Busan’s 16 Local Governments and the Fight for Fiscal Room

Busan’s 16 local governments operate under the same tax and transfer system, yet shrinking populations, rapid growth, inherited infrastructure and past commitments leave each with a different amount of money still open to choice.

Sasang-gu entered 2026 with a budget that appeared constrained but manageable. The district planned no new local government bond issuance, and the fiscal indicators accompanying its original budget gave little indication that an unusual financing problem would emerge within months. When officials prepared the first supplementary budget, however, the resources available for additional spending had fallen below the assumptions made at the beginning of the year, leaving a gap of roughly ₩14 billion. Sasang turned to its Integrated Fiscal Stabilization Fund and other available resources to cover the difference.

The gap did not follow a sudden collapse in local tax receipts. Resources expected to remain after the previous fiscal year closed were lower than the amount built into the original plan, even as projects already moving through Sasang’s aging industrial and residential areas continued to require funding. What changed was the amount of flexibility inside the budget. A district may begin the year with accounts that appear balanced and still lose room to maneuver once estimated revenue gives way to settled accounts and the resources actually available for execution.

Saha-gu, immediately to the south, reached a similar point through a different sequence. By late July, the district was projecting a 2026 funding gap of about ₩26 billion and had created a fiscal stabilization task force to identify savings in personnel-related expenditure, administrative costs and other parts of the budget. Its vulnerability arose from the combination of the projected shortfall and limited capacity to absorb it while continuing to finance welfare, environmental management and infrastructure across a district where older industrial zones, dense residential neighborhoods and a long coastline place overlapping demands on the same local finances.

Nam-gu faces the erosion of choice further ahead. The district government has said that investment commitments already in place could leave its 2027 finances short by ₩46.2 billion. Twenty-two projects carry combined costs of roughly ₩320 billion, with a substantial portion eventually requiring district funding. The concern is less a conventional debt problem than the convergence of obligations created in different years, as construction schedules, local matching shares and later operating expenses begin to claim the same future revenue and reduce the amount available for priorities that had not yet been conceived when those projects were approved.

Sasang, Saha and Nam entered the same fiscal year under the same national tax law and Busan Metropolitan City transfer system, yet the mechanism tightening each budget differs. Sasang discovered a gap when original assumptions met settlement and execution; Saha had relatively little cushioning once an emerging shortfall had to be absorbed; and Nam is confronting commitments accumulated across earlier budget cycles. Describing all three simply as fiscal shortages conceals the difference between a forecast that fails to hold, a buffer that offers limited protection and a future budget already carrying claims inherited from the past.

Geography adds constraints that annual fiscal ratios cannot capture. Yeongdo-gu may lose residents without losing the steep roads, shoreline, bridges and public facilities that still require maintenance. Across the Nakdonggang River, Gangseo-gu has to supply parts of a larger city as housing and population growth move faster than sections of the transport, medical and public-service network. Older industrial districts must finance renewal without abandoning infrastructure and neighborhoods that continue to support the existing economy, and the old urban core serves commuters, patients, tourists and shoppers who are largely absent from resident-population statistics.

Busan has one metropolitan government and sixteen gu and gun budgets, but the territory each government pays to operate has become increasingly different. The more useful measure of local fiscal capacity is therefore not simply the money recorded in an annual budget, but the portion that remains available after the city already in place has been paid for.

The Budget Is Not the Room

Much of a district budget already has a destination by the time the local council adopts it. Welfare programs, payroll, contracted services, ongoing construction, maintenance and local matching shares occupy large portions of expenditure before a new initiative enters the discussion. Revenue arrives with different degrees of flexibility as well: local taxes and non-tax income form the locally generated base, broadly usable transfers provide more discretion than earmarked subsidies, and some resources depend on settlement results or higher-level transfers that continue to change after the original budget has been adopted.

The headline size of a budget is consequently a poor proxy for policy space. Governments spending similar amounts may have markedly different abilities to absorb a revenue shock, postpone an existing project or initiate a new one.

