9 Sonic injustice
There is a tempting story about noise and justice, and it is half true. The tempting story says noise falls, like other environmental harms, most heavily on the poor — that the map of exposure is the map of deprivation. In many places it is: studies in the Netherlands and the United States find heavier burdens in deprived and minority neighbourhoods. But noise does not always behave like air pollution, and the honest version of the justice argument is more interesting than the slogan. In Barcelona between 2007 and 2014, and under the aircraft corridors over London, exposure ran higher among the more affluent — the inverse of the pollution pattern. That inversion is not an awkward exception to be explained away but the finding itself: the geography of noise is real, unequal, and irreducible to a single axis.
9.1 The expected pattern
The slogan earns its force honestly, because in a great many settings the burden really does track disadvantage. The most comprehensive single body of evidence comes from the United States, where a nationwide model of noise exposure found levels consistently higher in neighbourhoods with lower income and lower educational attainment, and found them higher again where the Black, Hispanic and Asian populations were larger. Racial segregation compounded the effect, so that the loudest places were the poor, segregated ones (Casey et al., 2017). The pattern recurs in Europe. A national study of the Netherlands, spanning more than nine thousand neighbourhoods, found environmental noise higher where incomes were lower, employment scarcer and the residents more often of non-Western migrant background — the same social gradient, on the other side of the Atlantic (Hayward & Helbich, 2024). Reviews of the wider literature record the same tendency, at least where deprivation is measured directly (Trudeau et al., 2023).
The mechanism is largely a matter of where the loud things were put. Major roads, rail corridors, flight paths and industrial sites tend to run through or beside the cheaper land — which is to say the land occupied by those with the least power to refuse them. That arrangement was produced over decades by the ordinary workings of land value and planning, and in some places by outright discrimination in how neighbourhoods were zoned, served and mortgaged. The infrastructure that carries a city’s noise was routed, again and again, away from the districts that could resist it and toward those that could not. What looks on a map like an accident of geography is often the residue of who held power when the road was drawn. The same asymmetry runs through the decisions taken since. Which streets are pedestrianised, where a barrier is raised, whose complaint is logged and acted upon — these are settled in planning processes to which the poorest residents have the least access and the faintest voice (Trudeau et al., 2023). The distribution of noise is shaped not only by where the powerless were placed, but by how little say they have had in moving it since.
Exposure, moreover, is only half of the disadvantage. Once the noise arrives, the capacity to keep it out is itself unequally distributed, and along the same lines. Older and cheaper housing comes with thinner walls and single glazing; the double-glazed window, the flat that faces the courtyard rather than the avenue, the move to a quieter street, the second home to retreat to — each is a way of buying quiet, and each is rationed by income. Tenure sharpens the divide, since a renter cannot readily retrofit a building that is not theirs, and time deepens it, since those with the fewest options to be elsewhere spend the most hours inside the exposure. The disadvantaged household therefore tends to receive more noise at the façade and to stop less of it at the wall — a compounded penalty that a map of outdoor exposure does not even register. Where this is the shape of things, the slogan is simply true, and the injustice is of the familiar kind.
9.2 The inversion
It is not always the shape of things. In Barcelona, environmental noise over the years 2007 to 2014 was distributed in the opposite direction from air pollution measured in the same city: the pollutants were heaviest in the more deprived neighbourhoods, as the intuition predicts, but noise exposure was higher in the more affluent ones (Saez & López-Casasnovas, 2019). A dedicated mapping of the city’s noise inequity located the loudest conditions in the central districts — the dense, mixed, sought-after core rather than the poorer periphery (Lagonigro et al., 2018). And the pattern is not peculiar to Spain. In London, the odds of living inside the fifty-decibel contour of aircraft noise were highest for residents with the highest household income, the lowest area-level deprivation, and white ethnicity — a clean reversal of the environmental-justice expectation. Tellingly, the same study found rail noise running the other way, more common among Black residents than white: two sources, two opposite social gradients, within a single city (Tonne et al., 2018). Reviews of the evidence describe the cross-city picture as frankly mixed, the social gradient of noise pointing in opposite directions in different places (Trudeau et al., 2023) — which is itself the point.
The reason noise can behave this way is that it attaches to the very things that make a location desirable. Density, centrality, street life, the constant traffic of a place where people want to be — these generate sound, and they are also what commands a premium. Central districts carry, on top of their traffic, a further layer the periphery is largely spared: the noise of leisure and nightlife, the terraces and venues that cluster where the crowds are. The lively inner district is loud precisely because it is lively; the calm lies at the edges, in the peripheral estates and the low-density, leafy uptown. To live where the action is, in most cities, is to pay for it partly in decibels, and those who can afford the centre are buying the noise along with the amenity. That premium is capitalised into rents and prices, which complicates the language of choice. The owner who buys into a loud central district has in some sense chosen its noise; the tenant priced into the same building, or the long-standing resident who cannot afford to leave as the district grows expensive around them, has chosen nothing. Even in the affluent centre, the exposure and the power to escape it need not be held by the same hands.
