Retirement Satisfaction and Life Satisfaction of Retirees

* This paper is an English translation of a paper originally released in Korean in the 27th issue of Monthly Labor Review published by the Korea Labor Institute in March 2007. Retirement Satisfaction and Life Satisfaction of Retirees Hyun-Goo Shin Senior Researcher Korea Labor Institute Email: shin@kli.re.kr May 2007 Ⅰ. Introduction The rapid aging of the population has brought consequences at different levels in Korea. Socially, the proportion of the old continues to rise; individually, the old-age period becomes longer, requiring a new approach to that particular period in one's life cycle. It was generally held that old age, accompanied by declining physical functions and deteriorating health, is a period of passivity, leaving one only to look back and tidy up the last strings of one's life. But there has been a growing call to take a more active view toward the old-age, as a period to start a second life. To determine the quality of life at old age requires not only indicators on the objective living environment, but also on the subjective perceptions that one has in such living environment. Any human being would desire to live a happy life, and to fulfill that desire necessitates both physical and mental/psychological elements. One good window into the subjective dimension of the quality of life is "life satisfaction." Researchers of quality of life acknowledge that the physical elements that make up a person's living environment are an important criteria for ensuring a higher quality of life. But they assert that there is another,

closely related dimension to quality of life, which is subjective perceptions of one's level of happiness or sense of value, and that such perceptions can be measured through life satisfaction indicators. In this context, this study aims to take a descriptive look into life satisfaction of retirees, using the data collected through the 1st basic survey of the Korean Longitudinal Study on Ageing (KLoSA). Their social/demographic characteristics and retirement satisfaction are used as main components. The reason why retirees have been selected as the subject of study on old-age life satisfaction is that retirement from the labor market is generally regarded as the start of old-age period. This study is structured as follows. First, in the quality of life study, the meaning of "quality of life" is investigated. The 1st KLoSA basic study dataset, the source of this study, is briefly introduced. Last, this dataset will be used to take a deeper look into the differences in life satisfaction by socio-demographic factors. Ⅱ. Life Satisfaction and Retirement Satisfaction Life satisfaction is one of important indicators to measure quality of life. "Quality of life" is a concept that comprises more than the objective conditions surrounding an individual or group. It places an equal emphasis on the subjective feelings that an individual or group experiences. (Tae-hyeon Kim et al., 1998). Other terminologies with similar meaning include well-being, subjective well-being, life satisfaction, or happiness (Hyeon-song Lee, 1997). At any rate, the concept of quality of life itself includes both the objective life conditions of an individual or a group, and the subjective feelings experienced by an individual or a group. This is why study on quality of life tended to take both the quantitative and qualitative approach. (Kyeong-sook Park, 2003). Quantitative approach measures the quality of life using different living

environment indicators for such elements as income, health, education, leisure, safety and environment. For example, literacy rate, average life expectancy, infant mortality rate or crime rate can be important indicators to quality of life. But the concept of quality of life includes a subjective dimension in addition to these objective indicators, for the living world is at once objective and subjective. Economic wealth or living conditions is no doubt an important element in determining an individual's quality of life, but subjective perceptions or interpretation of such objective conditions is as important an element. Related studies demonstrate that objective elements like economic wealth or living conditions are directly proportional to one's sense of happiness, only up to a certain level. Past that level, there is little association between objective conditions and subjective perceptions (Tae-hyeon Kim et al., 1998). A noteworthy trend of late is group-specific studies. As subjective judgement on one's life may exhibit different patterns depending on the individual or group characteristics, many of the recent studies have focused on indicators for particular groups, such as youth, the elderly, disabled or social minorities. (Gyeong-sook Park, 2003) From a traditional viewpoint, the old-age may be prone to lower sense of life satisfaction due to weakening biological functions, loss of social role and reducing social bonds. But as the society ages rapidly, old-age population grows and individuals' old-age period lengthens, interest in the elderly is growing both socially and individually. New perceptions challenge the traditional, asserting that more active approach must be taken regarding the old-age, rather than simply writing it off as a period of decline. In this context, subjective satisfaction on life will be an important indicator for the study on quality of life in the old age. Meanwhile, retirement is an important event that characterizes in an individual's later life. In fact, it is a turning point that requires a person has been employed so far to stop performing duties related to such employment and adapt to a new role. Particularly in Korea, retirement can be enormously traumatic to individuals, due to the relatively young retirement age, early non-voluntary retirement, lack of post-retirement

