Korea’s Innovation Cities Did Not Grow the Same Way

Busan retained more relocated public workers and their families than any other Korean Innovation City. Across the country, however, settlement, population growth, business formation, and local hiring followed markedly different paths, leaving South Korea with a more complicated record as it prepares a second round of public-institution relocation.

By June 2026, 86.4 percent of the employees originally designated to move with public institutions assigned to Busan had also moved their residence to the region, the highest share among South Korea’s 10 Innovation Cities. Among married employees, 81.4 percent had brought their families with them. Nationally, 34,214 of the 48,128 employees covered by the first relocation program had changed their residence to the receiving region, equivalent to 71.1 percent, while the family relocation rate among married employees stood at 60.6 percent.

The Busan figures matter because the relocation of an institution never guaranteed the relocation of the household attached to it. Employees could retain a home near Seoul, commute over long distances, or move alone while spouses and children remained in the capital region, preserving many of the residential and professional ties that the policy was intended to loosen. Busan offered a different set of choices. A transferred worker arrived in a metropolitan economy where housing, universities, hospitals, schools, rail transit, consumer services, and employment opportunities for other members of the family were already present on a scale that newly built regional centers could not always match during their formative years.

Had residential settlement been the principal objective of the first relocation round, the national comparison would be relatively easy to read. The Innovation City program carried a much wider set of expectations. Public institutions were to disperse employment and administrative capacity away from the Seoul metropolitan area; the cities receiving them were expected to anchor workers and their families, create opportunities for regional graduates, support new firms and industrial clusters, and ultimately weaken the concentration of population and economic power around the capital.

The record becomes less orderly once those ambitions are considered together. Chungbuk remained near the bottom of the employee-settlement table even as Jincheon County, one of the areas hosting its Innovation City, recorded some of the strongest population expansion of the relocation era. Naju, the center of the Gwangju–Jeonnam Innovation City, added residents and firms on a scale that bears little resemblance to the mature districts surrounding Busan’s relocated institutions. Ulsan expanded staffing at its relocated headquarters well beyond the level envisaged in the original plans while the host district continued to lose residents, and Jeonbuk produced substantially stronger results than Busan under the formal regional-talent hiring measure.

Those differences are not statistical noise around a single definition of success. A household establishing itself in Busan, a company entering newly developed business land in Naju, additional headquarters staff in Ulsan, and a regional graduate hired by a public institution in Jeonbuk describe different ways in which relocation can affect a local economy. Each matters, but none can stand in for the others, much less establish by itself that private investment, productivity, or national spatial balance improved.

South Korea will make its next relocation decisions with nearly two decades of evidence from the first round now available. The most interesting question is no longer whether the state can move offices and employees away from Seoul. The first round has instead become a national experiment in how very different regional economies absorb the same institutional shock, and Busan’s 86.4 percent settlement rate is most revealing when treated as one result inside that larger divergence.

The Scoreboard Breaks Apart

Chungbuk sits almost at the opposite end of Busan’s settlement ranking. Only 50.6 percent of the employees originally designated for relocation had moved their residence to the region by June 2026, and only 38.8 percent of married employees had relocated with their families. Yet Jincheon County had 35.7 percent more residents in 2024 than in 2015, according to National Assembly Budget Office data, while the number of firms counted in the Chungbuk Innovation City increased from 24 in 2018 to 664 in 2024. The low starting base and simultaneous creation of business land make those figures unsuitable for a direct league-table comparison with Busan, but they show that weak household relocation can coexist with rapid growth in the wider host area.

Gwangju–Jeonnam creates a similar inversion through a different combination of outcomes. Its worker-settlement rate reached 76.3 percent, below Busan’s, while Naju’s population increased 27.5 percent between 2015 and 2024. Firms associated with the Innovation City rose from 173 to 1,171 between 2018 and 2024, compared with an increase from 117 to 254 in Busan. The raw contrast looks dramatic until the physical structure of the two development models is taken into account: Naju was creating a new urban and business district at the same time that Busan was inserting public institutions into neighborhoods where companies, housing, and services already existed.

Population and business growth do not necessarily move together with the expansion of the transferred institutions themselves. In a National Assembly Budget Office comparison of current headquarters employment with the staffing envisioned in original relocation plans, Ulsan stood at roughly 143 percent of plan, the highest ratio among the regions examined; Busan was at about 106 percent. Ulsan’s host district nevertheless continued to lose population, illustrating how the growth of a public organization can occur inside a municipality whose demographic trajectory is being driven by much larger forces.