The terrace is not an unregulated gap; it is regulated and repeatedly re-litigated. Madrid’s 2022 ordinance, which relaxed the terms on which hospitality terraces could occupy public space, was annulled by the Madrid High Court for having skipped the environmental and budgetary assessments the law requires [ECLI:ES:TSJM:2024:3559], a ruling the Supreme Court left standing on appeal. A new ordinance took effect only in late 2025, with the interval governed by the older, less permissive 2013 rules (Ayuntamiento de Madrid, 2025; Tribunal Superior de Justicia de Madrid, Sala de lo Contencioso-Administrativo, Sección Segunda, 2024). Palma’s own ordinance names the tension in its own preamble, presenting itself as an attempt to reconcile the city’s economic development and tourist nature with residents’ right to rest (Ajuntament de Palma, 2014). And the friction is not confined to tourist capitals: a neighbourhood federation took Palencia’s terrace ordinance to the Castile and León courts on the same grounds, excessive occupation of public space at residents’ expense (Tribunal Superior de Justicia de Castilla y León, Sala de lo Contencioso-Administrativo, 2026). The underlying conflict is about public space and rest, and tourism only sharpens it where it is already most contested.
Where litigation is slow or costly, residents’ associations have increasingly gone to the ombudsman instead. In Zaragoza, the Stop Ruido Casco Histórico association took its complaint to the regional Justicia de Aragón in 2025, naming both hospitality terraces and the short-term tourist lets multiplying around them, clustered against heritage-listed buildings in the historic centre (European Network of Ombudsmen, 2025). In Alicante, a residents’ association’s complaint led the regional ombudsman to find the city’s 1991 noise ordinance inadequate to the leisure-district noise it now faced, and to recommend the city consider declaring an acoustically saturated zone under the Directive’s own domestic tool (Síndic de Greuges de la Comunitat Valenciana, 2023). And in 2023 Spain’s national ombudsman opened the same file simultaneously against Madrid and Barcelona, asking each for a programme of scheduled, unannounced inspections rather than one that waits on a resident’s complaint (European Network of Ombudsmen, 2023).
The pattern these disputes trace out has a name in the recent literature: overtourism does not arrive as noise alone, but noise is consistently among the costs it imposes on the people who live where it lands. A peer-reviewed study of Málaga found long-term residents identifying noise pollution alongside rent inflation and diminished access to public services as the burden of a tourism model whose gains accrue chiefly elsewhere. It linked that burden to a wider erosion of affordable housing, of neighbourhood continuity, of the historic fabric itself, which residents experienced as a slow expulsion from their own city (Siyamiyan Gorji et al., 2026). The terrace, in that light, is not a nuisance incidental to a successful tourist economy; it is one of the more audible ways that economy’s costs are distributed onto those least positioned to charge for them — the same distributive logic that runs through the whole city, heard here at the scale of a single street.
Barcelona’s own current figures make the shape of it concrete. Take the ten districts, and the share of each district’s residents exposed to a day-evening-night level at or above sixty-five decibels. The loudest by a wide margin is l’Eixample, the dense, central, relatively affluent grid, where roughly thirty-eight per cent of residents live above that level. The quietest include Nou Barris, one of the city’s humblest districts, at under fourteen per cent; the wealthy but peripheral Sarrià-Sant Gervasi sits well below l’Eixample despite having by far the highest incomes. Across the ten districts the correlation between household income and noise exposure is positive, and it does not soften as the threshold rises. Restrict attention to the very loudest exposure, above seventy or seventy-five decibels, and the concentration in the wealthier districts grows stronger, not weaker (Ajuntament de Barcelona, 2022; Ajuntament de Barcelona, Oficina Municipal de Dades, 2019).
None of which means that wealth produces noise, or that the gradient is clean. The driver is centrality and density; income enters only because the central, dense districts happen to be the desirable and expensive ones. The relationship frays at the edges accordingly. Ciutat Vella, the old town, is central and far from wealthy yet relatively quiet, its tight and partly pedestrianised streets carrying less traffic than the great avenues to the north; the wealthiest district is not the loudest, and one of the poorest is not the quietest. The inversion is a tendency, not a law — which is exactly why it resists the single-axis story, in either direction. There is a temporal dimension that a single snapshot hides. As heavy industry has left the inner city and the centre has gentrified, the geography of noise and the geography of deprivation, once more nearly aligned, have in places pulled apart: the factory that once punished a working-class district is gone, while the traffic and nightlife that replaced it now fall on a district that has since grown wealthier. The inversion is, in part, a record of what cities have become.
9.3 Why noise is not air pollution
Noise and air pollution share their principal source — the motor vehicle — and yet they come to rest in different places, and the difference is instructive. A pollutant, once emitted, disperses: it drifts on the wind, accumulates in still air, settles over whole districts, and concentrates where the dirty infrastructure was sited, which is disproportionately where the poor live. Noise neither drifts nor accumulates. It radiates from its source and falls away steeply with distance, halving and halving again over a few hundred metres, so that it attaches not to a region downwind but to a place — to the specific, dense, trafficked fabric that produces it. Pollution is a fog that pools in the low ground of the social landscape; noise is a property of the ground itself, and the desirable ground is often the loud ground.
