Gender and well-being of older persons in Cambodia - Population

Gender and Well-Being of Older Persons in Cambodia

John Knodel

Population Studies Center University of Michigan

Zachary Zimmer

Institute of Public and International Affairs University of Utah

Population Studies Center Research Report 09-665

January 2009

This report was prepared for presentation at the Workshop on Gender and Ageing sponsored by the Institute of South East Asia Studies, Singapore, February 10-11, 2009.

Gender and Well-Being of Older Persons in Cambodia 2

Abstract This study represents the first systematic attempt to assess the relation between gender and the well-being of older persons in Cambodia, with reference to health, demographic, social support and economic indicators. The context is one in which the numerical dominance of women among the older age population is unusually pronounced due to a past history of civil unrest and violence. Our results, based primarily on the 2004 Survey of Elderly in Cambodia (SEC), reveal both differences and similarities between the sexes. Elderly women are far less likely than men to have a surviving spouse or to be literate, although even for men educational levels are quite low. Women report worse self-assessed health and more health symptoms and physical functioning problems than men but have higher survival rates. Seeing and hearing problems are reported fairly equally. Older men and women differ little in terms of social contact with and material support from children and very few appear deserted by their family. Although men are more likely than women to have work or pension income, there is little gender difference in a number of indicators of material well being including housing quality, household possessions, and self assessed economic situation. Although a conclusive advantage or disadvantage is not evident for any one sex across most dimensions, there are considerable variations in characteristics and circumstances. Recognition of these variations can be useful for understanding the unique needs of men and women in a country in which many older people have lived difficult lives, having faced harsh circumstances related to war and poverty.


Introduction

Today’s population of older Cambodians lived through an exceptionally traumatic period of history during their adult years. Prolonged civil strife starting in the 1960s eventually led to the complete take over of the country by the brutal Khmer Rouge regime in 1975. During their four year rule, political violence, severe food shortages and lack of medical care resulted in deaths estimated to constitute as much as a fourth of the total population (Heuveline 1998). Many who died were the sons, daughters or spouses of today’s older-aged population. Social dislocation, continuing political conflict, pervasive poverty and, more recently, an AIDS epidemic took their toll during the succeeding years resulting in further losses of family members. These events undoubtedly impacted the core family support of older persons in a country which is among the poorest in the Asia and where formal channels of assistance are minimal (Knodel 2007; Zimmer et al. 2006).

Population ageing is at a very early stage in Cambodia with less than 6% of the current total population age 60 or older, the lowest proportion of any country within Southeast Asia (Mujahid 2006). Nevertheless, almost one in four Cambodian households has at least one member who is age 60 or older.1

Gender issues in general have received more attention, although the focus is almost exclusively on ones concerning interests of women. Again this is at least in part in response to the urging and support of international agencies (UNIFEM et al. 2004; UNDP Cambodia 2008). However specific concerns about older persons are virtually absent in the discourse concerning gender. For example, there is almost no mention of the needs and situation of either older men or women in the extensive gender assessment conducted on behalf of international agencies (UNIFEM et al. 2004).2 Yet, as detailed below, understanding issues concerning gender and ageing is particularly pertinent for Cambodia since one feature of the current older population is the unusually large predominance of women. In part this lack of attention to how gender and ageing relate to each other results from the paucity of relevant research on the topic for Cambodia. The present study is intended to help redress this situation by exploring potential interactions between gender and aging in Cambodia. The primary source of data is the 2004 Survey of Elderly in Cambodia (SEC), a representative survey of 1,273 persons age 60 and over conducted in Phnom Penh and the five largest provinces. In addition, we draw on the nationally representative 2004 Cambodian Inter-censal Population Survey and 2005 Demographic and Health Survey.3 In all cases, we present original tabulations based on the raw data.

The low proportion of elderly in the population is likely a key reason for the general lack of attention given to issues related to older persons in Cambodia by the government and major international agencies. More recently there are some signs that aging is beginning to emerge as an issue on the government's agenda, in part in response to efforts by the United Nations, especially the UNFPA, and HelpAge international. In most respects these efforts are modest and still in the planning stages (Office of the Council of Ministers 2007; UN ESCAP 2007). Moreover, explicit attention to gender specific issues in relation to the population aging appears to be largely lacking.

Country Setting

According to the UN, Cambodia is one of the world’s “least developed countries” and ranks low on the Human Development Index (UNAIDS 2006). Much of Cambodia’s human capital was depleted during the disastrous rule of the Khmer Rouge and its aftermath. During the Khmer Rouge reign educated and professional persons were particularly poorly treated and subject to targeted executions (de Walque 2005). Others died of the pervasive starvation and disease that existed at unprecedented levels. In addition, many took flight as refugees during and after the Khmer Rouge rule and never returned. Although the Khmer Rouge were dislodged from national power following the invasion of forces from Vietnam at the end of 1978 and early 1979, they maintained a presence in parts of the


countryside. Significant civil strife continued during the ten-year period of occupation by Vietnam and several years following it. UN sponsored elections in 1993 helped restore some semblance of normality although factional fighting at times among coalition partners undermined political stability.

In recent years the economy has been improving, fueled by economic growth driven largely by the expansion in the garment sector and tourism. Still, over 80% of the population lives in rural areas and most depends on agriculture for their living. Moreover, the future of economic growth in Cambodia is uncertain given the global economic crisis that is currently unfolding and the fact that the recent economic success has been uneven, coming from only a couple of industries including tourism which is especially unpredictable (CIA 2008; PRB 2008).

Culturally, the Cambodian population is relatively homogeneous at least with respect to ethnic and religious composition. Approximately 90% of the population identifies itself as ethnic Khmer and 95% profess Theravada Buddhism as their religion. Despite the decades of turmoil, according to the extensive Cambodia gender assessment by international agencies, hierarchical notions of power and status within the society persist, conditioning social relations and relegating women to a lower status relative to men (UNIFEM et al. 2004). Demographic Aspects and Characteristics

Predominance of women. In most countries, women exceed men at older ages due to the near universal advantage women have in survival chances throughout the life span. However, this excess is especially pronounced in Cambodia as a result of the disproportionate share of men who died during long period of civil strife, especially during the Khmer Rouge time (de Walque 2005).

Figure 1. Female sex ratio by age group, Cambodia, Thailand and South-East Asia, 2005

133

107 104

174

126 120

0

20

40

60

80

100

120

140

160

180

200

Cambodia Thailand South-EastAsia

Cambodia Thailand South-EastAsia

ages 50-59 ages 60+

wom

en p

er 1

00 m

en

Source: United Nations 2007a (medium variant)


Figure 1 clearly illustrates this feature of Cambodia's 2005 older age population, defined as those aged 60 and above.4 The ratio of women to men in older ages in Cambodia is unusually high and far in excess compared to either neighboring Thailand or Southeast Asia overall. Thus among the current older population, in Cambodia there are 174 women for every 100 men compared to 120 women per 100 men in the South East Asia region overall. Moreover, as Figure 1 also shows, this unusual imbalance also holds for Cambodians in their 50s. i.e. the age group who will be entering the older age span during the following decade.

Demographic Profile. More detailed statistics related to population aging in Cambodia are presented in Table 1 based on estimates and projections (medium variant) from the United Nations Demographic Division (UN 2007a). These estimates indicate that at the turn of the present century, of the almost 13 million in Cambodia’s population, fewer than 600 thousand were age 60 and older and constituted less than 5% of the total. Over the first half of the present century, the projections anticipate that the older population will increase rapidly reaching almost 4 million by 2050 and their share of the total population will have risen to 15%. Thus, largely as a result of expected declines in fertility, considerable ageing of the population is anticipated to occur. Still, population ageing in Cambodia will be well behind that anticipated for the region as a whole. For example, the UN projections anticipate that 24% of the region's population will be age 60 and over by mid-century.

Table 1. Selected statistics on population ageing in Cambodia, 2000-2050 2000 2010 2020 2030 2040 2050 Population size (in 1000s) total 12,780 15,224 18,102 20,761 23,089 25,114 population age 60+ 590 868 1,275 1,837 2,330 3,829 Population age 60+ as % of population all ages

total 4.6 5.7 7.0 8.8 10.1 15.2 male 3.4 4.4 5.6 7.4 8.8 14.0 female 5.7 7 8.4 10.3 11.4 16.5 % female in age group 60-64 62.1 59.1 57.6 55.2 52.6 51.1 65-69 65.5 61.4 59.1 58.1 54.4 52.5 70-74 66.1 65.3 61.7 60.2 57.9 55.3 75-79 66.7 68.9 63.8 62.1 61.2 58.1 80+ 59.0 70.7 68.2 66.7 64.9 65.5 all persons 60+ 64.1 62.6 60.2 58.6 56.5 54.2 % of 60+ population in age group 60-64 36.3 38.6 37.0 34.8 30.0 36.1 65-69 28.0 27.8 27.1 27.4 26.0 29.8 70-74 19.0 16.6 18.4 18.7 20.7 14.2 75-79 10.2 10.4 10.8 11.2 13.6 10.4 80+ 6.6 6.7 6.7 7.8 9.8 9.5 all persons 60+ 100 100 100 100 100 100 Potential support ratio 1 (population 15-59/population 60+)

11.6

10.6

8.8

7.1

6.5

4.1

Source: UN 2007a.


