THE COST OF the Gender  Gap IN AGRICULTURAL PRODUCTIVITY: FIVE AFRICAN COUNTRIES

Poverty-Environment Initiative

Empowered lives. Resilient nations.

THE COST OF the Gender Gap IN AGRICULTURAL PRODUCTIVITY: FIVE AFRICAN COUNTRIES This report is a joint product of the UN Women Regional Office for Eastern and Southern Africa (ESARO) and the United Nations Development Programme–United Nations Environment Programme Poverty- Environment Initiative (UNDP–UN Environment PEI) Africa.

© 2018 UN Women, UNDP and UN Environment.

Author: Yana van der Meulen Rodgers, Professor and Undergraduate Director of Women’s and Gender Studies in the School of Arts and Sciences, Rutgers University, New Brunswick, New Jersey

The views expressed in this publication are those of the author and do not necessarily reflect the views of UN Women or UNDP–UN Environment PEI.

The designation of geographical entities in this report, and the presentation of the material herein, does not imply the expression of any opinion whatsoever on the part of the publisher or the participating organizations concerning the legal status of any country, territory, or area, or of its authorities, or concerning the delimitation of its frontiers or boundaries. While reasonable efforts have been made to ensure that the contents of this publication are factually correct and properly referenced, UN Women, UNDP and UN Environment do not accept responsibility for the accuracy or completeness of the contents and shall not be liable for any loss or damage that may be occasioned directly or indirectly through the use of, or reliance on, the contents of this publication, including its translation into languages other than English. This publication may be reproduced in whole or in part and in any form for educational or nonprofit purposes without special permission from the copyright holder provided acknowledgment of the source is made. No use of this publication may be made for resale or for any other commercial purpose whatsoever without prior permission in writing from UN Women, UNDP and UN Environment.

Photos: cover, clockwise from left: © FAO/Petterik Wiggers (Uganda), © FAO/Giulio Napolitano (), A. Haroon Akram-Lodhi (Tanzania), © FAO/Amos Gumulira (Malawi); p. 1: © FAO/Giulio Napolitano (Ethiopia); p. 9: © UN Women (Rwanda); p. 15: © FAO/Petterik Wiggers (Uganda); p. 16: © UN Women (Tanzania); p. 19: © FAO (Uganda)

Editing and design: Nita Congress Contents

Acknowledgements iv

Executive summary v

Introduction 1

Methodology and scope 3 The gender gap in agricultural productivity in Africa 5 Factors driving the gender gap 9

Key factors for the five focus countries 10 Factor findings of comparable studies 12 Other significant factors of production 13 Gains from closing the gap 16

Closing the gap: remedies and recommendations 19

Increase women’s access to labour and time-saving equipment and services 20 Facilitate women’s shift to high-value crops 23 Improve women’s access to non-labour agricultural inputs 24 Strengthen women’s land rights 25 Other interventions for closing the gender gap 27 Notes 31

References 32

iii Acknowledgements

his regional report and its accom- We are also grateful for the country-level guid- panying policy brief are a joint ance and leadership provided by the country production of UN Women and representatives and technical staff of the five the United Nations Development countries in which the studies were carried TProgramme–United Nations Environment out: Ethiopia, Malawi, Rwanda, Tanzania and Programme Poverty-Environment Initiative Uganda. We also acknowledge the support (UNDP–UN Environment PEI) Africa. provided by Jack Onyisi Abebe and Valentine Waroga of UN Women. The regional report is based on the first quantitative study in 2015 on the Cost of the UN Women and UNDP–UN Environment PEI Gender Gap in Agricultural Productivity in especially wish to thank Yana V. Rodgers, the Malawi, Tanzania, and Uganda, and in 2017 consultant who developed the regional report in Ethiopia and Rwanda; and follow-up quali- and policy brief; and Haroon Akram-Lodhi, tative studies in 2017 in Malawi, Tanzania and the consultant who conducted the cost of the Uganda. gender gap studies in the five countries. We would also like to acknowledge Nita Congress UN Women and UNDP–UN Environment PEI for editing and designing the publications. would like to acknowledge Izeduwa Derex- Briggs of UN Women and David Smith of UNDP–UN Environment PEI for their lead- ership and guidance in finalizing this study. Appreciation is also due Fatmata L. Sesay of UN Women and Jacinta Okwaro and Moa Westman of UNDP–UN Environment PEI for their technical leadership in the development of the regional report and the policy brief.

iv Executive summary

omen represent over half Within this context, the UN Women Eastern of the agricultural labour and Southern Africa Regional Office, the force in Sub-Saharan Africa. United Nations Development Programme– Their substantive contri- United Nations Environment Programme Wbution to agriculture and their vital role in Poverty-Environment Initiative Africa, and the ensuring family food security have been World Bank commenced a collaboration to widely documented. However, gender- create evidence of the links between women’s based inequalities in access to and control economic empowerment, sustainable agricul- of productive and financial resources inhibit tural production and economic growth. The agricultural productivity and undermine resil- evidence clearly shows that gender gaps in ience and sustainability efforts. access to inputs have high economic costs and can affect the extent to which farmers A growing body of evidence points to a adopt new resource management practices salient feature of the agricultural sector across and technological innovations. Sub-Saharan Africa: lower rates of agricultural productivity for female cultivators than for This report reviews a number of studies male cultivators. Substantial gender gaps in to help policymakers diagnose and better productivity have arisen not because women understand the nature of these gaps so that are less efficient farmers, but because women agricultural interventions are more effective, experience inequitable access to land and to scalable and practical. These methods help agricultural inputs. Such unbalanced distribu- quantify the structure of constraints which tion frequently stems from and is bolstered prevent women from having full access to agri- by deeply entrenched sociocultural norms cultural resources. Dismantling this structure and traditional expectations of gender roles. of constraints is crucial for reducing women’s This structure of constraints is multifaceted. unpaid work burdens and raising the returns For example, women are more income- and to their labour. The means towards these ends time-constrained than men, which has reper- include policy reforms in five key areas: cussions on their ability to access credit, land and appropriate levels of inputs. These Increasing women’s access to labour and constraints thus lead to sizeable gender gaps time-saving equipment and services in the adoption of high-value crops and in the use of agricultural implements, male family Facilitating women’s shift to high-value labour, pesticides and fertilizer, among other crops elements. Improving women’s access to non-labour agricultural inputs

v Strengthening women’s land rights more equitable access to credit, reforms to combat gender-based violence and closing Pursuing other interventions to close the the gender gap in the application of climate- gender gap smart agricultural practices.

Examples of such reforms include the devel- The bottom line of most of these policies opment of infrastructure which reduces is that more effective targeting can work to women’s time in unpaid work, initiatives that reallocate constrained resources in socially promote gender-aware agricultural extension optimal ways. services and improved access to informa- tion, interventions to provide women with

vi The Cost of the Gender Gap in Agricultural Productivity: Five African Countries Introduction griculture continues to be an Sub-Saharan Africa: lower rates of agricultural important engine for growth in productivity for female cultivators relative to Africa’s local and regional econ- male. Substantial gender gaps in produc- omies; the sector employs a tivity have arisen not because women are Asubstantial proportion of the population and less efficient farmers, but because women is the basis for food security. experience inequitable access to land and to agricultural inputs. Such unbalanced distribu- Women represent over half of the agricul- tion frequently stems from and is bolstered tural labour force in Sub-Saharan Africa. Their by deeply entrenched sociocultural norms substantive contribution to agriculture and and traditional expectations of gender roles. their vital role in ensuring family food secu- This structure of constraints is multifaceted. rity have been widely documented. However, For example, women are more income- and gender-based inequalities in access to and time-constrained than men, which has reper- control of productive and financial resources cussions on their ability to access credit, inhibit agricultural productivity and under- land and appropriate levels of inputs. These mine resilience and sustainability efforts. constraints thus lead to sizeable gender gaps in the adoption of high-value crops and in the Empowering women in agriculture and use of agricultural implements, male family reducing gender disparities would be labour, pesticides and fertilizer, among other consistent not only with the Sustainable Devel- elements. opment Goals (SDGs) of the 2030 Agenda for Sustainable Development, but also with a host Within this context, the UN Women Eastern of regional and international conventions and and Southern Africa Regional Office, the frameworks related to women in agriculture. United Nations Development Programme– Notably, the African Union’s declaration of United Nations Environment Programme 2015 as the Year of Women’s Empowerment Poverty-Environment Initiative Africa, and the and Development towards Africa’s Agenda World Bank commenced a collaboration to 2063 and the 2003 adoption of the “Protocol create evidence of the links between women’s to the African Charter on Human and Peoples’ economic empowerment, sustainable agricul- Rights on the Rights of Women in Africa” tural production and economic growth. This (commonly known as the Maputo Protocol) effort resulted in an initial report in 2015 and both speak to gender in the context of Afri- two sets of studies by UN Women and PEI in ca’s long-term development 2017 on the cost of the gender gap in agri- Women represent agenda.1 Further, a number of cultural productivity. This report provides an over half of the International Labour Organi- overview of the key trends identified in these zation (ILO) conventions focus studies and how they compare with patterns agricultural on promoting equality in the documented in other published studies. The workplace and call for govern- UN Women–PEI studies focus on five Eastern labour force in ments to ensure decent work and Southern African countries: Ethiopia, Sub-Saharan Africa. conditions for women.2 These Malawi, Rwanda, Tanzania and Uganda. initiatives will help facilitate broader policy and institutional frameworks The present report also summarizes lessons supporting women farmers and correcting on best practices in eradicating inequities in gender imbalances in the agricultural sector. the agricultural sector emerging from the UN Women and PEI studies. These reforms and A growing body of evidence points to a best practices can be grouped into five areas, salient feature of the agricultural sector across each of which addresses an important driver

