In this blog post, we will examine how South Korea’s welfare system has evolved over time and, based on the pros and cons of universal and selective welfare, discuss the need for “targeted welfare” to utilize limited resources more accurately and efficiently.
Universal welfare has long been a key social agenda in South Korea. In particular, in the early 2010s, as elections approached, political parties and candidates put forward various welfare pledges, causing the expansion of welfare to emerge as a major political and social issue. At that time, there was also intense debate over the pros and cons of universal and selective welfare systems, particularly regarding issues such as free childcare and free school meals. Debates surrounding the scope of welfare and methods of funding—including the Seoul Metropolitan Government’s referendum on free school meals—demonstrated that welfare is not merely a policy issue but a question of how society allocates resources and shares risks.
One reason universal welfare gains support is that it provides benefits to a specific group without discrimination. By not separately screening beneficiaries, the likelihood of being stigmatized as a welfare recipient is reduced, and the administrative process for verifying eligibility can be made relatively simpler. Furthermore, by looking to countries with high levels of social security—such as the welfare systems in Nordic nations—the idea naturally took hold that economically developed nations should also expand their welfare systems. However, today’s welfare debate in South Korea should not be limited to the question of whether to expand welfare, as it was in the past. This is because pressure on welfare spending is increasing as the population ages rapidly. The OECD has also pointed out that by 2026, South Korea will need to address the fiscal pressures resulting from an aging population while ensuring long-term fiscal sustainability.
However, in advocating for universal welfare, we must be wary of an overly optimistic view of welfare itself. It is problematic to assume that expanding welfare will solve all of society’s problems. Welfare requires substantial financial resources, which are ultimately secured through various means, such as taxes, social insurance premiums, and the national budget. Therefore, it is just as important to consider who will receive support, how it will be provided, and to what extent, as it is to increase the scale of welfare spending. In particular, as South Korea faces both a low birth rate and an aging population simultaneously, the financial burden is likely to grow significantly in various areas—such as elder care, healthcare, and pensions—in the future. The OECD also emphasizes that, to ensure Korea’s long-term fiscal sustainability, it is necessary to manage the spending pressures resulting from an aging population and establish a long-term fiscal framework.
Some view today’s society as a structure in which the privileged and the wealthy monopolize resources and propose universal welfare as a solution to this problem. Indeed, various forms of inequality exist in Korea, including gaps in income and wealth, the burden of housing costs, and disparities in educational opportunities. However, when discussing the effectiveness of welfare policies, it is essential to specifically examine the extent of social inequality, identify which social groups face what problems, and assess how much a particular welfare policy can actually improve those issues. Expanding welfare policies should not be equated with solving social inequality in and of itself. This is because policies aimed at improving the quality of education, reducing labor market disparities, or expanding opportunities for housing and asset accumulation can sometimes provide more direct solutions.
Furthermore, evaluations of welfare policies should not be swayed by personal preferences or antipathy toward specific political factions or governments. Criticizing a particular government’s policies and evaluating the effectiveness of welfare policies are separate matters. In the past, there have been instances where distrust of the government or political conflict became intertwined with evaluations of welfare policies, causing partisan politics to take precedence over the substance of the policies. Welfare policies should be evaluated based on who actually benefits and how, whether the necessary funds can be secured on an ongoing basis, and whether there are any unforeseen side effects—regardless of one’s personal feelings toward specific politicians or political parties.
Blind faith in welfare is dangerous. If we are not careful, we may fail to achieve the expected improvements in living standards even after investing massive amounts of public funds. Even when expanding welfare, we must first adopt the right attitude toward it. Advocating for welfare based on vague optimism or by ignoring the complexities of reality increases the likelihood of making poor decisions. Welfare spending is not just a problem for the current generation. If we fail to establish a fiscally sustainable structure in the long term, future generations may have to shoulder an even greater burden. However, this does not mean that welfare should be cut unconditionally simply because of fiscal constraints. South Korea’s welfare system plays a crucial role in ensuring that society collectively shares the risks—such as poverty, illness, unemployment, disability, and old age—that individuals cannot bear alone. The issue is not whether to expand welfare, but rather how to design the right welfare programs and implement them effectively.
