WOMEN’S PARTICIPATION IN PAID EMPLOYMENT: A STUDY OF LOW INCOME HOUSEHOLDS IN BATTICALOA Ahila Thillainathan Abstract Women’s participation in income generating activities in Sri Lanka is relatively low, although its figures are better than other South Asian countries. According to the Central Bank of Sri Lanka’s socio-economic survey conducted in 2003/’04, female labour force participation has reduced from 32.5% in ‘96/’97 to 29.5% in ‘03/’04. The lowest figures, according to district-wise statistics, was seen to be in the Northern Province (17.5%) followed by the Eastern Province (18.2%). The study therefore focuses on identifying the reasons why women from low income households in the district of Batticaloa, Eastern Province, are/ are not involved in paid employment, as well as the environment that women prefer to be able to engage in income generating activities. Based on the capabilities approach, a preliminary survey was used to identify the key factors that are involved in determining whether women from low income households in income generation activities. The findings identified age and education as key factors under the capability variable and preference, attitude and child care under the motivation variable. A sample survey was then conducted to map the trends of each of the factors in order to construct the hypothesis model for calculating the index for each of these factors. It is recommended that the proposed composite index hypothesis model, for making assessments of the key indicators that determine women’s participation in paid employment in low income households, be tested next for consistency. Keywords: Women, Paid and unpaid work, Batticaloa, Capabilities approach, Arrow’s Impossibility Theorem

Introduction The South Asian region has relatively lower labour force participation rates for women, when compared to other regions in the world. Only around 40 women per 100 men in the labour force are economically active in South Asia (ILO, 2004) The Labour Force Participation of Sri Lanka, excluding Northern and Eastern provinces has remained constant around 47.6% in the past 8 years (Central Bank Sri Lanka, 2004). At the same time, male labour force participation has increased from 64% in ‘96/’97 to 66.2% in ‘03/’04, while female labour force participation has reduced from 32.5% to 30.9% in the respective years. When the Northern and Eastern provinces are included in the above statistics (excluding Killinochchi, Mullaitivu and Mannar), the figures change considerably, with an overall reduction in labour force participation from 47.6% in 1996/’97 to 46.4% in 2003/’04. This drop is not attributed to male labour force participation, as an increase is noted (65.3%) but to further reduction of female labour force participation (29.5%). When comparing province-wise information on labour force participation, the lowest figures for female labour force participation is seen to be in the Northern Province (17.5%) followed by the Eastern Province (18.2%). The long years of civil conflict which directly affected the North and East provinces will have played a major role in causing the above statistical figures. At the same time, it must be considered that the gap between male to female employment rates has widened in

1

these districts, even though the gap between male to female literacy rates is negligible. This observation is compounded by the fact that the unemployment rates of females is 7 times more than males in the eastern province and in the Northern Province, nearly four times that of males. The all island figures indicate that unemployment rate of females is roughly double that of males. The male labour force participation of Sri Lanka is seen to reflect a normal labour force participation path with the participation rate gradually rising with age, reaching a peak participation rate in the 35-44 age group and gradually declining with increasing age. There are relatively high levels of participation in the peak economically active period of 25 – 54 years, with considerable participation rates in the 18-24 and 55 - 61. When comparing women’s labour force participation rate by age group, the peak age for women’s participation in employment activities is observed to be in the 18-24 age group, following which there is a gradual drop across the 25-34 age group. The participation of women in income generation activities from the 35-44 age group does recover, before declining with further age. That these figures reflect a general norm in Sri Lankan society, where women stop engaging in income generating activities after marriage and the men work harder to increase the income of the family to support an expanding family is indisputable. Especially when the major reason quoted by females for not participating in the labour force of the country is housework followed by schooling and vocational training1. The Northern Eastern provinces have witnessed a gradual deterioration in the economic, social and physical conditions over the two decades prior to the signing of the ceasefire agreement. During these conflict years, the contributions of the regions to national GDP dropped from 15% in the 80s and to 4% in 97. Educational attainment rates have sharply fallen with higher school drop-out rates and prevalent malnutrition. “Loss of civilian life, physical and psychological trauma, the horror of forced displacement, the disintegration of community social networks, forcible recruitment into militant organizations, constant fear and uncertainty, and prolonged dependence on external relief are all facets of impoverishment in the north-east. No reliable information is yet available to indicate the depth of poverty in these two provinces.” (UNDP, _) An effect of the conflict on the agriculture sector in the conflict-affected areas is said to be the steady dominance of the sector on the region’s economy, while at the same time, the disrupted transportation and market venues converted the sector into a subsistence sector (Sarvananthan, 2003). In other words, this transformation led to the decrease in the agricultural production of the region which can be translated into less labour input in the sector and increase in unemployment. Whatever industries existed prior to conflict has been greatly decimated in the conflict-period. Access to economic infrastructures such as safe water, electricity, and telecommunications are below national averages. The 90s was also a period of great insecurity for youth of the North and East and thus the two inter-related conditions mentioned above led to young men migrating to other parts of the country or abroad, in search of both better economic prospects and security and the emergence of increasing single parent headed households, especially women headed households. The burden of single parenthood after the division of labour in the two parent families led to a drastic circumstance, where single parents found themselves unable to provide both the income and care of their families. This has increased the poverty levels in such households. In the case of women, they found themselves with little or no marketable skill, nor the knowledge or capital to start self-employment and the families had to depend mainly on food 1

Central Bank of Sri Lanka. 2004. Consumer Finances and Socio Economic Survey 2003/04.