2 The Cost of the Gender Gap in Agricultural Productivity: Five African Countries of the gender productivity gap for most if not Methodology and scope all of the focus countries: Beginning in 2015, UN Women, PEI and the Increase women’s access to labour and World Bank undertook a set of studies on time-saving equipment and services how gender differences in access to agri- cultural inputs affect agricultural production Facilitate women’s shift to high-value and economic growth. The studies ultimately crops covered five countries and utilized both quan- titative and qualitative methods. The first, Improve women’s access to non-labour “The Cost of the Gender agricultural inputs Gap in Agricultural Produc- Policy reforms can tivity in Malawi, Tanzania, Strengthen women’s land rights and Uganda,” was a quan- potentially make titative study conducted in women’s labour be Pursue other interventions which close 2015, with similar quantitative the gender gap, such as initiatives that studies undertaken in Ethi- valued on an equal improve women’s access to credit, provide opia and Rwanda in 2017. basis with that of skills training, and help end their isolation These five countries were from social and market networks facili- selected for several reasons, men. tating the flow of productivity-enhancing including their potential for information valuable policy lessons for the region as well as the availability of high-quality, nationally Implementing these reforms and best prac- representative household survey data with tices will require targeted interventions to information on household characteristics and address the deeper structural constraints on agricultural inputs and output. facing women in the agricultural sector. For instance, rectifying the debilitating effect of Another criterion in country selection was poor or absent infrastructure on women’s the relative importance of the agricultural unpaid work will have large positive effects sectors. As shown in Table 1, the agricultural on their productivity. Policymakers also sector has accounted for at least one-quarter need to consider creating an enabling envi- of gross domestic product (GDP) in the ronment for women in agriculture through five focus countries over the past six years. interventions such as protecting them from More dramatically, agricultural sector growth gender-based violence. Interventions need accounts for roughly 60 per cent of total to be designed which directly target women GDP growth on average in Ethiopia, Malawi farmers, encouraging the use of climate- and Rwanda; and for about 40 per cent of smart agricultural practices — particularly by total GDP growth in Tanzania and Uganda. enforcing gender-aware agricultural exten- Hence, agriculture remains a major driver of sion services — and correcting gender biases economic growth in these countries. in statutory and customary laws. These policy reforms, whether applied in isolation or in The quantitative studies provide estimates of tandem, have the potential to transform the monetary value of the gender gap in agri- women’s lives in developing countries such cultural productivity for each country and then that their labour is valued on an equal basis calculate the costs associated with gender with that of men. gaps in access to individual agricultural inputs. These gender gaps in agricultural productivity are calculated as the unconditional values of

Introduction 3 TABLE 1 GDP growth and agricultural sector growth in the five countries

Country 2010 2011 2012 2013 2014 2015 2016 Average Agriculture, value added (% of GDP)

Ethiopia 44.7 44.7 48.0 44.9 41.9 39.2 37.2 43.0

Malawi 31.9 31.2 30.6 30.8 30.8 29.7 28.1 30.4

Rwanda 30.6 30.4 31.0 30.8 30.9 30.2 31.5 30.8

Tanzania 32.0 31.3 33.2 33.3 31.4 31.5 31.5 32.0

Uganda 28.3 27.0 28.0 27.4 27.1 26.1 25.8 27.1 Agriculture, value added (annual % growth)

Ethiopia 5.1 9.0 4.9 7.1 5.4 6.4 2.3 5.8

Malawi 6.8 4.3 -0.1 6.6 5.9 -2.0 -2.3 2.7

Rwanda 5.0 4.7 6.5 3.3 6.6 5.0 3.9 5.0

Tanzania 2.7 3.5 3.2 3.2 3.4 2.3 2.1 2.9

Uganda 2.9 3.1 0.6 1.9 2.7 2.3 2.8 2.3 GDP growth (annual %)

Ethiopia 12.6 11.2 8.6 10.6 10.3 10.4 7.6 10.2

Malawi 6.9 4.9 1.9 5.2 5.7 2.8 2.5 4.3

Rwanda 7.3 7.8 8.8 4.7 7.6 8.9 5.9 7.3

Tanzania 6.4 7.9 5.1 7.3 7.0 7.0 7.0 6.8

Uganda 5.6 9.4 3.8 3.6 5.1 5.2 4.7 5.3

SOURCE: World Bank, World Development Indicators database, 2018, http://databank.worldbank.org/data/home.

the male-female differential in the value of agri- cultivated by female and male farmers and cultural output per hectare of cultivated land. combining that proportion with the gender The reports for Ethiopia, Malawi, Tanzania and productivity gap, and then calculating the size Uganda also report conditional gender gaps of the gap in relation to agricultural GDP and which take into account gender differences in overall GDP. plot size and regional variations in agro-climatic conditions across each country (the data for This report also summarizes results on how Rwanda did not have information on plot size). much various factors of production contribute The conditional gaps tend to be larger than to the overall gender productivity gap. These the unconditional gaps because, on average, estimates are based on a decomposition women work on smaller plots than men and procedure commonly used in the labour are subject to more variable climate condi- economics literature (e.g. Autor, 2015) to tions. The costs associated with these gender explain the drivers of the gender wage gap. gaps are calculated by first converting the agri- The procedure essentially decomposes the cultural output produced by female and male male-female agricultural productivity gap into farmers at the plot level into monetary values a portion explained by gender differences in by multiplying plot output by crop-specific access to various factors of production, and prices, then estimating the proportion of land a portion explained by gender differences

4 The Cost of the Gender Gap in Agricultural Productivity: Five African Countries in the market returns to those factors of in the empirical results, thus helping provide production. This report provides results from more comprehensive information to policy- this decomposition procedure for the six makers as they consider reforms to close these factors of production that explained most of productivity gaps. Changes in women’s agri- the gender gap: cultural productivity are highly dependent on local contexts, so survey-based data collection Access to male family labour techniques may be inadequate in capturing Planting of high-value crops some dimensions of well-being, status, Use of agricultural equipment/implements self-esteem, empowerment, social norms and Use of pesticides self-perceptions. Bringing together empirical Use of inorganic fertilizer analysis of original survey data with qualita- The wealth index tive methods helps yield a richer depiction of gender relations (Box 1). Of course, there are other challenges beyond these six factors which inhibit women’s agri- cultural productivity. Some of these challenges The gender gap are not reported here because they are not in agricultural easily measured and are not captured in the productivity in Africa underlying survey data; others are not reported because their contribution in the decomposi- Numerous Sub-Saharan Africa countries are tion procedure was relatively small.3 characterized by substantial gender differences in agricultural productivity. The gaps in the five To complete the analysis, this report draws focus countries are considerable, ranging from on a set of follow-up qualitative studies about 11 to 12 per cent in Ethiopia and Rwanda — “Factors Driving the Gender Gap in Agri- to 28 per cent in Malawi (Figure 1a). These yield cultural Productivity” — conducted by UN differences for male and female farmers vary Women and PEI in 2017 in Malawi, Tanzania across countries for a number of reasons which and Uganda. These studies provide field-level differ by country — including gender gaps in analysis of the factors in each country driving the use of fertilizers and improved seed varie- the gender gap in agricultural productivity and ties, access to and use of credit, ownership of the adoption of climate-resilient approaches land and availability of labour. to provide specific policy and programmatic recommendations. The qualitative evidence Gender gaps in agricultural productivity for the supplements the quantitative studies, as the five focus countries are within the range of those latter did not highlight all the factors that estimated in other published studies which could explain the gender gap, nor did they have used comparable methods (Figure 1b). go into much depth as to the socioeconomic, These gender gaps average around 20–30 per institutional and policy constraints which play cent across Sub-Saharan African countries for a role in determining inequitable access to which there are data, with the smallest gap in labour and non-labour agricultural inputs. Kenya (8 per cent) and the largest in northern (35 per cent). It would be interesting The qualitative data from the UN Women–PEI to explore why the gap appears to be so low studies were collected via a variety of methods, in Kenya. This 8 per cent estimate comes from including interviews and stakeholder consul- a summary chart published by the World Bank tations. Combining quantitative methods with (2012), which in turn drew from an older study qualitative can be useful for better under- by Saito, Mekonnen and Spurling (1994). The standing the dynamics of gender differences only more recent study to examine the gender

Introduction 5 BOX 1 The importance of good sex-disaggregated data

he empirical work in the UN Women and PEI differ with the reasons for female headship. As a Treports is largely based on data for male- and practical illustration of this issue, men and women female-controlled plots of land. Such data are not farmers in Ghana make the same maize adoption widely available, and proxies and extrapolations decisions if they are in male-headed households, typically must be made. In particular, the World but farmers in female-headed households are Bank’s Living Standards Measurement Study – Inte- less likely to adopt improved varieties (Doss and grated Surveys on Agriculture (LSMS-ISA), which Morris, 2001). This finding suggests that female- has been conducted in several Sub-Saharan African headed households face constraints not faced by countries, collects data on who in the household farmers in male-headed households. owns, manages and works each plot of land. While the LSMS and similar large-scale national surveys Understanding the significance of gender differ- are conducted at the household level, they include ences is best addressed by developing research some sex-specific information on individuals within designs based on sex-disaggregated data and by the survey design.4 Such surveys have the advan- adding an explicit gender dimension to the analysis. tage of allowing the researcher to draw conclusions Small-scale surveys are more manageable in terms at that scale, and can provide insights into gender of resource and time costs, but have disadvantages differences that can be used as a context for more for programmes targeting mega-environments detailed studies. spanning several regions and/or agro-ecolog- ical zones. One compromise is to use large-scale In most countries, large-scale surveys are conducted data sets to identify the types, levels and range at the household level without including any of different gender gaps or indicators of gender sex-disaggregated information on agricultural gaps (for example, in cell phone ownership or inputs and outputs. In such cases, gender is often access to credit). Based on this analysis, sub-pop- proxied by male- and female-headed households, ulations and areas with different types of gender which is not an optimal solution. In male-headed inequality can then be selected within which to households, both spouses are usually present; in collect more detailed agriculturally specific data female-headed households, a husband is usually that are disaggregated by sex. This approach was not present. This approach thus conflates meas- followed by the UN Women and PEI reports on the ures of household composition with the sex of the cost of the gender gap, with the 2015 quantitative household head. Moreover, comparisons between studies based on large-scale nationally represent- male- and female-headed households do not ative survey data, and the 2017 follow-up studies consider the endogeneity of female headship. in Malawi, Tanzania and Uganda taking a more Hence the welfare implications of female headship in-depth approach using qualitative methods.

productivity gap in agriculture in Kenya was The gender gap in agricultural productivity conducted by Alene et al. (2008), but this study is linked to unbalanced access to agricultural was limited to maize production in western inputs and to women’s relatively insecure land Kenya. The researchers did find a gender gap rights. Underlying this inequitable distribution of 19 per cent in western Kenya, which is more in the allocation of inputs are social norms in line with that of the other countries shown dictating gender roles and the gendered in Figure 1. division of labour in households and the

6 The Cost of the Gender Gap in Agricultural Productivity: Five African Countries FIGURE 1. Gender gap in agricultural productivity across selected countries

a. Focus countries b. Other countries

Percentage Percentage 35 35 30 30 25 25 20 20 15 15 10 10 5 5 0 0 Ethiopia Malawi Rwanda Tanzania Uganda Benin Ghana enya enya Niger Nigeria Nigeria (Central) (Nat’l) (Subnat’l)(Western) (Northern)

SOURCES: 1a: Ethiopia Ministry of Agriculture and Natural Resources and UNDP-UNEP PEI Africa, 2017; UN Women and UNDP-UNEP PEI, 2015; UN Women, UNDP-UNEP PEI and World Bank, 2015. 1b: Backiny-Yetna and McGee, 2015 (Niger); Oseni et al., 2015 (Nigeria); World Bank, 2012 (remainder).

