Modernization , Aging and Coresidence of Older Persons The Sri Lankan Experience

This paper examines the effects of the modernization on the living arrangements of elderly people in six selected communities representing urban, semi-urban, estate, rural, colonized settlement and fishing villages in Sri Lanka. The paper concludes that the modernization of the economy and society has exacerbated an intergenerational rift leading to an intensification of tensions between elderly people and other family members, despite the fact that the percentage of older people living with their children remains high. Such coresidence or intergenerational living comprises many types of living arrangements, and leads to mixed results for care of the elderly. Many elderly people have developed mechanisms to counteract the negative effects of coresidence: seeking independence during old age, by earning their own income and living alone or living with the spouse, indulging in behaviors such as drinking, spending time outside the home with friends of similar age, or creating their own living space within coresidence.

Sri Lanka has been aging faster than any other nation in South Asia with an annual average increase of 3.3% in the population aged 60 and above between 1981 and 2001 (Siddhisena 2005).Sri Lanka's elderly population of 60 years and older has increased from 11.8% in 2009 to 12.5% in 2013 (United Nations 2009 1 and 2013 2 ).According to the World Bank, 3 the population aged 65 and above in Sri Lanka was 8.41% of the total in 2011.As measured by the Department of Census and Statistics of Sri Lanka, in 2001, 8.9% of the population was aged 60 years and above (Gunasekera 2001), and this had increased to 12.3% in 2012 (Census and Statistics 2012 4 ).Population projections show that, by 2030, the proportion of the population aged 60 years and above is expected to increase to 21.5% (Siddhisena 2005).The population born between the 1950s and 1971-the baby boom in Sri Lanka-has contributed to the present as well as the projected increase in the elderly population.The achievements in economic, health and social development have contributed to higher survival rates among the elderly population in Sri Lanka.
The literature on aging in Sri Lanka that describes its socio-cultural implications often paints a gloomy picture with reference to the care of the elderly, the erosion of traditional homebased family care, inter-generational problems arising from aging (Amarabandu 2004), the lack of social security coverage, the reduced effectiveness of traditional family support systems for old people, and the disempowerment of old people in relation to family and community matters.Because of poverty, the elderly are often forced to work late in their lives (World Bank 2008).Often those negative developments are regarded as consequences of modernization.

