Spatial Distribution of Suicide in South Korea: County Differentials of Age, Gender Specific Suicide Rates In
Spatial distribution of Suicide in South Korea: County differentials of age, gender specific suicide rates in
2005 & 2010
1. Introduction
Recently, suicide is considered as one of the most prevalent causes of death in South Korea (NSO
2013). Suicide is the most frequent cause of death in 10~39 years old. Following cancer, suicide is the 2nd leading cause of death in 40~59 years old. Commonly, ‘death’ is what the most of human beings want to get away from. But suicide means some people are trying to choose death instead of life because of severe reasons disrupting their normal life. Therefore, suicide can be considered as a strong indicator of
‘abnormality’ of life. Also, it is obvious that death of suicide has huge negative impact not only for the individuals but also society. Furthermore, suicide generates bereavement impacts to their friends and family
(Mishara 1995), and to community with imitative suicides (Stack 1987). Among the countries in the world,
South Korea has been one of the top nations having high suicide rate. From 2003, suicide rate of South
Korea started to exceed that of other countries. It was due to sharp increase of suicide rate of South after great financial crisis in 1998. In 2010, suicide rate increased about three times higher than that of 1990, in contrast to the suicide rate of Japan has been stabilized from 1998, relatively. Overall suicide rate of OECD countries has been decreased yearly (OECD 2013), as stated in .
< Figure 1: Trends of suicide rate of South Korea, Japan and average of OECD countries>
Diverged suicide rate within country can be considered as severe issue in South Korea. In 1992, gaps among age group specific suicide rate were almost similar. Although there were steep increase in 1998, uprising trend was occurred in all of the age groups and stared to decrease soon. However, after the beginning of the 21st century, elderly suicide rate increased so sharply than younger age groups. The gap among suicide rate became more divergent until 2005. After then, highly increased elderly suicide rate was relented. But the gaps between elder and younger age groups is still lager than others. Also, according to the mortality data, the highest suicide rate by county (administrative units, similar to counties in the U.S.) was about 7.1 times higher than the lowest one in 2010. What is worse, elderly suicide rate shows more severely differentiated gap. The highest suicide rate by administrative units was about 14.2 times higher than the lowest one. It indicates that huge gap between regions really exist in South Korea. Some regions show more severe mortality, especially in suicide rate. We can expect that some risk factors exacerbate inequalities of mortality and morbidity in South Korea (Kang 2013, Wilkinson 1997).
140
120 age 0 - 9 age 10 - 19 100 age 20 - 29 80 age 30 - 39 60 age 40 - 49
40 age 50 - 59 age 60 - 69 20 age 70 - 79 0
age 80+
2010
2008
2007
2006
2005
2004
2003
2002
2001
2000
1999
1998
1997
1996
1995
1994
1993 2009 1992
2
< Figure 2: Pattern of Suicide rate by Age groups 1992-2010>1
Despite of sharply increased suicide rate in overall age groups, researches about suicide rate of various subgroups (age, gender, class and so on) were no conducted frequently. A large proportion of researches about suicide in South Korea usually focused on suicidal ideation and other psychological factors based on small survey (Choi, Kim and Suh 2009, 2011, Kwon, Um and Kim-Yu-Jin 2012, Lee and Lee
2009, Lee and Ha 2011, Yook et al. 2011). Partly, lack of studies about suicide caused by several limitations of data about suicide because it is hard to be collected by researchers (Park 2010) and unstable official records (Pescosolido and Mendelsohn 1986) . Even though some scholars tried to invest suicide rate as the unit of analysis, they just roughly utilized province level (similar to state level data in the U.S.) or overall suicide rate using only single year, despite this country is suffering from steep increase in completed suicide, diverged by age (Choi 2011, JeoungHee 2011, Kim and Kwon 2013, Kim and Kim 2011, Kim 2010, Shin
2007, Song et al. 2013). As a matter of fact, social factors do affect individual’s life, both positively and negatively. But social pressures don’t be applied to everyone in different social groups (Bruce et al. 2012,
Pescosolido and Mendelsohn 1986, Ryu 2008, Wilkinson 1997). Therefore, my main research questions in this study are like following ones: What kinds of social influential factors to differentials of suicide rate by age group, gender, class? Can these differentials of suicide rate contribute overall understanding to severe social disorder in current South Korea?
2. Literature Review
Without a doubt, the most famous and powerful sociological study of suicide is Durkheim's Suicide:
A Study of Sociology. Durkheim argued that suicide rate is a social fact and it should be included in the research area of sociology. Overall, the lack of social integration leads risk of committing suicide because
1 “ou e : Mo tality “tatisti s Natio al Offi e of “tatisti s of Ko ea,
3 of losing social regulation leads to anomic status which is pervasive in modern society (Durkheim 1951).
Although individuals choose to commit suicide with their decision, Durkheim considered suicide as a social phenomenon because suicide happens when society is not integrated to a proper degree. In other words, suicide should be understood as a social fact because it happens when society does not work well.
Durkheim’s perspective became basic theory for investigating suicide in Sociology (Wu 2010). Although many sociologists have tried to overcome Durkheim's study, they were considered as just adding virtually nothing of significance to Durkheim's theory (Evans and Farberow 2003). A number of studies about suicide still follow Durkheim’s perspective about social integration and related influential factors. But it is hard to deny that Durkheim is the most important scholar in the field of suicide studies. Although various studies on suicide has utilized different methods (qualitative or quantitative), suicide rate was considered to be affected by social isolation, aging, family security, human capital, living condition.