preparation, loss of social/economic role, reduced income, insufficient social welfare, and need for financial resources for children's education and wedding. It may be a major cause of decline in life satisfaction. According to Korean studies on retirees, retirement causes changes not only to the person's economic status, but also the physical, emotional and spiritual 1) status. Those who retire relatively early suffer in terms of adaptation to post-retirement life and life satisfaction in general. The study asserts that the stress caused by retirement has the biggest impact on these individuals' life satisfaction. (Yeon-gyeong Ji and Byeong-eun Cho, 1991; Hwa-yong Shin and Byeong-eun Cho, 2001; Hee-ja Han and Eun-shil Gang, 2001). It is thus reasonable to conclude that retirement is an important factor in determining life satisfaction in old age. In particular, retirement satisfaction was found to be much lower among retirees in Korea than in the U.S. Figure 1 shows the result of retirement satisfaction survey on retirees in Korea and the U.S. Among Korean respondents, 56% answered "Satisfied" (including "Very satisfied") and 44% "Not satisfied," the former slightly outweighing the latter. But only 8% indicated that post-retirement life was better, whereas nearly 55% answered that it was worse, the latter heavily outweighing the former. The results shows that 1 out of 2 retirees were less satisfied with post-retirement life than with pre-retirement. In contrast, 55% of American retirees were "Very satisfied" with retirement, and if including the "Satisfied," the number goes up to 91%. Only 9% indicated dissatisfaction, in good contrast with Korean respondents. 46% of the U.S. respondents preferred post-retirement, while only 19% indicated that life was worse after retirement. 1) By "spiritual" it is meant as an ultimate interest that leaves one to pursue something even after the physical, social and psychological needs have been fulfilled. It is generally measured through the person's level of interest in religious aspects. (Hee-ja Han and Eun-shil Gang, 2001).

Figure 1 Retirement Satisfaction Between Korean and U.S. Retirees 60 50 40 54.6 51.6 36.0 44.2 45.8 36.4 35.7 55.5 30 20 10 4.2 8.8 8.1 18.5 0 (%) Very satisfied Satisfied Not Satisfied Post-retirement is better No change Post-retirement is worse Korea U.S.(HRS) Source : KLoSA 1st Basic Survey (Korea), 2006. Health and Retirement of Study (U.S.), 2004. It is in this context that this study will assess, using the data from the KLoSA 1st Basic Survey, the social and demographic factors that impact the level of life satisfaction of retirees, with particular focus on the impact of retirement satisfaction on life satisfaction. Ⅲ. Data The data used in this study is from the 1st basic survey (wave 1) of the Korean Longitudinal Study of Ageing (KLoSA). The basic study for KLoSA was conducted from 2005 by the Korea Labor Institute (KLI) with funding by the Ministry of Labor. The 1st basic survey was launched in August 2006 and finished in December the same year. Middle/old-aged individuals aged 45 or over residing in ordinary households were the population for the 1st survey, and 10,254 individuals from 6,171 households across the nation not including islands were selected for the panel. 2)

The questionnaire included approximately 1,500 questions in seven categories of demographics, family, health, employment, income, assets, and subjective expectations and life satisfaction. Data was collected through personal interviews using computers (CAPI). (Hyun-Goo Shin, Kachung Boo, and Hye-jeong Lee, 2006; Hyun-Goo Shin and Hye-jeong Lee, 2006). The questionnaire covered the following categories that include the social and demographic factors associated with life satisfaction and retirement satisfaction that will be reviewed in this study. (See <Table 1>). <Table 1> Social and Demographic Factors of Life Satisfaction and Retirement Satisfaction Used in This Study Category Measurement factors Demographics Family Health Employment Subjective expectations and life satisfaction Gender, age, marital status, education, social activities Cohabitation with married children Subjective perceptions of health Retired or not, retirement age, retirement satisfaction Life satisfaction Questions regarding life satisfaction are 5 in total, asking about the respondent's level of satisfaction about his/her state of health, state of economy, relationship with spouse, relationship with children, and quality of life (sense of happiness), compared with peers. Responses include eleven 10-point score ranges, between 0 and 100. But in this study, life satisfaction was measured using only one out the 5 questions, the question on qualtiy of life (sense of happiness). The question is as follows. (See Figure 2 ) 2) Data from the 1st KLoSA basic survey will be distributed as a preliminary version in mid-march with only the minimum level of sorting. Relevant materials such as questionnaires and code books will be made available along with the data at the following website: http://www.klosa.re.kr or http://klosa.kli.re.kr. Information is made available for free, and is downloadable after user registration. Final version will be published in July after a more thorough review.