Modernization, old age and coresidence
It is generally accepted that the socio-political processes following modernization have fostered a devaluation of older people and that the effect of demographic transition and 'subtle and complicated sociological processes' has been their relative degradation (Cowgill 1974).Among the processes of change, urbanization, mobility and civil wars have affected older people, and have weakened extended family and communal ties.In modern times, families have become smaller and more dispersed.The young are employed in market-based employments, often away from the parental home (World Bank 1994), which can affect residential patterns, especially in relation to coresidence among the elderly.
Although there is no consensus as to the definition of 'modernization', it is often used to refer to the stages of social development which are based upon industrialization and their outcomes such as urbanization, economic development, increased education and the spread of mass communications and literacy (Cowgill 1974;Cowgill and Holmes 1972;Inkeles and Smith 1974;Lerner 1958;Palmore and Manton 1974;Bendix 1963).When dealing with processes of change from a traditional to a modern society, Cowgill's conceptualization, which is a synthesis of many definitions of modernization, seems most appropriate.Cowgill (1974: 127) defines modernization as 'the transformation of a total society from a relatively rural way of life based on animate power, limited technology, relatively undifferentiated institutions, parochial and traditional outlook and values, toward a predominantly urban way of life based on inanimate sources of power, highly developed scientific technology, highly differentiated institutions matched by segmented individual roles, and a cosmopolitan outlook which emphasizes efficiency and progress'.
Like the ideal type of Weber (Antonio, Shils, and Finch 2011), modernization theory derived in Europe and 'abstracted from the process of Western capitalist development' (Bendix 1963) is based on specific cultural experiences, and, if it is to be applied in developing countries, it must be adapted to suit their different cultural circumstances.Although it has been argued that modernization theory reflects 'the Western experience [of] the first historical example of modernization, and therefore can provide a paradigm for the study of the process in non-Western societies' (Bendix 1963), this paper holds the view that developing countries such as Sri Lanka, where the history of development has taken an entirely different trajectory, do not conform to the model of modernization as it developed in the West.Bendix's own description of urbanization in India (Bendix 1963) is an example that illustrates the problems of simply applying this Western model.
Regarding modernization as a social process emanating from socio-economic changes generated by scientific and technological advancement has heuristic advantages, particularly when analyzing social transformations taking place in developing countries (Inkeles 1969 and2013).It can be used to explain social marginalization, if modernization is treated as the main driver of change in a society.Such an approach helps explain why certain groups of people fall behind the main currents of social change prompted by the processes of modernization.If the theory of modernization is situated in its proper context in a particular society -in communities in Sri Lanka in this case -it can be used to examine development as a historical process emanating from the introduction of modernization, and also to examine the factors and forces or 'pressing needs of the situation' (Bernstein 1971) that led to the introduction of modernization.Thus, critiquing modernization theory, in this essay we try to explain the social outcomes of modernization among the elderly in Sri Lanka focusing on residential patterns and the quality of life in six different communities representing cultural ecological zones of the country.Cowgill and Holmes (1972), in their cross-cultural study of modernization and in subsequent studies in 1974 and 1986, postulate that, as societies modernize, the status of the aged declines.We disagree because this hypothesis is based on the romanticized idea that old people in pre-industrial or rural societies enjoyed high status.Bengtson et al (1975: 689) also critiqued the modernization and aging theory of Cowgill et al as suffering 'from a romanticized or naive portrayal of eldership in pre-industrial societies.'Most ethnographic studies on South Asia (for example, Goldstein and Beall 1981;Harlan 1968;Ranade 1982;Rajan and Kumar 2003) support the idea that old people in rural communities work hard physically, and do not depend on their progeny for support until they become frail.It is their ability to work hard, in their farms and households, which allows them to retain control of property and resources and their families.Hard physical labor among the elderly in pre-industrial societies, however, is not consonant with their prestige in society as postulated by modernization theorists.
Linda Martin (1990), in her review of the status of South Asia's elderly, states that there is no strong evidence that the status of the elderly in South Asia has declined nor that it was uniformly high in the past.Acceptance of Cowgill's assumption that the elderly in rural areas and in pre-industrial societies enjoy high status tends to overlook the difficulties and abuses that the elderly experience in their coresidential arrangements, extended families, neighborhoods and communities.Studies done in the 1940s do not confirm that the elderly in pre-industrialized societies enjoyed high status in extended family settings or coresidence (Simmons 1945).Goldstein and Beall (1981: 50) in their study of Sherpas in Nepal -a preindustrial society -found that those aged over 70 were physically fit, carried out heavy agricultural work and household chores.All but one owned their land and lived independently in the village.Although Sherpa culture confers high status on the elderly, the authors observe that 'despite their high level of economically productive activity, their relative fitness, and their ownership of agricultural land, many elderly Helambu Sherpas overtly expressed unhappiness with their lot.Several spontaneously volunteered that they wished they were dead, and others commented that their children had abandoned them and that the young in general did not care about the elderly.Almost all of those living alone drank home-brewed liquor daily and a few consumed three or more cups per day' (Goldstein and Beall 1981: 51).Harlan (1968: 475) notes that the elderly occupied precarious positions in the Indian villages which he studied and which he characterized as a 'pre-industrial agricultural community, little affected by urbanization '. Ranade (1982: 84) states that work participation rates of people older than 65 in India in 1982 were as high as 77% in rural areas and 55% in urban areas.He further observes that, in rural areas, most of the working elderly were engaged in heavy manual work connected with agriculture, and that they worked until their health deteriorated to such an extent that physical activity was not possible.
Coresidence of elderly people with their grownup and married children is seen as a preferred living arrangement for the elderly in South Asia (ESCAP 2007;Martin 1990).Uhlenberg (1996), in a study based in rural settings in Sri Lanka, found that 75% of the elderly coresided with spouses or children.Nugegoda and Balasuriya (1995), in their semi-urban area study, found that 85% lived in coresidence with children, while an urban and rural based study by Andrews and Hennink (1992) and the Marga Institute (1998) shows that the figure was above 83 per cent.Siddhisena (2005) observes that 'there are no significant rural urban differentials of this pattern' of coresidence, and that 'there is no strong preference for living with a son or daughter in later life'.However, findings of another study (Siddhisena and DeGraffe 2009) state, in conformity with Uhlenberg (1996) and Nugegoda and Balasuriya (1995) that 'Traditionally, and still the most common pattern by far, elderly parents live with one of their children, often the youngest son who typically inherits the family home and is expected to care for his parents in return'.It is hypothesized that 'coresidence rather than the kinship tie is more important in determining the pattern of caregiving' (Tennstedt et al 1993).The literature on coresidence in Sri Lanka does not examine how the educational achievements of children correlate with coresidence.In the US, however, it was found that children with college degrees are less likely to live in coresidence with or near their mothers (Compton and Pollak 2009).
Based on demographic data showing a growing percentage of the elderly among the population, a low rate of population growth, and increasing life expectancy, it has been postulated (Perera 1999, Siddhisena and Ratnayake 1998and Uhlenberg 1996) that the change from the extended family system to nuclear families in Sri Lanka has given rise to 'problems in providing better care to aged parents'.These issues are exacerbated by factors such as rapid urbanization including the shifting of residence from rural to urban locations, and women getting employment locally and abroad, all of which make it difficult for nuclear families to provide care for older people.
The early stages of modernization in Sri Lanka began to appear when the Europeans captured Sri Lanka (then Ceylon): they established new administrative and revenue systems, and introduced Christianity and new economic activities, plantations, and urban centers.These programs significantly changed the mode of production and land tenure, particularly the systems of inheritance and ownership of land.For the first time in the country's history, under Roman Dutch Law, women were regarded as legitimate heirs to parental property resulting in new authority structures and residential patterns.The new law made it possible to transfer and sell land, which had hitherto been considered family property to be passed down from one generation to the other.The introduction of waged labor, the creation of new employment outside the original villages and new forms of labor organization, technological innovations, road links with urban centers and cash crops converted subsistence agriculture into a market-oriented enterprise.These changes have had a tremendous impact on the elderly population, particularly in their residential patterns, living arrangements and work and income generation activities.An appreciation of the impact of modernization on residential patterns of the elderly requires a closer look at how the elderly live in their communities.The study employed an anthropological methodology aimed at capturing the effects of modernization on coresidential arrangements of the elderly in Sri Lanka.

