Household Analysis of Poverty and Gender Dynamics in Sri Lanka: Drivers of Exclusions

Transcription

1 Household Analysis of Poverty and Gender Dynamics in Sri Lanka: Drivers of Exclusions Market Development Facility March 2017 Draft Version

2

3 Table of Contents Executive Summary...vii Chapter 1: Introduction to the Study... 1 About the Market Development Facility... 1 Strategic Engagement Areas in Sri Lanka... 1 Purpose and Objectives of the Study... 3 Conceptual Framework Poverty... 4 Conceptual Framework Wo e s E o o i E po e e t... 5 Research Design and Methodology... 6 Research Questions... 6 Secondary Research Literature Review... 7 Primary Research Field-based Data Collection... 7 Field Research Teams... 7 Sampling Strategy... 8 Tools Used for Primary Research Data Management, Collation and Analysis Limitations of the Research General Limitations Specific Strategic Engagement Area Limitations Organisation of the Study Chapter 2: Overview of Poverty and Gender in Sri Lanka Poverty Reduction and Economic Growth in Sri Lanka Impact of Civil War on Poverty Features of Poverty in Sri Lanka Overview of Gender in Sri Lanka Chapter 3: Findings of the Poverty and Gender Study Diversifi atio of Sri La ka s Touris Methodology and Sample Characteristics Who is Poor in Tourism and Why? Key Constraints and Opportunities for Poor People in Tourism Pathways out of Poverty in Tourism Roles Performed by Women in Tourism Barriers for Women in Tourism Market Development Facility i

4 Recommendations to Address Barriers for Women in Tourism Chapter 4: Findings of the Poverty and Gender Study Improving Sri Lankan Produce and Products for Export and Tourism Markets Fisheries Methodology and Sample Characteristics Who is Poor in Fisheries and Why? Key Constraints and Opportunities for Poor People in Fisheries? Pathways out of Poverty in Fisheries Roles Performed by Women in Fisheries Barriers for Women in Fisheries Recommendations to Address Barriers for Women in Fisheries Horticulture Methodology and Sample Characteristics Who is Poor in Horticulture and Why? Key Constraints and Opportunities for Poor People in Horticulture Pathways out of Poverty in Horticulture Roles Performed by Women in Horticulture Barriers for Women in Horticulture Recommendation to Address Barriers for Women in Horticulture Chapter 5: Findings of the Poverty and Gender Study Stimulating Entrepreneurship and Business Activity in Former Conflict-affected Areas Methodology and Sample Characteristics Who is Poor in Former Conflict-affected Areas and Why? Key Constraints and Opportunities for Poor People in Former Conflict-affected Areas Pathways out of Poverty in Former Conflict-affected Areas Roles Performed by Women in Former Conflict-affected Areas Barriers for Women in Former Conflict-affected Areas Recommendations to Address Barriers for Women in Former Conflict-affected Areas Chapter 6: Findings of the Poverty and Gender Study Innovation in Digital Services Who is Poor in Digital Services? Pathways out of Poverty in Digital Services Roles Performed by Women in Digital Services ii Market Development Facility

5 Recommendations to Address Barriers for Women in Digital Services Chapter 7: Summary of Findings and Conclusions of the Poverty and Gender Study Snapshot of the Full Sample The Way Forward for MDF Recommendations for Tourism Recommendations for Fisheries Recommendations for Horticulture Recommendations for Former Conflict-affected Areas Recommendations for Digital Services Recommendations for WEE General Recommendations Annex 1: Final Sample Household Interviews Key Informant Interviews Focus Group Discussions Annex 2: Primary Research Tools Sample Household Questionnaire Diversifying Tourism Household Interview Summary Template Diversifying Tourism Focus Group Discussion Template Diversifying Tourism Annex 3: Sample Household Interview Summaries Household Interview Summary Diversifying Tourism Household Interview Summary Sri Lankan Produce and Products Household Interview Summary Entrepreneurship in Former Conflict-affected Areas Household Interview Summary Innovation in Digital Services List of Tables Table 1: Main Research Questions... 7 Table 2: Research Team Members... 8 Table 3: Poverty Categories Used in the Study... 8 Table 4: Details of Interviews according to SEAs, Geographic Areas and Respondent Profiles... 9 Table 5: Primary Research Tools Table 6: Primary Research Data Management Table 7: Primary Research Data Collation Market Development Facility iii

6 Table 8: Empirical Studies of the Cost of Conflict in Sri Lanka Table 9: Tourism Paid Employment Salaries Table 10: Type of SME and Corresponding Earnings List of Figures Figure 1: MDF Impact Logic Figure 2: District Official Poverty Lines 2016 Figure 3: Provincial Share of GDP ( ) Figure 4: Poverty incidence by district (2012/2013) Figure 5: Distribution of per capita income in horticulture households Figure 6: Workforce in ICT List of Boxes Box 1: Strategic Engagement Areas for Sri Lanka... 2 Box 2: References to Primary Research iv Market Development Facility

7 Abbreviations and Acronyms BRAC Bangladesh Rural Advancement Committee CCPI Colombo Consumer Price Index FGD Focus group discussion FTE Full-time equivalent GCE General Certificate of Education GDP Gross Domestic Product GII Gender Inequality Index HIES Household Income and Expenditure Survey ICT Information Communications Technology ICRW International Centre for Research on Women IDP Internally displaced persons IT Information technology KII Key informant interview LKR Sri Lankan rupee LTTE Liberation Tigers of Tamil Eelam MDF Market Development Facility MDGs Millennium Development Goals MFI Micro finance institute NGO Non-government organisation PNG Papua New Guinea PPI Progress out of Poverty Index PPP Purchasing Power Parity SLTDA Sri Lanka Tourism Development Authority SMEs Small and medium-sized enterprises SMMEs Small, micro and medium enterprises Market Development Facility v

8 SEA Strategic Engagement Area US United States USD United States dollar WEE Women's Economic Empowerment vi Market Development Facility

9 E e uti e Su ar Having emerged from three decades of civil conflict, Sri Lanka has made significant strides in poverty reduction and economic growth. Poverty has fell from about 22.7 percent in 2002 to 6.7 percent in 2012/13.1 Sri Lanka has made similarly impressive advances in achieving greater health and edu atio out o es fo o e, ut this has ot t a slated i to i p o ed o e s e o o i a d political participation. In the last decade female labour force participation has remained in the range of 32 to 40 percent, compared to a near-constant male participation of around 75 percent. While S i La ka s ad a es a e i p essi e, the ask po kets of po e t a d ul e a ilit that a e pronounced along gender, geographical, ethnic and sectoral lines. Poor women and men, with few assets and inadequate access to markets, are often hardest hit in economic shocks. But even the near poor women and men who are no longer classified as poor, but are below the 40th percentile are only slightly better off. They subsist on less than twice the poverty line and are similar to people categorised as poor in terms of consumption patterns and asset ownership.2 On the other hand, the richest 20 percent of Sri Lankan households e jo. pe e t of the ou t s atio al 3 income. There is a wealth of evidence that shows that structural inequalities lead to wasted productive potential in the long-run, and attention to equity is critical for sustained growth and eventual graduation into upper middle-income country status. As Sri Lanka gears itself to the meet the targets of the Sustainable Development Goals (SDGs) and its own ambitious development vision, it is vital that growth is inclusive of poor people, offers greater opportunities to women and integrates former conflict-affected areas In Sri Lanka, MDF works within four strategic engagement areas that represent a major part of the economy, have long-term growth prospects and are relevant for poverty reduction and inclusive g o th. These st ategi e gage e t a eas a e suppo ti g the di e sifi atio of S i La ka s tourism destinations, products and services, (2) improving Sri Lankan produce and products for export and tourism markets, (3) stimulating entrepreneurship and business activity in former 1 Household Income and Expenditure Survey 2012/13 Final Report, Department of Census and Statistics Ministry of Policy Planning Economic Affairs, Child Youth and Cultural Affairs, Sri Lanka 2 Newhouse, Suarez-Becerra, and Doan (2016). Sri Lanka - Poverty and Welfare: Recent progress and remaining challenges. 3 Sri Lanka Socio-economic Data (2016) Central Bank of Sri Lanka, June 2016 Vol. Xxxix Market Development Facility vii

10 conflict-affected areas, and (4) innovation in digital services to facilitate international tourism and export-led, services-oriented growth as well as inclusive growth. MDF complements its in-depth, ongoing research on market dynamics with an equally thorough examination and analysis of poverty to understand where it occurs, for whom, and for what reasons. Women, in their roles within households, in enterprises and as consumers, face additional barriers, and understanding those requires looking past the surface. This household-level study was undertaken to develop deeper understanding of poverty and gender dynamics. The objective of the study was two-fold. First, to guide MDF in designing interventions that address the heart of what keeps people t apped i po e t. Se o d, to u de sta d a d app e iate o e s oles i a kets and as a result, work with partners to adopt strategies that foster greater participation of women. The qualitative study is designed to complement statistical data, which, while offering a broad picture, does not answer crucial questions. What is it that makes some households poor, and what enables others to emerge out of poverty? The primary research looked at a number of households within an area, those that are poor, those that are vulnerable and those that are more dynamic, in order to understand factors that differentiate better-off households from poor households. The fi di gs of the stud sho that poo people a e passed S i La ka s e e t g o th. They exist on peripheries of booming sectors and belong to households that are low-skilled and vulnerable to economic shock. They live in remote, often conflict-affected areas, cut-off from opportunities, and are trying to recover from loss of assets and livelihoods. While relatively few households live under the national poverty line, most are not faring much better either. Most households are unable to meet basic needs and rely on formal and informal debts. The findings point towards strategies that can unlock opportunities for the poor. Tourism is promising as it has the potential to employ a decent number of people and reach those in remote areas. It was found that tourism jobs often are a pathway out of poverty. For MDF, this would mean to work towards expanding tourism to new areas. However, tourism is not automatically inclusive of the poor. A closed network where operators work with few service providers does not expand oppo tu it. MDF s st ateg should e to a ds ope i g up tou is to offer greater diversity of products, activities and services, and to improve quality by investing in in-demand skillsets. Among fishing communities, the crucial difference between poor and better-off households was the ability to catch more fish, which was determined by an ability to invest in nets and boats. This points MDF towards a direction of designing appropriate savings and financing schemes. Value-addition can offer jobs for women, and is also relevant for former conflict-affected areas. Export varieties of seafood a e hugel p ofita le, a d MDF s effo ts should e to a ds i easi g suppl of fingerlings and larvae. viii Market Development Facility

11 In agriculture, connections to markets are key, which can prompt farmers towards more productive cultivation. Traditional agriculture was found to offer limited prospects with farmers struggling to recover their investments. Problems have to do with poor irrigation, low productivity, expensive inputs, and loss caused by disease and weather. Diversification of crops and technological advancement is needed to build resilience in the livelihoods of farming communities. This change needs to be led by businesses markets can incentivize farmers to upgrade practices and offer them the inputs and information to make these changes. Challenges in former conflict-affected areas are greater. People are specially disadvantaged, having suffered loss of household members, loss of assets, dislocation and disrupted livelihoods. The vulnerable groups are war widows, women-headed households and unemployed youth. Business activity and entrepreneurship is rudimentary and will need to be nurtured. The unique challenges would require MDF to develop an adaptive and patient strategy. Wo e s oles are strong within households. They are the principle money managers and have decision-making roles within households. Outside of households, pernicious social norms dictate he e a d to hat deg ee o e s p ese e is a epta le. Household espo si ilities a d otio s of status, respectability, safety and reputation limit wo e s a ilit to a ess jo s. The fa to s that encourage greater participation of women were found to be proximity of the workspace, safe and secure transport, dedicated workspaces, presence of other women, and flexible arrangements around childcare. Financial literacy was found to be a gap across the study sample, underpinning all strategic engagement areas. Poor financial planning results in reliance on loans which leads to greater vulnerability. Most households are acutely aware of this and are hesitant to take loans doubting their ability to repay. Yet, risky investments and purchase of luxury goods are also common. MDF has a role to play in promoting better financial literacy to encourage investment in productive assets, and to work with partners to offer suitable saving, investment and insurance products. Lastl, the elatio ship et ee this stud a d MDF s o k is o ple e ta a d ite ati e. MDF s networks, prior fieldwork and interaction with a variety of players across value chains have greatly info ed this stud. The fi di gs of the stud ill guide MDF s futu e o k, i desig i g ta geted interventions to reach poor and vulnerable people, and encouraging more women to participate. As MDF extends its work and gains more experience, further nuance will develop in its understanding around markets and their interaction with poor men and women. In that regard, this study should not be taken as an isolated piece of research, but an ongoing, iterative process of constant learning informed by practice. Market Development Facility ix

12 x Market Development Facility

13 Chapter : I trodu tio to the Stud About the Market Development Facility The Market Development Facility (MDF) stimulates investment, business innovation and regulatory reform in order to create additional jobs and increase the income of poor women and men in rural and urban areas in the Indo-Pacific region. MDF began in Fiji in July 2011 and has since expanded to Timor-Leste (in 2012), Pakistan (2013) and Papua New Guinea (PNG) and Sri Lanka (both in 2015). Phase I of MDF ends in June Preparations for Phase II are underway. To stimulate investment, business innovation, and regulatory reform, MDF negotiates partnerships with strategically positioned private and public sector organisations in its countries of operation. Each partnership is comprised of a tailor- ade pa kage of a ti ities that e a les MDF s pa t e s to innovate, invest and/or undertake reforms in such a manner that small farms and firms benefit from better access to production inputs, services and end markets. This makes them more productive and helps them grow which, in turn, creates jobs and increases incomes for poor women and men. Each MDF partnership promotes business innovations or regulatory reforms, leverages private se to i est e t o pu li se to o e ship ost sha i g, has a de o st ated li k ith p o-poor growth, job creation and income generation, and helps the economy onto a pathway to transformative, systemic change see Figure 2: MDF Impact Logic. Household studies on poverty and gender such as this are undertaken by MDF to gain an in-depth view of income and expenditure trends and gender dynamics within households and to explain inclusivity barriers within markets and what can be done to address them. Understanding these ealities is a esse tial pa t of MDF s o k i desig i g a d i ple e ti g effe ti e pa t e ships that promote good, sustainable business practices and, at the same time, address key development halle ges fa ed ou t ies like S i La ka. These halle ges i lude po e t a d o e s economic empowerment. Strategic Engagement Areas in Sri Lanka In Sri Lanka, MDF has opted to work in the Tourism and Related Sectors in four Strategic E gage e t A eas SEAs. These a e: suppo ti g the di e sifi atio of S i La ka s tou is destinations, products and services, 2) improving Sri Lankan produce and products for export and tourism markets, 3) stimulating entrepreneurship and business activity in former conflict-affected areas and 4) innovation in digital services to facilitate international tourism and export-led, servicesoriented growth. These four SEAs were identified because: 1) they are export-led and drive much needed future growth in Sri Lanka growth that is broad-based and inclusive, and relevant for both poor women a d e, the add ess so e of S i La ka s ke de elop e t halle ges, a d the a e i flue ed a d e gaged i te s of the investment, innovation and reform needed to unlock this potential in a feasible and sustainable manner that promotes value for money by a business facilitation process implemented by MDF. Market Development Facility 1

14 Box 1: Strategic Engagement Areas for Sri Lanka MDF Strategic Engagement Areas (SEA) for Sri Lanka Suppo ti g the di e sifi atio of S i La ka s tou is desti atio s, p odu ts a d se i es 2) Improving Sri Lankan produce and products for export and tourism markets 3) Stimulating entrepreneurship and business activity in former conflict-affected areas I p o i g S i La ka s digital apa ilit to fa ilitate i te atio al tou is services-oriented growth. a d e po t-led, Following the impact logic outlined in Figure 1 below, MDF aims, in Sri Lanka, to introduce 23 innovations and reforms relevant for pro-poor growth, leverage USD 918,000 in private sector investment, make markets work better for the poor to the value of USD 2,754,000 and, in doing so, improve the livelihoods of over 25,092 poor women and men, create 2,040 new full-time equivalent (FTE) jobs, and create over USD 1,524,000 in additional income for the poor. These estimates are based on a two-year implementation window (2015 to 2017) and will be achieved by The decision to focus on the above four strategic engagement areas resulted from an intensive assessment phase that involved literature review and extensive fieldwork. The fieldwork was to validate and complement the available data, build up business networks and, above all, start identifying the rare, exceptional, innovative entrepreneurs and tentative investment plans or usi ess odels that ould eak e g ou d, ope up e a kets, se e poo lie ts ette, a d push the industry towards transformative and systemic change. This in-depth look at the household level of the poor, covered in this report, was then done to build profiles, ascertain poverty dynamics in regions and sub-sectors, and gender dynamics within these households, and discover why poor men and women remain excluded from the benefits of the growth. This in-depth stud p o ides MDF ith u i ue pe spe ti es o aspe ts of o e s economic empowerment (WEE) and social cohesion ambitions that are central to facilitate inclusive development in Sri Lanka. Understanding barriers and developing appropriate pathways to address them helps MDF in identifying, developing and designing relevant partnerships that can generate inclusive impact in a sustainable manner. 2 Market Development Facility

15 Figure 1: MDF Impact Logic Purpose and Objectives of the Study The objectives of the study were to: Identify the poor men and women involved in the SEAs and their roles, location and other general characteristics. Identify the factors that appear to be keeping them in poverty. Better understand their involvement in the SEAs and their broader livelihoods. Understand their constraints and opportunities in relation to the SEAs and their broader livelihoods. Identify specific pathways out of poverty for these men and women in relation to the SEAs. These pathways must be market-driven and feasible for a programme such as MDF to implement. Better understand the involvement of women, in particular, in the SEAs. Better understand the specific constraints and opportunities of women in relation to the SEAs and their broader economic power. Ide tif, he e eeded, additio al e o e datio s to p o ote o e s e o o i empowerment in relation to the SEAs. These recommendations must be market-driven and feasible for a programme such as MDF to implement. Market Development Facility 3

16 Conceptual Framework Poverty S i La ka s offi ial po e t li e as esta lished i a d set at LKR, eal total e pe ditu e 4 per person/per month for the year For the 2012/13 poverty incidence estimates, the 2002 poverty line was inflated for 2006/07 using the base 2002 Colombo Consumer Price Index (CCPI), and then inflated again using the base 2006/07 CCPI for the year 2012/13. Such calculations allow for comparability through the years. As such, a person was identified as being poor in the 2012/13 HIES if their expenditure fell below LKR 3,624 per month.5 As of July 2016, when the primary research for the Study took place, the poverty line was set at LKR 4,132 per person per month, or the equivalent of LKR per person per day. This figure is above the global poverty line of USD 1.25 a day in 2005 PPP terms, or the equivalent of LKR 59.8 per person per day.6 This MDF Study utilised both national and international poverty lines to establish household levels of poverty as these allow for consistency and comparison across diverse areas and communities in Sri La ka, as ell as a oss the diffe e t SEAs ithi MDF s po tfolio. It is i po ta t to e og ise that quantitative measures of poverty are often insufficient in capturing the multi-dimensional and complex nature of poverty.7 Standard methods of collecting information on income and expenditure allow only for the accounting of market goods, with less tangible areas being excluded such as health, nutrition, literacy, social apital, as ell as oi e, hu a se u it, isolatio, dig it, la k of time and subjective well- ei g. 8 4 The Department of Census and Statistics Announces the Official Poverty Line for Sri Lanka (June 2004) 5 Ministry of Finance and Planning, Department of Census and Statistics. Household Income and Expenditure Survey 2012/ MDF s al ulatio s a e ased o ost e e t figu es, i ludi g the o e sio fa to s a d o su e p i e i di es. While the international poverty line has been updated to USD 1.90 per person per day in 2011 Purchasing Power Parity (PPP), many studies have yet to update their calculations to this new line. As such, this study has chosen to remain with the 2005 line so that it can be compared with the existing evidence base. 6 7 Alternative measures of poverty have been proposed in the international development community, most notably the Human Development Index (HDI) which aggregates life expectancy, education and income, and the Multidimensional Poverty Index (MPI), which is comprised of health (malnutrition and child mortality), education (years of schooling and school enrollment) and living standards (cooking fuel, toilet, water, electricity, floor and assets). While these alternative measures better recognise the multidimensional nature of poverty, they nonetheless are forced to collapse complex dimensions into a single index for the sake of comparability and consistency (see Ravallion, M Poor, or Just Feeling Poor? On Using Subjective Data in Measuring Poverty). Furthermore, they can present valuation challenges in that certain indicators are weighted as more important than others, with limited justification as to why they are so. For e a ple, the HDI s eight o s hooli g is fa highe tha the eight pla ed o life e pe ta. I essence, any quantitative measure to define poverty is reductionist, and involves value judgements as to what constitutes a good or bad quality of life to determine whether a person is poor. Moreover, all measurement indices often overlook the relations of power that underpin poverty both at the intrahousehold level (e.g. gender and generational) and at the society-wide level (e.g. ethnic and religious). 8 ODI, Understanding Poverty and Wellbeing: A Note with Implications for Research and Policy Poverty Analysis Discussion Group. 4 Market Development Facility

17 Figure 2: District Official Poverty Lines 2016 I e og isi g the i o e ie t o ple it 9 of poverty, the team utilised qualitative research methods to explore the various dimensions of poverty, including how power, relationships and society produce the phenomenon that is poverty. The Study focused on gaining a more nuanced perspective on household economic circumstances, including how and who within the households manage their activities and income, the dynamics among members, and the risks and uncertainties that they face. Conceptual Framework Wo e s E o o i E po er e t WEE is e t al to MDF s po tfolio. The e is compelling global evidence that when women participate in the labour force, they become powerful agents of economic growth and poverty reduction. The International Center for Research on Women (ICRW) defines a woman as economically empowered he she has both the ability to succeed and advance economically and the power to make and act o de isio s. To su eed a d ad a e e o o i all, o e eed the skills a d esou es to compete in markets, as well as fair and equal access to institutions. To have the power and agency to benefit from economic activities, women need to have the ability to make and act on decisions and o t ol esou es a d p ofits. 10 A ess a d age a e i te o e ted a d oth a e e essa strategic framework for WEE can be found on its website11. to a hie e e po e e t. MDF s Norton, A., The shifting dynamics of poverty how well do we research poverty and wellbeing in a changing world? I te atio al Ce te fo Resea h o Wo e, Measu i g Wo e s E o o i E po e e t Defi itio, F a e o k a d I di ato s. Market Development Facility. Wo e s E o omic Empowerment: How Women Contribute to and Benefit from Growth Strategic Guidance Note Market Development Facility 5

18 In market systems, barriers and opportunities for WEE exist in multiple sub-systems in the household, community, workplace, supporting institutions, and the enabling environment.12 Both fo al ules a d i fo al ules su h as so ial o s, a i te a t to suppo t o u de i e o e s e po e e t. U de sta di g the fa to s that affe t o e s a ess to se i es a d a kets a d their decision-making power is important in contributing to faster and more inclusive growth. For MDF s po tfolio i S i La ka, su h u de sta di g a assist i dete i i g ho esou es a est e allocated and how to design inclusive interventions that are both realistic and practical. Research Design and Methodology MDF employed a mixed methods approach to the Study involving secondary research and extensive field-based primary research. Secondary research was first undertaken to get an understanding of poverty, gender and ethnicity dynamics within Sri Lanka. The primary data collection was u de take to aptu e the ua es of the po e t a d ge de d a i s ithi MDF s SEAs. While the primary research was qualitative in nature, significant quantitative elements were incorporated to establish poverty levels across the sample. Secondary research was used to contextualise, expand upon and triangulate findings from this primary data. As the primary research was non-representative in nature, the secondary research provided broader and better statistical evidence. Overall, the combined use of both quantitative and qualitative elements, as well as primary and secondary research allowed for one approach to address the limitations and enrich the findings of the others. This section describes the research process in greater detail, including the field research team composition, sampling strategy, research tools used, and the analytical process. Research Questions The study aimed to answer the following central research questions: Who are the poor within the Strategic Engagement Area? What are the ethnic dynamics within the Strategic Engagement Area? Why are they poor? What are suitable pathways out of poverty within each Strategic Engagement Area? Where are women represented within each Strategic Engagement Area? Which economic roles do women play? What particular constraints do women face? In what manner can economic activities within the strategic engagement areas contribute to g eate o e s e o o i e po e e t? Ma kel, E. a d Jo es, L., Pushi g the F o tie s of Wo e s Economic Empowerment in Market Systems Development Leveraging Economic Opportunities. USAID Market Development Facility

19 Secondary Research Literature Review Secondary research involved reviewing literature from relevant and credible organisations including the World Bank, UNDP the Sri Lanka Department of Statistics among others, and, as much as possible, collecting the most current data. Data compiled by the Department of Census and Statistics Sri Lanka was also used for analysis. All references to secondary research are noted in footnotes, wherever appropriate. This information provided as complete a picture as possible for each region. Primary Research Field-based Data Collection The main research questions posed are as presented above. More detailed research questions were developed for the field research conducted by a 12-person MDF team. These differed depending on the respective SEA in question, but all households included at least the following in Table 1. Table 1: Main Research Questions Question Category Question Category Details General household characteristics Family size, ethnicity, religion, location, educational background. Progress out of Poverty Index (PPI) Indicators Ten indictors from the PPI for Sri Lanka. These are unique to each country and based on a statisti all sig ifi a t o i atio of data f o S i La ka s HIES hi h, i this ase, i luded whether anyone in the household was engaged in a government job, the education level of the female household member and the total number of household members; whether the household as i possessio of a fa, TV/DVD, f idge o ooke ; the fa il s a ess to transport (bicycle, motorbike or car); and the construction materials used in their house. Income All types of income, both monetary and non-monetary sources; details of paid employment income including FTE jobs; details of farming activities, yields, sales and income; details of government and other support; and opportunities and constraints faced in increasing income. Asset ownership Property and land ownership; various forms of asset ownership; and how the household determines what assets to buy and what to sell. Expenditure patterns Major household expenditure, and management and decisions on expenditures. Wo e s e o o i roles and decisionmaking Wo e s oles i de isio -making regarding household expenditures, including major purchases (e.g. of assets); opportunities and constraints faced by women in participating in SEAs; division of labour between men and women, in economic and household activities, and in paid and unpaid labour; and control over mobility and time use. Field Research Teams Primary research was conducted in the field by a number of two-three person teams made up of MDF staff and supported by consultant Laura Kim from MarketShare Associates who also conducted the secondary research and led the analysis. She was assisted by Erin Markel from MarketShare Associates. The team members are listed in Table 2. Market Development Facility 7

20 Table 2: Research Team Members Primary Research Team Laura Kim Short Term Advisor, MarketShare Associates Victoria Carter MDF Fa ilit Wo e s E o o i E po e Mohammad Shahroz Jalil Deputy Team Leader and Country Representative, MDF Sri Lanka Paul Valemei Senior Results Measurement Specialist, MDF Fiji Minal Cabraal Business Adviser, MDF Sri Lanka Praveen Premkumar Business Adviser, MDF Sri Lanka Hashim Nazahim Business Adviser, MDF Sri Lanka Vijayalaya Karunanithie Business Adviser, MDF Sri Lanka Vindika Lokunarangodage Business Adviser, MDF Sri Lanka Ahmed Safiyulla Business Adviser, MDF Sri Lanka Shabana Ibrahim Business Adviser, MDF Sri Lanka Reneera Paul Business Adviser, MDF Sri Lanka e t a d Results Measurement Manager Sampling Strategy The primary research was qualitative, so purposeful sampling was utilised. The sampling process was done in two stages. Firstly, the Study Team focused on ensuring that the sample group in each SEA was deemed poor (Category 1 and Category 2) or near-poor/vulnerable (Category 3) by official (World Bank and Department of Census and Statistics) standards. Better-off households (Category 4), however, were not excluded from the sample as the Study attempted to be comparative. Further explanation of the poverty categories used is given in Table 3. Table 3: Poverty Categories Used in the Study Poverty levels Category 1 Below 100% international poverty line (USD 1.25/day 2005 PPP) Households that are classified as poor by global standards. The latest data for Sri Lanka is for 2011 shows that 4.1 percent of the population fell below the international poverty line of USD 1.25 per day in 2005 prices. This figure is equivalent to LKR 59.8 per person per day. Category 2 At or below 100% national poverty line Households that are classified as poor by Sri Lankan national standards, and comprise the bottom 6.7 percent of the population. The poverty line has been calculated as LKR per person per day. Category 3 Above 100% national poverty line but below 200% national poverty line Households that are not classified as poor by national official standards, but are classified as near-poor or vulnerable as they subsist on less than twice the poverty line. They are characterised by a high propensity to external shocks and/or high variability of income streams. They comprise the bottom 40th percentile of the population. The 200% poverty line has been calculated as per person per day. Category 4 Above 200% national poverty line Households that are neither classified as poor nor near-poor/vulnerable. They comprise, approximately, the highest 60th percentile of the population in terms of wealth. Secondly, MDF built heavily upon the knowledge of worker and household profiles gained during its initial scoping missions and strategy development to inform the selection of the types of respondents. The strategy was to ensure that the respondents were as heterogeneous as possible in terms of work profile, gender, ethnicity and region. In particular, the stimulating entrepreneurship 8 Market Development Facility

21 in former conflict-affected areas (SEA 3) sample looked specifically at the impact of conflict on the economic profile of these households. The team wanted to ensure samples included all ethnic and religious groups, although this tended to be sought along geographic lines (e.g. larger Tamil samples were taken during the SEA of stimulating entrepreneurship in former conflict-affected areas component of the research). Although questions regarding WEE were included in almost all interviews, the study wanted to ensure that both women household members and women directly involved in the SEA were taken into account. For these reasons, there was a dedicated effort to identify a higher sample of female workers in respective SEAs than would be representative. The details of who was interviewed in which region per SEA are presented in Table 4 as follows. The details of the final sample are presented in Annex 1. Table 4: Details of Interviews according to SEAs, Geographic Areas and Respondent Profiles SEA : Di ersifi atio of Sri La ka s touris desti atio s, produ ts a d ser i es Tourism worker and household interviews should include: Large, medium and small hotel workers (and homestays), including specialist/boutique providers to backpacker and emerging tourism markets. Tourism activity and transport providers, including tour guides, car/tuk-tuk for tourism operators, workers in small tourism activity businesses, especially in new tourism markets such as surfing and adventure travel. Souvenir and other merchandise sellers Food and beverage providers catering to tourists, including restaurant workers and snack/kiosk owners. Handicraft producers including handloom weavers, pottery makers, and traditional wooden mask makers. Geographic areas covered should include: Kandy (captures hotel workers in an established tourism area, from high-end to guesthouse) and Polonnaruwa (captures historical site workers). Arugam Bay and Mirrisa (captures workers and SME tourism businesses catering to new backpacker markets). Galle (captures workers and SMEs in an established beachside tourism location and crafts producers making pottery and masks). Maruthamunai (captures handloom producers and weavers). Ambalangoda (captures a mask producer). Trincomalee and Batticaloa (captures tourism workers and SMEs in former conflict-affected areas). SEA 2: Improving Sri Lankan produce and products for export and tourism markets Worker and household interviews should include: Improving Sri Lankan produce and products encompasses a number of different sub-sectors, namely horticulture and inland and coastal fisheries. It initially intended to cover spices, but this was excluded as it was discovered that spice farmers are markedly above the poverty lines set out above. In horticulture, both better and less well-off horticulture farmers were interviewed, with different land sizes, and within different climatic regions, and also included (to a lesser extent) agricultural retailers and nursery workers. In fisheries both reservoir fishermen and (to a lesser extent) coastal fishermen and workers of fish processing factories were interviewed. In spices, a spice farmer and a worker in a spice garden were interviewed. As it was discovered that spice households are often above the poverty line, these households were not separated out for a spice analysis and are instead considered in terms of their agricultural/horticultural/tourism characteristics. Market Development Facility 9

22 Geographic areas covered should include: Galle (captures horticulture nursery workers, vegetable farmers). Polonnaruwa (captures reservoir fishermen and women). Moneragala (captures fruit and vegetable farmers and farm labourers). Kandy (captures spice farmers). SEA 3: Stimulating entrepreneurship and business activity in former conflict-affected areas Worker and household interviews should include: Small, micro and medium enterprises (SMMEs) although there was a significant amount of overlap with SEAs 1 and 2 (diversification of S i La ka s tou is desti atio s, p odu ts a d se i es; a d i p o i g S i La ka p odu e and products). SMMEs involved in food and other small scale processing/production, including their workers. The sample included women-led enterprises and, in particular, war widow-led enterprises and Internally Displaced Persons (IDP) in camps and resettled individuals. Geographical areas covered should include: Mannar and Jaffna (for fishermen and women and fish processing facility workers). Batticaloa, Jaffna and Trincomalee (for horticulture farmers and labourers). Jaffna, Mullaitivu and Trincomalee (for micro and SME producers and workers). SEA : I pro i g Sri La ka s digital apa ilit Worker and household interviews should include: Small IT businesses and their workers. Geographical areas covered should include: Colombo and Kandy (for small IT businesses and their workers) as these are the central IT hubs. Tools Used for Primary Research Samples of the tools used during this study can be found in Annex 2. These included household questionnaires, key informant interviews (KIIs), and focus group discussions (FGDs) conducted in rural and urban settings. Details of the design of the tools used are presented in the table below. Table 5: Primary Research Tools Household Questionnaire The household questionnaire consisted of a semi-structured interview with open ended and close-ended questions. Its aim was to explore the key research questions for poverty and gender identified above. Open-ended questions allowed for exploration of new ideas o a eas, a d d illi g do o i te ie ees espo ses. The po e t se tio of the uestio ai e was aimed at: a) the male and female heads of households (together where possible) followed by a gender section that was divided into questions to which the female heads were to respond separately; and b) the male or female worker, with questions relating to their households. Members of the field team were trained and familiarised with the interview guide before the field research began at a training workshop, th ough use of the tool, a d du i g dail de iefs of the da s findings where adjustments to the questionnaire were made. Interviews were conducted in the local language and dialect, although the guide and note-taking was predominantly done in English. Examples of the Household Questionnaire used can be found in Annex 2. Focus Group Discussions FGDs were used along with the household questionnaire. FGDs were held with men and women in rural and urban settings in the four SEAs to provide additional information on poverty and gender and to validate and explore the information provided in the household interviews. These were based around themes identified from the household interviews 10 Market Development Facility

