The rent-market question: did homebound tenants start paying more to avoid road noise?
This paper asks a clean question with a physical mechanism. When COVID-19 pushed people into their homes, did renters start caring more about road traffic noise?
Yao-pei Wang, Yong Tu, and Yi Fan study Singapore, a compact city-state with developed transport networks and plenty of housing close to major roads. Their claim is not just that noisy homes rent for less. That was already plausible. The sharper point is that the noise discount became much larger after the pandemic, as tenants spent more of the day at home and as delivery traffic may have brought more vehicle activity into residential areas.
For practitioners, the useful part is the repricing channel. A pre-2020 apartment near a main road might have carried a modest rent discount. After work-from-home became common, the same physical flaw could become more costly to the tenant. The asset did not move. The tenant's use of the asset changed.
The identification setup: Singapore COVID shock, road-proximity treatment, and rental transactions
The authors use 46,980 rental transactions from Singapore's public open rental housing market, covering July 2006 through March 2022. That long pre-period matters because it lets them compare rent patterns for noisier and quieter units before COVID, then test whether the gap changed after the pandemic shock.
The treatment definition is simple: units within 100 meters of main roads are classified as more exposed to road traffic noise. Units farther away form the comparison group. The paper also checks alternative distance cutoffs and distance to the nearest main road, which helps reduce concern that the result hangs on one boundary.
The empirical design is a difference-in-differences setup. In plain English, the authors ask whether rents for road-exposed units fell relative to quieter units after COVID, after accounting for factors that affect rents across time and locations.
Singapore is a sensible setting for this question. The paper points to its compact urban form and developed transport networks, which create useful variation in exposure to main roads. That gives the authors a more granular source of variation than a broad city-center versus suburb comparison.
The size of the repricing: quietness premium rises from modest to economically large
The headline estimates are large enough to matter in underwriting. The paper finds that road traffic noise reduced rents by 3.8% immediately after the pandemic outbreak. In the subsequent year, the discount deepened to 12.7%, equal to about 186.7 U.S. dollars per month.
That is not a small amenity adjustment. A 12.7% rent penalty can change the economics of a marginal acquisition, especially in a low-yield residential market. It also suggests that some pre-COVID hedonic models may now be stale. If an appraiser or investor estimated road-noise discounts using old transaction data, the model may understate the rent drag for units where tenants expect to work, study, and take calls at home.
The timing is the interesting part. The quietness premium does not appear as a permanent fixed feature of the building. It widens after a behavioral shock. The price of quietness is state-dependent. It rises when the home absorbs more of daily life.
Why the mechanism is plausible: work-from-home, delivery traffic, and listing-text evidence
The authors point to two likely channels. First, work-from-home made daytime road noise more salient. A tenant who used to leave at 8 a.m. and return after 7 p.m. had limited exposure to weekday traffic. Once that tenant works from the living room, the same bus route or delivery lane can become a daily productivity cost.
Second, the pandemic increased reliance on delivery services. Food delivery, parcels, and e-commerce shifted more activity toward homes. Even if total city mobility fell during some periods, the mix and location of vehicle activity could still change in ways that made residential noise harder to ignore.
The paper adds text evidence from 10,425 rental advertisements. It finds that references consistent with tenant preference for quietness rose by roughly 10% from 2019 to 2020, reaching about one-third of advertisements. Listing text is not the same as revealed willingness to pay, but it is useful corroboration. Agents tend to advertise what they think tenants care about. If the word quiet becomes more prominent right when rent discounts widen near roads, the story is harder to dismiss as a statistical artifact.
What could break the result: road exposure proxies, selection, supply shifts, and Singapore-specific institutions
There are still places where the result could be weaker than it looks.
Distance to a main road is a proxy, not a decibel reading. Two units 80 meters from a road may have very different noise exposure depending on floor height, orientation, facade, trees, barriers, and whether another building blocks the sound. Measurement error often makes estimates harder to find, but it can also mix noise with other road-adjacent traits, such as air pollution or pedestrian access.
Selection is another issue. The tenant pool changed during COVID. Foreign worker flows, expatriate demand, student demand, and household formation all moved in unusual ways. If the composition of tenants near roads changed differently from tenants farther away, part of the rent gap could reflect who was renting rather than what they valued.
Supply could also matter. Landlords may have delayed listing certain units, renovated them, or changed asking behavior during the pandemic. If road-exposed units came to market under different conditions, the transaction sample may not fully capture the same asset mix before and after COVID.
Singapore's institutions help identification but limit portability. Its public rental market, compact estates, and transport planning make the setting unusual. A U.S. garden apartment portfolio, a London high street flat, and a Singapore rental unit do not have the same tenant base or lease structure. The direction of the effect may travel better than the exact percentage.
Why we could not backtest this on our data
We could not backtest this signal directly. The required data are property-level rental transactions, unit geocodes, distance to main roads or measured noise, and rental listing text. Our tradable datasets are mainly equities, ETFs, crypto, and standard company fundamentals. They do not include Singapore unit-level rental records or the location detail needed to classify road-noise exposure.
Mapping the result into REIT trades would also require property-by-property portfolio data, including each asset's road exposure and tenant mix. Most listed REIT disclosures do not provide that in a usable form. That does not reduce the paper's contribution. It just means the finding is better suited to underwriting and due diligence than to a clean systematic backtest.
How investors could use it despite weak direct tradability: underwriting, REIT diligence, and amenity risk
This is not an obvious long-short signal. It is still useful.
For direct real estate investors, road exposure deserves a fresh rent haircut in markets where hybrid work has stuck. The haircut may be largest for smaller units where the living room doubles as the office, for family units where home study matters, and for buildings with poor sound insulation.
For REIT analysts, the paper suggests a due diligence question that is easy to ask and often skipped: how much of the residential portfolio sits on noisy arterials, delivery corridors, or bus-heavy roads? Management teams may talk about transit access as a positive. That can be true for commuting. It can also be a noise liability when tenants spend more weekdays at home.
For developers, the implication is practical. Better windows, unit orientation, acoustic barriers, and internal layouts may have higher payback than pre-pandemic rent models imply. Quiet is not just a lifestyle word in a listing. In this paper, the estimated second-year post-COVID rent difference associated with road traffic noise is about 186.7 U.S. dollars per month.