Busan’s 2026 figures make the distinction visible in the gap between financial independence and financial autonomy. Financial independence measures the share of general-account revenue raised through local taxes and non-tax income. Financial autonomy expands that measure to include major forms of broadly usable intergovernmental revenue.

Jung-gu recorded a financial independence ratio of 22.05 percent and financial autonomy of 45.24 percent. Dong-gu moved from 17.15 percent to 39 percent, Seo-gu from 10.95 percent to 29.34 percent and Yeongdo-gu from 9.21 percent to 27 percent. General-purpose transfers materially expand the resources available for local decisions, although they do not strengthen the underlying local tax base in the same way as revenue generated within the district.

For Jung and Dong in particular, the gap reveals the fiscal structure behind districts whose economic roles extend well beyond their residential populations. Commercial activity, transport infrastructure, medical institutions and redevelopment are concentrated inside their boundaries, yet a substantial portion of the resources supporting district government arrives through redistribution from outside. Economic activity visible on the street and revenue controlled by the local government follow different maps.

A stronger own-source revenue base does not remove the distinction. Gijang-gun entered 2026 with the strongest verified financial independence and autonomy ratios among Busan’s gu and gun, at 27.38 percent and 50.05 percent, yet its consolidated fiscal balance was budgeted at minus 4.68 percent. The county raises a larger share of its revenue locally than most other districts, but its accounts also reflect investment and services spread across new towns, established settlements, industrial areas, coastal communities and rural territory.

The negative consolidated balance does not by itself indicate fiscal distress. Consolidated fiscal balance includes general accounts, special accounts and funds and may turn negative when accumulated resources are deliberately used for investment. Suyeong-gu provides the reverse warning because a small positive annual balance reveals little about future revenue already attached to facilities, programs or investment decisions that continue beyond the accounting year.

Welfare expenditure requires similar care. Buk-gu devoted 75.70 percent of its 2026 general-account spending to social welfare, the highest verified share among Busan’s districts. Yeonje-gu exceeded 70 percent, and Saha approached two-thirds. Those percentages combine programs financed through varying mixtures of national, metropolitan and district money and should not be mistaken for the share of locally raised revenue that has become unavailable.

Their importance comes from recurrence. Benefits, operating expenses and matching requirements return in subsequent budgets, and withdrawing a service already relied upon by residents is considerably harder than declining to begin one in the first place. Total expenditure may continue to rise even as the portion available for a genuinely new decision becomes smaller.

Original-budget ratios describe only the fiscal year’s starting point. Finance departments spend the following months incorporating revised transfers, settlement results, supplementary appropriations and changes in project schedules. Sasang’s 2026 experience showed how quickly the resources available for a supplementary budget could diverge from assumptions embedded in the original plan. Yeonje’s earlier accounts show movement in the other direction, with broadly usable revenue ultimately recognized during the year differing substantially from the amount recorded at the outset.

Fiscal room emerges from the interaction among recurring revenue, continuing obligations, usable reserves, transfers and commitments already in motion. Debt, annual balance and financial independence each answer important questions, but none reveals how much revenue remains after government has financed the expenditure it cannot readily stop.

Paying for the City You Have

Population changes faster than the physical city it leaves behind. Roads, drainage systems, retaining walls, public buildings and coastlines do not contract each time a resident moves away, and aging housing often requires greater intervention at the same time that fewer people remain to support the surrounding economy.

Yeongdo-gu shows the consequence more clearly than any other district in Busan. More than half of its housing stock falls within the older-housing measure used in the district comparison, at 57.81 percent, and vacant homes account for 14.35 percent. The district continues to maintain an inhabited island connected by bridges, threaded by steep roads and shaped by a shoreline where residential neighborhoods, shipbuilding activity and newer maritime uses occupy the same constrained territory.

A smaller population changes demand without removing most of that infrastructure. Schools may enroll fewer children, but their buildings do not become proportionately cheaper to maintain. Roads serving fewer households still need resurfacing, drainage and safety work. Older housing increases demands for redevelopment, fire prevention and neighborhood management even as the population supporting local commerce and consumption declines. The cost accumulates through ordinary maintenance and renewal rather than arriving as a single fiscal event.