The trend towards population ageing is expected to be relatively similar for both men and women although, for each year shown, the percentage of women over age 60 modestly exceeds that for men. Also indicated in Table 1 are the percent of each older age group that will be female. As noted in connection with the female sex ratio of the population discussed above, for all age groups within the older age span, women constitute a majority for each year shown. The magnitude of the female majority is expected to diminish considerably over time. Thus by mid century women are projected to constitute only 54% of the older population compared to 64% in 2000. This substantial decrease in the excess of females over males will result from the process of cohort succession. As younger cohorts who have not been subject to the distorting influence of the unusually high male mortality associated with past civil strife replace the current cohorts in the older age range, the gender imbalance will decline. In contrast, the UN projections anticipate only modest change in the sex imbalance for Southeast Asia as a whole, with a slight rise taking place in the percent female among the older population, from 54% to 55% between 2000 and 2050 (UN 2007a). Thus by mid-century, Cambodia should resemble other countries in the region in this respect and no longer be exceptional.

In most countries, mortality improvements at older ages are contributing to the ageing of the older population itself. Although the UN anticipates improving mortality at older ages in Cambodia as well, ageing of the older population will be very modest. It will also follow an irregular path reflecting past fluctuations in mortality and fertility associated with its turbulent history.5 Thus the age distribution of the older population in 2050 does not differ greatly from that in 2000 although some fluctuation is evident during the intervening years.

The final indicator shown in Table 1 is the potential support ratio. This ration is intended as a measure of the potential support base of persons in ages most likely to be economically productive relative to the population in older ages. Thus a falling ratio reflects a shrinking support base of adults on whom the old age population can depend. In the present study we define the measure as the ratio of persons 15-59 to persons aged 60 and older.6 A considerable decline in this ratio is projected over the 50 year period covered, falling from close to 12 in 2000 to just above 4 in 2050. This decline in the potential support ratio is mainly the result of projected reductions in fertility. If indeed fertility does continue to decline, older age Cambodians in the future will clearly have far fewer productive age persons per capita available to provide for their support.

Marital status. An elderly persons’ marital status has important implications for many aspects of their well-being. Spouses can be primary sources of material, social and emotional support and provide personal care during times of illness or frailty. Thus living with a spouse typically has advantages for older persons.

Nationally representative data on the marital status of the older population is available from the 2004 Cambodia Inter-Censal Population Survey and presented in Table 2. Very few older age Cambodians never married. Likewise only a small share our currently separated or divorced. In both cases, however, women are more prone than men to be in both situations. Far more striking gender differences, however, are evident with respect to the percent of older men and women who are currently married or are widowed. The large majority of elderly men (86%) compared to just under half of older women are currently married. In contrast, only 11% of men compared to close to half of older women are widowed (46%). This gender difference reflects a combination of higher male mortality, a tendency for men to marry women who were younger than themselves and higher remarriage rates among men than women in case of marital dissolution.


Table 2. Marital status (% distribution) by gender and age, Cambodia 2004

Never married Married Widowed Divorced/separated Age men women men women men women men women 60-64 0.3 2.3 93.3 56.5 5.2 37.4 1.1 3.8 65-69 0.6 1.1 89.2 49.8 8.9 44.2 1.3 4.8 70-74 0.2 1.7 83.5 44.5 14.4 50.8 1.8 2.9 75-79 1.1 1.6 78.9 38.1 16.9 56.5 3.2 3.7 80+ 1.8 1.9 63.1 32.2 32.0 62.5 3.2 3.4

Total 0.6 1.8 86.2 48.0 11.5 46.4 1.7 3.8

Source: 2004 Cambodian Inter-censal Population Survey

For both older women and men, the percent currently married substantially decreases with age while the percent widowed increases. Thus for the oldest old, i.e. those 80 and over, under a third of women and under two thirds of men are currently married while almost the reverse percentages are hold with regards to widowhood. However even for the younger elderly, substantial gender differences are also evident with far higher percentage of men than women being married and far more women than men being widowed.7

Number of living children. As documented below, adult children are important sources of support and other forms of assistance for their older age parents and undoubtedly play an important role in their psychological well-being. Those who have no children must rely on others for these forms of help. Beyond this, research suggests that coresidence with children and living nearby depends in part on the number of children available to provide such support (Zimmer & Korinek 2008). As Table 3 indicates, the vast majority of older persons in Cambodia have living children. However, a clear gender difference is evident with women being more likely to have no living children than men, both among those who are currently married and those who are not. In addition, among older persons who have living children, men average modestly more children than women. Family sizes remain large among elderly Cambodians who have children with two-thirds of men and close to half of women having 5 or more children still living.

Table 3. Number of living children, population age 60 and over, Cambodia 2004

Total Married Other

Men Wome

n Men Wome

n Men Wome

n Percent childless 0.9 5.3* 0.9 4.5* 1.1 5.6 Among persons with children Mean number 5.6 4.4* 5.8 5.2* 4.7 4.1* % distribution of number of children 1-2 11.4 22.9 9.9 14.7 18.3 26.6 3-4 21.5 30.3 19.3 23.7 31.2 33.3 5+ 67.1 46.8 70.8 61.6 50.5 40.1 Total 100 100 100 100 100 100 Statistical significance at .05 level yes yes no

Source: 2004 Survey of Elderly in Cambodia Notes: Children include own, step and adopted children; * Women significantly different from men at p<.05.


Loss of spouses and children in past. As noted in the introduction, Cambodian elders of today lived much of their life through a historical period characterized by severe social dislocation and wide scale civil conflict and political violence resulting in the death of family members for an enormous share of the population. Without doubt the most traumatic period was between 1975 and early 1979 when the Khmer Rouge held sway over the entire country and perhaps a fourth of the entire population perished. While adult mortality was high for both sexes, it was particularly high for men (de Walque 2005). Thus it is of interest to examine gender differences among today's elderly with respect to the loss of spouses and children in the past, particularly in relation to the Khmer Rouge years.

As Table 4 indicates large shares of today's elders lost a spouse during their lifetime. However, the percent that did so is twice as high among women than among men. Loss of a spouse due to violence is largely concentrated during the Khmer Rouge years and is confined mainly to deaths of husbands rather than of wives. Overall, the loss of spouses to illness is considerably more common than the loss to violence but during the Khmer Rouge years losses due to violence for men (i.e. spouses of women) were actually more common than losses attributed to illness.

Table 4. Percent of persons age 60 and over who experienced deaths of spouses and deaths of children age 11 or older by sex, cause of death, and period of death in relation to Khmer Rouge (KR) rule, Cambodian elders 2004 Loss of a spouse Loss of a child age 11+ Cause Period Men Women Men Women All causes (a) All periods 30.9 67.8* 42.8 52.5* Pre KR 8.0 12.9* 4.6 6.2 During KR 5.9 23.1* 19.5 32.4* Post KR 19.5 33.8* 25.9 23.3

Violence (b) All periods 2.1 16.6* 20.1 31.0* Pre KR 0.2 0.8 1.8 1.9 During KR 1.9 15.5* 16.1 28.0* Post KR 0.0 0.3 2.6 1.9

Illness All periods 27.3 45.4* 21.9 23.8 Pre KR 7.4 8.9 2.0 4.3* During KR 2.9 5.9* 3.6 4.6 Post KR 18.6 31.1* 18.0 17.0

Source: 2004 Survey of Elderly in Cambodia Notes: Pre KR =1970-75; During KR=1975-1979; post KR=1979-2004. (a) includes violence, disappearance, illness, accident and other causes (b) includes violence and disappearance * Women significantly different from men at p<.05.

Table 4 also examines the percent of today's elders who lost a child in the past. Results are shown regarding the loss of children age 11 and older since violent deaths are mainly concentrated from this age upwards. Both men and women attributed large shares of the deaths of their children age 11 and older to violence. The vast majority of such deaths occurred during the Khmer Rouge period. Despite the short four year duration of the Khmer Rouge era, substantial proportions of both elderly men and women lost at least one child during those years. Men were somewhat less likely than women to


report losing a child. The difference is mainly attributable to difference in losses due to violence during the Khmer Rouge period with only 16% of men compared to 28% of women experiencing such a loss. Given that men were more likely victims of violence than were women, this difference could possibly arise because those men who survived the period, and thus are available for interview in 2004, were selective for being disproportionately in areas during the Khmer Rouge period where violent death was less common. Unlike in the case of violent deaths, there is not a large difference in the proportion of men and women who reported losses of children age 11 and older due to illness.

Education and literacy. Substantial gender differences in education were pervasive in Cambodia in the past and especially during the time when the current older population was school age. As Figure 3 clearly shows, gender differences in formal education are stark among the current elderly population in Cambodia. Over three fourths of women aged 60 and over received no formal schooling compared to only slightly more than one fourth of men. Thus the percentage of men who attained each level of education shown is higher than that of women. Gender differences are extremely pronounced with respect to the percent who received secondary or higher education among the elderly population. Overall among women 60 and older, only 3% had received secondary education or beyond compared to 18% of men. Still, it is noteworthy that even among men, only a modest minority did so. Thus while there is a striking gender in education, neither elderly men nor women in Cambodia have much formal schooling .

Table 5 summarizes the educational attainment of older men and women in Cambodia by age based on the 2005 Cambodia Demographic and Health Survey. For both men and women the proportion with no education increases substantially with age. At the same time very large gender differences are apparent at every age within the elderly age span. Over 90% of women in the oldest age group (80 and over) lacked any schooling compared to 45% of men. Even among women in the younger elderly years, the substantial majority never attended school.