So, what are the limitations of the welfare system? First, it is fundamentally difficult for a welfare system to perfectly reflect reality. This is because welfare systems operate based on codified criteria, yet it is difficult to accurately assess an individual’s economic hardship based on criteria alone. Determining whether a person needs support is not a matter that can be clearly judged by a single number. Even with the same income, actual living conditions can vary greatly depending on factors such as debt, housing costs, family relationships, health status, caregiving burdens, and employment status. Therefore, the moment specific criteria are established, it becomes difficult to perfectly reflect the complex realities of every individual.
For example, South Korea’s National Basic Living Security System is a representative social safety net that provides benefits such as subsistence, medical care, housing, and education to citizens facing financial hardship. However, the system described in the 2012 original text has changed significantly from the current one. The current National Basic Living Security System has moved away from the previous practice of uniformly applying the minimum cost of living and now uses the median income standard to determine eligibility criteria for each type of benefit. The 2026 median income benchmark has been set at 2,564,238 won per month for a single-person household and 6,494,738 won per month for a four-person household. The eligibility thresholds are set at 32% of the median income benchmark for subsistence benefits, 40% for medical benefits, 48% for housing benefits, and 50% for educational benefits. The eligibility threshold for livelihood benefits for a four-person household in 2026 is 2,078,316 won per month.
The current system does not rely solely on income. Recognized income is calculated by combining assessed income and the income equivalent of assets, taking into account not only actual income but also assets, debts, and household-specific expenses. In other words, South Korea’s Basic Livelihood Security System has evolved to incorporate multiple factors in order to reflect reality as accurately as possible. However, no matter how diverse the criteria applied, it is impossible to perfectly measure an individual’s actual living conditions.
In particular, the “obligated support provider” criterion under the former National Basic Livelihood Security System has long been a subject of controversy. This is because, even if an individual’s income was low, they might not receive sufficient support if they were deemed to have a certain capacity to support their family. To address this issue, South Korea’s Basic Livelihood Security System has relaxed these criteria on several occasions. In 2026, the “support ratio” for medical aid was abolished for the first time in 26 years, a measure taken to alleviate the problem of low-income individuals who do not actually receive support from their families being excluded from the system. This demonstrates that the criteria for welfare systems are not fixed but must be continuously revised in accordance with changing realities.
There is nothing inherently wrong with the idea of verifying an individual’s income level or, if their income is low, determining whether family members are actually providing assistance. This is because certain criteria are necessary to prioritize the allocation of limited resources to those who need them most. The problem lies in the assumption that these criteria can perfectly capture reality. Any system must have certain criteria, and it is inevitable that those criteria cannot fully account for every circumstance in reality. Ultimately, what matters is not eliminating the criteria, but continuously identifying and correcting the errors that arise from them.
Of course, the problem can be mitigated through more precise criteria. The current Basic Livelihood Security System already applies different eligibility criteria for each benefit type, taking into account factors such as household size and recognized income. This is intended to reflect the reality that larger households require more living expenses. For example, even with the same income, the living conditions of a single-person household and a four-person household differ significantly, so it is reasonable to consider the number of household members. However, even these criteria cannot perfectly reflect the complexities of reality. In fact, if the criteria become overly complex, it can make it difficult for citizens to understand the system and lead to increased administrative costs.
In the past, there were cases where people altered their family relationships or household structures to circumvent the system’s criteria during the process of calculating income on a household basis. This was because, in situations where the combined income of family members resulted in exclusion from assistance, splitting the household could lead to a different income calculation. Ultimately, the household-based income criteria—designed to ensure more accurate assessments—had the potential to trigger unexpected behaviors in practice. Such cases demonstrate that when establishing criteria for welfare systems, we must also consider the outcomes that arise from the interplay between policy objectives and people’s actual behaviors.