Local recruitment creates another reversal. In the 2023 cross-city figures compiled by Busan Research Institute, Busan’s official relocation-region talent hiring rate was 34.5 percent, the lowest among the 10 Innovation Cities in the table, while Jeonbuk reached 61.5 percent. Busan subsequently reported a statutory rate of 35.7 percent for 2024, but the institutional universe and hiring denominator underlying that figure are much narrower than the full network of organizations normally associated with the city’s Innovation City.

A Breeze in Busan comparison of the 10 regions found no stable alignment between employee settlement and population growth, firm expansion, headquarters employment, or local hiring. Worker settlement and family relocation moved closely together, as would be expected from two measures describing related household decisions, but the other indicators produced rankings that shifted markedly from one measure to another. With only 10 regions, the exercise is descriptive rather than causal; its value lies in showing how difficult it is to compress the first relocation round into a composite score without first deciding which economic process is supposed to count as success.

The contrast also reveals why headline rankings can mislead even when every number is accurate. A firm count is particularly useful where newly developed land makes business entry visible; it is less capable of identifying higher sales or specialized employment inside companies that already existed before relocation. Population growth is conspicuous in a new residential district built from a small base, while several thousand additional households can disappear statistically inside a metropolitan borough already facing aging and long-term population decline. The same policy was therefore being observed through measures whose sensitivity differed according to the kind of place receiving it.

By the time those differences become visible, the term “Innovation City” has already concealed much of what matters. Busan, Naju, and Jincheon were never equivalent urban economies awaiting the same intervention.

Two Ways to Receive the State

Busan received its relocated institutions inside a city that had spent decades accumulating the infrastructure of metropolitan life. Financial organizations assigned to Munhyeon entered a district connected to an existing subway network and surrounded by a mature housing market, universities, hospitals, banks, accounting firms, legal services, and a labor market whose scale far exceeded the workforces of the transferred institutions. Marine research organizations sent to Dongsam joined a port economy built around shipping, logistics, maritime universities, fisheries, engineering, and scientific research, while the film and media agencies placed around Centum entered an area where broadcasters, production companies, cultural facilities, convention infrastructure, and private content businesses were already operating.

For a family deciding whether to follow a transferred employee, those inherited assets reduced the risks attached to relocation in ways that do not appear in a headquarters headcount. A spouse whose career had nothing to do with the public institution could search across a broad metropolitan labor market rather than depend on a small group of employers created by the new district; children could enter established schools; hospitals, shopping areas, rail transit, and cultural facilities were immediately available. Residential support in Daeyeon eased part of the transition, but the city’s deeper advantage was the range of institutions and markets that the relocation program itself did not have to build.

Naju and Jincheon absorbed public institutions while much of the urban environment around them was still taking shape. Apartment blocks, schools, commercial streets, public facilities, roads, corporate sites, and other infrastructure expanded alongside the new headquarters, intertwining institutional relocation with land development, housing supply, construction, migration, and the creation of a new local consumer market. The arrival of the state therefore left a more visible physical imprint because the city around it was changing at the same time.

A smaller population base amplifies that transformation. Several thousand additional residents can materially alter the demographic profile of a county or newly developed district in a way that the same absolute change cannot alter a mature metropolitan borough. Dedicated development land creates a similar effect in business statistics: a few hundred companies entering newly designated sites can produce enormous percentage growth, while an established city may absorb new institutional demand through thousands of firms that were already operating before the first headquarters arrived.

The provincial model had another advantage that becomes obvious on a map. New development made it possible to assemble contiguous land for business parks, research facilities, housing, corporate offices, parking, laboratories, and later expansion while roads and utilities were planned around those sites. Busan’s clusters were inserted into dense neighborhoods where land was more expensive, ownership more fragmented, and large industrial or research campuses much harder to create. The city possessed deeper existing markets but less room in which new activity could appear as a discrete addition to the Innovation City itself.

Firm-level research published in 2026 gives empirical weight to that difference. Using Statistics Korea’s Statistical Business Register and a multi-period difference-in-differences design, researchers found that metropolitan districts were more likely to register relocation effects through expansion among firms already present, including higher sales, whereas provincial cities and counties showed broader increases in firm entry and growth across micro, small, and medium-sized businesses. The national policy did not simply produce stronger effects in one kind of place and weaker effects in another; it interacted with two different forms of regional economic development.

