contents go

KDI - Korea Development Institute

KDI - Korea Development Institute



KDI FOCUS Assessing Old-Age Poverty with Income and Assets: Generational Insights and Policy Directions September 25, 2023


Assessing Old-Age Poverty with Income and Assets: Generational Insights and Policy Directions

September 25, 2023
  • 프로필
    Lee, Seunghee
This video provides English subtitles. Click on the video to watch more conveniently.
|   Related information   |
It is said that Korea is the worst internationally, surprisingly, the elderly poverty rate.
How did one out of two senior citizens end up in poverty?
We diagnosed elderly poverty by considering income and assets together!
|   Script   |
South Korea leads globally in a troubling metric: elderly poverty rates.

Based on income alone, the average elderly poverty rate in the OECD stands at 13.1%.
In stark contrast, South Korea records a daunting 43.4%.

Yet, there's more beneath the surface.
Might it be possible that many of the elderly have invested in assets, like real estate, to secure their later years?

Empirical data supports this notion.
A significant number of elderly households have predominantly invested in real estate,
and their total assets are valued between 300 and 500 million won.

This means relying solely on income data doesn't capture the complete financial situation of the elderly.

When assets are accounted for, the elderly poverty rate in Korea does indeed decrease compared to when only income is considered.
This suggests that a portion of the elderly population, even with limited incomes,
can leverage their assets as a means to escape poverty.

Nevertheless, even after accounting for assets,
Korea's elderly poverty rate remains alarmingly elevated on a global scale.

What's causing this?

According to the analysis by birth cohort, those born in the 1940s or earlier exhibit significantly high poverty rates.

Interestingly, even after accounting for age-related differences,
the economic position of younger seniors appears to be relatively better than that of older seniors.

Clearly, the high elderly poverty rate is largely influenced by these older cohorts.

This generational gap is also reflected among seniors classified as income-poor and asset-poor.
These disparities can be contextualized against the backdrop of Korea's rapid economic growth.
Specifically, the earlier generations earned income that paled in comparison to their successors,
resulting in noticeable income disparities.

Furthermore, the National Pension scheme, launched in 1988 and only becoming universally available by 1998,
has also contributed, as earlier generations had shorter enrollment durations, leading to reduced pension benefits.

Right now, Korea's Basic Pension covers about 70% of the elderly, given the high poverty rates.
But with more seniors and less impoverished new entrants, the system will face financial strains.
So, it's key to target the basic pension benefits at the most vulnerable with limited income and assets

To that effect, I suggest targeted benefits for seniors whose income, including their assets,
falls below a certain amount, and less support for those with more assets but limited income.

Those born before the 1950s should receive more support.
As less impoverished generations from the 1950s onward join the elderly demographic,
the size and strain on the Basic Pension system could naturally diminish.

Korea’s old-age poverty rate, based solely on income, ranks among the highest in the OECD. While the rate drops slightly when assets are considered, it remains persistently high. There are marked disparities in poverty rates across different elderly birth cohorts, with older groups facing higher rates. A greater share of the low-income, low-asset segment also marks the observed inequality. To effectively address poverty among the elderly, policy efforts should prioritize support for the oldest members of this vulnerable low-income, low-asset segment. Concurrently, financial assistance for seniors in the low-income, high-asset group should be scaled back. A practical move toward this direction would be revising the eligibility criteria for the Basic Pension to include asset-inclusive pensionable earnings.

Ⅰ. Old-Age Poverty in Korea

Korea’s old-age poverty rate1) is strikingly high, surpassing most other countries when measured by income metrics. In terms of disposable income, the poverty rate among the elderly decreased from 43.6% in 2016 to 37.7% in 2021, yet it remains above all other OECD countries. As shown in Figure 1, the OECD average stood at 13.1% in 2018, but Korea topped the chart with 43.4%. Korea’s high poverty rates among older people become more apparent when compared with its entire population: 17.6% in 2016 and 15.1% in 2021 in terms of disposable income. The disparity in poverty rates between the overall population and the elderly has narrowed slightly, moving from 26.0 percentage points (%p) in 2016 to 22.6%p in 2021. Nevertheless, this difference remains substantial.

Internationally, the old-age poverty rate typically exceeds the poverty rate for the broader population. However, Korea’s old-age poverty rate is exceptionally high, even in this global context. In spite of various policy measures introduced in Korea to combat this challenge, the issue of old-age poverty persists. Effectively addressing this requires policymaking informed by a detailed and accurate grasp of the current dynamics. With this perspective in mind, this study examines the issue in depth, segmenting the elderly demographic by generation, and offers a holistic assessment of their economic status based on both income and assets.