Methodology
In this study, we employed a qualitative -quantitative mix approach for data collection.The data were collected in early 2013 over a period of six months.The communities chosen for the study were taken from six cultural ecological zones in Sri Lanka: urban, semi-urban, traditional rural, estates, colonized schemes, 5 and fishing communities, and the results were weighted by population percentages obtained from 2011 census data for urban (18.3%), rural (77.3% 6 ) and estate (4.4%).
The selected areas were located in the Districts of Colombo, Kalutara, Rathnapura and Matara.One Village Headman Division or Grama Niladhari (GN) Division from each of the six cultural ecological zones was chosen for the study.The GN Divisions selected were Jambugasmulla in Colombo District, Alubomulla in Kalutara District, Kumbalgama, Rassagala, Budugala and Welipothayaaya (these two villages form one unit) in Rathnapura District, and Mirissa in Matara District.In each GN Division, 80 families with at least one elderly person were randomly selected (using a random number table) for the study, from a sampling frame of elderly persons over 60 years of age that were recorded with the respective Grama Niladhari.In addition to the sample of 480 households, a further 50 households were randomly chosen as potential replacements in case any respondents in the original sample were not available for interview at the time of the survey.As the village originally chosen to represent the rural community (Kumbalgama) had fewer than 80 families with elderly persons, the adjoining GN Division (Muththetuwegama) was also taken into account in selecting the rural community sample.
During the initial phase, 60 open-ended interviews were carried out in the six communities selected for the study with elderly people and their family members to enable the authors to get a thorough understanding of the socio-cultural aspects of old age as they apply in different communities in Sri Lanka.The older people for open-ended interviews were selected from the list of households maintained by the GN.However, in some cases, the randomly selected person in the household was replaced by another willing respondent and therefore, we ended up having more male than female interviewees (38 compared with 22).On many occasions early in the interview process, we found that it was difficult to isolate the interviewee from other members of the household.However, when we made appointments for a subsequent interview, the interviewee agreed to meet us individually, often outside the house under the shade of a tree or in an outhouse in the same garden.In the fishing village, most of such interviews were held on the beach.
Based on the qualitative data collected during the first phase of the study, specific topics and variables were identified for detailed examination using semi-structured interviews with a random sample of 480 persons or 19% of the total population aged 60 and over across the selected locations.Twenty-three households initially selected for the survey were replaced later for variety of reasons (for example, because they had moved out of the village or the elderly people were not present at the time the interviewers visited).Of the 23 households that were replaced from the initial list, 19 were located in the rural, estate and colonized communities.The data were computer analyzed using the SPSS program, while qualitative data were translated into English and manually analyzed for this paper.
The survey instrument used in this study comprised 12 major domains with variables identified under each domain for the collection of data.The major domains included in the survey instrument were household demography, physical arrangements and facilities, living arrangements, work and occupation, education, relationships with family and others, intergenerational aspects, attitudes towards life, time use and leisure, and household chores carried out by the elderly.This article concentrates on those variables from the survey data that pertain to coresidence and living arrangements.Nine variables on attitudes and behaviors of older people that were identified as crucial to elderly life in the qualitative interviews were enumerated in the survey using a 5-point Likert scale and analyzed using the reliability analysis program available in SPSS.The nine variables were used to create an anxiety scale using the multiple response analysis method in SPSS.

The traditional system of residence, inheritance and coresidence
In order to examine the changes that have taken place in residential patterns of the elderly in Sri Lanka, it is imperative to know about the traditional arrangements as documented in ethnographic research.In Sri Lanka, traditional residential arrangements involving living in one's own patrilocal residence (gedera) or with family (paula) are widely practiced among both the Sinhalese and Tamil populations.Paula comprises father, mother, children, and married children with their spouses and grandchildren living in the gedera or family house.Occasionally, there are close relatives such as sisters and brothers of the mother and the father also sharing the family house.Traditionally, a household or gedera 'is used …to refer to a group of people whose food is owned in common and who share a single cooking fireplace' (Robinson 1968).
Traditional coresidence arrangements typically took the form of horizontally extended families known as compound groups.Robinson (1968), studying a traditional village in Kandyan provinces in the late 1960s, shows that the traditional practice of village exogamy led to the creation of 'compounds' where the married sons took up residence in separate but adjacent houses on the same parental property.Robinson's work in Morapitiya, a Kandyan highland village, and Leach's research in Puleliya in North-Central province (Leach 1961) have documented such practices.Compounds of a somewhat similar nature comprising households established by married sons and sometimes daughters on parental property were found in most of the traditional villages in the low country as well.Robinson (1968) describes the practice of traditional coresidence of compound groups saying that 'People [become] members of a compound group if they live on a particular named piece of land.The patterns of kinship and inheritance of land determine an individual's right to the land and hence to compound group membership'.
There were two forms of marriage with different implications in relation to residence and inheritance.Many daughters tended to live in their husband's village, a practice known as diga marriage.However, other married daughters stayed in their own village, -a traditional practice (binna marriage) that was used to accommodate the married daughter and her children into the patrilocal family, especially when there were no sons in the family to inherit the parental property.In the latter arrangements, which were considered derogatory, the children took their mother's family name.Since the introduction of Roman Dutch law allowing females to inherit property, this practice of daughters living in the parental property has become an accepted practice, and did not require children born of such marriages to take their mother's family name.Living in compound groups is still regarded as the norm in rural Sri Lanka where married children tend live in newly constructed houses adjacent to their parental house or within the parental property.
It is against this background of traditional patterns of inheritance and residence that we examine the coresidence of elderly people as it is practiced now in the six different communities that represent various cultural ecological zones of Sri Lanka.