Figure 2 Question on Quality of Life in the 1st Basic Survey of KLoSA (G029) G029. Compared to your peers, what is your level of satisfaction on your quality of life (happiness) in general? [Interviewer: Please take a look at the response card and state your level of satisfaction. The closer it is to 0, the lower the satisfaction, and closer to 100, the higher the satisfaction.] 0 10 20 30 40 50 60 70 80 90 100 Incidentally, the KLoSA leaves the definition of "retirement" up to the individual respondents' judgement. Similar to the definition of the Korea Labor Panel Study (KLIPS), it generally indicates a state where the individual has stopped full-time income-earning activities and is currently not working or working only as a hobby, and barring any major changes, plans to work only as a hobby. Ⅳ. Results <Table 2> below shows employment status as indicated by the 1st KLoSA survey. Of the total 10,254 respondents, 3,848 or 38% were found to be working at the time of the survey, while 2,027 (20%) indicated themselves as retired. 3) The mean score for life satisfaction among total respondents was 61, with the retirees marking a lower-than-mean 59. But this score was not the lowest. It followed wage earners, self-employed and unpaid family workers, and was higher than job-seekers (55) and workers who are temporarily-on-leave (57). But a simple comparison of mean score shows that middle/old-aged individuals who are working enjoy a higher level of life satisfaction than those who retired. 3) Classification of employment status was solely left to the subjective judgement of the individual, and may be different from the classification used in the Economically Active Population Survey by the National Statistical Office.

<Table 2> Life Satisfaction by Employment Status Mean Standard deviation N(%) Wage workers 65.3 19.2 1,874( 18.3) Self-employed 66.1 18.4 1,697( 16.5) Unpaid family workers 66.3 17.3 277( 2.7) Job-seekers 55.2 24.5 422( 4.1) Retirees 59.3 23.3 2,027( 19.8) Temporary leave 56.6 25.3 152( 1.5) No experience in economic activities 58.7 22.1 3,805( 37.1) Total 61.3 21.6 10,254(100.0) Source : KLoSA 1st Basic Survey. Figure 3 below shows the distribution of retirees' life satisfaction based on the 11 scare scores. It shows the highest concentration in the score ranges of 70, 80 and 50, which may be interpreted as the retirees being somewhere in the middle ground in their general perceptions of life: i.e., neither very satisfied nor very dissatisfied. But as the slight tilt to the right after the 60 mark indicates, positive perceptions slightly outweigh the negative, a result that corresponds with other life satisfaction studies on the 60-or-older population. (Ik-gi Kim et al., 1999). Figure 3 Distribution of Life Satisfaction among Retirees 25 20 15 10 5 0 (%) 0 10 20 30 40 50 60 70 80 90 100 (Score) Source: KLoSA 1st Basic Survey.

