6. The case of Korea

While economic researchers have always had a great interest in income and wealth, multidimensional inequality assessments that move beyond traditional economic metrics are generally considered to be a more recent type of approach. In this spirit, this chapter will examine various inequalities in Korea such as housing, health, education, and regional inequalities.

The disposable income Gini coefficient of Korea was 0.331 in 2020, above the OECD average.

While Korea has a lower market income inequality level than the OECD average, the impact of its tax and transfer system on the Gini coefficient is smaller than in most OECD countries.

The market income Gini coefficient decreased slightly between 2011 and 2020, while the disposable income Gini coefficient for the same time period decreased at a greater rate (see figure 3). This may well reflect the fact that the 19th National Assembly, which was in place between 2012 and 2016, highlighted the importance of using tax policy to reduce inequality, and thus shifted the policy objective for corporate and value-added tax from economic growth to inequality reduction (Korean National Assembly Budget Office (국회예산정책처), 2017[1]). As a result, corporate tax increased from 12.2% in 2011 up to 33.6% in 2016, and property tax law from 10.5% to 49% (Korean National Assembly Budget Office (국회예산정책처), 2017[1]). This trend towards equity-oriented tax policies continued under President Moon Jae-in (문재인 in Korean), who reformed the income tax system in 2017 in order to further increase its redistributive impacts.

Such policies aiming to reduce income inequality have tended to focus on tax reforms rather than transfers. While a welfare system does exist in Korea, and indeed has expanded in the past decade (전규식, 정지수, 유경원 (Jeon G., Jeong, J., Yoo G.), 2016[2]), social welfare expenditure is lower than in many other countries – In addition, the degree to which fiscal spending contributes to reducing income inequality tends to be comparatively lower in Korea (Yeonhap News, 2018[3]).

In the past several decades, firm land ownership has increased while individual land ownership has decreased. In 1945, the Gini coefficient stood at 0.73, although this was reduced to 0.39 in 1960 after land reform saw redistribution of the land owned by former colonial landlords. Since then, wealthy farmers have started to gain more land, and during Korean industrialisation moved to metropolitan areas and were able to build factories and other infrastructure on this land. As such, the figure has climbed back up again steadily, reaching 0.811 in 2019. While this is certainly a major issue facing the Korean economy and merits further research, such an issue is related to wealth inequality and cannot be addressed directly through public expenditure and transfers, and is thus beyond the scope of this paper

The Korea Institute for Health and Social Affairs provides several indexes1 to measure health inequality in Korea, in order to be able to develop and monitor any health-oriented policies. Education inequality in Korea measures high-ranked university entrance, high-ranked science and engineering department entrance and postgraduate education. Most of these measures are highly dependent on parents’ education level, household income, gender, and region of origin. However, the influence of parents in education inequality does not show an increase (김준형 (Kim J.), 2018[4]). While again, these topics are important issues in Korea, an in-depth discussion of health or education inequality would go beyond this study’s focus on the direct income distributional aspects of public expenditure.

The Korean Statistical Information Service (KOSIS) provides data looking at living standards at the regional level. These data consist of 7 main dimensions, including employment and labour and social integration.2 They show that the majority of regions have seen an increase in GDP per capita between 2017 and 2020, with some of the highest percentage increases attributable to some of the lowest-income regions, including Daegu, Gwangju and Daejeon.

Between 31 December and 31 January, the Ministry of Economy and Finance receives submissions of medium-term business plans from each government office. The Ministry then has two months to send instructions on budget plans based on these business plans back to the offices, and formulate that year’s budget. This budget formulation is then submitted to parliament, who have 90 days to discuss it. The modified version is then given to the Minister of Economy and Finance, who assigns it to the relevant central government offices for execution. Once policies have been enacted, each head of a central government office must submit a settlement report of the fiscal year to the Minister of Economy and Finance, as required by the Public Finance Act. The Minister then aggregates tax revenue and expenditure and submits this overall settlement report to the Board of Audit and Inspection and the President, who then inspect the report before sending it to Parliament.