2

stamps, dry rations and ‘picchai sambalam’ (beggar’s wages). Women’s selection of income generation and employment patterns fall strictly into ad hoc patterns of temporary jobs. (Thiruchandiran, 1999). A major challenge indicated by the Millennium Development Goal Country Report 2005 Sri Lanka is providing an enabling environment that would allow women from low-income households to move out of poverty. While many development agencies, both government and non-government bodies, have been involved in the reconstruction and rebuilding efforts of the North and East conflict-affected areas since the ceasefire agreement was signed in 2002, there is a need for having a focused and productive livelihood strategy targeting women in order to increase the women’s participation in employment activities. The specific reasons of women in choosing/ not choosing to work must be analysed before designing any poverty alleviation programme for rural women and especially, the environment under which women currently not employed in income generation activities might choose to work must be taken into account. Given that the labour force participation of women is lower in the conflict-affected areas than in the rest of the country and as the eastern province displays a higher unemployment gap between females and males than the Northern Province, this study focuses on the Eastern Province. Given that there are three districts in the Eastern Province, the study next narrows its choice of focus to the district of Batticaloa, which is the central district in the Eastern province. This study therefore focuses on low-income households in a selected rural GN division in the district of Batticaloa in the eastern province of Sri Lanka and hopes to make a positive contribution to the stated objective, of identifying the reasons that women engage in paid work and the environment that will increase women’s participation in paid work. The study further proposes a model for the factors that determine the labour force participation rate of women, thereby providing a guideline for development agencies to assess current situation of women’s labour force participation in low-income households and also the factors that need to be targeted in order to promote women’s increased participation in income generating activities. Background of Batticaloa District, Manmunai Pattu Divisional Secretariat Division, Pudhukudiyirruppu (Central and South Grama Niladhari Divisions) The district of Batticaloa occupies the Central Part of the Eastern Province, covering an approx. land area of 2633.1 sq.km and internal waterway of 229 sq.km and accounts for 3.8% of the countries total land area. It is bordered in the north by Verugal Aru (river) and Trincomalee district, Bay of Bengal in the east, Amparai District in the south and south-west and Polonnaruwa in the west and north west. The district is divided into 14 administrative divisions (divisional secretariat divisions). The physical feature of the Batticaloa district is flat alluvial land in the west coast, bordering the lagoons and sandy soil in the east coast and the height of the land does not exceed 7.62 m above the sea level. The lagoon passes through the district and the town of Batticaloa, from Verugal in the north to Thuraineelavanai in the south, covering a total distance of 108.7 km. The population is mainly concentrated in the narrow strip between the sea and the lagoon in the east coast, while the west coast is sparsely populated and 90% of the paddy land is found here. The available mineral resources are rock deposits, sea shell, pan grass, cane, clay and forest timber. Agriculture is the major occupational sector (nearly 30,000 families) in the district with paddy cultivation (around 58,374 hectares), followed by coconut cultivation, onions, chillies, betel and vegetable cultivation. Cashew plantations are also an important agricultural activity. Fishing is the second major sector with nearly 16300 families engaged in both lagoon fishing and coastal fishing. The major industrial establishments are small-scale industries, which include weaving, rice-milling,

3

pottery, mat-weaving and brick-manufacturing, are located in the town of Batticaloa and Kattankudy. The only large-scale industrial factories are the National Paper Corporation, Valaichenai, and the garment factories, two in the town of Batticaloa and the other in Valaichenai. Other sources of employment are in Government, Corporation and private establishments. A teaching hospital in the town of Batticaloa caters to the entire district and the eastern province. In addition, five smaller private hospitals are also located in the town for the public. Though there are peripheral unit district hospitals and dispensaries functioning in most rural areas, some of the centers are not functioning due to the security situation. A student population of 115041 attend 308 schools, with 3712 teachers, in the district. The Eastern University with the faculties of agriculture, science, commerce and arts is located at Vantharamoolai, Chenkalady. Technical trainings are provided at the Technical College, Manchanthoduwai and the Eastern Technical Institute, Batticaloa town. During the conflict years between 1983 and 2002, investments dropped (District Planning Secretariat Batticaloa, 1999) resulting in the closure of the State Printing Complex at Kumburumoolai, Rice Milling Complex at Sittandy, Rice Mill at Batticaloa, Tile factories at Elluppaddichchenai and Mandur, re-trenchments at the National Paper Company Valaichchenai. The four straw collecting centers that provided raw materials for the paper factory were also closed during this period. More than 8000 persons, who engaged in paddy cultivation in the western area of the lagoon gave up their activity, as did thousands who reared cattle. Restrictions imposed on the fishing activity affected nearly 90% of those engaged in fishing and shifted them to the low income group. The highest female to male unemployment ratio was found to be at Manmunai North, Manmunai Pattu and Manmunai South and Eruvil. (District Planning Secretariat, 1999) As Manmunai North is an urban area, it was not considered. Further when the population per DS division in Batticaloa was considered, Manmunai Pattu was found to be the median DS Division with a population of 8262 families and a population size of 30284 and also the median DS Division, in terms of the percentage of Samurdhi beneficiary families of the total number of families (65.56%) living in the DS Division. Hence the DS Division of Manmunai Pattu was selected as a representative DS Division of Batticaloa. Manmunai Pattu DS Division has 27 GN Divisions and 89 villages, with a total of 8264 families living in the Division. Nearly 2% of the individuals living in this DS Division are disabled. In terms of the mainstay economic sector Batticaloa, agriculture, Manmunai Pattu had the highest cashew production and highest subsidy assistance for cashew cultivation under SLCC’s assistance in the 2003-2004 years. In terms of effects of the tsunami, Manmunai Pattu also suffered the most damage to its cashew plantation (nearly 40% of total cashew cultivating area was affected). Extent of Coconut cultivation is third largest in the district. The Division has also two palmyrah training centers, of the total 8 training centers in the district. With regard to permanent crop cultivation, Manmunai Pattu engages in the cultivation of oranges, limes, mango, jak, breadfruit, sugarcane, plantain, papaw and arecanut. Manmunai Pattu also had the third lowest harvested acreage amongst the DS Divisions engaged in paddy cultivation in the Maha season in Batticaloa. Other highland crops cultivated by Manmunai Pattu in the Maha season are manioc and sweet potato (medium ranking producers) and maize, ground nuts (lowest ranking producers). During the Yala season, the main highland crop cultivation are maize, green gram, black gram, ground nuts, manioc, sweet potato. The