Korea’s old-age poverty rate was 43.4% in 2018, the highest among OECD countries (average 13.1%).

Ⅱ. Old-Age Poverty by Generation

To more closely examine old-age poverty in Korea, this study categorizes its elderly population into five generations based on five-year birth intervals: those born in the late 1930s (1935-39), both the early and latter halves of the 1940s (1940-44, 1945-49) and the two halves of the 1950s (1950-54, 1955-59). As of 2016, these groups corresponded to age ranges of 77-81, 72-76, 67-71, 62-66, and 57-61, respectively. By 2021, each group had shifted to the age bracket occupied by its predecessor in 2016. This generational analysis suggests that poverty rates are notably higher for those born in the 1940s and earlier.

Figure 2 illustrates the trajectory of the old-age poverty rate by generation from 2016 to 2021.2) With the exception of those born before the early 1940s, generational old-age poverty rates tend to increase over time. This trend is anticipated, given the natural decline in income from work and business as individuals age.

The poverty rate also remains elevated for those born in the 1940s or earlier, with a distinct disparity in poverty rates across the birth cohorts. By 2021, those born in the 1940s or earlier exhibited a poverty rate exceeding 40%, in contrast to rates below 30% for those born in the 1950s. The rate difference between those born in the late 1940s and early 1950s stands at 16.7%p, suggesting the diverging paths of poverty for the pre- and post-1950 cohorts.

These intergenerational disparities in old-age poverty rates become clearer when age differences are equivalized. Those born in the late 1940s were 72-76 years old in 2021, which is the same age bracket the early 1940s cohort occupied in 2016. Accordingly, this study compares the late 1940s cohort in 2021 with the early 1940 cohort in 2016, taking into account changing working conditions.4) The group aged 72-76 in 2021, born in the late 1940s, exhibits a poverty rate of 44.5%. In contrast, the early 1940s group, when adjusted to the same agerange, has a higher rate of 51.3%. This finding suggests that younger
generations, when age-adjusted, are in a relatively better economic position than their predecessors. Such a trend can be observed across all generations, not just those born in the 1940s, indicating that more recently born elderly groups are better off in terms of poverty.


The old-age poverty rate varies markedly across the five elderly generations, with those born in the 1940s or earlier having the highest rates.

Korea’s old-age poverty rate after the 2010s has displayed two distinctive characteristics: (1) the modest decline in the overall oldage poverty rate and (2) a noticeable gap in the poverty rate between those aged 75+ (“older old”) and those aged 65-74 (“younger old”). Both these features relate directly to the generational differences in poverty rates discussed earlier. That is, the younger the elderly group, the lower their poverty intensity. In particular, the generation born in the 1950s displays a substantially lower poverty rate than the generations that came before them. The proportion of individuals born in the 1950s has been consistently rising since the 2010s. As they exhibit the lowest poverty depth among the elderly generations, their increasing numbers have substantially influenced the overall decline in the old-age poverty rate. Moreover, as the 1950s-born constitute the younger old, their transition into the older old, coupled with their growing proportion, explains the disparity in poverty rates among the elderly groups.

A more detailed look reveals that the old-age poverty rate, based on disposable income, dropped from 43.6% in 2016 to 37.7% in 2021. This 5.9%p decline can be broken down into intra-generational poverty trajectories (aging → income decline → higher poverty rate) and inter-generational population transitions (the growth of less impoverished generations). This decomposition indicates that demographic shifts contributed to a 7.5%p decrease in the poverty rate, offsetting the 1.6%p increase caused by inter-generational poverty dynamics.

Further looking into the demographic changes between 2016 and 2021 reveals that the share of older people from the 1940s and earlier decreased from 81.7% to 52.6%. In contrast, the percentage of the 1950s-born grew from 18.3% to 47.4% (Survey of Household Finances and Living Conditions). This substantial increase in the proportion of the relatively less impoverished 1950s-born cohort is a major factor driving the reduction in the overall old-age poverty rate during this timeframe.

During this period, the poverty rate for the younger old (65-74) decreased from 33.9% to 27.6%, while the rate for the older old (75+) marginally declined from 56.8% to 51.0%, though hovering around the 50% range. The gap between these two groups stands at roughly 24%p. The older old consists of individuals born in the 1940s or earlier, a group that typically faces higher poverty rates. As those born in the 1950s increasingly make up a greater portion of the younger old category, the poverty disparity between the two age segments has widened, even if only slightly.

The recent decrease in the old-age poverty rate and the disparities seen across age cohorts can likely be attributed to the growing proportion of individuals born in the 1950s who tend to be less impoverished.