Table 5. Educational attainment by gender and age, Cambodia 2005 % with no education % with any primary

education (incomplete or complete)

% with any secondary or higher

education Age men women men women men women 60-64 15.9 58.4 59.6 36.1 24.5 5.5 65-69 20.7 77.6 60.1 20.0 19.2 2.4 70-74 35.8 86.8 49.5 11.6 14.6 1.6 75-79 37.8 94.2 52.1 5.5 10.1 0.3 80+ 44.5 91.7 45.9 8.3 9.6 0.0

Total 25.9 77.3 55.8 20.1 18.3 2.7

Source: 2005 Cambodia Demographic and Health Survey

As a result of the lack of education, older women are far less likely than men to be literate. This is clearly evident in results from the 2004 Cambodia Inter-censal Population Survey. As Figure 2 shows, only a small minority of women in the oldest ages are literate compared to two thirds or so of the oldest men. Following independence, schooling became more widely available for women and the difference in literacy is far less pronounced for the prime age and younger cohorts. For both men and women, there is a dip in literacy among those in the 40-44 cohort compared to those five years younger and five years older in 2004. This likely reflects the lack of schooling during the Khmer Rouge period since persons in the 40-44 cohort would have been in primary school age at that time.


Figure 2. Percent literate by gender and age, Cambodia 2004

9086

8185 84

8084

87 8784

78

6964 64

83

74

6773

66

5864

60

46

29

2015

129

0

10

20

30

40

50

60

70

80

90

100

15-19

20-24

25-29

30-34

35-39

40-44

45-49

50-54

55-59

60-64

65-69

70-74

75-79

80+

Age

Perc

ent l

itera

te

MenWomen

Source: 2004 Cambodia Inter-censal Population Survey One important implication of the results in Figure 2 Is that future generations of elderly will be characterized by far less of a gender gap in literacy than the current generation. However it is clear that a significant gap, with women less literate than men, will persist among the elderly population for the foreseeable future. Health

Older age mortality. A common, almost universal paradox with respect to gender and health found in many settings is that older women are found to live longer than men but prevalence rates of morbidity based on cross-sectional data are found to be higher for women. In general, this also appears to be the situation in Cambodia among the older population. Table 6, shows both total and ‘active life expectancy’ estimates for older men and women in Cambodia. Active life expectancy is a term that refers to the expectancy of life across different statuses of functional health, for instance, with and without disability or other types of functional disorders (Robine et al., 2003).

Elderly women in Cambodia have a clear life expectancy advantage. At age 60, where men’s life expectancy is 15.0, the advantage for women is an additional two years, or about 13% longer life, and at age 80, where men’s life expectancy is 5.3 years, women live an extra 0.8 years, or 15% longer. However, dividing the years of life into those expected to be lived with and without disability suggests that the issue of who is advantaged is less clear. Here, disability is conceptualized as having a limitation or difficulty with at least one of four tasks classified as activities of daily living (ADL): getting up from lying down, eating, bathing and/or dressing. Having difficulty with an ADL is commonly considered to be an indication of a disability or inability to conduct some function that is


necessary for self maintenance (Verbrugge and Jette 1994). The expected number of years without disability is very similar for women and men from age 60 onward. The expected number of years with a disability is substantially higher for women. The net result is that the extra years of life that older women live in Cambodia are on average years of disabled life. For instance, as Table 6 shows, the two extra years gained at age 60 include 1.7 years of disabled life and only 0.3 years of life without disability.

Table 6. Total life expectancy, expected life without disability and expected life with disability by gender and age, Cambodia 2004 Age Total life

expectancy Expected life without

disability Expected life with

disability Men Women Men Women Men Women 60 15.0 17.0 11.9 12.2 3.1 4.8 65 12.0 13.6 9.0 9.1 3.0 4.5 70 9.3 10.6 6.5 6.5 2.8 4.1 75 7.1 8.1 4.5 4.5 2.6 3.6 80 5.3 6.1 3.0 2.9 2.4 3.2

Source: Zimmer, Zachary. 2006. Disability and active life expectancy among older Cambodians. Asian Population Studies 2(2): Table 4, p 142.

Physical functioning and health. More years of life with disability translates into higher prevalence rates of disability for women when observed with cross-section data. Table 7 shows this more specifically.

Table 7. Percent reporting any and most severe ADL and functional limitations, age 60 and over, by gender and specific limitation, Cambodia 2004 Has any limitation Has serious limitation Men Women Men Women ADL limitations Getting up from lying down 15.2 21.3* 5.5 8.6* Eating 7.2 9.7 2.9 4.1 Bathing 7.2 9.1 2.5 3.6 Dressing 5.9 6.4 2.0 2.6 At least one ADL limitation 19.1 25.8* 7.4 10.6* Functional limitations Lifting 51.9 75.7* 27.1 47.2* Walking 47.5 72.5* 22.9 40.5* Climbing 47.6 68.9* 19.1 32.5* Crouching 47.9 64.5* 16.6 28.5* Grasping 32.2 44.3* 9.8 17.0* At least one functional limitation 71.1 88.2* 37.8 61.9*

Source: 2004 Survey of Elderly in Cambodia Notes: a serious limitation refers to having a lot of difficulty or cannot do. * Women significantly different from men at p<.05.


The upper half of the table presents the percent of women and men reporting the four ADL limitations mentioned above as well as the percent reporting any one limitation. This latter measure can be considered as the percent with a disability. The SEC asked two survey questions about each task, which allows for the division into severities. First, respondents were asked if they have any difficulty conducting the task on their own, without help from another person or using a device. If they answered in the affirmative, they were asked whether they have some difficulty, a lot of difficulty or cannot do the task at all. Two severities are shown in the table: the percent reporting any difficulty (meaning some, a lot, or cannot do) and the percent reporting a serious difficulty (a lot or cannot do). Women are more likely to report all four specific limitations and are more likely to report at least one, and this is true for both levels of severity. However, gender differences for only two comparisons, getting up from lying down and reporting at least one limitation, are statistically significant.

The bottom half of Table 7 shows the percent reporting more general functional limitations. These are tasks that relate to conducting bodily movements: lifting things above one’s head, walking 200 meters, climbing a flight of stairs, crouching and grasping things with fingers. Because these are more general movement tasks, the percent reporting these difficulties is higher than the ADLs. While having a functional limitation, for example difficulty lifting things, can lead to an ADL limitation such as having difficulty bathing it does not necessarily do so. Thus a greater proportion will report the former rather than the latter (Freedman, Martin and Schoeni 2002). A significantly higher percent of women consistently report these limitations regardless of the severity. Overall, 71% of men report at least one functional limitation compared to 88% of women, while 38% of men report at least one functional problem of greater severity compared to 62% of women.

Figure 3 shows the association between gender and serious limitations in physical functioning across age. The top panel refers to ADL limitations and the bottom to functional limitations. Both kinds of limitations tend to increase with age. The gender differentiation also persists across ages, particularly for functional limitations, where the gap is between about 10% and 20% regardless of age.

Several other indicators of physical health are shown in Table 8. Respondents in the SEC were asked to rate their overall health as very good, good, fair, poor or very poor. The percent assessing their health as poor or very poor is statistically higher for women of all ages and for the two broad age groups shown. There is little gender difference however in either the percent reporting impaired vision or hearing. The SEC also asked about whether respondents had experienced a series of fifteen health ‘symptoms’ within the last month. These symptoms are not chronic conditions but could relate to specific health disorders. For instance, the symptoms included headaches, vomiting and fever. The table indicates that the average number of symptoms reported was lower for men and than for women regardless of age group.

Care assistance. Respondents who reported at least one ADL limitation can be considered to have a disability and require assistance in completing basic functional tasks necessary for self-maintenance. For this reason, having someone who provides care and assistance can be important for well-being. Respondents in the SEC who reported either an ADL or functional limitation were asked additional questions about whether they receive help in conducting daily tasks and who it is that helps. Table 9 summarizes the situation regarding receipt of care assistance according to gender as well as marital status. Division by marital status is critical since the spouse, where available, is likely to play a vital caregiving role. The absence of a spouse may therefore be a distinct disadvantage. Since women are more likely to be widowed than men, they may be disadvantaged with respect to receiving care. Given that results are conditioned on having at least one limitation, we note that the sample sizes are relatively small, especially when divided by marital status. Thus we omit reporting statistical significance and results should be considered only suggestive.


Figure 3. ADL and functional limitations by age and gender, Cambodia 2004

A. Percent with one or more serious ADL limitation

0

5

10

15

20

25

60 to 64 65 to 69 70 to 74 75 to 79 80 and older

Age

Per

cent

with

lim

itatio

ns

Women

Men

B. Percent with one or more serious functional limitation

0

20

40

60

80

100

60 to 64 65 to 69 70 to 74 75 to 79 80 and olderAge

Per

cent

with

lim

itatio

ns

Women

Men

Source: 2004 Survey of Elderly in Cambodia Note: Serious limitation refers to having of a lot of difficulty or unable to do


Table 8. Percent reporting selected health problems among persons age 60 and over, by gender and age, Cambodia 2004 All ages 60-69 70+ Males Females Males Females Males Females Unweighted N 463 810 312 470 151 340 % reporting poor self-assessed health (a)

64.4 78.7* 60.1 67.6* 69.9 83.9*

% reporting not seeing well without glasses

64.1 65.4 58.2 57.6 74.0 75.8

% reporting not hearing well without a hearing aid

31.4 30.6 25.0 23.1 41.8 40.7

Mean number health symptoms (b)

7.1 7.7* 6.7 7.4* 7.8 8.2*

Source: 2004 Survey of Elderly in Cambodia * Women significantly different from men at p<.05. (a) does not include 37 proxy respondents. (b) count of number of the following reported symptoms within the last month: headache, vomiting, fever, diarrhea, skin problem, chest pain, joint pain, dizziness, back pain, trembling hands, stomach ache, breathing problem, coughing, loss of bladder control, weakness.

