Article 43 Asks for More Than “2m of Road Frontage”

Article 43(1) of the Building Standards Act states that “the site of a building must adjoin a road for a length of two metres or more” (Act No. 201 of 1950, checked on e-Gov Law Search on September 9, 2026). A road here is as defined in Article 42, and in principle must be at least 4m wide. A way narrower than 4m is still deemed a road under Article 42(2) if buildings already lined it when the city-planning area was designated and the designated administrative agency has designated it; the line 2m horizontally from its centreline is then deemed the road boundary. These are the so-called “Paragraph 2 roads” and the setback that comes with them.

The problem is that many machiya and row houses deep inside alleys fit neither the basic rule nor the Paragraph 2 road provision. That is where Article 43(2) comes in. Article 43(2) contains two routes of an entirely different nature, and their practical timelines differ widely.

The permit side under Article 43(2)(ii) is somewhat more detailed. Kyoto City’s permit criteria define a “passage” as a way, other than a road, that existed on May 1, 1999, and call a site that reaches a road only through a flagpole-shaped strip at least 1.5m but less than 2m wide a “dedicated passage”. The criteria then include a provision permitting major repairs and remodelling of non-conforming existing machiya and similar buildings on sites fronting dead-end passages at least 1.8m wide that were lined with buildings on the reference date. Up to this point, the design supports restoring machiya deep in the alleys.

However, the uses that provision allows are limited to dedicated residences, combined residences, or specified-use buildings. A specified-use building means one with the combined-residence uses listed in Article 130-3 of the Building Standards Act Enforcement Order (offices, daily-goods shops, restaurants and cafés, barbershops, studios, etc.), where the total floor area of those parts is 50㎡ or less (75㎡ or less where the mayor finds the disaster-prevention burden to be small), the expected number of users is up to about 10 people, and a management system for evacuation guidance and fire-suppression measures is in place. Furthermore, where the distance from the road or specified passage to the far end of the passage is 35m or more, only dedicated or combined residences are allowed.

Lodging facilities operating under the Inns and Hotels Act are not among the uses listed in Article 130-3 of the Enforcement Order. It therefore reads that the path of “doing major repairs on an alley machiya under an Article 43(2) permit and then openly running it as an inn” is not provided for head-on in the text of Kyoto City’s criteria. That said, this does not decide whether any individual property can proceed. The criteria themselves state that buildings not covered by the criteria may still be certified or permitted after comprehensive consideration of use, scale, surroundings, urban-planning policy and other factors, leaving the decision to the designated administrative agency. In practice, every case must start with a prior consultation.

The Machiya That Remain Are the Ones in the Alleys — A Reversed Loss Rate

The FY2024 Kyoto Machiya Survey, conducted jointly by Kyoto City, the Kyoto Center for Community Collaboration and Ritsumeikan University, tracked the 40,146 machiya identified in the FY2016 survey. Over eight years, 5,566 were lost and 34,580 remained. The average annual loss rate was 1.73%, slightly up from 1.68% in the previous survey.

The breakdown by frontage road is what stands out. Over eight years, 16.2% of machiya on roads 4m or wider were lost, 13.2% of those on roads under 4m, and 10.2% of those on non-roads, evacuation passages or specified passages. The wider the road, the more were lost; the narrower the road, the more remain.

The reason is not hard to see. Rebuilding, or converting to apartments, is physically and legally feasible, in principle, only on sites that meet the frontage requirement. Of the 5,566 machiya lost, the land after demolition became vacant lots (33.5%), parking (5.2%), dedicated residences (30.4%), apartment buildings (5.3%) and lodging facilities (3.3%, 184 houses). Redevelopment pressure concentrates where frontage works, and the far ends of the alleys sit outside that pressure. The alleys have worked as a preservation device — and at the same time have accumulated a stock whose exit is closed.

