[Opinion] Revisiting Real Estate Statistics <1> The Urgent Need to Overhaul Apartment Price Indices

Professor Changmoo Lee, Department of Urban Engineering, Hanyang University

Controversies surrounding the apartment price index managed by the Korea Real Estate Board have persisted continuously since the Moon Jae-in administration. Why have these issues become particularly pronounced only under the current government? To understand the reality, if we look at the apartment price increase rates in Seoul during the Moon administration, the monthly housing price trend survey by the Korea Real Estate Board shows 21%, KB Kookmin Bank shows 50%, and Real Estate 114 shows 86%, with the Real Estate 114 index exhibiting a rise most similar to the Ministry of Land, Infrastructure and Transport’s actual transaction price index at 81%. The gap in cumulative increase rates over about four years is too large to simply attribute to the characteristics of each index.

Relatively smaller cumulative change rates are seen in the housing price trend survey and KB Kookmin Bank indices, which are sample-based indices targeting a portion of the population, whereas the Real Estate 114 index quickly reflects changes in housing stock, and the actual transaction price index incorporates as many observed transactions as possible. If the Kookmin Bank or Korea Real Estate Board indices were calculated including all price survey targets rather than samples, they would likely show cumulative increases similar to the actual transaction price index. This would be direct evidence that the underestimation in government indices stems from sample selection and changes.

But why do the problems of sample-based indices become so stark only under this administration? The reason is that market price fluctuations during this government have been extremely differentiated due to policy choices involving pinpoint regulations on submarkets. For example, if in the first year two houses rose by 10% (sample) and 20% (non-sample) respectively, and in the second year the sample was changed to a rapidly rising apartment, but that apartment only rose 5% (sample) due to strengthened regulations while the other apartment rose 15% (non-sample), the average increase for the entire stock would be 26.5%, but the cumulative increase for the sample would be underestimated at 15.5%. In markets where price changes vary unstably across submarkets, it is impossible to rationally select samples in advance. The best method is to move away from sample-based approaches and use stock-based market prices.

This leads to considerations about the method of price surveys. In fact, apartment prices are prices that are difficult to mismeasure over the long term. While short-term deviations may occur depending on the survey entity, prices inevitably converge in the medium to long term. Moreover, professional surveyors at the Korea Real Estate Board do not grasp price changes better than local real estate agents specializing in the relevant complexes who obtain actual transaction information earlier. Since market prices represent a range of prices for many similar houses, selecting and assigning specific complexes or units does not yield more accurate measurements than the average price changes of a certain size within that complex.

Accepting these facts, rather than struggling to maintain samples, it is judged best within the limits of a price index based on market prices that immediately reflect stock changes and are surveyed from real estate agents. This approach reduces costs, diminishes supervisory influence, and improves the representativeness of the index. One concern is that if serious underestimation caused by sample-based methods has been embedded in accumulated price indices, correcting past indices is essential. This would provide an opportunity to rectify government policy mistakes made by misjudging the intensity of housing market changes responding to medium- and long-term conditions.

Chang-Moo Lee, Professor, Department of Urban Engineering, Hanyang University