Seo-gu carries another version of the imbalance. Older housing and vacancies remain substantial, yet the district also contains one of Busan’s densest concentrations of medical capacity, with 45.26 hospital beds per 1,000 residents and a comparatively high concentration of welfare facilities. Major hospitals bring workers, patients and visitors into Seo from across the metropolitan area, giving the district a service role far larger than its residential population suggests. The district government still has to finance hillside neighborhoods, welfare demand and housing renewal around those institutions, and metropolitan medical importance does not generate an equivalent stream of flexible district revenue.

Dong-gu places metropolitan development beside older residential conditions even more visibly. Busan Station, port-related activity and the North Port corridor make the district a transport gateway and a major development zone, yet older housing and vacant homes remain embedded in nearby neighborhoods. The same local government is responsible for ordinary residential streets and services beside projects whose institutional and economic scale extends well beyond the district boundary.

Jung-gu compresses the contradiction into a smaller space. Nampo-dong, Gwangbok-dong and Jagalchi attract shoppers, tourists, patients and workers from across Busan, while the resident population is small and the housing stock remains old. Hospital-bed supply per resident appears exceptionally high, although the small denominator helps produce that result because metropolitan facilities are being measured against a limited residential population. Markets, streets and public spaces serve far more people than the resident count alone suggests.

The old core therefore carries costs that population decline cannot describe on its own. Yeongdo has island geography and steep terrain, Seo combines medical concentration with older hillside neighborhoods, Dong operates beside large-scale redevelopment and Jung supports metropolitan commercial functions within a small residential base. National and metropolitan transfers sustain much of that inherited urban system, but infrastructure and settlement patterns adjust far more slowly than annual revenue.

Gangseo-gu faces the reverse timing problem. Housing and population are growing, yet parts of the service network have not matured at the same rate. Myeongji and the western development corridor have added large residential populations, and Eco Delta City is building another major urban district on land that until recently contained farmland, industrial sites, wetlands and smaller settlements. Only 8.20 percent of Gangseo’s housing falls into the older-housing category used in the district comparison, giving it one of Busan’s newest residential profiles.

Health care has not developed with the same intensity. Gangseo recorded 3.94 hospital beds per 1,000 residents and 114.44 medical institutions per 100,000 people, both near the bottom of the Busan comparison. Residents use hospitals outside the district, and facility counts alone cannot measure quality or emergency access, but the gap illustrates how quickly residential development may move ahead of a complete urban-service network.

Transport and schools create fiscal pressure earlier in the development cycle. Families occupy new neighborhoods before every road, transit link, classroom, drainage system and park has been completed, requiring public investment before a mature urban economy has generated its full recurring revenue. Once the infrastructure opens, the capital expense is followed by staffing, maintenance, utilities and periodic repair.

The arrival of a new district administration in Gangseo also provided a smaller example of expenditure that lies closer to immediate political control. Shortly before Park Sang-jun took office as district mayor on July 1, the district spent ₩32.85 million refurbishing the existing mayoral suite, a roughly 230-square-meter space that had also been renovated when the previous administration began four years earlier. Public reporting identified wall finishes, tile and lighting among the work, and another ₩2.2 million was spent on 26 inauguration banners. Park had campaigned on relocating the mayor’s office and later proposed moving Gangseo District Office itself during his term.

Photographs taken by Breeze in Busan during the refurbishment show the existing sofa and chairs removed from the mayoral office as the space was being prepared for the new administration. The photographs do not establish their final disposition, and the money involved is negligible beside the cost of the transport infrastructure Gangseo is seeking. The episode is relevant at a narrower scale because it distinguishes expenditure driven by population growth and metropolitan infrastructure needs from spending that district officials are able to approve, defer or redesign themselves.

Major outside investment in Gangseo and scrutiny of its discretionary spending are compatible questions. Small administrative savings could never finance the district’s rail, road or medical needs, yet structural infrastructure pressure does not remove the obligation to explain expenditures government controls directly.

