One Tuition Plan, Four Regional Tests Across South Korea

Busan has gained residents for education and lost far more for work for three straight years. A proposed national-university scholarship would lower the price of entry, but its regional effect will depend on the jobs available after graduation.

Busan was the only South Korean province or metropolitan city in 2025 to record an education-related net inflow alongside an employment-related net outflow. Moves attributed to education brought 3,479 more residents into the city than they took out. Employment-related moves produced a net loss of 9,063.

The age distribution turned at roughly the same stage of life. Busan gained 927 residents ages 20 to 24, then lost 5,437 among those ages 25 to 29, the city’s largest net outflow in any five-year age group. Another 2,236 residents were lost between ages 30 and 34.

Neither table follows individual students from enrollment into employment. A 20-year-old moving to Busan for reasons unrelated to college still appears in the age data, and a graduate may leave without identifying work as the main reason. The two measures nevertheless locate Busan’s strongest population loss after the years most closely associated with entering higher education.

The direction has persisted. Busan gained 3,786 residents for education in 2023 and 3,383 in 2024, while employment produced net losses of 9,939 and 9,259. The city recorded positive balances among people ages 20 to 24 in both years and losses of more than 5,300 among those ages 25 to 29. Annual totals fluctuated, but the sequence remained intact: education drew people in, and employment drew substantially more out.

Daejeon followed a different path. It also gained residents for education and lost them for work in 2023 and 2024, but both its employment balance and its balance among people ages 25 to 29 turned positive in 2025. Migration records cannot identify which employers or industries produced that change. They do establish that a non-capital university city’s educational inflow does not inevitably end in early-career outflow.

The Education Ministry is now considering full-tuition scholarships for about 60,000 first-year students at roughly 30 national universities outside the Seoul metropolitan area. The plan could begin with the 2027 entering class. Covering the tuition of every eligible student would require an estimated 200 billion won a year, although about half of that amount is already being paid through existing national scholarships.

The proposal is still under budget review. The government has not finalized the exact level of support, how current grants would be counted or whether the award would eventually extend beyond the first year. Education officials have described the broader direction as settled: financial aid for regional national universities will increase, and assistance for high-performing students at regional private universities will also be expanded.

A tuition guarantee would provide a clear benefit at the moment a student chooses a university. Regional development requires more. Universities must offer programs students want to enter, employers must create positions that use those qualifications, and graduates must see enough value in the resulting careers to remain. South Korea’s migration data show that those conditions break down at different points across the country.

A scholarship inside a larger university strategy

The tuition proposal has emerged during a wider reorganization of regional higher education. The government is concentrating new funding in flagship national universities, linking degree programs more closely to regional industries and asking institutions to share courses, laboratories and research infrastructure across broader economic regions.

The Education Ministry’s 2026 package provides 544.8 billion won for flagship national universities. Three universities will receive additional support for industry-linked colleges and specialized research institutes, with companies participating in curriculum design and applied research. Separate funding will establish regional artificial-intelligence education centers, while a shared-university program is intended to extend courses, research assets and startup support beyond the flagship institutions.

The government also plans to expand employment-conditional departments, in which universities and companies design programs tied to future recruitment. Regional universities currently operate such departments on a smaller scale than universities in the Seoul metropolitan area. The ministry wants the average regional program to reach the current metropolitan level by 2030.

Those measures acknowledge that tuition alone cannot produce a regional career. They also clarify the political role assigned to the scholarship. Most of the existing regional-university strategy finances institutions, research capacity and industry partnerships. Full tuition would add a benefit that students and families can see before enrollment.

The financial incentive could make a regional national university more competitive even before new laboratories, curricula or employer partnerships produce measurable results. The government is therefore acting on two different timelines. Tuition can affect the 2027 admissions cycle. Research capacity, teaching quality and graduate employment will require several years.

Those timelines should not be merged when the policy is evaluated. An increase in applications could show that students responded to a lower price. It would not demonstrate that the university provided stronger education or that its graduates entered better jobs nearby.

What “full tuition” would add

The estimated 200 billion-won cost combines money already being spent with additional public support. About 100 billion won is expected to reach the eligible student population through the existing national scholarship system. The broadest version of the proposal would require roughly another 100 billion won each year.