The 2023 Korean governmental budget plan states that its main goal is to exert fiscal restraint to maintain financial stability, following the unexpected spike in expenditure over the past few years aimed due to covid-19 related policies. However, the budget plan will also incorporate considerations related to multidimensional inequality and redistribution.

Aggregate revenue in 2023 is predicted to be 625.9 trillion won, and aggregate expenditure is predicted to be 639 trillion won. This represents an increase of 8.7% since 2018, although it is a 5.2% reduction compared to last year’s expenditure levels.

The Korean government’s 2023 budget plan consists of several policies relating to inequality reduction3. These take the form of several overarching aims which then contain various policies within them.

One of these aims is to reinforce the social safety net. The Korean government plans to increase the standard median income by 5.47% -- a move that will alter the threshold for receiving basic livelihood security and medical benefits, and thus increase the number of recipients. Subsidies to support the vulnerable in social security living, education, medical benefits, and housing will also increase. Furthermore, social insurance will be modified to help 280 thousand blind spot workers4, involving new insurance for housing fraud and subsidies for poor housing environments. As a result, income, health, and housing inequalities are expected to decrease.

A second aim is to protect the socially underprivileged. To this end, the government plans to raise disability benefits and disability pension benefits, as well as provide funding for 8 hour-care for those with developmental disabilities. It will also subsidise call taxis for those with disabilities, and medical procedures such as disability pre-detection for children.

The third aim is to balance regional development. The budget outlines new subsidies for farmers and fishermen who have been in the blind spot, that is, should receive subsidies but do not. In addition, the government plans to increase the low-population rural area budget, and build infrastructure such as universities, smart cities, research institutes, and administration departments in rural areas.

The fourth aim is to increase the fertility rate. The government will subsidise parents who take care of infants, and will increase the level of support targeting low-income households and single parents through the free provision of diapers, powdered milk, and other products. For single parents and double-income households, it will build more childcare infrastructure. Finally, the government will increase parental leave subsidies for both employers and employees.

The Korea Institute of Public Finance (KIPF) houses two main microsimulation models, known as KIPFSIM08 and KIPFSIM10. KIPFSIM08 focuses on taxation and fiscal policy effects, especially income tax, consumption tax, pensions, and health insurance, while KIPFSIM10 is a variant of KIPFSIM08 with a greater number of social insurance programmes incorporated.

KIPFSIM08 was created in 2008, with the intention of creating greater capacity within government to verify the expected success and efficiency of proposed policies. It came alongside the Earned Income Tax Credit, a popular policy in Korea at the time, for which the Korean government was keen to create simulations. The types of tax and transfers it accounts for are indicated in Figure ‎6.6.

When the model is run, it first gathers demographic variables such as age, gender, education level, and income. At the intermediate step, the tax policy is parametrized and estimated. Given these estimated tax policies, parameters such as labour supply elasticity are subsequently estimated. The user can then choose a scenario, such as a social pension increase, and calculate the income distribution inequality index before and after the application of a scenario.

The models are updated on an ad-hoc basis, dependent on researcher wishes. Both KIPFSIM08 and KIPFSIM10 are static models, although KIPFSIM10 is able to go some way towards estimating labour supply elasticity by looking at labour market participation. It is worth noting that both rely entirely on survey data, and as such suffer from several measurement issues. Furthermore, microsimulation models are not often used to aid in government decision making.

This section introduces four key public data sources used in distributional impact analysis: the National Survey of Tax and Benefits, the Korean Longitudinal Study of Aging, the Korean Education and Employment Panel, and the Korea Welfare Panel Study.

  • The National Survey of Tax and Benefit data consists of survey data to analyse tax policy and welfare expenditure effects. It consists of a number of variables including income, housing, household expenditure, earned income tax credits (EITC), transfer income, vehicle, education, and pensions. It consists of 5 634 households and uses both individual- and household-level data.

  • The Korean Longitudinal Study of Aging data focuses on the middle and old-age population. It consists of activities of daily living (ADL), medical service usage, body function, self-evaluated health, high blood pressure, and diabetes, among other factors. It consists of 10 254 households and uses both individual- and household-level data.

