1. The Puzzle
A traveler moving between countries notices the difference before he can name it. In an American suburb, wooden poles march down the alley or the back lot line, each carrying a gray canister transformer and a fan of drop wires to nearby roofs. In a British housing estate of similar vintage and similar wealth, the sky is empty; power arrives from a cabinet at the corner through a trench under the pavement. In Tehran or in Delhi, the sky over a narrow lane is crossed and re-crossed by so many conductors, some utility-owned and some not, that a delivery van with a tall box has to judge clearance the way a boat judges a bridge.
These are not differences in technical sophistication. All four places deliver alternating current at reliable frequency to metered customers. They are differences in choice — and the choices were made by different actors, at different moments, under different constraints, and with different things counted as costs.
This paper sets out the factors that drive those choices. The aim is not to declare a winner. Undergrounding is not simply the “advanced” option and overhead is not simply the “backward” one; some of the wealthiest and most orderly countries on earth remain committed to poles, and some of the most tangled overhead networks are the product of rational decisions under real constraints.
2. What the Choice Actually Is
“Power lines” is a single phrase covering three distinct systems, and confusion about national practice usually comes from collapsing them.
Transmission carries bulk power at high voltage (roughly 110 kV and above) between generating stations and regional substations. It is overhead almost everywhere, in nearly every country, because the cost gap is at its widest here and because the routes cross open country where the visual cost is lowest. Estimates for constructing underground transmission range from four to fourteen times the cost of comparable overhead lines; a typical 69 kV overhead single-circuit line has been put at about $285,000 per mile against roughly $1.5 million per mile underground. Germany, often cited as an undergrounding leader, has only about 0.4 percent of its transmission grid below ground.
Medium-voltage distribution (roughly 4 kV to 35 kV) moves power from substations through neighborhoods to the local transformers.
Low-voltage distribution and the service drop — the last few hundred meters and the last few meters — is the segment the traveler actually sees from the sidewalk, and it is where national practice diverges most sharply.
The right question is therefore not “does country X bury its power lines” but “at what voltage level does country X stop tolerating an overhead conductor, and why there?”
3. The Technical Fact Behind Half the Difference: Voltage and Transformer Density
The single most consequential engineering decision, and the one least discussed in popular comparison, is the standard household voltage and the architecture it forces.
The United States settled on a 120/240-volt split-phase service. Low voltage means high current for the same power delivered, and high current means resistive losses that climb quickly with distance. The practical consequence is that a low-voltage American circuit cannot run far. The transformer must therefore be close to the customer — which is why the American streetscape carries a pole-mounted transformer roughly every four to ten houses, and why the American distribution system pushes medium voltage deep into residential streets.
Most of Europe, and much of the world that was wired later using European standards, adopted 230/400-volt three-phase service. Doubling the voltage quarters the losses for the same delivered power. A single ground-mounted substation can therefore serve one hundred to three hundred dwellings over a low-voltage network extending several hundred meters.
This changes the undergrounding arithmetic completely. In a 230-volt country, burying the last mile means trenching a handful of cable runs from one compact substation. In a 120-volt country, it means burying a dense capillary network plus dozens of pad-mounted transformers, each requiring a vault or a green cabinet, each needing access for replacement. The American system is not merely more visible; it has more objects in it. A country that chose the higher voltage effectively pre-purchased a cheaper path to undergrounding, decades before anyone framed the issue that way.
Iran, India, and Japan complicate the picture in instructive ways. Iran and India use 230-volt service but wired vast populations very rapidly and under capital scarcity, so they took the cheap construction path even where the architecture would have permitted the expensive one. Japan uses 100-volt service — the lowest common standard in the developed world — which multiplies transformer count and helps explain why the country carries roughly 35 million utility poles and continues adding on the order of 70,000 per year.
4. Cost, and the Question of Who Pays
The cost premium for undergrounding is real but is routinely quoted in a misleading way, because three different numbers get conflated.
New construction in a greenfield development carries the smallest premium. The trench is open anyway for water, sewer, gas, and telecommunications; there is no pavement to cut and restore; there is no existing service to keep energized during the work. In favorable rural conditions the gap can nearly close — one South Dakota cooperative estimated roughly $18,000 per mile overhead against $24,000 underground for a simple line.
Conversion of an existing overhead system carries the largest premium, because the utility pays for excavation through finished streets, restoration, coordination with buried utilities of unknown location, and temporary service. A Virginia regulatory review estimated that placing all existing distribution lines in that state underground would run about $83 billion, roughly $3,000 per customer per year, and could take decades.