4

vegetables grown in both the Maha and Yala season are red onions, chillies, ladies finger, brinjal, bittergourd, snakegourd, tomato, cucumber, ash pumpkin, red pumpkin, ash plantain and capsicum. Manmunai Pattu DS Division had the lowest fish production for the year 2004 but has the fifth highest percentage of families engaged in fishing, when compared to the rest of the Batticaloa DS Divisions. When comparing livestock statistics, the Division has also the lowest average monthly milk production (760 litres/ buffaloes and 11,123 litres/neat cattle) and the average monthly egg production of the DS Division was 3.4% of the total average monthly egg production for the year 2004. The registered cooperative societies in Manmunai Pattu include one multi-purpose cooperative society (MPCS), one secondary cooperative society, thirteen fisheries societies, twenty-nine thrift and cooperative societies and one industrial school cooperative society. In addition, there are 23 rural development societies, 10 women’s rural development societies, 13 fisheries’ societies, 27 samurdhi societies, 18 sports clubs, 12 social services societies and 51 other registered clubs. Manmunai Pattu has 27 pre-schools and 22 schools. In the 2004 results analysis pass rate for the students who sat for the Grade 5 scholarship was 8% (35), O/L 41% (113) and A/L 55% (40). A total of 5990 students are enrolled in Manmunai Pattu, with the maximum in primary school (2972) followed by secondary school enrollment (2298). Of the total unemployed graduates (100) recorded by the Divisional Secretary, 94 are art graduates. When considering the labour force potential of the different GN Divisions in Manmunai Pattu, calculated here in terms of the population over 18 with regard to the total population, the median division falls in Pudhukudiyiruppu South. When considering the female population, Pudhukudiyirruppu Central is the median division and hence for the purpose of this study, the two GN Divisions Pudukudiyirruppu Central and South were considered. The two GN divisions together comprise a total of 8 villages, having a total population of 1869 (521 families). With regard to basic facilities, there is a pre-school as well as a school up to A/L in the Pudhukudiyiruppu GN Divisions (North, Central and South). With regard to communications and transport, there is a post office and a communication center serving the population and buses pass the villages through the main road to Batticaloa and Kalmunai, every ten minutes or so. There is no internal public transport service within the villages but there are internal access roads and most villagers move within the village on foot or bicycles. With regard to essential health services, an ayurveda hospital in the GN Division serves the immediate needs of the villagers. For a western-medicine based hospital, the villagers have to travel to Araiyampathy, located nearly 4 miles away. The nearest market is also located in Araiyampathy. For banking services, the villagers have to travel to Kattankudy, beyond Araiyampathy. Agriculture and fishing has been the main sources of income generation for the male labour force, while for women, poultry-raising, home-gardening, cottage industries like coconut fibre making, palm leave weaving for mats, boxes etc and handweaving. There has been a high degree of migration of young men to the Middle eastern countries since the 90s, looking for better social and economic prospects. This has led to an increase in de facto women headed households. Poultry-raising has been a major source of supplementary income and nutrition but most of the households lost their livestock in the tsunami. According to the Village committee leader, a major income generating activity for women prior to 2000 was hand weaving but with the signing of the Indo-Lanka Free Trade Agreement and the influx of cheaper and higher quality Indian hand

5

woven products in Sri Lanka, the hand weavers of Pudhukudiyirruppu amongst other hand weavers lost their source of income.

Capabilities Approach “The key idea of the capability approach is that social arrangements should aim to expand people’s capabilities – their freedom to promote or achieve valuable beings and doings. The basic concepts under the capabilities approach are functionings (valuable activities and states that make up people’s wellbeing – such as a healthy body, being safe, being calm, having a warm friendship, an educated mind, a good job), capabilities (“the substantive freedoms he or she enjoys to lead the kind of life he or she has reason to value.”) and agency (a person’s ability to pursue and realize goals that he or she values and has reason to value)” (Alkire, 2005) Alkire (2005) further mentions that more often than not, capabilities will have to be selected by a community, by a team, or by a researcher. The key questions to keep in mind when selecting capabilities are: which capabilities do the people who will enjoy them value (and attach a high priority to) - this must be explored directly; which capabilities are relevant to the policy, project, or institution; which may be affected directly or indirectly. With the capabilities approach as the theoretical framework for defining the model for this study, a definite set of capabilities and motivation was identified and the trend for each of these factors was analysed for the sample population in order to define a model that would serve to determine the labour force participation rate of women from low-income households and which could be utilized to identify the areas that need to be targeted by development agencies, both individuals and organizations, when targeting the economic independence of women, through increasing women’s engagement in paid employment. For the purpose of this study, capabilities are defined as one’s inherent physical abilities, either those one is born with or acquired, such as age, level of educational attainment, physical health etc. at the time of the study and motivation is defined as one’s reaction to social conditions in the environment, such as preference for type of work, attitude towards work, child care, nett household income etc. While motivational factors are considered to be more malleable to changes in society, capability factors are considered to be less malleable to change. Also, motivational factors are less easy to measure and hence specific methods of assessment are required. Here, preferences and attitudes of the social class of women of low-income households are understood to be the aggregate of sample preference votes with the selected choice corresponding to the highest votes indicative of the preference of the women for the type of work and their attitude towards paid work. When deriving conclusions based on preference profiles of this sample of women, Arrow’s theorem for Social Choice or the Impossibility Theorem2, must be kept in perspective and that whenever there are two or more voters involved, then it is impossible to ensure that a completely just and fair preference ordering (unanimity or pareto-efficiency, non-dictatorship, irrelevance of 2

