Unlocking Value: Lessons from Korea's Land and Property Valuation and Taxation System

Summary

This October 2026 World Bank report (report number P506840, prepared by the Bank's Urban, Resilience and Land Korea Team) describes how South Korea built a single, publicly announced official valuation of every land parcel and house, and how that valuation became the base for local property tax, the national Comprehensive Real Estate Holding Tax (CREHT), development charges and welfare eligibility. It is a descriptive institutional report for policymakers in other countries, not a study of land value taxation: it values land per parcel, but since 2005 residential property tax has been levied on land and building combined, and the report does not discuss taxing land alone.

The System as the Report Describes It

Four values before 1989. Before 1989 four agencies each kept their own land value for their own purpose: the Taxable Standard Market Value (Ministry of Home Affairs, local tax), the Reference Market Value (National Tax Service, income and capital gains tax in designated speculative areas only), the National Land Value (Ministry of Construction, compensation for public acquisition) and the Appraised Market Value (Korea Appraisal Board, general market reference). The report says the Taxable Standard Market Value "consistently lagged actual market prices, typically at only 15-30 percent of prevailing transaction values," which it attributes to the absence of clear standards, irregular surveys and subjective decisions by untrained officials (PDF p. 15). The 15 to 30 per cent figure is stated in the body text for that one value; the Executive Summary generalises it to the pre-1989 values as a group, and the report's own Annex 4 (four parcels, compiled from a 1998 newspaper) shows the Reference Market Value at roughly 41 to 90 per cent of an appraiser's value against 5 to 23 per cent for the Taxable Standard Market Value (PDF p. 50).

The 1989 reform. The Public Notice of Values and Appraisal of Lands, etc. Act (1989) unified survey schedules, standards and responsible agencies and created an annual, standardised survey and public announcement, with 1 January as the base date; the earlier dual licences for land appraisers and "authorised" appraisers were merged into one certified-appraiser profession (PDF p. 16). The nationwide rollout in 1991 covered "more than 30 million parcels at that time" (PDF p. 13). The report ties the Act to a "public concept of land ownership" (see below).

Three steps from representative parcel to individual parcel (PDF pp. 17, 25-27):

1. The national government selects about 600,000 representative parcels out of roughly 35 million, which are surveyed and appraised by certified appraisers working under the Korea Real Estate Board (REB), and their values are published.
2. An index table is built for each smallest administrative unit (eup, myeon or dong) and each zoning type, using a hedonic price model (a multiple linear regression of price on characteristics such as zoning, land use, area, road access and terrain). The report's worked example: if the representative parcel has a mild slope and the individual parcel is flat, the index ratio of 1.16 means the individual parcel is 16 per cent more valuable; applied to a representative value of KRW 13.9 million per square metre it gives KRW 16,124,000.
3. Local governments survey the same characteristics for each individual parcel, apply the index ratios, verify the result and publish it for public inspection. Owners may object within 30 days of announcement, and the local government must respond within 30 days of the objection deadline.

The report states that the system has run annually since 1989: the representative and individual parcel values have been published 37 and 36 times respectively (PDF p. 17).

Houses and transaction prices. Land and buildings were valued separately until the Publicly Announced House Value System of 2005; multi-unit houses were first announced in April 2005 and all house types from 2006. Non-residential buildings are not publicly announced (PDF p. 18). From 1 January 2006 buyers, sellers and agents must report the actual price within 30 days of signing a contract, a rule introduced to curb "dual contracts" that under-declared prices (PDF p. 47). Reported prices feed back into valuation after statistical screening and expert review (PDF pp. 42-43).

Digital lineage. Computerised calculation of individual land values began with the Automatic Land Price Appraisal system (ALPA) in 1993; the function moved into the Korea Land Information System (KLIS) in 2005 and into the Korea Real Estate Administration Intelligence System (KRAS) in 2015; since January 2024 it runs on a dedicated system, KOREPS, hosted on the government cloud (PDF pp. 18, 23). The report's annexes give other dates for KLIS and KRAS; its main text gives 2005 and 2015.

Some 60 administrative uses. The announced values serve "some 60" separate functions, which Annex 7 breaks down as 10 welfare items, 4 charges, 19 administrative items, 8 taxation items and 19 appraisal-related items (PDF pp. 53-57). Two uses are worth singling out. The development charge is calculated from the publicly announced land value under the Act on the Restitution of Development Gains (PDF p. 54). And for the National Basic Livelihood Security System and the Basic Pension, announced land and house values are converted into a monthly "recognised income" that decides eligibility; for regional health-insurance subscribers they form the asset part of the premium (PDF p. 17).