Social Isolation and Aging
One of the core of Durkheim’s theory of suicide, less social integration treated as negative factor of sociology. Sainsbury (1955) argued that high mobility of residences (moving in and out in the area) and increased anonymous raise suicide rate of the area because of increased social isolation. In conventional society, there were lesser move in and out without historical, natural events. However, in current urban areas, there are high magnitude of mobility which is associated with lesser bonds with residences. Therefore,
Marris (1969) argued that the center of urban area has low cohesion and social network than suburban areas that decrease the regulation with lesser social relationship contacts. In their perspective, relatively low population density is positively associated with higher suicide rate due to level of isolation.
Currently, ‘Aging’ is becoming more and more important issue in South Korea. 11% of total population is consisted with old population over 65 years old, which means South Korea is still in ‘Aging
Society’. However, the aging population is growing so fast, with world widely low fertility rate. If the aging population grows so fast, South Korea will be the member of ‘aged country’ in the world in 2018 (Kim and
4
Kim 2011). It is the fastest transition from ‘Aging society’ to ‘Aged society’ in the world because it only took 18 years to be have doubled old population. Furthermore, the transition from ‘Aged Society’ to ‘Post-
Aged Society’ is projected to take only 8 years, which is also the rapidest speed in the world. Therefore, rapid aging will be severe social phenomena in the foreseeable future because it is hard mission to prepare impacts of growing aged population in just short time (Bae and Park 2006)
Although South Korea is facing to have larger proportion of aged people, current social welfare system for the elderly in Korea cannot be considered as well organized one. According to the OECD welfare expenditure statistics, South Korea has one of the lowest percentage of welfare expenditure compared to
GDP (OECD 2013) . Due to expanded life expectancy of population in South Korea, the elderly usually live longer than before. The average of life expectancy in 2012 is 80 (Finance 2013) . However, the mandatory retirement age is 58~60, which means people should survive without formal earnings about 20 years. Since South Korea is close to slack market, usually supply of workers is much higher than demand of workers, re-entering regular occupation are is not easy mission for the elderly. Also, social expenditure related to the elderly such as health insurance system and long-term care was largely increased. However, combined with changing family system from extended to nuclear one, care for the elderly by family has been decreasing in fast speed, hurting elderly’s stability of life (Choe and Jang 2009).
Family Security
Suicide rate can be accelerated by dramatic social change (Matt, Cynthia and Bernice 2011). How such changes made some groups more vulnerable than others to self-destruction? Divorce and bereavement also can be considered as risk factors on family security. In South Korea, worsened family insecurity harmfully affect risk of suicide (Kang 2013, Ryu 2008). Crude divorce rate in rural area was the most harmful factor on elderly Suicide. Relatively, male elderly persons were more sensitively affected by regional family insecurity.
Human Capital
5
Usually, education level plays important role on person’s whole life. It is generally accepted that education level has negative relations with mortality. A Higher level of education can be a predictor of lower mortality of the region. Some argue that education is the optimal measure because it is determined by early days in someone’s life. However, it is also very important for future income, information gathering, social network and health behavior (Hummer, Rogers and Eberstein 1982). Also, higher level of education can be related to access of good quality of life by using better goods, services, housings originated from higher quality of education (Walker 2009). In contrast, lower intelligence and education can be a risk factor of mental illness and suicide. As many human capitalists argue that, each individual’s capacity is originated from person’s education. Their accumulated investment, like schooling, can be signal of well-prepared candidates for good job positions in the future. Therefore, low education level would be risk factor of suicide. According to recent researches about suicide prevention, the elderly doesn't go to see psychiatric specialists or counselors to prevent their suicide. Furthermore, they tend to go to see primary stage of medical institution than advanced stage of medical institution that can provide specialized service of suicide prevention (Yook et al. 2011). Although primary stage of medical institution cannot afford to provide enough support to prevent suicide, more accessible medical institution and welfare facilities would be an effective buffer of suicidal risks.
Living Condition
Being worsened of inequality in health and well-being is a big issue in modern societies. Also, this issue is deeply related to class issue. Especially, mortality in developed countries is affected more by relative living standards than absolute living standards such as income equality and economic insecurity
(Wilkinson 1997). If the welfare system is not supportive enough for vulnerable groups, informal kinship support is important to protect them from dangerous events in life. Generally, socioeconomic status, economic and family insecurity and related mental illnesses are treated as risk factors for suicide (Atchley
1994, Chuang and Huang 1997, Pampel and Williamson 2001, Quadagno 2002). Because of worsened
6 insecurity in family relationship and economic status, they are more likely to be in worse mental and physical health results such as suicidal ideation and attempt (Gage 1971, Gunnell et al. 2003).
In many studies, residential classification such as urban and rural area can be considered as influential variables. There are various debates about significance of urbanism or level of development. In the early stages of urbanization, that region would have very high suicide rate influenced by social disruptions (Stack 1990). But better social infrastructure such as better environments has been reported as the factor of reducing regional suicide rate. Regional specific characteristics can be applied in sociological studies about suicide (Kowalski, Faupel and Starr 1987). In the cities, the better infrastructures are provided which makes people live in the better living conditions. In many scholarly works about suicide, authors just utilized secondary data such as survey that has small number of cases. Also, they just limited their issue in micro, psychological perspectives like suicidal ideation and opinions about suicides.