As such, it is difficult to accurately assess reality based solely on fixed criteria. It is akin to trying to wrap the uneven surface of an object precisely with a single sheet of wrapping paper. No matter how smooth or stretchy the paper is, it is difficult to perfectly cover every curve of the object. There will inevitably be parts that remain uncovered or where the paper bulges.
Criteria are used to ensure legal legitimacy and administrative consistency. They also serve as a tool for allocating limited budgets according to certain principles. Therefore, while overreliance on criteria can lead to problems, eliminating them entirely is not a realistic alternative. Setting strict criteria can reduce unnecessary spending, but it may also exclude those who truly need support. Conversely, relaxing criteria too much can significantly expand the pool of eligible recipients, thereby increasing the financial burden. Ultimately, criteria are useful only when they are appropriately designed and continuously refined to address errors that arise during actual implementation. This is why South Korea’s Basic Livelihood Security System continues to be adjusted in 2026 to reflect changes in the median income and system improvements.
Second, welfare systems can inherently have inefficient aspects. Even if a welfare system can identify eligible recipients with relative accuracy, it cannot always assist them in the most efficient manner. This is because operating the system incurs costs, administrative expenses are incurred to verify eligibility, and a one-size-fits-all approach to support makes it difficult to meet all the diverse needs of beneficiaries. However, this should not be generalized to mean that welfare systems themselves are inefficient. If the benefits of reducing social risks and alleviating poverty outweigh the administrative costs, then the costs of operating the system may be worth bearing for society.
First, there must be people to implement and manage the system. Personnel costs and administrative expenses are required to support their work. Verifying eligibility also incurs costs. For example, programs such as unemployment benefits are paid only after certain eligibility requirements and procedures are met. Recipients must submit relevant documents or materials to verify their job search activities, and the responsible agency reviews the submitted information before deciding whether to make a payment. While these procedures are necessary to prevent fraudulent claims and ensure the fairness of the system, they also represent costs associated with operating the program.
Even if these additional costs are covered, it may not be possible to fully satisfy the needs of beneficiaries. This is because everyone has different needs. When providing childcare support, for instance, some people may prefer daycare centers, while others may want assistance that allows their children to be cared for at home. When providing support for the elderly, some may want healthcare services, while others may prefer leisure activities or daily living assistance. Providing cash assistance can meet a variety of needs, but it raises the problem of verifying whether the money is being used in accordance with the policy’s objectives. Conversely, providing in-kind benefits or services clarifies the intended use but makes it difficult to fully reflect individual differences in needs. Furthermore, while beneficiaries’ needs change rapidly, the system must navigate laws, budgets, and administrative procedures, which can slow the pace of change.
In some cases, the system may even create new demand. When a specific service is provided for free or at a low cost, people who previously did not use it may begin to do so. While this is not necessarily a bad thing, it means that policymakers must account for unexpected increases in usage and the resulting financial burden when designing policies. Welfare systems not only fail to perfectly satisfy beneficiaries’ needs but, in some cases, can also alter people’s behavior and create new demand.
Welfare systems are not a panacea. No system can be. While establishing a system ensures a certain degree of fairness and predictability, it has limitations in reflecting individual circumstances in detail. Welfare systems have the advantage of providing swift and consistent welfare services to the entire population or specific target groups. However, in the process of applying eligibility criteria, those who are most in need may be excluded, while, conversely, those with relatively lower needs may be included. No matter how sophisticated the criteria, it is difficult to completely eliminate such shortcomings. A certain degree of error is inevitable when institutionalizing the complexities of human life. A system is just a system.
Furthermore, institutionalizing welfare can lead to inefficiencies in its operation. It incurs labor and administrative costs, and both administrators and beneficiaries must endure a certain amount of effort to verify eligibility. Even then, it may fail to fully satisfy the needs of recipients. These are all drawbacks inherent to institutional systems. While the market operates on the concept of the “invisible hand,” the process by which the government distributes welfare according to codified standards involves a different method of resource allocation than the market. Therefore, welfare systems should not be expanded or reduced unconditionally; they should be adopted only when the social benefits gained through the system sufficiently outweigh the administrative costs and side effects. A welfare system is not a wall one can lean on for everything.