Findings
Of the 480 elderly people surveyed, 50.2% were female while 49.8% were male.The average age of the people in the sample was 73.7 years with a standard deviation (SD) of 7.3 years.The highest mean ages were found in the fishing community (75.9 with 7.8 SD) and the traditional rural community (75.5 and 7.9 SD), while the urban community had the lowest (71.5 with 6.5 SD).The youngest respondent was 60 and the oldest was 90.The distribution by gender did not differ significantly across the communities.The young old (60-64 years) comprised 8.8% of the sample, while old (65-74) and old-old (75+) comprised 47.8% and 43.5% of the sample respectively.The percentage distribution of the three age groups did not differ significantly across the communities that were studied.
In this article, we report only a selection of the information collected, particularly examining the effects of modernization on the residential arrangements of the older people in the locations we studied.A decrease in coresidence?
A breakdown of our data using the sector-wide weighted average method shows that the percentage of elderly people living in their original homes with their spouse or children has not changed very much from 2003 (64.4% according to Siddhisena 2005) to 2014 (61.3% in the present study).This is contrary to the World Bank (2008) suggestion that 'as income and education increase, coresidence with children may decline in Sri Lanka'.
When analyzing the six communities as separate units in Table 2, it can be seen that a large percentage of elderly people lived in coresidence in the colonized (85%) estate (82%), urban (82%), and fishing communities (85%).Elderly people living in coresidence with children in Semi-urban and the rural community had 71% and 74% respectively.However, if proximity were used as a proxy for coresidence in rural areas, the percentage of coresidence would have increased to 90% as many children of older parents live in separate houses close by, sometimes built on land owned by the family.
The emergent popularity of coresidence in urban areas (Nugegoda and Balasuriya 1995;Andrews and Hennink 1992;Marga Institute 1998; and the present study) is shown in the qualitative data.An elderly person from an urban community explained that, after his son was selected for a new job in the university, they sold their village property and bought a house in Jambugasmulla, because his son did not have sufficient savings to purchase a house in the urban community.Another person said 'we came from Kegalle in 2001 and now we live here with my son in Jambugasmulla'; a retired bank officer said 'we were in Kamburugamuwa in Matara district, when I got the transfer to Colombo, which was the last phase of my job; we decided to buy a house in this area…my son also lives here with his wife and child.'Those who were living in their original family homes in the urban and semi-urban communities stated that their children continued to stay in the parental house after marriage, which is another reason for the high percentage of coresidence in urban and semi-urban communities.
…when my son finished university, he did not have a proper job … he conducted tuition classes for students doing the GCE AL examination 7 in a room at the temple.This class became popular, and my son wanted to build a classroom outside the temple which later became a larger building.
My son continues to do his tuition classes and live in this house after marriage… (Man, aged 78, in urban community).
…. Buying land in this area is very expensive… therefore I decided to build an extension to the original house of my parents… I got a bank loan and built my annex and we now live here... (Man, aged 45, son of an old person in urban community).
Most parents bequeath their property, preferably to the youngest son, keeping a life interest in the property, in the expectation that the child in question will live with them and look after them during their old age.Often, parents who retain a life interest in the property continue to be legal custodians of it.Some parents retain full ownership of their property until very late in life, or die intestate, which sometimes makes coresidence a necessary requirement for the designated child (usually the youngest son) to inherit the property.Usually in such situations, the house and property of the parents is given to the son or daughter who looked after them in their old age, and therefore, caring for old parents in such situations can be materialistically driven.
Traditional villages are different from the urban communities in terms of land availability and coresidence.When they get married, sons stay in their parental house for a year or two and then construct their own house close to the parental house on the same plot of land or nearby where they commence their new life.The need for coresidence for parents does not arise in such circumstances, as family members frequently visit each other -a reason for the lower percentage of coresidence of elderly people in traditional villages.This practice can be seen as a continuation of the traditional ideology of community living arrangement known as 'compounds' to suit the present circumstances.