  • The Korean Education and Employment Panel is focused on the student population. It consists of education experience, university entrance, labour market participation preparation, and job training, among other factors. It consists of 2 000 middle school students, 2 000 vocational school students, and 2 000 high school students.

  • The Korea Welfare Panel Study is focused on poverty. It consists of working poor and near-poor population characteristics, labour participation status, policy benefits, and income and wealth of the poor, among other factors. It consists of 7 000 households, where 3 500 are low-income and 3 500 are general.

While there are various ways to improve distributional impact analysis, this section focuses on two of the most urgent issues faced when addressing the information needs that arise as a result of distributional analysis.

A key issue in assessing the distributional impacts is caused by difficulty in combining different data sources. As evident from above, most data sources focus on a specific subject and thus are only able to be uses in analyses of narrow topics. Given the increasing prevalence and importance of multidimensional inequality analysis, economists often require large and heterogenous sets of information. To control such heterogeneity, Korea would benefit from a unified large data set that includes all relevant related information. A major reason why this has not yet occurred is due to privacy concerns, which could potentially be addressed by replacing social numbers with private identification numbers, so that people cannot be directly identified.

The second key issue is that many current policies are the result of a multiplicity of slight additions and modifications over the years and are thus very complicated. This can in turn make research and analysis of policies very difficult, as the intended effects of these policies are often not clear. Given the major changes that have occurred in real world situations, it would be better to overhaul these policies entirely, and simplify them in the process.

To solve both problems, there is a need for a more integrated approach, bringing together the data component and the policy component. Regarding data combination, a central institution tasked with unifying otherwise disparate data sets, similar to those found in Sweden and the Netherlands, would be highly valuable. Moreover, regarding the simplification of public policies, it could be useful for the government to appoint policy design experts who could remove redundant policies and rearrange the existing ones in a well-organised way.


[1] Korean National Assembly Budget Office (국회예산정책처) (2017), 트렌드 세법 (Trend tax laws).

[3] Yeonhap News (2018), “한국 재정지출 소득재 분배 효과 작다…OECD 평균의 ⅓ 불과 [The income redistribution effect of Korea’s fiscal expenditures is small… Only ⅓ of the OECD average]”, https://webzine.kacpta.or.kr/news/articleView.html?idxno=2152 (accessed on 23 October 2023).

[4] 김준형 (Kim J.) (2018), 부모배경에 따른 교육불평등은 심화되었는가?: OECD 5 국가의 교육불평등 비교 분석 (Socio-Economic Educational Inequality is increasing?: A Comparative Study on Inequality among 5 OECD Countries).

[5] 선명재, 전병목, 전병힐 (Seong M., Jeon B., Jeon B.) (2009), 조세재정모의실험 모형: KIPFSIM08 모형의 구축 (Model of Tax Finance Simulation: Establishment of KIPFSIM08).

[2] 전규식, 정지수, 유경원 (Jeon G., Jeong, J., Yoo G.) (2016), 조세 재정지출 정책의 소득재분배 효과 분석 (Income redistribution policy and income inequality in Korea and OECD countries).


← 1. These indexes consist of child well-being, smoking ratio, exercise level, breakfast, eating fruit, average sleep time, body image distortion, sexual experience, unmet healthcare needs, self-rated health, BMI, stress assessment, depression, suicidal ideation, and suicide attempts

← 2. The other dimensions are health and education, income, consumption, and growth, security and environment, population and household, and housing, leisure, and culture.

← 3. Examples included basic livelihood security and medical benefits, disability activity support, promotion of the ease of transportation for mobility-disadvantaged persons, youth housing, and military basic pay.

← 4. Defined as workers who have not been enrolled into social security programmes.

Legal and rights

This document, as well as any data and map included herein, are without prejudice to the status of or sovereignty over any territory, to the delimitation of international frontiers and boundaries and to the name of any territory, city or area. Extracts from publications may be subject to additional disclaimers, which are set out in the complete version of the publication, available at the link provided.

© OECD/KIPF 2024

The use of this work, whether digital or print, is governed by the Terms and Conditions to be found at https://www.oecd.org/termsandconditions.