Although there are few academic researches dealt with completed suicide, only used 2~3 variables(Ryu 2008) or applied simple correlation (Eun 2005) or used single cross-sectional study by using non-age & gender specific suicide rate by area (Kang 2013). Furthermore, due to lack of consensus of standard of defining urban, metropolitan and rural area, most of studies just divided areas by name of counties. If the counties have ěSiĜ and ěGuĜ as ending letter of name, they considered them as urban because that indicates city area. ěGoonĜ is considered as rural area. Those simple classification has limitations because some city areas doesnĜt have better living condition than rural area. Therefore, my analysis using repeated age & gender specified suicide rate with multiple regional indicators would be useful resource for suicide prevention policies with more detailed understandings about current trend of suicide in South Korea.
3. Data and Method
7
To conduct empirical examination of suicide rate, county level mortality statistics released by
National Statistics Office of South Korea can be utilized to test the relationship between macro socioeconomic statistics and suicide rate by age group in South Korea. Controlling effects of 2005 and
2010 will contribute to find time-invariant regional indicators affecting suicide rates (Judith and
Melinda 2011) .
Dependent Variable
It is effective way to find regional divergences by statistics of competed suicide rate in South
Korea. Key dependent variable of this study is the rate of total suicide per 100,000 in 2005 and 2010. I calculated regional total suicide rate as number of suicides divided by number of population and standardized as the rate per 100,000. Total suicide rate as well as age, gender specific suicide rate by county could be calculated by Mortality statistics data containing causes of death produced by National
Statistics Office. This data is collected by vital registration and official record of individuals along with life courses as death. Population data was calculated from ‘Number of registered population’ data collected by Ministry of Public Administration and Security, population of each area by age (NSO
2013) . Specific suicide rate was calculated like below. For gender and age group specific suicide rates, number of persons and committed suicides counted only within that specific gender and age groups, to prevent distortion of calculated suicide rate by population composition in that county.
Suicide rate: � � � � � ℎ � � � ℎ � ∗ , Also, I invested more about age specific suicide rate by dividing age groups as 1) 20 ~ 39 years old (education and early career period, 2) 40 ~ 59 years old (middle age), 3) 60 years old and older
(old age) by life course in South Korea. (Choi 2011). Major early events in life usually occurs in education and early career period (having academic degree or occupational training, starting career,
8 marriage, childbirth and so on). Especially for male, they are more likely to have higher risk of dying in middle and old age
Independent Variable
Korea Social Science Data Archive covering regional statistics from 2005 and 2011 (KOSSDA
2013) was released in 2013. Based on the dataset,they utilized and re-calculated with many sources of data form government agencies such as ‘Population Survey (Census)’, ‘Labor Demand Survey’,
‘Present Situation of Social Facilities’. Units of analysis are 251 areas of South Korea. However, due to lack of data, counties of BookJeju and Namjeju excluded from analysis. Also, due to changed boundary of range of the counties, several counties united to the consistency of analysis (Dongnam-gu and Subook-gu into Chenan-city, Echang-gu and Sungsan-gu into Changwon-city). Although size of basic units of analysis (areas) are much smaller than counties of U.S.A., it helps us to collect aggregated death record data from individuals. It is almost impossible to find information of deaths, especially by suicides. By using regional statistics, suicide rate of each region and related predictors by official records and statistics. Naturally, there are a lot of limitations of community level data such as ecological fallacy problems, unreliably reported dataset and banned to be opened to public.
For the predictor of social isolation and population mobility (Maris 1969, Sainsbury 1955), I utilized population density and population change rate. Most quantitative studies about suicide in Korea simply sorted counties by administrative names because she doesn’t have standard to divide areas.
Therefore, studies just classified with two (urban – rural) (Kang 2013)or three (metropolitan city, small city, rural) places (Ryu 2008). In contrast, I used quartile of population density to reduce the effect of weighted population. Percentage of population change is calculated as absolute value of ration between population mobility out of the population in previous period. So it implies ratio of mobility within the area, relative level of anonymous and density of social network. Also, as being old can be considered
9 as significant risk factor of suicide in Korea (Bae and Park 2006, Gage 1971, Lee 2006, Lee and Lee
2009, Ryu 2008), proportion of the elderly in county can be used as measure of aging.
For the predictor of family security, I picked up three major components of demographic variables, crude birth, marriage and divorce rate. Most of all, crude divorce rate has been considered as an influential community level factor to suicide rate (Bengtson et al. 2000, Choi, Kim and Suh 2009,
Gunnell et al. 2003, Kang 2013, Kim-Yu-Jin 2011, Mishara 1995, Ryu 2008, Stack 1990). As well as level of divorce, marriage and birth could be considered as influential factor to predict the level of suicide rate, by increasing cohesion with partners and motivation of life (Gunnell et al. 2003). Figures in crude measures implies occurrence of the event per 1,000 persons. Also, as a proxy indicator of economic security and human capital, education level was utilized as independent variable of this research.
