Chapter 201

Chapter 20

Intergenerational Resource Allocation
in the Republic of Korea

Chong-Bum An, Young-Jun Chun, Eul-Sik Gim,
Namhui Hwang, and Sang-Hyop Lee

The challenges resulting from rapid aging and extremely low fertility are of great concerns to Korea simply because no other societyhas faced so dramatic a demographic transition. This chapter provides some important features of the country’s intergenerational resource allocation, highlighting its unique features of old-age support. The results, based on 2000 data, show that, on the one hand, the lifecycle deficit for the elderly is smaller than in many other countries included in this study. On the other hand, the productivity of elderly Korean workers is low, although their labor force participation rates are high. The results also suggest that Korean households seem to do an adequate degree of consumption smoothing around the time of retirement, even without sufficient pension benefits. Whatever worked for Korean retirees in the past, however, will not work in the future, largely because of the rapidly changing population structure and fiscal and pension policies.

This chapter explores the extent of intergenerational resource allocation in Korea, using the NTA methodology. Along with presenting our estimates based on this new method, we highlight two unique features of the Korean economy in this chapter: the saving patterns of Korean elderly and the recent trend in Korea’s public-sector accounts.

Background

Korea’s population is aging very rapidly. The total fertility rate (TFR) decreased from 4.53 children per woman in 1970 to 1.08 in 2005, reaching the world’s lowest level. As of 2005, the share of the population aged 0–14 was 19.1%, while the share of population aged 65+ was 7.2%. According to the most recent UN projections (medium scenario), the elderly share in Korea will grow very rapidly, surpassing that of the United States and many European countries in less than two decades. About 19.6% of Korea’s population will be 65 or older in 2025, and this share is projected to exceed 37% by 2050.

Korea has had an extraordinary development experience, but it is now moving to a slow and stable stage of economic growth. Until the 1960s it was a poor and predominantly agricultural country. It then rapidly transformed itself into a modern, industrialized country, owing to the success of an export-oriented industrialization strategy. Between 1962 and 1996, real GDP per capita grew at an annual rate of 6.6%. At present, the country is focusing on maintaining stable growth while continuing its structural reform.

Although the familial support system is eroding in Korea, the public support system has been expanding rapidly since the late 1980s. The current social security system consists of three parts: social insurance, public assistance, and social welfare service. Social insurance comprises Industrial Accident Compensation Insurance, the National Health Insurance Scheme, the National Pension Scheme, and Employment Insurance. Depending upon their type of occupation, employees join the National Pension, Government Employees Pension, Military Personnel Pension, or Private School Teachers’ Pension plan.

The number of people insured through the National Pension Scheme has steadily increased, and by 2002 was covering more than 16 million people. There were 0.9 million beneficiaries in 2002, but the number has been growing by about a quarter million people annually and is expected to increase to 3 million by 2010. The National Health Insurance Scheme and its supplementary program covered almost all citizens as of 2002.

Public assistance comprises the National Basic Livelihood Security System, medical aid, veterans’ relief, and disaster relief. Public assistance has changed the paradigm of social policies for fighting poverty. All people who live under the poverty line are provided financial benefits regardless of their working ability. At the same time, the government provides a self-support program for those in the lower-income bracket who are able to work.

Intergenerational resource allocation

This chapter closely follows the NTA methodology described in Mason et al. (2009), and Lee, Lee, and Mason (2008). The primary sources of microdata sets used in our analysis are the National Survey of Household Income and Expenditure (NSHIE) and the Korean Labor and Income Panel Study (KLIPS) of 2000. We also used the Household Income and Expenditure Survey (HIES) to collect up-to-date information on asset income. In addition to the microdata, we used various sources to construct the aggregates. These included the national income and product accounts, the National Pension Statistical Yearbook, and the National Health Insurance Statistical Yearbook.

The economic lifecycle

Figure 20.1, which presents the lifecycle profile of per capita production and consumption in Korea, indicates that the lifecycle surplus extends for a period of 32 years, from age 24 to age 55. As in all economies, the production profile has a typical inverse-U shape.

[Figure 20.1 about here]

However, there are important differences between Korea and the other economies. One of them is that production in Korea appears to decrease from the late 40s, which is much earlier than in many other advanced countries. This is interesting because labor force participation rates for people in their 50s and 60s in Korea are not lower than in other advanced countries. The difference reflects the lower productivity (or fewer working hours) of older workers in relation to younger workers in Korea. It may be due to rapid changes in technological progress and educational attainment, which have put older workers at a disadvantage, as compared with those in many other advanced economies (Mason, Lee, and Russo 2005). This is a striking result, especially given that Korean companies used to rely on a seniority-based wage system. That unique system has deteriorated, however, especially since the 1997 financial crisis. How the financial crisis may have affected our results is an interesting question, but is beyond the scope of this chapter.