Heterogeneity of living arrangements
The families of coresidence typically take the form of beanpole families (Bengtson, Rosenthal, & Burton, 1990), vertically extended families with multiple generations but few people in each generation.The families we came across in the sample communities where old people lived comprised, on average, three generations, i.e. the old parent or parents, the son or daughter with his/her spouse, and children.The beanpole family in the Sri Lankan context is different from the traditional extended family because there is no horizontal expansion of the family with many members of the same generation living in the same household or on the same land in different houses and sharing family resources.
Our findings show that 80% of the elderly people in the sample reside in verticalized families with one of their sons or daughters, and grandchildren (Table 2).Our findings reaffirm what was found by Martin (1990) who states that 'family being the traditional social institution for care of the elderly persons' allows older parents to 'co-reside … in rural and urban areas' The family structures that we see in our sample are far from being nuclear families or extended families in the traditional sense, but a complex institution with different characteristics.They are social networks created by families to provide for the various requirements of the elderly as well as of their sons and daughters.The families include a vast range of structures such as single-person households where only old father or old mother lives alone, the old couple live by themselves (neolocal), living in the original family home or other house with one of the sons and his wife and children (filiolocal), living in their own home or other house with the daughter and her husband and children (filialocal), living in the original house or rented house with unmarried sons and daughters (bilocal).
The experiences of elderly persons in filiolocal settings mostly reported hostile relationships with family members in their households.An old woman, aged 82, in the semiurban community in filiolocal residence said that My daughter-in-law tries to impose restrictions and rules on me… I don't like that…This is my house and she must understand that she is here because she is married to my son… as long as I am here, I should be in control of things in this household… Of the 48% of our interviewees who live in filiolocal residence, most report authorityrelated issues at the household level.The issues have risen as the daughters-in law took over the management of the house, which was formerly run by the old woman.
Most women who are married and live with their husbands and parents-in-laws are employed and contribute to the family budget and household management.Their work might be in garment factories, government departments, or private sector institutions, while some are school teachers.As such women are busy with their employment and are out during the day, the mother-in-law is entrusted with household tasks such as cooking, washing, and childcare, and these are seen as tasks imposed on the older people by the daughter-in-law.The modernization process has thus created a renewed tension between older parents and the younger generation which correspond to mother-in-law and daughter-in-law conflict in traditional society.Some elderly persons in the (c) to (g) rows are living with their spouse as well as children.
In filialocal residence, one old woman aged 74 stated that I have no encumbrances such as home management etc. as they are all done by my daughter and son-in-law.I enjoy the company of my grandchildren… I look after them when my daughter is away at work.I have the freedom to do whatever I wish in this house although it is not my house.I observe religious meditation on full moon days, and frequently visit the nearby temple ….
About 21% of the elderly people we interviewed live in filialocal residence with their daughters and sons-in-law and their children.The tensions between mothers and the next generation were seldom reported in filialocal residences where the elderly lived with the daughter and son-in law where the ownership issue of property and authority issues did not arise.

'Kin in contact' and coresidence
The traditional linkages within extended families are now being performed through various newer forms of communications and relationships.Modern communication methods such as cell phones and internet-based social media programs have made it easy to link family members even when they live in different and distant locations.The extended networks of relationships or 'kin in contact' sometimes include distant relatives of the older persons who provide support and talk to them over the phone to inquire about their health situations and requirements etc.The survey data show that a large percentage of middle class elderly persons use telephone and internet-based social media facilities to contact their children and other relatives who live in distant places, and so communicate their grievances, seek help and complain about their problems.One woman in the semi-urban area said that when she runs short of money, instead of asking her son with whom he lives, she contacts her daughter who is working in the Middle East to send her money through a wire transfer.Another woman from the fishing community said …. when I have problems … when my daughter-in law scolds me, I do not go and fight with her… instead, I call my daughter who lives in Colombo and tell her about it and she calls my daughterin-law and settles the matter…when I tell my daughter-in-law that my daughter in Colombo is going to visit us (she usually calls me before she sets off) my daughter-in-law takes measures to wash the toilet, clean my room and make the home tidy… This is a good tactic for me to control my daughter-in-law… Another elderly woman from the semi-urban community said My children live in different places.One of them is in Korea.They all connect with me through the internet.They call me in the evening or in the morning at their convenience … they inquire about my health, what I've been eating, and so forth… if I need money, then they send it to my bank account which I can operate with a card.… I don't feel as though they are not nearby or in my house….They are always there for me to contact if I need to do so… my concern however is that I cannot feel and embrace my grandchildren… In the circumstances, physical distance is not crucial in distinguishing present family relationships from those in, for example, extended families.In the new era of communication technology, an older person has the ability to create his/her own network of relationships, which determines the nature and type of the family structure or 'kin in contact'.This, however, is an option available mostly to middle-class families; most rural families do not enjoy the benefits of new technology to improve their support systems.Some elderly people nevertheless felt that being together in one family or in close proximity is the best choice, as it allows them to see each other, and to personally attend to individual requirements.

Rural-Urban difference in coresidence
The percentage distribution of coresidence found in our analysis revealed a significant difference between rural and urban communities.Higher levels of education and smaller family sizes are important characteristics of elderly persons in urban and semi-urban areas who live in coresidence, as shown in Tables 3 and 4. Of the 19 graduates in the total sample of older people, 18 were found in the urban or semi-urban communities.Of those who had smaller families with one or two children, 67% were found in the urban or semi-urban communities.In urban and semi-urban communities, education level and number of children are the most robust predictors of coresidence.As we have not recorded the educational achievements of the children of the elderly people whom we have interviewed, it is not possible to comment on how children's education affects coresidence.Due to their close proximity to work and the exorbitant prices of land and houses, after marriage, the children of urban parents in our sample often make arrangements to share residence with their parents. 8Interview data show that elderly people in urban and semi-urban areas often visit libraries, meet friends, attend societies of which they are members, enjoy window-shopping in the recently established shopping malls in the city, occasionally go to the cinema or the theater, and attend the public gatherings and political meetings which are widely available in the city.Some old people in the urban areas are actively engaged in community work or religious societies, and participate in social welfare activities.The elderly people in rural communities (traditional, estate, colonized community and fishing community taken together) are involved in welfare and religious activities more frequently compared to those in the urban and semi-urban communities.Elderly people in filiolocal coresidence in urban and semi urban communities report relationship issues with inlaws living in the households more often than those in the rural setting.