Dividing the totals across 60,000 students produces two different averages. The complete package would amount to about 3.33 million won per student, while newly added expenditure would average roughly 1.67 million won. Neither figure represents the amount an individual would receive. Tuition, household income and current assistance vary too widely. The averages instead separate the value of the fees being covered from the additional cost to the government.

South Korea already pays the full tuition of students from basic-income and near-poor households. In 2026, the principal income-linked scholarship provides up to 6 million won a year for support bands 1 through 3, 4.4 million won for bands 4 through 6, 3.6 million won for bands 7 and 8, and 1 million won for band 9. Students in band 10 receive no award under that program. Multi-child grants, university scholarships and a second national scholarship can provide additional assistance.

Pusan National University’s 2026 engineering tuition shows how the proposed guarantee would interact with those payments. An engineering undergraduate is charged 2.364 million won per semester, producing an annual bill of 4.728 million won.

The principal national scholarship ceiling already exceeds that amount for a student in bands 1 through 3. Before any other grant is counted, the remaining annual balance would be 328,000 won in bands 4 through 6 and 1.128 million won in bands 7 and 8. A student in band 9 would face a balance of 3.728 million won, while a household outside the income-linked program could receive nearly the full annual tuition as new assistance.

Those calculations do not estimate what particular students currently pay. University awards and other government programs may reduce or eliminate the remaining balance. They show why two students whose final bill falls to zero could receive very different benefits from the new policy. One may already have been fully funded; another may gain several million won.

Academic field creates another difference. Average annual tuition at four-year general and education universities reached 7.27 million won in 2026. Humanities and social sciences averaged 6.43 million won, engineering 7.68 million won and medicine 10.33 million won. Income-linked aid applies across programs with sharply different prices, leaving larger residual charges in more expensive fields.

A serious distributional assessment would identify four amounts for each group of recipients: the tuition charged by the university, the grants already available, the additional scholarship created by the regional policy and the amount ultimately paid by the student. The totals should be reported by household category, institution and academic field.

The additional support could change more than the final invoice. Students with substantial remaining fees may borrow less, reduce paid work during the semester or choose a program that would otherwise have been unaffordable. A guaranteed public award could also provide more certainty than an institutional scholarship whose renewal depends on grades or annual university budgets.

Tuition remains only one part of university choice. Applicants also consider whether a university offers the desired major, how difficult admission will be, the institution’s reputation, housing and transportation costs, campus life and expected employment after graduation. A scholarship will exert the greatest influence where a regional national university is already a plausible academic choice and price remains one of the deciding factors.

The government’s decision to begin with first-year students reflects that logic. New entrants can change where they enroll; current students have already made that decision. The policy may therefore produce a visible admissions response even when its effect on student hardship is concentrated among higher-income households or students in more expensive programs.

The origin of the new students will determine what that response means. A student who chooses Pusan National University over a private university in Busan improves the national institution’s enrollment without adding another resident to the city. A student who chooses Busan instead of Seoul changes the regional distribution of enrollment. Someone who would not otherwise have entered college represents a third result.

Admissions totals cannot distinguish those paths. Participating universities should report the previous region of each student, the location of the high school attended and, in aggregate, the institutions applicants were also considering. Without that information, stronger recruitment could be described as regional attraction even when most of the movement occurred inside the same local university market.

One subsidy, four regional patterns

South Korea’s 2025 migration tables place the country’s regions into four broad groups. Seoul, Daejeon and North Chungcheong gained residents for both education and employment. Busan alone gained residents for education and lost them for work. Incheon, Ulsan, Sejong and South Chungcheong lost residents for education but gained them for employment. Daegu, Gwangju and most provinces recorded losses in both categories.

The classification compares direction rather than scale. A net movement of several hundred people in a smaller city and several thousand in a larger province do not carry the same demographic weight. Housing supply, family movement and local population size can also produce an overall balance that differs from the education and employment figures. The four groups identify the stage at which movement turned negative; they do not rank regional economies.

Daejeon provides the clearest comparison with Busan outside the Seoul metropolitan area. Education brought a net 5,679 residents into Daejeon in 2025, while employment added 712. The city gained 2,042 residents ages 20 to 24 and 206 ages 25 to 29.