As the indicator of standard of living in the county, I utilized average percentage of installed social infrastructure (average of percentage of paved road, water supply and disposal systems), fiscal independency ratio (percentage of independent budget of the county out of the total amount of budget), and number of beds per 1,000 persons as the indicator of medical facilities. Usually, having low level of social infrastructure and medical facilities increases the likelihood of dying compared with urban areas have relatively better living conditions (Chuang and Huang 1997, Gan et al. 2009, Kim and Kim
2011, Kowalski, Faupel and Starr 1987, Pampel and Williamson 2001, Park 2010, Ryu 2008)
As the method of analysis, I used least square multiple regression model to find the relationship between the predicted suicide rates of the county by the combination of social isolation, population mobility, aging (Model 1), family security (Model 2), human capital (Model 3), and living condition
(Model 4). I will also try to find the association of those factors to see the relationship between the factors to predict suicide rate of the county. In the last table analysis, I modified the full model (Model
10
5) with all of the age groups (20-39 years old, 40-59 years old, 60 years old and over) by gender
comparing 2005 and 2010.
4. Results
Overall trends of descriptive statistics
Overall, suicide rate of South Korea has been increased from 2005 to 2010 in every age groups as we can see in . In 2005, average of total suicide rate was 27.20. But in 2010, it was increased about 10 per 100,000. Although we could not find great increase of suicide rate in , we would expect worsened risk of suicide in 2010, compared with 2005. Also, compared with female, male has had higher suicide rate in every age groups. It may imply the higher risk of suicidal deaths among males in
South Korea than females. As an indicator of mobility, percentage of population change was decreased in
2010. However, increased standard deviation can imply more activated population changes within counties in South Korea.
Average crude divorce rate was a little bit decreased in 2010, but crude birth and marriage rate was increased, implying better family security in 2010 than 2005. As time goes by, the general education level
(16% -> 17.94%) , average percent of social infra-structure (75.62% -> 81.82), number of hospital beds per
1,000 persons (8.27 ->11.27). However, the fiscal independency ratio was decreased (32.43% -> 29.79%), implies lesser independency and capacity of welfare policies for the people in that county.
Also, as we can see in , and , the distribution of suicide rate was worsened in 2010 compared with 2005. In general, there were smaller number of areas which had darker color in the map of 2005, indicates lower suicide rate. However, in all of maps of the total, male and female suicide there are increased numbers of countries have darker color in 2010. Especially, eastern and southern
11 areas are more likely to have higher suicide rate than western and middle area. It may imply uneven development between more populated area (Seoul metropolitan city and other specialized cities) and less developed are with lower population density by rural to urban migration.
< Table 1: Descriptive statistics of independent and dependent variables 2005-2010 2>
2005 2010 Mean (standard deviation) Mean (standard deviation) 4327.92 (6576.00) 4385.04 (6616.37) Population density3 17.61 (8.16) 20.22 (8.70) Proportion of people older than 60 1.77 (1.73) 1.55 (1.96) Percentage of population change 8.23 (1.99) 8.73 (2.21) Crude Birth Rate 5.79 (1.12) 5.95 (1.20) Crude Marriage Rate 2.45 (0.54) 2.25 (0.40) Crude Divorce Rate 16.00 (10.17) 17.94 (10.13) Percentage of college educated or above 75.62 (20.22) 81.82 (16.02) Average percent of social infrastructure4 32.43 (18.84) 29.79 (17.68) Fiscal independency ratio 8.27 (5.16) 11.27 (17.68) Number of hospital beds per 1,000 27.20 (8.19) 31.30 (7.91) Suicide rate per 100,0005 40.00 (17.42) 49.02 (18.86) Male’s Suicide rate per 100,000 (total) 27.18 (17.67) 34.58 (18.51) Male’s Suicide rate per 100,000 (20-39) 51.82 (24.91) 57.40 (23.87) Male’s Suicide rate per 100,000 (40-59) 111.50 (51.15) 115.13 (48.58) Male’s Suicide rate per 100,000 (60+)
2 Cal ulated y atio le el. I y a alysis, those statisti s ill e di ided y ou ty le el as o u ity le el data. P o ided y Ko ea “o ial “ ie e Data A hi e KO““DA. . Regional Statistics of South Korea 5- . “eoul: Ko ea “o ial “ ie e Data A hi e. 3 pe so /K s ua e 4 A e age of pe e tages of pa ed oad, ate supply a d disposal syste s y ou ty 5 Total sui ide ate pe , . 12
18.58 (9.07) 24.23 (10.45) Fe ale’s “ui ide ate pe , (total) 16.99 (11.60) 23.98 (17.76) Fe ale’s Suicide rate per 100,000 (20-39) 17.11 (14.48) 22.08 (14.84) Fe ale’s Suicide rate per 100,000 (40-59) 41.83 (24.95) 44.88 (23.32) Fe ale’s Suicide rate per 100,000 (60+)
< Figure 3: Total suicide rate in 2005 and 2010 in South Korea>
13
< Figure : Males’ suicide rate in 2005 and 2010 in South Korea>
< Figure : Fe ales’ suicide rate i a d i South Korea>
14
presents the result of multiple regression model for total suicide rate of counties in South
Korea, 2005 and 2010. In the baseline model (Model 1), living in more populated counties were statistically significantly associated with lower suicide rate. This is opposite direction with classical studies of urban suicide in western countries (Maris 1969, Sainsbury 1955) due to less social integrity. However, in the context of South Korea, more populated countries means more developed areas, such as metropolitan areas in specified regions. Therefore, it can be the support of more social network and better living environment than less populated countries, which are more likely to be less developed area. In contrasts, proportion of elderly and population mobility was not significant. It may imply there are no significant differences of total suicide rate associated with differences of aging and population mobility in both of 2005 and 2010. In model 2, as a number of studies already invested (Judith and Melinda 2011, Kang 2013, Ryu 2008, Stack
1990), crude divorce rate was positively associated with suicide rate per 100,000 persons. Interestingly, crude birth rate as a conditions of cohesion and motivations in family security (Choe and Jang 2009), was negatively associated with total suicide rate. However, crude marriage rate as active marital behaviors was not negatively associated with total suicide rate in 2010. In 2005, crude marriage rate was positively associated with total suicide rate, which means more active county in marital behaviors has higher suicide total rates. It can imply unevenness and increased inequality in marriage market can be a risk factor of suicide in total suicide rate.