While this is a theoretical point, there are cases where people can be more accurate than systems. The judgment made after directly observing someone in need, understanding their circumstances, and then offering help can be more nuanced than a system that merely applies predetermined criteria. Even if it is difficult to logically explain why help is being provided, the human ability to assess a situation based on actual observations can reflect the complexities of reality. There are many people in Korean society who need assistance. Helping such people directly and helping them through the welfare system each have their own advantages and disadvantages. The state’s welfare system cannot completely replace voluntary assistance from individuals.
However, compared to calls for universal welfare, the argument for directly helping neighbors in need is often heard less frequently. Many people expect the government to operate welfare systems rather than taking action themselves. Of course, it is necessary for the state to operate welfare systems, because poverty, illness, unemployment, disability, and issues related to aging across society cannot be solved by the goodwill of individuals alone. However, it is also undesirable to assume that the state will solve every problem.
If South Korean citizens do not take action on behalf of the vulnerable, how can the state act sufficiently? The idea of creating a welfare system and entrusting all responsibility to it, rather than providing direct assistance, has its limitations. After all, the welfare system is ultimately funded by taxes and social insurance premiums paid by South Korean citizens. Therefore, as calls for expanding welfare grow, discussions must also take place regarding who will bear the financial burden and how. While the “everyone pays, everyone receives” model cannot be explained solely as a zero-sum game in the strictest sense, the relationship between funding sources and benefits cannot be ignored. Welfare policies must take into account the relationship between those who bear the financial burden and those who receive benefits, as well as intergenerational burdens and redistribution effects across income groups.
Successful welfare requires a social consensus in which people are willing to pay taxes and use those taxes to help those in need. What Korea may need even more than welfare is a revival of concern and solidarity toward the vulnerable. Before discussing free childcare, we need to examine what families who actually need care are facing; before discussing free school meals, we need to look at the conditions in which children are living at school. The state’s welfare system and citizens’ voluntary sharing should not be seen as mutually exclusive but rather as complementary.
This does not mean we should abolish the welfare system. Rather, it means acknowledging the system’s limitations and making full use of its beneficial aspects. We must abandon blind faith in welfare and utilize the system rationally. Social security systems, such as the National Basic Livelihood Security System, provide a certain level of welfare services based on nationwide standards and serve as an important social safety net in Korea. Basic social security measures that everyone can agree on should ideally be implemented at the national level. When the central government implements these measures using consistent standards, it can also reduce regional disparities and administrative duplication. However, existing welfare systems must be refined to better suit current realities.
South Korea’s current welfare system has changed significantly since 2012, when this article was originally written. The National Basic Livelihood Security System was reorganized into a tailored benefits system in 2015 and has since evolved to apply eligibility criteria for each benefit based on the median income. In 2026, the median income for a four-person household was set at 6,494,738 won, and the eligibility threshold for livelihood benefits is 2,078,316 won. Medical, housing, and educational benefits also apply different criteria. Furthermore, in 2026, the medical benefit dependency allowance was abolished for the first time in 26 years. Therefore, one should not evaluate current South Korean welfare policies based on the welfare system of 2012. The welfare system is also continuously being improved to keep pace with societal changes.
However, simply making the system more sophisticated does not solve all problems. If eligibility criteria are defined too simplistically, they fail to reflect the complex realities; conversely, if the criteria are made excessively complex, administrative costs rise and it becomes difficult for the public to understand the system. As the number of welfare programs increases, so too may the administrative and information management costs required to operate each program separately. Therefore, simply adding new programs is not the only solution to expanding welfare. It is necessary to evaluate what effects existing programs are actually having on specific individuals.