That distinction is crucial for Busan because a modest increase in the number of officially counted Innovation City firms can coexist with meaningful changes inside the existing business base. An accounting company near Munhyeon might gain public financial clients and hire specialists without ever appearing as a newly attracted firm. A maritime engineering company could acquire research work from Dongsam while remaining outside the formal Innovation City boundary. A production company in Centum could expand revenue or skilled employment through relationships with relocated content institutions without contributing anything to a measure based on firm entry.

The data needed to capture those changes are less convenient than a tenant count. Revenue growth among incumbent suppliers, changes in specialist employment, the geographic distribution of contracts, university-industry research, and the acquisition of new capabilities inside existing companies are harder to assemble consistently across cities, yet they are precisely the measures needed to determine whether the metropolitan model produced a different kind of growth rather than simply less of it.

Population statistics require the same caution. Nam-gu, Yeongdo-gu, and Haeundae-gu were already being reshaped by low fertility, aging, redevelopment, and the out-migration of younger residents when the public institutions arrived. Several thousand workers and family members could increase demand for housing and services, or strengthen a specialized employment cluster, without reversing demographic forces operating across much larger populations. Naju or Jincheon could register far more conspicuous gains as newly built housing attracted residents with no connection to public institutions and businesses responded to land prices, subsidies, infrastructure, and a growing local market.

The first relocation round therefore entered regions with different stocks of firms, workers, land, universities, housing, services, and transport, and those inherited conditions influenced what the policy looked like a decade later. For newly built centers, the institutional transfer was difficult to separate from the physical expansion of the city around it. In Busan, much of the relevant change was more likely to be absorbed by an economy already in place, which raises a different question once construction and the first wave of household relocation are no longer the main sources of growth.

What Lasted After the Move

A transferred institution begins affecting a local economy before any industrial cluster has time to form. Salaried employees rent or purchase homes, eat in restaurants, use shops and services, and pay taxes, while the institution itself contracts for maintenance, facilities, equipment, and other operating needs. In newly developed Innovation Cities, housing construction, roads, commercial development, and public infrastructure amplified that first wave, creating an early period in which the arrival of public employment and the expansion of the local built environment were difficult to separate.

More than a decade after the main transfers, the distinction between those immediate effects and the policy’s longer economic ambitions has become easier to observe. Research by the Korea Institute for Industrial Economics and Trade finds substantial employment gains in non-tradable industries serving local demand, with much of the increase concentrated in and around the Innovation City and weakening with distance from the host area. Employment in tradable sectors—activities capable of selling goods or services outside the immediate local market—developed more slowly and less consistently, while the strong demographic inflows associated with earlier stages of development have also weakened as the Innovation Cities moved beyond their construction and settlement phase.

The pattern follows a recognizable economic sequence. Retail, restaurants, housing services, and other locally consumed activities can expand as soon as a new concentration of salaries enters a district; a supplier capable of competing nationally, a scientific discovery converted into a commercial product, or a specialized financial-service market requires a much deeper interaction among skills, firms, capital, research institutions, and recurring demand. The first stage can be created relatively directly by moving an organization and its payroll. The second depends on the receiving economy’s capacity to turn that institutional presence into relationships that survive independently of the original transfer.

Busan’s local tax record captures the more direct effect with unusual clarity. Public institutions covered by the National Assembly Budget Office analysis paid a cumulative 287.36 billion won in local taxes between 2016 and 2024, with annual payments rising from 22.18 billion won to 48.84 billion won over that period. A separate 2026 difference-in-differences study reaches a similar conclusion from broader regional data: evidence of a uniform increase in total real gross regional domestic product was limited, while some receiving areas recorded clearer gains in public-administration output and real local tax revenue. The economic activity moved directly by policy—the institution, its employees, and the fiscal base attached to them—appears more consistently than the private transformation expected to develop around it.

Seen over time, the first relocation was therefore more reliable at transferring public-sector demand than at guaranteeing a new structure of private production. Government could decide where an institution employed staff and performed administrative functions; it could not decide with the same certainty whether a specialized supplier would emerge nearby, whether a research institute’s discoveries would become products made by local companies, or whether private firms would accumulate capabilities strong enough to compete beyond public-sector contracts.