Work and coresidence
Elderly people in traditional villages continue to work in paddy fields, cultivate vegetables and keep themselves physically active.Having their children resident nearby provides the labor required during the cultivation and harvesting seasons.On many occasions, we have observed that, although the sons and parents live in different houses, they cultivate the same paddy land that the family inherited from their forefathers and jointly own.52% of the elderly people (69% of men and 35% of women) in the sample were engaged in income-earning activities, as shown in Table 6.Most of the women who work do so as agricultural laborers in their own paddy and vegetable farms, while the men work as paddy farmers and share croppers or are engaged in wage labor and fishing.Most of the elderly people who are physically able to work continue to live in their own homes, even if their children living in urban areas have asked them to move in with them.Some older people have opted to live alone in order to avoid the problems that may occur when living in coresidence with children.and unable to look after my property, I will rent it out.I hope that the two of us will be able to look after ourselves, without troubling our children.
Appuhami, an energetic elderly person, aged about 80, is able to attend to his personal needs and care, and lives alone in his 10-room house in his village.He owns a large piece of land in the village.His daughter and son are both married and live in the urban area of Badulla.Appuhami lost his wife several years ago.Being a widower at 80 and living alone in the house, Appuhami now feels somewhat lonely.However, he doesn't like to stay with his son and daughter-in-law in their urban house, and doesn't want to live in a home for the elderly, because he regards such facilities as being for the poor and the destitute.His wish is that his son should come and live with him in his house, which is not possible due to the son's work.He thinks the best option for him would be for his son to visit him at least once a month and talk to him, show respect, and address him as 'appachchi' ('honored father').
If the husband is alive, very often, an old couple stays together in their own house as long as they can work and support each other.If the wife is dead, the widower suffers badly, as he is not accustomed to doing household chores or babysitting in the households of his sons and daughters.In the new domain of relationships with sons and daughters, the role played by the grandparents in urban and semi-urban communities has become a supportive one, rather than an authoritative or a production-oriented one.If the mother is a widow, she often stays in the household of a married son or daughter and looks after the grandchildren, prepares food, and attends to household chores such as washing clothes in the absence of the children who are out at work during the day.

Coresidence: dependency or independence?
Due to modern technology, in communities where paddy agriculture and fisheries are practiced, traditional knowledge has become redundant.The exploitation of fishing resources with new technology attracts young, physically fit and able fishermen, while the traditional fishermen who are now old and physically weak are marginalized.Some traditional old fishermen have become absentee owners of fishing boats which helps them economically.However, the majority of old fishermen, who did not have the money or political power to own boats, have become dependent on their sons, which they dislike.They continue to live in their own houses with married children, but they work hard to make themselves independent of their sons.Most of such old fishermen are engaged in inshore fishing, or work as night watchmen in boats, or work mending nets and repairing boats and engines.This pays for their daily expenses such as for tobacco and liquor.Although the majority of old fishermen live with their sons in the original house, or in the houses owned by sons or sons-in-laws, they lead a life of discontent.
… I meet all the needs for my father… I have told him not to go fishing or do any physical work... but he never listens to us… he comes home late for dinner and sometimes for lunch… he never uses his room to sleep… instead he sleeps on the verandah … every day he drinks … (Fisherman son of an old fisherman, Mirissa) Drinking toddy and smoking are habits of most elderly men in the fishing villages, as well as in the colonies and estate communities.After drinking, the old fishermen spend time on the beach until late in the evening.One old fisherman told me that he had owned three large fishing canoes and a boat, but now he cannot go deep-sea fishing due to his poor physical strength.However, he uses his small canoe (küda oruwa) and fishes in the shallow seas.The money he gets from selling those fish is used for tobacco and liquor.
We are nowhere (nannaththara vela)… we cannot go fishing in trawlers (tanki boattu)… they [young trawler fishermen] think we are physically weak… we are not important for the young people….
We are like pennies… (api sathe gaanata vatila)… I don't like to depend (yapenna) on my children for my needs… I somehow earn my money and I look after my needs… (Elderly fisherman, aged 71) Elderly people in agricultural colonies, traditional villages and tea plantations hold similar views.The majority of older people in such rural communities continue to work into their old age and earn money for their living.
The survey data show that 52 % of older people in the sample communities are involved in income generation activities, many of which demand physical labor.The research shows that physically demanding work among elderly people was more prevalent in agricultural, fishing and colonized communities.