Rather than that, we must design accurate and efficient welfare systems that fully reflect the complex reality, even if it is more difficult. In the past, search engines like Google garnered attention for their approach of analyzing numerous variables to improve the accuracy of search results. For web documents, they determined the order of search results by utilizing various factors, such as the position of search terms and the links between documents. Technologies like PageRank were representative methods for determining the importance of a page by analyzing the links between web pages. The key lesson from these examples is that synthesizing multiple pieces of information to make a judgment—rather than evaluating a complex subject based on a single, simple criterion—can yield more sophisticated results.
Welfare systems also need to be approached from this perspective. First, we must discuss sophisticated filtering methods. Without accurately distinguishing who is eligible for support, it is difficult to implement welfare effectively. Even a policy that is ideologically sound can lead to both fiscal waste and gaps in welfare coverage if it fails to accurately identify the intended beneficiaries. Rather than relying solely on a single criterion—such as “the bottom X percent of income” or “income below a certain amount”—we should consider methods that comprehensively take multiple factors into account. By utilizing various variables appropriate to the policy’s objectives—including not only income and assets but also household composition, housing costs, age, disability status, caregiving burdens, and employment status—we can assess the need for support with greater precision. In fact, South Korea’s Basic Livelihood Security System is currently operated by considering both assessed income and the income-equivalent value of assets.
In particular, there is a need to leverage information technology. While big data was introduced as an emerging technology in 2012, in South Korea in 2026, the use of data and artificial intelligence in public administration is no longer a story of the distant future. The Ministry of the Interior and Safety is promoting systems and guidelines for AI utilization in the public sector to expand AI- and data-driven administration by 2026, and in the welfare sector, it is proposing a direction to use AI to proactively guide citizens toward necessary services and provide personalized counseling.
By utilizing these technologies, it is possible to link various administrative datasets to identify those in need of support more quickly and help reduce duplication or omissions in welfare services. It is now possible to identify eligible recipients by synthesizing diverse information, moving beyond the traditional practice of simply verifying income and assets separately. However, applying data technology to welfare does not mean that all information about South Korean citizens can be collected and analyzed without restriction. When utilizing personal information in welfare administration, factors such as personal information protection, data accuracy, prevention of use for purposes other than those intended, algorithmic bias and discrimination, and procedures for appealing automated decisions must all be considered. This is because as technology becomes more sophisticated, there is a greater risk that incorrect data or biased criteria will be applied more rapidly and extensively. Therefore, IT technology should not be used merely to store and display citizens’ information but should also be utilized to analyze and improve the system, while simultaneously upholding the principles of transparency and personal data protection. To provide services suited to an increasingly complex era, South Korea’s welfare system must also embrace artificial intelligence and data technology.
Furthermore, we need to place greater emphasis on improving existing systems rather than creating new ones. If we focus solely on solving new problems, the issues within existing systems may remain unaddressed. Just as one’s skills improve by repeatedly solving the wrong problems, in policy-making, we must identify what went wrong and make iterative corrections to minimize the same mistakes. Even great inventions or works of art go through countless trials and revisions before they are completed. The same applies to the welfare system. It does not end with the creation of a system; rather, we must analyze the results of its implementation, identify unforeseen side effects, and continuously improve it by reflecting the experiences of actual beneficiaries. If we simply create systems blindly, the administrative system will become more complex, overlapping programs will increase, and waste may grow. It is important to steadily refine and develop a single system to nurture it into an effective one.
To achieve this, public attention must be directed toward improving these systems. While experts and relevant agencies are currently involved in the design and evaluation of South Korea’s welfare policies, it remains difficult for the general public to easily verify the specific details and outcomes of these systems or offer their opinions. When policy evaluations are centered on reports and academic papers, they can be inaccessible to the general public. However, there is no reason why the public cannot take an interest. We must develop these systems by combining the interest and ideas of the South Korean public with the analysis and concrete proposals of experts. Welfare debates on the internet should also move beyond merely repeating arguments for and against; they must evolve into discussions about which systems have produced what effects for whom, how much they have cost, and what areas need improvement. The government, too, needs to encourage public participation. Information should be consolidated to provide a clear overview of the welfare systems currently in place, and it should be disclosed in a way that is easily understandable to the general public. In particular, as the use of artificial intelligence in public administration expands, it is important to increase transparency so that citizens can understand what information is used to recommend specific services to them. The Ministry of the Interior and Safety’s plan to expand AI-powered, personalized public services by 2026 is one example of this trend.