Gijang-gun shows how the calculation changes after a stronger local revenue base has developed. Its revenue structure gives the county greater capacity to finance local investment, but medical provision remained relatively thin at 10.35 hospital beds per 1,000 residents, and services have to reach Jeonggwan, older coastal settlements, industrial areas, rural communities and development sites separated by considerable distance.

Greater revenue capacity helps finance infrastructure, but every road, public building and community facility expands the network future budgets have to operate. Gijang’s planned consolidated deficit, despite its comparatively strong revenue base, illustrates how fiscal capacity and annual investment demands may move in different directions.

Yeongdo, Gangseo and Gijang occupy different points on Busan’s demographic map, yet all three expose a timing problem in local finance. Yeongdo is maintaining infrastructure inherited from a larger past, Gangseo is building services for a larger future that has not fully matured, and Gijang has moved further through the growth cycle only to inherit the operating costs of a larger and more dispersed public network.

Where Value and Revenue Part Ways

Economic activity does not remain within the fiscal level of government that happens to contain it. Ports, railway stations, tourism districts and commercial centers may create enormous value inside one administrative boundary even though taxes, land ownership, regulatory authority and infrastructure responsibilities are divided among Busan Metropolitan City, the national government, public corporations and the district itself.

Jung-gu makes the separation particularly visible. Commercial and medical functions serve populations much larger than the number of residents recorded within the district. As the earlier fiscal ratios show, however, a substantial part of the district’s broadly usable revenue arrives through the transfer system rather than its own local tax base. That does not measure how much economic value Jung fails to capture, but it demonstrates that metropolitan commercial importance and locally generated district revenue are different things.

Dong-gu sits inside an even more fragmented institutional landscape. Busan Station, port land and the North Port redevelopment corridor involve Busan Metropolitan City, Busan Port Authority, public corporations, railway interests and national institutions. Economic activity around the station and port supports commerce, employment and land values, but those gains do not arrive at Dong-gu District Office as a directly equivalent pool of discretionary revenue.

Busanjin-gu shows the separation without the port. Seomyeon functions as one of Busan’s principal retail, office, medical and transport centers, and the district operates on a much larger fiscal scale than Jung or Dong, with a consolidated budget exceeding ₩1 trillion in 2026. Its financial independence and autonomy ratios remain far more modest than that economic scale might imply, and the own-project ratio remains below 10 percent. A large local economy can therefore coexist with a budget in which substantial expenditure depends on transfers or programs financed jointly with higher levels of government.

Suyeong-gu brings tourism and property value into the same discussion. Gwanganli supports one of Busan’s most visible waterfront economies, backed by restaurants, accommodation, expensive housing and heavy visitor traffic. Its district revenue structure remains more dependent on transfers than the visual prosperity of the waterfront might suggest. Tourism also produces public costs that resident-based indicators capture imperfectly because cleaning, traffic management, parking, public safety and environmental maintenance respond to populations that change sharply by hour and season.

No defensible formula can turn those four districts into a single measure of “uncaptured value.” Jung’s small resident base, Dong’s port and railway governance, Busanjin’s commercial concentration and Suyeong’s visitor economy create different relationships between economic activity and local revenue. Their common feature is a mismatch between the places where economic value becomes visible and the level of government controlling the public revenue associated with it.

Moving the analysis inward from the district to the neighborhood exposes another limitation. District averages are necessary for comparing governments, but residents encounter public services at a much finer scale.

Haeundae-gu illustrates the problem. Its beachfront, Centum City and high-value residential developments share one district budget with Banyeo, Bansong and older inland communities. Older housing accounts for 24.25 percent of the district’s stock, a relatively favorable figure within Busan, yet social-welfare facilities and hospital-bed provision rank less strongly in the districtwide comparison. The figures do not establish that a particular neighborhood is underserved, but they rule out the assumption that high property values describe service availability throughout Haeundae.