NOTE: Data points represent the unconditional values of the male-female gap in agricultural productivity, defined as the value of agricultural output per hectare. market place (Figure 2). Closely intertwined women are responsible for unpaid house- with these social norms are deeply engrained work places a large time burden on them gender stereotypes, laws and policies that and — together with high rates of economic discriminate by sex, and the lack of an activity in the poorest countries — results in enabling environment for women farmers. a double work burden for women. This extra Gendered norms and customs affect input burden can translate into gender-inequitable use, productivity and income generation; they distributions of inputs and resources within also affect producers’ adoption decisions and households and on farms. who benefits from innovations. In other words, the culture around gender norms interacts Primary endowments such as men’s and with the use of key productive inputs such women’s labour can have very different as land, and this interaction affects decisions productive roles in agricultural settings which will condition farmers’ adoption and which are subject to strong gender norms. investment decisions. Even if land, labour and Social norms around gender are not simply fertilizer were equally allocated between male constructs which individuals sense: they are and female farmers, gender gaps in agricul- tangible structures and impediments rooted tural productivity would still be observed due in organizations, economic transactions and to the interaction of gendered norms with group characteristics; and vary by region and many factors of production — some of which along social and demographic dimensions. are more easily measured and addressed than The existence of these norms and constraints others. means that development policies and projects in agriculture (such as training programmes Gender gaps in access to inputs and agri- and farmers associations), as well as techno- cultural productivity often stem from the logical innovations, may vary in their levels of perpetuation of cultural norms which dictate success across local settings. Dismantling a that child care and housework are primarily structure of constraints is crucial for closing a woman’s domain. The expectation that the gender gap in agricultural productivity.

Introduction 7 FIGURE 2 Path model of the gender gap in agricultural productivity

Economy-wide and Gendered norms environmental effects and customs Lower economic growth Division of unpaid work in Lesser poverty reduction the household Increased inequality Division of labour in the marketplace Inability to mitigate impacts of climate change Decision-making power, voice and agency Environmental degradation

Gender gap in agricultural productivity and profits

Gender differences in access to agricultural inputs Land; male labour; climate-smart fertilizer, pesticides and equipment; high-value crops Information, skills and extension services

Low agricultural productivity is also an envi- Ongoing processes of climate change are ronmental sustainability issue, and is further likely to have different impacts on men and exacerbated by climate change. Low produc- women farmers, who then may adopt differ- tivity levels lead to more intensive land use, ential coping and adaptation strategies in perpetuating a vicious cycle of environmental response. But gender gaps in agriculture and natural resource degradation and reduced affect how women and men are able to productivity. In many African countries, the access, participate in, adopt and benefit from productivity of major crops is significantly climate-smart agriculture. Taken together, below potential due to reduced soil fertility climate change and gender-differentiated and soil erosion driven by unsustainable and adaptation to climate change have signifi- changing land use practices, including ineffi- cant implications for poverty reduction. The cient use of chemical fertilizers and pesticides agriculture sector can play an important role (FAO, 2013). Additionally, increasing mean in gender-responsive climate mitigation and temperatures and changes in rain patterns adaptation, but environmentally sustainable will have profound impacts on agriculture and climate-smart approaches to agriculture (FAO, 2016). Food systems contribute up to a are not yet mainstream (FAO, 2013). third of greenhouse gas emissions and must therefore be part of the solution to reducing global warming (Vermeulen et al., 2012).

8 The Cost of the Gender Gap in Agricultural Productivity: Five African Countries Factors driving the gender gap n the five focus countries, women’s limited them time poor, with less ability than men to access to agricultural inputs and factors of invest their own labour in agricultural work. production is the main culprit behind the In Tanzania, women’s relatively lower access gender gap in agricultural productivity. to male family labour explains virtually the IAnalysis of the costs associated with unbal- entire gap in agricultural productivity, costing anced access to agricultural inputs can help the economy about $102 million. In Malawi policymakers set priorities among alternative and Ethiopia, inequitable access to male interventions. To ensure the biggest “bang for family labour accounts for about 45 per cent the buck” in policy interventions, this section of the agricultural productivity gap, with an aims to identify and focus on the most costly accompanying cost to the overall economy constraints to women’s productivity by exam- of $45.1 million in Malawi and $89.0 million ining which factors of production account for in Ethiopia. These results are consistent with the largest portion of the gender gap in agri- findings in the qualitative reports that the cultural productivity. The discussion draws on quantity and quality of labour to which women estimates in the quantitative UN Women and have access is poor compared to men’s PEI reports. access. Moreover, the time women spend on unpaid care work limits the time and quality of labour they can devote to their farms. It is not Key factors for the clear why access to male labour appears to five focus countries play little to no role in explaining the gender gap in Rwanda and Uganda. One possible Table 2 reports results from this decomposi- explanation is that Rwanda and Uganda are tion procedure for the six factors of production among the few countries in Africa which have which explain most of the gender gap in implemented gender-responsive budgeting. Ethiopia, Malawi, Tanzania and Uganda. The Such budgeting efforts often target agricul- decomposition procedure for Rwanda yielded ture and poverty reduction, which might have substantially different results, partly because provided some cushion for women in these the structure of the underlying survey data countries (Budlender, 2000). differed from that used for the other coun- tries. Therefore, the factors of production differ accordingly and are reported sepa- Women’s limited use of agricultural implements and rately, as described below. machinery

Limited access to male family Among the other imbalances, the gender labour differential in the use of implements and machinery explains 18 per cent of the gender As shown in Table 2, one of the largest gap in Malawi, 9 per cent in Uganda and contributors to the gender gap in agricultural 8 per cent in Tanzania. This gender imbal- productivity is access to male family labour, ance in access to equipment costs as much especially in Ethiopia, Malawi and Tanzania. as $17.7 million in lost GDP in Malawi. (For Typically, women have less access to male Ethiopia and Rwanda, it appears that the family labour in cases of divorce, separation underlying data sets did not contain varia- and widowhood. Women are also subjected bles which directly measure women’s and to social norms around gender that make it men’s access to implements and machinery.) very difficult to hire male wage labour. More- As discussed in the qualitative reports, one of over, women’s relatively high burdens of the main reasons for this gender imbalance in unpaid care work and domestic work leave

10 The Cost of the Gender Gap in Agricultural Productivity: Five African Countries TABLE 2 Decomposing the gender productivity gap in agriculture in five countries

Ethiopia Malawi Rwanda Tanzania Uganda Gap Gap in GDP Gap Gap in GDP Gap Gap in GDP Gap Gap in GDP Gap Gap in GDP Driver (%) (mil. $) (%) (mil. $) (%) (mil. $) (%) (mil. $) (%) (mil. $) Access to male family labour 43.7 89.0 45.2 45.1 97.3 102.1 — — Planting of high-value crops — — 28.4 28.4 3.0 3.2 13.3 8.9 Use of equipment/implements — — 17.8 17.7 8.2 8.6 9.0 6.0 Use of pesticides 45.3 92.2 1.0 1.0 12.0 12.6 4.5 3.0 Use of inorganic fertilizer 25.1 51.0 5.3 5.3 6.4 6.7 3.0 2.0 Wealth index — — 5.3 3.3 −0.1 −0.1 — — Sales of household farm production 67.0 280.5 Household size 22.5 94.1 Spending on insecticides 12.8 53.7 Years of education 11.5 48.3

SOURCES: Ethiopia Ministry of Agriculture and Natural Resources and UNDP-UNEP PEI Africa, 2017; UN Women and UNDP-UNEP PEI, 2015; UN Women, UNDP-UNEP PEI and World Bank, 2015. NOTE: Decomposition categories differ for Rwanda because of differences in underlying survey data. Percentages do not sum to 100 because the decompositions include other categories with negative values. Percentages do not sum to 100 because a number of drivers can be negative. Only a selection of those that reduce the gender gap are shown here, and together, they may overcompensate.

access to equipment is women’s lack of cash or ownership. Women’s lower likelihood of income given their responsibility for meeting planting high-value crops may also be an household maintenance needs. outcome of limited access to climate change adaptation tools and extension services; this lack of access is related not only to cash Limited access to high-value crops income constraints but also to women’s rela- enterprise tive time poverty given their need to perform Gender differences in the planting of high- domestic work and care labour. value crops account for another 28 per cent, 3 per cent and 13 per cent of the gender gaps The gender differential in high-value crops in Malawi, Tanzania and Uganda, respec- comprises the highest proportion of the tively; with a cost to GDP of over $28 million overall gender gap in agricul- in Malawi, $3.2 million in Tanzania and almost tural productivity in Uganda. Women are more Women in Uganda appear to $9 million in Uganda. High-value crops likely to plant include cash crops and exported crops, which be particularly burdened by are typically farmed by men; women are more the cash and time constraints subsistence crops. likely to plant subsistence crops. The quali- of their unpaid care and tative reports attribute these disparities to domestic work — which results in their not social norms which dictate that women have growing higher-value crops or having the primary responsibility for household food requisite cash to access better agrarian tech- production. Moreover, women may be unable nologies. Uganda is distinct among the five to scale up to the level required for high-value countries in farm production practices in that crops if they are constrained by plot size and/ women are relatively less available to work on

