When hotel development approvals are laid out on a timeline, most feasibility documents stop at “how many years from groundbreaking to completion.” What an investment committee actually wants to know at the land-assembly stage, however, is how much time stacks up before that point. Zoning, floor-area ratios, frontage road widths, scenic districts, forest land development permits — pushed to their essence, these are all rules that determine what shape you are allowed to build. Time itself is not their subject. What this article covers is ordinance-based environmental impact assessment: the procedural time that accumulates before construction can begin.

The conclusion first. Lodging facilities are not among the project types covered by Japan’s Environmental Impact Assessment Act. The substance of the regulation sits instead with the 68 ordinance and guideline regimes held by prefectures, government-designated cities and similar bodies — and 57 of those regimes list recreational facilities as a covered project type. Most thresholds are written in terms of “how many hectares of site preparation,” so converting them into room counts shows that, depending on conditions, the procedure can reach down to projects of around 50 rooms, while an ordinary standalone resort hotel of 100–200 rooms will generally fall well short. And where the procedure does apply, the waiting cost is dominated not by the carrying cost of the land but by the delay to opening itself.

Lodging is absent from the Act’s 13 project types — the substance sits in 68 ordinance regimes

The Environmental Impact Assessment Act covers 13 project types: roads; rivers (dams, weirs, floodways, lake and marsh development); railways; airfields; power plants; final waste disposal sites; reclamation and land drainage; land readjustment projects; new residential urban development projects; industrial park development projects; new urban infrastructure development projects; distribution business park development projects; and residential land preparation projects (plus port and harbor plans). Hotels, ryokan and resort facilities appear nowhere in that list. So if you check only the Act to answer “does assessment apply,” you will get the determination wrong.

The substance sits on the local government side. The Ministry of the Environment’s “List of Projects Covered by Local Government Environmental Impact Assessment Systems” (as of March 31, 2025) lists 68 regimes. Counting the breakdown: 57 regimes (83.8%) list recreational facilities as a covered project type; 42 have provisions for composite projects that combine multiple projects in making the determination; and 29 cover the construction of buildings itself. Conversely, the 11 regimes that do not place recreational facilities among covered projects include Tokyo, Aichi, Osaka, Ibaraki, Gifu, Shiga and Tottori — these are designed to catch projects instead through building-side requirements such as height and total floor area.

Two structural points follow. First, in resort development it is “site preparation area” that bites. Second, for high-rise and mixed-use projects in urban areas it is “height and total floor area.” The phrase “environmental impact assessment” is the same in both cases, but the entry point differs entirely between the mountains and the city.

Scale requirements in major jurisdictions — read area-based and building-based separately

Pulling the scale requirements from the actual ordinances and regulations gives the following. All are subject to amendment, so always confirm the latest local statutes with the responsible department at the time of your review.

Converting thresholds into room counts — what triggers at 50 rooms, and what still misses at 769

Area requirements as such do not fit into a business plan. So let us restate the thresholds as room counts. Site area per room varies widely with development density, so we set three cases: 1,000m²/room for a low-rise, low-density resort; 600m²/room for medium density; and 300m²/room for a relatively compact layout.

The 50ha that constitutes a Class 2 project in Hokkaido equates to 500 rooms even at 1,000m²/room, and 833 rooms at 600m²/room. Nagano’s “site area of 30ha” in forest zones and similar sits in the same 300–1,000-room band. In other words, the 100–200-room resort hotel that is normally envisaged will not, on its own, come close to the site-preparation threshold. Consistent with this, Hokkaido publishes five projects as covered under its environmental impact assessment ordinance, and all five are power plants and roads — no recreational or composite facility projects appear. Okinawa’s page on ordinance procedures likewise indicates that no projects are currently in progress. Both are circumstantial evidence that the thresholds are set at a level that does not normally catch a standalone resort hotel. That said, even at a scale below the environmental assessment threshold, cutting and filling are themselves regulated under a separate permit regime. How much time and cost site preparation on sloped land adds is covered separately in our review of the Embankment Regulation Act, under which fills over 1m require the governor’s permission.

So when does it apply? In practice, there are four entry points.

A note on points 3 and 4. Converting a total floor area threshold into a room count requires floor area per room. Extracting from our building plan data those projects with “hotel” as the sole use and 50 rooms or more, floor area per room had a median of 49.0m² and a third quartile of 65.0m² (N=15 projects). The distribution of this per-room unit itself, and how it feeds through to construction cost, is examined in detail in our estimate of ¥20.67 million per room derived from 31 plans. Even applying that 65m²/room figure, Shizuoka’s 50,000m² equates to 769 rooms and Tokyo’s 100,000m² to 1,538 rooms. These are not numbers a lodging-only project reaches.