Busan attracted fewer residents for education and lost far more for work. Its modest gain among people in their early twenties was followed by a loss of more than 5,400 in the next age group. A scholarship in the two cities would therefore enter university systems surrounded by different recent employment conditions.

Daejeon’s research economy provides useful context without explaining the annual migration change on its own. Daedeok Innopolis contained 3,063 organizations and nearly 100,000 workers in 2024. Its institutions and companies reported more than 27 trillion won in sales and almost 10 trillion won in research and development spending. Government-funded research institutes, technology companies, universities and public agencies form a dense professional labor market that Busan does not replicate in the same form.

Daejeon’s 2025 improvement cannot be attributed to that ecosystem from migration data alone. The research complex existed when the city was still losing residents for employment in 2023 and 2024. Its presence nevertheless gives science and engineering graduates access to a concentration of research institutions, technical employers and public-sector careers within a relatively compact region. Further analysis would need to identify the occupations and employers that absorbed people in their late twenties when Daejeon’s balance turned positive.

Busan has a larger and more diverse economy, spanning port logistics, maritime services, manufacturing, health care, tourism, finance and public services. Economic diversity does not guarantee that the available positions match the expectations of university graduates. A city can have many jobs and still offer too few entry points into the professional tracks that young adults associate with long-term advancement.

Incheon, Ulsan, Sejong and South Chungcheong face the reverse problem. Their labor markets attracted residents in 2025 even as education-related moves produced net losses. A regional national-university scholarship could address an earlier point of weakness by making local study less expensive.

Migration records cannot establish why students left those regions. Tuition may have mattered, but so may the absence of a particular major, university reputation, the wish to live in another city or the expectation that studying near a preferred labor market will improve access to its employers. A reduction in education-related departures would still be a relevant early result, provided the additional national-university enrollment came from students who would otherwise have studied outside the region.

Daegu and Gwangju lost residents at both stages. Daegu recorded an education-related net loss of 581 people and an employment-related loss of 11,876 in 2025. Gwangju lost 1,131 for education and 10,117 for work.

Cheaper national universities could improve recruitment in those cities, but admissions assistance does not directly change the employment movement recorded later. Their results will need to separate the performance of the university from the performance of the surrounding labor market. A higher first-year enrollment figure cannot demonstrate that both losses have been reversed.

The government can apply broadly similar financial-aid rules across the four groups. The same performance measure would be misleading. Application growth and fewer unfilled places are particularly relevant where education-related movement is negative. Busan requires evidence from the transition into work because education has produced a net inflow for three consecutive years. Daejeon requires more time to establish whether its 2025 turn will persist.

The national comparison exposes the weakness of treating university enrollment as a substitute for regional settlement. Regions attract students and workers through different institutions, at different ages and for different reasons. A tuition subsidy operates at the entrance to that process. Its regional effect depends on the part that follows.

Busan’s break comes after education

Busan’s internal geography reflects the same division. Geumjeong District, home to Pusan National University’s main campus, gained 2,921 residents for education in 2025 and lost 1,887 for employment. Nam District, which contains Pukyong National University’s main campuses and several private institutions, gained 2,196 for education and lost 1,754 for work.

Both districts ended the year with overall population losses after housing, family and other stated reasons were included. The figures cover all residents crossing administrative boundaries and cannot be assigned to individual universities. They show education-related arrivals and job-related departures occurring in the same year around two of the city’s largest concentrations of higher education.

Busanjin District offers a different combination. It gained residents for both education and employment, while an even larger housing-related inflow accounted for most of its overall population gain. Campus location alone does not determine a district’s result. Employment centers and residential development can reinforce or outweigh educational movement.

Busan’s external migration also extends beyond a one-way departure for Seoul. The city’s largest net loss in 2025 was to Seoul, at 5,493 residents, followed by Gyeonggi Province at 3,241 and neighboring South Gyeongsang Province at 2,421. Exchanges among Busan, South Gyeongsang and Ulsan remained substantial in both directions.