Factors driving the gender gap 11 their own plots of land due to social norms farm inputs, especially organic fertilizer. that create the expectation that women work While organic fertilizer has beneficial effects on plots jointly controlled by their husbands. for soil quality, women’s over-reliance on this In fact, women are expected to work on their input reduces the productivity of their plots husbands’ plots before working their own compared to those of men using chemical land. These norms reduce women’s likelihood fertilizers. Gender differentials in the alloca- of investing in higher-value crops on their own tion of pesticides and fertilizers also play a plots of land. relatively large role in Tanzania, collectively explaining almost one-fifth of the overall agri- cultural productivity gap. Women’s limited access to agricultural inputs The results in Table 2 are not directly compa- Female-controlled plots have relatively lower rable for Rwanda, because the underlying data yields because important inputs such as set did not contain information for the same inorganic fertilizer and pesticides are used categories and because different agricultural mostly on male-controlled plots. Organic ferti- input categories were economically mean- lizers are usually produced by livestock owned ingful and statistically significant. That said, by a household, while inorganic fertilizers are the results for Rwanda indicate that women’s purchased in the marketplace. In most coun- inequitable access to cash crops explains tries, women tend to have less access to both about two-thirds of the gender gap (with a types of fertilizer. One of the primary expla- resounding $280.5 million cost to the coun- nations for women’s relative try’s GDP), while access to pesticides accounts for another 13 per cent. The remaining two Inputs such as lack of access to fertilizers and pesticides is their rela- categories relate to gender differences in inorganic fertilizer tively lower cash income, household size (which is largely picking up a higher dependency ratio for women) and in and pesticides are which is related to heavy demands on their time in average years of education. used mostly on performing unpaid work at home. As shown in Table 2, male-controlled gender differentials in the use Factor findings of plots. of pesticides and fertilizer are comparable studies particularly large in Ethiopia, explaining 45 per cent and 25 per cent of the The decomposition results for the five focus total agricultural productivity gap, respec- countries are consistent with those published tively. The cost in terms of lost GDP in Ethiopia for other countries using the same meth- is enormous: $92.2 million lost due to ineffi- odology. In particular, Backiny-Yetna and cient allocations of pesticides and $51 million McGee (2015) find that in Niger, the most lost due to inefficient allocations of fertilizer. important gender differentials explaining the The report for Ethiopia points to particu- total productivity gap occur in access to male larly large income and wealth constraints for labour, the use of inorganic fertilizer and land women which prevent them from being able ownership. to purchase as many modern plant protection technologies as men, including pesticides, These factors also matter in Nigeria, albeit herbicides and fungicides. In an apparent with different emphases. In northern Nigeria, attempt to compensate for these deficien- gender differences in the market returns to cies, Ethiopian women are more likely to rely factors of production (the explained part of on self-provided non-land and non-labour the gap) matter more than gender differences

12 The Cost of the Gender Gap in Agricultural Productivity: Five African Countries in access to these factors of production. The authors find that rainfall shocks which increase opposite holds in southern Nigeria, where the output of crops predominantly cultivated virtually the entire gender gap in agricultural by women shift expenditures towards food, productivity is explained by gender differ- while rainfall shocks affecting cash crops ences in access to inputs, especially land cultivated by men have no effect on food holdings (Oseni et al., 2015). The authors point expenditures. Hence, the potential economic to cultural differences between the north and gains from reducing the gender gap translate south that could explain these differences. into poverty reduction and improved nutri- Notably, the north is predominantly Muslim, tional outcomes. so women may be more constrained in hiring male labour and accessing markets. In Kenya, female farmers are found to be as responsive to price incentives in terms World Bank (2012) decomposition results of output supply and input for Benin, Ghana and Kenya indicate that demand as male farmers, Inefficient gender differences in access to inputs account and as economically effi- for most if not all of the total gender gap in cient when agricultural inputs allocation of agricultural productivity. and human capital factors of production are taken into agricultural inputs, Numerous published studies across devel- account (Alene et al., 2008). not inefficiency of oping regions have documented that once access to inputs such as land, fertilizer, labour In Malawi, although men use women farmers, and credit is taken into account, women are more fertilizer, enjoy greater is the main as productive and technically efficient as men. access to extension services That is, the main explanation for the gender and devote relatively more explanation for gap in agricultural productivity is not that land to cash crops, experi- women are less efficient cultivators. Rather, mental results indicate that the gender gap there is an inefficient allocation of land, labour female farmers are no less in agricultural and other agricultural inputs among house- efficient than male in terms hold members. Much of the evidence to of crop yields when they are productivity support this claim comes from Sub-Saharan provided with equal access to Africa. For example, in Burkina Faso, plots inputs (Gilbert, Sakala and Benson, 2002). controlled by women have lower yields than similar plots simultaneously planted with the As a final example of the large number of same crop but controlled by men within the studies on the gender gap in agricultural household. The main explanation is that inef- productivity in Sub-Saharan Africa, in the ficient allocations of land, labour and inputs Osun state of Nigeria, female rice farmers are such as fertilizers contribute to lower yields actually more technically efficient than male, for plots controlled by women. particularly when age and years of education are controlled for (Oladeebo and Fajuyigbe, Results indicate that reallocating factors of 2007). production in a more efficient manner could increase output by 6 per cent (Udry 1996). Duflo and Udry (2004) find an inefficient allo- Other significant cation of resources in Côte d’Ivoire, where factors of production strong gender norms dictate that men and women farm their own plots without trying Although the decomposition results summa- to maximize joint household production. The rized in Table 2 do not include access to

Factors driving the gender gap 13 credit and land rights as the primary drivers Across developing regions, women own and of the gender gap in the focus countries, control substantially less land than men.5 these factors of production do play a role In most African countries, where women’s in influencing crop choice and access to key control over land depends on customary agricultural inputs. tenure systems, the variation in women’s control over land is more substantial than Women have faced difficulties in accessing in other regions. This circumstance partially formal credit through commercial banks due reflects variations across countries in the to their lack of collateral; this problem is exac- importance of community-held land and in erbated by weak or non-existent property the extent to which countries have patrilineal rights for women. Credit market imperfec- versus matrilineal communities. In all Sub-Sa- tions in turn can have greater adverse effects haran African countries for which there are on women’s ability to engage in income-gen- data on land holdings and ownership, about erating activities and purchase farm inputs 19 per cent of agricultural land holders are (Palacios-López and López, 2015). women. As shown in Figure 3, this average hides considerable variation. At one extreme Although access to land may not always be is Cabo Verde, where about 50 per cent of all a binding constraint, it has implications for agricultural holders are women; at the other the scale of farming women can engage in extreme, the share of land held by women is and how productive they can be. This, in turn, no more than 6 per cent in Guinea and Mali. affects their ability to generate a marketable surplus. Hence, greater gender equality in Looking more closely at some of the reasons access to credit and in land ownership could for women’s lack of land rights, Goldstein and increase women’s engagement in agricultural Udry (2008) find that in rural Ghana, women markets. have relatively less social and political power

FIGURE 3 Percentage of individual landholders in various African countries who are women

Percentage 60

50

40

30

20

10

0 Mali Guinea Malawi Nigeria ambia Uganda Lesotho Ethiopia Tanzania Comoros Botswana DR Congo Seychelles Madagascar The Gambia Cabo erde Burkina Faso C te d’Ivoire Mozambique

SOURCE: Food and Agriculture Organization of the United Nations, Gender and Land Rights Database, 2018, www.fao.org/gender-landrights- database/en/.

14 The Cost of the Gender Gap in Agricultural Productivity: Five African Countries in villages, are less likely to have secure land land expropriated. The implications of these rights, and consequently, are less likely to gender inequities in land holdings for agricul- invest in improving land fertility. The authors tural investments and output are enormous: attribute women’s substantially lower profits insecure land tenure reduces the incentive of per hectare compared to men primarily farmers to invest either in their land or in high- to women’s insecure land tenure and the value crops. heightened risk women face of having their

Factors driving the gender gap 15 Gains from closing the gap ender gaps in agricultural produc- — which has the highest gender productivity tivity in Sub-Saharan Africa are gap among the five focus countries (28 per sizeable. The potential economic cent) — leads to a relatively smaller increase gains from eliminating them in crop production than for Rwanda, whose Gwould translate into meaningful reductions productivity gap is about 11 per cent. Thus, in poverty and improvements in nutritional the gains in closing the gender gap in agri- outcomes — and in the overall macro- cultural productivity in Rwanda appear to economy. A growing body of literature shows be greater. These results could be pointing that the marginalization of women’s labour to already favourable conditions for gender impedes poverty reduction efforts, dampens equality and for the agricultural sector in productivity and reduces economic growth Rwanda, such that making women more (DFID et al., 2013). In agricultural economies, productive has relatively high returns at the the effect of gender on growth prospects is macroeconomic level. linked to the gender division of labour and to gender inequality in land ownership and Although the percentage increases in crop loan access. production in the other three focus countries are smaller, the boosts to both agricultural As demonstrated by the UN Women and PEI GDP and overall GDP are country case studies and as supported by still large. For example, in Gender equality cross-country evidence in numerous other Ethiopia, the 1.4 per cent in access to land, published studies, gender equality in access increase in crop production to land, technology and agricultural inputs translates into a whopping technology and holds the key to increasing productivity in $221 million increase in agri- agricultural food production. Not only is gender inequality cultural GDP. Much of the economically inefficient, the sizeable macro- reason for this large increase inputs holds the economic repercussions of its presence is that crop output accounts provide a rationale for better understanding for 71 per cent of agricultural key to increasing and eliminating the barriers which prevent GDP in Ethiopia, which serves productivity in food women from having full access to agricultural as a strong rationale for resources and inputs. The predicted benefits promoting gender equality production. from removing the barriers women face in in agricultural productivity. terms of increased productivity of land and Rwanda has the largest increase in agricultural overall agricultural output are substantial GDP among the five countries at over $345 (Croppenstedt, Goldstein and Rosas 2013). million. Even Uganda’s increase in agricultural GDP of $58 million — the lowest among the Increasing access to labour, land, equipment countries — is still substantial. With multiplier such as ploughs and other implements at effects factored in arising from the contribu- appropriate times in the crop cycle would tion of agriculture to GDP, the increases in greatly improve productivity, crop produc- overall GDP across countries are even higher, tion and income. As shown in Figure 4, the ranging from $67 million in Uganda to $419 increases in crop production, agricultural GDP million in Rwanda. and overall GDP are substantial. Closing the gender gap in agricultural productivity corre- Closing the gender gap in agricultural produc- sponds to an estimated increase of almost tivity corresponds with a substantial reduction 19 per cent in crop production in Rwanda in poverty. At least 80,000 people in Tanzania and a 7 per cent increase in Malawi. It is inter- and as many as 238,000 people in Rwanda esting that closing the gender gap in Malawi could be lifted out of poverty per year over