Nam-gu contains an equally mixed urban geography. Universities account for more than 46,000 students in the comparative dataset, major investment projects are creating future district obligations, and hospital and welfare-facility provision remains relatively weak by Busan standards. University populations support retail, rental housing and cultural activity without necessarily becoming permanent residents, and redevelopment may raise land values at the same time that roads, public facilities and other costs are added to future budgets. A district average compresses those different processes into one fiscal profile.

Suyeong’s service indicators point in several directions. Medical institutions are comparatively plentiful, whereas elementary-school provision, senior leisure facilities and green space rank lower. Tourism requires waterfront management and public safety, and residents still require schools, parks and neighborhood services whose geography and schedules bear little relation to peak visitor demand. Both arrive inside the same district budget.

Dongnae-gu offers a quieter version of the problem. Housing and medical provision are comparatively strong, with green space and childcare less abundant. The fiscal challenge increasingly concerns the renewal of an established city rather than outward expansion, as roads, public buildings and neighborhood facilities already in place require maintenance and replacement through numerous ordinary budget lines.

Geumjeong-gu combines more than 37,000 university students, extensive green land and comparatively strong welfare provision with a different demographic problem: whether those assets support a durable residential and economic base as the population ages. Students sustain commercial activity without necessarily settling permanently, and extensive green land contributes to environmental quality without producing the same fiscal effects as developable property. Their budgetary value depends partly on whether they ultimately contribute to population retention and recurring economic activity.

The comparison across Haeundae, Nam, Suyeong, Dongnae and Geumjeong is more useful as a warning about averages than as a ranking. Financial independence describes revenue origin, welfare ratios classify expenditure, and hospital, school or green-space measures capture selected dimensions of the service environment. Each indicator remains valid within its scope, but none can substitute for the geography residents actually experience.

The Budget Already Spoken For

Large public projects are most visible when governments announce them. Total investment, funding shares and construction schedules dominate public debate, although a project’s fiscal life extends well beyond the announcement and often well beyond the opening ceremony.

Nam-gu’s projected 2027 funding gap shows why the timing matters. The district government says twenty-two major investment projects carry combined costs of about ₩320 billion, with roughly ₩160 billion ultimately attributable to the district and approximately ₩20 billion in additional district funding required each year over the next four years. The projected ₩46.2 billion gap reflects the point at which projects approved at different times begin drawing on the same future budgets.

Total project cost says relatively little about the amount of district money needed in a particular year. National and metropolitan contributions, expenditure already completed and financing outside the district’s discretionary budget may all be included in the headline figure. The constraint lies in the schedule of local contributions, particularly when projects that once occupied separate budget years begin to overlap.

Nam-gu’s 2024 closing accounts recorded no outstanding local-government debt, a fact entirely compatible with pressure projected for 2027. Debt reflects obligations recognized under a particular accounting category at a particular time and does not capture every matching share, construction commitment or future operating expense that may compete for later district revenue.

The obligation changes once construction is complete. Personnel, electricity, security, cleaning, maintenance contracts, programming and periodic repair enter recurring budgets without attracting the same attention as the original capital appropriation. Outside grants may reduce construction costs without paying to operate a facility indefinitely, and user fees often cover only part of the expense. A project may therefore continue narrowing future policy space long after its construction budget has disappeared.

Gijang-gun has greater capacity than most Busan districts to absorb such obligations because its revenue base and own-project ratio are stronger. Greater capacity also permits more investment, however, and spending on transport, education, sports and cultural infrastructure expands the stock of facilities later budgets must maintain. Long-term affordability depends on the cumulative burden rather than the apparent affordability of each project at the moment of approval.

Haeundae confronts the same issue within a more mature urban environment. Tourism, expensive housing and newer development support economic activity and raise expectations for public space, cultural infrastructure and visitor management. New facilities enter a budget already responsible for older neighborhoods and established welfare services, adding another obligation to a system that is already extensive.