Busan’s three principal clusters provide a concrete way to see why those later effects cannot be measured with one common indicator. A deeper financial effect in Munhyeon would appear through private financial employment, corporate funding, asset management, fintech, accounting, legal services, and other specialist work generated around major public financial institutions. Dongsam’s marine research organizations would leave a different trace through university research, licensing, commercialization, specialist engineering, and the circulation of scientists between public laboratories and private maritime firms, while Centum’s film, game, and media institutions would be better assessed through production expenditure, intellectual-property ownership, project finance, private-company revenue, and the ability of skilled creative workers to build careers in Busan.

A conventional count of firms or public employees can contribute to all three assessments without resolving any of them. Munhyeon unquestionably gained a concentration of public financial employment through institutions such as Korea Asset Management Corporation, Korea Housing Finance Corporation, Korea Securities Depository, and Korea Housing and Urban Guarantee Corporation, but demonstrating that the private financial economy deepened as a consequence would require evidence on specialist jobs, local professional-service contracts, private investment, and financing relationships with companies in Busan and the wider southeast. Dongsam may generate economic value without hundreds of new firms appearing around the research institutes if university laboratories receive funding, companies license technology, researchers move into commercial roles, or engineering businesses acquire expertise that raises productivity. Centum, similarly, can only be understood by following projects, intellectual property, money, and skilled labor rather than headquarters addresses alone.

The relevant evidence becomes still more demanding as the first relocation recedes in time. Construction and housing development can create a temporary surge that fades once buildings are completed; firms drawn by new land or incentives can enter without becoming durable businesses; local services can survive around a stable public payroll while remaining dependent on consumption that never extends beyond the district. A stronger claim of structural change requires companies that survive and expand, higher wages in specialized occupations, technologies that reach commercial markets, suppliers that acquire customers beyond the relocated institutions, and labor markets deep enough for skilled workers to change employers without returning to Seoul.

The geography of the effect matters as much as its persistence. KIET’s finding that much of the employment response remained close to the Innovation Cities means that a policy conceived in national terms can produce its clearest gains within a relatively narrow local area unless commuting networks, universities, suppliers, finance, and related industries connect that concentration to a larger regional economy. Busan entered the relocation program with many of those wider networks already in place, which helps explain its ability to absorb workers and families but also raises the standard of evidence needed to show that public institutions altered the trajectory of private firms rather than simply joining an economy that was already there.

By this stage of the policy cycle, the weakness is no longer a shortage of administrative information. Korea can count where institutions moved, how many designated employees changed residence, how much local tax was paid, and whether recruitment or procurement rules were satisfied with considerable precision. The evidence becomes thinner when the analysis reaches the part of the original ambition that matters most for long-run regional development: whether firms, workers, research, and capital acquired new relationships capable of sustaining growth once the novelty of relocation had passed.

That gap becomes unusually visible in Busan because several familiar performance figures change meaning as soon as their denominators are examined.

When the Denominator Changes

Busan’s Innovation City is normally described as a network of 13 relocated public institutions. The local-talent hiring statistics describe a much smaller institutional universe.

Busan Research Institute’s reconstruction of the hiring framework shows how the group narrows as legal status and recruitment exemptions are applied. National administrative agencies do not all fall under the same statutory regime, while research-oriented positions, small recruitment rounds, and other categories can be excluded from the calculation. By the time the headline local-talent rate is produced, only six institutions materially contribute to the result, leaving much of the measured hiring concentrated among organizations in the Munhyeon financial cluster even though Dongsam and Centum remain central to Busan’s Innovation City strategy.

That narrowing changes more than the institutional count. Busan’s official regional-talent recruitment rate reached 35.7 percent in 2024 after legally exempt hires were removed from the denominator, while regional talent accounted for 22.6 percent when measured against all regular new hires. The first percentage is a valid measure of performance among positions subject to the statutory regime; the second describes how much of the broader regular hiring pool went to workers counted as regional talent. A regulator interested in compliance and a graduate interested in the actual share of available jobs are therefore looking at different questions even when both are told they are examining the “local-talent hiring rate.”

The volume of hiring complicates the comparison further. Busan Research Institute recorded 665.5 regular new hires at the relocated institutions in 2018 and 433.5 in 2023; the 2024 count reached 509 by October, although a partial-year total cannot be compared directly with completed annual figures. A quota-based percentage can improve while the number or composition of jobs changes, which makes the absolute hiring pool indispensable for understanding whether regional graduates gained more employment opportunities in practice.