Gains from closing the gap 17 FIGURE 4 Gains from closing the gender gap in agricultural productivity

a. Increase in crop production b. Increase in agricultural GDP

Percentage Million $ 20 400

15 300

10 200

5 100

0 0 Ethiopia Malawi Rwanda Tanzania Uganda Ethiopia Malawi Rwanda Tanzania Uganda

c. Increase in total GDP d. People lifted out of poverty each year

Million $ Thousands 500 250

400 200

300 150

200 100

100 50

0 0 Ethiopia Malawi Rwanda Tanzania Uganda Ethiopia Malawi Rwanda Tanzania Uganda

SOURCES: Ethiopia Ministry of Agriculture and Natural Resources and UNDP-UNEP PEI Africa, 2017; UN Women and UNDP-UNEP PEI, 2015; UN Women, UNDP-UNEP PEI and World Bank, 2015.

a 10-year period if the total gender gap in Overall, the results in Figure 4 suggest that agricultural productivity in each country were Rwanda would gain the most from closing closed. The benefits in Rwanda are again the gender gap in agricultural productivity large, with over 200,000 people lifted out of in terms of the greatest increases in food poverty annually. These absolute shifts in the production and GDP and a substantial reduc- number of people living in poverty represent tion in poverty, with Ethiopia and Malawi not meaningful changes in the poverty headcount far behind. Although the gains in Tanzania ratio (the proportion of the population living and Uganda appear to be relatively smaller below the poverty line). These ratios ranged within this sample, in absolute terms the from 20 per cent in Uganda to 51 per cent benefits are extensive. in Malawi at approximately the time the UN Women–PEI surveys were conducted.6

18 The Cost of the Gender Gap in Agricultural Productivity: Five African Countries Closing the gap: remedies and recommendations he results reported provide an due to their disproportionately large work- objective framework within which to loads caring for children and engaging in evaluate alternative policy options domestic work. Inequitable access to agri- to close the agricultural productivity cultural inputs is often reinforced by deeply Tgap between men and women. As shown in entrenched social norms which devalue Table 3, these policy reforms can be classified women’s work. Transformative interventions into five broad areas: are needed to change gender norms so other proposed innovations in agricultural produc- Increasing women’s access to labour and tion can have optimal outcomes. time-saving equipment and services One avenue towards rectifying this situation Facilitating women’s shift to high-value is to consider policies which help women to crops access hired labour. A large hurdle is that strong social norms may prevent women from Improving women’s access to non-labour hiring male labour — especially when these agricultural inputs norms dictate that men and women engage in different agricultural tasks. These norms Strengthening women’s land rights would need to be challenged with educa- tional campaigns to help ensure the success Pursuing other interventions which close of subsequent policies aimed at strength- the gender gap ening women’s ability to hire wage labourers, male and female. These reforms will have a bigger impact in some countries than others depending on A somewhat innovative, but potentially effec- the importance of the underlying factors of tive, approach could be to launch edutainment production in explaining the gender gap as programmes on national television with the measured by the decomposition analysis. objective of educating the general public Table 3 provides an intuitive representation of about the benefits of relaxing rigid cultural the “bang for the buck” afforded by various norms in farming work. This would be policy reforms and programme interventions supported by evidence-based advocacy to according to whether each country is likely to policymakers and community leaders on the experience high, medium or low economic importance of women’s economic empower- gains. Governments can choose among ment in households — and specifically, the these policy reforms and priorities according importance of engaging women in sustain- to those that would yield the most benefit for able agriculture. As discussed in the UN their country. Women–PEI reports, policies to assist women in gaining access to hired labour include cash vouchers for hiring labour, as well as doorstep Increase women’s delivery of equipment and training. access to labour and time-saving equipment Reforms which provide women with greater access to time-saving household and farming and services equipment will help reduce their time poverty and free up their own labour to engage in One of the key results from the decompo- productive agricultural work. Household sition analysis and the follow-up qualitative equipment would include energy-efficient and studies is that women lack access to male environmentally friendly improved cooking family labour and are time-poor themselves

20 The Cost of the Gender Gap in Agricultural Productivity: Five African Countries TABLE 3 Policy reform options and potential gains across the five countries

Ethiopia Malawi Rwanda Tanzania Uganda Increase women’s access to labour and time-saving equipment and services

§§Support women’s access to hired labour §§Confront gender norms around labour High High Medium High Low §§Introduce labour-saving devices §§Invest more in public infrastructure

Facilitate women’s shift to high-value crops

§§Improve access to info & markets §§Promote gender-awareness in R&D Low Medium High Low Medium §§Confront gender norms

Improve women’s access to non-labour agricultural inputs

§§Improve access to fertilizers & pesticides §§Encourage wider range of machinery use High Medium Medium Medium Medium §§Increase women’s access to credit

Strengthen women’s land rights

§§Consider large-scale land titling reforms Medium High Medium High High §§Correct gender biases in laws

Pursue other interventions which close the gender gap

§§Reduce gender-based violence §§Promote climate-smart agricultural practices §§Gender-responsive extension services High High High High High §§Provide agricultural skills training §§Representation in cooperatives §§Encourage knowledge dissemination

SOURCE: Adapted and updated from Ethiopia Ministry of Agriculture and Natural Resources and UNDP-UNEP PEI Africa, 2017; UN Women and UNDP-UNEP PEI, 2015; UN Women, UNDP-UNEP PEI and World Bank, 2015. stoves to reduce the amount of time women equipment would include replacing tradi- spend fetching firewood, thus freeing up their tional sowing, storage and land preparation time for productive work. Digging wells has tools with modern machinery and equipment, also helped reduce women’s domestic work which would contribute to sustainable agricul- burdens. In Burkina Faso, various organiza- tural production in a changing climate. This tions undertook initiatives to construct wells, equipment would include modern irrigation supply carts to villages for hauling wood, build facilities, solar driers for grains and tractors fuel-efficient ovens, and introduce hullers for large-scale land preparation. and grain mills to convert grain into flour. Kompaoré, McSweeney and Frisanco (2007) Another set of interventions to save (unre- present evidence which suggests the intro- munerated) time for women includes duction of these new technologies reduced infrastructure improvement in the water sector, women’s workloads and helped them use their electrification, road construction, better trans- freed-up time to engage in more economi- portation options and sanitation services. The cally productive activities. Time-saving farm absence of infrastructure, especially in rural

Closing the gap: remedies and recommendations 21 areas, is an important driver of the relatively and by liberating women from home produc- high burden of unpaid domestic work borne tion. With electrification, households used by women. Infrastructure improvement is less wood within the home and increasingly considered an important gender-smart policy adopted electric lighting and cooking. Electri- contributing to both economic growth and fication thus served as a labour-saving device gender equality. These investments work in the rural landscape allowing women to at the macroeconomic level by improving move from home production to market work. several important outcomes, including health, access to markets, labour productivity and provides another example of a successful investment. They also help narrow the gender intervention which both improved infrastruc- gap in agricultural productivity by increasing ture to reduce women’s unpaid work and women’s access to time-saving services; contributed to environmental conservation. this reduces their unpaid work burdens and The country’s 2005 National Rural Employ- increases their ability to expend their own ment Guarantee Act (NREGA) provides each labour on agricultural work. rural household a minimum of 100 days of work per year. Although there are few eligi- In a study of time use by bility requirements, it was believed that the Planners and men and women in Tanzania, nature of the work and the level of wages Fontana and Natali (2008) were such that only the poor would self- policymakers do not find that planners and poli- target access to the programme. According cymakers do not adequately to the government, the main aim of NREGA always recognize recognize the role of is to increase wage employment and wage how improving improving infrastructure in security in rural India. Secondary objectives transforming the market included addressing underlying reasons for infrastructure economy to work more effec- poverty such as deforestation, soil erosion can transform the tively. The authors calculate and drought. As part of these secondary that provision of better infra- services, sponsored projects involved road market economy. structure in Tanzania would, construction, improved irrigation and water on average, result in a 24 per conservation. cent increase in total annual cash earnings from reducing food preparation times, and Women were particularly attracted to NREGA in increases of 4 per cent and 2 per cent in work for several reasons. First, NREGA wages total annual cash earnings, respectively, from do not differentiate by sex — which means reductions in water and fuel collection times. that the offered wage differential between Since most of these gains would accrue to rates for alternative options and NREGA work women, there may be additional beneficial is especially attractive for women. Second, spillover effects in terms of poverty reduc- much of the work can be undertaken locally, tion, investments in children and increased a stipulation that reduces travel times to work economic well-being within the home. sites and reduces the burden for women who bear most of the responsibility for housework. Another infrastructure-focused develop- NREGA worksites also have mandated child ment scheme yielding significant benefits for care facilities. While this provision was not women is ’s rural electrification fully enforced at all sites, it eased the burden programme. Dinkelman (2011) finds that the of child care in areas where this functioned introduction of rural electrification increased relatively well (Narayanan, 2008). female employment within five years, primarily by allowing for micro-enterprises

22 The Cost of the Gender Gap in Agricultural Productivity: Five African Countries Programme evaluation results suggest that Inequitable access to land also helps NREGA has had positive impacts on employ- to explain why women plant relatively ment and wages — especially for women fewer high-value crops. Women may be (Azam, 2012). The wage increases for women unable to scale up if they own or manage may have resulted in important feedback just small plots of land, or they may be effects such as stronger bargaining power unwilling to invest in cash within the household, just as the programme crop cultivation if their There are many itself led to the construction of much-needed land tenure is insecure or rural infrastructure. they do not own the land reasons why women at all. Reforms to land decide not to grow ownership laws (discussed Facilitate women’s shift below) will help in this high-value crops. to high-value crops regard.

The decomposition results indicate that Women may also avoid investing in cash women are less likely than men to grow high- crops for market sale if they have trouble value crops, and this explains a substantial accessing markets due to restrictions in portion of the agricultural productivity gap, their mobility arising from social norms especially in Malawi, Rwanda and Uganda. A — and, in many countries, relatively less number of issues underlie what appears to be access to motorized forms of transpor- a choice to not grow high-value crops: tation. Challenging social norms around women’s mobility and improving public Gender norms in most countries in the infrastructure and means of transportation region dictate that women are largely will help relieve some of these constraints. responsible for food crop production, so they are constrained by these norms to Another constraint is women’s relatively focus on subsistence farming in order to lower access to new farming techniques and ensure the food security of their house- seed technologies. Specific policy reforms holds. Switching to cash crops could to address this constraint include improving involve changes in the nutritional content women’s access to information, promoting of household food consumption. gender awareness in research and develop- ment on new technologies, and confronting Women may avoid planting higher-value social norms around crop choice and crops if their cultivation entails heavy marketing. The diffusion of new seed tech- demands on time, and the women are nologies that match women’s preferences time-poor or have relatively less access to will help. In planning agricultural innovations, hired labour. Policy options described in it is important to differentiate between male the previous section to increase women’s and female preferences affecting technology access to labour will help address this choice and adoption decisions, rather than problem. assume uniform household preferences. A large number of studies have disproven the Women may also be constrained in terms unitary household model with evidence that of cash, meaning that greater access to households do not pool resources nor share credit (discussed in the next section) or to the same preferences. In many countries, some sort of funding programmes would individual men and women in a household help. manage different resources with different criteria for making decisions about farming.