Table 9. Care assistance among persons age 60 and over who have an ADL problem or functional limitation, by gender, Cambodia 2004 Total sample Married Unmarried Men Women Men Women Men Women Unweighted N 75 203 47 28 28 175

% who received help 46.4 49.5 34.8 43.8 75.0 52.0

% not receiving help who report they need help

21.2 24.5 17.8 29.6 42.9 22.2

% receiving help who report they do not receive enough

40.0 36.7 28.0 45.0 57.1 34.2

Source: 2004 Survey of Elderly in Cambodia a) Will not add to 100% because individuals could name multiple sources

A little less than half of both men and women with limitations receive care assistance. Despite women being more likely to be widowed, there is not much evidence that they have a care-receiving disadvantage or a greater level of ‘unmet need’. A little less than a quarter of both men and women not receiving help report needing help, while around two-fifths of both sexes who receive help report that the help is not enough. Results differ somewhat by marital status. While the unmarried are more likely, overall, to report receiving help, they are also more likely than those married to report not receiving help but needing it and not receiving adequate help. Among the unmarried, men are more likely to be in this unfavorable situation. For instance, about 57% of unmarried men who receive help


report they do not receive enough compared to about 34% of women. Again, these results are based on small samples and are only suggestive.

Table 10 examines the source from where help is obtained. Respondents in the SEC who received help were asked to name each person providing help as well as to identify who is the main person providing care. There is some difference between women and men. Men are much more likely to name a spouse than are women as proving any care and as the main carer. Further division of the sample into married and unmarried indicates that this tendency is not a function of differences in martial status. For instance, 46% of married men say they receive help from a spouse compared to only 5% of married women.

Table 10. Sources of care assistance among persons age 60 and over who receive assistance, by gender, Cambodia 2004 Total sample Married Unmarried Men Women Men Women Men Women Percent who receive help from the following sources

Spouse 24.4 2.1 45.8 4.8 -- -- Son 20.0 23.7 29.1 55.0 9.5 15.6 Daughter 60.0 65.3 45.8 55.0 76.2 67.5 Son in law 4.4 11.3 0.0 19.0 9.5 9.2 Daughter in law 0.0 10.2 0.0 0.0 0.0 13.0 Grandchild 51.1 48.0 45.8 25.0 61.9 54.5 Other person 15.6 9.3 12.5 0.0 19.0 13.2 Percent distribution of the main caregiver

Spouse 15.6 2.1 29.2 5.0 -- -- Son 11.1 8.2 16.7 20.0 4.8 4.0 Daughter 40.0 50.5 25.0 45.0 57.1 53.3 Son in law 2.2 5.2 0.0 20.0 4.8 0.0 Daughter in law 0.0 5.2 0.0 0.0 0.0 6.7 Grandchild 31.1 25.8 29.2 10.0 33.3 30.7 Other person 0.0 3.1 0.0 0.0 0.0 5.3 Total 100 100 100 100 100 100

Source: 2004 Survey of Elderly in Cambodia a) Will not add to 100% because individuals could name multiple sources

There is little difference between men and women overall in the percent who receive help from their own sons and daughters or from grandchildren although differences appear when the care recipients are grouped by marital status. For women overall and married women, however, their children are more likely to serve as the main carer than for men thus making up for the lesser role of their spouses. Still for both men and women, the most typical main carer is a daughter or a grandchild. The only exception is among married men, who name a spouse as the main caregiver about as frequently as a daughter or grandchild. The importance of the grandchild is one notable feature of these results. In total, about 50% of both men and women report receiving help from a grandchild, and about one in three men and one in four women name the grandchild as the main caregiver. Only a daughter is more likely to play the main caregiving role.


Emotional well-being. The SEC included two sets of survey questions intended to measure emotional well-being. The first includes six items selected and translated from the standard CES-D depression scale, which has been used frequently in the United States to indicate levels of depression and the connection between levels of depression and other health-related outcomes (Radloff 1977). The second consists of four general questions about life satisfaction, referring to satisfaction with family relationships, housing, respect received from the younger generation, and overall life. Results of individual items, and a summary measure of each series of items, are shown in Table 11.

Table 11. Indicators of emotional well-being and satisfaction among persons age 60 and over, by gender, Cambodia 2004 Men Women Indicators of emotional well being Percent in last month who report: feeling not at all happy 34.9 43.2* had difficulty sleeping most of the time 34.7 36.6 not able to eat most of the time 15.3 20.7* feeling depressed most of the time 7.8 15.4* feeling unsuccessful with life most of the time 4.3 6.5 feeling lonely most of the time 4.0 12.5*

Mean number of negative emotional well being responses 1.01 1.35* Indicators of satisfaction Percent who are very or somewhat satisfied with: family relationships 49.3 52.8 housing situation 44.2 51.8* respect received from younger generation 23.0 28.9 life overall 38.2 42.4

Mean number of satisfied responses 1.59 1.94*

Source: 2004 Survey of Elderly in Cambodia * Women significantly different from men at p<.05. Notes: 31 proxy respondents excluded from all items. A small number of respondents answering don’t know individual items are excluded from percents for individual items. 59 cases answering don’t know to any emotional well-being item are excluded from summary measure for emotional well being. 81 cases answering don’t know to any satisfaction item are excluded from summary measure for satisfaction.

Women are more likely to report each of the depressive symptoms but are also more likely to give favorable responses to the satisfaction items. This result presents somewhat of a paradox. We can only speculate on an explanation. It may be that clinical levels of depression are indeed higher among older women in Cambodia but that this is compensated for by better inter-personal relationships and higher levels of network support that result in higher levels of life satisfaction. Alternatively, men may be more hesitant to report depressive symptoms as it goes against a culturally defined masculine image. Given these results it is not possible to state a definitive conclusion regarding gender differences in psychological well-being among Cambodian elders.

Risk Behaviors. Gender differentials in life expectancy, physical functioning, and other health disorders may in part be a function of differences in risk related behaviors. The SEC asked respondents about three such behaviors: tobacco smoking, alcohol drinking, and betel nut chewing.


The latter is a mild narcotic plant associated with mouth and throat cancer (Reichart 1995). Table 12 shows several clear gender patterns in these behaviors. Men are by far more likely to be smokers and drinkers, particularly smokers, while women are by far more likely to chew betel nut.

Table 12. Percent engaging in risk behaviors among persons age 60 and over, by gender and rural/urban residence, Cambodia 2004 All regions Urban Rural Men Women Men Women Men Women N (unweighted) 463 810 81 181 382 629 Smoking currently smokes 68.4 14.7* 39.5 3.7* 70.7 16.1* smokes daily 61.5 11.7* 35.9 2.4* 63.7 12.8* ever smoked 88.5 18.6* 63.2 6.1* 90.5 20.2* Alcohol drinking currently drinks 10.7 2.8* 7.9 3.7* 10.8 2.7* drinks daily 3.5 1.2* 0.0 0.0 3.6 1.2* ever drank 62.5 12.0* 41.0 7.4* 64.3 12.5* Betel nut chewing currently chews 5.7 63.9* 5.1 34.1* 5.9 67.6* chews daily 2.9 57.4* 2.6 29.3* 3.0 60.8* ever chewed 8.4 71.2* 7.7 41.5* 8.4 74.7*

Source: 2004 Survey of Elderly in Cambodia * Women significantly different from men at p<.05. Urban includes city of Phnom Penh. Rural includes semi-rural peripheral areas of Phnom Penh and rural areas in other provinces.

The SEC also asked about past behaviors for respondents who were not currently engaging in these behaviors. The percent of men who said they used to drink but no longer do is quite high, especially in comparison to the percent who are current drinkers. A second important point to note about risk behaviors is the difference in tendencies by place of residence. While the gender differentials are similar in both rural and urban areas, high risk behavior tends to be much more common in rural areas. For instance, 71% of rural men compared to only 36% of urban men are daily smokers. Also, 61% of rural women compared to 29% of urban women reported chew betel nut daily. Living Arrangements

Many aspects of well-being of older persons are influenced by their living arrangements. In the Asian context, and specifically in Cambodia, living with an adult child, especially a daughter, has been the traditional pattern (Kato 2000). While household composition is the most common and readily available indicator of living arrangements, it is important to recognize that the meaning and implications of particular configurations defined by such information can be ambiguous. One limitation is that such measures do not encompass information about others who live nearby but may still play an important role in the lives of elderly members (Knodel and Saengtienchai 1999). Another difficulty arises because the function of living arrangements can not be inferred with any certainty simply from their form (Hermalin, Roan, and Chang 1997). Thus although measures of the living arrangements based on household composition can be suggestive, they need to be interpreted cautiously.


With that said, coresidence with one or more adult children (or a functionally equivalent arrangement) often meets the needs of both generations. In contrast, living alone likely leads to less frequent interpersonal interactions, and hence potential feelings of loneliness. In addition, there is also a greater chance that urgent needs for assistance created by an acute health crisis or accident will go unnoticed longer than if others are present in the same household. Living only with a spouse in the absence of children is generally viewed as less problematic since a spouse can be a principal source of emotional and material support and provide personal care during illness or frailty.

Table 13 summarizes living arrangements based on the SEC. We note that the sample was limited to elders who are members of private households and does not cover persons living in institutional settings. While old-age homes in Cambodia are almost nonexistent, some elderly lived in the temples and are thus excluded from our sample. Unfortunately little systematic information is available on the extent to which elderly live in temples.

