The office changed. The city felt it.
The “donut effect” describes a geographical redistribution: weaker demand in some central areas alongside stronger demand in surrounding residential rings. Remote work can loosen the connection between a job's office address and the place a household chooses to live. It can also move weekday spending from a city-center café to a neighborhood business.
Ramani and Bloom used the term in research on the COVID-19 shock to U.S. cities.[1] That is a dated empirical setting, not evidence that an identical process unfolded everywhere for a decade. This article examines the mechanism and its limits rather than extending the label into an unsupported universal forecast.
“The city” is also too broad a unit for some questions. A central business district, an inner residential neighborhood and a suburb can change in opposite directions. An apparent metropolitan recovery may coexist with a difficult adjustment for a particular block, transit route or municipal revenue stream.
Follow the trip that no longer happens
- Solid cobalt center: central business district (CBD).
- Pale inner ring: inner residential neighborhoods.
- Blue outer ring: outer residential and suburban areas.
Fewer routine commuting trips can redistribute housing demand and daytime spending. Rings show relationships, not measured distances or predicted prices.
Housing and spending move through different channels.
Fewer office days can make a longer commute tolerable, expanding the set of homes a worker considers. A desire for a home office may also increase demand for space. Neither mechanism implies an unlimited move away from employment centers: hybrid attendance, schools, family networks and amenities still constrain location.
Daytime spending can change without a household moving at all. Someone who remains in the same apartment but works locally may stop buying lunch near the office. A downtown shop can therefore lose demand even when the metropolitan population is stable. Conversely, a destination that draws residents and visitors for reasons beyond work may retain demand.
The later research on how working from home reshapes cities examines these spatial adjustments while acknowledging the difficulty of isolating remote work from the pandemic's other effects.[2] Lockdowns, health concerns and changes in preferences arrived together. A correlation between remote-capable jobs and urban change is informative, but it is not permission to assign every change to one cause.
Commercial adjustment
Office landlords and nearby businesses confront changed patterns of use. A fall in occupied office space need not translate immediately into cheap homes: conversion can face physical, financial and regulatory barriers.
Public adjustment
Transit and local services face changes in the timing and geography of demand. A system designed around a dense weekday rush may need a different service pattern.
The useful response is adaptation, not a slogan.
Diagnosis distinguishes changes in residents, commuters, spending per visit and building use. An attendance subsidy addresses a different mechanism from investment in housing, learning or recreation.
Distribution matters as much as aggregate activity. A remote worker may gain flexibility while a cleaner or food-service worker loses hours. New demand in an outer neighborhood can help businesses and raise housing pressure on existing residents. These effects should be traced, not bundled into a single claim that remote work is good or bad for cities.
The donut effect describes redistribution within urban networks. It does not establish that a city center must disappear.
Sources & further reading
Follow the evidence.
- 01Ramani & Bloom: The Donut Effect of Covid-19 on Cities
NBER Working Paper 28876, first issued 2021; U.S. pandemic-era spatial change.
- 02How working from home reshapes cities
Research article with comparative evidence and discussion of causal identification; 2024.