Ⅲ. Old-Age Poverty Measured by Income and Assets

While income assessment offers insights, income alone may not capture the complete economic conditions of the older population. Korea’s National Pension Service, established in 1988, is still evolving as the primary safety net for seniors. It was only in 1998 that the national pension scheme extended coverage to all citizens. As a result, many in the elderly demographic have only participated in the pension scheme for a short period, receiving modest pension payouts. As of 2018, public transfers constituted only 25.9% of older people’s income, not even half of the OECD average of 57.1%. Since it is difficult to secure adequate retirement income through pension benefits, there is a strong possibility that many older Koreans rely on asset accumulation for their retirement strategies. Supporting this notion, Statistics Korea’s Social Survey indicates a heavy reliance on assets, especially real estate, for retirement planning. Such a tendency suggests that low-income seniors with substantial assets might leverage those assets to make up for income shortfalls. Thus, an analysis of old-age poverty requires a holistic understanding of their financial situations, considering both their income and assets. 

Table 1 presents the asset composition for households led by individuals aged 65+ based on the 2016-2021 Survey of Household Finances and Living Conditions.5) Their portfolio is heavily weighted towards real estate, with the share of the overall asset mix remaining consistent throughout the survey period. On average, older households in Korea typically possess assets valued between 350-500 million won, accompanied by modest liabilities ranging from 30-40 million won. Liquid financial assets constitute approximately 16% of the total assets, with the majority being tangible, chiefly real estate. Real estate represents over 80% of the assets for older households—a percentage that is notably high compared to other major economies. Table 2 provides a comparative asset breakdown for older households in major countries, referencing the Luxembourg Income Study. While real estate is commonly a substantial asset class in many countries, those other than Italy and Germany tend to have a higher proportion of liquid assets than Korea. Other tangible assets,7) such as automobiles and expensive durables, account for just 2-3% of the total assets.


Given the nascent state of public transfers in Korea, asset accumulation is the likely channel of retirement planning, meaning income alone does not provide a complete picture of old-age poverty.

In poverty analysis of both income and assets, a typical approach involves converting assets (stock) to income-equivalent (flow).8) This study employs comprehensive incomization and annuitization for such transformation. The term “comprehensive income” covers actual earnings and implicit sources of income like imputed rent.9) Within this context, “income” refers to the total value consumable annually without drawing down on assets. In other words, comprehensive incomization reveals benefits not captured from cash income, such as the implicit rental income from imputed rent.10) In applying comprehensive incomization, this study adopts asset-specific methodologies. Real estate, for instance, is categorized into residential and non-residential types. The rent equivalent method11) is used for the comprehensive incomization of residential properties, whereas the user cost method12) is chosen for non-residential properties. For financial assets, this study relies on the interest income data from the Survey of Household Finances and Living Conditions.

Annuitization, on the other hand, diverges from comprehensive incomization as it involves asset consumption. One representative method of annuitization transforms net worth, calculated as the difference between total assets and liabilities, into regular pension payouts, which are then classified as income. In essence, annuitization captures benefits that, while not actively realized, become accessible through mechanisms that leverage owned assets like reverse mortgages. Analyzing old-age poverty using this approach offers insights into the possible economic standing attainable when fully leveraging one’s assets.

Measuring the old-age poverty rate by incomizing assets, as shown in Figure 3, substantially lowers the rate compared to calculations using income alone. The old-age poverty rate after applying comprehensive incomization shows a yearly reduction of 7-8%p compared to the rate based on disposable income. This decline suggests that when assets are factored into a broader economic assessment, a substantial number of older individuals remain above the poverty line. In the case of annuitization, the decrease is even more pronounced, with a reduction of 14-16%p annually. These findings underscore the ability of older people to effectively utilize their assets to self-navigate out of poverty.


When incorporating both income and assets into the analysis, the old-age poverty rate determined by asset incomization is markedly lower than that derived from income alone but remains elevated.

To thoroughly examine the economic conditions of the older population, this study segments them into four income-asset categories. It should be noted that the definition of “poor” can vary depending on whether it is based on disposable income or comprehensive income. Specifically, a senior might be classified as poor based on disposable income but not when evaluated using comprehensive income. Such a scenario arises when a senior, despite a low income, has cash flows from owned assets that push them above the poverty threshold. Older people deemed poor according to disposable income are labeled “low-income, low-asset” if they also fall below the poverty line based on comprehensive income. Conversely, if they remain above the comprehensive income threshold, they are categorized as “low-income, high-asset.” The classification is completed with the additional categories “high-income, high-asset” and “high-income, low-asset.”14)

Figure 4 depicts the trends of these four elderly groups based on their income and assets. The share of the low-income, high-asset group remains steady at about 10% throughout the entire survey period. The percentage of the low-income, low-asset group has decreased from 33.8% in 2016 to 27.7% in 2021, yet remains elevated. In terms of income, both the low-income, high-asset and low-income, low-asset groups belong to the low-income category, but their assets differ considerably in size. Given that economic status hinges on both income and assets, the economically vulnerable are best identified as those in the low-income, low-asset category within the low-income bracket. Older people in this group face financial challenges due to scarcity in both income and assets, underscoring the need for additional government support.