Table 13. Selected measures of living arrangements, by gender and marital status among persons aged 60 and over, Cambodia 2004 Total Married Other

Men Women Men Women Men Women Among all elderly, percent living alone 0.8 5.0* -- -- 4.3 7.2 only with spouse 7.2 4.1* 8.8 13.2 -- -- with any child 84.7 76.1* 83.7 73.0* 89.0 77.5* with or next to a child 87.0 77.8* 86.3 74.8* 90.1 79.1* in a skip generation household (a) 11.4 14.1 11.5 11.5 11.0 15.3 Among elderly with at least one child of the specified type, percent living with any child 85.4 80.4* 84.4 76.5* 90.0 82.1 an ever-married child of either sex 53.7 66.7* 49.0 56.6 74.1 71.3 an ever-married son 15.2 18.2 13.2 15.0 24.6 19.8 an ever-married daughter 49.6 60.5* 45.9 49.5 65.7 65.6

Source: 2004 Survey of Elderly in Cambodia Notes: (a) A skip generation household is one in which there is a coresident grandchild but no coresident ever married child or child-in-law. * Women significantly different from men at p<.05.

As the results show, living alone is rare among Cambodian elders, although it is more common among women than men. In contrast, living only with a spouse is more common among men but this is a function of gender differences in marital status. Among current married older Cambodians, more women than men live only with their spouse but because so many more women than men are not currently married (and thus can not live with a spouse), for the total population of elders, the reverse is true.

The vast majority of older Cambodians live with at least one child and even higher shares live with or next to a child. Regardless of marital status, men are modestly more likely than women to live with a child. In part this reflects the fact that women are more likely to have no children than men (see Table 3). When living with a child is conditioned on having at least one living child the difference contracts.


One type of living arrangement involving older persons of particular interest is the so-called ‘skip generation household’. Such households refer to situations in which grandparents live together with dependent grandchildren but in the absence of any of their adult children who have all either migrated or died. For the purpose of the present study, skip generation households are defined as those that have one or more grandchildren but no married child or child in-law in the household. Overall, older women are modestly more likely than men to live in skip generation households. We note that the differences between men and women are not statistically significant. Still they likely reflect gender role differences with respect to care for young children. This would explain why gender differences do not show up for married persons (since a woman is also present in cases of married men) but is evident for respondents who are not currently married (as a wife is not present to take responsibility for the grandchild in case of non-married men).

Coresidence does not necessarily imply the parents are primarily dependent on the children in the household and, especially in the case where the children are unmarried, the opposite may be true. However, coresidence with ever-married children is more likely to involve more important roles played by children in household support. It reflects a mature stage of intergenerational living arrangements that evolves after single children leave the household or marry and take adult responsibilities. Such coresidence is clearly related to the gender of the child. Older Cambodians are far more likely to live with a married daughter than a married son, which reflects a traditional matrilocal tendency among Cambodians. In contrast to the gender difference with respect to coresidence with children in general, women are more likely than men to coreside with an ever married child (conditioned on having an ever child). The difference arises, however, from gender difference in marital status of the respondent. Once marital status is controlled, no statistically significant difference between men and women is found. Social Contact with Children

Maintaining contact with children who leave the household is important for the social and emotional well-being of many older persons, especially if they do not have children living with them or very nearby. The migration of children out of the local community reduces opportunities for face-to-face interactions and thus can undermine intergenerational social support if contact is not maintained through other means. As Table 14 shows, only a relatively small minority of older Cambodians who have children do not live with or relatively near at least one child. Moreover, there is not much difference between parents who are currently married and those who are not.

Table 14. Location of nearest adult child among parents age 60 and older who have at least one adult child

All parents Parents not currently married

Men Women Men Women Has at least one adult child living in: Same household 82.4 80.4 89.2 82.2 Next door 85.0 82.1 90.3 83.8 Same village 95.1 94.3 97.8 94.9 Same district 95.3 95.3 97.8 95.1 Same province 97.2 98.1 98.9 97.8

Source: 2004 Survey of Elderly in Cambodia Notes: An adult child is defined as being age 16 or older. No gender difference is statistically significant at the .05 level.


Thus well over 90% of Cambodian parents have at least one of their children living within the same village and in most cases living in the same household. Less than 3% of older age parents have all their children living outside the province in which they live.8

In the rapid spread of cell phones recent years is likely greatly improving the ability of parents and migrant children to keep in contact. Although no information on telephone contact was collected in the SEC, only 16% of respondents lived in a household with either a cell or landline phone so at the time of the survey telephone contact would have been still fairly limited. Respondents were asked how often each non-coresident child age 16 or older visited and how often the respondent visited the child. Table 15 summarizes the frequency of such visits. Unlike in other tables, the unit of analysis in Table 15 is the child rather than the respondent. Since most respondents have more than one non-coresident child, the number of cases is substantially greater than the number of respondents with non-coresident children.

Table 15. Frequency of visits between non-coresident children and parents age 60 and older

All non-coresident children

Non-coresident children of not currently married parent

Father Mother Father Mother Daily 32.6 32.8 33.8 31.2 At least weekly 50.3 47.3 54.1 45.4* At least monthly 65.5 65.5 68.5 63.6 At least yearly 82.3 81.7 82.8 80.6 N of cases 1543 2202 314 1726

Source: 2004 Survey of Elderly in Cambodia Note: Non-coresident children are the unit of analysis in this table. Results are not weighted and based on the combined information about visits in either direction (i.e. child to parent and parent to child). * Women significantly different from men at p<.05.

As the results show, approximately a third of non-coresident children see their parents on a daily basis, reflecting the fact that many live next door or very nearby to the parents. Approximately half of non-coresident children see their parents at least weekly and almost two-thirds see their parents at least monthly. Thus Cambodian parents and their children who live outside the household tend to be in relatively frequent contact with each other. Less than one fifth of the non-coresident children do not see their parents at least yearly. The frequency of visits between non-coresident children and parents who are not married differs little from the overall pattern. In addition, the frequency of visits between children and their fathers and mothers are very similar. For only one of the comparisons shown in Table 15 is the difference between fathers and mothers statistically significant.

Figure 4 examines social contacts through visits using the parent as the unit of analysis. It shows the percent of fathers and mothers who see at least one child daily, weekly, monthly and yearly. We assume that respondents who coreside with a child see a child daily.9 For parents who do not live with a child, the frequency is based on reported visits with children in either direction. As the results indicate, few Cambodian older age parents do not have contact with at least one of their children on a relatively frequent basis. Moreover there is little evidence a difference between mothers and fathers in this respect with 97% of both seeing a child at least monthly. None of the fathers and only 2% the mothers reported that they did not see a child at least on a yearly basis. Moreover, even among the


few parents who did not see a child at least yearly, just over half reported that they had received money or a gift from a child during the past year. Thus few Cambodian older age parents are totally deserted by all their children and indeed most of them have frequent contact.

Figure 4. Social contact with at least one adult child among parents age 60 and older who have at least one adult child, Cambodia 2004

9296

97100

92

95 9798

50

60

70

80

90

100

Daily At least weekly At least monthly At least yearly

Perc

ent s

eein

g at

leas

t one

adu

lt ch

ild

Men Women

Source: 2004 Survey of Elderly in Cambodia Notes: an adult child is defined as being age 16 or older. No gender difference is significant at the .05 level. Sources of Material Support

Economic activity. The official retirement age in Cambodia for civil servants ranges from 55 to 60 depending on the level of the position. Employees of some private sector firms may also be subject to compulsory retirement ages. However for the vast majority of the population, who are mostly engaged in agriculture or are self employed in the informal sector, there is no particular discrete age at which work ceases. In addition, even those who are compelled to retire at a specific age may take on a different job and continue to work. Nevertheless, there are numerous reasons that lead older Cambodians to disengage from economic activities as they grow old. One of the most important is likely the decline in physical strength needed to carry out farming in which the majority is engaged.

Table 16 indicates the percentage of older men and women according to their work status during the prior year based on the 2004 Inter-censal Survey. The survey distinguished between being employed (including self-employment) in economic activities and doing housework. Overall, just over two thirds of older men are reported to be employed compared to only just over two fifths of older


women. In contrast, women were substantially more likely to report being engaged in housework than men. However, even if housework is treated as work, a substantially higher percentage of older men than women worked during the prior year. Moreover, pronounced gender differences are evident for each age shown. At the same time, the percentages of both men and women who are not working increases rapidly with age within the older age range. Thus, among those aged 80 and above, fully 80% men and 88% of women are no longer employed or doing housework. Clearly those who are no longer engaged in some economically productive activity, unless they have sufficient savings or investments of their own, will need to depend on others for their material support.

Table 16. Work status during prior year by gender and age, Cambodia 2004 Men Women working housework not working working housework not working 60-64 87.7 2.3 10.0 65.3 12.7 22.0 65-69 78.6 3.7 17.7 46.5 17.0 36.5 70-74 59.4 4.6 36.0 26.4 14.9 58.8 75-79 40.0 6.0 54.0 16.3 8.6 75.1 80+ 17.1 2.7 80.2 5.6 6.5 87.9 Total 68.6 3.5 27.9 41.4 13.2 45.4

Source: 2004 Cambodia Inter-censal Population Survey

Sources of income and support. Table 17 indicates the percentage of older men and women who receive income from several types of sources. Given that married couples are likely to share their income with each other, both the percent who reported that they themselves as well as the percent who reported either themselves or their spouse receive income from the source are presented.