Lifetime cost is closer than either figure suggests. The same Virginia work found routine operation and maintenance costs for typical underground systems nearly identical to overhead — about $920 per mile per year against $917 — though duct-bank systems used in dense commercial districts ran roughly four times higher. What shifts is the shape of the risk: underground faults are rarer but each one is worse. Restoring an overhead outage has been put at roughly $3,000 to $5,000 per mile, while underground repairs can exceed $20,000 per mile, and locating a buried fault takes time that a lineman with binoculars does not need.
The decisive question is therefore not “what does it cost” but “who is permitted to pay for it, and over what horizon.” A regulated utility earning a return on capital invested has an incentive toward the more expensive asset if the regulator will approve it. A utility under a strict price cap does not. A municipality that can levy an assessment district on benefiting property owners will underground; one that cannot, will not. California’s Rule 20 framework is a clean illustration: it sorts undergrounding projects by who bears the cost — the full ratepayer body, a partial mix, or the property owner directly — and the volume of work done under each tier tracks the funding, not the engineering merit.
5. Physical Geography
Ground conditions set the floor on cost and sometimes veto the choice outright.
Soil and rock. Trenching through the alluvium of the Netherlands is a different undertaking from trenching through the granite of the Canadian Shield or the coral limestone of parts of Florida.
Water table. High groundwater degrades direct-buried cable and floods vaults. Coastal and delta cities that underground must invest in sealed systems.
Frost. Northern countries must bury below the frost line regardless, which raises the marginal cost of depth very little once the trench exists — one reason Nordic undergrounding programs pencil out.
Vegetation. This is the great hidden driver. Tree contact is the leading cause of distribution outages in forested temperate countries, and vegetation management is a permanent operating expense that undergrounding eliminates. Finland’s program is explicitly storm-and-snow-driven: after a history of severe outages from storms and heavy snow loads, an updated Electricity Market Act imposed outage requirements that pushed distribution operators toward cabling, with one major operator moving from 38 percent underground in 2016 toward a 75 percent target for 2028.
The converse matters just as much. Much of Iran is arid to semi-arid, with sparse tree cover, few ice storms, and no hurricanes. The largest single argument for undergrounding in the American Southeast or in Scandinavia simply does not apply across much of the Iranian plateau. Overhead lines there are not a sign of neglect; they are a rational response to a benign environment for conductors.
Seismicity. This cuts both ways and is genuinely contested. Poles fall in earthquakes, block emergency routes, and pull down live conductors. But buried cable in liquefiable soil is torn by ground displacement and is far slower to locate and repair when the roads above it are damaged. Japan has argued the point with itself for forty years. The historical instinct favored overhead lines precisely because in an earthquake-prone country it has been easier to repair lines that can be seen. The modern policy has swung the other way, focused on keeping roads clear: the infrastructure ministry has aimed at removing poles along roughly 21,826 kilometers of urban emergency transport routes, where new pole construction is already banned because fallen poles obstruct emergency vehicles.
6. Timing: When the Country Was Wired, and When It Was Rebuilt
Distribution networks are among the most persistent artifacts a society builds. A pole line installed in 1935 and maintained ever since is still a pole line in 2026. The decisive variable is often simply when the streets were last open.
Britain and much of continental Europe wired densely in the interwar period and then had large portions of their urban fabric destroyed and rebuilt between 1940 and 1960. Rebuilding a bombed street meant opening the ground anyway. Post-war planning regimes, working with newly nationalized electricity boards, standardized underground service in new development as a matter of course. The result is that London and Paris have essentially all of their distribution below ground, while the same countries retain ordinary overhead lines across rural districts, where the case never justified itself.
The United States wired continuously and was never rebuilt. American suburban expansion after 1945 was rapid, private, and cost-competitive at the margin, and the overhead standard was carried outward with it. Undergrounding became common in American new construction only from roughly the 1970s onward, and mostly where developers or homeowners’ associations valued the appearance enough to absorb the cost.
Japan is the sharpest case of path dependence. Post-war reconstruction prioritized speed and cheapness above all, poles went back up, and the resulting network became so dense and so entangled with property rights and cost-sharing arrangements that unwinding it is now extraordinarily expensive: on the order of ¥100 million to ¥500 million per kilometer, with costs conventionally split in thirds among national government, local government, and the utilities. Tokyo has been trying since 1986 and has reached roughly 7 to 8 percent of cables underground in the 23 wards. The remedy now being pursued is to stop the problem at the source: a proposed metropolitan ordinance would in principle prohibit new utility poles in newly developed residential areas — a first for Japan.