California Institute of Technology. Implications of Arrow’s Impossibility Theorem for Social Choice Methods.

http://alumnus.caltech.edu/~seppley/Arrow's%20Impossibility%20Theorem%20for%20Social%20Choice%20Methods .htm#Proof%20of%20Arrow's%20theorem

6

independent alternatives) is obtained. A major limitation however is that the choices made available for voting is bound and determined by the survey designer’s perceptions of the available options. Preliminary survey In order to gain an initial understanding of the village of Pudhukudiyirruppu and the women’s work therein, a focus group discussion with a purposive and biased sample of twenty-three members of the revolving loan fund, set up by the Working Women’s Development Foundation, for starting or expanding income generating activities, was held on December 24th, 2005. All the participants owned their own houses and land, but the tsunami had destroyed their houses and at the time of the meeting, lived in temporary shelters constructed on their own land and sharing communal toilets. Drinking water was still brought in a bowser daily from the town of Araiyampathy. Almost all participants had engaged in raising poultry and other livestock, before the tsunami and one participant even mentioned that she had owned up to 50 poultry which had generated considerable income but the tsunami had destroyed the entire livestock. A year afterwards, the women mentioned that they still did not feel like raising poultry or any livestock again. “I lived with my mother and a goat. The tsunami came and now my mother and goat have gone. I do not have anyone anymore.” Similar feelings were prevalent amongst other women whose family members had died in the tsunami. At the same time, it was observed that the women, for whom the damage by the tsunami was limited to material loss, were more interested in improving their standard of living. The women’s perception of wealth mainly focused on the material commodities of land, house, vehicle and money. The discussion was then focused on women in paid work. The participants estimated that around 40 women in the village were engaged in paid work, provided by NGOs. Nearly 50% of these women were engaged in handcraft and cottage industry. It also evolved that many women did not consider poultry farming or home gardening as work, even though it may have generated irregular income. Neither did they consider home care as work. With regard to women engaged in paid work, it was observed that while it was universally seen by the villagers as something undertaken for survival, the drive to increase income was observed among those women who were solely responsible for providing economically for their families, children, parents or both. Where the male member of the house was present and an income receiver, who earned sufficient income for the family, the women would not have opted to engage in paid work but limited themselves to domestic caring. The reasons given for other women for not engaging in paid work was that there were lots of children to take care of and thus they did not have time to engage in paid work and also in cases where the husband was working, he did not wish his wife to engage in paid work. It was considered an issue of his dignity, if his wife had to engage in paid work, which seemed to imply that he was unable to provide for his family and had to send his wife to work. The available job opportunities in the village for women were considered to be sewing, cane product-making, mat weaving, agriculture – weeding, seeding and harvesting and small business like buying and selling rice, coconut and other essentials and selling food. Most non-working women were interested in the above-mentioned jobs, as a source of additional income, but were

7

unable to engage in them due to lack of capital and tools and those already working expressed interest in any work opportunities that would provide them with a higher income than what they currently earned, provided the work place was located in the village. Box 2.1. The stories of the women Mehala, 22 years old studied up to her O/L and her sister, 23, completed her A/L. Both of them live with their parents. Both of them do not engage in any kind of paid work and depend on their parents. Her 70 year old father continues to be the primary income earner, through his masonry work generating 200 rupees per day, while her 58 year old mother buys and sells coconuts. They were affected by the tsunami and currently live in a temporary shelter.

Uthayamalar is a 33 year old woman who lives with her 58 year old mother and 24 year old brother. Two of her siblings died during the conflict years of the 90s and her father died in the tsunami of 2004. She and her brother and mother live in a temporary shelter after the tsunami washed away their house. Uthayamalar studied up to 11th Grade but had to discontinue her studies due to displacement. Neither her mother nor she work. The family depends on the male member, though the youngest, for the household income, in addition to the welfare assistance of Samurdhi and the tsunami stamp. Her brother works as a mason, after undergoing training provided by an NGO and now earns a daily income of 150 rupees, the total income depending on the number of days he works. Uthayamalar is now planning on investing in home-gardening, after the mobilization of WWDF.

Thirumangai is a 50 year old woman who has three children, two daughters aged 28 and 16 and an 18 year old son. Her daughters dropped out of school after completing their 10th and O/L, while the son completed his A/L. Thirumangai and her husband both dropped out of school in their primary schooling years, due to the poverty situation in their respective homes. Her husband is engaged in the cane business and the nature of his work is such that there is no regular income and his work very much depends on his ability to go to the lagoon and cut cane, which was considered a very difficult task, with little income generated. The family lost their home in the tsunami and now live in a temporary shelter. Thirumangai buys rice from the market and sells to people in her village. Her reason for engaging in income generating activity is that her husband’s work does not generate sufficient income for the family and so she needs to contribute to take care of her family.