<Table 3> below is a summary of the mean score for life satisfaction among retirees by socio-demographic characteristics. As indicated by One-Way ANOVA, the difference in the mean score by group depending on the socio-demographic characteristics (with the exception of years after retirement) was also found to be significant in the population. By gender, men indicated a slightly higher level of satisfaction, 60, than women, 58. But existing studies have yet to demonstrate consistent results in terms of difference in life satisfaction by gender. For example, some studies, as this one, indicate a higher sense of satisfaction among men than women (Joong-dong Gwon and Joo-yeon Cho, 2000; Hyeong-shil Lee, 2003), while some others reveal the opposite (Gyeong-sook Park, 2003; Gi-nam Park, 2004), and yet others conclude that gender is not a factor in determining life satisfaction (Tae-hyeon Kim et al., 1999; Cheol Gwon et al., 2006). Thus, the result of this study, that men are more satisfied with life than women, should be generalized with caution. Meanwhile, by age, life satisfaction among retirees was found to decline as age goes up. But broader studies (including youth) have indicated a higher level of life satisfaction among the elderly than youth or middle-aged. (Myeong-sook Jeong, 2005). But in this study, looking only at the middle/old-aged retirees by three age groups, 50s or younger, 60s and 70s, the mean score for life satisfaction was found to drop with age, at 62, 60 and 57 respectively. Looking at life satisfaction by household characteristics, it was found that those with spouse have a distinctly higher level of satisfaction than those without (either by death or divorce). Such difference also corresponds with common understanding. Cohabitation with married children was also found to have significant impact on life satisfaction of the elderly. The retirees who live with their married children exhibited a higher level of satisfaction, a result that also corresponds with common understanding on account of Korea's traditional family value of supporting the elderly. The fact that retirees' life satisfaction is partially determined by cohabitation with children is also indicative of the fact that most seniors in Korea still prefer to live with their

children. If, similar to western societies, the elderly value privacy and independence but must live with their children owing to financial or health reasons, their life satisfaction are likely to be lower than those who live separately from their children (Gyeong-sook Park, 2003). Meanwhile, it was found that the higher the educational level, the higher the life satisfaction. The mean score by education group (primary school or lower, middle school, high school, college or higher) was 54, 60, 65, and 70 respectively, moving higher along with educational level. Higher educational attainment tends to allow for a more proactive and active post-retirement life, as it makes more resources available, such as economic or social resources. 4) Thus, higher life satisfaction along with higher educational attainment also corresponds with common understanding. The state of health is one of the most important factors in life satisfaction, not only for the elderly but for all individuals in general. The results show differing levels of life satisfaction by subjective perceptions of health, with individuals with positive views being more satisfied. By five categories of health, from very dissatisfied to very satisfied, the mean score for life satisfaction went up from 37, to 55, 65, 71, and 74. Especially noteworthy is the big difference between the "Very dissatisfied" and "Very satisfied." By social activities such as religious or friendly gatherings, there was a big difference in life satisfaction between those who are active and those not active at all. Such result is also in line with previous studies on the association between social activities and life satisfaction. Last factor is retirement. As discussed earlier, little difference was found in life satisfaction by years after retirement. Looking at years after retirement in four categories (with an interval of 5 years), the score was almost even at 59 for all four. But aside from years of retirement, if the individual is satisfied with retirement in general, his life satisfaction tended to be higher as well. Individuals not satisfied with retirement at all yielded a mean score of life satisfaction of 49, 4) One of the important factors of life satisfaction is financial resources, such as income, assets or spending money. But the financial resource factor could not be used in this study as the data was not completely organized. It may be partially reflected through the factor of educational background.

those relatively satisfied 67, and those highly satisfied, 72. As the fact of retirement is a reflection of one's current reality, satisfaction with retirement is inevitably closely associated with satisfaction with life in general. <Table 3> Mean Score for Life Satisfaction by Major Socio-demographic Factor Gender Age Marital status Education Subjective perceptions of health Cohabitation with married children Social activities Years after retirement Retirement satisfaction Note : ** p<0.01, *** p<0.001 Source: KLoSA 1st Basic Survey. Mean Standard deviation N Female 57.7 24.7 866 Male 60.4 22.3 1,161 50s or younger 62.0 23.7 443 60s 60.3 22.9 725 70s or older 56.9 23.5 859 Have spouse 61.7 22.2 1,526 No spouse 51.7 25.2 501 Primary school or less 53.7 23.6 1,014 Middle school 59.6 23.0 299 High school 65.3 21.1 449 College or higher 70.0 20.1 263 Very dissatisfied 37.1 24.8 297 Dissatisfied 54.8 22.4 627 Average 64.6 19.6 576 Satisfied 70.9 16.0 480 Very satisfied 74.0 18.3 47 No 58.4 24.0 1,699 Yes 63.8 19.1 328 None 50.1 24.9 621 Yes 60.7 21.1 495 5 years or less 58.9 23.7 708 6~10 years 59.9 23.0 577 11~15 years 59.4 24.1 236 16 years or longer 59.9 22.3 477 Very satisfied 71.8 19.2 87 Relatively satisfied 67.1 18.9 1,058 Not satisfied at all 48.6 24.3 882 F 6.70** 8.04*** 71.28*** 51.40*** 147.86*** 14.98*** 57.10*** 0.24 385.89*** It may be reasonable to assume that the socio-demographic factors reviewed so far also have close association with each other in relation to life satisfaction. To identify the impact of each factor on life satisfaction while controlling for other factor, multiple linear regression analysis was performed. Age, years after retirement and social activities were not grouped into the categories as in <Table 3>, but were converted into age, years after retirement and extent of social activities and input into the regression model as a continuous variable.