23 o du ted MDF, as ell as the MDF pilla s fo WEE: o e s oles and controls including issues of access, decisionmaking and workload. An example of the FGD template used can be found in Annex 2. Key Informant Interviews Key informant interviews (KIIs) were used to triangulate the information being gathered in the Household Questionnaire a d the FGDs. These e e sele ted ased o a o i atio of the espo de t s positio, lo ge it a d k o ledge i relation to the four SEAs. KIIs gave a broader overview of the involvement of men and women, including e.g. views on employment from the point of view of those involved in hiring, and trends over time (e.g. one key informant had been involved in the spice sector for over 19 years). A complete list of the KIIs undertaken can be found in Annex 1. Box 2: References to Primary Research Data Management, Collation and Analysis How interviews are referenced in the study: Primary research is referenced in footnotes throughout the document. Codes are assigned to all household interviews, key informant interviews and focus group discussions. The following codes are used to refer to interviews: T-HH-XX refers to interviews conducted under the Tourism SEA. HH refers to household interviews. All i te ie s a e u e ed, ep ese ted XX i the ode. Ke i fo a t interviews and FGDs are referred as T-KII-XX and T-FGD-XX respectively. SLMG-HH-XX refers to household interviews conducted under the Sri Lankan Produce and Products SEA. This code also covers interviews for the crafts aspect of tourism. BILR-HH-XX refers to household interviews conducted in former conflict-affected areas. DC-HH-XX refers to household interviews conducted under Innovation in Digital Services SEA. All codes cited can be corresponded against the list of full sample, presented in Annex 1. Each household interview and FGD was conducted by a team of two MDF business advisers, in which one person acted as the interviewer or facilitator, and the other person acted as the note taker. For the household interviews and FGDs, all primary data were recorded on the questionnaires. Afterwards, the team undertook a joint writing exercise to record information and observations from the interview or discussion. KIIs were often conducted by one person who did a writing and analysis exercise to reflect on the session. The summary notes were then used for a detailed and nuanced qualitative analysis of respondents. The details are provided in the table below: Table 6: Primary Research Data Management Household Summaries The primary research teams prepared a narrative poverty and gender summary from each household interview. A tabular, but narrative format (see Annex 2 for a sample) guided this process supporting the teams to reflect on the interviews and to report back on numeric information as well as observations of the village, household and individuals within the household which are included in the poverty analysis for each region below. One of these summaries are included for each SEA (four in total) in Annex 3 with identifying details of respondents redacted. Market Development Facility 11

24 FGD Summaries FGD notes were incorporated into a FGD summary template. This allowed for further qualitative analysis regarding sectors a d o e s oles i the SEAs, people s halle ges a d oppo tu ities, potential solutions, and issues around decisionmaking, and control over income and workloads. Information from the FGDs was used to deepen understanding of people s, pa ti ula l o e s, situatio s, a d to t ia gulate, la if o o t adi t i fo atio p ovided by the head of the household. This in turn led to a deeper review and discussion around household activities and dynamics. KII Summaries KII summaries were write-ups of the notes during these interviews. The content of these was then analysed in alongside the FGD summaries, to deepen, strengthen and understand the data sets collected. Given the breadth of the quantitative data collected, collation of multiple indicators and figures was required to provide a systematic and easily digestible analysis of findings. Details are as follows: Table 7: Primary Research Data Collation Household Quantitative Data Responses from the interviews were collated into spreadsheets related to: SEA specific data entry; SEA specific data summary; and summarised study data summary. The data summary sheets provided calculations such as net income from specific agricultural activities minus cost of production and sales, total income from all sources, number of children in the household 13 years of age and under, and livestock numbers among others. This collation offered a detailed look at individual sector findings (e.g. dynamics of fish prices) through to an overview of the SEAs and overall sample (e.g. an overview of food pricing across all households) via easy to access quantitative data per household and/or worker and across a region or SEA. The SEA-specific and summarised Study spreadsheets collated data across households to get a sense of ranges and averages (while recognising that this information is not statistically significant due to the small sample size, ut suggesti e of the households e o o i situatio. This i luded data from each of the households or workers. The goal as to i di ate the le el of a household s i o e i elatio to othe households. Co su ptio of o p odu tio, pensions and remittances, non-farm income, and gifts were all included in income calculations to get a reasonably comparative figure from one household to the next within each region. Household Progress out of Poverty Index Data was collected according to the PPI that is used as a proxy of poverty likelihood in Sri Lanka. This served as an additio al i di ato o household po e t le els a d o t i uted to dis ussio a d a al sis of i di idual households situations. For example, if a house had a low PPI but relatively higher per capita income, the Study Team could review that household and discuss the variables. Given the outdated nature of the PPI (which often happens when the questions are technology-related, as is the case for Sri Lanka) this proxy was not used in the analysis. Limitations of the Research The research was not without limitations, which are described below. Despite these limitations, a significant quantity and quality of information was able to be collected that has been useful in understanding the nature of poverty and gender in the targeted regions and SEAs. General Limitations While as much detailed information on income, expenditure and assets was gathered, the Team understood that the majority of respondents interviewed did not keep written records and relied on memory to answer questions, leading to recall bias. For example, farm expenses and income could 12 Market Development Facility

25 be complicated as respondents tried to remember costs associated with a specific crop that may be only grown for part of the year. The Stud Tea sought to o k a ou d espo de ts a aila ilit a d to a oid dis upting or putting pressure on them due to the limited time the Study Team had to explore each SEA. Time constraints also affe ted the tea s a ilit to do thei a ati e su a ies a d a al ses, hi h also led to e all bias. It was not always easy to assess from the beginning of an interview what the quality of that interview would be. In some interviews the value of the information received was of concern. This was particularly the case when gathering information on WEE as the male head of household was often present when the WEE component was being administered to the female head of household, and his presence may have inhibited candid responses from the woman. While the Team tried to minimise these scenarios as much as possible, it was often difficult to maintain privacy in small households. These factors were taken into consideration in the analysis, and interviewers were trained to take note of such observations. A higher sample of female workers and households was taken in the respective SEAs than would statisti all e e pe ted to e take. This as i te tio al to e su e o e s u e t a d pote tial future involvement in the SEAs was properly understood, thus the Study was intentionally skewed towards female households. Specific Strategic Engagement Area Limitations In households involved in generating agricultural incomes (including horticulture households, but also other households where agriculture forms part of the overall income stream), crop yields to measure farm productivity was an important factor in understanding why rural households were poor or vulnerable. However, as with sales and income, reliable and consistent information on productivity or yields was difficult to collect due to: a) households not keeping records on this information and relying on memory b) yield fluctuations from year to year, and many harvested crops being used for home consumption or as partial payments for on-farm employed labour; and c) crops being harvested at irregular intervals for sales, making it difficult to estimate the total seasonal yield. SEA : i p o i g S i La ka s digital apa ilit p o ed diffi ult to e te due to st i t a a ge e ts around confidentiality. It was discovered during the Study that most workers in this industry were required to sign non-disclosure agreements as part of their employment or services contracts. This meant the sample size was smaller than originally anticipated. If necessary, the team will undertake further household research when its relationships with this sector are stronger and there may be some relaxing of confidentiality restrictions. Organisation of the Study Chapter 2 offers an overview of poverty and gender in Sri Lanka, drawing from statistical data and secondary research, and offers context for the primary research that follows, considering dimensions of poverty, gender and a history of conflict in Sri Lanka. Market Development Facility 13

26 Chapters 3-6 of this report address the four SEAs identified as focus areas for Sri Lanka. They are organised as follows: The scope of the SEA is described for all sub-sectors that it addresses A section on description of the methodology and the characteristics of the respondents Analysis of primary and secondary research to describe the poor and their specific characteristics A al sis a d e o Roles and specific constraints faced by women are underlined Finally, recommendations are given to improve barriers. e datio s fo i te e tio s a e p ese ted i Path a s out of Po e t o e s pa ti ipatio a d to address WEE Chapter 7 provides an overview of the characteristics of poverty in the overall sample and offers o lusio s to guide MDF s futu e o k. 14 Market Development Facility

27 Chapter : O er ie of Po ert a d Ge der i Sri La ka Poverty Reduction and Economic Growth in Sri Lanka Having emerged from three decades of civil conflict, Sri Lanka has made significant strides in poverty reduction and economic growth. Data from the HIES indicated that the national poverty headcount rate fell from about 22.7 percent in 2002 to 6.7 percent in 2012/13.13 The island nation outpaced other South Asian countries in the 2014 Human Development Index, ranking 73 out of 188 countries, positioning the country in the high human development category.14 It also met most targets for the Millennium Development Goals (MDGs), including achieving the target for halving poverty seven years before the 2015 deadline. S i La ka s e o d i po e t edu tio has la gel ee d i e se e al o ple e ta fa to s: its gradual transformation from a predominantly agriculture economy towards a more service-driven economy; 2) urbanisation and agglomeration; 3) increased domestic consumption and 4) an increase in international prices of rice and tea, which raised agricultural wages.15 This economic shift was evidenced by the annual growth rate of industrial and service sectors, whereas the agriculture sector accounts for a declining share of GDP. In 2015, the service sector a ou ted fo. pe e t of the ou t s GDP, followed by the industrial sector at 26.2 percent and agriculture at 7.9 percent.16 While the agriculture sector also experienced growth over the last few years, it was mainly due to higher international prices of rice and tea between 2006 and The subsequent fall in prices is unlikely to sustain agricultural wages, and has contributed to an overall performance slowdown in the agriculture sector in At the same time, the agriculture sector continues to employ 28.7 percent of the population19 and thus remains critical to the economy and poverty reduction. While S i La ka s ad a es a e i p essi e o the atio al le el, the ask po kets of po e t a d vulnerability that are pronounced along gender, geographical, ethnic and sectoral lines. Poor women and men, with few assets and inadequate access to markets, are often hardest hit in economic shocks. But even the near-poor women and men who are no longer classified as poor but are below the 40th percentile are only slightly better off. They subsist on less than twice the poverty 13 World Bank, Global Poverty Working Group. Data are compiled from official government sources or are computed by World Bank staff using national (i.e. country specific) poverty lines. 14 UNDP. Human Development Report Newhouse, et al, Sri Lanka - Poverty and Welfare Central Bank of Sri Lanka. Economic and Social Statistics of Sri Lanka Newhouse, et al, Sri Lanka - Poverty and Welfare Ministry of Finance, Sri Lanka. Annual Report Central Bank of Sri Lanka. Economic and Social Statistics Market Development Facility 15

28 line and are similar to people categorized as poor in terms of consumption patterns and asset ownership.20 On the other hand, the richest 20 percent of Sri Lankan households enjoy 52.9 percent of the ou t s atio al i o e.21 Such inequality is reflected in the stagnant Gini coefficient, which fell from 0.41 in 2002 to 0.37 in 2009/10, but rose back to 0.40 in 2012/2013 (based on HIES data). There is a wealth of evidence that shows that structural inequalities lead to wasted productive potential in the long run, and attention to equity is critical for sustained growth and eventual graduation into upper middle-income country status. With 40 percent of the population highly vulnerable to shocks, the important challenge for Sri Lanka is to ensure inclusive growth, particularly as it gears itself to the meet the targets of the Sustainable Development Goals (SDGs) and its own ambitious development vision. Impact of Civil War on Poverty Sri Lanka is a multi-ethnic country with an estimated population of 20.3 million people.22 The three major ethnic groups are the Sinhalese, the Sri Lankan Tamils and the Moors. The Sinhalese, who are predominantly Buddhist, comprise approximately 75 percent of the population. The Sri Lankan Tamils are the largest ethnic minority comprising around 11 percent of the population. They are concentrated in the northern and eastern portions of the island. Approximately 78 percent are Hindus and 20 percent Christians. The Moors comprise 9 percent of the population and are almost exclusively Muslim. The largest concentration of Moors is found in the Eastern Province. Other minority groups include the Indian Tamils, Malay, Burghers, and others. From 1983 to 2009, the country was ravaged by a civil war between the Government of Sri Lanka a d the Li e atio Tige s of Ta il Eela LTTE ho a ted a i depe de t state fo the ou t s Tamil minority. No official figures exist, however it is believed that up to 100,000 people died as a result of the conflict.23 The northern and eastern provinces, which served as the main theatre for the war, suffered the most through the destruction of infrastructure and livelihoods due to a lack of security and investment. In the last few years of the war ( ) the Government of Sri Lanka launched a military offensive against the LTTE and more than 300,000 people were displaced.24 Many families and communities experienced extreme loss due to the conflict, and gender dynamics were greatly affected. In a survey undertaken in war-torn regions; 80 percent of women were either 20 Newhouse, et al, Sri Lanka - Poverty and Welfare Central Bank of Sri Lanka. Economic and Social Statistics of Sri Lanka Census of Population and Housing 23 Al Jazeera. Sri Lanka Starts Count of Civil War Dead Saparamadu, C. and Lall, A., Resettlement of the conflict-induced IDPs in Northern Sri Lanka: Political Economy of State Policy and Practice Working Paper 10: Secure Livelihoods Research Consortium. Center for Poverty Analysis (CEPA) Market Development Facility

29 widowed, unmarried, divorced or living separately from their husbands. Among them, 70 percent were heading households and the majority had no jobs.25 I additio to its hu a osts, a da ages a ou t s e o o in multiple other ways through the destruction of resources, diversion of public expenditures toward combat, disruption of social order, dissaving and portfolio substitution.26 While some costs are fairly straightforward to measure, such as those due to destruction or increased government spending, other costs such as death, injury, changes in social dynamics, capital flight or reduced tourism pose significant measurement challenges. As such, there is little agreement on the relevance or reliability of particular variables. Nonetheless, estimates are useful in providing a general idea of the economic impact of conflict. One study used a sample of 92 countries to estimate a reduction in GDP growth rate by 2.2 percent per civil war year. Another study of 211 countries also had similar estimates: a five-year war reduces the average growth rate by 2.4 percent per year.27 Table 8: Empirical Studies of the Cost of Conflict in Sri Lanka Author Year of Publication Period Observed Cost of Conflict Richardson and Samarasinghe % of GDP 1988 Grobar and Gnanaselvam % of GDP 1988 Harris 1997, % of GDP 1982 Kelegama % of GDP 1995 Arunatilake, Jayasuriya, and Kelegama %, 168%, 205% of GDP 1996 (at different interest rates of %) Ganegodage and Rambaldi % annual GDP loss Significant efforts have been made to analyse data on the Sri Lankan conflict. One estimate placed the economic cost of the co fli t f o to at t i e S i La ka s GDP usi g the ost 28 conservative figures. The most recent evidence showed that the impact of war during the period amounted to a negative annual loss of 9 percent of GDP.29 Following conflict, rapid and high growth rates are not unusual, given the low base and therefore high etu s to i est e ts du i g e o e. I deed, i ea s i ediatel afte the a s e d, nominal GDP growth for the Northern and Eastern provinces surged. In 2011, growth rates for the 25 Survey conducted by the Jaffna-based Center for Women and Development, as cited by AWAW. S i La ka Wo e s Age da o Pea e, Security and Development Pages Collier, P., On the Economic Consequences of Civil War Oxford Economic Papers 51: Hoeffler, A. and Reynal-Querol, M., Measuring the costs of conflict Washington, DC: World Bank. 28 Arunatilake, N., Jayasuriya, S. and Kelegama, S., The Economic Cost of the War in Sri Lanka In World Development, 29(9). 29 Ganegodage, K. R. and Rambaldi, A. N., Economic consequences of war: evidence from Sri Lanka Journal of Asian Economics. Market Development Facility 17

30 provinces were reported at 27.1 and 12.4, respectively.30 In 2012, growth rates were at 25.9 and 25.0 percent, respectively Central Bank of Sri Lanka. Economic and Social Statistics of Sri Lanka Central Bank of Sri Lanka. Economic and Social Statistics of Sri Lanka Market Development Facility

31 Figure 3: Provincial Share of GDP ( ) Source: Central Bank of Sri Lanka (2016). Sri Lanka Socioeconomic Data 2016 While the effects of conflict may even out at the macroeconomic level through consistently high growth rates, microeconomic studies show that economic and poverty traps persist for the affected populations. Such legacies of war remain visible in the northern and eastern provinces. While growth rates in these provinces have remained impressive, they nonetheless remain among the lowest contributors to overall GDP (Figure 3). Furthermore, poverty incidences are higher if compared to the overall country. In 2012/13, the Northern and Eastern provinces reported a 10.9 percent and 11.0 percent poverty rate respectively, as compared to a nationwide average of 6.7 percent.32 Blattman and Miguel have stressed that the most crucial questions about conflict are not whether it harms the e o o, ut athe i hat a s, ho u h, fo ho, a d ho pe siste tl Central Bank of Sri Lanka. Economic and Social Statistics of Sri Lanka Blattman and Miguel, Civil War National Bureau of Economic Research. Working Paper Series, Market Development Facility 19

32 Features of Poverty in Sri Lanka Poverty in Sri Lanka is characterised by spatial diversity. As evidenced in the 2012/13 poverty map in Figure 4, poverty rates in Uva, Eastern and Northern provinces are significantly higher than the rest of the country. In particular, the poverty indices for certain districts are particularly high, including in Mullaitivu (28.8 percent), Mannar (20.1 percent) and Kilinochchi (12.7 percent) districts in the Northern Province, Batticaloa (19.4 percent) in the Eastern Province, and Moneragala (20.8 percent) in Uva Province. Figure 4: Poverty incidence by district (2012/2013) That being said, high poverty rates do not necessarily translate to high populations of poverty. For example, while Mullaitivu and Ma a dist i ts po e t ates a e e e among the highest in the country, they collectively only accounted for 3.4 percent of poor people in the country. On the other Source: Sri Lanka Department of Census and Statistics and the World Bank (2015). Spatial Distribution of Poverty hand, districts such as Ratnapura in Sabagarawuma Province and Galle in the Southern Province may exhibit lower poverty rates, but they represented a large number of people living under the poverty line. Poverty in Sri Lanka is also characterised by its political-administrative divisions, comprised of the urban, rural and estate sectors. Figures from the 2012 census indicated that the urban sector, or areas governed by either municipal or urban councils, comprised 18.8 percent of the population. The urban sector remained lowest in terms of the poverty headcount index, having cut it from 7.9 percent in 2002 to about 2 percent in 2012/ Mu h of this is ot su p isi g, gi e that u h of S i La ka s e o o i activity is concentrated in the capital city of Colombo and surrounding areas in the Western Province. That being said, there are sizable populations living in settlements that may exhibit urban characteristics but are officially defined as rural areas, with some estimates suggesting that as much as one-third of its entire population may be living in unrecognised urban settlements. The rural sector comprises the majority of the population at 77.3 percent.35 While the poverty rate is 6.8 percent, the rural sector accounts for over 85 percent of poor Sri Lankans nationwide.36 As such, u h of the ou t s edu tio i e t e e po e t an be attributed to reductions in rural poverty. 34 Ellis, P. and Roberts, M., Leveraging Urbanization in South Asia: Managing Spatial Transformation for Prosperity and Livability Washington, DC: World Bank 35 Department of Census and Statistics Census of Population and Housing. 36 Newhouse, et al, Sri Lanka - Poverty and Welfare Market Development Facility

33 The estate sector i.e. plantation areas of more than 20 acres and more than 10 residential labourers with a single administration body, comprised only 4.4 percent of the population. However, the estate sector had a poverty incidence of 10 percent and housed some of the most marginalised groups in the country. The estate sector came into being in the British colonial period, during which indentured labourers were brought from South India to grow tea, rubber and coconut.37 Estates are primarily located in the central and southern portions of the country, with Central Province comprising the largest share of estates (at 18.9 percent), followed by Uva Province at 12.8 percent.38 There are no estate populations in the Northern and Eastern provinces. Finally, the poor are disproportionately concentrated in the agriculture sector; half of the poor and almost half of the near-poor are employed in this sector.39 The majority are small-scale farmers who face low productivity and economic returns, and remain highly vulnerable to climate change.40 Overview of Gender in Sri Lanka Women make up the majority of the Sri Lanka population, accounting for 51.6 percent of the population in the 2012 National Census. Their status in society and economy provides a mixed picture; although Sri Lanka has made many advancements toward gender equality, women continue to face systematic disadvantages due to inequitable social norms in formal and informal institutions which creates barriers to further empowerment. Health and education data is promising with regards to gender equality. Sri Lankan girls are more likely than boys to attend and complete primary and secondary schools.41 Sri Lanka reached gender parity in primary education, with the ratio of girls to boys in education reaching 99.4 percent in 2009/10.42 Maternal mortality rates also continue to drop, significantly from 43 deaths per 100,000 live births in 2005 to 30 deaths per 100,000 live births in Women enjoy improved standards of health and education, but this has not translated into i p o ed o e s la ou fo e a d politi al participation. Female labour force participation has o siste tl e ai ed stag a t at a ou d. pe e t, as o pa ed to e s pa ti ipatio ates of 74.6 percent (see Figure 6).44 Political participation, as represented by the proportion of women in 37 Center for Poverty Analysis. Moving Out of Poverty in the Estate Sector in Sri Lanka: Understanding Growth and Freedom from the Bottom Up Department of Census and Statistics Census of Population and Housing. 39 Newhouse, et al, Sri Lanka - Poverty and Welfare Menike, L. and Arachchi, K., Adaptation to Climate Change by Smallholder Farmers in Rural Communities: Evidence from Sri Lanka International Conference of Sabaragamuwa University of Sri Lanka Asian Development Bank. Sri Lanka Country Gender Assessment Government of Sri Lanka and the United Nations in Sri Lanka. Sri Lanka Millennium Development Goals Country Report WHO, UNICEF, UNFPA, World Bank Group and UNPD. Trends in estimates of maternal mortality ratio (MMR; maternal deaths per 100,000 live births) Sri Lanka Department of Census and Statistics. Labour Force Survey Market Development Facility 21

34 national parliaments, has also remained stagnant at about 5.8 percent, as compared to 12.0 percent for India, 20 percent for Bangladesh, and 20.6 percent for Pakistan.45 The 2015 Global Gender Gap Index46 ranks Sri Lanka 85 out of 145 countries. The index ranks countries based on key demographic and economic indicators, namely, economic participation and opportunity, educational attainment, health and survival and political empowerment. In fact, Sri Lanka forms 6 percent of countries that have widened their gender gaps over the past ten years, and is one of two countries in which that gap has widened by more than 1 percent. The 2014 Human Development Report ranked Sri Lanka similarly at 72 out of 155 countries in its Gender Inequality Index based on reproductive health, empowerment and e o o i a ti it i di ato s. S i La ka s rankings would seem to indicate that they were disproportionately affected by its scores on political and economic participation and opportunity. Recent analytical work done by MDF on WEE shed light on the key factors that compromise the level of o e s a ess to se i es a d a kets. Fi stl, as is the t e d o ld ide, o e s oles at home (and especially as related to childcare responsibilities) restrict their opportunities to participate in the labour market. According to the Labour Force Survey of , about 62.8 percent of o e ho did ot pa ti ipate i the la ou fo e ited e gaged i house o k as the ai reason for non-participation as compared to 6.8 percent of men a figure that is consistent with earlier years. Second, although women are less likely to participate in the labour market, when they look for work, they are more likely than men to be unemployed or underemployed. According to the 2014 Labour Force Survey, women have an overall unemployment rate of 6.5 as compared to 3.1 percent of men, and an overall underemployment rate of 3.4 as compared to 2.4 of men. Furthermore, women earn less than men earning 84 percent of male wages in 2012, and the gap cannot be explained by differences in observable characteristics or types of jobs.48 Third, women face systematic discrimination in a working environment that is not conducive to their empowerment. In paid employment, labour laws prevent women from taking up night work or parttime work in the growing service sector,49 and the laws governing maternity benefits require employers to bear the full cost which is a disincentive for them in hiring women.50 This is compounded by a lack of family-friendly policies in private firms which has encouraged women to seek work in the public sector where they receive relatively better gender parity in earnings and 45 Inter-Parliamentary Union. Statistical Archive for September The Global Gender Index was designed to measure gender equality and forms part of the Global Gender Report published annually by the World Economic Forum. The highest possible score is 1 (equality or better for women, except for lifespan, 106% or better for women) and the lowest possible score is Sri Lanka Department of Census and Statistics. Labour Force Survey World Bank Sri Lanka - Ending poverty and promoting shared prosperity: a systematic country diagnostic. Washington, D.C: World Bank Group The Shop and Office Employees Act No. 19 of Maternity Benefits Ordinance No. 32 of 1939 (for women employed in any trade) and the Shop and Office Employees Act No. 19 of 1954 (for women employed in shops and offices) (from Abeykoon et al in World Bank 2015). In 1993, Sri Lanka ratified the ILO Maternity Protection Convention (Revised) 1952 (No. 103), but there remain discrepancies between the national legislation and the convention. 22 Market Development Facility

35 benefits (such as maternity leave). The public sector has shorter days and this allows women to gain a greater balance between paid o k a d household duties a d to ai tai thei pe eptio of espe ta ilit spe di g a so iall a epta le pe iod of ti e o household duties. Ho e e, oth in private and public sectors, only a minority of women have been able to rise through the ranks; they often get stuck in entry-level positions.51 In Sri Lanka, the majority of working women own and operate micro-enterprises,52 often because the work environment as an employee can be inflexible, frustrating or they are unable to break through the glass ceiling. At other times, women are compelled to start their own businesses due to unemployment, lay-offs or lack of marketable skills.53 Evidence has shown that while women microe t ep e eu s fa e ultiple a ess o st ai ts, the i itial fou datio s required to encourage fe ale e t ep e eu ship a e i pla e, i.e. fa ilies a e ge e all suppo ti e of fe ale entrepreneurship and tend to assist in finding physical and human resources.54 Gender-based violence rates in Sri Lanka are also not insignificant. Forty-one percent of Sri Lankan men reported perpetration of at least one emotionally abusive act against female partners, and 31 percent of women reported experiencing physical violence.55 The relationship between genderbased violence and work, is more complex. WEE, if done conscientiously, has been shown to reduce gender-based violence because as a woman gains economic power she garners more respect in her household, thereby decreasing her risk of violence. As will be shown in the course of this study, entrance to the workforce also provides women with additional support mechanisms in terms of their own monetary and human capital, and in the form of new social networks. Overall, prospects of increasing WEE are promising in Sri Lanka, but it needs to be carefully managed, and initiatives must remain sensitive and attuned to normative patterns and cultural values. 51 Asian Development Bank, Deutsche and Gesellschaft für Internationale Zusammenarbeit (GIZ) GmbH, Country gender assessment, Sri Lanka: An update. (Mandaluyong City, Philippines: Asian Development Bank, 2015) 52 It is important to note that Sri Lanka does not have a nationally accepted definition of SMEs. 53 Winn, J., Women Entrepreneurs: Can We Remove the Barriers? International Entrepreneurship and Management Journal Vol 1(3). 54 Attygalle, A. et al. Female Entrpreneurship and the Role of Business Development Services in Promoting Small and Medium Enterprises in Sri Lanka Institute of Policy Studies and Oxfam Sri Lanka Fulu et al. Why Do Some Men Use Violence Against Women and How Can We Prevent It? Quantitative Findings from the United Nations Multi-Country Study on Men and Violence in Asia and the Pacific Market Development Facility 23

36 Chapter : Fi di gs of the Po ert a d Ge der Stud Di ersifi atio of Sri La ka s Touris Tourism is one of the fastest growing sectors of the Sri Lankan economy employing a workforce of around 130,000, 26 percent of whom are women. For the last five years the sector has been growing at an average of 18 percent with the number of tourist arrivals now standing at 1.79 million people per year (it was only 0.44 million in 2009). In 2015, total receipts amounted to LKR 405,492 million or USD 2980 million, with the direct GDP contribution from tourism at 4.8 percent.56 As a result, tourism is considered an engine of growth for the Sri Lankan economy. Despite the scope for tourism throughout Sri Lanka, in terms of landscapes, beaches, wildlife, religious and cultural appeal, the bulk of tourism is limited to traditional hotspots, established circuits and narrowly-focused product baskets. Hence a large population that could benefit from tourism is left out. Even within the tourism-focused areas, poverty and vulnerability are concerns among lowly-paid entry level workers who work for small-scale tourism operators. MDF s o k ill help di e sif the tou ism base and open it up to more areas of the country. In doing so, it will help create sustainable market linkages, new products and new skills which would o t i ute to eati g o e i o e a d jo s fo poo o e a d e i S i La ka. MDF s perspective will not only be economic, but will also focus on the conservation of tourism assets to promote sustainable tourism. F o MDF s s opi g o k i to tou is i, it as fou d that the ajo o st ai ts i pi gi g o this sector were: inadequate integration of hotels and resorts with activity providers, low skills of staff and providers, poor travel information and booking services, and inadequate product development and promotion. In the areas of souvenirs, crafts and fashion sales, Sri Lanka has over 25 categories of products (e.g. Palmyrah tree and reed products, handloom weaving and wooden crafts) and considerable potential to improve product design, packaging and promotion. The government has emphasised tourism and is leading on creating the Sri Lankan destination brand but it is still early days. More can be done to develop tourism in all regions of Sri Lanka rather than focusing only on the traditional hotspots in the central and south provinces. There are potential growth opportunities in service and activity provision, and in enhancing arrival and spending opportunities for tourists on Sri Lankan-made products. MDF is building its tourism portfolio around these areas of focus. Tourism is expected to generate greater employment in the coming years given increasing numbers of visitor arrivals and increased demand for tourism products and services. However, while tourism can reduce poverty in developing countries, it is not universally and automatically inclusive of the poor. In fact, research examining the economic effects of the tourism industry on the poor is limited.57 Different types of tourism can have varying impacts on the poor, both positive and All statistics except contribution to GDP from Sri Lanka Tourist Development Authority. Annual Report Tou is from World Travel and Tourism Council, Travel & Tourism Economic Impact 2015 Sri Lanka Mitchell, J. and Ashley, C., Tourism and Poverty Reduction: Pathways to Prosperity Market Development Facility s GDP contribution

37 negative, so it is important to understand these differences. Through this understanding, MDF can influence the industry to grow in a manner that is sustainable and inclusive of poor men and women. Methodology and Sample Characteristics In order to understand the poverty dimension of the tourism industry in Sri Lanka, the study team conducted 22 in-depth household interviews across the Central, North-Central, Eastern and Southern provinces. Respondents for the household interviews included one or two household members who were directly employed or self-employed within tourism, but represented a diverse range of jobs from restaurant cooks and cleaners to front-end hotel receptionists, tour guides, street vendors and national park shopkeepers. These interviews were supplemented with one FGD, and seven KIIs. Respondents for the KIIs included management staff at a number of hotels, restaurants and a spice garden. One FGD was held with independent tour operators to understand the dynamics of tourism service provision. A fairly even distribution between the employment and self-employment (SME) sectors was chosen; 10 respondents were engaged as SME owners and 12 were engaged as employees. Nine female respondents and 13 male respondents were interviewed. The SME sector was represented by three female respondents and seven male respondents, whereas the employment sector had an even distribution between men and women, both with six respondents. Geographically, the sample covered seven districts across four provinces, however tourism-related SMEs were concentrated in the Eastern Province and the sampled households engaged in tourism employment were concentrated in the Central and Southern provinces. For the crafts component of the study, MDF conducted six household interviews: two with household microenterprises engaged in traditional pottery, two in wooden carvings, and two in textile production. One FGD was conducted with female handloom textile weavers employed at a medium-sized factory. The interviews took place in Maruthamunai, Ampara in the Eastern Province, an area that is traditionally known for handloom products, as well in in Galle, Southern Province, famous for pottery and mask-making. Who is Poor in Tourism and Why? The characteristics of the tourism households are explored in detail below, including variances and trends within both the poor and less poor households in the sample. Profile The poor people involved with tourism tended to be workers in elementary positions, such as, informal street vendors, cleaners, and women running micro-enterprises with poor quality products operating street side juice bars, selling low-quality beach sarongs, or providing badly marketed tourism products. Craft and souvenir producers, cultural artists, workers in basic tourism accommodation (guesthouses, inns, and lodges) located in less-trafficked tourism belts also tended to struggle to get decent incomes and remained trapped in poverty. Beyond these, Sri Lanka has a growing number of tourism-related activity providers (cooking classes, glass-bottom boats workers and guides) whose income levels vary according to the tourism seasons but, on the whole, hovered in the near-poor category. Market Development Facility 25

38 26 Market Development Facility

39 Poverty and per capita income Extreme poverty in the tourism sector was not prevalent, tending to exist only at the periphery of the sector. This was an overwhelmingly positive message, given that roles the Study Team expected to have been poor (hotel bell boys and other lower end staff) proved to be pathways out of poverty for many households. The exception was craft producers who, although not below the poverty line, were living close to it. Their earnings were constrained because these families also had other adult dependents (e.g. a sister abandoned by her husband and her small children) to support. The sample included large households relying entirely on the crafts income stream for their overall household income, and others whose crafts income was only a part of their overall income. Following the overall study trend, the poorest income-earning households were predominantly female-headed households due to widowhood, desertion by spouses, or the low income-earning ability of the male household head.58 These households were also low skilled and were working in the tourism sector out of desperation, existing on the periphery of the sector, rather than planning to be in tourism per se. All lived in debt, ranging from an annual LKR 41,750 shortfall to LKR 200,400 shortfall, to meet their basic needs. In contrast with poor households, a common feature of the less poor households is having multiple and regular sources of stable income. Tourism workers in these households were either engaged in profitable SMEs or had stable, skilled jobs in the tourism industry. Their income was complemented by other income generators in the family, who provided other sources of dependable income e.g. a government salary or pension. This did not suggest that income poverty was associated with households solely engaged with tourism. The poorest households also had multiple sources of income. The key difference was that these were either poorly performing SMEs or unskilled workers with seasonal, irregular and fluctuating income. Other sources of income for these households also came from unskilled, unstable roles, hence contributing to poverty and vulnerability. Samurdhi support59 was received in monthly cash payments as savings and utilised for loans. The poorest households were not always able to secure this support, despite households seeking it out. The study found that households with elderly family members appeared to be more easily able to secure Samurdhi support. Other monetary and non-monetary support came from families. Better off family members, typically siblings, assisted poor households. Siblings provided non-monetary support in the form of childcare e.g. in one household the woman headed household was especially reliant on this assistance, as the female head of household would not be able to work away from the home if this 58 T-HH-06 and T-HH Samurdhi Relief Program is a public welfare initiative of the Sri Lankan Government which provides financial and nutritional assistance to low income households. Market Development Facility 27

40 help was not available.60 Well off family members also assisted in purchasing assets such as tuktuks, as a means of securing a livelihood, or even land or a house. Expenditure, debt and borrowings As the Study Team expected, basic consumption costs were the predominant income expenditure for these households. Food costs amounted to 60 percent of household expenditure on average. (this compared with 90 percent for households who derived their income from fisheries). Unlike other sub-sectors in this study, there was no difference between the proportions of expenditure on food by poor or non-poor households. Poor households spent an average of 62.1 percent of their income on food, and the non-poor spent an average of 59.0 percent of their total income on food. Poor households were more likely to take loans to fulfil basic household needs and to compensate for months when income from tourism work was lean. Even for households that did not have a loan, reliance on seasonal tourism work necessitated taking a loan during off-season months. In contrast, less poor households took loans to purchase assets such as land or vehicles, which were used to improve mobility or as a productive asset e.g. tuk-tuks or taxis. The situation was slightly different for craft producers who took loans to generate income or i ease p odu ti it e.g. fo loo s o, i o e ase, to ake a side- a hi e to shape a ask s ears. The interest rate paid by the mask maker was 12 percent for 36 months which he settled on ti e e ause, he said, the a ou t, though ot o e hel i g, as a u de that he did ot a t t ou li g hi.61 There was an aversion to obtaining loans to expand businesses despite the availability of small business loans from public banks by most households even if they had previously used loan services. E e if I take a loa I ha e to pa it off a d, ith u e t o ligatio s, I si pl to sustai a loa pa e t, o e espo de t e plai ed. ill ot e a le The majority of the poor households interviewed had increasing debt levels. They were found to have more expenditure then income and were subsequently in debt. On average their incomes fell below the 200 percent national poverty line, and their average monthly shortfall amounted to LKR The extent of shortfall was found to be higher for poorer households. Households not found to be poor or vulnerable were still taking out loans to pay for basic household needs throughout the year. This indicated that those who were not, on the surface, poor were still struggling to meet basic needs. Ethnicity There appeared to be no correlation between ethnicity and poverty in the tourism sample. Generally, work in the tourism industry was not preferred by a certain ethnicity or caste, as 60 T-HH SLMG-HH Market Development Facility

