- Top of page
- Projection Model
South Korea has been undergoing significant change in its population structure over the past three decades. Within 10 years, South Korean baby-boomers will reach the age of 65 years and accelerate this change. This trend in population structure is crucial, because an aging population may increase medical demand, especially that for long-term care (LTC) services, which would create a financial burden on society.
This study estimates total LTC expenditure in South Korea from 2015 to 2050 by modifying the method proposed by the UK Personal Social Science Research Unit, the seminal study on projecting costs of LTC services. Using population data from the projections of the Korean Statistical Information Service, I stratify the projected population by gender and age, using the groups 65–69, 70–74, 75–79 and 80 or over and divide LTC services into two categories, namely facility and home care.
South Korea's total LTC expenditure is predicted to continuously increase and then reach 4.2% of GDP in 2050. Expenditure on LTC services for women is higher than that for men. Moreover, the increase in total expenditure is dramatic after 2040 for home-based services but is constant for facility services. This study shows that the presence of baby-boomers heavily influences LTC expenditure in South Korea. Copyright © 2013 John Wiley & Sons, Ltd.
- Top of page
- Projection Model
South Korea has been experiencing a significant change in its population structure over the last three decades. Industrialization and changes to values concerning children have meant fertility has decreased. According to the Marriage and Fertility Survey (Korea Institute for Health and Social Affairs, 2005 and 2009) the proportion of unmarried men and women who will definitely have children after marrying has fallen from 54% and 42%, respectively, to 24% for both, between 2005 and 2009. In addition, advanced medical technologies have increased life expectancy (Kim et al., 2010; Sunwoo et al., 2010; Kang et al., 2010). As a result, the proportion of the population aged 65 years or over has increased from 3.8% in 1980 to 10.7% in 2009. Moreover, it will reach 25.9% between 2030 and 2035, because South Korean baby-boomers will accelerate the change in the population structure.
The Korean War between 1950 and 1953 led to a population decrease of about 2 million. To replenish the labor force after the war, the Korean government enforced various policies to encourage childbirth, as well as to improve public health conditions. As a result, approximately 8 million babies were born between 1955 and 1963, while infant mortality dropped steeply. For a decade, between 0.7 and 1 million babies were born every year. These were enormous numbers compared with previous years, in which 0.3 to 0.5 million babies were born (Sunwoo et al., 2010). As of 2010, the number of baby-boomers in South Korea is 7.12 million, 14.6% of the whole population. In 2025, these baby-boomers will start entering their 60s and 70s. Therefore, the number of over 60s will rise rapidly between 2025 and 2035, as mentioned before. However, from 2040 the proportion of the population aged 60–64 years will decrease as the effect of the baby-boomers fades away; yet, it will still be greater in size than in 2015. Meanwhile, in 2050, we can predict a higher proportion of the population aged 75 and over (Figure 1).
Although life expectancy at birth has reached the age 81.2 years, as of 2011, there is an 11-year difference between life expectancy and healthy span, meaning that many older adults might suffer from chronic illness for years. Along with an increasing number of unhealthy older people, there is concern that demand for medical services—especially long-term care (LTC) services—will increase. The increasing demand for LTC services might affect the total health expenditure and, via increasing pension payments, public finances (Colombier and Weber, 2011). As shown in Figure 2, medical expenditure in each age-gender group has been increasing consistently in South Korea since 2006. For women aged 60–69 and 70–79 years, medical expenditure was about 1.5 billion won (\)1 in 2006, but it grew to about 3.0 billion won in 2008. As well as the increase in medical expenditure, figures for treated days of chronic illness (TDCI) further demonstrate that an older population uses more medical services. As you can see in Figure 3, the amount of TDCI that older adults experience has been increasing. For example, the number of days suffering from hypertension, diabetes and mental illness has gradually increased since 2006. The increase in hypertension is of particular concern, as this can cause a stroke, which would then demand LTC. According to a survey (Ministry of Health and Welfare, and Seoul National University Hospital, 2009) as of 2008, 420 000 older adults suffer from dementia; its prevalence rate was 8.4% in 2008, but this will reach 9.7% in 2020.
