Pith. sign in

REVIEW 5 minor 31 references

KRAFT releases 5.32 million Korean apartment sales with aligned socioeconomic context for reproducible housing research.

Reviewed by Pith at T0; open to challenge. T0 means a machine referee read the full paper against a public rubric. the ladder, T0–T4 →

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

T0 review · grok-4.5

2026-07-15 09:10 UTC pith:UEANTXNW

load-bearing objection Solid data-descriptor infrastructure: 5.32M cleaned Korean apartment sales plus carefully separated auxiliaries; novelty is packaging, not new measurement.

arxiv 2607.11961 v1 pith:UEANTXNW submitted 2026-07-12 econ.EM

KRAFT: A Transaction-Level Dataset for Korean Apartment Sales Integrated with Contextual Indicators

Sejin Myung , Hyungjoon Kim This is my paper
classification econ.EM
keywords housing marketsapartment transactionsSouth Koreatransaction-level datasocioeconomic indicatorsdata descriptorregional economicsmacro-financial transmission
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

Housing studies need individual sale records and the surrounding economic and social conditions, but those sources usually live in separate systems with mismatched labels and time stamps. This paper presents KRAFT, a public dataset of more than five million non-cancelled South Korean apartment sales from 2015 through 2024, covering every first-level administrative region. Each sale carries timing, location, size, price, floor, and construction year. Alongside the sales sit cleaned macro-financial, demographic, school, private-education, house-price-index, consumer-sentiment, and policy-uncertainty series, kept at their original spatial and temporal grain rather than forced into one table. The release therefore lets researchers build transaction-level or aggregated panels without first reconciling fragmented official files.

Core claim

KRAFT is a nationwide, transaction-level dataset of 5,320,379 non-cancelled South Korean apartment sales from January 2015 to December 2024, released together with harmonized auxiliary indicators for macro-financial conditions, demographics, education, housing-price indices, consumer sentiment, and economic policy uncertainty, organized so that original resolutions are preserved and multi-resolution analyses remain reproducible.

What carries the argument

Component-level release structure: year-specific transaction CSVs plus separate auxiliary tables joined by standardized YearMonth (and Sido/Sigungu where applicable), after administrative-label harmonization and documented spatial mappings.

Load-bearing premise

The mapped consumer-sentiment and private-education variables are treated as usable local context even though they were expanded from coarser regional groupings rather than measured independently in every municipality.

What would settle it

Check whether the documented CSI-to-Sido and private-education-to-Sigungu mappings produce systematically biased coefficients or coverage gaps when the same housing models are re-estimated on unmapped original source tables for overlapping subperiods.

Watch this falsifier — get emailed when new claim-graph text bears on it.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit.

Referee Report

0 major / 5 minor

Summary. This Data Descriptor presents KRAFT, a nationwide transaction-level dataset of 5,320,379 non-cancelled South Korean apartment sales from January 2015 to December 2024, covering all 17 Sido regions. Each record includes YearMonth, Sido/Sigungu/Beopjeongdong, exclusive area, reported price (10,000 KRW), floor, and construction year. The package also releases separate auxiliary tables for macro-financial series, demographics, school counts, mapped private-education expenditure, Sido-level housing price indices, regionalized consumer survey indices, and Korean economic policy uncertainty, with English variable names, YearMonth keys, and region-harmonization metadata. Files are organized as annual transaction CSVs plus component-level auxiliaries so that original spatial and temporal resolutions are preserved; Zenodo deposit and validation code are provided.

