10 The economics of quiet
Quiet has quietly become a commodity. It is sold as a premium on a flat away from the road, as the business model of a wellness retreat, as the added cost of triple glazing. What is priced, in each case, is the absence of a harm measured in the earlier chapters. Both sides of that ledger can be put in numbers: the cost noise imposes (premature mortality, lost sleep and productivity, devalued property, healthy years consumed) and the cost of buying it down. Setting the two against each other is the ordinary business of policy. But it leaves a question a purely economic treatment would pass over: what it means, ethically, that tranquillity has become a scarce resource, and that it is rationed by the ability to pay for it.
10.1 What noise costs
The headline figure is large and recent. The European Environment Agency estimates that transport noise costs Europe at least 95.6 billion euros a year — around 0.6 per cent of the region’s gross domestic product. The agency derives that sum by valuing the loss of some 1.3 million healthy life-years annually (European Environment Agency, 2025). The unit behind the valuation is the disability-adjusted life-year, or DALY: one DALY is one year of full health lost, whether to premature death or to living with a disabling condition. It is a device for putting the incommensurable on a single axis — a death and a year of broken sleep, added together. Its convenience is bought at the price of some violence to both. Divide the euros by the years and the agency’s implied price of a healthy year comes out near 73,500 euros — a number worth holding onto, because everything downstream depends on it.
The 1.3 million years are not evenly sourced. The largest single share, in the World Health Organization’s foundational accounting, comes not from death but from disturbed sleep, followed by annoyance, with cardiovascular disease a smaller but graver third (World Health Organization Regional Office for Europe, 2011). The agency’s own component tally is blunt about the human content of the figure: some 66,000 premature deaths a year, around 50,000 new cases of cardiovascular disease, 22,000 of type-2 diabetes, and, beneath the mortality, 4.6 million people suffering severe sleep disturbance and nearly 17 million chronically annoyed (European Environment Agency, 2025). The words the agency uses are “at least”, and they are doing real work. Several outcomes are excluded for want of firm evidence, and the disability weights applied are conservative, so the true burden is very likely larger than the headline. Set beside other environmental hazards, the figure places transport noise among the top handful of environmental health risks in Europe, behind air pollution and temperature but ahead of most else — a ranking that sits oddly with how little regulatory weight it carries (European Environment Agency, 2025). And the health tally understates the economic one in a further way, since the sleep it breaks and the concentration it interrupts are paid for again in absence and lost output that a count of healthy years does not fully reach.
The figure has a longer pedigree at smaller scales. Britain’s environment department has for years put the social cost of noise in England at somewhere between seven and ten billion pounds a year, arrived at by similar means (Department for Environment, Food and Rural Affairs, 2014). And the method scales down to a single city. Barcelona’s own noise map records eight hundred and sixty thousand residents living above 55 decibels, distributed across the exposure bands. Applying the World Health Organization’s exposure-response functions to that distribution yields an annual health cost close to 700 million euros, or roughly 440 euros for every exposed resident, with disturbed sleep again the dominant term (Ajuntament de Barcelona, 2022; World Health Organization Regional Office for Europe, 2018). It is a useful figure to carry into the abatement question below: a city spending a fraction of that annual sum on quieter surfaces or a night-time freight ban is not spending against a vague public good but buying down a cost it can put a number to. None of these numbers is a measurement in the way a decibel is. Each is an estimate, assembled from an exposure map, a set of exposure-response curves, and a price placed on a year of health — and each is only as solid as the weakest of those three.
10.2 How you put a price on silence
There are three main ways to attach a number to quiet, and they answer subtly different questions.
The first reads the price out of the housing market. Buyers pay less for a home on a loud road than for an otherwise identical one on a quiet street. The gap, expressed as the percentage by which a property’s value falls for each additional decibel, is the Noise Depreciation Index. The most careful synthesis of the evidence puts it around 0.4 per cent per decibel for road traffic and closer to 0.6 per cent for aircraft noise, though the figure varies with country and method (Nelson, 2004). This hedonic approach has the great merit of resting on real money changing hands rather than on what people say they would pay. Its limits are the limits of the market it reads. It captures the preferences of those with the means to choose where they live, and is largely silent about the renter who cannot move, the harm that never shows up in a sale price, and the value of quiet to people priced out of it altogether.