41 opposed to work in fisheries and agriculture for instance. This is a consequence of tourism being a elati el e se to, i hi h su h o s o eliefs ha e ot et set i. Income from tourism employment Monthly salaries of respondents in paid employment, ranged from as little as LKR 2,250 a month to as high as LKR 45,000 a month. As expected, the difference in tourism pay was based on the type of job and the skill required (See Table 5). The salaries did not account for years worked nor the geographical disparities in the industries, but they did illustrate how skill levels explained large differences in wages. The noticeable disparity between two respondents, a cleaner at a small restaurant and a 4-star hotel worker could likely be explained by the nature of the business hiring, whether a small guesthouse or an established hotel. The hotel worker began working there five years ago and received an LKR 1,000 increment in salary per year.62 The cleaner at the restaurant had been working as a cleaner for nine years, with no change in salary.63 Table 9: Tourism Paid Employment Salaries Job position Skill level Monthly salary Annual salary 4-star kitchen staff Skilled LKR 45,000 LKR 540,000 Hostel manager Skilled LKR 21,000 LKR 252,000 Hostel manager Skilled LKR 21,000 LKR 252,000 Yoga centre receptionist Skilled LKR 20,000 LKR 240,000 Restaurant worker Skilled LKR 20,000 LKR 240,000 Small restaurant chef Skilled LKR 18,000 LKR 216,000 Unskilled LKR 15,000 LKR 180,000 4-star hotel maintenance Skilled LKR 14,800 LKR 177,600 Whale watching assistant Unskilled LKR 9,000 LKR 108,000 Room boy at hotel Unskilled LKR 8,000 LKR 96,000 Small shop assistant Unskilled LKR 2,250 LKR 27,000 Small shop assistant Unskilled LKR 2,250 LKR 27,000 Cleaner at small restaurant Typically, workers in the sector found placements through someone they knew or industry networks. One respondent worked as a room boy at one of the large hotels in Kandy.64 He learned of a job opening from someone at the hotel, and reported that most jobs at this hotel were filled in this manner. Workers who attend vocational training courses were able to find employment through industry networks. 62 T-HH-01. Hotel worker, Kandy, Central Province. 63 T-HH-20. Restaurant cleaner, Galle, Southern Province 64 T-HH-01 Market Development Facility 29

42 People also searched job advertisements to find openings particularly if they were looking for work in other cities. Respondents felt that there were no apparent barriers to entry in the tourism sector which may indicate that the sector is growing and capable of absorbing more workers. Income from tourism SMEs (self-employment) For SMEs, the potential to earn was higher for those who were unskilled, but the risks were correspondingly high, as there was little income security. The highest earner was an entrepreneurial boat service provider who had received a boat from the government and turned it into a business. The lowest earners were, as expected, the informal vendors who sold small refreshments or souvenirs on the streets. As shown in Table 6, a licensed freelance tour guide demonstrated very low earnings, comparable to that of other tour guides who stated during a FGD that earnings could be as high as LKR 40, ,000 a month, or between LKR 480,000 to 600,000 a year. As will be elaborated in the next section, the tour guide s st uggles ould e see as ep ese tati e of the changing dynamics in the industry, particularly in the North Central Province, where SME service providers faced competition from larger international businesses. Table 10: Type of SME and Corresponding Earnings Type of business Skilled or unskilled Annual earnings Boat services (snorkelling, swimming) Skilled LKR 1,260,000 Three-wheeler driver and informal tour guide Unskilled LKR 527,800 Three-wheeler driver for surfers Unskilled LKR 282,200 Snake charmer Unskilled LKR 268,000 Small shop owner Unskilled LKR 256,000 Three-wheeler driver for surfers Unskilled LKR 248,000 Street vendor (sarong) Unskilled LKR 76,000 Street vendor (juice) Unskilled LKR 40,000 Freelance tour guide Skilled LKR 34,000 Unskilled LKR 30,000 Street vendor (handheld fans) Income from crafts employment Two respondents, a mask maker and a handloom weaver,65 reported earnings that were moderate and equivalent to that of a day labourer rate (LKR 1000 and LKR 1100 per day respectively). Their employment was full time and fairly stable (25 and 21 days worked per month respectively). One of the respondents was a disabled worker, who said his job was ideal given his physical limitations. 65 SLMG-HH17 and SLMG-HH Market Development Facility

43 He said he found it difficult to get to and from places of employment because of the travel component.66 The interviewed handloom households67 sold their production to a single buyer. One household interviewed was not interested in selling to suppliers in Colombo as the transportation costs were high, but might consider it in future. A female weaver in Ampara68 and her husband, a police officer, earned the income for their 12member household. She earned a profit of 50 LKR for making one sarong. The buyer, delivered the raw-material to her and collected the finished products. On average, the buyer paid her 15,000 LKR per month plus 5000 LKR for labour. Labour rates were low and unlikely to improve as cheaper imports from India threaten the viability of local weaving. During April, the weaver produced sarees to meet a high demand for the Sinhalese and Tamil New Year season. She produced around eight sarees and for each was paid LKR Another loom producing household said that during the festival months of January, April, August, and December, there were opportunities to sell more sarongs as long as they could produce sufficient quantities to meet demands for pre-orders and informal sales. They said they are not able to attract additional workers to meet this demand as they were required to pay them a LKR 10,000 advance, which they cannot afford.70 Similarly, pottery production was also seasonal. Demand was highest in the tourist seasons, and around cultural and religious holidays. Prior to these times, the household produced in bulk. Outside of these seasons, the household sold their produce to three traders in the market. The traders operated shops and usually bought pots a month. The household could get up to four orders a month of up to 1000 pots. Craft production costs, labour and other expenses of SMEs For handloom producers, raw materials for production were provided by the buyers, and did not amount to a high part of the production cost (around 9-10 percent of the total costs).71 Additional workers were hired but, as this required LKR 10,000 upfront payment to workers, the households were limited in the number of workers they could hire. Expansion costs were a key issue because of the high price of looms. Wood and other material costs (e.g. paint) for mask production came to around 8 percent of the total production costs. 66 SLMG-HH SLMG- HH-8 and SLMG-HH SLMH-HH09 69 SLMG-HH SLMG-HH SLMG-HH-09 Market Development Facility 31

44 Material costs for pottery businesses were said to be so high that the majority of pottery businesses in Galle district had gone out of business. Other craft producers (mask makers) also faced high material prices but had been able to overcome this by finding affordable loan options, albeit the amounts required were significantly smaller than those required for setting up a new handloom, for instance (e.g. LKR 3,500, obtained from relatives, then LKR 25,000 from Samurdhi).72 Key Constraints and Opportunities for Poor People in Tourism SMEs are less able to compete with larger central operators SMEs, particularly in established tourism areas (such as the Cultural Triangle) were less able to compete with larger central operators out of Colombo and Kandy as they were not able to keep up with the changing composition and needs of tourists (e.g. a shift from European to more Asian and Middle Eastern tourists). As such, those who ran micro-enterprises were often unable to compete with larger domestic and/or foreign businesses. This was because small independent operators were unable to target the upper segment of tourists who pre-purchased package deals, often driving up directly from Colombo. These SMEs were also unable to target lower segments of tourists such as backpackers as they did not understand and cater to the needs of this market. An independent tour guide in the Cultural Triangle said he was unable to find work most days because most tourists came to the area on package tours and had a guide already assigned to them by a travel agency. This sentiment was echoed during a FGD with multiple tour guides. The guides said they had witnessed an increase in the number of Chinese and Eastern European tourists who would arrive with their own interpreters. When asked about a different segment of tourists, such as backpackers, the respondents said backpackers were often unwilling, or could not afford, to pay fees for a guide in addition to the high entrance ticket prices to sites (between USD per site).73 A similar sentiment was provided by the manager of a spice garden, who said the main challenge in the tourism industry was that the costs of travel and accommodation in Sri Lanka are so high that tourists were unwilling to spend additional money once they arrive in a location.74 There was a ele e t of ei g left ehi d i these households. Si e the e d of the o fli t, the boom in tourists to Sri Lanka has led to overwhelming opportunities, both for workers and for enterprising individuals. Low skilled SMEs were in a situation where they were able to benefit from the growth in the tourism industry, but could also be hurt by it. With the growth of tour-operator tourism, providers have tended to offer goods and services in-house in order to maintain quality and ensure a reliable level of service. The less equipped SMEs were disadvantaged by the growth of the tour operator business model, as they could not offer tourists the same level of quality and service and, hence, were unable to compete. Conversely, those offering a better product or service could now connect with tourism operators and gain access to their customers. 72 SLMG-HH T-FGD T-KII Market Development Facility

45 Tour guides need a broader range of skills to compete Despite being licensed tour guides who were trained and spoke English, respondents observed that their businesses had been steadily decreasing over the past five years. They all agreed that being an independent tour guide was of depreciating value, and did not wish their children to follow in their footsteps.75 In order to secure more steady customers, one tour guide mentioned that he would need to be associated with a tour company or travel agency, but lacked the networks to obtain such work. The tour guides were frustrated as they saw themselves as highly knowledgeable about the cultural history of the sites, having received formal guide training in Kandy. They felt that the unlicensed guides did not have this knowledge, but despite this, they did not have any sort of edge above them. This was an example of how a supposedly 'higher' level of skill did not mean much in the context of the market, and in fact 'skill' level has often as much to do with softer skills and relatability as harder skills such as technical skills or accrued knowledge. Tourists were often less interested in the detailed information on the site, and were keener to go with a guide who could converse comfortably in their language and relate to them. A stronger example of this was provided by the beach side surfers, who had strong English skills and a high level of relatability to the young backpackers, yet fairly low technical skills, so they were easily able to attract customers based more on shared camaraderie. Across the board, language skills were possibly the most crucial factor in working with tourists. On top of this, the surfers tapped into the need to u de sta d tou ists cultural contexts. Location was a factor in being able to benefit from the growth in tourism People who lived or worked far away from the main hotspots did not get a lot of tourist traffic therefore earned little from tourism, regardless of the quality and value of their products or services. Lesser known tourist attractions were not well publicised and did not get enough exposure to attract tourists. Information in guidebooks, websites and blogs about lesser known tourist areas was limited so few tourists visited them. Locals who attempted to draw livelihoods from the tourism sector received marginal or unstable incomes. There was a tendency for products and services to be developed with poor understanding of the tou ist s eeds Products and services were offered that did not meet customer needs/ were unattractive to them e.g. in crafts, large masks that were heavy and awkward to carry or sarongs with a lot of detailing which were not appealing to beach tourists. There was a tendency to copy what others were doing resulting in an oversupply of a limited range of products that pulled prices down. It inhibited the evolution of new and diverse products and services. In crafts, attitudes towards trader relationships were conservative and risk-adverse People in crafts feared upsetting the status quo by adding new relationships into the mix. Relationships with traders were noted to be positive, on the whole, and offered the advantage of a constant and reliable sales point. One producer noted he was happy with traders or else he would 75 T-FGD-01. Market Development Facility 33

46 not sell to them, and that this was because they come straight to the house and paid cash on the same day as the sale. Another respondent said: E e si e I sta ted o usi ess, I e ee dealing with the sa e t ade s. The e oth f ie ds of i e a d the e al a s ought asks, even during the off-seaso [ he sales slo ed do ] If I did t ha e egula t ade s, I ould have to work twice as hard, sell twice as much and earn twice as less as I do during the peak-season. When discussing a potential new buyer, he noted that, E e afte I spe d ti e a d o e o the samples, this buyer may only make a one-time purchase and, if at the expense of this order I neglect my regular buyers, I will end up losing them as well and I cannot afford to let that happen as they give me regular business during the off-seaso. This behaviour perhaps exemplified the risk-aversion of these households, who feared the unknown and feared damaging their carefully maintained relationships. In terms of paid employment, those who were poor were employed in unskilled roles. The poor did not receive on-the-job training, and respondents stated that there were no prospects for them in tourism beyond their current positions. For example, a female shop assistant near a atio al pa k said that, e e though he jo as o e hou s t a el th ee-wheeler from her house, she preferred it to her only other alternative, which was casual labour.76 Another female shop assistant near a national park said she began work out of desperation after her husband deserted her and her family.77 Both respondents noted that no on-the-job training was provided, and they did not seem to express any plans for a future working in other aspects of the tourism industry. Similarly, a cleaner at a beachside restaurant said he received his job through connections and was lucky to have it as it was the best he felt he could realistically have in tourism, given his low education and lack of skills. He said employers were looking for cleaners with higher levels of education and seeking to recruit those with more experience and skills (including English language skills) so they could be used for multiple tasks. He seemed grateful to have his job and said his employer was generous to him. Without his job he would be forced to work in casual labour to support his family.78 Where tourism was still seasonal, with clear and often long low seasons, people could not secure good tourism related jobs Workers in these areas noted they had been let go from hotel jobs and did not see a long-term future in the industry, and were looking to go overseas. A room boy who worked at a hotel stated that, while there was some potential to increase his income in tourism, he did not see many prospects for himself. Rather, he was looking to move abroad, as have his friends, for better opportunities. He echoed other respondents in stating that opportunities in tourism were limited for those who lacked skills, experience and training T-HH-13. Ampara, Eastern Province. 77 T-HH-12. Shop assistant. Ampara, Eastern Province 78 T-HH-20. Beachside restaurant cleaner in Galle, Southern Province. 79 BILR-HH-16. Room boy, Trincomalee, Eastern Province. 34 Market Development Facility

47 The very poorest households experienced economic shock from the loss or incapacity of the main breadwinner At the household level, the major constraint of the very poorest households was economic shock, predominantly in the form of the loss or incapacity of the main breadwinner. This was due to desertion of wives, death or health-related limitations of the male household head. In some cases, the female household head was working out of necessity rather than choice, and was often poorly skilled. A hostel manager had to take up work after her husband passed away, and the household lost their only source of income from his private bus operator salary.80 Another household respondent became involved in driving three-wheelers for tourists after his trade business failed causing his family to have to sell their house and other assets.81 Added to this, was the community pe eptio that tou is as egati e a d asso iated ith ad ha its. I a a s, tou is as slowly providing a pathway out of poverty from the aforementioned economic shocks, given that respondents had no skills and experience in tourism prior to their current roles. Financial literacy was low in poor households The poor households suffered from chronic debt, in that their expenditures consistently outweighed their income. The one e eptio as a estau a t lea e ho said he did ot spe d a additio al pe of his LKR, o thl sala u less it as a solutel e essa. A o o the e as the use of loans for the purchase of luxury assets (a reflection of poor financial literacy) and a reluctance to obtain loans from commercial institutions due to the fear of the inability to repay. There was a preference to purchase household expenses on credit or borrow informally from neighbours or family members, for which interest was not required to be paid. Those ho did ell i the tou is i dust te ded to e highl tou ist fa i g, i positio s that enabled them to receive service charges and tips. Receiving tips and service charges was a reason why tourism workers in relatively lower level positions, such as bellboys and cleaners, did well in larger hotel establishments where service charges were factored into their pay. Those who had less front-end contact in this type of role had fewer opportunities to gain direct access to tou ists o e e.g. o e e e i di e t o ta t ith tourists on the beach in Galle selling juice and sarongs, but tipping was not usual in this situation thus they did not benefit despite their proximity to tourists. Tourism workers in established tourism areas and in established hotels were finding a pathway out of poverty The overwhelmingly positive message from the sample was that tourism workers engaged in formal paid tourism roles, particularly in established tourism areas and in established hotels were finding a pathway out of poverty. Most telling during the Study was the difficulty experienced by the Team in locating tourism income-based households that fell below the poverty line. Even bellboys and kitchen helpers and other seemingly lower tier tourism jobs were found, in fact, not to be poor and 80 T-HH-02. Hostel manager. Kandy, Central Province. 81 T-HH-11. Three-wheeler driver. Ampara, Eastern Province. Market Development Facility 35

48 unskilled workers could and did, in the right circumstances (where there was an upward career path and on-the-job training), rise up through the industry. It was those on the very periphery of the sector, that were likely to stay poor and who struggled the most. When tourists could operate all year around, household incomes benefited significantly. Pathways out of Poverty in Tourism The Stud ide tified a u Pathway 1 e of path a s out of po e t fo Tourism that can inform MDF s o k. Expansion of tourism to enable the absorption of workers into untapped tourist regions, particularly in the former conflict-affected Eastern and Northern provinces There were a number of conditions that, if present, could provide clear pathways out of poverty for the households involved, namely, stable income sources, the availability of service charges, and on-the-job skills training. The positive news was that, by expanding tourism, it would play an important role in increasing incomes potential for households, as new job opportunities were created (i.e. it is, for the majority, less about transforming the sector and more about increasing the options for absorption of workers). There were many jobs suitable for the poor in tourism because of the low initial skills requirements and the opportunity for on-the-job training e.g. as services workers in hotels, guides and vendors. Opportunities existed in the expanding tourism areas opening up along the eastern coastline, particularly in the former conflict-affected Eastern and Northern provinces and where there was room to absorb large numbers of unskilled or low-skilled workers. Expansion of tourism opportunities in emerging areas provided a clear pathway for poor households to benefit from a growing and promising industry. Pathway 2 Diversification of tourism activities, particularly within these emerging tourism destinations Diversification is seen as beneficial in two ways. Firstly, the emerging tourism areas will need to have a competitive advantage to draw tourists to them, particularly as they are located at a considerable distance from the capital and main travel-hub Colombo. Tourism activities in these regions will have to offer something distinct and different. Secondly, the SME models offered by small-scale tourism providers are undiversified and unoriginal, tending to copycat one another, and they often lack durability. Fostering more diversified options should enhance the appeal and sustainability of SMEs and foster additional employment activities for men and women. Pathway 3 Stronger links between large operators, SMEs and communities Beyond strengthening the product and service of small and medium tourism providers, there is also a space for MDF to focus on strengthening SME capacity and networks so that local tourism services do not disappear from tourist attractions. The more interconnected these activities become the better with each other, the better will be the tourist experience to a particular destination (e.g. dive shops connecting up with rafting providers, specialist lodges and so on). 36 Market Development Facility

49 These businesses will need to be robust, offer unique products and services, and occupy a distinct position in the market so as to not be absorbed by in-house activities of large hotels. MDF can look at facilitating links with international tourist operators to provide support to SMMEs in training, and to promote supply linkages for small operators to large operators, social enterprises and ethical training, and joint regional marketing. Pathway 4 Offset seasonality by diversifying activities and services The opportunity to earn income through work in tourism work is constrained where tourism is highly seasonal and dependent on one or a few activities. Time and again, it was seen that workers and SMEs did all their business during the tourist season and did not engage in any other income generating activities in the tourism off-season.82 In Arugam Bay the tourist season was very short as it depended solely on surfing. MDF can help broaden the range of activities that are available to encourage a wider tourist segment to visit destinations by working with innovative businesses and communities who recognize potential for the development of new attractions. This can be an effective pathway out of poverty for SMEs associated with tourism and service providers, as well as workers, as incomes drawn from the sector can be sustained for a longer period over the year. Pathway 5 Open up tourism by improving connectivity and information Pre-booked tours, travel and accommodation run by large tourist operators restricts opportunities for small service providers such as guides, activity providers, tuk-tuk operators, souvenir companies and craftspeople to sell to these tourists. The tour group model tended to centralise operations and to offer all products and services through its own closed network, effectively closing out other small businesses and service providers. MDF interventions which open up the sector by making better information available online about destinations, routes, and accommodation, by offering convenient booking services, and improving road, air and rail connectivity could encourage travellers to venture out on their own. This would create pathways for competitive service providers and SMEs to increase their income as greater numbers of tourists became willing to seek out their products and services. Pathway 6 Develop flexible, transferable and soft skills Providing outdated skill sets (as exemplified by tour guides in Kandy, who had a high degree of knowledge regarding the sites, but weak language skills) leads to models of training that were unsuitable for the market. A focus on adaptable skills training to increase the quality of interactions between tourists and workers, in conjunction with foundation skills (such as hospitality training) may yield more for those working in tourism than building technical capability. The effectiveness of different training types should be explored, looking at softer skills development, enabling workers to better understand the needs and interests of tourists. This was done well by the surfing trainers of Arugam Bay, who understood the needs and cultural 82 T-HH-09, 10, 11 and 12. Market Development Facility 37

50 backgrounds of the backpackers, and were thus making good incomes by adding value to basic goods and services. At the household level, avenues could be explored to promote financial literacy skills and help SMMEs better manage their finance and debt levels. Pathway 7 Better connect crafts producers to tourism Crafts producers have poor connections with end markets and, as a result, they are unaware of the needs of such markets and produce unsuitable items. By better connecting crafts producers to these markets, in a manner that also enables crafts producers to upgrade their products (through better design and packaging), would open them up to more lucrative and sustainable markets. In some cases, this might be more cost effective e.g. rather than producing large, ornate and outdated masks that did not sell well and took a lot of time and resources to produce, smaller, smarter and simpler products could fetch good prices in the tourist market if appropriately branded and positioned. Pathway 8 Improve product design and branding by focusing on tourist needs, product differentiation and product experience Often overlooked was the fact that tourists were buying the experience and memory of a product as much as the item itself. For example, tourists were more likely to buy a mask when visiting an area that was famous for masks, or to buy a souvenir specific to a particular place or experience/activity. In this manner, the experiential aspect of tourism product branding could become very important, and help to overcome cheaper imports which are not specific to the particular place. The story and positioning associated with a product could help to differentiate products, rather than, the complexity, ornateness or price. These changes posed no risks to households, and should not raise production costs; resources would just be used more smartly. Research on local brand values might need to be explored, as well as consumer preferences. Tourism provides a favourable platform for handicraft growth, particularly as souvenir shopping experiences in Sri Lanka are limited.83 As such, a reorientation of the handicraft sector from the domestic market to the international tourism market may be promising. Roles Performed by Women in Tourism Below are snapshots that denote some of the characteristics of roles that women occupy in tourism. Women in households In general, poor women working in the tourism sector (other than highly educated women in hotels and other paid employment who earned more) were doing so out of financial necessity. Where men were present in the households, women were not found to be highly engaged in SMEs, but were very active in household money management. In general, the roles of women and 83 Report of the Task Force for a sustainable development of The Handloom Industry in Sri Lanka. Textile Industry Development Division, Ministry of Industry & Commerce Market Development Facility

51 men were well defined in this regard, with women managing the income from these activities, and men bringing that income home. Women in these households, it was explained, were never expected to work.84 Only in one SME (a small shop) did the wife assist with food preparation.85 One exception to this was in the crafts industry where women were highly active, working as handloom and basket weavers, and pottery makers. Two households were involved in family businesses run by the husband and wife working together, and both were in these businesses because they had been handed down f o the ife s fa il.86 Wo e s de isio aki g a d i flue e i households In tourism-related households where women worked or stayed at home, there was a high level of influence and control by women over household finances. In Sri Lankan households, it is common for men to hand over the vast majority of their income, when they got home each day, to their ife. Me oted the sa it as thei ole to ea the o e a d thei ife s ole to a age it, a d they believed their wives were the best people to manage the money for the household.87 In handicraft households women were found to be at the centre of decision making in terms of the family business as well e.g. one woman noted, I de ided to u the loo (for her handloom business) and then informed m hus a d a out it a d e t ith hi to u it.88 One specifically oted she e jo ed afts o k e ause, ou a o k e uall et ee e a d o e. Women of female-headed households managed their own incomes, though noted that they consulted with adult children when spending their money. The picture was slightly different in Muslim households where the men generally did the household purchasing, although the women told their husbands what they needed. This restriction was related more to the more restricted mobility of women in these households rather than o e s a ilit to ake pu hases. Women at work Women were found in a wide range of tourism-related employment roles. It was said that the presence of women in tourism-related jobs was increasing, particularly in front office, bookings, accounts and management jobs. The choice to enter paid work was often to do with necessity, although financial independence and the desire for the freedom to support themselves and others was a huge driving factor, along with not wanting to be a burden on parents. Women in managerial level roles generally entered with a high level of education and language skills e.g. a female respondent and one of the highest tourism income earners in the study had shifted 84 T-HH T-HH SLMG-HH-14 and SLMG-HH T-HH-08, T-HH-10, T-HH SLMG-HH-09 Market Development Facility 39

52 from a managerial position in graphic design, and others had gained valuable experience through spending years prior in the Middle East.89 Societal perceptions are generally averse to women working in the tourism sector. Women were e pe ted to e i espe ta le oles su h as i hospitals, a ks, the go e e t a d s hools i.e. 90 generally pensioned roles which were regarded as being secure. However, not all jobs taken up by women are perceived in the same way, and there were conditions related to the jobs that impact this. This is further discussed in the concluding chapter. Women did take up jobs outside of these traditional female roles however there were limits placed on them or they were viewed unfavourably e.g. jobs where women were required to travel away from home for long periods came with social stigma. A respondent noted that she, like other women she knew, would rather set up her own business and work from home. Undoubtedly these wider social views influenced household members in the study. Women seemed more concerned with the views of their husbands as opposed to other family members or other women in the community. During the FGD with female weavers, it was observed that women generally worked in groups in dedicated workspaces in factories while weaving and this was a critical factor in terms of enjoying work and convenience. In many cases families had initially been resistant to women working in tourism-related jobs ho e e, o e the sa the fi a ial e efits, the usuall a e a ou d to the idea.91 Women as entrepreneurs Poor self-employed women in tourism usually became so out of necessity. Women worked on the periphery of the sector in whatever would make them an income e.g. a married woman whose husband had few income sources sold juice informally by the beach in Galle; a sarong seller, also on the beach in Galle, was forced to trade as her husband spent most of his income on alcohol; and a widowed fan seller in Kandy. All were selling poor quality products, had low educational attainment and few alternative work sources. None saw a future for themselves in the tourism sector. Family-run handicraft enterprises, thrive through skills being passed down through the generations. In handloom textiles and pottery making microenterprises, both men and women were engaged in production, but men managed the front end of the business such as management and sales. An interview with a small handloom textile household revealed that, while the trade was i he ited f o the ife s fa il, it as the hus a d ho i te a ted ith u e s a d e uited weavers. The wife said the production work was shared equally between the two of them.92 Another textile household presented an unusual story in that the female handloom weaver was in 89 T-HH-01, T-HH-03 and T-HH T-HH T-HH-12 and T-HH SLMG-HH-08, Ampara, Eastern Province 40 Market Development Facility

53 charge of the dealing with the buyer because she was the only household member involved in textile production.93 The two study households involved in pottery also demonstrated that the production was equally shared between men and women. In one household, the wife was responsible for the moulding and decoration, whereas the husband worked the kiln. However, when it came to dealing with business outside of the home, it was the husband who negotiated purchases of inputs and sales with traders, all of whom were male.94 It would seem that there is a social tolerance for a woman to undertake management and sales work, if she has no husband or other male family member able to assume these roles. Barriers for Women in Tourism Women faced significant social barriers to entering the tourism industry. This was primarily due to so ial a d ultu al o s a out o e s eputatio s a d espo si ilities. Su h thi ki g has manifested into households discouraging females from working in tourism, discriminatory practices f o tou is i stitutio s, a d o e s elu ta e to o k i tou is. Men held negative attitudes about women working in tourism and women are concerned for o e s safet Most male respondents said they did not encourage women working in tourism, but were reticent to explain why, saying simply that women would not be able to attract tourists.95 Female respondents e e o e illi g to e plai so e of the so ial o s that go e ed o e s eha iou i tou is. A female manager at a beachside restaurant in Galle made it clear that she did not hire women in her restaurant, as she was o e ed fo o e s safet : Whe sea e o e, fa il p o le s a e o i g up. It is te i le, the a e d i ki g, the do t behave properly. They ask o e to d i k ith the. When probed whether safety issues could be overcome by hiring women to work in the back-end of the usi ess, su h as i the kit he, she fu the ade it lea that o e s eputatio s ould e at risk if they worked in tourism: Safet is ot e ough. The thi ki g patte he e [i S i La ka] is diffe e t. A tuall, a lot of p oposed a iages a e happe i g i the ou t. If the fi d out that ou o k i tou is, the do t a t to marry you anymore because they thi k ou a e o ki g i ot su h good thi gs. 96 She said that if she herself went to work in another tourism-related industry, her husband would not allow it and it would cause problems in their marriage. The only reason she was working in the restaurant was because it was owned by her husband, and she co-managed it with her sister. 93 SLMG-HH-09, Ampara Eastern Province 94 SLMG-HH-14, Galle, Southern Province 95 T-FGD-1, Tour guides, Sigiriya, North Central Province. 96 T-KII-7, Beachside restaurant manager. Galle, Southern Province. Market Development Facility 41

54 Having said this, as noted above, women were also quick to point out that these views played a limited role when there were bills to be paid, and they were generally more concerned with the views of their husbands than with others. Women working out of necessity were keen to state that the a ed little fo the thoughts o this of othe o e i the o u it, si e it as t these 97 women that had to pay the bills. Others were fou d to e fi st o e s, hoosi g to joi fo reasons of financial independence. However, the same respondent who had rejected the stigma of the community noted that while her husband was around (he had since deserted her) he would never have allowed her to take the job she has now. Thus the opinion of husbands, as would be expected, played perhaps the most central role in whether women could or could not work. Women worked in paid employment in formal tourism establishments where they were limited to the front desk There were few reception jobs relative to the overall staff size. For example, a famous 4-star hotel in Kandy said they had five female employees out of a total of 230, all of who were working in reception. Observations around other hotel establishments also confirmed gender disparity in larger institutions. When asked why there was such a dearth of female employees despite provision of a competitive salary, lodging and other benefits, the representative said that women tended to leave, likely as a result of marriage or to pursue other jobs such as teaching.98 Thus while the supply of jobs may be low, the demand for these jobs seemed to be even lower such that if a qualified woman were to enter tourism, she would not face much competition. Attitudes to women were more habitual responses than genuine resistance It as o se ed that espo de t s ea tio s to hat o e ould a d ould ot do as elated to a habitual or normative acceptance of roles, rather to a resistance attributed to a particular reason. Tou is jo s i this se se e e e e a d hat the e tailed as ot ell k o. Pe eptio s around what these women did for a living were often quickly made, based on few facts, and made out of a sense of discomfort. Once the initial hurdle of acceptance was overcome and women were bringing home a significant income, families did generally come to adjust their thinking. Having said this, the stud fou d that the espe ta ilit aspe t to o k as al ost as ital as the i o e potential, and thus sensitivity to factors such as time away from home, distance from home, perception of whether the women were working with men too closely in hospitality roles, whether they engaged too much with strangers (particularly in close contact), needed to be considered. Interventions need to be designed with care, and the perceptions around a job must not be overlooked. Households e e ot suppo ti e of o e s de isio to o k, ut ele ted out of e essit Female respondents in paid employment stated that their respective households were not enthusiastic or very supportive of their decisions to take on a job, but that they relented out of financial necessity. Concerns surrounding childcare seemed to colour household impressions of married women and work. A shop assistant stated that her family, including her husband, were 97 T-HH T-KII-1, Human resources executive. Kandy, Central Province. 42 Market Development Facility

55 against the general idea of her working, because they felt that her children would not be properly cared for. The shop assistant explained that this view of a woman caring for her household was widely shared in the community and, unless there was an absolute financial necessity, women were discouraged from working. Despite claiming that women should be able to work as much as men, she said she would prefer not to have to work.99 In crafts, employee women weavers were less productive than their male counterparts due to their limited mobility and household responsibilities, and they earned less During the FGD, women handloom weavers said that, while men and women were paid equally per piece at the factory, men were more productive. Women would produce between six to ten metres of textiles a day, whereas men would produce about fifteen metres. When asked why men were more productive, the women responded that men had more hours to work, whereas they had household and childcare responsibilities. In other words, men could spend all day and night at work, whereas the women might get interrupted, have to leave work early, or take a break during work hours for household reasons. In a medium-sized factory, such flexibility was allowed, and the women were easily able to cover for one another.100 In one household microenterprise, this situation had led to explicit discrimination against hiring women weavers, because the company could not afford the flexibility. The fe ale head of a household said that o e s fa ilies e e agai st the o ki g lo g hou s away from home, which translated into more absences. She stated that they tried to hire male weavers as much as possible. That being said, they hired some women to help her spin, given that spinning was less labour- and time-intensive.101 Older women appeared to be less stigmatised for their work in the tourism sector A female small hostel manager said she was offered her job, in part, because she was older and the efo e o e t ust o th. She e plai ed the e as a pe eptio that ou g o e s eputatio s would be tainted if they worked in hotels.102 Further, there appeared to be less of a cultural stigma around women with older children working, whereas there was a strong expectation that those with young children would stay at home to care for them. Female respondents who were self-employed informal vendors did not exhibit as many concerns about gender dynamics. This is likely due to the fact that they were the heads of households because of widowhood, divorce, or a non-functioning male spouse. Further, given that desperation had driven them to take advantage of the tourism industry, and given their work was on the fringes of the industry, it might have been the case that they did not see themselves as working within tourism. Having said this, self-e plo ed fe ale tou is o ke s did e p ess o e s a out o ilit, a d o e s o kloads 99 T-HH SLMG-FGD-2, Ampara, Eastern Province. 101 SLMG-HH-08, Ampara, Eastern Province 102 T-HH-03. Market Development Facility 43