Significance. For housing-market, household-finance, and regional-economics research on Korea, a cleaned, documented, multi-resolution package that reduces the usual multi-institution preprocessing burden is of clear practical value. Strengths that raise the contribution above a simple data dump include: (i) explicit integrity validation (schema-consistent annual files, 120 months, 17 Sido, zero exact duplicates, no non-positive prices/areas, full Sido-month coverage); (ii) transparent handling of administrative changes (e.g., Gunwi-gun, Incheon Nam-gu/Michuhol-gu) via Region_Mapping.csv; (iii) public Zenodo release and validation code; and (iv) clear Usage Notes that flag mapped CSI and private-education variables as contextual proxies rather than independent local surveys. If the released files match the reported checks, the dataset is immediately usable for price modeling, regional comparison, and macro-financial transmission studies.

minor comments (5)
  1. Several reference access dates are listed as '5 July 2026' (e.g., [1], [5]–[14]). These appear to be future or placeholder dates and should be corrected to the actual access dates used.
  2. §4.2 and Table 4: the 225 records with construction year later than contract year (0.0042%) are retained as source-consistent; a one-sentence note in Variable_Dictionary.csv or Usage Notes on how users might filter pre-completion sales would further reduce misuse risk.
  3. Usage Notes correctly warn that CSI and private-education variables are mapped from coarser source categories. Adding a short cross-reference in §§2.3.2 and 2.3.5 to the exact mapping rules in Region_Mapping.csv (or a dedicated mapping appendix) would make the documentation fully self-contained.
  4. Table 2 and §3.1: education merge keys mention a 'normalized parent-city Sigungu' join for cases such as Yongin-si Sujigu. A one-line example of the parent-city key construction in the README or code repository would help users avoid silent merge failures.
  5. Minor typographical consistency: 'exc lusive', 'macro -financial', and similar spaced hyphens appear in the abstract and early sections; a final copy-edit pass would clean these.

Circularity Check

0 steps flagged

No circular derivation: KRAFT is a data-descriptor compilation with integrity checks, not a predictive model that reduces outputs to fitted inputs.

full rationale

This paper is a Scientific Data–style data descriptor. Its central claim is the existence, cleaning, harmonization, and public release of 5,320,379 non-cancelled apartment sale records (2015–2024) plus separately released auxiliary tables that preserve original spatial/temporal resolutions. There is no theoretical derivation, no fitted parameter that is later called a prediction, no uniqueness theorem, and no ansatz smuggled in via self-citation. Technical Validation (Section 4) consists of file completeness, schema consistency, coverage counts, descriptive statistics (means, quartiles, ranges), and metadata consistency checks; these are integrity diagnostics, not forecasts forced by construction. The only self-reference is the Zenodo deposit of the same dataset [32] and the authors’ validation code repository—standard for data descriptors and not load-bearing for any claimed result. Mapped CSI and private-education variables are explicitly flagged by the authors as contextual proxies (Methods 2.3.2, 2.3.5; Usage Notes), not as independently surveyed local values; that is a transparency note, not circularity. Consequently the derivation chain is empty of circular steps and the circularity score is 0.

Axiom & Free-Parameter Ledger

0 free parameters · 4 axioms · 1 invented entities

As a data-descriptor paper, KRAFT rests on standard assumptions of official statistical systems rather than free parameters or invented physical entities. The load-bearing choices are the decision to retain source quirks, the documented region-label crosswalk, and the spatial expansion rules for CSI and private-education series. No numerical constants are fitted to produce a scientific claim.

axioms (4)
  • domain assumption Official MOLIT Actual Transaction Price Disclosure System records, after removal of canceled and exact-duplicate rows, constitute a usable census of reported apartment sales for 2015–2024.
    Stated in Methods §2.2; the entire transaction count and coverage claims rest on this source fidelity.
  • domain assumption Documented administrative-region crosswalk (Incheon Nam-gu/Michuhol-gu, Hwaseong-si, Gunwi-gun transfer, Sejong labeling) correctly aligns multi-source labels for longitudinal joins.
    Methods §2.4 and Region_Mapping.csv; incorrect crosswalk would break multi-year regional panels.
  • ad hoc to paper CSI regional groups (Seoul / six metros / other) and private-education region-type categories can be expanded to Sido/Sigungu as contextual indicators without claiming independent local surveys.
    Methods §§2.3.2, 2.3.5 and Usage Notes; the paper itself flags these as mapped proxies.
  • domain assumption Year-end YearMonth keys are an acceptable indexing convention for annual indicators when users are warned not to treat them as December observations.
    Usage Notes; standard practice but still a design choice that affects monthly panel construction.
invented entities (1)
  • KRAFT dataset package independent evidence
    purpose: Unified, English-named, merge-keyed release of Korean apartment transactions plus multi-domain contextual indicators.
    The sole new object introduced; it is a data product, not a theoretical entity. Independent evidence is the public Zenodo deposit and validation outputs.

