an average (or median) farmer in Tanzania, Kenya and Zambia?. • Profile of Smallholder Farmers 3.1 Demographic Profile: • Kenyan SHFs are slightly older on average, followed by SHFs in Zambia and Tanzania. • In Tanzania 18% of SHFs are between 16 and 24. Average age Tanzania:40 • Kenya:44 Zambia:41 Median age Tanzania:38 • Kenya::42 Zambia:38

19. Profile of Smallholder Farmers Research Question 1: What is the profile of an average (or median) farmer in Tanzania, Kenya and Zambia? 3.1 Economic Profile (1): Absolute income is highest in Kenya - both in average and median terms. Zambian and Tanzanian SHFs have similar median incomes. It is important to note here that both for FinScope and for the benchmark studies extreme outliers are observed. Excluding these from the analysis could change the overall picture. Agriculture makes up the largest proportion of income (80%) for SHFs in Tanzania. Disclaimer: Note that comparing SHF’s income levels does not account for national differences in income levels and is therefore constraint in describing the relative wealth of SHFs.

20. Research Question 1: What is the profile of an average farmer in Tanzania, Kenya and Zambia? • Profile of Smallholder Farmers 3.1 Socio-economic profile: • Kenya’s SHFs have the largest income gender gap both in total and relative terms. • Zambia’s SHFs education income gap is most pronounced: the median SHF with tertiary education earns $121, while the median SHF with no formal education or primary education only earns $9 or $12 a month. *Income levels might be skewed due to a small sample size in the tertiary category.

21. Profile of Smallholder Farmers 3.1 Research Question 1: What is the profile of an average (or median) farmer in Tanzania, Kenya and Zambia? Economic Profile (2) – income distribution of SHFs: • Income distributions in all three countries point towards a large share of SHFs living below $1/day. • Zambia observes the most unequal distribution: while it has overall fewer SHFs living below $2 /day than Tanzania, 48% of Zambian farmers still live below $0.5/day. 84% < $2 / day 69% < $2 / day 75% < $2 / day Disclaimer: Note that comparing SHF’s average income does not account for national differences in income levels. We can however interpret the income distribution for each country.

22. Profile of Smallholder Farmers 3.1 Research Question 1: What is the profile of an average (or median) farmer in Tanzania, Kenya and Zambia? Land size in acres: Kenya’s SHFs are the most productive group, cultivating the smallest areas of land but generating the highest income from agriculture per month. Zambia’s SHFs cultivate the most land but earn the least, making them the least productive group in terms of income in USD per acre of land cultivated. $9 / month $15 / month $20 / month Median monthly income from agriculture

23. Profile of Smallholder Farmers Research Question 1: What is the profile of an average (or median) farmer in Tanzania, Kenya and Zambia? 3.1 Engagement in Value Chains: • Kenya: Many SHFs engage in livestock and dairy. Cultivation of cash and food crops is lower than in other countries. • Tanzania: Most SHFs are involved in food crops and complement this with either cash crops or livelihoods. • Zambia: Few SHFs engage in livestock. The emphasis is on food and cash crops.

24. Comparative Analysis of smallholder farmers in Kenya, Zambia and Tanzania • Uptake and Usage of Financial Services 3.2 Research Question 2: What is the uptake and usage of financial services by SHFs in Tanzania, Kenya and Zambia?

25. Research Question 2: What is the uptake and usage of financial services by SHFs in Tanzania, Kenya and Zambia? • Uptake and Usage of Financial Services 3.2 Nature of Financial Uptake: • Kenya’s SHFs have the highest financial inclusion levels with only 9% who are excluded. • Of those who are financially included in Kenya, 33% are banked as opposed to 16% in Zambia and 7% in Tanzania. • Mobile money penetration is 56% in Kenya, while it is lowest in Zambia (16%) and 41% in Tanzania. Base: All SHFs

26. Research Question 2: What is the uptake and usage of financial services by SHFs in Tanzania, Kenya and Zambia? • Uptake and Usage of Financial Services 3.2 Nature of Financial Uptake: • SHFs in Tanzania make most use of informal services and are catching-up with insurance and mobile money. • 91% of Kenyan SHFs are formally included. Kenya has the highest take up across most types of financial services, except informal services and MFIs. Base: All SHFs

27. Research Question 2: What is the uptake and usage of financial services by SHFs in Tanzania, Kenya and Zambia? • Uptake and Usage of Financial Services 3.2 Gender gap: For Kenyan farmers the gender gap seems to be closing with respect to mobile money but is pronounced for banking services for all countries. Mobile money uptake in Zambia and Tanzania has a larger gender bias in favour of men. Note that formal includes commercial banks, post bank, mobile money, MFIs, SACCOs. Base: All SHFs