The general lesson: the cheapest moment to bury a cable is the moment the ground is already open. Countries that had such moments and used them look different forever after from countries that did not.
7. Land Tenure, Street Width, and the Right to the Air
Overhead distribution requires a linear right-of-way and clearance. Underground distribution requires a corridor beneath a street that someone controls and that is not already full.
Where street ownership is public, wide, and mapped — most of Western Europe, most planned American subdivisions — the underground corridor is a matter of coordination. Where streets are narrow, ancient, unmapped, or partly private, the underground corridor may not exist at all. In the older quarters of Middle Eastern and South Asian cities, a lane may be three meters wide, already carrying water and sewer of uncertain age and position, bounded by buildings whose foundations are directly beneath the pavement, and subject to property claims that predate the utility. In such a place, the air between the buildings is the only unencumbered corridor available. Slinging conductors from wall bracket to wall bracket is not laziness; it is the exploitation of the only easement nobody contests.
This is the direct answer to the alley observation. The wires strung across lanes in Tehran, Delhi, and Mumbai are the visible record of a network that could not go under the street and so went over it — and then, because the corridor was free and unregulated, accumulated telecommunications, cable television, and informal connections in the same space.
Japan’s version of this constraint is different but related: land values and fragmented ownership make sub-surface corridors costly to establish, which is part of why the 2001 deep-underground law was created to allow public use below about forty meters without compensating surface owners, though it has been used mainly for tunnels rather than for distribution.
8. Theft, Metering, and the Wire as a Social Fact
In a number of countries the overhead network is not only an engineering artifact but a site of contested revenue, and this changes the calculus in a way that has no counterpart in Northern Europe.
An exposed low-voltage conductor at reachable height is an invitation to an unmetered connection. India’s distribution utilities track this through aggregate technical and commercial (AT&C) losses, which combine physical line losses with theft and billing failure. The trajectory is instructive: the Ministry of Power reported AT&C losses falling from 36.64 percent in 2002-03 to 27.15 percent in 2009-10, and more recently to 15.04 percent in FY25, though the Power Finance Corporation’s 2023-24 report showed a worsening from 15.11 to 16.12 percent, with state-level spread mattering more than the national average. At the extreme, Uttar Pradesh was reported to have lost 36 percent of its electric power to theft in 2012, and electricity theft has been estimated at around 1.5 percent of Indian GDP.
This turns undergrounding from an aesthetic program into a revenue-protection program. Reporting on India’s IPDS-era work in Uttar Pradesh describes exactly that logic — loss rates falling from roughly forty-three paise per rupee to under ten once the network went underground and metered, with underground and aerial bundled cabling rolled out at scale in Ayodhya ahead of its emergence as a major pilgrimage destination.
Iran presents a different distortion. Heavily subsidized electricity tariffs suppress the revenue per customer, which weakens the utility’s capacity to fund capital-intensive distribution, while simultaneously encouraging high consumption and rapid load growth. Combine that with sanctions-constrained access to capital equipment and a benign climate for overhead conductors, and the persistence of visible pole-and-wire construction becomes straightforwardly explicable. Where the price signal does not fund the asset, the cheap asset wins.
9. What a Country Counts as a Failure
Reliability statistics look objective and are not. A country that measures average interruption duration per customer and ties utility revenue to it will drive capital toward whatever reduces that number. A country that measures nothing in particular, or that treats scheduled load-shedding as normal rather than as failure, will not.
Germany’s outcome illustrates the linkage: roughly 73 percent of medium-voltage and 87 percent of low-voltage cable underground, associated with very high grid reliability. The Netherlands went further and made the ratchet explicit — under changes adopted in 2010, every new kilometer of aerial line must be offset by converting a corresponding length elsewhere from aerial to underground, a compensation principle that permits growth while forbidding net increase in visible line.
Notice what that policy really is. It is not a reliability rule. It is a landscape rule enforced through the grid code. The Dutch decided that overhead line is a stock to be capped, in the same way a country might cap emissions or building height. Once you can name the thing you are managing, you can manage it.
10. Aesthetics and Amenity as Genuine Policy Variables
English-speaking discussion tends to treat appearance as a soft preference that loses to hard economics. Several countries have declined to treat it that way.