Elsewhere, future budgets are constrained less by individual megaprojects than by recurring commitments accumulated through ordinary programs. Buk-gu entered 2026 with a weak own-source revenue base and a general-account budget dominated by welfare spending. Yeonje-gu manages a dense administrative and residential district while incorporating revenue that changes during the fiscal year. Sasang-gu is financing industrial and residential renewal while adjusting to resources that did not match assumptions in the original budget, and Saha-gu carries welfare, industrial, environmental and coastal obligations with comparatively little protection against a large forecasting error.

The work of a budget office therefore extends far beyond allocating money. Revenue estimates must be updated, transfers incorporated, projects sequenced, matching shares secured, procurement completed and existing services maintained at the same time. A program may appear fully financed on paper and still reach residents late if a contract stalls, land acquisition takes longer than expected or staffing is unavailable when a facility opens.

Simple comparisons of expenditure rates or staff totals offer little reliable guidance on administrative competence. Programs with straightforward payment schedules naturally spend faster than complex construction projects, and staffing requirements vary with the service burden assigned to each district. A more useful test is whether an appropriation survives the administrative process and becomes the service, facility or infrastructure residents were promised.

Reserves also have an operational value beyond their nominal balance. Drawing on a stabilization fund can prevent an immediate disruption, but the same resources cannot absorb the next one until the fund is replenished. A surplus carried into the following year initially widens available choices, yet that flexibility disappears once the money is committed. Buffers provide time and maneuvering room as much as cash.

Viewed from that perspective, Busan’s sixteen districts resemble different operating systems placed inside the same legal framework. Yeongdo maintains a city built for a larger population, Gangseo assembles services for one still expanding, Nam sequences commitments inherited from earlier years, Sasang adjusts when projected resources fail to materialize as expected, and Buk preserves everyday services inside a budget dominated by welfare. The structure of the city each government must operate matters long before the annual balance is calculated.

Governing the Unequal City

Population decline eventually raises a question about administrative scale itself. Jung-gu, Seo-gu, Dong-gu and Yeongdo-gu each maintain a separate autonomous district government with an executive structure, budget process, personnel system and administrative departments of its own. Their terrain and service needs differ substantially, yet all four govern parts of an old urban core where resident populations have declined and many of the largest policy problems cross district boundaries.

Using the population counts published with Busan Metropolitan City’s staffing disclosure, dated Dec. 31, 2024, Jung, Seo, Dong and Yeongdo had a combined 335,188 residents, fewer than Busanjin-gu’s 369,533. Busan’s authorized staffing figures, updated March 31, 2026, list 505 positions in Jung, 667 in Seo, 646 in Dong and 655 in Yeongdo, producing a combined total of 2,473 compared with 1,069 in Busanjin.

Population alone cannot determine how many employees a district needs. Yeongdo’s island geography, Seo’s medical concentration and hillside neighborhoods, Dong’s redevelopment pressures and Jung’s commercial functions create workloads that differ markedly from Busanjin’s. The comparison nevertheless shows that administrative systems carry fixed organizational costs that do not fall in proportion to resident population.

Separating neighborhood-level staffing from the rest of the organization makes the scale question more concrete. Jung, Seo, Dong and Yeongdo together report 1,932 positions outside the neighborhood-level administrative structure, compared with 765 in Busanjin. The difference is not a count of redundant positions; it identifies the portion of government that would require function-by-function examination before any credible estimate of consolidation savings could be made.

Some administrative functions are more plausible candidates for sharing than others. Budget systems, payroll, procurement, information technology and parts of planning or facility management may serve a larger territory without forcing residents to travel farther for routine services. Welfare counseling, civil-service counters, neighborhood administration and field maintenance depend much more heavily on local accessibility. Treating those categories as interchangeable would mistake administrative overhead for frontline government.

Busan examined the merger question in 2017, when research commissioned from the then-Busan Development Institute and the Korean Association for Local Government Studies argued that combining Jung, Dong, Seo and Yeongdo could improve fiscal and administrative efficiency. The study reported a benefit-cost ratio of 1.35 and projected substantial savings, but the estimate belongs to an earlier fiscal and urban context and later became part of a controversy over citation and plagiarism allegations concerning portions of the research. It cannot serve as a current estimate of what consolidation would save.