Inference from focus group discussion Work pattern differs across age groups. The women in the 20s and 30s , with an emphasis on the under 35 age group, was observed to have more diverse income generating work and skill level than women in the 40s and 50s age group, who were focused on basic buying and selling of essential food products. This change could be attributed to the vocational training skills imparted to the younger age groups, through local NGO interventions, in the post-conflict reconstruction environment. Level of educational attainment pattern is different across age groups. The level of educational attainment was seen to steadily rise across the strata as the age group decreased and women in their 20s are observed to be educated up to a mean education level of O/Ls, in contrast with the

8

Grade 4 mean education level of women in their 50s. An increasing prominence to education was observed in the younger age groups and women in the age range of 35 – 45, who had school going children, were observed to be keen on providing a good education for their children, even investing on tuition classes. All school going children of respondents below the age of 40 were attending school but school drop-outs due to poverty were observed amongst children of respondents above 40 years. This behaviour may indicate a change in living standards for the better amongst respondents or the change in attitudes towards education. Especially as most respondents cited poverty as the major reason for disrupting their own studies. Economic responsibility pattern differs across age groups. A larger proportion of economically women-headed households were observed in the 30s and 40s age group. Dependency ratio is seen to rise with age. i.e. for every working member of the household, the number of family members dependent on his or her earnings is seen to increase with the age of the respondent. Very few respondents mentioned that they had a monthly savings habit. Motivation level differs across age groups. Women in the 35 – 45 age group expressed determination in working and interest in opportunities in improving income, as they wished to provide a better living standard for their children Hypothesis formulation: Based on the preliminary survey and the range of factors identified that would influence the female labour force participation, a simple hypothesis model was developed, considering capabilities (age, education level) and motivation (Earning of male member of household, child care, attitude and preference towards work). As child care is a difficult criteria to measure, especially when it is difficult to gauge the number of hours that women engage in domestic care, for the purpose of this study child care shall represent the number of children in the household. The number of women engaged in paid work in low income household increases under the following circumstances: (1) Motivation (M): a. If the monthly income earned by the male head of the household is less than the amount above which welfare assistance would become inapplicable (< 2000) b. If daily hours of work spent in domestic care and child care is less than eight hours c. If attitude and preference of the women towards paid work is positive (2) Capability (C): a. If age is in the peak range of physical ability b. Education Based on the above factors, the proposed hypothesis model for determining female labour force participation in low income household is a composite index, which transforms a raw variable into a unit free index between 0 and 1. This transformation draws on the Human Development Index3, 3

UNDP. 2004. Technical Note: Calculating the human development indices.

http://hdr.undp.org/docs/statistics/indices/technote_1.pdf

9

but uses a normal distribution curve to assign values between 0 and 1 for each of the sub-variable range of possibilities. The resulting index would provide an indication of whether women would be likely to participate in an income generation programme, with values closer to 1, indicating greater inclination than values closer to 0. The proposed composite index model,

LW = (C + M ) / 2 where, C = and, M =

1 ∑ (C xi + C yi ) 2 i =1

1 ∑ (M xi + M yi + M zi ) 3 i =1

C x = age, C y = educational level, M x = child care, M y = Preference in relation to work environment, and M z = Attitude towards work The ranking for each of the sub-factors, is valued on a normal distribution curve on a scale of 0 – 1, with 0 indicating tendency towards reduced women’s labour force participation and 1 indicating a higher tendency of women to engage in paid employment. The sample survey therefore analyses the trend of the two capability and three motivation subfactors mentioned above and attempts to formulate a normal distribution curve for each, based on the trend. The questionnaire for the sample survey was designed, based on the Sri Lanka Labour force survey questionnaire and the Central Bank of Sri Lanka’s Consumer Finances and Socioeconomic survey questionnaire, so as to map the trends in age and educational level, work status of household members, working environment preference and attitude towards women and paid work. Sample Selection: The voters list was obtained from the Grama Niladhari (GN) Officer, Pudhukudiyirruppu GN Division and the population size (1247) was obtained. The required sample size4 was calculated as follows: Confidence limit = 1.96 Estimating standard deviation: 15 – 64 age group Male/ Female ratio5 = 0.94 Average number of females in population, according to above ratio = 1247 x 0.53 = 661 Preliminary estimate within which range for total female population = 661 – (1247/2) = 37 Range = 624 to 698 Estimated standard deviation = 74/4 = 18.5 Minimum precision that would be acceptable = 5 2

 1.96 x18.5   = 49 5  

Required sample size = 

4 5

_______. Sample Size Determination. Sampling and Sample Size. 402-204 CIA. The World Factbook Sri Lanka. https://www.cia.gov/cia/publications/factbook/geos/ce.html

10

Sampling fraction = 49/1247 = 0.04 A random number generator6 software programme was used to generate the numbers of persons to be interviewed. Six volunteer surveyors were provided with the number lists and corresponding names from the voters’ list and data collected for 60 women. As low income households were the target of the study and as it was realized from the preliminary survey that it would be difficult to ascertain the real incomes of the household, for consistency of this study, households receiving Samurdhi assistance was considered to be a lowincome household, irrespective of the real income earned by the household members. Where a number in the random number list indicated a household not receiving Samurdhi assistance, that number and corresponding household was deleted from the study sample. Also, where the number in the voter’s list indicated a male member, the corresponding female member in the household was interviewed, thereby ensuring the randomness of the sampling. The sixty women thus interviewed were considered to be a representative sample 50 of the female labour force of 40 the GN divisions of Pudukudiyirruppu Central and 30 South. After decoding the data 20 collected and discounting the forms with data collection 10 errors, the actual sample size 0 used for the study was 47. The 20.5 24.5 28.5 32.5 36.5 40.5 44.5 48.5 52.5 age distribution of the Age respondents is illustrated in figure 3.1. The cumulative frequency distribution indicates the highest frequency in the 25 -34 age group followed by the 35 – 44 with the modal class being in the 35 – 38 range. Cumulative frequency