Taxes built on the values. Local property tax is based on the announced value multiplied by a Fair Market Value Ratio, 60 per cent for residential property (43 to 45 per cent for eligible single-home households in 2026) and 70 per cent for land and non-residential buildings; rates for residential property run from 0.1 to 0.4 per cent, and local governments may vary them by up to 50 per cent either way (PDF pp. 30, 33-34). Land is classed as comprehensively aggregated (idle or non-business, 0.2 to 0.5 per cent), separately aggregated (business, 0.2 to 0.4 per cent) or separately taxed (farmland, forest and some others, 0.07 to 0.4 per cent) (PDF p. 31). The national CREHT applies to owners whose holdings exceed thresholds, and the report says its entire revenue, USD 2.94 billion in 2023 (citing the Korea Development Institute), is redistributed to local governments with priority to fiscally weaker ones (PDF p. 28).

How much local revenue. Property tax was 13.2 per cent of local tax revenue in 2023 (Ministry of the Interior and Safety). The report's "nearly 30 percent" for "property-related taxes" is property tax plus acquisition tax, not property tax alone, and the report gives the acquisition share inconsistently: 16.1 per cent in the main text (PDF p. 31) and 21.6 per cent in Section 4.2.4 and Table 9 (PDF p. 36). With 16.1 per cent the two taxes sum to about 29 per cent.

Assessment ratios and the 2020 plan. In 2020 the government launched a Real Estate Value Realization Plan to raise the ratio of announced to market value to 90 per cent (land by 2028, apartments by 2030, single-family houses by 2035). The report gives the Ministry of Land, Infrastructure and Transport's 2020 ratios as 69.0 per cent for multi-unit housing, 53.6 per cent for single-family houses and 65.5 per cent for land, and says these are the baseline of the original plan, not current ratios (PDF p. 34). In 2023 the government revised the plan, reverting to the 2020 levels and keeping reduced Fair Market Value Ratios in response to concerns over tax burdens; announced multi-unit values fell 18.6 per cent and property tax revenue 13.4 per cent (PDF p. 39).

Four "lessons". The report closes with four lessons for other countries: separate government-produced public values from private appraisals under a mandated central authority and a professional workforce; connect data systems, with common standards and phased pilots; build on standardised spatial data and mandatory, verified transaction reporting; and keep valuation separate from tax-rate decisions by using an adjustment ratio with statutory floors and ceilings and a 30-day objection right (PDF pp. 9-10, 40-44). These are the authors' judgement, and the report itself says "Korea's experience is not a blueprint" (PDF p. 9). Its further claims about growth, jobs, collateral and credit are asserted rather than evidenced.

Relation to the Georgist Case

The strongest link is to the land-cannot-be-assessed objection. Korea publishes a parcel-level land value for essentially every parcel in the country every year, produced by a representative-parcel method with a hedonic index table that a local official can apply without hand calculation, and subject to public inspection and a 30-day objection right. That is a national-scale demonstration that annually updated parcel values are administratively achievable, in a country of about 35 million parcels, and that a unified official valuation replaced one that, for the Taxable Standard Market Value, had run at 15 to 30 per cent of market prices. The report records, however, that the ratio of announced to market value still varies by region and property type and that the disparities "have not been fully resolved" (PDF p. 29).

The limits matter for a land value tax reading. The report does not say whether Korean parcel values are estimated as if the land were vacant (a site-value basis), and its list of land characteristics includes actual land use (PDF p. 27). Since 2005 the residential tax base combines land and building, non-residential buildings and their land are appraised separately, and the report says nothing about taxing land alone. Korea is therefore evidence about valuation capacity, not an example of a land value tax.

The closest the report comes to Georgist policy is its account of the late-1980s "public concept of land ownership," defined in Annex 6 as: "Land is a shared asset of the state or society as a whole. Private ownership is recognized, but restrictions on use and disposition may be imposed to prioritize the public interest" (PDF p. 52). The report dates three measures to 1989: the Act on Ceiling on Residential Land Ownership, the Land Excess Profit Tax Act (a progressive tax on land holdings above a threshold) and the Development Gain Reclamation (Restitution) Act, which reclaims part of development gains. The Constitutional Court ruled the Land Excess Profit Tax Act unconstitutional in 1994, citing "excessive burden, double taxation, and irrational tax rates," while upholding the public concept itself; the Ceiling Act and the tax were repealed in 1998, and the development gain law remains in force, amended in 2023 (PDF p. 52). Elsewhere the report records that the local Comprehensive Land Tax, a progressive tax on each owner's aggregate landholding, was abolished, and that the CREHT of 2005 added a progressive national tax on top of the existing local taxes (PDF pp. 15, 47). For the wider picture of Korean land policy see South Korea; for the general idea of recovering land value gains see land value capture.