The administrative border does not match the labor market. A graduate living in Busan and working in Gimhae would not appear as an out-migrant. A graduate moving to Changwon for a job would be counted as a loss to Busan even though the work remained within the southeastern economy. City and regional retention must therefore be measured separately, and residence cannot be used as a substitute for workplace.

Busan’s employment indicators do not describe a city without work. In June 2026, it had 1.71 million employed residents, an employment rate of 59.0 percent and an unemployment rate of 2.2 percent. The employment rate was unchanged from a year earlier, while unemployment had fallen.

Those figures can coexist with a large employment-related migration loss because the statistical systems observe different events. The monthly labor survey asks whether current residents are working. Migration statistics record changes of address and the primary reason supplied for the move. Busan can maintain employment for its existing population while losing residents whose occupations, employers or career plans lead elsewhere.

A 2025 Busan Development Institute study gives the employment loss a more concrete setting. The city produces a relatively large supply of four-year university graduates, but its business structure contains comparatively few workplaces with 300 or more employees and a high concentration of companies with fewer than 50 workers. Many of the positions examined in the study paid less than the monthly wage of at least 2.5 million won sought by young job seekers.

Local employers described another side of the mismatch. Businesses reported difficulty finding applicants with the required occupational skills and experience, while also pointing to young workers’ reluctance to join smaller companies and high turnover after recruitment. The gap therefore does not run in one direction. Graduates may find too few positions offering the pay, stability and advancement they expect, while employers may struggle to recruit candidates prepared for the work they have available.

The study does not establish that those conditions caused the 9,063 employment-related net loss. Its data cover broader groups of young workers and firms rather than the individuals recorded in the migration tables. The findings do identify a plausible mechanism that a tuition scholarship does not reach: Busan can produce university graduates and maintain open positions while failing to connect enough of those graduates with jobs they regard as viable careers.

The city’s own employment programs increasingly recognize that recruitment is not the end of the problem. Busan has selected companies for youth appeal using measures that include starting pay, benefits, work-life balance, stability and growth prospects. A newer package offers financial support when young workers remain at local small and midsize companies for six months, one year and longer.

Such programs address job quality, information and retention more directly than tuition support. Their scale, however, remains limited compared with the number of university students and young adults moving through the regional labor market. They also illustrate why the city’s problem cannot be reduced to vacancies. A position contributes to regional settlement only when a worker accepts it, remains and sees a path beyond the first contract.

National graduate statistics reveal another limit in the available evidence. Among the 2024 graduating cohort, the employment rate was 69.5 percent across higher-education institutions and 62.8 percent among graduates of four-year general universities. Institutions in the Seoul metropolitan area recorded a rate of 71.3 percent, compared with 67.7 percent outside it. Eleven months later, employment retention stood at 83.1 percent and 80.5 percent, respectively.

Public ownership explained little of the national difference. Public institutions recorded an employment rate of 68.9 percent, compared with 69.6 percent at private institutions. Academic field produced much larger gaps: 79.4 percent in medicine and related fields, 70.4 percent in engineering and 61.1 percent in the humanities.

The results weaken the assumption that public status alone improves a graduate’s labor-market position. A regional national university may charge less and possess stronger public research capacity, but employment still depends heavily on what the student studied and which industries are hiring.

The regional labels in the graduate survey refer to the university’s location, not the graduate’s workplace. A Pusan National University graduate working in Seoul remains part of the non-metropolitan university rate. A Seoul graduate employed in Busan remains part of the metropolitan rate. The statistics show how graduates of universities in different locations fare in the national labor market; they do not show whether regional universities supply workers to their own regions.

South Korea already links graduate records with employment insurance, income, industry and employer information. Regional employment details are available within the underlying administrative system. Public reporting could therefore connect the university and field studied with the location, industry, employer size, pay and duration of the job that followed.

That connection is essential for Busan. A higher graduate employment rate would have limited regional meaning if most of the improvement came from jobs in Seoul or Gyeonggi. A lower city retention rate could also understate success if graduates built careers in nearby parts of South Gyeongsang while continuing to participate in the southeastern economy. Both workplace and residence must be reported.

The university market may move before the population

The tuition proposal would enter a higher-education market with a large price gap already in place. Average annual tuition at private four-year universities reached 8.23 million won in 2026, compared with 4.25 million won at national and public institutions. Covering the remaining tuition of first-year students at regional national universities would widen that difference before new teaching, research or employment initiatives had time to produce results.