56 in tourism SMEs was high. One respondent noted she felt male street vendors had greater mobility a d fe e issues ith se u it tha fe ale st eet e do s. Wo e s o kloads i this sphe e e e highly demanding, with one street vendor noting she got up at 5a.m. to take care of her children, arrived to sell by the gardens at 1 : a.., pa ked up to go ho e at p.. a d the did t fi ish he household chores to sleep until 10p.m. or 11p.m. at night.103 This lack of flexibility and mobility prevented women from taking up formal training opportunities A female handloom weaver had an opportunity to train as a master weaver, but could not do so as the training centre was five kilometre from her house.104 Such a constraint seemed heightened within the Muslim community, where mobility was found to be even more restricted. In these circumstances, women did not have the freedom to move about on their own, or to travel far from home. The Study Team noticed there were few women, other than those accompanied by a man, visible in the main town. Employment and engagement in crafts were feasible only when certain preconditions were met, including proximity, dedicated spaces, and flexible hours to engage in household duties Women in the sample were able to work because a number of conditions prevailed, without which they would may not have been able to work in crafts. These conditions included living nearby the factory, being able to operate in women-only dedicated spaces and have the flexibility to leave the factory and attend to duties at home if required. Such factors were as critical for women in sectors other than crafts. The sense that crafts was suitable for women could be misleading; factors such as proximity and time balance must be factored into designs for all sectors. Institutional norms within the tourism industry were unfriendly to women Institutions in tourism appeared to be gender agnostic in their operation but, in doing so, they failed to e og ise iti al a ie s to o e s pa ti ipatio, su h as a la k of t a spo tatio, ope o k spaces and non-congenial working hours. In factory settings, there were only a few companies that provided day care facilities. Legislation restricting the employment of females at night limited the participation of women. Recommendations to Address Barriers for Women in Tourism The role of women interacting with tourism, either directly, or as members of households associated with tourism, could be encouraged and improved in a number of ways. The study identified the following areas. Expansion of opportunities for women in paid employment in tourism and establishing work conditions that enable women to overcome social acceptability barriers to entry Research findings suggested there were increasing opportunities for women to work more visibly in tourism, initially for the outliers (or first movers) who sought to participate in the sector and 103 T-HH SLMG-HH-09. Ampara, Eastern Province 44 Market Development Facility

57 who, in turn, increased acceptance among the community and allowed other women to enter in time. It may be beneficial to partner with larger international tour operators, hotels and local recruitment companies to promote gender-sensitive recruiting and hiring practices, as well as to facilitate the development of a safer, and more conducive environment for women (e.g. look at workspace requirements, transport options, working hour flexibility, and what made roles more or less acceptable and tailor them accordingly). There may also be an underestimated public relations angle to these jobs i.e. working women could be promoted as role models of women of good position and character, to try and change the negative perceptions around women in tourism. These perceptions were which are often knee-jerk reactions and may actually be overcome with the right kind of branding. This could be built into the training programme, with material targeted to women workers, families, and more broadly, by working with regional and national providers to publicise positive images of women in tourism (e.g. through popular TV series and careful advertising and branding within the industry). Nurturing of competitive female-operated SMEs The Study revealed that female-operated SMEs in tourism were highly informal, highly peripheral, and the lowest earning and poorest in the sample. However, given a low number of SMEs in the Study sample it was unclear the extent to which successful tourism-centric female-operated SMEs were present in the industry. MDF may benefit from undertaking a feasibility study to determine the extent to which such SMEs are present or have the potential to be successful. While social and cultural norms presented significant barriers, pockets of opportunities may exist, particularly in the Northern and Eastern provinces where female-headed households were more prevalent. While building and supporting the tourism SME market across the board, MDF should ensure additional constraints, such as market access and softer factors such as limited networks, faced more by women, are addressed as part of the intervention design. Training of women within the tourism sector must not only look at building technical skills, but also at building soft skills and confidence Soft skills relate here to assisting women to overcome cultural and structural barriers and beliefs which prevent them from entering and rising in the industry. For instance, a woman may have or be capable of building the technical ability to perform a role, but she may be inhibited by feeling that she could not perform the function, or was uneasy about stepping into a new or higher position. Technical training providers should look at the need for both hard and soft skills. These skills could be provided in numerous ways, such as, through mentoring schemes, modifying e isti g t ai i g p og a es o thei deli e to add ess a ie s to o e s advancement, certifying training to enhance the value of it to the women receiving it, or making the training relevant to the o e s o kpla e. Soft skills a e ofte o e looked, ut a ake a iti al difference between women applying or not for roles. Smart soft strategies need to be developed to overcome issues around stigma. Mindfulness of the limited mobility and time constraints of women, in jobs, SMEs and training While the afts se to is ofte see as ei g s o ous ith o e s o k, i fa t, o e fa e the same issues in this industry as they do in others i.e. they have to manage their work around Market Development Facility 45

58 home tasks, and getting to and from jobs and training can be limited due to social reasons related to mobility. For these reasons, training and design upgradation programs, should be developed in a way that is inclusive of women. Training and other capacity-building initiatives must be sensitive to o e s li ited obility and time constraints. If not, there is the risk of creating segmentation within the market where only men are able to produce higher-end products and women continue to produce simple, lower-end products. Other factors, such as the provision for outsourced homebased work and then final finishing done separately in a central point, may be cost effective for the business and allow women to produce concurrently with other tasks, such as childcare. There will be no single solution in these cases, and the choice of whether home-based work, childcare options, and dedicated workspaces are developed will depend on the specific requirements of the particular group of women in question (e.g. some communities might already have solid childcare family support networks, and this might pose less of a problem). Working in a communal setting as a means of mutual support for women It was observed that dedicated female workspaces that allowed for children to enter and that were close to the community, were highly favoured by women. These workplaces appeared to overcome some of the constraints around time management and stigma seen in other work places. They enabled women to give and gain support, and to cover for each other in the event that household emergencies arose. Supporti g ha pio s to reate role odels for ha i g ore o e frie dl ork spa es While there were many cases that illustrated the barriers to women participating in tourism, there were also a few where some businesses managed to 'break the mould' and hire and retain a female workforce. These institutions were found to have used innovative recruitment channels, provided convenient facilities for women workers, namely, transportation, segregated work spaces, and flexibility in work hours. Such institutions, supported by MDF, could act as role models for other organisations to follow which could facilitate greater participation of women in tourism. 46 Market Development Facility

59 Chapter : Fi di gs of the Po ert a d Ge der Stud I pro i g Sri La ka Produ e a d Produ ts for E port a d Touris Markets The scope of this SEA included products and services made in Sri Lanka for international consumers, including tourists and export markets. It covered fisheries (inland and coastal), fruits (banana, papaya, watermelon, pineapple and other tropical fruit), vegetables (chili, brinjal (aubergine), and onion among others) and spice (where there was the potential to involve poor households). Poverty was more widespread in this SEA. The majority of the farmers, fishers and producers operated on a subsistence basis, with varying degrees of commercial orientation. Growth prospects for these households lay in linking them to growing markets like tourism and export. Produce for export, albeit a small portion of total production, had shown a marked increase of over 500 percent in the last ten years.105 It was sold to markets in the Gulf, Maldives, Europe and the USA among others. The produce market also catered to the tourism market. With tourism arrivals increasing, prospects remained encouraging to link small producers to the hospitality industry. Small farmers and fishers faced low yields and outputs in the face of rising costs, outdated practices and poor crop choice. Those affected by the war had low productivity due to constrained access to inputs, markets and information. Women were part of the agriculture household and often performed post-harvest activities in fishing households. Linking these households to growing end markets like tourism and export, provided a pathway for them to emerge from poverty. However, this will need an understanding of the barriers faced by the poor and what trapped them into poverty. The Study, covering the major production zones across the country, will help MDF answer these questions and enable it to design and implement appropriate partnerships that contribute to poverty alleviation. This chapter is divided in two sections covering fisheries and horticulture respectively. 105 Based on calculations from Central Bank of Sri Lanka 2015 data. Market Development Facility 47

60 Fisheries Methodology and Sample Characteristics To understand the poverty and gender dynamics within reservoir (inland) fishing households, interviews were conducted with five households in Polonnaruwa, North Central Province. Both the men and women of the household were interviewed where feasible. Three households fished on the Parakrama Samudra reservoir, one in the Minneriya Tank, and one in the Girithale Tank. In addition, a FGD as o du ted ith e e s of the Gi itale Rese oi Fishe e s Asso iatio to u de sta d the dynamics among fishing communities. Household interviews were also conducted with a coastal fishing household in Mannar and an employee at a canning factory in Mannar. Findings from these interviews are analysed here and in Chapter 5 which covers the findings for the stimulating entrepreneurship and business activity in former conflict-affected areas SEA. Who is Poor in Fisheries and Why? The characteristics of the fisheries households explored below, include variances and trends within both the poor and less poor households in the sample. Profile Fishers tend to have vulnerable incomes due to seasonality and variation in their catches. Fishe s ho do ot o oats, a d o k o othe s oats, e e fou d to e li i g o e step below, nearer the poverty line. Fishers relying on inland fishing were more exposed to variation in earnings and were therefore seen as more vulnerable. The case for coastal fishers was similar, with deep-sea workers having slightly higher incomes. In terms of ethnicity, ethnic lines were blurred with fishing being determined more by geographical location and access to water bodies (which are spread across the country) than ethnicity. Despite this, Tamil Christians of lower castes were found to be associated with fishing in some areas. In the Northern Province and the Eastern Province, there were a number of conflict-affected fishing households that had low-income levels. Per capita income and reliance on fishing as an income stream Fishing households made comparatively good incomes from their fishing income streams, with no household in the sample falling below the 100 percent national poverty line. However, households were found to be income vulnerable (falling below the 200 percent national poverty line) with even the better off households struggling to acquire basic resources such as fishing equipment. The less poor households shared almost all the characteristics of the poorer households i.e. relating to education levels, degree of reliance on fishing income streams, and amount of time dedicated to the activity. The only apparent distinguishing factors for income appeared to be that they were able to catch more fish (due to stocking issues at their reservoir points rather 48 Market Development Facility

61 than equipment or technique differences) and, to a lesser extent, the price received for the fish (discussed further below). Fishi g a ou ted for arou d 9 per e t of fishi g households i o e. There appeared to be three reasons for this reliance on a single income source. Firstly, fishing was time-intensive. Fishermen stated that time constraints prevented them from pursuing any other type of incomegenerating activity, as nets must be placed early in the morning and then taken down in the evenings, in addition to harvesting, post-harvesting and sales.106 Secondly, while all households had land, none of them cultivated crops or kept livestock for income. They said they are unable to do so as the weather was too hot however they grew a small quantity of food (such as coconuts and bananas) for household consumption. Thirdly, fishing is labour-intensive requiring the participation of the majority of the household in activities. In this manner it is a hole fa il e te p ise.107 Women were not always involved in fishing themselves, but all of the interviewed women assisted in the fishing economic activity through preparation, cleaning and undertaking other post-harvest activities. Many fishers regarded fishing as the only skillset they had, and that it was a lifestyle economic option, with many interviewees stating that, aside from the financial limitations, they enjoyed the work.108 There was some limited practice of seeking additional cash injections beyond fishing e.g. from driving work and day labour.109 Expenditure, borrowings, debt and savings Despite providing a good overall income, income from fishing appeared to be unstable, with lean and flush cash months110 where, for six months of the year households were able to earn enough to make ends meet and for the other six months they had to borrow for basic needs (such as buying groceries on credit and pawning jewellery) and for netting costs.111 Poorer fishing households were barely covering basic needs costs, spending their income predominantly on food (roughly 70 percent of total household expenditure) and they struggled at times to cover basic costs, and spoke of times when they had to skip meals. (The poorest household was spending 91 percent of its income on food).112 For less poor households, despite making relatively good incomes, financial planning did not appear to be balanced, and a large proportion of their expenditure went on social obligations rather than reinvestment in economic activities e.g.one household spent LKR 84,000 a year on weddings, and another spent LKR 30,000 in a year on church fetes and religious pilgrimages113 i.e. spending almost as much on 106 SLMG-FGD SLMG-HH SLMG-HH SLMG-HH-05 and SLMG-HH SLMG-HH SLMG-HH SLMG-HH SLMG-HH-04 Market Development Facility 49

62 social obligations as it did on food expenditure, leaving limited room for reinvestment into items such as fishing kayaks. Poorer households were not able to gain individual financing for critical items such as nets, and were instead taking up group loans with three individuals coming together to obtain a loan from a microfinance institute. Respondents said this was convenient in that agents came to the house to collect repayments, but ideally these fishermen would prefer to obtain the loan individually.114 Less poor households were taking out loans for investment e.g. the least poor household had borrowed from the Bangladesh Rural Advancement Committee (BRAC) Sri Lanka to invest in a grocery store business and taken a housing loan from the Co-operative Savings Bank.115 In contrast, the two most vulnerable households took out formal loans to pay for essential fishing needs, such as nets116 or to purchase a fishing boat.117 As expected there was a general reluctance to take out commercial formal loans and a preference for informal lending or welfare options. The most vulnerable household purchased its household groceries on credit or pawned jewellery to cover expenses.118 The other vulnerable household in the sample said they would take out a loan from Samurdhi.119 However, even one better-off household borrowed from its friends and family (interest-free) to meet expenses.120 Men hand over most of the money they receive from fishing to women,121 who save some of it in in a bank account in their own name.122 Much of the investment for the household (of slightly better-off households with additional income) appeared to be female-driven, with women in the two best-off fishing households taking out loans from BRAC Sri Lanka and the Cooperative Bank for income generating activities (i.e. net fixing and setting up a grocery store, and another had plans to make tilapia (a type of fish) nets).123 Income from fishing Fish catch and sales (produce, prices, trading and markets) The sample households were predominantly catching tilapia, koi and carp. They were not targeting particular species of fish with any particular nets, and there were no specific zones in the ponds that were better than others, with fishermen generally determining their catch 114 SLMG-HH SLMG-HH-04 she was depositing LKR 500 per month into the bank account. 116 SLMG-HH SLMG-HH SLMG-HH SLMG-HH SLMG-HH SLMG-HH-04, SLMG-HH-05, SLMG-HH SLMG-HH SLMG-HH Market Development Facility

63 locations depending on where others were setting up their nets on a given day.124 The most critical differentiating factor in better-off households was their ability to catch a greater quantity of fish. There was little difference in terms of the number of days and months households were catching fish this was even overall and there was no correlation between time spent fishing and fishing incomes. This would support fieldwork observations that stocking of fingerlings and larvae was a core issue. Even more than price received for catch, the most critical variable impacting household incomes was the quantity of fish caught. Fish was sold either immediately on the lake bank to local traders, or (if available) or via motorbike deliveries to established customers, in small trades or for a lower price for bulk trades (4-5 kg at a time). No fishing household had a problem selling fish and spoke of customers frequently waiting for the catch.125all households sold the majority of what they caught and no household had issues with wastage or spoilage. Any remaining fish were consumed by the household (around 5kg on average per week).126 However, few poor households sold their fish for high prices. The top income from fishing households in the sample was based on an average price of LKR 360 per kg, while the lowest income from the poorest household was based on an average price of just LKR 185 per kg i.e. it varied by nearly 50 percent, for the same species. The fishing committee managing the reservoir or lake would also take a cut of the fish price, usually around LKR 10 per kg. Prawns were lucrative, selling at LKR 1000per kg, ut for ed a s all part of households incomes between 2 percent to 6 percent of fishing revenue. At these quantities (around 2-3 kg per month), they did not add significant additional income potential to these households. Prawns were always sold and were not consumed by the households.127 At the Parakrama Samudra Reservoir, it was said that prawns from all 400 fishermen were sold to a single buyer. These families also sold dried fish for approximately LKR 280 per kg generally to customers coming to the house (by the women of the household).128 Overall production costs Fishing nets cost, on average, LKR 40,000-50,000 per annum in four of five households. These expenses included repairs and purchases. The Study was unable to capture how many nets the households had when they started their fishing enterprise, but in looking at the consistency of expenses across households, those who were vulnerable spent a disproportionate amount of time ensuring they had nets or that their nets were usable. For example, for the wealthiest household, net expenditures accounted for just 3 percent of their overall revenue from fishing per annum, but for the most vulnerable household, this expense rose to 15 percent. This is discussed under constraints below. 124 SLMG-HH SLMG-HH SMLG-HH-03, 04, and SLMG-HH SLMG-HH-03 Market Development Facility 51

64 Licensing were allocated fishermen on an individual basis, and the licence fee was used for restocking fingerlings, with the government subsidising remaining costs. The fee was reported to be LKR (of which LKR 50 went to the provincial fishing association, and where possible, LKR 50 was for the local wing). The fishing associations also determine the depth of catch and the timings of when fishing is permitted. Post-harvest practices As noted above, the turnaround time for fish sales was fast with sales taking place either at the lakeshore or via motorbike trades, and no real post-harvest practices were used by these fishers. It could be the case that the lack of any post-harvest management led to the immediate sales at the bank, and that this was perhaps a benefit and a curse for the fishers. While lakeshore buyers were convenient for the fisher, it was possible that the lack of post-harvest management limited fishers to these few buyers and made it impossible to seek other buyers further away (for instance, by supplying to restaurants) who could potentially pay higher. It took approximately three to four days for small dried fish to be peeled, dried and prepared for sale. Women were predominantly involved in this process, although men were known to assist if they had the time.129 Income from coastal fishing boat labour Only one household was interviewed involved in coastal fishing as labour, limiting the Study Tea s u de sta di g of i o e f o this aspe t of the fishi g i dust. Withi this o e household, it was said that, as with reservoir fishing, the income was highly irregular because the fishers needed to catch enough fish to cover the costs of the trip (e.g. a minimum of 15kg of prawn) and, if there was a shortfall they made no income for that trip and would have to cover the shortfall in the next trip. It was said there were approximately eight to ten trips per month, five to seven of which would be profitable, and each successful trip would result in an earning of LKR Both the husband and wife were engaged with the fishing boat, with the wife being paid LKR 200 per boat to assist with unloading and sorting. Income from fish factory labour One household was interviewed in relation to fish factory labour. The respondent reported that this work was often irregular and depended on the fish factory receiving a steady stream of raw materials. As this did not always occur, the workers were called in only if materials arrived. Within the fishmeal sector, the day rate for women was said to be LKR 500. The woman interviewed worked up to 10 days per month for which she earned LKR Compared to this, in the canning factory she would receive, she said, a monthly salary of LKR 16, SLMG-HH Market Development Facility

65 Key Constraints and Opportunities for Poor People in Fisheries? The government and fishing societies struggle to manage fishing stocks and licences effectively The overarching constraints for these households were a lack of effective resource management by the government and mismanaged membership of fishing associations. The government stocked fish intermittently but lacked the resources to do it consistently and in the required quantities. Private companies sometimes stocked the reservoirs, but there were few with hatcheries producing larvae or fingerlings. This meant that an increasing number of households were competing in the inadequately stocked reservoirs. An annual fee was paid to an association or a cooperative to ensure fair access to fingerlings, licensing and time allocation, however households said there were not enough fish to go around, and an influx of new fishermen. One respondent stated: Despite the fa t that ou i o es ha e i eased as a esult of i eases i the ua tit of the catch, the fact that the parent committee keeps on accepting new members and we already have four hundred is now putting the size of our catch at risk. We do not make a good haul of prawns because the hatchlings do not breed and the supply has now depleted to such a level that e o l at h a ou d t o kilos pe a u. 130 A othe fishe a said, If e at h fi e kilograms of fish today, it means we have five kg less to o o, i.e. e ause fish sto ks e e ot ei g epla ed.131 Historically, fishing cooperatives in Sri Lanka have been oriented toward promoting welfare and inclusivity rather than resource management.132 When asked why it was that cooperatives were unable to limit new licenses, respondents said that fishing cooperatives lacked the appropriate influence with government agencies that were responsible for stocking, or that they were unable to manage the increasing influx of new members and their needs. Fishing households were unable to cover the cost of inputs, in particular, net repairs and replacement. With the increased costs of licences and low stocking resulting in low catches, households increasingly struggled to cover the cost of nets. Nets had to be purchased with a lump sum payment which households could not afford given their meagre monthly earnings, and this often resulted in them taking out unfavourable loans, further compounding their costs. The FGD with fishermen revealed that a finished net (including materials and construction) could cost LKR 13,000 with each fisher needing approximately six to seven nets to fish optimally.133 The cost of nets could amount to LKR 91,000 in a season. In a household interview, one fisherman stated that he spent LKR 50,000 a 130 SLMG-HH-04, Polonnaruwa, North Central Province 131 SLMG-HH Amarasinghe, O., and Bavinck, M., Building resilience: fisheries cooperatives in southern Sri Lanka. In Poverty mosaics: realities and prospects in small-scale fisheries (pp ). Springer Netherlands. 133 SLMG-FGD--01, Polonnaruwa, North Central Province Market Development Facility 53

66 year on nets which amounted to almost 18 percent of his annual fishing income and that he had to take out a LKR 30,000 group loan from a micro finance institution (MFI).134 Another household respondent said she needed to take out a loan just to repair nets and that this problem with nets was common throughout the community: My husband only catches tilapia and he uses ten nets to do this. To repair the ten nets we require close to 30 bundles of nylon thread each costing LKR 1,500, and since we did not have the finances, I had to take a loan or else we would have lost our primary income. This has happened to a few families in the area. They own a kayak but, because they cannot afford to repair the nets, they can o lo ge fish a d ofte e ai at ho e allo i g othe s to use thei ka aks. 135 Fishers were forced to lease nets or borrow kayaks Another household respondent stated that the lack of resources to purchase nets (which he calculated at LKR 13,000 per net) forced him to lease nets from another individual, which meant he had to give 25 percent of his income to the lessor. He had little negotiating power with the lessor as he could not sustain his livelihood without nets.136 Another fisherman said he had to give half of his catch to his lessor.137 Similar issues were found with boats which had, in some cases, not been provided since government support in the 1980s.138 Even the two most well-off households had applied to the go e e t s Ce -Nor Development Programme to obtain one of the four kayaks that was available, with one household noting that it was very difficult to fix a damaged kayak and damaged kayaks were often better off simply cast aside.139 In the Parakrama Samudra Reservoir, it was observed that fishers lent each other their kayaks at no cost to support one another.140 Even with increased stocking, poor and vulnerable households may not be able to reap the benefits from this, until they can recoup their initial set-up netting costs. There was a risk that less wealthy households may get sidelined, or that their catches would be disproportionately smaller. Opportunities for diversified incomes were few in fishing communities Fishing was a time-consuming, family activity requiring families to be close to the fishing location throughout the day and thus left them unable to pursue other income generating options. The land surrounding fishing locations was generally poor and unsuitable for agriculture, and families could not use it to diversify their incomes. Some were able to supplement their fishing income as day labourers or with driving jobs. There was a need to strengthen the income potential of fishing, as it appeared there were few alternatives for income diversification. 134 SLMG-HH-03, Polonnaruwa, North Central Province 135 SLMG-HH-05, Polonnaruwa, North Central Province 136 SLMG-HH-07, Polonnaruwa, North Central Province 137 SLMG-FGD-01, Polonnaruwa, North Central Province 138 SLMG-FGD SLMG-HH SLMG-HH-04 and SLMG-HH Market Development Facility

67 Irregular incomes restricted the ability to forecast expenditure and borrow Fishing incomes from both reservoir and marine fishing were irregular, leading to limitations on financial planning and borrowing practices, compounded by poor financial management practices in fishing households. Fishing incomes were dependent on the size and quality of the catch, and these were unpredictable. Households were therefore unable to effectively forecast their income in a given period. For marine fishing this was exacerbated by sharecropping arrangements which regularly resulted in losses that had to be made up over the course of the month, leading to a need to borrow money. This means these households were paying more for basic items such as fishing nets. Families making comparatively better incomes did not appear to be making measured financial decisions with the additional incomes, choosing instead to spend on social obligations rather than investing little into economic activities. Pathways out of Poverty in Fisheries The pathways out of poverty for households associated with inland and coastal fishing, and seafood processing were determined as follows. Pathway 1 Facilitate private companies and/or public-private partnerships to address the poor management structures that have resulted in low stock levels in reservoirs As discussed, the overarching problem for these households was the increasing competition from an ever-increasing pool of licensed fishermen and the inability of fishing associations to keep up with the additional stocking this required. There was a place for the private sector and potential public-private partnerships to address the stocking issue in particular, and to look at some of the constraints leading to the poor functioning of the fishing associations. Private seafood buyers and vendors were likely to be more efficient in stocking the reservoirs as they had an incentive to purchase more fish from the households. Pathway 2 The impact of the unplanned fishing net and boat purchases could be mitigated with buy-back schemes and better credit systems The Study revealed that poor fishermen in particular struggled with the unit cost of nets as a proportion of their overall revenue, as well as the large payment required for such nets. Similar problems existed around the kayak market. Private companies may be incentivised to work with fishers to provide nets at a more affordable payback rate based on a guarantee of a proportion of sales so as to ensure that poorer fishers can take advantage of the reservoir restocking of fish. Pathway 3 Alternatives to the direct purchasing from the lakeside could be explored to better connect fishermen with more lucrative markets There may be potential to build stronger linkages between fishing and the tourist and export markets, rather than going through local traders. There may also be the potential for better market segmentation for fish (whereas some fish could continue to be locally consumed, others Market Development Facility 55

68 could be linked to more specialist traders e.g. prawns which attracted a high price and were geared toward high-end consumers). This would require improved sorting and grading practices. Pathway 4 Improved post-harvest practices around fish storage and drying Fish sales were immediate and transacted on a daily basis. Dried fish practices were rudimentary, but, by extending the shelf life of the produce, both fresh and dried fish, this could create new markets and higher value sales. Roles Performed by Women in Fisheries Below are snapshots that denote some of the characteristics of roles that women occupy in fisheries. Women in households The majority of women were engaged in preparation and post-harvest activities, including preparing and repairing fishing nets, harvesting fish from the nets, and making and selling dried fish. As was the case with men, female respondents had taken up fishing-relating work as handed down to them by previous generations. A large proportion of women appeared to be engaged in dried fish141 a process which required gutting, washing and brining a small fish called binthulayo and drying it on a wooden frame for three days, before selling it for approximately LKR per kg. The practice appeared to be an increasing trend, with one female respondent stating that when she started drying fish 22 years ago, only five or other six women in her community were doing it. She now believed that 90 percent of households were engaged in fish drying.142 As fishing activities were very much family-based, women did not work in isolation; men were also seen assisti g thei i es ith o e o a e plai i g, M hus a d ofte assists e ith d ied fish aki g he usuall helps i getti g the fish ead fo i i g. 143 There was sometimes a flippant perception by male household members that fish drying was more of an activity that could fill o e s ti e, athe tha ou t as a ea i gful i o e st ea, despite the sa i gs pote tial it could bring.144 Within coastal fishing (around boats), women were said to work in groups assisting in sorting and loading and would typically work one to two shifts consecutively (a shift would be around three hours) and be paid LKR 300 per shift. It was common for women to liaise directly with boat owners to get their husba d s oat sala i additio to thei o age.145 Wo e s de isio aki g a d i flue e regardi g i o e fro 141 The fish is dried because it is considered valueless in its fresh form. 142 SLMG-HH-03, Polonnaruwa, North Central Province 143 SLMG-HH-05, Polonnaruwa, North Central Province 144 SLMG-HH-03, Polonnaruwa, North Central Province 145 BILR-HH Market Development Facility household fish produ tio

69 Withi fishi g households it as o o fo e to ha d o e thei i o e f o the da s at h to their wives, and for the wives to manage the household expenditure.146 In this way, fishing o u ities follo ed the household e te p ise t pe set-up whereby the husband and wife worked together to manage the fishing labour and household cash management.147 This made women responsible for the savings from the central household income, and this was ost o o l kept i a ja at ho e, o pla ed i hild e s sa i gs a ou ts i the a k.148 Women in these fishing communities were also found to be engaged in savings schemes with other women149 such as BRAC savings schemes.150 Women gained an additional, small income from binthulayo sales the use of which they had direct control over. One woman spoke of how she had used this to purchase a refrigerator, something her husband would not have approved of, but had helped her to cover the costs for.151 Women as lead fishers and at work There was a small, albeit notable, number of women who partook in fishing tending to catch small fish for drying, and not in competition with men for the bigger catches. One female respondent and her sister were two of only eight women in their community who were registered members of the fishing cooperative. She decided to fish out of necessity. Interestingly, she only fished for binthulayo for drying purposes because her husband was not able to catch them with his nets. She suggested that other female fishers also only caught binthulayo. When asked why other women did not fish, she said their husbands bring in binthulayo from their catch.152 This suggested that women fishers existed in small numbers and they operated in isolation from the male fishers. Another female household member said she learned to fish as a precaution in the event that her husband could not work, and thus her normal daily activities involved assisting her husband, rather than fishing with him.153 An interesting opportunity came from a female respondent who was making LKR 60,000 from working in a fish-canning factory. From an overall household income of LKR 354,000, she was contributing about 17 percent of the household income. Here it could be seen how a woman's skills and experience with fish preparation could transfer into paid employment. This factory had the conditions that dedicated female workspaces were provided so that women had socially acceptable working conditions. 146 SLMG-HH McGoodwin, J. (2001). Understanding the Cultures of Fishing Communities: A key to fisheries management and food security. Food and Agriculture Organization of the United Nations (FAO). Technical Paper SLMG-HH SLMG-HH-03 SLMG-HH SLMG-HH SLMG-HH-05, Polonnaruwa, North Central Province. 153 SLMG-HH-07 Market Development Facility 57

70 Barriers for Women in Fisheries Fishing is a male dominated industry so women are unlikely to be involved in catching or selling fish. Fisheries work was considered 'risky' due to the labour involved, so women (except for some outliers) were unlikely to be involved in mainstream fishing. Those engaged in it were more often seen to be outsiders undertaking the activity alone (away from men and other women). With low fish stocks and competition for and a limited number of fishing licences available, licences would go to the male household members first. Even if women overcame the hurdles of the male nature of the industry, they would be unlikely to be granted a fishing licence, which are effectively rationed between families. Fishi g o k spa es, i ludi g t ade spa es, a e ale It is difficult for women to enter the industry without being conspicuous, as selling occurs from the fishing boats or at the shore. There was little value that women could add beyond repairing fishing ets. A FGD ith the i es of fishe e e ealed that a o a s p ese e at fishing areas was culturally unacceptable.154 So ial o s ega di g o e s eputatio s posed diffi ulties fo o e i selli g thei d ied fish or expanding their post-harvest businesses (in particular, in accessing and liaising with buyers). Selling dried binthulayo was an entirely home-based operation and sales were to buyers coming to their homes several times a month. 155 It was also an isolated business in that women engaged in binthulayo drying operated alone. In one Study household, the female respondent said she wanted to expand her business by packaging her dried binthulayo and selling them in the market and to shops. However, this required her to leave the house and her husband was opposed to this. He clarified that he was concerned that people would talk about his wife and impugn her reputation as they lived in a small community. Further, if his wife was to undertake this business, their older daughter would have to accompany her, and put her reputation at risk.156 Few opportunities existed for women to apply their cutting and preparation skills in production and processing workplaces. As seen in the case of a female worker transitioning from her household-based gutting and drying te h i ue to the o kpla e, o e s skills i the p epa atio of fish an transition well to the workplace, however there are few fish processing type set-ups that would offer women such opportunities. 154 SLMG-FGD SLMG-HH-04, Polonnaruwa, North Central Province. 156 SLMG-HH-03, Polonnaruwa, North Central Province 58 Market Development Facility

71 Recommendations to Address Barriers for Women in Fisheries The role of women interacting with fisheries either directly, or as members of households associated with fisheries could be encouraged and improved in a number of ways. The study identified the following areas. Expand opportunities for women to engage in post-harvest processing by supporting fish processing facilities which may lead to employment generation The promising example of a female respondent who worked in a canning factory in Mannar whose experience in cutting up fish could be transferred into an income earning opportunity with a significant contribution to her household income (17 percent).157 If mobility was an issue, then ideas regarding dedicated workspaces and transport options might be explored. Supporting such industries could be positive in generating employment opportunities for women in this sector, where opportunities were limited. Seek to alue add a d i pro e the sales of dried fish a d o e s a ess to traders Women were engaged in fisheries, but tended to work within their own households and therefore dealt in isolation from buyers. There was no community of women fishers as there was for men. There was a possibility to explore networking or an association that focused on dried fish to turn it into a less ad-hoc business for the community. Binthulayo could potentially increase their income and diversify the markets for these households. Ensure assistance for lending and household financial management practices are aimed at women Women are the financial managers in fishing households and are those likely to make use of loan services. Any practice aimed at improving the cash flow or borrowing within these households should target the women of the house, as they are the drivers of financial habits in these communities. Households identified small retail shops, particularly around fish netting, as something they would like to invest in. As this would be located close to home it could be possible for women to be involved if set up in an appropriate manner. Explore the potential for women who wish to be engaged directly in fishing This would be contingent on unlocking the fish stock issue but, if resolved, it appeared there was a community of women who wanted to engage in or extend their engagement with fishing. Should there be more opportunities for fishing, it would be important to ensure women as well as men could avail themselves of these. This might require adjustments to the way work was organised, such as women working in groups to fish, and selling taking place separately (as noted, this work area was male dominated). 157 BILR-HH-02. Market Development Facility 59

72 60 Market Development Facility

73 Horticulture Methodology and Sample Characteristics To understand poverty dynamics within horticulture, household interviews were conducted with 18 farming households. Geographically, the sample covered six districts across five provinces of Uva, Eastern, Southern, Northern and Central provinces. In terms of gender, the primary respondents comprised 14 men and 13 women (i.e. both husbands and wives were interviewed where possible). The study focused on interviewing smallholder farming households, but also included two outlier households with over 20 acres of holdings for comparative purposes. The Study attempted to capture as much diversity of crops as possible, including rice paddy, chilies, coconuts, black pepper, banana, ground nuts, papaya, onions, okra, gourds, peanuts and eggplants. The household interviews were supplemented by a KII with a horticulture trader to gain further insights into the horticulture market dynamics. Who is Poor in Horticulture and Why? The characteristics of households associated with horticulture are explored in detail below, including variances and trends within both the poor and less poor households in the sample. Profile While farming was considered a noble profession in Sinhalese culture, the majority of farmers were engaged in low yielding, rain-fed paddy production and lived below or the near poverty line. Even for those that diversified into other crops, low productivity and high input and cultivation costs meant they generally had low earnings. Crop wise, farmers in Sri Lanka cultivating paddy, maize, black gram, ground nut and cabbage were all living in poverty owing to poor productivity and limited output. Farmers doing spice, onions, chili, brinjal were better off and had decent levels of incomes. Geographically, farmers in low country, dry zones had a higher likelihood of being poor. Agricultural workers in some cases were better off then farmers, however they too operated in the near poor category and often relied on non-farm income (construction) for more steady, yearround earnings. Ethnicity had less bearing on who the poor were in this segment. Poverty and per capita income The majority of horticulture households interviewed were poor or vulnerable. The wealthiest households were doing markedly better in comparison to others; the top two wealthiest households had larger land areas in comparison to the rest of the sample, and exceeded the size of holdings that would qualify them as smallholders BILR-HH-21 and BILR-HH-19. Market Development Facility 61