Site size matters too. Of the remaining machiya, 29.6% sit on sites of 15 tsubo or less and 49.2% on 15–30 tsubo, so 78.8% are on sites of 30 tsubo (about 99㎡) or less. Above 30 tsubo, the loss rate exceeds the overall average of 13.9%, reaching 18.8% above 105 tsubo. Larger sites are being redeveloped first. The machiya that come to market are therefore more likely to be small, on narrow roads, and vacant. Incidentally, machiya judged vacant in the previous survey had a loss rate of 27.0% — more than double the 11.5% for occupied ones. Timing an acquisition is itself a way of preserving the asset.

The Appraisal Benchmarks Are Not Standing in the Alleys

The fairness of an acquisition price is normally tested against nearby transaction comparables and official valuations. When frontage is the issue, however, that procedure has a structural blind spot. Tallying the 283 Land Price Publication and Prefectural Land Price Survey points in Kyoto City (2025 and 2026 releases, with frontage road width recorded) from the MLIT Real Estate Information Library, only 6 benchmark sites — just 2.1% — face roads under 4m wide. The median is 6.2m, and 31.1% of sites face wide roads of 12m or more. In Kanazawa City, too, only 5 of 95 sites (5.3%) face roads under 4m, with a median of 7.4m.

Because benchmark sites are chosen to represent typical land in an area, fronting a legal road is almost a given. On the side of the road where 62.6% of Kyoto machiya stand, there are almost no comparables. Describing an alley property as “X% below the nearby price per ㎡” risks subtracting from a baseline that actually belongs to a different asset class.

83.6% of Machiya Lodgings Have One Room — The 200㎡ Line Shapes the Product

Within the coverage of MetroEngines Research & Consulting, there are 1,363 machiya-category lodging facilities with 2,328 rooms in total within Kyoto City (latitude 34.87–35.10°N, longitude 135.65–135.87°E). Of the 1,359 properties with a known room count, 1,136 (83.6%) have one room and 1,251 (92.1%) have three rooms or fewer, averaging 1.71 rooms. Of the 1,558 machiya-category properties nationwide, 87.5% are concentrated in Kyoto City.

This extreme skew toward tiny properties reads as a consequence of the legal framework rather than design preference. Article 6(1)(i) of the Building Standards Act requires a building confirmation application for special buildings listed in column (i) of Appended Table 1 where the total floor area used for that purpose exceeds 200㎡, and Article 87(1) applies that provision mutatis mutandis when a building’s use is changed so that it becomes such a special building. Put the other way round, a change of use of 200㎡ or less needs no building confirmation (provided it involves no major repairs or remodelling). Kyoto City’s page for businesses planning to open lodging facilities likewise explains that no building confirmation procedure is needed where the area concerned is 200㎡ or less and no major repairs or remodelling are carried out.

And 78.8% of remaining machiya sit on sites of 30 tsubo (about 99㎡) or less. A typical Kyoto machiya does not exceed the 200㎡ line as it stands. The rational design answer for running one as lodging while avoiding a head-on frontage determination takes the form of a small, one-building, one-room product. Of the 1,494 lodging facilities confirmed as opening in 2026, the machiya category accounted for 44 properties with 89 rooms (2.0 rooms on average); 34 of the 44 had one room, and 34 (77.3%) were in Kyoto Prefecture. (Opening counts are based on OTA listings; because listings start several months before opening, figures for recent and future months may rise as more listings appear.) The same year, guesthouses (84) and vacation rentals (719) — small formats that tend to sit in alleys — made up the core of new supply.

In Kyoto City, the reading of those additional restrictions is stricter still. The Kyoto City Building Law Practice Handbook (April 2023) interprets a “road that is currently 4m or more wide” under Article 6 of the Kyoto City Building Standards Ordinance as a road that runs through at a width of 4m along the site’s frontage and the stretch until at least one end reaches another road 4m or wider. Even for a site on a Paragraph 2 road, if there is a single un-setback section before the road reaches another road, it is not a “road that is currently 4m or more wide”. Setting back only your own site is not enough — this is the kind of constraint you cannot resolve alone without the cooperation of neighbouring owners. The handbook also reads the ordinance’s 6m frontage rule as requiring 6m or more at a single point, not a combined length across two frontages.

