Poland must create 7.85 million shelter places. The towns that have to plan them have no one who can.
Shelternode finds, costs and funds the buildings, on official data, for the 2,477 municipalities the 2024 civil-protection law binds. Poland first, then Europe.
The money is arriving. The plans are not.
Under art. 90 of the 2024 Act, Poland needs 7.85 million places in protective structures. The 2023 inventory counts 1.43 million, a fifth of the target and 3.8% of the population. Closing the gap costs around PLN 112 billion and, at the pace of 2025–26, about 37 years.
The work falls on small authorities: 56% of the requirement sits with municipalities under 50,000 people. Since January 2025 only 158 of 2,477 have published a shelter tender, half of those for documents, and 28% of tenders end without a contract.
Nationally that is 18,000 to 90,000 separate building projects, each to be identified, inspected, designed, funded, registered and kept current. That is a planning problem, and it repeats every year as money arrives.
A planning system a town runs itself
A consultant's study is correct the day it is printed. The law asks for continuous review. Shelternode is the cycle a municipality keeps running: Survey reads the official cadastre and applies the Act's exclusions; Reach computes coverage under both legal distance definitions; Plan ranks and costs the buildings that close the gap fastest within the available grant; Documents produce the plan, the funding application and the registry export. Monitoring reopens the plan when a permit is issued, a building changes or a deadline runs. Officials decide; the software shows its reasoning.
A regional view gives the voivode office one comparable picture of every municipality. A developer module computes the shelter capacity the Act now requires in new buildings.
Money without plans
Municipal planning capacity as the binding constraint on Europe's shelter build-out. Evidence from Poland. Working paper, October 2026.
Poland's December 2024 civil-protection law obliges every municipality to plan shelter places for a fixed share of residents, and spending has followed on a scale not seen in decades. Using only public data, we estimate the statutory requirement for every municipality, model the cost and time to meet it, analyse all 351 shelter-related public tenders published nationally since January 2025, simulate the contribution of new building rules, and solve a capacitated facility-location problem for one town on the official cadastre. We conclude that the scarce resource is continuous planning at municipal level, state six requirements for systems that provide it, and describe one.