74 Income per capita The households analysed included some of the poorest households captured in the overall study. The poorest household presented somewhat of an anomaly. Its situation had recently changed drastically because, despite the respondent having saved a significant Figure 5: Distribution of per capita income in horticulture households amount of money while working in the Middle East, he had made some unwise LKR LKR investment decisions on his farm, LKR leading to an income loss of almost LKR LKR million.159 Other households LKR presented more conventional LKR LKR circumstances, including a sharecropping farmer in the rain-fed Number of households dry-zone of Monaragala who cultivated mainly for household consumption and could only farm in the Maha season due to water restrictions.160 A farmer in Monaragala supplemented his income with day labouring jobs161 and a farming household in Jaffna lost all its assets and savings in the war.162 A female-headed household where a wife had been abandoned by her husband had, in 2015, been resettled after the war on a temporary shelter and small piece of land;163 and a paddy sharecropper in Galle also grew vegetables.164 Expenditure, debt and borrowings The poor and vulnerable households were deeply in debt. The poor households were deeper in debt than the vulnerable households. The poorest household had an average monthly shortfall of LKR 205,853 due to a failed investment.165 Even when excluding this household, the remaining poor indebted households still had an average monthly shortfall of LKR 12,333 (a three-fold difference between the poor and the vulnerable) and an expenditure-to-income ratio of 3:1.166 As expected, the largest percentage of household expenditure was for basic costs, with more than half of the income going toward food costs. Poor households spent an average of 65.3 percent of their expenditure on food, whereas vulnerable households spent 44.4 percent. This was in comparison to the wealthiest two outlier households that spent 24.4 percent of their expenditures on food. Poor families confirmed this in interviews, reporting there were usually a few months in the year when they struggled to meet basic costs such as food expenses and were 159 BILR-HH SLMG-HH SLMG-HH BILR-HH BILR-HH SLMG-HH BILR-HH BILR-HH-03, 10 and Market Development Facility

75 unable to meet daily expenses.167 One family described how only recently they had been able to increase their food intake from two to three meals a day.168 Notably, however, non-essential items were being bought even by the poorest households. One household had no furniture, but a TV, DVD player and an electric fan,169 and this scenario was evidenced in other households sampled. Most households across the sample had borrowed money, exhibiting no correlation with poverty levels. The three wealthiest horticulture households borrowed large sums to invest in agricultural activities, ranging from LKR 850,000 to LKR 3 million.170 All three of these households had taken large loans through formal financing e.g. bank loans or leases, related to their agricultural activities such as, in one household, to buy a tractor.171 Vulnerable households and poor households, on the other hand, tended to be reliant on informal financing or on Samurdhi sources, and used their borrowings for a range of expenses, including farming-related activities to basic household expense and items such as bikes and TVs.172 Smaller households were more reliant on informal financing173 e.g. a female respondent noted that, like others working in the horticulture sector, she did not have access to formal lending services from private institutions such as banks, but that there were alternatives such as MFIs and government loans. Both vulnerable and poor households borrowed money from friends and family members.174 Smaller farmers seemed to prefer to undertake day labour in times of need rather than use a loan service.175 An agri-inputs shop interviewed as part of the stimulating entrepreneurship in former conflict-affe ted SEA sa ple said it as o o fo ediu -size farmers (especially paddy fa e s to u i puts o edit a d settle afte the ha est, 176 and another paddy farmer confirmed he obtained a loan of LKR 25,000 from a trader to purchase seeds, as his crops had been damaged during flooding.177 In Moor communities, moneylenders charged no interest for religious reasons, and there were no real repercussions for late payments, apart from a s oldi g.178 One woman had borrowed from Samurdhi for business purposes (LKR 50,000).179 Government welfare support appeared to be reaching poor horticulture families, especially in the North and the East. Most of the households receiving support (including Samurdhi) were located 167 BILR-HH-17, BILR-HH BILR-HH BILR-HH BILR-HH-04, BILR-HH-19, BILR-HH BILR-HH SLMG-HH SLMG-HH-21 Samurdhi loans and SLMG-HH-13 informal financing 174 BILR-HH-17 LKR 50,000 from friends. SLMG-HH-20 from friends, and BILR-HH SLMG-HH BILR-HH BILR-HH-20. BILR-HH-12 The family borrowed LKR 100,000 from Samurdi three years ago and this had helped with expenses when he was detained and the family was struggling. 178 BILR-HH BILR-HH-15 Borrowed to purchase a sewing machine. Market Development Facility 63

76 in these regions, and support appeared to be reaching the poorest families in the sample.180 In a Moor village in Batticaloa, however, one respondent was bitter that his family had received very little Samurdhi support compared to, in his opinion, the Tamil communities surrounding them.181 Non-monetary support was not common in these horticulture households, and was only received in the former conflict-affected areas (e.g. one household had received goats f o the Ca ee De elop e t Fou datio.182 Assets (including land) Beyond large landholders being wealthier, there was little correlation between smallholder land size ranges and poverty. There appeared to be little correlation between land size ranging from 0.03 acre up to 4 acres and per capita income. The average plot size for households falling under the poverty line was Again, excluding the outlier poorest household, the average land size was 0.7 acres. Notably, among the poorest households was the only sharecropper of the sample. Discounting those with very minimal costs for land, few households were paying costs in relation to their land leasing/ownership, the highest being LKR 50,000 per year by one a householder for leasing from his mother (he was involved in a land dispute case with her because of it),183 with the others ranging from LKR 5,000 LRK 25,000 for various costs, such as leasing from other farmers.184 In the Northern and Eastern provinces, poorer farming households had few or no assets,185 as many pre-existing assets had been lost in the war, and current incomes were not enabling families to save.186 One family noted they did not have any savings at the moment as a result of their recent business expansion.187 This is particularly concerning given the inability of a big proportion of families to meet basic monthly expenses, and indicated how vulnerable they were, even where they apparently sat above or around the poverty lines. 180 BILR-HH-17 Get LKR 650 per month BILR-HH-07 He paid in LKR each month and gets a lump sum once a year, which varies in amount but he got LKR 8000 this year. BILR-HH-20 Paying in LKR 700 to Samurdih every month. BILR-HH-06 He is getting per month in Samurdhi support. 181 BILR-HH BILR-HH-12, BILR-HH-20 and BILR-HH-15 had all received non-monetary support. 183 SLMG-HH BILR-HH Assets were not explicitly covered in this study and we only discuss them briefly here in passing according to questions asked around the PPI 186 BILR-HH-03 lost all savings in the war. Now they have a fixed income and plan to increase their coconut plantation. BILR-HH-04 had no savings. BILR-HH-12 dad jewellery, but it was lost in the conflict, and now have little savings to speak of. BILR-HH-17 had high savings from the Middle East, but invested it all in the business i.e. on goats, over 100 of which died. 187 BILR-HH-17 and BILR-HH Market Development Facility

77 Both men and women were holding bank accounts.188 One woman had prioritised savings for her children even though she could not save for herself189 and all tended to invest in gold jewellery in the rare cases they had extra funds, which they pawned in cases of need.190 Poverty and other basic household characteristics Within the sample, there was no correlation observed between geography and poverty. Given the small sample size, it was difficult to draw any conclusions regarding regional or provincial differences. Similarly, there was no correlation observed between household size and poverty. The average household size was four, which aligned with the average household size in Sri Lanka (at 3.9 nationally, and 3.8 in the rural sector). Respondents were generally well educated with most female household heads having been educated to Year 8 or higher, and a significant number having been educated to GCE level and above. Reliance on the horticulture sector as an income stream There was no real correlation between reliance on horticulture as an income source and per capita income although notably, households relied on agriculture for about 70 percent of their income. Other sources were casual manual labour,191 tailoring,192 driving,193 salaries from a government job,194or receiving international remittances,195alimony,196 pension,197 or payments from Samurdhi, especially in the northern and eastern provinces. Income from horticulture production Crops and pricing trends The households in this sample were found to be cultivating between one to three horticultural crops each on an average land size of acres (excluding the very large farmers). Net incomes from farming amongst the smaller farmers (excluding the larger farmers in the sample) ranged considerably, from LKR 12,500 per year all the way up to LKR 323,250 per year. 188 BILR-HH BILR-HH-02 Mother had deposited LKR 10,000 in a bank for each of their three children. 190 BILR-HH BILR-HH BILR-HH SLMG-HH BILR-HH-19 comprised a farmer whose wife was a teacher. 195 SLMG-HH SLMG-HH SLMG-HH-18 Market Development Facility 65

78 Looking specifically at the crops being grown by the smallholder households (excluding crops grown by large land holding farmers that were not within the feasible scope of what small farmers could grow), the Study found that crops grown with the highest potential prices were chilies (LKR 1000 per kg being the highest price paid, although this ranged considerably), black pepper (LKR 420 per kg the highest price paid), and peanuts (LKR 250/200 per kg highest price paid). Beyond this, the highest price for the remaining crops dropped to LKR 100 and below, with paddy and corn at the bottom of the pricing. The Study Team noted that farmers were getting a high price range for their crops, and this appeared to be due to seasonal variations related to rainfall. For example, the price for chilies ranged from LKR 70 to LKR 1000 per kg, and even within the same household the price varied by up to 100 percent. Farming households were aware of which crops could offer the best profit margins but they could not grow those for various reasons. A farmer in Galle noted bitter gourd (although it was said to be labour intensive), lady fingers (taking about a year to harvest), and ridge gourd (three months to harvest) were more profitable.198 Cinnamon was also said to be in good demand, but did not suit the climatic conditions in Galle.199 Black pepper was noted by two farming families in Monaragala to bring farmers a good income,200 but the crop had not done well in the region that year, with one farmer also saying he could not keep his pepper plants alive during the dry season.201 Pumpkin and watermelon were said to provide good yields and a better income, and one farmer in Monaragala said he would grow these if he had the capital.202 Looking at the specific horticulture sector picture for these households, farmers clearly recognised the potential of growing horticulture crops, and tended to pick horticulture crops to grow based on a o i atio of thei skills k o ledge a d the op s p ofita ilit.203 One farmer based in Galle noted that vegetable prices were rising and there was a good opportunity to promote a variety of vegetable crops around his property.204 One family spoke of their plans to build a nursery where they could sell their plants, and the estimated cost for the materials required for this was said to be around LKR 30, There was also a high market demand for organically produced crops.206 Overall production costs Poorer households tended not to purchase inputs, especially seeds, and spent low amounts per year on fertiliser. One farmer household noted that, although they had access to agri-inputs, they 198 SLMG-HH SLMG-HH SLMG-HH-21 and SLMG-HH SLMG-HH SLMG-HH SLMG-HH BILR-HH SLMG-HH SLMG-HH Market Development Facility

79 preferred to produce their own home-grown inputs and seeds in order to save on costs.207 They were spending LKR 7,500 per year on fertilisers and pesticides and in rare cases hired extra labour at a cost of LKR 1200 per day. The household said they only bought these when they were able to earn extra money, and if not they produced their own fertilisers.208 Farmers spoke poorly of traders in this regard noting that, due to a lack of competition, they felt that traders were charging them high prices (this, of course, was their point of view only, and perhaps did not take i to a ou t the t ade s ealities. Si ila l, poo e fa e s e e u a le to ope ith the i itial investment and subsequent running costs of key irrigation measures (such as building wells and paying for electricity for the pump), in turn vastly reducing their potential growing seasons. Post-harvest practices Equipment and labour was hired, with even some of the poorer farmers utilising a tractor (costing LKR 1500 to hire per day in Galle and at an overall cost per year of LKR 6,000 for another farmer in Monaragala).209 Trucks and three wheelers were often hired to transport goods to market after harvest.210 One farmer described how produce was packed in gooney bags and then transported to the wholesale vendors via motorbike for which he was paid in cash.211 The harvest was consumed as well as sold externally.212 At times when a household could not meet its regular living costs, crops grown for sale were substituted for purchased food.213 Livestock production Livestock rearing on farms was found to be uncommon with only four households rearing poultry and/or goats.214 Distance from markets In terms of distance from markets, notably the best performing farm was located in the agricultural heartland of Dambulla,215 whereas the poorest farmers were either very distant from markets or close to a local market only, where they could not negotiate prices. Proximity to such markets on one hand was said to reduce transportation costs, but on the other was said to make it difficult to expand, because of the difficulty in being able to afford the leasing costs of the land around markets SLMG-HH SLMG-HH SLMG-HH-20 and SLMG-HH BILR-HH SLMG-HH SLMG-HH-20, SLMG-HH-22 and SLMG-HH SLMG-HH BILR-HH-21, BILR-HH-15 and BILR-HH BILR-HH SLMG-HH-11 Market Development Facility 67

80 Smallholder farmer use of farm labour Farmers themselves hired day labour, particularly during preparation and harvest seasons. For example, one farmer said he required four to five labourers for two days to harvest rice and chilies.217 Labourers were also hired for paddy. Labour expenses ranged from LKR 3,000 a year to LKR 8,760,000 a year. On average, the amount spent on labour was about 17 percent of the overall revenue. As discussed below under constraints, access to and costs associated with labour hire were often high. Access to agricultural information In terms of sector-specific information provision and support, reactions to government agricultural extension support varied. In Galle, agricultural extension offers described as being well qualified visited the village from time to time and, because of this, plantation procedures improved.218 Contrastingly, in Monaragala there were said to be agricultural extension officers for e e illage, ho e e the household oted the did t feel o fo ta le talki g to thei offi e because most of the requests they had made related to water were not successful.219 One family noted they had received (previous) government support for their plantation business via the Di i agu a s he e he e his p odu e as pu hased the go e e t.220 Another householder mentioned he had used government soil testing services.221 Income from farm labour Poor households supplemented their income by providing farm labour to others in their communities, particularly during harvest seasons. Multiple study respondents stated that on days that they were unable to cultivate their land, they provided labour services to others in the community for additional income.222 This was particularly the case when they struggled in a particular week or month, the male household member found work on another farm.223 Those from destitute backgrounds, on the other hand, relied on farm labour as their primary source of income and primarily work in weeding, staking and harvesting.224 These labourers were hired at a fixed daily rate of between LKR for women and LKR for men.225 Although 217 SLMG-HH SLMG-HH SLMG-HH SLMG-HH SLMG-HH For example, see SLMH-HH-13, where it was said that this a common practice, and SLMG-HH SLMG-HH-13, Badulla, Uva Province. BILR-HH-12, Trincomalee, Eastern Province. SLMG-HH-12, Monaragala, Uva Province. SLMG-HH21, Monaragala, Uva Province, and SLMG-HH BILR-HH-06, Jaffna, Northern Province. 225 BILR-HH-17, Batticaloa, Eastern Province. SLMG-HH-20, Galle, Southern Province (1200 LKR for labour). BILR-KII-1, Batticaloa, Eastern Province. 68 Market Development Facility

81 being heavy physical work, there were advantages, with one farmer noting this income source allowed for a quick release of funds until the crops are ready to be harvested.226 Income from nursery production labour Two female nursery farmers were interviewed as part of the sample. The nursery was said to hire a lot of female labour. Both respondents said there were few opportunities for formal horticulture labour and the work provided on-the-job training.227 Notably, their training was provided by the wife (of a husband and wife nursery business), who not only helped to train the women about nursery techniques, but also helped teach the women about household management and money management. The female staff spoke fondly of her, and she was respected in her position in the business, making her a suitable trainer (such in-house family-run training provision has been witnessed in other MDF countries, such as Pakistan). Nursery jobs entailed preparing bags, planting, watering, pruning, harvesting and cleaning. A woman would earn an average of LKR 550 a day, with no additional benefits, which was comparable to a day la ou e s ate fo a o a, ut i ol i g less ph si al e e tio. Income from agri-inputs dealing Only one agri-inputs dealer was interviewed for this Study (in Trincomalee), so the sample is naturally restricted. The household was doing comparatively well with a per capita income of LKR 482, the fourth highest of the horticulture households interviewed in this sample. This dealer sold seeds, saplings, fertilisers and small agro-equipment, obtaining his items from the largest distributor in Dambulla. He noted that the farming industry around him was in decline due to unpredictable climatic changes and the cost of production. All the medium-size farm enterprises which bought inputs from his shops always bought on credits and settled their accounts after the harvest. This was a common practice for paddy farmers. Sixty percent of the customers who bought agri-inputs were women. They seemed to know the agri-inputs well and the respondent said he gave them advice on how to apply/use the agri-inputs. Key Constraints and Opportunities for Poor People in Horticulture Lack of irrigation systems and the means to buy it affected productivity and incomes In areas with significant dry seasons (Moneragala, Trincomalee, Batticaloa and Jaffna), poor smallholder farming households relied primarily on rain-fed agriculture, and had little access to irrigation technologies. Rain-fed agriculture limited outputs of small plot sizes that characterised the majority of Stud households. The ate o st ai t e ai s u esol ed due to households i a ilit to affo d upfront investment costs. Lack of any/sufficient irrigation systems was a problem during Sirum Poham (the irrigated season) in Trincomalee and Batticaloa228 and was cited as the largest constraint 226 SLMG-HH SLMG-HH BILR-HH-15. Market Development Facility 69

82 to increased productivity and diversification of crops in the Uva Province and in Moneragala. One respondent stated that the cost of digging a well would be around LKR 20, Another household said the labour cost alone to dig a well would be approximately LKR 28,000, and the overall cost would likely amount to over LKR 100, As o e household e plai ed, If I had a well, I could do some crops and take care of my household expenses. 231 In the Eastern Province, several study households irrigated their land through a canal system that was built to channel water from a tank.232 One village in Batticaloa utilised a village rainwater pond and canal system for which they paid a LKR 20 annual usage fee, but said this was still insufficient.233 One slightly better-off farmer was paying LKR 15,000 per month for associated water pump costs.234 Another farmer in Jaffna noted that only well-off farmers were able to afford irrigation.235 Beyond the upfront investment, the electricity costs associated with irrigation could make this a prohibitively high cost for farmers, who required lower cost energy solutions to remain viable. Cropping practices did not appear to maximise land potential and advise was not available Because of the dry seasons, farmers seemed to have very basic cropping pattern practices, doing little to maximise the potential of seasonality. The Study Team sought to find cropping recommendations by government agricultural departments, but these were entirely lacking. It would be realistic to conclude that farmers received little or no support in relation to cropping pattern options. Flooding leads to wastage and crops losses Excess rain and flooding posed problems during the rain-fed seasons in Batticaloa and Jaffna,236 and sporadic rain and risk of flooding caused problems in Moneragala.237 In Moneragala, a farming household reported that the sporadic rain caused their expensive fertiliser to be wasted, leading to a net loss of LKR 1000 in cashews and LKR 10,000 in corn.238 In Jaffna, another farming household reported that the weather patterns caused them to lose their entire potato cultivation.239 Unpredictable weather changes were also cited to be a problem in Jaffna, with one farmer having recently lost his whole potato harvest because of it240 and another losing 100 goats to the rains in 229 SLMG-HH-13. Another household in Monaragala said it would cost him LKR 28,000 in labour costs alone, and the overall price would amount to more than LKR 100, SLMG-HH BILR-HH BILR-KII BILR-HH BILR-HH BILR-HH BILR-HH SLMG=HH SLMG-HH BILR-HH BILR-HH Market Development Facility

83 Batticaloa.241 Such challenges faced in the last year alone spoke to the level of household vulnerability, particularly given that the Study households only grew two or three crops on their plots. The fear of bad weather conditions could deter households from taking out loans, investing in inputs, higher (albeit riskier) productivity crops and affect their capacity to generate income. An agriinput supplier in Trincomalee observed that unpredictable climate changes and high production costs were leading to reduced farming activities in his local area.242 Low technology and poor extension services meant farmers were not adapting to climate change Smallholder farms appeared to be characterised by low levels of technology and limited knowledge on, climate-resilient resource management and technology, with the poorest farmers in particular gaining little to no extension support (private or public). Farmers were unaware of the reasons for major crop failures. For example, one household in Moneragala stated that their black pepper cultivation was unsuccessful, leading to zero agricultural income. However, when probed as to the reason behind crop failure, they claimed this was a widespread occurrence throughout the area, but did not know precisely what caused it.243 The poorest farming households noted they did not receive any extension service support, either in a direct or embedded form. Particularly for farmers in former conflict-affected areas, who were returning and starting up in farming for the first time or after a long gap, this lack of information could lead to poor investment decisions that were unsuitable for the climate. As noted above, one farmer had made the highly unwise decision to purchase 200 goats with money saved from overseas earnings but, with no farming experience, over 100 of these were killed in just one rainy season.244 Labour shortages meant land could not be cleared and cultivated There seemed to be a mismatch in the farm labour market. Despite evidence of the supply of labour, many households stated that a combination of labour shortages, and by extension the rising cost of labour, posed challenges in Trincomalee, Jaffna, Dambulla and Galle.245 While the extent of manpower needed varied from crop to crop, horticulture production overall is labour-intensive. There was a consistent demand for farm labour that was unmet.246 One farming household mentioned that one would need to travel four to five km in search of labourers.247 Another s allholde fa e had left a a e of la d u ulti ated e ause he had t ee a le to ha e it cleared since the previous harvest.248 Farmers in Moneragala dealt with this via farmer labour exchanges, said to be a common practice, particularly during the harvest season.249 Due to increased 241 BILR-HH BILR-HH-14, Trincomalee, Eastern Province 243 SLMG-HH-22, Monaragala, Uva Province. The household had to rely on international remittances, alimony payments and pension funds. 244 BILR-HH BILR-HH-14, BILR-HH-04 and BILR-HH BILR-HH-14, Trincomalee, Eastern Province. BIL-HH-20, Jaffna, Northern Province 247 BILR-HH-03, Jaffna, Northern Province 248 SLMG-HH SLMG-HH-21 Market Development Facility 71

84 urban migration and an increasing number of youth seeking higher education, obtaining labour for agriculture was becoming difficult and expensive.250 This information, however, contrasted with Batticaloa, where a farmer said he was unable to find agricultural work for more than five days a month,251 obtaining it through family contacts.252 The poorest farmers lacked access to major markets The poorest farmers tended to be distant from major markets, relying on small, local markets to sell their produce. While they were aware of downstream prices (since they themselves had access to supermarkets), they felt they were in no position to gain better prices because they could not afford the costs associated with getting produce to markets. This was particularly apparent in the North, the East and in the Uva Province. Famers had little option but to sell to local traders and low rates While farming households exhibited knowledge about market prices, they appeared to have limited bargaining power and they often complained that they sold their produce at discounted prices to traders and wholesalers. Farming households appeared to have access to end market pricing (e.g. one farmer knew of general price levels via visiting the local supermarket chain, Food City). Study households said they had knowledge of market demand and prices, but they lacked negotiating power with traders.253 One Study household respondent expressed frustration that he would have to spend considerable time and money transporting his crops to a main marketplace and get disappointing prices. While he could sell his goods to a nearby smaller market or to farm-gate traders, he would get much lower prices and still have no room to negotiate.254 Of note, one respondent household was aware of the rising demand for organic products, but they had to sell all their products to wholesale traders at a low rate, and did not see much profit.255 Those farmers who did not have farm gate sales faced heavy transport to get produce to distant markers In former conflict-affected areas in particular, long distances to buy and sell goods resulted in heavy transport costs, and poorer households complained that market prices were controlled by traders and they lacked bargaining power or alternative options.256 For example, one farmer had to travel 30 km to purchase agri-inputs257 and another said these were only available in the main city.258 Having to travel led to them to keeping back much of their production for household consumption rather 250 SLMG-HH-20, Galle, Southern Province 251 BILR-HH BILR-HH BILR-HH-20, Batticaloa, Eastern Province. BILR-HH-03, Jaffna, Northern Province. SLMG-HH-21. ssmg-hh BILR-HH-17, Batticaloa, Eastern Province 255 SLMG-HH BILR-HH-17 and BILR-HH-03, the two poorest farming households in this sample, both complained of this. BILR-HH-04, a better off household, also noted the market is controlled by traders connected to the Dambulla market. 257 BILR-HH BILR-HH Market Development Facility

85 than sale. Another paddy producing household said prices were controlled by the mills and farmers received a lower rate than was set by the government.259 Farm gate pick-up was thus preferred.260 Increased prices of inputs and loss of fertilizer subsidies inhibited farm growth Prices of some agricultural inputs, that were affordable before the war, were high in former conflictaffected regions. For example, red onion seeds, once available prior to the war, were now so expensive that farmers had switched to papaya cultivation as the seeds were more readily available.261 High input prices were repeatedly cited as a reason that farms were prevented from expanding.262 Lack of knowledge of cultivation techniques, lack of investment capital, and lack of access to inputs were cited as reasons why paddy farmers did not go into more lucrative crops in former conflict-affected areas.263 The discontinuance of the government subsidy for fertilisers (previously LKR 350 until April 2016) was said to be particularly challenging for farmers, who were paying LKR 2,500 for the same fertilisers.264 War-affected areas no longer had access to some materials Due to displacement caused by the war, raw materials and resources once found in the former conflict-affected areas had moved to other regions, which affected producers in these areas. This dynamic was evident in the sesame oil industry.265 Pathways out of Poverty in Horticulture The Pathways out of Poverty for households associated with horticulture are as follows. Pathway 1 Affordable irrigation options need to be made available in areas such as Trincomalee, Batticaloa, and Jaffna and Moneragala Farmers spoke often of lower yields or the inability to farm being associated with no/low access to water in Sirum Poham (in Trincomalee, Batticaloa and Jaffna) and Maha (in Moneragala), and noted this as a major restriction to farm income growth. Low cost irrigation that was not heavily reliant on fuel (with its associated high costs) would need to be considered, in conjunction with an environmental impact assessment to ensure that farmers were not sourcing water from vital and slow-to-be-renewed groundwater sources that were needed by communities. Other options could be drought-resistant crop varieties. 259 BILR-HH BILR-HH BILR-HH BILR-HH BILR-HH-20 The reasons why one paddy farmer had not diversified, and noted this was common. 264 BILR-HH BILR-HH-07. Market Development Facility 73

86 Pathway 2 Improved cropping pattern information should be developed and made available via appropriate channels There appeared to be an institutional gap in that cropping pattern information was not collated and analysed by agricultural departments, and in turn farmers did not receive assistance on appropriate cropping patterns. Inefficient cropping patterns, compounded by dry seasons, appeared to be resulting in highly inefficient use of farming seasons, and thus farmers would benefit greatly from information about how to better utilise seasonality. The mechanisms for this delivery would need to be explored further i.e. whether this would more appropriately be achieved through a public or private channel, and what the incentives for this might be. Pathway 3 In general, information on inputs, techniques, and investments could be improved, potentially via embedded information provision or via contract farming Farmers only interface with immediate traders, who it appeared, rarely provided much information. Extension services were also fairly lacking or non-existent. Farmers, especially those in former conflict-affected areas who were returning to the industry/entering for the first time, were unclear on how to invest. There was ample space to provide farmers with embedded information in a number of areas. The incentives for this needed further exploration, as this was tied to the types of products that would be able to be marketed to farmers, and the increased costs of new investments given they are already struggling to meet costs. In this regard, contract farming options could be explored in conjunction with market access. Beyond technical farm measures, this pathway should also consider financial literacy support, as cash flow issues are connected to poor investment planning and can hinder long-term growth. Pathway 4 Alternatives to the use of expensive fertilisers could be explored Compost is a cheap and easily accessible alternative to fertiliser, given that the government had removed its subsidy the price of fertilisers had increased substantially. The commercial case for this would need to be developed. Pathway 5 Crop and livestock insurance could be explored in Batticaloa and Jaffna, along with measures to increase agricultural resilience, particularly diversification and the spreading of crop cycling. Floods and heavy rains made crops and livestock a high risk investment in some areas in the North and the East. Crop insurance might be explored as a measure to mitigate risk. Other measures could include increasing agricultural resilience i.e. staggering planting and harvesting throughout the year, by diversification of crops or crop varieties, or looking into more resilient varieties (that can survive waterlogging or heavy winds. Pathway 6 Low-cost mechanisation and other labour-reducing measures may be worth exploring in areas where there are labour shortages. Smallholder farms could be incentivised to adopt technologies that could replace farm labour. 74 Market Development Facility

87 Pathway 7 Improved access to markets for different crops to different traders to motivate farmers to grow higher value crops Farmers were experimenting with higher value horticulture crops, had knowledge of crops they wished to grow, and/or wanted to set up nurseries. However, trader relationships thwarted these ambitions as the traders purchased from farmers in a manner that did not differentiate between the quality of produce. Improved and diversified market access would encourage and enable farmers to work on higher value crops, and yield the benefits of this (e.g. some are already growing organic produce, but it is not recognised and valued by traders). Roles Performed by Women in Horticulture The role of women in horticulture could be encouraged and improved in a number of ways. The study identified the following areas. Women in households Women are significantly involved in agriculture production within their households. Some women played prominent roles in cultivation,266 in terms of physical labour and farm decisionmaking. One woman noted that when she was younger she had been involved in negotiating pricing (she would talk to traders to get the prices),267 whereas others played more supportive roles, such as cleaning crops and harvesting without having a say or knowledge in production or in the business.268 There were no observable characteristics among the households that could e plai the diffe i g e te t of o e s pa ti ipatio. Ho e e, it as lea that o e e e ot idle and that they performed work duties throughout the day in addition to childcare, eldercare and household responsibilities. U like a othe ou t ies, the e appea ed to e o fe ale ops. Wo e e e i ol ed ith all crops but at various stages of production. Assigning contributions to agriculture by gender was problematic because both men and women were involved in crop production through a collaborative process. Women were notably more present in taking the lead in backyard farming. Women on farms preferred to work with family members to avoid becoming targets of gossip. One female farmer mentioned that she faced more challenges when she worked alone when her husband was imprisoned on suspicion of being involved with the LTTE. Other women in the community looked down on her and spread rumours, believing that a woman who worked was below a woman who did not.269 A female farmer in a separate household said she would only be in the field when her husband or his aunt were present thus never alone and never with another man present. She did not want to be seen as promiscuous, particularly as she was a transmigrant 266 SLMG-HH SLMG-HH SLMG-HH BILR-HH-12 Market Development Facility 75

88 from another province.270 Another respondent noted that the women of the house never went to the shops.271 Wo e s de isio aki g a d i fluence in household agricultural production Women throughout the sample exhibited a high level of control over household cash flows and household related decision-making. Women, who had no role in agricultural production, had little knowledge of or influence on it.272 Study data suggests that women who were involved in agricultural production had joint decision-making power with their husbands or they were completely uninvolved in the business decision-making. In one household, the wife was in charge of household expenses, but she had no involvement in business expenses, including deciding what inputs to purchase, what crops to grow, or how to interact with buyers or choosing traders.273 On the other hand, a widowed household farmer stated that, when his wife was alive, all decisions regarding the farm were jointly made.274 Women at work It was not uncommon for women, as well as men, to work as casual labourers on other farms. Other non-farm-related supplementary income generation by women included the daughter in one household contributing to the income as a tailor.275 One female respondent noted that women had to work to support their households, particularly to get past seasonal dependencies.276 Or as another female respondent put it: The ajo it a e da o ke s o help out o thei o fa s. If the do t o k, it s eithe e ause the do t eed to supple e t the household i o e o the a e ph si all i apa le of su h o k. 277 Employment opportunities in formal horticulture were limited, but were taken up by women. Wages seemed to be competitive with casual labour rates for women, and the working conditions were preferable. A female nursery worker said there was a lack of formal horticulture jobs in her area, other than at a nursery, and that her only option was to work as a casual farm labourer. She would have preferred working at the nursery as there was a support system of other female employees who could help one another with work, particularly when households faced illnesses 270 SLMG-HH SLMG-HH SLMG-HH-13, SLMG-HH BILR-HH BILR-HH SLMG-HH SLMG-HH-11 & SLMG-HH SLMG-HH Market Development Facility

89 and difficult circumstances.278 Another nursery worker said she preferred working in the nursery as it allowed her to determine how many days a month she could work.279 Barriers for Women in Horticulture As farms operated as small family enterprises, the constraints listed above affected women as much as they did men, thus measures to rectify these constraints should be aimed at men and women in farming households. Nevertheless, there were additional constraints that were more specifically elated to o e s i ol e e t i ho ti ultu e. Farms naturally have cash flush and cash low seasons, and did not always manage their financial planning, and investment decisions well. Women were the key financial actors in these farming households and, involved in agricultural purchasing. In these households there was few signs of considered long-term financial planning, and this limited their ability to reinvest. This was in part due to the limited financial situation of some of these households which, out of necessity, had to make short term decisions and buy items on credit when cash was low. Even in the poorest households, at times it appeared that hire purchase of large TVs or bikes was favoured over investments, such as, in a BRAC investment scheme that the respondent knew would be a good income source for the household.280 It was not always culturally acceptable for women to be involved in farm sales I o se ati e households, as ith fishi g o u ities, t ade as p edo i a tl a and it was not always culturally acceptable for women to be involved. a s pu ie This was an important limitation because women were the household money managers and responsible for making purchases, so their lack of access to buyers could negatively impact on the household cash flow. This exclusion from trade also limited o e s a ilit to t ade should the wish to do so. In some areas, women working on farms were stigmatised and looked down upon In conservative households it was regarded as inappropriate for women to undertake paid work. If a woman worked then she would be seen as poor and stigmatised for it, and it would have a negative impact on her reputation. This stigmatisation appeared to be worse for women who were working in isolation, especially when away or apart from their husbands. Wo e s o kloads i fa i g households e e ti e o su i g a d ph si all de a di g As is typical of for farming women worldwide, workloads were heavy, and women often spoke of the challenge of taking care of the household as well as their farm responsibilities. Women said they could not help on the farm as much as was needed because of their competing household duties. 278 SLMG-HH SLMG-HH SLMG-HH-21 Market Development Facility 77

90 This was often only mitigated when there was more than one woman or an older female child in the house who could step in to assist. Wo e s ag i ultu al la ou negative factor. ages e e poo l diffe e tiated, hi h as oth a positi e a d In day labour, wage differentials between men and women were fixed, despite the fact that men and women may do the same work. (There is a proportion of work that men did that required more manpower than women may be capable of, but this argument for lower pay does not hold true much of the time). For example, one farming household said a man was paid LKR 1,000 per day whereas a woman was paid LKR 900 per day. When asked whether a woman would earn the same as a man if she were equally productive, the respondent claimed that these were set market rates and could not be changed. 281 A female nursery worker explained the reason women dominated the nursery workforce was because men were not willing to take the offered salaries. The men who worked in the nursery were paid to do heavier physical work, including loading and unloading of plants. The Study Team considered that the competitive pricing edge women offered could enable them to obtain work, so such differentiation and its impact on women was complex. Recommendation to Address Barriers for Women in Horticulture Measures aimed at improving household financial literacy should ensure the information is provided to women (equally or perhaps even more so than men) Women were central to the financial management, cash flow planning, investment decisions and savings of these households. As good or bad financial management is central to whether farms were able to afford additional products and services that enabled them to expand and take risks, all too often the financial cash flow cycle of these households was overlooked as a major constraint. By working on measures to help farms better manage their income, particularly in cash flush months, this could unlock opportunities in other areas. The mechanism for this would need to be explored further, but could be encompassed as part of embedded information in a contract farming scheme e.g. or built into financial products (e.g. loans) that were already reaching these households, with the commercial incentive of decreasing default rates. Improved marketing of agricultural products and services to women, and the design of appropriate agricultural products for women Women were found in many places to play a key role in purchasing agricultural inputs, even if ultimately they would be used by both members of the household. It makes sense that marketing should be targeted in a manner that attracts women as well as men, and it should not be assumed that marketing tactics are gender neutral (there are differences between how men and women come to decisions on what they buy). Further, when developing products and services in general, MDF should be aware of the physical limitations women face. As discussed above, when providing services, MDF should factor in the mobility limitations women face in the North by not, for 281 SLMG-HH Market Development Facility