Local representation also carries value that does not appear on a payroll line. Smaller governments provide direct political representation and may keep services closer to neighborhoods where populations are older and mobility is limited. A larger jurisdiction could reduce duplication in administrative functions and still create other costs if decision-making and service access moved farther from the communities affected.

The choice therefore extends beyond full merger or complete preservation of the status quo. Districts could share procurement, information systems, specialist staff or selected planning functions without abolishing their governments, and Korea’s legal framework permits special-purpose local governments to perform certain responsibilities jointly across jurisdictional boundaries. For Busan, the practical issue is which functions genuinely benefit from four independent administrative systems and which could be delivered across a larger service area without weakening local access.

The OECD’s 2026 review of Korea identifies the same issue at a wider scale. As populations contract, per-person service costs rise where economies of scale matter, even though transport, housing, land use and other urban functions often operate across administrative boundaries that remain institutionally separate. Shared-service platforms and inter-municipal cooperation provide possible ways to reduce those scale costs without requiring every political boundary to disappear.

Busan’s old core is a useful test because the mismatch operates in both directions. Transport, redevelopment, tourism and housing renewal frequently extend across several districts and may be poorly suited to fragmented administration. Within each district, however, averages can conceal neighborhoods with sharply different service needs, making some boundaries too small for metropolitan problems and too large to describe everyday local conditions.

Any serious redesign would have to address both. Back-office functions would need examination for duplication, neighborhood services for accessibility and demand, and the transition itself for the cost of integrating information systems, personnel structures, offices and political institutions. Current staffing figures establish a reason for undertaking that analysis rather than supplying a predetermined conclusion.

Local Government Runs Out of Room Before It Runs Out of Cash

The comparison across Busan reveals a problem that conventional fiscal rankings describe poorly. A district may balance its annual accounts and carry little debt even as future governments inherit a narrower range of choices, whereas another may run a planned deficit while retaining stronger revenues and accumulated resources. The decisive difference lies in how much of the budget already has a claim on it before a new proposal reaches the council chamber.

Existing welfare programs, payroll, infrastructure maintenance, local matching shares and capital commitments enter the spending structure first. Geography determines how expensive those obligations are to deliver, demographic change alters the population using them, and decisions approved years earlier may reserve revenue long before a future council begins debating its own priorities.

District governments remain responsible for the choices they control, including which projects to initiate, when to use reserves, which public assets to replace and how much recurring expenditure to create. Much of the wider structure lies beyond the district level, however, because Korea assigns substantial spending responsibilities to local governments while leaving them with comparatively limited independent authority over revenue.

For Busan Metropolitan City, the resulting policy question concerns what the metropolitan redistribution system is intended to equalize. Tax capacity and resident population remain indispensable measures, but neither fully captures the cost of maintaining a shrinking island district, supplying infrastructure to a fast-growing new town, supporting a commercial center used by the wider metropolitan population or operating a facility after outside construction funding has ended.

A transfer system more closely aligned with expenditure needs would have to consider demographic decline, density, fixed territorial costs, pre-investment in growth areas, metropolitan service functions and future operating commitments alongside each district’s revenue-raising capacity. The OECD has similarly argued that fiscal equalization in Korea needs to respond more explicitly to ageing, depopulation, density and service costs while preserving incentives for local revenue effort and careful spending.

No transfer formula can remove the choices local governments still have to make. Structural disadvantages do not justify every project, and additional revenue does not remove the obligation to scrutinize discretionary expenditure. Administrative consolidation requires the same discipline because savings achieved through shared back-office functions would have little value if they simply transferred costs to residents through weaker access to frontline services.

Busan’s sixteen districts are therefore facing different versions of the same underlying constraint, shaped by revenue forecasts, existing welfare systems, inherited infrastructure, economic geography, development and commitments made years earlier. The most revealing measure of local fiscal strength is the portion of future revenue that remains uncommitted after those claims have been financed, because a local government may continue paying every bill even after it has begun to lose the capacity to choose what comes next.