Closing the gap: remedies and recommendations 23 In other words, households do not plough, for both seed and fertilizer through the FISP plant, fertilize, weed or irrigate crops — indi- has reduced the gender gap in adoption of viduals do. Even when individual men and modern maize in Malawi.7 women in farm households jointly own some of the productive resources used for farming Women may also face gender-based and make decisions together, they may differ constraints in their access to natural resources. in how they share the costs and distribute the Common resources such as water may not benefits of production decisions. be shared equally in patriarchal societies. A case in point is southern Tanzania, where farmers depend on irrigation to grow their Improve women’s crops, and an informal set of codes governs access to non-labour how much water farmers divert to their own agricultural inputs fields and how much they leave for farmers downstream. A field experiment conducted More widespread adoption of improved by Lecoutere, d’Exelle and Van Campenhout technologies and climate-smart agricultural (2015) shows that high-status men keep more approaches may not require changes in the than half of the available water for themselves innovation process, but rather better access — regardless of whether the water is scarce for women to complementary inputs. One or in abundance — while high-status women of the overriding constraints women face share altruistically, no matter what. Gender in accessing such inputs is and social status thus influence how users of An overriding their spending power. In self-governed common watersheds distribute their water. Policies assuming fair and equi- constraint for particular, women’s respon- sibilities for meeting cash table use of common pooled resources can women in accessing household needs constrains be misguided if gender and social standing their spending on agricultural are not taken into account. needed agricultural inputs. Policy-based instru- inputs is their ments to increase their cash Another important institutional reform is holdings and reduce input increasing women’s access to credit. The spending power. prices will help alleviate this security of land ownership is associated with constraint. In some cases, higher agricultural productivity; however, agricultural input subsidies can help reduce women may face more obstacles than men the gender gap in the adoption of new seed in obtaining credit. These obstacles can technologies. weaken the potential of land titling in bene- fiting women. One of the consequences of Malawi provides a useful case study because remaining isolated from networks is restricted it has had a long-term commitment to agricul- access to credit markets. This restricted access tural input subsidies. Notably, Malawi’s Farm also results from women’s relatively meagre Input Subsidy Programme (FISP) has helped control over assets that can be used as collat- close the gender gap in modern maize adop- eral. Further, formal lending institutions often tion (Fisher and Kandiwa, 2014). Receipt of view women as risky clients because of their subsidized input coupons had no discernible lower levels of education and skills, which effect on modern maize adoption for male increases the perception that they cannot be farmers, but adoption by female household “banked.” heads was positively influenced by receipt of a FISP package consisting of both maize seed These restrictions handicap women in and fertilizer coupons. Receiving a subsidy their ability to purchase agricultural inputs.

24 The Cost of the Gender Gap in Agricultural Productivity: Five African Countries Research suggests that the targeted provision management and help ensure that capital is of small-scale loans through microfinance initi- spent on productivity-enhancing agricultural atives can support and incentivize women’s inputs. agricultural activities and promote economic welfare. The success of initiatives such as the Grameen Bank model in rural Bangladesh Strengthen women’s has contributed to the proliferation of microf- land rights inance initiatives across countries and regions throughout the world. This movement has Another critical intervention in the agricul- demonstrated the extent of the unmet need tural sector is formalization of women’s land for credit among poor women and the poten- ownership. When female farmers have more tial for commercial banks to play a larger role formal control over land, their productivity in providing access to formal credit to margin- increases. Land rights need to be guaran- alized women. teed in such a way that women can exchange, lease, bequeath, sell or mortgage land in an On the other hand, several studies have enforceable manner. Recommendations for found that microloans may bring only modest policy reforms supported by findings in the advantages to women (Banerjee, Karlan and UN Women–PEI reports center on changing Zinman, 2015). An alternative to microfinance the legal structures governing with demonstrated success in reaching women women’s land rights. This When female in agriculture is rural banking policy reforms objective can be achieved which incentivize the opening of commer- through improved documen- farmers have more cial bank branches in remote areas that were tation, stronger communal formal control previously unbanked. A good example is rights, constitutional revisions India’s rural banking reform in which commer- to inheritance rights and land over land, their cial banks were required to open branches in titling programmes. four previously unbanked locations (often in productivity rural areas) in order to obtain a license for A growing number of govern- increases. opening a branch in an already banked loca- ments have implemented tion. The government also set deposit and large-scale land titling programmes, whose lending policies to provide incentives for results indicate that joint titling of land for people to use the new banks. The savings married couples serves as an effective way rates were higher and lending rates lower for more women to gain legal land rights. than in urban areas, and there were targets Mandatory joint titling in particular raises set for lending in priority areas that included the likelihood of women gaining secure land agriculture and small-scale entrepreneurship. rights; voluntary joint titling is somewhat less Results in Menon and Rodgers (2011) indi- effective in providing large proportions of cate that the increased availability of credit women with secure rights, especially in coun- afforded by this programme was of particular tries with strong patriarchal social norms. significance to women. A common problem across developing Finally, another approach to help low-in- regions is that land titles are often distrib- come women emerge from subsistence-level uted to heads of household only. Since the production is capital bundled together majority of household heads in most devel- with training and financial services (Buvini´c oping countries are male, this practice has and Furst-Nichols, 2016). Such services will led to an overwhelmingly disproportionate help build women’s capacities in financial number of land titles allocated to men. A

Closing the gap: remedies and recommendations 25 number of countries have revised their land Recommendations for policy reforms inheritance rights in recent years to guar- supported by the literature centre on ensuring antee equal inheritance of land to sons and that statutory and customary laws are in daughters and to ensure joint land ownership accordance with legal guarantees — espe- between husband and wives. cially in Africa, where both customary land redistribution schemes and official land law Policy recommendations center on improve- reforms have led to reallocations frequently ments in implementation of land law reforms, biased against women. Government agencies especially when uneven implementation have typically distributed land to household results in gender disparities in the issuance heads. This process favours men in general, of land titles. For example, Ethiopia had a and specifically the more senior men who successful first-time process already have power through customary land Many Sub-Saharan for land registration which holdings. Even though government efforts to was rapid, participatory reform land laws may give female household countries still have and effective in achieving heads the right to receive land, in practice the objective of increasing they may not receive plots in the reallocation somewhat arbitrary household investments in process when local officials redefine them as inheritance and land (Deininger et al., 2008). dependants of male relatives. If customary However, there was substan- laws in Sub-Saharan countries have placed property rights for tial variation across regions in restrictions on women’s access to land or women. whether land was registered prevented women from pursuing title, these in the name of the husband laws need to be revised so they are in accord- only, wife only, or jointly — largely due to ance with constitutional provisions governing regional variation in programme require- equality (Joireman, 2008). ments for joint certification and photographs of landowners. Also needing improvement Many Sub-Saharan African countries still have was women’s representation on village-level somewhat arbitrary inheritance and property land-use committees, with distant overnight rights for women, due to competing provisions meetings serving as a substantial disincentive in laws and legal statutes. Greater attention for women’s participation in these political needs to be paid to gender relations and bodies. power structures in rural areas which disad- vantage women in their attempts to own land. Rwanda’s large-scale Land Tenure Regular- State-sponsored efforts to formalize property ization programme resulted in greater land rights need to address not only customary tenure security and large positive effects laws and women’s right to property, but also on agricultural investment, especially in enforcement, especially in under-resourced female-headed households (Ali, Deininger areas. and Goldstein, 2014). The programme clar- ified land rights, reduced tribal conflicts Examples of relevant policy lessons are found and reduced gender discrimination in land in Kenya, Senegal, Tanzania and Zambia. In access. These accomplishments contributed Senegal, land redistribution policies did not to increased land access for married women, correct gender inequality in access to land, improved documentation of inheritance as only a small proportion of women could rights and large increases in soil conservation obtain land rights as household heads. efforts. Following extensive droughts in the 1970s, the government distributed equal-sized plots to all heads of household. Yet by 1990, women

26 The Cost of the Gender Gap in Agricultural Productivity: Five African Countries — all household heads — held only 6 per cent In Zambia, the HIV/AIDS epidemic has caused of irrigated plots. Moreover, married women enormous suffering and loss of life. Chapoto, in Senegal did not have customary land-use Jayne and Mason (2011) found that more than rights, and many widows and divorcees who two-thirds of the households which experi- were entitled to plots were denied this right enced the death of the male head after 2001 because they were classified as dependants had less land within a three-year span, and of male relatives (Koopman, 2009). Conse- about one third of the widow-led households quently, tiny plots and tenuous land rights controlled less than half the land they had for women in the Senegal River Valley have controlled before the male head had died. constrained food output and marketable The authors note that government decrees to surpluses. protect the security of land access by female widows have had little to no impact when local In the River Njoro Watershed area of Kenya, authorities were not included policy reforms have led to a shift away from in the decree arrangements. Government communal land ownership towards private This situation highlights the decrees to protect individual ownership, with indigenous Maasai importance of working with pastoralist communities having to adjust to local community authorities widows’ security further changes introduced by agricultural in developing programmes in land access settlements. Interview data in Willya and to property rights for widows. Chiuri (2010) indicate that property rights have had little to reforms have caused substantial changes in While the evidence from gendered tasks and workloads, with women empirical studies provides no impact when seeing increased domestic workloads and a clear rationale for proce- local authorities fewer user rights to resources such as cows, dures to increase women’s small farm animals and firewood. Local men, land ownership, the litera- were not included however, have gained from cash payments ture also suggests that such in the decree they have earned from migratory livestock procedures would have more herders. The authors conclude that commu- potent impacts if they were arrangements. nities as a whole have gained from finding embedded in a framework ways to blend indigenous and modern insti- which sought to widen the scope of institu- tutions, but women are bearing the burden of tional structures so they assist women living the transition. on the margin.