Busan contains national and private universities recruiting from overlapping pools in engineering, business, humanities, natural sciences and other widely offered fields. An applicant who views two programs as academically comparable would face a powerful financial reason to choose the tuition-free option.

Price will not override every preference. Students may still leave for a major that is unavailable locally, a university with stronger recognition or a city where they expect to work after graduation. The scholarship’s immediate effect is more likely to appear among applicants who already regard a regional national university as competitive but remain undecided about cost.

In a shrinking market, those decisions have consequences beyond the recipient. The policy does not increase the number of students graduating from high school. National universities may become more selective and fill their places more easily while nearby private universities lose applicants, tuition revenue and the enrollment needed to maintain programs.

The movement may be rational from a student’s perspective and still leave the regional system no larger. A student shifting from a private university in Busan to Pusan National University creates an institutional gain and loss within the same city. Counting only the subsidized university would turn redistribution into apparent regional growth.

Regional private universities are not absent from the government’s plan. Education officials have proposed expanding merit scholarships and revising regional-talent assistance so that students from the Seoul metropolitan area can qualify when they attend private universities outside it. Shared-university funding is also meant to spread courses, research equipment and startup support across institutional boundaries.

The effectiveness of those safeguards will depend on their design and scale. A smaller scholarship at a private university may not offset a full national-university guarantee. Access to a shared laboratory can improve a specific department without solving institution-wide enrollment losses. Joint courses may broaden academic choice while leaving tuition revenue and staffing costs with the weaker university.

The government’s employment-conditional departments face a similar test. In the most recent count, metropolitan universities enrolled an average of 86 students per institution in programs linked to future recruitment. Regional national universities averaged 41 and regional private universities 59. Expanding those departments could provide a more direct bridge into work, but the number of seats is not the final outcome.

A contract with a company must be followed through recruitment, job location, pay and retention. Graduates placed outside the region may succeed individually without strengthening the local labor market. A company that reduces hiring before graduation may leave students with specialized training and no promised position.

Public investment in strong national universities can still create broad regional value. Flagship institutions can support advanced laboratories, specialized faculty and research partnerships that smaller universities cannot finance independently. Concentration may be necessary where declining enrollment has left too many similar programs competing for too few students.

The public account becomes incomplete when gains at a national university are reported without the losses and adjustments around it. Applications, unfilled places, transfers, program closures and institutional scholarships should be tracked across the entire regional market. Several years later, graduate destinations should determine whether the restructuring created stronger regional careers or merely changed the name of the university on the degree.

What the class of 2027 can show

The first results will appear as soon as the 2027 admissions cycle ends. The Education Ministry will know how many students received support, how much new assistance was added after existing grants and whether applications increased at eligible universities.

Those records can establish whether the scholarship improved affordability and changed student choice. Borrowing, paid work during the semester, withdrawals and transfers would show whether the financial benefit affected the conditions under which students studied. The previous regions and alternative university choices of incoming students would reveal whether the policy attracted people from elsewhere or redistributed applicants locally.

Regional results will arrive on a different schedule. The 2027 cohort must be followed through graduation and into employment. Each graduate’s field of study should be connected with the location, industry, employer size and pay of the first job. Employment after 11 months will indicate whether that position lasted. Records three years after graduation should show whether the person remained in the university’s city, worked elsewhere in the surrounding economic region or moved to the Seoul metropolitan area.

Busan cannot be judged primarily by the early indicator that would matter most in a region losing students for education. The city has attracted residents for study and lost far more for work in each of the past three years. Higher national-university enrollment would show that students responded to the scholarship. It would not show that Busan had improved the transition from college into a regional career.

Daejeon’s recent turn illustrates the value of following the later stages. Its education-related inflow had existed before its employment balance became positive. The change became visible only when workers and people in their late twenties also began to register a net gain. Whether that improvement lasts, and which employers sustained it, will determine how much can be learned from the comparison.

The tuition bill will show whether the proposal helped students in 2027. The education, employment and migration records that follow will show whether it also helped the regions where they studied.