Coresidence, preference for sons and care during old age
Table 7 shows that 59% of the elderly people in the sample villages live with their sons, while about 32% live with their daughters.However, a higher percentage of elderly women (39%) live with daughters in coresidence, compared with only 26% of older men who live in coresidence with their daughters.Of the 101 widows in our sample, 60 live with married daughters and 13 live with married sons while others live on their own and some with unmarried children.Of the 11% of all women who live with both son and daughter [row (g)], some live with their unmarried children.Some elderly persons in the (c) to (g) rows are living with their spouse as well as children.
Widows are found in (a) and (c) to (h) rows.
The finding of this study that elderly people more often live with sons is consonant with the cultural practice of bequeathing the parental house to the youngest son of the family.With regards provision of food, monetary assistance, and emotional support to elderly people living in coresidence, around 30% of the elderly surveyed denied that they received such support from their children.The breakdown of elderly care by gender shows that there is a significant difference between male and female in receiving care from children.Compared to male older people, female elderly are provided with more support and care by their children.Abuses and poor respect by children towards the elderly were mostly reported by the female elderly.
Often coresidence is seen in aging research as the best form of residential arrangement or 'aging in place' 11 for the elderly as it is assumed that it ensures the care and wellbeing of the elderly people.In Sri Lankan communities, coresidence cannot be seen as a substitute for paid nursing care, simply because paid nursing care in not a widely available option for the care of the elderly. 12As our data shows, coresidence is often a kinship and family based arrangement negotiated between parents and their children.Our study further shows that coresidence does not guarantee good care for the elderly (Table 8), and the presence of coresidence therefore has not necessarily resulted in high quality elderly care in Sri Lanka.
Reciprocity in care between parents and children is a widely established cultural norm among the Sinhalese, and this is an important factor leading to the practice of coresidence.There are many adages, sayings and proverbs among the Sinhalese that reinforce these behavioral obligations.Children who fail to look after their elderly parents are socially condemned in the villages, more so than in urban areas.
Our study reveals that 'reciprocity' (cultural agreement) and residing in one household (physical arrangement) are fast becoming two separate concepts and are not necessarily linked as they used to be in the past.In one semi-urban family, a widow aged 87 lives with her son, daughter-in-law and granddaughter, but, in practice, she lives by herself in a separate room in the house, cooks for herself in a shared kitchen, and manages her own expenditure.This shows that coresidence does not necessarily mean that children spend time with their parents or take care of them.
Living in coresidence often gives rise to additional work for elderly people as they have to attend to household chores and the needs of their grandchildren.As one elderly widow from the urban community explained … My daughter and son-in-law are working.They are not at home during the day.Their two children spend time with me.I take their daughter to her Montessori class and I keep the young boy on my lap until the class is over.Then we come home and I feed the two children.I prepare lunch before we leave for the Montessori class, as it is easier for me.You know children -they want to eat as soon as they come home… (Widow, aged 68) One elderly woman in the fishing community said that I love these children so much… I take full care of them when my daughter, a teacher, goes to school in the morning … my son-in-law leaves home early in the morning to go fishing in the boat (tanki boattuwa), and comes home sometimes after several days.These two boys are very naughty … they never sleep in the afternoon and never allow me to have a nap… my only solace is sleeping at night after 10pm after the boys have gone to bed….(Widow, aged 71) An old woman living with her son and daughter-in-law in a semi-urban area said that My daughter-in-law is not good with me… she never shares her cooked food with me… I cook my food when they are not using the kitchen… when I don't have vegetables, my daughter who lives elsewhere helps me to buy vegetables.When coconuts are plucked from trees in this garden, they keep them for their own consumption and don't give me any… after all, this is my house although my son owns it now after my husband's death… my daughter-in-law turns off the main tap during the day because she doesn't like me washing my clothes and bathing… she is more concerned with the water bill than my cleanliness and health… when my other children call me, she disconnects the line before I come to the phone…' (Widow, aged 90).
In one family in a rural community, the elderly father said that he does not speak with his daughter-in-law and never eats food that she has cooked because she had plotted and married his son without the consent of his parents.He said 'I never allow her [the daughter-in law's] parents or kinsfolk to visit my house… they have cheated us…' Our research reveals that the generation gap and cultural differences can have an impact on the expected reciprocity.Old parents' habit of getting up in the middle of the night and early in the morning, their lack of attention to the clean use of toilets, their adherence to traditional ways of eating and preferences in relation to food and clothing, their use of tobacco and chewing betel leaves and spitting are some of the aspects of their behavior that were detested by the younger members of their households.In urban and semi-urban communities where parents live in the home of their son or daughter, these problems are more frequent than in situations where they live in their own homes.Sons or daughters, migrating to urban areas and aspiring to middle-class lifestyles, expect their elderly parents also to live by such mores.The imposition of behavioral changes on parents in such instances creates misunderstandings and problems for the older people living with their children.
A high-ranking government officer took up residence in an urban area after he was appointed to a new position, and brought his elderly father to live with him in his new house.The researchers met the father one early morning walking with difficulty towards the bus station and asked him where he was going.The father had come to live with his son, and his daughterin-law had brought him to their house a week earlier, but he did not like living with them as he felt it was like a prison (hirakuduwak).They were trying to control his behavior: they did not allow him to walk around in the neighborhood, or to wake up early in the morning; they did not prepare his morning cup of tea until quite late, and did not allow him to chew betel, and so on with many more complaints.Therefore, he had decided to go back to his own house in the village and live by himself, and this was why he had left his son's house in secret early that morning.
One elderly man living in a semi-urban area was critical of the way that young people behave towards the elderly and said: Some young people oppose what we say.People of the new generation keep their distance, and they have attitude 'who cares about the old farts'.They are arrogant and neglect us.... It's true that the new generation does not respect the old.Now see how they pay their respects: just bend, or touch the feet in the 'Indian' style.It is not the way elders should be bowed to in our culture.This bad behavior has spread through television dramas (Man, aged 68) A middle-aged person who followed us on our way back home after interviewing an old couple in the semi-urban community said that they were badly treated by their children.Their son picks all the coconuts and sells them to the nearby boutique, never allowing his mother to use them.The mother cooks food for the whole family at her very old age and at her own expense, and her daughter-in-law frequently scolds her for not keeping the house tidy.The old parents were not allowed to switch on their TV or radio on full moon days (poya days) to listen to Buddhist sermons as they would very much have liked to, on the grounds that it disturbed the school work of their granddaughter who was preparing for the GCE AL examination.The old couple did not disclose any of these problems to us in our interview, although they looked somewhat depressed and miserable.
An urban elderly widow who lives alone in her own house expressed her view that older people have their own ways of thinking and doing things, which are different from those of the new generation.She said: I do not expect my children to be around me 24 hours a day.I have to entertain myself.I am responsible for my own happiness.I have friends on Facebook.It's so much fun!My niece puts pictures on Facebook.I do not depend much on them [her children] for my happiness.I am not bitter.I know a lot of old people … they complain about their children.If there is a roof over my head …that's enough for me (Woman,aged 78) Analysis of Likert scale data on attitudes, behaviors and life satisfaction of older people shows that 85.0% of elderly women and 89.7% of elderly men were highly concerned with (1) feeling lonely, (2) loss of desires (including the desire to eat tasty food, wear attractive clothing, and sexual desires), (3) feeling bored, (4) loss of appetite, (5) anxiety about death, (6) getting respect from young people, (7) feeling happy and content most of the time, (8) feeling that other people live a better life than me, and (9) the presence of orthopedic problems.Based on this data we hypothesize that the prevalence of depression among the elderly in coresidence is much higher than what is reported by Malhotra et al in 2010, who found that the prevalence of depressive symptoms among the elderly in Sri Lanka was 27.8% overall (24.0% for men and 30.8% for women), which is considered a high rate for an Asian population.We observed that levels of anxiety increased with the duration of stay in coresidence