Table 17. Sources of income by gender and marital status, persons aged 60 and over, Cambodia 2004 Total Married Other Men Women Men Women Men Women % with work income Self 46.9 27.4* 51.8 39.0* 24.7 22.3 self or spouse 56.4 32.6* 63.5 55.7 n.a. n.a. % with pension income Self 8.1 1.1* 9.0 1.4* 3.8 0.9* self or spouse 8.1 2.8* 9.0 7.0 n.a. n.a. % receiving rent or investment income Self 5.9 5.6 6.5 5.6 3.0 5.6 self or spouse 6.9 5.9 7.8 6.4 n.a. n.a. % receiving welfare income Self 0.4 1.0 0.5 0.9 0.0 1.1 self or spouse 0.4 1.0 0.5 0.9 n.a. n.a.

Source: 2004 Survey of Elderly in Cambodia Note: n.a.=not applicable *=statistically significant at the .05 level


For the overall population of older persons, men are much more likely than women to report income from their own work, reflecting their higher level of economic activity. For both men and women, the percent who report self income from work is lower than the percent who report themselves as working. This indicates that a fair proportion of older persons’ economic activity is for nonpaid work such as subsistence farming. When the combined work of self and spouse is used as a source of income, the gender difference is even more pronounced, reflecting the far higher proportion of women compared to men who are not currently married and thus cannot receive income from a spouse. When married couples are considered, including income from spouse considerably reduces the gender difference in work as a source of income. Among non-married older persons there is little difference in the percent who receive income from their own work. For both sexes, the percentage who have income from work is considerably lower for non-married than for married respondents, reflecting the fact they are on average considerably older than those who are still married.10

Income is far less common from pension, welfare, and rent or investment than work. Income from pensions is far more common among men than women reflecting the substantial gender gap, especially during the lifetime of the current elderly, in civil service and other types of occupations that provide pensions. For married persons, the gender gap in pension income is greatly reduced when receipt of pensions by a spouse is combined with ones own receipt of pensions. Income from rent or investment, while rare, is roughly as common for women as men, regardless of marital status. Income from welfare is extremely rare for elderly of either sex.

Although older persons may receive income from various sources, some are more important that others. Respondents in the SEC where asked to identify who was the main person who supported their household as well as which source they considered as the most important for their own support. As results presented in Table 18 show, the majority of both older aged men and women in Cambodia say that their children are the main contributors to overall household support.

Table 18. Main source of support and main contributor to household support, by gender and marital status, persons aged 60 and over, Cambodia 2004

Total Married Other Men Women Men Women Men Women Main contributor to supporting household (% distribution) Self 11.3 10.2 10.7 5.1 13.8 12.5 Spouse 9.4 5.1 11.2 15.7 (a) (a) self and spouse 21.1 6.6 25.8 21.3 0.0 0.0 Children 55.1 71.6 50.1 54.0 77.7 79.5 Other 3.1 6.4 2.1 3.8 8.5 8.0 Total 100 100.0 100 100 100 100 Statistical significance at .05 level yes yes no Main source of support (% distribution) work (own or spouse) 41.9 19.5 47.5 37.7 17.0 11.2 Children or children-in-law 53.0 70.8 47.7 52.5 76.6 78.9 other family members 1.8 5.8 0.7 3.8 6.4 6.7 non-family sources 3.3 3.9 4.1 5.9 0.0 3.2 Total 100 100 100 100 100 100 Statistical significance at .05 level yes yes no

Source: 2004 Survey of Elderly in Cambodia Note: A small number of non-married persons who reported their spouse as their main contributor to supporting the household included in the category ‘other”.


While this is true for a greater share of women than men in the total elderly population, little gender difference is evident when marital status is controlled. Overall, elderly Cambodians who are not currently married are substantially more likely to cite their children as their main contributor to their household support than those who are married. Since more women are non-married than men, children are a more common main source of support for women than for men overall. Among those for whom children are not the main source of support, the respondent, the respondent's spouse, or both together are the main contributors to their household support.

Responses to the question about the main source of the respondent's own support largely mirror the findings with respect to who is the main contributor to supporting the household. Children are again cited as the most important source. Among the overall elderly population, women more often than men named children as their main source of support. Again this difference is a function of marital status differences combined with the fact that children are more likely to be the main source of support for non-married than married elders. The second most important source of support was the respondent's or the respondent's spouse's economic activity. Among the total population of older persons, work or spouse’s work is far more common as a main source of support for men than women but again differences disappear when marital status is controlled.

Intergenerational exchanges. In much of the developing world, including Southeast Asia, intergenerational exchanges of material support is an integral part of the family system and particularly crucial for the welfare of elderly parents (World Bank 1994). As noted above and as previous analyses have shown, children are the main source of support for the majority of Cambodian elders (Zimmer et al. 2008). Children can provide material support both in the form of money and through giving food and other goods. In the SEC, respondents were asked about the contributions made to them by each of their children in terms of money, food or clothing, and general support of the household during the previous year. Respondents were also asked about the contributions they made to each of their children. Since some contributions may be too modest to make any real contribution to the material well-being of the recipient, information was also solicited about the value of the contributions.

Table 19, summarizes the percent of older aged Cambodians who receive assistance from children. The type of support provided by a child and the use to which it is put may differ according to whether or not the child lives with the parent. For example, money or food that is provided by coresident children may be made mainly to cover communal household expenses rather than being given the parent’s for their own use. This is less likely to be so for support given by non-coresident children. Thus results are shown separately for contributions from coresident and non-coresident children (conditioned on having a child of each particular type) as well as for all children combined.

Among elderly parents, the vast majority report that a child contributes to the support of the household, provides some money, and provides some food or clothing. However the amount of support is typically quite modest. Only about a fourth of elderly parents reported that they received combined contributions of money and food or clothing during the previous year worth at least $25. In general women are more likely than men to report receiving the types of support being considered, especially from coresident children. This may reflect a greater tendency for women than men to manage the household budget, a common pattern in Southeast Asia. This interpretation gains some support from the fact that gender differences in receipt of material support from children are largely a function of differences among married respondents. Among respondents who are not currently married, there is little difference between men and women in the percent who receive the various types or amounts of material support from children. Since married respondents may be sharing the contributions with their spouses, it is not clear that women benefit more than men even though overall they are more likely than men to be the recipient of contributions by children.


Table 19. Receipt of material support from children during past year, by gender and marital status, persons aged 60 and over, Cambodia 2004

Total Married Other Men Women Men Women Men Women From any child (those with living children) Support of household 89.6 91.2 88.7 86.6 93.5 93.1 Money 94.5 93.6 94.4 93.3 94.6 93.8 Food/clothes etc. 82.8 87.3* 81.7 86.7 87.1 87.5 Money and/or food/clothes worth $25+ 24.1 27.8 23.4 28.6 28.0 27.4 From any coresident child (among elders with coresident children) Support of household 79.0 91.6* 74.9 83.0* 96.4 95.3 Money 78.2 80.4* 67.1 76.0* 89.3 90.9 Food/clothes etc. 62.3 79.2* 58.4 67.4* 78.6 84.1 Money and/or food/clothes worth $25+ 15.6 19.1 13.7 14.5 22.9 21.1 From any non-coresident child (among elders with non-coresident children) Support of household 61.7 63.2 64.4 65.2 49.4 62.3* Money 89.2 90.2 90.0 95.7* 85.5 87.7 Food/clothes etc. 72.7 76.3 73.1 82.4* 71.1 73.5 Money and/or food/clothes worth $25+ 14.8 19.2* 15.3 24.8* 13.1 16.4

Source: 2004 Survey of Elderly in Cambodia *=statistically significant at the .05 level

Older age parents may also provide material support to their children. In two thirds of the cases where parents and children coreside, the parents own the dwelling and thus are providing living quarters for the children who live with them. In half of these cases, the house is owned jointly by husband and wife reflecting the fact that joint ownership is the typical situation for married respondents. Among the remainder, a mother is more likely the owner of the house then a father reflecting the dominance of women among elders who are not currently married.

Some Cambodian elders also provide direct material support to their children in the form of money and food or goods. Table 20 summarizes the percent of elderly parents who provided such support to at least one of their children during the previous year. Again because the use of such support may differ considerably between children who live with the parents and those who do not, results are also shown separately in relation to coresident and non-coresident children (contingent on having a child of the particular type).

Clearly parents are less likely to provide material assistance to their children than the reverse. Very few report providing amounts of support valued at $25 or more. Support of each type shown that is given is far more common to coresident children than to non-coresident children. Only a very small fraction of elders reported giving substantial amounts of material aid to any non-coresident child. To the extent gender differences are apparent, men are more likely to provide each type of support than are women. This may reflect the fact that older age men are more likely to have cash income through work or pensions than older women. Within marital status categories, gender differences are quite modest and not statistically significant. Married respondents are more likely to provide material assistance to children than non-married respondents. In turn, women are more likely to be in the non-married category, which accounts for the overall gender differences.


Table 20. Provision of material support to children during past year, by gender and marital status, persons aged 60 and over, Cambodia 2004 Total Married Other Men Women Men Women Men Women Support given to at least one child (among elders with living children) Money 39.5 24.5* 42.9 37.6 24.2 18.6 Food/clothes etc. 33.7 17.9* 37.9 32.6 14.8 11.3 Money and/or food/clothes worth $25+ 7.7 3.6* 9.1 6.3 1.2 2.4 Support given to at least one coresident child (among elders with coresident children) Money 42.1 24.5* 46.7 41.8 22.8 17.1 Food/clothes etc. 34.7 17.2* 39.9 34.6 12.8 9.9 Money and/or food/clothes worth $25+ 8.0 3.8* 9.7 7.8 0.9 2.1 Support given to at least one non-coresident child (among elders with non-coresident children) Money 13.1 9.0* 14.2 12.3 8.4 7.4 Food/clothes etc. 10.0 7.1 11.1 11.1 5.0 5.2 Money and/or food/clothes worth $25+ 1.8 1.0 2.1 0.9 0.5 1.1

Source: 2004 Survey of Elderly in Cambodia Note: *=statistically significant at the .05 level

Economic Well-being

The SEC provides considerable information relevant to assessing the economic well-being of respondents. This includes information on assets, debt, household possessions, and housing quality that can serve as objective indicators as well subjective assessments by the respondents themselves and by the interviewer.