Korea’s international ranking in old-age poverty stays high when assessed through asset incomization. For a global comparison, this study analyzes old-age poverty rates, determined by asset incomization from different countries, using data from the Luxembourg Income Study.15) Figure 5 presents the old-age poverty rates for eight countries,16) including Korea, based on disposable income, comprehensive incomization, and annuitization. Anglo-American nations show a marked reduction in old-age poverty rates when asset incomization is considered. In contrast, continental European regions, known for their generous public transfers, see minimal change in rates due to asset incomization. When assets are considered, Korea experiences a notable reduction in its rate, narrowing the disparity with other countries, though it continues to rank higher. These findings underscore the marked differences in old-age poverty levels between Korea and other countries, even when taking a holistic view that includes both income and assets.


Korea’s old-age poverty rate, evaluated using both income and assets, still ranks among the highest in the world.

The analysis presented earlier, which focused on inter-generational poverty rates among older adults, requires reevaluation by incorporating assets into the equation. When factoring in assets, the differences within the low-income, low-asset group become more pronounced across generations, especially among those born in the 1940s and earlier.

Figure 6 displays the proportions of the vulnerable low-income, low-asset group broken down by generation. The generational disparities in Figure 6 are more apparent than those in Figure 2. As of 2021, individuals born in the late 1930s and early 1940s exhibit poverty rates above 50% when measured by disposable income alone. However, when assets are taken into consideration, these rates decline to 45.9% for the late 1930s-born and 37.2% for the early 1940s-born. This pattern persists across other generations, with the later-born generations showing a smaller proportion of vulnerable individuals than their predecessors.


When incorporating assets into the analysis, the share of the low-income, low-asset varies significantly across generations, particularly among those born in the 1940s and earlier.

The variation in the proportion of vulnerable elderly across generations, when considering both income and assets, indicates that economic circumstances are not consistent among older generations. Such a gap becomes particularly evident when comparing those born before 1950 to their later-born counterparts. As of 2021, the proportion of the low-income, low-asset group from the 1940s and earlier exceeds 30%, whereas that of those from the 1950s stands below 20%. This data suggests a distinct difference in old-age poverty patterns for cohorts born before versus after 1950. 

Differences in old-age poverty across generations, even when accounting for income and assets, can be traced back to Korea’s rapid economic growth and the varying maturity levels of the oldage security system among elderly cohorts. Figure 7 displays the lifelong GNI per capita for five specific birth cohorts: 1935, 1940, 1945, 1950, and 1955. Though separated only by five years, these cohorts exhibit substantial differences in GNI per capita throughout their lives. For instance, at age 30, the 1945 cohort had a GNI per capita of $613, whereas the 1950 cohort’s GNI per capita was nearly threefold at $1,699. This disparity continues throughout their respective lifespans.17) The data consistently shows that each generation’s income level is generally lower than the one that follows. The 1960s and 1970s marked a period of soaring land prices and stark income inequality (Lee and Hwang, 1998), which suggests that those born before 1950 were more likely to encounter higher rates of poverty due to their lower income levels and obstacles in asset accumulation compared to subsequent generations. In addition, the varying progression of the old-age security system’s development across these cohorts has influenced these generational disparities. With the national pension only being extended to the entire population in 1998, earlier generations had shorter enrollment durations and consequently received reduced benefits, leading to inadequate retirement income.18)


The difference in levels of old-age poverty across generations can be attributed to the intergenerational income disparity caused by rapid economic growth and the differing maturity levels of retirement security systems for each generation.

Ⅳ. Policy Directions for Old-Age Poverty

Severe old-age poverty, ranking among the highest in the OECD, poses a grave social issue for Korea. While factoring in asset incomization reduces the poverty rate compared to measurements based solely on disposable income, the rate remains alarmingly high on an international scale. The crux of Korean old-age poverty is the high proportion of the older demographic that falls into the vulnerable low-income, low-asset category. Most of these individuals were born in the 1940s or earlier, which means they are now advanced in age and in urgent need of government intervention.