Fig 3.1.Age distibution across sample

Analysis The data gathered from the questionnaires is presented here as two categories: capability and motivation. The work trend based on age and education was analyzed under capability. Work trend based on child care responsibilities, preference and attitude was analyzed under motivation. Capability: Work trend based on age The cumulative frequency distribution for women engaged in paid work and unpaid work indicates a narrow gap between the two for the age groups 18 – 34 (fig 4.1) with an increasing gap beyond 34 years, with a positive inclination towards paid work. Further, when considering the type of work that women engage in, more emphasis of women who are engaged in paid work is on self-employment. The trend (fig 4.2) also suggests that there is a narrow gap between selfemployment and casual labour in the 18 – 24 age group and the gap starts widening after 24

Paid and unpaid work

40

Cumulative frequency

fig 4.2. Type of work across age

50

25 20

Regular

Haahr. 30 M. 1999. True random numbers. http://www.random.org/nform.html 15 Unpaid work Total potential

10

11 .5

.5

.5 52

48

.5

Age

44

40

.5

.5

.5

.5 36

32

28

24

.5

0

Self-employment Casual

10

Unpaid work

5 0

20 .5 24 .5 28 .5 32 .5 36 .5 40 .5 44 .5 48 .5 52 .5

20

20

6

Cum ulative frequency

fig 4.1. Work trend across age

Age

years, with a peak in the 45 – 54 age group. Regular/ permanent employment is a negligible factor in the given sample. Also, the frequency distribution table indicates that the number of women engaged in unpaid work when compared to each category of paid work. Here, the categories considered under paid work are regular or permanent employment, self-employment and casual labour. What this indicates is that women are more inclined towards paid work during their 35 – 54 age period and that there is increasing tendency towards self-employment. Capability: Work trend based on education The cumulative frequency distribution curve for the educational pattern of women across each age group (fig 4.3) indicates that the majority of women in the 45 – 54 age group have been educated up to primary school level but with the highest school dropouts. In the age group 25 – 44, there is an equal number of women who have had no schooling. Also, a shift towards secondary school attainment levels is seen to increase with each younger generation and in the current 18 – 24 age group, it is seen to be the minimum level of educational attainment. Further, the trend of participation of women in paid work (fig 4.4) is observed to have a peak when the 25 education attainment is at primary 20 No education level and the peak for women Primary 15 choosing to stay at home and engage Secondary in unpaid work is at the secondary 10 A/L school level. While at each level of University 5 education, the number of women 0 engaged in paid work is observed to 18-24 25-34 35-44 45-54 be higher than those choosing to Age group engage in unpaid work, the exceptions are the secondary school leavers and the graduates, where there is a strong bias towards unpaid family work. Cumulative frequency

fig 4.3. Patterns of education across age group

When considering the type of paid work across different levels of education (fig 4.5), while those with no schooling are seen to have equal preference for self-employment, casual labour and unpaid work, there is a shift amongst those choosing either category with increasing levels of education. Also, while self-employment remains the most popular form of paid work, the only exception to the case is found to be the primary school leavers, where there are more women engaged in casual labour than self-employment. As those dropping out of school at the primary school level do so mostly due to poverty conditions, there is a tendency to engage in paid work amongst this category. Further, as one of the major form of employment for women in Pudukudiyirruppu is agricultural casual labour, the primary school leavers tend to engage themselves more in casual labour. This finding is in line

Educational Attainment

Self-employment Casual

Educational level

12

re e De g

A/ L

Unpaid work O /L

re e

A/ L

O /L

De g

No

No

sc ho ol in

g Pr im ar y Se co nd ar y

Unpaid work

Regular

g Pr im ar y S ec on da ry

Paid + Unpaid work

12 10 8 6 4 2 0 sc ho ol in

12 10 8 6 4 2 0

fig 4.5. Effect of education on type of work Distribution across type of work

Participation in work

fig 4.4. Effect of education on work

with the hypothesis that education is a capability factor, which determines what type of paid work women will engage in, when considering income generating activities. Further, the observation that indicates that women who have dropped out of school in the secondary level are more inclined to unpaid work than paid work may be based on the fact that with increasing number of women studying up to the secondary school level means that more of the labour force are in the secondary school level of educational attainment could be indicative of the employment opportunity rather than the choice of the women concerned. However, the observation needs to be kept in perspective, especially as it is also observed that there is a drastic drop in the number of women engaging in casual labour, compared to self-employment, under this educational category. The secondary school level could be viewed as the current bar of educational attainment that determines whether the individual would be inclined to engage in casual labour or self-employment, when considering paid work.

