[The Pitfalls of Statistics]② Lost Trust in 'House Price Statistics'... "Fluctuates with Each Administration"

Korea Real Estate Board 'Weekly Apartment Price' Statistics
Prices Reported Lower When Soaring, Higher When Falling
Experts Say "Inevitable to Be Swayed by Government Influence"
Calls to Abolish Weekly Statistics and Increase Reflection of Actual Transaction Prices

Trust in the weekly apartment price statistics released by the Korea Real Estate Board is steadily declining. During the Moon Jae-in administration, when housing prices surged, the statistics showed a smaller increase than the perceived price rise, sparking controversy over 'statistical manipulation.' Recently, during the housing price decline, the statistics have rebounded faster than private statistics, showing an 'outlier' pattern. Amid conflicting forecasts about the recent direction of housing prices, the loss of trust in the statistics is further increasing market confusion. Real estate experts emphasize the need to reform the weekly statistics and enhance the independence of the institution, as the Korea Real Estate Board, a government-affiliated agency, has lost credibility.

Government Statistics Show Rising Housing Prices, While Private Statistics Indicate a Decline

According to the Korea Real Estate Board's statistics on the 4th, apartment prices in Seoul have shown an upward trend for six consecutive weeks. Based on weekly statistics, the increases were 0.03% on May 22, 0.04% on May 29, 0.04% on June 5, 0.03% on June 12, 0.04% on June 19, and 0.04% on June 26. Although the cumulative increase of 0.22% during this period is not large, the market perceives this as a shift from a prolonged slump to a rising trend in Seoul apartment prices, raising expectations for a price rebound and causing homeowners to gradually raise their asking prices.

In contrast, the weekly survey by KB Real Estate, a representative private statistic, still shows a decline in Seoul apartment prices. During the same period, KB Real Estate's price change rates were -0.109%, -0.040%, -0.055%, -0.082%, -0.070%, and -0.054%. Compared to the Korea Real Estate Board's statistics, not only the direction but also the trend does not match. As a result, both the general public and experts describe the situation as 'confusing.' A professor of economics in Seoul said, "There is a significant difference between the recent statistics of the Korea Real Estate Board and KB Real Estate. I am very concerned about which data to use for analysis."

The fact that housing price trends appear higher in the Korea Real Estate Board's data is the exact opposite of the situation during the Moon Jae-in administration when prices surged. During the 'boom' years of 2020-2021, the Seoul apartment price increase rate (based on weekly surveys) was 31.319% according to KB Real Estate, while the Korea Real Estate Board reported only 7.5%. Although both institutions produce statistics through their own surveys, the market often points out that the Korea Real Estate Board's statistics differ from the perceived housing prices. In July 2020, then Minister of Land, Infrastructure and Transport Kim Hyun-mi stated that Seoul housing prices had risen 11.28% over three years since the Moon administration began, but civic groups cited KB statistics to counter that prices had increased by 52%, sparking controversy over alleged 'statistical manipulation' by the government.

Statistics Fluctuate According to Government Needs?..."Lost Trust"

Many point out that the Korea Real Estate Board's statistics showing discrepancies with perceived housing prices during both rising and falling markets stem from its status as an agency under the Ministry of Land, Infrastructure and Transport. When housing prices surge, there is a need to lower statistics to mitigate public criticism of the government's real estate policies, while during price declines, there is an incentive to adjust statistics upward to alleviate concerns about the economy and reverse jeonse (key money deposit) issues. Professor Han Moon-do of Seoul Cyber University’s Department of Real Estate said, "The Korea Real Estate Board's statistics inevitably fall under political influence. This was true under the previous administration and remains so now."

So where does the credibility issue of the Korea Real Estate Board's statistics originate? It can be found in the method of statistical calculation. According to the survey method posted on their website, if there is an actual transaction case for a sample house, the Korea Real Estate Board reviews it and calculates a 'sample price.' If there is no transaction, they use similar transaction cases within the same complex to estimate the 'sample price.' In other words, they calculate a sample price regardless of whether there is an actual transaction. In contrast, KB Real Estate uses the actual transaction price if available; if not, they derive the sample price through surveys of local real estate agents.