In model 3, having more college educated people in the county was negatively associated with total suicide rate, implies higher human capital is associated with lower suicide rate in counties. It is consistent with previous studies that more socio-economic resources by human capital can be predictor of suicide
(Choi 2011, Eun 2005, Innamorati et al. 2009). In model 4, I invested living condition with baseline model.
In 2005, living condition was not statistically significant with total suicide rate. However, in 2010, having better living condition as higher independence of budget in county and better social infrastructure was associated with lower suicide rate, marginally. In the full model (model 5), living in more populated counties, higher crude birth rate were associated with lower total suicide rate. However, higher crude
15 divorce rate was associated with higher suicide rate in both of 2005 and 2010. Number of beds, proportion of elderly, crude marriage rate were marginally significant in either of 2005 or 2010. Therefore, for total suicide rate, living in more populated city after controlling other variables (human capital, living condition and family security) was still statistically significant, which implies unevenness of developmental stages among more and less populated areas. Also, having more family security was negatively associated with total suicide rate. After controlling all of other variables, living condition and human capital was not statistically significantly associated with total suicide rate.
Crude birth rate -1.05* -0.69+ -1.39** -0.85* Crude marriage rate 1.27+ 0.74 1.64* 0.97 Crude divorce rate 6.04*** 5.48*** 4.63*** 4.30** Proportion of more than college - degree -0.35*** 0.30*** -0.13 -0.1 fiscal self-reliance ratio -0.04 -0.07+ -0.03 -0.05 Average percent of installed social infra -0.07 -0.11+ -0.08 -0.11+ Number of hospital bed per 1,000 0.12 0.04 0.05 0.00 Constant 34.49*** 33.25*** 20.66** 24.30*** 42.29*** 40.69*** 43.06*** 46.68*** 38.66*** 42.76***
presents the regression model for males’ total suicide rate of countries in 2005 and 2010.
Similar to the models of total suicide rate, living in more populated countries such as metropolitan area was associated with lower suicide rate in 2005 and 2010, except of 2nd quartile of population density in 2010.
In addition, more aged persons was positively and significantly associated with higher suicide rate. This result can be support of the current highest high suicide rate of elderly, especially for males’ suicidal deaths
(Gan et al. 2009, Innamorati et al. 2009, Yook et al. 2011). Interestingly, more active population mobility in countries was significantly associated with lower suicide rate in 2005, other than 2010. It may imply more active areas has motivations of mobility has lower suicide rate, in contrast to the classical theories of suicide rate of urban areas (Maris 1969, Sainsbury 1955). In model 2, after controlling variables in baseline model, crude marriage rate and crude divorce rate were statistically significantly associated with higher suicide rate of male in the counties in 2006. Crude marriage rate was not significant in 2010. Also, crude birth rate was associated with lower males’ suicide rate in 2005, other than 2010. Therefore, family security model was not that statistically significant in 2010 except crude divorce rate. In model 3, similar to the effect of human capital to total suicide rate, proportion of college educated persons was negatively associated with males’ suicide rate in both of 2005 and 2010. It can imply higher level of education could be positive factor of males’ lower level of suicide rate of the counties. In model 4, living conditions were not statistically significantly related with males’ suicide rate. In the full model (model 5), living in more populated area was still statistically significant after controlling family security, human capital and living condition. Crude birth rate (negatively), crude marriage and divorce rate (positively) were significantly associated with males’ suicide rate in 2005. However, in 2010, only more aging level was statistically significant after controlling other variables. Although it was not statistically significant, the coefficient of aging level (proportion of elderly) in 2005 had opposite sign, compared with that of 2010. It may imply multicollinearity between aging level and other variables, potentially population density and human capital, living standard variables because of concentrated population of elderly in less developed areas, such as rural areas.
Crude birth rate -2.55** -0.88 -2.91** -1.03 Crude marriage rate 2.67* 1.29 3.23* 1.75 Crude divorce rate 9.67*** 7.89** 8.44*** 4.89 Proportion of more than college degree -0.50*** -0.47*** -0.09 -0.24 fiscal self-reliance ratio -0.06 -0.09 -0.07 -0.05 Average percent of installed social infra -0.05 -0.04 -0.1 -0.02 Number of hospital bed per 1,000 0.21 0.12 0.1 0.08 Constant 45.38*** 40.64*** 30.45* 24.42* 56.45*** 52.33*** 52.59*** 49.07*** 49.06* 40.28*
presents the influential factors on females’ suicide rate of counties in 2005 and 2010.