28. Research Question 2: What is the uptake and usage of financial services by SHFs in Tanzania, Kenya and Zambia? • Uptake and Usage of Financial Services 3.2 Financial uptake based on education: • Uptake of financial services is positively correlated with education for all three countries. • Informal services increase significantly with education in Tanzania. • For Kenyan farmers the gap is closing for mobile money while there is a larger correlation between higher education and uptake for Zambia and Tanzania. Base: All SHFs

29. Research Question 2: What is the uptake and usage of financial services by SHFs in Tanzania, Kenya and Zambia? • Uptake and Usage of Financial Services 3.2 Mobile money services uptake by age: • In Kenya, mobile money uptake is similar for SHFs between 16 and 54 but decreases in groups older than this. • In Tanzania, mobile money uptake is bell-shaped with low values for the youngest and oldest SHF segments, peaking for SHFs between 45 and 54. • In Zambia, SHFs of all age groups have similar mobile money uptake. Base: All SHFs

30. Research Question 2: What is the uptake and usage of financial services by SHFs in Tanzania, Kenya and Zambia? • Uptake and Usage of Financial Services 3.2 Median Monthly Total Income by uptake of financial services: Users of banks and mobile money have significantly higher incomes than non-users across all three countries. Users of informal services have similar incomes to non-users (Note informal income in Kenya is driven by an outlier).

31. Comparative Analysis of smallholder farmers in Kenya, Zambia and Tanzania • Uptake and Usage of Financial Services 3.2 Research Question 3: How do SHFs save and borrow in Tanzania, Kenya and Zambia?

32. Uptake and Usage of Financial Services 3.2 Research Question 3: How do SHFs save and borrow in Tanzania, Kenya and Zambia? Savings and borrowing behaviour across countries: • Savings behaviour is similar across all three countries, with Kenya leading in terms of the highest proportion of savers. • Borrowing behaviour varies strongly. However this could be largely driven by methodological differences between FinScope and the Benchmark studies. Base: All SHFs

33. Uptake and Usage of Financial Services 3.2 Research Question 3: How do SHFs save and borrow in Tanzania, Kenya and Zambia? Saving behaviour – Tanzania: 44% of SHFs in Tanzania save money. Most of these SHFs still keep cash at home (52%) as opposed to using mobile money (18%) despite the relatively high uptake of mobile money across the sample. Base: SHFs who save

34. 3.2 • Uptake and Usage of Financial Services Research Question 3: How do SHFs save and borrow in Tanzania, Kenya and Zambia? Borrowing behaviour - Tanzania: 45% of SHFs borrow but 73% of these do so informally by asking family or friends. 20% use savings groups and only 2% borrow from a mobile money service provider. Base: SHFs who borrow

35. Uptake and Usage of Financial Services 3.2 Research Question 3: How do SHFs save and borrow in Tanzania, Kenya and Zambia? Saving behaviour: Kenyan SHFs who save (58%) often do so by using mobile money (41% use KCB M-Pesa, 15% use MPesa and 12% use M-shwari). Savings groups are also used frequently (33%) and most SHFs have moved away from savings at home (‘savings in a hidden place’ = 1.2%). Base: SHFs who save

36. Uptake and Usage of Financial Services 3.2 Research Question 3: How do SHFs save and borrow in Tanzania, Kenya and Zambia? Borrowing behaviour: SHFs in Kenya borrow are less likely to borrow than those in Tanzania (23% vs 45%). Those that do have mostly moved away from borrowing from friends or neighbour (5%) and are using informal groups, such as Chama (30%) or SACCOs (14%), or mobile money services such as Mshwari (20%). Base: SHFs who borrow

37. Uptake and Usage of Financial Services 3.2 Research Question 3: How do SHFs save and borrow in Tanzania, Kenya and Zambia? Saving behaviour: The most common way of saving for SHFs in Zambia is savings money in a hidden place, e.g. at home (40% of SHFs who save). Mobile money uptake for saving purposes is relatively high (12%), considering the overall low uptake of 26% across SHFs in Zambia. Base: SHFs who save

38. Uptake and Usage of Financial Services 3.2 Research Question 3: How do SHFs save and borrow in Tanzania, Kenya and Zambia? Borrowing behaviour: Only a small proportion of SHFs in Zambia claim that they borrow money. Of those who do, 33% do not indicate a source and 32% borrow from family and friends. Base: SHFs who borrow

39. Comparative Analysis of smallholder farmers in Kenya, Zambia and Tanzania • Digital Services 3.3 Research Question 4: How prepared are SHFs to use digital services in Tanzania, Kenya and Zambia?