The second asks people directly what avoiding noise is worth to them — the stated-preference or willingness-to-pay method. It can reach the non-market value that prices miss, and can price a quiet that no market currently sells; but answers to hypothetical questions are known to drift from behaviour, and what people say they would pay depends heavily on how the question is framed. European appraisal frameworks nonetheless lean on such values to put a price on each decibel avoided, per household per year, and to feed the cost-benefit tests that decide whether a bypass is dug or a barrier raised. The values are indispensable and quietly consequential. A scheme can live or die by them, and they carry into the arithmetic all the biases of who was asked, and how.
The third is the health route already used above: count the cases of each outcome, weight them by how disabling each is, and value the resulting healthy years. Its machinery is transparent, which is also its vulnerability, because every step is a choice open to challenge. The exposure-response functions are averages drawn from varied populations. The disability weights are contested. The 0.02 assigned to annoyance and the 0.07 assigned to sleep disturbance in the standard method have both been argued down towards lower values, and since sleep supplies the largest share of the total, a small revision there moves the headline by billions (World Health Organization Regional Office for Europe, 2011). Halve the weight given to annoyance, as the World Health Organization’s own remeasurement of it suggests it should be, and a double-digit share of the 95.6 billion simply evaporates — not because the noise changed, but because a number in a spreadsheet did.1 And the price of a healthy year is a policy variable, not a natural constant.
The result is that the cost of noise is, unavoidably, a construction: not false, but built. A reader shown only the final euro figure has been shown the least informative part of it. The three methods do not converge on a single number, either, and their disagreement is informative rather than embarrassing. The property market prices what buyers can see and act on, the survey prices what people believe they feel, and the health method prices what epidemiology can attribute. A careful account triangulates among them and treats the spread between them as a measure of what remains genuinely uncertain, not as a failure to find the one right answer. One thing none of the three does on its own is say who bears the cost. An aggregate (95.6 billion, 700 million) is silent on distribution, and a figure silent on distribution can be satisfied by lowering the average while leaving the worst-exposed exactly where they were. The methods measure how much; Chapter 9 was needed to ask on whom.
10.3 The abatement ledger
If the cost of noise is large, the more practical question is what it costs to bring it down, and here the accounting is clarifying. Measures fall into a natural hierarchy — at the source, along the path, at the receiver. Cost-effectiveness falls away sharply on the way down.
At the source, quieter tyres, quieter engines and tighter vehicle standards treat the entire fleet at once, and are for that reason the cheapest way to remove a decibel from a whole city. A tighter continental limit on tyre noise, applied to every vehicle sold, lowers the background of every street at once, for no municipal outlay at all. The electric vehicle belongs here too, with the qualification entered in Chapter 6: below about 20 km/h its mandated warning sound gives much of the quiet back. Along the path, the options are more visible and less efficient. A low-noise road surface removes perhaps one to five decibels, and costs on the order of a tenth as much per decibel as a barrier — near nothing if it is laid during a resurfacing that was going to happen anyway. Its benefit fades within around six years, though, as the pores clog (Federal Highway Administration, 2023). A noise barrier is the most effective single intervention along the path, cutting levels by ten to fifteen decibels, but at one to five million dollars a mile and a considerable take of land, and it protects only the first rows of housing behind it (Federal Highway Administration, 2023).
Rail carries its own path measures — resilient fastenings and wheel or rail dampers that shave a few decibels at modest cost — and cities reach, too, for interventions that are not engineering at all: night-time lorry bans, speed reductions, the rerouting of heavy traffic, the pedestrianisation that leaves the old core of a city quieter than its wealthier grid (Chauvineau et al., 2025; Klein Fiorentin et al., 2026; Lyons et al., 2017). These cost little and can do much, but several of them redistribute noise as much as remove it, shifting it from one street onto the next. At the receiver, insulation and glazing work reliably, but house by house, and they place the cost and the effort on the exposed household itself (Yang & Jeon, 2020).
Against a health cost measured in tens of billions, even modest abatement tends to pay for itself. Noise control is cheap prevention, and the case for it is an economic case before it is anything else. The transition sharpens the point. Quiet bought as a by-product of work already funded — the resurfacing already scheduled, the bus fleet already turning electric — is close to free, so that the cheapest decibels of all are the ones removed while doing something else. The returns are as uneven as the measures, though. A barrier delivers a large benefit to a few households at high cost; a source standard delivers a thin benefit to millions at almost none; a resurfacing repays handsomely if it rides on maintenance already budgeted and poorly if it is bought for quiet alone. Cost-effectiveness, in noise as in medicine, is less about whether to act than about choosing the measure whose benefit per euro is highest, which is seldom the most visible one.