Interestingly, the sign of coefficients of population density was opposite, compared with 2005 and 2010. It may also implies probability of multicolinearity between population density, proportion of elderly and population mobility. In 2010, living in more populated area was negatively associated with lower suicide rate. However, in 2005, living in 2nd quartile of population density had higher suicide rate compared to countries with bottom 25% of population density. In model 2, crude birth rate (negatively) and crude divorce rate (positively) were statistically significant, after controlling other variables. Crude marriage rate was not statistically significantly associated with females’ suicide rate. As other models in total and males’ suicide rate, human capital was negatively associated with suicide rate of females. In model 4, only average level of installed social-infrastructure was marginally significant with suicide rate (negatively). In the full model, crude birth rate, average percentage of social-infrastructure and number of hospital beds per 1,000 were statistically significant either in 2005 or 2010.
Finally,
presents age-group specific suicide rates by gender. Originally, education and early career age group (20-39) was included in the models. However, all of the indicators were not statistically significant in both of the male and female, in 2005 and 2010. Therefore, I just included full models of middle age (40-59) and old age (60 years old and over) in 2005 and 2010. In 2005, living in more populated area, more population mobility were negatively associated with suicide rate of male in middle age. Crude divorce rate was significantly associated with higher suicide rate. In 2010, living in more populated area, having higher crude birth rate, higher education level were associated with lower suicide rate of male in middle age. Interestingly, instead of crude divorce rate, crude marriage rate was statistically significantly associated with higher suicide rate of male in middle age in 2010. It can imply noise and unexpected relationship between crude divorce and marriage rate in 2010. In contrast, indicators of suicide rate of female were not statistically significant in both of 2005 and 2010. In this result, I expect social- macro indicators are not effective to predict suicide rate of female in middle ages of the counties in South
In old age, the age group has highest suicide rate in both of male and female, showed similar associations between predictors and suicide rate. At first, living in more populated countries, more aging level were statistically significant, controlling other variables in other models. It can imply the macro factors related to living areas with more density of social network can lower the suicide rate of the male elderly, which is one of the most vulnerable group in current South Korea. Usually, aging level as the proportion of elderly was positively relative with suicide rate, which means more aged people in the counties. However, more elderly in the counties after controlling other variables were statistically significantly associated with lower suicide rate of male elderly. It may imply more opportunities of social network with homogeneous age group can be beneficial effect to male elderly. Crude birth rate, average percent of social-infrastructure
(negatively) and crude divorce rate (positively) were significant in 2005, but not significant in 2010. Similar to male elderly, living with more homogenously aged population was negatively associated with lower suicide rate of elderly females. Also, living in better living condition (average percentage of installed social- infrastructure) was negatively associated with suicide rate of the female elderly. In both of 2005 and 2010, crude divorce rate was positively associated with suicide rate. Therefore, I could find strong support of family security in most of the age groups in various models with different age-specific suicide rate.
5. Discussion
Recently, suicide is one of the top leading causes of death in middle and old ages, especially for old males (Kim-Yu-Jin 2011, Kim and Kim 2011, Lee 2006, NSO 2013). Along with sharply increased total suicide rate of South Korea within last decades (OECD 2013), more diverged gap of suicide rates by age groups and gender (Park 2010, Ryu 2008) need to be studies. Because it means some groups are suffering from micro and macro risk factors differently. However, due to lack of qualified dataset of individuals committed suicides, most of the studies concentrate of psychological factors to risk behaviors such as suicidal ideation and attempts (Choi, Kim and Suh 2009, Gan et al. 2009, Innamorati et al. 2009,
Lee and Lee 2009, Walker 2009). Even though some scholars tried to conduct researches by using
23 community level data (with completed suicidal deaths by official records), they just used few number of variables (Ryu 2008), a single year with only suicide rate (Kang 2013), province (similar to state in the
U.S.) level of analysis (Kim and Kim 2011), and simple correlation (Eun 2005) to find the relationship between social factors and suicide rate in current South Korea. Compared with previous studies, I differentiated age-group specific suicide rate with more diverse categories of indicators. By community level statistics as indicators, I expected to find general not only the trend of total suicide rate but also that of age-group specific suicide rate of counties in 2005 and 2010.
I have invested to find influential social factors to suicide rate of counties in South Korea in 2005 and 2010. As various studies stated, less social integration with higher social isolation (Durkheim 1951,
Kwon, Um and Kim-Yu-Jin 2012, Maris 1969, Park 2010, Sainsbury 1955, Walker 2009), less secure family (Bengtson et al. 2000, Judith and Melinda 2011, Kang 2013, Mishara 1995, Ryu 2008, Stack 1990), living in less developed area with less human capital and living standard (Bruce et al. 2012, Chuang and
Huang 1997, Kim and Kwon 2013, Kim and Kim 2011, Lee and Ha 2011, Pritchard 1967) were positively associated with higher suicide rate, which means they are risk factor of suicide rate of the counties.
Interestingly, crude marriage rate was more likely to be positively associated with suicide rate, which means that is a kind of risk factor of suicide rate.