Trend

Motivation: Work trend vs. Child Care One of the assumptions made in this study was that increasing domestic responsibilities was equated to increasing children and caring responsibilities and was considered to be an inherent factor contributing to women’s reduced participation in paid work. The cumulative frequency distribution graph (fig 4.6) of women fig 4.6. Work trend vs. Child care engaged in both paid work and home care and those limited to home care 50 presents an unexpected finding in the Paid and unpaid 40 case of the low income households work 30 Unpaid work studied. 20 Total potential for

What is observed is that though there is 10 paid work an equal likelihood for women to either 0 0 1 2 3 4 5 6 7 engage in paid or unpaid work when No. of children they have 0 – 1 children and then there is a shift towards unpaid work where there are two children to care for. While logic would indicate that beyond this, any further children in the household would indicate a reduced participation of women in paid work, the opposite scenario is observed in reality. There is a sharp increase in the number of women engaged in paid work when the number of children increases to three and four and then drops with further increase in the number of children, until it reaches the number of women engaged in unpaid work. The reason for this behaviour can be the fact that we are considering a low-income household and with increasing number of children, the total earned income of the household plus the welfare assistance becomes insufficient to provide for the entire family and hence women are forced to take up paid work to ensure that the family gets some basic necessities. Motivation: Preference In order to gauge the preference of the sample population in terms of the ideal conditions for work, a social welfare function was considered to assess the preference of the concerned women. N The social welfare function F : L( A) → L( A) which aggregates voters' preferences into a single preference order on A was developed and the preference profile, n-tuple (R1x5,...RNx5) was drawn, each preference was 1x5 vector of voter preferences in the available choice orderings was drawn.

13

The preference profile on the preferred type of employment incorporated preference in relation to type of employment (self-employment/government/other), type of work (paid/unpaid), number of days of work (1-2/2-3/3-4/more than 4), frequency of salary payment (daily/weekly/monthly) and distance willing to travel for work/ market (home-based/adjacent to home/within village/up to next village/ further) and the single preference order on A, the preferred form of employment for women in the low-income households. Each woman was given the choice to select any of the options provided under each of the five categories that together constituted the individual’s preference for environment for paid work. The number of preference elements under the preference profile that the individual is free to choose from is 600 (5x2x4x3x5). The actual nominated choices by the women came down to 16 different preference elements and neglecting the irrelevant alternatives (having a total of one or two votes), the remaining three choices were also found to be the top three choices (table 4.1). Table 4.1. Ideal work environment - the top three preferences Type of employment

Type of work

No. of days of work

Frequency of Salary payment

Selfemployment

Paid work

3-4 More than 4

Weekly

Market/ Distance willing to travel for market Home-based Adjacent to home Home-based

Frequency

20 5 5

It must be added though that should the 17 discounted voters have modified their choices in any of the options, the nett social preference profile given above may have been altered, depending on the ranking. However, from the list, we can observe that the options that are slightly diversified is the number of days of work and distance willing to travel for work/market. The voted preference of the women in the sample therefore indicates that if considering paid work, the women would prefer self-employment, which is home-based. This is an important factor when considering employment generation programmes for rural women. Motivation: Attitude A social welfare function of nine possible alternative attitude towards work was presented to the respondents and respondents were requested to select sentence (s) they identified with most. According to their selection, the list was ordered to reflect the prevalent attitudes existing amongst the women towards paid work. The nine possible alternatives presented statements on paid work as an important criterion for men and women on one end and traditional roles of men as income-earners and women as domestic carers at the other end. The responses received have been presented in the attitude chart (table 4.2). Table 4.2. Attitude chart Attitude indicators

Frequency

Sum-total of similar indicators

A woman’s work is to look after her house The male head of household has the responsibility of earning an income for the family

21 16

Both the man and woman have equal responsibility for generating an income for the family

5

14

37

Each man and woman should have the capability to stand on their own feet The aim of work is to generate an income Both the man and women have equal responsibility in undertaking domestic work/care

2 2 1

10

What is observed as the prevalent attitude amongst the women is that they consider traditional gender roles as the system of belief and there is a tendency for women to be home-based. This is in positive correlation to the choice of women to prefer home-based income generating activities over those requiring more mobility. Further, when the women were asked to mention the total number of working members in the household to cross-check with the information they provided earlier on the household composition and occupation, all the women responded that there was only one working member in the household, where an adult male member was present and zero, where absent. This response was irrespective of the fact that more than 50% of the women had mentioned that they either raised poultry or undertook casual labour or self-employment income generating activities. This corresponds to the above observation that the prevalent attitudes amongst the women was that a woman’s responsibility was home care. Home and home-related income generating activities were encompassed under one category of domestic work and hence most women responded that they did not work, while in reality they did.

Conclusions The study therefore makes the following conclusions with respect to the hypothesis on the factors that contribute towards participation of women in paid employment in low income households: Capability: • Age: A sharp increase in women engaged in paid work is seen to occur in the 35 – 44 age group, while a sharp increase in the number of women engaged in domestic care is seen to occur in the 25 – 34 age group. What we can therefore conclude is that age is indeed a positive correlation factor for considering paid employment. • Education: The secondary school level of educational attainment is seen as the current bar of education that determines whether the individual will opt for casual labour (low status) or self-employment (moderate status), when considering paid work. Motivation: • Child care: The findings of this study conclude that women from low-income households, who have more children, are inclined to engage in paid work, than those who have one or two children. Therefore, the number of children is indicative of the tendency of the woman to engage in paid work and that women having three or more children are more inclined to paid work. • Preference for the ideal environment for engaging in paid work: The preference option of the sample indicates that women from low income households consider home-based, selfemployment as the ideal form of employment, with a regular weekly salary for 3-4 days of work. • Attitude towards women and work: An exceptional majority of the sample indicates that women from low income households have adapted to the social conditioning that

15

women’s major role is in the domestic hemisphere and that the man is responsible for generating income for the household. The above factors provide a clear picture of the current capabilities and motivation levels of the women in the sample. For example: A very strong indication that women’s attitudes towards the concept of women responsible for domestic care would increase the tendency for women to refrain from paid work or drop out, with increasing family responsibilities. Also, income generation projects designed such that women have to travel outside their village for the purposes of work is likely to have less participation or higher drop-out rates than having a project within the village. Accordingly, the composite index model for making assessments on women’s participation in paid work in low income households can be described as follows,

LW = (C + M ) / 2 where, C = and, M =

1 ∑ (C xi + C yi ) 2 i =1

1 ∑ (M xi + M yi + M zi ) 3 i =1

The ranking for each of the sub-factors, is valued on a normal distribution curve on a scale of 0 – 1, with 0 indicating tendency towards reduced women’s labour force participation and 1 indicating a higher tendency of women to engage in paid employment.