91 instance, simply designing a training session in a location that many women would be unlikely or unwilling to get to. Similarly, factors such as product size or weight can play a significant role in whether women are able to use it. Such practical factors are often overlooked and, with little or o feed a k loops i pla e to gauge o e s opi io s o the ualit of products and services, designs are unlikely to improve. Mi dful ess of o e s high le el of o trol o er household i o e, their role i pur hasi g decisions, and the implications for product and service design. Women in Sri Lanka have a high level of control over household cash flows and household related decision making. They also often play very active roles in family businesses. Thus when working on new product and service offers for these households, MDF and its partners need to keep in mind that the decisions to take up these offers is being undertaken by two household members, not one. Information on the products and services should reach both members of the household because not catering to women in marketing and advertising changed decision making dynamics, even if otherwise women would have been the one to make purchase decision (for instance a fa e s t ai i g o ga ised a pesti ide o pa i hi h o l e pa ti ipated ould lead to men exerting themselves more as decision makers). As a result of this, MDF does not wish to u de i e o e s a ilities to ai tai thei positio e o i g the o edu i g thei oles in this process. As women were often the money managers, product suppliers, in particular, should be mindful that it is ofte o e u i g ite s f o shops, ot thei ale ta get audie e e.g. i o e etail shop i the Stud, pe e t of the usto e s e e fe ale. It is often women at the front line determining whether a product is purchased or not. The implication is that more needs to be done to better understand product marketing in relation to women as well as men. In some cases, women need direct access to market information and training, rather than receiving second-hand information from husbands or others. Smallholder farming households are producing for increasingly competitive markets. New skills and knowledge are required to remain competitive and to enter higher-value (i.e. export markets), including information on national and international standards and requirements, bureaucratic procedures and consumer preferences. Such learning is expected to take place through supplier- u e i te a tio s, ut gi e o e s disti t ag i ultu al p odu tio fu tio s, they may often receive filtered or partial knowledge through their husbands or male relatives in the household. This is ofte due to o e s te de ies to ot pa ti ipate i e te sio se i es, or because they are not engaged in supplier-buyer relationships where knowledge may get transferred. Private sector partners can prove to be instrumental in ensuring that technical guidance is provided to both men and women. In other MDF country contexts it has been observed that, right from the outset of the training design, this must be considered (e.g. in TimorLeste, nursery management was not initially identified as being a key area for farms because the needs assessment was conducted only with the male household members and yet, when women were consulted, they identified this as an area they needed greater skills in. This had an effect on overall production and incomes as productivity/yields increased when women had better information Market Development Facility 79

92 For female-headed households, fa ilitatio of o e s effe ti e parti ipatio i arkets is necessary to ensure that improvements in productivity translate to profitable returns. Bringing women into markets requires targeted interventions, particularly for female-headed households. Simply increasing farm production does not automatically translate into increasing a woma fa e s i o e even if women and men access the same marketing channels, women can still derive less benefits. Such circumstances affect their ability to take risks, identify opportunities, and to invest when it comes to market channels. Emerging evidence shows that accessing market opportunities through existing groups (whether social, community, religious or financial) has been a means to make the process easier for women and allows them to enhance thei a gai i g po e. Bei g pa t of a a keti g g oup a edu e o e s u it osts, su h as in transportation.282 Introducing interventions may unwittingly compromise gender relations, and thus any approaches must be made in a sensitive manner. Initiatives in South Asia have included enabling women to market produce by introducing special market areas that are safe trading environments for women. Such areas can also be seen as non-threatening to men. The allotment of shops in holesale a kets o e e ship i a ket e do s asso iatio s ha e sig ifi a tly improved 283 o e s pa ti ipatio i I dia. Co side i g the stig a atta hed to e te i g ale t ade spa es in Sri Lanka, these approaches could be worth exploring in the Sri Lankan context. 282 Hill, R. V., and Vigneri, M., Mainstreaming gender sensitivity in cash crop market supply chains In Gender in Agriculture (pp ). Springer Netherlands. 283 World Bank. India: Taking Agriculture to the Markets. World Bank, Washington, DC; World Bank, FAO, IFAD Module 5: Gender and Agricultural Markets. In Gender in Agriculture Sourcebook Market Development Facility

93 Chapter : Fi di gs of the Po ert a d Ge der Stud Sti ulati g E trepre eurship a d Busi ess A ti it i For er Co fli t-affe ted Areas The rationale for working in the former conflict-affected areas is to facilitate the return of business activity and promote entrepreneurship, i o de to help e i e a economy that existed prior to the conflict, but was wiped out, and is now struggling to establish itself and grow (for instance, the case of stagnant SMEs in Jaffna).284 The major geographic focus of this engagement area is the northern and eastern provinces where the effects of the conflicts were most pronounced. MDF sees itself nurturing the growth of entrepreneurship, improving access to business mentoring and support services, promoting investment and product development, and facilitating better access to technology. The economic landscape within these provinces is largely agrarian but, following the end of conflict, there are signs that the private sector is setting up manufacturing and service enterprises. The Stud Tea s esea h i fo e o fli t-affected areas focused on the poverty situation of SMMEs and their employees. Interviews were also conducted with people in the fisheries, horticulture and tourism sectors. These interviews are also reported and analysed within the Chapters 3 and 4. Methodology and Sample Characteristics To understand poverty dynamics in former conflict-affected areas, household interviews were conducted with 22 households across the horticulture, fisheries, tourism and SME sectors. Geographically, the sample ranged from Jaffna, Mullaitivu and Mannar in the Northern Province to Trincomalee and Batticaloa in the Eastern Province. Within horticulture, interviews were conducted with smallholder fruit and vegetable farmers of different sizes, paddy farmers, a poultry farmer, and farm labourers. Within fisheries, interviews with a coastal fishing boat worker and a fish canning factory worker were conducted. Within tourism, interviews took place with a tourism activity provider and hotel worker. A range of small enterprises were captured under the SME sample, including four small enterprise owners (an agri-inputs shop, waste collector, food processor, and a sesame oil processor). Among the worker samples, four SME workers (from a plastics recycling factory, a wholesale shop, a micro food processor and a jam factory) were interviewed. It was difficult to locate medium-sized enterprises to interview, and there appeared to be an absence of these in the targeted regions. 284 Sinnathurai, Vijaykumar. (2013). Growth and Issues of Small and Medium Enterprises in Post Conflict Jaffna. Research Gate. Market Development Facility 81

94 Who is Poor in Former Conflict-affected Areas and Why? The characteristics of households in former conflict-affected areas are explored in detail below, including variances and trends within the poor and less poor households in the sample. Profile The poor were concentrated in former conflict-affected areas in the North and the East, home to ethnic and religious minorities. They included vulnerable groups such as female-headed singleincome households (who were forced to earn after the main breadwinner died or became incapacitated), war widows, orphans, people with a disability and unemployed youths who could not find a job in an economy that was decimated by the war. The poor were also concentrated in (unproductive) agriculture and fisheries sectors and/or were employed as unskilled labourers. People lack access to the skills that provide access to more rewarding employment. Poverty and per capita income Just over half the households interviewed were income poor or vulnerable and, of the remainder that were not poor, half fell close to the 200 percent poverty line. A slight majority of those interviewed in this area were found to be struggling to earn a decent income and operated within the poor or near poor category. The poorest households analysed in this sample included some of the poorest households captured in the overall Study. Other households with per capita incomes below the poverty line included two cases of female-headed households where women had been abandoned by their husbands285 and another where the family lost all its assets and savings in the war.286 Samurdhi support, and other monetary and non-monetary support, was commonly received by poor families in former conflict-affected areas in lieu of wages. Samurdhi support was provided as monthly cash payments, as yearly pay-offs against monthly savings, and as loans.287 In addition to Samurdhi, Poor Hindu families could receive up to LKR 1500 per month from their temple.288 Poverty pay from the government (outside of Samurdhi) was also listed as a source of income and this amounted to just LKR 300 per month.289 The government and NGOs also provided non-monetary support, including coconut plants,290 free electricity for short periods291and 285 BILR-HH-15 and BILR-HH10 respectively. 286 BILR-HH BILR-HH-17 gets LKR 650 per month. BILR-HH-07 paid LKR 200 each month and gets a lump sum once a year, which varies in amount but he got LKR 8000 this year. BILR-HH-20 is paying LKR 700 to Samurdi every month. BILR-HH-06 is getting per month in Samurdi support. 288 BILR-HH BILR-HH BILR-HH-02 Received 20 coconut plants. 291 BILR-HH Market Development Facility

95 goats.292 CARE provided LKR 175,000 for the establishment of a micro-foods factory, and training.293 Expenditure, debt and borrowings It was seen that the poorer households were those struggling the most with debt, and falling deeper into debt. Households with both higher and lower per capita incomes spent high proportions of their income on borrowings. However, although the poorer families tend to use loans for basic living costs to make ends meet, the relatively more well-off families tended to use it for income earning related borrowings.294 Again, proportionately, the biggest expenditure for smaller households appeared to be food purchasing, which unsurprisingly became a significantly higher proportion of income expenditure for poorer households. Poor families confirmed this in interviews, reporting that there were usually a few months a year when they struggled to meet basic costs such as food expenses and were unable to meet daily expenses.295 As a result they tended to manage expenses on a day-to-day basis.296 One family described how only recently they had been able to increase their food intake from two to three meals a day.297 Notably, nonessential items were bought even by the poorest households, such as televisions and DVD players.298 One household had no furniture, but still had a TV, DVD player and an electric fan.299 Poorer families mostly borrowed money from friends300 sometimes paying a small interest rate (known as ete atti ). One lender required the house deed be given to the lender to secure the loan.301 The agri-inputs shop interviewed said it was common for medium-size farm (especially paddy farmers) to buy inputs on credit and settle after the harvest302 and a paddy farmer said he borrowed LKR 25,000 from a trader to purchase seeds, as his crops had been damaged during flooding.303 In Moor communities, moneylenders charged no interest for religious reasons and the e e e o eal epe ussio s fo late pa e ts, apa t f o a s oldi g.304 One woman had borrowed from Samurdhi for business purposes (LKR 50,000),305 another paid off a loan for a 292 BILR-HH-12 From the Career Development Foundation 293 BILR-HH BILR-HH BILR-HH-17, BILR-HH-02, BILR-HH-15, BILR-HH-10, BILR-HH-05, i.e. the poorer households, all discuss this. 296 BILR-HH BILR-HH BILR-HH-02 The third poorest ranked household of the BILR sample bought a television and DVD player, as did BILR-HH-07, which is also earning just LKR 177 per capita income per day. 299 BILR-HH BILR-HH-17 LKR50,000 from friends. BILR-HH-07 From friends. 301 BILR-HH10 LKR 50 for a LKR 1000 loan. 302 BILR-HH BILR-HH-20. BILR-HH-12 The family borrowed LKR 100,000 from Samurdi three years ago and this had helped with expenses when he was detained and the family was struggling. 304 BILR-HH BILR-HH-15 Borrowed to purchase a sewing machine. Market Development Facility 83

96 hild s su ge take out th ough the seetu s ste a t aditio al s ste of o e s sa i g a d 306 credit). and a third woman was part of a set-up similar to the seetu s ste k o as u ited o eta i to hi h she deposited a a ou t e e o th, a d the asso iatio helped he ope 307 a savings account and obtain loans. As was common elsewhere, jewellery pawning was a means of obtaining a funds.308 Assets (including land) in business Most families, in particular the poorer households, held little or no assets, having lost many of these in the war, and incomes were not enabling them to save.309 Two of three families spoke of having any real savings, two of which fell in the top per capita income households,310 and the third was a young woman living with her aunt and uncle, who managed to save LKR 100,000 via a o e s oope ati e, despite he fa il s dail pe apita i o e of LKR.311 Two families did not have any savings as a result of recent business expansions.312 This was concerning given the inability of a big proportion of families to meet basic monthly expenses, and indicated just how vulnerable they were, even when they sat above or around the poverty lines. Both men and women had banks accounts.313with one woman prioritising savings for her children although she could not save for herself.314 As noted, people tended to invest in gold jewellery at the rare times that they had extra funds, which they then pawned as needed.315 Ethnicity The vast majority of households sampled in former conflict-affected areas were Tamil Hindus. The study sample was too small to make any clear connection between ethnicity and poverty. The least poor household in the Study was Sinhalese; two Moor households were among the poorest households, but no conclusions should be drawn from this. 306 BILR-HH-10 and BILR-HH BILR-HH BILR-HH BILR-HH-01 No apparent savings and buying food on credit. Fishing income for two trips of the month falls short and this had to be made up in further trips. BILR-HH-02 saves for the children. BILR-HH-03 Lost all savings in the war, now they have a fixed income, and plan to increase their coconut plantation. BILR-HH-04 no savings. BILR-HH-05 She managed to sa e LKR, ia a o e s asso iatio a d saves LKR 20 every Sunday as her back-up fund. BILR-HH-06 They cannot save there is no money to spare. BILR-HH-07 They had no savings and had to pawn their jewellery (they do own a TV and DVD). BILR-HH- The o all a t sa e. He a d his ife ha e a a k account set up. BILR-HH-09 Do have savings which they intend to use on a motorbike. BILR-HH-10 No savings and no bank account. BILRHH- Do t ake e ough o e to e a le to sa e, espe iall si e their business expansion. BILR-HH-12 Had jewellery but it was lost in the conflict, and now had little savings. BILR-HH-13 They had no savings, assets or equipment. BILR-HH-16 They hade enough money to make ends meet, but not to save. BILR-HH-17 Had high savings from the Middle East, but invested it all in the business, e.g. on goats, over 100 of which had since died. BILR-HH-18 Are able to save and, when they can, they invest this in jewellery. BILR-HH-19 Due to debts they had to pawn LKR 700,000 in jewellery, which they have been unable to redeem. 310 BILR-HH-18 and BILR-HH BILR-HH BILR-HH-17 and BILR-HH BILR-HH BILR-HH-02 Mother had deposited LKR 10,000 for each of their three children. 315 BILR-HH Market Development Facility

97 Poverty and other basic household characteristics The Study Team noted a higher number of people with a low level of education (i.e. years 7 and below) in former conflict-affected areas compared with the overall study sample. This was unsurprising given the high amount of displacement, with one woman noting that she had been unable to complete her O-levels as a result of the conflict BILR-HH-08. Market Development Facility 85

98 Reliance on the SME sector as an income stream Reliance on income from an SME as a proportion of overall household income ranged considerably, from 15 to 100 percent, with no real trend emerging i.e. more or less reliance on the sector here was not correlated with lower levels of income poverty. A notable income source that was helping to keep a fa il out of po e t as a siste s food etail job.317 Other additional income sources of the poorer households included a factory labourer, sesame oil production, and a security guard job.318 Income from SME employment Incomes of SME workers were close to or lower than the national minimum wage. Wage rates a oss households e e lo a d fell lose to o elo S i La ka s atio al i i u age of LKR 400. The only wage earner who did better than average in this small sample was the only one on a permanent contract, and even then his earning were just LKR 600 per day.319 Income from SME self-employment In relation to SME owners, the vast majority of work was undertaken by the respondents themselves, with the exception of a food microprocessor in Mullaitivu, which employed three staff in 0.25 FTE positions at a very low rate of LKR 200 per day. Key Constraints and Opportunities for Poor People in Former Conflict-affected Areas Households had to recover after losing economically active years and assets during the war The first major constraint was the loss of economically active years as a result of displacement in the war. This was compounded by the loss of economic assets (such as land, property and savings) and the ps holog of u e tai t aused o ti ued dis uptio a d t au a hi h affe ted people s decisions to re-invest and re-build). Figures as of July 2015 estimated there were 73,700 internally displaced people in Sri Lanka, mostly situated in the Northern, Eastern and North Western Provinces, most of whom are Tamil and Muslim, and 794,000 that had registered as having returned to their homes.320 Echoing the macro statistics, the Study research found people were only beginning to rebuild their lives having only recently returned, some from years in refugee camps and/or as refugees in India. During this time as refugees, most people reported that their only income sources had been labour/odd jobs, and that their livelihoods had been affected through loss of income (monetary and home consumption) from household farming activities321 and the loss of their 317 BILR-HH BILR-HH-07 and BILR-HH BILR-HH Internal Displacement Monitoring Centre, Sri Lanka IDPs Figures Analysis, BILR-HH-12 Noted farming activities were disrupted in the war years. 86 Market Development Facility

99 assets.322 One family who had been involved in ocean tourism had switched to fishing during the war.323 Only one (a Moor family) reported that one male head and another male relative had worked in the Middle East as labourers for six years as a result of limited opportunities during the war.324 This was so across religious lines, and with many people having not returned (one respondent noted that of 450 Moor families previously living in the area, only 31 had returned).325 Families that had lost homes and land were still living in temporary circumstances Many respondents had lost their homes in the war.326 Many were living in government, donor or NGO programme funded temporary housing, and some had been given land. One of the poorest households said this was one of their major constraints.327 Others were reliant on the goodwill of their relatives and328 others leased land. It was common for people to live in relati es ho es, however, according to one respondent this gave a misleading picture of stability as property owners could sell their property at any time and leave them homeless.329 In the worst cases, people reported they were squatting on land and living in fear that they may be asked to vacate at any moment.330 Beyond loss of land it was many properties were damaged by the war and were still under construction331 leaving many families living in temporary shelters while the houses were being 322 BILR-HH-01 Spent time in Indian refugee camps or displaced to other areas from until fairly recently during which they worked in a small labour works. BILR-HH-02 Was in India - she returned in 2004 but the rest of her family remain in India lived in IDP camp for 2 years. BILR-HH-03 Was in a refugee camp in Northern Sri Lanka. BILR-HH-04 Spent time up in the North before being resettled. BILR-HH-09 He and his family had to migrate to a camp inside vanni for seven months. BILR-HH-11 The family left to India during the conflict and returned in BILR-HH-12 The family had to migrate to India in 2004 due to the conflict. They lived in a refugee camp in Madurai until While they were there he did odd jobs in order to survive. After returning to Sri Lanka in 2010, he was arrested on suspicion of being involved with the LTTE and was released only in March BILR-HH-13 Family was displaced and during the war they fled to India as refugees in 2006 and returned after the war in BILR-HH-15 During the war time they displaced in multiple locations and they all resettled back in their own land back in last year. They resettled in 2015 from the sampoor refugee camp located in east of Sri Lanka. BILR-HH-17 They had to migrate temporarily when there was heavy conflict in the area. 323 BILR-HH BILR-HH-17 In Saudi and Qatar. 325 BILR-HH-17 The surrounding villages we re comprised of Tamils and were resettled post-conflict. Prior to the war, Faiz and his family lived in the same area and the Moor village was larger then. However, during the conflict, about 450 families migrated to other areas such as E a u, i ludi g Faiz s fa il. Of these, o l Moo fa ilies etu ed to the a ea post-conflict. They all came back voluntarily as this is their own land. BILR-HH-15 before the war they lived in their own house and farmed. They had to abandon their home and land when they were displaced. BILR-HH-03 Now land is in a military High Security zone with no public access along ith his ife s do la d cited as the major constraint in this HH. They lost motor bikes and their savings. BILR-HH-08 She lost her gold jewellery during the war BILR-HH BILR-HH BILR-HH-01 NERP built the house and they paid LKR 27,000 for the land. BILR-HH-02 They got the house from World Vision in 2004 along with 20 coconut plants but, due to the war, the house was only partially constructed. BILR-HH-11 They were given two acres of land by the LTTE on their return, but this was taken from them after the end of the conflict. They were given other land 80 perches under her name. BILR-HH-11 After being released from the army, he was given 15 perches of land and LKR 755,000 to build a house in June 2016 by the government. BILR-HH-03 Refle ted this fa il s situatio a d the oted it as o o a d that f o the outside the p ope ties looked alright, but this was often a temporary situation until the owner returned or sold the property. 330 BILR-HH-06 They did not have their own land, and they lived in fear that they may be asked to vacate at any moment. No-one in that village had received grants for housing. 331 BILR-HH-04 Respondent noted it was common for properties to have been damaged in the war and still be under reconstruction. Market Development Facility 87

100 rebuilt or repaired.332 Services in the area were poor, especially for tap and drinking water which had to be sourced and/or purchased from elsewhere.333 Road conditions were extremely poor around paddy farms in Batticaloa, with one farmer having spent his own money to connect roads to his paddy field.334 Trauma, disability and the loss of breadwinners are ongoing effects of the war There was deeper personal loss, with multiple respondents having lost relatives to the conflict (due to death, injury or detention), some of whom had been key breadwinners. One respondent had ee o pha ed a d spe t ti e i a hild e s ho e u til she as fou d a d etu ed to li e ith relatives. Another had lost her father in the conflict.335 Two respondents were disabled as a result of injuries suffered in the war.336 Those living by the coast had faced the double hit of losing livelihoods and family members to the tsunami in Return transitional issues were also noted, with one respondent saying it was hard for her to return because she had grown up in India, and it was hard for the community to accept her because of the way she spoke, but she had been able to reintegrate over time.338 O e espo de t su ed up the e pe ie e sa i g, War destroyed our whole life a d e a e sta ti g to uild agai our life changed drastically and we see a limited [bleak] future for our family. 339 Lost savings meant families had no cushion against the impact of shocks Loss of savings (mostly due to the war) and the inability to save left people highly vulnerable to shock, even where their incomes were above the poverty line. This was worrying as this was pushing families further into poverty. This situation made any new investments by these families impossible or highly risky. Industries had moved from areas affected by the conflict and not returned Due to displacement caused by the war, raw materials and sources once found in the region had now moved to other regions, which affected producers. This dynamic was found in the sesame oil industry.340 A worker at a sesame oil processor reported that previously sesame was grown in the 332BILR-HH-17 Most of the settled families werere living in temporary shelter in huts. BILR-HH-15 At present they are resettled on their own land in a temporary shelter. BILR-HH-17 They are living in two one-room huts on the land. BILR-HH-12 The family was living in a hut with no partitions/rooms. 333 BILR-HH-01 Got from neighbour although it was saline so they bought tap water from a vendor at LKR 2 per litre. BILR-HH-02 Water only from well consumption. BILR-HH-06 No access to electricity or running water. BILR-HH-11 Access to electricity, however no access to drinking water. BILR-HH-15 Resettled in their own land in a temporary shelter with no access to water and electricity and using firewood for cooking. BILR-HH0-10 They are using firewood for cooking and get water from a neighbour. BILR-HH-16 using firewood for cooking. BILR-HH-07 well water source. BILR-HH-13 They have no water in the house so access it from street taps. BILR-HH-08 They spend LKR 60 a month on water (i.e. the cost of accessing a well). 334 BILR-HH BILR-HH-05 Orphan who as an adult went to live with relatives. BILR-HH-02 lost father, lost family members due to the war. 336 BILR-HH-03 He was disabled in the war and receivied a small amount of disability support. BILR-HH-05 Young female disabled factory worker injured in the war from sight loss from artilleryand who could nott lift due to birth defect. 337 BILR-HH BILR-HH. It as ha d fo he as she had g o BILR-HH-. Quoted, a dest o ed ou BILR-HH Market Development Facility up i I dia a d as t a epted due to he a e t, ut o e ti e as a le to i tegrate. hole life a d e a e sta ti g to uild agai.

101 region and sourced locally, but the conflict had disrupted that. This is the case for other industries as well which were forced to suspend operations because of the war. For instance, the Northern province previously had glass and cement manufacturing industry, seafood processing and rice processing. These industries had not recovered after the return of peace. The lack of documentation (such as national identification or land title deeds lost during the war) made it difficult for people to access formal systems/the bureaucracy It is estimated that over 100,000 people lost key documentation during the war making it difficult for them to access public services).341 One respondent had not been able to obtain a bank account or obtain loans as her lost her National ID card in the war. This issue had recently been resolved.342 Those who returned with money after the war had no sources of advice on investments With many newly established income sources and businesses receiving little guidance and support, poor investment decisions were common. BILR-HH-17 was a clear example of this; returning to Sri Lanka from the Middle East in 2014, the respondent started a farming business. He had no prior experience of farming but bought 200 goats, 100 of which died in a single rainy season, resulting in the loss of a big proportion of his investment in a year. He noted that farmers in the former conflictaffected areas did not receive advice from the government or extension officers. Lack of access to loan funding and reluctance to borrow constrained expansion Poorer households cited lack of capital to expand as a constraint to their businesses,343 yet were unable to access formal loans or unwilling to do so due to unstable income sources and fears about the risk of defaulting. One family had applied for a loan with the Bank of Ceylon but had it rejected.344 Some families simply did not wish to rely on loans because of their inability to manage household expenses,345 with one woman explaining, It s eas to a ess a loa f o a i o-finance organisations in the region, but I fear to take it as I will be unable to repay. A othe spoke of the risk of high interest rates.346 Others noted formal finance was not available, so they relied on friends instead.347 Another said it was difficult to obtain a loan because of the documentation requirements, and the precondition that you usually needed to be in a permanent job.348 Another respondent o fi ed the ould o l appl ia thei siste e ause the la d as i thei siste s a e.349 Farmers seemed to be getting around this by accessing group loans and applying collectively (with the risk that if one defaulted, the others, unless they covered it, will not be able to access a loan for 341 BBC. Ju e S i La ka s Ta ils fa e ide tit 342 BILR-HH BILR-HH BILR-HH BILR-HH-03 and BILR-HH BILR-HH BILR-HH BILR-HH BILR-HH-12. isis. Market Development Facility 89

102 some time into the future).350 Given the high debt levels and high proportion of household expenditure, these families were not in a strong position to borrow for investment and expansion. Agri-inputs costs had become preclusively high and access was a problem Prices of some agricultural inputs, once affordable before the war, were prohibitively high in the region. For example, red onion seeds, once available prior to the war, were unaffordable so the farmers switched to papaya cultivation as the seeds were cheaper and more readily available.351 The high price of inputs was repeatedly cited as the reason farms were not expanding.352 Lack of knowledge of cultivation techniques, lack of investment capital, and lack of access to inputs were cited as reasons why paddy farmers were not going into more lucrative crops.353 The discontinuance of the government subsidy for fertilisers (previously LKR 350 until April 2016) was said to be particularly challenging for farmers, who now paid LKR 2,500 for the same fertilisers.354 One farmer had to travel 30 km for agri-inputs,355 and another noted these inputs were only available out of the main city.356 Farm labour shortages affected some areas A lack of farm labour was said to be a problem in Trincomalee, Jaffna and Dambulla.357 This contrasted with Batticaloa, where a farmer said he was unable to find jobs of this type for more than five days a month,358 which he obtained via contacts of family members.359 Lack of any/sufficient irrigation systems was a problem during Sirum Poham (the irrigated season) in Tricomalee and in Batticaloa360 Yields were said to be low in this season in Batticaloa.361 O e householde e tio ed, If I had a well, I could do some crops and take care of household e pe ses. 362 Another village in Batticaloa was getting around this with a village rainwater pond and canal system for which they paid a LKR 20 annual usage fee, but explained that this was still insufficient.363 One slightly better off 350 BILR-HH BILR-HH BILR-HH BILR-HH-20. The reasons why one paddy farmer had not diversified, and noted this was common. 354 BILR-HH BILR-HH BILR-HH BILR-HH-14, BILR-HH-04 and BILR-HH BILR-HH BILR-HH BILR-HH BILR-HH BILR-HH BILR-HH Market Development Facility

103 farmer was paying LKR 15,000 per month for associated water pump costs.364 Another farmer in Jaffna noted that only well-off farmers were able to afford irrigation.365 Excess rain and flooding were problems in the rain-fed season in Batticaloa and Jaffna366 Unpredictable weather changes were cited as a problem in Jaffna, with one farmer having lost his whole potato harvest because of it,367 and another lost 100 goats to the rains in Batticaloa.368 Market access for producers was limited by high transportation costs and little bargaining power with traders In horticulture, the long distances to buy and sell goods mean heavy transport costs, and poorer households complained that market prices were controlled by traders and they had little bargaining power of alternative options for selling their goods.369 A paddy-producing household said prices were controlled by the mills and farmers received a lower rate than was assigned by the government.370 Farm gate pick-up was preferred.371 Child e s tuitio fees e e a heavy burden on poorer families Two families explained they felt they would be in a better position once their children completed their education; until that time it would be difficult to make plans and save money.372 Perhaps as a reaction to their own lack of opportunities caused by their poor education, one respondent e plai ed, the hild e s edu atio is the ai fo us, e e o e tha food. 373 According to HIES data, the proportion of households now spending on private tuition rose from percent in the data to percent in the data. Reasons for this included stagnation/slight dropping of public expenditure on education, limited places for higher education (and thus high competition) and an overall increase in income (thus more ability to afford this).374 Education in Sri Lanka education is free, however families still incur costs associated with school attendance. Drinking alcohol came at high social and economic costs for families Aside from the social costs of drinking, additional financial costs were related to alcohol, in terms of the inability to work/lost work hours, and the high and regular costs of alcohol. One household 364 BILR-HH BILR-HH BILR-HH BILR-HH BILR-HH BILR-HH-17 and BILR-HH-03, the two poorest farming households in this sample, both complained of this. BILR-HH-04, a better off household, also noted the market was controlled by traders connected to the Dambulla market. 370 BILR-HH BILR-HH BILR-HH-02. BILR-HH-15 Also spent LKR 5,300 on education related costs and BILR-HH-07 spend LKR 10,000 per month on education costs. 373 BILR-HH-07. Pallegedara, A. Risi g De a d fo P i ate Tuto i g Despite S i La ka s F ee Edu atio. The Colo o Teleg aph Market Development Facility 91

104 noted that the male head was spending LKR on alcohol every day. In Mannar, empty alcohol bottles were frequently seen, and during one interview a drunken argument with a neighbour took place which the respondent reacted to as a normal event.375 Female workers in a fish processing factory said that many worked there out of necessity because their husbands were alcoholics, with one respondent estimating that 10 of the 75 families in her community included an alcoholic male head where altercations were the norm, and roughly 50 percent of the men in the community consumed alcohol to some degree. Another noted that half of the households in her village were women-headed and in the rest of the households, men drank alcohol. She oted, I a ot pla a futu e ith hi e ause of his d i ki g ha its, so I aki g o pla s. She explained that her husband only provided her with half of his income he spent the other half on alcohol.376 A 2014 report by the Institute of Policy Studies found there was a shift in alcohol consumption post-war, and that this unfortunately appeared to be worse in poorer households where the men were unemployed or engaged in crippling day labour work. There was a subsequent impact on women, as the alcohol invoked household conflict.377 Work done by poor families was low skilled, informal, sporadic and poorly paid Farmers supplemented their incomes by working as labourers or other manual workers but in some places work was scarce and poorly paid.378 This resulted in low and irregular incomes379 e.g. farm labour and masonry/odd jobs paid LKR 1000 per day. One farmer noted he was only able to find work of this kind for a maximum of five days per month.380 Women were usually paid about half of this sum.381 Farmers often resorted to this kind of work out of necessity when they were struggling in a particular week or month.382 Information about this work came via word of mouth from friends and neighbours.383 Children were also required to work to support the family in hard times and this meant sacrificing their education and setting up a vicious cycle whereby no/low qualifications and skills led to poor quality jobs and pay. The children of one of the poorest families were forced to 375 BILR-HH BILR-HH-01 Spending LKR on alcohol for the male HH head every day. BILR-HH-02 Respondent said women working in the fish processing factory did so out of necessity because husbands were alcoholics). BILR-HH-02 Respndent said in her community approximately 10 of the 75 families where the men are alcoholics and there are incidents and around 50% of the community where men consume alcohol to some lesser degree. BILR-HH-06 Many of the men in her neighbourhood are alcoholics. 377 Perera, A., Former War Zone Drinking its Troubles Away Institute of Policy Studies BILR-HH BILR-HH BILR-HH-20. BILR-HH-12 Also confirmed this seems to be the norm, as he gets LKR 1000 for labour work and LKR 100 for every overtime hour worked. 381 BILR-HH BILR-HH BILR-HH-02 and BILR-HH Market Development Facility

105 forgo their O-level education to go to work in unskilled positions to support the family out of necessity.384 Other children had migrated to Colombo for work385 and overseas.386 Micro and small businesses were ill-equipped to grow. The two least profitable businesses in this sample matched what seemed to be a trend in these regions; NGO-funded enterprises that, while admirable for the employment they offered women, generally sold poor quality items that were reliant on the same NGOs or Sri Lankan diaspora for their markets, with little chance of future growth.387 Few medium-sized businesses meant there were few jobs and poor pay With an absence of medium size enterprises in former conflict-affected areas, there were few jobs available in the sector, and the jobs that did exist offered poor pay and limited upward mobility. The Study Team found it difficult to identify medium size enterprises to interview which was consistent with the experiences of the earlier MDF scoping team in these regions. Wage rates across the SME jobs sampled were low, with th ee of fou falli g eithe lose to o elo S i La ka s atio al minimum wage of LKR 400. The only wage earner doing better than average in this small sample was also the only one on a permanent contract, and even then he was earning just LKR 600 per day (she took this job because her previous one was only paying a salary of LKR 3000 per month).388 Microenterprises were paid less, if they were able to hire at all with one paying its staff LKR 200 per day.389 One respondent, who was working as a plastics collector, noted that although he could work at factories, they offered low pay so he preferred to continue to work freelance.390 Another confirmed that there were few opportunities for promotion or growth.391 Pathways out of Poverty in Former Conflict-affected Areas The opportunities for pathways out of poverty within the former conflict-affected areas were considered to be as follows: Pathway 1 Foster entrepreneurial and SME growth particularly around product design, new markets and long-term investment strategies. 384 BILR-HH BILR-HH BILR-HH BILR-HH-07 and BILR-HH BILR-HH BILR-HH BILR-HH BILR-HH-08. Market Development Facility 93

106 For the men and women interviewed in the sample, SME growth seemed most constrained by poor product design and a reliance on immediate markets or, for women in particular, NGOdriven marketing strategies, poor investment strategies, leading to poor/non-existent growth plans. Access to formal loans seemed to be a problem, but without stability in these businesses, people were unwilling to borrow. Reinvestment in existing business was not just a matter of securing loans, but a matter of ensuring the businesses planning was robust enough for them to be able to sustain grow and meet financing costs in the long term. It would be better to look at this more broadly than just at loans. MDF needs to ensure businesses are able to better structure, plan and organise for growth so they can afford the cost of loans in the long term Pathway 2 Some SMEs could focus on fostering affordable, locally available products and services, to stop-gap the losses of local production that were destroyed or prevented from growing due to the war. Businesses that serve to reduce transaction costs in and around these regions, such as logistics, should be explored. Respondents in the Study spoke of products and services having been available before the war that were now unavailable or more expensive, or that production had moved to other regions. For example, one farmer no longer grew onions because the seeds were no longer affordable as they had been before the war. A sesame producer spoke of the lack of sesame production in the region, which had once been a feature of the area. SMEs in this region could be supported to play roles in revitalising and re-injecting money into the area, contributing to the local labour market, and of lowering affordability constraints surrounding products and services that prevent people from being able to grow other corresponding businesses. SMES could do this by focusing on local production or on lowering the cost of importing items by offering competitive alternatives or reducing supply chain costs (e.g. the logistics of how products reached these areas could be considered). MDF could work on some or a combination of these areas. Pathway 3 SMEs need to offer built-in embedded services to assist people to make suitable investment decisions. Farmers in particular spoke of having little or no access to extension services. SMEs can play a role in providing embedded services in a commercially sustainable manner that enable people to make informed decisions about the products and services they buy. Larger providers from other regions e.g. from the agricultural market from Dambulla, might be explored to see why there was a gap in this provision. Pathway 4 SME growth needs to be designed to foster local hiring and sufficient job opportunities for local workers. There was a dearth of jobs in these regions, and many of those available did not pay a living wage. This had significant economic knock-on impacts; men and women were unable to make purchases or to reinvest in the area until they could improve their financial situation. Employment related incomes would serve to help pull these men and women up from poverty and help generate the cash injection needed to supply small-scale local products and services that the area needed. 94 Market Development Facility