Tanzania and Zambia offer further examples of the need to address customary laws and local Other interventions for practices in Sub-Saharan Africa. In Tanzania, closing the gender gap the “double-safeguarding of land rights” — which involves legitimizing customary land Several additional interventions can directly inheritance practices and contemporary land target women’s engagement in the agricul- rights — may lead to the exclusion of marginal tural sector and contribute towards narrowing groups, including women. As discussed in the gap in accessing multiple key inputs. One the qualitative UN Women and PEI report for of the most important such interventions Tanzania, women’s groups in Tanzania have is legislative reform around gender-based successfully challenged the exclusionary violence, stronger enforcement of such laws, effect of certificates of customary occupation. and other programmes to change attitudes towards and raise awareness about gender- based violence.

Closing the gap: remedies and recommendations 27 The Malawi, Tanzania and Uganda reports that closing gender gaps in access to factors clearly show that gender-based violence and of production, along with the adoption of traditional attitudes towards masculinity serve climate-smart agricultural practices, will help as an important contributing factor to the close gender gaps in farm crop productivity gender gap in agricultural productivity (UN and improve household livelihoods. Such Women and UNDP-UNEP PEI, 2019a, 2019b climate-smart agricultural practices include and 2019c). Gender-based violence is a viola- selecting crop varieties which are planted tion of civil and political rights as well as social earlier, mature earlier, and are high-yielding and economic rights, leading to numerous and drought resistant; using cover crops to adverse outcomes for women’s health, well- improve crop husbandry; preserving natural being and productivity — and for overall grass and tree coverage on uncultivated agricultural output. As noted in the country land; promoting the use of organic ferti- reports, policymakers need to consider not lizers; conserving water through rainwater only stronger laws and enforcement to prevent harvesting and precision irrigation; rotating gender-based violence, crops; introducing agroforestry methods Closing gender gaps but should also develop which integrate trees with crops (or livestock) programmes to educate men on plots of land; using contour ridging and in access to factors and women on the economic markers; and introducing soil management of production, benefits of cooperation within techniques which limit disruptions to soil the household, so agricultural composition and structure. along with the productivity improvements at adoption of climate- the farm level can be lever- Extension services are one of the most common aged. Such programmes vehicles for disseminating knowledge and smart agricultural would help women gain training in climate-smart agricultural tech- access to more male house- niques. However, women in agriculture often practices, will hold labour, one of the key have limited access to agricultural support help close gender drivers of the agricultural services. For one thing, extension services productivity gap in several of are often dominated by men, resulting in the gaps in farm crop the focus countries. potential for gender bias in the diffusion of productivity and knowledge of new seed technologies and The country reports also high- training in new farming techniques. improve household light the importance of closing gender gaps in climate-smart Female extension agents may be more likely livelihoods. agricultural practices. As the to reach women farmers with new training reports note, agricultural techniques and information, especially in productivity hinges on the effective applica- societies that are highly segregated by tion of various factors of production, which gender (Quisumbing et al., 2012). This argu- are in turn linked to the agronomic prac- ment is supported by evidence in Kondylis tices being followed. These practices affect et al. (2016). That study used a randomized soil fertility and serve as the foundation of control trial conducted in Mozambique successful adoption of climate-smart agricul- to estimate the impact of training female tural practices. The reports were not able to messengers in sustainable land manage- quantify the extent to which these practices ment (SLM) techniques on the awareness, explain the total gender gap in agricultural knowledge and adoption of SLM prac- productivity because the underlying survey tices by female farmers. In this experiment, data did not include this information. The communities were randomly selected to qualitative research, however, clearly shows have a female messenger trained in SLM who

28 The Cost of the Gender Gap in Agricultural Productivity: Five African Countries was encouraged to teach other women the collective approaches have the potential to techniques. The authors found that women’s increase women’s access to resources. Such awareness of pit planting farming techniques approaches include group investments in increased by 9 percentage points and their capital inputs, individual ownership combined adoption of the technology increased by with group cultivation and the distribution of 5 percentage points one year later in commu- group rights by governments to poor rural nities with female messengers compared to women (ICRW, 2005). Although collective communities in the experimental areas which action groups can be helpful in mobilizing did not have female messengers. The results communities to engage in imply that placing women in extension posi- development, they need to Collective tions can help women farmers overcome be monitored carefully to approaches have the barriers to adoption posed by inequitable ensure existing inequities are access to agricultural extension services and not exacerbated (Kinkingnin- potential to increase information on sustainable farming tech- houn-Medagbe et al., 2010). women’s access to niques. Women also tend to be resources. Policy interventions which empower women absent from professional through the provision of agricultural skills and service networks which allow for the training opportunities are vital. A training quick diffusion of information. Much of this programme which has successfully improved is because sociocultural norms limit their farmer skills and training — especially of mobility and social circles. Improving access women farmers — is Junior Farmer Field and to information for women so as to improve Life Schools. In an effort to reach some of the their productivity and success in the market most vulnerable individuals, this programme place is of special importance. India’s Gyan- has targeted orphaned children in countries doot network is a good example of how with a high prevalence of HIV/AIDS (Came- women’s access to information has been roon, Kenya, Malawi, Mozambique, Namibia, fostered (DFID, FAO and ODI, 2002). This Sudan, Swaziland, Tanzania, Uganda, Zambia network is a system of village “kiosks,” linked and Zimbabwe). Both boys and girls are through an intranet system, where people taught agricultural concepts including may access public information such as land sowing, weeding, conservation, processing of records, technical advice, marketing informa- crops and marketing. This training is bundled tion for agricultural commodities and prices, with basic education and learning, and other and details on government projects. It also concepts of equality, health and nutrition. serves as a communication outlet, as individ- Programme evaluations have been positive, uals may submit complaints directly to the especially in terms of their impact on the local government; and as a means of e-mar- empowerment of women and girls who have keting. Gyandoot’s innovations — especially benefited from the new methods and innova- the ability to post questions to agricultural tive means of skills delivery (FAO, IFAD and extension staff and to gain information on ILO, 2010). market prices of grains and other commodi- ties — are invaluable to India’s female farmers. Similarly, measures to connect small women The technology has the power to bring meas- farmers to marketing networks and local agri- urable benefits to a segment of farmers who cultural value chains improve their overall are often not recognized in their own right, viability and help build their confidence. and who are often curtailed in their everyday Women are not typically represented in agri- dealings with the market. cultural cooperatives to a large degree. Yet

Closing the gap: remedies and recommendations 29 Mobile phone technology is another excel- areas. Much of this effect is accounted for by lent means by which women’s access to women who do not have extensive child care information and services has been fostered. responsibilities within the home. Most women The introduction of this technology has had found work in wage employment, which trans- a widespread impact on economic develop- lated into increases in household income and ment across Africa (Aker and Mbiti, 2010). measurable declines in extreme poverty in But there is evidence that mobile phones South Africa. Hence, improving access to have had disproportionately positive effects services and information can bring substantial for women. For example, Klonner and Nolen targeted benefits to women. (2008) find that with the advent of a mobile phone network in South Africa, employment increased by 15 percentage points in rural

30 The Cost of the Gender Gap in Agricultural Productivity: Five African Countries Notes

1. See https://au.int/sites/default/files/ 4. In addition to the World Bank’s LSMS (http:// documents/31358-doc-au_echo_ surveys.worldbank.org/lsms), sex-disag- january_2015.pdf for the African Union gregated information is captured in the Commission’s report on the Year of Women’s databases of the Global Findex Study Empowerment and Development towards (https://globalfindex.worldbank.org/), the Africa’s Agenda 2063. The devotion of World Bank’s World Development Indi- a chapter on “Women in Agriculture: cators (http://datatopics.worldbank.org/ Addressing the Gender Gap through Rele- world-development-indicators/) and the vant Agricultural Frameworks” highlights International Food Policy Research Institute’s the importance of reforms to reduce gender Women’s Empowerment in Agriculture Index inequality in the agricultural sector. The (WEIA; http://www.ifpri.org/project/weai). African Commission’s protocol on the rights of women can be found here: http://www. 5. The data in this paragraph are from Food achpr.org/instruments/women-protocol/. and Agriculture Organization of the United Nations, Gender and Land Rights Database, 2. See in particular ILO Convention No. 100 2018, www.fao.org/gender-landrights- on Equal Remuneration (1951), Convention database/en/. The database defines an No. 111 on Discrimination in Employment agricultural holder as “the civil or juridical and Occupation (1958), Convention No. person who makes the major decisions 156 on Workers with Family Responsibilities regarding resource use and exercises (1981) and Convention No. 183 on Maternity management control over the agricultural Protection (2000). Conventions 100 and 111 holding”; this definition includes land owners, are among the eight fundamental conven- producers and managers. Additionally, most tions of the ILO Declaration on Fundamental country time use surveys and demographic Principles and Rights at Work. Other relevant and health surveys collect at least some ILO instruments include Convention No. 175 sex-disaggregated information. on Part-time Work (1994), Convention No. 177 on Home Work (1996), Convention No. 182 on 6. Source: World Bank, World Development the Worst Forms of Child Labour (1999), the Indicators database, 2018, http://databank. Declaration on Equality of Opportunity and worldbank.org/data/home. Treatment for Women Workers (1975) and the Declaration on Fundamental Principles and 7. Note that the Fisher and Kandiwa (2014) Rights at Work (1998). study uses the same data set as that on which the UN Women–PEI productivity gap study 3. The quantitative studies upon which this for Malawi is based — the 2010–2011 Inte- report is based only reported results for those grated Household Survey. This implies that factors of production which, for at least one the gender gap in agricultural productivity in of the countries, (i) had a substantially large Malawi would have been even higher than the impact on the gender gap in agricultural reported 28 per cent if the FISP were not in productivity, and (ii) had a statistically signif- effect. icant coefficient in the regression analysis used to derive the decomposition results.

31 Bibliography

Aker, Jenny, and Isaac Mbiti (2010). Mobile phones Chapoto, Antony, T. S. Jayne and Nicole Mason and economic development in Africa. Journal of (2011). Widows’ land security in the era of HIV/ Economic Perspectives 24 (3): 207–232. AIDS: panel survey evidence from Zambia,” Economic Development and Cultural Change 59 Alene, Arega, Victor Manyong, Gospel Omanya, (3): 511–547. Hodeba Mignouna, Mpoko Bokanga and George Odhiambo (2008). Economic efficiency and supply Croppenstedt, Andre, Markus Goldstein and response of women as farm managers: compar- Nina Rosas (2013). Gender and agriculture: ineffi- ative evidence from Western Kenya. World ciencies, segregation, and low productivity traps. Development 36 (7): 1247–1260. World Bank Research Observer 28 (1): 79–109.