To cope with issues relating to LTC in an aging society and to protect the value of life for older adults, the government of South Korea had been discussing and reviewing a proposal for an LTC insurance system. After being refined for 7 years, Long-term Care Insurance Law was passed in April 2007, and the LTC insurance scheme was enforced from 1 July 2008. Although financial aid for LTC services had previously been selectively provided to low income older adults, the newly established LTC insurance system is a universal coverage scheme. Table 1 represents an overview of the Korean LTC insurance system. The entire population is policyholders, meanwhile older adults of age 65 years or over and people who develop a senile disease when or before they are 45 are the beneficiaries. They are graded 1, 2 and 3. A research visit is used to evaluate the grade on the basis of four areas: physical function, cognitive function, behavioral changes and nursing treatment. The benefits are facility-based services, home-based services and a special cash allowance. Policyholders pay their deductible at a rate of 15% or 20%, depending on the type of services. The LTC insurance system is financed by social insurance and taxes. A policyholder's LTC insurance premium is calculated by multiplying their medical insurance premium by the rate of LTC insurance, 4.05% in 2008. The rate of LTC insurance is decided by the LTC Committee, who considers its financial condition. In addition, the government provides a National Health Insurance Corporation with 20% of expected returns from the LTC premium. The whole system is managed by the Minister of Health and Welfare; therefore, the Ministry takes overall responsibility for policy-making, providing a certain standard of service, and so on.
Table 1. Overview of long-term care system in Korea
|Type of benefit||Facility-based service, home-based service, special cash allowance|
|Funding sources||Social insurance and tax|
|Share of expenses||Deductible rate*:|
|– Facility-based service: 20%|
|– Home-based service: 15%|
|Beneficiary||The aged 65 years or over, or under the age 45 years with senile disease|
|Provider||National Health Insurance Corporation|
There were 147 000 beneficiaries in the first month of enforcement, but this number grew to 280 000 by October 2009 (Chang et al., 2010). Two years since the LTC insurance scheme was established, it is necessary to evaluate the long-term financial sustainability of the current system. In this context, this study predicts LTC expenditure in South Korea until 2050, using a model modified from the UK Personal Social Science Research Unit (PSSRU, Wittenberg et al., 2006). The study is organized as follows: First, it describes the data and the model, then results are reported, and finally, it presents a discussion.
- Top of page
- Projection Model
The projection model referred to in this study is derived from Chung et al. (2009), which modifies the PSSRU method. Population data are from the projection of the Korean Statistical Information Service (KOSIS). I stratify it by gender and age: 65–69, 70–74, 75–79 and 80 years or over. This is instead of using 60–64, 65–69, …, 80 years or over (Chung et al., 2009). These different strata may be more informative for an aged society, because typically older adults are considered as being 65 or over. Data used for the study came from a survey conducted by the Ministry of Health and Welfare: “2008 Living Profiles and Welfare Services Needs of Older Persons in Korea.” It surveyed 15 146 people aged 60 years or over, and 2698 of their children who were living with them, in 16 major cities and provinces. It contains information on their physical, mental and psychological characteristics, and their socioeconomic status, and so it indicates their demand for health services and welfare.
To estimate the utilization rate of LTC services for every age group, a logit model for each gender is first constructed using the experience of long-term facility/home care services as the dependent variable and including education, income, number of chronic diseases, whether living with someone or not, housing status and type of physical assistance (Table 2).
Table 2. Characteristics of sample
|Variable|| || |
|Gender||Male (= 1)||51.1%|
|Mean (SD)||1.49 (0.49)|
|80 years or over||18.6%|
|Mean (SD)||73.89 (6.08)|
|Education||Elementary or below (= 1)||71.7%|
|Middle school grad||11.4%|
|High school grad||10.4%|
|Community college grad||1.2%|
|University grad or over||5.3%|
|Mean (SD)||3.03 (1.57)|
|Income (per month, 10 000 Korean won)||0–30||41.5%|
|500 or over||0.3%|
|Mean (SD)||58.45 (75.98)|
|Number of chronic disease (by doctor's diagnosis)||0||19.1%|
|3 or over||30.2v|
|Mean (SD)||1.90 (1.54)|
|Live alone||No (= 0)||54.9%|
|Housing status||Own (=1)||74.3%|
|Private insurance||Yes (= 1)||0.4%|
|Physical assistance from children living together (during last year)||Yes (= 1)||74.6%|
|Physical assistance from children living separately (during last year)||Yes (= 1)||48.3%|
|Physical assistance from spouse (during last year)||Yes (= 1)||84.8%|
|Physical assistance from siblings (during last year)||Yes (= 1)||7.1%|
|Physical assistance from friends or neighbors (during last year)||Yes (= 1)||14.3%|
|Used long-term care facility||Never (= 0)||99.2%|
To carry out a sensitivity analysis, the probability of LTC service utilization is predicted in two ways. One includes the variables education, income, number of chronic diseases and whether living alone to estimate utilization. The second model, by contrast, additionally includes housing status and whether purchasing private insurance or not.
where Y is the utilization of service j, i is the group of age, j is the type of LTC services (facility or home-based) and xk is the covariable.