Let’s now summarize the discussion using the terms “mass-destruction-style welfare” and “precision-strike-style welfare.” In 2003, the United States invaded Iraq based on the assessment that Iraq possessed weapons of mass destruction. However, no stockpiles of weapons of mass destruction sufficient to justify the war were subsequently discovered, and criticism regarding the war’s legitimacy and the intelligence assessments continued. In 2011, the United States officially ended U.S. military operations in Iraq and completed its troop withdrawal. Therefore, the phrase “officially declared an end to the war last year,” which appeared in the original text, is no longer applicable at this point in time. Rather than serving as a current example to illustrate welfare policy, the Iraq War should now be understood in a metaphorical context—as a reminder that judgments regarding specific objectives and the enormous costs associated with them must be carefully examined.
Weapons of mass destruction refer to weapons—such as nuclear, biological, or chemical weapons—capable of causing massive damage in a single strike. The 1945 U.S. atomic bombings of Hiroshima and Nagasaki in Japan demonstrate that nuclear weapons can cause massive loss of life and urban destruction in a short period of time. The atomic bomb dropped on Hiroshima was called “Little Boy,” and the one dropped on Nagasaki was called “Fat Man.” Japan declared its surrender on August 15, 1945, and World War II officially came to an end in September of that year with Japan’s signing of the instrument of surrender.
In contrast, precision-strike weapons are designed to strike specific targets as accurately as possible. The U.S. Tomahawk cruise missile is often cited as a prime example. The Tomahawk is a cruise missile capable of long-range flight and precise navigation; it utilizes various navigation technologies, such as satellite navigation and inertial navigation, to fly toward its target. Describing the Tomahawk’s performance using the single figure provided in the source text—“up to 1,600 km with a margin of error within 10 meters”—does not adequately reflect current technology and the various operational variants. Therefore, rather than presenting a specific range and margin of error as a single fixed number, it is more appropriate here to focus on the fact that it is a weapon designed to precisely strike specific targets over long distances. The Tomahawk was also used in the Iraq War and was employed to strike a variety of targets, including military facilities and command-and-control centers.
Such precision-strike weapons are difficult to develop. This is because highly advanced navigation technology, sensors, control systems, and software are required to navigate accurately to the target location. In some cases, a single sophisticated weapon can be more effective against a specific target than an indiscriminate, large-scale attack. The same applies to social welfare. It is difficult to solve complex social problems with systems that are overly simplistic to reflect reality. If the numerous welfare programs created in this way fail to properly address the gaps in reality, simply increasing the number of programs is not enough. What is needed is to meticulously design each program to meet actual needs and to continuously improve them by analyzing the results of their implementation. Utilizing IT and artificial intelligence technologies can also be one approach. At the same time, public attention must shift from merely creating new systems to improving existing ones. In particular, given that the OECD has recommended that South Korea prioritize targeted support for vulnerable households and businesses by 2026, the central thesis of this article—that support should be concentrated where it is needed—can be linked to current policy discussions.
Before embarking on any of these efforts, we must first abandon blind faith in welfare. Systems are just systems. Blind faith in welfare can easily lead to misguided policy choices. Just as even the most powerful military cannot solve every problem on its own, the welfare system is not a panacea for all social issues. That said, we must not view welfare merely as unnecessary spending. Welfare is a vital social mechanism designed to mitigate social risks—such as poverty, illness, unemployment, disability, and old age—and to guarantee a dignified life. The key is not to deny the necessity of welfare, but to ensure that it is properly delivered to those who truly need it. Welfare is not something to be blindly trusted, but rather a system that must be rationally designed and utilized to achieve social objectives. Given this, shouldn’t we now discuss how to create “precision-targeted welfare”—which provides the right amount of support to those who need it, while continuously verifying results and making improvements—instead of “mass-destruction-style welfare”?