107 Waged employment is suitable for men and women with little or no assets and a high pre-existing debt balance, as the risks are lower than investing in assets for small enterprises. Often they are unable to access the financing for such options, and with little assets other than their land as collateral, it may well be wise for them to avoid doing so. MDF could explore support for new SMEs, or incentives for employers to take on staff, amongst other policy support. These incentives would need to be sustainable beyond initial period of support. Market Development Facility 95

108 Pathway 5 Affordable irrigation options need to be made available for the Sirum Poham season (irrigated season) in areas such as Trincomalee, Batticaloa and Jaffna. Farmers spoke of lower yields and an inability to farm due to no/low access to water in Sirum Poham, and noted this as a major restriction to farm income growth. Low cost irrigation not heavily reliant on fuel would need to be considered in conjunction with an environmental impact assessment, to ensure that farmers sourcing water from vital and slow to be renewed groundwater sources, that are needed by communities. Pathway 6 Mindfulness of the limiting factor of the absence/lack of farm labour in areas such as Trincomalee and Jaffna. Farmers and traders in some regions were limited by not being able to access enough farm labour. MDF should keep this in mind when looking at production-related measures that result in additional farm labour needs, as this may be a preclusive factor to production growth. Affordable alternative mechanised and other labour-reducing measures could be explored. Labour issues needs to be explored further in Batticaloa which has opposite problem of not being able to find sufficient work. Pathway 7 Crop and livestock insurance should be explored in Batticaloa and Jaffna, along with measures to increase agricultural resilience, particularly through diversification and crop cycling. Floods and heavy rains made crops and livestock a high risk investment in some regions. Crop insurance might be explored as one measure to reduce risk. Other measures might include looking at ways to increase agricultural resilience by reviewing crop planting and harvesting schedules throughout the year, diversification of crops or crop varieties grown, or looking into more resilient varieties (i.e. varieties that can survive flood waterlogging or heavy winds). Roles Performed by Women in Former Conflict-affected Areas The ha a te isti s of o e s oles in households in former conflict-affected areas are explored below. Wo e s de isio aki g a d i flue e Consistent with the overall Study sample, households in former conflict-affected areas indicated that women in Sri Lanka had a strong degree of financial control over household income. Only a few households reported the male head of the household solely managing finances. It was notable that, where men were managing the regular purchasing, this was most often because women were no longer present or unable to do so; in one case this was because his wife had died 96 Market Development Facility

109 (she had managed everything before),392 in another because his wife was too ill to do so,393 and because the household male head managed this for his mother who was without a husband.394 In one household in a Moor community, it was mentioned that men made all the decisions around the o e despite o ega dless of his ife s apa it to do this.395 Generally, it was a regular practice for men to hand over their salaries or earnings on a weekly basis, only holding back a small amount for their personal expenses.396 The trend of women being the financial decision makers held true even in respect to larger purchases. Women were often the initiators of the idea of what major item to buy, researching where to get it, and going out to buy it (or going with the assistance of another relative), giving them a high level of autonomy over the process.397 I had the idea to u a ele t i fa a d I mentioned it to him and I went and bought it and he had to pay for it, o e o e e plai ed, laughing.398 E e egotiatio of the hus a d s i o e ages o p i es ould e o a -led, with one woman noting that she was the one to negotiate with the boat keepers, and she believed that in roughly 30 percent of cases women would do this.399 Women also helped to make business de isio s, ith o e a agi g the fa il s etail shop a d all the a ou ts a d e pe ses fo the farm, as well as the farm labour.400 Women business owners tended to lead on decisions around their own business.401 When women earned their own income, they tended to prioritise expenditure on their children. One woman prioritised savings for her children even when she could not save for herself BILR-HH BILR-HH BILR-HH BILR-HH BILR-HH-01 The wife takes the lead and does the overall management. BILR-HH-03 He hands his sales money over to his wife and she controls expenses. BILR-HH-07 Major purchases decided by the children and their mother. BILR-HH-09 He gives all the money he earns to his wife, retaining only LKR for personal expenses his wife manages personal expenses, and most major purchase decisions are made together. BILR-HH-08 She hands over most of her income to her sister, keeping a little aside for personal expenses her sister manages everything and makes all the decisions in relation to the household. BILR-HH-12 In general, she manages the household expenses, especially as she had to manage without him for 4 years when he was in jail. BILR-HH-13 In terms of decision making over the income over the HH expenses, she hands over her entire salary to her mother. Mother budgets and makes the purchase decisions. BILRHH-15 She is the breadwinner of the family, she manages the expenses and makes the household decision on her own. 397 BILR-HH-03 His mother handles all of the expenses. BILR-HH-11 When it comes to household and big purchases, she makes them jointly with her husband. BILR-HH-12 When it comes to bigger purchases, the decisions are usually joint. BILR-HH-13 Big purchases are shared with the father and the family makes a joint decision. BILR-HH-09 He gives all the money he earns to his wife, retaining only LKR for personal expenses his wife manages personal expenses, and most major purchase decisions are made together BILR-HH-01. BILR-HH-01 - Boat keeper negotiation done by woman % a d olle tio of hus a d s sala 400 BILR-HH BILR-HH BILR-HH-02 Mother had deposited LKR 10,000 for each of their three children. fo the oats. Market Development Facility 97

110 Women at work Women in former conflict-affected areas were more commonly found working than in other regions in the sample, which is unsurprising given that there are estimated to be 40,000 female headed households in Northern Sri Lanka, many of whom exist in this form because of the conflict.403 Prior to the war, the concept of a female breadwinner was very rare, and the job market was said to have done little to catch up, and thus women in the North were found to rely on a variety of strategies to support themselves, from family networks and kinship structures, community practices of solidarity and resistance, and finding ways to normalise their extraordinary circumstances.404 Despite the enormous challenges they faced, a high level of vulnerability, they were repeatedly found to be empowered in their own lives.405 The households sampled reflected this secondary research, in that women were seen and said to be working in micro-businesses, in garments factories as day labourers, as domestic help, in retail shops and agriculture, and it was viewed as common for women in Sri Lanka to work.406 Women-headed micro businesses were also found to be fairly common, and many women owned their own small retail shops.407 Consistent with the secondary research, women were found, in general, to work out of necessity, citing issues such as difficulties in making ends meet, family problems, not being financially sound, a d that less edu ated o e o ked e ause of po e t.408 One woman, the daughter of a war widow, noted that when she was growing up, she never imagined she would have to work a day in her life.409 A othe o a s hus a d had dese ted the fa il, a d he othe had also ee widowed. It was observed that the majority of the households in that village were women headed as the men were either missing or had lost their lives in the war.410 Women who are adult dependents in households because their husband left were found to add to the burden of the 403 Raksha, Vasudevan (2013), Female Headed Households in Northern Sri Lanka. Graduate Institute Publications. IRIN, Bleak Futu e fo S i La ka s Fe ale Headed Households, sri-lanka%e2%80%99s-female-headed-households) , fou d at: Raksha Vasudevan (2013), Female Headed Households in Northern Sri Lanka. Graduate Institute Publications. 406 BILR-HH-05 This was common although it was said there is a lack of opportunities. Said the majority of unmarried women work in factories and the married women work as domestic help unmarried women prefer not to do this due to the risk of violence. BILR-HH-08 Most of the other women in her community do work, either in factories, as domestic help, in retail shops or in agriculture. BILR-HH-11 Both husband and wife are primary breadwinners. BILR-HH-11 She knows some women who work many who own a micro business or go to garment factories to operate as day workers. BILR-HH-12 Most of her friends and women in the community do wo k. the ajo it a e da o ke s o help out o thei o fa s. 407 BILR-HH-15 Retail shop ownership of women was said to be common. BILR-HH-21 His wife manages a retail shop. 408 BILR-HH-02 Woman working in fish meal factory due to difficulties in making ends meet. BILR-HH-05 The majority of women work because of family difficulties. BILR-HH-08 Most women in her community do work, these women all work out of necessity. BILR-HH-11 Ma o e do t o k e ause the a e fi a iall sou d. BILR-HH-13 The other women (uneducated) work because of poverty. 409 BILR-HH BILR-HH Market Development Facility

111 income earners.411 As in other regions studied, the preference was that women not work where they had sufficient means so as to not have to work.412 Once they do work, women talked of coming to enjoy the work because of the social support system it added and because of the sense of independence it brought along with the satisfaction of being able to support their children and families. One woman stated, I like ei g e plo ed, it gives me independence and a social life and I would advise other women not to borrow or take loa s ut to o k if the a t to get thei fa il out of po e t. 413 A othe said, Wo e should stand on their own feet and try to edu ate thei hild e. 414 Another working woman had been able to open bank accounts for herself and her children.415 Women often worked in groups or with their husbands or a family member as a means of social protection. Women said this was because it increased their security. In one case, a woman said there was a noticeable difference in incidences of harassment when her husband was around to when he was away and this was not only harassment from men, but in the form of judgement from, and rumours spread, by other women.416 There was a different perception of what jobs married and unmarried women could do e.g. one woman believed that unmarried women preferred not to work as domestic help due to the risk of violence from employers.417 Female support systems at work were found to be important. One woman worked in a group of seven women and said they met to talk about their work and their personal lives, and helped each other out where possible (e.g. if one woman was sick, they would share the workload).418 Relationships with the boss and other colleagues were also cited as important forms of support.419 There was a range of forms of community support that enabled women to work, from childminding, watching the business,420 to husbands helping to run the business BILR-HH year-old daughter now at home because husband left. BILR-HH-13 Most of the educated women work and hold good positions specially in the government sector. 412 BILR-HH-11 She believes that families in the area prefer it when women do not work.in general she believes that women face more constraints than men she k o s a ho do t o k as the a e fi a iall sou d. BILR-HH- If the do t o k, it s eithe e ause the do t eed to supple e t the household i o e o the a e ph si all i apa le of su h o k. BILR-HH- I like ei g e plo ed, it gi es e i depe de e a d a so ial life a d I ould ad ise othe loans but to work if the a t to get thei fa il out of po e t BILR-HH BILR-HH-02 Woman opened this for the children. o e ot to o o o take 416 BILR-HH13 She feels secure as the owner and his wife are always in the shop. (BILR-HH-01 fishing wives in Mannar on grading and sorting. BILR-HH-12 one interesting point she mentioned is that when she was working alone (when her husband away) she faced a lot more issues. The other women in the community spoke badly of her, were judgmental and spread rumours. 417 BILR-HH-05 said unmarried women prefer not to work as domestic help due to the risk of violence. 418 BILR-HH BILR-HHhome. She sa s the o e is good a d the p o ide eakfast e e da. A d she gets o e-hour lunch break so she is able to go 420 BILR-HH-01 Adult daughter maintains the household work. BILR-HH-11 husband helps cultivate the centella and babysits when she has full orders. BILR-HH-06 6-month-old daughter and other children are looked after by her neighbours and her mother. 421 BILR-HH-11. Market Development Facility 99

112 Family reactions to women working were mixed, with most families coming to accept it in times of necessity. Manual labour work was not looked upon well by the family, but where it was done out of necessity families did understand and accommodated it.422 One woman said that when she first decided to take up a job in a food production cottage industry, her mother refused to give her permission, but she convinced her to allow her to do so to support the family. The mother agreed after speaking to the owner of the business.423community reactions were mixed; this included discrimination and judgement from other women, even where the woman was married. A woman explained that, even though her husband had returned, she still faced problems, and the o e ho did t o k looked down upon those who do.424 Government jobs were always regarded as the ones people wanted to go into and were seen as the jobs that educated women took as they had security (a pension) and they were seen as respectable.425 Teaching was regarded as suitable for women because it was said to allow women more family time so as to balance their workloads.426 This balance seemed to be preferred over other seemingly espe ta le jo s, ith o e o a oti g she a ted he daughte to e a tea he ut he daughter wanted to e a a ke. 427 Some women chose, or were encouraged to stay at home due to perceived and real safety concerns. Working outside the home was seen to have risks, with one woman stating that she had been subject to abuse for working, but she was not worried about it. Another woman said that safety and security were big considerations when women chose their jobs.428 Women as entrepreneurs In the small sample, women were found to be highly active as wage workers and sole traders. Women were active in decision-making and management roles in small businesses. In the agriinputs business, the female household head helped to manage the shop. She was highly knowledgeable about the agri-inputs sold by the shop, from fertilisers, seeds, to pesticides, and provided advice to customers about the products and their application. She was aware of shop finances and, during the interview, she frequently interrupted and corrected the male head of household on questions around the running of the business. Beyond her support of the shop, she 422 BILR-HH-06 and BILR-HH-13 The family does not oppose her working due to the difficulty the household faces. 423 BILR-HH BILR-HH BILR-HH-04 They are seen as security. BILR-HH-09 He would like his daughter to secure a government job and so she is remaining at home until she can find one. He has similar aspirations for his son. 426 BILR-HH-19 Teaching allows women to have more family time and to balance workload. She prefer her daughter to be a teacher but the daughter wants to be a banker. 427 BILR-HH-19. BILR-HH-08 Most women in the community work - fo the o e ho do t o k, it is eithe e ause thei o kload is too u h o thei husband/family has concerns about their safety. BILR-HH-11 Her business involves her going out alone and she has been subject to abuse due to this, ut she is t too o ied a out it. BILR-HH-13 Safety and security is a big concern when they choose their jobs. BILR-HH-10 In this region, due to the security and risk, most households refuse to send their women to work, it was said Market Development Facility

113 also supported the family by doing small tailoring work for which she received LKR per month.429 The micro-food processing business was run by a woman with the support of her husband. She had received support from CARE to establish the business430 which she started out of necessity on returning from India in 2005 and when, at one point, the household had gone without food for five days straight. She said she knew then that she had to start something, and cooking offered the best opportunity at that time. She was aware of prices and market information due to her close association with local bodies and organisations. This role was at times intensive, as is the case with small businesses, and there were up to ten days a month when there were big orders that meant she only got a couple of hours of sleep at night. She said she never thought she would be where she is now. In the remaining two households, the women were not involved in the business and stayed at home.431 Women were found working as a cashiers/packing assistants, recycling plant workers and food production factory workers for which they had received short periods of basic training and were being paid close to and below the minimum wage.432 Barriers for Women in Former Conflict-affected Areas Harassment in and on the way to the workplace was a concern for these women, and often prevented them from working at all. As discussed above, women did get harassed in a variety of way, and this was not limited to male harassment and safety risks, but also to the negative reputation-tarnishing harassment of other women. Unmarried women were said to be unable to take certain roles (such has housekeeping) because of the risk of being harassed. Linked to safety, lack of proximity to the workplace and a lack of transport were commonly cited constraints for women. P o i it to ho e as a fa to i o e s hoi e of jo s. O e o a said she liked the a ilit to go home during her lunch break and that extensive travelling for work was not preferred by women. Work location and transport were restricting factors for women; in particular few women travelled outside of Jaffna for work.433 Another said that women in the area were attracted to working in a 429 BILR-HH BILR-HH BILR-HH BILR-HH-10 Five days training on labelling and packing. 433 Women also work in roles because the location is conveniently located near their house. BILR-HH-01 fishing sorting because it was just behind her house. BILR-HH-01 Respondent said women were commonly found in this sorting role. BILR-HH-02 Respondent stated women prefer to work closer to home and that extensive travelling is not preferred by women or their families. Women will refuse good jobs and significant training opportunities because workplaces are too far. BILR-HH-02 Government factory 25km away with female workers but women in that specific village were not willing to travel that far due to their household chores. BILR-HH-05 Distance to work and transportation were listed as restrictions to women working. BILR-HH-05 Transportation and distance matter to unmarried women in their Market Development Facility 101

114 recently opened factory in part because it provided transport.434 Mobility issues seemed stronger in Muslim households, where women were not seen to leave their houses during the course of the interviews.435 Proximity was also linked to the ability to manage home-based tasks. Social stigma around working was negative Families preferred women not to work although they accepted it in times of dire necessity. There were firm perceptions that women could not work outside the home because of young children436 with women often working when her children were older.437 Few opportunities Consistent with the secondary research, there were said to be few employment opportunities for women.438 Lack of skills was cited as a constraint women faced more than men, and transferability of skills seemed to be a factor in their success.439 Women in this sample tended to be paid less than men for similar roles.440 As with the secondary research441 women did not seem to have problems setting up businesses, but these tended to be NGO-driven and funded, producing poor quality products and limited knowledge of, or chances to grow or enter, new markets.442 Alcoholism, was a regular and major financial drain The impact of alcoholism, which was a regular and major financial drain, at worst disabled men from regular work, forced women into employment, and increased the risk of home-based violence. Alcoholism was witnessed more openly in the former conflict-affected regions, with women adjusting their lives around it in an attempt to bring stability. Women led households more so when the husband was an invalid or an alcoholic, where they tended to make all decisions themselves. choice of work the majority of women avoid travelling out of Jaffna to avoid risk. Very few travel to Colombo. BILR-HH-13 Women prefer to work in close proximity to where they live. BILR-HH-13 She walks to work and lives 5-10 mins away from the shop. 434 BILR-HH BILR-HH-17. BILR-HH-03. BILR-HH- His ife does t o k si e she has to look afte the family and the household. From 6am-6pm his wife has to remain home because they cannot leave their16-year-old daughter alone BILR-HH BILR-HH-05 Respondent noted this was common although it was said there is a lack of opportunities. 439 BILR-HH-05 Women still face more constraints than men. Lack of skills and qualifications is seen as a constraint. BILR-HH-11 Her transferrable skills helped her succeed in diversifying her business. 440 BILR-HH-02 Respondent noted jobs women are involved in do not require skills and qualifications and are the lowest paid roles. BILRHH- LKR pe da o pa ed ith e s LKR pe da. BILR-HH-17 LKR1000 for women compared to LKR1500 for men in another and they are paid in cash. BILR-HH-19 Per day men are paid LKR 1200 and women are paid LKR Attygalle, A. and Hirimuthugodage, D. (2014). Female Entrepreneurship and the Role of Business Development Services in Promoting Small and Medium Women Entrepreneurs in Sri Lanka. Institute of Policy Studies of Sri Lanka and Oxfam International, Sri Lanka. 442 BILR-HH-07 and BILR-HH Market Development Facility

115 O e o a said the e is o poi t i dis ussi g it ith hi save and manage everything.443 a d that the e ti e u de fell o he to 443BILR-HH-06 major purchases are up to her vision she says there is no point in discussing it with her alcoholic husband the entire burden is on her she has to save up and manage everything on her own. She makes purchase decisions solely. Market Development Facility 103

116 Recommendations to Address Barriers for Women in Former Conflict-affected Areas More employment opportunities for women in SMEs in conjunction with fostering a labour market that is adapted to the high female workforce in the North. As noted, there were 40,000 female-headed households in the North, but the labour market had not evolved to provide for this labour pool in terms of creating new jobs for women that paid them a decent salary. The labour market fostering is in part associated with looking at whether SMEs at all factor in hiring policies conducive to women, or at least have policies that do not inadvertently prejudice women. The development of safe and affordable transportation options both in and around places of employment. Women need safe and affordable options for transportation in the North to enable them to get to and from workplaces and undertake other activities in their daily lives. This could either be done by working with dedicated transportation providers. Relating to employment, MDF could look into working with companies to support in-house transportation options (either as an element of or constituting the whole partnership design), or to help these companies obtain suitable outsourced transportation options. MDF should ensure companies seeking to hire women in the North were aware of the key importance of safety in transit to and from the workplace. Dedicated female workspaces with a critical mass of female employees. To reduce the social stigma and safety concerns of women working in mixed floor spaces, MDF should work with its partners to create dedicated women-only workspace options. This would not improve their physical safety or reputation and foster an environment for learning but encourage other women to join. There was evidence from the Study that dedicated workspaces enabled women to assist each in times of emergencies such as childminding, and provided them with a greater social support system. Proximate workplaces (including home-based workspaces) with flexible hours and/or day care options. Employers need to be mindful of the dual work roles women play, especially those with small children. Women interviewed wanted to work closer to home so that they could check in on their hild e. Thus o kspa es that a e lose to o e s ho es, o ased out of o e s ho es/a community space, should be considered. This would allow for work to be done concurrently, such as weaving while watching children. MDF should make its business partners aware of the importance of the proximity factor if they wish to hire a female workforce. Supporting female entrepreneurs with the marketing and product development aspects of their businesses. Women appeared to be able to obtain financing to set up small businesses, but these businesses tended to stay at the micro-business level, with poor guidance available as to product quality and 104 Market Development Facility

117 markets. More needs to be done to ensure small female-led businesses with growth potential do not stagnate and are able to grow, to support themselves and to enable them to hire staff to support the labour market. A deeper analysis of the enabling environment for female-led businesses was beyond the scope of this Study, but further work should be done to better understand the position and limitations of female-led enterprises in the North. Mindfulness of the physical limitations of women when designing products, services, and jobs. When developing products and services, MDF should be aware of the physical limitations faced by women. It should factor in the mobility limitations women face in the North e.g. by designing a training session in a location that women would be most likely to attend. Factors such as product size or weight can play a significant role in whether women are able to use it (as exemplified in the plastics recycling centre cases). Such practical factors are often overlooked in designs and few or no feedback loops put i pla e to gauge o e s opi io s o the ualit of p odu ts a d services Social stigma needs to be factored into design in highly pragmatic ways. MDF needs to remain aware that the social stigma of working in certain environments, roles, and more broadly of working at all, plays a strong role in where (and whether) women work. There are many cases where women need to work, and signs that they enjoy this given the right set of circumstances. The Study showed that stigma could be reduced by employing a combination of p a ti al easu es i ludi g i easi g o e s safet a d p o idi g dedi ated o kspa es measures. While stigma will continue to run deep, MDF should not shy away from working towards practical measures to reduce it to support women to begin work. Social changes can come from changing practices; with more women able to join the workforce, this may contribute to a gradual change in their co u ities thi ki g a ou d the app op iate ess a d alue of women at work. Ensuring low visibility around conflict-inducing factors (e.g. money and assets). Situations where women suddenly began to make significant sums of money, through sales or wages, have been triggers for domestic conflict if there was alcoholism in the home. Alcoholism was prevalent amongst males in the sample in the North. Reducing the visibility of money and assets has been found, in other MDF countries, to be a key factor in reducing the risk of women losing her wealth. Workplaces can assist women to set up a bank account through which they can pay her rather than paying women in cash. This ill li it thei hus a d s a ess to the o e, reduce the risk of arguments at home and the risk of being robbed on their way home from work. Women with low social protection should be provided with safe-guarding measures before they are given cash or valuable assets. Mi dful ess of o e s high-level control over household cash flows and their role decision making around purchases and the implications of this for product and service design. Women in Sri Lanka have a high level of control over household cash flows and household-related decision making. They also often play very active roles in family businesses. Thus when working on new product and service offers for these households, MDF and its partners need to keep in mind that the decisions to take up these offers are being undertaken by two household members, not Market Development Facility 105

118 one. As women are often the money managers, product providers should be mindful that it is often women buying items from shops, ot thei ale ta get a ket e.g. i o e etail shop i the Study 60 percent of the customers were female). Women are often at the front line deciding whether the product was purchased or not. The implication was that more needed to be done to better understand product marketing in relation to women as well as men. 106 Market Development Facility

119 Chapter : Fi di gs of the Po ert a d Ge der Stud I o atio i Digital Ser i es Information and communication technology (ICT) plays a powerful role in economic and social development by revolutionising the ways in which government, the private sector and society interact. The Sri Lankan government has directed its national policies toward a knowledge economy, pushing initiatives to ensure the sector is dynamic and globally competiti e. As su h, S i La ka s fi st mover advantage in opening and transforming its economy has placed it in the fortuitous position of ei g a le to take a holisti a d o p ehe si e app oa h to ICT de elop e t. 444 Recovery of the global economy and improvements in domestic conditions post-2009 have contributed to an increase in the size of the industry, as reflected by the increase in workforce. Figure 6: Workforce in ICT Number of workforce Year Source: Information & Communication Technology Agency of Sri Lanka (2013). National ICT Workforce Survey That being said, ICT services are not universally available throughout the country, and the economic and geographical divide on access is stark. Information literacy is estimated to be 36.9 percent in urban areas, 22.0 percent in rural areas, and 8.6 percent in the estates sector.445 The ICT sector is comprised of a highly edu ated o kfo e, ith pe e t ha i g deg ees. Ba helo s deg ee a e the standard entry level qualification for recruitment in many job categories.446 The gender divide in ICT is sobering and is reflected in multiple indicators. In tertiary education, despite the fact that women comprise 57.8 percent of students, they only account for 34.2 percent 444 World Bank (2015) Sri Lanka: Innovation, ICT and Competitiveness Asian Development Bank, Deutsche and Gesellschaft für Internationale Zusammenarbeit (GIZ) GmbH, Country gender assessment, Sri Lanka: An update. (Mandaluyong City, Philippines: Asian Development Bank, 2015) 446 Sri Lanka Association of Software and Service Companies, SLASSCOM (2014). Sri Lankan IT/BPM Industry Review. Market Development Facility 107

120 of those enrolled in computer and ICT courses.447 Female participation in the workforce is at 29.7 percent as of While women are making inroads into working in ICT, they tend to be disproportionately concentrated in lower skilled jobs in business process outsourcing (BPO) companies in roles related to word processing or data entry. 448 Despite this, a e dotal e ide e suggests that digital at hi g fi s e.g. peer-to-peer platforms like Uber, Airbnb, and PickMe around the world can provide positive economic benefits for those who are typically underrepresented in the sector, particularly for those who need additional income and/or cannot work traditional hou s: lo a ie s of e t a d the utilizatio of u i uitous, common capital assets, such as cars, bicycles, and extra bedrooms allow individuals to work during off hou s o hile the e othe ise u e plo ed. 449 Given this context, the Study conducted three household interviews with respondents who worked in the ICT sector as managers or employees in companies in Kandy. To understand the impact of digital matching firms, the Study also conducted a FGD with female three-wheeler drivers who used the taxi-hailing mobile application PickMe in Colombo City. As noted in the outset of this report, it was difficult to gain access to workers in this industry due to staff being required to sign nondisclosure agreements. Who is Poor in Digital Services? Household poverty of ICT workers directly employed in the ICT sector is rare. Given the composition of the ICT sector and the workforce, the Study was unable to locate households that lived under the poverty line. Nonetheless, for learning purposes, interviews were conducted with three IT employees to get an illustrative idea of thei households e o o i i u sta es. One of the households interviewed comprised of a husband and wife who were both working in IT. The husband also owned a separate tour company and an marketing firm.450 Another household interview was conducted with an operations manager at an IT company, who alone had a monthly salary of LKR 80,000, along with medical benefits and holiday leave. When asked what the lowest salary was at his company, he said that entry-level graphic designers were paid a monthly salary of LKR 15,000, but that figure could easily rise to LKR40,000 with two years of experience.451 A third household interview was conducted with a low-level employee who earned LKR 11,900 a year. While the salary was low compared to that of others in the ICT sector, it was nonetheless at 300 percent above the national poverty line. Further, the respondent had the privilege of having a 447 Asian Development Bank (2015). Ibid. 448 Asian Development Bank (2015). Ibid. Telles, Jr., R. (20. Digital Mat hi g Fi s: a Ne Defi itio i the Sha i g E o o. U ited States Depa t e t of Co e e, Economics and Statistics Administration Issue Brief # DC-HH-02, Kandy, Central Province. 451 DC-HH-03, Kandy, Central Province. 108 Market Development Facility

121 wealthy family that could provide him with a safety net. As such, he was able to save his earnings toward pursuing tertiary qualifications.452 To look only at direct employment would take a very narrow view of the scope of the ICT industries and underestimate its potential to reduce poverty. There were two categories of beneficiaries in this regard: those that were engaged indirectly as a result of increased digital capabilities, and those engaged indirectly. Examples of the latter include: a) drivers and other workers who were mobilised as a result of information exchange sharing apps, such as PickMe drivers; b) staff recruited by employers, such as fast food operators for food delivery; and c) user/customers of these services who benefited from increased information exchange and an enhanced ability to access products and services. This resulted in higher incomes or cost savings for these households (e.g. by reducing information search costs throughout the value chain). These categories have the potential to reach and impact large numbers of men and women. While a detailed view of the media habits of the poor would require a separate study, the Study Team did enquire as to whether the respondents of this study had access to a mobile phone, and the results were far higher than national statistics on mobile phone use would suggest. In this sample, 59 of 63 households had access to mobile phones, and of these just five stated that women did not have access to that phone. In other words, both poor and less poor households alike had access to mobile phone services. This contrasts with the data which suggests that 50 percent of the Sri Lankan population have access to a mobile phone, which is low in the region.453 When asked about access to information, particularly around agricultural information, no household listed mobile-based services as an information channel. This looks to be a significant market opportunity, as a look at the secondary research in this sphere would suggest that ICT for agriculture, despite being a government priority since 2003, remains fairly limited.454 Pathways out of Poverty in Digital Services The pathways out of pove t offe ed Pathway 1 I p o i g S i La ka s Digital Capa ilities a e as follo s: Development of ICT services for rural markets, particularly around ICT services for agriculture As noted, it appeared that the mobile penetration gap for Sri Lanka in relation to its South Asian neighbours was narrowing, and use of ICT for agriculture was still quite thin. MDF can draw from lessons of successful neighbouring countries such as India and Bangladesh regarding ICT measures for agriculture that would, for instance, disseminate information to farmers. This would first require a look at the ICT industry to see what the picture of rural penetration currently looked like 452 DC-HH-01, Kandy, Central Province. Business Wire (28 September 2016). A Country & Industry Analysis Sri Lanka Telecommunications Malsha, APS Jayasinghe, APR Wijeratne, M (2011) Effects of ICT on Agricultural Production: Sri Lankan case study. Joint National Conference on Information Technology in Agriculture Market Development Facility 109

122 and whether there were incentives for the ICT industry to target this sector, as upkeep of the content of such services is crucial. Pathway 2 Collaborate with digital matching firms to match suitable workers to the right employers ICT services can act as matchmaking services between companies and their employees, and between companies and customers. Services e.g. to help locate and hire tourism or other hospitality workers, might be an effective way of matching skills demands with supply. This could be particularly effective for women workers given the hours are flexible. (This is discussed below under WEE Pathways). Roles Performed by Women in Digital Services Unfortunately, the study was unable to collect data on female employees at ICT companies. However, a FGD was conducted with four female drivers who had begun providing services via PickMe over the previous two to six months. The work allowed them to work to their own schedule, so they could to deal with household responsibilities and childcare without compromising their income-generating activities. One participant stated that while there was an expectation that men would apply to be drivers, there was no discrimination against her applying for and working with the company. She liked the training program on proper driving and customer service, and she felt that this gave her a sense of respect and pride to work for the company. Another respondent stated that male three-wheeler drivers did not give her any trouble, as they understood that she was a PickMe driver and thus was not competing for customers off the street. The largest constraint for women was identified as safety or the perception of safety. One participant noted that people were always staring at her and pointing out that a woman was driving a three- heele. O the othe ha d, these o e s e pe ie es poi ted to the fa t that usto e s themselves felt more satisfied with using women drivers. They received fewer complaints than their male counterparts, as they were perceived to be more disciplined. One respondent stated that female customers, in particular, tipped more and praised her for working to create a safe environment for customers, particularly late in the night. Recommendations to Address Barriers for Women in Digital Services Collaborating with digital matching firms to incentivise women to join the sharing economy that allows them job security and flexibility. The increase in economic opportunities through applications like PickMe did not necessarily exclude women, but women tended to self-select themselves out, mainly due to the perception that it as a a s jo. But fi di gs i di ated that fe ale d i e s ould be valuable for companies, as they would appeal to a largely untapped female market. Female passengers, who also shared safety and harassment concerns, seemed to welcome female drivers, particularly in the evenings. Furthermore, families were likely to feel more comfortable putting their children 110 Market Development Facility

123 into vehicles operated by women. Indeed, concerns about safety and harassment were important, but as long as appropriate training and protocols were set in place, many more women may feel compelled to join the sharing economy. Such schemes were beginning to emerge in other South Asia countries such as Careem (which holds a bigger market share than Uber) in Pakistan and which is now hiring female taxi drivers in areas with notorious mobility problems for women, such as Karachi Bukhari, M. and Sattar, W., Cab-haling Company Careem Launches women Drivers in Conservative Pakistan, Reuters (December 2016). Market Development Facility 111

124 Chapter : Su ar of Fi di gs a d Co lusio s of the Po ert a d Ge der Stud Snapshot of the Full Sample This section provides a snapshot of some of the key household characteristics of the 63 household sample, looking at the poverty level in general terms after looking at the poverty in relation to the respective SEAs. Poverty and per capita income In tourism sector, poverty was found to be less prevalent, with even the most basic jobs (hotel cleaners, gardeners and bellboys, of both large and small establishments) earning above the poverty line. In these cases, the poor tended to be those on the fringes of the sector. In tourism and other sectors, the poorest households tended to be those trying to eke out livelihoods on the fringes of the sectors. The poorest households relied heavily on agriculture incomes in the poorer regions which had poor climatic conditions (often with one major dry season). This included five households in the former conflict-affected regions and Uva provinces, where irrigation was a major problem, and where households had few economic alternatives. Households relied on debt to finance household expenses across all income levels. High levels of debt are indicative of vulnerability. Many households reported struggling to meet basic costs of living. Even where families were making relatively better incomes, expenditure choices were not always well considered and seldom geared around reinvestment into income earning opportunities. Throughout the Study it was found that the poorest households were using hire purchase schemes to buy items such as three wheelers and TVs, often at the expense of making investments into income earning activities which would have enabled them to increase their incomes. In many cases hire purchase placed an unnecessary additional burden on these households. Other social commitments, such contributions to wedding costs, were placing a high financial burden on those working in the fisheries sector. In many household, the male breadwinners had abandoned the household, passed away, were incapacitated due to illness or alcohol or working in poorly paid or unreliable work (e.g. as casual day labourers with little work offered in a month). In these cases, the breadwinners were women, who often support elderly and infirm relatives in addition to their immediate families. Major economic shocks were also a feature for the poorest households, often coupled with unstable income streams. These economic shocks ranged from the loss of key income earners to the loss of assets from the war456 or the loss of a family business.457 I o e household a ido s 456 BILR-HH-17, BILR-HH T-HH Market Development Facility