pith-pipeline@v1.1.0-grok45 · 23487 in / 2900 out tokens · 23778 ms · 2026-07-15T09:10:46.484447+00:00 · methodology

0 comments
read the original abstract

Apartment transaction records are useful for studying housing markets, household finance, regional economics, and macro-financial transmission, but transaction data are often distributed separately from contextual socioeconomic indicators. We present KRAFT, a nationwide transaction-level dataset of South Korean apartment sales from January 2015 to December 2024. The dataset contains 5,320,379 apartment sale transactions across all 17 Sido regions and includes transaction timing, administrative location, exclusive residential area, reported transaction price, floor level, and construction year. KRAFT also provides auxiliary indicators covering macro-financial conditions, demographic structure, education infrastructure, private education expenditure, housing price indices, consumer sentiment, and economic policy uncertainty. The released files are organized as year-specific transaction files and separate auxiliary data tables to preserve the original temporal and spatial resolution of each source. KRAFT supports reproducible research on apartment price modeling, regional housing-market comparison, housing-demand analysis, and links between housing transactions and socioeconomic context.

Figures

Figures reproduced from arXiv: 2607.11961 by Hyungjoon Kim, Sejin Myung.

Figure 1
Figure 1. Figure 1: Overview of the KRAFT dataset construction workflow [PITH_FULL_IMAGE:figures/full_fig_p004_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: Annual number of apartment transactions from 2015 to 2024 [PITH_FULL_IMAGE:figures/full_fig_p016_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: Number of apartment transactions by Sido [PITH_FULL_IMAGE:figures/full_fig_p017_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: Distribution of validation-purpose transaction price per square meter Together, the annual, regional, Sido-month, and price-distribution checks support the completeness and plausibility of the KRAFT transaction records. The generated figures and count tables are provided together with the validation code to support reproducibility [PITH_FULL_IMAGE:figures/full_fig_p017_4.png] view at source ↗

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

31 extracted references · 16 canonical work pages

  1. [1]

    2024 Population and Housing Census (Register-based Census)

    Ministry of Data and Statistics. 2024 Population and Housing Census (Register-based Census). Press release, 29 July 2025. Ministry of Data and Statistics, Republic of Korea. https://mods.go.kr/board.es?act=view&bid=11739&list_no=439063&mid=a20107020000 (ac cessed 5 July 2026)

    2024

  2. [2]

    Case, K. E. & Shiller, R. J. The efficiency of the market for single -family homes. American Economic Review 79(1), 125–137 (1989)

    1989

  3. [3]

    Mankiw, N. G. & Weil, D. N. The baby boom, the baby bust, and the housing market. Regional Science and Urban Economics 19(2), 235–258 (1989). https://doi.org/10.1016/0166-0462(89)90005-7

    work page doi:10.1016/0166-0462(89)90005-7 1989

  4. [4]

    & Sinai, T

    Himmelberg, C., Mayer, C. & Sinai, T. Assessing high house prices: Bubbles, fundamentals and misperceptions. Journal of Economic Perspectives 19(4), 67 –92 (2005). https://doi.org/10.1257/089533005775196769

    work page doi:10.1257/089533005775196769 2005

  5. [5]

    Actual Transaction Price Disclosure System

    Ministry of Land, Infrastructure and Transport. Actual Transaction Price Disclosure System. Ministry of Land, Infrastructure and Transport, Republic of Korea. https://rt.molit.go.kr/ (accessed 5 July 2026)

    2026

  6. [6]

    Economic Statistics System (ECOS): exchange rates, interest rates, and monetary and liquidity aggregates

    Bank of Korea. Economic Statistics System (ECOS): exchange rates, interest rates, and monetary and liquidity aggregates. Bank of Korea. https://ecos.bok.or.kr/ (accessed 5 July 2026)

    2026

  7. [7]