Especially for elderly, suicide rate has been increased like a skyrocket from 2000 to 2010. Although the uprising trend was stalled after 2005, it started to arise from 2009. Still, in contrast to some western countries (Gunnell et al. 2003), suicide rate of the elderly is the highest in South Korea. As South Korea’s social welfare systems and policies for the elderly are not organized well, the elderly lost their sources of support such as devaluated asset and informal kin support for their adult children and relatives (Lee 2006).
In contrast to other age groups, more proportion of elderly people in that county was negatively associated with the suicide rate of the elderly in both of the genders. It may imply more homogeneous age group after controlling other social factors can be benefit to elderly persons to prevent higher suicide rate compared to other counties with less proportion of elderly, which can imply higher social isolation level in those counties.
24
Although this study could contribute understanding of recent trend of suicide rate (completed suicides) in South Korea, there are several limitations. First of all, as Durkheim was criticized by other scholars by his study of suicide by regional statistics, this study also have high risk of ecological fallacy. I need to use the countries as the basic unit of analysis, assuming the homogeneity of individuals in the area.
Because county level is the smallest unit of analysis provided by Korean government and institutions in
South Korea. However, as counties are still big units to assume same characteristic of individuals, it is challenging one to overcome the possibility of ecological fallacy.
Moreover, as I used counties as unit of analysis, relative effect of counties cannot be same naturally.
For example, some more populated counties may affect more into total suicide rate in national level, other than less populated counties. Although I applied categorical variable to the models, I still need to find some other ways to reduce the effect of weights of population. Also, as some counties has relatively sensitive number of age-group specific suicide rate (ex. Female suicide rate in elderly), it may distort the relationship in the regression analysis, because I just used two years in my research. However, in my next research, if I add and pool more waves from 2006 to 2009, I expect to have more stable result than I did in this research, with more affluent number of cases.
25
< References >
Atchley, Robert C. 1994. Social Forces & Aging: An Introduction to Social Gerontology: Wadworth Publishing Company. Bae, Ji-Yeon and Hyun-Sik Park. 2006. "An Exploratory Study on the Elderly Risk Behaviors." Journal of the Korean Gerontological Society 26(4):785-800. Bengtson, Vern L. , Kyong-Dong Kim, George C. Myers and Ki-Soo Eun. 2000. Aging in East and West: Families, States, and the Elderly. New Yrok: Springer Publishing Company. Bruce, Western, Deirdre Bloome, Benjamin Sosnaud and Laura Tach. 2012. "Economic Insecurity and Social Stratification." Annual Review of Sociology 38:341-59. Choe, Jung-Moon and Sang-Hee Jang. 2009. Families in Contemporary Society. Seoul: Acanet. Choi, In, Young-Sook Kim and Kyung-Hyun Suh. 2009. "The Psychological Factors Affecting Suicidal Ideation of Elderly." HEALTH AND SOCIAL SCIENCE 25. Choi, Yong-Hwhan. 2011. "Socio-Economic Analysis of Suicide." Issue and Diagnosis 3:1-25. Chuang, Hwei-Lin and Wei-Chiao Huang. 1997. "Economic and Social Correlates of Regional Suicide Rates: A Pooled Cross-Section and Time-Series Analysis." Journal of Socio-Economics 26(3):277-89. Durkheim, Emile. 1951. Suicide: A Study in Sociology. London: McMillan. Eun, Ki-Soo. 2005. "Socioeconomic Determinants of Suicide Rate in South Korea." Korean Journal of Population Studies 28(2):97-129. Evans, Glen and Norman L. Farberow. 2003. The Encyclopedia of Suicide: Second Edition: Facts on File Inc. Finance, Ministry of Strategy and. 2013. Economic and Environments Report. Seoul: Ministry of Strategy and Finance of Korea. Gage, Frances Boland. 1971. "Suicide in the Aged." The American Journal of Nursing 71(11):2153-55. Gan, Sam-Sung, Wang-Keun Yoo, Sung-Wook Kang and Young-Gyu Phee. 2009. "A Study of the Determinants of Suicidal Ideation among the Elderly in Korea." Health and Social Welfare Review 29(1). Gunnell, David, Nicos Middleton, Elise Whitley, Daniel Dorling and Stephen Franke. 2003. "Why Are Suicide Rates Rising in Young Men but Falling in the Elderly? A Time Series Analysis of Trends in England and Wales 1950-1998." Social Science & Medicine 53:595-661. Hummer, Robert A., Richard G. Rogers and Issac W. Eberstein. 1982. "Sociodemographic Differentials in Adult Mortality: A Review of Analytic Approaches." Population and Development Review 24(3):553-78. Innamorati, Marco, Antonio Tamburello, David Lester, Silvia Rigucci, Mario Amore, Cristina Di Vittorio, Paolo Girardi, Roberto Tatarelli and Maurizio Pompili. 