Mapping Without Taking: Five Questions Before You Draw Another Displacement Route
Humanitarian mapping has an uncomfortable inheritance. Its tools descend from colonial survey, its aesthetics from military cartography, and its funding from institutions that prefer clean geometries to messy politics. None of that makes mapping wrong. It does mean that the decision to publish a route, a shelter location, or a density surface is never neutral, and that the burden of justification sits with the person holding the software license.
We have spent the past several years watching displacement maps proliferate — some rigorous, some careless, a few actively dangerous. What follows is not a compliance framework. It is a short list of questions we now require our own team to answer in writing before anything goes live. We offer it to researchers, advocates, and newsroom graphics desks in the same spirit: not as a verdict, but as a delay long enough to think.
1. Who consented, and to what, specifically?
Consent in mapping is usually treated as a formality absorbed into an intake interview. It should instead be a discrete negotiation about foreseeable uses. A family may agree to be counted in a needs assessment and object, reasonably, to appearing as a point on a public web map that can be filtered by neighborhood.
Ask the narrower question: does the person understand that this dataset may be downloaded, recombined with property records, and republished by someone we have never met? A 2026 project geocoding temporary shelter capacity in a southwestern border city illustrates the problem well. Enrollment staff gathered locations for internal planning; the resulting spreadsheet was later folded into an open data release with no second conversation. Nothing illegal occurred. Something avoidable did.
Consent worth the name is specific, revocable, and documented at the level of the use, not the person.
2. Who owns the data when the grant ends?
Most mapping projects answer this question by accident, which means the answer defaults to whoever controls the server. Before fieldwork begins, we want a written clause covering four things: who holds the archive, who may veto publication, how long identifiers are retained, and what happens to the dataset if the funder disappears or the organization dissolves.
Community data licenses are no longer exotic. Ethnic community-based organizations increasingly ask for co-ownership as a condition of participation, and they are right to. A dataset about a neighborhood is an asset, and assets that leave the neighborhood rarely return.
3. What harm does precision create?
Granularity is a value judgment disguised as a technical setting. A point layer of informal transit nodes, water distribution sites, or unregistered shelters can be read by people who wish the occupants harm — traffickers, militias, landlords seeking leverage, and, in the American context, enforcement agencies that purchase commercial location data precisely because they cannot always obtain it directly.
Our working rule is to map at the smallest resolution that still answers the research question. If a county-level choropleth supports the argument, a rooftop pin adds risk without adding insight. If a route corridor is needed, a corridor of appropriate width beats a polyline through specific checkpoints. Precision should be earned rather than assumed, and every point layer should carry a sentence explaining why it could not be coarser.
4. Who benefits from the map?
A map that travels only upward — from the settlement to the capital, from the field site to the journal — is an extraction diagram with a legend. Benefit-sharing is often dismissed as idealistic, but the practical versions are modest and verifiable: translation of the final product into the languages spoken where the data originated, delivery of the map to local officials before it reaches donors, paid roles for community researchers, co-authorship for local partners, and hosting arrangements that keep the dataset accessible in the region.
Ask what the community receives that it did not have before. If the honest answer is visibility to outsiders, that is a benefit for the outsiders.
5. Who is speaking in the legend?
Categories are arguments. Refugee, migrant, asylum seeker, internally displaced person, irregular arrival, crisis, wave, surge — each of these words carries a policy lineage and a political sponsor. A legend that sorts people into one set of boxes necessarily erases those who fit none of them: the returned, the unregistered, the displaced who never crossed a recognized border, the family that moved three times within a single city and therefore never appears as mobile at all.
Representation also concerns denominators. A density map that normalizes arrivals against county population tells a different story than one normalized against housing stock. Neither is dishonest. Both should be labeled with the choice that was made and by whom. When possible, the people being mapped should review the categories before they harden into a published figure.
A one-page ritual
We now attach a short annex to every map we publish: a consent statement describing how participation was negotiated, an ownership clause naming the archive holder, a justification of resolution, a benefit-sharing plan with named recipients, and a provenance note for every category in the legend. It takes an afternoon to complete and has killed two projects outright.
That is the point. A checklist that never stops a map is decoration.
Mapping as stewardship
Maps shape appropriations, deployment decisions, court filings, and the ambient assumptions of readers who will never visit the places they depict. That power does not require mappers to be neutral — it requires them to be accountable. The five questions above are not obstacles to good humanitarian cartography. They are the difference between documenting displacement and consuming it.