125 income had dropped from LKR 70,000 to LKR 5,000 per month after the death of her husband.458 Other family-related shocks put families under pressure, such as the numerous reported cases of abandoned sisters and their children, and frail relatives to take care of, none of whom were o ki g a d all ho o side a l affe ted the household s o e all pe apita i o e.459 Some households received supplementary monetary support from non-income sources such as Samurdhi support. For the poorest households this was a significant component of the household income (representing around 27 percent of total income). This support appeared to be reaching the right households, with all those that had received support falling below the 100 percent national poverty line or the 200 percent national poverty line. Beyond Samurdhi support, people also received support from relatives (most frequently), NGOs (seen frequently in the former conflict-affected areas sample), and from religious organisations (e.g. from temples). This support was both monetary (monthly payments, as is the case for Samurdhi) and non-monetary (land, livestock, and assets). Expenditure, debt and borrowing For the majority of households, expenditures outweighed incomes. This debt imbalance was over 15 percent greater than the household income for most households, and for some this expenditure was more than 50 percent greater than income. The households suffering the most from this debt balance were poorer households whose financial situation appeared to be worsening. These households spoke of the struggle to cover their basic expenses and of the practice of regularly buying groceries on credit. For them, loan sources were generally limited to family members, credit from a local store or, in only a handful of cases, microfinance or Samurdhi loan support. Some spoke of not contemplating borrowing for anything other than necessities, as they would have no idea how to pay the money back.460 For these households, food expenses equalled or surpassed the income gained (i.e. for these households, the most basic needs were leading them to debt). One woman explained that her power had been off for three months, as she could not afford her electricity bills.461 The outlook for these households looked dim, even without additional shock factors. Assets (including land) in business For the households in former conflict-affected areas in particular, loss of all assets (land, savings, jewellery, and so on) was a major factor underpinning poverty levels. These households had lost almost everything during the war, compounded by the loss of economically active years spent as refugees. Bank accounts were held by most of the Study sample, and these were frequently held and managed by women. Given the vulnerability of a lot of the households, however, few funds were 458 T-HH For example, SLMG-HH-09 and SLMG-HH-08, respectively, with an abandoned sister, her daughter and mother, and a frail father who was unable to work (and thus a mother that also required supporting along with him) 460 T-HH-06, BILR-HH15, T-HH T-HH-06 Market Development Facility 113

126 held in these accounts. Beyond bank savings, many families held jewellery as a form of savings and pawned it in times of need. Beyond bank accounts, households had access to microfinance individual and group loans. However, few households reported access to other services, such as insurance or commercial business loans. Ethnicity The tou is se to as o side ed to e la gel eth i all eut al i.e. the e e e people f o all ethnic groups involved in the sector and there were no apparent barriers to entry or operation specific to any ethnic group. However, in other sectors there are inter-ethnic differences and intra-ethnic disparities and these should be taken into account in planning interventions. For instance, the sample showed that Christian Tamils of lower castes were predominantly fishermen, whereas Tamil Hindus tended to work in agriculture and breed livestock. Sensitivity to ethnicity is a major consideration in the former conflict-affected areas. The conflict in these areas was borne out of a sense of ethnicity-based exclusion, and some of the ethnic issues that contributed to the conflict still remain. MDF work in this area will need to be cognisant of the need to promote ethnic harmony and social cohesion through its interventions. Other household characteristics The Study indicated that female-headed households were often disproportionately poor because women did not often work for money unless out of dire financial necessity. Female-headed households were found to be synonymous with the loss or incapacity of a male household member. There were fewer trends noted in relation to general characteristics such as geographical location of the household, or the types of economic activity the household was engaged in. In certain sectors and sub-sectors lack of proximity to key markets (e.g. tourism, horticulture and crafts) correlated with having lower incomes. In terms of education level of the sample, the Study showed that households in former-conflict affected areas had a higher proportion of male and female household heads with lower (Year 2-7) education levels. This was a result of years of displacement due to the war which caused disruption to both schooling and income earning. Another trend noted was that poorer households often had extended families with non-working adults to support, many of who were unable to work (either they were women who had been abandoned, or incapacitated relatives). Disability was seen in two forms across the sample: a) disabled workers (seen in households involved in crafts and located in former conflict-affected areas) and b) disabled or frail dependents that were unable to earn an income. Disability is not only a challenge for the disabled individual, but non-productive disabled dependents placed further financial limitations on the household with medical bills often adding significant costs to these households. While s hooli g is f ee i S i La ka, households still i u ed osts fo thei hild e s edu atio such as for private tuitions, uniforms, transport and school materials. These costs were often significant for these households who could only envisage themselves emerging from poverty when their children had grown up. 114 Market Development Facility

127 Women in households Across Sri Lanka women play the role of the household money manager, regardless of their level of i ol e e t ith the hus a d s o the fa il s i o e ea i g a ti ities. The ge e al practice across the Study sample was for men to hand their income to women on a daily or weekly basis, only holding back a small sum for their own personal spending. Household purchases were generally made by the women, who may consult their husbands. This was slightly different in Muslim households where it appeared that, for mobility reasons, men made household purchases, although here too there was often consultation with their wives. Women have a high level of influence over major spending. It was common for purchasing decisions to be determined jointly between the husband and wife. The initiators of ideas of what to buy and what was required for the household were often women. Women not only initiated the purchasing, but subsequently researched and found funding sources for it (such as from a MFI). This was more prominent where women played the dominant household earning role and the man had a lesser income or could not work. When women did not play such a prominent role, this was said to be due to a lack of confidence or belief in their own ability. Women were active in agriculture and the household functioned as a farming enterprise with men and women sharing tasks and jointly making decisions. In fishing households, women performed support functions, such as mending nets, cleaning fish. Drying fish and selling it was a women-led activity. Women were generally not involved in trade transactions. In households involved with fisheries and crafts, women were often cut out at the point of trade because they were uneasy in being in trade spaces which were dominated by men and it was seen as culturally inappropriate for women to be there. Sri Lankan women in the Study were most often found exerting their influence at the household level and assisting with tasks related to the business within the home. There were exceptions with women in households involved in horticulture in the South being highly active in input purchasing. Limited physical mobility played a strong role in the influence women had around trade functions. Women at work In the majority of cases, women worked out of necessity, citing issues such as difficulties in aki g e ds eet, fa il p o le s, ot ei g fi a iall sou d, a d that less edu ated o e worked because of poverty.462 The poorest households in the sample all involved women where the male breadwinner was absent or incapacitated. Once they do work, women talked in various forms of coming to enjoy the work because of the social support system it provided them, the sense of independence it brought, and the satisfaction of being able to support their children and families. 462 BILR-HH-02 Woman working in fish meal factory due to difficulties in making ends meet. BILR-HH-05 The majority of women work because of family difficulties. BILR-HH-08 Most women in her community do work, these women all work out of necessity. BILR-HH-11 Ma o e do t o k e ause the a e fi a iall sou d. BILR-HH-13 The other women (uneducated) work because of poverty. Market Development Facility 115

128 Women often worked in groups or with their husbands or a family member as a means of social protection. Women said this was because it increased their security. There was a different perception of what jobs married and unmarried women could do e.g. one woman believed that unmarried women preferred not to work as domestic help due to the risk of violence from employers.463 Female support systems at work were important. Relationships with the boss and other colleagues were also cited as important support systems.464 As was seen in all areas, it was common for women to work in groups, as this lowered the stigma around work and increased safety levels. Community support enabled women to do paid work and included help with child-minding, watching the business,465 and from husbands helping to run the business.466 Family reactions to women working were mixed, with most families coming to accept it in times of necessity.467 Community reactions to women working were also mixed and included discrimination and judgement from other Government jobs were regarded as more desirable, appropriate for educated women having high security (a pension) and respectable.468 Teaching was also regarded as suitable for women because it allowed women more family time and a balanced workload.469 This balance seemed to e p efe ed o e othe see i gl espe ta le jo s, ith o e o a oti g she a ted he daughter to be a teacher, but her daughter wanted to be a banker.470 Some women chose to stay, or were encouraged to stay, at home due to perceived and real safety concerns. Being outside the home for work seemed to have its risks, with one woman stating that she had been subject to abuse because of this, but she was not worried about it, and another stated that safety and security were big concerns when women chose their jobs.471 When women earned their own income, they tended to prioritise expenditure on their children. One woman prioritised savings for her children even when she could not save for herself.472 As BILR-HH-05 said unmarried women prefer not to work as domestic help due to the risk of violence. BILR-HHhome. She sa s the o e is good a d the p o ide eakfast e e da. A d she gets o e-hour lunch break so she is able to go 465 BILR-HH-01 Adult daughter maintains the household work. BILR-HH-11 husband helps cultivate the centella and babysits when she has full orders. BILR-HH-06 6-month-old daughter and other children are looked after by her neighbours and her mother. 466 BILR-HH BILR-HH-06 and BILR-HH-13 The family does not oppose her working due to the difficulty the household faces. 468 BILR-HH-04 They are seen as security. BILR-HH-09 He would like his daughter to secure a government job and so she is remaining at home until she can find one. He has similar aspirations for his son. 469 BILR-HH-19 Teaching allows women to have more family time and to balance workload. She prefers her daughter to be a teacher but the daughter wants to be a banker. 470 BILR-HH-19. BILR-HH-08 Most women in the community work - fo the o e ho do t o k, it is eithe e ause thei o kload is too u h o their husband/family has concerns about their safety. BILR-HH-11 Her business involves her going out alone and she has been subject to a use due to this, ut she is t too o ied a out it. BILR-HH-13 Safety and security is a big concern when they choose their jobs. BILR-HH10 In this region, due to the security and risk, most households refuse to send their women to work, it was said BILR-HH-02 Mother had deposited LKR 10,000 for each of their three children. 116 Market Development Facility

129 noted above, one of the main reasons women were found to be working was to enable them to pay to support others. Women as entrepreneurs Wo e s le el of i ol e e t i fa il u usi esses a ied idel. I so e se to s, su h as tourism and horticulture, where the male head was present, women often played less active roles around the family business and, in general, their position was more centred around the money management of the income gained. In other words, lack of direct involvement in income earning did not equate with less financial control in the household. However, in many other cases, particularly in the former conflict-affected areas and in crafts, women were active in decision-making and management roles in small businesses. The Way Forward for MDF As mentioned at the onset of the Study, Sri Lanka has made remarkable progress in addressing poverty, with the national poverty levels falling to 6.7 percent in 2013 from more than 22 percent in What these numbers obfuscate are the many Sri Lankan households, which, though not falling below the poverty line, subsist not much above it either. This segment of the population those that are not technically poor but vulnerable to poverty constitute 40 percent of S i La ka s population.473 The snapshot above paints a picture of some of these poor and vulnerable households across Sri Lanka. This picture shows that there is only cosmetic difference in incomes, spending patterns, savings and debt levels between the poor and the vulnerable. The reason why these households are trapped in poverty are multifaceted. For conflict-affected areas, poverty is inevitably a function of the conflict and the ensuing toll on life and livelihoods. In the dynamic tourism sector, those operating on the margins, informal operators, and new entrants in recently opened up areas, all struggle to eke out a share from the recent growth of this sector. For agriculture and fisheries, the reasons why these households are poor are again varied. The solutions that MDF offers to address poverty must be equally inventive, innovative and targeted. MDF s o je ti es a e t o-fold, to promote economic growth, but always in a manner that is inclusive of poor and disadvantaged groups and opens up opportunities for women. Economic growth, in itself, is not always inclusive. This drives MDF to select strategic areas of work, where involvement of poor people is reasonable, and where opportunities exist for enhanced markets to improve their lives. The analysis of growth constraints and poverty dynamics goes hand-in-hand for MDF at all stages of the program. At one level, there is an expansive analysis of broader economic and sectoral trends, growth constraints and identification of potential business opportunities. Complimentary to this, there is a need to zoom-in to the household level and understand what keeps the people engaging with these sub-sectors poor. This household-level poverty and gender study offers the 473 The World Bank, Sri Lanka Ending Poverty and Promoting Shared Prosperity: A Systematic Country Diagnostic. (Washington: World Bank Group,2015), Page Market Development Facility 117

130 learnings from this extensive and involved look at households, with the objective to guide MDF in developing specific solutions that enable poor households to participate more effectively in the subsectors and regions where MDF works. The findings point towards strategies, summarised as follows, that can unlock opportunities for the poor. Recommendations for Tourism Recognising that tourism creates many opportunities, the fo us of MDF s o k should e to a ds expanding tourism to new areas, so a more geographically diverse segment of Sri Lanka s populatio can benefit from growth in tou is. At the sa e ti e, MDF should also i est i a o e ope tourism sector that offers diversity of products, services and experiences. In order to influence this change, MDF can work in the following areas: Expansion of tourism to enable the absorption of workers in more untapped tourist regions, particularly in the post-conflict Eastern and Northern provinces. MDF can work directly with in-bound tour operators, hotels, activity providers to drive more of the increasing tourists in Sri Lanka towards these new areas. This will create opportunities directly and indirectly in tourism industry, for example, for hotel staff and farmers supplying to these respectively. Diversification of tourism activities, particularly within these emerging tourism destinations is necessary to develop and sustain increased tourism, and will also generate more employment in itself. MDF can work with activity providers to offer a broader range of activities, and assist them in developing local expertise which is able to cater to tourists by offering quality service. Stronger links between large operators and individual SMEs and communities will lead to a more inclusive and vibrant tourism offering in Sri Lanka. This is opposed to the la d tou us odels he e tou ists a e shuttled f o o e desti atio to a othe ith p e ediated stops. MDF will need to work with the large tour-operators to establish these linkages, while also supporting SMEs (individually or collectively) in improving their skills, standards and level of service so that they can engage large operators and their clients. Offsetting seasonality by diversifying activities and services will increase opportunities in areas where tourism season is short. The season can be lengthened by working with activity providers to offer diverse services, and working with the government and the private sector to develop and market diverse attractions. Households associated with tourism in these areas will automatically benefit from the expanded tourism season. Look carefully at skills building so as to enable skills provision to offer flexible, transferrable and softer skills sets. MDF s ole ould e to ide tif the eeds of the i dust, the i demand skill-sets, and work with private sector players to develop business models that allow them to train individuals for the industry. MDF can better connect crafts producers to growing markets such as tourism. Crafts offer many benefits for poor households such as the ability to o k f o ho e a d at o e s o s schedule. MDF can work with craft-sellers and producers to understand the demands of catering to tourists, and enabling them to target this market. This will lead to a bigger market for local crafts producers and increase opportunities for them. MDF can also look to 118 Market Development Facility

131 improve product design and associated branding by focusing on target segment suitability, differentiation and product experience. Recommendations for Fisheries For fisheries, the findings of the study were deceptively simple. The more fish a household caught the less poor it would be. However, complexities belie this simple fact. It is increasingly difficult for a poor fishing household to ramp up its catch, both due to the investments required and the amount of fish available to catch. MDF will need to address both factors. It can work in the following areas: Facilitate private companies and/or public private partnerships to address the poor management structures and the resultant low stocking levels in reservoirs. Increasing stocking will inevitably benefit inland fishers as this would enable them to catch more. This is the central challenge where inland fishing is concerned. MDF will need to work with a diverse range of partners ranging from the government to input providers and exporters to address this issue. The impact of the sudden net purchase and boats can be mitigated with buy-back schemes and better credit systems. This is the primary reason poor households are not able to increase their catch. The ability to purchase these crucial inputs on credit or instalment will provide these households the ability to increase their productivity and income. MDF will need to work with financial institutions to understand this problem and offer suitable solutions. Value-addition and processing of seafood to create new job opportunities is also an area which can benefit poor workers. This is a tremendous opportunity in the North, where almost all medium and large scale employment opportunities were wiped out due to the conflict. MDF can work with seafood processors to set up facilities that buy the fish from local fishers, processes it by employing local labour, and then export or retail the valueadded product. Improved post-harvest practices around fish storage and drying may have an impact on the income fishers are able to get from traders and buyers. MDF will need to identify the actor with the right incentives to invest in this area. Recommendations for Horticulture In agriculture, choice of crops, investment in inputs, availability of irrigation and connections to markets are all crucial factors which determine whether a crop would be profitable or not. Productivity and technical advancement in agriculture is low in Sri Lanka, which keeps farmers producing sub-optimally, and hence traps them in poverty. Some areas MDF can work in are as follows: Affordable irrigation options need to be made available for the Sirum Poham season (irrigated season) in areas such as Trincomalee, Batticaloa, and Jaffna and for the Maha season in Monaragala. MDF can work with companies offering solutions such as pumps and drip irrigation to better market their products and increase awareness. MDF needs to ensure that this is done in a way which does not stress underground water resources. Market Development Facility 119

132 Improved cropping pattern information should be developed and made available via appropriate channels. This finding reverberates across regions that farmers are often not aware of alternative, more profitable horticultural crops and keep producing traditional crops like paddy. MDF can work with a number of partners including government extension services, input companies, telecom companies to develop business models to disseminate this information. Related to the previous point, information on inputs, techniques, and investments could be improved, potentially via embedded information provision or via contract farming. The entry point for MDF can be input companies, agri and food processors, exporters, government and telecom providers. Lo - ost e ha isatio and other labour-reducing measures may be worth exploring in areas with labour shortages. This will allow landholders to scale up production, and for others to diversify into farming. For instance, the study found that fishers having land do not cultivate anything, because they are not able to devote any time to crops or afford to hire labour. The entry point for MDF can be local machine manufacturers or importers and interventions can target introducing new machinery as well as better marketing of existing machinery. Recommendations for Former Conflict-affected Areas Poor people in former conflict-affected areas face a compendium of challenges. The conflict caused disruption and trauma which had both economic and psychological consequences. Poor people are specially disadvantaged, having suffered loss of household members, loss of assets, dislocation and disrupted livelihoods. Constraints present in other engagement areas, for instance, in tourism and agriculture, are exacerbated in former conflict-affected areas. These are not separately discussed here. In terms of promoting business activity and broadening the economic base, MDF will need to support nascent local entrepreneurship while also encouraging more established businesses to expand their activities in these areas. MDF can work in the following areas: SME growth and development needs to be fostered, especially to fill the gap in medium size enterprises. MDF will need to be opportunity driven in this regard, identifying partners with the ability and willingness to make investments. Crucial areas for investments would include local manufacturing, re-establishing services, and value addition in produce, specially in fisheries. MDF can work with partners to offer services to assist people to make suitable investment decisions. The need for business mentorship is great along with other business services such as accounting and taxation consultancies. MDF will need to work from the ground-up with new businesses who are interested in offering these services. Correspondingly, SME growth needs to be designed to foster local hiring and sufficient job opportunities for local workers. MDF can influence its partners in former-conflict affected areas to hire locally which can lead to a more cohesive workplace and reduce tensions between the business and the community in which it exists. 120 Market Development Facility

133 Recommendations for Digital Services Digital services cover a broader realm than just workers employed directly in the ICT industry. While ICT workers tend to be fairly well-off, earning high salaries and able to progress in their roles. The scope for MDF, in this regard, is in reaching poor people across the country through ICT enabled services. This is a e a li g fa to i MDF s othe st ategi e gage e t a eas su h as tou is (relating to information dissemination, reservation services) and horticulture (information provision). Sri Lanka is well-positioned for ICT services to reach a large number of poor people and enable them to improve their lives. Ownership of cell phones is high among the population, as is the penetration of internet across the island. Based on this, MDF can leverage the following opportunities: Using ICT services there is potential to reach marginalised, disconnected groups farmers and fishers in remote locations, women in former conflict-affected areas which still remain relatively disconnected. MDF can work with the private sector to offer much needed ICT services to the broader population, but also to these groups, reaching whom would otherwise be cost-prohibitive. MDF can work with telecom providers to provide relevant and practical agricultural advice to farmers. In former-conflict affected areas, businesses can explore ICT based models to provide business services and consultancies. The peer-to-peer services, such as PickMe in transport and Airbnb in tourism lower the barrier to entry and provide a simple, user friendly way for men and women to offer services. Some of these services may be ideal for women, as they might provide flexibility of o ki g a ou d o e s o s s hedule. MDF a pla a ole i aki g these services more inclusive, by working with businesses to lower barriers which prevent more people from using these platforms. In transportation, for instance, more women could be encouraged to join by appropriate marketing and recruitment approaches. Similarly, homestay owners might need an intermediary to interact with services like Airbnb. Recommendations for WEE Across different regions, and value chains, the study found that women were grappling with similar constraints and were up against the same pervasive social norms. Following are the parameters which can contribute to greater WEE. With this understanding, MDF can influence its partners to eate the ight o ditio s hi h fa ilitate o e s pa ti ipatio. Market Development Facility 121

134 1. Proximity of workspace to home Proximity of the workspace to home could be addressed with MDF partners by initially reviewing close by workspace options, hiring from the immediate community, or setting up business operations close to the community. Home-based contracted or piecework options could be explored. This links to the next pathway safe transportation. The need for workspaces that are close to or easily accessible with reliable transportation comes a out e ause of o e s eed to pla dual oles at home and work, and it especially an issue for those with young children who need to be near the children or their schools if an emergency arises. This need to be available at home and at work feeds into the perceptions of others (often the husband) who are only willing to approve of women working if the women will be home at the end of the day. This attitude, for better or worse, is best accommodated than overlooked and in ti e it a ha ge o e the sig ifi a e of the o a s i o e to the household e o es lea e. A edgi g the doo ope app oa h is likel to e o e effe ti e tha t i g to fa ilitate a wholesale change. Proximate workspaces allow for work to be done concurrently, such as weaving while watching children. As it does in other countries in which it works, MDF should make its business partners aware of the importance of the proximity issue if they plan to hire a female workforce. 122 Market Development Facility

135 2. Provision of safe transport options Across the sample, mobility issues were found to be key and linked closely to safety. This issue should be tackled initially by checking if it is a problem at all perhaps the family resolves this with its own transport. Businesses could provide safe and affordable transport options with staff directly of hiring a contractor to provide this service. MDF could work with secondary service providers (such as Uber or Pick Me) to provide transport. More generally, this could either be done by working with dedicated transportation providers. Specifically relating to employment, MDF could look into working with companies to supporting in-house transportation options (either as an element of or constituting the whole partnership design), or to help these companies to obtain suitable outsourced transportation options. MDF should ensure companies seeking to hire women are aware of the key importance of safety around the movement of women to and from workplaces. Market Development Facility 123

136 &. Dedi ated orkspa e a d riti al ass hiri g optio s In most strategic engagement areas, it was found that women working together in groups reduced the stigma of women working and provided a key support system for them. They helped each other manage home-based tasks and care for children. These groups can be used to provide other support functions such as health promotion, banking or edu atio. The otio of st e gth i u e s is useful i add essi g o e s safet a d the a eptance of their roles as employees thei fa ilies. The e is also a iti al ass fa to, i that o e a e likel to e te a workplace if there are already women there so as to not bring themselves into disrepute. 124 Market Development Facility

137 5 & 6. Workforce age group and child care provision One of the groups most likely to drop out of the workforce is women with young children. There were few or no childcare options in most of the sectors studied, and how women juggled childcare was not well understood. MDF should consider childcare as a factor when in employing women, and to check if this can be accommodated at home (e.g. by other relatives) or whether it is a preclusive factor for women to employment. If so, MDF can work with the partner directly to set up embedded childcare, or look into other options such as contracted childcare facilities between the partner and an external provider or other standalone childcare facilities in the market. In taking a flexible approach appropriate options will emerge in a given case. Market Development Facility 125

138 7 & 8. Social stigma needs to be factored into the design of partnerships in highly pragmatic ways Social stigma around women working was an issue across the board in Sri Lanka, and plays a major role in when, how and even if women take up employment (or even consider taking up employment). MDF needs to remain aware that the social stigma of working in certain environments, in certain roles, and more broadly of working in general, play strong and silent roles in where (and whether) women work. Despite this stigma, there are many cases where women need to work, and signs that they enjoy this given the right set of circumstances. Stigma can be reduced by a combination of the practical measures namely, introducing safety measures, providing dedicated female work spaces and ensuring women have access to their homes through flexible transportation and home based work options. While stigma will continue to run deep in communities, MDF should not shy away from working to reduce it so that women can begin to enter work. Social change comes from changing practices and, with more women joining the workforce in suitable conditions, this may contribute to a gradual change in thinking around women at work. 9. Supporting female entrepreneurs with marketing and product development aspects of their businesses Women were able to obtain finance to set up small businesses on their own, but these businesses tend to stay at the micro level. They often lacked good guidance about product quality and markets. More needs to be done to ensure small female-led businesses with growth potential do not stagnate but are able to grow to support themselves and to enable them to hire staff to support the labour market. A deeper analysis of the enabling environment for female-led 126 Market Development Facility

139 businesses could be undertaken to better understand the position and limitations of female-led enterprises. 10. Mindfulness of the physical limitations of women when designing products, services, and jobs MDF should be aware of the physical limitations women face when they develop products and services. Factors such as product size or weight can play a significant role in whether women are able to use it or sell it. These practical factors are often overlooked even though women are often the ai u e s i a se to, a d the e a e ofte o feed a k loops i pla e to gauge o e s opinions on the quality of products and services. 11. Ensuring low visibility around conflict-inducing factors (e.g. money and assets). Situations where women suddenly began to make significant sums of money, through sales or wages, have been triggers for domestic conflict if there was alcoholism in the home. Alcoholism was prevalent amongst males in the sample in the North. Reducing the visibility of money and assets has been found, in other MDF countries, to be a key factor in reducing the risk of women losing her wealth. Workplaces can assist women to set up a bank account through which they can pay her rather than paying women in cash. This ill li it thei hus a d s a ess to thei o e, reduce the risk of arguments at home and the risk of being robbed on their way home from work.. Mi dful ess of o e s high le el o trol o er household cash flows and role in purchase decision making, and the implications for product and service design. Wo e i S i La ka s laggi g egio s ha e a high le el of o t ol o e household cash flows and household related decision-making. They often play very active roles in family businesses. When working on new product and service offers for these households, MDF and its partners need to keep in mind that the decisions to take up these offers is being undertaken by two household members, not one. Information on the products and services should effectively reach both the man and the woman of the household, a) because this naturally reflects the dynamics in hand and so as ot to u de i e o e s positio e o i g the o edu i g thei ole i this process. As women are often the money managers, product providers should be mindful that it is ofte o e u i g ite s f o shops, ot thei ale ta get a ket e.g. i o e etail shop i the Study, 60 percent of the customers were female). Women are often at the front line deciding whether the product is purchased or not. The implication is that more needs to be done to better understand product marketing in relation to women as well as men. General Recommendations Beyond recommendations for each strategic engagement area and WEE, the study found that management of household finances across all regions was lacking. Households, specially those having lean and flush season for income, struggled to fulfil their expenses, and ultimately relied on loans from family, friends, banks and other sources. They were also forced to sell assets and deplete savings (such as gold) in order to meet their needs. For these households, investments in productive Market Development Facility 127

140 assets were virtually impossible, and there was little financial cushion for them in the case of any economic shock. Despite the precarious position most households in the Study found themselves in, poor investments and purchasing of luxury items was common. This suggests an urgent need for financial literacy across poor and vulnerable communities in Sri Lanka. MDF has a role to play by working through its partners to offer information on financial planning, budgeting, saving and investment to its beneficiaries. In this manner, as households begin to earn more, they will also be in a better position to manage this extra income. At the same time, MDF can work with businesses such as banks, microfinance institutes, insurance companies, and other private companies to offer financial products that are suitable to the needs of poor and vulnerable clients. For instance, being able to buy agricultural machinery, nets and boats on easy instalments would enable households to increase their incomes. There is a need for appropriate insurance products and saving and investment schemes which understand and serve the needs of low-income clients. This Stud has gai ed a lot f o MDF s p io o k. The tea utilised its networks, prior fieldwork and analysis in order to inform the Study and guide it. In turn, the process of conducting the research, both primary and secondary, and the ensuing analysis has led to a richer, more nuanced understanding of poverty and gender dynamics within the strategic engagement areas that MDF works in. This understanding is already being utilised by the team in its day-to-day work. The findings and the conclusions drawn from this study will influence the strategic direction of MDF in Sri Lanka along with the design of its interventions. The Study will be instrumental in enabling MDF to design more sophisticated and targeted interventions. Through ongoing on-the-ground work and analysis of markets and their interaction with poor and vulnerable men and women, the team will develop further richness in its knowledge which would lead to a more nuanced analysis. In this regard, this Study should not be taken as an isolated, freestanding piece of research. MDF in its ongoing work will continue asking and analysing how poor a d ul e a le people a e ette se ed a kets. I this a e, MDF s o k a d this research are complimentary and iterative, underpinned by principles of constant learning informed by practice. 128 Market Development Facility

141 A e : Fi al Sa ple ANNEX 1 Final Sample Market Development Facility 129

142 Household Interviews 130 Market Development Facility

143 Code Location Respondent Gender of respondent(s) T-HH-01 Kandy Room boy from Amaya Male T-HH-02 Kandy Older female co-manager in hostel Female T-HH-03 Kandy Older female lead manager in hostel Female T-HH-04 Kandy Young employee at yoga center Female T-HH-05 Kandy Restaurant worker in Amaya Hills Female T-HH-06 Kandy Street fan vendor Female T-HH-07 Kandy Snake charmer/gypsy Male T-HH-08 Polonnaruwa Tour guide Male & Female T-HH-09 Arugam Bay Tuk tuk driver - informal guide Male T-HH-10 Arugam Bay Tuk tuk driver - transports surfboard Male & Female T-HH-11 Arugam Bay Tuk tuk driver for surfers Male T-HH-12 Arugam Bay National park shop seller (lead manager) Female T-HH-13 Arugam Bay National park shop seller (non-lead) Female T-HH-14 Arugam Bay Shopowner Male T-HH-15 Mirisa Whale watcher / kitchen worker Male T-HH-16 Mirisa Waiter Male T-HH-17 Galle Restaurant cook Male T-HH-18 Galle Juice seller Female T-HH-19 Galle Garment seller on beach Female T-HH-20 Galle Restaurant cleaner Male SLMG-HH-01 Kandy Labourer in spice farm Female SLMG-HH-02 Kandy Elderly HH spice farmers Male & Female SLMG-HH-03 Polonnaruwa Fishing HH - Parakama Samudra Male & Female SLMG-HH-04 Polonnaruwa Fishing HH - Parakama Samudra Male & Female SLMG-HH-05 Polonnaruwa Fishing HH - Parakama Samudra Male & Female SLMG-HH-06 Polonnaruwa Fishing HH - Minneriya Male & Female SLMG-HH-07 Polonnaruwa Fishing HH - Girithela Male & Female SLMG-HH-08 Maruthamunai Handloom sarong SME Male & Female SLMG-HH-09 Maruthamunai Handloom sarong + sari SME Female SLMG-HH-10 Moneragala Horticulture HH (corn, cashew) Male & Female SLMG-HH-11 Moneragala Horticulture HH (banana) Male & Female SLMG-HH-12 Moneragala Horticulture HH (vegetable) Male SLMG-HH-13 Moneragala Horticulture HH (paddy, chili) Male & Female SLMG-HH-14 Galle Pottery SME HH Male & Female SLMG-HH-15 Galle Pottery HH Male SLMG-HH-16 Ambalangoda SME woodcrafter Male SLMG-HH-17 Galle Disabled laborer in mask factory Male Market Development Facility 131

144 Code Location Respondent Gender of respondent(s) SLMG-HH-18 Galle Nursery worker Female SLMG-HH-19 Galle Nursery worker Female SLMG-HH-20 Galle Horticulture HH (vegetable) Male & Female SLMG-HH-21 Moneragala Small Horticulture farmer- Paddy & Vegetables Male & Female SLMG-HH-22 Moneragala Female Horticulture farmer- Pepper and Corn Female BILR-HH-01 Mannar Aquaculture-Fisher/worker Male BILR-HH-02 Mannar Fish Canning worker-coleman Female BILR-HH-03 Jaffna Fruit & Vegetable Farmer Male BILR-HH-04 Jaffna Fruit & Vegetable Farmer Male BILR-HH-05 Jaffna Disabled worker at Jam factory Female BILR-HH-06 Jaffna Female vegatable farm labor Female BILR-HH-07 Jaffna Worker at Sesame Oil processing factory Male BILR-HH-08 Jaffna Factory worker at Plastic recycling company Female BILR-HH-09 Jaffna Independent waste collector Male BILR-HH-10 Mullaitivu Female BILR-HH-11 Mullaitivu BILR-HH-12 Trincomalee Worker at a micro-food processor- IDP resettlement area Owner of a micro-food processor- IDP resettlement area Papaya farmer- prisoner of war- IDP refugee BILR-HH-13 Trincomalee Worker at Trader/Wholesaler- all goods Female BILR-HH-14 Trincomalee Owner- Agro input trader/distributor Male BILR-HH-15 Trincomalee Poultry farmer- IDP resettlement area Female BILR-HH-16 Trincomalee Hotel worker at a Guest House Male BILR-HH-17 Batticaloa Horticultural & Livestock Farmer- Muslim- Moor Male BILR-HH-18 Batticaloa Tourism Activity Provider- OWNER Male BILR-HH-19 Batticaloa Paddy & Peanuts farmer Male BILR-HH-20 Batticaloa Paddy farmer Male BILR-HH-21 Dambulla Fruits and Vegetables- Large commercial farmer Male DC-HH-01 Colombo IT worker- back office Male DC-HH-02 Kandy Owner- multiple IT companies Male DC-HH-03 Kandy Operations Manager- IT company Male 132 Market Development Facility Female Male

145 Key Informant Interviews Code Location Gender of respondent(s) Respondent T-KI-01 Kandy Female executive at Amaya Female T-KI-02 Kandy Manager/hotel owner Male T-KI-03 Polonnaruwa GN manager at hotel Male T-KI-04 Galle Shopkeeper Female T-KI-05 Galle Spice garden manager Male T-KI-06 Galle Tour guide GH manager Male T-KI-07 Galle Beach restaurant manager Female SLMG-KI-01 Kandy Spice garden tour guide Male SLMG-KI-02 Galle Group of Women Farmers- Ginger Female BILR-KI-01 Jaffna Organic mushroom farmer Male BILR-KI-02 Nilaveli Red onion farmer Male BILR-KI-03 Batticaloa Trader- Paddy and Horti crops Male Focus Group Discussions Code T-FGD-01 Location Sigira SLMG-FGD-01 SLMG-FGD-02 Maylambavadi SLMG-FGD-03 Gender of respondent(s) Respondent Eight tour guides (n=8) Eight fishe e a d o e fishe e s ife member of an association in Girithela Fourteen palmyrah weavers, mostly Tamil Hindus and few Tamil Christians in a fishing village Six female handloom weavers in a factory and the male factory owner, all Tamil Muslims. Male Male Female Female BILR-FGD-01 Trincomalee Tuktuk service- 3 drivers Male DC-FGD-01 Colombo Pick-Me-Drivers 2 female drivers Female Market Development Facility 133

146 A e : Pri ar Resear h Tools ANNEX 2 Primary Research Tools 134 Market Development Facility