Other indicators carry similar definitional boundaries without being invalid. Busan’s 75.2 score in the 2024 national survey of living conditions was the highest among the Innovation Cities, but the survey covered roughly 5,000 adult residents rather than transferred public employees alone. The result therefore supports a claim about the quality of the residential environment as experienced by Innovation City residents, not a narrower claim that relocated employees themselves were more satisfied with the move than workers elsewhere.

Regional procurement contains an even more important denominator. Busan institutions reported executing 99.7 percent of their planned purchases of regional products in 2024, spending roughly 33.9 billion won against a plan of about 34 billion won. The figure measures how fully the institutions executed their own regional procurement plans; it does not mean that 99.7 percent of their total purchasing went to businesses in Busan.

A measure designed to examine the depth of local supplier relationships would instead require total procurement as the denominator and would ideally distinguish office goods from construction, technology, professional services, engineering, and research contracts. A local purchase of routine supplies and a multiyear contract with a regional technology firm both satisfy forms of procurement, yet only the latter is likely to reveal much about the development of specialized industrial capability.

Firm statistics encounter a related problem of geographic boundaries. A company entering dedicated cluster land in Naju or Jincheon can be identified relatively easily as part of Innovation City growth, whereas a financial firm near Munhyeon, a maritime engineering company outside Dongsam’s formal boundary, or a production company elsewhere in Centum can acquire new work from relocated institutions without appearing as an Innovation City entrant. Physical entry is easier to count than an economic relationship extending through an existing metropolitan market.

None of these indicators was necessarily poorly designed for the administrative function it originally served. Government needed to know whether institutions had transferred, whether designated employees had changed residence, whether legal recruitment obligations were being met, and whether planned local procurement was being executed. The difficulty arose when compliance measures built for those tasks were asked to carry much broader claims about private-sector productivity, supplier development, technology commercialization, professional labor markets, and long-term regional growth.

Those latter outcomes do not fit naturally into a single annual percentage. They require baselines established before relocation, repeated observation across many years, and data following firms, workers, research relationships, and supply chains beyond the administrative boundary of the public institution itself. A policy can therefore appear highly successful under one set of figures and far less decisive under another without either interpretation requiring the underlying statistics to be wrong; the discrepancy lies in the questions the indicators were constructed to answer.

That distinction becomes consequential as South Korea prepares to use the evidence of the first relocation to design a second round whose spatial logic is already beginning to look different.

From Dispersion to Agglomeration

The government’s 2026 economic strategy calls for a second public-institution relocation plan in the second half of the year and the first transfers, centered on lead institutions, from 2027, while the broader five-pole, three-special-region agenda seeks stronger centers of employment, industry, research, and advanced services outside the Seoul metropolitan area. The emerging framework is taking shape under economic and demographic conditions markedly different from those that surrounded the original Innovation Cities.

The first round carried a strong distributive logic. Public institutions concentrated around Seoul would be spread across the country, giving regions that had lost population and economic functions to the capital a direct share of nationally funded employment, budgets, and administrative capacity. Geographic redistribution was therefore part of the policy’s purpose rather than merely an instrument for producing some later economic effect.

Nearly two decades of implementation have shifted attention toward the scale and composition of what is moved. KIET’s research indicates that relocation effects tend to become stronger with larger concentrations of employment while remaining spatially limited unless they connect to wider urban and industrial systems. The National Assembly Futures Institute has similarly questioned whether another round of broadly even territorial distribution would generate the strongest results, emphasizing regional centers with substantial labor markets, education, health care, transport, and existing economic functions.

For municipalities that have lost residents, employers, and public functions to the capital, the case for wider distribution remains politically powerful. A national institution brings stable jobs, tax revenue, construction, visitors, and status, all of which are difficult for a shrinking locality to dismiss as secondary benefits. Yet the economic relationships that make a cluster self-sustaining require scale of a different kind. A specialist supplier needs enough customers to survive after one contract ends; a worker needs several employers if changing jobs is not to require a move back to Seoul; a university research group gains commercial relevance when firms capable of using its work are nearby; accountants, lawyers, engineers, investors, and technology companies deepen their capabilities only when recurring demand justifies specialization.

Distributing a fixed quantity of public employment among many places can therefore extend the geographic reach of the policy while reducing the concentration available in any single labor market. Concentrating institutions in fewer regional centers offers a better prospect of creating dense networks of firms and workers, but it also carries an obvious risk: the city strong enough to compete with Seoul may achieve part of that strength by drawing people and economic functions from smaller communities in its own region.