There are also important differences between Korea and other countries in their consumption patterns. In Korea, consumption peaks in the late teens and decreases until the 40s. Such a decline in consumption was also found in other countries as a result of decreasing expenditure on education (see Chapter 6, on consumption, by An-Chi Tung in this volume). However, the decline seems to come earlier in Korea, immediately after high school graduation. This result may be related to the high level of private education consumption among high school students in Korea. Private education in the form of tutoring is very common because of the difficulty of college entrance examinations.

Another difference between Korea and other countries participating in the NTA study is that consumption bythe elderly appears to be much lower in Korea. This outcome is in large part due to very low consumption of private heath care by elderly Koreans. To investigate whether this finding was due to a measurement error or a data problem, we compared our results with other publicly available sources of information; but the results were not qualitatively different.The literature on health systems identifies several factors that explain why some groups of people in a society have low medical expenditure.The factors include limited access to health care, low income, or physical difficulty in visiting health care institutions. In Korea, a nontrivial portion of very old people also still die at home, requiring low medical expenditure (Kwon 2008).The low spending may be peculiar to this cohort and their life experience rather than characteristic of Koreans in general or Korean institutions. If so, then spending late in life may be much higher in the future.

Transfers

Public transfers accounted for 27% of all transfers to children in Korea in 2000 (Figure 20.2). Although this figure is somewhat lower than in more developed countries—for example, it is 40% in the US—most of the difference between the US and Korea is due to the difference in public education spending. Children’s pro rata share of other public goods in Korea was roughly equal to that in the other countries. Public transfers were also an important component of the reallocation system for the elderly in Korea. Although they were far less prominent than in most advanced economies, public transfers still accounted for 37% of age reallocations for the elderly.

[Figure 20.2 about here]

The shape of public health care transfers was similar to that in the other countries, whereas the shape of public pension transfers was not (not shown in Figure 20.2). Because 2008 was the first year of normal benefit disbursement from the National Pension Scheme, our age profiles of public pension transfers based on 2000 data show very small inflows of pension transfers to the elderly (ages 60+).

Asset-based reallocations

Because asset income was negligible in the public sector, the age profile of public asset-based reallocations was similar to that of general taxes (Figure 20.3)and almost a mirror image of the public savings profiles (not shown in the figure).

[Figure 20.3 about here]

Private asset-based reallocations were positive at most ages except for ages 15–32. At those ages, asset income was smaller than saving. Private asset income was rising steeply among people in their 40s and 50s, but their private savings were quite low compared with their asset income. Perhaps the reason for their low savings is that people in this age group were financing much of the consumption of children and the elderly.[i] A striking result was the increase in asset income as well as savings among people in their mid-70s and older. The increase was due in part to the net inflow of interest income for those age groups, while capital income was still a dominant form of asset income for Koreans of all age groups.

It is far from clear why the asset-based reallocations in Korea show these age profiles. However, it is worth mentioning that the Korean housing system has several unusual features that may have a significant effect on the private sector through private transfers and private asset-based reallocations. A house is the major form of assets for homeowners in Korea, accounting for 50% of total assets held by all households, significantly more than the 30% in the US (Cho 2005). But it is not easy to liquidate housing assets by taking out a reverse mortgage or selling a home, which often leads homeowners to transfer expensive property to other family members as a bequest when they die.[ii]

A popular alternative to homeownership is to rent a house or an apartment under a system unique to Korea, called the key money system (chonsei). The tenant pays a deposit up front (without having to make any subsequent payments of rent) and receives the nominal value of the deposit from the landlord upon maturation (Cho 2005). Given this unique structure of the rental market system, renters in Korea have a portion of their assets tied up in housing and thus cannot diversify them. This system is likely to have affected the age profiles. For example, many elderly Koreans do not have income sources other than their houses. The source of interest income could be the key money from renting their houses. Estimating the extent of how this system affects the intergenerational economy in Korea is an interesting research question, but is left for future research.