This is why an environmental-justice framework built on the model of pollution does not transfer cleanly to sound. The pollution model assumes that a hazard flows toward the powerless, and prescribes moving the burden back up the gradient. Noise can flow toward the powerful — not because it seeks them out, but because they have chosen to live amid the density that makes it. A politics of noise that simply imports the pollution script, hunting for the burdened poor and redistributing the harm upward, will in a city like Barcelona be aiming in the wrong direction, and will miss the people it most needs to reach.
For the deeper complication is that being more exposed and being more harmed are not the same thing. The same Barcelona analysis that found noise heavier in the affluent neighbourhoods also found the risk of dying from environmental hazards markedly lower there, on the order of thirty per cent lower in the most affluent areas. The residents of the poorer neighbourhoods were more susceptible to a given exposure, carrying worse baseline health, more overlapping exposures and less capacity to recover from any of them (Saez & López-Casasnovas, 2019). The susceptibility is not mysterious. A body already worn by insecure work, poorer diet, untreated illness and the chronic stress of scarcity has less reserve to absorb a night of broken sleep. It rarely meets the noise alone, either, for the loud street is often also the polluted one, the overcrowded flat, the neighbourhood with the fewest green places to recover in. The harms stack, and they stack on the same people. Exposure and impact can therefore diverge: the affluent may take in more noise and suffer less from it, while the deprived take in less and suffer more. A justice defined purely on exposure would then reach a perverse destination: the first duty would be to shield the already-advantaged from their self-chosen decibels, while the people bearing the heavier bodily cost went unattended.
There is a further dimension the exposure map cannot see at all. Sonic injustice, on the fullest account, is not only the unequal distribution of unwanted sound but the unequal distribution of access to wanted sound — to quiet, to restorative and pleasant acoustic environments, to the plain ability to govern what reaches the ear (Trudeau et al., 2023). On that axis the advantage runs the ordinary way, and runs sharply. The affluent household, even where the street outside is loud, can buy its way inward to quiet through glazing and a courtyard-facing room, and outward to the good soundscapes that the poor cannot reach — the garden, the calm park, the house in the hills. Its exposure is, in a sense, chosen and escapable; the exposure of the poor is neither. Measured only at the façade, the injustice looks scrambled; measured by who can command quiet when they need it, it resolves back toward the familiar direction. Both are true at once, and neither alone is the whole of it. It is also the half of the problem least examined. The evidence on who bears unwanted sound is by now substantial; the evidence on who is denied access to wanted sound barely exists — a single study set against dozens (Trudeau et al., 2023). What is not measured tends not to be governed, and the access gap has so far been largely neither.
9.4 Exposure against income
The two quantities can be held against each other directly. In the map below, each of Barcelona’s ten districts carries both at once: household income, and the share of its residents above a chosen noise level. Moving the threshold recolours the districts and recomputes the correlation between them. What emerges is the overall alignment, noise and affluence rising together against the pollution gradient, and also the places where it breaks — the central-but-modest and the wealthy-but-peripheral that keep the pattern from being a straight line.
Source · household disposable income per capita by district, indexed to Barcelona = 1.00 (2019), after Ajuntament de Barcelona, Oficina Municipal de Dades (2019); share of residents exposed to \(L_\text{den}\) by district, all sources (2022), after Ajuntament de Barcelona (2022); district boundaries from CartoBCN (Ajuntament de Barcelona). Accessed 30 July 2026.
9.5 Justice without a single axis
If the burden of noise runs one way in one city and the other way in the next, and if being more exposed is not the same as being more harmed, and if the power to obtain quiet is distributed differently again, then it is no longer obvious what a just distribution of quiet would even be. It cannot be simply the redistribution of noise away from the poor, because in some cities the poor are already the quieter. It cannot be equal exposure for all, because equal exposure would ignore both the unequal susceptibility that makes the same decibel more damaging to some, and the unequal access to quiet that persists whatever the façade reading says. Each simplification, pressed against the evidence, gives way.
What is left, once they fall away, is a more demanding and more honest object. A just soundscape would protect the most susceptible before the most exposed, since it is the susceptible who convert a given noise into the greater loss of health. It would treat access to quiet as a good that ought not to be rationed by income — a restful bedroom, a nearby calm park, the plain ability to escape sound at will. And it would refuse to load the heaviest exposure onto those least able either to keep it out or to recover from it, even where, on the raw map, they are not the most exposed of all. That is a distributional standard with several axes rather than one, and no single map will satisfy it. In practice it would pull policy in more than one direction at once: toward glazing and barriers for the exposed poor, toward new quiet places in the districts that have none, toward protecting the susceptible neighbourhood even when its façade readings are not the city’s highest. And it would judge success not by the average decibel brought down but by whether those least able to secure quiet for themselves had at last been helped to it.
The distribution, though, is only the first of three questions. It does not settle whether protecting quiet is worth its cost, nor why instruments tuned to a single averaged number at the façade should so reliably miss both the susceptibility and the access that no exposure figure records. The answer to neither follows from a map.