    Consumer Price Survey: Consumer Price Index

    Ministry of Data and Statistics. Consumer Price Survey: Consumer Price Index. Ministry of Data and Statistics, Republic of Korea. https://mods.go.kr/menu.es?mid=a20209010000 (accessed 5 July 2026)

    2026

  8. [8]

    Number of internal migrants for city, county, and district (M,Q,Y 1970.01 – 2026.05)

    Korean Statistical Information Service. KOSIS Statistical Database: population, households, and internal migration statistics, including “Number of internal migrants for city, county, and district (M,Q,Y 1970.01 – 2026.05)” and ‘Netmigration(Administrative reports) (Person)’. Ministry of Data and Statistics, Republic of Korea. https://kosis.kr/eng/ (acces...

    1970

  9. [9]

    Private Education Expenditures Survey of Elementary, Middle and High School Students

    Ministry of Data and Statistics. Private Education Expenditures Survey of Elementary, Middle and High School Students. Ministry of Data and Statistics, Republic of Korea. https://mods.go.kr/menu.es?mid=a20111020000 (accessed 5 July 2026)

    2026

  10. [10]

    National Survey of House Price Trends: housing price trend statistics and apartment housing price index

    Korea Real Estate Board. National Survey of House Price Trends: housing price trend statistics and apartment housing price index. Korea Real Estate Board. https://www.reb.or.kr/rebEng/cm/cntnts/cntntsView.do?cntntsId=1372&mi=10057 (accessed 5 July 2026)

    2026

  11. [11]

    School Information Disclosure and public disclosure data services

    Korea Education and Research Information Service. School Information Disclosure and public disclosure data services. Korea Education and Research Information Service. https://www.keris.or.kr/eng/cm/cntnts/cntntsView.do?cntntsId=1332&mi=1381 (accessed 5 July 2026)

    2026

  12. [12]

    Consumer Survey Index (CSI)

    Bank of Korea. Consumer Survey Index (CSI). Economic Statistics System (ECOS), Bank of Korea. https://ecos.bok.or.kr/ (accessed 5 July 2026)

    2026

  13. [13]

    Uncertainty Indices for South Korea

    Economic Policy Uncertainty. Uncertainty Indices for South Korea. https://www.policyuncertainty.com/korea_monthly.html (accessed 5 July 2026)

    2026

  14. [14]

    R., Bloom, N

    Baker, S. R., Bloom, N. & Davis, S. J. Measuring economic policy uncertainty. Quarterly Journal of Economics 131(4), 1593–1636 (2016). https://doi.org/10.1093/qje/qjw024

    doi:10.1093/qje/qjw024 2016

  15. [16]

    M., Garcia, N

    Pires, I. M., Garcia, N. M., Zdravevski, E. et al. Daily motionless activities: A dataset with accelerometer, magnetometer, gyroscope, environment, and GPS data. Scientific Data 9, 105 (2022). doi:10.1038/s41597 - 022-01213-9

    doi:10.1038/s41597 2022

  16. [17]

    Tasgaonkar, P., Zade, D., Ehsan, S. et al. Indoor heat measurement data from low-income households in rural and urban South Asia. Scientific Data 9, 285 (2022). doi:10.1038/s41597-022-01314-5

    work page doi:10.1038/s41597-022-01314-5 2022

  17. [18]

    Hettiarachchi, H., Dridi, A., Gaber, M. M. et al. CODE-ACCORD: A Corpus of building regulatory data for rule generation towards automatic compliance checking. Scientific Data 12, 170 (2025). doi:10.1038/s41597- 024-04320-x

    doi:10.1038/s41597- 2025

  18. [19]

    & Jang, H

    Song, Y ., Ahn, K., An, S. & Jang, H. Hedonic dataset of the metropolitan housing market – Cases in South Korea. Data in Brief 35, 106877 (2021). doi:10.1016/j.dib.2021.106877

    work page doi:10.1016/j.dib.2021.106877 2021

  19. [20]