Migration into the first Innovation City host areas makes the problem tangible. In 2024, the municipalities containing Busan’s Innovation City functions received 6,983 movers from the Seoul metropolitan area and 13,504 from neighboring regions, according to the National Assembly Futures Institute’s analysis of domestic migration microdata. The figures describe all movers rather than transferred public employees and cannot establish that the relocation program caused their decisions, yet they reveal the spatial environment in which a hub strategy operates. Busan can reduce dependence on Seoul while simultaneously increasing its pull over surrounding parts of the southeast.

Whether that process advances national balance depends on what occurs across the wider regional economy. A financial cluster in Munhyeon has a different national significance if companies throughout the southeast gain better access to capital and specialized financial services; marine research in Dongsam matters more if firms in Busan, Ulsan, and South Gyeongsang can use the resulting knowledge; a deeper metropolitan labor market contributes more if skilled workers can move among several regional employers rather than face a choice between one local institution and Seoul. A regional center that merely accumulates population and functions from its neighbors would alter the map of concentration without resolving the economic hierarchy that made decentralization necessary.

Busan illustrates why established metropolitan centers are attractive in that debate without providing a simple argument for sending them more institutions. The city’s labor and housing markets were sufficiently deep to absorb transferred households at a rate unmatched elsewhere, and finance in Munhyeon, maritime research in Dongsam, and content industries in Centum offer recognizable economic bases around which related functions could potentially accumulate. The first-round evidence does not show that metropolitan scale alone guarantees a productive cluster, because proximity among headquarters matters economically only when institutions share workers, suppliers, research, capital, and decision-making relationships with the private economy around them.

The location of authority inside the institution becomes important at precisely that point. National Assembly Budget Office data show that relocated organizations can retain substantial operations in the capital region for legitimate reasons: Korea Securities Depository, for example, had 267 employees in capital-region facilities in the relevant analysis compared with 458 at its Busan headquarters, while Korea Asset Management Corporation had 108 employees in the capital region compared with 835 in Busan. Market access, government coordination, client relationships, or specialist labor can justify such arrangements, which means the headcounts alone reveal little about the completeness of relocation.

What matters for the regional economy is the work attached to those positions. Procurement authority, capital allocation, research leadership, executive decision-making, technology operations, and strategic planning influence firms and labor markets differently from routine staffing, so two headquarters with similar employment totals can produce very different local effects depending on where those functions are actually performed. The same principle will apply to any regional center chosen for the next round: colocating institutions creates administrative concentration, while agglomeration emerges only if workers can move among related employers, suppliers acquire capabilities they can sell to multiple customers, research reaches firms, capital circulates locally, and private investment eventually acquires reasons to remain without relying on the original transfer.

South Korea’s second relocation therefore presents a choice that cannot be resolved by counting institutions alone. A strategy driven primarily by territorial distribution may spread public employment too thinly to support deep labor markets and specialist industries, while a strategy driven exclusively by concentration could reproduce within the regions the hierarchy of dominant centers and peripheral communities that the original policy was meant to counter. The five-pole framework implicitly acknowledges that a country facing demographic contraction cannot reproduce advanced research, finance, medical services, universities, and specialized employment at equal scale in every municipality; the unresolved issue is whether strong non-capital centers can become credible alternatives to Seoul while extending their economic reach beyond the cities in which those functions are concentrated.

Busan still leads the national settlement table at 86.4 percent, and the wider evidence does not diminish the significance of that result. A large metropolitan region demonstrated that the relocation of a public institution can become a genuine household move rather than an office transfer accompanied by continued residential dependence on the capital region, which answers one of the practical questions that troubled the first round from its inception.

The national record narrows what the number can support. Residential settlement cannot establish whether Munhyeon developed a deeper private financial market, whether Dongsam’s research institutions generated durable commercial capabilities, whether Centum retained more of the value created by Korea’s content industries, or whether Busan’s institutional gains strengthened the southeastern economy beyond the city itself. Those outcomes develop through firms, investment, research, specialist labor, procurement, and professional networks whose effects appear on a different timetable and were only partially captured by the first-round scorecards.

South Korea has already demonstrated that a substantial part of the state’s institutional footprint can be redistributed and, under favorable urban conditions, that many of the households attached to those institutions can move with it. The second round begins with a more demanding standard because the long-term case for relocation will increasingly depend on whether public employment and institutional authority help create regional economies in which private firms, workers, capital, and knowledge acquire durable reasons of their own to remain outside Seoul.