Korea’s hallmark poverty alleviation measure for the elderly is the Basic Pension program. This initiative provides pension payouts to about 70% of the elderly population, in line with its policy goal of countering the high rate of old-age poverty. However, given Korea’s rapidly aging population, the Basic Pension is set to face escalating fiscal challenges due to the rising number of beneficiaries. Moreover, the program fails to effectively supplement the incomes of the most vulnerable elderly. Future strategies should thus focus on providing targeted assistance to these high-risk groups, ensuring they receive more robust support.

To this end, the income support mechanism should downsize the eligibility criteria for beneficiaries, placing emphasis on their capacity to monetize assets. Seniors in the low-income, high-asset group can leverage policies like housing or farmland pensions to pull themselves out of poverty. Accurately determining one’s financial standing based on income and asset securitization becomes vital in this regard. Specifically, in addition to efforts to publicize and inform the elderly about asset-backed pension arrangements, the beneficiary pool of the Basic Pension should be scaled down. Such a reduction involves replacing the converted value of assets into income within pensionable earnings with potential income that might arise from asset securitization during the pension application process.

In order to prioritize financial assistance to those in most need, the Basic Pension should be earmarked solely for elderly individuals below a specific threshold of asset-adjusted pensionable earnings. Adopting this strategy would particularly benefit those born in the 1940s or earlier. Besides typically being less affluent, these individuals have not fully reaped the benefits of the National Pension in the way that later generations have, largely due to Korea’s rapid economic growth. To guarantee stable retirement income for them, the Basic Pension needs adjustment. Eligibility should depend on a set

percentage of pensionable income, which would enhance benefits for qualified individuals. This method would also naturally shrink the Basic Pension’s recipient pool as wealthier generations from the 1950s and onward become eligible. Redirecting a significant portion of the funds set aside for the first-tier pension to other elderly care programs would be a strategic move to elevate the overall well-being of the older demographic

To provide targeted assistance for the most vulnerable, the Basic Pension should be based on asset-adjusted pensionable earnings, reducing the support for seniors in the low-income, high-asset category.

Choi, Kyung-jin and Byungkwon Lim, “Comparative Analysis of Old-Age Income Security Utilizing Owner-Occupied House” Financial Planning Review, 13(1), 2020, pp.1~28 (in Korean).
Lee, Joungwoo and Seonghyeon Whang, “The Problems of Income Distribution and Related Policy Issues in Korea” Policy Study, 20(1), 1998, pp.153~230 (in Korean).
Lee, Seunghee, “Analysis of Elderly Poverty Considering Income and Assets” in Youngwook Lee, et al., Current Status of Elderly Poverty Considering Assets and Policy Implications, Ministry of Economy and Finance, 2023 (in Korean).
Lee, Soowook, et al., What Is to Be Done for the Mid-20s to 30s in an Era of Low Growth? : Dilemmas of Housing Market, Korea Research Institute for Human Settlements, 2015 (in Korean).
Lee, Youngwook, et al., Current Status of Elderly Poverty Considering Assets and Policy Implications, Ministry of Economy and Finance, 2023 (in Korean).
OECD, Pensions at a Glance 2021: OECD and G20 Indicators, OECD Publishing, Paris, 2021(
Wolff, E. N. and A. Zacharias, “Household Wealth and the Measurement of Economic Well-being in the United States,” Journal of Economic Inequality, 7, 2009, pp.83~115.
Yun, Suk-myung, et al., A Study on The Assessment of Various Elderly Poverty Indices (Ⅰ), Korea Institute for Health and Social Affairs, 2017 (in Korean).

< Sources >
Banks of Korea
, Economic Statistics System (, Accessed: Sept. 2023).
Luxembourg Income Study, “LWS (Luxembourg Wealth Study) DB” (Obtained: Aug. 2023).
National Pension Service, “National Pension Statistics,” 2023. Statistics Korea, “Social Survey,” 2023.
Statistics Korea, “Survey of Household Finances and Living Conditions,” 2017~21.

  • Ⅰ. Old-Age Poverty in Korea

    Ⅱ. Old-Age Poverty by Generation

    Ⅲ. Old-Age Poverty Measured by Income and Assets

    Ⅳ. Policy Directions for Old-Age Poverty
related materials ( 9 )
  • Key related materials
Join our Newsletter

World's Leading Think Tank, Korea Development Institute

Security code

We reject unauthorized collection of email addresses posted on our website by using email address collecting programs or other technical devices. To access the email address, please type in the characters exactly as they appear in the box below.

KDI Staff Information

Please enter the security code to prevent unauthorized information collection.

KDI Staff Information

Please check the contact information.

KDI Staff Information

Please check the contact information.