The probability of LTC service utilization predicted for each age group is derived from the following equation, using coefficients from Equation (1):
where is the predicted probability of LTC service utilization by age group.2
To take into consideration changes in disability-free life expectancy, the probability is assumed to reduce by 1% and 2% every 5 years. The volume of LTC service, S, is obtained by multiplying P, the predicted probability of LTC service utilization by age group, by N, the projected population of each demographic. This predicts the quantity of facility or home-based services, S, needed for each age group.
where S is the needed volume of LTC services, P is the probability of use and N is the projected population estimates.
Finally, to derive the total expenditure, TE, the costs of facility and home-based services are needed. In this study, I use the average LTC cost spent by beneficiaries in all three ratings,3 based on KOSIS data, to calculate the (expected) LTC cost, C, for every 5-year period between 2015 and 2050. The costs are calculated regardless of the age group, but with consideration of the service type, facility or home-based.
where Cjt is the LTC cost of type j in year t, R is the inflation rate4 and t = 2015, 2020, …, 2050.
Then, TE is derived by multiplying the volume of service, S, by the average LTC cost, C.
where TE is the total expenditures and t = 2015, 2020, …, 2050.
- Top of page
- Projection Model
Table 2 summarizes the sample characteristics. The average age of the sample population is 73.89 years, and women make up 48.9%. Only 28.3% of the sample graduated from middle school or higher. The majority, 80.8%, earn 900 000 won or less per month; 74.3% of them have their own houses, whereas 45.1% live alone. In addition, most receive physical help from their spouses or children. Approximately 81% have been diagnosed with a chronic illness, and most, 99.2%, have never visited a facility that provides LTC services.
Figure 4 shows the predicted LTC expenditure for women from 2015 to 2050. Panels (a) and (b) display facility services and home-based services, respectively. In a comparison between the two service types, based on total predicted expenditures with model 1, the estimations on home-based services are approximately 16 times higher than those of facility services in 2015 (1288 billion won for facility services and 20 893 billion won for home-based services). In the case of facility services, the age group 70–74 years shows the highest expenditures from 2015 to 2040, but after that, the age group 80 years or over shows the highest level. For home-based services, the age group 80 years or over tends to represent the highest expenditure throughout the investigated period. In addition, facility services show higher expenditure with model 2 than with model 1, whereas home-based services show lower expenditure with model 2.
Figure 5 presents the estimations for men. Similar to the case for women, expenditure on home-based services is approximately 12 times higher than that on facility services in 2015 (774 billion won for facility services and 9115 billion won for home-based services). The age group 75–79 years has the highest expenditure on facility, regardless of model. In the case of home-based services, the age groups 65–69 years (years 2015–2025) and 80 years or over (years 2030–2050) have the highest expenditure. In addition, model 1 has a lower level of expenditure compared with model 2 for both service types.
Taken together, both genders show an increasing trend throughout the study period. Women, however, have significantly higher LTC expenditure for both facility services and home-based services compared with men. Further, as older adults spend more on home-based services than (relatively) younger adults do, the difference in expenditure between the two service types grows. As shown in Figure 6, as a proportion of GDP, this expenditure is predicted to increase steadily but dramatically later; total LTC expenditure is 2.5% of GDP in 2015 and 4.2% in 2050. Changes in disability-free life expectancy, however, would reduce its proportion. In 2050 total spending would reach 3.9% and 3.6%, respectively, for the decrease of the probability of LTC service utilization by 1% and 2%.
- Top of page
- Projection Model
South Korea will become an aged society in the near future. Life expectancy has been increasing, from 62.15 in 1970 to 79.4 in 2007, but the fertility level has declined, from 4.53 in 1970 to 1.23 in 2010. In addition, baby-boomers will reach the age of 60 years in 2015. The aging population trend might increase the demand for medical care, especially LTC for chronic illnesses, leading to a financial burden on society. The findings of the present study indicate that South Korea's total LTC expenditure will increase continuously from 2015 to 2050.
