Geospatial Data Systems: A Lossy and Destructive Journey from Data Production to Data Use

When geospatial data cause harm, the problem is usually diagnosed as inaccuracy — the map was wrong, the resolution too coarse, the classification imprecise. This paper identifies a different mechanism: harm produced not by what spatial data misrepresent but by what they lose as they travel from production to governance use, and by how thinned outputs acquire institutional authority over those who had no part in their construction. Drawing on Leonelli's data journeys framework and approximately one hundred documented cases of geospatial governance harm, we develop the concept of the operationalization squeeze — the compression of situated categories into portable proxies that harden into decision gates — and identify two analytically independent dimensions along which it operates: topology (how data circulate or are prevented from circulating) and warrant (how thinned outputs exercise authority at landing). A recurring asymmetry structures these journeys: institutions act on probabilistic proxies while exporting the cost of uncertainty onto those they govern. Three traced cases — parametric drought insurance in Malawi, slum demolitions in Delhi, and agricultural subsidy classification in Spain — demonstrate the framework, and a cross-case matrix confirms that circulation shape and mode of authority vary independently. The paper proposes three design principles for journey-aware governance, unified by a common axiom: the burden of uncertainty must follow the direction of power, not flow away from it.