Conclusions
Modernization theories postulate the spread of nuclear families, a reduction in the proportion of adults living with their parents, and the disassembling of extended families.However, contrary to this, in Sri Lanka we have witnessed a large percentage of the elderly coresiding with their adult married children with very little change over the past decade.The proportion of elderly people living in their original house with their children and/or spouse has not significantly changed from 64% in 2003 (Siddhisena 2005) to 61% in 2014 (current study).Weighted averages show that in rural areas (all the communities except the urban and semiurban and estate communities), coresidence of the elderly was 82%.In the urban and semi-urban communities, it was 77%.
The proportion of living alone or living with the spouse has increased from 11% (Siddhisena 2005) to 20% in the present study, which can be seen as a recent development particularly due to the difficulties in coresidence experienced by the elderly.Those who have opted to live alone lead an independent life, but at the risk of poorer care, support and security.As shown in this article, married children of older parents in urban communities live with their parents in parental homes due to their inability to purchase or build a house of their own.Such forced coresidence has given rise to intergenerational problems and relationship issues making living in coresidence problematic for the older people.The older people also find that there are less elderly homes in Sri Lanka with facilities suitable for them to live an independent lifeanother reason for coerced living in coresidence.The changes in coresidential arrangements and many problems in such living arrangements including exacerbation of tension between the daughter-in-law and mother-in-law that existed in traditional times, and the increase in the proportion of living alone witnessed in this study can be seen as results of the limited nature of modernization in Sri Lanka particularly in relation to elderly care and management and provision of alternative living arrangements.
Conceptualizing coresidence as a means for the care of the elderly is somewhat misleading in Sri Lankan context.As shown in our research, coresidence occurs for a variety of reasons.Often old parents are used to take care of houses and grandchildren while their married sons and daughters are out during the day.Living in coresidence also assures the property rights of sons and daughters: elderly parents transfer their property into the name of their children, while retaining a life interest in the property, and this forces them to live with their children in their old age.The elderly enjoy little freedom to decide on their residential arrangements as few options are available.Many elderly agree to live in coresidence because of the lack of any alternative, although it does not help them to fulfil their expectations for life during old age.
Coresidence is often the only way that an elderly person can live out their old age and it is the 'aging in place' in Sri Lanka.A number of social forces -the low availability and high price of land, the need for both husband and wife to go out to work, and the need to look after grandchildren in the absence of their parents -have all contributed towards forced coresidence between elderly people and their descendants.This coerced living in coresidence has given rise to many different arrangements (such as living alone or with the spouse, cooking separately and living in a room of the main house) that have been developed by the elderly and their children in order to overcome the tensions and relationship issues that have emerged as a consequence.
An old person, aged 69, from the tea growing community said that I have three acres of tea and two amunas 9 of paddy in my village.I get about 25,000 rupees a month from my tea estate.My wife and I both work on the tea estate as well as in the paddy field.The paddy gives us enough rice for the two of us as well as for my children, who are married and live in Kandy [117 km away].My day-to-day expenses are covered by the money I get from the tea estate.I sell all my raw tea leaves to the nearby tea factory at Rs. 40 a kilo.My children, two daughters and a son are employed in Kandy and are married and live there... when I am very old

Table 1 :
Older people and gender distribution in the sample locations

Table 2 :
Living arrangements of older persons (60 years and over) in sample communities

Table 3 :
Level of education of elderly persons in urban / semi-urban communities vs the other

Table 5 :
Average agreement with the statement 'do you

Table 6 :
Current income earning activity of elderly persons by gender

Table 7 :
Type of residence of older people by gender

Table 8 :
The provision by gender of elderly care to those who live in coresidence