The LTC expenditure difference between genders shows an interesting result. For the case of facility services for women, the age group 70–74 years has relatively high expenditure until 2040 in comparison with the other age groups. If the effect of baby-boomers who will become 80 or over after 2040 is considered, then the age group 70–74 years still maintains a relatively high level of expenditure after 2040. In addition, the expenditure of that group increases dramatically after 2025. The reason for this surge can be found in population changes; people born in 1955 will become 70 years old during that period.
In terms of facility services for men, 75- to 79-year-old adults show relatively high expenditure compared with the other age groups. Both women and men at age 80 years or over show a dramatic increase in expenditure on home-based services, growing by approximately 35% between 2045 and 2050. As people age, the probability of LTC service utilization mostly increases, although there is no large probability difference between model 1 and model 2. This increase is more certain for home-based services. Therefore, the increase results from their high home-based services utilization rate in comparison with the other age groups.
According to my estimations, the probabilities of home-based service utilization are 0.58 and 0.46 for women and men aged 80 years or over, respectively, but 0.45 and 0.23 for those aged 65–69 years. Moreover, the total volume of LTC services increases using model 1 by 32% (female) and by 56% (male) from 2030 to 2050, meaning that the increment in total expenditure in this period becomes relatively large. Second, expenditure on home-based services shows a constant increase for each age group regardless of gender. The oldest group shows a dramatic increase after 2040 because of the influx of baby-boomers.
Older adults prefer home-based services to facility services. First, this is because cost may matter to them; home care services are cheaper than facility care services. Facility care can cost patients 500 000–600 000 won per month. However, for home-based services, older adults need 100 000–200 000 won per month. This includes nurse visits, auxiliary medical appliances and day care services. In addition, a cultural factor may contribute to the soaring expenditure on home-based care services. The majority of older adults prefer to spend their remaining years at home with their spouses or families, which, they believe, is the traditional way to face death. According to a survey by Jung et al. (2005), 64.5% of those aged 80 years or over do not want to use facility care services when in need of them. These reasons explain why the difference between facility and home-based services is increasing. Indeed, despite having enough assets, such as owning their own houses or having private insurance, expenditure on home-based services is declining for both genders. For facility services, a decrease also appears for men.
As shown by these results, LTC expenditure in South Korea is continuously growing, from 2.6% of GDP in 2020 to 4.2% by 2050.5 Therefore, South Korea needs to seek a way to reduce this cost burden. To reduce demand for LTC services, one can first consider changing the LTC service paradigm. The LTC insurance system currently focuses on maximizing beneficiaries' functional abilities to achieve independence in their daily lives, maintaining a better quality of life and reducing the financial burden of care on family members. However, to maintain a reasonable level of LTC expenditure, policies should be designed to include more prevention-oriented LTC. For example, the LTC insurance system should consider not only supporting the provision of intensive care for serious illnesses but also supporting the prevention of diseases and management of minor illnesses.
By examining how poor the health of older adults is and managing cases, the LTC insurance system might assist older adults avoid serious illnesses and improve their health (Lee, 2008; Kim et al., 2007). To provide preventive care effectively, a dedicated center should be established. These efforts might eventually help decrease demand for LTC services and thus LTC expenditure, allowing financial stability. For instance, in Japan the LTC policy direction changed to prevention from April 2006 in order to reduce the continuous growth in the financial burden. As a result, Japan has experienced a decrease in expenditure on LTC services (Lee, 2010).
The strengths of this study are twofold. First, it reports expenditure by gender and age group, meaning that its findings can be used by policy makers to design more effective LTC policies. Moreover, it shows how the presence of baby-boomers greatly affects LTC expenditure and thus how examining by age group is crucial to establishing related policies. Second, it applies up-to-date empirical data. Although some research studies project the cost of LTC services in South Korea, they were published before the LTC insurance system was established, and thus, their ability to suggest suitable policy implications is limited.
However, this study also has some limitations. First, data limitations mean that it fails to consider more variables that could change in the future and apply different probabilities of utilization to age groups. Second, although three ratings are used to classify claimants, the study only uses average cost to calculate costs. In a future study, it would be worthwhile considering all ratings of expenditure to provide more accurate projection figures.