The Civic Commons Access to tools for building community where you live
Framework · from the local civic tech collection

Design for Care, Not Enforcement

Every 'report a problem' tool can become 'report a neighbor.' The design line that keeps neighborhood tech on the right side: point it at conditions and needs, never at people — and add friction where harm is easy.

Neighborhood technology has a documented shadow. Nextdoor's early 'crime and safety' posts became a notorious channel for racial profiling — neighbors reporting people, not incidents — until organized pressure (notably from Oakland's Neighbors for Racial Justice) pushed the company to redesign the posting flow in August 2016, forcing reporters to describe behavior rather than appearance; the company claimed the new form cut racial-profiling posts by about 75% in its test markets — Nextdoor's own figure (its CEO had told NPR roughly 50% the day before launch, and reporting a year later found profiling persisted). Peer-reviewed research on 311-style reporting finds the same shape from both directions: requests are over-reported in higher-income, higher-homeownership, whiter tracts (Minkoff 2016, *Urban Affairs Review*), while low-income and minority neighborhoods under-report street problems relative to actual need (Kontokosta & Hong 2021, *Sustainable Cities and Society*) — so complaint data is never a neutral map of conditions (White & Trump 2018, *Urban Affairs Review*). And 'report a problem' features aimed at encampments or 'suspicious activity' convert civic energy into enforcement against the most vulnerable neighbors. SeeClickFix's own record carries the critique that anonymous reporting can disproportionately target homeless residents.

The design responses, drawn from projects on this shelf:

  • Report conditions and needs, never people. A map can hold broken streetlights and icy ramps; it must not hold 'suspicious person.' (Taiwan's Disfactory went further: anonymity *protecting the reporter's village relationships*, for reports about land, not people.)
  • Add friction where harm is easy. Nextdoor's multi-step form is the proof that a well-placed speed bump changes behavior.
  • Route reports to care first. 'Ms. Ruth's ramp is icy' goes to the neighbor with the salt, not to code enforcement.
  • Collect the minimum, and know who could subpoena it. COVID mutual aid networks learned to treat neighbor data as a liability to minimize, not an asset to grow.
  • Name a human steward. Moderation is a relationship, not a filter.

The test for any build: imagine the tool in the hands of the block's most punitive resident, pointed at its most vulnerable one. If that picture worries you, redesign before you ship.

Details

  • design moves: conditions and needs, never people, friction where harm is easy, route to care first, data minimization, named human steward
  • practice kind: framework
  • verification note: Citations verified 2026-08-13. Nextdoor: coverage of the Aug 24, 2016 launch (BuzzFeed News, Slate, CBS — 75% company-claimed in test markets; NPR quoted ~50% pre-launch; BuzzFeed News May 2017 found profiling persisted). 311 equity: Minkoff 2016, Urban Affairs Review 52(2), doi:10.1177/1078087415577796; Kontokosta & Hong 2021, Sustainable Cities and Society 64, doi:10.1016/j.scs.2020.102503; White & Trump 2018, Urban Affairs Review 54(4), doi:10.1177/1078087416673202.
Related by topic

Contributed by Relational Technology Project · Drawing on reporting about Nextdoor's 2016 redesign, 311-equity research, and mutual aid data practices

See something to fix or add? Suggest an edit on GitHub — the commons lives in git too, and its history is the changelog.

local-civic-techcivic-tech-frameworkequitysafetyguardrails

Remix this in Relational Builder  Browse the commons