Assets and debt. Table 21 provides information on assets and debt of older Cambodians. Since spouses may jointly benefit form assets of either, results regarding asset ownership are presented both in relation to the respondents themselves as well as to the combined situation of respondents and their spouse. Most older Cambodians either own their dwelling units by themselves or jointly with a spouse. Among those who do not, the majority live in dwellings owned by their children. Overall, women are more likely than men to be the owner of the house themselves. At the same time, women are also more likely than men to live in a house owned by their children. In contrast, men are more likely to jointly own their house with a spouse. When marital status is controlled, however, the gender differences largely disappear. Although among non-married elders, women are less likely than men to own the house themselves and more like to live in a house owned by their children, the difference is not statistically significant.

With respect to other assets, men are more likely than women to own land and livestock while women are more likely to own jewelry. Very few elderly men or women have bank accounts. At the same time, men are more likely to be in debt than women although this is limited to those who are married. Thus while the husband may be the one responsible for the debt, it may nevertheless detract from the wife's economic well-being as well. In any event, in most cases debt is not a major problem as indicated by the very low percentages who say debt poses a serious burden for them.


Table 21. Assets and debt by gender and marital status, persons aged 60 and over, Cambodia 2004 Total Married Other Men Women Men Women Men Women House ownership (% distribution) Self 19.7 37.7 9.5 4.3 64.9 53.0 spouse 5.9 1.6 7.2 5.1 -- -- self and spouse together 53.9 21.7 65.9 69.4 -- -- children 17.0 31.3 14.3 16.6 29.8 37.8 Other 3.5 7.8 3.1 4.7 5.3 9.1 Total 100 100 100 100 100 100 Statistical significance at .05 level yes no no

Other assets % own land Self 25.5 14.3* 28.4 8.8* 20.3 15.4 self or spouse 28.3 16.0* 32.7 19.1 -- -- % own livestock Self 41.3 22.0* 42.5 34.9 36.3 16.2* self or spouse 47.1 23.3* 49.5 39.3* -- -- % having bank account Self 0.1 0.5 0.1 0.9 0.0 0.3 self or spouse 0.3 0.5 0.3 0.9 -- -- % having jewelry Self 13.4 27.0* 14.7 27.1* 7.8 27.0* self or spouse 28.8 28.0 33.5 30.1 -- --

Debt % in debt 26.3 19.9* 28.4 21.2* 16.3 19.2 % for whom debt is a serious burden 5.2 3.9 5.8 5.2 2.3 3.4

Source: 2004 Survey of Elderly in Cambodia Note: *=statistically significant at the .05 level

Housing quality and household possessions. Few elders of either sex live in a house not owned by themselves, their spouse, or their children. Thus the gender differences in the distributions of house ownership do not necessarily confer advantage to either elderly men or women. A different issue, however, is whether there is a gender difference in the quality of housing of older Cambodians. Figure 5 examines several characteristics of the dwellings in which older men and women reside. Women are somewhat more likely than men to live in a house that has electricity but with respect to the other characteristic shown gender differences are minimal. Both men and women are almost equally likely to live in a house with a finished floor, that has a better quality roof, and that has a flush toilet. Thus overall there does not appear to be a major gender difference with respect to housing quality for Cambodian elders.


Figure 5. Housing quality indicators among persons 60 and older by gender, Cambodia 2004

26.7

16.1

49.7

33.234.3

19.2

46.0

33.0

0.0

10.0

20.0

30.0

40.0

50.0

60.0

Has electricity Has finished floor Has better qualityroof

Has flush toilet

Perc

ent

MenWomen

Source: 2004 Survey of Elderly in Cambodia. Note: Only the difference in electricity is statistically significant at the .05 level.

Besides assets and housing, wealth is also reflected in the possessions of a household. Figure 6 compares older men and women in Cambodia with respect to a series of such possessions. Even if a household member other than the elderly respondent is the owner of a particular item, elderly members are likely to benefit from its presence. For example, for most appliances in a household, all members are likely able to use or benefit from them. Moreover, the presence of household possessions reflects the overall economic status of the household. In general, there is little difference in the percent of elderly Cambodian men and women who live in households with each of the items shown. Also to the extent differences are evident, no consistent pattern is evident. Thus there appears to gender equality in this respect.


Figure 6. Household possessions among persons aged 60 and over, Cambodia 2004

1%

5%

5%

14%

14%

32%

65%

77%

2%

4%

4%

18%

21%

28%

63%

65%

0% 10% 20% 30% 40% 50% 60% 70% 80% 90%

Air conditioning

Refrigerator

Car or truck

Phone

Electric Fan

Motorcycle

Television

Radio

Percent living in a household with the stated possession

WomenMen

Source: 2004 Survey of Elderly in Cambodia Note: Only the difference in radio and fan are statistically significant at the .05 level.

Summary measures. Information on characteristics of the dwelling unit and the presence of household possessions can be combined using principal component analysis to provide an overall wealth score (Filmer & Pritchett 2001). Given the items on which this score is based, it can be considered as an objective measure of the economic well-being of respondents. Other questions in the SEC provide subjective judgments. Respondents themselves were asked to rate their own economic status relative to others in the community and to assess the sufficiency of their income and their satisfaction with their economic situation. An overall self-assessed economic well-being score can be derived by combining responses to these three questions.11 In addition, interviewers were asked to judge the economic status of the household on a five-point scale based on the appearance of the respondent's house. In order to facilitate interpretation of these three measures, the scores have been converted to percentiles in such a way that higher percentiles signify better economic well-being. The mean percentile scores are presented for men and women in Table 22.

Overall, there is little gender difference in the mean percentile scores for the objective measure of wealth based on housing characteristics and household possessions and for the self-assessed economic well-being scores. However interviewers rated the socioeconomic status of women respondents somewhat lower than men with the difference being statistically significant. The overall difference in this measure, however, is largely attributable to the fact that interviewers rated non-married respondents lower than those who are currently married. Thus within the two marital status


categories no statistically significant difference is apparent. Thus it appears that on both the objective and subjective summary measures, there is little substantial difference in economic well-being between older Cambodian men and women. Table 22. Objective and subjective summary measures of economic well-being, by gender and marital status, persons aged 60 and over, Cambodia 2004 Measure of economic well being (expressed as mean percentile)

Total Married Other Men Wome

n Men Women Men Women

Wealth score based on house characteristics and household possessions 50.5 49.7 51.7 49.4 45.1 49.9

Self assessed economic well-being score 50.5 49.7 51.0 52.1 48.1 48.6

Interviewer assessed economic status score 52.7 48.2* 54.0 51.4 47.2 46.8

Source: 2004 Survey of Elderly in Cambodia Note: *=statistically significant at the .05 level Discussion and Conclusions

This study represents the first systematic attempt to assess the relation between gender and the well-being of older persons in Cambodia, with reference to health, demographic, social support and economic indicators. The context is one in which the numerical dominance of women among the older age population is unusually pronounced due to a past history of civil unrest and violence, especially during the rule of the Khmer Rouge, that disproportionately took its toll on adult men among the cohorts who now constitute Cambodia’s elderly. An additional impact on the current generation of older women is that they far more likely than the surviving men to have experienced the loss of a spouse during period of civil disturbances.

The results of our analyses, based primarily on the 2004 Survey of Elderly in Cambodia (SEC), reveal both differences and similarities between elderly men and women. On some measures older Cambodian women clearly face disadvantages compared to their male counterparts. They are far less likely to be currently married and thus to have a spouse to provide social and material support and assistance. Elderly women are also far less educated than men and far less likely to be literate. At the same time, although most elderly men had some formal schooling, their educational levels are also quite low.

There are numerous gender differences in terms of the health dimensions of older Cambodians examined but the direction of the differences do not consistently favor one sex over the other. Elderly women, for instance, report poor self-assessed health more often than men and also report a greater number of health symptoms although seeing and hearing problems are reported fairly equally. Elderly women are also more likely to suffer from physical functioning problems. It should be noted that earlier analysis of SEC data indicated that the overall level of disability among older men and women in Cambodia is extremely high in comparison to other countries (Zimmer 2006) and that part of this may be explained by high levels of poverty among the population (Zimmer 2008). On the most crucial measure of health, survival rates, older Cambodian women are clearly advantaged compared to men since they live longer. For instance, life expectancy for a 60 year old Cambodian woman is two years greater than for a man of the same age. At the same time, most of these extra years are years lived with a disability.


Men and women both engage in risk behaviors. Older men are more likely to smoke and drink while women are more likely to chew betel nut. Most likely the risk behaviors of men are more serious and thus contribute to their higher mortality. Women are more likely to report depressive symptoms but are also more likely to report satisfaction with life. Women are unlikely to name a spouse as a caregiver even when married, but men and women with physical limitations appear equally likely to have unmet caregiving needs. Overall, older men and women in Cambodia face some similar and some contrasting health challenges that need to be understood within the context of a poor developing country.