    Hedonic prices and implicit markets: Product differentiation in pure competition

    Rosen, S. Hedonic prices and implicit markets: Product differentiation in pure competition. Journal of Political Economy 82, 34–55 (1974). doi:10.1086/260169

    doi:10.1086/260169 1974

  20. [21]

    Black, S. E. Do better schools matter? Parental valuation of elementary education. The Quarterly Journal of Economics 114, 577–599 (1999). doi:10.1162/003355399556070

    work page doi:10.1162/003355399556070 1999

  21. [22]

    & McMillan, R

    Bayer, P., Ferreira, F. & McMillan, R. A unified framework for measuring preferences for schools and neighborhoods. Journal of Political Economy 115, 588–638 (2007). doi:10.1086/522381

    work page doi:10.1086/522381 2007

  22. [23]

    House prices, borrowing constraints, and monetary policy in the business cycle

    Iacoviello, M. House prices, borrowing constraints, and monetary policy in the business cycle. American Economic Review 95, 739–764 (2005). doi:10.1257/0002828054201477

    work page doi:10.1257/0002828054201477 2005

  23. [24]

    E., Quigley, J

    Case, K. E., Quigley, J. M. & Shiller, R. J. Comparing wealth effects: The stock market versus the housing market. The B.E. Journal of Macroeconomics 5, Article 1 (2005). doi:10.2202/1534-6013.1235

    work page doi:10.2202/1534-6013.1235 2005

  24. [25]

    & Sufi, A

    Mian, A., Rao, K. & Sufi, A. Household balance sheets, consumption, and the economic slump. The Quarterly Journal of Economics 128, 1687–1726 (2013). doi:10.1093/qje/qjt020

    work page doi:10.1093/qje/qjt020 2013

  25. [26]

    The geographic determinants of housing supply

    Saiz, A. The geographic determinants of housing supply. The Quarterly Journal of Economics 125, 1253 – 1296 (2010). doi:10.1162/qjec.2010.125.3.1253

    work page doi:10.1162/qjec.2010.125.3.1253 2010

  26. [27]

    & Sinai, T

    Gyourko, J., Mayer, C. & Sinai, T. Superstar cities. American Economic Journal: Economic Policy 5, 167 – 199 (2013). doi:10.1257/pol.5.4.167

    work page doi:10.1257/pol.5.4.167 2013

  27. [28]

    L., Gyourko, J

    Glaeser, E. L., Gyourko, J. & Saks, R. E. Why is Manhattan so expensive? Regulation and the rise in housing prices. Journal of Law and Economics 48, 331–369 (2005). doi:10.1086/429979

    work page doi:10.1086/429979 2005

  28. [29]

    U., Park, S

    Sarwar, A., Bukhari, F., Sattar, A., Bukhari, M., Rehman, Z. U., Park, S. & Rho, S. Macroeconomic context- aware graph topology learning for stock price forecasting using graph neural network. ICT Express 12(1), 13–19 (2026). doi:10.1016/j.icte.2025.12.002

    work page doi:10.1016/j.icte.2025.12.002 2026

  29. [30]

    U., Park, S

    Bukhari, F., Sarwar, A., Sattar, A., Rehman, Z. U., Park, S. & Rho, S. Regime -aware static and dynamic graph structure learning for stock price forecasting. Humanities and Social Sciences Communications (2026). doi:10.1057/s41599-026-07653-7

    work page doi:10.1057/s41599-026-07653-7 2026

  30. [31]

    & Rho, S

    Ali, Z., Ansari, Y ., Bukhari, M., Maqsood, M., Park, S. & Rho, S. CMGM: A novel cross-market assets and multi-market modeling graph neural networks for financial market forecasting leveraging market states dependencies. Alexandria Engineering Journal 128, 1101–1124 (2025). doi:10.1016/j.aej.2025.08.024

    work page doi:10.1016/j.aej.2025.08.024 2025

  31. [32]

    & Kim, H

    Myung, S. & Kim, H. J. KRAFT: Korean Residential Apartment Fundamentals and Transactions. Version 1.0. Zenodo. https://doi.org/10.5281/zenodo.21320079 (2026)

    work page doi:10.5281/zenodo.21320079 2026