The difference between ModelⅠ and Ⅱ is whether retirement satisfaction was included as an independent variable or not. <Table 4> summarizes the results. First, looking at the results from Model Ⅰ, while gender and age exhibited negative impact, other factors all showed positive impact. But when controlling for other factors, gender and age did not demonstrate significant impact, unlike the result from the One-way ANOVA. The factors that had statistical impact within 0.05 significance level were marital status, years after retirement, education, subjective perceptions of health, cohabitation with married children and social activities. Of these, subjective perceptions of health had the biggest impact ( β=0.37) on life satisfaction felt by retirees, followed by social activities ( β=0.16), presence of spouse ( β=0.12), and cohabitation with married children ( β =0.12). Years after retirement was not found to have a major impact (β =0.06). As for education, there was no statistically significant difference between the primary school and middle school graduates, but there was a big difference between high school/college graduates and primary school graduates. Looking at the results from Model II, which includes retirement satisfaction, gender and age showed negative impact and others positive impact, a result that is consistent with that of Model I. Likewise, all factors with the exception of gender, age, and years after retirement were found to have statistical significance. The difference with Model Ⅰ is that years of retirement has no statistical significance. Instead, retirement satisfaction was found to have a major impact on life satisfaction. Comparing the standardized regression coefficient of each factor with Model I, subjective perceptions of health had the biggest impact but at a 20% lower level ( β=0.37 β=0.30), followed by retirement satisfaction ( β=0.22). The pattern of reduced impact was found among other factors as well, such as social activities ( β=0.16 β =0.13), presence of spouse ( β=0.12 β=0.11), and cohabitation with married children ( β=0.12 β=0.09), while still affecting life satisfaction. In sum, with the exception of gender and age, family-related factors such as presence of spouse, cohabitation with married children, and

socio-economic factors such as education, health-related factors such as subjective perceptions of health, social relations factors such as social activities, and retirement factors such as retirement satisfaction were all found to be associated with life satisfaction. <Table 4> Result of Regression Analysis on Life Satisfaction and Socio-demographic Factors Unstandardized coefficients (B) ModelⅠ Standard error Standardized coefficients (β ) Unstandardized coefficients (B) ModelⅡ Constant 26.13*** 4.26 30.33*** 4.26 Standard error Standardized coefficients (β ) Male (Reference group: female) -1.81*** 1.12-0.04-0.93*** 1.09-0.02 Age -0.01*** 0.06-0.01-0.07*** 0.06-0.03 House spouse (Reference group: no spouse) Middle school (Reference group: primary school or lower) 6.66*** 1.20 0.12 5.93*** 1.12 0.11 2.44*** 1.37 0.04 2.02*** 1.34 0.03 High school 3.92*** 1.27 0.07 2.75*** 1.24 0.05 College or higher 7.65*** 1.54 0.11 5.85*** 1.51 0.09 Subjective perceptions of health 7.98*** 0.45 0.37 6.49*** 0.46 0.30 Cohabitation with married children 7.24*** 1.27 0.12 5.80*** 1.24 0.09 Years after retirement 0.14*** 0.05 0.06 0.08*** 0.05 0.03 Social activities 0.57*** 0.07 0.16 0.47*** 0.07 0.13 Retirement satisfaction (very satisfied and satisfied) 10.38*** 0.97 0.22 F 75.47*** 82.83*** R² 0.28 0.31*** N 1,998 1,998*** Note : * p<0.05, ** p<0.01, *** p<0.001 Source : KLoSA 1st basic survey. Ⅴ. Conclusion As the society continues to age, it entails the social consequence of growth in the aged population and the individual consequence of lengthened old-age period, issues that have attracted growing attention in old-age life. Against such trend, what determines the quality of life of elderly individuals is not only the absolute conditions surrounding

