As anyone who regularly reads this blog knows, we have issues with the Japanese media’s coverage of real estate prices, since they almost exclusively cover Tokyo. For sure, Tokyo properties mean a lot in the scheme of things, but Tokyo real estate is exceptional in the sense that the city still charts growth while the rest of Japan doesn’t; or, at least, not at the exceptional rate that Tokyo does. Consequently, the huge increases in property values that Tokyo is now seeing don’t really represent Japan, something the media neglects to point out. What you get in the mainstream press when it comes to coverage of real estate is skyrocketing condo prices in Tokyo and akiya—vacant properties—everywhere else, and nothing much in between. So we were intrigued by a recent series in the Asahi Shimbun about how the government assesses property values, since they do it all over Japan. Though the articles don’t explain anything about national trends in property values, they imply that one of the reasons for the lopsided coverage mentioned above is that it’s difficult to trust any related statistics released by the government. 

The series was prefaced by an article about an announcement from the municipal government of Chofu, Tokyo, that its assessments for property taxes and so-called city planning taxes were incorrect in 166 cases; specifically, for 88 properties the assessments overestimated their value while for 78 the assessments underestimated their value. The amount of refunds due for the overestimates comes to ¥80.95 million, while the additional imposition of taxes for the underestimated assessments add up to ¥52.61 million.

The mistakes mostly had to do with how land was categorized. For instance, property taxes for land categorized as residential can be reduced by five-sixths if it contains a structure of some kind. Other variables include additional structure, additions to the main structure, and demolition work, all of which would require that the land be reassessed. Apparently, Chofu has yet to determine exactly how they got it wrong, but in any case, according to the law if an assessment is found to be too low, the municipality can demand compensation no more than five years after the original assessment. But if an assessment is too big, then there is no time limit for compensating the property owner.

Obviously, land value assessment is a tricky business that’s open to a lot of subjective factors, so in its Keizai Plus section, Asahi followed up the Chofu item by looking into the broader “mystery” surrounding the land ministry’s public real estate assessments, which are carried out every year for 26,000 locations throughout the country. Two appraisers are assigned to each location, with each one making their own separate assessment, and if there are discrepancies, the ministry mediates. The “correct” valuations are then published on the ministry’s home page. 

It sounds simple enough, but when Asahi checked the original valuations of the paired assessors using AI, they found that in 69 percent of the locations assessed, the two appraisals matched exactly, and in the remaining cases the difference was “within 1 percent.” The location with the highest assessment for residential land is Akasaka in Tokyo, and the assessors’ valuations have matched exactly for the last five years. The prefecture with the highest matching rate was Tokushima at 97 percent. Asahi’s initial reaction was that if the valuations are so predictable, then why do they cost so much? The ministry spends ¥4 billion a year on assessments.

But the question goes deeper. When Asahi asked a private assessor who works for real estate companies they were told that “under normal circumstances” with two assessors valuating the same property, you’ll get a difference of about 10 percent, but “if you follow the ministry manual” for assessment, the valuations will almost always match.

Asahi then acquired the ministry’s 271-page assessment manual, which explains the process. At the end of every June, the land appraisal committee assigns 2,200 appraisers from 167 ministry branches to the work of valuating land, with each appraiser being assigned to 20 locations. Each location is appraised by two people, who are charged with “projecting the value” of the land in question on Jan. 1 of the following year. In October their valuations are sent to the relevant section in the land ministry, where they are discussed three times using past cases as references. The final valuations are then approved and a report is issued.

When Asahi asked the ministry for documents related to  projecting the values of land for the last two years, they were delivered heavily redacted. The ministry explained that there’s a “danger” that publicly releasing such information could “influence subsequent valuations.” When Asahi showed the documents to an outside appraiser who once did appraisal work for the ministry on a contract basis, the outside appraiser said that the two appraisers assigned to valuate the same land always looked at past cases so that their valuations “made sense” when they wrote up the final report, which is why they usually match. A different outside appraiser said that whenever the two valuations did not match, the ministry automatically investigates the difference, since public valuations directly affect property taxes and therefore “shouldn’t be different.” This appraiser said the ministry wanted the reports to reflect a “beautiful system.” 

The land ministry distributes questionnaires to anyone who bought property in a given year in order to understand “the full scope of land transactions,” but usually only 10-20 percent of the questionnaires are returned. Before visiting the land that is to be appraised, the two appraisers refer to data supplied by these questionnaires over the years and use them to come up with valuations. As one outside appraiser describes it, the appraisers base their valuation on their “imagination, which is then adjusted by data.” The difference between the ministry appraisers and industry appraisers is that the latter “wants to reflect the newest transactions as much as possible,” while the ministry appraisers mostly goes by a gut feeling influenced by the ministry itself.

Asahi wondered about the “fairness” of a system whose methodology is a “mere formality,” meaning that it would seem that the valuations are decided beforehand based on past valuations and how the ministry thinks they should change. When asked about this, the ministry simply admitted that there are cases where they ask the appraisers to explain their differences in valuations, “but we don’t demand that they be identical.” 

Real estate people told Asahi that they never refer to the ministry’s valuations when buying and selling properties because they don’t reflect “reality.” 

This gap in perception actually explains a lot about the media’s coverage of both overheated Tokyo real estate prices and the akiya problem. One appraiser told Asahi that the publicly listed values of properties in central Tokyo are only about half their actual value. When public land in the capital is sold, it’s much cheaper because the ministry valuation is used to decide the selling price. Buyers benefit greatly. 

On the other hand, ministry valuations of land in rural areas tend to be higher than their actual value, so in both cases the public assessments are not of much use to real estate companies. As the appraiser sees it, the land ministry’s assessment is an “empty gesture” to come up with circumstances that fit a desired outcome. The purpose of that outcome is to control property taxes, which are set by the central government even if they’re administered by local governments. In other words, keep a lid on property taxes in urban areas where property values are very high, while increase property taxes in rural areas where property values are often worth hardly anything, the idea being that there should be at least some semblance of balance nationwide in terms of how municipalities collect revenue to run their operations. 

This dynamic shouldn’t be surprising and, in a way, makes sense. The real value of a property, in theory, is what the owner can get for it if they sell it. As an extreme example, when it comes to akiya and land that has been vacant for a long time, the value is effectively zero if the owner has tried to sell the property but can’t. However, such valuations negatively affect the bottom lines of municipalities where there are many akiya, so some public value has to be attached to those properties. Whether the owner of such a property actually pays their property tax is a whole other issue.