A more awkward conclusion follows too. The cheapest measures, the ones that act at the source, are the ones a city cannot enact alone, since they depend on vehicle and tyre standards set nationally or continentally; the measures a city can enact on its own authority, the barriers and the insulation grants, are the least cost-effective of all. The tools are misaligned with the powers, and the gap between them is a matter of governance.
10.4 The cost-of-noise calculator
Nothing in the accounting below is hidden. Every coefficient sits on the page, and any of them can be replaced. The residents entered in each day-evening-night band drive an annual cost that assembles from the three best-established components: annoyance, sleep disturbance and heart disease. Alongside them sits the one frankly political parameter — the price placed on a healthy year — and moving it shows how much of the headline rests on that single choice. The Barcelona preset carries the city’s real exposure distribution. The property panel reports, separately, the capital value discounted from the housing stock, which is a stock rather than an annual flow and must not be added to it.
Source · exposure-response functions for annoyance and sleep disturbance after World Health Organization Regional Office for Europe (2018); disability weights and burden method after World Health Organization Regional Office for Europe (2011); monetary values after European Environment Agency (2025); property depreciation after Nelson (2004); Barcelona exposure from Ajuntament de Barcelona (2022). Accessed 31 July 2026.
10.5 The commodification of tranquillity
That quiet can be priced at all is itself part of the diagnosis. The hedonic method does not merely estimate a cost; it reads, off the housing market, a going rate for tranquillity. In doing so it records that quiet is already a commodity, bought and sold, and distributed like any other by the capacity to pay. The premium on the flat away from the road is the same fact seen from the buyer’s side: those with means purchase their way out of an exposure that others cannot leave. Chapter 9 found that pattern in the geography of who is exposed; the price of quiet is what makes it a market rather than merely a misfortune.
That has a sharp consequence for how this economic case should be used. Once quiet is treated as a commodity, the natural policy test becomes whether abatement is “worth it”, measured against what people are willing to pay for it. Willingness to pay is, by construction, a measure weighted by ability to pay. A cost-benefit rule fed on those values will find quiet most valuable exactly where residents are wealthy enough to bid for it, and least valuable where they are not. The valuation methods that make the case for action can, applied without care, entrench the very inequity the case is meant to address. None of which makes cost-benefit analysis illegitimate. A method reading value off the market inherits the market’s distribution of voice, and will register the quiet-seeking of a wealthy suburb more loudly than the identical wish in a poor one. Kept as a servant — to show that abatement pays, to rank measures by their return — it is the most powerful instrument noise policy has. Made a sovereign — to decide, by aggregate willingness to pay, where quiet is worth providing at all — it reproduces the very map of exposure Chapter 9 set out to question.
The way out of that circle is not to abandon the numbers, which are the strongest argument noise policy has, but to recognise what they are and what they are not. They establish, beyond reasonable dispute, that quiet is valuable and that protecting it more than pays for itself. What no price captures is the difference between a quiet that is earned in the market and a quiet that is owed as of right — between the tranquillity a household buys and the tranquillity a city ensures for everyone within it. What they cannot settle is whether tranquillity ought to be a good distributed by price at all, or whether it belongs, like clean air and safe water before it, among the things a society guarantees rather than sells. That is no longer a question of economics. It is a question of rights, and of the design of a quieter environment.
The remeasurement, commissioned after the 2018 guidelines and conducted by paired comparison in four European countries, put severe annoyance at 0.011 and moderate annoyance at 0.006, against the 0.02 carried forward from the 2011 accounting (Charalampous et al., 2024; World Health Organization Regional Office for Europe, 2024). Its revision of the sleep-disturbance weight is larger still, from 0.07 to 0.009, and since sleep supplies the greater part of the total, the consequence for the headline would exceed the illustration above. Two limits apply. The health-state descriptions used in the remeasurement differ from those of 2011, so the new weights are not a drop-in substitution for the old. And a disability weight is not an exposure-response function: revising it changes what the burden is valued at, leaving the number of people exposed unchanged.↩︎