Women are more likely than men to live alone, although only a small minority does so. Women are also less likely to coreside with a child in part because women are more likely to have no living children. At the same time, among elders who have married children, women are more likely to coreside with one and thus be in a living arrangement where a mature adult child is readily available to provide assistance within the household.

Several gender differences are evident that potentially but not necessarily convey advantage or disadvantage. For example, older women are less likely than older men to be in the labor force and more likely to depend on children as their main source of income. The latter is related to the fact that women are not only less likely to be economically active but also comprise a larger proportion of non-married older persons, who are most economically dependent on their children. Men are more likely to have debt but this is only among those who are married and thus may disadvantage their spouses as well. Men are also more likely to provide material support to children although rarely in an amount that is substantial.

On several important dimensions of well-being, we find little association with gender. With respect to social contact with children, differences between elderly fathers and mothers are not pronounced and few of either sex appear to be deserted by their children. Likewise the large majority of elderly parents, regardless of sex, receive some material support from their children. Older men and women also appear similar in their material well-being as measured by the quality of their housing and the number of household possessions. In addition, they differ little in their self-assessed economic situation. This is particularly interesting given that men are more likely to have income from work and pensions. One reason may be that these differences in income sources are largely among married elders and thus both spouses may benefit even if only one is the income recipient. While the finding that major gender differences in a number of aspects of material well-being appear to be absent is encouraging, it must be interpreted with the context of widespread poverty among both elderly Cambodian women and men.

In sum, our comprehensive analysis of gender differences across several critical dimensions of well-being shows various important differences in the lives of older men versus women in Cambodia as well as similarities. While we do not see a conclusive advantage or disadvantage existing for any one sex across any single dimension, we do see variations in characteristics and circumstances. Recognition of these variations can be useful for understanding the unique needs of men and women in a country in which many older people have lived difficult lives, having faced harsh circumstances related to war and poverty.


_________________________________________________________

Endnotes 1 Based on original tabulations of the 2000 Cambodia Demographic and Health Survey. 2 The only mention of older persons is in reference to the fact that older women together with younger women shoulder much of the burden of caring for AIDS patients and AIDS orphans. 3 A detailed description of the Survey of Elderly in Cambodia including the questionnaire, sample design, and weighting procedure is available in Knodel et al. 2005. Detailed descriptions of the 2004 Cambodian Inter-censal Population Survey and 2005 Demographic and Health Survey are available in National Institute of Statistics 2005 and National Institute of Public Health, National Institute of Statistics and ORC Macro. 2006. 4 In accordance with conventional definitions of the old age span in Cambodia and in many developing countries, we consider the older age population to be those age 60 and older. 5 For example, the UN projects that life expectancy at age 60 in Cambodia will increase about by about 3 years between 2005-10 and 2045-50 (United Nations 2007b). 6 Conventionally the potential support ratio is defined as the ratio of the population age 15-64 to that 65 and older but to be consistent with the definition of the older population used in the present study, we use ages 15-59 and 60 and older. 7 The marital status distribution of the Survey of Elderly in Cambodia (SEC) sample, on which most subsequent results in this report are based, differs somewhat from that found in the larger nationally representative Cambodia Inter-censal Population Survey (CIPS) even though both surveys occurred in 2004. The difference is particularly pronounced for women, with the SEC indicating substantially lower proportions of older women married and higher proportions widowed than is true in the CIPS. The reason for this is unknown. This is unlikely to be the result of the fact that the SEC is based on only six provinces since comparisons controlling for the sample area with the 1999 socioeconomic survey finds similar contrasts (Knodel et al. 2005). Thus there is some likelihood that the SEC is overly weighted towards widowed persons, particularly women and underweighted with those who are currently married. 8 For only 3 cases of the 1220 parents of adult children interviewed in the SEC are all the children living outside of Cambodia (not shown). 9 The SEC asked about frequency of visits only in relation to children who live outside the household. 10 In the SEC sample, non-married men are on average more than six years older than a married man and non-married women are more than four years older than married women. 11 Given the number of categories in which responses to these three questions were recorded, the combined sum (i.e. score) varied between 0 and 9, with 0 assigned to persons who rated themselves in the worst off category for each question and 9 for those who rated themselves highest on each.

References Central Intelligence Agency (CIA). 2008. The World Factbook: Cambodia (updated as of 4 December

2008). <www.cis.gov/library/publications/the-world-factbook/geos/cb.html>

de Walque, Damien. 2005. Selective Mortality During the Khmer Rouge Period in Cambodia. Population and Development Review 31(2):351-368

Filmer, D. & L. H. Pritchett. 2001. Estimating Wealth Effects without Expenditure Data - or Tears: an Application to Educational Enrollments in States of India. Demography 38(1):115-132.

Freedman, Vicki A., Martin, Linda G. and Schoeni, Robert F. 2002. Recent trends in disability and functioning among older Americans: A critical review of the literature. Journal of the American Medical Association. 288 (24): 3137-3146.

Hermalin, Albert I., Carol L. Roan & Ming-Cheng Chang. 1997. Plans for Moving among the Elderly in Taiwan and Their Outcomes: A Panel Study. In Annual Meeting of the Population Association of America. Washington, DC.

Heuveline, P. 1998. 'Between One and Three Million': Towards the Demographic Reconstruction of a Decade of Cambodian History (1970-79). Population Studies-a Journal of Demography 52 (1): 49-65 (1998).

Kato, E. U. 2000. Ageing in Cambodia: Tradition, change and challenges. In Ageing in the Asia-Pacific Region. Edited by D. R. Phillips. New York, NY : Routledge.

Knodel, John. 2007. Poverty and the impact of AIDS on Older Persons: Evidence from Cambodia and Thailand. Economic Development and Cultural Change 56(2):441-475.

Knodel, John & Chanpen Saengtienchai. 1999. "Studying Living Arrangements of the Elderly: Lessons from a Quasi Qualitative Case Study Approach in Thailand," Journal of Cross-Cultural Gerontology 14 (3): 197-220.

Knodel, John, Souvan Kiry Kim, Zachary Zimmer & Sina Puch. 2005. Older persons in Cambodia: A profile from the 2004 Survey of Elderly. Research Report 05-576, Ann Arbor, MI, Population Studies Center.

National Institute of Public Health, National Institute of Statistics [Cambodia] and ORC Macro. 2006. Cambodia. Demographic and Health Survey 2005. Phnom Penh, Cambodia and Calverton, Md, USA: National Institute of Public Health, National Institute of Statistics and ORC Macro.

National Institute of Statistics, 2005. Cambodia Inter-Censal Population Survey 2004, General Report. Phnom Penh: UNFPA.

Office of the Council of Ministers. Secretariat of the Committee on Population and Development. 2007. Population Aging in Cambodia: Planning for Social Protection. Phnom Penh: UNFPA.

Population Reference Bureau (PRB). 2008. 2008 World Population Data Sheet. Washington, DC: Population Reference Bureau

Radloff, Lenore Sawyer. 1977. The CES-D Scale. Applied Psychological Measurement 1 (3) : 385-401.


Reichart, P.A. 1995. Oral cancer and precancer related to betel and miang chewing in Thailand: A review. Oral Pathology and Medicine. 24: 241-243.

Robine, J-M, C. Jagger, CD Mathers, EM Crimmins and RM Suzman (Eds). 2003. Determining Health Expectancies. Hoboken, NJ: Wiley.

Verbrugge, Lois M. & Jette, Alan M. 1994. The disablement process. Social Science and Medicine 38 (1) : 1-14.

United Nations ESCAP. 2007. Country Report: Review and Appraisal of the Progress on Implementation of the Madrid International Plan of Action on Ageing (MIPAA) in Kingdom of Cambodia. Presented at High-level Meeting on the Regional Review of the Madrid International Plan of Action on Ageing (MIPAA) Macao, China 9-11- October 2007. Submitted by the Royal Government of Cambodia. [www.unescap.org/ESID/psis/meetings/AgeingMipaa2007/Cambodia.pdf]

United Nations (UN). 2007a. World Population Prospects: The 2006 Revision. New York, United Nations.

United Nations (UN). 2007b. World Population Ageing 2007. New York, United Nations.

UNDP Cambodia. 2008. Project Fact Sheet 01/2008 (project 00037213) Partnership for Gender Equity. UNDP: Phnom Penh.

UNIFEM, WB, ADB, UNDP and DFID/UK. 2004. A Fair Share for Women: Cambodia Gender Assessment. UNIFEM, WB, ADB, UNDP, DFID/UK: Phnom Pen

World Bank. (1994). Averting the Old Age Crisis. New York: Oxford University Press

Zimmer, Zachary. 2006. Disability and active life expectancy among older Cambodians. Asian Population Studies, 2 (2) : 133-148.

Zimmer, Zachary. 2008. Poverty, wealth inequality and health among older adults in rural Cambodia. Social Science and Medicine, 66 (1): 57-71.

Zimmer, Zachary, John Knodel, Kiry S. Kim, & Sina Puch. 2006. The Impact of Past Conflicts and Social Disruption in Cambodia on the Current Generation of Older Adults. Population and Development Review 32(2):333-360.

Zimmer, Zachary & Kim Korinek 2008. Does Family Size Predict Whether an Older Adult Lives with or Proximate to an Adult Child in the Asia-Pacific Region? Asian Population Studies 4(2):135-159.

Zimmer, Zachary, Kim Korinek, John Knodel and Napaporn Chayovan. 2008. Migrant Interactions with Elderly Parents in Rural Cambodia and Thailand. Journal of Marriage and Family 70:585–598.

Gender and well-being of older persons in Cambodia - Population

Documents

Transcript of Gender and well-being of older persons in Cambodia - Population