Subjectivity in the Survey Process...Limitations of Sample Surveys

The biggest drawback of the Korea Real Estate Board's 'sample survey' is the potential for subjectivity in the survey process. Professor Han Moon-do said, "The Korea Real Estate Board's weekly statistics include excessive qualitative factors, but KB Real Estate fills in qualitative factors more with asking prices rather than personal opinions of surveyors, making it more accurate." He added, "Financial institutions use KB Real Estate prices for mortgage loans, so their statistics align more closely with actual prices. On the other hand, the Korea Real Estate Board, being a government agency, has limitations."

Private statistics like KB Real Estate have fewer conflicts of interest, so the possibility of intentional bias in sample surveys is low, but government statistics are different. Especially, while KB Real Estate mainly reflects real estate agents' asking prices for sample prices without actual transactions, the Korea Real Estate Board has professional surveyors input 'transactionable prices' for sample houses, allowing more subjective 'qualitative factors' to intervene. Generally, in the 'sample survey' method, which selects representative apartment complexes to investigate housing prices, statistics tend to show weaker trends than actual price movements, but the Korea Real Estate Board's housing price statistics are criticized for exceeding this range.

Professor Lim Jae-man of Sejong University's Department of Real Estate said, "Even if you create a price index only from actual transactions, it does not represent the entire market trend. How can anyone know the exact price when sample houses number in the tens of thousands and transactions are not occurring? Surveyors have strong incentives to lower figures when prices are high and raise them when prices are low." The Korea Real Estate Board increased the sample size to 32,900 apartments for weekly surveys earlier this year amid repeated credibility controversies, but this is still only about half of KB Real Estate's 62,220 units.

Abolish Weekly Statistics... "No Other Country Like Ours"

Some voices call for changing the real estate statistical survey method to be based on actual transaction prices. The aim is to eliminate qualitative factors and increase the reflection of actual housing prices in the statistics. Of course, using actual transaction prices can cause large fluctuations and distortions due to false transaction reports intended to manipulate prices. Therefore, experts explain that actual transaction statistics should exclude canceled contracts and abnormal transactions with excessively high or low prices. To do this, weekly statistics, which lack sufficient time to filter out false transactions, should be abolished.

In fact, many major countries such as the United States, the United Kingdom, Germany, Japan, and France create and publish housing price indices based on actual transaction prices, and rarely release weekly data like South Korea. Professor Lim Jae-man said, "Statistics should be made using actual transaction data, but many transactions are reported for price manipulation without registration. Statistics should be based only on registered transactions." He added, "Such statistics inevitably come out after some time has passed. I have never seen a country that publishes housing price levels on a weekly basis like ours."

The Korea Real Estate Board's weekly real estate statistics problems are not new. The Statistics Korea pointed out in its regular statistical quality diagnosis report released at the end of 2020 that "it is difficult to survey apartment prices (sale and jeonse) on a weekly basis," and "although reference materials such as actual transaction prices are used to survey apartment prices, there are difficulties in using them, and the sample size for weekly apartment price surveys is smaller than monthly surveys, which can cause errors."

Supplementing with Private Statistics When Needed for Loans or Policies

Weekly statistics may be necessary for mortgage loan operations or understanding real estate market trends. However, these can be sufficiently supplemented by private statistics such as KB Real Estate. Professor Lim Jae-man said, "There are calls in the U.S. to create daily statistics to quickly grasp market trends, but the government does not need to produce short-term statistics. Even if the government internally monitors weekly trends to respond quickly to the real estate market, it should not release inaccurate statistics."

Researcher Lee Eun-hyung of the Korea Institute of Construction Policy said, "KB Kookmin Bank needs weekly statistics for loan operations, but the Korea Real Estate Board does not necessarily have to release them. The current weekly fluctuation of around 0.01% is not very meaningful." He added, "If the government takes the lead in statistics and real estate policies, public attention will focus excessively on real estate, causing side effects."