The time the procedure stacks up — about a year in statutory periods alone, into the low two years with four-season surveys

The skeleton of the procedure runs in this order: the consideration document (planning stage), the scoping document (deciding the methods of survey, prediction and evaluation), conducting the survey, prediction and evaluation, the draft assessment statement (on which opinions are heard), the assessment statement (finalized reflecting those opinions), and then the follow-up survey plan and report. Each stage has periods fixed by ordinance and regulation. Below are the periods that can be confirmed from Shizuoka’s procedural flow and the text of the Tokyo Metropolitan Environmental Impact Assessment Ordinance.

Simply adding up the statutory periods — 60 days for the determination; 30 days of public inspection, 45 days for opinions and 90 days for the governor’s opinion at the scoping stage; 30 days of public inspection, 45 days for opinions and 120 days for the governor’s opinion at the draft statement stage — reaches roughly one year without allowing any time at all for preparing and revising the documents. On top of that come a year of field surveys spanning the four seasons and several months to prepare the draft statement. Where a consideration document procedure precedes it for a Class 1 project, a further front-end stage stacks up. Even assuming the developer’s own work proceeds without a hitch, the practical way to hold it is a range of 1.5 to 3 years.

Time our data captures, and time it does not — high-rise mixed-use runs 5.41 years after groundbreaking alone

Plan data derived from the Ministry of Land, Infrastructure, Transport and Tourism’s “Building Dynamics Statistics Survey” allows the construction period to be read for projects where both the scheduled start date and the scheduled completion date are available. What is recorded there, however, begins at the building confirmation application — so procedural time before groundbreaking, including environmental impact assessment, is in principle invisible in this data. To estimate what is invisible, let us first confirm what is visible.

Of the 121 most recent plans (on a project basis, excluding duplicates), 17 fell within the scale band of Tokyo’s ordinance for high-rise buildings — 30 stories or more together with a total floor area of 100,000m² or more. Groundbreaking to completion for those 17 had a median of 5.41 years (N=16, projects with an identifiable start date). For the rest of the same dataset the median is 2.25 years (N=97). The gap is 3.16 years.

Stratifying by total floor area produces the same pattern. The 50,000m²-and-above band has a median of 4.92 years (N=25); 20,000–50,000m², 3.33 years (N=6); 10,000–20,000m², 2.42 years (N=11); 5,000–10,000m², 2.51 years (N=38); and under 5,000m² or unknown, 1.90 years (N=112). Splitting by use, projects with “hotel” alone in the use field had a median of 1.90 years (N=124), while mixed-use projects listing multiple uses had 3.00 years (N=68). We have already examined post-groundbreaking time in a review covering 67 planned projects, and the gap seen there between standalone and mixed-use is reproduced in this stratification too.

Looking at individual cases: a redevelopment in Nodayacho, Kita-ku, Okayama City (300 rooms, 40,400m² total floor area, 12 stories) broke ground in April 2021 against a scheduled completion of December 2028 — 7.75 years for groundbreaking to completion alone. A project in Hanazono, Kutchan Town, Hokkaido (500 rooms, 120,000m², 3 stories) started in January 2022 with completion scheduled for December 2025, at 3.99 years. A project in Shimogoryo, Furano City, Hokkaido (200 rooms, 13,000m², 10 stories) is scheduled for completion in December 2027, but the start date is not available in our data.

How the plan data itself presents also warrants care. Confirmation-application-based plans normally appear one to two years before opening, so counts for future years come out structurally low. Of the 200 projects extracted here, 163 were in Tokyo, 8 in Osaka, 8 in Chiba and 3 in Hokkaido, with Nagano, Shizuoka and Okinawa at one each. Resort development is almost invisible in plan data not because supply is thin, but because the scale is small and the confirmation application has not yet come due.

The structure of waiting costs — land carrying cost is only 14% of the total

Costs accrue during the procedural period as well. But what bites in an investment decision is not the carrying cost itself. First, the assumptions behind the estimate.

On these assumptions, annual revenue per room is ¥4.11 million, and applying the 31% GOP margin for 200 rooms gives GOP of ¥1.275 million per room. The cost of one year of waiting breaks down into two parts.

The carrying cost of the land comes to only ¥197,000 per room even at an annual rate of 5.9%. The loss in project value from pushing the opening back by a year, by contrast, is ¥1.209 million — more than six times as much. Of the ¥1.41 million total, carrying cost accounts for 14%; the body of the waiting cost is the start date of the cash flow itself. How holding taxes erode NOI yields, and how the level of interest rates feeds through to project viability, each matter in their own right — but confined to the pre-groundbreaking waiting phase, shortening the procedure by a few months does more than moving interest rates or tax rates by a percentage point.

Read on two axes, what sets the level of the waiting cost is the procedural duration, not the carrying cost rate. Holding the procedural duration at 2.5 years and moving the carrying cost rate 2.5 points, from 4.5% to 7.0%, swings the per-room figure only from ¥3.28 million to ¥3.49 million — a range of ¥210,000. Holding the carrying cost rate at 5.9% and extending the procedural duration by two years, from 1.5 to 3.5, moves it from ¥2.08 million to ¥4.65 million — ¥2.57 million. The ratio of sensitivities is twelvefold. Establishing up front whether ordinance-based assessment applies to the site does more for the precision of a business plan than refining assumptions about interest rates or tax rates.

