Why Japan’s Utility Poles Are Not Going Away

Japan has a new national undergrounding plan, stronger disaster evidence and cities that want open skies. It also has roughly 36 million poles, narrow streets packed with buried infrastructure, millions of individual service connections, and a system in which digging one kilometer can take years. The future is fewer poles in important places—not a pole-free archipelago.

The short answer: Japan is prioritizing, not erasing

Japan is accelerating mudentchūka—literally, making streets “without utility poles.” But the third national undergrounding plan, adopted on June 2, 2026, does not promise to remove every pole. Through fiscal 2030 it seeks about 1,000 kilometers of completed work and plans for roughly 4,000 additional kilometers, concentrated where falling poles would block emergency access, endanger pedestrians or spoil nationally important landscapes.

That distinction matters. Undergrounding a main evacuation road, the route to a hospital or a dense school street can deliver large public benefits. Burying every lightly traveled rural spur or every already reliable residential line may cost far more per household while producing less safety or resilience. Japan’s policy has therefore moved from a visual ideal—clear every skyline—to a risk-based infrastructure program.

The scale of the installed network explains the caution. Government material commonly describes about 36 million electricity and NTT communications poles. Even after a 2016 law declared that no-pole streets should be promoted, the total continued to rise. In fiscal 2023, about 237,000 electricity and communications poles were installed and about 195,000 removed: a net increase of roughly 42,000.

First lesson: “undergrounding” is not just putting wire in soil

An overhead distribution street is a visible system. Poles hold 6.6-kilovolt distribution lines, lower-voltage service wires, transformers, switches, streetlights and multiple telecommunications cables. From there, individual drops reach each house, shop, apartment and traffic device.

A buried system must reproduce those functions below ground while keeping equipment inspectable, cool, dry and electrically separated. Crews construct conduits or a shared utility duct, cable chambers, joints and service connections. Transformers and switching gear usually do not disappear: they move into cabinets at ground level or, where space permits, into buildings. The result is a clearer sky, not an equipment-free street.

Why one kilometer costs so much

The cable is not usually the dominant expense. The street is. A project must survey undocumented conditions; design around water, sewer, gas and existing telecom ducts; move conflicting facilities; negotiate cabinet locations; maintain access to homes and businesses; manage traffic; excavate; install ducts and chambers; connect every customer; remove old lines and poles; and restore pavement.

Conventional work is often described as costing around ¥500 million per kilometer of road—about $3.1 million at the exchange rate shown on this page. The number varies sharply with road width, demand density, geology, traffic controls and what already lies underground. It is an order of magnitude, not a universal tariff.

Cost is divided because the project creates several kinds of value. Under the common-duct model, the road authority builds much of the shared space while power and communications companies install cables and related facilities. National and local government and utility customers ultimately carry different portions through budgets, subsidies and regulated network charges. The economic question is not simply “Can Japan afford it?” but “Which streets justify the opportunity cost before bridges, water pipes, schools and other aging assets?”

Narrow Japanese streets create a three-dimensional puzzle

Many neighborhoods grew from paths and pre-automobile streets. Buildings sit close to the road, sidewalks are absent or thin, and the shallow underground space is already occupied. A common duct needs horizontal and vertical clearance. Ground-mounted transformers need places where they do not block wheelchairs, sightlines, shop entrances or fire access.

This is why a wide boulevard can be easier than a quiet lane. On a boulevard, the work is expensive and traffic management difficult, but there may be a real sidewalk and repeatable geometry. On a narrow residential street, a cabinet 1.45 meters high and roughly 0.45 meters deep can consume the very pedestrian space pole removal was meant to recover. Placing equipment on private land requires easements and owner consent.

The underground is also institutionally crowded. The road manager, electric distributor, NTT and other telecom carriers, gas and water utilities, sewer operator, police, property owners and contractors have distinct assets, schedules and legal duties. A design change by one party can force changes by several others.

Why projects take years even when everyone agrees

A Ministry of Land, Infrastructure, Transport and Tourism review put the average common-duct project at about seven years. The sequence explains why: preliminary coordination, survey and design, relocation of existing utilities, duct construction, installation and testing of each operator’s cables, transfer of thousands of customer connections, and finally removal of poles.

Old poles cannot be pulled when the main duct is finished. Every active electricity, telephone, fiber, cable television and streetlight connection must first be migrated. Some owners may be absent; buildings may need work on private property; operators cannot all switch on the same day. A street can look “finished” while poles remain as temporary support for the last unswitched service.

Japan is trying simultaneous construction, combined procurement, three-dimensional digital surveys and standardized low-cost designs. Shallow burial, compact boxes, roadside drainage space and direct burial can reduce excavation. They expand the feasible set, but no technique removes the need to know what is underground and to keep the network safe for decades.

Disaster resilience: underground is better at some failures, worse at others

The strongest case for removal is often disaster access. The 2024 Noto Peninsula earthquake damaged or toppled about 3,480 poles, according to the national plan, obstructing road clearance as well as electricity and communications restoration. Typhoons, trees and flying debris can bring down overhead networks and leave narrow streets impassable.

Buried cables are protected from wind and most falling objects, and they cannot topple across an evacuation route. Evidence from the 1995 Great Hanshin-Awaji Earthquake also showed a lower damage incidence for underground distribution equipment than for poles in the most strongly shaken areas.

But “underground equals disaster-proof” is wrong. Earth movement, liquefaction, flooding, salt water and damaged joints can harm cables and ducts. A broken overhead conductor can often be seen and bypassed quickly. An underground fault must be located, excavated and worked on in a confined environment. The Agency for Natural Resources and Energy has noted that restoration of underground equipment may take about twice as long as overhead equipment once damage occurs.