40. Digital Services 3.3 Research Question 4: How prepared are SHFs to use digital services in Tanzania, Kenya and Zambia? Access to digital technologies (1): Kenyan SHF’s are leading in terms of access to digital technologies. However Zambia’s SHFs are more likely to own a mobile phone or a smart phone than Tanzanian SHFs. This offers an interesting opportunity for mobile money uptake. Base: All SHFs

41. Digital Services 3.3 Research Question 4: How prepared are SHFs to use digital services in Tanzania, Kenya and Zambia? Access to digital technologies – gender gap: • While the gender gap in mobile phone access is closing in Kenya, it is still more pronounced in Zambia and Tanzania. Base: All SHFs

42. Digital Services 3.3 Research Question 4: How prepared are SHFs to use digital services in Tanzania, Kenya and Zambia? Access to digital technologies – age: • In Kenya, all age groups apart from those above 65 have similarly high mobile phone uptake. • In Tanzania and Zambia, the distribution is more bell-shaped with younger and older SHFs being less likely to own a mobile phone. Base: All SHFs

43. Digital Services 3.3 Research Question 4: How prepared are SHFs to use digital services in Tanzania, Kenya and Zambia? Mobile Money Readiness Index We develop a mobile money readiness index that scores each SHF HH between [1,10]* along the following criteria: Next we apply Ward’s Linkage clustering method to segment the sample into two groups: *See the Annex of this presentation for more detail on the scoring methodology. We find that despite higher mobile phone ownership likelihood, Zambia’s SHFs score lowest in terms of the mobile money readiness index.

44. Digital Services 3.3 Research Question 4: How prepared are SHFs to use digital services in Tanzania, Kenya and Zambia? Mobile money readiness index by gender age: • In Kenya, we find that despite the closing gender gap in mobile phone ownership, the percentage of females with high mobile money readiness scores is still 10% below males • The gender gap is most pronounced in Tanzania, where only 39% of females score highly compared to 65% males. Base: All SHFs

45. Comparative Analysis of smallholder farmers in Kenya, Zambia and Tanzania • Digital Services 3.3 Research Question 5: What is the uptake and usage of digital financial services in Tanzania, Kenya and Zambia?

46. Digital Services 3.3 Research Question 5: What is the uptake and usage of digital financial services in Tanzania, Kenya and Zambia? Mobile money uptake: • Mobile money uptake varies in in each country, underlining their different progress in terms of market evolution. • Although the gender gap is closing in mobile money in Kenya, it remains significant in Tanzania and Zambia.

47. Digital Services 3.3 Research Question 5: What is the uptake and usage of digital financial services in Tanzania, Kenya and Zambia? Percentage of SHFs that take-up mobile money by mobile money readiness index (high/low)* Group 1: SHFs with high index values that do not access mobile money What are the barriers that prevent SHFs from using mobile money where technology access should not be an issue? Group 2 SHFs with low index values that do not access mobile money Which interventions could support SHFs to move from a ‘low readiness index’ to accessing and using mobile money? *Note that proximity to financial access point is not included in the index hence this could be a potential barriers

48. Digital Services 3.3 Research Question 5: What is the uptake and usage of digital financial services in Tanzania, Kenya and Zambia? Percentage of SHFs that take-up mobile money by mobile money readiness index (high/low)* • 2,347,790 Population size of SHFs with high index values Population size of SHFs with high values that do not take-up MM • 1,408,674 • 18,481,711 Population size of SHFs with high index values Population size of SHFs with high values that do not take-up MM • 1,108,903 • 4,214,895 Population size of SHFs with high index values Population size of SHFs with high values that do not take-up MM • 1,393,377

49. Digital Services 3.3 Research Question 5: What is the uptake and usage of digital financial services in Tanzania, Kenya and Zambia? Group 1: SHFs with high index values that do not access mobile money What are the barriers that prevent SHFs from using mobile money where technology access should not be an issue? Segmentation by gender • Females with high index values take-up MM less frequently than males. • This reflects earlier findings re MM uptake (see slide 46) Segmentation by age • The average age of those that take-up MM is not significantly different from those that don’t in the high index value segment.

50. Digital Services 3.3 Research Question 5: What is the uptake and usage of digital financial services in Tanzania, Kenya and Zambia? Group 1: SHFs with high index values that do not access mobile money What are the barriers that prevent SHFs from using mobile money where technology access should not be an issue? Segmentation by median income • SHFs that take-up mobile money have higher median incomes. This holds for all three countries. Segmentation by average income • The difference between SHFs that take-up MM and those that don’t is even more pronounced when considering average income.