Under the 2.5-year scenario, ¥3.40 million per room. Against construction cost of ¥38.38 million, that is 8.9%. This is not an unforeseen outlay; it is the price of time, and it belongs in the business plan from the outset. Put the other way: if the presence or absence of an assessment can be determined in advance, that ¥3.40 million can be reflected in advance in the acquisition price negotiation and the project schedule. It becomes an unforeseen cost only when the determination is deferred.

Stay under the threshold, or go through it — the break-even is 6.6 years

A choice frequently considered in practice is to hold the room count or the site preparation area just below the scale requirement and avoid the procedure altogether. Time can be bought, but revenue scale falls. Which is better can be calculated.

Compare a 200-room scheme that goes through the procedure with a 140-room scheme shrunk to stay under the threshold. Reflecting economies of scale, the GOP margin is set at 31% for 200 rooms and 28.6% for 140.

The GOP difference between 200 and 140 rooms is ¥90 million a year. Capitalized at 5.5%, that is a project value difference of ¥1.64 billion. The waiting cost where the 200-room scheme takes 2.5 years in procedure, by contrast, is ¥679 million, and ¥806 million even at 3.0 years. Going through the procedure comes out roughly ¥900 million ahead.

That relationship reverses — that is, shrinking becomes the better option — once the procedural duration exceeds 6.61 years (at a carrying cost rate of 5.9%; 6.91 years at 4.5% and 6.39 years at 7.0%). That is far outside the 1.5–3-year practical range seen in the previous section. Change the assumptions and the numbers move, but at least on the basis used here, cutting rooms to avoid a scale requirement is unlikely to pay.

Two conditions attach, however. One is that funding can be secured for the waiting period. The other is that the opening does not end up structurally out of step with the peak of demand. The latter is the subject of the next section.

Shift the opening year by one, and the competitive line-up changes

If the estimate of procedural duration is off by a year, the opening year moves by a year. How new supply coming on stream in the same area at the same time overlaps changes accordingly. Lining up new openings nationwide (on a confirmed-listing basis on major booking sites) by year gives 2,012 properties in 2023, 1,979 in 2024, 1,746 in 2025 and 1,509 in 2026.

What requires attention is the breakdown by category. Of the 1,509 properties in 2026, 729 (48.3%) are rental villas, while resort hotels number just 90. At 93 in 2024 and 101 in 2025, that is roughly flat — a national level of 90–100 a year, or an average of two per prefecture. How the picture of supply flips depending on whether you count properties or rooms is set out in our breakdown of 2026’s 1,242 openings and 33,749 rooms. That new resort hotel supply is this thin means a one-year shift essentially swaps out the competitive line-up. In a market with many openings, a year either way is absorbed into the average; in a market of two a year, who ends up next door changes.

Put the other way, the accuracy of the estimate matters more to the competitive environment than the length of the procedure itself. Assume 2.5 years and finish in 2.5, and the opening year does not move; assume 1.5 and take 2.5, and it slips by a year. The value of mapping out the pre-groundbreaking permitting layer in advance lies less on the cost side than on the side of schedule certainty.

In summary — where to pin down the time before groundbreaking

Pulling it together:

What to do at the land-assembly stage is simple. Check the ordinance of the prefecture or designated city in which the site sits; see whether there is a covered project type corresponding to recreational facilities, composite facilities or new building construction; and read what unit the threshold is written in (site area, land-form change area or total floor area). Then overlay whether the site falls within a special zone under the Natural Parks Act, or a special zone or district under a prefectural natural park or nature conservation ordinance. These two steps largely settle whether procedural time will stack up before groundbreaking. As with forest land development permits, buried cultural properties and soil contamination, the determination itself is the kind of work that can be finished before the land is acquired.

Time cannot be cut, but it can be pinned down. If the business plan says 2.5 years, that is an assumption, not a delay.

Related reading

• Hotel Build Times: 2.80-Year Median, 1.96 Solo vs 4.00 Mixed-Use
• Japan Forest Law: 1ha Permit Trigger, 50% Forest Cover for Hotels
• 61% of 2026 Hotel Openings Fall Outside Urbanization Promotion Areas
• Japan Scenic Districts: 20% Coverage Cap Means 5x Land Cost Per Room
• Hotel Construction Cost ¥8.7M-71.4M per Room: 3 Tiers, 12 Markets
• Japan Earthwork Permits Cover 81.8% of Land: Slope Resort Costs 2026
• Japan 2026: 1,242 Openings, 33,749 Rooms — 47% Villas, 42% Business
• Hotel Construction Cost ¥20.67M/Room: 35.0m² Floor Area in 31 Plans