The British approach is the most developed. Rural overhead lines remain common across Britain, but the regulator provides a dedicated allowance for undergrounding existing lines in national parks and designated landscapes, on grounds of visual amenity alone, with no reliability justification required. This is an explicit purchase of a view with ratepayer money, made transparently and capped.
Japan’s program similarly braids together disaster resilience, sidewalk width, and landscape. The stated national aim has repeatedly been framed in terms of both safety and townscape, and the metropolitan strategy has been articulated as a goal of a Tokyo without utility poles.
The point for comparative purposes is that a country cannot pursue a preference it has no institutional means of expressing. The United States has no national mechanism for valuing streetscape in utility ratemaking; the question is settled locally, subdivision by subdivision, which is exactly why American practice looks patchy — undergrounded in the newer and wealthier developments, overhead everywhere else.
11. The Middle Paths
The overhead/underground binary obscures a set of intermediate technologies that increasingly dominate real decisions.
Aerial bundled conductor (ABC) — insulated low-voltage conductors twisted into a single cable and hung from poles or wall brackets. It is far cheaper than trenching, largely eliminates tree-contact faults and accidental contact, and — critically for the theft problem — is much harder to tap without detection. It is the workhorse of upgrade programs in South Asia, Africa, and Southern Europe, and it explains why some cities look dramatically less tangled than they did fifteen years ago without a single meter of trench being dug.
Covered conductor and spacer cable at medium voltage serve the analogous function for wildfire ignition risk, now a dominant driver in the American West.
Targeted undergrounding — burying only the highest-risk segments identified by ignition modeling or outage history — is displacing wholesale conversion programs, because the marginal benefit per dollar is enormously uneven across a network.
Distributed generation and storage change the question rather than answering it. Rooftop generation with local storage reduces dependence on the last mile but rarely removes the connection, and in most regulatory schemes it shifts fixed network costs onto remaining customers, which affects what a utility can afford to build.
12. A Decision Framework
Drawing the strands together, a country’s practice can be predicted reasonably well from eight variables:
- Standard service voltage, which sets transformer density and therefore the number of objects to be buried.
- Vegetation and storm exposure, which sets the reliability value of burial.
- Ground conditions and seismic risk, which set the cost floor and the repair penalty.
- Timing of urban construction and reconstruction, which determines whether the ground was ever open at low marginal cost.
- Street geometry and land tenure, which determine whether a sub-surface corridor is available at all.
- Regulatory cost-recovery architecture, which determines whether anyone is allowed to pay for the expensive option.
- Non-technical loss exposure, which converts insulation and burial into a revenue instrument.
- Whether landscape has an institutional voice, which determines whether appearance can enter the calculation as anything other than a private preference.
Run the United States through this list — low service voltage, mixed vegetation exposure, continuous unreconstructed build-out, fragmented local cost recovery, negligible theft, no national amenity mechanism — and the observed outcome of a heavily overhead network with undergrounded pockets is precisely what the variables predict. Run Britain through it and the observed outcome of undergrounded towns with overhead countryside and a special allowance for protected landscapes is likewise predicted. Run Iran through it — arid climate, subsidized tariffs, capital scarcity, rapid growth, dense old street fabric — and the pole lines and the alley crossings follow.
13. Conclusion
Nothing in the physics of alternating current requires a wire to be visible or hidden. What determines the answer is a chain of decisions, most of them made before anyone framed the question as a choice: what voltage to standardize, when to build, whether the ground was already open, who owns the street, whether the regulator will fund a return on buried assets, whether an unmetered tap is a live risk, and whether anyone has standing to speak for the look of a street.
Two conclusions follow for anyone weighing a change.
First, the cheapest intervention is almost always a rule about new construction rather than a program of conversion. Tokyo’s proposed prohibition on new poles in new residential development will do more per yen than decades of retrofitting, and the Dutch compensation principle achieves a similar result by capping the stock rather than attacking it. A country that stops adding overhead line has already solved the problem on a fifty-year horizon.
Second, the comparison should be made honestly. Overhead construction is not a failure of will. It is faster to build, faster to repair, cheaper to modify as load grows, and better suited to sparse settlement, arid climates, and unstable ground. A network of poles in rural Oregon or on the Iranian plateau is a reasonable answer to the conditions it faces. What is not reasonable — and what the alley wires of Delhi and Tehran genuinely represent — is an overhead corridor that has been allowed to accumulate without ownership, inspection, clearance standards, or removal of dead conductors. The problem there is not that the wires are in the air. It is that nobody is responsible for the air.
