Cities pay for green roofs at installation. The roofs die at handover.

Researchers who tracked 46 green roofs across the southeastern United States report that nearly half are no longer functioning as intended: nine were abandoned, with vegetation loss and no maintenance, and twelve had been removed completely ^[1][3][4][5]. That matters because the public subsidy for these roofs is structured as a construction payment while the cooling they are bought for is a service that only exists while someone funds upkeep ^[11][13].

The physical case for the roofs is not the weak point. Plants on a roof absorb sunlight that would otherwise heat the building, cool surrounding air through evapotranspiration, and act as insulation that cuts heating and cooling loads ^[15]. They also hold rainfall back from storm drains, which reduces flood risk and improves water quality downstream ^[16]. On that basis the European Commission and the US Environmental Protection Agency have both promoted green roofs as urban heat island mitigation and extreme weather resilience ^[14]. The form is old enough to have a track record: Rockefeller Center's terrace gardens were designed in the 1930s to carry trees, grass and shrubs ^[19].

The financing is where it breaks. Cities offer developers tax credits, stormwater fee reductions and accelerated permitting to get roofs installed ^[11]. Philadelphia's business tax credit can cover up to 50% of green roof construction costs ^[12]. All of that sits on the capital side of the ledger. Maintenance sits on the operating side, and the study's authors note that when a building's operational budget is cut, landscape elements including green roofs are commonly altered or have maintenance scaled back to save money ^[8]. Their read is that neglect driven by maintenance cost, rather than bad design, is why many roofs do not last beyond a few years ^[7]. By the arithmetic in their own sample, the 21 abandoned or removed roofs are about 46% of the 46 studied ^[6].

The predictor list is the part worth pricing. LEED certification did not predict whether a roof survived ^[9]. Neither did whether the owner was a private corporation or a nonprofit ^[9]. What did was a change in building ownership: according to the authors, every time a building changed hands the green roof was either abandoned or removed, possibly because knowledge of the roof left with the previous owner ^[10]. That makes the failure trigger an ordinary commercial transaction that incentive programs do not track, since those programs are built around getting roofs constructed rather than keeping them working ^[13].

Two limits on the finding. It covers 46 buildings in one region ^[1]. And it audits whether roofs were still functioning and maintained years after construction, not how much cooling they delivered when new ^[2], so it bounds the persistence of the benefit rather than its size.

Watch whether any jurisdiction ties the back half of a green roof credit to an inspection several years out, or requires the roof to be disclosed and its maintenance plan transferred at sale ^[11][12]. Watch whether stormwater fee reductions get re-verified against vegetation cover, since that is the one incentive with an annual billing cycle already attached ^[11]. And watch for replication outside the Southeast before treating the ownership-transfer result as general ^[1].