Ali, Daniel Ayalew, Klaus Deininger and Markus Deininger, Klaus, Daniel Ayalew Ali, Stein Holde Goldstein (2014). Environmental and gender and Jaap Zevenbergen (2008). Rural land certi- impacts of land tenure regularization in Africa: pilot fication in Ethiopia: process, initial impact, and evidence from Rwanda. Journal of Development implications for other African countries. World Economics 110: 262–275. Development 36 (10): 1786–1812.

Autor, David (2015). Lecture Note 1: Wage density DFID, FAO and ODI (Department for International decompositions. MIT Graduate Labor Economics Development, Food and Agriculture Organization 14.662. of the United Nations and Overseas Develop- ment Institute) (2002). Program for Information on Azam, Mehtabul (2012). The impact of Indian job Sustainable Livelihoods: India case study. DFID, guarantee scheme on labor market outcomes: London. evidence from a natural experiment. IZA Working Paper No. 6548. DFID (Department for International Develop- ment), William and Flora Hewlett Foundation Backiny-Yetna, Prospere, and Kevin McGee (2015). and IDRC (International Development Research Gender differentials in agricultural productivity in Centre ) (2013). Growth and economic Niger. Policy Research Working Paper No. 7199. opportunities for women: literature review to Washington, D.C.: World Bank. Inform the DFID-IDRC-Hewlett Foundation research program on women’s economic Banerjee, Abhijit, Dean Karlan and Jonathan empowerment, gender equality and growth in low Zinman (2015). Six randomized evaluations of income countries. microcredit: introduction and further steps. Amer- ican Economic Journal: Applied Economics 7(1): Dinkelman, Taryn (2011). The effects of rural elec- 1–21. trification on employment: new evidence from South Africa. American Economic Review 101(7): Budlender, Debbie (2000). A global assessment of 3078–3108. gender responsive budget initiatives. Doss, Cheryl, and Michael Morris (2001). How does Buvini´c, Mayra, and Rebecca Furst-Nichols (2016). gender affect the adoption of agricultural innova- Promoting women’s economic empowerment: tions? The case of improved maize technology in what works? World Bank Research Observer 31 (1): Ghana. Agricultural Economics 25 (1): 27–39. 59–101.

32 Duflo, Esther, and Christopher Udry (2004). Intra- Joireman, Sandra Fullerton (2008). The mystery of household resource allocation in Côte d’Ivoire: capital formation in Sub-Saharan Africa: women, social norms, separate accounts and consumption property rights and customary law. World Devel- choices. National Bureau of Economic Research opment 36 (7): 1233–1246. Working Paper No. 10498, Cambridge, MA: NBER. Kinkingninhoun-Medagbe, Florent, Aliou Diagne, Ethiopia Ministry of Agriculture and Natural Franklin Simtowe, Afiavi Abgoh-Noameshie and Resources and UNDP-UNEP PEI (United Nations Patrice Adegbola (2010). Gender discrimination Development Programme–United Nations Envi- and its impact on income, productivity, and tech- ronment Programme Poverty-Environment nical efficiency: evidence from Benin. Agricultural Initiative) Africa (2017). The Cost of the Gender and Human Values 27 (1): 57–69. Gap in Agricultural Productivity in Ethiopia. Klonner, Stefan, and Patrick Nolan (2008). Does FAO (Food and Agriculture Organization of the ICT benefit the poor? Evidence from South Africa. United Nations) (2013). Climate-Smart Agriculture University of Essex. Sourcebook. Rome: FAO. Kompaoré, Scholastique, Brenda Gael McSweeney FAO (Food and Agriculture Organization of the and Jennifer Hilda Frisanco (2007). The quest for United Nations) (2016). The State of Food and gender equality in Burkina Faso: female workloads, Agriculture Climate Change, Agriculture and Food education and empowerment. UNESCO, Geneva. Security. Rome: FAO. Kondylis, Florence, Valerie Mueller, Glenn Sheriff FAO, IFAD and ILO (Food and Agriculture Organ- and Siyao Zhu (2016). Do female instructors ization of the United Nations, International Fund reduce gender bias in diffusion of sustainable for Agricultural Development, and International land management techniques? Experimental Labour Organization) (2010). Gender Dimensions evidence from Mozambique. World Development of Agricultural and Rural Employment: Differenti- 78: 436–449. ated Pathways Out of Poverty: Status, Trends and Gaps. Rome: FAO. Koopman, Jeanne (2009). Globalization, gender, and poverty in the Senegal River Valley. Feminist Fisher, Monica, and Vongai Kandiwa (2014). Can Economics 15 (3): 253–285. agricultural input subsidies reduce the gender gap in modern maize adoption? Evidence from Malawi. Lecoutere, Els, Ben d’Exelle, and Bjorn Van Camp- Food Policy 45: 101–111. enhout (2015). Sharing common resources in patriarchal and status-based societies: evidence Fontana, Marzia, and Luisa Natali (2008). Gendered from Tanzania. Feminist Economics 21 (3): 142–167. patterns of time use in Tanzania: public investment in infrastructure can help. Paper prepared for the Menon, Nidhiya, and Yana Rodgers (2011). How IFPRI Project on Evaluating the Long-Term Impact access to credit affects self-employment: differ- of Gender-focused Policy Interventions. University ences by gender during India’s rural banking of Sussex, Brighton, UK. reform. Journal of Development Studies 47 (1): 48–69. Gilbert, Robert, Webster Sakala and Todd Benson (2002). Gender analysis of a nationwide cropping Narayanan, Sudha (2008). Employment guarantee, system trial survey in Malawi. African Studies Quar- women’s work and childcare. Economic and Polit- terly 6 (1): 1–21. ical Weekly 43 (9): 10–13.

Goldstein, Markus, and Christopher Udry (2008). Oladeebo, Job O., and A. A. Fajuyigbe (2007). The profits of power: land rights and agricultural Technical efficiency of men and women upland rice investment in Ghana. Journal of Political Economy farmers in Osun State, Nigeria. Journal of Human 116 (6): 981–1022. Ecology 22 (2): 93–100.

ICRW (International Center for Research on Oseni, Gbemisola, Paul Corral, Markus Gold- Women) (2005). Property ownership for women stein and Paul Winters (2015). Explaining gender enriches, empowers and protects. ICRW, Wash- differentials in agricultural production in Nigeria. ington, D.C. Agricultural Economics 46 (3): 285–310.

Bibliography 33 Palacios-López, Amparo, and Ramon López (2015). UN Women and UNDP-UNEP PEI (United Nations The gender gap in agricultural productivity: the Development Programme–United Nations Envi- role of market imperfections. Journal of Develop- ronment Programme Poverty-Environment ment Studies 51 (9): 1175–1192. Initiative) (2019b, forthcoming). Factors driving the gender gap in agricultural productivity: Tanzania. Quisumbing, Agnes, Ruth Meinzen-Dick, Terri Raney, André Croppenstedt, Jere Behrman UN Women and UNDP-UNEP PEI (United Nations and Amber Peterman, eds. (2014). Gender in Development Programme–United Nations Envi- Agriculture: Closing the Knowledge Gap. Wash- ronment Programme Poverty-Environment ington, D.C.: Food and Agriculture Organization Initiative) (2019c, forthcoming). Factors driving the of the United Nations. gender gap in agricultural productivity: Uganda.

Saito, Katrine, Hailu Mekonnen and Daphne UN Women, UNDP-UNEP PEI (United Nations Spurling (1994). Raising the productivity of women Development Programme–United Nations Envi- farmers in Sub-Saharan Africa. Africa Technical ronment Programme Poverty-Environment Department Discussion Paper Series 230, World Initiative) and World Bank (2015). The Cost of the Bank, Washington, D.C. Gender Gap in Agricultural Productivity in Malawi, Tanzania, and Uganda. Singhal, Arvind, and Everett M. Rogers (1999). Entertainment-Education: A Communication Vermeulen, S., R. Zougmore, E. Wollenberg, Strategy for Social Change. Mahwah, New Jersey: P. Thornton, G. Nelson, P. Kristjanson, J. Kinyangi, L. Erlbaum Associates. A. Jarvis, J. Hansen, A. Challinor, B. Campbell and P. Aggarwal (2012). Climate change, agriculture and Udry, Christopher (1996). Gender, agricultural food security: a global partnership to link research production, and the theory of the household. and action for low-income agricultural producers Journal of Political Economy 104 (5): 1010–1046. and consumers. Current Opinion in Environmental Sustainability 4: 128–133. UN Women and UNDP-UNEP PEI (United Nations Development Programme–United Nations Envi- Willya, Daniel, and Wanjiku Chiuri. 2010. New ronment Programme Poverty-Environment common ground in pastoral and settled agricultural Initiative) (2017). Equally productive? Assessing the communities in Kenya: renegotiated institutions gender gap in agricultural productivity in Rwanda. and the gender implications.” European Journal Policy Brief. of Development Research 22 (5): 733–750.

UN Women and UNDP-UNEP PEI (United Nations World Bank (2012). World Development Report Development Programme–United Nations Envi- 2012: Gender Equality and Development. Wash- ronment Programme Poverty-Environment ington, D.C.: World Bank. Initiative) (2019a, forthcoming). Factors driving the gender gap in agricultural productivity: Malawi.

34 The Cost of the Gender Gap in Agricultural Productivity: Five African Countries

UN Women, grounded in the vision of equality as enshrined in the Charter of the United Nations, works for the elimination of discrimination against women and girls; the empowerment of women; and the achievement of equality between women and men as partners and beneficiaries of development, human rights, humanitarian action, and peace and security. Placing women’s rights at the centre of all its efforts, UN Women leads and coordinates United Nations system efforts to ensure that commitments on gender equality and gender mainstreaming translate into action throughout the world. It provides strong and coherent leadership in support of Member States’ priorities and efforts, building effective partnership with civil society and other relevant actors.

The Poverty-Environment Initiative (PEI) of the United Nations Development Programme (UNDP) and the United Nations Environment Programme (UN Environment) supports country-led efforts to mainstream poverty-environment linkages into national development planning and budgeting. PEI provides financial and technical assistance to government partners to set up institutional and capacity-strengthening programmes and carry out activities to address the particular poverty-environment context. PEI is funded by the governments of Norway, , Sweden, the United Kingdom, and the European Union and with core funding of UNDP and UN Environment.