one's living environment but also the subjective level of satisfaction in life. Retirement from the labor market serves as a turning point for the individual, requiring that he stop performing the role accompanied by the employment position and adapt to new roles. Traditionally, retirement - and ensuing loss of labor, loss of economic/social resources - could be a source of tremendous stress to individuals, so much so that it is one of the major causes of dampened life satisfaction. In the case of Korea in particular, retirement has a major impact on life satisfaction, as it is often the result of early removal from the labor market, and accompanied by lack of post-retirement planning, loss of social/economic roles, reduced income, insufficient social welfare programs and ongoing financial needs for children's education and wedding. Against backdrop, this study reviewed life satisfaction of retirees using the data from KLoSA 1st Basic Survey, with focus on retirement satisfaction and socio-demographic factors. As described in the paper, the socio-demographic factors that affect life satisfaction of retirees were found to be subjective perceptions of health, retirement satisfaction, extent of participation in social activities, cohabitation with married children, marital status (presence of spouse) and education. Of these factors, subjective perceptions of health was found to have the biggest impact, while retirement satisfaction also had a major impact on life satisfaction in general.

<References> Joong-dong Gwon Joo-yeon Cho (2000), "Factors that Affect Life Satisfaction in Old Age," Journal of the Korea Gerontological Society, 20(3), pp.61~76. Cheol Gwon, Cheol-hyeon Nam, Hae-seok Hwang, Han-tae Jeong, Soon-hee Ha, Gyeong-jin Min, and Seok-soon Kim (2006), "Factors that Affect Life Satisfaction of the Elderly," Korea Health Study, 32(1), pp.66~74. Ik-gi Kim, Dong-bae Kim, Seon-hee Moh, Gyeong-sook Park, Young-hee Won, Yeon-sook Lee, and Seong-nam Cho (1999), Life of Korean Elderly: Overview and Outlook, Future Human Resource Series. Tae-hyeon Kim, Dong-bae Kim, Mi-hye Kim, Young-jin Lee and Ae-soon Kim (1998), "Study on Enhancing the Quality of Life at Old Age," Journal of the Korea Gerontological Society 18(1), pp.150~169. (1999), "Study on Enhancing the Quality of Life at Old Age ( Ⅱ)," Journal of the Korea Gerontological Society, 19(1), pp.61~81. Gyeong-sook Park (2003), Population Aging - a Current Future, Euiam Publishing. Gi-nam Park (2004), "Difference in Life Satisfaction at Old Age," Journal of the Korea Gerontological Society, 24(3), pp.13~29. Hwa-yong Shin and Byeong-eun Cho (2001), "Resources and Life Satisfaction of Married Couples with Retired Husbands," Journal of the Korea Home Economics Association, 39(1), pp.39~51. Hyun-Goo Shin, kachung Boo and Hye-jeong Lee (2006), "Introduction to KLoSA 1st Basic Survey," Labor Review, Sept. (no. 21), pp.90~99. Hyun-Goo Shin and Hye-jeong Lee (2006), "Impact of CAPI on the Quality of Survey Data," Survey Study, 7(2), pp.71~95. Hyeon-song Lee (1997), "Koreans' Quality of Life: with Focus on the Objective Factors," Korean Sociology, no. 31, Summer edition, pp.269~301.

Hyeong-shil Lee (2003), "Inter-generation Exchange of Social Support and Life Satisfaction among Rural Elderly: Comparison by Gender and Age Group," Journal of the Korea Home Economics Association, 41(4), pp.57~69. Myeong-sook Jeong (2005), "Age and Life Satisfaction," Journal of the Korean Psychological Association: Development, 18(4), pp.87~108. Yeon-gyeong Ji and Byeong-eun Cho (1991), "Inter-external Locus of Control and Perceptions of Stress and Life Satisfaction following Retirement," Journal of the Korea Home Economics Association, 29(2), pp.217~240. Hee-ja Han and Eun-shil Gang (2001), "Changes in Life after Retirement," Journal of Korea Psychology Nursing Association, 10(2), pp.203~219.