2009. "Socioeconomic Risk Factors in the Precipitation of Suicide in the Elderly." The Open Geriatric Medicine Journal 2:28-33. JeoungHee, Kim. 2011. "Concerning Suicide Rates in Korea: An Expected Role of Korean Scholarly World." HEALTH AND SOCIAL SCIENCE 30:5-28. Judith, K. Hellerstein and Sandler Morrill Melinda. 2011. "Booms, Busts, and Divorce." The B.E. Journal of Economic Analysis & Policy 11(1). Kang, Eun-Jeong 2013. "Compositional and Contextual Factors Related to Area Differentials in Suicide." Korean Journal of Health Education Promotions 30(1):41-52. Kim-Yu-Jin. 2011. "Moderating Effects of Life Problems, Social Support on Relationship between Depression and Suicidal Ideation of Older People." Health and Social Welfare Review 31(4). Kim, Hyoung-Soo and Lee-Kyung Kwon. 2013. "Relationship between Elderly Suicide Rates and Socio- Economic Factors in Korea: Centering around the Trend of Changes in 1990-2010." The Journal of the Korea Contents Association 13(6):236-45. doi: 10.5392/jkca.2013.13.06.236. Kim, Ki-Won and Han-Gon Kim. 2011. "Macro Analysis of Factors Impacting on the Elderly's Suicide Rates in the Republic of Korea." Korean Journal of Population Studies 34(3):31-54. Kim, Wang-bae. 2010. "Suicide and the Breakdown of Society." Korean Studies Quarterly 33(2):195-224. 26
KOSSDA. 2013. Regional Statistics of South Korea 2005-2010. Seoul: Korea Social Science Data Archive. Kowalski, Gregory S., Charles E. Faupel and Paul D. Starr. 1987. "Urbanism and Suicide: A Study of American Countries." Social Forces 66(1):85-101. Kwon, Jung-Don, Tae-Young Um and Kim-Yu-Jin. 2012. "A Study on Factors Affeecting Suicide Risk among Seniors Who Live Alone : The Effect of Loneliness and the Mediating Effect of Hopelessness." Health and Social Welfare Review 32(1):89-144. Lee, Cheol-Woo. 2006. The Aging and Elderly Policy in Korean Society. Seoul: Korean Academic Research Information. Lee, So-Jung and Soo-Hyuong Lee. 2009. "A Study of Elderly Suicide Preventive Measures." HEALTH·WELFARE POLICYFORUM:32-42. Lee, Won-Cheol and Jae-Hyeok Ha. 2011. "The Association between Nonstandard Employment and Suicidal Ideation: Data from the First~Forth Korea National Health and Nutrition Examination Surveys." Korean Journal of Occupation, Environment and Medicine 23(1). Maris, Ronald W. 1969. Social Forces in Urban Suicide. Dorsey: Homewood. Matt, Wray, Colen Cynthia and Pescosolido Bernice. 2011. "The Sociology of Suicide." Annual Review of Sociology 37:1-24. Mishara, Brian L. 1995. Beyond Suicide: The Impact on Family and Friends, Helpers, Indiivduals and Society, Edited by B. L. Mishara. New York: Springer. NSO. 2013. Mortality Statistics of South Korea: National Statistics Office. OECD. 2013. Oecd Factbook: OECD. Pampel, Fred C. and John B. Williamson. 2001. "Age Patterns of Suicide and Homicide Mortality Rates in High-Income Nations." Social Forces 80(1):251-82. Park, Hyung-Min. 2010. Suicide, a Lesser Evil Choice. Seoul: Ihaksa publishing company. Pescosolido, Bernice and Robert Mendelsohn. 1986. "Social Causation or Social Construction of Suicide? An Investigation into the Social Organization of Official Records." American Sociological Review 51(1). Preti, Antonio. and Paola Miotto. 1999. "Suicide and Unemployment in Italy, 1982-1994." Journal of Epidemiology and Community Health 53(11):674-701. Pritchard, Colin. 1967. "The Economic Cycle and the Social Suicide Rate." American Sociological Review 32(3):457-62. Quadagno, Jill. 2002. Aging and the Life Course: Mcgraw-Hill Publishing Company. Ryu, Jung-Kyun 2008. "Regional Differntials in Elderly Suicide Rate: The Focusing on Effects on Family Instability, 1995-2005. ." Korean Journal of Population Studies 31(2):21-44. Sainsbury, P. . 1955. Suicide in London: An Ecological Study. London: Chanmp and Hall. Shin, Dong-Jun. 2007. "Suicide - Social Pathology of Disarrangement." Hwanghae Review:98-124. Song, Tae-Min, Jin-Dal-lae, Ju-Young Song, Ji-Young An and Young-Ja Cho. 2013. "A Comparative Analysis of Suicide Characteristics by Age Group in Korea." Journal of The Korea Society of Health Informatics and Statistics 38(1):122-42. Stack, Steven. 1987. "Celebrities and Suicide: A Taxonomy and Analysis, 1948-1983." American Sociological Review 52(3). Stack, Steven. 1990. "The Effect of Divorce on Suicide in Denmark, 1951-1980." The Sociological Quarterly 31(3):359-70. Walker, Jacob T. 2009. "County Level Suicide Rates and Social Integration : Urbanicity and Its Role in the Relationship." Sociological Spectrum 29:101-35. Wilkinson, Richard G. 1997. "Health Iinequalities: Relative or Absolute Material Standards?". British Medical Journal 22(314):591-95. Wu, Fei. 2010. Suicide and Justice: A Chinese Perspective: Routledge. Yook, Sung-Phil, Khwang-Hyun Choi, Eun-Joo Kim and He-Sun Lee. 2011. Elders in Crisis: Suicide Awareness, Prevention, and Intervention for the Elderly. Seoul: Hakji Publishing Company.