Nuances and Limits

• Not an LVT case. The valuations are per parcel, but the taxes are on land plus building for houses, with separate rate schedules by land use; the report's lessons concern valuation administration, not the choice of tax base.
• The 15 to 30 per cent figure is narrower than it looks. It is the body-text statement for one pre-1989 value; the Executive Summary generalises it, and Annex 4 shows a much higher ratio for the National Tax Service's Reference Market Value.
• Two ways of counting local revenue. Property tax alone is 13.2 per cent of local tax revenue; "nearly 30 per cent" adds acquisition tax, and the report's two figures for acquisition tax (16.1 and 21.6 per cent) cannot both be right.
• The 2020 ratios are a starting point. The 69.0, 53.6 and 65.5 per cent figures are the baseline of a plan that was walked back in 2023; they should not be read as current assessment ratios, and for multi-unit housing the report only says the 2026 ratio was held at 69 per cent.
• Inconsistent dates. The report gives KLIS and KRAS dates that differ between the main text and some annexes (the main text says 2005 and 2015); it also says in Annex 12 that the 2016 Act introduced a non-residential announcement while the main text says non-residential values are not announced.
• Thin profession. Only 4,877 certified appraisers are reported for the whole profession in 2025 (PDF p. 21); the report does not say how many work on public valuation.
• Secondary and third-party figures. Pre-1989 coverage data come from Chae (1992) and the 1998 newspaper table; the CREHT revenue comes from KDI; the claim that the 2019 and 2021 CREHT rate rises reduced housing-price disparities and asset inequality rests on one study (Hong and Hwang 2024), and the OECD (2024) is cited for the view that Korea's progressive national property taxation balances redistribution with fiscal sustainability.
• Reform politics are thinly described. The report says CREHT reaching single-home households "sparked public and political debate" and that the plan was revised "in response to concerns over tax burdens"; it names no opponents.
• Dated. The report describes 2026 policy settings, which change annually.

Bears On

• Objection: Land value can't be assessed accurately
• Method: Mass Appraisal Methods
• Place: South Korea

See Also

• World Bank, Determinants of Property Tax Revenue — the cross-country finding that valuation and administrative capacity drive property-tax yield
• IMF, Taxing Immovable Property
• Almy, Valuation and Assessment of Immovable Property (OECD)
• Land Value Tax
• Site Value
• Land Value Capture

Sources

1. World Bank Group, Urban, Resilience and Land Korea Team (October 2026), Unlocking Value: Lessons from Korea's Land and Property Valuation and Taxation System, report P506840 — used for: the four pre-1989 values and the 15-30 per cent figure (PDF p. 15; Executive Summary generalisation at p. 6 and Introduction at p. 11; Annex 4 at p. 50); the 1989 Act, the 1991 rollout and the certified-appraiser merger (PDF pp. 13, 16); the 600,000-of-35-million design, the index table and its 1.16 worked example, the 30-day objection window and the 1 January base date (PDF pp. 17, 25-27); houses from 2005-2006 (PDF p. 18); the 2006 transaction-reporting rule (PDF pp. 13-14, 47); the ALPA, KLIS, KRAS and KOREPS dates (PDF pp. 18, 23); the roughly 60 uses and Annex 7's 10/4/19/8/19 breakdown, including the development charge and recognised income (PDF pp. 17, 53-57); the land and building tax structure, Fair Market Value Ratios and rates (PDF pp. 30-31, 33-34); the CREHT redistribution (PDF p. 28); the 2023 local-tax shares and their inconsistency (PDF pp. 8, 10, 31, 36); the 2020 ratios, the 2020 realisation plan and its 2023 revision (PDF pp. 33-34, 39); the public concept of land ownership, the 1989 laws, the 1994 Constitutional Court ruling and the 1998 repeals (PDF p. 52); the 4,877 appraisers (PDF p. 21); and the four lessons (PDF pp. 9-10, 40-44). Page numbers are PDF pages (printed page = PDF page minus 10). Claim grades: the report's descriptions of statutes, agency roles, dates and procedures are A/B-claims, with figures taken from Korean ministries as B; the lessons and the growth, jobs and credit assertions are C-claims (authors' judgement) or D-claims (asserted without evidence). The World Bank is a Tier 1 institution under the wiki's attribution rules. The report cites the Ministry of Land, Infrastructure and Transport (2020 assessment ratios, valuation procedures), the Ministry of the Interior and Safety (2023 local-tax shares), the Korea Development Institute (CREHT revenue), Chae (1992) (pre-1989 coverage), the Maeil Business Newspaper (1998) (price comparison), Hong and Hwang (2024) and the OECD (2024). The report carries the usual disclaimer that its findings do not necessarily reflect the views of the World Bank's Executive Directors. Full text read from the PDF at last review (2026-11-08); the landing page is JavaScript-rendered. Landing page