Kanazawa Lifts Article 43(1) Itself by Ordinance

Blanket-consent criteria differ by municipality. Beyond that, the provision being relied on can itself be different.

Kanazawa City’s Ordinance on Relaxation of Building Standards Act Restrictions in the Kazuemachi Preservation District for Groups of Traditional Buildings (Ordinance No. 2 of March 23, 2015), based on Article 85-3 of the Building Standards Act, relaxes for buildings in the district Article 43(1) (the frontage requirement) as well as the main text of Article 44(1), Article 52, Article 53 and Article 56(1)(i). For traditional buildings, a permit under Article 4(1) of the preservation ordinance or a concluded consultation under Article 6 is required; for other buildings, conformity with the landscape-restoration standards is also required. Meanwhile, the enforcement rules for the Higashiyama Higashi Preservation District for Groups of Traditional Buildings (Rule No. 76 of June 27, 2001) deal with escape openings, wall positions and height, and do not cover Article 43. Even within preservation districts in the same city, the provisions relaxed differ by district.

Kyoto City’s route is certification or a permit under Article 43(2) — a case-by-case decision. Kanazawa’s Kazuemachi route is a general relaxation by ordinance under Article 85-3, which applies as a system to any property in the district that meets the requirements. That difference changes how due diligence is designed. The former requires building prior consultation and Review Board time into the schedule; for the latter, the issues are the boundaries of the designated district and whether the landscape standards can be met. Properties nationwide cannot be discussed in the general terms of “will the Article 43 proviso go through?”.

Three Categories to Sort Properties Into Before Acquisition, and Nine Points to Check

In practice, it is efficient to sort properties into three categories — “cannot be rebuilt”, “certification or permit required” and “compliant” — and vary the checks by category. The following does not determine whether any specific property can proceed; it organises the issues you can resolve yourself before taking a case to prior consultation.

Note that blanket-consent criteria for Article 43 permits and certifications differ in content from one municipality to another, so the same “deep in an alley” property is not treated uniformly. The Kyoto City and Kanazawa City criteria cited here are based on materials published as of writing (September 9, 2026), and they are revised frequently. Whether an individual property can be rebuilt must always be confirmed through prior consultation with the designated administrative agency.

Three Exit Scenarios — A 27.2% Discount Is the Tipping Point

We use a simplified model to test how the constraint should be reflected in price. The model is a whole-house machiya rental deep in an alley in central Kyoto (site about 83㎡, total floor area about 110㎡), compared against a same-size main-street property that can be rebuilt. It does not refer to any specific real property.

Side by side, the three scenarios reveal the structure. C improves the exit cap rate to 5.5% by securing frontage, but the ¥21M spent on the neighbouring lot inflates the capital invested, so its IRR falls below B’s. Securing frontage is not a move that always pays; whether it succeeds flips on the price of the neighbouring lot. Conversely, a property where the adjacent lot can be acquired on reasonable terms can be an exceptional opportunity even within the same alley.

The other line of reading is how large an acquisition discount is needed. Taking the compliant main-street property’s IRR of 8.31% as the benchmark and back-solving the acquisition price at which the alley property (major renovation, 7.5% exit) reaches the same IRR gives about ¥40.1M — 27.2% below ¥55M. With only a 20% discount the IRR is 7.43%, below the compliant property; only at a 36% discount does it clearly exceed it, at 9.50%.

Viewed on two axes, the discount alone does not close the argument. If the exit cap rate widens to 8.0% (that is, the constraint is weighed more heavily at sale), even a 27.2% discount does not reach the compliant property’s IRR. Conversely, if the exit cap rate can be held to 7.0%, a 20% discount is enough to match it. Establishing the outlook for certification or a permit through prior consultation, and settling the consent of passage right-holders before acquisition, does more for IRR — via the exit cap rate — than haggling over the purchase price. This two-axis relationship is what we mean when we say “the discount is the price of the constraint”.