Resilience is the product of two questions: How likely is a failure, and how hard is it to recover? Undergrounding often lowers the first number while raising the second.

The rational strategy is therefore selective and redundant. Bury lines where pole collapse would cause cascading harm—emergency transport routes, hospitals, evacuation sites, isolated communities—and pair them with sectionalizing switches, alternate feeds, mobile generation, flood protection and repair plans. A cable’s location alone does not make a resilient grid.

How Japan became a country of poles

The international comparison is true—and incomplete

Government presentations often contrast Tokyo’s 23 wards at about 8% and Osaka at about 6% with London, Paris and Hong Kong at or near 100% on the selected road measure. The gap is real, but the comparison needs context. Definitions, geographic boundaries, reconstruction history, road ownership, density and whether low-voltage or telecommunications lines are counted can differ.

Many European centers installed underground networks while streets were rebuilt or before modern demand filled the subsurface. New developments can coordinate utilities before residents arrive. Retrofitting a live Japanese lane is more expensive than designing a pole-free subdivision on empty land. This is why preventing new poles can be more cost-effective than removing old ones later.

Climate and hazard mix also matter. Undergrounding performs especially well against wind and ice, but flood-prone cabinets, earthquake deformation and tsunami exposure require different designs. “Other cities did it” supports ambition; it does not eliminate local engineering.

Ashiya shows both possibility and patience

Ashiya, the compact and affluent Hyogo city between Kobe and Osaka, has pursued “a city without poles and wires” more vigorously than most municipalities. Its official planning documents reported 32.98 km of pole-free municipal road and a 14.9% rate as of April 2020, then the highest among Japanese municipalities by that measure.

Its experience demonstrates that leadership, landscape policy, resident support and sustained budgets can produce visible results. It also demonstrates the time horizon. Even a wealthy, small city with a clear identity remains far from universal completion. Projects continue district by district, and the city has used hometown-tax donations to support the effort. A showcase is evidence that progress is possible, not that the national problem is easy.

Tokyo’s new plan: go where failure matters most

Tokyo revised its plan on June 30, 2026 around “defending the capital.” Over five years it intends to begin about 320 km and advance roughly 720 km when existing projects are included. Priorities expand toward emergency bases and inside Ring Road No. 8, with more support for wards and municipalities, digital construction, simultaneous works, stronger island resilience and restrictions on new poles in housing development.

This is a portfolio approach. Central Tokyo boulevards, first-tier emergency roads, station areas and island lifelines receive different treatments because their failure consequences differ. Local ward roads remain the hardest frontier: they are numerous, narrow and under local budgets, yet often the places where poles most obstruct pedestrians.

Who pays, who decides, and why consent matters

A pole may stand in a public road but carry assets owned by several companies and serve private buildings. The road authority can designate projects and impose certain occupation limits, yet it cannot simply cut active networks. Utilities must preserve universal service and recover prudent costs. Residents want open sidewalks but may reject a transformer cabinet outside a window or years of night construction.

Distribution costs ultimately appear somewhere: taxes, municipal debt, developer prices or regulated electricity and telecom charges. If all ratepayers fund an expensive beautification project in one wealthy neighborhood, distributional questions arise. If a municipality pays only for scenery, it may underinvest in resilience. Transparent benefit categories—disaster, accessibility, landscape, renewal opportunity—make cost sharing more defensible.

Where removal produces the most value

What would make undergrounding faster and cheaper?

Dig once

Coordinate cable ducts with water, sewer, gas, road renewal and redevelopment. The cheapest excavation is the one another necessary project already opened.

Standardize, but preserve engineering judgment

Shallow conduits, compact boxes and shared tubes can reduce material and excavation. Standards should make repeatable conditions routine without forcing a design into soil, flood or demand conditions where it is unsafe.

Map the underground

Accurate three-dimensional records reduce surprises, redesign and utility strikes. Digital models are not glamorous, but uncertainty is a major cost.

Prevent additions

Require pole-free layouts in suitable new subdivisions and coordinate renewable-energy connections. Avoiding one future retrofit is often more valuable than accelerating one current excavation.

Measure outcomes, not announcements

“Consultation started,” “duct completed,” “cables transferred” and “pole removed” are different milestones. The 2021–25 program began coordination on roughly 3,700 km but completed conduit on a much smaller share. Reporting should show what residents actually receive.

The lesson: open sky is a scarce infrastructure outcome

Japan’s poles are not evidence that the country lacks engineering skill. They are the accumulated result of rational decisions made under earlier conditions: rapid electrification, cheap expansion, dense settlement, repeated rebuilding and fast communications growth. Today’s conditions are different. Aging cities, stronger storms, accessibility, tourism and seismic road clearance increase the value of removal. Higher construction costs, aging water and transport assets, and complex live networks limit the pace.

The right objective is therefore not a dramatic national pole count. It is fewer catastrophic blockages, shorter outages, wider usable sidewalks, better historic streets, fewer unnecessary new poles and lower lifecycle cost. Sometimes underground cable is the answer. Sometimes rear routing, redundancy, stronger poles or local generation will deliver more resilience per yen.

Japan’s skyline will change block by block and decade by decade. The most important wires will disappear first; many ordinary ones will remain visible. That is not policy failure. It is what happens when a beautiful urban aspiration meets a functioning, inhabited, earthquake-prone country that cannot be switched off while it rebuilds itself.