C x = age (a positively skewed distribution curve for the age group with the peak at age 35)

C y = educational level (a normal distribution curve for the age group with the peak at secondary school level of education) M x = child care ( a negatively skewed normal distribution curve, with the peak at 3 children)

M y = Preference in relation to work environment child care (the mean of the combination of the values indicated by the individual distribution curves for each of the sub-factors – type of employment, market etc.) M z = Attitude towards work (A positive skewed distribution curve with the peak for responses stating that women and men need to equally participate in income generation and domestic care)

Recommendation The study concludes that the proposed composite index model would provide development agencies a relatively clear view of the current situation in the female labour force participation as well as areas that need to be targeted for increasing women’s participation in paid employment. It must be noted that the findings from the model are not recommended to be used as the standard assessment tool for designing income generation programmes for rural women, rather that it be used as model for providing a mapping of the current capability of the individual to engage in paid work under each of these factors and then to compare with the actual potential and what measures need to be undertaken to realize the potential. Considering that one of the findings was that women were inclined to home-based employment and that one of the reason for women’s reluctance to move out of their village for purposes of income generation is security, encouraging

16

women to move in groups to and from the work-place to their own village would be better in empowering women rather than designing a project where home-based employment activities are encouraged. A major limitation of this model is that it does not take into account the range of employment opportunities and the relative increase in income that would be entailed when engaged in the respective form of employment. Also, an important factor, real income of household is omitted from the model as it is difficult to gauge the real incomes of the household, beyond a certain generalization of all households coming under Samurdhi welfare assistance as a low income household, which in itself is not a strong foundation given that the district profile report states that Samurdhi officers had claimed it was difficult to assess the real income and hence there was misuse of the welfare assistance claim. Further, the attitude and preference for engaging in paid employment under the motivation factor is ranked according to the author’s perception of the ideal social condition for women to engage in paid work and therefore is biased. It is recommended that as a next step the model is tested on various low-income household samples from different communities, within the North and East and then compared with the results of the model obtained when comparing with the rest of Sri Lanka. Acknowledgements The author would like to thank Dr. Athula Ranasinghe, for his guidance and support in conducting the study and the writing of the thesis and subsequent paper. The input of the volunteers Sithiradevi, Chandra, Tharshini, Vasanthy and Naheswary under the supervision of Rajani Kandasamy (Working Women’s Development Foundation) who helped me in the collection of primary data for my study and Rahini Kandasamy (CARE Batticaloa) for organizing the preliminary survey meeting is acknowledged. The author also gratefully acknowledges the input of the women of Pudukudiyirruppu, the village committee leader and the Grama Niladhari Officer in the preliminary focus group discussions and the interviews. References Alkire. S. 2005. Briefing Note. Capabilities and Functionings Definition and Justification. Human Development and Capabilities Association. http://fas.harvard.edu/~freedoms/documents/HDCA_Briefing_Concepts.pdf California Institute of Technology. Implications of Arrow’s Impossibility Theorem for Social Choice Methods. 1996. http://alumnus.caltech.edu/~seppley/Arrow's%20Impossibility%20Theorem%20for%20Social%2 0Choice%20Methods.htm#Proof%20of%20Arrow's%20theorem Central Bank of Sri Lanka. 2004. The Consumer Finances and Socio-Economic Survey Report 2003/’04 – Part I. CIA. The World Factbook Sri Lanka. 2006. https://www.cia.gov/cia/publications/factbook/geos/ce.html Department of Census and Statistics. Sri Lanka. 2005. Labour force survey questionnaire. http://www.statistics.gov.lk/samplesurvey/lfs_questionnare.pdf

17

District Planning Secretariat, Batticaloa. 2004. Statistical Handbook. _______. 1999. Resource Profile Batticaloa District. Government of Sri Lanka and UNDP. 2005. Millennium Development Goal Country Report 2005 Sri Lanka. http://www.mdg.lk/mdg_country_report.htm Haahr. M. 1999. True random numbers. http://www.random.org/nform.html (May, 2006) ILO. 2004b. Global Employment Trends for Women 2004 (Geneva), March. http://www.ilo.org/public/english/employment/strat/download/trendsw.pdf Sarvanananthan. M. 2003. An introduction to the conflict time economy of the North and East province of Sri Lanka. International Center of Ethnic Studies Working Paper Series. Colombo. Thiruchandiran. S. 1999. The other victims of war: Emergence of female headed households in Eastern Sri Lanka Women as .Victims: Class and Gender, Social Isolation. Vol II (19-32). WERC. UNDP. 2004. Technical Note: Calculating the human development indices. http://hdr.undp.org/docs/statistics/indices/technote_1.pdf (_______). Sample Size Determination. Sampling and Sample Size. 402-204. Class note handout.

18