The 27.2% figure naturally moves with the assumptions, but what it shows is robust. If the discount on an alley property is only around 20%, it is not enough to pay for the constraint. Much of the margin that feels like “buying cheap” at acquisition is demanded back at exit in the form of the cap rate. A property where prior consultation has established the certification or permit outlook, and where the most time-consuming issue — passage right-holders’ consent — has been settled, has room to narrow that cap-rate gap. That is where due diligence pays for itself. An IRR-based breakdown of whole-house rental economics by city is also covered in Simple Lodging × Vacant Home Reuse 2026: Kyoto, Osaka & Naha Whole-House IRR.

Machiya Lodging Listed Prices Beat Their Ward Averages — Demand Gathers on the Constrained Side

On the revenue side, this asset class actually scores well. For the 199 machiya lodgings in Kyoto City with at least six months of listings in the latest 12 months (September 2025 – August 2026), we took each property’s 12-month average listed price and the median by ward. The comparison is the ward-wide listed price (all property types, all-plan average) over the same period.

The “one room” of a whole-house rental is the price of an entire building and cannot be placed on a par with the per-room rate of multi-room properties. Even with that caveat, machiya lodging listed prices exceed the ward-wide level in every ward. And the wards with the densest concentration — Shimogyo (371 properties), Higashiyama (310) and Nakagyo (211) — are the very core of the pre-war urbanised area covered by the Kyoto Machiya Survey, overlapping the zone where machiya on roads under 4m and non-roads are concentrated.

By room-count band, the median listed price is ¥48,100 for one room (whole-house; N=115), ¥42,800 for 2–3 rooms (N=29), ¥42,200 for 4–9 rooms (N=37) and ¥37,400 for 10+ rooms (N=18). One-room properties rank highest because the price covers the whole building; this does not indicate superiority on a per-room basis. It does, however, confirm that the whole-house format is not capping prices. That the regulatory sweet spot — under 200㎡, one building, one room — is not a revenue handicap matters when considering the growth potential of this asset class.

Conclusion — The Discount Is Not “Cheapness” but “the Price of the Constraint”

To sum up. First, 62.6% of the 34,580 remaining Kyoto machiya face roads under 4m or non-roads, and this share has actually edged up from 61.6% in the previous survey. Behind it is the reversed loss rate: machiya on wider roads get redeveloped, while those on narrower roads remain. The machiya that come to market are structurally skewed toward the constrained side.

Second, official valuations offer almost no comparables for pricing that constraint. Of the 283 benchmark sites in Kyoto City, just 2.1% face roads under 4m, with a median of 6.2m. For this asset class, the usual procedure of discounting from the nearby price per ㎡ uses a reference point that sits somewhere else.

Third, certification and permits under Article 43(2) are different things, and even when granted they come with conditions on use and scale. The uses Kyoto City’s criteria envisage are dedicated residences, combined residences and specified-use buildings; lodging facilities are not envisaged head-on. What works in practice is a different route — under 200㎡ and short of major repairs — and the fact that 83.6% of machiya lodgings have one room bears this out. Some districts, like Kazuemachi in Kanazawa, relax Article 43(1) itself by ordinance, so municipal differences exist at the level of how the system is designed.

Fourth, in the simplified model, an alley property needed a 27.2% acquisition discount to reach the same IRR as a compliant main-street property. A discount of only 20% does not pay for the constraint. Conversely, a property where prior consultation has established the outlook and the time-consuming issue of passage right-holders’ consent has been settled can narrow the exit cap-rate gap. In this asset class, the upside lies not in price negotiation but in the process of resolving regulatory risk.

To repeat in closing: this analysis stays at the level of regulation and statistics and does not determine whether any individual property can be rebuilt. Blanket-consent criteria for Article 43 permits and certifications differ by municipality and are revised often, and passage rights cannot be fully read from the registry alone. At the stage of considering an acquisition, always hold a prior consultation with the designated administrative agency. That is the only practical way into the “deep in the alley” asset class.