National transfer flow accounts and sources of funding

Table 20.1 presents the national transfer flow account for Korea in a highly summarized form, reporting aggregate lifecycle deficits and aggregate age reallocations by age in billions of won. Total age reallocations and their major components are shown in the bottom panel, with positive values representing inflows and negative values representing outflows. As a group, people under the age of 20 and those over the age of 65 were consuming more than they were producing in 2000. The lifecycle deficit of young children was 7.9 trillion won, about 25% of total national consumption. The lifecycle deficit of people 65 and older was 1.7 trillion won, about 5.3% of total national consumption. The lifecycle surplus among prime-age adults (ages 20–49) was 7.4 trillion won, or 23% of total national consumption. Clearly, the resources being shifted across age groups are enormous relative to the size of Korea’s economy.

[Table 20.1 about here]

Korean children under the age of 20 were depending almost entirely on transfers to meet their lifecycle deficit. Of those age reallocations, about 73% were private transfers. Transfers provided about 53% of the lifecycle deficit of the elderly and asset-based reallocations about 47%. The elderly relied heavily on asset income, but they did not dissave. Rather, they had modest saving.

Two uniquely Korean phenomena

Two unique features of the Korean data that are related to its economic lifecycle and the old-age support system are the saving patterns of the Korean elderly and the recent trend in Korea’s public-sector accounts. The first offers clues to understanding Korea’s retirement-consumption puzzle. An examination of the second can partially compensate for the snapshot feature of the Korean NTA, at least for the public sector.

The retirement-consumption puzzle

According to the lifecycle hypothesis, which predicts consumption smoothing by people in the work force, there is no reason for consumers to reduce consumption suddenly at retirement if the time of retirement is fully anticipated. Thus researchers find it puzzling when consumption declines after retirement, as is the case in Korea. There may be factors as pensions that contribute to this anomaly. Although Korea’s social security program has been in effect since 1988, pension benefits for normal retirees have not been paid yet because the required 20 years of coverage were reached only in 2008, the first year of normal pension benefit disbursement. Thus Korean retirees have not benefited from the public pension as much as retirees in other countries. Moreover, many Korean elderly are still living in extended households, and for them the public pension may be less important as a source of support than it is for retirees in many Western countries.

Empirical studies have attempted to explainthe puzzle. Banks, Blundell, and Tanner (1998) found that British households significantly curtailed their consumption around age 60. Reduced work-related expenses could not explain the observed fall in consumption at retirement. In the US, researchers found that individuals decreased their consumption by 7% to 20% at the time of retirement (Bernheim, Skinner, and Weinberg 2001; Hurd and Rohwedder 2003). Several studies have offered alternatives to the lifecycle hypothesis to explain the consumption decline, such as the unknown timing of death (Hammermesh 1984), the uncertainty of retirement timing (Carroll 1994), or a decrease in the bargaining power of husbands at retirement (Lundberg, Startz, and Stillman 2003).

We tested the puzzle for Korea, using the KLIPS data from 1998 to 2003 and the 2003 special module (the survey of elderly Koreans).[iii] The results show that only unexpected retirement supported the retirement-consumption puzzle, whereas planned retirement did not decrease consumption at retirement. In fact, the results show that consumption actually increased slightly at retirement in the case of planned retirement. Thus we found no supporting evidence for the retirement-consumption puzzle in Korea in cases where there was a reasonable level of foresight.

Recent trends in the social welfare expenditure

Another feature of the Korean cross-sectional data is that there have been recent changes in public transfers. The snapshot for year 2000 has limited validity in portraying the current status of the Korean public support system simply because the system has been changing rapidly.

The most striking feature of government expenditure in the last decade was the rapid increase in social welfare expenditure (Table 20.2). Public pension benefits grew by 9% per annum between 2000 and 2005, while the medical insurance benefits rose 15.3% annually during the same period. The National Basic Livelihood Security System (NBLSS) benefits consist of cash and in-kind benefits to households with incomes below the poverty line. It had the highest growth rate, 21.7% per annum, between 1997 and 2005, but the rate of growth slowed somewhat after 2000. Social service expenditure also rose dramatically, recording 15.4% per annum between 2000 and 2005.

[Table 20.2 about here]

Figure 20.4 shows a large increase in social welfare benefits for the elderly during that short time span.[iv]The upward shift of the age profile is most conspicuous for those over age 55, owing to increases in the public pension and medical insurance benefits. The sharp increase in the public pension benefits for those aged 60–64 reflects the increase in the number of newly entitled national pension benefit recipients. The upward shift of the benefits for those aged 65 and older was due mainly to the increase in the benefits of the occupational pensions.

[Figure 20.4 about here]